PROCEEDINGS Academy of Natural Sciences PHILADELPHIA VOLUME LVIII 1906 philadelphia : The Academy of Natural Sciences LOGAN SQUARE 1906-1907 ,A The Academy of Natural Sciences of Philadelphia, February 23, 1907. I hereby certify that printed copies of the Proceedings for 190C have been mailed as follows : — Pages 1- 32 mailed May 3 - 64. 65-112. 113-160. 161-260. 261-292. 293-356 357-420. 421-428. 429-460. 461-508. 509-556. 557-588. May 18, 1906. May 19, 1906. May 29. 1906. June 20, 1906. July 24, 1906. August 21, 1906. Septembei • 25, 1906. October 19, 1906. Novembe " 6, 1906 January 7, 1907. January 16, 1907 February 19, 1907 March 8, 1907. EDWARD J. NOLAN, M.D., Recording Secretary. PUBLICATION committee: Henry Skinner, M.D., Philip P. Calvert, Ph.D. Henry A. Pilsbry, Sc.D., Witmer Stone, Edavard J. Nolan, M.D. The President, Samuel G. Dixon, M.D., cx-officio. EDITOR: Edavard J. Nolan, M.D. 7n r CONTENTS. For Announcements, Reports, etc., see General Index. PAGE Baxks, Nathan. New Oribatidte from the United States (Plates XIV-XVIII), 490 Blatchley, W. S. On Some Reptilian Freaks from Indiana, 419 Brown, Stewardson. Botanizing in the Canadian Rockies, . 429 Chamberlain, Ralph V. A New Lithobius from Colorado, . 3 Fowler, Henry W. Some Cold-blooded Vertebrates of the Florida Keys (Plates III and IV), 77 Further Knowledge of Some Heterognathous Fishes, Part I, 293 Note on the Dusky Salamander (Plate XIII), ... 356 Description of Two New Species of Centropomus, . 423 Further Knowledge of Some Heterognathous Fishes, Part II, 431 Some New and Little-known Percoid Fishes, 510 McIndoo, N. E. On Some Fishes of Western Cuba, .... 484 Moore, J. Percy. Additional New Species of Polychaeta from the North Pacific (Plates X, XI, XII), 217 Descriptions of Two New Polychseta from Mexico, 352 Descriptions of New Species of Polychseta from the South- eastern Coast of Massachusetts (Plate XIX), . .501 Oberholser, Harry C. A Monograph of the Genus Collocaha, . 177 Pilsbry, Henry A. Description of a New Australian Glycymeris, 213 Pilsbry, Henry A., and C. M. Cooke, Jr. On Hawaiian Species of Sphyradium, 215 I'lLSBRY, Henry A., and J. H. Ferriss. Mollusca of the South- western States, II (Plates V-IX), 123 Mollusca of the Ozarkian Faima (Plates XX, XXI. XXII), . 529 Rehn, James A. G. Studies in South and Central American Acridinse (Orthoptera), with the Descriptions of a New Genus and Six New Species, 10 Records and Descriptions of Non-Saltatorial Orthoptera from British Guiana, 262 Descriptions of Five'New Species of Orthoptera from Tonkin, 279 IV CONTEXTS. PAGE Rehn, James A. G., and Morgan Hebard. A Contribution to the Knowledge of the Orthoptera of Montana, Yellow- stone Park, Utah and Colorado, 358 Smith, Burnett, Ph.D. Phylogeny of the Races of Volutilithes petrosus (Plate II), 52 Stevens, N. M., and A. M. Boring. Planaria Morgani n. sp. (Plate I), 7 Vaux, George, Jr.,. and William S., Jr. Observations on Glaciers in Alberta and British Columbia (Plates XXIII et seq.), 568 Watts, Harvey M. Weather Predictions, 430 PROCEEDINGS OF THE ACADEMY OF NATURAL SCIENCES OF PHILADELPHIA. 1906. January 2. Mr. Arthur Erwin Brown, Vice-President, in the Chair. Thirty-four persons present. The Council reported that the following Standing Committees had been appointed to serve during the ensuing year: Finance. — John Cadwalader, Edwin S. Dixon, Effingham B. Morris, Horatio C. Wood, M.D., and George Vaux, Jr., Treasurer. Publications. — Henry Skinner, M.D., Henry A. Pilsbry, D.Sc, Witmer Stone, Philip P. Calvert, Ph.D., and Edward J. Nolan, M.D. Library. — Dr. C. Newlin Peirce, Thomas A. Robinson, Thomas Biddle, Jr., M.D., Benjamin Sharp, M.D., and George Vaux, Jr. Instruction. — Benjamin Smith Lyman, Henry A. PilsJjry, D.Sc, Charles Morris, Philip P. Calvert, Ph.D., and Dr. C. Newlin Peirce. Committee of Council on By-Ia\ws, — Arthur Erwin Brown, Thomas H. Fenton, M.D., John Cadwalader and Charles B. Pen- rose, M.D. 2 PROCEEDINGS OF THE ACADEMY OF [Jan., Th e Chair announced the deaths of the following : — Charles T. Yerkes, a member, December 29, 1905. George Bowdler Biickton, a correspondent, September 25, 1905. J. B. Ellis, a correspondent, December 30, 1905. Mr . Stewardson Brown made a communication on a botanical trip to the Bermudas taken during the last summer. (No abstract.) January 16. Mr. Arthur Erwin Brown, Vice-President, in the Chair. Thirty-four persons present. Dr. Sharp having taken the Chair, Mr. A. E. Brown made a commu- nication on theories of evolution since Darwin. (No abstract.) The following w^ere elected members: H. S. Jennings, Robert G. LeConte, M.D., and Henry Leffman, M.D. The following was ordered to be printed : 1906.1 NATURAL SCIENCES OF PHILADELPHIA. A NEW LITHOBIUS FROM COLOEADO. BY RALPH V. CHAMBERLIN. Lithobius harrietse sp. nov. Diagnosis. — Angles of the 9th, 11th, and 13th dorsal plates produced ; articles of antennae 36; ocelli 11-12 in 3 series; prosternal teeth 2-2, 3-3; spines of first legs beneath 1,3,1, of penult 1,3,3,1 with 2 claws, of anal 1,3,3,0 with 2 claws; coxal pores in 1 series, round, 4,6,6,5; length 11.5 mm. Description. — Dorsum brown with a l^lack or blackish median longi- tudinal stripe; head paler especially anteriorly; antennae brown or yellowish-brown, paler distally; legs yellow; venter yellow, darkest over middle portion. Head wider than long (85 : 78), widely truncate or slightly incurved behind, lateral margins moderately convex behind posterior ehd of eye patch, then converging to bases of antennae, anterior margin of head between antennae substraight; two furrows extending trans- versely between eyes concave anteriorly, the posterior one usually giving off two branches which converge caudally and meet in a point on the median line; each posterior lateral portion of head marked with one or several sharply impressed curved lines. Ocelli arranged in an oblong patch in three series ; 11 or 12 in number, 1 -1-4,4,2(3); single ocellus large, vertically elliptical; first eye of upper row larger than others excepting the single one. Antennae short, consisting of 36 articles of short length, the ultimate conical, in length about equalling the two preceding together; articles in specimens examined nearly glabrous, possibly due to rubbing. Prosternal teeth 2-2, 3-3; when 3-3 the external tooth on each side more widely separated from the median than the latter from the in- ternal. Angles of the 9th, 11th, and 13th dorsal plates produced; 1st and 3d scuta considerably sinuate posteriorly, 5th slightly so 7th straight, 8th, 10th and 12th sinuate. Spines of first legs ^j^;o' ^^ *^^® penult ^^^{, the claw with 1 spine; of the anal qi's 30' ^^^^ ^^^^^ v^'Wh 1 spine. 4 PROCEEDINGS OF THE ACADEMY OF [Jan., Coxal pores in 1 series, round, 4,6,6,5. Gonopods of female with claw tripartite, the lobes rather short; basal spines 2-2, or 3-3, stout, acutely pointed above middle, the inner smallest. Length of body 11.5 mm.; width 10th plate 1.6 mm.; length of an- tennae 3.3 mm. ; of anal legs 4 mm. Locality. — Near Glenwood Springs, Col. The types consist of four adults. 1906.] NATURAL SCIENCES OF PHILADELPHIA. February 6. Mr. Arthur Erwin Brown, Vice-President, in the Chair. Fourteen persons present. The Piibhcation Committee reported that papers under the follow- ing titles had been presented for publication: "Mollusca of the Southwestern States, III," by Henry A. Pilsbry (January 24). "A New Lithobius from Canada," by Ralph V. Chamberlain (Janu- ary 26). ''Some Cold-blooded Vertebrates of the Florida Keys," by Henry W. Fowler (January 29), "Studies in South and Central American Acrididae (Orthoptera), with the descriptions of a new genus and six new species," by James A. G. Rehn (February 2). "Further Knowledge of some Heterognathus Fishes," by Henry W. Fowler (February 6). Mr. Witmer Stone made a communication on certain orni- thologists of the past, giving incidents in the lives and commenting on the works of Catesby, William Bartram, Barton, Wilson, Charles Lucien Bonaparte, Audubon, John K. Townsend, William Gamble, John Cassin and Samuel W. Woodhouse. (No abstract.) The Council reported that the President, Samuel G. Dixon, M.D., had been appointed to represent the Academy at the approaching celebration of the 200th anniversary of the birth of Benjamin Franklin by the American Philosophical Society. February 20. Mr. Arthur Erwin Brown, Vice-President, in the Chair. One hundred and six persons present. 6 PROCEEDINGS OF THE ACADEMY OF [Feb., The Publication Committee reported that a paper entitled 'Tlanaria morgani n. sp.," by N. M. Stevens and A. M. Boring, had been pre- sented for publication (February 14). The deaths of Charles Smith, a member, March, 1905, and of John C. Wilson, a member, February 19, 1906, were announced. Mrs. Charles Schaeffer made a communication on the caves and flowers of the Cougar Valley in the Selkirks. (No abstract.) The following were elected members: D. H. Tennent, R. E. B. McKenny, and Edward S. Miles. The following were ordered to be printed : 1906.1 NATURAL SCIENCES OF PHILADELPHIA. PLANARIA MOEGANI n. sp BY N. M. STEVENS AND A. M. BORING. Size. — Length 10-12 mm., width in the region of the pharynx 1.6 mm., at the level of the eyes 1 mm. Form. — Head wedge-shaped, no auricular appendages, posterior end obtuse, thickness slight compared with Planaria maculata or P. simplicissima (Plate I, figs. 1 and 2). Color. — ^White and translucent; digestive tract often colored green, yellow, brown, or red by food recently ingested. Eyes. — ^Two, crescent-shaped, situated far back and near together (PI. I, figs. 1-4). Nervous Systejn. — Cephalic ganglia large and distinct, connected by a broad commissure, lateral nerve cords as in P. maculata, two promi- nent nerves extending forward from below the eyes (Plate I, fig. 4). Pharynx. — Single, much elongated, centrally located in forms with functional sexual organs, posterior to the center in forms which are reproducing asexually (Plate I, figs. 1 and 2). Digestive tract as in the typical triclads with little anastomosis of branches (fig. 3). Reproductive Organs. — Ovaries two, situated ventrall}^ between the anterior axial division of the gut and its first lateral branches, some- times lobed, and often extending nearly to the dorsal epidermis (Plate I, fig. 4). Oviducts following a spiral course dorsal and slightly lateral to the nerve cords, back to the region of the atrial organs, where they converge and unite into a short common oviduct which enters the atrium dorsal to the opening of the uterus stalk, instead of entering a vagina as in P. maculata (Cm-tis, '02). Spermatozoa are often found massed at the anterior end of the oviducts. Yolk glands lie between all of the main branches of the gut dorsal to the testes. Testes many, irregular in form and variable in size, scattered between the branches of the gut, for the most part ventral and median to the nerve cords, and extending from the region of the ovaries to the posterior end of the pharynx (fewer and larger than in P. maculata). Distinct seminal vesicles extending posteriorly from the last pair of testes (y.s., Plate I, figs. 5 and 6). Vasa deferentia broad, short and 8 PROCEEDINGS OF THE ACADEMY OF [Feb., only slightly coiled, entering the anterior enlargement of the lumen of the penis nearly at right angles. Penis short and broad, nearly filling the atrium (similar to that of P. maculata). Uterus large, dorsal, ex- tending forward nearly to the pharynx chamber; stalk of the uterus passing to the left side, dorsal to the vasa deferentia and oviducts, and entering the atrium laterally and ventral to the entrance of the common oviduct (Plate I, figs. 5 and 6). Reproduction by Fission. — Like Planaria maculata, this species reproduces both sexually and asexually. Indications that fission had occurred were observed in May, 1901-1905. In May, 1905, while all the individuals in one branch of a certain stream were sexually mature and depositing egg-capsules, those in another branch of the same stream were found in all stages of fission and regeneration (Plate I, figs. 7-10). During the last week of May and first week of June a few individuals showed regenerating atrial organs. The same difference, as to sexual matiu-ity and fission, was observed in collections from the two branches of the stream in October, and at the date of writing (February 7th) there are no sexually mature worms among the thirty to forty specimens which were reproducing by fission when brought into the laboratory in October. About half of the number have divided in the aquarium. Fission occurs at a point nearer the pharynx than in P. maculata (fig. 7), Regeneration. — Both anterior and posterior regeneration at all levels proceeds much as in P. simplicissima, but both regeneration and form regulation are more rapid. Habitat. — Planaria morgani is found on the under side of stones and dead leaves in a small stream, one branch of which flows through the Bryn Mawr College campus. The authors have not found it elsewhere. Explanation of Plate I. Fig. 1. Freehand sketch of living planarian, magnified five times. Atrial organs present (a). Limits of the digestive tract indicated by dotted out- line. Fig. 2. Same of an individual without atrial organs. Fig. 3. Camera drawing of whole mount, showing pharynx and branches of the digestive tract. Magnification 10 diameters. Fig. 4. Reconstruction from camera drawings of frontal sections, sho^ving cerebral gangUa, lateral nerve cords, cephalic nerves, eyes, and ovaries. Mag. 35 d. Fig. 5. Reconstruction of reproductive system from a series of frontal sections, dorsal aspect. p. =pharynx. /. =testis. od. =oviduct. v.s. =vesicula seminalis. ■u.=uterus. p.Z. = penis lumen, pe. =penis. a. =atrium. c.o. = common oviduct. ?•.= reproductive pore. m.s. =uterus stalk. Mag. 10 d. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 9 Fig. 6. Similar reconstruction from sagittal sections. Left-hand aspect. Ox-i- duct and vas deferens of right side omitted. Lettering as above. Mag. 10 d. Fig. 7. Freehand sketch of the anterior piece of a living specimen soon after fission. Mag. 5 d. Fig. 8. Similar drawing from a larger anterior piece several davs after fission, showing considerable regeneration. Fig. 9. Posterior piece soon after fission, showing the two separate branches of the digestive tract. Mag. 10 d. Fig. 10. Similar posterior piece after several days' regeneration (probably about a week). Branches of digestive tract united and new pharj-nx and eyes formed. Mag. 10 d. 10 PROCEEDINGS OF THE ACADEMY OF [Feb;, STUDIES IN SOUTH AND CENTEAL AMERICAN ACRIDINiE (ORTHOPTERA), WITH THE DESCRIPTIONS OF A NEW GENUS AND SIX NEW SPECIES. BY JAMES A. G. REHN. The subfamily treated in the following pages is usually termed the Truxalinse, but the resurrection of the Old World Linnsean genus Acrida necessitates the use of the name Acridinae. Of this subfamily by far the majority of the South and Central American genera have been studied in this connection and numerous opinions and conclusions regarding their relationship are given in the following pages, while several genera, or subgenera, have been con- sidered untenable and one new one proposed. The consideration of several wrongly identified or overlooked species caused a shifting of generic names, which necessitated in one case a new generic name. Twenty-two genera have been considered in more or less detail, and sixty-three species, of which six are new, have been examined. The material examined numbered six hundred and twenty specimens, of which two hundred and sixty-seven were from the Academy Collection, three hundred and twenty-eight from the Hebard Collection, and twenty-five from various sources. I wish to express my thanks to Mr. Hebard for the privilege of examining the material from his collection. HYALOPTERYGES. HYALOPTERYX Charpentier. 1845. Hyalopieryx Charpentier, Orthopt. Descr. et Depict., tab. 46. Type. — H. rufipennis Charpentier. From its closest allies Hyalopteryx can be separated as follows: from Radinotatum by the presence of well developed tegmina and wings, the caudal expansion of the pronotum and the much shorter rostrum and less produced head; from Achurum by the more robust form, the caudal expansion of the pronotum (more or less pronounced ac- cording to the sex), the much broader tegmina and wings as well as the shorter, blunter rostrum and less produced head ; from Eutryxalis in the longer, slender limbs, the somewhat broader tegmina, the shorter 1906.] NATURAL SCIENCES OF PHILADELPHIA. 11 rostrum, and the more fenestrate wings of the male ; from Truxalis in the longer, more produced rostrum, the more angulate caudal margin of the pronotum, the slender caudal limbs and the broader, less subequal tegmina; from Orphula in the characters which separate it from Truxalis. The genus Hyalopteryx appears to con- nect two extremes of the series, one extreme represented by Radi- notatum, a quite peculiar type, and Achurum, and the other by Truxalis and Orphula, Eutryxalis apparently sharing the intermediate position, but really being closer to Truxalis. The sequence of genera which appears most natural in the Hyalopteryges is- Radinotatum McNeill. Achurum Saussure. Hyalopteryx Charpentier. Eutryxalis Bruner. Truxalis Fabricius. Orphula Stal. Hyalopteryx rufipennis Charpentier. 184.5. Hyalopteryx rufipennis Charpentier, Ortliopt. Descr. et Depict., tab. 46. [Brazil.] Sapucay, Paraguay. December 13, 16, 19, 1904. March 7, 1902. February 13 and 15, March 6, 7 and 9, 1905. (Foster, Hebard Coll.) Eleven d'd', eleven 9 9. These specimens appear from Charpentier's description and figure to be the same form as he called rufipennis. The description particu- larly mentions the distinct lines on the metazona, and states that the lateral carinse are similar to the median, which would hardly have been said if the metazona was irregularly lineato-rugose and the lateral carinse faint and strongly expanded on the metazona, as in the new species here described as H. asinus. Of the caudal femora he says: "Die Hinterschenkel .... oben eine fein erhabene Linie, die vorn am Knie in eine ganz kleine Spitze ausgeht, woselbst seitwiirts zwey weit grossere hervorragen, wie Stacheln." From this and the com- parative proportions of the structures mentioned above and the sub- genital plate as figured in his plate it is very evident that the Sapucay specimens are either Charpentier's rufipennis or a very closely related new species. Burr's H. exaggerata,^ the only other previously de- scribed species, possesses elongated genicular structures as in my new asinus. The series examined is quite uniform, a little variation in size being * Trans. Ent. Soc. London, 1902, pt. II, p. 183. [Chiquitos, Bolivia.] 12 PROCEEDINGS OF THE ACADEMY OF [Feb., noticed and a little in the intensity of fine blackish maculations on the dorsum, a few specimens being almost free from the latter while others have a '^salt and pepper" effect. Hyalopteryx asinus n. sp. Types: c^ and 9 ; Sao Paulo, Sao Paulo (c?), Jundiahy, Sao Paulo ( 9 ), Brazil. September 14, 1900 (c?), February 24, 1899 ( 9 ). (Hem- pel [d^l Schrottky [ ? ].) [Acad. Nat. Sci. Phila.] Allied to H. rufipennis Charpentier, but differing in the larger size and slenderer limbs in both sexes, and in the male in the bullate meta- zona, the broader, more acute tegmina, the elongate internal genic- ular processes of the caudal femora and the extremely elongate sub- genital plate. The Bolivian H. exaggerata Burr is a close ally as Fig. 1. — Hyalopteryx asinus n. sp. Lateral view of male type. (X 2.) shown by the genicular structures mentioned in the brief original description, but it can be separated by the yellow anal field of the wings and by inference by other characters, as the author says : " In venation and structure, this form resembles H. rufipennis. ' ' Size rather large ; form elongate, very distinctly compressed, greatly so in the male ; surface rugulose and covered with scattered whitish hairs, both the character of the surface and the hairs being much more pro- nounced in the male than the female. Head about equal to the pro- notum in length in the male, distinctly shorter in the female, the occiput much more ascendent in the male than in the female ; f astigium projecting beyond the eyes a distance equal to the width at the cephalic margin of the eyes, lateral margins parallel in the male, slightly con- verging in the female, the apex with a rounded angle, lateral sections more distinctly depressed and the short medial carina more apparent in the female than in the male; face very strongly retreating in the male, slightly less in the female; frontal costa narrow, moderately 1906. NATURAL SCIENCES OF PHILADELPHIA. 13 excavated dorsad and mesad, very strongly compressed dorsad, slightly broader and parallel to the ocellus, gently but regularly diverging to the clypeus, the margins in the male all more or less sinuate; antennae of the male distinctly but not greatly exceeding the head and prono- tum in length, strongly ensiform, depressed, greatest width not con- tained more than seven times in length, the proximal section slender and the apex acute; eyes elongate subovoid, much more acute cephalad in the female than in the male, the infraocular sulcus slightly greater than the length of the eye in both sexes. Pronotum of the male distinctly, but not very greatly, expanded and sub-bullate on the metazona, of the female with the metazona but very slightly broader than the prozona, the lateral carinae slightly and regularly expanding caudad ; cephalic margin subtruncate, caudal margin obtuse-angulate in the male, obtuse in the female but with the angle more acute and the side margins subemarginate ; median carina distinct in both sexes, but apparently more prominent in the female than in the male; lateral carinae of the male parallel on the cephalic portion of the prozona and following the "shoulder" on the metazona; prozona equal to the length of the metazona in the male, Hyaloptcri/x a'^inufi n. sp. Fig. 2. — Dorsal view of head and pronotum of male type. Fig. 3. — Lateral view of genicular region of caudal femur of male type. Fig. 4. — Dorsal view of head and pronotum of female type. (X 2.) very slightly shorter in the female; metazona in the male with the dorsal rugae broken, irregular, and not strictly longitudinal, in the female more regular than in the male but not strongly marked ; lateral lobes with the dorsal length very considerably greater than the depth, cephalic and caudal margins converging except for a short parallel ventral section, the ventral margin sinuate-oblique. Tegmina of the male broad, the greatest width about a fourth the distance from the 14 PROCEEDINGS OF THE ACADEMY OF [Feb., apex and contained slightly less than four times in the length ; costal margin moderately arcuate, apex acute, the sutural margin straight with the apical fourth obtusely deflected to the apex, which consider- ably exceeds the tip of the subgenital plate. Tegmina of the female lanceolate, exceeding the tips of the caudal femora by about the length of the head, greatest width about a third the distance from the apex, which is acute ; greatest width contained about six and a half times in the length; costal margin arcuate distad, the sutural margin straight for about four-fifths the length, the apical fifth obliquely deflected to the apex. Wing of the male very nearly two-thirds the length of the tegmen; costal margin strongly arcuate, the apex rectangulate ; hyaline ulnar area with the width contained about two and a half times in the width of the entire wing, complete transverse veins eight in number. Interspace between the mesosternal lobes very distinctly longitudinal in the male, subquadrate in the female; interspace be- tween the metasternal lobes narrow, slightly longitudinal in the male, transverse quadrate in the female. Abdomen more distinctly com- pressed in the male than in the female; male subgenital plate enor- mously produced, compressed, the apical section needle-like, the length equal to that of the pronotum. Cephalic and median limbs slender, femora very slightly arcuate in the male. Caudal femora very slender and elongate, the greatest width in the male being about nine times in the length, about eight and a half in the female, exceeding the tip of the subgenital plate proximal portion hardly inflated, but almost regularly tapering to the apex, carinse distinct, pattern of the pagina distinct, acute-angulate and but shallowly impressed, genicular lobes in both sexes produced, the dorsal angles of the genicular region produced into broad, somewhat flattened, acute-angulate processes, in the male that of the internal face over twice the length of the external and as long as the femoral depth, in the female the processes of the two sides subequal; caudal tibise slender, very slightly shorter than the femur, lateral margins with seventeen to eighteen spines in the male, nineteen in the female; tarsi with rather small arolia. General color dorsad russet sprinkled and washed with mummy brown, ventral color pale cinnamon. Head in the male with the cari- nse spotted with mummy brown and two rather faint postocular lines of the same color present; eyes in the male mars brown, in the female Vandyke brown. Pronotum with a pair of irregular dark bars on the lateral carinse. Tegmina in the male liberally sprinkled with small mummy brown and bistre maculations. Wings with the costal margin 1906.] NATURAL SCIENCES OF PHILADELPHIA. 15 of the hyaline iihiar area and the cross veins of the same canary yellow, the costal section of the wing wood brown clouded with vandyke brown; anal field proximad rather pale scarlet, distal section smoky hyaline, between the two a very distinct and moderately broad bar of hazel. Caudal limbs of the general color of the ventral surface. Measurements. & 9 Length of body, 31.2 mm. 37.5 mm. Length of pronotum, 5.5 " 7.5 " Length of tegmen, 27 " 34.5 " Greatest width of tegmen, 7 " 5.1 " Length of caudal femur, 18.3 " 22 " A paratypic series of three males has been examined in addition to the type. One is from Jundiah}^ (Schrottky), and the others from Sao Paulo (September 14, 1900; Hempel). These specimens do not differ appreciably except very slightly in color and in the slightly greater size of the Jundiahy male. EUTRYXALIS Bruiier. 1900. Eutryxalis Bruner, Ace. Gen. and Spec. Locusts Argent., pp. 22, 24. Included Metaleptea minor Bruner (not of Giglio-Tos) and Eutryxa- lis strigata Bruner, of which the former ( = Hyalopteryx gracilis Giglio- Tos) is the type. This genus is closely related to Hyalopteryx, but differs in the char- acters given under that genus. Eutryxalis gracilis (Giglio-Tos). 1S97. H[;/alopterij.r] gracilis Giglio-Tos, BoUett. Mus. Zool. ed Anat. Comp. Torino, XII, No. 302, p. 22. [San Lorenzo, Jujuy, Argentina ; Caiza, Bolivian Chaco.] 1900. Eu[tryxalis] minor Bruner, Ace. Genera and Spec. Locusts Argent, p. 24. (Not Metaleptea minor Giglio-Tos.) [Argentina ; common through- out the provinces north of the Rio Colorado.] Sapucay, Paraguay. February 10-15, March 2-10, 1905. Seven c^c?, ten ? ? . [Foster, Hebard Coll.] These specimens are quite uniform in size, and in color have the browns and greens distributed in the bicolored individuals as seen in Truxalis. No uniform brown specimen has been examined. The possession of two males and three females of this species from Carcarafia, Argentina, received from Prof. Bruner and labelled Eutryxa- lis minor Giglio-Tos, enabled the author to clear up a rather unfortunate and complicated question of misidentification. A male and female 16 PROCEEDINGS OF THE ACADP:MY OF [Feb., specimen of Metaleptea minor Giglio-Tos, received through Dr. Borelli (the collector of the types) from the Turin Museum, show that the species really is an Orphula and not closely related to Truxalis {Meta- leptea Bruner), while the specimens determined by Bruner really are Hyalopteryx gracilis, answering the description very well. The species is quite distinct from Hyalopteryx and well worthy of generic separation. From the known records this species appears to be distributed over a large area, extending from Caiza in the Bolivian Chaco to the Rio Colorado, and east to the Paraguay river region. TETIXALIS Fabricius. Truxalis brevicornis (Johansson). Bartica, British Guiana. May 25, 1901. d" 9 . (Crew, A. N. S. P.) Sao Paulo, Brazil. September 7 and 14, 1900. 4 d" d", 2 9 9 . (Hem- pel, A.N. S. P.) Jundiahy, Brazil. 9. (Schrottky, A.N.S.P.) Corumba, Brazil (lowland). March. 5&&. (H. H. Smith, U. S. N. M.) Sa- pucay, Paraguay. February 10-17, March 2-21, 1905. 10 d'd', 17 9 9 . (Foster, Hebard Coll.) This series presents examples of all the color phases noticed in this widely distributed species. Records for this species include Buenos Ayres and San Lorenzo, Argentina, Villa Rica and Asuncion, Paraguay, and Caiza, Bolivia, as well as numerous localities north of Brazil. Bruner says it is " found throughout the Republic [Argentine] north of the Rio Colorado, especially along the eastern border."- ORPHULA St&l. 1873. Orphula Stal, Recensio Orthopterorum, I, p. 105. Included pagana (Stal), plebeia (Stal), intricata Stal and punctata (De Geer), of which pagana has been selected as the type by Giglio-Tos.^ Orphula pagana (St&l). 1860. Gomphocerus (Hyalopteryx) paganus Stal, Kongliga Svenska Fregatt. Eugenics Resa, Zool., I, Ins., p. 3.39. [Rio Janeiro, Brazil.] Sao Paulo, Brazil. September 1 and 5, 1900. 2 9 9. (Hempel, A. N. S. P.) Chapada, Brazil. April. 2 c?, 1 9 . (H. H. Smith, U. S. N. M.) Corumba, Brazil (lowland and highland). March. 2 9 9. (H. H. Smith, U. S. N. M.) Sapucay, Paraguay. February 6-13, ^ Ace. Gener. and Spec. Locusts Argent., p. 23, 1900. 3 Bollett. Mus. Zool. Anat. Comp. Torino, IX, No. 184, p. 9. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 17 March 6-19, 1905. Id" (^,10 9 9. (Foster, Hebarcl Coll.) Asun- cion, Paraguay. 2 d^c^, 2 9 9 . (A. N. S. P.) The specimens from Corumba are slightly larger than individuals of the same sex from the other localities. Considerable variation also exists in the form of the fastigium, the 9 from Chapada having it rather broad, but otherwise inseparable, while one individual from Sapucay has the same portion more acute than usual, in a considerable measure approaching 0. minor. The tips of the tegmina are sharper and more acute in individuals from Chapada, Corumba and Sao Paulo than in Asuncion specimens, but the series from Sapucay includes practically both extremes. Some specimens have the dorsal dark lines absent and the coloration but little varied. The species has previously been recorded from Santos, Brazil, Formosa and Resistencia nel Chaco, Argentina. Orphula minor (Giglio-Tos).-' 1897. M[etaleptea] minor Giglio-Tos, BoUett. Mus. Zool. Anat. Comp. Torino, XII, No. 302, p. 23. [San Lorenzo, Jujuy, Argentina; Caiza and Aguairenda, Bolivian Chaco.] Caiza, Bolivia. (Borelli, A. N. S. P.) One 6". Chaco. (Borelli, A. N. S. P.) One 9 . These authentic specimens, received from the Turin Museum, show that the species is a member of the genus Orphula and closely related to 0. pagana. From the latter species it differs in the slenderer head, with less prominent and somewhat longer eyes, and the more acute fastigium. MERMIRI^. SYRBULA Stai. The genus Syrhula presents two extreme types, one represented by admirahilis, and the other by montezuma and eslavoe. These extremes might with justice be separated, as has provisionally been done by the author (i.e., subgenus Hems), but for the presence of a type like S. acuticornis, which is clearly an annectant form. Syrbula montezuma (Saussure). The specimens from Cuernavaca, ^Mexico, previously recorded by the author as *S. valida and eslavoe, and from La Joya, San Luis Potosi, Mexico, are referable to this species, which is closely related to S. eslavoe and differs chiefly in the less constricted lateral carinae of the pronotum. * Bolivar's Orphula jucunda {Adas Soc. Espan. Hist Nat., XXV, p. 15) from the Rio Atalapo is probably a member of this genus, but apparently quite dis- tinct from either 0. pagana, with which it was originally compared, or 0. minor. 2 18 PROCEEDINGS OF THE ACADEMY OF [Feb., Syrbula eslavse Rehn. 1900. Syrbula eslavoe Rehn, Trans. Amer. Ent. Soc, XXYII, p. 90. [Es- lava, D. F., Mexico.] 1900. Syrbula (Herus) valida Rehn, Ibid, p. 91. [Eslava, D. F., Mexico.] These two names were applied to the sexes of the same species. AMBLYTROPIDIJE. AMBLYTROPIDIA St&l. Amblytropidia ferruginosa Stdl. 1873. A[mblyiropklia] ferruginosa Stal, Recensio Orthopterorum, I, p. 107. [Brazil.] Sapucay, Paraguay. December 19 and 20, 1901. April 30, 1902. February 13-27, March 2-11, 1905. 9 S'd', 12 9 9. (Foster, Hebard Coll.) These specimens are assigned here with some little doubt, as they are larger than Stal's measvirement of the female type (length 25 millimeters). The females all show more or less blacki.sh-brown maculations on the tegmina, and the males have the caudal femora strongly rose-colored with the genicular regions blackish. In the latter respect they appear to approach the very brief characterization of australis, but the an- tennae are longer and the general coloration rather different. The latter character is of very uncertain value in this genus, as the extremes of one species are qviite different, considering A. occidentalis a repre- sentative form. Amblytropidia australis Bruner. 1904. [Amblytropidia] australis Bruner, Biol. Cent.-Amer., Orth., II, pp. 62, 64. [Argentina.] Chapada, Brazil. April and June. 2 d'd", 1 9. (H. H. Smith, U. S. N. M.) As the original description of this species is extremely brief, little can be gleaned from it to aid in identifying material. In the material examined the tegmen are darker near the costal margin than elsewhere, and a median dark line is present on the head and pronotum of the female and the pronotum of one male. The genicular regions are obscured with dark color in both sexes, but more strongly in the male than in the female. The range of the species is here extended north of its previous limit, Caiza, Bolivian Chaco. Amblytropidia vittata Giglio-Tos. 1894. A[mblytropidia] vittata Giglio-Tos, Bollett. Mus. Zool. Anat. Comp. Torino, IX, No. 184, p. 13. [Luque, Paraguay.] 1906.] NATURAL SCIENCES OF PHILADELPHIA. 19 Sao Paulo, Sao Paulo, Brazil. September 1 and 7, 1900. 2 ? 9 . (Hempel, A. N. S. P.) The coloration of these two specimens is generally similar, but one has a longitudinal pale green bar on the proximal portion of each teg- men. This specimen also has a median line on the head and pronotum quite dark, while the other individual is devoid of any striking markings. Amblytropidia trinitatis Bruner. 1904. Amhhjiropidia trinitatis Bruner, Biol. Cent-Amer., Orth., II, pp. 63, 65. [Demerara, British Guiana; Trinidad.] Trinidad, West Indies. 1 o^. (H. D. Chipman, through Bruner, A. N. S. P.) Venezuela. 1 c?. A. N. S. P. The Venezuela male is inseparable from the topotypic specimen. The range of the species is considerably extended by the ^^enezuela record. Amblytropidia auriventris McNeill. 1897. [Amblytropidia] auriventris McNeill, Proc. Davenport Acad. Xat. Sci., VI, p. 227. [Orizaba, Mexico.] Atoyac, Vera Cruz, Mexico. December. 1 cJ*. (Bruner, A. N. S. P.) Amblytropidia mysteca (Saussure). The material previously recorded by the author as A. mysteca from Cuernavaca, Texolo, Uruapan and Patzcuaro, Mexico, has been re- examined, and in all cases represents mysteca as understood by Bruner. The genus Sinipta Stal is a rather aberrant member of this group, chiefly differing in the depressed, subensiform antennoe, which, how- ever, does not appear to be sufficient to remove it from association with Amblytropidia, some species of which have the antennae some- what depressed. ORPHULELL^. The genera generally accredited to this group are with two excep- tions, Calephorus and Comacris, American. The genera examined by the author in this connection are given below, with their apparent positions as far as can be expressed in a linear arrangement. Calephorus Fieber { = Oxycoryphus Fisher). Parorphula Bruner. Sisantum Bruner. OrphuUna Giglio-Tos. Orphulella Giglio-Tos. Clinocephalus Morse. Dichromorpha Morse. Chloealtis Harris. Cocytotettix n. n. { = Fenestra Bruner, not of Giglio-Tos). Toxopterus Bolivar. 20 PROCEEDINGS OF THE ACADEMY OF [Feb., Of these ten genera a total of* thirty-six species has been examined. Several genera like Chloealtis and Toxopterus are rather aberrant and really occupy isolated positions, that of Chloealtis, however, being somewhat modified by Cocytotettix, which is in a way a transition type. The European Calephorus is also somewhat removed from the nearest ally Parorphula, but its position appears to be in this group and its affinity is clearly with the above-mentioned genus. The rela- tionship of Sisantum and Orphulina is close, as an examination of the type species of each shows, but, for the present at least, I have con- sidered them distinct. The only American genus not examined is (Eonomus Scudder, from California. CALEPHORUS Fieber. 1853. Calephorus Fieber, Lotos, III, p. 97. May, 1853. Included C. elegans Fieber and Gryllus dubius Rambur, both equalling Acrydium com- pressicornis LatreiUe. 1854. Oxycoryphus Fischer, Orthoptera Europisea, p. 311. Type, Acry- dium compressicornis Latreille. This genus is mentioned merely to show its relationship to Paror- phula from which it appears to be an offshoot. Aside from the pres- ence of a distinct intercalary vein no character of great weight is ap- parent to distinguish Calephorus from Parorphula. Three females of C. compressicornis from France and Egypt have been examined. PARORPHULA Bruner. 1900. Parorphula Bruner, Ace. Gen. Spec. Locusts Argent., pp. 22, 25. Included P. graminea, pallidinota and strigata Bruner, of which the first, graminea, can be selected as the type. This genus appears to form a transition type between Calephorus and the Sisantum-Orphulina group. Parorphula graminea Bruner. 1900. P[arorphida] graminea Bruner, Ace. Gen. Spec. Locusts Argent., p. 26. [Provinces of Buenos Aires, Santa Fe and Cordoba, Argentina.] Carcarana, Santa Fe, Argentina. 3 c? c^ , 3 ? 9 . (Through Law- rence Bruner, A. N. S. Phila., and U. S. N. M.) SISANTUM Bruner. 1904. Sisantum Bruner, Biol. Cent.-Amer., Orth., II, pp. 30, 69. Type. — S. notochloris Bruner. Sisantum notochloris Bruner. 1904. Sisantum notochloris Bruner, Biol. Cent.-Amer., Orth., II, p. 69. [Medellin, Vera Cruz, Mexico.] Medellin, Vera Cruz, Mexico. 1 (5^. (T. Heyde through L. Bruner, A. N. S. P.) 1906.] NATURAL SCIENCES OF PHILADELPHIA. 21 The genus Sisantum is closely related to Orphulina, and the two form a group intermediate between the Calephorus-Parorphula type and Orphulella. ORPHULINA Giglio-Tos. 1S94. Orphulina Giglio-Tos, BoUett. Mus. Zool. .\nat. Comp. Torino, IX, No. 184, pp. 8, 9. Type. — 0. pidchella Giglio-Tos. This genus can be recognized without difficulty by the characters given by Giglio-Tos. The structure of the frontal costa is c[uite dis- tinctive, at least in several species. Orphulina pulchella Giglio-Tos. 1894. 0[rphnlina] pulchella Giglio-Tos, Bollett. Mus. Zool. Anat. Comp. Torino, IX, No. 184, p. 10. [San Pedro, Paraguay.] Sapucay, Paraguay. March 7-11, 1905. 4 9 9. (Foster, Hebard Coll.) Two of these specimens have the dorsum green or tinted with green- ish, the other two being dull ochre in the same region. Orphulina balloui (Rehn). 1905. Orphulella balloui Rehn, Ent. News. XVI, p. 178, pi. VIII, figs. 2 and 3. [Bay Estate, Barbados, West Indies.] This species is really a member of the genus Orphulina and rather closely related to 0. pulchella. It can be separated by the blunter fastigium, which has both the dorsal and lateral aspects showing a more rounded angle, the slightly more elongate ovoid eye, the slightly less longitudinal and more quadrate interspace between the meso- sternal lobes, and the different coloration. The National Museum Collection contains a single female, labelled "West Indies," collected by the U. S. Fish Commission. Orphulina veteratoria n. sp. Types: c? and ? ; Sao Paulo, Sao Paulo, Brazil. September 5 ( ? ) and 19, 1900. (Hempel; No. 216 part.) A. N. S. P. Allied to both pulchella and balloui, but separated from the former by the more robust form, the less compressed eyes, the slightly less pronounced expansion of the lateral carinse on the metazona and the slightly more acute fastigium. The male has the interspaces between the sternal lobes much as in Orphulella. From 0. balloui it can be separated by the smaller size, more acute fastigium, the more promi- nent eyes and the heavier caudal femora. This species is not closely related to 0. acuta. Size small ; form moderately robust. Head slightly shorter than the 22 PROCEEDINGS OF THE ACADEMY OF [Feb., pronotum, slightly but distinctly ascending; fastigium about as long as the width at the cephalic angle of the eye ( cJ' ) or very considerably Orphulina reteraforia n. sp. Lateral view of female tj-pe. (X 3.) shorter ( 9 ), acute-angulate in the male, rectangulate in the female, an intermarginal depressed lunate area in both sexes; lateral foveolse distinct in the female, much less distinct in the male, cephalic, elongate- lanceolate; face moderately retreating with the apex rather truncate; frontal costa rather narrow, sub- parallel to below the ocellus, very slightly expanded between the antennae, the margins moderately diver- gent ventrad, moderately sulcate from the antennae ventrad; eyes subovate in the male, ovoid in the female, quite prominent in the male, very slightty longer than the infraocular space in both sexes; antennae slightly longer than ( cJ* ) or subequal to ( ? ) the head and pronotum together, moderately de- pressed, slightly expanded proximad and very faintly subensiform. Pronotum with the caudal width con- tained about once and a half in the -length; cephalic margin arcuato-truncate, caudal margin rounded obtuse-angulate ; median carina distinct, lateral cari- nas slightly narrowed at the second sulcus and moder- ately divergent caudad in the male, subparallel to the second sulcus and very slightly divergent caudad in the female; first transverse sulcus not intersecting any of the carina, second intersecting the lateral, and the third intersecting the lateral and median carinae, metazona slightly longer than the prozona in the male, subequal in the female; lateral lobes very distinctly longer than the depth, ventral margin with a considerable oblique cephalic emargination. Interspace between the mesosternal lobes slightly transverse in the male, and slightly longitudinal in the Fig. 6.— Orphu- lina veteratoria n. sp. Dorsal view of liead and pronotum of female type. (X 3.) 1906.] NATURAL SCIENCES OF PHILADELPHIA. 23 female ; metasternal lobes separated by a very narrow space in both sexes. Tegmina slightly exceeding the tips of the caudal femora, the caudal margin with a distinct but rather small proximal dilation, apex rotundato-truncate; intercalary vein distinct in the female, irregular in the male. Abdomen moderately compressed ; subgenital plate of male blunt. Caudal femora moderately robust, considerabh' dilated in the proximal two-thirds, pagina rather deeply sculptvu-ed; caudal tibiae somewhat shorter than the femora, lateral margins bearing ten or eleven spines, arolia small, subtrigonal. General color of the male bistre, the anal area of the tegmina and the dorsum of the pronotum burnt umber, the lateral angles of the pro- notum ventrad and slightly mesad on the pronotum marked with black; eyes walnut brown; ventral siu-face buffy becoming buff-yellow on the abdomen; caudal femora ochre yellow becoming tawny, ochra- ceous and raw umber distad, caudal tibise dull heliotrope purple, the spines yellowish narrowly tipped with black. General color of female bistre becoming wood brown ventrad; dor- sum of the head, pronotum and anal area of the tegmina pale apple green, the lateral carinse of the pronotum marked with blackish as in the male but more narrowly; eyes raw umber blotched with bistre. Measure77ients. Length of body, 14 mm. 19 mm. Length of pronotum, 3 " 4 " Length of legmen, 12 " 15 " Length of caudal femur, 10 " 11.2 " The types are the only specimens of this species which have been examined. It is possible that the sexes here described do not repre- sent the same species, and that my association of them is erroneous. Several characters of the c? do not appear to be in accord with one's ideas as to what the opposite sex of the ? type should be, but as neither belongs to an old species of the genus I have thought best to associate them, pending the acquisition of further material. In case the sexes should prove to represent distinct species I would restrict the name veteratoria to the 9 . Orphulina acuta n. sp. Types: d^ and ? ; Sao Paulo, Sao Paulo, Brazil. (Hempel; No. 216 part.) A. N. S. P. Allied to 0. pulchella, but a heavier species with the fastigiiun more acute, the eyes longer, and more compressed when viewed dorsad, the 24 PROCEEDINGS OF THE ACADEMY OF [Feb., tegmina and caudal femora shorter, the latter more robust, and the whole insect lacking the prominent lateral stripe of 0. pulchella. Fig. 7. — Orphnlina acuta n. sp. Lateral view of female type. (X 3.) Size small; form somewhat robust, moderately compressed. Head distinctly shorter than the pronotum, considerably elevated in both sexes, the occiput more rounded in the male than in the female, the latter having it almost straight; fastigium acute- angulate in both sexes, but slightly more so in the male than in the female, the width at the cephalic angle of the eye equal to ( c? ) or slightly greater ( 9 ) than the length, margins distinct, a moderately de- pressed intermarginal area present on the dorsum, no median carina; lateral foveolae distinct, deep, cephalic, blunt elongate-lanceolate in shape; the front dorsad forming a right angle with the fastigium, ventrad of the antennae considerably retreating; frontal costa narrow in the male regularly but slightly expand- ing from the fastigium to the clyj^eus, deeply sulcate, particularly dorsad of the ocellus; frontal costa in the female similar to the male but subparallel from between the antennse to the ocellus, the width being greater and the sulcation shallower than in the male; eyes regularly ovoid in the male, acute ovoid flattened cephalad in the female, in both sexes very distinctly exceeding the infraocular sulcus in the length; antennae slightly longer (c?) or slightly shorter ( 9 ) than the length of the head and pronotum, slightly depressed and expanded proxim ad , being thus very slightly ensif orm . Pronotum with the great- est caudal width of the dorsum contained once and a half in the length; median carina very distinct, severed in the middle, lateral carinae Una acuta n. sp. Dorsal view of head and prono- tum ot female type. (X 3.) 190G.] NATURAL SCIENCES OF PHILADELPHIA. 25 parallel to the second sulcus, considerably expanding between the second and third, very slightly expanding to the caudal margin; cephaHc margin subtruncate, caudal margin obtuse-angulate, more rounded in the male than in the female; first transverse sulcus obsolete; lateral lobes of the pronotum with the dorsal length slightly greater than the depth, ventral margin with a considerable cephalic emargina- tion. Interspace between the mesosternal lobes moderately longi- tudinal in the male, slightly so in the female; metasternal lobes con- tiguous caudad in the male, separated by a space about a fourth the width of the mesosternal interspace in the female. Tegmina consid- erably exceeding the tips of the caudal femora in length, apex rounded, costal margin with a very narrow and slight proximal lobe; hyaline in the proximal third ; intercalary vein irregular and extending the whole length of the area, becoming lost in irregular reticulations ; ulnar area with a distinct longitudinal dividing vein in both sexes. Wings ample. Abdomen considerably compressed; subgenital plate compressed with a blunt apical tubercle. Caudal femora robust, the pagina deeply and strongly sculptured; caudal tibise distinctly, but not greatly, shorter than the femora, the lateral margins armed with ten or eleven spines ; tarsi with large arolia. General color shading from tawny-olive to burnt umber. Head and pronotum with postocular bars of vandyke brown, broad and dark in the male, quite narrow and light in the female. Dorsum of the head, pronotum and closed tegmina lighter than the lateral aspects, tawny-olive in the female, bay in the male. Venter wood brown. Limbs varying shades of the general color; tarsi very pale viridian green, with the base of the first, the whole second, the apex of the third, arolium and tips of the claws black; spines on the tibiae tipped with black. Measurements. & 9 Length of body, 14 mm. 1S.5 mm. Length of pronotum, 3 " 4 Length of tegmen 12.5 " 16.2 Length of caudal femur, 8.5 " 10.7 A paratypic series of two females have also been examined (Septem- ber 7 and 14, 1900), and differ in no important characters from the type. One specimen represents a darker and more sooty type of color- ation than the other. ORPHULELLA GisHo-Tos. 1894. Orphulella Giglio-Tos, BoUett. Musei Zoolog. Anat. Comp. Torino, IX, No. 184, pp. 8, 10. Included 0. gracilis Giglio-Tos, punctata (De Geer), intricata Stal 26 PROCEEDINGS OF THE ACADEMY OF [Feb., and elegans Giglio-Tos. Of these punctata is the oldest known and the most widely distributed species, and may be considered the type. The species of this genus are quite difficult to separate and the recogni- tion of almost all the forms can be considered only tentative. The extent of individual variation is great, both in structure and color, and reliable characters for the separation of species are iew. The work of Bruner in the Biologia is by far the best treatment of the genus so far published. Orphulella negleota Rehn. 1900. Orphulella neglecta Rehn, Trans. Amer. Entom. Soc, XXVII, p. 94. [Orizaba, Vera Cruz, Mexico.] Orizaba, Mexico. June. 1 d ; type. (0. W. Barrett, A. N. S. P.) 2 cJ^cJ*. (Sumichrast, A. N. S. P. received through Dr. Saussure.) Jalapa, Mexico. August and September. 2 cTc?. (Barrett, A. N. S.P.) Presidio, Mexico. June. Id. (Barrett, A.N. S. P.) Coatepec, Mexico. August. 1 d", 1 9 . (Barrett, A. N. S. P.) This series shows that the species varies somewhat in size (males from 16 to 21 millimeters in total length), but little in color. All except one d^ from Jalapa are of the same colors, allowance being made, of course, for fading in the Sumichrast specimens, while the Jalapa exception has the usual green dorsum replaced by a dull olive tint, while the pronotum near the lateral lobes is not as distiactly marked as in the other specimens. The ? specimen measures as follows: length of body, 22 mm.; length of pronotum, 4.1 ; length of tegmen, 17; length of caudal femur, 12.9. This species seems to be a form limited to the Eastern Cordilleras. Orphulella aouleata Rehn. 1900. Orphulella aculeata Rehn, Trans. Amer. Entom. Soc, XXVII, p. 92. [Cuernavaca, Morelos, Mexico.] Cuernavaca, Mexico. June. 1 c^, 3 9 9 including types. (0. W. Barrett, A. N. S. P.) This species appears to be quite distinct, and is not an Orphulina as has been thought possible by Bruner.^ Superficially this form considerably resembles 0. viridescens Scudder, but the structural char- acters are quite distinctive and show that no close relationship exists. Orphulella gracilis Giglio-Tos. 1894. Orphulella gracilis Giglio-Tos, Bollett. Mus. Zool. Anat. Comp. Torino, IX, No, 184, p. 11. [Province of San Pedro, Villa Rica and Asuncion, Paraguay.] Sapucay, Paraguay. December 20, 1901 (one specimen). 2 dd. (Foster, Hebard Coll.) 5 Biol. Cent. -Amer., Orth., II, p. 74, 1904. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 27 The validity of this species appears to be somewhat doubtful, Giglio- Tos himself ^ considering it a synonym of the North American 0. maculipennis { = pelidna Burmeister). In this he is of course in error, but it is extremely close to 0. punctata, which is so variable that gracilis may only be an extreme with the lateral carinae more or less obliterated between the first and third transverse sulci. The form is rather more elongate, however, than in the average individual of 0. punctata. Orphulella punctata (De Geer). San Rafael, Vera Cruz, Mexico. cJ^, 9. (Townsend through L. Bruner, A. N. S. P.) San Marcos, Nicaragua. 2 9 9. (C. F. Baker, A. N. S. P.) Trinidad. d,9.' A. N. S. P. Bartica, British Guiana. May 1-27, 1901. 23 d'd', 14 ? 9 . (Crew, A. N. S. P.) Sao Paulo, Brazil. September 1-19, 1900. 5^^(^,5 9 9. (Hempel, A. N. S. P.) Sapucay, Paraguay. February 3-27, March 6-12, 1905. lOd'cJ', 22 9 9 . (Foster, Hebard Coll.) This large series of specimens, representing a verj'- considerable range, I find cannot be divided into species on any one or number of the innumerable characters of variation exhibited. The green and brown phases of coloration, varied with maculations and distinct bars or uni- form with practically no markings, and variations in the shape and strength of the lateral carinae of the pronotum, in the length of the tegmina and wings, in the form of the fastigium and eyes, some slight, others pronounced, are all present, and on superficial examination would appear to furnish good characters for dividing the series. At- tempts along these lines, however, soon show how hopeless it would be to erect species on even the most striking types, as intermediates are present to connect them with any other extreme. Two specimens from Gualaquiza and Valle del Lamora, Equador, received from the Turin Museum, labelled "Orphula olivacea" and recorded as such by Giglio-Tos,^ are provisionally referred to this species. They are clearly not 0. olivacea, as comparison with para- typic material shows, and having been preserved in alcohol are badly shrivelled in consequence.^ Orphulella elegans Giglio-Tos. 1S94. 0[rphidella] elegans Giglio-Tos, BoUett. Mus. Zool. Anat. Comp. Torino, IX, No. 184, p. 12. [Resistencia nel Chaco, Argentina; Province of San Pedro, Villa Rica and Asuncion, Paraguay.] 8 BoUett. Mus. Zool. Anat. Comp. Torino, XII, No. 302, p. 24, 1897. ' BoUett. Mus. Zool. Anat. Comp. Torino, XIII, No. 311, p. 39 1898. ^ Bolivar's Orphula pafruelis (Adas Sac. Espan. Hist. Nat., XXV, p. 15) from the Rio Atalapo I have not recognized, and can give no idea of its proper position. 28 PROCEEDINGS OF THE ACADEMY OF [Feb., Four specimens, two of each sex, from Carcaraiia, Argentina, de- termined as this species by Bruner, I am unable to separate from my series of 'punctata, which includes material determined as that species by Bruner. Individuals of the latter from Nicaragua, Costa Rica and British Guiana resemble the Carcaraiia specimens very closely. Orphulella intricata (Stil). 1873. T[ruxalis\ intricata Stal, Recensio Orthopteromm, I, p. lOG. [Buenos Ayres.] Two specimens, c? and 9 , one from Cordoba, the other from Car- caraiia, Argentina, determined as this species by Bruner, are avail- able for study. I am of the opinion that this species will prove to be merely a phase of the very variable 0. punctata. Orphulella mexicana (Saussure). 1861. Ox[ycoryphus] mexicanus Saussure, Revue et Magasin de Zoologie, 2e ser., XIII, p. 314. [Mexico.] Jalapa, Vera Cruz, Mexico. August. 2 9 9. (0. W. Barrett, A. N. S. P.) Coatepec, Mexico. August. 1 9 . (Barrett, A. N. S. P.) Texolo, Mexico. March 13, 1899. 5 d'c?. (S. N. Rhoads, A. N. S. P.) The above records, with the Jalapa record given by Bruner in the Biologia,^ constitute the sum of our knowledge of this species; It appears to be restricted in its range, as suggested by Bruner, and the taking of five specimens in one day shows it can hardly be called rare. The species is apparently confined to the country on the lower eastern slope of the Cofre de Perote. The compressed form of this insect will readily separate it from allied species. The compression of the male abdomen with its appendages is quite pronounced. Orphulella tepaneca (Saussure). 1861. St[enoboth]r[us] tepanecus Saussure, Revue et Magasin de Zoologie, 2e ser., XIII, p. 319. [Mexico.] Orizaba, Vera Cruz, Mexico. July 14, 1891. 1 d", 1 9. (W. S. Blatchley, A. X. S. P.) This pair was received from Blatchley labelled ' ' Orphulella zapoteca, ' ' but they are clearly not that species and answer Saussm^e's description of tepanecus very well. This species would appear to be limited in range to the vicinity of Orizaba, Bruner having also recorded it from there. Orphulella meridionalis Bruner. San Marcos, Nicaragua. 1 c?. (C. F. Baker, A. N. S. P.) Chinan- dega, Nicaragua. 16". (C. F. Baker, A. N. S. P.) » Biol. Cent.-Amer., Orth., II, p. 80, 1904. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 29 These specimens have been compared with the Costa Rican individ- uals previously recorded by the author.^" Orphulella zapoteca (Saussure). 1861. Ox[ycorijphus] zapoteca Saussure, Revue et Magasin de Zoologie, 2e ser., XIII, p. 316. [Mexico.] Texolo, Vera Cruz, Mexico. March 3 and 13, 1899. 8 (^(^. (S. N. Rhoads, A. N. S. P.) The specimens from Alta Mira, Tamaulipas, previously recorded by the author as 0. tepaneca are referable to this species. The frontal costa, however, is moderately sulcate in all the Tamaulipas individuals. Orphulella viridescens Scudder. 1899. Orphulella viridescens Scudder, Canad. Entom., XXXI, p. 187. [Mt. Alvarez, Mexico.] The specimens previously recorded by the author from Eslava, D. F., Mexico," are now before me. This species is quite distinct from any other mainland form, and rather closely related to 0. scudderi. DICHROMORPHA Morse. 1S96. Dichromorpha Morse, Psyche, MI, pp. 326, 383. Type. — Chloealtis viridis Scudder. Dicliromorplia mexicana Bruner. 1904. DicJiromorpha viridis Rehn (not of Scudder), Proc. Acad. Nat. Sci. Pliila., 1904, p. 518. [Guadalajara, Jalisco, Mexico.] 1904. Dichromorpha mexicana Bruner, Biol. Cent.-Amer., Orth., II, p. 87. [Sinaloa and Tepic, Mexico.] A^ide from the longer tegmina and wings this species differs from viridis in the slightly more robust form and ■^he more acute fastigium in both sexes. The range of this species now covers three States or Territories on the west coast of Mexico — Sinaloa, Tepic and Jalisco. Dichromorpha longipennis Bruner. 1904. Dichromorpha longipennis Bruner, Biol. Cent.-Amer., Orth., II, p. 87. [Mexico.] Guadalajara, Jalisco, Mexico. August 20. 1 d^. (J. F. Mc- Clendon, A. N. S. P.) This specimen is uniform brown and does not exhibit the green dorsum mentioned by Bruner as present in the type d^. This the first definite record of the species. 1" Proc. Acad. Nat. Sci. Phila.. 1905, p. 802. " Trans. Amer. Entom. Soc, XXVII, p. 95, 1900 30 PROCEEDINGS OF THE ACADEMY OF [Feb., COCYTOTETTIX'^n. n. 1900. Fenestra Bruner (not of Giglio-Tos), Ace. Gen. Spec. Locusts Argent., pp. 22, 30." Included F. pulchripennis, intermedia and argentina Bruner, of which the first may be considered the type. This genus is quite distinct from any other, but is nearer to the North American Chloealtis than any South American type. The Matto Grosso C. linearis " is rather aberrant in the slender form, narrow tegmina, weaker carina and more hyaline wings. Cocytotettix pulchripennis (Bruner). 1900. F[enestra] pulchripennis Bruner, Ace. Gen, Spec. Locusts Argent., p. 30. [Argentina; open camp.] Carcarana, Santa Fe, Argentina. 2 cJ^d^. (Through L. Bruner, A. N. S. P.) Cocytotettix intermedins (Bruner) 1900. F[enestra] intermedia Bruner, Ace. Gen. Spec. Locusts Argent., p. 31. [Carcarana, Argentina.] Carcarana, Santa Fe, Argentina. 2 (^(j, \ 9. (Through L. Bruner, A. N. S. P.) The cJ* of this species superficially remind one of the males of the certain species of the genus Syrbula. TOXOPTERUS Bolivar. 1890. Toxopterus Bolivar, Anales Soc. Espan. Hist. Xat., XIX, p. 313. Type. — T. miniatus Bolivar. This genus is rather aberrant and isolated in position in consequence, but its position is in the Orphulellse, and nearer to Cocytotettix than any other genus. The structure of the antennae is similar to that of Gomphocerus, but such resemblance is also found in Eritettix, which is a member of the Amblytropidise, and are superficial and not of great taxonomic value. Toxopterus miniatus Bolivar. 1890. Toxopterus miniatus Bolivar, Anales Soc. Espaii. Hist. Xat., XIX, p. 314. [Cumbase, Peru.] Sapucay, Paraguay. January 18, 1903. January 28, February 6- 25, March 2-10, 1905. 7 d'd', 13 9 9. (Foster, Hebard Coll.) 12 -pj-jg genus Fenestra is invalid from Bruner {Ann. Mas. Civ. Stor. Nat. Genova, XXXIII, p. 120, 1893), as he designates no named species for type or included forms. Tlie first species named under the generic term Fenestra was F. bohlsii GigUo-Tos (Zooi. Jahrb., Syst., VIII, p. 807, 1895), which, on a subse- quent page (p. 32), I will show to be the species later named Dichroatettix viridi- frons Bruner, and for which the generic name Fenestra must be used. 13 Kui^vTo^, shrieking, tettie, grasshopper; in allusion to the loud stridulations . 1* Proc. U. S. Nat. Mus., XXX, p. 374. 1906.] NATURAL SCIENCES OF PHILADELPHIA, 31 This series exhibits considerable variation in the intensity of the coloration, the dorsal aspect ranging in color from dull blackish brown to mottled ochre and umber and again washed with dull green. The paler lateral bands, including the ventral portion of the lateral lobes, pleura and caudal femora are distinct in all the specimens examined and in several are decidedly greenish; in these cases the gense are also colored. Some individuals are distinctly maculate, or might really be called tessellate, ochre and umber being the two colors. Considerable variation in the sharpness of the angle of the fastigium is also noticeable, and in both sexes. In some females the angle is almost as acute as in some males. The species has previously been recorded by Giglio-Tos ^^ from Gualaquiza, Cuchipamba and the valley of Santiago, Ecuador. GOMPHOCERI. The South American genera of this group are six in number, three previously known and three recognized as the result of the present study. Of these genera two will be described in a subsequent paper, but they are mentioned here to show their position in the series. The genera would stand as follows: Fenestra Giglio-Tos {=Dichroatettix Bruner). Staurorhectus Giglio-Tos. Isonyx n. gen. Borellia n. gen. Stereoteitix n. gen. Tristira Bruner. The first genus is quite distinct, as is the second. The next three genera, Isonyx, Borellia and Stereoteitix, show some characters in common, but differ in a number of others, such as the position of the lateral foveolae and the relative proportion of the spurs. The last, Tristira, is a quite distinct type, much isolated from the others and immediately recognizable by its peculiar facies. FENESTRA Giglio-Tos. 1895. Fenestra Giglio-Tos, Zoolog. Jahrbiicher, Syst. Abth., VIII, p. 807. Type.— F. bohlsii Giglio-Tos. 1900. Dichroatettix Bruner, Ace. Genera Spec. Locusts Argent., pp. 22, 32. As I have stated on a previous page (p. 30) these names apply to the same genus and species. As there stated Fenestra is invalid from Bruner, 1893,^® on account of the lack of a type or included named '^Bollett. Mus. Zool. Anat. Comp. Torino, XIII, No. 311, p. 39. »« Ann. Mus. Civ. Stor. Nat. Geneva, XXXIII, p. 120. 32 PROCEEDINGS OF THE ACADEMY OF [Feb., species. The only information we have is a footnote to the extent that ' ' Ce genre est fonde sur une espece de Buenos Aires. ' ' The next use of the name is by Gigho-Tos, who placed the species in hand in that genus, in spite of the sanguineous wings, which appendages were stated to be hyaline by Bruner. Bruner was apparently unacquainted with the paper of Giglio-Tos and erected a new genus Dichroatettix for a species, which is undoubtedly the same as F. hohlsii Giglio-Tos, while Fenestra was applied to three species, which therefore require a new generic name, a want supplied on a jM-evious page of this paper. ^^ This genus is related to the North American genera Napaia and Horesidotes. It can be distinguished from the former by the less expanded antennae, the strongly constricted lateral carinas of the pro- notum, the subequal prozona and metazona, the distinctly angulate caudal margin of the pronotum, the longer tegmina with more definite venation and the very peculiar cerci. From Horesidotes (which has not been examined) it appears to differ in that the occiput has no distinct median carina, the caudal margin is more angulate and the lateral lobes of the pronotum are deeper than long. Fenestra bohlsii Giglio-Tos. 1895. F[enestra] bohlsii Giglio-Tos, Zoolog. Jahrbucher, Syst. Abth., VIII, p. 807. [Paraguay.] 1900. Dichroatettix viridifrons Bruner, Ace. Gen. Spec. Locusts Argent., p. 33, figs. 9 and 10. [Cordoba and Carcarana, Argentina.] Cordoba, Argentina. 1 d. (Through L. Bruner, U. S. N. M.) Sapucay, Paraguay. January 24, 26, February 13 and March 2- 21, 1905. 5 d'd^, 10 ? ? . (Foster, Hebard Coll.) This series shows that the green color of the face is not present in some of the females, and in such as it is present the lateral lobes of the pronotum are always, and generally also the dorsum, the pleura and the caudal femora, except the apices, of the same color. All the males examined have the face green. In some specimens the dorsum of the pronotum is without the velvety black patches seen in others, in fact about half the specimens are without them. Size, as exhibited " The name Dichroatettix might with justice be used for this genus if the un- identifiabihty of the Brunnerian genus Fenestra precludes its use by GigUo-Tos. As an unidentifiable name requires different treatment from a nomen nudum the use of Fenestra once in such a way would by some be considered to preclude its subsequent use, its existence terminating as a name when considered unidentifi- able. A nomen nudum, being considered to express nothing definite, can be properly used by a later author from whom it dates. An unidentifiable name, on the other hand, represents something imperfectly and improperly presented and should be discarded for all time. I have presented both views, the proper presentation of the name by a later author and the complete elimination of the name, and leave others to choose for themselves. In any case Giglio-Tos's species bohlsii would stand. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 33 by this series, is subject to little variation. Some of the females show considerable variation in the angle of the fastigium. STAURORHECTUS Giglio-Tos.i^ 1897. Staurorhectus Giglio-Tos, Bollett. Mus. Zool. Anat. Comp. Torino, XII, No. 302, p. 25. Type. — S. longicornis Giglio-Tos. This genus is apparently closer related to Chorthippus (Stenohothrns) than to any other genus of the group, the tegmina of the male being rather similar in some species, but the lateral foveolse are very weak, the lateral carinae of the pronotum are much less distinct and the caudal femora slenderer. staurorhectus longicornis Giglio-Tos. 1897. S[taurorhectus] longicornis Giglio-Tos, Bollett. Mus. Zool. Anat. Comp. Torino, XII, No. 302, p. 26. [San Lorenzo and Tala, Argentina; Caiza, Bolivia.] Campo Santo, Bolivia.^^ 1 d^. (BorelU, A. N. S. P.) San Lorenzo, Jujuy, Argentina. 1 $ . (Borelli, A. N. S. P.) Sapucay, Paraguay. March 2-27, 1905. 12^^(^,37 9 9. (Foster, Hebard Coll.) The interesting series listed above exhibits a very great range of variation with four distinct color forms in the ? ; the cJ^ is more uni- formly colored. These forms I will designate a, h, c and d, lettering from the most diversified to the most uniform type. Type a. A median longitudinal line from fastigium to tip of teg- mina buff, flanked by lines of blackish brown on pronotum and teg- mina, on the latter occupying all except a narrow proximal costal line of pale yellow. Lateral carinse distinct, light, slightly constricted. Light bar on head slightly margined with brownish. Remainder of lateral aspect, and caudal femora except the dorsal blackish line and the reddish distal section, grass green. Represented by twelve speci- mens. Type 6. Similar to type a, but green replaced by ochraceous brown. Represented by eight specimens. Type c. The dorsal median line in this type is almost lost and nearly concolorous with the lateral bars, so that dorsum from the fastigium caudad is nearly uniform. Represented by six specimens, ^* It is quite possible this may prove to be the same as Compsacris Bolivar {Anales Sac. Espaii. Hist. Nat., XIX, p. 314, 1890) based on one species — C. pulcher — from "Villa Bella en el Peru," which is apparently ViUa Bella, Bolivia, at the junction of the Beni and Mamore rivers. If such should prove to be the case Compsacris would replace Staurorhectus. None of the species examined by the author are closely related to C. pulcher judging from the description. ^'Labelled as above, but Campo Santo, Salta Province, Argentina, is probably intended. 34 PROCEEDINGS OF THE ACADEMY OF [Feb. Type d. Tegmina colored much as in type a, but head, pronotum, pleiu-a and Hmbs suffused with rose red, the usual markings being only faintly indicated. Represented by eleven specimens. The largest 9 specimen (38 millimeters in length of body) belongs to type c, the smallest (30 millimeters) to type a. The typical ? re- ceived from the Tiu'in Museum has been immersed in a liquid preserva- tive and it is impossible to tell to which type it belongs. One d^ individual has a coloration which would be con4dered type c. while the other males are of a type which approximates closer to the 9 type h with, however, some greenish, but not on the caudal femora. The curve of the fastigium of the female varies from arcuate to dis- tinctly angulate, and is more excavated in some than in others. Staurorhectus glauoipes n. sp. Type: 9 ; Sapucay, Paraguay. March 8, 1905. (Foster; No. 32.) Hebard Collection. Allied to S. longicornis Giglio-Tos, but considerably smaller and slenderer with the fastigium more acute, the frontal costa more sul- cate, the lateral foveolse of the vertex practically suppressed, the ulnar Fig. 9. — Staurorhedus glaucipes n. sp. Lateral view of type. (X 2.) area of the tegmina without a distinct longitudinal dividing vein, no distinct continuous intercalary vein present, and possessing a char- acteristic coloration which does not appear to vary appreciably in a series of ten females. Size rather small ; form moderately slender. Head distinctly shorter than the pronotum, the occiput not elevated and gently arched longi- tudinally; fastigium shorter than the width at the cephalic angle of the eyes, shghtly acute, the apex rounded, margins with a narrow semicircular depression, no median carina present; lateral foveolse not distinct, ventrad ; face considerably retreating, the apex acute when viewed laterad; frontal costa moderately wide, margins subparallel, reaching to the clypeus, sulcate from dorsad of the antennae ventrad; eyes elongate-ovoid, very slightly longer than the infraocular sulcus; antennse considerably exceeding the head and pronotimi together in 1906. NATURAL SCIENCES OF PHILADELPHIA. 35 length, nearly two-thirds the length of the tegmen, moderately depressed proximad and very slightly dilated. Pronotiim moderately rounded but with a perceptible dorsal flattening; cephalic margin slightly arcuate, caudal margin broadly obtuse-angulate with the angle rounded, no lateral carinse present, but a moderate shoulder developed on the metazona; greatest caudal width contained about one and two- thirds in the length, metazona regularly and closely, but not deeply, punctate; three distinct transverse sulci present on the dorsum, the caudal only intersecting the median carina, which is distinct and regular ; lateral lobes with the dorsal length very considerably greater than the depth, ventral margin sinuato-angulate. Interspace between the mesosternal lobes slightly longitudinal ; interspace between the meta- sternal lobes hardly narrower than between the mesosternal lobes, pentagonal. Tegmina very slightly exceeding the apex of the abdo- men, but falling considerably short of the caudal femora; marginal field with a very slight proximal dilation, apex rather narrowly rounded, no intercalary vein present. Wings ample. Abdomen consider- ably compressed. Caudal femora considerably inflated proximad, the distal section quite slender, pattern of the pagina very distinct, regu- lar; caudal tibiae very slightly shorter than the femora, gently sinuate, the external margins with eleven or twelve spines, internal spurs slightly unequal; tarsi with tlistinct, rounded arolia. General color apple green, becoming more oil green on the caudal femora. Two broad lines, one on each side, of vandyke brown start from the margins of the fastigium, cross the dorsum of the eyes, extend over the pronotum, and on the tegmina broaden out and occupy the entire discoidal and all except the proximal portion of the marginal fields. On the head the color of these bars is decidedly blackish brown, about true vandyke brown on the pronotum, becoming dilute and fainter as the stripe broadens on the tegmina, while along the ventral margins of these bars the green is touched with yellowish, while the light proximal section of the marginal field is cream color. Antennse vandyke brown with the proxi- mal joint green; eyes tawny-olive and bistre mot- tled; caudal femora with the genicular regions laterad and ventrad black, the pagina with a dorsal longitudinal bar of rhedus glaucipes n.sp. Dorsal view of head and pro- notum of type. (X 3.) ' 36 PROCEEDINGS OF THE ACADEMY OF [Feb., olive, except for a pregeniciilar annuliis of clear green; caudal tibise bice green, the genicular section blackish and the distal portion slightly suffused with brownish, spines and spurs tipped with black. Measurements. Length of body, 26 mm. Length of pronotum, 5 " Length of tegmen, 19 " Length of caudal femur, 16.3 " A paratypic series of nine males have also been examined, the dates being as follows: December 21, 1904; January 26, February 15, March 7, 18 and 21, 1905. This series is rather uniform in size, two females, however, being appreciably smaller than the others. In color one specimen has the green more glaucous, while several others show a tendency in the same direction. One specimen has the caudal femora, the lateral lobes and the face suffused with orange vermilion. The intensity of the longitudinal bars appears to vary very little; on the tegmina, however, some specimens are more strongly colored than others. ISONYX n. gen.20 This genus probably is closer related to Borellia Rehn ^^ than to Staurorhectus, and may possibly be considered a distinct and separate type with no close relationship to either the above mentioned genera. The characters are so contradictory and the facies so different that the author does not feel jvistified in making detailed comparisons, but prefers to present the rather striking features of the form in the general description. It might be added that this position is assigned to Isonyx after studying all the South American Gomphoceri, both genera and species. Isonyx paraguayensis n. sp. Types: d^ and ?. Sapucay, Paraguay. February 12 ( 9 ) and 27 (cJ^), 1905. (Foster; Nos. 24 and 165.) Hebard Collection. Size rather small; form moderately compressed and slender; surface subsericeous. Head slightly (cJ^) or considerably (9) shorter than the pronotum, the occiput very slightly elevated and rounded; fasti- gium subrectangulate in the male, obtuse-angulate in the female, very much shorter than the width at the cephalic angle of the eyes, margins 20 I(Tof , equal; owx, claw. -1 Proceedings of the U. S. iSational Museum, XXX p. 379. 1906. NATURAL SCIENCES OF PHILADELPHIA. 37 distinct with a broad sublunate intermarginal depression, no median carina present, fastigium when viewed laterad moderately decUvent: -Isonyx paraguayensis n. gen. and sp. female type. (X 3.) Lateral view of dorsal section of the face vertical, ventrad of the antennae consider- ably (c?) or gently (9) retreating; lateral foveolse distinct, cephalic well impressed and excavated, sublanceolate; frontal costa strongly compressed dorsad, regularly expanding ventrad of the antennae to the clypeus, with an additional rotundate expansion at the ocellus, sHghtly sulcate ventrad of the ocellus; eyes subovate in the male, subovoid in the female, flattened cephalad in the latter sex, moderately prominent Fig. 12. — Isonyx paraguayensis n. gen. and sp. Dosal \'iew of head and pronotum of female type. (X 3.) Fig. 13. — Isonyx paraguayensis'n. gen. and sp. Lateral view of abdomen of male type, (x 6.) in the male when viewed dorsad ; antennse subequal to the head and pronotum in length in the male, slightly shorter in the female, Pro- notum deplanate dorsad, the greatest caudal width contained about once and a quarter in the length; cephalic margin subtruncate, caudal 38 PROCEEDINGS OF THE ACADEMY OF [Feb., margin obtiise-angulate in both sexes ; median carina distinct but not kigh, severed by the third sulcus shghtly cephalad of the middle; lateral carinse clepsydral, strongly constricted to the first sulcus, ab- sent between the first and second, straight between the second and third, wider spaced caudad of the third sulcus than cephalad of it, arcuate diverging caudad ; lateral lobes with the greatest dorsal length slightly less than the greatest depth, the ventral margin sinuato- angulate, a weak shoulder running diagonally ventro-cephalad on the prozona. Interspace between the mesosternal lobes sub- quadrate in the male, slightly transverse in the female; metasternal lobes contiguous caudad in the male, separated by a small ovate space in the female. Tegmina shghtly exceeding the tips of the caudal femora, the tips rounded, costal margin with the proximal lobe small and low; intercalary vein present, distinct but irregular, and becoming lost proximad in the reticulations; ulnar area with a longitudinal dividing vein as distinct as the intercalary in the male, more distinct and regular in the female. Wings ample. Abdomen considerably compressed; cerci of the male styliform, apex rather blunt; subgenital plate of the male somewhat contracted and turned in dorsad, apex moderately acute and but slightly produced. Cephalic and median limbs rather slender, more robust and somewhat inflated in the male. Caudal femora quite robust, considerably inflated, pagina sharply but not very deeply sculptured ; caudal tibiae slightly shorter than the femora and with a hardly perceptible sinuation, lateral margins with ten spines in the male and nine in the female, internal spurs moderately arcuate, equal; tarsi with distinct arolia. General color vandyke brown, mottled and overlaid with seal brown, clove brown and blackish. Eyes cinnamon (d^) or mummy brown ( ? ) ; occiput with two dark arcuate, diverging bars extending caudad from between the eyes. Pronotum with the lateral carinse, lateral sections of the dorsum of the metazona and the extreme upper portions of the lateral lobes seal brown or blackish, leaving a unicolor median bar extending from the fastigium to the caudal margin of the prono- tum, which is more or less hazel in the male and apple green in the female; remaining section of the lateral lobes mottled and hned. Tegmina with a longitudinal discoidal series of quadrate clove brown spots, the marginal field in the female touched proximad with apple green, and the sutural margin in the male with a touch of vinaceous- cinnamon. Limbs mottled and faintly annulate with the general colors, the caudal femora blackish in the genicular region and with the dorso-lateral face marked with several velvety seal brown spots. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 39 one submesad and another nearer the base; caudal tibiae dull olive with a faint proximal lighter annulus, the spines and spurs tipped with black; tarsi soiled greenish with the proximal part of the first joint, the whole second and the tip of the third blackish. Measurements. Length of body, _ 17.7 mm. 21 mm. Length of pronotum, 3.2 " 4.3 " Length of tegmen, 14.5 " 18.3 " Length of caudal femur, 10.7 " 13 " A paratj^ic series of six females has also been examined (February 13, 15 and 27, March 17 and 21, 1905). This series shows that the species varies somewhat but not greatly in size, and the coloration presents a considerable range of base tones and finer definition of markings without much modification of such pattern as is found in the types. Two specimens are touched with green as in the ? type, but both have it weaker on the head and tegmina, and the lateral carinse are broadly marked with the same tint on the metazona, and the caudal femora have distinct oblique blackish bars. Other specimens have the lighter general colors replaced by ferruginous or wood brown, giving a warm rufescent type or a rather sandy form, while the femoral bars are present as variations irrespectively of color form All the females except the type have the lateral carinse marked with lighter on the metazona. One of the rufescent type has the dark colors all blackish, presenting a strong contrast with the other ferruginous patches. SOYLLIN^. After examining individuals of all the genera of the Scyllinoe except Eupnigodcs McNeill and Zapata Bruner, the following arrangement of the genera seems to be desirable as it appears to express their natural relations : Boopedon Thomas. Euplectrotettix Bruner. Scyllina Stal. Eupnigodcs McNeill. Zapata Bruner. Psoloessa Scudcler, Stirapleura Scudder. Ageneotettix McNeill. Aulocara Scudder. Ligurotettix McNeill. 40 PROCEEDINGS OF THE ACADEMY OF [Feb., The position of Eupnigodes and Zapata is taken from other authors, while Ligurotettix is placed here instead of in the Epacromioe, as it is quite different from Mecostethus and Epacromia, approaching Aulocara in some respects. While no doubt somewhat aberrant Ligurotettix is clearly a member of the Scyllinse, the intercalary vein, for example, being no stronger than is the case in many specimens of Aulocara. EUPLECTROTETTIX Bruner. 1900. Ewpledrotettix Bruner, Ace. Gen. Spec. Locusts Argent., pp.23; 38 (Eupletrotettix laps. p. 38). Included E. ferrugineus, conspersus, schulzi and prasinus Bruner, of which ferrugineus msty be considered the type. Eupleotrotettix ferrugineus Bruner. 1900. Euplectrotettix ferrugineus Bruner, Ace. Gen. Spec. Locusts Argent., p. 39. [Sandy knolls about Asuncion, Paraguay; Territory of Formosa, Argentina.] Sapucay, Paraguay. February 13, 1903; December 16, 1904; Janu- ary 26, February 10-25, March 2-19, 1905. 9 6'd',lO 9 9 . (Foster, Hebard Coll.) These specimens are all larger than the measurements given by Bruner, but otherwise do not appear to materially differ. The series is quite uniform in size, and an average c? and 9 measure as follows : Length of body, 18.5 mm. 25 mm. Length of pronotum, 3.5 " 4.8 " Length of tegmen, 15.5 " 21 " Length of caudal femur, 11.5 " 15.2 '' Two distinct types of coloration are present, one irregularly mottled and washed with fuscous, the other with a light orange ochraceous median bar on the head and pronotum, flanked by blackish lateral lines, which are filiform on the head and broader on the pro- notum. The tegmina in the latter type also possess a pair of light parallel longitudinal lines on the anal areas. The indications of femoral bars mentioned by Bruner are quite strong in nearly all the females and several of the males examined, w^hile the blackish genicular region of the cJ' and the orange-red abdomen of the same sex are quite striking. SCYLLINA St&l. 187 . Scyllina Stal, Recensio Orthopterorum, I, p. 112. Included S. peragrans (Stal) and S. viatoria (Saussure) of which 1906.] NATURAL SCIENCES OF PHILADELPHIA. 41 the former is the type, the latter having been removed to Plectro- phorus {=Plectrotettix) by McNeill. 1895. Pseudostauronohis Giglio-Tos, Zool. Jahrb., Abth. Syst., VIII, p. 808. Type.— P. brunneri Giglio-Tos. 1897. Plectrophorus McNeill, Proc. Davenp. Acad. Nat. Sci., VI, p. 198, 251, pi. 4, fig. 21 [not of Ferussac, 1819]. Included Stenobothrus viatorius and gregarms Saussure. 1897. Plectrotettix ]\IcNeill, Psyche, VIII, p. 71. Of the various species here considered a series of one hundred and fifty-six specimens have been examined, including typical material of the five species described by Bruner. Thorough consideration of this material shows that no tangible characters exist for separating Scyllina, Pseudostauronotus and Plectrotettix even as subgenera as proposed by Bruner.^^ The characters on which the subdivisions were there made are: the presence or absence of an intercalary vein in the post-radial area, the cells of this region being accordingly arranged in two series or else irregularly reticulate; the caudal lobe of the pronotum either subequal to or longer than the cephalic; the inner claw of the hind tibiae more or not more than twice the length of the outer, and the number of spines on the outer margin of the hind tibise (9 to 12 or 1.3 to 16). The first of these characters can usually be depended upon as stable, but in this genus the intercalary vein is by no means a stable character being present or absent in individuals of the same species, and in some cases varying in an individual to the extent that it is distinct on one tegmen and absent on the other. The proportions of the lobes of the pronotum are also found to vary, as in species belonging unquestion- ably to the section supposed to have subequal divisions the caudal section distinctly exceeds the cephalic in length, while the reverse is the case in other species. The length and shape of the tibial claws while diagnostic when the type species alone are compared, is without value when other forms are considered ; specimens of varipes, which is allied to conspersa, has the spurs no longer than is found in Mexican forms referred to Plectrotettix. The number of tibial spines is a char- acter which appears unworthy of use in separating subgenera, especially when the extent of variation is such as to cause great doubt as to which division a particular specimen belongs. Wliile the type of the genus Gomphocerus {Epacromia) peragrans Stal,^^ has not been recognized since the original description, a cir- cumstance probably due to the fact that the type locality is an unfre- ^^Biol. Cent. Amer., Orth., II, pp. 99-100. ^' Kongliga Svenska Fregatten Eugenies Resa, Zool., I, p. 343, 1860. [Puna. 42 PROCEEDINGS OF THE ACADEMY OF [Feb., quented one, the description is such that it could be identified with Httle difficulty. The name Pseudostauronotus was proposed by Giglio-Tos under the impression that the type, hrunneri, was a member of the very insuffi- ciently described genus to which Brunner applied the same name two years previously .^^ As Brunner's name rests on a description of seven words without included species or tjrpe mentioned, it is really a nomen nudum and as such not invalidating the later use of the name by Giglio- Tos, even when used in misapprehension. Scudder has shown ^^ from the evidence of material sent him by Brunner that Pseudostauronotus of that author equals his much older Stirapleura. Some species of the genus appear to resemble species of Boopedon, others forms of Stirapleura and several have a striking superficial resemblance to forms of the Locustine genus Schistocerca. The fol- lowing arrangement appears to present their relationship as clearly as possible in a linear arrangement. Species but little variegated, the coloration rather uniform; lateral carinas of the pronotum moderately arcuate, not prominent; super- ficially resembling Boopedon. Scyllina uniformis Rehn. Scyllina instabilis n. sp. Types: 6^ and 9 ; Sao Paulo, Sao Paulo, Brazil. September 5, 1900. (Adolph Hempel; No. 202.) A. N. S. Phila. This species is closely allied to S. uniformis, but differs in the sharper caudal angle of the pronotum, the longer metazona of the pronotum (much longer than the prozona), the slightly narrower interspace be- tween the eyes and the rather different type of coloration. Size medium ; form robust. Head with the f astigium broad, rounded rectangulate, depressed area crescentic, interspace between the eyes equal to the interantennal width of the frontal costa, lateral foveolse oblong, punctate, slightly impressed; viewed laterad the fastigium rounds into the frontal costa without' angle; frontal costa broad, sub- equal, slightly constricted above the antennae, two rows of punctures dorsad; supplementary facial carinas indistinct ventrad; eyes acute reniform, slightly longer than the infraocular portion of the genae; antennae slender, slightly depressed proximad. Pronotum with the cephalic margin arcuato-truncate, caudal margin obtuse- angulate, sharper in the male than in the female; median carina rather high, distinct, cut slightly cephalad of the middle by the " A?w. Mus. Civ. Stor. Nat. Genova, XXXIII, p. 123, 1893. « Canad. Entom., XXIX, p. 76, 1897. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 43 distinct transverse sulcus; lateral carinse distinct cephalad and caudad, subobsolete mesad with the constriction slight ; lateral lobes slightly over half as deep again as long, cephalic and caudal margins subparallel, ventral margin rounded obtuse-angulate. Tegmina distinctly (d^) or moderately exceeding the apex of the abdomen, but only slightly exceeding the tips of the caudal femora; costal expansion slight, ow, apex rotundato-truncate; postradial area irregularly areolate. Ceph- alic and median limbs rather slender. Caudal femora robust, inflated, the distal portion slenderer and with the pregenicular portion con- stricted, carinse strongly ribbed, pattern of the pagina distinct and regular; caudal tibiae with eleven spines laterad, longer internal spur slightly more than twice the length of the shorter, sharply curved distad but not distinctly hooked. Figs. 14 and 15. — Scyllina instabilis n. sp. Female type. Dorsal view of head and pronotum and lateral view. (X 1^.) General color ranging from ochraceous-rufous (male) to dull hazel and clay color (female). Head with a blackish postocular bar distinct in the male, slight in the female ; infraocular line very distinct in the male, absent in the female; lateral margins of the frontal costa and ventral portion of the lateral foveolse lined with blackish, more distinct in the male than in the female; face and sides of the head blotched with a darker brown in the female; eyes rather cinnamon in the male, pale Isabella color in the female; antennae of the general color. Pro- notum of the male clear ochraceous-rufous, with the lateral carinae slightly marked cephalad, the lateral lobes with a broad 44 PROCEEDINGS OF THE ACADEMY OF jFeb., blackish bar extending from the principal transverse sulcus dorsad obliquely ventrad to the ventro-cephalic angle; of the female with a dorsal cross imperfectly indicated by dull hazel on a clove brown ground, lateral lobes reddish clay color with weak longi- tudinal blotches of broccoli brown. Tegmina with the anal area obsolete maculations in both sexes, those of the female more distinct than those of the male, remainder with rather narrow rather regularly disposed transverse bars of clove brown, which are weaker distad than proximad in the female, but uniform in strength in the male. Limbs cream buff marked with clove brown ; caudal femora cream ochra- ceous-rufous dorsad in the male, not barred, in the female hazel with four rather weak transverse bars of dull brown, carinse of the lateral face dotted with blackish, the pattern of the pagina outline in brown- ish; ventral face very dark verditer blue; caudal tibise ranging from orange-vermilion (male) or Chinese orange (female) proximad to ma- roon purple (male) or burnt carmine (female), spine blackish at the tips. Measurements. <^ 2 Length of body, 24.5 mm. 33 mm. Length of pronotum, 5.4 " 6.2 " Length of tegmen 23 '• 28.5 " Length of caudal femur, 18.3 " 21.3 " In addition to the types a paratypic series of five males (September 5 and 14, 1900) have been examined. Considerable color variation is presented by this series, which is roughly divisible into three types; one (a) with the dorsum of the pronotum unicolorous, which is repre- sented by the type alone, another (h) wdth a rather pale median pro- notal line flanked laterad by blackish, represented by three individuals, and a third (c) wath a pronotal cross as in the 9 wdth the caudal margin of the pronotum as pale as the cross, represented by two specimens. The median pale line is distinctly carried to the tips of the tegmina in one of type b, while the transverse tegminal bars are rather regular and distinct, though not complete in some specimens ; a pale proximal tegminal line is present in two individuals of type h. The dorsiun of the caudal femora is distinctly barred in both of type c, and faintly in one of type b, while the lateral face shows distinct oblique bars in several specimens. The color of the caudal tibise in all males other than the type is as in the ? type. A single 6^ specimen from Sapucay, Para- guay (III 9.05; Foster, No. 195 part), in the Hebard Collection, is referred to this species. It presents a rather different appearance 1906.] NATURAL SCIENCES OF PHILADELPHIA. 45 when compared with the typical series, but the differences are wholly of color. The dorsum of the pronotum is to a large extent velvety black, broken at the sulci, while the lateral carinse are marked as in specimens of type c, but darker, with the median longitudinal bar present. Scyllina picta (Bniner). 1900. P[lectrotettix] pidus Bruner, Ace. Genera Spec. Locusts Argent., p. 37, fig. 13. [Cordoba and Sante Fe Provinces, Carcarana and Rosario, Argentina.] A topotypic series of three males and three females from Carcarana, collected by Bruner, have been examined. This species is only known from Cordoba and Sante F6 Provinces, Argentina. Scyllina brunneri (Giglio-Tos). Sapucay, Paraguay. February 13, 1905. 1 d^. (Foster, Hebard Coll.) This specimen is considerably darker than any strongly marked specimen of this species seen. The whole insect is quite dark except for the light femoral bars, but the pattern is still distinctly visible, the medio-dorsal and lateral bars being present, and the 'tegminal streak clear apple. green. The size of this specimen is slightly greater than Matto Grosso specimens. Scyllina pratensis (Bruner). 1904. [Plectrotettix] fn-atensis Bruner, Biol. Cent.-Amer., Orth., II, p. 100. [Pernambuco, Brazil.] Paratypic specimens, a d^ and 9 . of this species, loaned by Prof. Bruner, have been examined. They differ from brasiliensis in the slightly smaller size, rather slenderer form, less divergent carinae of the pronotum and the more acute fastigium. The color of the caudal tibiae is of little diagnostic value, as in other species of the genus, the male in hand having them wholly red, the female ochraceous with the dorsal surface purplish distad and reddish proximad. Measurements of paratypcs. & 9 Length of body, 20.2 mm. 26.5 mm. Length of pronotum, 4 " 5 " Length of tegmen, 18.5 " 23.5 " Length of caudal femur, 15 " 17.5 " Scyllina gregaria (Saussure). 1861. St[enobothrus] gregarius Saussure, Revue et Magasin de Zoologie, 2e ser., XIII, p. 318. [St. Thomas; Haiti.] 1903. Plectrotettix gregarius Rehn, Trans. Amer. Entoin. Soc, XXIX, p, 46 PROCEEDINGS OF THE ACADEMY OF [Feb. 133. [Utuado, Arroyo, Bayamon and Mayaguez, Porto Rico ; Culebra and Vieques Islands.] A ? from Vieques Island has been examined in this connection. The species is closely related to pratensis, but differs in the narrower space between the eyes and the greater interantennal constriction of the frontal costa. It is interesting to note that the known ranges of these very closely related species are separated by a distance of over two thousand miles, in the land areas of which, as far as at present known, no species of the genus is found. Scyllina brasiliensis (Bruncr). 1904. [Pledrotettix] brasiliensis Bruner, Biol. Cent.-Amer., Orth., II, p. 100. [Southern Brazil.] Sapucay, Paraguay. 1 cJ^, 1 $. Prof. Bruner's Coll. Sapucay, Paraguay. January 28, February 2-27, March 6-10, 1905. 11 d'd', 20 9 9 . (Foster, Hebard Coll.) Sao Paulo, Sao Paulo, Brazil. Sep- tember 1-14, 1900. 4 c^c^, 6 ? 9 . (Hempel, A. N. S. P.) The Sapucay specimens loaned by Prof. Bruner are marked as types, although the locality southern Brazil is given with the original brief description. The variation in the intensity of the markings of this species is quite considerable. The males appear to be, for this genus, rather uniform in coloration, and the variation in size is not very great. The females, however, vary greatly in the amount and shade of green coloring, some having the dorsal face of the caudal femora, head and the greater portion of the pronotum and pleura rather pale apple green, while others have these parts mottled and overlaid more or less strongly with dull brown. In a few specimens the green, except the tegminal bar, is wholly replaced with a shade of brown, but in such cases the pattern remains the same. In the material examined but very few specimens have the characteristic pattern obscured and even in those cases it is not obliterated. As a series the females are quite uniform in size, one Sapucay female, however, being quite large. This latter individual is also very deeply colored. The measurements of Bruner's Sapucay specimens and the large ? mentioned above are as follows : Bruner. Bruner. Large. 6" 9 9 Length of body, 23.5 mm. 30.3 mm. 34.5 mm. Length of pronotum, 4.5 " 6 " 6.7 " Length of tegmen, 21 " 26.8 " 29.5 " Length of caudal femur, .... 15 " 20 " 22.3 '' 1906.] NATURAL SCIENCES OF PHILADELPHIA. 47 The records of this species given above cover the known region, from Sao Paulo, Brazil, to western Paraguay. Scyllina conspersa (Bruner). 1904. [Plectrotettix] conspersus Bruner, Biol. Cent.-Amer. Orth., II, p. 100 [No locality.] Sapucay, Paraguay. December 17, 1901; December 16, 1904. January 26 and 28, February 8-15, 1905. 9 d'c?, 11 ? ? . (Foster, Hebard Coll.) Through the kindness of Prof. Bruner I have before me the type of this species, a 9 from Sapucay, Paraguay. From the material in hand it appears that this species has several color forms, as in brunneri and other species. The type has the dorsal surface of the head and pronotum uniform wood brown without cruciform markings or longi- tudinal median stripe, and a majority of the females and several of the males examined belong to this type. Several males and females have the dorsum dark with weak, but distinct, cruciform pronotal markings and a longitudinal light bar of variable prominence, in some cases absent. In about half the specimens the heavy blackish markings on the lateral lobes of the pronotum and caudad and ventrad of the eyes are distinct, while the dorsal bars of the caudal femora are dis- tinct in the greater number, weak and interrupted in several others mcluding the type, dorso-lateral carina of the femora marked ventrad with blackish in all specimens, very distinct in some, broken and faint in others, the genicular lobes also blackish. Measurements of the type. Length of body, 34.7 mm. Length of pronotum, 6.9 " Length of tegmen, 32.2 Length of caudal lemur, 22.5 Scyllina sufFusa Rehn. Scyllina varipes (Bruner). 1905. Plectrotettix varipes Bruner, Entom. News, XVI, p. 214. [Sapucay, Paraguay.] Sapucay, Paraguay. ]\Iarch 5 and 15, 1905. 4 6^6^, 5 9 9. (Foster, Hebard Coll.) These specimens have been compared with typical individuals kindly loaned by Prof. Bruner. This species is very richly colored and one of the more easily recognized forms, the distinct femoral bars and the entirely black genicular region being quite striking in all the specimens examined. Considerable superficial resemblance exists to *S. conspersa, which is found in the same locality, but the angle of the 48 PROCEEDINGS OF THE ACADEMY OF [Feb., face and the shorter and heavier caudal femora, as well as the color of the caudal tibise, serve to separate it without difficulty. In the case of this species and S. conspersa the colors of the tibiae appear to be constant, all the specimens of conspersa examined having the distal section dark bluish, while the deep rich crimson of the distal two-thirds of varipes is clear and uniform in the whole series. Scyllina smith! Rehn. Scyllina borellii Giglio-Tos. Scyllina schistocercoides Rehn. Scyllina viator ia (Saussure). Specimens recorded by the author as this species from Texolo, La Joya, San Luis Potosi and Alta Mira, Tamaulipas, when re-examined in the light of recent work on the Mexican and Central American forms of the genus, prove to represent true viatoria. The males have the dorsal face of the caudal femora distinctly barred, while the females have these parts almost uniform green. Scyllina calida (Bruner). 1904. Plectrotettix calidus Bruner, Biol. Cent.-Amer., Orth., II, p. 101 [Cuernavaca, Morelos and Guerrero, Mexico, Nicaragua, Costa Rica.] Previous records of the author of viatoria from Cuernavaca, Uruapan, Gualajara, Zapotlanejo and Zapotlan should be referred to this species. The range of variation in size and coloration in this form is very great, some individuals having the colors weak and poorly defined, while other individuals from the same locality are quite richly colored. The width, number and intensity of the transverse bars of the tegmina also varies greatly as in some individuals they are as broad as the inter- vening sections, while in others they are broken, imperfect and rather pardaline in character and distribution. Scyllina excelsa (Bruner). 1904. Plectrotettix excelsus Bruner, Biol. Cent.-Amer., Orth., II, pp. 101-102. [Tlalpam and Tacubaya, Mexico.] A male and two females of this species from Tacubaya show that this is probably the most striking Mexican species. These individuals have been previously recorded by the author as viatoria. The rather short tegmina, more robust form and heavier caudal femora will assist in separating this from the allied species. These specimens show practically no green, the light shades being ochres. PSOLOESSA Scudder. 1875. Paoloessa Scudder, Proc. Boston Soc. Nat. Hist., XVII, p. 512 Included P. texana, jerrughiea and maculipemiis Scudder, of which maculipennis can be considered the type. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 49 Psoloessa maculipennis Scudder. 1875. Psoloessa maculipeiinis Scudder, Proc. Boston Soc. Nat. Hist., XVII, p. 513. San Luis Potosi, Mexico. August 5, 1903. 1 9 . (M. E. Hoag. A. N. S. P.) This species has l^een recorded from Texas, New i\Iexico, Arizona and Cahfornia. Psoloessa buddiana Brmier. 1889. Psoloessa Buddiana Bruner, Proc. U. S. Nat. Mus., XII, p. 61, t. I, fig. G. [Carrizo Springs, Southwest Texas.] Zapotlan, Jahsco, Mexico, July 7, 1902. d", 9 . (C. H. T. Town- send, A. N. S. P.) These specimens are referred here with some Httle doubt. The species has been recorded from Montelovez, Coahuila. STIRAPLEURA Scudder.^s 1876. Stiraplcura Scudder, Ann. Rep. Chief of U. S. Engineers, 1876, pt. 3, p. 510. Type. — Stirapleura decussata Scudder. Specimens of all the South American forms of this genus have been examined and the species are here listed to show their relationship. Stirapleura variabilis Bruner. Stirapleura signatipennis (Blanchard). 1851. (Edipoda signatipennis Blanchard, in Gav, Hist. Fis. y Polit. de Chile, Zool., VI, p. 79. [Coquimbo, Chili.] Penco, Chili. December, 1903. 7 9 9. (C. S. Reed, A. N. S. P.) Concepcion, Chili. 19. (C. S. Reed, A. N. S. P.) These specimens fully agree with Blanchard's original description. This is the largest species of the genus, exceeding even the North American S. decussata. An average 9 measures as follows: Length of body, 25 mm. Length of pronotum, 4.8 " Length of tegmen, 19 " Length of caudal femur, 16.5 " Stirapleura bruneri n. n. 1900. Sltirapleura] signatipennis Bruner (not QiJdipoda signatipennis Blanchard), Gen. Sp. Locusts Argent., p. 34. [Argentina from the Pampa Central and extending into Uruguay to the eastward.] This species is quite distinct from signatipennis Blanchard, as six ^* Saussure's Stenobothrus chilensis {Revue et Magasin de Zoologie, 2e ser., XIII, p. 319, 1861) is no doubt a member of this genus, and possibly the male of signatipennis, the female being the only sex known to the author. The size given by him (lengtli with elytra 15 mm.) is much too large for the male of humilis. 50 PROCEEDINGS OF THE ACADEMY OF [Feb., specimens from Carcarana, Argentina, determined by Bruner show. The Argentine form is smaller, with the lateral carinse of the pronotum more constricted and the coloration more variable than in signatipennis. This is apparently one of the species on which Brunner erected his insufficiently characterized Pseudostcnironotus,'^'' as specimens received from Saussure from Buenos Ayres bear the manuscript name Pseudo- stauronotus occidentalis. Stirapleura humilis (Blanchard). 1851. Q^dipoda huynilis Blanchard, in Gay, Hist. Fis. y Polit. de Chile, Zool., VI, p. 79. [Coquimbo, ChiU.] Penco, Chili. November and December, 1903, and February, 1904. 17 d'c^, 16 9 ? , 5 immat. (C. S. Reed, A. N. S. P.) Concepcion, Chili. November and December, 1903, and February, 1904. 20 d'd', 7 9 9,3 immat. (Reed, A. N. S. P.) Coronel, Chili. January, 1904. 1 d". (Reed, A. N. S. P.) Longuen, Chili. February, 1904. ^2d(^, 4 9 9. (Reed, A. N. S. P.) Guiltio, ChiU. December, 1903. 2 d" (^ , 6 9 9 . (Reed, A. N. S. P.) Lota, Chili. October, 1903. 1 (^ . (Reed, A. N. S. P.) This large series exliibits a very great amount of individual variation in size and color. The females range in total length from 13 to 20 millimeters, the males varying proportionately. The color ranges from a mottled black-brown and gray with femoral bars and weakly lateral carinse of the pronotum, to another with the general color pale with broad dark lateral bars on the dorsum of the head and pronotum with a broad pale median bar, a longitudinal dark bar on the tegmina flanked toward the costa by a pale green line, and the femora with bars and suffused along the dorso-lateral carinse with blackish. The latter type has the markings of the lateral lobes of the pronotum quite distinct and the genicular regions of the caudal femora and tibise are paler than in the other type, in which these portions are blackish. These differences are irrespective of locality and date, and are con- nected by dozens of intermediates. This species appears to fit Blanchard's form, some specimens agree- ing very well with the rather vague description. It is apparently a common species. Stirapleura pallida Bruner. Stirapleura obscura Bruner. The position of this species is a little doubtful. In some respects it resembles hnnieri, but its general characters approach brunnea. Stirapleura brunnea Rchn. " Ann. Mus. Civ. Slor. Nat. Genord, XXXIII, p. 123, 1893. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 51 March 6. The President, Samuel G. Dixon, M.D., in the Chair. One hundred and twenty-seven persons present. The death of William W. Jefferis, a member, February 24, 1906, was announced. The reception of a paper entitled "Phylogeny of the Races of Volutilithes petrosus," by Burnett Smith, Ph.D. (February 23), was reported. Dr. William P. Wilson made a communication on the Philippine Islands, their resources, their inhabitants and present progress toward independence. (No abstract.) March 20. Mr. Arthur Erwin Brow'n, \'ice-President, in the Chair. Sixty-nine persons present. Miss Mary S. Holmes made a communication on the Dos Pueblos Caiion, California. (No abstract.) Francis William Rawle was elected a member. The following were ordered to be printed : 52 PROCEEDINGS OF THE ACADEMY OF [March, PHYLOGENY OF THE EACES OF VOLUTILITHES PETROSTJS. by burnett smith, ph.d. Introduction. In a recent article * the author has described some of the shell char- acters which, occurring in the later stages of Gastropod ontogeny and phylogeny, mark the gerontic or senile condition of the individual or of the race. It was seen that these senile features may at times be as useful to the student of phylogeny as those of the earlier stages, which latter have always received the more attention. In the paper referred to, the attempt was made to show that the modern Fulgur carica of our New Jersey coast is the descendant of a main F. carica stock, which originated in Miocene time ; and that such forms as F. maxivium and its allies, which had heretofore been regarded as ancestral by Grabau ^ and others, are in reality not so at all, but senile offshoots which at an early period in the history of the stock diverged from the main line of descent. The paucity of individuals representing the forms considered ancestral, and the extreme abundance of the speci mens of senile forms, was a source of some embarrassment to the author. It is therefore with considerable satisfaction that I am now able, though in another genus {Volutilithes), to present a series in which the forms comprising the main ancestral stock are nearly as plentiful as those which represent the senile offshoots. The history of certain of the species of Volutilithes can be well traced in the Eocene of our Gulf States, but nowhere better than in Alabama, where the fine sections along the Alabama and Tombigbee rivers have furnished abundant material for the study of the geological and geo- graphical distribution of these forms. The races and species studied range from the Matthew's Landing horizon in Alabama to the Jackson horizon in Mississippi, and therefore represent a fair proportion of the phylogenetic units through nearly the whole of the Eocene. They are likewise restricted to a small geographical area, and the geological sections in which they appear are unusually good. . t * Senility among Gastropods, Proc. Acad. Nat. Sci. Phila., May, 1905. 2 Studies of Gastropoda, II — Fulgur and Sycotypus, Am. Nat., Vol. 37. 1006.] NATURAL SCIENCES OF PHILADELPHIA. 53 The material at my disposal is very full and complete,^ and has enabled me to study large series of individuals. These series in each case represent the forms in the proportions in which they occur in the field, and therefore offer every opportunity for obtaining the average racial characters in any particular locality or horizon. Morphology. The genus Volutilithes is widely distributed throughout our Southern Eocene. At some localities several sharp and distinct species occur, but at most of the localities Volutilithes is represented by one species and by a particular race of that species. The term race is used in place of variety, for in these assemblages the individuals frequently differ widely. There is, however, always an average of characters which stamps the stage of evolution which the race has attained. Of course, extreme specimens often have more of the features of some other race than of the race to which they belong, but these individuals are always few; and there is never any difficulty in recognizing the general race characters for any particular geographical point, and the phase of phylogenetic development found in the faunule at that point. In cases of this sort, the introduction of varietal and specific names is to be deprecated, and the adoption of the system of race analysis is strongly urged by the author. In the forms under consideration, the first two or three whorls are smooth and rounded, constituting the Smooth Stage. The first orna- mental feature to appear on the smooth, rounded whorl is the trans- verse rib, that is, a slight elevation on the whorl which runs across it from suture to suture. These early ribs are invariably curved slightly, and each one is simple and uniform from suture to suture. The curved ribs persist as a rule for about a quarter or a half of a whorl, or even for a much less space; in fact, sometimes we have only one or two of them. This Curved Rib Stage is short but remarkably constant, and though occasionally much suppressed, has been found in every species and race dealt with in this paper. The curved ribs, after about one-third of a whorl, change abruptly into the straight ribs of what has been designated the Cancellated * ' It forms part of the Isaac Lea Collection of Eocene MoUusca at the Academy of Natural Sciences of Philadelphia, and was brought together largely by Mr. Charles W. Johnson, now of the Boston Society of Natural History. The col- lections are a tribute to his skill and perseverance as a collector, and are an example of what museum research collections should be. * See Dall, Trans. Wag. Inst., Vol. Ill, p. 68. 5i PROCEEDINGS OF THE ACADEMY OF [March, Stage, which begins by the appearance of two swelhngs or tubercles on each rib. One of these tubercles is near the suture, and the other at the position which is later occupied by the shoulder angle. They will be referred to as the suture and shoulder tubercles respectively. These two tubercles are often of nearly equal size at the start, though frequently the suture tubercle is the larger. Tubercles other than these two mentioned are seldom visible on an early rib ; but when a sufficiently small shell is obtainable, other smaller tubercles can be seen upon each rib, and the tubercles diminish in size quite regularly from the suture tubercle downward. Each tubercle is connected with the corresponding tubercle of the next rib by a faint spiral, the first appearance of the definite cancellation which later becomes more pronounced. In most specialized forms the shoulder angle develops shortly after the appearance of the tubercles. The cancellated condition is found more or less well developed in all the different races. In primitive species ^ it may persist as a con- stant feature to the end of the individual's life; but in most forms it covers only a few whorls and is more variable than the preceding Curved Rib Stage, in that certain of its features change as the shell grows. These changes are gradual, but they render the end of the Cancellated Stage much less definite than its beginning. They may be briefly summarized as follows : The tubercles, with the exception of the shoulder tubercle, usually tend to degenerate, and soon disappear, though their corresponding faint spirals may sometimes persist. The shoulder tubercle, on the contrary, becomes stronger and sharper. First the ribbing above the shoulder disappears, and later the ribbing below the shoulder shortens, and with its practical disappearance what might be termed the Spiny Stage is inaugurated. In the Spiny Stage the shoulder tubercle is now sharp and spine-like. Other tubercles have disappeared and the spirals are, as a rule, re- stricted to the region of the anterior siphon. Ribbing has practically gone, but very short rudiments may sometimes still be present below the shoulder spines. The commencement of this stage.is never sharp, for it is the result of gradual change from the preceding one.® It is 5 In A771. Nat. for 1902, Vol. 36, p. 926, Grabau says: "It is perhaps not too much to say that in the majority of the larger phyletic series, except those highly specialized, the radicle is a smooth, round-whorled form, succeeded by types in which the adults are ribbed, and later cancellated, after which progressive modi- fication may be carried further." 'No morphological distinction is made between spine and tubercle. "The writer considers that the difference between the small rounded tubercle of the early whorls and the large, sharp spine of the later whorl is one merely of degree. 1906.] NATURAL SCIENCES OF PHILADELPHIA. ^5 not reached in all species, but occurs usually as the end term of normal progressive development in the ontogeny, and of slow and even evolu- tion in the phylogeny. The Spiny Stage, then, may sometimes be present as a well-defined stage, but it may also be-accompanied by extreme individual variation and by features of senility. These latter characterize the Senile Stage, and may follow regularly after the Spiny Stage; but, as stated above, they not infrequently are thrown back by acceleration, and found together in the same whorl with the features of the Spiny Stage. The Senile Stage, whether following normally after the Spiny Stage or associated with it, is always easily distinguished. In the forms under consideration its most important features are the following : ^ 1. Tendency for the shoulder spines to pass into a shoulder keel. 2. Irregularity of growth lines. 3. Thickening of the shell. 4. Encroachment of the anal siphon upward on the preceding whorl. 5. Protrusion of the mantle m the region of the anal siphon, pro- ducing a smoothing of the preceding whorl by a shelly overgrowth. These five characters are all sometimes found together in one in- dividual or race, but, as a rule, only two or three of them are so asso- ciated. In studying the development of the shell features, each whorl has been taken up in detail, and its particular ornamental characteristics noted. Of course, this system is more or less arbitrary, and at times K3 Fig. 1. — Line shows position taken for the end of the first whorl. inaccurate, on account of the acceleration of features in certain in- dividuals. In the following study, the writer has endeavored to get the average characters for a certain whorl, except where individual variation is great. In this latter case the variations are noted. The word spine is therefore used, throughout this paper, simply in a descriptive sense, and implies no difference from the early tubercle, except that it is larger and sharper. Both are produced by the same fold in the mantle, and every gradation between the two is observable." Smith, Burnett, Proc. Acad. Nat. Sci. Phila., May, 1905, p. 347. This opinion is quite different from that held by Grabau (see Am. Nat., Vols. 36, 37). ^ For a discussion of senility see Smith, Burnett, Proc. Acad. N^at. Sci. Phila., May, 1905. 56 PROCEEDINGS OF THE ACADEMY OF [March, In counting the whorls, it has been attempted to have the end of the first whorl come in about the position shown in the figure. It is impossible to make an absolutely exact count, but if care is used there will seldom be much of an error in the last whorl. As stated above, the method is purely arbitrary, but it seemed to the writer to be the only practical way in which the statistics of the development of orna- mental features could be arranged. Fig. 2. — Volutilithes limopsis Conrad. Matthew's Landing, Ala. Apex enlarged, showing smooth stage followed by the curved rib stage, which begins in the latter part of whorl four. Whorl five shows the cancellated stage. Length = 1.6 mm. Volutilithes limopsis Conrad. Locality — Matthew's Landing, Ala. This species was originally described by Conrad in Jour. Acad. Nat. Sci. Phila., 2d Series, Vol. IV, p. 292, PI. 47, fig. 24. Whorls 1, 2, 3 — Smooth and rounded. Whorl 4 — Smooth and rounded, except toward the end, where the curved rib stage may start. Whorl 5 — The curved rib stage often commences early in five, per- sisting as a rule for about a quarter of a whorl. The cancellated stage follows abruptly with its straight, tubercled ribs. These tubercles are largest at the suture, decreasing regularly in size anteriorl}^ on each rib. Each tubercle is connected with the corresponding tubercle of the next rib by a faint spiral. A shoulder angle is never developed. Whorls 6, 7 — Characterized by the cancellated stage and much like the latter part of five. Whorl 8 — Characterized by the cancellated stage and much like six and seven. The tubercles nearest the suture are the strongest, and they diminish in size regularly as the rib is followed anteriorly. The ribs are also less prominent anteriorly, disappearing entirely on the branchial siphon. Spirals cover the entire whorl. Whorls 9, 10 — jNIuch as in eight, the cancellated stage still persisting. li)06.] NATURAL SCIENCES OF PHILADELPHIA. 57 The tubercles are a little sharp, and the protruded mantle shows a very slight tendency to smooth the preceding whorl — a tendency which in later forms results in a senile feature. Remarks. — This form possesses no stage later than the cancellated stage and never has a shoulder angle. The suture tubercle is through- out life the dominant one. The specimens are remarkably uniform, no perceptible individual variation occurring. The smooth stage is very long. From its primitive characteristics, and its geological posi- tion near the base of the Eocene, the wTiter has no hesitation in assign- ing to it an ancestral position among the races and species dealt with in this paper. Fig. 3. — Volutilithes rugatus Conrad. Matthew's I>anding, Ala. Adult indi- vidual. Length = 44 mm. Volutilithes rugatus Conrad. Locality — Matthew's Landing, Ala. This species was originally described by Conrad as Volutilithes rugata in Jour. Acad. Nat. Sci. Phila., 2d Series, Vol. IV, p. 292, PI, 47, fig. 32. Whorls 1,2, 3 — Smooth and rounded. Whorl 4 — Smooth and rounded, except near the end, where in many individuals the ciu'ved rib stage begins. Whorl 5 — The curved rib stage usually occupies the latter part of four and early part of five. Its ribs are wide apart and it persists for about a quarter of a whorl. It is followed by the straight, tubercled ribs of the cancellated stage. The two upper tubercles are the largest, those below becoming fainter anteriorly. The ribs die away anteriorly on the branchial siphon. Each tubercle is connected with the corre- sponding one of the next whorl by a spiral. 58 PROCEEDINGS OF THE ACADEMY OF [March, Whorls 6, 7, 8 — All characterized by the cancellated stage. Changes, however, take place during these whorls. At first the ornamentation, is much like that of the cancellated stage in V. limopsis, but later the ribs become relatively farther apart and broader (in a spiral direction). At the same time many fine riblets are introduced between the pri- mary ribs. The tubercles are largest near the suture and become smaller as the ribs die away anteriorly. There is never a distinct shoulder angle. The whorls are covered with spirals. Whorl 9 — Individual variation now sets in. In some specimens the whorl is much as in eight. In others the rib is swollen into a very rounded, transversely long angle which hardly occupies the position of the shoulder angle of later forms, and in addition the shells become thick. In all individuals the mantle protrudes in the region of the anal siphon, and tends to smooth the preceding whorl with a shelly overgrowth. This latter is never extreme. Wliorl 10 — Much as in nine, though here the forms with swollen ribs predominate. In many specimens the shells become thick, the growth lines irregular, and the smoothing of the preceding whorl by the shelly overgrowth is quite well marked, though it never becomes extreme as in some forms of later time. Remarks. — This species in its earlier stages of growth closely re- sembles V. limopsis. It differs radically, however, from that form with the progress of its ontogeny. Though its individuals vary greatly, it is nevertheless entitled (in the opinion of the author) to rank as a distinct species. In its later whorls there is great individual variation, accompanied by senile features. These latter, though never extreme, are well marked. They comprise protrusion of the mantle in the region of the anal siphon, producing a smoothing of the preceding whorl by a shelly overgrowth; thickening of the shell, and irregularity of growth lines. There is no spiny stage. V. rugatus is considered to be a short senile offshoot from V. limopsis. In the absence of intermediate forms this evolution must have taken place either at some other locality or at some geologically earlier date. Volutilithes petrosus Conrad. The forms described at the present day under this name represent an assemblage of races. If every gradation of evolutional develop- ment did not exist among them, many of these races would doubtless now be regarded as distinct species. In fact, there are several dis- carded specific names which were used in this way by Conrad and Lea. Each race is an assemblage of individuals, showing more or less in- 1906.] NATURAL SCIENCES OF PHILADELPHIA. 59 dividual variation, yet nevertheless possessing on the average definite racial characters. In other words, the majority of specimens from any one locality or horizon show a certain stage of evolution. Some of the specimens may show developmental features characterizing either a more primitive or a less primitive race; but these specimens are few, and are not typical of the average of individuals from the locality or horizon to which they belong. The races of V. petrosus are known to range from the A^anafalia beds to the Jackson beds inclusive. The original specific description '^ is accompanied by a figure of a member of the Claiborne race. Though this particular race is perhaps not the most typical expression of the V. petrosus main stock, the writer feels that no better method can be employed than to use the name V. petrosus to designate the assemblage of races under discussion. Several of the races differ in the development of senile characters from the V. petrosus main stock, but there is every gradation between these and the representatives of the main stock. On account of this gradation the senile forms are included as races of V. petrosus, in spite of the fact that their old individuals differ greatly in appearance from those of the more normal type. The young of all the races, senile and otherwise, are remarkably uniform and constant. The early whorls indicate clearly that they are all descended from a cancellated ancestor, and bear a strong re- semblance, especially in the cancellated stage, to the characters of V. limopsis. In the absence of any record of a V. petrosus being found below the horizon of the Nanafalia beds, the author believes that it is perfectly safe to conclude that all these races in question are de- scended from the V. limopsis of the Matthew's Landing horizon. Volutilithes petrosus Conrad. Nanafalia Race. Locality — Nanafalia, Ala. The only specimen known to the writer from this horizon is not suf- ficiently well preserved to admit of its accurate study whorl for whorl. It is, however, a perfectly normal specimen of moderate size and ex- presses the general characters of the V. petrosus main stock. It is about the size of the eight-whorled individuals of the Gregg's Landing or the Jackson races. When it is placed in line with such specimens, the three shells exhibit very few differences. This Nanafalia individual * Valuta petrosa Conrad, "Fossil Shells of the Tertiary Formations of North America. ' ' 60 PROCEEDINGS OF THE ACADEMY OF [March, has spines in its last whorl which make it resemble the Jackson race quite closely. In this respect it is nearer the Jackson specimens than the Gregg's Landing specimens. Its mantle perhaps smooths the preceding whorl a trifle more than does the mantle in the Jackson race. However, in its general features, it is very close to the Jackson form. The amount which the mantle has smoothed the preceding whorl is about equal to that found in shells of the same size from Gregg's Landing. Volutilitlies petrosus Conrad. Gregg's Landing Race. Locality — Gregg's Landing, Alabama River, Ala. Whorls 1, 2, 3 — Smooth and rounded. Whorl 4 — The curved rib stage begins early in four. It persists for a quarter or a third of a whorl, but changes abruptly into the can- cellated stage with its straight, tubercled ribs. The shoulder and suture tubercles are of about the same size at first. By the end of four the shoulder angle is quite well developed. Whorl 5 — Characterized by the cancellated stage and much like the latter part of four. Suture and shoulder tubercles are of about the same size. The shoulder angle is well developed. Tubercles are connected by fine spirals. Whorl 6 — Characterized by the cancellated stage, but changes occur. The shoulder tubercle is becoming stronger, while the suture tubercle is getting weaker. The ribbing above the shoulder angle is also becoming weaker. Whorl 7 — Still characterized by the cancellated stage, but here the suture tul^ercle has disappeared and is only represented by its spiral. The ribbing above the shoulder angle has gone, though it is still well represented below the shoulder angle. Whorl 8 — Still characterized by the cancellated stage and much like whorl seven. Below the shoulder angle the whorl is covered with fine spirals. There is a very slight smoothing of the preceding whorl by the protruded mantle, for the end of the cancellated stage is not sharp like that of the curved rib stage. Whorl 9 — From here on individual variation is great. Some forms have a whorl much as in eight, with the ribbing still well developed below the shoulder, though the shoulder tubercles are more spine- like. In other specimens the ribbing below the shoulder is gone, the shoulder spines are more extreme, and the anal siphon shows a tend- ency to encroach upward on the preceding whorl. In addition, these 1906.] NATURAL SCIENCES OF PHILADELPHIA. 61 latter forms have a heavier shell, and the shelly smoothing of the pre- ceding whorl by the protruded mantle is more marked. Whorl 10 — Here again we have every gradation between forms char- acterized by the spiny stage, in which the preceding whorl has been only slightly smoothed by the mantle, and forms which show a moder- ate development of senile features. These latter show the tendency for the spines to pass into a shoulder keel. They also possess a thick shell with irregular growth lines, the anal siphon encroaches upward, and the preceding whorl is smoothed by the shelly overgrowth of the protruded mantle. Though these latter forms possess senile features, these features are only slightly developed. They are, in fact, just the beginnings of those characteristics which become so extreme in some of the races of later time. Remarks. — This race occurs in the Gregg's Landing marl. The greater number of individuals of the Gregg's Landing race have a perfectly normal ontogeny, and in their later whorls have the spiny stage well marked. These individuals belong therefore to the V. petrosus main stock, and as they form a majority of the assemblage, they determine the position of the race to be on the main stock. A minority of individuals, however, show senile featiu'es. These senile features are not well marked, but they indicate that here in this race was inaugurated the tendency which resulted in the development of the senile races of later time. In other words, we have here the point where a senile offshoot is starting to branch from the main ancestral stock. The marks of this tendency are but slight, and the individuals are few; yet, nevertheless, the tendency is plainly recognizable. Volutilithes petrosus Conrad. Bell's Landing Race. Locality — Bell's Landing, Alabama River, Ala. Whorls 1, 2, 3 — Smooth and rounded. Whorl 4 — The curved rib stage commences early in four and per- sists for about a quarter of a whorl. It is followed by the straight, tubercled ribs of the cancellated stage. The suture and shoulder tubercles are about equal in size, and the shoulder angle soon becomes well marked. Spirals connect the tubercles. Whorl 5 — Characterized by the cancellated stage and closely re- sembles the latter part of foiu'. Whorl 6 — Characterized by the cancellated stage. Suture tubercles grow w^eaker, shoulder tubercles grow stronger. Ribbing above the shoulder also grows weaker, but it is well marked below the shoulder. Whorl 7 — Much as in six. 62 PROCEEDINGS OF THE ACADEMY OF [March, Whorl 8 — Still characterized by the cancellated stage, but the suture tubercles and the ribbing above the shoulder have now gone. Shoul- der tubercles are stronger. Ribbing below the shoulder is weaker. The beginning of the tendency to form a shelly overgrowth is seen, and the cancellated stage comes to an end. Whorl 9 — Ribbing is about gone and the shoulder tubercles have become spines. In other words, the spiny stage has set in. In most individuals, however, the senile characters are also mingled with those of the spiny stage. The shell gets thicker, and the mantle protrudes in the region of the anal siphon, and smooths the preceding whorl with a shelly overgrowth. This latter, though not extreme, often covers the spines of eight, which, however, are visible through it. Whorls 10, 11, 12 — In these whorls we simply have an exaggeration of the characters of nine. The features of the spiny stage and the marks of senility are found together in the same whorl. Toward the end senile characters are quite well marked. The shell becomes thick, the growth lines irregular, the anal siphon encroaches upward, and the shoulder spines tend to pass into a shoulder keel. In addition, the shelly overgrowth produced by the protruded mantle is well marked, and in some specimens may be quite thick. As a rule, how- ever, the spines of the preceding whorl can be counted through the shelly envelope. Remarks. — ^This race occurs in the Bell's Landing marl. Though occasionally a rare specimen may nearly approach the V. petrosus main stock in its general features, the vast majority of individuals show such a development of senile characters that it is necessary to consider the race to be a senile offshoot from the main stem. These senile marks though typical are not highly developed. This race unquestionably sprang from the senile individuals which occurred as a minority in the otherwise normal, geologically earlier Gregg's Landing race. Volutilithes petrosus Conrad. Wood's Bluff Race. Locality — ^Wood's Bluff, Tombigbee River, Ala. Whorls 1, 2, 3 — Smooth and rounded. Whorl 4 — The curved rib stage appears early in four and persists for about a quarter of a whorl. It is followed by the straight, tubercled ribs of the cancellated stage. The suture and shoulder tubercles are about equal in size. A shoulder angle soon develops. The tubercles are connected by spirals. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 63 Whorl 5 — Characterized by the cancellated stage and closely resem- bling the latter part of four. Whorl 6 — Characterized by the cancellated stage. Ribbing above Fig. 4. — Volutilithes petrosus Conrad. Wood's Bluff Race. Wood's Bluff, Ala. Apex enlarged. The first three whorls are characterized by the smooth stage. In whorl four the passage from the curved rib stage to the cancellated stage can be seen. Whorl five shows tlie cancellated stage. It will be noted that the stages are all more accelerated than in text fig. 2. Length = 1.. 5 mm. the shoulder growing weaker, though well marked below the shoulder. Suture tubercles weaker, shoulder tubercles stronger. Whorl 7 — Much as in latter part of six. Whorl 8 — Still characterized by the cancellated stage, but the suture tubercles are gone, and are represented only by their spiral. Ribbing above the shoulder has disappeared, though it is still present below the shoulder. The ribbing below the shoulder is, however, weaker. The tendency to form a shelly overgrowth is plainly seen. Though its end is not w^ell marked, the cancellated stage comes to a close with this whorl. Wliorl 9 — The spiny stage usually begins with this whorl. It is, however, accompanied by well-developed senile features, the most marked of wliich is the shelly overgrowth, which is produced by the protrusion of the mantle on to the preceding whorl. Whorls 10, 11, 12 — Here we have simply an exaggeration of what occurs in nine. The shelly overgrowth is extreme, obscuring the ornamentation of several of the earlier whorls. The shell becomes thick, the growth lines irregular, the anal siphon encroaches upward, and the shoulder spines tend to pass into a shoulder keel. Remarks. — This is typically a senile race. No individuals of more than eight whorls ever approach in their general appearance the V. petrosus main stock. All specimens exhibit senile characters. These latter are much more extreme than those found in the geologically 64 PROCEEDINGS OF THE ACADEMY OF [Mai'Ch, earlier Bell's Landing race. It has unquestionably been evolved from the Bell's Landing race by an exaggeration of senile characters. Volutilithes petrosus Conrad. Hatchetigbee Bluff Race. Locality — Hatchetigbee Bluff, Tombigbee River, Ala. The series of young individuals at hand is so small, and the ghelly overgrowth produced by the protruded mantle is so great in the old individuals, that a detailed study of the specimens, whorl for whorl, is not practical. These forms are extraordinarily senile, and have been derived from the geologically earlier Wood's Bluff race by an exaggeration of the senile features. The shoulder spines tend to pass into a shoulder keel, the shell is thick, the growth lines irregular, the anal siphon encroaches upward, and the preceding whorl is encumbered with the shelly overgrowth. This last feature is very extreme. In addition the race is, on the whole, a dwarfed race, the full-grown individuals being smaller than those of the earlier less senile races. This race is the end term of the senile offshoot which diverged from the V. petrosus main stock in Bell's Landing time. The author knows of no authentic abnormal senile specimen from beds higher than the Hatchetigbee horizon in Alabama or Mississippi.^ Volutilithes petrosus Conrad. Yellow Bluff Race. Locality — Yellow Bluff, Alabama River, Ala. Unfortunately the material at hand is not sufficiently well preserved to give a detailed description, whorl for whorl. Nevertheless, there is no difficulty in making out the following points. The great majority of individuals possess, in general, the characters of the V. petrosus main stock, and have arrived at little more than that evolutional stage which is typified by the Gregg's Landing race. Though a few individuals exhibit senile characters, these latter are only very slightly developed, and we are undoubtedly warranted in placing the Yellow Bluff race on the Volutilithes petrosus main stock as the descendant of the Gregg's Landing race. The exact stratigraphical horizon for this race is unknown to the 'See Trans. Wag. Inst., Vol. Ill, part 1, p. 7.5. The statement made on this page by Dall is not borne out by a study of full collections. He says that Athleta "is foreshadowed by the distorted specimens of Volutilithes petrosa to which I have already referred, and which occur from the Lower Eocene of Alabama (Wood's Bluff) up to the Claiborne Sands and the beds known as Jackson, over- lying the Claiborne, associated with the undistorted normal form, which is always more numerous." 1906.] NATURAL SCIENCES OF PHILADELPHIA. 65 author. It cannot, however, be lower than that of the Bell's Landing marl. If it is at the same horizon, we have an example of a slightly senile race (the Bell's Landing) living at the same period of time not far from a race (the Yellow Bluff) which exhibits, on the whole, main stock characteristics. If, on the other hand, the Yellow Bluff race occupies a position above the Bell's Landing marl, we have a primitive race overlying a more specialized race. If this latter is the case, a discordance between the geological and phylogentic successions has been produced by the migration into the section of the Yellow Bluff race. Volutilithes petrosus Conrad. Lower Claiborne Horizon, Claiborne, Ala. Very poorly preserved specimens are found in these beds. They are determinable as representatives of the V. petrosus main stock, but a detailed study, whorl for whorl, is not practical. Volutilithes petrosus Conrad. Claiborne Race. Locality — Claiborne, Ala. Whorls 1, 2, 3 — Smooth and rounded. Whorl 4 — Early in four the curved rib stage appears. It is not well marked, persisting for but a short time, and is soon followed by the straight, tubercled ribs of the cancellated stage. The shoulder angle soon develops. The suture and shoulder tubercles are of about the same size. Wliorl 5 — Characterized by the cancellated stage and much as in the latter part of four. Spirals not strong. Suture tubercle weakens, shoulder tubercle becomes stronger. Whorl 6 — Characterized by the cancellated stage. The suture tubercle and the ribbing above the shoulder are almost gone. Shoul- der tubercle stronger. Ribbing below shoulder is weaker. Whorl 7 — The ribbing and the suture tubercles disappear, and the shoulder tubercles develop into spines during this whorl. In other words, we have the transition from the cancellated to the spiny stage. W^iorls 8, 9, 10— All characterized by the spiny stage, but individual variation is great. Senile features do not appear. Remarks. — Occurs in the Claiborne Fossiliferous Sand. This race exhibits in general the characters of the V. petrosus main stock, though it is probably not the most typical expression of that stock as it existed at the time of the Claiborne deposits. The most notable features are the shortening of the curved rib stage and the acceleration of the spiny 66 PROCEEDINGS OF THE ACADEMY OF [March, stage, which is long and well developed though accompanied by con- siderable individual variation. The curved rib and cancellated stages are very short. Fig. 5 — Volutilithes sayana Conrad. Claiborne, Ala. Apex enlarged, show- ing ornamental features thrown far back by acceleration. The smooth stage is almost entirely confined to the large first whorl. In whorl two the transition from the straight untubercled ribs (suppressed curved rib stage) to the can- cellated stage is seen. Whorl three shows the cancellated stage. Length = 1.8 mm. Fig. 6. — Volutilithes sayaria Conrad. Claiborne, Ala. Adult individual. Length = 48 mm. Volutilithes sayana Conrad. Locality — Claiborne, Ala. This species was originally described as Valuta sayana by Conrad. ^° Whorl 1 — Smooth and rounded. Whorl 2 — Smooth and rounded at first, but soon ribbing begins. Untubercled ribs are the first ribs, but they are not curved. Neverthe- 1" Fossil Shells of the Tertiary Formations of North America. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 67 less they probably represent the curved rib stage and will be assigned to it. They persist for only a short while and are followed by the tuber- cled ribs of the cancellated stage. At first the svitiire tubercle is con- siderably larger than the shoulder tubercle. A shoulder angle soon develops, and by the end of the whorl the suture and shoulder tubercles are about of the same size. Wliorls 3, 4 — Characterized by the cancellated stage and much as in latter part of two. Changes however occur, for the ribbing above the shoulder and the suture tubercles weaken, though the shoulder tubercle becomes stronger. The spirals are not well marked. After whorl four great individual variation sets in. The cancellated stage may persist through the sixth whorl. After this the individual variation is still more marked. Many specimens pass through the spiny stage, though it is considerably modified by the development of numerous secondary spirals and riblets. In many specimens the shoulder spines are long in a transverse direction and notched by several spirals. Sometimes the whorl may be of a rounded shape, with the spines blunt, weak and obscure. Reyyiarks. — Occurs in the Claiborne Fossiliferous Sand. This most variable species has an apex very different from those of the other forms considered in this paper. The smooth stage is nearly all con- fined to the first whorl, which is extraordinarily large for Volutilithes. By a careful study of its ontogeny it is plain, however, that this species was derived by acceleration from the Claiborne race of V. petrosus, or else from some form verj'' close to it. We are forced, then, to the con- clusion that shells with rather different apices may sometimes be very nearly related ; in spite of the fact that, as a rule, this condition pre- cludes close affinity. Though this variable species may have given rise to species and races of later age than the Claiborne, the writer has no evidence to that effect. The anal siphon does not encroach up- ward, and the protruded mantle smooths the preceding whorl but little; nevertheless, the thickness of the shell and the irregularity of growth lines in many of the older individuals point to the fact that it is a senile race. It is well developed in the Claiborne horizon at several points, where it is always variable and more numerous than the Clai- borne race of V. petrosus. From this latter race it was probably evolved during or just before Claiborne time, and may therefore be con- sidered as a senile offshoot from the T^ petrosus main stock. Volutilithes petrosus Conrad. Jackson Race. Locality — Jackson, ]\Iiss. 68 PROCEEDINGS OF THE ACADEMY OF [Maich, An individual of this race was originally described by Conrad as Volutalithes symmetrica}^ Whorls 1, 2 — Smooth and rounded. ^V\rhQj.l 3 — ^xhe curved rib stage commences early in three and per- sists for about a quarter of a whorl. It is followed by the straight tubercled ribs of the cancellated stage. The suture and shoulder tubercles are about equal in size. By the end of the whorl the shoulder angle is well marked. Wliorls 4, 5 — Characterized by the cancellated stage. Ribbing above the shoulder weakens , as does also the suture tubercle. Shoulder tubercle gets stronger. Ribbing below the shoulder well marked and crossed by fine spirals. Whorl 6 — Characterized by the cancellated stage and much as in five. Fig. 7—Volutilithes petrostis Conrad. Jackson Race. Jackson, Miss. Apex enlarged. Whorls one and two show the smooth stage. Whorl three shows the curved rib stage and whorl four the cancellated stage. It will be noted that the stages are all more accelerated than in text fig. 4, but less accelerated than in text fig. 5. Length = 1.3 mm. Wliorl 7 — Characterized by the cancellated stage. Ribbing gone above the shoulder, but well developed below the shoulder, and crossed by fine spirals. Suture tubercles almost gone though represented by a spiral. Shoulder tubercle stronger and sharper. Whorl 8 — Much like seven, but toward the end the ribs disappear, the shoulder tubercle becomes a spine, and the cancellated stage comes to a close. Whorl 9 — Characterized by the spiny stage. High shoulder spines and no ribs. Spirals well developed, but strongest on the anterior siphon. Wliorl 10 — In many individuals the shoulder spines tend to pass into a shoulder keel, the shell thickens, and the growth lines become irregular. The shelly smoothing of the preceding whorl by the pro- truded mantle is only slight. This whorl is characterized by the senile stage. " Proc. Acad. Nat. Sci. Phila., Vol. VII, p. 260. In this case "Volutalithes" is probably a typographical error. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 69 Remarks. — This race shows a regular and even ontogeny. The different stages are quite distinct. The senile characters are not thrown back into the spiny stage but follow regularlj^ after it, and constitute the culminating feature of individual development. These marks of old age are never extreme. Acceleration early in life has placed the curved rib stage at the beginning of the third whorl, and the cancellated stage has been correspondingly lengthened. The cancellated stage is longer here than in any of the other forms con- sidered in this paper except V. limopsis. This assemblage is the last representative of Volutilithes in the Alabama and Mississippi Eocene. It fulfills every requirement for position on the V. petrosus inain stem, it is the culminant race of the stock, and has attained its characters by a slow and even phylogenetic development. Volutilithes haleanus Whitfield. Locality — Lisbon, Ala. Whorls 1, 2 — Smooth and rounded. Whorl 3 — Smooth and rounded at fii'st, but later the curved rib stage begins. This persists for about a quarter of a whorl, and is then fol- lowed by the straight, tubercled ribs of the cancellated stage. The shoulder tubercle is just a trifle larger than the suture tubercle. The shoulder angle is well marked by the end of the whorl. Wliorl 4 — Characterized by the cancellated stage. Shoulder angle. Shoulder tubercle stronger than the suture tubercle. Whorls 5, 6 — Characterized by the cancellated stage and much like whorl four. During these whorls the suture tubercle tends to attain the same strength as the shoulder tubercle. Spirals not very well marked. Whorl 7 — From here on the primitive cancellated stage is replaced by what may be called a modified cancellated stage. The suture and shoulder tubercles are now of about the same size. The ribs are crowded so closely together that they are almost in contact, being separated by a mere depressed line. The shoulder tubercles are so close together that the general effect is that of a whorl with a shoulder keel. Whorls 8, 9 — ]\Iuch as in seven, but toward the end the shell becomes thick and the growth lines slightly irregular. The protruded mantle produces only an insignificant smoothing of the preceding whorl. Remarks. — By a study of its early whorls it is evident that this form was at some time derived from the V. petrosus main stock. Its can- cellated stage closely resembles that of the races of V. petrosus until we reach the seventh whorl. After that it is quite different in its general 70 PROCEEDINGS OF THE ACADEMY OF [March, characters from every other American species of Volutilithes. Just when it diverged from the V. petrosus main stock the writer is not prepared to say. This race has, properly speaking, no senile stage. In order to show more clearly some of the important correlations between the ontogeny and phylogeny, the following tables are intro- duced. The figures stand for the number of the whorl, while the letters are used for the different ontogenetic stages as follows : A — Smooth Stage. B — Curved Rib Stage. C — Cancellated Stage. D — Spiny Stage. E — Senile Stage. I — Individual Variation. This last may appear anjnA^here, but it usually follows stage C. B is exaggerated in all the tables. V. limopsis. Matthew's Ldg. V. petrosus. Gregg's Lclg. Race. V. petrosals. Jackson Race. A f ^1 c A B| 1 C t).„ a / \ B| 1 c b ix.i^ 1 E. is very slight. E. is never extreme Table I. — Showing slow, even evolution in the main stock, accompanied by acceleration. V. limopsis. Matthew's Ldg. V. petrosus. Gregg's Lag. Race. V. petrosus. Bell's Ldg. Race. V. petrosus. Wood's Bluff' Race. 1 2 3 4 5 6 7 s 9 10 11 12 A 1^ 1 c ,iE A 1 1 c Da, al A 1 ^1 c Da A 1 Pi c Ea .dh 1 E. is very slight. E. is slight. E. is extreme. Table. II. — Showing senility in an offshoot. The senility increases, not by an acceleration, but by an exaggeration of senile features. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 71 10 11 12 T''. limopsis. Matthew's Ldg. T'. petrosus. Gregg's Ldg. Race. V. petrosus. Claiborne Race. V. sayana. Claiborne. A I il c A \ ^1 1 c D.,if A 1 ( . 1 D.. I A ft 1 c 1 , Daadf 1 E. is very slight. I. is great. D. is much modi- fied. I. is great. Table III. — Evolution of F. sayana through the Claiborne race of V. petrosus. Shows general acceleration, condensation of B, and reduction in the number of whorls. 12 3 4 5 (•, 7 8 0 10 11 12 V. limopsis. Matthew's Ldg. V. rugaius. Matthew's Ldg. A fi \i c. A 1 1 '1 C I E 1 1 r^ r- I. is the transition between C. and E Table IV.— Shows evolution of V. rugatus from the primitive ancestral stock by the appearance of senile features in the last whorl or in the last two whorls. Range and Distribution. Of the forms considered in this paper Volutilitkes limopsis stands out distinctly as fulfilling all the requirements of an ancestor. Among other characters, the long smooth stage and the absence of senile fea- tures show its primitive state. The general likeness of the full-grown shells to the young of many later species, together with its low geo- logical position, justifies us in considering it to be the form from which were descended the races which followed Matthew's Landing time. In the same horizon with V. limopsis we find V. rugatus. The young of the latter show many of the features of the adult V. limopsis, while the old individuals are often slightly senile and quite different in ap- pearance from the small specimens. It was undoubtedly derived from V. limopsis, or some closely allied form, and may be considered a short senile offshoot which did not survive the Matthew's Landing. 72 PROCEEDINGS OF THE ACADEMY OF [Mai'Ch, In the Nanafalia the first member of the Volutilithes petrosus assem- blage makes its appearance, and is representative of the V. petrosus main stock. After the Nanafalia the races of V. petrosus are a promi- nent feature of the faunas. Though the adults of these races may differ greatly, the young resemble each other closely, and indicate their common derivation from V. limopsis. In the Gregg's Landing marl, the Gregg's Landing race represents the V. petrosus main stock, though a few of its individuals show slightly senile tendencies. In the Bell's Landing marl the Bell's Landing race has separated from the main stock with the evolution of senile features. It is the first member of a senile offshoot which in later time produces the Wood's Bluff and Hatchetigbee Bluff races. These latter are evolved from the Bell's Landing race, not so much by an acceleration as by an exaggeration of the marks of senility. The Hatchetigbee Bluff race is without descendants and the last term of this senile series. Its individuals are dwarfed and stunted, reflecting plainly the unfavor- able conditions under which they lived, and which soon caused the extinction of the race. While in some localities the members of the Volutilithes petrosus assemblage were following out a line of evolution which resulted in their elimination from the geological record, the main stock was still flourishing in nearby regions. For instance, at Yellow Bluff on the Alabama river we have a representative of the V. petrosus main stock. The YeUow Bluff race is indicative of favorable conditions at that locality. In the Lower Claiborne the V. petrosus main stock occurs, though it is represented only by a few poorly preserved specimens. The Clai- borne, however, not only gives us the Claiborne race but an offshoot from it, namely, V. sayana. This Claiborne race is an assemblage of rather accelerated and variable forms which probably express the effect of shallow water marine conditions upon the main stock. By an acceleration of ornamental features V. sayana was derived either from the Claiborne race or from some very closely related form. Passing now into the Jackson, we find the last member of this phy- logenetic series which we have traced up from the Matthew's Landing horizon. It is the Jackson race of V. petrosus. This form is the result of the slow and even evolution of the ancestral race which took place in the favorable conditions of a more open sea. Its ontogeny is nor- mal, each stage following its predecessor in regular fashion. Though this form is accelerated, there is little mingling of the senile and spiny characters by unequal acceleration. The senile stage closes a well- 1906.] NATURAL SCIENCES OF PHILADELPHIA. 73 developed spiny stage. This race of V. petrosus differs but slightly from the form which we saw in the Nanafalia beds, and demonstrates clearly how little a stock may vary through a long period of time if it is living under favorable conditions.^- The phylogeny of the Alabama and Mississippi species of Volutilithcs may be represented by the table on p. 74. The Yellow Bluff race is provisionally placed in the Bell's Landing horizon. Summary. From the Matthew's Landing to the Hatchetigbee inclusive the strata belong to the so-called Lignitic formation. They consist of thick deposits of cross-bedded sands and clays, often glauconitic, and of lignites alternating with thin beds of marine fossils." It is evident that these deposits were formed in a great shallow arm of the sea, whose w^aters were sometimes fresh or blackish and at others salt. The marine faunas which from time to time invaded this Lignitic gulf brought with them, at first, species and races of Volutilithes with normal aspect. These, however, only represent the dwellers on the edge of the respective Volutilithes communities, the centres of which were farther out in the open sea. These forms which were subjected to the conditions of the Lignitic eventually followed out a course of evolution which was a direct reflection of their unfavorable environ- ment. The races of Volutilithes petrosus at Bell's Landing, Wood's Bluff, and Hatchetigbee Bluff make such a series, in which the senility becomes more and more extreme with the course of time." Occasionally throughout the later Lignitic, as at Yellow Bluff, we find a race which has migrated from a more favorable environment, and which resembles the primitive races of earlier Lignitic waters, as those of Gregg's Landing and Nanafalia. The grade of phylogenetic development which the Yellow Bluff assemblage has attained also '^ The author has seen forms purporting to come from the Vicksburg horizon which appear to be poorly preserved examples of the Jackson race of V. petrosus. He does not, however, feel justified (without further evidence) in carrying the range of V. petrosus above the Jackson horizon. " Bull. 43, U. S. G. S., "Tertiary and Cretaceous Strata of the Tuscaloosa, Tombigbee, and Alabama Rivers," Eugene A. Smith and Lawrence C. Johnson. ^* It is to be noted in this connection that forms occur in the Miocene of Europe which appear to belong to Volutilithes and to possess the shelly overgrowth. The abnormal American races are produced by local conditions, that is, by those of the Lignitic, and are widely separated not only geographically but geologically from the European specimens. It is therefore highly improbable that the Eu- ropean forms (if they are Volutilithes at all) are descended from the American ones, In the absence of more definite information, we are justified in regarding the phenomenon as one of parallelism. 74 PROCEEDINGS OF THE ACADEMY OF [March, ^ '^^ « a 02 fcfl bC'O 5j S^K 1^ G ^^ : SPi s 'Stl 9 t^ = W ^ ^ 1906.] NATURAL SCIENCES OF PHILADELPHIA. 75 indicates that a normal, slow, and even evolution has been going on in one locality ; while at the same time rapid, senile evolution has taken place among forms subjected to unfavorable conditions in another region. Where, then, was the centre of this normal phylogeny? We may well look toward the open sea to find this favorable environment, and the deposits which follow the Hatchetigbee beds furnish us with the desired answer. With the inauguration of the Lower Claiborne began that movement which finally resulted in the deepening of the sea and the production of the white limestone of the Jackson. The brackish water conditions disappear and are replaced by those of a shallow water (but typically marine) environment. The Claiborne race of Volutilithes petrosus shows the effect of such conditions upon the main stock. The sinking of the sea-bottom proceeded throughout Claiborne time, and finally the Jackson limestone with its relatively deep water fauna is intro- duced. The Jackson race of V. petrosus represents the result upon the main stock of slow and even evolution in a favorable environment. It follows, then, that every stock has some particular set of condi- tions in which it develops normally. Contemporaneously with this normal phylogeny a senile evolution in the same stock may occur at some less favored locality. In the forms we have considered the environment most favorable to a slow and even evolution is an open sea one, in which limy deposits are forming. It is only, then, in a thick limestone deposit that we could expect to find all the phyloge- netic stages of the Volutilithes petrosus main stock and other primitive species. In conclusion it is interesting to note that one of the only two existing species of the genus, namely, Volutilithes abyssicola Adams and Reeve,^^ is a cancellated form which apparently is quite close in its grade of evolutional development to the ancestral V. limop- sis of remote Eocene time. This recent species is found in relatively deep water off the Cape of Good Hope, and probably is a member of a primitive stock, which iri the favorable environment of a deep and open sea has been able to persist long after its nearest relatives have become extinct.^^ The writer wishes to express his indebtedness to Prof. Henry A. Pilsbry, of the Academy of Natural Sciences of iPhiladelphia, and to Prof. Amos P. Brown, of the University of Penns3dvania, for many ^^ Zoology of the Voyage of the Samarang. i« See Dall, Trans. Wag. Inst., Vol. Ill, p. 74. 76 PROCEEDINGS OF THE ACADEMY OF [Marcll, helpful suggestions, and to Miss Helen Winchester for care exercised in making the drawings. In addition the writer wishes to especially acknowledge the kindness of Mr. Charles W. Johnson, of the Boston Society of Natural History, in giving him much valuable information in regard to the range and distribution of the races and species. Explanation of Plate II. Fig. 1. — Volutilithes limopsis Conrad. Matthew's Landing, Ala. Adult of moderate size showing the persistence of the cancellated stage throughout the later whorls. Length = 18.5 mm. Fig. 2. — Volutilithes rugatiis Conrad. Matthew's Landing, Ala. Young individ- ual showing the cancellated stage. With the exception of the last whorl, the sculpture is very much like that of V. limopsis. Length = 19.5 mm. Fig. 3. — Volutilithes petrosus Conrad. Bell's Landing Race. Bell's Landing, Ala. Young individual showing the cancellated stage. Length = 20 mm. Fig. 4. — Volutilithes petrosus Conrad. Jackson Race. Jackson, Miss. Young individual showing the cancellated stage. Length^lQ mm. Fig. 5. — Volutilithes petrosus Conrad. Gregg's Landing Race. Gregg's Landing, Ala. Adult individual showing the general characters of the V. petrosus main stock. The later whorls show the spiny stage. Length = 50 mm. Fig. 6. — Volutilithes petrosus Conrad. Yellow Bluff Race. Yellow Bluff, Ala. Adult individual showing the general characters of the V. petrosus main stock. This is, however, not as typical an expression of the features of the main stock as is seen in fig. 5. The later whorls show the spiny stage. Length = 44 mm. Fig. 7. — Volutilithes petrosus Conrad. Jackson Race. Jackson, Miss. Adult individual showing the general characters of the V. petrosus main stock. The later whorls show the spiny stage. This is the culminant type of the main stock. Length = 52 mm. Fig. 8. — Volutilithes petrosus Conrad. Bell's Landing Race. Bell's Landing, Ala. Adult individual showing the slightly developed senile characters. Length = .52 mm. Fig. 9. — Volutilithes petrosus Conrad. Wood's Bluff Race. Wood's Bluff, Ala. Adult individual showing strongly developed senile characters. Length = 58 mm. Fig. 10. — Volutilithes petrosus Conrad. Hatchetigbee Bluff Race. Hatchetig- bee Bluff, Ala. Adult individual showing extreme senility and dwarfed size. This is the culminant type of the senile offshoot shown in figs. 8 and 9. Length = 43 mm. 1906.1 NATURAL SCIENCES OF PHILADELPHIA. 77 SOME COLD-BLOODED VERTEBRATES OF THE FLORIDA KEYS. BY HENRY W. FOWLER. During June of 1904 the wi'iter, accompanied by Mr. Stewardson Brown, of the Botanical Section of the Academy, was enabled through the generosity of Mr. Clarence B. Moore to visit the Florida Keys. The chief object of the expedition was to explore the islands for land snails of the genus Liguus. Incidentally we were enabled to study the flora and fauna of this most interesting region. We availed ourselves of every opportunity to gather information, l^ut only data which I feel to be absolutely reliable is here included. The vernacular names of the different species, so far as I could judge, are those of the fisher- men and residents. Our operations extended from between a point directly south of Cape Sable, or from Grassy Key west to the Marquesas, covering an area about fifty miles in length. We did not visit the Tortugas, which are the most western of the Keys and more isolated than the Mar- quesas. Making Key West our headquarters, we set out after explor- ing that island for the IMarquesas, stopping at Boca Grande Key on our way. The Marquesas bank comprises a number of islands, dis- posed in a more or less circular or atoll-like manner with great shallows all about, though mostly inside. Few places we visited repaid us as well as these islands. On our return we skirted the southern shores of the Keys toward Key West, stopping at Ballast Key. We then left Key West again and travelled along the southern shores of the Keys eastward to Grassy Key, stopping at Boca Chica, Sugar Loaf, Big Pine, No Name, Cud joe's. Knight, Vaca and Grassy Keys. On our return, which was along the northern shores of these islands, we stopped at Vaca, Bahia Honda, Hailer's Rock, Little Pine, Big Pine, Summer- land, Cudjoe's, Sugar Loaf, West Cudjoe's, Riding, Snipe and Jewfish Kej^s, Most of the Keys are more or less covered with mangroves; others, such as Big and Little Pine, have large tracts of pine forest with an undergrowth of scattered palmettos. The latter grow several feet in height, and in combination with the pines present a beautiful con- trast after the monotony of mangroves. All of the islands are low, rising but a few feet above the sea-level, and most all are furnished 78 PROCEEDINGS OF THE ACADEMY OF [Marcll, with rather broad white sandy beaches. Some places along these beaches are studded with mangroves, affording excellent shelter, when overhanging, for many fishes. Along a number of the Keys w^ere flat embankments covered with long grass, and in other places coconut groves, as that on Boca Chica Key. jNIany Keys were covered with more fertile areas in which various tropical plants, shrubs and trees were found in abundance, often studded here and there with cacti, and sometimes presenting an impenetrable barrier. Throughout the Keys we visited the water is comparatively shallow. In some places reefs are formed where the most beautiful of tropical marine animal life abounds in luxuriance. Most of these places, we were in- formed, are constantly though gradually shifting, so that new charts are needed from time to time to show the changes in depth and to- pography. Some of the shallows are left entirely dry by the tides, and others are covered with large beds of sea-wrack. In such places many wading-birds find an abundant supply of food. We found a few pools on certain of the larger Keys, such as Big Pine, Summerland and Little Pine. They are of rather brackish-water, though I did not taste of all. Hailer's Rock is a small flat or low island of sand-heaped rocks, with a growth here and there of a few low bushes. The southern end is of bare dry rocks running well out to sea and leaving numerous tide- pools. These pools vary in size and depth from some of a few inches in extent to others of several feet. Here such gaudj^ and tropical fishes as the Cock Eye Pilot, Sheep's Head, Rock Fish, and School- master were found. The Rock Fish, or Gobies and Blennies, are colored much like their surroundings, and trust in great part to this for concealment. Some are pale and others are dark. They dart swiftly from place to place when disturbed, much after the manner of our own Darters, or else hide in crevices. Though they were active, I secured a number by hand with a little trouble. While off Key West we saw cetaceans, evidently a species of Del- phinus. The Opossum and Raccoon were reported from Key West. and the latter also from Sugar Loaf, Big and Little Pine Keys. I saw a Rabbit on Big Pine Key and a Deer on Sugar Loaf. Deer were found former!}' on Little Pine Key, but were now believed to be absent. Though I did not see the Wood Rat on Summerland Key, I was in- formed that they sometimes climb the trees to feed on fruits, such as the sapadilla. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 79 FISHES. GINGLYMOSTOMID^. 1. Ginglymostoma cirratum (Gmelin). Shark. Very abundant in all warm shallows about the Marquesas. At this season (June) they come in these places to copulate, and I was informed that they do not occur here, or at least were rarely seen, at other times for this purpose. I saw a male and female in coitus. On one occasion about four males were seen swimming close about a female, whose whitish belly could be seen quite a distance, as she was back down. She was also evidently awaiting their initiative. I approached in a small boat till almost directly above, when it was possible to prod them with an oar as they swam slowly about and but little agitated. A pair will remain fastened together some little time, and then suddenly part without much commotion. Their loca- tion may easily be determined in these places by a portion of the dorsal fin, and frequently also a part of the caudal, projecting above the surface. The term "cooting," employed by some fishermen for the act of copulation among these sharks, may possibly have been derived from the word courting. Examples vary somewhat in color, some appearing pale cinnamon, though they are always distinguishable quite a distance, even when entirely under water. All I observed were large, nearly twelve or fourteen feet in length, and were not at all shy. Occasionally in shal- low water, especially like that along the southern shores of Ballast Key, they were sometimes more or less exposed or with both caudal fins well out of the water when copulating. ALOPIIDiE. 2. Alopias vulpes (Gmelin). Fish Shark. Reported to be rather numerous at times in and about the Keys. They were said to feed almost entirely on fish. GALEID^. (Galeus Klein, in Walbaum, Pet. Arted. Gen. Pise, III, 1792, p. 580. Type Squalus galeus Linna'us = Galeorhinus Blainville, 1816-17.) 3. Cynais canis (Mite hill). A small Shark was seen in the shallow water about Sugar Loaf Key. 4. Galeocerdo tigrinus Miiller and Henle. Reported as sometimes taken during smnmer. 80 PROCEEDINGS OF THE ACADEMY OF [March, 5. Carcharliinus lamia (Rafinesque). Large examples of some species of Carcharhinus, most likely this species, were reported under the name ''Man Eater." I did not see any examples or hear any reports of Scoliodon terroe-novoe (Richard- son), unless it was confounded with the present form. OESTRACIONTID^. (Cestracion Klein, in Walbaum, Pet. Arted. Gen. Pise, III, 1792, p. 580. Type Sqimhis zygoejia Linnaeus = Sphyrna Rafinesque, 1810.) 6. Cestracion tiburo (Linna?us). Bonnet Head Shark. Reported as rather abundant. 7. Cestracion zygaena (Linnseus;. Hammer Head. Hammer Head Shark. The large examples which occur about Bahia Honda Key are said to attain at least fifteen or more feet in length. C. tiburo is said to be small. Both species were reported as more abundant in summer. RHINID^. (Rhina IQein, in Walbaum, Pet. Arted. Gen. Pise, III, 1792, p. 580. Type Squalus squatina Linnaeus = Squatina Dumeril, 1806.) 8. Rhina squatina (Linnwus). Occasionally taken in smnmer. PRISTID^. 9. Pristis pectinatus Latham. Saw Fish. Reaches a considerable size. Young rather common in the shallows. NARCACIONTID^. {Narcadon Klein, in Walbaum, Pet. Arted. Gen. Pise, III, 1792, p. 580. Type Raja torpedo Linnaeus = Narcobatus Blainville, 1816.) 10. Tetronarce occidentalis (Storer). Electric Fish. Found occasionally associated with species of Dasybatus. DASYBATID^. {Dasybatus Klein, in Walbaum, Pet. Arted. Gen. Pise, III, 1792, p. 581. Type Raja pastinaca Linnaeus = Dasyatis Rafinesque.) 11. Dasybatus hastatus (DeKay). Stmg Ray. Color in life pale hair-brown above, margin of disk paler or more or less narrowly whitish, and submarginally a slightly darker shade than general color of body. Tail blackish-brown, spine dark. Lower surface of body white. Iris brown. This species was seen several times about the shallows of the Mar- 1906.] NATURAL SCIENCES OF PHILADELPHIA. 81 quesas. When swimming they appear to ghde over the sand, from which they are distinguished with difficulty when quiet. They usually swim rapidly away upon the approach of a small boat, though we captured a single example with the harpoon. They were found in pairs, and seldom attained more than a couple of feet in width of disk. PS ALLISOSTOMID^ . {Psallisostomus Klein, in Walbaum, Pet. Arted. Gen. Pise, III, 1792, p. 581. Type Esox osse,us Linnaeus.) 12. Psallisostomus osseus (Linnaeus). Common Gar. Reported from about the Keys south of Cape Sable, but not seen by me. ELOPID^. 13. Tarpon atlanticus (Valenciennes). Tarpon. This fish is reported to be used for food by the negroes at times. It is the custom of anglers to throw their bodies on the beach or shore, as it is said they attract the sharks if allowed to stay in the water, thus interfering with the fishing. The young were said to be not quite so common, though it is possible that they may have been over- looked. 14. Elops saurus Linnaeus. Ten Pounder. Said to be aljundant at times. ALBULIDiE. 15. Albula vulpes (Linnaeus'. Bone Fish. Mostly abundant. DUSSUMIERIID^. 16. Jenkinsia stolifera (Jordan and Gilbert). Fig. 1. Head Sf; depth 5; D. iii, 8; A. i, 14; P. ii, 11; V. i, 6; scales 30? (according to pockets) in a lateral series to base of caudal, and about 4 more on latter; about 8? series of scales transversely between origin of dorsal obliquely back behind origin of ventral; width of head about 2 in its length, and depth of head a trifle less; length of depressed dorsal 1^; base of anal 2; length of caudal (damaged) 1^; pectoral If; ventral 2; least depth of caudal peduncle 3f ; mandible 2|-; snout 3^; orbit 3; interorbital space 4; maxillary 2^. Body long, compressed, profiles similar and not trenchant. Great- est depth at origin of dorsal. Caudal peduncle compressed, and its length about equal to its depth. 82 PROCEEDINGS OF THE ACADEMY OF [March, Head large, deep, compressed, upper profile curved a little from tip of upper jaw and then straight after middle of orbit, and lower profile forming a broad angle below front of orbit. Snout short and rounded. Eye large, rounded, almost impinging on upper profile and anterior. Adipose eyelid covering greater part of orbit. Mouth ter- minal and a little superior, small, and mandible projecting a little in front. Maxillary well inclined, distally broadly expanded till about equal to f of pupil, and its posterior edge not quite reaching opposite same. Teeth in jaws minute, and if present on roof of mouth and tongue very small or indistinct. Tongue small, a little elongate, and with free rounded tip. Each ramus of mandible well elevated inside mouth. Interorbital space rather narrow and flat. Nostrils close together, and about midway in length of snout. Preorbital rather ''"i* •if«'^i^^|4#i#^ ^ ■^'4 Fig. 1. Jenkinsia stolifera (Jordan and Gilbert). narrow and together with infraorbital, preopercle and opercle furnished with rather large arborescent flutings, those on latter with more or less adipose substance above. Opercle with a gash in upper posterior margin. Articular surface or mandible below its articulation with a broad obtuse spine or denticle. Gill-opening carried forward till about opposite front margin of orbit. Rakers fine, slender, longer than filaments. Filaments and pseudobranchise about equally well developed. Isthmus trenchant in front, and flattened rather broadly posteriorly. Scales cycloid, caducous or but few remaining on alcoholic examples and those mostly comprising dorsal and anal sheaths, and on base of caudal. A patch of adipose tissue on shoulder behind opercle above. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 83 Origin of dorsal nearly midway between tip of snout and base of caudal and first developed ray longest. Anal low, inserted a little behind tip of depressed dorsal or a little nearer origin of ventral than base of caudal. Caudal forked and lobes pointed. Pectoral low, inner rays of each fin approximating and reaching about half way to ventral. Ventral inserted a little behind origin of dorsal or a little nearer base of caudal than tip of snout and reaching about f of dis- tance to anal. Vent close in front of origin of anal. Color in alcohol faded pale brownish, a little darker on back, and lower surface paler. Head more or less silvery, top brownish. Fins all pale or whitish, dorsal and caudal a little grayish. ■Median line of back with a dusky streak. Along course of what would be lateral line posteriorly a brownish streak. Iris silvery, a little slaty above. Length If inches. Hailer's Rock. Altogether 17 examples were taken associated with the large schools of Anchovies seen about the tide pools. In life they are colored transparent or translucent greenish similar to the Ancho- vies, One example contained a crustacean parasite within the gill- opening, the long spiral egg-cases well protruding. CLUPEID^. 17. Harengula sardina (I'oey). Fig. 2. Harengula sardina (Poey), . Two small examples taken at Hailer's Rock. 8. Harengula humeralis (Valenciennes). Shiner. Fig. 3. Color in life bright or glistening silvery-white on sides and lower surface of body. Back bluish-green. Snout brownish. Eye silvery. 84 PROCEEDINGS OF THE ACADEMY OF [March, upper margin brownish bhie-green. Dorsal and caudal gray-brown, other fins transparent whitish. In the shallows of the Marquesas 30 examples were seined, asso- ciated with the Anchovies and other small fishes. In a tide-pool on Fig. 3. Harengnla humeralis (Valenciennes) Hailer's Rock 2 were also taken. After being in alcohol some time most of the examples have the peritoneum showing plainly through the body-wall. 19. Clupanodon oglina (Linnaus).^ Thread Herring. Occasional. No examples taken. BNGRAULIDID^. 20. Anchovia brownii (Gmelin). Sardine. Color in life, above clear transparent hyaline-greenish washed with 1 Drs. Jordan and Gilbert, in Proc. U. S. Nat. Mus., V, 1SS2 (1883), p. 574, have restricted Lacepede's genus Clupanodon to jussieui. Later Drs. Jordan and Evermann, in Bull. U. S. Nat. Mus., No. 47, 1, 1896, p. 422, apparently con- sider pilchardus as typical. These precepts do not fall within the rule of elimina- tion as the last name for which a generic name is proposed, so far as I can find, is thrissa. For this Dr. Gill proposed Opisthonema. At any rate Sardinia cannot be so considered as it is a synonym of Thrissa Rafinesque. Lacepede's species thrissa may be considered restricted to the American Thread Herring, as the Chinese fish included under Osbeck's reference is one of the Dorosomatidoe. Opis- thonema Gill is thus to be superseded by Clupanodon Lacepede The True Sar- dines formerly placed under the latter name, will stand as species of Thrissa Rafinesque, with Clupea pilchardus Linnseus as the type. Messrs. Jordan and Seal have pertinent remarks in Proc. U. S. Nat. Mus., XXVIII, 1905 p. 771. 1906. NATURAL SCIENCES OF PHILADELPHIA. 85 bright silvery, and sides and lower regions of the brightest silvery. Fins pale and more or less transparent. Iris silvery. Reaches a length of 3 inches. Found throughout the Keys, the above notes from examples from Hailer's Rock. Very large schools were seen both in this locality and about the Marquesas. They swim in large shoals associated with the other small fishes, haunting the shal- lower as well as the deeper places along shore. They were unusually numerous about the broken rocky shores, where free ingress and egress could be had with the outer waters. Here the schools were seen more or less quiescent and with their heads all directed one way, so that upon any disturbance the whole mass would move more or less as if by one impulse. They appear mostly transparent in the water. I found them occasionally in the siu"f , where they were probably more abundant than I was able to determine. When taken from the water they soon die. I never observed them in any tide-pools which were cut off from the sea during low tide. 21. Anchovia choerostoma cayorum subsp. nov. Fig. 4. Head 3f ; depth 5; D. in, 9; A. iii, 25; P. i, 12; V. i, 6; scales 36 (squamation injured) in a lateral series to base of caudal with several .t^^^^^xx,. Fig. 4. Anchovia choerostoma cayorum Fowler. more on latter ; about 8 series of transverse scales from above origin of anal; width of head 2| in its length; depth of head H; snout 5; eye 3^; interorbital space 3f ; maxillary 1^; length of depressed dorsal If; caudal 1^; pectoral If; ventral 2f ; base of anal 3^ in head and trunk. Body strongly compressed, elongate in form, rather slender, and lower profile a little more convex than upper, edges not trenchant though rounded, and greatest depth about origin of dorsal. Caudal 86 PROCEEDINGS OF THE ACADEMY OF [March, peduncle strongly compressed, rather deep, and least depth about f its length. Head strongly compressed, rather deep, and profiles more or less similarly convex. Snout rather blunt, rounded, convex, and upper profile pronouncedly convex. Eye circular, rather large and well anterior. Adipose tissue developed. Mouth a little inclined and gape extending nearly f length of head. Maxillary slender, long, and furnished with a single series of fine or minute teeth to its distal ex- tremity which reaches almost to gill-opening in front of origin of pectoral. A similar series, though very minute, along margin of mandible. Vomerine teeth a little larger than those on palatines which are very small and uniserial. Also a still smaller series on each pterygoid. Symphysis of mandible reaching about opposite posterior nostril. Tongue small, little free, rounded, reaching forward till a little in front of anterior orbital rim and with a free asperous patch above. Nostrils close together, superior and a little nearer front rim of orbit than tip of snout. Interorbital space with a slightly elevated median ridge so that it is a little convex. Opercle and top of head posteriorly with varied shallow flutings. Cheek with shallow perfora- tions, and some above eye posteriorly. Gill-opening extending forward till opposite front rim of pupil. Rakers about 14 + 20?, slender, long, and longest much longer than filaments. Pseud obranchise smaller than filaments. Isthmus rather broadly rounded though compressed above. Shoulder-girdle notched once on each side below for reception of interopercle, and above behind opercle an adipose-like tract. Scales cycloid, large, and caducous. Dorsal and anal with scaly basal sheaths. Scales on base of caudal rather small. Pointed axillary flap at base of pectoral f length of fin and inner rays of each fin approximated. A similar ventral flap. Origin of dorsal nearer base of caudal than tip of snout and first branched ray longest. Anal inserted about opposite first fourth in length of depressed dorsal or much nearer origin of pectoral than base of caudal, and its anterior rays elevated. Caudal forked and lobes pointed. Pectoral low and reaching origin of ventral. Tip of de- pressed ventral reaching at least opposite origin of dorsal. Vent close in front of anal. Color in alcohol pale brownish, paler or whitish below. Back with edge of each scale sprinkled with a series of brownish dots. A double series of brownish dots down middle of back. Top of head, snout, and opercle brownish with darker brownish dots. Upper posterior portion 1906.] NATURAL SCIENCES OF PHILADELPHIA. 87 of head with gill-opening showing through dark. Head otherwise bright silvery. Iris silvery. A broad brownish band overlaid with silvery and nearly as wide as pupil extends from above gill-opening to middle of base of caudal. It is broader in middle than at either end. A series of brownish spots along base of anal. Dorsal and caudal pale grayish marked with numerous pale brownish dots, other fins whitish. Color in life with body more or less silvered and back above with a translucent appearance. Length 2f inches. Type No. 30,613, A. N. S. P. Hailer's Rock, Florida Keys. June, 1904. H. W. Fowler. Also Nos. 30,614 to 30,620, A. N. S. P., paratypes with same data. Found only at the above locality in company with A. hrownii. Only these few were secured among the multitudes of the latter. From A. chcerostoma (Goode), with which this form may prove iden- tical, it appears at present to differ a little, upon comparison with Porto Rican material, in the more elongate form, larger eye and head, and more blunt or rounded snout. Young examples have a smaller eye, nearly 4 in head. (Cayo, in Spanish meaning a ridge of small rocks or islands in the sea, like those off southern Florida.) SYNODONTID^. 22. Synodus foetens (Liniifeus). Lizard Fish. Color in life clear pale brownish above with about 10 wampum-like blotches of darker brown. In some examples this produces a pepper and salt appearance. A slightly dark double series of blackish dots across each blotch on back. On side just below back a series of zig- zag W-shaped markings also of same darker color. Same color ac- cents end of each blotch in form of a trifle darker speck. A somewhat diamond-like blotch or ring between each median lateral marking and somewhat connected. Another blotch alternately below each of these, and still lower more paler tints of diffuse markings. A deep brown blotch at base of tail and another of pale brown at base of each lobe. Head brownish above, and variegated with deeper spots and more or less olivaceous-golden on sides above. Jaws and mandible variegated with olivaceous-brown. Iris golden-olive, in some lights pm-e golden, pupil dark green. Region of isthmus dull citron. Tints of colors on sides all more or less golden. Dorsal and caudal variegated 88 PROCEEDINGS OF THE ACADEMY OF [March, with golden-brown tints, otherwise pale grayish. Other fins dull whitish. Length about 3 inches. A number of small examples were taken in the sandy shallows of the Marquesas. MURiBNID^. 23. Gymnothorax funebris (Ranzani). Moray, Large examples were seen about the reefs off Snipe Key, lurking in the crevices and holes of the coral rocks. They were deep olive in color. SILURID^. 24. Galeichthys felis (Linnaus). Cat Fish. Salt Water Cat. Very abundant and a very foul feeder, eating any refuse or filth thrown overboard. They were so ravenous that it was possible to catch them on any kind of bait even on a bent pin. At times great quantities were said to have been taken in seines, in which cases the fishermen would cut the leads and floats, leaving the whole mass to extricate themselves as best they may. Their spines are very much dreaded by the fishermen, who claim they are able to produce dangerous and ugly wounds. Most all the examples I observed were about Big Pine Key. PCEJOILIID^. 25. Fundulus similis (BairdandGirard). Sac-^-Lait. Color in life with a whitish lateral band from opercle above to base of caudal interrupting slaty vertical bars which were 9 in number. Lower surface of body white. Dorsal, anal and caudal grayish, former with a well-defined pale base. An olive shade on opercle. Iris silvery, a little brownish above and below, and latter also with a little pale orange. Jaws pale. Length 3f inches. Only four examples obtained in the shallow pools on Boca Chica Key. The larger ones have as many as sixteen dark vertical bars. Found associated with this species were all of the following PceciliidcB 26. Cyprinodon variegatus riverendi (Poey). Three examples from the Marquesas and Boca Chica Key. 27. Cyprinodon mydrus Goode and Bean. Puss Gut. Fig. 5 (male). Color in life nearly milky-white, upper surface or back above slightly brownish. Lower surface or body milky-white and fins very pale brownish. After death some examples turned pale olive on back and top of head, and on dorsal and caudal. These fins, back and side 1906.] NATURAL SCIENCES OF PHILADELPHIA. 89 speckled with golden-brown or coppery in the males. Lines of a similar shade on side of head, one across chin, one on preorbital, another from corner of mouth to lower rim of eye, and two others below and parallel up over opercle. Iris golden-brownish. Side of body with indistinct traces of several pale slaty vertical streaks. Lower fins very pale brownish-golden. Ventral and anal with a more or less regular series of golden spots, less distinct on latter. Females Fig. 5. Cijprinodon mydrus Goode and Bean. were paler and had no spots on dorsal and caudal, and those on body more yellowish. In spirits young examples show large blotches of grayish on side and lack brilliant colors of the adults noted above. They were more or less milky-white in life. This species is very abundant in pools and shallows of the Keys. They were met with in great numbers all about the Marquesas, and in the pools on Boca Chica Key. Young were abundant in the tide- pools on West Cud joe's Key. Great numbers were also found about Big Pine Key. Cyprinodon carpio Giinther ^ may be identical, but the indefinite locality "America" admits a doubt. The figure given by Drs. Jordan 2 Cat. Fish. Brit. Mus., VI, 1866, p. 306. 90 PROCEEDINGS OF THE ACADEMY OF [March, and Evermann ^ differs in that there are five rows of scales on the cheek. 28. Mollienisia latipinna Le Sueur. Mud Fish. Color in life brown. Side of head with gray and same color on costal region. Iris brown. Lower surface of body white. Dorsal and caudal with grayish-brown spots and other markings of brown, blackish-brown and dull amber-color. Some examples with caudal dull amber in middle basally and a lower series of longitudinal lines of dots of pale amber. Ventral and anal white. This species was found to be very abundant in the pools on Boca Chica Key, associated with other small fishes. Many are males with the dorsal fin enlarged and the edge of the caudal blackish. MASTAOCEMBELID^. 29. Tylosurus notatus (Poey). Gar. Fig. 6. Color in life clear hyaline-green overshot with silvery. Greenish about edge of opercle. Iris silvery. A narrow greenish line from origin of pectoral above to base of caudal. Vertical fins pale greenish- brown, tinted slightly with ruddy. Boca Grande, Big Pine and Boca Chica Keys, also Hailer's Rock and the Marquesas. Tylosurus notatus (Poey). Found in small pools on some of the Keys. They usually associate in small schools in the shallows or near the shore. When in the water their bodies have a transparent appearance and in some lights are difficult to distinguish. They are also fond of lurking about over- hanging mangroves. ^Bull. U. S. Nat. Mus., No. 47, IV, 1900, PI. 112, fig. 2&7. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 91 HEMIRAMPHID^.* 30. Cliriodorus atherinoides Goode and Bean. Hard Head. Color in life bright silvery. A narrow bright silvery-white line from gill-opening to base of caudal. Upper surface of body and back pale translucent brownish, each scale with a small patch of dark dots forming longitudinal series and about three above lateral line. Down middle of back three narrow lines of same color, median one pale or less distinct than others, and all close together. Dorsal and caudal pale grayish-brown and other fins translucent whitish. Iris silvery- white. Length 4^ inches. Alcoholic examples show lateral silvery band slaty. Found associated with Atherina laticeps in the sandy shallows of the Marquesas and about Hailer's Rock. 31. Hemiramphus brasiliensis (Linnjeus). Bala6. Reported as frequently seen. ATHERINID^. 32. Atherina laticeps Poey. Silver Sides. A'ery abundant. Large schools were seen about the Marquesas, Boca Chica and Snipe Keys. MUGILID^. 33. Mugil curema Valenciennes. Mullet. Abundant in the shallows near the shore where they associate in large schools. When disturbed they rush away, causing a loud noise in the water. Many were taken in cast-nets by fishermen, as many as a dozen at one time. They average several pounds each, and I found their flesh excellent eating. All examined appeared to be this species. They were most abundant in the INIarquesas shallows. 34. Mugil trichodon Poey. Fig. 7. Head 3^^ ; depth 3i; D. IV-I, 8; A. Ill, 8; scales about 36 to base of caudal in a lateral series; about 12 series of scales between origin of spinous dorsal and middle of belly; 19 scales before spinous dorsal; snout 3| in head; eye 4; interorbital space 2^; pectoral If; ventral If. Teeth large, conspicuous, uniserial, ciliform, and a little larger in upper jaw. Pectoral not quite reaching opposite origin of spinous dorsal. Color in life with back bluish-slaty. Side and lower surface silvery- white. Base and axil of pectoral dark slaty. Dorsal, caudal and *A number of Flying Fish, Exoccetidoe, were seen off Big Pine Key. They were all small. 92 PROCEEDINGS OF THE ACADEMY OF [March , anal pale brownish. Side of head with slight golden reflections."^ Iris brownish, a silver margin close to pupil. Length of largest example 2| inches, and though only three were taken with other small fishes, many others w^ere seen. They frequent the shallows like M. curema, but so far as I could observe never asso- ciate. Marquesas Keys. Fig. 7. Mugil trichodon Poey. SPHYR^NIDJE. 35. Sphyraena barracuda (Walbaum). Barracuda. Color in life more or less silvery. Above pale brow^nish with about six pairs of deep olivaceous double saddles meeting as many pairs of rounded blotches on side of body just above or over lateral line. A similar colored bar from end of snout to eye continued behind and across opercle. Iris pale olivaceous, slightly orange around pupil. Dorsals and caudal slightly grayish or a trifle dusky marginally, other fins pale whitish. These were all young examples. In the smallest, which are a little over an inch long, saddles of back are separated from lateral blotches by a pale or translucent brownish area. Dorsal and caudal but little darker than other fins. A brownish-olive spot at base of caudal in all stages of young. Many examples obtained at the Marquesas and Boca Chica Key, largest about 3^ inches. An adult, several feet long, was harpooned off Channel Key, and a number of large examples were noted off Bahia Honda Key. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 93 SYNQNATHIDiE.s 36. Syngnathus mackayi (Swain and Meek). Pipe Fish. Fig. 8. Color in alcohol brownish marked with small whitish clots or spots, each ring with a pale or whitish blotch on each keel, usually small and on side at intervals of four rings extending down in form of narrow bars. Back grayish with pale brown vermiculations medianly. On Fig. 8. Syngnathus mackayi (^wain and Meek). lower suface of tail small white spots producing a more or less con- fluent appearance. Head brownish, mottled grayish above, and snout and under surface with more or less whitish spots. Dorsal pale brown- ish-white mottled with pale brown. Caudal similar, only blotched and darker. Pectoral pale transparent brownish-white. Iris pale slatv. Color in life resembling the following species. Back whitish, finely vermiculated or specked with lavender, brownish and olivaceous. Lower surface a Httle more green than sides which are marked with bluish-green. Snout slightly brownish. Iris pearl-colored with brownish. Length 6^ inches. JMarquesas and Big Pine Ke}-. This species is subject to color variation, some examples inclining more to green and others to brown than the one noted above. 37. Syngnathus elucens Poey. Pipe Fish. Fig. 9. Color in alcohol brownish, each ring with a blotch of whitish pos- teriorly on each keel, those on sides of about every fifth ring producing about a dozen vertical lateral bands. Whitish blotches on lower sur- face of caudal becoming more or less confluent posteriorly. Head brownish, snout with a few w^hite spots and under surface with a number of w^hitish blotches. Dorsal, caudal and pectoral whitish, former fins with brownish specks. Iris dull slaty. Color in life with back whitish finely specked or dotted with lavender, brownish and olive medianly. Lower surface more clearly green than ^ I follow Profs. Jordan and Snyder in retaining Syngnathus for the species formerlj' called Siphostoma. Of the four species available by elimination, viz., typhle, acus, pelagicus and ophidion, the second {acus) remains and therefore must be regarded as the type. It is congeneric with Siphostoma Rafinesque. For typle Rafinesque also prop^osed Typle, and for ophidion, Nerophis. The latter name must therefore i-e used for the Ocean Pipe Fishes. 94 PROCEEDINGS OF THE ACADEMY OF [March, sides and each ring marked by a whitish or Hght vertical line. Some examples are darker, others have more white or brown, and still others are marked below with very pronounced white spots below, with gen- eral color of body pale greenish-brown and tail paler. Fins trans- parent, dorsal and caudal marked with brownish. Iris pearl-colored with brownish spots. Length 4^ inches. Fig. 9. Syngnathus elucens Poey. A nimiber of examples from the Marquesas, taken with many small fishes found in the sea-weed and marine vegetation of the shallows. One also from Big Pine Key. 38. Syngnathus louisianae Giinther. Hpe Fish. Fig. 10. Color in alcohol rather pale brown, upper side of back dark l^rown and middle of back much paler. Several series of brown wavy longi- tudinal lines more or less irregular, but parallel, on middle of back and caudal above. Along side, just below each keel, a short brown horizontal line on each ring. In dark color of upper half of side several others appear on some rings. On lower half of side brown lines are sometimes more or less convex, and on caudal they assume various vermiculate patterns. Lower or under surface of both head and body dull or pale brownish-white without markings. Head brown- ish, pale above, and on side of snout a dark brownish streak through eye and back across opercle above, giving rise on side of body to hori- Fig. 10. Syngnathus louisianae Giinther. zontal dark line on each scute just below uppermost keel. Dorsal, caudal and pectoral pale brownish-white, caudal wath some darker or brownish markings. Iris pale slaty, except brownish streak, and grayish below. Color in life similar. Back grayish-white above or medianly with brown lines. Upper side olivaceous-brown. Lower surface pale whitish, lines on lower side brownish, and under sin-face of abdomen 1906.J NATURAL SCIENCES OF PHILADELPHIA. 95 pale greenish. Dorsal whitish with very pale lines, other fins more or less transparent. Iris shell-color with a brown horizontal cross-bar. Length lOf inches. The Marquesas and Big Pine Key. This is the largest and strong- est species met with. It was found associated with the others in the sea-weed. It varies in color when seen in life. One example now in alcohol is dark blackish-brown on upper half of side. A young example taken later at Hailer's Rock has about sixteen dark annu- lations. 39. Hippocampus punctulatus Guichenot. Sea Horse. Color in life deep dull olivaceous, tubercles or joints pale at bases and ruddy at tips. Body everyivhere more or less variegated with narrow whitish vermiculating lines, and those about eye radiating from pupil, which is blackish. Pectoral and dorsal pale grayish^ latter finely spotted at base with small dusky or olive specks. Sea-weed of the i\Iarquesas. 40. Hippocampus zosterae Jordan and Gilbert. Sea Horse. One living example was pale brownish-white, side speckled with pale yellowish. A submarginal dorsal line of blackish. ]\Iarginal portion of belly dusky-olive. Rings white. Pupil blackish. Another was a little more brownish. Dorsal margined with pale orange, and a submarginal black band. Brown lines radiating from eye and on side of head slightly variegated with pale olive. Another was a little more highly colored and more of a golden hue. Rings at intervals more grayish-white. Coronet with a dusky fila- ment. One had a whitish snout. General color more brownish. Brownish lines, edged with white, radiate from eye. Tubercles on trunk and caudal rings at intervals more whitish. An example was similar to pale brownish-white one, but with minute specks. A pale brownish line from eye to end of snout. "Wliitish caudal rings at intervals, and like most of preceding with a more or less evident submarginal dusky line on dorsal, fins otherwise plain and pale-colored. One with pronounced tubercles all more or less pure white, inter- spaces on rings pale olive, those on ventral surface darker, and all more or less variegated with whitish dots. Whitish lines radiate from eye. Dorsal with a submarginal brownish band, fins other«'ise pale whitish. Rings on caudal white and pronounced at intervals. Brown band from eye to tip of snout crossed by several whitish fasciae. 96 PROCEEDINGS OF THE ACADEMY OF [March, Another less tuberculate, more brownish, and variegated with white and pale brown. Darker examples were almost blackish, without vermiculating lines. Side somewhat brownish. Snout whitish. Dorsal and caudal plain- colored. Other examples were more variable. Many others were noted, some olivaceous, brownish, coral-white, or almost entirely white, and others pale lavender. From the above notes it is easily seen that this species is subject to great color variation in life. Some examples are much more brilliant than others. There is also variation in the tubercles. All my ex- amples collected in the sea-weed (Zostera) of the iMarquesas shallows. SCOMBRID^. 41. Scomberomorus cavalla (Cuvier). King Fish. Abundant. TRIOHIURID^. 42. Trichiurus lepturus Liiin;' Sometimes taken. ISTIOPHORID^. 43. Istiophorus nigricans (Lac6pSde). Sail Fish. Almost every season a large example is reported from these Keys. OARANGID^. 44. Trachinotus falcatus (Linnreus). Pampano. Color of young in life, dusky-lavender. A diffuse vertical patch of brownish dots on side from middle of dorsal to anal. Dorsal and anal more or less sprinkled blackish, former otherwise with whitish and latter with pale orange adjoining blackish of anal. Margin of anal and entire caudal white. Caudal peduncle white. Spinous anal red, spotted with black. Breast rosy, with brownish specks. Top of head brown. Jaws and lower surface of head whitish. Cheek specked with brown. Iris red, with brownish margin. One ly\ inches long, from the Marquesas. SERRANIDiE. 45. Petrometopon cruentatus (Lacopede). Red Hind. No examples were seen, but it is reported from the Marquesas and other places. 46. Epinephelus striatus (Bloch). Nassau Grouper. Marquesas Islands. Seen in the fish-pots. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 97 47- Epinephelus maculosus (Valenciennes). Reported. 48. Epinephelus drummond-hayi Goode and Bean. Speckled Hind. Reported from about the Marquesas. 49. Epinephelus morio (Valenciennes). Red Grouper. Color in life olivaceous-brown, branchiostegal region and cavities of jaws salmon-color. A brown streak opposite upper margin of maxillary in preorbital depression. Inside of pharynx vermilion, and inside of mouth fleshy-pink. Fins, except pectoral, dusky or blackish, especially submarginally. Narrow margins of rayed vertical fins whitish. Pectoral golden-brow^n, most noticeable marginally. Base of ventral pale whitish. Trunk marked with pale diffuse or indistinct scattered pale spots. A few dots about size of pin-head on preorbital. A salmon-pink line on base of anal. Iris golden or brassy-brown. Two examples from Bahia Honda Key. About the Marquesas I saw a number of these fishes in fish-pots and in live-boxes. At Snipe Key they were frequently seen about the reefs. 50. Promicrops gnttatus (Linnajus). Jew Fish. One taken in a pool on Boca Chica Key, and a large one off West Cud joe's Key. They were reported as not at all uncommon about the other Keys. Several were also seen in the reef near Snipe Key. LUTIANIDiE. 51. Lutianus griseus (Linnitus"). Mangrove Snapper. Marquesas and Boca Grande Key. Found about the roots and usually under overhanging mangroves, from which places they dart or swim quickly out to the open water when disturbed. They asso- ciate with the other snappers, and in some places are very abundant. Some attain quite a large size. 52. Lutianus apodus (Walbaum). Schoolmaster. A very young example in alcohol which I take to be this species has the body dark brownish with about six pairs oif dark transverse bands, the paler spaces between each much narrower alternately. Vertical fins more or less pale or whitish, spinous portions blackish basally. Pectoral and ventral dusky. A brownish streak from tip of snout to eye and then continued posteriorly along upper side of head to gill- opening above. Iris slaty. Eye about three in head, Preopercle with a short spine at lower posterior corner. Length one inch. Mar- quesas Islands. 7 98 PROCEEDINGS OF THE ACADEMY OF [March, Other examples were found at Big Pine Key. 53. Lutianus aya (Bloch). Red Snapper. Reported from the Marquesas, but no examples taken. 54. Lutianus synagris (Linnaeus). Marquesas Islands. 55. Ocyurus chrysurus (Bloch). Yellow Tail, Color in life purplish-brown above, side rosy-purplish, becoming more purplish below. Jaws rosy. Iris brassy-white, and a red bar behind pupil. A gilt-green streak from tip of snout to caudal, broaden- ing behind and including upper half of caudal peduncle and finally including all of tail. Edge of tail narrowly dull red-orange. A num- ber of large gilt-green blotches above lateral streak. Narrow pale longitudinal gilt lines, lower ones very pale, below lateral band. Dorsal gilt-green. Margin of soft dorsal and base of spinous fin dull diffuse plumbeous. Anal pale milky, medianly pale greenish-yellow. Pectoral dilute pinkish. Ventral milky, first ray dull greenish-yellow\ Lower surface of body wiiite. Inside of mouth and gill-opening white. Length 8 inches. Hailer's Rock. One of the most abundant food-fishes, large numbers being in the markets of Key West during all of our visit. It has been reported from all the Keys by fishermen. IMost of our examples were taken in the deeper channels, like some off Bahia Honda Key. H^MULIDJE. 56. Haemulon sciurus (Shaw). Grunt. Yellow Grunt. Fig. 11 (young). Color in life sky-blue above, inclining or fading to plumbeous below. Lines on snout and upper back in front more or less violaceous. Inside of mouth bright scarlet. Iris silvery. Dorsal plumbeous gray-brown, rayed fin and caudal with slightly more plumbeous and latter becoming pale brownish marginally. Margins of dorsal dull amber, also same shade at base of spinous fin and about seven dull amber blotches basaUy on radii of second dorsal. Pectoral brownish-amber. Ven- tral and anal golden-amber, and streaks on side same. Axil of pectoral pale. The figure represents a young example from the Marquesas. Jew Fish and Bahia Honda Keys, and found very abundant through- out our trip. 67. Haemulon plumieri (Lac(SpMe). Grunt. Boar Grunt. Color in life of young, all olivaceous above, more or less golden or gilded, especially laterally. Head with bright blue lines, on trunk 1906.1 NATURAL SCIENCES OF PHILADELPHIA. 99 immediately becoming pkmibeous and fading. Fins all more or less pale olive-brown, margin of spinous dorsal somewhat pinkish. Base of caudal with a blackish spot. Inside of lower jaw pale orange. Fig. 11. Hcemulon sciurus (Shaw). itm ^ Iris silvery tinted with pale orange above. Length 2f inches. An- other was similar except gilt bands were a little broader, upper surface of body a little more olivaceous, ventral and anal more golden, and mandible white. Very abundant about the ^Marquesas where many w-ere seined. 58. Bathystoma rimator (Jordan and Swain). Youiis Grunt. Grassy Key, Hailer's Rock and the Marquesas. Very abundant in these localities, also about the reefs near Snipe Key, where large schools were seen swimming about in the quiet or still waters. SPARIDiE 59. Calamus penna (Valenciennes). Speckled Porgj-. Fig. 12 (young). Color in life brownish, pale purple above on back, and side and lower surface silvery. About six golden-olive vertical bands with edges of 100 PROCEEDINGS OP THE ACADEMY OF [March, each scale within more or less dusky. Markings extending on fins which are very pale brownish. Iris golden. Length 3^ inches. The Marquesas and Boca Chica Key. Fig. 12. Calamus penna (Valenciennes) 60. Lagodon rhomboides (Linn;tus). Bream. Color in life pale bluish-olive above, below silvery white. Gilt lines above parallel with lateral line, and below horizontal. Vertical lines dusky slate-color, and forming a black blotch at beginning of lateral line. Dorsal spines gilded silvery slate-color. A median broad gilt band entire length of dorsal and anal. Edge of spinous dorsal gilt. Caudal brownish with golden tint. Pectoral and ventral whitish, latter with a gilt line in middle of its length. Inside of gill-opening silvery. Iris silvery with golden and brown. Length 3f inches. Marquesas, Grassy and Big Pine Keys, and Hailer's Rock. GERRIDiB. Shad. 61, Eucinostomus gula (Valenciennes). Color in life bright silvery, back tinted with very pale gray-blue. About four indistinct grayish vertical bands on side of back, most 1906.] NATURAL SCIENCES OF PHILADELPHIA. 101 distinct in young examples. Dorsal and caudal grayish, other fins paler and translucent. Iris brownish with a brownish vertical bar. The Marquesas, Boca Chica, West Cudjoe's and Grassy Keys, and Hailer's Rock. POMAOENTRID^. 62. Pomacentrus leucostictus Miiller and Troschel. Cock Eye Pilot. Color in life with upper half of body anteriorly brilliant violet, this color changing insensibly into yellow, and becoming intense orange- yellow posteriorly as on "caudal and on caudal peduncle. Spinous dorsal like back in front. Rayed dorsal yellowish like back posteriorly. A black blue-edged ocellus at junction of dorsals basally. Lower margin of anal olivaceous-yellow^, rest of fin like abdomen. Pectoral and ventral orange. Three brilliant, or light, blue lines from eyes and snout to occiput. Iris brown. Two small examples, a little less than an inch in length, were taken in the tide-pools off Hailer's Rock. When in the water they are very conspicuous on account of their brilliant coloration. They are not especially active swimmers though shy, the smaller ones usually escap- ing capture on account of their size. Found associated with Blennms cristatus, Cyprinodon mydrus and Glyphisodon mauritii. 63. Glyphisodon mauritii (Bloch). Sheepshead. Color in life, back shaded with lemon-yellow. Five distinct vertical bars of deep plumbeous, edges of scales within each dusky-plmnbeous. Lower siuface of body chalky-white. Fins brownish and transverse bars extending on them. Scales on base of anal specked with dusky, otherwise plumbeous-white. Base of pectoral above, blackish. Each ray and spine of ventral pale dusky, axillary scale white. Head above plumbeous with a deep golden shade. Side of head plumbeous. Iris brown, front rim yellow. One example 4|- inches long from the Mar- quesas shallows. A young example from Hailer's Rock was more greenish-yellow above in life. Below slaty-plumbeous, becoming more or less silvery on ventral region. Vertical bars dusky. Greenish-yellow and dark bars on spinous dorsal, other vertical fins pale-grayish. Pectoral and ventral whitish. Iris brown. Length 1^ inches. S0ARID.ZE3. 64. Cryptotomus beryllinus Jordan and Swain. Parrot Fish. Color in life brown, edges of some scales darker, those on lower sur- face of a light chestnut tint. Dorsal and anal slightly brownish, with 102 PROCEEDINGS OF THE ACADEMY OF [March, a ruddy tint and mottled with sHghtly darker spots. Caudal brown- ish with slightly brownish cross-bars. Pectoral and ventral pale brownish-white with slight cross-markings of a chestnut tint on latter. Eye silvery. Marquesas. Another, from same locality, had ventral and anal more rosy-red. Small ones were speckled with dusky and tinted orange-purple on anal below. Spots on belly pale purple. 66. Sparisoma flavescens (Schneider). Parrot Fish. Marquesas, Big Pine and Boca Chica Keys. . 66. Pseudoscarus guacamaia (Cuvier). Parrot Fish. Green Parrot Fish. I saw a large example of this species swimming about the reef near Snipe Key. In the water it appeared to be entirely dark green. CH^TODONTID^. 67. Chaetodon capistratus LinnEeus. Butterfly Fish. One example seen in the reef about Snipe Key. It was very evi- dently this species and measured about 5 inches in length, as far as I could judge. 68. Pomacaiitlius arcuatus (Linnceus). Black Angel. Color in life blackish. Bars bright lemon-yellow. Each scale on side with a black spot. Iris deep blackish-brown, marginally blackish. Length 4 y g" inches. Only one example was taken about the roots of a red mangrove in the Marquesas shallows. It was also reported by several fishermen as rather common about the roots of overhanging mangroves of several Keys. 69. Angelichthys ciliaris (Linnseus). Yellow Angel. Golden Angel. Blue Angel. Several of these beautiful fishes were seen about the reefs near Snipe Key. They like the protection of large rocks and are exceed- ingly brilliant among their equally showy surroundings. MONAOANTHID^. 70. Stephanolepis hispidus (Linnseus). File Fish. Leather Fish. Color in life green mottled with whitish and brown. Iris pale yellowish. Some examples with fine dusky dots and others with pubic flap edged with pale orange. Subject to great color variation, especially with respect to pattern. Very abundant in the sea-wrack about the Marquesas. Also taken about Hailer's Rock. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 103 OSTRAOIONTIDiB. 71. Lactophrys trigonus (Linnteus). Trunk Fish. Shell Fish. Color in life pale olive with blue spots or dots. Blotch of blackish on side. Fins pale yellowish-brown, caudal more brownish. Base of dorsal blackish. Iris pale yellowish. Length 6^ inches. Marquesas Islands. 72. Acanthostracion tricornis (Liimseus). Horned Trunk Fish. Color in life with belly pale sulphury. Spots on body blackish- brown. Fins pale brownish, caudal with a few spots. Chin whitish. Iris yellowish-white with brown spots. Length 3| inches. Marquesas Islands. TETRODONTID^. 73. Spheroides nephelus Goode and Bean. Puffer. Swelling Fish. Color in life with side of body citron-yellow. Back variegated with pale vermiculating blue lines of dots and marked with black spots. Blue lines also forming ocelli of black lateral blotches. Fins dilute brownish. Iris dilute yellowish. Length 2^ inches. Marquesas Islands. 74. Spheroides plumieri (Schneider). Puffer. Swelling Fish. Color in life brownish above tinted with pale olive and marbled with blackish-brown. Tentacles on back whitish. A dark expanded triangular blotch between eyes. Side of body olive, blotches blackish. Fins pale or dilute brownish. Caudal whitish, and basally and sub- marginally with a broad transverse blackish-brown bar. Lower sur- face of body chalky-white. Iris pale orange-yellow. Length Sc- inches. Marquesas Islands and a number from Boca Chica Kej^ DIODONTID^. 75. Diodon hystrix Lmnseus. Porcupine Fish. This, and sometimes the following, are taken and dried as curios. Nothing further than that both species occur was gathered from the fishermen. 76. Diodon holocanthus Linnrtus. Porcupine Fish. 77. Lyosphaera gibbosa Evermann and Kendall. Marble Fish. Color in life pale olivaceous above. Black below forming reticulat- ing lines with bright orange spots. Black lines reticulating above in olivaceous. Jaws pale brownish. Fins whitish. Iris pale golden. Length i^ inches. 104 PROCEEDINGS OF THE ACADEMY OF [March, This species was reported not at all uncommon in the shallows of the Marquesas about the Zostera, though I only found one example. It is a very curious and interesting little animal and I watched it inflate and collapse several times before death. The inflation was slight, much less than that of Spheroides. This specimen agrees best with the larger figure given by Drs. Evermann and Kendall,^ though the dark reticulating lines are more distinct. Those on top of the head are narrow, or fine, and others are broken a little. Those all over abdomen are broader and darker. SOORP^NID.^. 78. Scorpaena grandicomis Cuvier. Sculpin. Color in life brownish variegated with darker and whitish. A broad brownish band from eye down over cheek. Eye brownish. Fins blackish-brown and whitish. Body marked most everyw^iere with small whitish spots. Length 2^^ inches. Two examples were seined in the sea-wrack of the Marquesas shal- lows. GOBIID^. 79. Mapo soporator (Valenciennes). Color in life mottled brownish above, sides with darker brownish spots. Back also with dusky and olivaceous markings. Lower sur- face white. ]\Iargin of anal narrowly whitish. Ventral white. Other fins more or less grayish and transparent. Iris brownish. Length 2yV inches. Hailer's Rock, in a tide-pool. After being in alcohol there is an evident small bluish shoulder-spot. Another example was brownish in life with a dull bluish tint, spots more dull or pale bluish. Markings on back brownish. Fins brown- ish, spots pale bluish. Outer portions of dorsal pale yellowish, bas- ally spotted with pale bluish, and brownish between. Caudal same. Anal paler bluish than on dorsal, becoming somewhat dilute brownish- yellow distally with a submarginal brownish line, and tip of each ray whitish. Pupil greenish, edged narrowly with golden. One example 4 inches long from the sea- wrack of the Marquesas shallows. PLEURONEOTID^. 80. Platophrys ocellatus (Agassiz). Flounder. Color in life pale sandy-brown, mottled with white, deep brown and gray, white forming distinct spots. Fins also spotted with same '^Bull. U. S. Fish. Comm., XVII, 1897 (1898), p. 131, PI. 9, figs. 11 and 12. Rappahannock River, near the mouth of Windmill Creek, Virginia. (W. C. KendaU. Type No. 48,794, U. S. Nat. Mus.) 1906.] NATURAL SCIENCES OF PHILADELPHIA. 105 colors, dark brownish spots at regular intervals and showing through on dextral side, which is whitish. Iris pinkish. A single small example from the ^larquesas. BLENNIID^. CLIXIN.E. CONGKAMMUS gen. nov. Type Congrammus moorei sp. nov. Body elongate and tapering from head. Head deep, large and broadened posteriorly. Snout short. Eyes small, on top of head, directed upwards, close together, and near tip of snout. Mouth su- perior and much inclined. Lips with a series of large cutaneous flaps or cirri. Teeth in jaws in fine bands. No teeth on vomer or pala- tines. Interorbital space very narrow. No gill-rakers. No small upturned hook on shoulder-girdle. Scales rather large and cycloid. Lateral line continuous and arched till over pectoral. Dorsal not divided, radii similar or no apparent spines, and about 40 in number. Anal extending well forward on abdomen. Caudal free and rounded. Pectoral with tips of lower rays, and those of other fins, with exception of caudal, free. Color pale or brownish-white. Small Blennies found in tropical waters or in shallows with sandy bottoms, and with something of the appearance of the Star Gazers or Uranoscopidce. (Kdyypoc^ or Toyypog^ Congnis, the ancient name of the Conger Eel ; ('fiftocj sand.) 81. Congrammus moorei sp. nov. Sand Conger. Fig. 13. Head 3%; depth of trunk at tip of pectoral 6; least depth of body 7; D. 41; A. Ill, 33; P. 14; V. I, 3; scales 46 in lateral line to base of ~1\ Fig. 13. Congrammus moorei Fowler. caudal, and 1 more on latter; 5 scales obliquely between dorsal and lateral line at tip of pectoral, and 4 scales obliquely back and 106 PROCEEDINGS OP THE ACADEMY OF [March, down from same point to anal; width of head ly^ in its length; depth of head 1^; mandible 2f;. caudal 2^; pectoral If; third ray of anal 3^; snout about 8 in head from tip of upper jaw; eye about 7^; inter- orbital space about half of eye. Body elongate, swollen about branchial region in width but depth at this point less than farther back, and rest of trunk or long tail taper- ing. Upper profile convex, especially above pectoral posteriorly. Caudal peduncle compressed, and its depth nearly a fifth of length of head. Head large, compressed anteriorly and swollen laterally posteriorly, and its greatest depth greater than that of trunk. Upper profile nearly straight, and but Uttle inclined. Lower profile well inclined. Snout very short and a little broad. Eyes superior, directed upwards, close together near tip of snout and longer than broad. Mouth small, superior, and with jaws protruding. Mandible produced beyond upper jaw\ Teeth in jaws, only apparent when mouth is open, fine and in bands. Apparently no teeth on vomer or on tongue. Each lip with a fringe of long cutaneous flaps of even length at regular inter- vals and conspicuous at all times. Tongue thick, rounded, a little posterior in mouth, and free in front. Nostrils directly in front of eye, separated, and anterior with a small cutaneous rim or in a short tube. Interorbital space very narrow and a little concave. Opercle with radiate strise above, each of which is produced beyond posterior margin as a slender cutaneous process over gill-flap and beyond. Gill-opening large, extending forward about midway in length of head. Rakers absent. Filaments small, about equal to orbital diam- eter. Isthmus compressed, and surmounted with broadened sur- face which is concave medianly. Scales only on trunk, except several on base of caudal, wliere they are large. Scales cycloid, large, except on nuchal region anteriorly, and in parallel series. Lateral line of large simple tubes and con- tinuous, at first high or after fourteenth scale of its course midway in depth of body. Dorsal of simple flexible rays, none apparently evident as spines, tips of each free anteriorly, but becoming gradually more restricted by membranes posteriorly where it is free from caudal. Origin of dorsal beginning a trifle before posterior margin of preopercle. Anal with three rather pungent short spines, anteriorly separated from rest of fin, and nearly opposite origin of pectoral. Rest of anal, except first few elongated rays which are more or less erect with first two inclined forward and with well-developed membranes, similar to dorsal. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 107 Caudal rounded and small, free from dorsal and anal. Pectoral large, broad, tips of lower rays free, and median rays longest. Ventrals jugular, each with three graduated rays with more or less free tips. Color in alcohol pale brownish, whitish below, and a trifle darker on top of head. Latter, and back, with many minute pale brown specks or dots. Fins all transparent or brownish-white. Iris whitish, pupil slaty. Color in life white, below brighter, above more translucent and marked on back alternately with pale brown and white. Fins trans- parent whitish, dorsal and caudal hardly darker. Iris pale greenish- white, becoming silvery marginally, and pupil black. Length 2 inches. Type No. 30,621, A. N. S. P. Hailer's Rock, Florida Keys. June 23, 1904. H. W. Fowler. Only a single example, the type, was taken in very shallow water on the sandy shore of Hailer's Rock. (Named for Mr. Clarence B. ^loore, of Philadelphia, well known for his valuable researches in Archaeology, and through whose interest in Zoology the expedition to the Florida Keys was realized.) • BLEXNIIN.E. 82. Blennius cristatus Linnaeus. Kock Fish. Color in life olive-brown above, saddles darker than ground-color. Body speckled or minutely spotted with deeper brown and whitish, markings becoming pale plumbeous below. Lines of dorsal brown. Edge of fins whitish. Caudal tinted a little with pale green. Ab- domen dilute greenish-white and translucent. Fins all very pale transparent brownish, and pale markings dusky or whitish. Lower surface of head pale dusky. Eye brown. Crest on head dull deep red with dusky spots. A larger example with dilute carmine tints to margins of dorsals and upper edge of caudal, posterior margin of latter and tips of dorsal radii whitish. A whitish-plumbeous spot at base of each anal ray and also at tip of each, submarginally dusky. Spots and lines on body dilute plumbeous or pearly-blue. Bars on caudal dusky. Occipital filaments reddish. Axil of pectoral dusky. Brownish bars on lower half of pectoral. Iris brown. Small examples have brownish lines on dorsal more pale brownish. Found very abundant in all the tide-pools on Hailer's Rock and West Cud joe's Key. In such places they are often found completely entrapped, darting quickly about when disturbed, however, soon find- 108 PROCEEDINGS OF THE ACADEMY OF [March , ing shelter in the crevices. They are colored much in keeping with their surroundings, so that it is difficult to make out their presence when quiet. 83. Auohenopterus marmoratus (Steindachner). Rock Fish. Color in life brown, darker markings dark brown. A pale spot on side of head tinted greenish-white. Pale markings on body whitish or pale brown. Caudal whitish. Blotches on dorsals and anals blackish-brown. Iris reddish. One small example from the sea- wrack in the ]\Iarc{uesas shallows. BROTULID^. 84. Ogilbia oayorum Evermann and Kendall. Color in life nearly uniform or pale isabella-color, finely dusted with minute darker specks or dots. Eye slaty. Four examples from West Cud joe's Key were taken in the tide- pools. They all show a distinct opercular spine and have very incon- spicuous or imbedded scales, not arranged transversely or otherwise in indistinct series, the fish having the appearance of being naked when alive. BATRACHID^. 85- Opsanus tau (Linnteus). Color in life brownish, below whitish. Dorsal and caudal pale brown, markings darker. Paler markings creamy. Iris brownish. One 3f inches from the Marquesas. Another, 1^ inches long, from West Cud joe's Key has no ventral sucking-disk. Color-pattern suggesting that of certain Cottidae. Blotches on body and vertical fins posteriorly very large and confluent, so that contrast between light and pale markings is pronounced. Belly and abdomen immaculate whitish. OGCOCEPHALID^. 86. Ogcocephalus radiatus (Mitchill). Toad. Plate III. Color in life deep slaty-brown, without any distinct markings. Dorsal, caudal, anal and pectorals same color above and basally below, and marked with a beautiful network of whitish lines forming more or less distinct ocelli. Sides of trunk similarly colored. Lower distal portion of pectoral, ventral, anal and caudal tinged with golden. Under surface of body otherwise pale or soiled whitish. Iris brownish with pale radiating line. Length 7|- inches. Big Pine and Ballast Keys, about the shallows. They do not move fast and are easily captured in the hand or in a small dip-net. They 1906.] NATURAL SCIENCES OF PHILADELPHIA. 109 lurk about the rocks or sea-wrack, and their protective colors make it difficult to distinguish them when quiet or not moving about. AMPHIBIANS. BUFONID^. 87. Bufo quericus Holbrook. In Hfe plumbeous-gray, markings dusky slate-color. Lower surface of body translucent whitish. Soles of feet and toes brownish. Iris with yellowish next to pupil. Abundant in the open grass-land on Big Pine Key. They progress by a series of short jumps. HYLID^. 88. Hyla squirella Daudin. In life plumbeous above speckled wdth dull brownish. A blackish- brown line from tip of snout along side of head and including eye in its course. Lower surface of body pale brownish-white. Legs pale translucent brown, whitish below and posterior feet dusky. Iris brown. This species was found on the ground about shady hammock-land on Boca Chica Key. I also saw one on a Gumbo Limbo tree on Vaca Key. 89. Hyla carolinensis (Pennant). A small green frog, evidently this species, was seen on Key West and Boca Chica Key. REPTILES. COLUBRID^. 90. Natrix fasciata (Linntens). Moccasin. One example was taken in the mangroves of Boca Grande Key. Here they live in the pools among the roots and short sprouts, and are thus difficult to see on account of the seclusion afforded. I did not meet with it in any other locality, though it doubtless occm-s as it was reported from No Name Key. The vernacular is ''Moccasin," as distinguished from the "Cotton Mouth Moccasin" (Agkistrodon contortrix). 91. Bascanion constrictor (Linnaus). Black Snake. Reported from No Name Key. Seen on Knight and Summerland Keys. 110 PROCEEDINGS OF THE ACADEMY OF [March, 92. Lampropeltis doliatus (Limiffius). Red Snake. An example, reddish or coppery-red in color, was found on Boca Chica Key. Others were reported from No Name, Smnmerland and Cudjoe's Keys. OROTALID^. 93. Crotalus adamanteus Beauvois. Rattler. Rattle Snake. Diamond Rattle Snake. Reported from Boca Chica, Big Pine and No Name Keys. I ob- served it on Summerland and Sugar Loaf Keys, and on the latter one was secured. SCINOID^. 94. Eumeces fasciatus (Linnaeus). Lion. General color in life waxy-black. Lines on head and neck deep brick-orange, becoming rather more orange on trunk, and finally fad- ing into deep ultramarine-violet tail, though latter at first azure. Line along side of head paler or more ochraceous. Lower surface of head pale or translucent, mandible somewhat dilute purplish and throat greenish. A tinge of pale brick-color on lower side of head. Iris brownish. Lower surface of trunk dusky with brassy reflections. Feet and legs black. This beautiful little lizard was first noticed in the hammock-land on the east end of Boca Chica Key. They were also rather plentiful on Vaca Key, where a rather large one was secured. TBID^. 95. Cnemidophorus sexlineatus (Linnteus). Lagarto. Lion. In life upper surface brownish, and from above hind legs to tip of tail grayish-brown. Lower sm-face entirely chalky or livid dilute lavender-blue, inclining more to bluish laterally. A median pale streak from occiput along upper sm-face of tail to become joined with its fellows. They unite with a slightly ochraceous one on each side, though becoming pale gray on tail, and extending parallel from occiput. A bright yellow line similar in shape and course from above posterior margin of eye and extending back along side of tail where it also be- comes gray. From lower eyelid below another of pale lemon to hip and fading out on femur. Still another, and paler, fading out on ribs in front. It is indistinctly defined. Eyelid pale yellowish. Iris brown. Lower side of head pale bluish. Top of head pale brown. Inside of mouth flesh-colored. Upper surfaces of fore legs grayish- brown. In sutures of sciuamation on posterior costal region, hind 1906.] NATURAL SCIENCES OF PHILADELPHIA. Ill legs, and front of tail, brick-colored patches. Lines fade out on tail which is almost entirely gray a short distance from its base. A brown- ish-gray line on posterior side of hind leg, leaving a white one between it and color above. It extends along lower side of tail w^here it is blue- green, and though continued a good distance back is not well defined. Toes and soles of feet pale brownish flesh-color. Key West, Boca Chica, Big Pine, Knight, Vaca and Grassy Keys. The above notes are from an example from the former locality. They were abundant on most Keys, where they were seen running swiftly through the grass. They soon hide in holes or among rocks. About the lagoon on Key West I found them most common under low grassy vegetation where numerous insects, such as small grasshoppers, were seen. IGUANIDiE. 96. Anolis principalis (Linnaus). Lion. Key West, Boca Chica, Marquesas, Big Pine, No Name, Vaca, Grassy and Summerland Keys. Many bright green examples were seen on Key West among grasses and other green vegetation, from which they are distinguished with difficulty. Those on the ground, or on brown sticks, etc., were brown in color. I never saw the animal change to the negative of its surroundings. Thej^ are quick, though more easily captured than the other species. CROCODILID^. 97. Alligator mississippiensis (Daudin). Alligator. Reported from several pools we saw on Little Pine Key, thougli I did not see any example during our stay. They were also reportefl from pools on Summerland Key, and were said to occasionally reach six feet in length. DERMOOHELYDID^ . 98. Dermoclielys coriaoea (Vaiuk-lli). Trunk Back. Frequently taken about the Keys. Several shells were seen about Key West. CHELONIID^. 99. Thalassochelys caretta (Linnaeus). Logger Head Turtle. Logger Head. Reported as al^undant in the past and still frequently taken. It, like both of the following species, is much sought after, and bids fair to become exterminated in a short time if the ruthless depredations of the turtle-fisher are not speedily checked. These animals, all of which 112 PROCEEDINGS OF THE ACADEMY OF [March. breed in suitable places in the Keys, are thus subjected to addi- tional danger as their eggs are also held in demand. These are eagerly gathered when found and used as food. 100. Eretmochelys imbricata (Linnseus). Hawk's Bill Turtle. Shells of this animal were seen in Key West. It was reported as of frequent occurrence in past times. 101. Chelonia mydas (Llnnseus). Green Turtle. Turtle. Hunted for food and the most familiar of the family. Several times newly laid eggs, most likely of this species, were found on the sandy beaches of the different Keys. EMYDID^. 102. Malaclemmys littoralis rhizophorarum subsp. nov. Terrapiu. Diamond Back Ter- rapin. Plate IV. Head large, heavy, flattened somewhat above, and below, and a little swollen laterally behind each eye. Eye large. Snout short, and rather obtuse. Jaws each with a broad horny covering. ]\Iouth large, and its profile together with mandible convex. Legs and feet strong, furnished with stout claws, five on each anterior limb and four on each posterior. Claws on fore feet rather broad, and those on hind feet narrow and longer. Scales on limbs well developed supe- riorly. Tail short, conic, and rather weak. Carapace slightly ovoid when viewed from above, with its greatest width opposite posterior portion of fourth vertebral plate. Anterior carapace a little emarginate. Carapace deepest when viewed laterally about middle of its length, or tops of crests of second and third verte- bral plates highest with profile sloping down gradually in front and behind. Lower margin of carapace sloping gently down to a point above suture of seventh and eighth marginal plates, then turns up- ward and outward a little to descend again behind. Vertical and horizontal measurements of marginal plates forming bridge with plastron, or fifth, sixth and seventh, about equal. Edges of marginal plates from sixth backwards sharp and becoming more revolute to most posterior. Axillary and inguinal plates well developed. Each vertebral plate, except posterior which is evenly convex, with a median keel surmounted by a more or less well developed osseous tubercle posteriorly, and best developed proceeding to last. Each plate of carapace with conspicuous concentric rings. Plastron a trifle ovoid, and its greatest width would fall a trifle in advance of its middle or about middle of bridge. Front of plastron nearly truncate or its anterior edge only very slightly emarginate. Behind bridge sides of plastron nearly parallel to posterior margin of femoral plate where there is a prominent notch, and then more convergent to ends of anal 1906.] NATURAL SCIENCES OF PHILADELPHIA. 113 plates which are furnished with a deep notch on median Une. Epider- mal plates of plastron smooth. Color in life. Carapace deep dusky, with perhaps faint or obscure darker mottlings. Inferior portion of marginal plates of carapace of the same shade of pale yellowish-brown as the plastron, though at each sutiu-e a broad blackish blotch. These blotches are most distinct and conspicuous on the bridge, those of posterior plates very narrow. Plas- tron dull or pale yellowish-brown, and articulation of each plate with a broad irregular blackish margin, at least anteriorly. Head, legs, feet and tail pale-slaty or plumbeous, and former speckled or spotted with a dark or dusky-slaty. On legs this is evident as each scale has a dark or slaty center. Jaws whitish or of a whitish horn-color. Top of head pale or light. Eye grayish. Claws horn-color, dark or brownish above. Length of carapace 7^ inches. Type, No. 16,570, A. N. S. P. Boca Grande Key, Florida. June 18, 1904. H. W. Fowler. Clarence B. Moore Collection. This form is known to me only from the above described example, which was collected in a pool among the mangroves of Boca Grande Key together with Natrix. It did not appear to me to be especially abundant on the islands we visited, whether for seclusive habits or other reasons, I could not tell. It was reported from other islands within the limits of our trip, however. As it is highly valued as food it is much sought after. Provisionally, as a form of Malaclemmys littoralis of Prof. Hay,^ from the Texas coasts, it may be recognized as a distinct subspecies. Although he had a very large series of individuals, the sutures of the marginal plates of the carapace are not described as black, which is the case in the example before me. His figures do not show this, and certainly the colored one of the adult is different. I have recently had the opportunity of examining a large series of several hundred living examples of Malaclemmys centrata (Latreille), through Dr. Samuel G. Dixon, in the markets of Philadelphia. Most of these were said to have been procured near Savannah, Georgia. Great variations in color were noticed, some individuals approaching the above described example very closely, wdiile others had perfectly uniform plastrons, etc. Most of the variations, like those figm-ed under Emys concentrica by Sowerby and Lear,* were found. {Rhizophora, Mangrove.) ' Bull. Bureau of Fish., XXIV, 1905, p. 18. « Tort. Terrap. Turt., 1872. Four figures referable to Pis. 33-55, and 36, are not numbered in the copy before me. 8 114 PROCEEDINGS OF THE ACADE1VI.Y OF [March, NEW, EARE OR LITTLE KNOWN SCOMBROIDS, NO. 3. BY HENRY W. ^O^VLER. CHLOROSCOMBRIN^. Chloroscombrus hesperius sp. nov. Fig. 1. Head 4; depth 2i; D. i, VIII-I, 26; A. II-I, 25; P.'ii, 17; V. I, 5; scale.s 98? (squamation injured) in lateral line to base of caudal; depth of head, over middle of orbit, Ij^ in its length; width of head 2^; mandible 2; third dorsal spine 2y\; second dorsal ray 2; first anal ray 2^^; ventral 2^; snout 3f in head, measiu*ed from tip of upper jaw; eye 2f; maxillary 2^; interorbital space 3f ; least depth of caudal peduncle 5^; upper caudal lobe 3^ in head and trunk; pectoral 2f . Body rather shorter than in C. chrysurus, strongly compressed, Chloroftcombrns hesperms Fowler. ovoid, lower profile much more convex than upper and greatest depth at base of spinous anal. Caudal peduncle long, slender, compressed, and its least depth about 3 in its length. Head deep, compressed. Snout short, forming obtuse profile, upper surface convex and upper jaw little produced. Eye rather large, 1906.1 NATURAL SCIENCES OF PHILADELPHIA. 115 circular, and a little high and anterior. Adipose eyehd broad poster- iorly, and covering a good portion of iris. Mouth obliquely vertical, rather small, and mandible well protruded. Maxillary reaching op- posite front margin of eye, and its distal expanded extremity a trifle less than pupil or about f of orbit. Teeth minute, in bands in jaws, and also present on vomer, palatines and tongue. Tongue long, rather narrow, with rounded tip, and free in front. Lips thin, except lower on side of mandible. Interorbital space elevated and with a median elevated trenchant keel extending to occiput. Nostrils adjoin- ing, near upper .front rim of orbit and posterior larger. Gill-opening extending forward about opposite posterior nostril. Rakers 10+33, slender, compressed, and longest f of orbit. Fila- ments about -f- length of longest rakers, and pseudobranchiae about same size. Isthmus narrowly compressed, and its edge not trenchant. Scales small, cycloid, and forming convex series bulging posteriorly over costal region. A few small scales on occipital region, postocular region, and on opercle above, head otherwise naked. Spinous dorsal and anal depressible in narrow scaly sheaths. Soft dorsal and anal anteriorly with basal scaly sheaths and rays more or less depressible within. Lateral line convex anteriorly, its bulge about equal to If in straight portion, and no keel. Spinous dorsal small, inserted nearly midway between tips of upper jaw and pectoral, and fom-th or fifth spine probably longest (damaged). Soft dorsal inserted nearly midway between front rim of orbit and base of last dorsal ray, a little in front of straight part of lateral line, and first few rays elongated a little. Soft anal similar, inserted a little in advance of soft dorsal. Origin of spinous anal about last third in space between origin of ventral and that of soft anal, second spine probably a little longer (first damaged). Caudal rather large, lobes slender, pointed, and upper much longer. Pectoral long, slender, and reaching about opposite first third of base of anal. Ventral in- serted a little behind origin of pectoral, small, both fins depressible in a groove extending to and containing vent, and when depressed thus reaching J of space to spinous anal. Color in alcohol more or less silvery, and back above brownish. A small dull or pale brown spot on upper edge of opercle. Head pale brown above. Eye dull brassy. Fins pale brownish, spinous dorsal and upper anterior dorsal rays with a little brownish. A blackish blotch at base of upper caudal lobe. Length 6f inches. Type No. 11,198, A. N. S. P. West Africa. Dr. H. E. Savage. 116 PROCEEDINGS OF THE ACADEMY OF [March, This species is closely related to Scomber chloris Bloch, with which it may possibly prove identical. However, that form, if the original figure is to be trusted, is very close to Scomber chrysurus Linnaeus. It will be seen to differ from C. chrysurus in the slightly deeper form, larger eye, more anterior insertion of the anal, and longer pectoral. The gill-rakers are about the same. {'EfTTTipco?, western, with reference to the type locality.) TRACHINOTIN^. GLATJCUS Walbaum. Klein, in Walbaum, Pet. Arted. Gen. Pise, III, 1792, p. 585, with Scomber amia Linnseus as type, has priority over Hypacantus Rafin- esque, Carat. Alcun. Nuov. Gen. Spec. An. Piant. Sicilia, 1810 (April 1st), p. 43 {=Hypacanthus =Hypacantha =Hypodis =Hypodys Rafinesque = Lichia Cuvier ^Porthmeus Valenciennes). Glaucus amia (Linnseus) examined. CAMPOGEAMMA Regan. Campogramma Regan is distinct, as an examination of Campo- gramma glaucus (Linnseus) shows. TEACHINOTUS Lac6p6de. PAMPANOA subgen. nov. Type Choetodon glaucus Bloch. Anterior rays of dorsal and anal prolonged so that they form long falcate lobes. Caudal long and triangular. Side with several dark transverse vertical bars, at least above. Trachinotus glaucus (Bloch) examined. (Pdmpano [Spanish], a common vernacular for the fishes of this genus and mostly written Pompano.) Subgenus TRACHINOTUS Lacep^de. T. ovatus (Linnseus), T. falcatus (Linnaeus), T. goreensis Cuvier and T. carolinus (Linnseus) examined. STROM ATEIDiE. STROMATEIN^. Subgenus STROM ATEUS Linnfeus.i Stromateus brasiliensis sp. nov. Fig. 2. Head 4J; depth 2i; D. V, 40; A. V, 37; P. II, 18; width of head 1^% in its length; snout 3f ; eye 4f; maxillary 3f ; interorbital space 2f : ^Stromateus fiatola Linnseus examined. 1906. NATURAL SCIENCES OF PHILADELPHIA. 117 least depth of caudal peduncle 2|; pectoral (damaged), a little less than head; upper lobe of caudal (damaged) about 3^ in head and trunk; about 125? scales in a lateral series from gill-opening below lateral line to base of caudal. Body ovate, deep, compressed, and greatest depth about origins of rayed dorsal and anal. Back rather trenchant before dorsal. Caudal peduncle compressed, its least depth about f of its length. Head deep, compressed, blunt in front, profiles steep, and upper a little concave in front. Snout blunt, convex and profile in front nearly vertical. Eye rather small, circular, a little low, and posterior margin -4^^^^^. Fig. 2. — Stromateus hrasiliensis Fowler. about midway in length of head. Adipose eyelid well developed, and concealing a good portion of eye. Mouth a little inclined, small, and mandible hardly protruded in front. Maxillary not quite reaching front rim of orbit, and distally expanded till about equal to pupil. Teeth minute, uniserial, and forming a rather weak cutting edge. Palatines with minute asperites. Vomer edentulous. Tongue smooth, broad, rounded, and free in front. Nostrils adjoining, a little nearer front of snout than front margin of eye, and posterior larger. Inter- orbital space elevated, trenchant, a keel beginning at the internasal region. Preopercle broad and with radiating flutings. Opercle striate. 118 PROCEEDINGS OF THE ACADEMY OF """' [March, Gill-opening extending forward about opposite front margin of eye. Rakers 5 + 13, pointed, longest about | of longest filaments which are f of orbit. Pseudobranchise a trifle longer than longest raker. Isthmus rounded. Scales small, and extending well out on most of fins, almost entirely on dorsals and anals where they are greatly reduced and crowded. Head naked except cheek and opercle. Lateral line damaged, convex at least anteriorly. Dorsal apparently over middle of pectoral, spines graduated to last which is longest though not as high as longest anterior rays. Anals similar though origin of spinous well posterior, a little before end of pectoral. Only first 7 or 8 rays of rayed dorsal and anal a little longer than others. Caudal large, forked, and lobes rather broad. Pectoral broad, its origin a little above lower rim of orbit. Anus placed nearly opposite origin of spinous dorsal. Color in alcohol pale brown, mostly uniform, or the back only^a trifle darker and marked with indistinct traces of numerous crowded slightly darker or grayish spots. Fins pale brown, pectoral a trifle darker. Iris dull brassy. Length 13f inches. Type No. 11,354, A. N. S. P. Rio Grande do Sul, Brazil. Mus. Comp. Zool. Cambr., Mass. One example which does not agree with Gay's figure of Stromateus maculatus? This is very poor, as no squamation whatever is indi- cated, the fins are naked, also the head, there is no adipose eyelid, opercle and preopercle without striae, and the spots are greenish- golden, the lower ones brighter. The dorsal shows VI, 38, and the anal III, 32. Jenyn's Chiloe example ^ had D. VII, 41, and A. V, 50, and the one from Port St. Julien in Central Patagonia is possibly identical with S. hrasiliensis. The original account of S. maculatus by Valenciennes states that the spots are yellow. {Brasiliensis, of Brazil.) * PTERORHOMBUS subgen. nov. Type Fiatola fasciata Risso. Fiatola Risso, Hist. Nat. Eur. Merid., Ill, 1S26, p. 289 (fasciata). (Not of Cuvier.) Small ventrals present. Strornateus fasciata (Risso) examined. 2 Ail. Hist. Chile, Zool., 1854, "lam 3bis/' fig. i. ' Zool. Voy. Beagle, IV, Fish, 1842, p. 74. * Hist. Nat. Poiss., IX, 1833, p. 296. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 119 (Urspov^ fin, with reference to ventral; 'popi^oi, Rhombus, a rhomb or diamond, and formerly applied to species of Seserinus, Palometa and Poronotus.) SESEKINTJS Quoy and Gaimard. Seserinus Cuvier, in Quoy and Gaimard, Voyage de TUranie, ZooL, Jan.-Mar., 1825, p. 384 {xanthurus), has priority over Peprilus Cuvier, Regne Animal, Ed. II, II, 1829, p. 214 {longipennis = paru), now in use. Cuvier's account in I.e. . Ed. I, II, 1817, p. 342 {alepidotus = paru) cannot be used on account of its vernacular form. Seserinus xanthurus Quoy and Gaimard examined. NOMEID^. Psenes chapmani sp. nov. Fig. 3. Head 3; depth H; D. X, I, 23; A. Ill, 25; P. ii, 17; V. I, 5; scales 60 in lateral line to base of caudal ; 9^ scales vertically between origin of soft dorsal and lateral line, and about 25 vertically between latter and origin of soft anal ; width of head 2 in its length ; depth of head over middle of eye about equal to its length ; snout 3i ; eye 3 ; maxillary 3 ; interorbital space 3; least depth of caudal peduncle 3; pectoral Ij; ventral If. Body deep, compressed, rather ovoid, profiles similar, and greatest depth about origin of soft anal. Caudal peduncle compressed, and its least depth about 1\ in its length. Head deep, compressed, profiles steep, and similar, blunt in front. Snout short, blunt, convex, upper jaw but slightly projecting. Eye rather large, circular, anterior, and its lower margin about midway in depth of head. Mouth small, jaws about even, and maxillary extend- ing obliquely down barely opposite front margin of orbit. Teeth in jaws uniserial, rather pronounced, well separated and sharp. No teeth on roof of mouth. Tongue rounded, and free in front. Nostrils small, very close together, and well in front of eye above. Inter- orbital space convexly elevated. Gill-opening extending forward about opposite front rim of orbit. Rakers 11+20, clavate, a little shorter than filaments which are aljout 2 in orbit. Pseudobranchise nearly as large as filaments. Scales small, those above lateral line forming series parallel with its course, and those below forming horizontal longitudinal series. Except interorbital space, supraocular region, nasal region, snout, mandible, and branchiostegal region, head is covered with scales, those on middle of cheek and opercle largest. About 7 series on cheek. Scales in front of spinous dorsal small and crowded. Vertical fins 120 PROCEEDINGS OF THE ACADEMY OF [March, almost entirely scaled over most of basal regions which form sheaths for spinous dorsal and anal and also rayed dorsal and anal at least anteriorly. Only ends of caudal lobes naked. Scales on bases of these fins also small and crowded. Base of pectoral scaly. Ventral with innermost ray united to abdomen by a membrane and both fins de- pressible in a groove. Lateral line cm-ved similarly with profile of back, also approaching it more closely posteriorly, and extending a little above middle of side of caudal peduncle. Lateral line not extending on base of caudal and tubes simple. Spinous dorsal inserted a trifle in advance of origin of pectoral, spines flexible, and graduated to foiuth which is longest. Soft dorsal Fig. 3. — Psenes chapmani Fowler. inserted a little nearer tip of snout than base of caudal or about over middle of pectoral, and fin a little higher anteriorly. Anal similar, spines graduated to third which is longest, and rayed fin inserted per- haps a trifle in advance of rayed dorsal. Caudal forked, lobes rather long and pointed. Pectoral broad, hardly reaching half way to tip of last anal ray. Ventral inserted a little behind origin of pectoral, and reaching second anal spine. Anus in ventral groove before tip of ventral fin. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 121 Color in alcohol rich warm brown, lower surface paler, evidently- silvery in life. Along ea^'h series of scales on side of body a narrow streak or line of darker color than general body-color. Above lateral line these also follow series of scales parallel with it, and below it they are also horizontal. Lower side of head including cheek and opercle spotted or dotted with brown. Spinous dorsal and ventrals duskj^- brown. Rayed dorsal with more or less pale dusky-brown on its outer portion. Caudal and anal paler. Pectoral pale brownish. Inside of mouth pale. Iris brassy. Peritoneum whitish. Length 3f inches. Type No. 14,621, A. N. S. P. Taken on the surface of the open Atlantic in Lat. 24° 21' N., Long. 34° 32' W. (northwest of Cape de Verde Islands). Dr. W. H. Jones. Also seven paratypes with same data. The very young differ in ha^dng the body marked with large blotches, no longitudinal lines on side, and lower parts of head together with thorax marked with numerous brown spots. Larger example with 28 rays. Young examples also from the open Atlantic's siu-face in Lat. 23° 53' N., Long. 53° 06' W. (northwest of the Lesser Antilles). Another also from the Atlantic in the Saragossa Sea. All from Dr. William H. Jones. The East Indian form ^ needs comparison. Psenes guttatus appa- rently differs in coloration, especially the spotted thorax and lower surface of the head. None of my examples show a trace of the bluish bar indicated by Valenciennes. Psenes edwardsii Eigenmann* has a little different radial formula and about 140 scales in the lateral hne. (Dedicated to Dr. Henry C. Chapman, Professor of Institutes of Medicine and Medical Jurisprudence in the Jefferson Medical College of Philadelphia, who has contributed many valuable additions to the ichthyological collections of the Academy.) LBPODID-ffi fam. nom. nov. LEPODUS Rafinesque. Carat. Alcun. Gen. Spec. An. Plant . Sicilia, 1810, p. 53 {saragus = raii) is to be used, as Brama Klein, in Walbaum, Pet. Arted. Gen. Pise., Ill, 1792, p. 586, with Cyprinus brama Linnaeus as type, preoccupies Brama Schneider, Syst. Ichth., Bloch, 1801, p. 98, based on Sparus rail Bloch. The Pomfret, Lepodus rail (Bloch), examined. 6 Hist. Nat. Poiss., IX, 1833, p. 193, PI. 265. ^Bull. U. S. Fish. Comm., XXI, 1901 (1902), p. 35, fig. From under a medusa, 30 miles south of Newport, Rhode Island. 122 PROCEEDINGS OF THE ACADEMY OF [March, OORYPHu^NIDuE. Coryphoena equiselis Linnaeus is of perfectly correct form, and not to be emended to C. equisetis as some writers seem to suppose. Further, it is not a misprint for the latter but simply a different form of the same. A small example, 25 mm. in length, from between the Cape de Verde Islands and Montevideo (Uruguay), examined. PEMPHERIDJB. PEMPHEEIS Cuvier. PRIACANTHOPSIS subgen. nov. Type Petnpheris mulleri Poey. Anal rays 25 to 32. Pempheris mulleri Poey examined. (Upcwv, saw; aKavOa, spine; 6(pi^, appearance. So named as these fishes resemble the young of Priacanthus.) This paper concludes the series. 1906.1 NATURAL SCIENCES OF PHILADELPHIA. 123 MOLLITSCA OF THE SOUTHWESTERN STATES. II. BY H. A. PILSBRY AND J. H. FERRISS. The present paper deals with molliisks of Texas, New Mexico and Arizona, collected chiefly by the authors in 1903, and not included in the first paper of this series.^ No attempt has been made to present complete faunas ; but in connection with the previous paper, about all the reliable data extant upon the snail faunas of the Chiricahua and Huachuca ranges in southeast Arizona and the eastern and southern borders of the Lower Sonoran area in Texas will be found herein.^ It has not been thought expedient to repeat data elsewhere accessible. For the student of molluscan distribution, the life zones of the United States as mapped by Dr. Merriam ^ emphasize the secondary and not the primary facts of distribution. The laws of temperatui'e control, which he has developed with keen insight, do not define transcon- tinental zones of primary import zoologically. These zones are sec- ondary divisions of vertical life areas of which the molluscan faunas were evolved in large part independently. The Sonoran fauna is probably intermingling more now with that of eastern North America than at any former time, at least so far as such sedentary forms as land mollusks are concerned. Similar conclusions have been reached by Mr. A. E. Brown in deal- ing with Texan reptiles.* The results of his study "establish three facts, hitherto not wholly free from imcertainty : first, that the bound- ary between the Austroriparian and Sonoran reptilian faunas lies ap- proximately between the 96th and 98th meridians of longitude in Texas; second, that the restricted Texan district of Cope is not Aus- troriparian but Sonoran; third, that transcontinental zones of distribu- tion cannot be maintained in the Medicolumbian region for reptiles. ' ' 1 Proc. A.N.S. Phila., 1905, p. 211. ^ Some New Mexican records, chiefly from material collected by Prof. T. D. A. Cockerel! in tlie upper Pecos valley, are added. ' Biological Survey, U. S. Dept. of Agriculture, Bulletin No. 11, Map. North American Fauna, No. 25, Plate I. ^ Texas Reptiles and their Faunal Relations, Proc. A. N. S. Phila., 1903, pp. .543-558. Post-Glacial Nearctic Centers of Dispersal for Reptiles, Proc. A.N.S. Phila 1904, p. 464. 124 PROCEEDINGS OF THE ACADEMY OF [March, ". . . . the community in reptiles between the Sonoran and the Austroriparian is much less than that between the Sonoran and the Central; it is, in fact, limited to widely ranging genera .... the separate identity of these two faunas [Sonoran and Austroriparian], as far back as they can be traced, confirms the conclusion already reached, from their present distribution, that the association of the Sonoran with the Austroriparian into one transcontinental zone is unnatural." This terse statement only requires the substitution of ' ' mollusks ' ' for "reptiles" to serve as a summary of the results of oiu- present in- vestigation. In land mollusks, as in reptiles, the Sonoran types have pushed farther into the humid Austroriparian than the eastern types into the Sonoran area. Bulimulus and the texasiana group of Poly- gyra are cases in point ; while Eastern subgenera of Polygyra, and prac- tically all Austroriparian forms except the minute, widely ranging genera, stop short at the rise marking the approximate limit of the Cretaceous formation and the Sonoran area in Texas. The common boundary of the Austroriparian and Sonoran, while surprisingly sharp for continuous areas, is deeply accidented by the river valleys, which carry very narrow Austroriparian ramifications into Sonoran borders. Thus, along the San Marcos river in Hays county, the Guadalupe river in Comal county, and the San Antonio river in Bexar county, an Austroriparian fauna with such Eastern forms as Omphalina, Poly- gyra thyroides, P. monodon fraterna, P. auriformis, P. roemeri, Bidimu- lits d. liquahilis, etc., is found, while the fauna on the bluffs or away from the streams is frankly Sonoran. Farther southwest, the Aus- troriparian forms are wholly wanting, even where moist local condi- tions prevail, as in the inunediate vicinity of the streams of Val Verde county. A list of the forms characteristic of the eastern or lower portion of the Sonoran in Texas (Texan District of Cope) follows : Praticolella berlandieriana Strohilops I. texasianus Thysanophora hornii Zonitoides nummus Polygyra hippocrepis Vitrea indentata umbilicata Polygyra texasiana hyperolia Vitrea dalliana roemeri Polygyra t. texasensis Euconulus cher sinus trochulus Polygyra mooreana^ Helicodiscus eigenmanni Bulimulus d. mooreanus Planorhis carus Bulimulus d. ragsdalei Planorbula obstructa * This species extends somewhat into the Austroriparian. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 125 Bulimulus d. pecosensis Goniohasis comalensis Bulimulus alternatus marioe Palvdestrina seemani Holospira roemeri Paludestrina diaholi Holospira goldfussi Ammicola comalensis Microceramus texanus Cochliopa riograndensis Euglandina singleyana Potamopyrgus spinosus Bifidaria procera cristata Valvata micra Nine genera of this list are not known to occur in the Austro- riparian or humid region of Texas. For list of the latter fauna, the student is referred to the catalogue of Mr. J. A. Singley, cited below, from which a long list of Austroriparian forms of Eastern type may readily be compiled. The fii'st hst of Texan mollusks of any extent was published by Romer in his excellent work on Texas, 1849. Numerous references to the terrestrial mollusks will of course be found in Binney's successive volumes. In 1878 Mr. A. G. Wetherby^ published some notes on the forms he found in eastern Texas {American Naturalist for 1878, pp. 184, 254). The principal source of information, however, is Mr. J. A. Singley's Contributions to the Natural History of Texas, part I, Texas Mollusca, published in the Fourth Annvxil Rep. Geol. Survey of Texas, 1893. In this list Mr. Singley has included with the records of his own extensive collecting, others from many sources, so that the records are of unequal value. The Texan list stands much in need of revision and a good many names thereon are doubtless to be rejected, either because the forms do not occur in Texas, as in the case of Ampullaria, or because of wrong identifications ; yet the work cannot be done until resident naturalists take it up. HELICINID^. Helicina orbiculata tropica 'Jan' Ptr. Texas: San Marcos, Hays county; Comal county, around New Braunfels; San Antonio, Bexar county; two miles north of Hondo, Medina county; Del Rio, Devil's river and High Bridge of the Pecos, Val Verde county. Some colonies are all white; others are mingled with red or blue shells. HELIOID^. Praticolella berlandieriana (Moric). Figs. 1. 2. Texas: San Marcos, Hays county; Guadalupe river above New ' By error Mr. Wetherby's name was printed "W. G. Weatherby 126 PROCEEDINGS OF THE ACADEMY OF [March, Braunfels, Comal county; San Antonio, Bexar county; Hondo river, two miles north of Hondo, Medina county; Del Rio, Val Verde county. Mr. Ferriss took some thin, translucent specimens, 8x10 mm., at Smith ville. The shell varies from 9.8 to 11.7 mm. diam. The pub- Fig. 1. Fig. 2. lished figm-es do not show that there is frequently a short lamellar tooth or callous on the parietal wall near the suture, a short distance within, as shown in the ^figured specimen from the Guadalupe river near New Braunfels. Thysanophora hornii (Gabb). Texas: Hondo river, near Hondo, Medina county; Rio San Filipe, near Del Rio; Devil's river, west of Del Rio, and Pecos river at the High Bridge, Val Verde county, all in river drift. New Mexico: Florida mountains, near Deming, Grant county. Arizona: Cochise county; Cave creek canyon in the Chiricahua mountains; Fort Bowie; drift of San Pedro river at Benson, and Manilla mine, at the west end of the Huachuca mountains. This small species is a typical Thysanophora, closely related to the type of the genus. It varies but little throughout its extensive range, unless it be in the cuticular lines and hairs, which are worn off of the river-drift specimens. It has not before been reported in this country from east of Cook's, Grant county, N. M. Our collection of 1903 extended its known range in the United States more than 500 miles eastward, to the border of the Staked Plains in central-southern Texas. Further localities in Arizona are given in Nautilus, XII, p. 99. In Mexico, T. hornii has been collected at Topo Chico, near Mon- terey, Nuevo Leon, and at Victoria, Tamaulipas {Proc. A. N. S. Phila., 1903, p. 763). It is therefore one of the widely distributed species of the States both north and south of the international boundary in the Sonoran area. Polygyra auriformis (Bid.)- Helix auriformis Bland, Ann. Lye. Nat. Hist, of N. Y., VII, p. 37 (1858). ? Helix sayii Wood, Index Testaceologicus, Suppl., p. 22, pi. 7, p. 34 n (1828) . Texas: Galveston (J. H. Ferriss); Calhoun county (Hubbard); 1906. NATURAL SCIENCES OF PHILADELPHIA. 127 Austin and San Antonio (Pilsbry, 1885); along the Guadalupe river, three or four miles above New Braunfels (Ferriss and Pilsbry). Sing- ley adds the localities Bastrop and Burleson counties. It is an Austro- riparian species which reaches into the borders of the Lower Sonoran area along the river bottoms. While the identity of Wood's H. sayi with auriforfnis is not estab- lished with sufficient certainty to make a substitution advisable, in our opinion, yet there can be no doubt that Wood had either that species or one of its immediate allies. Helix sayi of Binney, 1840, being a homonym, may be changed to Polygyra sayana {Terrestrial Mollusks, III, pi. XXIII). Polygyra hippocrepis (Pfr.)- Figs. 3, 4. The known range of this curious snail is exceedingly restricted. It has been found only near New Braunfels, in Comal county, Texas. We found it in abundance in April, 1903, on the west side of the Guadalupe river about four to six miles north of New Braunfels, under stones near and at the foot of the bluff, with Holospira goldfussi. Another place much nearer the town is on the rocky wooded hillside above the springs of Comal creek, where we found a few living ones. This place is just beyond the pleasure gardens. Fig. 4. The structm-e of the aperture of P. hippocrepis has never been fully described or figured. The parietal tooth is not V-shaped as in other PolygjTas, but U-shaped, hence the name hippocrepis — horseshoe. There is an internal tubercle on the columella, as in P. mooreana. The upper and lower lip-teeth enter and arch towards each other, forming a U-shaped curve, which stands close to but a little deeper than that formed by the parietal tooth. Where the two entering teeth join there is a notch and a delicate slender hook with the point curved towards the adjacent basal wall projecting forward. The immersion of the originally lower lip-tooth gives room for a secondary callous ridge along the basal lip, as shown in fig. 4. 128 PROCEEDINGS OF THE ACADEMY OF [March, These structures form in their way a more perfect barrier perhaps than that produced by other means in P. auriculata and uvulifera. The species stands isolated at present. It is perhaps an aberrant and highly evolved relative of the Tennessee-Arkansas group of P. jack- soni, etc. Polygyra texasiana (Moricand). PI. V, flgs. 16, 17, 20. Specimens of typical P. texasiana were taken by us in Hays, Comal and Bexar counties, Texas. Ferriss took it at Galveston. The type locality is ''Texas." In this widespread form the last two whorls are strongly rib-striate above, the riblets rapidly diminishing on pre- ceding whorls ; the embryonic whorl is smooth and glossy. The ribs are strongest just behind the lip, and here continue upon the periphery or to the base, which is elsewhere nearly smooth or merely rippled. Fresh shells show a reddish peripheral band on the pale brownish- corneous surface. Specimens figured are from the west side of the Guadalupe river above New Braunfels, Texas. Alt. 5, diam. 11 mm. Along the Rio Grande P. texasiana occurs with transition forms to P. t. hyperolia. See below. Form with striate base. In some localities the ribs of the upper sur- face continue upon the base (pi. V, figs. 18, 19, Calhoun county, Texas), the other characters being unchanged. There are transitions to the normal sculpture of texasiana in some specimens, and we do not think it desirable to distinguish this form at present by a special name. Its distribution must be more fully worked out than we have been able to do. Calhoun county is on the Gulf coast near the southern angle of the State. P. texasiana hyperolia n. subsp. PI. V, figs. 13 14, 15. Shell more depressed than texasiana, glossy, very finely striate, almost smooth, above and below, with several riblets behind the lip-constriction. Uniform brownish-corneous or paler beneath, without a peripheral haxid. Aperture smaller and slightly more oblique than in texasiana. Alt. 4, diam. 9.3 to 10 mm. The type locality is the high land west of Devil's river. This is the common Polygyra along the Rio Grande in Val Verde county, extend- ing north and northwest. The specimens from down the river, at Hidalgo, .... county, and Laredo, Webb county (collected by Singley), are either texasiana or transitional between texasiana and hyperolia in sculpture. At Del Rio, along the Rio San Filipe, Ferriss and I found still the texasiana and transition forms. On the high land west of the Devil's river, Val Verde county, we found hyperolia in some numbers, under prostrate Yucca trunks and 1906.] NATURAL SCIENCES OF PHILADELPHIA, 129 sometimes concealed in the shelter of the downward-drooping dead leaves on standing Yuccas. From this region they have been washed into the Devil's river, where bleached shells are abundantly found in the drift debris. We also took specimens in the high land along the Pecos river, near the High Bridge (Viaduct, on some maps), east of the river, and in drift debris in the canyon. In the interior counties of Texas we took specimens in the drift of the Hondo river, Medina county, about two miles north of Hondo. Much farther northward Mr. Ferriss found hyperolia at Colorado City, Mitchell county, Texas. In New Mexico Prof. J. D. Tinsley collected it on South Spring creek, near Roswell, in the Pecos valley, where it occurred fossil in a bed of white marl, three to four feet below the surface. It is not now found living in that locality. We are indebted to Prof. CockereU for these specimens.'' P. t. hyperolia varies about as much as texasiana in size. Specimens from west of Devil's river measure from 3.8x8 mm., with 4^ whorls, to 4.2 X 11 mm., with 5 whorls. This variety is not the Helix {Poly gyro) tamaulipasensis of Lea, which is typical P. texasiana, as I have ascertained by a comparison of the type kindly made for me by Dr. W. H. Dall. Polygyra texasiana texasensis (Pils.). PI. V, figs. 11, 12. Polygyra texasensis Pils., Nautilus, XVI, p. 31, July, 1902. Similar to P. t. hyperolia in the depressed shape and smooth surface, without riblets above; but decidedly larger, with about 5^ whorls, the umbilicus broader than is usual in the other forms of the species. Alt. 5.2, diam. 12.5 mm. Alt. 5.8, diam. 12.5 mm. Alt. 5.8, diam. 13.7 mm. Colorado City, Mitchell county, Texas ; types collected by James H. Ferriss in 1902, No. 83,258, A. N. S. Pliila. This is a large edition of P. t. hyperolia, and further collections from the almost unknown northwestern half of Texas are needed to deter- mine whether it is really distinct enough from that subspecies to call for recognition of the latter by name. In the large series of hyperolia collected there were, however, no forins as large as texasensis. Polygyra mooreana (W. (3. Binn). PI. V, figs. 4-10. This species stands close to P. texasiana, but the shell is smaller than ' Reported as P. triodontoides in The Nautilus, XIII, November, 1899, p. 84. 9 130 PROCEEDINGS OF THE ACADEMY OF [March, any but the smallest texasiana, and with about the same number of whorls (5 to 5^) appears more closely coiled. It is constantly dis- tinct by the elongated columellar tubercle within the last whorl (shown in fig. 9). This tubercle seems to be what Binney erroneously describes as "an internal transverse tubercle on the base of the shell" — an ex- pression which would lead one to expect such a structure as that found in the Stenotremas. Binney states that the color is ''white," but fresh shells are brownish-corneous, a little paler and somewhat transparent at the base. The diameter varies from 6.5 to 8 mm. Variation in the degree of elevation is shown in the figures. P. mooreana was originally described from Washington county, Texas. We took specimens at Smithville, Bastrop county, San Mar- cos, Hays county, around New Braunfels, Comal county (figs 4^7), and on the Hondo river north of Hondo, Medina county (figs. 8, 9, 10), ever3nvhere in copious quantity. There are also specimens before us from Washington county, Belton, Fort Worth, Waco, Austin and Lee county. Further records are given by Mr. Singley in his valuable catalogue. It was taken at Galveston by Ferriss. Polygyra mooreana tholus (W. G. Binn). PI. V, figs, l, 2, 3. Larger than mooreana, with the last whorl far more openly coiled below, exposing more of the penultimate whorl in the much wider umbilicus. Specimens from Washington county measure : Alt. 5.5, diam. 11, width of umbilicus 4.2 mm.; whorls 7. Alt. 4, diam. 8.2, width of umbilicus, 3 mm.; whorls 6. This form holds such a relation to P. mooreana as P. d. sampsoni to P. dorjeuilliana, being very openly coiled beneath, with a deep and very long groove on the last whorl within the umbilicus. The internal ridge is like that of P. mooreana. It seems to be comparatively local in distribution and is certainly rare. We have never seen a fresh specimen. It was described from Washington county, Texas, the locality of the specimens figured. It is in the collection of the Academy also from Calhoun county. Ferriss took a few specimens at Galveston, and Sing- ley found it in Fort Bend and Brazos counties. Von Martens reports a worn specimen found by Fried el at Vera Cruz, Mexico. Polygyra roemeri (Pfr.)- Smithville, Bastrop county, Texas (Ferriss); San Antonio, Bexar county (Pilsbry and Ferriss) ; near New Braunfels, Comal county, one specimen (Pilsbry). BULIMULID^. The genus Bulimulus in Texas has caused much perplexity to stu- dents. Many and diverse have been the views held as to the rank 1906.] NATURAL SCIENCES OF PHILADELPHIA. 131 and identity of the several forms. Several times since the first visit of one of us to Texas, in the winter of 1885-86, they have been carefully studied anew, with each time larger materials and better knowledge of the country. To the specimens already in the museum of the Academy about 1,000 were added by our expedition of 1903, from central and western Texas and Indian Territory. The series is divisible into two species, B. dealhatus and B. alternatus marioe; the first further split into numerous local races or subspecies. The typical forms of B. alternatus and B. schiedeanus do not, so far as we know, extend into Texan territory; both were described from Mexico where they range over a large area. B. patriarcha W. G. B., unquestionably a form of schiedeanus, was also described from Mexico (Buena Vista), and we have seen no specimen of it from within our borders.^ In the United States, Bulimulus has not been found west of the neighborhood of El Paso. We know^ nothing of what forms live in the State northwest of a line from Austin to San Antonio, or north of the Southern Pacific Railroad from San Antonio westward to El Paso. Beyond some extension of the ranges of B. dealhatus mooreanus and B. alternatus marioe, little is to be expected from the great unknown area. The following forms are now recognized : 1. B. dealhatus mooreanus W. G. B., Pfr. Arid region of central and south Texas. 2. B. d. liquahilis Rve. Eastern and southeastern Texas. 3. B. dealhatus Ssiy. Alabama to Kentucky, west to Kansas. 4. B. d. ozarkensis P. and F. Northern and western borders of the Ozark uplift. 5. B. d. ragsdalei Pilsbry. Bluffs of Red river and southwestern Texas. 6. B. d. pecosensis P. and F. Southwestern Texas. 7. B. d. pasonis Pilsbry. El Paso, western Texas. 8. B. alternatus marioe (Albers). Southern Texas. Of these forms, the anatomy of B. d. mooreanus, B. d. liquahilis, B. d. ragsdalei, B. d. pecosensis and the Val Verde county race of B. alternatus marioe has been examined more or less fully. All agree in having a rather short, fusiform penis with a basal sheath which in- * We do not mean to deny that the forms mentioned occur in Texas. The coun- ties along the Rio Grande west of the mouth of the Pecos are still unexplored for shells, and B. schiedeanus especially may turn up in this region. 132 PROCEEDINGS OF THE ACADEMY OF [March, eludes the vas deferens; a rather long and slender epiphallus, and a flagellum. An atrium can scarcely be said to be developed, but the vagina is rather long in all the forms. The globular spermatheca is borne on a long duct, which is lightly bound to the oviduct and is usually somewhat swollen near the middle. The right eye-retractor passes between the d^ and 9 branches, and there is an excessively weak and short penial retractor about I mm. from the end of the flagellum, and inserted on the hmg floor. B. a. marice differs strongly from all forms of B. dealbatus by the great length of the penis with its appendages, and of the duct of the spermatheca. Moreover, the penial organs are longer than the spermathecal duct, while in mooreanus and pecosensis the spermathecal duct is the longer. These differences confirm the opinion, derived Fig. 5. — A, Bulimulus a. marice. B, B. d. liquabilis. C, B. d. pecosensis. D, penis of a full_v mature individual of same E, B. d. ragsdalei. F, penis of another individual. G, B. d. mooreanus. Fig. D X 2§; the other figures X 2. from a study of the shells, that B. alter nat us and dealbatus are well differentiated specifically. The measurements of the genitalia in millimeters follows: 1906. NATURAL SCIENCES OF PHILADELPHIA. 133 Total lengthl Length \ of penis, I of penis Length epiphallus \ (to insertion of and of vas vagina, flagellum. \ deferens'). Length ! of sperm a- Museum theca with number, duct. B a. maria Pecos river Del Rio 58 45 34 29 7 5 84,627 84,635 B. d. liquabilis San Marcos 36 29 11 27 91,396 B. d. mooreanus Guadelupe river 23 16 6 30 i 84,628 B. d. ragsdalei Devil's river j 16 10.5 5 17 84,638 B. d. pecosensis 19 24 11.5 .5 j 6 20 84,618 Proceedings of the Academy of Natural Sciences of Pliiladelphia, 1906. Page 183 — Under heading Length of Spermatheea with duct, first line of table, change 5 to 50. On last Hue of table change 1.5 to 15. Distribution, eastern division of the Lower Sonoran in Texas. Originally described from ''Washington and DeWitt counties, Texas." Specimens from both of these are before us, and also from Hays, Comal, Guadalupe, Bexar, Medina, Frio, Uvalde and Victoria counties. Some shells from Fort Worth and Waco, in northern Texas, are also like mooreanus, while others resemble B. d. liquabilis. B. d. mooreanus is smaller, thinner and smoother than the Mexican B. schiedeanus (Pfr.), but intergrades with B. d. liquahilis Rve. through certain specimens having ragged dark streaks, chiefly on the spire, such as fig. 5 of pi. VI. In the main, the subspecies is quite uniform 132 PROCEEDINGS OF THE ACADEMY OF [March, eludes the vas deferens; a rather long and slender epiphalliis, and a flagellum. An atrium can scarcely be said to be developed, but the vagina is rather long in all the forms. The globular spermatheca is borne on a long duct, which is lightly bound to the oviduct and is usually somewhat swollen near the middle. The right eye-retractor passes between the <3^ and ? branches, and there is an excessively weak and short penial retractor about 1 mm. from the end of the flagellum, and inserted on the lung floor. B. a. marice differs strongly from all forms of B. dealbatus by the great length of the penis with its appendages, and of the duct of the spermatheca. Moreover, the penial organs are longer than the spermathecal duct, while in mooreanus and pecosensis the spermathecal duct is the longer. These differences confirm the opinion, derived ©r\ Fig. 5. — A, Bulimulus a. mariae. B, B. d. liquabilis. C, B. d. pecosensis. D, penis of a fully mature individual of same E, B. d. ragsdalei. F, penis of another individual. G, B. d. mooreanus. Fig. D X 2§; the other figures X 2. from a study of the shells, that B. alter natus and dealbatus are well differentiated specifically. The measurements of the genitalia in millimeters follows: 1906.1 NATURAL SCIENCES OF PHILADELPHIA. 133 Wotal length j of penis, epiphallus and flagellum. ■ Length \ of penis {to insertion' of vas deferens) . Length vagina. Length of sperma- theca with duct. Museum number. B a. marioe Pecos river Del Rio 58 45 34 29 7 5 84,627 84,635 B. d. liquabilis San Marcos 36 29 11 1 27 1 91,396 B. d. mooreanus Guadelupe river 23 16 6 30 84,628 B. d. ragsdalei Devil's river . ! 10.5 5 17 84,638 B. d. pecosensis 19 1 24 11.5 1.5 j 6 j 1 20 84,618 B. dealbatus mooreanus (' W. G. B.,' Pfr.). PI. VI, figs. 1-6. Pfr., Monographia Hel. Viv., VI, p. 143 (1868). Bulimusschiedeanus in part, W. G. Binnej^, Terr. Moll., IV, p. 129, pi. 80, fig. 8. Bulimulus schiedeanus var. mooreanus W. G. Binnej^, Terr. MoU., V, p. 392, figs. 277, 278; Man. Amer. Land Shells, p. 400, figs. 438, 440. The shell is thin, ovate-conic, opaque white above, coffee-with-cream col- ored below the periphery, or sometimes either white or coffee-tinted throughout ; sometimes varied with a few gray streaks, or some scattered translucent-gray dots; surface smooth, with the spire more or less striate, apical whorls waxen or dark. Umbilicus narrow. Interior cream-white. Five specimens from New Braunfels measure: Alt. 26 22.5 25.5 23 22.5 mm. Diam. 16.5 13.5 14 12 13 Apertiu-e 15 12.8 12.6 11.8 12.8 " Distribution, eastern division of the Lower Sonoran in Texas. Originally described from ' ' Washington and DeWitt counties, Texas. ' ' Specimens from both of these are before us, and also from Hays, Comal, Guadalupe, Bexar, Medina, Frio, Uvalde and Victoria counties. Some shells from Fort Worth and Waco, in northern Texas, are also like mooreanus, while others resemble B. d. liquabilis. B. d. mooreanus is smaller, thinner and smoother than the Mexican B. schiedeanus (Pfr.), but intergrades with B. d. liquabilis Rve. through certain specimens having ragged dark streaks, chiefly on the spire, such as fig. 5 of pi. VI. In the main, the subspecies is quite uniform 134 PROCEEDINGS OF THE ACADEMY OF [March, in characters, and easily recognized. It is the only Bulimulus found in the greater part of the region it inhabits, and is one of the com- monest forms in collections. The specimens seen from Fort Worth are not unlike those referred to B. d. liquabilis from Waco, and it is not easy to decide upon which race they belong to. They are undifferentiated forms. In Comal, Guadalupe, Bexar and Medina counties we found only typical mooreanus. In Frio county the shells are somewhat more solid and often whiter. In a series of 313 specimens from along the Guadalupe river above New Braunfels (pi. VI, figs. 2, 3, 4) I could find but one shell with any of the ragged stripes of B. d. liquabilis. In 165 taken along the river below San Antonio (pi. VI, fig. 1) there were 7 with some ragged stripes, at least on the upper whorls. A series of 55 mooreanus from Victoria had 4 striped shells (pi. VI, figs. 5, 6). The percentage of intergrading specimens is therefore small. B. dealbatus liquabilis (Reeve). PI. VI, figs. 7-:2. Bxdimus liquabilis Reeve, Conch. Icon., V, pi. 57, fig. 387 (Dec, 1848). Bulimus confinis Reeve, Conch. Icon., V, pi. 86, fig. 643 (Feb., 1850). Bulimulus schiedeanus Pfr., W. G. Binney, Terr. Moll., V, p. 391, fig. 276, but not the description. The shell is thin, variable in shape but usually obese, the aperture half the total length or more. Translucent-corneous or brownish- corneous, more or less profusely marked with opaque whitish ragged streaks. Interior whitish or colored like the outside. This is the form of the humid Austroriparian zone in Texas, as B. d. mooreanus is of the arid division. Specimens are before us from the following counties: Dallas, McLennan, Coryell, Bell, Travis, Lee, Hays, Jackson, Nueces. Also from Limestone Gap, Indian Territory. Where the range of liquabilis adjoins that of mooreanus there is a belt of overlapping ; but so far as we know the two do not actually occur together. At San Marcos, Hays county, for instance, B. d. liquabilis lives on the moist low alluvium along the San Marcos river, while B. d. mooreanus is found on the Cretaceous limestone hills above the town. It seems that the one form extends finger-like up some of the streams, while the other may occupy intervening upland or calcareous stations. B. liquabilis and B. confinis were both described from "Texas." B. d. liquxibilis differs from mooreanus by the predominance of corneous-brown coloring. It is also usually more globose and less smooth. It is more globose than B. dealbatus, with less convex whorls, as a general rule, yet there seems to be practically complete intergradation between the races, and some Texas shells are not dis- 1906.] NATURAL SCIENCES OF PHILADELPHIA. 135 tinguishable from those of Alabama, though most others could not be matched from east of the Mississippi. On the whole, the race has intimate relations to those east and west of it, yet seems in its own area to be sufficiently differentiated to be recognized. Specimens from Jackson county (collected by J. D. Mitchell) are typical of liquahilis. They are small (pi. VI, figs. 9, 10, 11), well striped as usual, variable in the length of the spire, and generally have a strong, narrow lip-rib. ram. Alt. 18.7 20.5 18.5 18 21 Diam. 12 12.5 11 10 12 Aperture 11.5 11.5 10 9.5 11.5 Whorls 5i 5f 6 6 6* Shells from Lee county (J. A. Singley), are larger, more globose, cor- responding to Reeve's B. confinis. The corneous-brown tint largely predominates over the reduced whitish streaks. Lip-rib present in adults (pi. VI, fig. 12). Alt. 19 22 25 17.6 mm, Diam. 14 14 15.2 11.5 " Aperture 12 13 14.8 10.3 " Whorls — 6 6* 6 Waco specimens mostly have the spire longer, as do those from Belton, Austin and San Marcos. In the series from the latter locality (Ferriss and Pilsbry, 1903) there are about equal numbers of typical striped shells and nearly unicolored corneous-brown ones, with indis- tinct whitish streaks alternating with pale reddish on the spire (pi. VI, figs. 7, 8). They measure: .It. 25.5 25.5 26 mm. •iam. 14 15 14.5 " .perture 14 14 14 The genitaha of one of the unicolored shells are figured (fig. B). The individual supplying the preparation was that shown in fig. 7 of plate VI. Fig. 8 of the plate, a streaked shell, was found similar anatomically. Judging by these shells, the race liquahilis is quite appreciably different from mooreanus in the proportions of the geni- talia. Some specimens from Limestone Gap, Indian Territory, evidently belong to this race. 136 PROCEEDINGS OF THE ACADEMY OF [March' The shells figured by Binney as B. schiedeanus {Manual of American Land Shells, p. 399, fig. 437) are apparently B. d. liquobilis. They are certainly not the true schiedeanus, of which figures may be found in the Manual of Conchology. Bulimulus dealbatus (Say). Pl.VI, fig. 13. Helix dealbata Say, Journ. Acad. Nat. Sci. Phila., II, p. 159, 1821. Say gave the localities "Missouri and Alabama." His specimens in the Academy Museiun are labelled Alabama. One of these three, agreeing with the dimensions given by him, is figured (pi. VI, fig. 13), as the type of the species. The Alabama form has a globose last whorl and rather large umbUicus. The aperture is a little more than half the total length of the shell, but much less than the diameter of the shell. The shell is profusely striped with ragged white stripes on a pale brownish ground. The type measures, alt. 19.4, diam. 12.2, length of aperture 10.6 mm. The same form occm-s in Tennessee, Kentucky (Warren county), and west to Kansas (Shawnee county). Some specimens from the humid portion of Texas also seem to belong here. B. d. ozarkensis n. subsp. PI. VI, fig. 14, 15. B. dealbatus Say, Pilsbry, Proc. A. N. S. Phila., 1903, p. 204 (Seligman, Mo.). On the northern and western borders of the Ozark uplift this form has been differentiated. The whole shell, and especially the last whorl, is narrower, the aperture is smaDer, ordinarily half the total length or less, and the umbilicus is narrower. The shape is about that of B. d. ragsdalei, but the last whorl is not rib-striate, though the spire is weakly so. Coloration as in dealbatus. Specimens measure: Locality. Seligman. Rogers Mam. Spring. Alt. 26 21.6 21 mm. Diam. 13.2 10.5 11 Aperture 12.8 10 10.8 " Whorls 7 7 6i Limestone Gap, I. T. Alt. 22 21 20 20 17 mm. Diam. 11.3 11.2 9.7 10.8 9.4 " Aperture 11 10.2 9.5 10.8 9 Whorls H 6i 6| 6i 6 Distribution, northern and western outliers of the Ozark system: Mammoth Spring, Fulton county, Arkansas ; Sehgman, Barry county, southwestern Missouri (fig. 14); Rogers, Benton county, Arkansas (fig. 15) ; Lmaestone Gap, Choctaw Nation, Indian Territory. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 137 While only weakly ciiaracterized, this form seems to range over a considerable area, and apparently deserves recognition by name. B. dealbatus ragsdalei (Pils ). PI. VI, flgs. 16, 17. Nautilus, III, p. 122; Man. of Conch., XI, p. 129. The shell varies from the ovate shape of typical dealbatus to a more lengthened and slender form, and is conspicuously rib-striate, the striae white on a tawny or white-blotched ground and weaker on the base of the shell. The lip-rib is strongly developed. Three adult specimens of the type lot measure: Alt. 21.5 18.5 16.5 mm. Diam. 10.8 10.3 8.3 " Aperture 10 9.3 8 Whorls 6* 6i 6 Only dead, more or less bleached shells have been taken from the top of the Red river bluff at the southern end of Warren's Bend, twenty-five miles northwest of Gainesville, Cooke county, and a mile north of St. Jo, Montague county, Texas. Across the whole State of Texas nearly 400 miles distant, this form reappears on the Rio Grande river, in Val Verde county, in an area inhabited also by B. alternatus marice. We are quite unable to find any constant differences between these Southwestern shells and the Red river types, although the apparent absence of the form in the inter- vening territory suggests that the similar forms of the two areas are independent parallel modifications of dealbatus stocks, rather than actually connected genetically. Yet it is quite possible that the widely sundered colonies have been or still are connected through the great conchologically unknown area northwest of the oblique line across the State marking the limit of our explorations. Dead, bleached shells were found in abundance on the high land west of Devil's river, but the living ones for some time eluded us. Finally we found them hidden under the dead reversed leaves which thatch the trunks of Yuccas, and sometimes under prostrate dead Yuccas — retreats they share with the smooth race of Polygyra texasiana (pi. VI, figs. 18 to 22). The proportions vary a good deal, a series of adult shells measuring : 17 mm. 6^ Alt. 26.5 20 20 21.5 19 16.5 Diam. 12.5 9.2 10.3 10 10.8 10 Aperture 12.5 9 10 10.5 10.7 9.3 Whorls 7i 7 61 6^ 6 5i 138 PROCEEDINGS OF THE ACADEMY OF [March, The average size of 78 living shells is about 19 x 9 to 10 mm. Fig. 18 represents the largest shell taken, a dead one 26.5 mm. long. There is but little variation in sculpture among shells from this place. Along the Rio San Filipe, not far from the Rio Grande, in chapparal on the east side, we found numerous specimens differing from those of Devil's river by having the rib-striae nearly obsolete on the last whorl except just below the suture. A large number of dead shells were found, but only very few living ones (pi. VI, figs. 23, 24). They have the dull reddish, white-streaked coloration and the shape of the Devil's river ragsdalei. Bulimulus dealbatus peoosensis n. subsp. PI. VI, figs. 26, 27. B. d. schiedeanus var., Pilsbry, Man. of Conch., XI, p. 132, pi. 17, fig. 6. The shell is conspicuously calcareous, whitish with some fleshy or sometimes corneous or ochraceous streaks ; upper whorls striate, the last somewhat roughened by irregular growth- wrinkles. Spire long, composed of numerous short convex whorls, the suture nearly hori- zontal; apex white or pale; aperture small, usually ochre-tinted in the throat, lip strengthened by a rib within. Alt. 31 29.7 26.5 24 22.8 21 mm. Diam. 14.8 14 12.8 12 12.7 10.7 " Aperture 15 14 12.7 10.3 11.5 10 " Whorls 7i 7i 7i 7i 7 7 T)rpe locality, on the mesa about 1^ miles southeast of the eastern end of the High Bridge of the Pecos (Southern Pacific Railroad), Val Verde county, Texas. We found one small colony of this form, in the midst of the large dark-mouthed B. alternatus marioe, and like that chiefly living on Agave. The extent of the colony was perhaps not more than 50 yards, but as the sun had already set, and we had just emerged from the labyrinthine side canyons of the Pecos, we had time to collect only about thirty-five specimens, each, most of them dead. Everywhere else in the region around the High Bridge we found only B. a. marice. This form is clearly a stunted race of the larger and less slender B. schiedeanus of the Mexican fauna. B. schiedeanus has been con- sidered specifically distinct from dealbatus by Binney and all the older authorities, as well as by von Martens, who gives a series of good figures in the Biologia Centrali Americana; but while the typical schiedeanus is distinct enough, there are not lacking specimeiLs sug- gesting intergradation with some forms of dealbatus. If schiedeanus 1906.] NATURAL SCIENCES OF PHILADELPHIA. 139 be retained as a distinct species, then tl^e race pecosensis will be rele- gated to it as a variety. Some specimens of pecosensis taken about thirty years ago by Dr. H. C. Wood, probably, as he informs' me, from somewhere in the "Great Bend" of the Rio Grande, were noticed in the Manual of Conchology, Vol. XI. Bulimulus dealbatus pasonis Pils. PI. VI, flg. 25 Pilsbry, Nautilus, XVI, July 1902, p. 32. The shell is smaller and more slender than any other form of deal- batus, the diameter about half, the aperture less than half the total length of the shell; nearly smooth, being sculptured with irregular growth-wrinkles only. The dead specimens are dull reddish-corneous with some streaks and mottling of opaque white. Wliorls 5f to 6, quite convex. Aperture small, ovate, the insertions of the lip and columella markedly approaching, without an internal lip-rib in the specimens seen. Umbilicus comparatively large. Alt. Diam. Aperture Franklin mountain, near El Paso, Texas. (J. H. Ferriss.) This is the most distinct, as it is the most remote, of the races of B. dealbatus, and unless connecting forms come to light, it may well be given specific rank. The absence of regular striation on the spire, the converging ends of the lip and the diminutive size combine to give it individuality. As yet but few specimens have been found, and in a single place. Some individuals of the Val Verde county B. d. ragsdalei approach pasonis in size, but in a series of about 200 examined none approach the other characters of the present race. Bulimulus alternatus mariae (Albers). PI. VII. Die Heliceen, p. 162 (1850). Binney, Terr. Moll., V, p. 390, figs. 272, 273, pi. li a, upper and lower figs.; pi. li b, all figs. This species is recognized by its dense, solid, calcareous texture, oblong shape and colored, usually dark brown or purplish interior, and by the great length of the penis and spermathecal duct. Its range extends in a wide belt along the Rio Grande, from the Gulf at least to the Pecos river. How much farther west we do not know. B. marioe was named by Albers for his daughter Mary, having been recognized as distinct on her birthday. The type, figured by Pfeiffer, is a ragged-striped shell with only a low prominence, hardly to be called 15.7 15.3 mm. 7.5 7.8 " 6.8 6.7 " 140 PROCEEDINGS OF THE ACADEMY OF [March. a tooth, on the columella. Shells of this exact form and coloring^occur at Laredo, Webb county, on the Rio Grande (pi. VII, fig. 28). In a general way, the eastern (typical) form of marice from Frio county and Laredo eastward differs from the western (Val Verde county) form in several respects: Southeastern Form. Western Form. Smaller, more oblong; lip more Larger, more conic; lip less thick- thickened; columella frequent- ened within; no columellar ly toothed; often with ragged tooth ; of ten streaked , but with- streaks, at least on the earlier out ragged or mottled streaks. whorls. PI. VII, lower three PI. VII, upper two lines of lines of figures. figures. We do not think it advisable at present to make a subspecifie separa- tion, especially since none of the names proposed apply to the western race. The genitalia of this race are figured (fig. A). While almost every colony of the eastern race has slight peculiarities in shape, relative abundance or absence of the several color-forms, etc., yet with large series the intergradation is seen to be so complete that we can find no ground for dividing them into several races, as Prof, CockereU proposes. His plan {Journal de Conchyliologie, 1891, pp. 23, 24) is as follows : 1, With gray or brownish markings: a. Without a columellar tooth, alternatus Say. b. Columellar tooth present, marice Alb. 2, Without color markings : a. With no columellar tooth, alhidus Taylor, 6. A columellar tooth, . . hinneyanus Pfr. MS. W. G. B. [ = in- termedius Singh, Ckll., Corpus Christi.] The true B. alternatus (type in coll. A. N. S.) has not been found north of the Rio Grande, and in aU probability does not reach our limits. The name hinneyanus is preoccupied, so that were the race valid, we should use for that form the name intermedius SingleyMS. CkU., based on Corpus Christi shells. The type locality of alhidus Taylor MS. Ckll. is Derby, Frio county, Texas. Many perfectly adult Corpus Christi shells have no columellar tooth, and hence we should have to divide them between intermedius and alhidus. Similarly, various specimens selected from the Derby, Frio county, series before me would be either alternatus, alhidus or inter- medius; while from the Hidalgo series, all four of the supposed varieties 1906.] NATURAL SCIENCES OF PHILADELPHIA. 141 may be selected. The study of good series of shells taken at random and not selected shows that all lots of streaked shells are more or less mingled with white ones, and in colonies of toothed shells, fully adult and old individuals may be found without a tooth. On the other hand there seem, however, to be colonies without ragged-streaked individuals, and also communities in which no toothed shells are to be found. Perhaps the colonies containing mottle-streaked or toothed individ- uals are mixed or hybrid communities. It is a case where experiments on Mendelian lines by some one on the ground might be productive of valuable results. In illustration of the foregoing remarks, a few of the colonies repre- sented in the collection of the Academy may be noticed in more detail. In studying these races it is absolutely essential that the snails be col- lected alive. The colors, particularly of the interior, are evanescent and fade quickly on exposure to the sun and weather, though they change very little if at all in the museum. Corpus Christi, Nueces county (pi. VII, figs. 13, 14, 15, 16), collected by J. A. Singley. Of thick-set compact contour, strong, with a thick cream-tinted lip-rib, which is sometimes brown stained. Pinkish white, almost uniform with some very inconspicuous grayish or fleshy streaks. Not one in a lot of 43 is marked with hrown streaks outside. Interior varying from pale yellow to reddish brown, the darker tint exceptional. Columella varying from strongly toothed to nearly straight. Whorls 6, the earliest post-embryonic not distinctly striated. This form is the "binneyanus Pfr. " of Binney, intermedins Singley MSS. of Cockerell. Alt. Diam. Apertm-e At Hidalgo, Hidalgo county (pi. VII, figs. 17, 18, 19), the shells are more slender, white, rarely streaked throughout (fig. 17), but almost all of a series of 41 are variegated on the earlier whorls (fig. 18). Some are strongly toothed, but most shells have no columellar tooth. Interior brown. Alt. 30 29.5 29 28.5 mm. Diam. 13.5 13.5 13 14 Aperture 14 14.5 13.5 14.5 " A series from Brownsville consists of similar but less elongated shells. 29 27.3 30 25 23 mm, 16 14.5 15 14 12 16.3 14 15.3 14 11.5 " 142 PROCEEDINGS OF THE ACADEMY OF [March, At Derby, Frio county (pi. VII, figs. 20, 21, 22, 23, 24), the shells are bluish white with bluish streaks, a few brown tinted with brown streaks. The columellar tooth is small or wanting. This form has been called albidus Taylor, Ckll. At Laredo, Webb county (pi. VII, lower line of figs.), the shells are large and mostly long. The lot consists of pm-e white, and of coffee- tinted, brown-streaked shells, with all intermediate forms. A colu- mellar tooth is exceptionally developed. At Del Rio, Val Verde county (pi. VII, figs. 9, 10, 11, 12), we found a more conic i^ce. The shells are white, often with some very faint grayish streaks. Interior varying from ochre-yellow to dark brown. In the series of about sixty living shells before us, none has mottled markings, and while in a few there is a weak indication of the colu- mellar tooth, it is as a rule absent. The larger ones measure 29 to 33 mm. long, 16 wide. Some of the dead shells found in the same place are larger, 35 x 18.5 to 37 x 17 mm. A single albino was taken alive (fig. 12). These shells are from the plain along the San Fihpe river, on the east side, not far from the Rio Grande. B. d. ragsdalei was found in the same place. On the mesa west of Devil's river we found large dead shells like those from Del Rio but even larger, often with a low lump on the parietal wall, but no tooth on the columella. 39 x 20.5 to 35 x 16 mm. At the Pecos High Bridge, on agaves, north of the railroad near the eastern end of the bridge, we found Bidimulus abundant and finely developed (pi. VII, figs. 1 to 7). The shape varies, but is always more conic than in the eastern localities. It is either nearly uniform white, or coffee-tinted varied with oblique brown or purplish streaks; these two color-forms in about equal numbers, found together on the same plants and connected by intermediate examples. Apex white. The interior is very dark purple-brown. The outer lip is usually but little thickened within, and the columella has no tooth, though often it is slightly salient in the middle. Whorls 7 to 7^. Alt. 36 37.5 33.3 34.5 36 33.3 37 mm. Diam. 19 19 17 17 17 16 15 Aperture 17.8 17.8 16 16 16 16 14.3 " The last two measurements show the extremes of shape in a series of 150 Hving specimens (No. 84,627 A. N. S.). Another set from east of the Pecos canyon about a mile from the Rio Grande is similar. The largest one measures, alt. 38, diam. 20.3, aperture 19 mm. (pi. VII, fig. 8). 1906.] NATURAL SCIENCES OF PHILADELPHIA, 143 OLEAOINID^. Euglandina singleyana (W. G. Binn.). Glandina singleyana W. G. B., Fourth Supplement Terr. Moll., V, p. 163 (1891). Singley, Contrib. Nat. Hist. Texas, p. 302. Sinking Spring, near San Marcos, Hays county, Texas, young speci- mens only; New Braunfels, Comal county, and its environs; San Antonio, Bexar county; Hondo river drift, near Hondo, Medina county, and in the canyon of the Pecos at High Bridge, Val Verde county. The last locahty is the extreme southwestern point known for the species. Mr. Singley reports it also from Guadalupe, Qohad, Gon- zales, Travis, CaldweU and Frio counties. PUPILLID^. Pupilla muscorum (L.). Benson, Arizona. Pupilla blandi (Morse). Texas: Flood debris of Guadalupe river, about four miles above New Braunfels, Comal county. New Mexico: Pecos river drift at Pecos (CkU.). The occurrence of this species near New Braunfels is anomalous; that place lies far below its normal zone, and the shortness of the Guadalupe river precludes the idea that it could have drifted any great distance. Mr. Singley also found one specimen at or near the same place. Pupilla hebes (Anc). Proc. A. N. S. Phila., 1900, p. 589, pi. 22, figs. 9, 10. Huachuca mountains, Cochise county, Arizona. Two out of about a dozen specimens are sinistral. They are readily separable from the sinistral P. syngenes by the absence of a crest behind the outer lip. Pupoides marginatus (Say). In Texas we took this species at San Marcos, Hays county; New Braunfels, Comal county; Hondo river, Medina county, and near the Rio Grande at Del Rio, Devil's river four miles from mouth, and drift of Pecos river near High Bridge, Val Verde county. Arizona : Drift of San Pedro river at Benson. Bifidaria pellucida hordeacella (Pils.). Texas: Drift of Sinking Spring, San Marcos, Hays county; Guada- lupe river about four miles above New Braunfels, Comal county; Hondo river, two miles north of Hondo, Medina county; Rio San Filipe near Del Rio, Devil's river near its mouth, and Pecos river near 144 PROCEEDINGS OF THE ACADEMY OF [March, the High Bridge, all in Val Verde county. Arizona: Benson. New Mexico: Drift of Pecos river at Pecos (Ckll). It varies somewhat widely in size, and is found in great quantity in drift debris along all the Texan rivers explored. Bifidaria pellucida parvidens (Sterki). Drift debris of Pecos river at Pecos, New Mexico (Ckll.). This sub- species has not before been reported from so far east. Bifidaria procera (Gld.). Texas : Drift of Sinking Spring, San Marcos ; Hondo river drift, two miles north of Hondo, Medina county. The var. cristata Pils. and Van. was found copiously in the drift of Guadalupe river, Comal county; near Del Rio; Devil's river; and Pecos river near the High Bridge, Val Verde county. Also Benson, Arizona. Bifidaria contracta (Say). Texas: San Marcos; Guadalupe river drift, near New Braunfels; Hondo river, near Hondo; Rio San Filipe near Del Rio; Devil's river; Pecos river near High Bridge. Rather abundant in the river debris except on the Pecos, where but three shells were found. In 1885 Pilsbry took specimens at Galveston. Bifidaria armifera (Say). Texas: Drift of Guadalupe river four miles above New Braunfels, Comal county; Pecos river near High Bridge. New Mexico: Pecos, on the Pecos river, collected by T. D. A. Cockerell. Bifidaria pentodon (Say). Texas: Sinking Spring, San Marcos, Hays county; Hondo river, Medina county. Vanatta has figured one of the specimens from San Marcos. Nau- tilus, XIX, pi. 6, fig. 11. Bifidaria tappaniana (C. B. Ad.) Pupa pentodon of authors, not of Sav. Texas: Guadalupe river about four miles above New Braunfels, Comal county (figured by Vanatta, Nautilus, XIX, pi. 6, fig. 12) ; Hondo river, Medina county; Devil's river, Val Verde county. Bifidaria perversa Sterki. Drift debris of the San Pedro river at Benson, Arizona (Ferriss, 1904). Hitherto known from Nogales, Arizona. (E. H. Ashmun.) Bifidaria ashmuni Sterki. Huachuca mountains, Cochise county, Arizona (J. H. Ferriss.) Also from the following places, collected by E. H. Ashmun. Arizona: 1906,1/ NATURAL SCIENCES OF PHILADELPHIA. 145 Near Jerome at Mescal Gulch, Walnut Gulch, Page's ranch, Kirwagen's ranch, and drift of Verde river; Navajo Springs ; Santa Rita mountains; Oak Creek at Owensbys; drift of Little Colorado river at Holbrook; Nogales, both north and south of the international boundary. New Mexico: Mountain station, Oscura mountains, Socorro county; San Rafael, Valencia county; White Oaks, Lincoln county. Subgenus CH^NAXIS nov. The shell has a large hollow axis, open below, and about one- third the total diameter of the shell, the structure otherwise being like Bifidaria s. str. ; peristome continuous and free. Type, B. tuba. The Bolivian Infundibular ia infundibuliformis (Orb.)^ resembles B. iuha in having a very large umbilicus, and we at first thought to asso- ciate the Arizonian species with it; but Infundibularia differs by the strictly conic shell, and in the aperture, which shows but one lamella, a very large angulo-parietal. Other lamellae or plicae, if they exist, must be very deeply immersed, as in the group Immersidens. The type of d'Orbigny's species has been lost, and its internal structure is unknown. It was found with Bifidaria nodosaria (Orb.), :a minute species, imperfectly described and figured, but probably not distinct from B. -pelludda (Pfr.). Bifidaria tuba n. sp. Fig. 6. Shell cylindric with a short apical cone, openly umbilicate, the um- bilicus nearly one-third the diameter of the shell, penetrating well- like to the apex. Pale brown, smooth, with light growth-lines only. Whorls 5^^, convex, the apex obtuse; last three whorls forming the cylindrical part of the shell. The last whorl is compressed around the umbilicus, and scarcely straightened in front. The aperture is short- oval, the peristome continuous, thin and well expanded. The angular and parietal lamellae are combined into one long lamella, notched on the summit; where the two join, the inner end of the angular projects a little on the right side. The columellar lamella is massive, slightly bifid, deeply placed, and enters about as deeply as the parietal. There are small, short, upper and lower palatal and basal folds, in the typical positions, a short distance within the lip, usually with a minute denticle between them, and another at the base. Length 3, diam. L5 mm. Drift debris of the San Pedro river, Benson, Cochise county, Arizona. 'Types No. 87,062 A. N. S. Phila.; cot5^es in Ferriss collection. A '^Voyage dans V Amerique Miridionale, Mollusques, p. 323, pi. 41 bis, fig. 7-10. 10 146 PROCEEDINGS OF THE ACADEMY OF [Marchj Fig. 6. single immature specimen was taken by Rev. E. H. Ashmun in drift of the Salt river, at Tempe, Arizona. Remarkable for its ample umbilicus. The freshest specimens are about the color of Pupoides marginatus. None were taken alive. In the adult shell opened (fig. 6) the parietal and columellar lamellae do not enter deeply, though they go so far that the inner end cannot be seen in the mouth. One l^roken shell from Tempe, taken by Mr. Ashmun, and which has been in coll. A. N. S. P. for some years, has a strong lamella on the columella, ap- parently at least a half-whorl long, but not pene- trating much farther than fig. 7 shows. This lamella may perhaps be developed during the latter part of the neanic stage, to be resor]:)ed in the fully adult condition. Further material is needed to demonstrate this. It is possible that the Tempe shell represents a distinct species. It tapers slightly more than the types from Benson. Fig. 7. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 147 Vertigo ovata Say. Benson, Arizona: Drift of Devil's river, Val Verde county, Texas. Vertigo binneyana Sterki. Benson, Arizona (Ferriss). Vertigo oscariana Sterki. Drift debris of the Guadalupe river, about four miles above New Braunfels, Texas, a single specimen, a little larger and stronger than Eastern (Austroriparian) examples. Vertigo milium Old. San Marcos, Hays county; Guadalupe river above New Braunfels, Comal county, and on the Hondo river, Medina county, Texas, in flood debris. Only one specimen from each place, among thousands of other Pupillidce, etc. Strobilops labyrinthica texasiana n. subsp. Shell moderately elevated with dome-shaped spire, brown, whorls 5^, the first 1^ smooth, pale-corneous, the rest regularly ribbed obliquel^^ the last whorl rounded periplierally or a trifle and obtusely subangular in front, the riblets passing over undiminished upon the base, which is as strongly sculptured as the upper surface (or sometimes smoothish just in front of the aperture). Aperture with expanded, thickened peristome and strong parietal callus, a single strong parietal lamella emerging to the edge of the callus, a second weak one visible within. About half a whorl inw^ard there is a series of about six laminae, the inner one upon the columella, the next short, strong and tongue- shaped, bending outward; the third nearly twice as long, high and sinuous; the fourth very minute and low, often wanting, leaving a space; the fifth and sixth long and low; and just above the periphery on the outer wall a very weak, low, long seventh plica may usually be traced. Umbilicus rather large. Alt. 1.5, diam. 2.2 mm. Types No. 91,330 A. N. S. Phila., from drift of the Guadalupe river about four miles above New Braunfels, collected by Pilsbry and Ferriss, 1903. Other localities in Texas are Austin (Pilsbry), San Marcos (Pilsbry and Ferriss), New Braunfels (Ferriss, Pilsbry and Sing- ley), Guadalupe river bottom, Victoria county, and Lavaca river, Jackson county (J. D. Mitchell), Lee county (Singley), Calhoun county (E. W. Hubbard), Gainesville (J. B. Quintard). A smaller form, diam. 2 mm., was taken in drift debris of the Hondo river about two miles north of Hondo, Medina county (Ferriss and Pilsbry). It also ranges northward into Indian Territory and to Kansas. 148 PROCEEDINGS OF THE ACADEMY OF [March, This form, which for the present we subordinate to the old S. laby- rinthica, is the only Strohilops except S. hubbardi which I have seen from Texas. It may be distinguished from *S. virgo by the costulate base, wider umbilicus and far weaker inner parietal lamella. S. strebeli is a much more depressed cone. The statement by Woodward,^" copied into American works, that H. labyrinthica occurs in pretertiary Eiuropean strata, is misleading, since the American species is really not identical with any from Euro- pean strata, though there are numerous allied forms in the European tertiaries. The genus became extinct in Europe, but survives in China, Helix diodontina Heude being a Strobilops. In America it has not been found in the Northwest or Pacific States, but extends south to Mexico, the West Indies and Venezuela, and a species apparently belonging to the genus has been described from the Galapagos. VALLONIID^. Vallonia excentrioa Sterki. Galveston, under boards in a vacant lot. (Pilsbry, December, 1885.) Vallonia perspeotiva Sterki. Texas : Drift debris of Devil's river, and of Pecos river near the High Bridge, Val Verde county. Arizona: Benson, in drift of San Pedro river. Vallonia gracilicosta Reinh. New Mexico: Drift of Pecos river, at Pecos (Cockerell!). Vallonia cyclophorella Anc. Arizona : Drift of San Pedro river, Benson, Cochise county, a single specimen. COOHLIOOPIDiE. Shell oblong, cylindric-oblong or narrowly tapering, smooth and glossy, with imperforate axis ; aperture ovate or acuminate, the columella notched below or continuous with the basal lip. Foot without pedal grooves. Kidney with direct ureter, of the Basommatophorous type. Genitalia with a long appendix on the penis, as in Achatinella and the Pupillidoe. Jaw and radula about as in Pupillidoe. This group has usually been included in the Achatinidoe or Stenogy- ridoe, but the direct ureter removes it to a group of primitive snails represented only by minute species in America, but by the beautiful Achatinellidce and Partulidoe in Polynesia. Coecilianella (Cecilioides) I" Manual of the Mollusca, p. 286, edit. 2, 1868. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 149 belongs to the same family, and probably Glessvla also, but the pallial organs of that Indian genus are unknown. The group is not related to the Achatinidce. Cochlicopa lubrica (Mull.). Bear Park and Cave creek canyon, Chiricahua mountains, and Fort Bowie (Ferriss) ; Carr canyon, Huachuca mountains (Dr. H. Skin- ner); all in Cochise county, Arizona. Drift debris of Pecos river^ Pecos, New Mexico (Ckll.). ZONITIDiE. Omphalina extends to the western border of the Austroriparian area in Texas. The form taken by us will be discussed in the conclud- ing paper on Southwestern mollusks. Zonitoides minuscula (Binn.i. Texas: San Marcos, New Braunfels, near Hondo, Del Rio, Devil's river and Pecos river. Everywhere common in drift debris. The specimens all show a tendency to be more widely umbilicate than typical Northern minuscula, a large part of them being typical Z. m. alachuana (Dall). Those from San Marcos and Comal county are of the size of Northern minuscula, but westward the shells reach a decid- edly larger size, with the exception of the lot taken at Devil's river, which show but little tendency towards a wide umbilicus. In Arizona, Ferriss took specimens of var. alachuana at Bear Park and Cave creek canyon, Chiricahua mountains, in the drift of San Pedro river at Benson, and in the Huachucas. Like other minutiae, these shells are very rare in the Chiricahua and Huachuca mountains. Helix mauriniana Orb., from Cuba, which has been put in the synonymy of minuscula, seems to be a Thysanophora close to or identical with T. saxicola (Pfr.), as Arango has already stated. Z. minuscula occurs also in Japan. Zonitoides minuscula neomexioana Pils and Ckll. This form is distinguished by the possession of minute and shallow spiral striation. It seems to be of somewhat common occurrence in New Mexico, and upon examining a set of seven specimens taken by me in Galveston in 1885 I find that they are similarly sculptured. They came from under boards in a lumber yard, and it may be that they were brought from New Mexico with lumber, though I do not know that any lumber was shipped from New Mexico twenty years ago. 150 PROCEEDINGS OF THE ACADEMY OF [March. Zonitoides singleyana (Pils.)- Zonites singleyanus Pils., Proc. A. N. S. Phila., 1SS9, p. 84; 1888, pi. 17, fig. M. (New Braunfels). Hyalinia Iceviuscula Sterki, Nautilus, VI, p. 53, Sept., 1892 (New Braunfels). Texas: San Marcos, New Braunfels, Del Rio. Devil's river and Pecos river above the High Bridge; ever}n\'here in river debris. Arizona: Drift of San Pedro river at Benson. Zonitoides nummus Vanatta. Proc. A. N. S. Phila., 1899, p. 524, figs. (New Braunfels). This species seems to be confined to the Texan Lower Sonoran. We took it at San Marcos, Hays county; Guadalupe river above New Braunfels; Hondo river, Medina county; and in Val Verde county near Del Rio; along the Devil's river, and in the Pecos canyon above the High Bridge; everywhere in drift debris. Zonitoides arborea (Say). Texas: Galveston; Smith ville, Bastrop county; Sinking Spring, near San Marcos, Hays county; near New Braunfels, Comal county. New Mexico: Drift of Pecos river at Pecos (Cockerell). Arizona: Cave creek canyon and Bear Park, Chiricahua mountains, Cochise county. Vitrea indentata (Say). Drift of Pecos river, Pecos, New Mexico (Ckll.). As usual, it is the Canadian and Carolinian form of the species which extends down the Rocky mountains into New Mexico, and not the Sonoran subspecies. Vitrea indentata umbilicata ('Singl.,' Ckll). CkU., Nautilus, XII, p. 120, Feb., 1899. Texas: San Marcos, Hays county; around New Braunfels, Comal county; Hondo river two miles north of Hondo, Medina county; Del Rio, Devil's river and Pecos river at the High Bridge, Val Verde county; Alpine, Brewster county, Arizona: Cave creek canyon and Bear Park, Chiricahua mountains; Fort Bowie. Also Florida mountains, Grant county, New Mexico. Large specimens of this race are probably what has been reported from Texas as sculptilis Bid., — a species which does not, we believe, occm- in that State. This .Sonoran race differs from indentata by its distinctly perforate axis and larger average size, yet the perforation varies so much in size in specimens from the Carolinian zone that I would not myself have named the Southwestern form. The name is ill-chosen, since the shells are not ' ' umbilicate, ' ' as that term is technically used, but ' ' per- forate. ' ' 1906.1 NATURAL SCIENCES OF PHILADELPHIA. 151 Vitrea hammouis (Strom). Drift of the Pecos river at Pecos, New Mexico (Cockerell). The specimens scarely show a trace of spiral hues. The species is unknown in the Austroriparian and Lower Sonoran zones. Vitrea dalliana roemori n. subsp. Fig. 8. Shell openly umbilicate, the width of umbilicus contained about 4f times in the diameter of the shell, pale whitish-corneous, in general shape resembling V. dalliana, V. wheatleyi and V. petrophila. Sculp- ture of very close and regular radial grooves, on the last whorl of large specimens becoming crowded and less regular, giving a striate appear- ance. The flat intervals between the grooves show no spiral striae. The base is nearly smooth. Whorls 4^, but slightly convex, slowly widening, the last about double the width of the preceding. Suture scarcely impressed, translucent-margined below. Base convex. Aper- ture lunate, slightly oblique ; the insertions of the peristome are distant. Alt. 2, diam. 4 mm. ; umbilicus .85 mm.; aperture 1.7 x 1.65 mm. Fig. 8. — \'iirea tkdliana roemeri. Sinking Spring creek, near San Marcos, Hays county, Texas. Types No. 91,318, A. N. S. Phila., collected by Pilsbry and Ferriss, 1903. Also taken in several places around New Braunfels, Comal county; in the drift debris of the Hondo river, two miles north of Hondo, Medina county; in drift of the Rio San Filipe near Del Rio, and of the Devil's river, Val Verde county. It has about the distribution of Holospira goldfussi and Helicodiscus eigenmanni. This very pretty little species has a slightly more ample umbilicus than V. wheatleyi or petrophila, and the sculpture is closer and more regular than in either. The last whorl, in dorsal view, is wider than in V . wheatleyi. It is much smaller than V. hammonis. V. dalliana roemeri attains a larger size than the Floridian V. dalliana and the shells have somewhat more regular and crowded grooves on 152 PROCEEDINGS OF THE ACADEMY OF [March, the last whorl; it is less depressed, and the aperture is perceptibly less broad, more roundly lunate. The differences seem sufficient to call for subspecific separation. The Texan and Floridian areas of distribution seem to be separated, so far as our present data indicates. V. daUiana should be looked for along the northern border of the Gulf. It is known now from peninsular Florida only. Vitrea dalliana " and roemeri are much smaller than V. hammonis, and seem to replace that in the Austroriparian and Lower Sonoran zones. Wlien originally described it was compared with Zonitoides arhorea (Say), but it is not really related to that but to the hammonis group. In fact V. hammonis, hinneyana and dalliana form a group of very closely related species. In a large number of V. dalliana examined from several localities, the largest shell measures, alt. 1.6, diam. 3.2, width of umbihcus .75, aperture 1.4 mm. wide, 1.2 high. This shell, from Osprey, Manatee county, Fla., is here figured. The figures do not represent the fine and beautiful sculpture of the surface. Fig. 9.— F. dalliana. Vitrea milium meridionalis n. subsp. Similar to V. milium but larger, diam. about 1.75 mm., with nearly 3^ whorls, the first one finely, distinctly Urate spirally, the last whorl with obhque wrinkles much coarser than in milium, more or less an- astomosing, and fine spiral striae, the latter distinct on the base. V. milium with the same number of whorls is smaller and more finely wrinkled, and in Maine and Ohio shells spirals on the first whorl are excessively weak or wanting, not deeply engraved to the tip, as in Texas shells. Texas: San Marcos, in drift of Sinking creek, in the limestone hills; along the Guadalupe river above New Braunfels (t^^^e locahty); Hondo river, Medina county ; drift of Pecos river. (Pilsbry and Ferriss.) New Mexico: Cloud croft, Sacramento mountains (Viereck); Santa Fe (Ashmun). Arizona: Huachuca mountains (Ferriss); Walnut Gulch near Jer- ome (Ashmun), Specimens from Baldwin and Clarke counties, Ala. (C. B. Moore), " Zonites dallianus Simpson, Pilsbry, Proc. A. N. S. Phila., 1889, p. 83, pi. 3, figs. 9-11. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 153 seem to be referable to this race. In the Northwest another form of the species, Z. milium pugetensis Dall, replaces the typical milium. Neither of the subspecies differs much from milium, but what differ- entiation there is seems to be correlated with geographic range. This species was erroneously placed in Zonitoides in the Classified Catalogue of 1898. We are now convinced that it belongs, as Morse demonstrated, to the subgenus Striatura of Vitrea. Vitrina alaskana l>all. V. pfeifferi Newc, Proc. Cal. Acad., II, p. 92, 1861 ; not of Deshayes, 1852. Vitrina alaskana Dall, Land and Fresh-Water Mollusca of Alaska and Ad- joining Regions, Harriman Alaska Exped., XIII, p. 37. Arizona: Huachuca mountains (Ferriss), numerous rather small specimens, the only ones we have seen from Arizona. It seems to be a common species of the Canadian and Transition zones eastward, speci- mens being before us from the following places in New Mexico : Chi- corico canyon near Raton (Cockerell); Las Huastus canyon, Sandia mountains, near Albuquerque (Miss Maud Ellis) ; near Las Vegas (Miss Mary Cooper) ; White Oaks and Gilmore's Ranch, Sierra Blanca (Ash- mun); Fort Wingate (Dr. E. Palmer); James canyon, Cloudcroft, Sacramento mountains (H. L. Viereck). The tyipe locality of V. alaskana is Carson valley, Nevada; but it has a wide range, from Alaska to the Mexican boundary and probably beyond, southward occurring only at high elevations. Euconulus fulvus (Miill.). Drift of Pecos river, Pecos, New Mexico (Cockerell). Cave creek canyon, Cochise county, Arizona (Ferriss). Euoonulus chersinus trochulus (Reinh.). Nautilus, XII, p. 116. Texas: Sinking Spring, San Marcos, Hays county; New Braunfels and vicinity, Comal county; Hondo river, north of Hondo, Medina county; Rio San Filipe near Del Rio, and Devil's river, Val Verde county; ever3rwhere in drift debris. BNDODONTID^. Pyramidula cronkhitei anthonyi Pilsbry. n. n. Helix striatella Anthony, Boston Journ. of Nat. Hist., Ill, p. 278, pi. 3, fig. 2' 1840. Not Helix striatella Rang, 1831. This shell, well known under the preoccupied name H. striatella Anth., has typically a rounded periphery and moderately developed oblique and sigmoid rib-strise, 4 or 5 in the space of a mm. on the front of the last whorl at the periphery. There are 3^ to 3f whorls. 154 PROCEEDINGS OF THE ACADEMY OF [March, Alt. 2.7, diam. 5.25 mm. Type locality, Fairmount Park, Philadelphia, near "Strawberry Mansion," No. 68,899, A. N. S. P., collected by E. G. Vanatta. In my opinion the widespread Eastern race is not specifically dis- tinct from Helix cronkhitei Newc, of northern California, etc., of which specimens from Dr. Newcomb are before me, but it evidently requires separation as a subspecies. Arizona: Chiricahua mountains, Cochise county, at Fort Bowie, Bear Park and Cave creek canyon (Ferriss); Carr canyon, Huachuca mountains (H. Skinner). New Mexico: Drift of Pecos river at Pecos (Ckll.). Texas : In 1885 I dug a single specimen of this species from the bank of Comal creek, New Braunfels, where it occurred with several other land and fresh-water shells. Singley reports it from the Pleistocene of Swisher county. It is not known to occur living in Texas, or any- where in the Austroriparian zone. Its occurrence in the Texan Pleisto- cene is anomalous, like the presence of Pupilla blandi at New Braunfels. RADIODISCUS Pilsbry, n. gen. Minute, discoidal, openly umbilicate Patuloid snails with the first 1^ whorls minutely engraved spirally, the rest of the shell densely radi- ally costulate; aperture lunate, but slightly oblique and as high as wide. Type, R. millecostatus. This genus is proposed for a group of tiny Pyramidula-like snails, various members of which have been found in Arizona, Mexico and South America as far south as Patagonia. In the spiral sculpture of the embryonic whorls these shells are like Helicodiscus ; in shape and size they resemble Planogyra, but in that the embryonic shell is smooth. There are also some similar Polynesian forms, referred to Charopa, etc. In the Endodontidce, where small differences in the shell characterize extensive series of species, it seems desirable to recognize as of generic value such readily recognizaljle groups as Radiodiscus. The species are chiefly distinguished by their dimensions and the degree of fineness of the sculpture. There are several undescribed forms in the collection of the Academy. Eadiodiscus millecostatus n. sp. Fig. 10. Shell very small, disk-shaped, chestnut-brown, the first whorl bluish- white. Whorls 3f , slowly widening, separated by a very deep, chan- nelled suture. The first whorl projects a little, and is microscopically striate spirally, the rest of the whorls are radially very densely costu- late, the riblets nearly straight, delicate and much narrower than 1906.] NATURAL SCIENCES OF PHILADELPHIA. 155 their intervals, and about 21 in number in a millimeter measm-ed at the periphery, in the front of the last whorl. The last whorl is rounded throughout. The umbilicus is widely open, its width contained about three times in the diameter of the shell. The aperture is slightly oblique and deeply lunate. Alt. 1.1, diam. 2 mm.; diam. of umbilicus .6 to .7 nam. Huachuca mountains, Cochise county, Arizona, (J. H. Ferriss) ; Carr canyon (Dr. Henry Skinner). Also in the State of Michoacan, Mexico, at Patzcuaro and Morelia (S. N. Rhoads, 1899). Fig. 10. This tiny Patuloid, of a group new to our fauna, is one of the most interesting recent finds. It is related to R. hermanni (Helix hermanni Pfr.), of central and eastern Mexico, but that species is larger and more coarsely sculptured. Thfe two species occur together at the Mexican localities mentioned above, but are readily separable. Dr. Henry Skinner found one broken specimen of R. millecostatus in Carr canyon, Huachucas. Genus HELICODISCUS Morse. This genus has hitherto been consitlered to include two species, lineatus (Say) and fimbriatus Wetherby. The accessions of Southwest- ern material show that several other forms must be recognized. The species are not very conspicuously differentiated, and young shells by themselves are not always readily placed ; yet with adults we find no difficulty. The salient characters of the forms follow: a. — Diam. of shell 3 to 3.5 mm., whorls 4 to 4^; teeth almost always present. Eastern United States, . . . H. parallelus (Say). a^. — Diam. of shell 4.5 to 5 mm., whorls 4^ to 5+; often deficient in teeth. b. — Spiral striae obsolete. Idaho. . . H. salmonensis (Hemph.). 156 PROCEEDINGS OF THE ACADEMY OF [March, b^. — Spiral strise well developed, subequal; last whorl not deflexed at aperture. c. — Umbilicus very wide; last whorl narrow, aperture very- small. Arizona and New Mexico, H. e. arizonensis n. subsp. c^^Umbilicus smaller, more cup-shaped, last whorl wider. Texas, H. eigenmanni Pils. 6^. — Spi'-als coarse, some of them more conspicuous, with a cutic- ular fringe ; last whorl in full}^ adult shells abruptly deflexed in front. East Tennessee, North Georgia, //. fimhriatus Weth. Helioodisous parallelus (Say). PI. VIII, figs. 7. 8, 9, 10. Helix lineata Say, Journ. Acad. Nat. Sci. Phila., I, p. 18 (1817) ; II, p. 373. Not Helix lineata Olivi, Zool. Adriatico, p. 177 (1792). Planorbis araUeUuf< Say, Journ. A. N. S. Phila., II, p. 164, (1821), corrected to paraUellus in the Index, p. 407. (Upper Missouri.) Helicodisciis lineatus Morse, Journ. Portland Soc, I, p. 25, figs. 61, 62, pi. 2, fig. 3; pi. 7, fig. 63 (1864). Binney, Man. Amer. Land Shells, p. 75. This common species has been well described by Binney and others. It has ordinarily four whorls, but there may be as many as 4J in excep- tionally large shells. At least one pair of tubercular teeth may be seen in most specimens. Shells of maximum size measure: Alt. 1.3, diam. 3.5 mm. (Grand Rapids, Mich.) Alt. 1.25, diam. 8 mm. (Philadelphia.) The typical form of H. 'parallelus is before me from localities in Ontario, Maine, Massachusetts, Connecticut, New York, New Jersey, Pennsylvania, Maryland, District of Columbia, Virginia, North Carolina, Ohio, Michigan, Minnesota, Indiana, Iowa, Kentucky, Tennessee, Alabama, Florida, Missouri, Kansas, Arkansas, Indian Territory. I have not seen Texan specimens of typical lineatus, but it doubtless occurs in northern and eastern parts of the State. In New Mexico, at Pecos, Las Vegas, Mesilla, Sandia mountains and other places, and in the Pecos river canyon near its mouth, Val Verde county, Texas, there is a form of H. parallelus with rather weak, sparse spirals, the intervals more distinctly striate radially than in typical parallelus. This seems to be a form of the southeastern Rocky moun- tains and southward to the Rio Grande, occupying territory between the ranges of H. eigenmanni and H. arizonensis, with some overlapping on the territory of the latter. Say's first name. Helix lineata, was preoccupied, but his Planorbis parallelus applies to the same species. In the text of the Journal the first letter of the name did not print up, but the space in place of it shows that it had been there, and the p is correctly supplied in the index. The type used in that volume of the Journal was old and full 1906.] NATURAL SCIENCES OF PHILADELPHIA. 157 of ''bad letters." All authorities agree that such obvious typograph- ical errors should be corrected. Helicodiscus eigenmanni Pils. PI. VIII, figs 1, 2, 3. Nautilus, XIV, p. 41. August, 1900 (Beaver cave, near San Marcos, Texas). This fine species is easily recognized by its large size, the shell in all stages of growth being very much more robust than H. parallelus. It is sculptured with numerous spiral threads as in H. parallelus, and when fresh is pale yellow. Alt. 2, diam. 5 mm., with 5^ whorls. We took specimens in the drift debris of Sinking creek at San Mar- cos, Hays county; on the Guadalupe river above New Braunfels, Comal county; on the Hondo river, Medina county, and in the drift of Devil's river, Val Verde county; Alpine, Brewster county. I have seen specimens also from Calhoun county (Hubbard), Lee county (Singley), and Navidad river bottom, Jackson county (J. D. Mitchell). It has not been reported from any place out of Texas, and seems to have about the same distribution as Holospira goldfussi. Helicodiscus eigenmanni arizonensis n. subsp. PI. VIII, figs. 4, 5, 6. Shell larger than H. parallelus with 4^ to nearly 5^ whorls, the spire flat or convex, the umbilicus wider than in parallelus or eigenmanni. Surface closely and strongly lirulate spirally. Aperture oblique, lunate, very small. Alt. 2, diam. 5 nmi., whorls 5^. (Fort Bowie.) Alt. 1.5, diam. 4.6 mm., whorls 4^. (Cave creek canyon.) Arizona: Fort Bowie (type locality); Bear Park and Cave creek canyon, Chiricahua mountains; Huachuca mountains (Ferriss); Santa Rita mountains (Ashmun). New Mexico: Florida mountains, Grant county (Ferriss); Bland, Sandoval county (Ashmun) ; debris of Arroyo Pecos at Las Vegas (Ckll.). Readily distinguished from H. parallelus when mature by the greater size, with wider umbilicus, more whorls and proportionately smaller mouth. Some specimens are two-toothed. H. eigenmanni has a smaller umbilicus, whorls of markedly greater calibre and a larger aperture. Helicodiscus salmonensis (Hemphill). HeUcodiscu^ fimbriatus Wetherbv (salmonacca Hemphill), W. G. Binney, Tliird Supplement to Terr. MoU., V, BuU. Mus. Comp. Zool., XIX, No. 4, p. 189 (May, 1890). Helicodiscus fimbriatus var. salmonensis Hempliill, in Binney, t. c, p. 220 (May, 1890). Helicodiscus fimbriatus Wetherby, var. salmonaceus Hemphill, Binney, Fourth Supplement, Bull. Mus. Comp. Zool., XXII, No. 4, p. 177, pi. 3, fig. 8. Distinguished by the absence of spiral striae, according to Hemphill. 158 PROCEEDINGS OP THE ACADEMY OF [March, I have not seen specimens. Mr. Binney in his first note considers the Salmon river form identical with what Mr. Simpson reported as H. fimbriatus from Indian Territory, and gives no separate or definite description of it, though he mentions that Hemphill had given the (MSS.) name "salmonacea." The figure of one of the original speci- mens, given in Binney's Fourth Supplement, represents a shell with wide umbilicus and small aperture, like H. arizonensis, from which it differs, according to published information, by the smoother surface, arizonensis being constantly very well sculptured. Punctum pygmaeum (Drap.). San Marcos, Hays county; Comal county; Hondo river, Medina county; Devil's river, Val Verde county. The form in this region is slightly larger than northeastern speci- mens, and is more strongly sculptured. There are barely four whorls, the first H smooth, the next striate; the last two whorls have striae at regular intervals much larger, with about six fine striae in each space, and the basal spirals are very distinct. This sculpture reminds one of the west coast forms, conspectum, pasadenoe and calif ornicum, which however are decidedly larger and coarser shells of a dark brown color. SUCOINEID^. Succinea luteola Gld. Gould, Proc. Bost. Soc. N. H., Ill, p. 37, June, 1848 (Texas); Terr. Moll., II, p. 75, pi. 67c, fig. 1. (Florida; Texas, especially Galveston.) Bin- ney, Terr. Moll., V, p. 419; Man. Amer. Land Shells, p. 441. Succinea texasiana Pfr., Monogr., II, 526; Roemer's Texas, p. 456, 1849 (Galveston). Succinea lutescens Sowerbv, Conchologia Iconica, XVIII, pi. 10, fig. 67 a, b, 1872 (Texas). We took specimens in Texas along the Guadalupe river above New Braunfels, Comal county; San Antonio, Bexar county; near Hondo river about two miles north of Hondo, Medina county, and in Val Verde county at Del Rio, high land west of Devil's river, and in the canyon of the Pecos near the High Bridge. In Gould's original description the only locality given was Texas. In the Terrestrial Mollusks he states "found in Florida, and more abundantly in Texas, especially in the region of Galveston." Speci- mens collected at Galveston by the author in 1885 agree perfectly with Gould's figures, and that place may be taken to be the type locality. I have seen no Florida shells which I would refer with cer- tainty to luteola, though *S. floridana is closely related. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 159 Succinea concordialis Gld- Figs. 11, 12. Gould, in Terr. Moll. U. S., II, p. 82 (Lake Concordia). Binney, Terr. MoU. U. S., V, p. 419; Man. Amer. Land Shells, p. 44L Succinea forsheyi Lea, Proc. A. N. S. Phila., 1864, p. 109; Obs. Gen. Unio XI, 134 (Rutersville, Texas). Succinea Imleana Lea, Proc. A. N. S. Phila., 1864, p. 109 (Alexandria, La.). Succinea halei Lea, Obs., XI, 136 (n. n. for S. ' ~ Distribution, Gulf States from Florida to the Rio Grande, on mud or herbage near the water's edge. Common from Louisiana west- ward, probably rare and local eastward. The type locality, Lake Concordia, is not in Texas, as Gould and Binney supposed, but in Louisiana. The lake is an abandoned ox- bow of the Mississippi river, opposite Naches, Mississippi. Some of Lea's original lot of S. halei (haleana) before me show that to be merely the young of concordialis. S. forsheyi Lea, of which two cotypes are in the Philadelphia collection, is surely identical with concordialis. An adequate knowledge of the distribution of S. concordialis east- ward awaits further exploration of the Gulf coastal peneplain, which in Mississippi, Alabama, Georgia and Florida has been very imper- fectly examined for land molliLsks. I have not seen S. wilsoni Lea, described from Darien, Ga., the figure of which looks a good deal like concordialis, though it seems to be less swollen basally. Specimens col- lected by Mr. A. A. Hinkley at Cypress creek, Ala., in 1895 are evi- dently concordialis ; and a set of very pale shells, corneous instead of amber-colored, before me from Mayport, Florida, collected by i\L A. Mitchell about twenty years ago, seems to agree with concordialis in everything but color. Northward it extends to Frierson (L. S. Frier- son) and Bayou Pierre (George Williamson), in northwestern Louisiana, the specimens from these places being rather small. .The species must also extend in a long lobe up the Mississippi and its tributaries, for typical specimens have lately been sent by Mr. A. A. Hinkley from Dubois, Illinois, and by Mr. T. Van Hyning from Des Moines, Iowa. Mr, Van Hyning notes that ' ' the animal is black with small yellow dots. ' ' These Northern shells may be distinguished from »S. retusa by their pot-bellied figure and reddish apex. In Texas, specimens were taken by us in April at San Marcos, Hays county; New Braunfels, Comal county; San Antonio, Bexar county; along the Rio Grande near and San Filipe river, at Del Rio. and along the Devil's river, Val Verde county. We have seen it also from Lee county (Singley) and Spring creek, Victoria county (J. D. Mitchell). It lives on the moist earth immediately adjacent to the water's edge, and where found is usually abundant. It is a thin shell, rather 160 PROCEEDINGS OF THE ACADEMY OF [March, deeply amber-colored, with the apical whorls darker, reddish-orange. Whorls 3^, the last deeply descending, somewhat flattened above, very convex basally, having thus a more sack-like contour than S. retusa; this being its chief peculiarity. The sculpture consists of rather coarse wrinkles and often some indistinct spiral impressions on the last whorl. The aperture is symmetrically ovate, the columella con- cave throughout, with a delicate fold above. Large specimens from San Marcos measure: Length 16.8, diam. 9, length of aperture 11.5, width 6.7 mm. Length 16, diam. 8, length of aperture 10.5, width 6 mm. The mantle is intensely black, dappled throughout in the last whorl with rounded yellowish spots. Those above the kidney are brighter, more conspicuous and usually larger, often more or less confluent. Fig. 11. Fig. 12. Towards the edge of the mantle the spots are large and lengthened. The foot including head and eye-stalks is pale grayish-white, speckled with irregular grayish-black spots. The sole is pale yellow. Figs. 11 and 12 were drawn from specimens taken at San Marcos, Texas, near the river. SpecimeiLS from the Rio Grande and Devil's river are of almost as large size; but in some situations it is much smaller, a set from along the San FiUpe river at Del Rio measuring, length 12, diam. 6.7, length of aperture 8 mm. Those taken at San Antonio and New Braunfels are also of small size. Such variation in size is apparently not racial, but dependent upon local conditions of the food supply, etc. The color and markings of the mantle are substantially the same in all colonies we found, though there is individual variation in the size of the light spots, figure 12 representing one of the darker individuals. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 161 Succinea grosvenori Lea. Proc. A. X. S. Phila., 1864, p. 109 (Santa Rita Valley, Kansas?, and Alex- andria, Louisiana) ; Obs. Gen. Unio, etc., XI, p. 13.5, pi. 24, fig. 108. S. mooresiana Lea, Proc. A. N. S. Phila., 1864, p. 109 (Court House Rock, Platte river, on the California route) ; Obs., XI, p. 136. pi. 24, fig. 109. S. Hneata W. G. Binnev, Proc. A. X. S. Phila., 1857, p. 19; Man. Amer. Land Shells, p. 174. This species is characterized by its very full, rounded whorls and irregular sculpture, often with traces of spiral lines, though as frequently without them. Having part of the original specimens of the above synonyms before us, we are convinced that the three supposed species cannot be separated, although one would hardly expect the same form to range from the lower Mississippi to and throughout the arid great plains and the mountain region of Colorado and New Mexico. Succinea greerii Tryon, described from Vicksburg, Mississippi, is a little less swollen, with the suture not quite so deep, but it is doubtful whether the slight differences shown by the lot of eight specimens will prove constant. The species belongs to the campestris group. We took a few specimens of S. grosvenori at San Antonio, Texas, with S. concordialis and a large form of *S. avara. Succinea avara Say. Texas: !>an Marcos, Hays county; Comal county; San Antonio, Bexar county; Devil's river, Val ^'erde county. Arizona: Benson, Cochise county. AURICULID^. Caryohium exile H. C. Lea. San Marcos, Hays county; New Braunfels, Comal county; Hondo river, Medina county, and Devil's river, Val Verde county, Texas. Common in drift debris. Carychiuin exiguum (Say). Drift of Hondo river, Medina county; of Devil's river, Val Verde county, and of Guadalupe river, Comal county, Texas. LYMN^IDJE. Lymnaea columella Say. New Braunfels, Comal county, Texas. A single slender specimen. Lymnaea desidiosa Say. Texas: San Marcos, Hays county; New Braunfels, Comal county; San Antonio, Bexar county ; Rio San Filipe, Val Verde county. Lymnaea humilis Say. Guadalupe river, Comal county. Texas. 11 i 162 PROCEEDINGS OF THE ACADEMY OF [March, Lymnaea bulimoides cockerelli n. subsp. Figs. 13-17. Shell subglobose, pale yellowish-corneous, composed of 4^ convex whorls which are finely striate but without spiral lines or malleation. Spire very short, last whorl and aperture very large. Aperture short- ovate, its length three-fifths to two-thirds that of the shell. Columella broadly expanded, not folded. Umbilicus large. Fig. 13. Fig. 14 Fig. 16. Fig. 17. Las Vegas, N M. Arroyo Pecos. Ogalalla, Neb. Alt. 10 8.3 8 8.7 8.5 12 8 9 mm Diam. 7.2 6 6.2 6 6 7.8 6 6.6 " Aperture 6.7 5.3 6 5.7 5 7.5 5.8 6.1 " New Mexico: Las Vegas (type locality. Miss Mary Cooper), and in the charcoal zone of the Pleistocene of Arroyo Pecos (T. D. A. Cockerell) ; near Farmington (George H. Pepper) ; Acama (Dr. E. Palmer). Colorado : Pool southeast of Denver (J. D. Putnam). Nebraska: Ogalalla (C. T. Simpson). Dakota: Lake Hermann (P. C. Truman). Fig. 18. Fig. 19. This form differs from L. bulimoides and L. techeUa by its more glo- hose shape and shorter spire, and so far as we have seen is readily separ- able from both. Figs. 13-15 represent the types from Las Vegas ; 16, 17 are larger shells from Farmington, sent by Mr. George H. Clapp. L. hiilimoides sonomaensis Hemphill , from Sonoma county, California , approaches cockerelli, but differs by the more rapidly expanding last 1906.] NATURAL SCIENCES OF PHILADELPHIA. 163 whorl; narrower flat columella and narrower umbilicus, which is like that of typical bulimoides. A specimen measures, alt. 10, diam. 7.7, aperture 7.2 mm. (figs. 18, 19). Lymnaea bulimoides techella (Hald.). Figs. 20-23. Limncea techella Hald., Amer. Journ. of Conch., Ill, p. 194, pi. 6, fig. 4. (Texas.) Shell obese, with acutely conic spire, of five or six convex whorls; pale yellowish or light brown, finely striate and usuxilly malleated, the flattened facets obliquely descending. Last whorl very ventricose, umbilicus large. Aperture short-ovate, about three-fifths the total length; basal lip expanded, colimiellar lip broadly dilated, without a fold. Umbilicus large. Fig. 23. San Marcos, Texas. Tempe, Ariz. 8.5 11.8 13 14 12.7 12.5 mm. 5.1 7.3 7.6 9 7.8 7 5.1 7.3 13.7 8.8 7 6.8 " Cotype. Length 8 Diam. 5.1 Aperture 4.9 In the area under consideration we have seen specimens from the following places: Texas: Fort Worth (Sampson); Royse, Rockwall county (Ragsdale); Dallas (E. Hall); Houston (Pilsbry); San Marcos, Hays county (Pilsbry and Ferriss); mouth of Nueces river (Singley). New Mexico: Albuquerque, and McCarty, Valencia county (Ash- mun) ; Rio Grande at Mesilla (Cockerell). Arizona: Salt river at Tempe (Ashmun). Other records could probably be supplied from the literature, but it seems usually to have been referred to as L. bulimoides. Figs. 20-22 represent specimens from San Marcos, Hays county, Texas; fig. 23 is ajnore malleate shell from Salt river, Tempe, Arizona. 164 PROCEEDINGS OF THE ACADEMY OF [Marcll, The young of one season have a comparatively shorter spire than old shells, the rate of descent of the suture progressively increasing somewhat. As usual, there are only traces of malleation at this stage. The erosion of the apices in all but one of the type lot (as mentioned by Haldeman) makes them shorter than they would otherwise be, giving much the appearance of L. h. cockerelli. Abundant series from Houston and San Marcos collected by Pilsbry, with others from various naturalists, demonstrate the identity of Haldeman's shells with the larger-spired adult form figured above. L. tecJiella was formerly con- sidered by one of us to be a synonym or race of L. cubensis Pfr., and L. hulimoides was treated as a variety of the same species. They are cer- tainly very similar, but cubensis has a more triangular and less broadly developed columellar expansion; and in view of the way Lymnoea is being split up into species and races, it may be best to retain the several forms as distinct species or races until the subject can be taken up with ample material and time for the study of Antillean and Mexican forms together with our own. Planorbis trivolvis Say. Specimens referable to typical P. trivolvis were taken in Comal county, and at Del Rio, Devil's river and the Pecos river, Val Verde county, Texas. Planorbis bicarinatus Say. Guadalupe river, Comal county, Texas; abundant and typical. Planorbis cams n. sp. PI. IX, figs. 4, 5. Shell discoidal, biconcave, the spiral on the left side slightly more sunken and narrower than on the right. Whorls 3^, convex, the last round peripherally and on both -sides, curving more abruptly into the concavity on the left side. Sculpture of close, very regular obliquely radial rounded striije separated by slightly narrower deep grooves. Pale brown in color. Aperture but slightly oblique, heart-shaped, peristome thin, acute, a trifle dilated at its insertions. Diam. 3.3, alt. (thickness) 1 mm. "Sinking Spring" near San Marcos, Hays county, Texas; Guada- lupe river about four miles above New Braunfels, Comal county. Rio San Filipe and Devil's river, and canyon of the Pecos river about a mile above the High Bridge, Val Verde county; everywhere in drift debris. Types from the last locality. This little Planorbis is very distinct by its beautiful sculpture, con- stant in numerous specimens from five rivers in central and western Texas. It is much more abundant in the Rio Grande drainage than 1906.] NATURAL SCIENCES OF PHILADELPHIA. 165 in Hays and Comal counties. It is about the size of P. parvus, but the aperture is less oblique and the sculpture differs. It was found with parvus in Comal county and in the Pecos canyon. Planorbis filocinctus n. sp. PI. IX, figs, l, 2, 3. Shell very small, biconcave, the spiral narrower and more deeply sunken on the left than on the right side. Whorls about 2f , the last broadly rounded peripherally, rounded also on both sides, but less convex on the right than on the left side. Sculpture of inconspicuous growth-lines and numerous thread-like spiral striae. Color of bleached specimens white or faintly yellowish. Aperture oblique, heart-shaped, about as long as wide, excised moderately by the preceding whorl. Greatest diam. of the disk 2.4, alt. (thickness) nearly 1 mm. San Pedro river, Benson, Arizona, in drift debris. Types collected by J. H. Ferriss, 1904. This little species may be at once recognized by its spiral striation, which is far stronger than in any other North American species. It is flattened less than P. parvus. Only five specimens were taken, but it is so unlike other known Mexican or United States Planorbes that there seems no doubt of its specific distinctness. Planorbis parvus Say. Guadalupe river al^out four miles above New Braunfels, Comal county; Devil's river, Val Verde county, Texas. San Pedro river, Benson, Arizona. Planorbis cultratus Orb. Pilsbry, Nautilus, III, p. 63, pi. 1, figs. 1, 2, 3. This species is thin and delicate, very acutely carinate at the periphery which is close to the left side. In 1889 one of us reported it from Hidalgo, Texas, where it was taken by Mr. Singley. In 1903 we took a single young shell in the drift debris of Devil's river, Val Verde county. The young have a much less acute keel, in fact are angular rather than carinate, and they are less compressed in proportion. P. cultratus is found also in Florida (collected at Miami by S. N. Rhoads), Central America and the West Indies. Few of the specimens thus far known from Texas are fully mature or in good condition. Planorbis liebmanni Dkr. Canal at New Orleans, Louisiana (H. Hemphill !). In Texas at Waco (Hemphill), Austin (E. Hall), San Marcos, New Braunfels, Hondo river, Del Rio, Devil's river and Pecos river (Ferriss and Pilsbry); also in the southeastern part of the state in Victoria county (J. D. Mitchell) and near Brownsville, Cameron county (Clapp). Also in ^lexico. 166 PROCEEDINGS OF THE ACADEMY OF [March, Well-grown specimens have a diameter of 9 to 10 mm., with about 5 whorls. Segmentina obstructa (Morel.)- Except for the teeth, this species is not distinguishable from Plan- orbis liebmanni Dkr. It does not, however, attain quite so large a size, the largest we have seen measuring slightly less than 9 mm. diameter. The teeth are found in very young shells also ; but never more than one set is present at any stage of growth, so far as seen. Specimens are before us from the following places, all in Texas : Austin (E. Hall); San Marcos, Hays county; New Braunfels, Comal county; Hondo river north of Hondo, Medina county (Pilsbry and Ferriss) ; Brownsville, Cameron county (sent by G. H. Clapp) ; Hidalgo, Hidalgo county (Singley) ; Rio San Filipe, Devil's river and Pecos river, Val Verde county (Ferriss and Pilsbry). It has also a wide range in Mexico. The genus Segmentina was based upon the European species S. nitida Miill. This is a very glossy, flattened shell with acutely angular periphery, simple thin lip, deeply embracing whorls, and barriers com- posed of three laminae (parietal, basal and upper) transverse to the whorl, leaving a narrow, three-branched space between them. In eastern Asia a modification of this type is found in such species as S. largillierti (Phil.), forming the subgenus Polypylis Pils. The shell is less compressed and not carinate, but glossy with deeply clasping whorls. The parietal lamina is obliquely transverse, the others transverse, basal long, a shorter one in the outer wall, and one or two in the upper margin. There are several or many barriers. The American forms, subgenus Planorbula Hald., have less smooth and much less compressed shells, the whorls only slightly clasping, often angular or subangular on the right side but rounded peripherally. There are six laminae: a sigmoid, obliquely entering parietal with a small tubercular denticle near its lower or left end ; a transverse basal ; an obliquely entering outer lamina with a transverse one above it, and a small entering lamina in the upper margin. In S. armigera the entering lamella in the outer margin curves up- ward slightly at its inner end. In S. wheatleyi all of the laminae are much more strongly developed, and the entering outer one is much longer, running up in a long curve behind the transverse lamina above it. The structures are, however, fundamentally identical in the two species. In the Antillean and Mexican group to which S. obstructa belongs 1906.] NATURAL SCIENCES OF PHILADELPHIA. 167 the whorls are rounded, the parietal laminse are as in Planorbula ex- cept that the larger one stands more obliquely, the basal lamina i transverse, but all the rest on the outer wall are entering plicce, the larger lower one curving downward a little at its inner end, not upward as in armigera and wheatleyi. As in all American forms of the genus only one set of laminae seems to be present in any individual, though the earliest set is formed at a very early age. In the European and Asiatic species several sets are usuall}' present. PLEUROCERATID^. Goniobasis comalensis Pilsbry. Figs. 24-28. Melania rufa Lea?, Romer, Texas, p. 457 ("In den Quellen des Comal- Spring bei New IBraunfels sehr haufig"). Melania pleuristriata Say, A. G. Weatherby [Wetherby] American Naturalist, April, 1878, p. 254, with var. marmocki (springs of southwestern Texas)! Goniobasis comalensis Pils., Nautilus, IV, p. 49, Sept., 1890 (Comal creek, New Braunfels). G. 'pleuristriata Say and G. cornalensis Pils., Single/, Contrib. Nat. Hist. Texas, Geol. Surv. Tex. Ann. Rep., 1892, pp. 311, "312. Shell conic-turrite, thin but strong, covered with an olive-brown cuticle. Whorls of the spire with a distinct keel which projects a short distance above the suture, and is usually wanting on the last two Fig. 24. Fig. 25 Fig. 27. Fig. 28. whorls; the whole surface showing fine sigmoid growth-striae, and in the best specimens very faint, minute, spiral striae. Aperture ovate, the outer lip thin, sigmoid, retracted below the upper insertion ; basal lip rounded or subangular. Columella arcuate, somewhat thickened. Wliorls about 7 in the most perfect shells, but usually fewer, the upper ones being eroded. Length 18, diam. 7.3, aperture 7.3 mm. Comal creek at New Braunfels, Comal county, Texas. Also in the 168 PROCEEDINGS OF THE ACADEMY OF [March Guadalupe river 6 to 4 miles to the north; San Marcos river at San Marcos, Hays county, on the dam above the fish hatchery (Pilsbry and Ferriss; Singley); a variety from Bexar county (A. G. Wetherby). This form differs from Melania pluristriata Say (M. rubida Lea) of central Mexico, by its far smaller size and less ample aperture. The type of pluristriata measured 1.25x.oo inches; of rubida 1.30x.57 inches; and the specimens -from Lake Chapala examined by Dr. von Martens from 26 x 12 to 35 x 13 mm., while no Texan shell we have seen, in many hundreds examined, reaches 1 inch long. Von Martens refers the Mexican species to Pachycheilus. Dr. William H. Dall, to whom we submitted specimens of G. comalensis for comparison with the type of M. rubida Lea, writes that "though the general form is the same, the rubida is very much larger and perfectly distinct." The examples figured are from Comal creek. New Braunfels, near the mill. The ascertained distribution of G. comalensis comprises only the short rivers of a single small system emptying into Espirito Santo Bay, intercalated between the Colorado and Nueces basins, and comprising the Guadalupe and San Antonio rivers and their tributaries. The Goniobases are known from these streams only at the edge of the "Edwards Plateau." As they live on rocks, etc., in swift water, it is not likely that they approach much nearer to the Gulf. In the Fig. 29. Fig. 30. Fig. 31. Colorado river, where Pilsbry collected in 1885, nothing was seen of the species, nor has it been reported from the Nueces or its branches. We saw no sign of Goniobasis in the tributaries of the Rio Grande, where our collecting was extensive enough to have found it if it existed.^' The operculum (fig. 29) consists of about four whorls, the nucleus being situated at about the lower third of its length. It is closely ^^ In the collection of the Academy there are several specimens labelled ' ' Dal- las county," received from Mr. J. A. Single^^ This is in the Trinity river drainage, but Singley in his catalogue of 1893, issued long after these specimens were placed in thejcollection, expressly states that he found the species nowhere but in Comal and Hays counties. There was probably some error in labelling the^specimens. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 169 marked with growth-wrinkles, and shows some striae at right angles to them. The radula (fig. 30) is normal for Goniobasis. The central tooth has 4, 1, 4 denticles, the inner lateral 2, 1, 4, inner uncinus 7, outer about 20 denticles. An inner uncinus in profile is drawn on the ex- treme right side (fig. 31), to show the length of the denticles, which are foreshortened in fig. 30. G. comalensis is dimorphic, like G. virginica and many other species. At New Braunfels the smooth form described as typical predominates, but there are also some individuals with numerous acute spiral ridges, about fifteen on the last whorl, but often fewer by suppression of those just below the periphery. At San Marcos only the smooth phase was found, in several hundred specimens taken. A series of 12 labelled "Bexar county," received from A. G. Wetherby, consists wholly of multistriate shells, which moreover, although collected alive, are of a whitish or livid whitish tint, in this respect approaching the Mexican pluristriatus. The largest of this lot is 23.5 mm. long, 9.3 wide, exceeding in size any seen from Comal or Hays counties. These are part of the lot collected by Mr. G. W. Marmock, of Bexar coimty, and commented on by Wetherby in the American Naturalist for 1878. The "variety marmocki" mentioned by him, but without a word of definition, may have been the smooth form of the species, but there is nothing to indicate this either in Wetherby's note or the set of shells he sent to Tryon.^^ P'orm fontinalis, nov. Figs. 32-35. In a small spring in the pleasure garden near New Braunfels, one of the fountains of Comal creek, we found only very small shells, the largest 7 to 8.3 mm. long, 4.3 to 4.7 mm. wide, and of a markedly short, conic shape. The old ones are much eroded, and none are of the multi- striate phase. This race inhabits only the springs and the rapid streams from them for a short distance down. In another, much larger spring, and the stream from it for about fifty yards down, the shells are also dwarfs, though somewhat larger than those from the smaller spring. These springs flow out of the limestone rock, the water being beauti- fully clear. It is not cold, being perceptibly warmer than the river at the time of our visit, about the middle of April. There is very little vegetation upon the rocks, and the small size of the snails may be due " Mr. Wetherby also mentions that "Helix photus Pfr." was collected by Mr. Marmock. This name may be an error for H. tholus W. G. B. 170 PROCEEDINGS OF THE ACADEMY OF [Mavch, ^^ Fig. 32. Fig 33. Fig. 34. Fig. 3' to insufficient food supply. Individuals are, however, very numerous. Associated with them are numerous Physas, also of pygmy proportions, though evidently adult. The snails of these springs evidently consti- tute physiologic rather than morphologic varieties. AMNICOLID^ Paludestrina seemanni (Ffld.). New Mexico: South Spring creek, near Roswell, in a Pleistocene deposit (Cockerell and Tinsley, 1899). Texas: Drift debris of Pecos river, about a mile above the High Bridge, Val Verde county (Pilsbry and Ferriss, 1903). Dr. R. E. C. Stearns has recorded this species from Death valley, Inyo county, California. The specimens, some of which are before me,, are somewhat more robust than those from New Mexico and Texas. The above records largely increase the eastward range of the species. Paludestrina diaboli n. sp. Fig. 36. Shell very slender, turrite, shaped about like P. seemanni; composed of 4^ very convex whorls separated by a deep suture. Surface smooth. Aperture vertical, oval, a trifle nar- rower above than below, but not angular there. Peris- tome continuous, barely in contact with the preceding w^horl for a short distance near the upper end. Umbili- cus small but distinct. Length 1.3, diam. .62 mm. Drift debris of the Devil's river, about four miles from its mouth, Val Verde county, Texas. A single shell was also found on the Rio San Filipe near Del Rio, in the same county. The shells were all taken dead and bleached. It is readily separable from P. seemanni by its diminutive size. This is the smallest species, of its familv known from North America. Fig. 36. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 171 Amnicola comalensis n. sp. Fig. 37. Shell distinctly perforate, ovate, thin, corneous, faintly marked with growth -lines. Spire regularly conic, the apex obtuse. Whorls 4^, regularly con- vex, not shouldered, the suture w-ell im- pressed. Aperture ovate, subangular above, the peristome adnate for a short distance above the perforation. Length 3.9, diam. 2, length of aperture 1.3 mm. Comal creek, near New Braunfels, Comal county, Texas. Also from the Guadalupe river about four miles above New Braunfels. This species is much smaller than A. limosa, Fig. 37. decisa, or other forms resembling it in color and shape. A. cincinnatiensis Anth. and the very closely related A. peracuta P. and W. both have more shouldered whorls, and are much larger than A. comalensis. CocMiopa riograndensis n. sp. PI. IX, figs. 10, 11, 12, 13. Shell of the usual depressed-turbinate shape, openly umbilicate, of a slightly olivaceous corneous tint. Surface faintly marked with growth-lines, and sculptured with unequal spiral threads, a few of the larger ones dark colored. One thread at the shoulder is usually the most prominent. In some shells the spirals are very weak, hardly perceptible. Whorls 3^, moderately convex, flattened and sloping below the suture, elsewhere rounded. Aperture quite oblique, rotundly ovate, the peristome thin, equably arched except near the outer and columellar insertions where it is noticeably straightened. The ends are connected across the parietal wall by a thin or thick callus. The columella is not noticeably thickened. Alt. 2, diam. 2.8 mm. Alt. 1.65, diam. 2.65 mm. Found in drift debris of Rio San Filipe near the Rio Grande, Val Verde county, Texas, thirty-six specimens. Pilsbry and Ferriss, 1903. Types, 91,324, A. N. S. Phila. A single shell was also picked out of similar debris on the Devil's river, about four miles from the Rio Grande, in the same county. This discovery of this form adds a new genus to the fauna of the United States and greatly extends the distribution of the group, the 172 PROCEEDINGS OF THE ACADEMY OF [March, other species being Central American. ^^ The shell has much the ap- pearance of Valvata, but is readily distinguished from that by its ovate, not circular aperture, which is more or less conspicuously angular above, and by having the whorls more compactly coiled. It should, however, be said that we could not verify the generic reference of C. riograndensis by examination of the operculum or radula, as unfor- tunately none of the shells we obtained were living, though some were fresh in appearance, and the operculum was not found. It consists of few whorls in Cochliopa, as in Amnicola, while that of Valvata is many-whorled with central nucleus. C. riograndensis is not closely related to any of the known species, being of lighter texture, much more openly umbilicate, and with some- what different sculpture. VALVATID^. Valvata micra n. sp. PI. IX, figs. 7, 8, 9. Shell excessively small, composed of 2^ tubular whorls ; spire nearly flat; the last whorl is nearly round, barely or not quite in contact with the preceding at the aperture, near which it enlarges more rapidly. Suture deep. Surface finely, weakly striate. Aperture moderately oblique, subcircular, the peristome simple, continuous. Umbilicus ample but rapidly narrowing within. Alt. .48, diam. 1.15 to 1.2 mm. Drift debris of Guadalupe river about four miles above New Braun- fels, collected by Pilsbry and Ferriss, 1903. This is one of the smallest mollusks known, yet the dilation of the " Prof, von Martens defines four species of Cochliopa in the Biologia Centrali Americana — guatemalensis Morel., tryoniana Pils., trochulus and infundibulum Marts, (pp. 428, 429). He omits C. roivelli Tryon, which was originally described from Clear Lake, California. Rev. J. Rowell, who found the original specimens, now states (m litt.) that " Cochlicopa Rowellii was named from shells collected by me near Baulinas Bay (not Clear Lake), Marin county, California." There are authentic specimens from Central America in the collection of the Academy, so that I still doubt whether the species really was actually collected in Cali- fornia. Prof, von Martens surmises that C. guatemalensis (Morel.), which he had not seen, may not be different specifically from C. tryoniana ; but it is in fact not even closely related to that species, being very much smaller, more depressed, and evenly sculptured with threadlike spirals." The suture descends shortly in front in fully mature shells. An additional locality is Polvon, in western Nicara- gua (McNiel in coll. A. N. S. Phila.). I have examined the radula of C. guate- malensis which proves to be Amnicoloid, and demonstrates the species to be a Cochliopa, not a Valvata as Morelet supposed. The central tooth has the formula ~, tjie inner lateral has 6 denticles, the third from the inside being largest, and its body has the usual boss or projection below. The inner uncinus has 13, the outer very many excessi^'ely minute denticles. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 173 whorl at the aperture shows it to be adult. Five specimens were taken. The shells are all more or less bleached, the freshest being of a pale corneous tint. The round whorls, deep suture and general appear- ance are quite unlike any Planorbis or Vallonia, and indicate, we think, either that it is a Valvata or an Amnicoloid snail comparable to Horatia Bgt. or Daudehardiella Bttg. in the Paltearctic fauna. Until fresh specimens with the soft parts or operculum are found, the position of this molluscan atom will remain uncertain. Valvata micra nugax (pi. IX, fig. 6), a slightly larger form, alt. .9 diam. 1.5 mm., with three whorls and a projecting spire, may prove to be a distinct species; but for the present, until more specimens are found, it may be placed under T^. micra as a variety or form. CYRENIDu^]. Pisidium singleyi Sterki. Nautilus, XI, 1S9S, p. 112 (type loc. Guadalupe river, Comal county, Texas). Drift debris of Guadalupe river about four miles above New Braun- fcls, and of Devil's river, \'al Verde county, Texas. A closely related form was found near Del Rio. Pisidium compressum Prime. New Braunfels, Comal county, Texas. Mr. J. A. Singley found P. trapezoideum Sterki at the same place (coll. A. N. S. Phila. No. 60,127). Pisidium abditum huachucanum n. subsp. The shell is quite inflated, dark brownish-olive, irregularly striate and marked with several conspicuous dark growth-arrest streaks; very inequilateral, the beaks low and near the anterior end. Anterior end abruptly truncate, posterior end produced and rounded. Hinge rather narrow, the lateral teeth in the right valve single, short and high, triangular ; in the left double. Length 5. 1 , alt. 4.3, diam. 3.4 mm. Stream in Carr canyon, Reef, Cochise county, Arizona, collected by C. R. Biedermann, February 8, 1904. Specimens were submitted to Dr. V. Sterki, who could not identify the form with any known species. About half of the shells are more compressed than those described as typical, one measuring, length 4.1, alt. 3.4, diam. 2 mm. The very inequilateral, anteriorly truncate out- line and low beaks are characteristic of both the obese and com- pressed forms. Eupera singleyi (Pils.). Sphoerium {Limosina) dngleyi Pils., Proc. A. N. S. Phila., 1889, p. 88, pi. 3, figs. 14, 1.5 (May 14, 1889). iCyclas macvJata Morelet, 1859, not of Anton, 1839 =Sphoerium yacatanense Fischer and Crosse, Miss. Sci. Mex., Moll., II, p. 653 (1894). This species is now known to us by specimens from the following 174 PROCEEDINGS OF THE ACADEMY OF [March, places, all in Texas: White Oak bayou, Houston (Singley); Cedar creek, Hudson county (G. C. Heron, type locality) ; Lavaca river, Jackson county (J. D. Mitchell) ; New Braunfels, Comal county (Sing- ley, Ferriss and Pilsbry) ; Guadalupe river about four miles above New Braunfels (Singley, Ferriss and Pilsbry) ; Devil's river, Val Verde county (Ferriss and Pilsbry). It is a prettily maculate little clam, widely distributed in Texan waters. The largest shell I have seen measures 6 mm. long. It may prove to be specifically identical with E. maculata Morelet. of Yucatan, but that name is preoccupied by Anton for another species of the same group. Both were described as Cyclas and belong to Eupera. For this reason Crosse and Fischer proposed to substitute the name yiica- tanense; but five years previously I had described and figured S. sing- ley i. The latter name will therefore stand. UNIONID^. A list of species taken by us in Arkansas, Indian Territory and Texas will be published elsewhere. As our work in these States was done in the early spring, we found most of the streams too high for effective collecting of bivalves. Explanation of Plates V-IX. Plate V. — Figs. 1, 2, 3. — Polygyra mooreana thohis (W. G. B.). Washington county, Texas. No. 251, A. N. S. Pliila. Figs. 4, 5, 6, 7. — Polygyra mooreana (W. G. B.). Guadalupe river, Comal county, Texas. No. 91,364. Figs. 8, 9, 10. — Hondo river near Hondo, Texas. No. 91.361. Figs. 11, 12. — Polygyra texasiana texensis Pils. Tvpes. Colorado City, Texas. No. 83,258. Figs. 13, 14, 15. — Polygyra texasiana hyperolia Pils. and Ferr. West of Devil's river, Texas. No. 91,363. Figs. 16, 17, 20. — Polygyra texasiana (Moric). Typical. Guadalupe river Comal county, Texas, No. 91,362. Figs. 18, 19. — Polygyra texasiana (Moric). Race with striate base, Calhoun county, Texas. No. 229, A. N. S. Phila. Plate VI. — Fig. 1. — Bulimidus dealbatus mooreanus (W. G. B.). San Antonio, Texas. No. 84,626. Figs. 2, 3, 4. — B. d. mooreanus. Guadalupe river above New Braunfels. No. 84,628. Figs. 5, 6. — B. d. mooreanus. Victoria, Texas. No. 76,210. Figs. 7, 8. — B. d. Uquabilis (Rve.). San Marcos, Texas. Nos. 91,396 and 91,397. Figs. 9, 10, 11. — B. d. Uquabilis. Jackson county, Texas. No. 76,286. Fig. 12. — B. d. Uquabilis. Lee county, Texas. No. 58,379 Fig IS.— Bulimulus dealbatus (Say). Type. No, 58,381, A. N. S. Phila. Alabama. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 175 Fig. 14. — BulimuluH d. ozarkensis Pils. and Ferr. Seligman, Missouri. Xo. 83,132 Fig. 15. — B. d. ozarkensis Pils. and Ferr. Rogers, Arkansas. Xo. 91,358. Figs. 16, 17. — Bulimulus d. ragsdalei Pils. Types. Montague county, Texas. Xo. 58,380. Figs. 18, 19, 20, 21, 22.— B. d. ragsdalei. Southwestern herd. West of Devil's river, Val Verde county, Texas. Xos. 91,356 and 84,638. Figs. 23, 24. — B. d. ragsdalei Southwestern herd. Del Rio, Texas. Xos. 87,486 and 91,355. Fig. 25. — Bulimulus d. pasonis Pils. Tvpe. Xear El Paso, Texas. Xo. 83,259. Figs. 26, 27. — Bulimulus d. pecosensis Pils. and Ferr. Xear tlie Pecos river. Xos. 91,359 and 84,618. Plate VII. — Varieties of Bulimulus alternatus marioe (Alb.). Figs. 1-7.— Xear the High Bridge of the Pecos. Xo. 84,627, A. X. S. Phila. Figs. 4 and 5 represent young shells. Fig. 8. — Xear the Rio Grande, east of the Pecos river. Xo. 84,625. Figs 9, 10, 11.— Xear Rio San Filipe below Del Rio. Xo. 84,635. Fig. 12. — Xear Rio San Filipe below Del Rio. An albino, taken alive. Figs. 13, 14, 15, 16.— Corpus Christi, Texas. Xo. 60,136. Figs. 17, 18, 19.— Hidalgo, Texas. Xo. 60,094. Figs. 20-24.— Derby, Frio countv, Texas. Xo. 60,501. Figs. 2.5-30.— Laredo, Webb county, Texas. Xo. 60,502. Plate VIII. — Figs. 1, 2, 3. — Helicodiscus eigenmanni Pils. San Marcos, Texas. Xo. 91,320. Figs. 4, 6. — H. e. arizonensis Pils. and Ferr. Fort Bowie, Cochise county, Arizona. Xo. 87,077, A. X. S. Phila. Fig. 5. — H. e. arizonensis. A more depressed specimen. Cave Creek Canyon, Chiricahua Mountains Xo. 87,076. A. S. X. Phila. Figs. 7, 8, 9, 10. — Helicodiscus parallelus (Say). Philadelphia, Pa. Xo. 78,272, A. X. S. Phila. Plate IX. — Figs. 1, 2, 3. — Planorbis filocincius Pils. and Ferr. Type. Figs. 4, 5. — Planorbis carus Pils. and Ferr. Type. Fig. 6. — Valvata micra nugax Pils. and Ferr. Type. Figs. 7, 8, 9. — Valvata niicra Pils. and Ferr. Type. Figs. 10, 11, 12, 13. — Cochliopa riograndensis Pils. and Ferr. Cotypes. 176 PROCEEDINGS OF THE ACADEMY OF f Aprils April 3. Mr. Arthur Erwin Brown, Vice-President, in the Chair. Eighty-nine persons present. The reception of a paper entitled "New, Rare and Little-known Scombridse, No. 3," by Henry W. Fowler (March 22), was reported by the Publication Committee. The death of Dr. Ogden Doremus, a correspondent, March 22, was announced. Dr. E. G. Conklin made a communication on recent views as to the cause of sex. (No abstract.) April 17. Mr. Frank J. Keeley in the Chair. Seventy persons present. The Publication Committee reported the reception of papers under the following t'tles: "Description of a new Australian Glycymeris," by Henry A. Pilsbry (April 10). "A Monograph of the Genus Cohocalia," by Harry C. Oberholser (April 12). "On Hawaiian Species of Sphyradium," by Henry A. Pilsbry and C. M. Cooke, Jr. Mr. George Vaux, Jr., exhibited a number of beautiful lantern illustrations before the reading of a paper by William S. Vaux, Jr., giving the results of recent studies of glaciers 'of the Canadian Rockies and the Selkirks, in continuation of a series of observations extending over a period of eight years. The paper will be published later. Messrs. Ogilsby Paul and Norton Downs, M.D., were elected mem- bers. The following were ordered to be printed[: 1906.] NATURAL SCIENCES OF PHILADELPHIA. 177 A MONOGRAPH OF THE GENUS COLLOCALIA. BY HARRY C. OBERHOLSER. There is perhaps no other group of Swifts that equals in general interest the genus Collocalia. Some of the species build edible nests wholly or partly of saliva which is secreted in large quantities by salivary glands of unusual size; and the gathering and marketing of these nests has in some places and at some times grown to a traffic of considerable importance. Owing to the great similarity of many of the species, and the ob- scurity of the characters separating them, the group is one of the most difficult in the entire family Micropodidse. Published descriptions,^ particularly those of the older authors, are often so brief and so lacking in mention of really diagnostic characters that it is hard, sometimes even impossible, to determine with certainty the form intended. For- tunately, perhaps, the sexes, except in a very few cases, are practically identical both in size and coloration. Notwithstanding the fact that some authors have minimized the importance of difference in size, ^ The most important papers on the genus are as follows: HoRSFiELD, T., and Moore, F. — Genus Collocalia. A Catalogue of the Birds in the Museum of the Hon. East India Company, I, 1854, pp. 98-106. Bonaparte, C. L. — Note sur les Salanganes et sur leur nids. Comptes Rendus de r Academie des Sciences de Paris, XLI, 1855, pp. 976-979. Bernstein, H. A. — Beitrage zur naheren Kenntniss der Gattung Collocalia Gr. Nova Acta Academiae Leopoldinae Carolinae Naturae Curiosorum, XXVI, 1, 1857, pp. 13-32. Bernstein, H. A. — Ueber die Nester der Salanganen. Journal fur Ornithologie, 1859, pp. 111-119. Wallace, A. R. — On the Identification of the Hirundo esculenta of Linnaeus, with a Synopsis of the Described Species of Collocalia. Proceedings of the Zoological Society of London, 1863, pp. 382-385. Gray, G. R. — A Synopsis of the Species of the Genus Collocalia, with Descriptions of New Species. Annals and Magazine of Natural History, Series 3, XVII, 1866, pp. 118-128. Salvadori, T. — Osservazioni intorno ad alcune Specie del Genere Collocalia, G. R. Gr. Atti delta R. Accademia della Scienze di Torino, XV, 1879, pp. 343-350. Hartert, E. — Collocalia. Catalogue of the Birds in the British Museum, XVI, 1892, pp. 496-511. Hartert, E.— Gen. Collocalia G. R. Gray. Das Tierreich, I, 1897, pp. 66-70. 12 178 PROCEEDINGS OF THE ACADEMY OF [April, this is often one of the very best and most appreciable distinctions between both species and subspecies. One of the most sm-prising as well as interesting developments of our study has been the discovery of so many undescribed members of the genus, among which two entirely distinct species with wholly brownish lores like Collocalia leucophaea seem especially notable, as does also one from New Caledonia allied to Collocalia leucopygia. Most of the new forms are here elaborated from material recently col- lected, and upon which no published report has yet been made. Much confusion has hitherto arisen because of failure properly to discriminate those birds which have the tarsus more or less feathered from those in which it is entirely bare. In some cases forms with feathered and others with unfeathered tarsi have been considered sub- specifically related or even identical ! The difficulty vanishes, however, if all the birds with any feathering on the tarsus be considered apart from the rest and placed, as they apparently should be, and as they are in the following pages, in a separate subgeneric group. Indeed, it may be best to regard them as even generically distinct, though it is not here so done, because this character of tarsal feathering is the sole one separating the groups, and while very marked in such forms as Col- localia innominata, it is but slight, sometimes difficult to appreciate, therefore in a sense intermediate, in Collocalia fuciphaga and its allies ; and because if there be any logical difference between a genus and a subgenus, the criterion of practical intergradation of characters through intermediate species should be so considered. The proper generic name for the whole group seems to be Collocalia — the one of current usage. The name Salangana St.-Hilaire has been recently substituted ^ because supposedly of earlier date, but Salangana now proves first to have been employed simply in a vernacular sense,^ and to be citable as a generic term only from a later article ^ which is posterior to the work containing Collocalia Gray.^ There are no other synonyms. The material used in the preparation of the present memoir aggre- gates 159 specimens, principally of recent collection, and represents very nearly all the recognized forms. It is in large part that of the United States National Museum, which is now, through the many donations from Dr. W. L. Abbott, of considerable extent and includes good series of many of the species. This has been supplemented by 2 Richmond, Proc. U. S. Nat. Mus., XXV, 1902, p. 301. ^ " Salanganes, " Saint-Hilaire, I' Echo du Monde Savant, IV, 1837, p. 84. *Rev. Zool, 1840, p. 145. List Gen. Birds, 1840, p. 8. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 179 the birds obtained by Mr. Charles H. Townsend during the Pacific cruise of the United States Fish Commission steamer "Albatross," in 1899-1900, and by specimens from the Academy of Natural Sciences of Philadelphia, the American Museum of Natm-al History, and the Bernice Pauahi Bishop Museum at Honolulu. To the authorities of these institutions, to Mr. Townsend, Dr. Charles W. Richmond, and Mr. Witmer Stone the -\\Titer's thanks are due for the courtesies which have made this paper possible. Genus COLLOCALIA Gray. CoUocalia Gray, List Gen. Birds, 1840, p. 8 (type, Hirundo esculenta Linnaeus). Salangana Lesson, Rev. Zool., 1840, p. 145 (type, Hirundo esculenta Linnaeus). Chars, gen. — Small and mostly rather plainly colored Swifts; outer and middle toes with normal number of phalanges; tarsus not at all or but sparsely feathered, and not shorter than the middle toe; wings long, reaching when closed far beyond end of the tail ; tail short, nearly square to somewhat deeply emarginate, the shafts of the rectrices without spiny tips. Type. — Hirundo esculenta Linnaeus. Geographical range. — India and Ceylon, south to the Seychelles Islands, east through the Malay Peninsula, the East India Islands, and Polynesia, to the Philippine, Mariana, ^larquesas, Society and Tonga groups, and northern Australia, Key to the Species and Subspecies of Collocalia. A. — Tarsus more or less feathered (subgenus Aerodramus). a. — Rump with a conspicuous white or brownish white band. h. — Centre of abdomen brownish gray like the breast; wing more than 120 mm., .... Collocalia innominata. b^. — Centre of abdomen white or nearly so; wing less than 120 mm. c. — Upper parts brownish black; throat and breast nearly pure white like the abdomen; posterior lower surface without conspicuous dark shaft lines, Collocalia agnota. c\ — Upper parts dull black, scarcely brownish; throat and breast brownish gray like chin, and much darker than abdomen; posterior lower surface with conspicuous dark shaft lines, .... Collocalia leucopygia. a^. — Rump without a conspicuous white or brownish white band. h. — Feathers of lores with brownish bases, . Collocalia ocista. ¥. — Feathers of lores with pure white bases. c. — Wing not less than 124 mm. d. — Larger; upper surface lighter; lower surface with conspicuous blackish shaft lines, Collocalia lowi. 180 PEOCEEDINGS OF THE ACADEMY OF [April, d^. — Smaller; upper surface darker; lower surface without conspicuous blackish shaft lines, Collocalia brevirostris. c^ — Wing less than 124 mm. d. — Larger (wing about 121 mm.); upper surface more brownish; rump noticeably paler than back, Collocalia fuciphaga elaphra. d^. — Smaller (wing 108-118 mm.); upper surface more blackish; rump not noticeably paler than back. e. — Lower surface paler; upper surface somewhat more brownish, Collocalia fuciphaga vestita. -•e^ — Lower surface darker; upper surface somewhat more blackish, Collocalia fuciphaga fuciphaga. B. — Tarsus entirely unfeathered (subgenus Collocalia). a. — Abdomen not white in sharp contrast to the breast, the lower surface nearly uniform brownish gray. h. — Rump without a conspicuous white or brownish white band, c. — Feathers of the lores with brownish bases, Collocalia leucophaea. c^. — Feathers of the lores with pure white bases. d. — Larger (wing 127-140 mm.). e. — Upper surface darker, much more blackish ; lower surface darker, and uniform, Collocalia origenis. eV — Upper surface lighter, much more brownish; lower surface lighter, the throat paler than the rest, Collocalia whiteheadi. rfi.— Smaller (wing 109.5-123 mm.). e. — Upper parts more brownish, more uniform-, the pileum scarcely darker, the rump not lighter, than the back; throat not paler than abdomen. Collocalia unicolor unicolor. e^. — Upper parts less brownish, less uniform, the pileum darker, the rump usually lighter, than the back; throat paler than abdomen, Colloccdia unicolor amelis. 51. — Rimip with a conspicuous white or brownish white band, c. — Feathers of lores with brownish bases, Collocalia thespesia. c^ — Feathers of lores with pure white bases. d. — Smaller (wing 110-114 mm.). e. — Abdomen and upper surface darker; light rump band with more conspicuous dark shafts ; size smaller (wing 110-111 mm.), Collocalia francica terraereginae. gi. — Abdomen and upper surface paler; light rump band with less conspicuous dark shafts; size larger (wing 112-114), Collocalia francica francica. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 181 d\ — Larger (wing 114-124 mm.)- e.— Lower surface with distinct blackish shaft lines ; light rump band with dark shaft lines very conspicuous. /.—Upper and lower parts much paler; light rump band more conspicuous, Collocalia francica germani. f\ — Upper and lower parts much darker; light rump band less conspicuous, Collocalia francica inexpectata. gi .—Lower surface without distinct blackish shaft lines; light rump band with dark shaft lines almost obsolete. /.—Upper surface with little or no greenish gloss ; rump band more brownish ; posterior lower parts darker, Collocalia francica spodiopijgia. /I. —Upper surface with a distinct greenish gloss ; rump band more whitish; posterior lower parts paler, Collocalia francica townsendi. a\—At least the middle of abdomen white, in sharp contrast to the gray breast. h. — Rump with a white band. c— Rectrices with white spots; white rump band without blackish shaft lines, . . . Collocalia uropygialis. ci.— Rectrices without white spots; white rump band with conspicuous blackish shaft lines, Collocalia troglodytes. b^. — Rump without a white band. c. — Rectrices wath white spots. ^.— Upper surface brighter and more bluish or purplish ; rump, wing-quills, and wing-coverts without light grayish tips, Collocalia escidenta. rfi.— Upper surface duller and more greenish; rump, wdng-quills, and wing-coverts with light grayish tips. g.— White spots on tail-feathers sharply defined ; chm and upper throat paler; blackish shaft-mark- ings of short lower tail-coverts smaller, Collocalia neglecta. gi.— Wiiite spots on tail-feathers not sharply defined ; chin and upper throat darker ; blackish shaft- markings of short lower tail-coverts larger, Collocalia natalis. c^ — Rectrices without white spots. f/.— Feathers of rump with broad white edgings, Collocalia marginata. f^i.— Feathers of rump without white edgings. g._\Ving less than 95 mm.; upper surface much duller, Collocalia dodgei. 182 PROCEEDINGS OF THE ACADEMY OF [April, e^. — Wing more than 95 mm.; upper surface much brighter. [/. — Upper parts much more bluish or purplish. g. — Larger (wing averaging 104.6 mm.), Collocalia linchi cyanoptila. g^. — Smaller (wing averaging 99.2 mm.), Collocalia linchi afjinis. /'. — Upper parts much more greenish. g. — Larger (wing 103-105 mm.), Collocalia Ihichi linchi. g^. — Smaller (wing 97-101.5 mm.).^ h. — Upper surface duller, somewhat less bluish or purplish; size slightly greater, Collocalia linchi isonota. h^. — Upper surface brighter, somewhat more bluish or purplish ; size slightly less, Collocalia linchi elachyptera. AERODKAMUS.e subgenus nov. Chars, subgen. — Similar in i3roportions to the subgenus Collocalia, but tarsus more or less feathered. Type. — Collocalia innominata Hume. Collocalia leucopygia Wallace. Collocalia leucopygia Wallace, Proc. Zool. Soc. Lond., 1863, p. 384. Chars, sp. — Upper surface dull, slightly sooty black, with a weak metallic gloss, most evident on crown, the rump with a broad white band whose feathers have conspicuous blackish shaft lines ; wings and tail black with some metallic sheen, the latter without white spots ; sides of head and neck clove brown, the loral feathers with white bases; chin and throat brownish gray; remaining lower parts pale grayish anteriorly, becoming brownish on sides of body, and whitish on abdo- men, all the feathers with distinct darker shaft lines; longest under tail-coverts glossy blackish brown; thighs and lining of wdng brownish black. Wing, 95-99.5; tail, 45; exposed culmen, 4; tarsus, 8.5 mm. Type locality. — New Caledonia. Geographical distribution. — Loyalty Islands, New Hebrides Islands, and New Caledonia. This species may readily be distinguished from all its congeners with feathered tarsi, excepting Collocalia innominata and C. agnota, by the white band on the rump ; and from all but Collocalia agnota by the " arjp, aer; Spafidv, cursare. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 183 white or whitish abdomen. In color it approaches rather near some species of the subgenus Collocalia, notably some forms of Collocalia francica, but the whitish abdomen is of course diagnostic. Collocalia agnota sp. nov. Chars, sp. — Similar to Collocalia leucopygia, but upper surface lighter, much more brownish; chin, upper throat, and sides of head much paler brown ; ventral surface from the lower throat to anal region more purely white, the jugulum and breast white like the abdomen, and contrasting sharply with the grayish brown of the chin and upper throat; feathers of breast and abdomen without noticeably darker shaft lines ; basal portion of rectrices much paler. Geographical distribution. — Island of New Caledonia. Description. — Type, adult female. No. 8,757, American Museum of Natural History; New Caledonia; Verreaux Collection. Upper parts brownish black, the pileum with some metallic greenish gloss, the l^ack with a faint suggestion of the same, the cervix a little lighter and more brownish ; rump with a broad band of pure white, the feathers of which have dark brown shaft lines; rectrices externally brownish black, with a slight bluish metallic sheen, internally more brownish, and proxi- mally growing paler until at base they are quite light brown, but without white spots; wings brownish black with some metallic sheen, the inner margins of quills fuscous; lores, orbital region, and a narrow line across the forehead sepia brown, the feathers of the lores with pure white bases; sides of neck and breast grayish brown, much lighter than the cervix; chin and upper throat dull broccoli brown; rest of lower parts medially almost pure white, though just appreciably brownish an- teriorly, the sides of body washed with brownish, the shorter lower tail-coverts pale brownish gray with darker shaft lines and whitish margins, the longest ones dark brown, with distally a metallic sheen; thighs brownish black; lining of wings dark brown. Wing, 105; tail, 46; exposed culmen, 4; tarsus, 9 mm. The relationships of this evidently distinct and apparently unde- scribed species are undoubtedly with Collocalia leucopygia, as is indi- cated by the conspicuously feathered tarsi, the white abdomen, and the white rump band with dark shaft lines. The only specimen known is the above described one in the American Museum of Natural History. Collocalia innominata Hume. Collocalia innominata Hume, Stray Feathers, I, 1873, p. 294. Collocalia maxima Hume, Stray Feathers, IV, 1876, p. 223 (Mergui and Bank- asoon, Tenasserim) (nomen nudum). Chars, sp. — Quite different from Collocalia leucopygia: much larger; 18:1 PROCEEDINGS OF THE ACADEMY OP [April, entire lower surface dark brownish gray; upper parts more brownish, the light rump band decidedly less conspicuous. Wing, 130-137; tail, 50-56; exposed culmen, 6; tarsus 12-12.5 mm. Type locality.— Fort Mouat, South Andaman Island, Andaman Islands. Geographical distribution.— Sonthern Tenasserim to Lower Siam and Perak, Malay Peninsula; Mergui Archipelago; accidental(?) on South Andaman Island. The large size, feathered tarsi, light brownish rump band with dark shaft lines, and the conspicuously darker mesial lines of the lower surface, which last are present in only a few forms of CoZZocaha, 'suffic- iently distinguish this very strongly characterized species. An adult female taken by Dr. W. L. Abbott on September 5, 1896, at Trong, Lower Siam, proves its occurrence in that region. Collocalia ocista sp. nov. Chars, sp. — Similar to Collocalia leucophaeae, but tarsi feathered ; slightly smaller, except the tail which is longer ; upper surface, includ- ing wings and tail, darker, more blackish (less brownish), and on head and back as well as elsewhere with a decided greenish metallic gloss; lower parts more grayish. Geographical distribution. — Marquesas and Society Islands. Description. — Type, adult female; Nukahiva Island, Marquesas Islands, September 16, 1899; Charles H. Townsend. Upper surface almost uniform dark sooty brown, slightly deeper on the pileum, a little lighter on the rump — neither of these differences very noticeable — and everywhere with a greenish metallic sheen; wings and tail darker, more blackish, with a bluish or purplish metallic gloss, the wing-coverts slightly more greenish, the innermost secondaries and inner margins of the quills, at least basally, lighter and more brownish — about the same color as the back; sides of head clove brown almost as dark as the crown, the lores quite so, the loral feathers with lighter brown bases; lower parts nearly uniform deep brownish gray, only the chin and longest under tail-coverts somewhat darker; lining of wing clove brown. This new and very interesting bird is superficially so much like Collocalia leucophaea, with the type of which it has been compared, that a specimen in the Academy of Natural Sciences, Philadelphia, collected long ago in the Marquesas Islands, and presented by Dr. Gambel, was so labelled ; but the species may readily be separated by the characters above given. It is fully as dark below as C. leucophaea, and like that species is peculiar in possessing light brown bases to the 1906.] NATURAL SCIENCES OF PHILADELPHIA. 185 feathers of the lores. From Collocalia fuciphaga fuciphaga, which also to some extent it superficially resembles, it may be distinguished by its larger size and more brownish upper parts, as well as most trench- antly by the light brown instead of pure white bases of the loral feathers. From Collocalia fuciphaga elaphra, a subspecies hereinafter described,^ with which it agrees in size, and which it more closely approaches in the general color of the upper surface, it differs in being rather more blackish and more uniform above, the rump not so appreciably paler than the back; in having much darker lower surface; also, and most decidedly, in having the bases of the feathers of the lores light brown. This last character alone will separate it from all forms of the genus excepting C. leucophaea and C. thespesia.^ Three specimens of this new species were collected by ilr. Charles H. Townsend in 1899, during his recent Pacific cruise on the U. S. Fish Commission steamer ''Albatross." One of these, the single specimen from Tahiti, is rather more brownish above, and a little paler on the crissum than the type, but is very different from C. leucophaea ; another, from the Marquesas Islands, is more blackish above and slightly darker below; but the old specimen in the collection of the Academy of Natural Sciences of Philadelphia is practically identical with our type. Measurements of Collocalia ocista are as follows: Sex. Locality. Date. Wing. Tail. Culmen. Tarsus. 9 Niikahiva I., Marquesas Islands' Sept. 16, 1899 Sept. 15, 1899 121 119 117.5 119 123 65 61 61 61 59 5 4 4.5 4.5 4 8.5 9 8.5 9 9 — Nukahiva I., Marquesas Islands Marquesas Islands 6^ Nukahiva I., Marquesas Islands Tahiti I., Society Islands. July 18, 1902 Nov. 13, 1899 Average 119.9 61.4 4.4 8 8 Collocalia fuciphaga fuciphaga (Thunberg). Hirundo fuciphaga Thunberg, K. Vet. Akad. Nva Handl., XXXIII, 1812, p. 153, pi. 4 (Java). Hirundo vanikorensis Quoy and Gaimard, Voy. Astrolabe, Zool., I, 1830, p. 206, pi. XII, fig. 3 (Vanikoro Island, Santa Cruz Islands, Pacific Ocean). Postea, p. 188. ' Postea, p. 195. ' Type. 186 PROCEEDINGS OF THE ACADEMY OF [April, Cotyle ranicorensis Boie, Isis, 1844, p. 170 (nom. emend, pro Hirundo vani- korensis Quoy and Gaimard). Hemiprocne salangana Streubel, Isis, 1848, p. 368 (East Indies). Cypsehis inquietus Kittlitz, Denlvwurd. Reise, II, 1858, p. 26 (Ualan Island, Caroline Islands, Pacific Ocean). Chars, subsp. — Similar to Collocalia leucophaea, but decidedly smaller; tarsus sparsely feathered ; bases of the loral feathers pure white ; upper parts very much more blackish, and rather more uniform, the rump not appreciably lighter than the back. Type locality. — Java. Geographical distribution. — East India Islands and western Poly- nesia, from Nias, Java, Borneo, and the Natuna Islands, northeast to the Philippine, Mariana, and Caroline Islands, east and southeast to New Guinea, Duke of York Island, Tonga (Friendly) and Loyalty Islands. The birds of a good series, consisting of specimens from Java, the Philippine and Caroline Islands, and from Nias Island, off the western coast of Sumatra, are very uniform in size as well as in the dark lower surface and glossy blackish upper parts, indicating thus that the name Hirundo vanikorensis Quoy and Gaimard,^" which was based on the bird from the Santa Cruz Islands, is synonymous with Hirundo fuci- phaga Thunberg,^^ from Java. A single specimen from Guam Island is, however, much more brownish above than any of the others ex- amined, and may represent an und escribed race. Much of the difficulty heretofore experienced in identifying this species and in segregating its various forms has arisen from confusing with it the birds with light lower surface and unfeathered tarsi, which occur in various localities, and most if not all of which are Collocalia unicolor amelis. The tail in Collocalia fuciphaga is usually quite deeply emarginate, but this character varies greatly. Some specimens in our series have the tail almost even, and there is every gradation evident between this condition and that of greatest emargination, differences apparently to be accounted for only by individual variation. Care must therefore be exercised in using the shape of the tail as an absolute character to distinguish this species. Measurements of Collocalia fuciphaga fuciphaga are as follows: »" Voy. Astrolabe, Zool., I, 1830, p. 206, pi. XII, fig. 3. " K. Vet. Akad. Nya Handl., XXXIII, 1812, p. 153, pi. 4. 1906. NATURAL SCIENCES OF PHILADELPHIA. 187 Sex. Locality. Date. Wing. Tail. Exposed Culmen. Tarsus. d" Nias Island Mar. 16, 1903 d" Mercedes, Mindanao, P. I. Dec. 19, 1903 9 Pantar, Mindanao, P. I. Aug. 13, 1903 — Java ' d Ualan Island, Caroline Is. Feb. 8,1900 — Ualan Island, Caroline Is. Feb. 16, 1900 114.5 114 108 113 114 109 54 50 48 51 54 54 4.5 5 4.5 4.5 4 4 8 10 8.5 9.5 9 9 Average 112.1 51.8 4.4 9 CoUocalia fuciphaga vestita (Lesson). Salangana vestita Lesson, I'Echo du Monde Savant, ser. 2, VIII, 1843, p. 134. CoUocalia nidifica Gray, Genera Birds, I, 1845, p. 55 (Sumatra). Chars, suhsp. — Similar to CoUocalia fuciphaga fuciphaga, but entire lower surface much paler; upper parts rather lighter and more brownish. Type locality. — Sumatra. Geographical distribution. — Sumatra, Simalur Island, and southern part of Malay Peninsula. The original CoUocalia fuciphaga was based on the bird from Java, from which this western form differs as above said. For the latter the name Salangana vestita Lesson ^^ seems to be available. This is founded on " Hirundo esculenia Lath. Gen. Syn. pi. Hab. les iles de la Malasie," ^^ which in both description and locality (Sumatra) is perti- nent to the present form. Moreover, the only Sumatran specimen examined agrees with the birds from Simalur Island, upon which latter the present separation is primarily based. The Nias Island bird, how- ever, is identical with that of Java, as already noted, and therefore must be called CoUocalia fuciphaga fuciphaga. A single adult from Tanjong Silantei, on the east coast of Johore, southern Malay Peninsula, seems to be exactly like the birds from Simalur Island in color, but is considerably larger, the wing measuring 118 mm. ; notwithstanding this it seems best referred, for the present at least, to C. f. vestita, although additional specimens may show it to be representative of a recognizable subspecies. Measurements of CoUocalia fuciphaga vestita are as follows : 12 VEcho du Monde Savant, ser. 2, VIII, 1843, p. 134. >3 Latham, Gen. Synop., Suppl. II, 1801, p. 257, pi. CXXXV. 188 PROCEEDINGS OF THE ACADEMY OP [April, Sex. Localihj. Date. Wing. Tail. Exposed Culmen. Tarsus. 6" Tanjong Silantei, E. Coast Johore Simalur Island Simalur Island Sumatra.. July 26, 1901 Dec. 7, 1901 Dec. 7, 1901 118 111 112.5 115 51 51 48 52 5 5 4.5 4 9 9 9 9.5 Average 114.1 50.5 4.6 9.1 Collocalia fuciphaga elaphra subsp. nov. Chars, subsp. — Similar to Collocalia fuciphaga fuciphaga, but de- cidedly larger; upper parts much more brownish, and less uniform, the rump being noticeably lighter than the back; lower surface decidedly paler. Geographical distribution . — Seychelles Islands; Anaml^a Islands. Description. — Type, adult, sex unknown, No. 119,779, U. S. N. M.; Mahe Island, Seychelles Islands, April 17, 1890; Dr. W. L. Abbott. Upper surface sooty brown, with a very slight greenish gloss, the pileum rather darker and more greenish, the rump much lighter brownish — but not whitish — this due partially to the lighter bases of the feathers; wings and tail darker, more blackish, with a dull bluish, greenish or purplish sheen, the wing-coverts decidedly greenish; sides of head and neck dark brown, darkest on the lores, where the feathers have pure white bases; entire lower surface rather deep brownish gray, somewhat paler posteriorly, darker on chin, the longest under tail- coverts slightly glossed with greenish ; lining of wing blackish brown. The specimens on which this race is principally established were collected by Dr. W. L. Abbott on Mahe, one of the Seychelles Islands. They differ so much from typical Collocalia fuciphaga, of which we have a considerable series from several localities, that it seems necessary to regard them as representatives of another subspecies. The lighter colored rump, which seems to be one of the best char- acters of this race, is not sufficiently decided to give the impression of a whitish band such as exists in Collocalia francica and its allies, but it is nevertheless quite different from the uniformly dark condition ob- taining on the upper surface of Collocalia fuciphaga fuciphaga. From Collocalia fuciphaga vestita, which in some respects it resembles more closely than it does true fuciphaga, it may readily be distinguished by its greater size, lighter, more brownish, and less uniform upper surface, with the rump noticeably paler than the back. This new race, how- 1906. NATURAL SCIENCES OF PHILADELPHIA. 189 ever, does not need special comparison with any further forms of the genus, even those possessing feathered tarsi. Two adults from Pulo Jimaja, Anamba Islands, are somewhat less brownish above, as well as very slightly darker below, but nevertheless agree in all details of coloration quite closely with the birds from the Seychelles Islands, and differ thus quite remarkably from Collocalia fuciphaga fuciphaga, by whose range the Anamba Islands are pretty well surrounded. In all measurements excepting that of the wing these two specimens exceed the measurements of fuciphaga and equal or even surpass those of elaphra; but the wing-quills are molting and not fully grown, and there are indications from the relative length of the primaries that the length of the wing would have become when perfect almost if not quite as great as that of elaphra. In light of pre- sent knowledge, therefore, it seems best to refer to elaphra, with the above explanation, these two examples from the Anamba Islands, even though by so doing elaphra presents an anomalous geographical distribution. Should, however, a satisfactory series show the Anamba bird to be really much smaller than that from the Seychelles Islands, and reasonably constant in its slight color differences, it ought probably to be separated subspecifically. Measurements of two adults of Collocalia fuciphaga elaphra are here given : Sex. Locality. Date. Wing. Tail. Exposed rp Culmen. ? Mahe I., Seychelles Is..... April 17, 1890 ? Mah6 I., Seychelles Is-'^J April 17, 1890 120.5 121 51 51 4.5 4 9 9.5 Averaaje 120.8 51 4.3 9.3 Collocalia brevirostris (McClelland). Hirundo brevirostris McClelland, Proc. Zool. Soc. Lond., 1839, p. 155. Chars, sp. — Similar to Collocalia fuciphaga fuciphaga, but much larger; tail less deeply emarginate; and rump noticeably lighter than the very dark back. Wing, 124-127; tail, 55-59; exposed culmen, 5.5; tarsus, 10 mm. Type locality. — Assam. Geographical distribution. — Himalaya Mountains from Dalhousie, about 76° east longitude, east through Nepal and Sikhim to Assam and Manipur. This form has been until comparatively recent years usually consid- i^Type. 190 PROCEEDINGS OF THE ACADEMY OF [April ered identical with Collocalia unicolor, from which, however, it may readily be distinguished by its feathered tarsi, darker, less brownish upper surface, with rump appreciably lighter than the back. It is most closely allied to Collocalia fuciphaga, and by Dr. Hartert is con- sidered a subspecies of this; but its large size and other characters, combined with its isolated range, so far as C. fuciphaga is concerned, quite clearly indicate its specific distinctness. Furthermore, there is, so far as is known at present, no intergradation in size between Collocalia brevirostris and any form of Collocalia fuciphaga; and the race of the latter, C. f. elaphra, which is nearest in size is most different in color. The so-called intermediate specimens from the western Himalayas, mentioned by Dr. Hartert,^^ have proved to be examples of Collocalia unicolor, a form belonging to the group without feathers on the tarsus, and have thus no bearing on the question. Therefore, until actual proof of intergradation with Collocalia fuciphaga be forthcoming, the present form should stand as a full species. Collocalia lowi (Sharpe). Cypselus lowi Sharpe, Proc. Zool. Soc. Lond., 1879, p. 333. Cypselus labuanensis , Ibis, 1879, p. 116, in text {noinen nudurn). Chars, sp. — Like Collocalia innominata, but rump without a well- defined light band. Wing, 127-134; tail, 50-57; exposed culmen, 5; tarsus, 10-12 mm. Type locality .—\^2ih\\2in Island, northern Borneo. Geographical distribution. — Northern Borneo; Palaw^an Island, Philip- pine Islands; Anamba Islands; accidental in Sumatra and on Nias Island (Hartert). This species resembles Collocalia leucopygia and Collocalia innomi- nata in possessing dark shaft lines on the lower surface, and in this it differs with them from all the other members of the subgenus. In color above and below — except for the lack of a light rump band — as w^ell as in size and proportions, it is identical with Collocalia innomi- nata. It is so much larger, and usually has the tail so much less deeply emarginate, -than Collocalia fuciphaga and all subspecies of the latter that further comparison is unnecessary. One specimen in the United States National Museum has the tail by no means square, but considerably emarginated, from which it is evident that there is con- siderable individual variation in this regard. Subgenus COLLOCALIA Gray. Chars, suhgen. — Tarsus entirely without feathers. Type. — Hirundo esculenta Linnaeus. 15 Cat. Birds Brit. Mus., XVI, 1892, p. .502. 1906. NATURAL SCIENCES OF PHILADELPHIA. 191 CoUocalia origenis sp. nov. Chars, sp. — Resembling CoUocalia ivhiteheadi, but upper parts much darker, more blackish, and more uniform, the rump not appreciably lighter than the back; under surface darker, and throat not decidedly paler than abdomen. Geographical distribution. — Mindanao, Philippine Islands. Description.— Type, adult male. No. 192,162 U. S. N. M. ; Mount Apo, 4,000 feet, Mindanao, Philippine Islands, July 4, 1904; Dr. E. A. Mearns. Upper siu-face uniform brownish black, the rump not lighter, but wings and the distinctly forked tail more brownish, their feathers paler along the inner margins; entire ventral sm'face uniform smoky hair brown, the throat not paler; sides of head and neck darker brown than the under parts; a blackish spot in front of the eye; lining of wing blackish brown. This new species was discovered by Dr. Mearns during his recent trip to the island of Mindanao, the four adults obtained having been brought by natives who had found them in a cave on Mount Apo ; and we are indebted to him for the privilege of describing them. In its large size, deeply emarginate tail, and lack of tarsal feathering CoUocalia origenis agrees with CoUocalia whiteheadi, but is readily dis- tinguishable by the dark colors, particularly on the upper parts which are even more blackish than in CoUocalia fiiciphaga fuciphaga from the Philippines. There is a slight sexual difference in this species, appa- rent in our series of two adult males and two adult females, but it is possibly not constantly distinctive: the females are somewhat more brownish above, and slightly paler below. A very young bird, with wings and tail but little grown, taken by Dr. Mearns, July 11, 1904, at Todaya, altitude 4,000 feet, on ]\Iount Apo, is fully as l^lackish on the upper parts as the adults, and somewhat darker, decidedly more grayish on the ventral surface. Measurements of the adults are as follows!: Sex. Locality. Date. ? Mt. Apo, Mindanao, P. I. July 4, 1904 129 138 134 130 53 60 51 54 6 5.5 5.5 5.5 13 14 12.5 14 Average. 132.8 54.5 5. 13.4 Type. 192 PROCEEDINGS OF THE ACADEMY OF [April, Collocalia whitehead! Grant. Collocalia whiteheadi Grant, Ibis, 1895, p. 459. Chars, sp. — Similar in color to Collocalia loici, but dark shaft streaks on lower parts less distinct ; averaging slightly larger, with unf eathered tarsi, and usually more deeply emarginate tail. Wing, 127-140; tail, 50-57 ; exposed culmen, 5; tarsus, 11.5-13 mm. Type locality. — Monte Data, highlands of Lepanto, northern Luzon, Philippine Islands. Geographical distribution. — Islands of Luzon and Palawan, Philippine Islands. This very distinct species may be easily distinguished from all the others with unfeathered tarsi, excepting Collocalia origenis, by its ^-ery large size, and from that form by the well-marked color characters already detailed. Specimens from Palawan are smaller than those from Luzon, touching the minimum of measurements above given, and also appear to be more brownish on the upper parts — differences which, should they prove reasonably constant, would entitle the Palawan bird to subspecific rank. Collocalia unicolor unicolor (Jerdon). Hirundo unicolor Jerdon, Madras Journ., XI, 1840, p. 238. Cypselus concolor Blyth, Journ. As. Soc. Bengal, XI, pt. 2, 1842, p. 886 (nom. nov. pro Hirundo unicolor Jerdon). Chars, subsp. — Resembling Collocalia ichiteheadi, but much smaller; more brownish and more uniform above, the rump not appreciably paler than the back, the pileum but little if any darker; lower surface rather more brownish and more uniform, the throat usually not lighter than the abdomen; lining of wing lighter. Wing, 112-120.5; tail, 50-56; exposed culmen, 4-4.5; tarsus, 9-10 ,mm. Type locality. — Coonoor Pass, Nilghiri Hills, southern India. Geographical cUstrihution. — Ceylon; and the western coast region of southern India, north to Vengmia; western Himalayas. Although this species has commonly been either synonymized with Collocalia fuciphaga or treated as a subspecies of it, the entirelj^ un- feathered tarsi are a character fully sufficient for specific recognition; and furthermore the upper surface is more brownish than in even the least blackish forms of Colloc*lia fuciphaga. In fact Collocalia uni- color is really more closely allied to Collocalia ichiteheadi than to Col- localia fuciphaga! Dr. Hartert has given" as a reason for considering C. unicolor a subspecies of C. fuciphaga that some Celebes birds similar to C. unicolor in color, which he refers to C. fuciphaga, have no tarsal 1 Ibis, 1896, p. 369 1906.] NATURAL SCIENCES OF PHILADELPHIA. 193 feathers ; but these have hardly bearing on the question, for they without much doubt belong to Collocalia unicolor amelis. The present form is said to occur in the western Himalaya Mountains, but careful compari- son of a satisfactory series from this locality with typical examples from southern India would possibly show subspecific differences. Im- mature birds of Collocalia unicolor are rather darker, more sooty than adults, both above and below, Collocalia unicolor amelis subsp. nov. Chars, subsp. — Similar to Collocalia unicolor unicolor, but less brown- ish and less uniform above, the rump usually appreciably paler than the back, the pileiun decidedly darker; lower surface rather less brown- ish and less uniform, the throat usually decidedly lighter than the abdomen; lining of wing darker; tail usually rather less deeply emarginate. Geographical distribution. — Philippine Islands; Island of Guam; ? Celebes; ? Louisiade Archipelago. Description.— Type, adult male, No. 189,931 U. S. N. M. ; Irisan, Ben- guet, Luzon, Philippine Islands, May 19, 1903; R. C. McGregor and A. Celestino. Upper parts dark sooty brown with a greenish tinge and but little gloss, the head darker, the rump slightly paler; wings and tail darker than the back — blackish brown with a greenish sheen, the inner webs of the wing-quills dull and paler brown; lower surface grayish brown, darkest on the lower tail-coverts, palest on the throat and jugulum; lining of wing blackish brown. "Iris brown, bill and claws black; legs dark reddish brown." Several specimens of this new form, identified as Collocalia white- headi, were some time since received from the Philippine Museum by the U. S. National Museum. Notwithstanding the similarity of pro- portions (including the forking of the tail), the very great difference in size renders it easily distinguishable from C. whiteheadi on even superficial examination. From Collocalia francica and other light- rumped species Collocalia unicolor amelis is at once to be separated by its lack of the whitish band on the rump. Its nearest relative is, of course, Collocalia unicolor unicolor of southern India, from which, though similar in size, it differs as above mentioned ; but com- parison of a series of each of these forms shows none^of the characters to be quite constant, hence amelis must rank as a subspecies. The tarsi in C. u. amelis are entirely devoid of feathers, therefore no special comparison is really necessary with Collocalia fuciphaga and its allies, except for specimens of fuciphaga that have accidentally lost the tarsal 13 194 PROCEEDINGS OF THE ACADEMY OF [April, feathering. In color C. u. amelis differs from C. fuciphaga fuciphaga by reason of paler ventral surface, particularly the throat, and more brownish upper parts, with lighter rump. It is paler, more brownish above than C. f. vestita, with the rump noticeably lighter than the back. Compared with C. /. elaphra it has the upper parts less brown- ish, the crown particularly more blackish, the rump less different from the back; lower surface less uniform, the posterior portion darker; and size somewhat less. The small Celebes birds without tarsal feathering, mentioned by Mr. Grant,^^ belong probably to this subspecies, for they are evidently not Collocalia fuciphaga. Also the birds from St. Aigan Island, in the Louisiade Archipelago, said by Dr. Hartert ^^ to be light below, to have unfeathered tarsi, and to build nests different from C. fuciphaga, are probably to be referred to C. u. amelis, though possibly subspecifically distinct. Aside from the above, Collocalia u. amelis is not at present certainly known except from the Island of Guam, where taken by Dr. E. A. Mearns, July 20, 1905, and from the Philippine archipelago, where it has been obtained on the Islands of Luzon, Verde, Cagayancillo, Sibuyan, Panay, Mindoro, Mindanao, and Palawan. It may, however, be quite generally distributed over the East India Islands, and may have been recorded as Collocalia fuciphaga from various localities, since hitherto it seems to have been confused with that species, because its unfeathered tarsi were not considered significant. Measurements of part of the U. S. National Museum series of Collo- calia unicolor amelis are as follows: Sex. Locality. Date. Wing. Tail. Culmen. Tar sits. 6" 9 Irisan, Benguet, Luzon, [P.I.'» Mindanao, P. I Palawan, P.I May 19, 1903 June 3, 1903 Nov. 28, 1887 Sept. 5, 1887 June 3, 1903 113 113 111.5 120 118 118 117 123 109.5 116.5 53 50 50 49.5 46 49.5 50 50 50 50 5 5 4.5 5 5 5 5 5 4.5 4 9.5 10 . 9.5 9.5 10 9.5 9 Irisan, Benguet, Luzon, . . tP. I. 10 10.5 10 10 Average 116 49.8 4.8 9.9 '8 Ibis, 1895, p. 462. '" Novit. Zool, VI, 1899, p. 211. ^» Type. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 195 CoUocalia leucophaea (Peale). Macropteryx leiicophaeus Feale, United States Explor. Exped., VIII, 1848, p. 178, pi. XLIX, fig. 3. CoUocalia cinerea Cassin, United States Explor. Exped., Mamm. and Ornith., 1858, p. 183, pi. XII, fig. 4 (Tahiti Island, Society Islands) {nee Gmelin). Chars, sp. — Similar to CoUocalia unicolor unicolor, but larger; upper parts, including wings and tail, more brownish, with less metallic sheen, and less uniform, the pileum somewhat darker, the rump paler, than the back; feathers of the lores with light brownish instead of white bases. Wing, 120-127; tail, 56-59; exposed culmen, 4.5-5; tarsus, 9.5-10 mm. Type locality. — Tahiti Island, Society Islands. Geographical distribution. — Tahiti Island, Society Islands, Pacific Ocean. This very distinct species is superficially somewhat like CoUocalia fuciphaga clophra from the Seychelles, but lacks the tarsal feathers; is larger, more brownish above, with less metallic gloss; much darker, duller, more uniform on the lower surface; and has brownish in place of white bases to the loral feathers. It is of course still more different from CoUocalia fuciphaga fuciphaga, being much larger, decidedly more brownish, with scarcely any metallic gloss above, and having neither feathers on the tarsi nor white bases to the feathers of the lores. On the Island of Tahiti alone is this species positively known to occur, since all the specimens from the ^larquesas Islands now at hand prove to belong to CoUocalia ocista. All previous records of this species from the Marcjuesas Islands therefore need verification, though of course its occurrence there is by no means improbable. Even if the name given to this species by Cassin — CoUocalia cinerea ^^ — were not, as has already been show^n by Dr. Hartert," a mistaken identification of his specimen with the Hirundo cinerea of Gmelin,^^ which is a swallow — Atticora cinerea — it would still not be the tenable name for the present species, since it is ten years posterior to the Ma- cropteryx leucophaeus of Peale,^^ and was based on the same type speci- men which is yet in the U. S. National Museum. CoUocalia thespesia sp. nov. Chars, sp. — Similar to CoUocalia francica germani, but the wing slightly, the tail very much longer; upper parts lighter, more brownish; " United States Explor. Exped., Mamm. and Ornith., 1858, p. 183, pi. XII, fig. 4. " Cat. Birds. Brit. Mus., XVI, 1892, p. 502. ^^ Syst. Nat., I, ii, 1788, p. 1026. 2^ United States Explor. Exped., VIII, 1848, p. 178, pi. XLIX, fig. 3. 196 PROCEEDINGS OF THE ACADEMY OF [April, bases of the loral feathers pale brown instead of pure white; sides of head and neck, together with entire lower svu'face, excepting the anal region, much darker and more uniform. Geographical distribution. — Tahiti Island, Society Islands. Description. — Type, female adult; Tahiti Island, Society Islands, November 14, 1899; C. H. Townsend. Pileum clove brown; cervix, back, upper tail-coverts, and part of rump rather lighter and slightly more rufescent with scarcely any metallic gloss; rump with a very light brown band, slightly whitish along the edges of the feathers, and similar to that of CoUocalia francica francica, though apparently not quite so broad or so pale, the feathers with somewhat though not very conspicuously defined deeper brown shafts; wings and tail darker than upper parts, and somewhat blackish with a slight purplish or bluish sheen, but the basal portion of the rectrices and the inner edge of the wing-quills decidedly more brownish — much like the back; median and lesser wing-coverts with the tertials also rather lighter, more brownish than the quills, and somewhat glossed with greenish; sides of head clove brown almost as dark as the crown, the feathers of the lores with lighter brown bases ; sides of neck brown like the cervix ; lower surface the same, but lighter, becoming still a little paler on the abdomen, considerably so and appreciably more rufescent on the crissum; lining of wing clove brown. Wing, 122; tail, 58; exposed culmen, 5; tarsus, 10 mm. The single specimen of this new and remarkable species was ob- tained by Mr. Townsend during the cruise of the "Albatross" already mentioned. It differs very strongly from CoUocalia francica spodio- pygia, the light-rumped form from the Samoan Islands, in its larger size, lighter, much more brownish upper parts, rather more narrow rump band, darker, more uniform ventral surface, and especially, as from all other species of the entire genus excepting C. leucophaea and C. ocista, by the brownish instead of pure white bases of the loral feathers. From CoUocalia leucophaea it may of course be easily dis- tinguished by the pale band across the rump, as well as by somewhat less brownish upper surface and decidedly paler anal region. From CoUocalia ocista it differs in lack of tarsal feathers, in more brownish upper surface, pale rump band, and lighter anal region. The only published name that can by any possibility apply to this species is Hirundo peruviana Forster,-^ based on a specimen said to have been obtained on Tahiti. But unless the description of this bird is very erroneous it cannot refer to thespesia, as the following quotation ^^Descript. Anim., 1844, p. 240. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 197 will show :" Corpus totum supra cum remigibus, rectricibusque nigro- nitens. Uropygium, gula, pectus, abdomen, crissum cinereo-fuliginosa."-^ The upper surface in thespesia is by no means "nigro-nitens," being dark brown, scarcely more blackish than the same part in Collocalia leucophaea; while the rump is not of the same color as the dark gray ventral surface, but is very much paler, even whitish. The description of peruviana very much better fits Collocalia francica spodiopygia, to which it should probably be referred. Collooalia francica francica (Gmelin). Hirundo francica Gmelin, Syst. Nat., I, ii, 1788, p. 1017. Hirundo francisca Vieillot, Nouv. Diet. d'Hist. Nat., XIV, 1817, p. 525 (Mauritius). Hirundo francicB Lesson, Traite d'Ornith., 1831, p. 270 (Mauritius). Chars, sp. — Like Collocalia unicolor amelis, but somewhat smaller; rather more brownish above, with a distinct broad whitish or brownish white band across the rump; paler below, especially on the abdomen and crissum; lining of wing paler, more brownish. Wing, 112-114; tail, 51-52; exposed culmen, 4-5; tarsus, 8.5-10 mm. Type locality. — Mauritius Island. Geographical distribution. — Islands of INIauritius and Bourbon; ? Ceylon. This form bears some resemblance to Collocalia innominata, but is of course much smaller ; more brownish above, with a paler rump band which has not conspicuously darker shafts ; darker below, particularly posteriorly, without noticeably darker shaft lines; and lacks feathers on the tarsus. It seems to be confined to the Islands of Mauritius and Bourbon, for although several times accredited to Madagascar its occurrence there remains to be confirmed; and all the East Indian records belong to other subspecies. Collocalia francica townsendi subsp. nov. Chars, subsp. — Similar to Collocalia francica francica, but wing longer; upper surface darker, more blackish, and more glossed with greenish, the rump band on the contrary more whitish; low^er surface darker. Geographical distribution. — Tonga Islands. Description. — Type, adult female; Eua Island, Tonga (Friendly) Islands, November 28, 1899; C. H. Townsend. Upper parts, including wings and tail, brownish black with a greenish metallic sheen, the inner webs of the rectrices, especially on their basal portions, and particularly the inner webs of the wing-quills, more brownish; a conspicuous 28 /bid., p. 241. 198 PROCEEDINGS OF THE ACADEMY OF [April, brownish white — almost pure white — band on the rump, its feathers with slightly though not very noticeably dusky shafts; sides of head clove brown, the lores darker, with white bases ; sides of neck somewhat lighter brown; lower surface brownish gray, darkest on chin and the longest lower tail-coverts, decidedly paler, even almost whitish, on lower abdomen and anal region; lining of wing clove brown. From Collocalia francica spodiopygia, to which this form has hereto- fore been referred, it differs in considerably more greenish glossy upper surface, much more whitish rump, and paler posterior lower parts. It has the most whitish rump of any subspecies of C. francica, and seems furthermore to be one of the best marked of these. It is another of the novelties brought back by Mr. C. H. Townsend from his Pacific cruise in 1899-1900, and was obtained in only the Tonga group — on Eua, Vavau and Nine Islands. Measurements of these specimens are as follows: Sex. Locality. Date. Wing. ^-■^- ^^X7n: ^«^^^'«- (^ Vavau I., Tonga Islands. JDec. 4,1899; 116 d^ Niue I., " " !Nov. 25, 1899i 116.5 9 Eua I., " "" Nov. 28, 1899 117 52 54 50 4 9 4 9 4 9 Average 116.5 52 4 9 Collocalia francica terraereginae (Ramsay). Cypselus terrce-regince Ramsay, Proc. Zool. Soc. Lond., 1874, p. 601. Collocalia infuscata Salvadori, Atti R. Accad. Sci. Torino, XV, 1880, p. 348 (Ternate Island, Molucca Islands). Chars, subsp. — Similar to Collocalia francica francica, but somewhat smaller; upper surface darker; posterior lower parts more deeply colored, imparting thus a more uniform appearance to the ventral surface; feathers of the light rump band with more conspicuously blackish shaft lines. Wing, 110-111; tail, 48-53; exposed culmen, 5; tarsus, 8.7-10 mm. Type locality. — Cardwell, Rockingham Bay, Queensland, Australia. Geographical distribution. — Northern Queensland and southern New Guinea to Jampea Island, Ternate Island and probably also other islands of the Molucca group. This race may be distinguished from Collocalia francica spodiopygia chiefly by its much smaller size, more conspicuously blackish shaft " Type. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 199 lines on the feathers of the light rump band, and by more glossy upper surface. From Collocalia francica townsendi it differs in shorter wing, less whitish rump band with conspicuous blackish shaft lines, and darker abdomen. The subspecific separation of the present form is made principally on the basis of the bird from Ternate Island, described by Count Salvador! as Collocalia injuscata^^ which is certainly different from both francica and spodiopygia, although considered by Dr. Hartert as inseparable from the latter.^^ Although no specimens from northern Queensland, the Cypselus terraereginae of Ramsay j^^have been examined in the present connection, the characters of size and coloration assigned them agree apparently best with this form, for which therefore terrae- reginae by priority becomes the proper subspecific name. Should, however, the birds from Queensland, w^hich are undoubtedly not to be referred to spodiopygia, and probably not to francica, ultimately prove to be different from those of the Molucca Islands, the latter must stand as Collocalia francica infuscata Salvador!. Collocalia francica spodiopygia (Feale). Hirundo peruviana Forster, Descript. Anim., 1844, p. 240 (Tahiti Island, Society Islands) {nee Gmelin). Macropteryx spodiopygius Peale, United States Explor. Exped., VIII, 1848, p. 176, pi. XLIX, fig. 2 (Upolu Island, Samoa Islands). Herse forsteri Hartlaub, Journ. f. Ornith., 1854, p. 169 (nom. nov. pro Hirundo peruviana Forster). Chars, subsp. — Like Collocalia francica francica, but wing longer; upper parts more blackish and less glossy; also lower surface, partic- ularly the abdomen, darker; rump band less whitish. Wing, 116-117; tail, 48-54; exposed culmen, 4-5; tarsus, 8-9 mm. Type locality. — Upolu Island, Samoa Islands. Geographical distribution. — Solomon, Samoan, and Fiji Islands; Tahiti Island. This dark form from western Polynesia certainly deserves recogni- tion on good average characters as distinguished from Collocalia fran- cica francica with which it has heretofore been synonymized. It is, indeed, both in color and size, apparently nearer Collocalia francica inexpectata, from which it differs chiefly in its smaller size, less glossy upper surface, lighter rump band, and obsolescence, usually absence, of dark shaft lines on rump and under parts. Peale's type of 28 Atti R. Accad. Sci. Torino, XV, 1880, p. 348. " Cat. Birds Brit. Mus., XVI, 1892, p. 504. ^'>Proc. Zool. Sac. Lond., 1874, p. 601. 2C0 PROCEEDINGS OF THE ACADEMY OF [April, spodiopygia is still in the United States National Museum, and fairly well preserved. The name Hirundo peruviana Forster,^^ based on specimens from Tahiti, seems recently to have been pretty generally ignored or over- looked; but that it in all probability applies to this form, as some authors have indicated, notwithstanding that there is no other record from Tahiti, may easily be seen by examination of the excellent orig- inal description.^- It is, however, unavailable for use in the present case, since it is preoccupied by Hirundo peruviana Gmelin,^^ which is certainly some other bird. The Herse forsteri of Hartlaub ^* is, as he states, a renaming of Forster's Hirundo peruviana, but is antedated by Macropteryx spodiopygius Peale.^ CoUocalia francioa inexpeotata Hume. Collocalia inexpedata Hume, Stray Feathers, I, 1873, p. 296. Chars, subsp. — Similar to CoUocalia francica francica, but larger; upper surface decidedly darker, less brownish, and with more metallic greenish gloss; wings and tail with more metallic bluish and purplish tinge; rump much darker, making its light band less well defined, the feathers with much more distinct blackish shaft lines; lower parts, particularly abdomen and crissum, darker, more uniform, and with more evident dusky shaft lines. Wing, 114-124; tail, 50-54; exposed culmen, 4.5-5; tarsus, 9.5-10.5 mm. Type locality. — Button Island, Andaman Islands. Geographical distribution. — Southern Andaman Islands; Nicobar Islands; Pulo Tioman; and both coasts of the southern part of the Malay Peninsula; ? accidental in Amherst, Tenasserim (Hartert). This race may be distinguished from Collocalia f. townsendi by some- what larger average size, much darker rump band with blackish shaft lines more conspicuous, and decidedly more deeply colored posterior lower parts, with evident darker shaft lines; from Collocalia f. terrae- reg'mae by much greater size, darker rump band, and more noticeable dark shaft lines on abdomen and crissum. It differs from Collocalia francica spodiopygia in its somewhat greater size, more glossy upper surface, more bluish and purplish metallic sheen on wings and tail, presence of appreciably dusky shaft lines on the posterior lower parts, ^^ Descript. Anim., 1844, p. 240. 32 Ihid. 33 Syst. Nat., I, ii, 1788, p. 1025. 3* Journ. f. Ornith., 1854, p. 169. 35 United States Explor. Exped., VIII, 1848, p. 176, pi. XLIX, fig. 2. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 201 and in duller, less well-defined rump band, the feathers of which have much more conspicuous dark shaft lines. Birds from the southern part of the Malay Peninsula — both eastern and western sides— as well as from Pulo Tioman, off the east coast, average larger than those from the Andaman and Nicobar Islands; but this is apparently too slight and inconstant, in the absence of any accompanying color difference, to warrant recognition by name. In C. f. inexpectata as in other forms there is considerable individual color variation, evident in this case chiefly on the lower parts and the light rump band. The single specimen of inexpectata recorded by Dr. Har- tert ^^ from Amherst, Tenasserim, is possibly rather to be referred to Collocalia f. germani, for it is more probably an unusually dark example of the latter, whose home is near by, than a stray from the far-off south- ern Andaman Islands. Collocalia franoica germani Oustalet. Collocalia germani Oustalet, Bull. Soc. Philom. Paris, 1876, pp. 1-3. Collocalia francica, subsp. /3. merguiensis Hartert, Cat. Birds Brit. Mus., XVI, 1892, p. 506 (Mergui, Tenasserim). Chars, subsp. — Similar to Collocalia francica francica, but larger; upper surface more glossed with metallic greenish; light rump band with more evident blackish shaft lines; posterior lower surface darker, and with more conspicuous dark shaft lines. Wing, 115-121.5 (average, 118.3); tail, 49-52.5 (average, 51.3); exposed culmen, 4.5-5 (average, 4.9); tarsus, 9-11 (average, 10) mm.^^ Type locality. — Condore Island, China Sea, off the southeastern coast of Cochin China. Geographical distribution. — Mergui Archipelago; Tenasserim; Malay Peninsula south to Lower Siam; Condore Island, Cochin China; Philip- pine Islands. Like many of the forms of Collocalia, this race is, on account of individual variation, difficult to determine satisfactorily without a sufficient series, but it is nevertheless worthy of recognition. It is distinguishable from Collocalia f. spodiopygia by its lighter, more greenish glossy upper surface, more metallic bluish and purplish sheen on wings and tail; well-defined blackish shaft streaks on the feathers of the light rump band ; paler lower parts, with dark shaft streaks on abdomen; and somewhat larger size. From Collocalia f. townsendi it may be separated by rather greater size; lighter, more brownish upper 3« Cat. Birds Brit. Mus., XVI, 1892, p. 505; Tierreich, I, 1897, p. 69. ^' Ten specimens. 202 PROCEEDINGS OF THE ACADEMY OF [April, parts; less whitish rump band and darker abdomen, both with more distinct dusky or blackish shaft lines. From Collocalia f. terraereginae it may be distinguished by its much larger size, lighter, more brownish upper surface, and more conspicuous dark shaft lines on the posterior lower parts. From Collocalia f. inexpectata it differs in its decidedly paler under surface; lighter, more brownish upper parts; and more prominent, less brownish, light rump band ;but occasional intermediate individuals occur that are difficult to distinguish. The dark shaft lines of the lower surface vary much in different individuals, being apparently most evident in immature birds. The measurements of this form given by Hartert ^^ are rather too large, as may be seen by reference to the above averages, but were probably taken from a limited number of specimens. A series of birds from Trong, Lower Siam, and another from the Philippine archipelago seem to be identical, and together are certainly all referable to this race, although a few of those from Cagayan Sulu, in the southwestern Philippines, are more blackish above and show in this, but not in size, some vergence toward Collocalia francica terraereginae. There are examples in the United States National Museum from the following islands of the Philippines: Panay; Cagayan Sulu; and Cagayancillo, in the Cagayanes group. It has also been recorded as Collocalia francica from Cuyo, Negros and Calamianes. All Philippine records of both francica and inexpectata refer of course to germani. The identity of Collocalia germani Oustalet and Collocalia francica merguiensis Hartert may be considered as conclusively established, for Dr. Hartert assures us ^^ that he has compared the types. That germani is rightly considered a subspecies of Collocalia francica seems to be equally certain, for Dr. R. Bowdler Sharpe, who at our request carefully examined the series of merguiensis in the British Museum, informs us ^*' that there is not the slightest indication of tarsal feather- ing on any of the specimens. Collocalia troglodytes Gray. Collocalia troglodytes Gray, Gen. Birds, I, 1845, p. 55, pi. 19. Chars, sp. — Very much smaller than Collocalia francica francica; upper parts, including wings and tail, black with a metallic greenish or bluish gloss; a pure white, well-defined rump band, the feathers of which have blackish shafts and most of them dark tips; abdomen 38 Cat. Birds Brit. Mus., XVI, 1892, p. 506; Tierreich, I, 1897, p. 69. 3» In litt. « In litt. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 203 white or whitish in contrast to most of the remaining more or less deeply brownish lower parts; under tail-coverts like the back; lining of the wing rather more brownish. Wing, 86-96; tail, 38-42; exposed culmen, 3.5; tarsus, 9-9.5 mm. Type locality. — Philippine Islands. Geographical distribution. — Philippine Islafids. This very distinct species differs markedly from all the preceding forms with unfeathered tarsi in its small size, deeply blackish upper surface, white abdomen, and peculiar white rump band. It is super- ficially very similar to Collocalia leucopygia, but aside from the lack of tarsal feathers is easily distinguished from that species by its more greenish upper surface, broad blackish tips to the white feathers of the rump band ; and more extensively blackish lower tail-coverts. There is considerable variation in the wing-length of C. troglodytes, but this does not seem to be correlated with different geographical areas. The species has been obtained on the following islands of the Philip- pine archipelago: Mindanao, Palawan, Siquijor, Cebu, Masbate, Negros, Guimaras, Panay, Romblon, Sibuyan, Mindoro, Luzon, Marinduque, Samar, Leyte, and Ticao. Collocalia uropypialis Gray. Collocalia uropygialis Gray, Ann. and Mag. Nat. Hist., ser. 3, XVII, 1S66> p. 123. Chars, sp. — Similar to Collocalia troglodytes, but rather larger; feath- ers of the white band on the rump without blackish shafts or tips ; most of the rectrices usually with white spots on the basal portion of inner webs; some of the under wing-coverts and lower tail-coverts with whitish tips or margins. Wing, 89-99; tail, 38-44; tarsus, 7 mm. Type locality. — Aneiteum Island, New Hebrides Islands. Geographical distribution. — New Caledonia and New Hebrides Islands. Readily distinguished from all the other light-rumped forms of the genus by the white spots on the tail-feathers; with the remaining species it needs no special comparison. Collocalia marginata Salvador!. Collocalia marginata Salvadori, Atti R. Acad. Sci. Torino, XVII, 1882, p. 448. Collocalia cebuensis Kutter, Journ. f. Ornith., 1882, p. 171 (Cebu Island, Philippine Islands). Chars, sp. — Somewhat like Collocalia uropygialis, but larger; rather more greenish above; rump without a white band, but its feathers with conspicuous white edgings; rectrices never with white spots. 204 PROCEEDINGS OF THE ACADEMY OF [April, Wing, 103-107; tail, 41-43; exposed ciilmen, 3.5-4; tarsus, 7.5-9 mm. Ty-pe locality.— Cehu Island, Philippine Islands. Geographical distribution. — Philippine Islands. This interesting bird, supposed until recently to be very rare, bids fair to be found distributed pretty generally throughout the Philip- pine archipelago. Up to the present time it has been ascertained to occur on the following islands: Cebu, Masjjate, Luzon (McGregor), Mind or o, Calayan, and Sibuyan. It scarcely needs close comparison with any other member of the genus except Collocalia linchi, and to distinguish it from this there should be no difficulty. The white margins of the brownish gray feath- ers of the throat and breast are usually broad and conspicuous, but occasionally in even unworn specimens are almost obsolete. Collocalia linchi linchi Horsfield and Moore. 1 Hemiprocne fucirora Streubel, Lsis, 1848, p. 369 (East Indies). Collocalia linchi Horsfield and Moore, Cat. Birds Mus. East Ind. Comp., I, 1854, p. 100 (Java). Chars, subsp. — Very much hke Collocalia marginata, but feathers of the rump without white margins ; upper surface usually more brightly colored; throat and breast generally with less conspicuous white edgings. Type locality. — Java. Geographical distribution. — Java ; Lombok ; Kangean Islands ; Suma- tra, including the islands along its western coast; Singapore, and the southern part of the Malay Peninsula; Borneo(?). This species is shown by the considerable amount of material from various parts of its range now in the United States National Museum to be divisible into several recognizable races which are diagnosed below. Aside from the evident geographical variation there are some differences that seem to be due to wear of the plumage and to obtain alike in all the forms. The color of the upper parts changes little through sex, age, or season, although there is a normal but not very considerable amount of individual variation. The deep gray of the anterior lower parts is decidedly more brownish when the feathers become old ; and the white or whitish margins that in fresh plumage are often very conspicuous on throat, breast, and under wing-coverts sometimes almost entirely disappear through abrasion. Likewise the broadly w^hite edgings of sides, flanks, and lower breast wear off and ultimately allow much of the slate gray bases of the feathers to show, in this way decidedly darkening the general appearance of the parts. This form of Collocalia linchi, as compared with the other sub- 1906.] NATURAL SCIENCES OF PHILADELPHIA. 205 species, is characterized by large size, and by very greenish upper parts, inchiding wings and tail, with but slight tinge and very little or no admixture of bluish. The United States National ^luseum possesses a fair series from Sikakap Strait, North Pagi Island, off the western coast of Sumatra ; and a specimen from Singapore is also identical. Borneo is doubtfully included in the range of this race, as we have examined no specimens from that island, and its bird may be like that of either the Natuna Islands or even the Philippines, or may prove to be an endemic form. Measurements of Collocalia linchi linchi are as follows : Sex. Locality. Date. Wing. ^-^- fxr/rf Tarms. ? ? ? Sikakap Str., No. Pagi I. Singapore Nov. 12, 1902 Aug. 28, 1879 105 103 105.5 104.5 103.5 105 44 ! 4.5 43 1 4.5 42 4.5 42 4.5 42.5 4.5 42 4.5 8.5 9 9 8.5 8.5 9 Average 104.4 42.6 4.5 8 8 CoUooalia linchi cyanoptila subsp. nov. Chars, subsp. — Similar to Collocalia linchi linchi, with which it agrees in size, but upper surface of head and body much more bluish in shade, and with considerable admixture of bluish purple; wings, and particularly the tail, decidedly more bluish. Geographical distribution. — Natuna Islands and Linga Island. Description.— Type, adult female. No. 174,688, U. S. N. M.; Bung- uran Island, Natuna Islands, July 1, 1900; Dr. W. L. Abbott. Upper parts metallic dark bluish green, much mixed with bluish purple, most noticeably so on posterior portion; tail bluish green decidedly tinged with purplish; wings sepia brown, dull metallic purplish on most of their exposed portions, the coverts more brightly colored, and mixed with dark green; sides of head and neck, throat, breast, flanks, and sides of body brownish slate, the feathers of the lores with cottony white bases, those of throat and breast with inconspicuous pale gray or whitish tips, those of lower breast, sides, and flanks broadly margined and terminated with white; median portion of abdomen white with some narrow slaty shaft streaks; lower tail-coverts dark bluish metallic green, the longer ones immaculate, the shorter ones margined with white; lining of wing dull dark metallic green. 206 PROCEEDINGS OF THE ACADEMY OF [April, This new race is quite different from true Collocalia linchi, the only form of the species with which it agrees in size, and it may readily be distinguished on even superficial comparison by the difference in the color of the upper parts. The best and most uniform character is the always much more bluish or purplish color of the upper surface of the tail. A single adult male from Linga Island, southeast of Singapore, is intermediate between cyanoptila and linchi; but there is so much purplish and bluish in the color of the upper parts, and the tail is so nearly like that of cyanoptila, that the specimen must apparently be referred to the latter, although this island is far from the Natuna group, and in a location where linchi is the form we should naturally expect to find. Further examples from Linga may, however, show that the average characters of the bird found there will place it with linchi. Measurements of the adults of Collocalia linchi cyanoptila are as follows : Sex. Locality. Date. Wing. Tail. Exposed Culmen. Tarsus. Bunguran I., Natuna Is." July 1,1900: 107 " j " i 102 July 11, 1900l 108 Linga Island 'July 10, 1899 101.5 41 41 43 40 4.5 4.5 4.5 4 7.5 8.5 8 7.5 Average 104.6 i 41.3 4.4 7.9 Collocalia linchi affinis Beavan. Collocalia affinis Beavan, Ibis, 1867, p. 318 (ex Tytler MS.). Chars, suhsp. — Similar to Collocalia linchi cyanoptila, l^ut very much smaller. Type locality. — Port Blair, South Andaman Island, Andaman Islands. GeograpJiical distribution. — Andaman and Nicobar Islands. This form, long ago described from the southern Andaman Islands by Beavan, has hitherto usually been considered the same as linchi. The satisfactory series obtained by Dr. W. L. Abbott in the Nicobar Islands, however, now conclusively establishes its claim to recognition. It differs from true linchi in its much smaller size, and the decidedly bluish or purplish shade of the upper surface, particularly the tail- coverts and tail. " Type. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 207 Measurements of Collocalia linchi affinis are as below : Sex. Locality. Date. Wing. Tail. Exposed Culmen. Tarsus. 9 Little NicobarI.,Nicobar [Is. Mar. 2, 1901 98 40 4 8.5 9 tl U X 98 41.5 4 8.5 $ Feb. 26, 1901 99 39 4 8 9 " " " 100 41 4 8 9 jFeb. 27, 1901 101 39.5 4 8 9 " i 99 38 4 8 Average 99.2 39.8 4 8.2 Collocalia linchi elachyptera subsp. nov. Chars, subsp. — Similar to Collocalia linchi affinis, but more greenish on all the upper parts, particularly on back, wings, and tail. Geographical distribution. — Islands of the Mergui Archipelago. Description.— Type, adult male. No. 173,028, U. S. N. M.; Bentinck Island, Mergui Archipelago, March 9, 1900; Dr. W. L. Abbott. Upper parts deep metallic green, with an appreciable bluish purple tinge, particularly on the upper tail-coverts ; tail the same, but more bluish ; wings fuscous, the exposed portion of quills and primary coverts with a dull metallic purplish sheen, the other superior coverts like the back; lores dark brown, the feathers with pure white bases; rest of sides of head and neck, with chin, throat, and breast, dull brownish slate color, the feathers of the throat and upper breast with pale grayish or whitish tips; lower breast, sides, and flanks brownish slate with a slight metallic greenish sheen, the feathers all broadly margined with white which much obscures the darker color ; middle of abdomen white with shaft markings of dusky; exposed portions of lower tail-coverts dark metallic greenish, the shortest brownish, all but the longest broadly bordered with white; lining of wing dark brown, glossed with metallic green. "Bill black; iris dark brown; feet dark fleshy brown." This race is of course intermediate between Collocalia linchi affinis, from the Andaman Islands, and Collocalia I. isonota,'"' from the Philip- pines; but the differences that characterize it are readily appreciable in a series, as well as in a large proportion of individual speci- mens, and seem, particularly when its isolated range is taken into account, quite sufficient for its recognition by name. It is readily dis- tinguishable from true linchi by reason of its much inferior dimensions « Postea. 208 PROCEEDINGS OF THE ACADEMY OF [April, and more bluish or purplish upper parts, especially wings and tail; from cyanoptila by its small size and more greenish upper surface, including wings and tail. So far as known it does not occur outside the Mergui Archipelago, where it has been taken on only Bentinck Island; but it undoubtedly is to be found on other islands of the group, while there would seem to be no improbability of its occasional if not regular occvirrence on the adjacent mainland. Measurements are subjoined: Sex} Locality. Date. Wing. Tail. Exposed Culmen. Tarsus. j\ Bentinck I., Mergui Arch. Mar. 10, 1900 Mar. 9, 1900 101 98.5 100 100.5 97 41 39 42.5 41 39 4 4.5 4.5 4.5 4.5 8.5 8 8.5 9 8.5 Average 99.4 40.5 4.4 8.5 CoUocalia linchi isonota su-bsp. nov. Chars, suhsp. — Resembling CoUocalia linchi elachyptera, but aver- aging slightly larger; color of upper parts decidedly duller and some- what less bluish or purplish. Geographical distribution. — Philippine Islands. Description. — ^Type, adult male, No. 192,610, U. S. N. M.; Irisan, Benguet, Luzon, Philippine Islands, June 3, 1903; R. C. McGregor and A. Celestino. Upper sm-face dull, dark, metallic green, the crown and upper tail-coverts with a slight bluish tinge; tail of the same color, scarcely more bluish; wings fuscous, the exposed surface of the quills dull greenish blue with but little metallic sheen, the supe- rior coverts like the back; lores dark brown, the feathers with pure white bases; sides of head and neck, chin, throat, breast, sides, and flanks brownish slate color, the throat and breast feathers with margins of pale grayish or whitish, those of lower breast, of sides and flanks so broadly bordered with white that the ground color is largely over- laid; median portion of abdomen white, with very fine dusky shaft lines; longest lower tail-coverts dull, dark, metallic green, slightly or not at all margined with whitish, the shorter ones brownish with ample white edges, or nearly all white with a dusky shaft-stripe; lining of wing dark brown, slightly glossed with metallic greenish. This race differs from CoUocalia linchi linchi in its smaller size and « Type, 1906.) NATURAL SCIENCES OF PHILADELPHIA. 209 much duller, slightly more bluish upper parts ; from C. I. cyanoptila in inferior size and much duller, more greenish upper surface; and from C. I. affinis in somewhat longer wing, slightly shorter tail, and decidedly duller, more greenish upper parts, particularly the tail. Birds in juvenal plumage, even before their wings and tails are fully grown, seem to be exactly like the adults in color above, and not ap- preciably different below except for possibly broader, more evident white margins on lower breast, sides, flanks, and lining of wing. This form of Collocalia linchi appears to be confined to the Philip- pine archipelago, and has been taken on only Luzon, Mindoro, Min- danao, and Bongao. Measurements of three adults are as under : Sex. Locality. Date. Wing. Tail. Exposed Culmen. Tarsus. 1 Irisan, Benguet, Luzon, [P.I. 11 a 44 May 17, 1903 June 6, 1903 Mav 8, 1903 101.5 100 100.5 37 39.5 36.5 3.5 4 4 8.5 9 9 100.7 37.7 3.8 8.8 Collocalia dodgei Richmond. Collocalia dodgei Richmond, Smithson. Quart., II, 1905, p. 431. Chars, sp. — Resembling Collocalia linchi isonota, but very much smaller; upper parts still duller, more sooty, with an even more green- ish metallic gloss. Wing, 88.5;'^ tail, 33; exposed culmen, 3.5; tarsus, 7 mm. Type locality. — Mount Kina Balu, northeastern Borneo. Geographical distribution. — Mount Kina Balu, northeastern Borneo, Description. — Type, adult, sex unknown, No. 191,575, U. S. N. M.; Mount Kina Balu, Borneo, spring of 1904; George H. Goss and H. D. Dodge. Upper parts dull brownish black with an appreciable oil- green metallic gloss, the pileum slightly darker; tail and superior tail- coverts darker greenish, somewhat shaded with bluish and without trace of white markings; wings like the tail, but inner margin of quills brownish with little or no metallic sheen, and upper wing-coverts, with innermost secondaries, more greenish like the back; sides of head and neck, chin, throat, breast, sides, and flanks brownish slate, darkest " Type. *^ Dr. Richmond's measurement of the wing is too great (loc. cit.). 14 210 PROCEEDINGS OF THE ACADEMY OF [April, on sides of head, the feathers of lores with pure white bases, those of middle of lower throat and breast slightly, of the sides and flanks more or less broadly, margined with whitish; centre of abdomen white; lower tail-coverts dusky with a slight greenish gloss, the longest ones nar- rowly, the shorter ones conspicuously edged all around with white, some of the shortest entirely white save for a narrow dusky shaft line ; under wing-coverts blackish with the dull oil-green gloss of the upper surface; a few of the dark brown axillars tipped with white. This remarkably distinct species continues to be known from only the type specimen, which was obtained by Messrs. Goss and Dodge during their recent expedition to Mount Kina Balu. This example is fully adult and in perfect plumage. By reason of its white abdomen, uniform greenish upper parts, with lack of any white on the rump and also of white markings on the tail, Collocalia dodgei needs special comparison with none of its congeners unless it be C. linchi. From the typical form of the latter species it differs even more than from C. I. isonota, the one to which it seems to be most closely allied, but from which, however, it is specifically dis- tinct. Whether the previous records of Collocalia linchi from ]\Iount Kina Balu and from other parts of Borneo belong to this newly de- scribed species or to some form of linchi we are unable with our pre- sent material to determine, and this will be an interesting question for future investigation. Collocalia esculenta (LinniBus). Hirundo esculenta Linnseus, Syst. Nat., ed. 10, I, 1758, p. 191 ("China" locality wrong; Amboina is the proper type locality). Collocalia hypoleuca Gray, Proc. Zool. Soc. Lond., 1858, p. 170 (Aru Islands). Collocalia viridinitens Gray, Ann. and Mag. Nat. Hist., ser. 3, XVII, 1866, p. 120 (Celebes). Collocalia spilura Gray, Ann. and Mag. Nat. Hist., ser. 3, XVII, 1866, p. 120 (Batchian). Chars, sp. — Resembling Collocalia linchi, but averaging decidedly smaller ; most of the rectrices with large white spots on the basal por- tion of their inner webs. Wing, 69-103; tail, 38-44; exposed culmen, 3.5-4; tarsus, 8 mm. Type locality. — Amboina Island, Molucca Islands. Geographical distribution. — East India Islands from Celebes and Sumbawa Island to the Solomon Islands and the Louisiade Archi- pelago, including New Guinea; Cape York, Queensland, Australia. Dr. Hartert is undoubtedly correct in his identification and adoption of Linnseus' Hirundo esculenta, ^^ even though the latter author gives " Syst. Nat., ed. 10, 1, 1758, p. 191. 1906.] NATURAL SCIENCES OP PHILADELPHIA. 211 thelocality wrong, since the description given by Rumphius/^ whom Linnaeus cites, refers without doubt to the present species. There seem to be at least two forms inchided under the name escu- lenta as now current, but the material at our disposal does not [permit a satisfactory segregation of these. Collooalia neglecta Gray. CoUocalia neglecta Gray, Ann. and Mag. Nat. Hist., ser. 3, XVII, 1S66, p. 121. Chars, sp. — Similar to CoUocalia esculenta, from which it differs in the much duller, and greenish instead of bluish or purplish upper sur- face; rump, wing-quills, and wing-coverts with light grayish tips that are, however, sometimes obsolescent, particularly in old or worn birds. Wing, 88-97; tail, 41-43; exposed culmen, 3.5-4; tarsus, 6.5 mm. Type locality. — Eastern Timor. Geographical distribution. — East Indian Islands of Timor, Savu, Alor, Dammer, Roma, Kisser, and Wetter. In the specimens from Timor examined the upper parts are dull grayish metallic green, the wings and tail slightly more bluish; the anterior lower parts, including the breast, brownish gray, the feathers tipped with grayish white, producing a mottled appearance, the middle of the breast and jugulum having most white, the chin scarcely any; abdomen nearly pure white; feathers of the sides with brownish gray centres and whitish margins much like the breast; lining of wing dark brown slightly mixed with whitish. Judging from Dr. Hartert's remarks,^^ birds of this species from Timor are not just like those on the neighboring islands, and the latter may as he suggests be subspecifically separable. This species is very different from all members of the genus excepting CoUocalia natalis and C. esculenta, and may easily be distinguished by its dull greenish upper parts combined with white-spotted tail-feathers. It is of course near CoUocalia esculenta, but apparently quite distinct enough to stand as a species, although Dr. Hartert is of opposite opinion.*^ CoUocalia natalis Lister. CoUocalia natalis Lister, Proc. Zool. Soc, Lond., 18S8, p. 520. Chars, sp. — Similar to CoUocalia neglecta, but white markings of tail not sharply defined; chin and upper throat darker, almost black- " Herb. Amboin., VI, 1750, p. 183. ^» Novit. Zool, IV, 1897, p. 268. ^' Novit. Zool, XI, 1904, p. 203. 212 PROCEEDINGS OF THE ACADEMY OF [April, ish; blackish shaft markings of the shorter under tail-coverts larger; feathers of the rump with somewhat broader white edgings. Wing, 122 mm. Type locality. — Christmas Island, Indian Ocean, south of western Java. Geographical distribution. — Christmas Island, Indian Ocean. This little known form we have not seen ; but it is apparently recog- nizable, though very much like Collocalia neglecta, of which it is possibly but a subspecies. 1906.1 NATURAL SCIENCES OF PHILADELPHIA. 213 DESCRIPTION OF A NEW ATJSTEALIAN GLYCYMEKIS- BY HENRY A. PILSBRY, Glycymeris insignis n. sp. The shell is rounded-cuneate, the valves rather thick and strong, inequilateral, the posterior outline being longer and subangular, the anterior end rounded. The posterior dorsal slope is flat, and the beaks project well above the dorsal outline of the valve-edges. The specimens are worn and without cuticle, white in the middle and ante- riorly, pinkish orange near the posterior end. Sculpture of about 17 low rounded radial ribs parted by narrower intervals, and about six strong broad but unequal concentric waves separated by deep con- strictions. The waves and radial ribs are weak near the anterior end, and wanting on the flat posterior slope. The hinge-line is very strongly curved, the two sides at an angle of about 90 degrees to each other. The hinge-plate is moderately wide, with 12 anterior and 11 posterior teeth, three or four near the middle being very small and somewhat irregular. There is a small triangular area or lozenge below the beaks, sculptured en chevron with about six widely diverging grooves. The basal margin is coarsely crenulate inside in correspondence with the external ribs, and the concentric waves of the exterior are very weakly indicated inside. 214 ' PROCEEDINGS OF THE ACADEMY OF [April, Length 23, height 24, diameter 15.6 mm. Length 23.5, height 23.3, diameter 16.7 mm. Geographe Bay, West Australia. Cotypes two valves, right and left, received from Dr. J. C. Cox, of Sydney, New South Wales. The strongly developed concentric sculpture of this species is unlike any described Pectunculus known to me. The flat posterior dorsal slope is another peculiar feature. Dr. Cox writes that it has never been found in fresh condition or with the valves attached. 1906. NATURAL SCIENCES OF PHILADELPHIA. 215 ON HAWAIIAN SPECIES OF SPHYRADITIM. BY HENRY A. PILSBRY AND C. M. COOKE, JR. The genus Sphyradium occurs over practically the whole Palsearctic and Nearctic realms, with one species in the Neotropical, but it has not hitherto been reported from Polynesia. The two Hawaiian species to be described below have all the shell characters of the genus, but since the jaw and teeth of neither is known, their reference to Sphyradium is provisional. Sphyradium sharpi P. and C, n. sp. Figs. 1, 2. The shell is narrowly umbilicate, tapering-cylindric, the summit obtusely conic; thin; chestnut brown, paler near the apex, the surface somewhat shining and distinctly but not closely striate. Whorls slightly over 5, quite convex, the suture deeply impressed. The aper- ture is strongly oblique, rounded lunate. Peristome thin and simple, the basal and outer margins well arched ; columella vertical with broadly dilated edge. The insertions of the lips are remote. The umbilicus is circular and deep. Length 1.8, diam. 1.36 mm. Hawaii : Crest of the Kilauea crater, about a half mile south of the hotel. Types No. 91,699, A. N. S. Phila., collected by Dr. Benjamin Sharp. Cotype in the Bernice Pauahi Bishop Museum at Honolulu. This species is closely related to the following, from which it differs in being somewhat larger but with a smaller number of whorls. The Fig. 1. 0 Fig. 2. Fig. 3. shape also differs perceptibly, S. alexanderi being noticeably more egg- shaped, with a rounded summit. It is also less strongly striate and 216 PROCEEDINGS OF THE ACADEMY OF [April, more polished, S. sharpi occurred with various species of Tornatellina and Nesopupa. Sphyradium alexanderi C. and P., n. sp. Fig. 3. The shell is openly perforate, dextral, broadly ovate, subtruncate at the base, the apex blunt; nearly smooth, minutely striate especially below the sutures, shining, very thin, slightly diaphanous, chestnut, apical whorls much lighter; whorls 5^5f, well rounded, increasing very slowly; sutures simple, well impressed. The aperture is oblique, quadrate-lunate; columella vertical, lip thin, reflexed at the columella, the margins remote; umbihcus circular, deep. Length 1.7, diam. 1.17 mm. West Maui, at the top of Mt. Kukui, elevation about 6,000 feet. This shell, in company with a species of Tornatellina and another of Auriculella, was found by C. M. Cooke, Jr., at the very apex of West Maui overlooking lao valley. It was found only on low shrubs, two or three feet in height, at the base of leaves. Collected in February, 1904. Type in Bishop Museum, M. P. 8. Cotype in A. N. S. Phila., No. 91,292. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 217 ADDITIONAL NEW SPECIES OF POLYCH^TA FROM THE NORTH PACIFIC. BY J. PERCY MOORE. In the following pages are described a number of new species of Polychaeta, belonging to several families, from the collections of the Alaskan Salmon Commission of 1903. Three papers noticing other novelties in these collections have already been published in these Proceedings with the approval of the Commissioner of Fisheries, through whose kindness the material was placed in my hands for study. The full report on all of the species represented will appear in the Bureau of Fisheries Bulletin. Notophyllum imbricatum sp. nov. Plate X, figs. 1-3. The two examples upon which this species is founded bear a re- markable resemblance in proportions and general aspect to a Polynoe. The body is depressed and the large, scale-like notopodial cirri are imbricated and, except for a short space near the anterior end, com- pletely conceal the back from above. The type and larger specimen is a trifle over 30 mm. long, and the greatest width between tips of the parapodia a little anterior to the middle of the body is 6.5 mm. The small individual is 12 mm. long and 1.8 mm. wide, and the back is concealed by the elytra even more completely than in the larger one. The prostomium is flattened and wedge-shaped, with the slightly concave or straight base about two-thirds of the length, and the angles rounded. A slight emargination of the anterior portion of the sides is filled by an irregular low tubercle from which the frontal cirri arise. A pair of large brown eyes occupy the greater part of the posterior region of the prostomium laterad of the median tentacle (fig. !)• The frontal tentacles arise in lateral and ventral pairs, the latter slightly in advance, from the slight emargination above mentioned. They nearly equal the prostomium in length, are stout and somewhat swollen above the base, and taper to fine tips. The median tentacle arises between the eyes and nearly fills this space. It is one and one- half times as long as the prostomium, stout and somewhat enlarged 218 PROCEEDINGS OF THE ACADEMY OF [April, above the base and tapers to a slender end. From behind its base a low ridge runs to the posterior margin of the prostomium. In both specimens the proboscis is retracted and the large mouth is bounded by a rather prominent lip, formed by the union of the peristomium with the succeeding somite. The peristomium and two succeeding segments are much crowded forward so that the tentacular cirri arise beneath the head. The first or peristomial cirrus is a ventral cirrus and arises from the forwardly directed portion of the peristomium beneath the eye. It resembles the median cephalic tentacle in form and size, but is a trifle longer and arises from a distinct basal article. Arising from beneath the pos- terior dorsal margin of the prostomium, apparently from the peristo- mium, are three somewhat flattened appendages on each side, the outer- most of which is the longest and connected with the corresponding tentacular cirrus by a slight web. The middle one is nearly as long, and the inner one is minute (fig. 1). Somite II bears two pairs of tentacular cirri, a dorsal and a ventral one, separated by a considerable interval in which arises a small papilla probably representing a parapodium. The ventral cirrus occupies a position below the peristomial cirrus, which it resembles, but which it exceeds decidedly in both length and thickness. The dorsal cirrus of this somite arises at a higher level than any others in the body from beneath the postero-lateral angle of the prostomium, and its rather stout but long and tapering style reaches to IX or X. Somite III bears a dorsal tentacular cirrus only, which, with the com- plete parapodium to which it belongs, is depressed to a position more ventral than usual, in marked contrast to the dorsal cirrus of II. It arises from beneath and slightly behind the latter, which it resembles in form and size (fig. 1). As stated before, the first three somites are carried well forward, the peristomium and II being coalesced ventrally but distinct, though very short, above. Remaining somites are well differentiated and obscurely biannular, the anterior and decidedly larger annulus bearing the parapodia. There are about 78 somites, the posterior end of the body being in a state of regeneration and the somites of that region as a consequence very small and tapered to a minvite pygidium. The body, excluding the parapodia, is slender, about two-fifths of the total width, and of nearly uniform diameter, except near the taper- ing ends. It is somewhat arched dorsally ; flattened and with a shallow neural groove below. Parapodia (fig. 2) are prominent and well developed throughout. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 219 In the middle region they exceed one-fourth of the entire width. They consist of a broad base, a somewhat flattened, tongue-shaped neuro- podium, slightly cleft at the tip and bearing a rounded and swollen prominence on the ventral side of its base, and a minute papilliform notopodium, which diverges dorsally from the neuropodium. Each division is supported by a single aciculus, the notopodial being very slender and curved, the neuropodial stouter and straight. Along the posterior face of the base of the neuropodium above the papilla mentioned arises a flange-like ceratophore bearing a large, foliaceous palette-shaped ventral cirrus which curves upward behind the neuropodium and completely conceals both it and its setae from behind. The notopodial ceratophore is very large, causing the notopodium itself to appear as a mere appendage. At its dorsal side the free distal border is prominently produced, thus prolonging the surface of attachment of the notopodial cirrus. The cirrus itself is very prominent, of a somew^hat irregular reniform outline and attached by the marginal sinus. All of the notopodial cirri are turned nearly horizontally and overlap on the back in an imbricate fashion, closely similar to the elytra of the Polynoidoe. Anteriorly the dorsal cirri diminish in size so that they fail to cover the back completely and their ceratophores become more slender and elevated. The ventral cirri also become smaller, but remain prominent as far forward as III without essential change of form or position. A minute neuropodial tubercle exists on II, but it is uncertain if setae are present thereon. The eyes are purplish brown and the general color of the body a distinct dull greenish olive. Little pigment remains in any part of the body, but the notopodial cirri are more or less marked wdth dusky streaks and spots. A few small ova float free in the body cavity. The notopodium usually bears but a single slender, curved and simple seta, and even this appears to be absent from several of the anterior parapodia. Neuropodial setae are numerous, upwards of 20 to 30 occurring in the subacicular and 12 to 18 in the supraacicular groups. They are colorless and transparent, compound, with the stem gently curved and slightly enlarged at the end, where each side of the socket is provided with 6 to 8 very long, slender teeth and several shorter ones. The blade is slightly curved and tapers to an acute tip, and is striated and provided with minute marginal denticulations (fig. 3). The only specimens are the two from Station 4,269, Afoqnak Bay, 14 to 19 fathoms, hard gray sand and rocks. 220 PROCEEDINGS OF THE ACADEMY OF [April, Eulalia quadrioculata sp. nov. Plate X, figs. 4-6. This description is based upon a single much contracted entire specimen (type) and a fragment of the anterior end of another in every- way similar. The type is 27 mm. long, the protruded proboscis 2 mm. additional, the width without parapodia 2 mm., with parapodia 3 mm., the diameter of the distal end of the proboscis 2.5 mm., and the number of segments 106. The evidently much contracted prostomium (fig. 4) is about two- thirds as long as wide, scarcely emarginated in the median line pos- teriorly, rather tumid in the posterior lateral part and then slightly concave to a small truncate median anterior lobe. The usual pair of eyes are transversely elliptical, situated on the dorsal surface of the prostomium about three times their diameter apart, nearly twice their diameter from the lateral margins of the head and not more than their diameter from the posterior margin. They have distinct lenses. In nearly the same transverse line or very slightly in advance and half way between the dorsal eyes and the margin, or just within the nuchal organs, is a second pair of eye-like spots of black pigment, but lenseless and smaller and more irregular than the dorsal eyes. The four frontal tentacles are subequal, about as long as the head, and rather thick, with acute tips. The ventral pair project somewhat downwards, the dorsal directly outwards. From the middle of the small lobe lying between the frontal tentacles a shallow longitudinal groove passes to the median tentacle, which arises from a point just anterior to the eyes. It equals the frontal tentacles in length, but is slightly more slender in its distal part. The four tentacular cirri arise in the positions usual in the genus from somites I, II and III, and are short and subequal, being about twice the length of the prostomium. The segments are all well differentiated and dorsally are strongly arched; anteriorly they are strictly simple smooth rings, but in the posterior third become biannulate. Ventrally the body is marked by a neural groove and lateral ridges bearing glandular areas ventral to the parapodia. On this surface the biannulation extends nearly to the anterior end. The anterior segments are contracted and extremely short, farther back they become relatively longer, and toward the posterior end the body is distinctly flattened. The pygidium, which is provided with a thickened welt-like rim surrounding the anus, bears on the ventral side a pair of prominent acuminate cirri resembling the ventral cirri in size. Parapodia are located at the level of the ventral surface. That on 1906.] NATURAL SCIENCES OF PHILADELPHIA. 221 II, related to the second tentacular cirrus, is rudimentary and achae- tous ; that on III is larger, perfectly formed and bears setse. The others have the form shown in fig. 5 and bear foliaceous dorsal cirri. When fully developed they are sharply marked off from the ventral glandular swellings by distinct lateral grooves. The notopodium is of course wanting, and the neuropodium is a small somewhat flattened process divided distally into a very small postsetal and a decidedly longer presetal lobe, from between which the single vertical series of setae projects. Each lobe is notched on the edge at the point where the single .straight slender aciculum reaches the surface. The neuropodial cirrus is prominent from III back and is a short, rather thick process attached to the posterior side of the base of the neuropodium. In the middle region it is somewhat foliaceous and reaches to the tip of the neuropodium; in form it is triangular with the broad, somewhat convex base ventral and the apex dorsal, while the attachment is by one of the short sides. Posteriorly they become more slender and project distinctly beyond the end of the neuropodium. The dorsal cirri (fig, 5) are prominent and foliaceous throughout. They arise from stout bases situated a short distance dorsal to the neuropodia, which they in most cases exceed in size. All of the cirri have their fibrous and glandular structures arranged pinnately along an axial core. Anteriorly the cirri are rather broadly lanceolate with acuminate, somewhat recurved tips; posteriorly they become more slender and elongated; and finally are very narrowly lanceolate and of a length exceeding the diameter of the body. At the same time their foliaceous character is gradually lost. About thirty setae, equally divided between the supraacicular and subacicular groups, form the vertical fan-like fascicle. They are of the usual compound form (fig. 6) with the transparent, colorless stems rather stouter than those of E^ longicornuta, the thickened and nearly truncate end furnished on each side with seven or eight slender teeth, one of which is much larger than the others. The blades are short, broad at the base, and rather conspicuously striated and fringed. The type specimen, a female filled with large eggs, retains a dull olive color throughout the l^ody, becoming brown on the dorsal cirri. Be- sides the black pigment in the eyes, there is a diffuse spot near the tip of each dorsal cirrus, and a very minute spot beneath each ganglion of the ventral chain. The cotype shows some indications of a faint transverse band across the dorsum of each segment. The type and cotype are from Quarantine Rock, Port Townsend, Washington, June 27, 1903. 222 PROCEEDINGS OF THE ACADEMY OF [ApHl^ Eulalia longicornuta sp. nov. Plate X, figs. 7, 8. A complete example, much contracted, measures 15 mm. in length, with a body width of 1.5 mm. at the widest part, and a total of 73 segments. The broad prostomium is very slightly cordate, about as wide as long, and has no distinct concavity or constriction behind the frontal tentacles. The latter arise from the extreme anterior end of the prostomium, are slightly longer than the head and very slender. The median tentacle arises immediately in front of a line connecting the anterior border of the eyes or very close to the centre of the prostomium. It is If times as long as the head, very slender, and tapers regularly from the base to the tip. The single pair of eyes are circular, black, situated nearly their diameter from the posterior margin of the prostomium and twice their diameter from each other. Both specimens have the proboscis retracted, in which condition the mouth is bounded below by a some- what swollen, longitudinally furrowed lip. The peristomial somite appears on the dorsum as a slightly elevated lenticular area overlapping the prostomium. The first ventral tentacular cirrus arises directly beneath the eye, the second one from somite II ; both are slender and subulate and reach back to somite VII or to a length 2+ times the head. The two dorsal tentacular cirri arise from II and III respect- ively, and are about twice the length of the ventral cirrus, very slender and regularly tapered to a delicate tip. Dorsally the body is strongly arched above, below flattened and slightly grooved. It is widest at the middle and tapers regularly and nearly equally both ways. The segments are well marked throughout and show scarcely a trace of biannulation, even posteriorly. The anal cirri are missing from both type and cotype. Although the specimens are smaller, the parapodia (fig. 7) are even more prominent than in E. quadrioculata, owing to the projecting character of the dorsal angle of the presetal lobe, but otherwise they are similar. The dorsal cirri are, however, very different from those of that species, being folded against the sides of the body instead of held erect. They are strongly fohaceous throughout, the anterior and middle ones having a broadly pyriform outline, and those of the latter region being especially broad ; the posterior ones are more slender and rather cuneate-ovate. Moreover, the central area is always broad and thick and the gland ducts and other markings radiate from it in all directions and not in a pinnate manner. The ventral cirrus is also prominent and projects beyond the dorsal lobe of the parapodium. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 223 There are about 10 supraacicular and 12 subacicular setae in all, but fewer on the extreme anterior and posterior parapodia and, except for the usual variations in relative length of the blade, all have the form exhibited in fig. 8. The slender stem ends in an acutely oblique enlargement with 7 to 9 nearly equal, closely appressed teeth on each side. The blades are slender, flexible and very acute, with very fine marginal fringe. In both specimens the color has faded to a nearly uniform olive with some brown striations on the dorsal cirri. The type is a female filled with eggs. These specimens were found among serpulid tubes taken at the Quarantine Station dock near Port Townsend, Washington, on June 27, 1903. Pionosyllis magnifica sp. nov. Plate X, figs. 9-11. This large syllid is described from two specimens, one of which (the type) measures 48 mm. long and nearly 2 mm. wide, exclusive of cirri and setae, in the middle of the body. The form is much depressed,, especially in the widened middle region, from which it tapers to the very small head and pygidium. As just indicated the prostomium is small, its width little exceeding three-fifths of the width of the second segment and one-fifth of the maximum breadth of the body. It is depressed, somewhat quad- rangular in form, widest anteriorly where the width is about double the length. The somewhat flattened palpi project forward and some- what downward and are very slightly connate at the base (fig. 9). All three of the cephalic cirri are decidedly slender and arise in a transverse row from almost the extreme anterior margin of the prosto- mium. In one specimen they are subequal and about three times the length of the prostomium. In the other the median one is three and one-half times, the lateral about twice the prostomial length. The ends may be partly sloughed away. The eyes, though small, are very conspicuous and nearly black, the anterior slightly the larger. Together they form a quadrate figure conforming to the outline of the head and about two-thirds as large. Dorsally the peristomium is very short, but ventrally it projects as a prominent lip surrounding the large mouth. A fragment of the anterior end of an example of this species from Station 4,235 has the proboscis protruded as a short bell-shaped structure, bearing ten prominent papillae at the end and just behind them a conical, dorsal median tooth which appears to be quite soft. On the roof of the pharynx just behind the everted portion there appears to be, however, a hardened, horny elliptical area. This specimen also has the eyes. 224 PROCEEDINGS OF THE ACADEMY OF [April, larger than the others and nearly connate, but no swimming setae exist on any of the small number of anterior segments present. Of the tentacular cirri (fig. 9) the dorsal peristomial is about four times as long as the prostomium, that of somite II is slightly longer and of III as much shorter ; the ventral peristomial cirrus is about two and one-half times the prostomial length. All of the tentacular cirri resemble the cephalic cirri in being slender, delicate, and not monili- form, or with indistinct irregular furrows only. After the second, the segments increase very gradually in length and more rapidly in width and soon assume the depressed form char- acterizing the middle region where they are five or six times as wide as long. The type specimen has about 150 somites, the cotype only 110. Posteriorly the body becomes slender and ends in a minute annular pygidium, bearing on its ventral side a pair of very slender caudal cirri equalling the last twelve segments and exceeding any of the cephalic appendages. > The parapodia are of the form usual in Syllidse and project promi- nently from the sides at a low level. Although there is no distinct notopodium, a slender notopodial aciculus is always present just be- neath the notopodial cirrophore (fig. 10). The well-developed neu- ropodium terminates in a broadly rounded, more dorsal, postsetal process and a longer, rather prominent, and more ventral presetal process. The three or four aciculi terminate at the upper outer angle of the former. Ventral cirri are always short, stout and bluntly rounded, with oblique bases passing into the ventral surface of the body, and in size about equal the neuropodia. Notopodial cirri arise from very large and prominent cirrophores. Except for their slightly larger size anteriorly, where they are about three times the length of the prostomium, the notopodial styles are similar throughout. They are probably somewhat contracted and in life would be longer and more slender. In the middle of the body they are scarcely one-half of the total width. As shown in the figm-e they are rather stout and coarse, and, though more or less deeply marked wdth irregular transverse fur- rows, are never regularly articulated or moniliform. The only varia- tion in the parapodia is that they become more prominent in all their parts posteriorly. Notopodial aciculi are slender, curved and acutely pointed; the neuropodial are stouter, nearly straight and knobbed at the end. All setae (fig. 11) are compound and all are subacicular in position. In middle parapodia they are numerous, arranged in about ten horizontal rows of three to five each, or about forty in all. They are colorless. 1906.] NATURAL SCIENCES OF PHILADELPHIA, 225 with long, slender, curved shafts, the ends of which are rather abruptly enlarged and not very oblique, and are provided with only a few small teeth at the apex. The appendages or blades are strongly hooked and bifid at the end and distinctly fringed. The shortest posterior ones have a length of about one and one-third times the diameter of the end of the shaft, the longest about four and one-half times that diameter. Both specimens are entirely colorless and the large one is filled with masses of sperm. The type comes from Admiralty Inlet, near Port Townsend, Wash- ington, Station 4,219, 16 to 26 fathoms, on a bottom of green mud with sand and broken shells. A fragment was taken at Yes Ba}^, Behm Canal, 130 to 193 fathoms, bottom of gray mud. Stauronereis annulatus sp. nov. Plate X, figs. 12, 13, and Plate XI, figs. 18-22. The larger example (much contracted) is 13 mm. long and about 1 mm. wide exclusive of the setse, and has 72 segments. The other is 9 mm. long with 62 segments. The prostomium (fig. 12) consists of a broad shovel-like anterior process and two short segments, each bearing a pair of tentacles and a pair of eyes. Although these two divisions or rings are about equal, both the eyes and tentacles of the anterior one are much the larger. The anterior tentacles (palpi) arise from the ventro-lateral region of the first annulus. Each consists of a large and very stout basal piece strongly curved backwards by the sides of the head and bearing on its end a very small ellipsoidal terminal article. Just above the base of each, on the dorso-lateral region of the head, is a large very dark brown eye. Immediately behind and slightly above the large eyes the second pair of tentacles are borne on the second annulus. They are about as long but not so stout as the anterior ones, are cylindrical in form and consist of six or seven nearly spherical articles. The second pair of eyes are minute dark brown spots about one-fifth the diameter of the anterior pair, and are situated on the dorsum of the second annulus about half as far apart as the anterior pair. The mouth is small and bounded behind by the second somite, and from it project the ends of the jaws (fig. 18). Owing to the retracted state of the proboscis the jaw apparatus is not all visible. A specimen cleared in glycerine shows on the dorsal side a lenticular area having a very dark brown border and a paler interior enclosing a small central space. In ventral view are seen the pair of dark brown mandibles (fig. 18) with slender bowed bases and curved, divergent, tapering end-plates bearing about seven strong teeth along the medial margin. 15 226 PROCEEDINGS OF THE ACADEMY OF [April, The prostomium and somite II are apodoiis, short simple rings. All others bear parapodia, but are not otherwise more complicated in structure. Up to about the fifteenth they increase in size, but remain uniannular throughout. The body is strongly arched above and nearly flat below, and terminates in a simple ring-like pygidium bear- ing a pair of small ventral anal cirri as long as the diameter of the pygidium. Parapodia'are small, slender and strictly lateral in position. Except that they correspond in size with the segment bearing them and are consequently largest at the middle of the body, they are quite similar throughout. The neuropodium (fig. 13) is slender, nearly cylindrical and slightly enlarged distally, where it terminates in a presetal lobe divided into a larger ventral and a minute acicular process, and a postsetal lobe which begins just below the aciculum and runs to a rather prominent dorsal angle. Its dorsal surface bears a group of long cilia. The neuropodial cirrus is a simple finger-like papilla arising in or near the distal third of the ventro-posterior siu"face of the neuropodium and reaching nearly to the end of the latter. The dorsal cirrus arises immediately above the base of the neuro- podium and consists of a slender, elongated cirrophore nearly as long as the latter and slightly diverging from it. Usually but not always it is slightly constricted about the middle and the end is a little thick- ened. Probably it represents the notopodium, as what appears to be a slender aciculus penetrates about half of its length. A tuft of long cilia resembling that on the dorsal surface of the neuropodium is usually present on its ventral and sometimes on its dorsal surface also. On the end of this basal piece is borne a terminal style of about half its length and of a slender, conical form. The first parapodium altogether lacks a dorsal cirrus. Setse are numerous and of four forms, all very delicate, transparent and colorless. They are arranged in well differentiated supra- and subacicular fascicles, each comprising two kinds, and'all of which occur throughout the body, except perhaps in a few anterior segments, where the long compound setse have not been found. All of the supraacicular setae are simple ; those most numerous being very delicate but stiff capillary bristles, tapered to very acute tips, slightly curved and rather strongly serrated in an antrorse manner along one side (fig. 20). Their exposed parts are fully three times as long as the entire neuropodium. The other form has generally stouter stems, straight and tapering until near the end (fig. 19), where they present a gentle ventral curvature and at the same time become 1906.] NATURAL SCIENCES OP PHILADELPHIA, 227 broadened and end in two divergent, slightly curved points, the ventral of which is the longer; w^hile below the dorsal is a serrated sheath, the rather coarse teeth of which are directed outward. Seldom more than three or four of the latter form occur in a parapodium along with ten or fifteen of the former, Subacicular sets are all compound. The most numerous kind occur to the number of about fifteen. They have moderately stout, rather strongly curved stems, the end being very unequally and obliquely bifurcated to form a socket, the dorsal border of which is provided with a few teeth (fig. 20), while the blades are comparatively short, but increase in length from the ventral to the dorsal margin of the bundle, and have one border fringed and the end terminated by a pair of distinct but not widely separated and nearly parallel teeth. The second kind of compound setse (fig. 22) seldom exceeds three in numl^er. They have the same construction as the more numerous form but are much more slender and delicate in all their parts, and the blade often equals the entire length of the neuropodium or about three times the length of the longest blades of the other type. On the most anterior segments the two forms appear to grade into each other. The setse of this species differ decidedly from those of typical members of the genus and are more nearly like those of Prionognathus ciliatus Keferstein. These worms are quite colorless and their form and histological structure indicates that they may be pelagic in habit. The type and a somewhat smaller cotype were obtained at Quaran- tine Rock, Port Townsend, Washington, June 27, 1903. Notomastus giganteus. Plate X, figs. 24, 25. The only complete specimen measures 140 mm. in length and 7 mm., in maximum diameter in the thoracic region, but a second incomplete example is much larger. Even taking into account the contracted state of the specimens, this species is much stouter than usual for the genus. The body is nearly terete or slightly depressed and for the first fifteen or twenty millimeters increases in diameter, and then falls off to the posterior end which is two millimeters in diameter within ten segments of the anus. The prostomium is a small rounded lobe bearing a minute conical palpode and is completely retracted within the peristomium, which, except in being slightly longer, resembles the immediately following segments. The protruded and collapsed proboscis "forms a discoid, wrinkled structure fully 8 mm. in diameter. All of the thoracic seg- 228 PROCEEDINGS OF THE ACADEMY OF [April, ments are strongly biannulate, the anterior annulus being slightly shorter than the posterior. Each annulus is marked out in irregular areas, of which there is generally but one series to each annulus, though on somites IV to VII inclusive they become arranged irregularly in two rows. Behind VII the thoracic segments become smooth and more glandular and decrease in length, though even the last is fully one-half longer than the first abdominal. Beginning with II, each thoracic segment bears small notopodial and neuropodial tufts of very delicate, narrowly winged capillary setae in the usual positions, but neither the lateral sense organs nor the genital pores can be detected in surface views of these specimens. The abdominal segments, of which there are 190, are very short, with shallow, ill-defined furrows, and are either simple rings or ante- riorly obscurely biannulate. The surface is smooth, and the integu- ment provided with a thick glandular coat on the anterior and a much thinner coat on the posterior segments. The notopodial tori are very long, but not at all elevated above the surface in these specimens, and bear a great number of uncini. The notopodial torus is much more elevated and prominent, especially on posterior segments, but is much shorter and contains a much smaller number of uncini. The gills are low, rather long, inconspicuous folds. The pygidium forms a narrow circumanal welt bearing two longer prominently protruding ventral cirri and four much smaller ones in two pairs more dorsally placed. The uncini (figs. 24, 25) are numerous and very delicate crochets of a peculiar form. They are f-shaped with the densely fibrillated core exhibiting a slight spiral turn, and the tip provided with a single bent terminal process, somewhat flattened and at the base swollen and over- arched by a depressed hood, the margin of which is denticulated with eight or ten teeth. The type comes from Station 4,264, off Freshwater Bay, in Chatham Strait, at a depth of 282-293 fathoms, and on a bottom of green mud ; the larger but incomplete cotype was taken at Station 4,197, in the Gulf of Georgia, at a depth of 31 to 90 fathoms, on a bottom of sticky green mud and fine sand. Travisia pupa sp. nov. Plate XI, fig. 23. This is a thick, stout, grub-shaped worm tapering nearly" equally both ways but having the anterior end rather blunter and thicker. The considerably number of examples in the collection measure from 24 mm. to 82 mm. in length, with corresponding diameters at the middle of from 8 mm. to 32 mm. Exclusive of the pygidium there are thirty. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 229 one or thirty-two somites, but the number is not correlated with the size of the worm, the largest two having thirty-one and the smallest thirty-two. The prostomium is a minute conical organ, thin-skinned, weak- walled and hollow, and apparently capable of distention by internal fluid. It is followed by a short, uniannular segment continuing the general conical form of the anterior end. Dorsally this segment is crenulated on its posterior margin; below it is flattened and slightly grooved in the middle of the posterior part ; and on each side it is pro- vided with a dorso-lateral groove which terminates anteriorly in a deep sensory pit at the posterior margin of the prostomium. The next segment is biannular, with the anterior ring distinctly larger. Below it forms the upper lip and is thrown into a number of deep longitudinal folds and furrows which pass into the mouth. On each side, above a rather prominent swelling, is a quadrangular area opposite the groove on the first segment and itself bounded by a longitudinal groove above and another below, both of which cut the segment for its entire length. This area bears the small tufts of notopodial and neuropodial setse and between them an elliptical clear spot or pit. Dorsally each ring is marked by longitudinal grooves which effect a peculiar lobed and crenulated arrangement of the posterior margin. The third somite is triannulate, but the anterior two rings are somewhat united and on the ventral surface the first enters the sides of the mouth and the second forms the longitudinally grooved lower lip. At the sides this segment is constructed like the second, except that the posterior third of the quadrate area is depressed and smooth and that a cirrus is borne on each side immediately above and behind the notopodial setae. Above the same longitudinal sulcation and posterior lobing appear. The next twelve segments are formed of three equal rings completely separated by continuous furrows, except for the short interruptions at the setigerous areas on each side. Dorsal longitudinal sulcations are wanting and a second sensory pit, which first appears on the middle of the second ring below the setigerous area of somite IV, becomes very conspicuous on the posterior segments of this region, and on every specimen finally ceases on XV. Somites XVI, XVII and XVIII each consist of a large posterior and a small anterior ring, the furrow separ- ating which becomes successively more and more restricted to the dorsal and ventral regions. Traces of a short anterior ring, differen- tiated only dorsally and ventrally, still continue on XIX and XX, but all remaining somites are strictly uniannulate with prominent over- lapping posterior margins, which finally become telescopic. Somite 230 PROCEEDINGS OF THE ACADEMY OF [April, XIV, which is exactly at the middle of the body, is the longest; but the reduction in l^oth length and diameter of the segments is slight until near the ends, where it becomes more rapid. The pygidium is a short squarely truncated tube marked externally by longitudinal grooves which correspond with the clefts between the nine to twelve uneven lobes into which its margin is divided. The surface of the body is vesiculated or finely pustular in the follow- ing manner: Generally over the posterior half of the body, in all of the intersequental furrows and on all except the most anterior setig- erous areas, the pustules are very small and, though numerous, not crowded. In macerated specimens they are collapsed and appear as punctations. From the middle of the body they gradually increase in size forward. Each annulus of anterior somites is provided on its highest part with an irregular transverse series of very large vesicles which usually lie nearer to the posterior margin and overlap the suc- ceeding ring, when the worm is contracted, as a rough and irregular fold interrupted in the median dorsal region. These folds and their vesicles are usually best marked on the middle ring of each somite, on which they increase in prominence from the median break laterally to the setigerous areas, below and even more above which they form rough lobes overlapping the third ring. The ventral lobe bears the ventral sensory pit referred to above. Anteriorly the transverse series of enlarged vesicles tends to form several rows and all of the vesicles to increase in size. The first three segments bounding and anterior to the mouth are covered nearly uniformly with vesicles of moderate size; and similar ones cover the anterior two-thirds of the setigerous areas of the first ten or twelve segments. The prostomium is perfectly smooth and lacks surface vesicles altogether. Posteriorly the ridges of enlarged vesicles and the lateral lappets become gradually flattened out and the entire surface much smoother. The lappets above and below the setigerous areas remain, however, and may become even more prominent on the last few segments, where, however, they are not vesicular. Neuropodial cirri appear on III and continue on every segment to near the posterior end, ceasing on from XXIV to XXVII in different specimens. They arise at the dorsal margin of the setigerous area, chiefly from the third annulus but in part from the second also. At the base their diameter nearly equals the length of the third annulus, but they quickly become slender. The surface is strongly wrinkled trans- versely, much like the contracted tentacle of a jelly-fish, and doubtless they are in life capable of great extension;, but in the preserved speci- 1906.] NATURAL SCIENCES OF PHILADELPHIA. 231 mens, even where longest (in the middle of the body), they barely equal one-half or one-third of the body diameter. Toward the ends they are reduced to one- third or even one-foiu-th of this length. Lateral sense organs appear as a pair between the prostomium and peristomium and occur between the setae tufts of every succeeding segment, except that they are occasionally absent from XXXI or XXXII, In shape they are elleptical with the long axis vertical. Ventral sensory pits appear on IV and continue without exception to XV in the position indicated above. At fu'st very small, they increase rapidly until they exceed the lateral organs, unlike which they are always circular. Small notopodial and neuropodial tufts of setae occur on all somites from II caudad at the junction of the second and third rings. Both tufts are retractile into pits and the notopodial setae are somewhat longer than the neuropodial. All setae are very slender, flexible and thread-like, of various lengths in each tuft, and have each margin provided with a fringe of appressed hairs. Many of the specimens are enclosed in a very tough mucous mem- brane more or less coated with silt, and often inliabited by small nematodes. This species bears much resemblance to Travisia olens Ehlers, which has only thirty segments and rather distinct parapodial papillae. It is apparently an abundant worm, conspicuous from its large size, and widespread on muddy bottoms. Specimens were collected from the following stations: 4,192, Gulf of Georgia, 18 to 23 fathoms, green mud and fine sand; 4,194, Gulf of Georgia (tj^e locality), 111 to 170 fathoms, soft green mud; 4,197, Gulf of Georgia, 31 to 90 fathoms, sticky green mud and fine sand; 4,230, Behm Canal, 108 to 240 fathoms, rocky; 4,235, Behm Canal, 130 to 193 fathoms, gray mud; 4,237, Behm Canal, 192 fathoms, green mud; 4,246, Kasaan Bay, Prince of Wales Island, 101 to 123 fathoms, gray and green mud, coarse sand and shells. Brada pilosa sp. nov. Plate X, figs. 14-17. This well-marked species is represented by about a dozen specimens varying in length from 15 to 30 mm., the largest having a maximum diameter of 5 mm. With the prostomium and tentacles retracted, which is the condition of all of the specimens, the form of the body is slender clavate, bluntly rounded at the anterior end and gradually thickening to a])out XI or XII, from which point it tapers into the rather slender, gracefully formed caudal region. 232 PROCEEDINGS OF THE ACADEMY OF [April, The exact form of the prostomium and mouth is not apparent, inas- much as this region is retracted, leaving a conspicuous trifid opening which has been sometimes indicated in descriptions of other species as the true mouth. In this condition the peristomium is trilobate. All of the segments are simple rings, separated from one another by clearly defined but not conspicuous intersegmental furrows which become more distinct posteriorly. The segments pass regularly into one another without any conspicuous breaks in contour, and increase in length, as they do in diameter, to about XII, then undergo little change to the middle of the body, behind which they become again gradually shorter. Toward the posterior end they diminish to a minute pygi- dium which contains a small, vertical, slit-like anus, but appears to lack cirri or other appendages. The number of segments varies from 31 to 33, the latter number being present in the type. All, including the peristomium, bear both notopodial and neuropodial setse in tufts upon minute tubercles. The chief characteristic of the species is the richness of its papillation. The entire dorsal surface is thickly covered with filiform papillae espe- cially numerous toward the ends, where they are so densely arranged that they actually touch and crowd one another. On the middle segments they are more widely separated, but are still so numerous that where Brada villosa bears 3 or 4 in the length of a somite, this species bears 8 to 12, They are not disposed in regular rows, but are arranged more or less irregularly at nearly equal intervals in all direc- tions. The intersegmental furrows lack papillae and appear as smooth lines, like narrow avenues through a grass field. Just anterior to each parapodium the papillae become few or nearly disappear, leaving a wider open space at this point. Those papillae remaining in this region are collected about the setigerous tubercles, but instead of be- coming enlarged and forming rosettes they are actually smaller than the dorsal ones. All of the papillae (fig. 14) are slender, elongated and filiform with a small terminal knob. Many of them collect a thick girdle of sedi- ment in a zone near the base, which gives the appearance of a bulbous enlargement. They differ considerably in length and toward the anterior end of the body exceed the length of the segment bearing them, so that this region has the appearance of being thickly coated with fine hairs. At the level of the ventral margin of the neuropodial tubercles the dorsal papillation ceases abruptly, and is replaced by the much smaller papillae (fig. 15) which cover this surface. Although their number is nearly as great, these papillae are so much smaller than 1906.] NATURAL SCIENCES OF PHILADELPHIA. 233 those found on the dorsal surface that to the naked eye they appear merely as a fine granulation. Every somite, including the peristomium, bears both notopodial and neuropodial setse in small lateral tufts. The setae of both fascicles have the same character, all being slender and transversely jointed, except at the acutely pointed tip. The former (fig. 16) are very slender and delicate and the internodes increase in length toward the tip. The latter (fig. 17) are more deeply colored, much stouter, distinctly curved and the much shorter joints decrease in length toward the tip. In both tufts the number of setse is small, about eight notopodials and eight or ten neuropodials being the rule. On the peristomium the notopodials are very long, equalling four or five segments, and they project forward far beyond the mouth. On succeeding segments they project outward and upward and decrease in length until they are about equal to the segment bearing them. The neuropodials are rudiment- ary on I, but increase in both length and thickness on succeeding anterior segments, those at the posterior end becoming again more slender but without diminution in length. As stated above the prostomium is in all cases retracted. A dissec- tion shows that the tentacles are fine and very numerous, numbering upwards of thirty on each side. They are borne on a pair of bosses which are about twice as wide dorsally as ventrally, where they curve around the mouth and nearly meet. The palpi are very short and broad and marked by a longitudinal groove and transverse wrinkles on the ventral side, the dorsal surface being smooth. The skin is gray, but the papillae impart to the dorsal surface a buffy yellow color. Brada pilosa very closely resembles Trophonia hirsuta Theel, but is distinguished by the reduction in size and number of the papillae form- ing the setal rosettes and by the un jointed tip and other peculiarities of the setae. The species is not uncommon northward. Examples occur from the following stations : Station 4,251 (type locality), Stephens Passage, 198 fathoms, rocky bottom; 4,235, Yes Bay, Behm Canal, 130 to 193 fathoms, gray mud; 4,252, Stephens Passage, 198 to 201 fathoms, gray mud; 4,258, Lynn Canal, 300 to 313 fathoms, mud. Maldane similis sp. nov. Plate XI, figs. 26-30. The type and largest example is 56 mm. long and 2.5 mm. wide, the latter being nearly constant throughout the entire length. This species belongs to the M. biceps group in having the cephalic and caudal plates of nearly the same form and size. The cephalic plate ■234 PROCEEDINGS OF THE ACADEMY OF [April, is very broadly elliptical in outline, the margin little limbate, the pos- terior two-fifths separated by a deep notch on each side from the anterior three-fifths, the former erect and with its margin finely den- ticulated with from twenty to twenty-five teeth. One-fourth of the remaining margin anterior to the notch forms on each side a rather thick, narrow, spreading rim, the margin of which is quite entire or merely slightly crenulated, not conspicuously toothed as in M. hiceps. Anteriorly the cephalic rim is separated from the postero-lateral mar- gins of the palpode by a pair of distinct radial furrows. The palpode is very large and contributes easily three-tenths of the entire margin. It is very broad and flat, with a smoothly curved anterior border and rather more than one-third of the middle of its posterior border pro- duced backward on to the head plate as a slightly elevated median welt scarcely rising to the height of a ridge. On each side of this, forming its lateral boundaries and the posterior boundary of the lateral portions of the palpode are the deep sensory slits, U-shaped or hooked, with the lateral limb the shorter. No furrows or other mark- ings occur on the surface of the cephalic plate. The mouth is large and the lips prominent and pouting. A short ■distance behind it is a transverse groove encircling the ventral half of the peristomium and joining a conspicuous longitudinal groove which begins at the lateral notch in the cephalic margin and passes along the side of the peristomium, to end posteriorly in the circular furrow which separates a complete narrow ring from the hinder part of the •segment. Inasmuch as the posterior part of the peristomium is re- tracted within the anterior margin of II, this ring is completely con- cealed ventrally by the prominent half-collar developed in that posi- tion on the latter segment. The next segment (II) is very short, its length not more than one- half of its diameter; the posterior third is completely separated as &, distinct ring; and the anterior margin is somewhat produced into a collar, the ventral half of which springs into especial prominence abruptly at the level of the setse. Somite III is about one-third longer than II and similarly biannulated, but its anterior margin is not col- lared. Thus far the skin is entirely glandularly thickened. The following six segments (IV to IX) are decidedly longer, about equalling their own diameter, and the secondary furrow, while always present, is in a more anterior position not far behind the middle. While the three anterior segments are glandular throughout, the ones under discussion have the glandular area confined to the ventral half of the bod}^ and especially to the anterior ring in the vicinity of the 1906.] NATURAL SCIENCES OF PHILADELPHIA. 235 tori. Somites IX and X are indistinguishable, and on the latter the now prominent tori have shifted to the posterior end of the segment. From this point the segments continue to lengthen to XIV, which is about three times as long as thick, after which they again decrease. Throughout the middle region the integuments are soft and translucent, except for the swollen, oval, glandular areas surrounding each torus. Somites XVIII, XIX and XX are again much thicker than long and the tori are correspondingly large and prominent; XX is al^out one-fourth as long as thick with a prominent pair of achaetous tori much below the usual level and meeting ventrally. The pygidium consists of a very short basal ring bearing a pair of coalesced glandular thickenings corresponding to the tori of XX. The anus is conspicuous and in a dorsal position at the base of the dorsal membrane. The limbate margin of the pygidium bears a re- markable reseml^lance to the cephalic plate, but is oblique in a reverse ■direction. Its ventral two-fifths are separated rom the dorsal three- fifths by a deep rounded notch, with thickened margins which nearly meet externally and constrict its opening. The ventral portion has its margin marked by four very broad, shallow crenulations and a pair of prominent triangular lateral lobes. The dorsal plate is more prominent and flaring, with a smooth and regular margin marked only by a broad and extremely shallow median emargination. Somites I and II and the pygidium are achsetous; II bears strictly lateral setaj only, arranged in a vertical tuft just above the dorsal ends of the ventral collar; III and IV bear similar setse tufts and very short series of uncini disposed in the same plane and both strictly lateral and sessile. On succeeding segments the setae tufts have short, slightly oblique bases placed a little in advance of the uncini, which form lines five or six times as long as those on the preceding somites, more ventral in position, and elevated upon distinct tori. Proceeding caudad the uncinial lines increase slightly in length and become more ventral in position. The number of uncini in a torus is about 8 on III, 13 on V, 28 on X, 32 on XV and 35 on XIX. The small tuft on II contains setse of two kinds, the one small with a rather wide wing and abruptly tapered stem terminating in a slender tip ; the other very much longer, with the stem conspicuously striated the tips rigid and less slender and the wing very narrow (fig. 26). On follomng somites the number of setse increases to 10 or 12 pairs, the larger ones become stouter and the smaller more slender. By somite XV the number is further reduced ; the small setse exhibit only minute pointed tips and the ends of the larger ones are elongated and 236 PROCEEDINGS OF THE ACADEMY OF [April, provided with ensheatliing awns which appear to be disposed in sym- metrical pairs. Anterior uncini (on III) (fig. 27) have the head little enlarged and somewhat thrown back, the stout, blunt, nearly straight beak some- what elevated, the crest composed of a single large tooth with a cluster of small ones surrounding its base, and the guard rudimentary or ab- sent (fig. 28). In a succeeding tori there is a gradual transition to the typical form which appears at about VI. Such uncini (fig. 29) have a distinct shoulder, well-defined neck and head, and a stout, tapering, hooked beak with an" acute, slightly recurved tip. The crest is well developed and formed of two transverse rows of numerous teeth which are largest at the vertex and become rapidly reduced laterally. The guard is strong and arising well below the beak sweeps boldly beyond and above its tip (fig. 30). The type and one other specimen were taken at Station 4,264, off Freshwater Bay, Chatham Strait, 282 to 293 fathoms, on a bottom of green mud. Maldanella robusta sp. nov. Plate XI, figs. 31, 32. No complete specimen of this species occurs in the collection, but fortunately there are several heads and one posterior end and it is possible to so fit the fragments together as to secure a complete descrip- tion. • The worm thus reconstructed is a fine large one, measuring about 195 mm. long and 7.5 mm. in diameter at the middle part. Excluding the pygidium there is twenty-one segments, of which the peristomium and first preanal segment are achsetous, II bears setse only, XX setse only on one side, and III to XIX inclusive both setse and uncini. The prostomium and peristomium are completely coalesced and exhibit no trace of a dividing furrow or suture. The cephalic plate meets the dorsal profile of the peristomium at an angle of approxi- mately 135° to 150°, so that the ventral length of the head is nearly or quite twice the dorsal. The cephalic plate has a nearly regular ellip- soidal outline, with a length of about twice the width. Its thin margin is elevated all round, highest and most erect behind, thence slightly diminishing in height to a point anterior to the middle where it is folded and sometimes slightly notched. Anterior to this point it again becomes higher and more flaring until it curves into the base of the palpode, from which it is separated by a slight cleft. Except for a faint crenulation, the margin is entire. The median ridge is rather broad and low, but extends through the anterior half or less only of the 1906.] NATURAL SCIENCES OF PHILADELPHIA. 237 cephalic plate, ending abruptly behind; anteriorly it widens slightly and passes partly into the marginal rim, partly into the palpode. On each side of the ridge are deep sensory slits which anteriorly bend somewhat sharply outward and backward, and continue along the base of the lateral fold nearly as far as the posterior end of the median limb. From near the posterior end of the median ridge to the lateral fold or notch in the marginal membrane passes a strictly transverse furrow on each side, leaving an extensive area in the posterior region of the head marked onl}^ by a few crescentic furrows parallel with the pos- terior margin. The palpode is a short, broadly rounded, rather thick, tongue-like structure which is continued backward on the ventral side as a broad welt to the mouth, within which it bifurcates to form a pair of ridges separated by a deep cleft. The mouth is relatively small and bounded by a nearly circular fold, elevated and furrowed somewhat like a piece' of rope and which is open only anteriorly to admit the posterior extension of the palpode. As before mentioned there is absolutely no visible line of separation between prostomium and peristomium, but the two together constitute a continuous head, shaped somewhat like a horse's hoof; that is, it spreads anteriorly, where it is truncated obliquely by the cephalic plate. The seven somites next following are cylindrical, with a nearly uniform diameter about equalling the posterior diameter of the peristomium, and a length but little greater. All have traces of anterior collars, which are best developed on IV to ^^II, on which also the glandular layer of the skin is thick and extensive. The skin of the head and somites II to I\' is very smooth, iridescent and marked by fine furrows crossing in various directions, like those on the human skin. Somites V to ^TII are of a dull, opaque, non-iridescent white. Following this the body is distinctly depressed and the segments elongated to two or three times their diameter. They bear prominent tori situated along broad elevated longitudinal muscle bands. While the greater part of the surface of the segments is smooth and has a conspicuous bluish iridescence, the longitudinal muscle ridges are vertically furrowed and, when the segments are much contracted, these furrows are extended nearly around the segment. With the exception of -^, the segments bounding which are continuous and the place of transition of the parapodia from an anterior to a posterior position, all of the furrows are well developed. Somite IX is the last exhibiting a distinct glandular region, which is confined to a narrow anterior zone. For nearly the entire length of the worm there appears in the median ventral line a neural cleft in the muscles, having 238 PROCEEDINGS OF TPIE ACADEMY OF [April, the aspect of a clear, translucent, bluish line. There is no diminution in the length of the posterior setigerous segments, but the achsetous segments are considerably reduced in both length and diameter. They are terete, about twice as long as thick, have the surface deeply wrin- kled transversely and bear rather prominent tori in the posterior one- third. Following these is the campanulate pygidium which has a narrow ringed base and a deep cup-shaped body, the margin of which is divided into thirty-nine very regular bluntly rounded teeth, the four ventralmost of which are considerably broader than the others, while at three other points one of the latter has been replaced by two smaller ones. Both without and within the surface of the cup is longitudinally fluted, and in the deeper part of the interior numerous fine ribs, usually two to each marginal tooth, pass to the margin of the very large anus. Somite II bears capillary setse only, which arise as a narrow vertical tuft from a slit-like cleft into which they are retractile. Succeeding segments, to XIX inclusive, bear both setse and uncini which are strictly lateral in position, the dorsal interval between the setse scarcely exceeding the ventral interval between the uncini. As far as somite VIII the setae continue to occur in the form of vertical tufts retractile within slit-like pockets; and the uncini, which begin immediately below the setse and lie in the same plane, form strictly linear series sessile or even depressed below the surface. On IX and all subsequent segments the setse are situated on rather prominent wart-like papillse in the form of crescentic tufts open below ; and the uncini are elevated on the crests of swollen tori. On II, III and IV the parapodia are situated in the anterior one-third; on V and VI they are but little anterior to the middle; on VII, VIII and IX they are again near the anterior end ; on X they shift abruptly to the posterior one-third and so remain to the last. The posterior tori are especially prominent. Whether XX is normally setigerous cannot of course be determined until additional specimens are known. The number of uncini increases toward the posterior end, the counts being 22 to 26 on III, 35 to 42 on V, 45 to 50 on X, 53 to 57 on XV and 60 on XIX. The setse are numerous and form dense tufts. On II those of one series are shorter and about three times as thick as the others and have well-developed wings. Those in the other series are slender and nearly wingless. On succeeding somites all of the setse become much more elongated and more slender, but do not differ otherwise, and appa- rently lack altogether any lateral hairs or awns. As indicated above uncini (fig. 31) are numerous on all segments; 1906.] NATURAL SCIENCES OF PHILADELPHIA. 239 and are stout, striated, and of a deep yellow color throughout. They differ very little in form on the different segments, the only noticeable distinction being that the extreme anterior ones have the crest teeth less well developed and the guard hairs fewer. The rather long, slender, curved stem has a distinct but tapering shoulder, a rather long, erect neck and an enlarged head, below which is a very prominent square guard process (fig. 32). The beak is stout and hooked, the crest teeth three or four, the lowermost very large, the upper small or obso- lete; sometimes, especially on anterior uncini, a pair of small but dis- tinct lateral teeth is present. The guard is strong and consists of about sixteen stiff hairs which arise from a scale-like base ensheathing the front of the guard process, and spread regularly in an even curve around the end of the beak, above which they arise convergingly to a considerable height. The body of the alcoholics is generally colorless or pale yellow, but the cuticle has a strong bluish iridescence throughout. A portion of a tube is soft and flexible, consisting of a thick mucoid substance covered with a stratum of moderate thickness of soft grayish brown silt. This species is evidently related to the three species of Maldanella described by Mcintosh from the deep waters of the oceans of the Southern hemisphere. The Japanese maldanid Clymene harai Izuka { = Axiothea campanulata Moore) also belongs to this genus. The sources of the examples of M. robusta are Station 4,197, Gulf of Georgia, 31 to 90 fathoms, bottom of sticky green mud and fine sand; Station 4,230, Behm Canal, 108 to 240 fathoms, rocky bottom; and Station 4,246, the type locality, Kasaan Bay, Prince of Wales Island, 101 to 123 fathoms, bottom of green mud with coarse sand and shell fragments. Clymenella tentaculata sp. nov. Plate XI, figs. 33-35. This very interesting species is unfortunately imperfectly known, the following description being based upon one anterior and one posterior piece which may be parts of the same individual and which together represent nearly an entire worm. The former measiu-es 22 mm. long and 2 mm. wide at IX and con- sists of the prostomium and nine segments ; the latter is much twisted and the five setigerous segments, four achsetous preanal segments and pygidium measure about 20 mm. long. The cephalic plate is very much expanded and flares widely at the margin. Its outline is very broadly oval. The hinder third of the rim is separated from the anterior two-thirds by a pair of small lateral 240 PROCEEDINGS OF THE ACADEMY OF [April, incisions, behind which it is more erect and diminishes in height to a minute median posterior notch. Anterior to the lateral incisions the rather abruptly widened and flaring margins continue undiminished almost to the palpode which they join on each side. The cephalic margin is everywhere smooth and its margin entire. A pair of con- spicuous sensory slits divide the central disk of the head for about the anterior five-sixths of its length into three narrow longitudinal areas of equal width which are united behind. The central one is somewhat ridged and widens almost imperceptibly as it passes into the palpode anteriorly. The palpode consists of a short rounded base bearing a slender, elongated finger-like process on its median anterior margin. The cephalic plate forms a dorsal angle of about 120° or less with the peristomium, which is indistinguishably coalesced with the prostomium. It is little more than one-half as long as the cephalic plate and its sur- face is slightly granulated but unwrinkled. Owing to the protrusion of the proboscis, which has a depressed acorn-shape, with the basal division thickly papillated, the mouth is invisible. Somite II (the first setigerous) has a length about equalling the width at the anterior end, from which it gradually diminishes in dia- meter caudad. The next two segments are narrowerer, after which the diameter increases gradually to VIII, though the length remains nearly constant ; IX has the same diameter, but if complete is scarcely half as long as VIII, All of these segments are transversely wrinkled superficially and are provided with a distinct, raised neural line. No prominent collars but merely a low free rim, most distinct on V, are developed on their anterior ends. Owing to the much coiled and twisted condition the real proportions of the distorted posterior segments cannot be easily ascertained. They are evidently three or four times as long as wide, slender and thin- skinned, except posteriorly where the prominent, swollen and glandu- lar parapodia are developed. There are no especially developed glandular zones or muscular ridges and the neural line is elevated throughout. The last six or seven segments decrease in length and the entire region tapers to the pygidium. The first of the posterior achae- tous segments is about one and one-half times as long as wide and of a shape similar to those preceding it, having fully developed but naked tori on the posterior end. The next three are simple rings without tori and of rapidly decreasing length. Terminating the body is a remarkably small, top-shaped pygidium lacking any limbate margin. Instead there arises around the base of the anal papilla, which con- 1906.] NATURAL SCIENCES OF PHILADELPHIA. 241 stitutes its greater part, a circle of twenty-three separate and distinct cirri, all of which are slender, regular and equal, and not, as in many species of the genus, alternately longer and shorter. Apparently they increase slightly in size from the dorsal to the ventral side; and the median ventral one is much elongated, its length equalling that of the four achaetous segments combined or about ten times the length of the other marginal papillae, and it is very slender distally. On the first three setigerous somites (II to TV) about thirty seta? occur in each group and are disposed in small vertical tufts just above the lateral line and about one-third of the length of the somite from its anterior end. The small number of uncini (3 to 5) are sessile in short, transverse lines just below the seta?. On succeeding segments the setae are more numerous and project upward and outward in tufts from small tubercles. The uncini are more numerous (13 on Y, 20 on IX, and still more posteriorly) and form longer lines widely separ- ated below by a space of twice or more their own length. The tori have become prominent swellings. On IX they have become trans- ferred to the posterior end, although the boundary between this seg- ment and VIII is not clearly defined. This condition of the tori con- tinues throughout the body, though they become even more prominent posteriorly and are united across the dorsum of each segment by a glandular band. The dorsal interval between the setae is about equal to the ventral interval between the uncini. The first achaetous seg- ment at the posterior end bears a pair of perfectly normal tori, but no setae or uncini. .The setae are often imperfect and their distribution is worked out only incompletely. Anterior segments have them all slightly curved,, delicate and narrowly winged, with very slender, tapering tips. They occur in two series, one of finer, the other of coarser setae. Farther back these two kinds become further differentiated. Both become longer and the slender ones provided with short basal wings, beyond which is a delicate capillary tip doubly fringed with strongly divergent, very- fine hairs. All uncini are yellow and have the stems longitudinally striated.. Those of somites II to IV, in which the number is small, have the form shown in fig. 33. The stems are slender, slightly curved, regularly enlarged, but with no distinct nodulus, then slightly constricted to a neck, bearing a scarcely enlarged head (fig. 34), with a simple, tapering unhooked beak, a small crest of three minute teeth and apparently no guards. These uncini increase regularly to the dorsalmost, which 16 242 PROCEEDINGS OF THE ACADEMY OF [April, also has by far the most prominent beak. On the remaining somites the uncini differ decidedly in the form of the head (fig. 35) which is much enlarged backward. The smaller beak is more curved and hooked, the crest high and prominent, with five or sometimes six non- fibrous, imbricated teeth of diminishing size. The guard consists of several (about 8) fine tapering hairs, which arise from a plate just beneath the beak, the inferior outline of which they follow to the tip, above which the}^ then rise as curled inarched filaments. The alcoholic specimen is colorless, but the greatly developed para- podial plexuses of bloodvessels indicate red bands during life. Some fragments of tubes are 3 mm. in diameter. Their flexible walls are composed of a soft mucoid membrane covered with a thin coating of very fine neatly deposited sand. This species presents interesting resemblances to Praxilla gracilis (Sars) Malmgren in the form of the head and tentaculiform palpode, to Clymenella catenata (Malmgren) in the number of achsetous preanal somites, and to C. ruhrocincta Johnson in the elongated median ventral pygidial cirrus. It is known only from Station 4,264, off Freshwater Bay, Chatham Strait, July 25, 1903, 282 to 293 fathoms, bottom of green mud. Uicomache carinata sp. nov. Plate XI, figs 36-39, and Plate XII, figs. 43 and 44. A well-preserved but somewhat contracted specimen (the type) is 65 mm. long and 3 mm. in diameter at the thickest part. Another in- complete but more fully extended example must have exceeded twice this length when complete. The prostomium and peristomium are coalesced, forming a continu- ous head about If times as long as wide. The prostomial region is nearly vertical with about its medial ^ formed by a ridge, which forms a prominent profile, somewhat arched above and ending below and anteriorly in a transverse crescent, separated by a slight groove from the short, somewhat thickened palpode. The latter is continuous with the similarly thickened lateral margins of the head, within which is a pair of shallow longitudinal depressions, while within these again and close to the median ridge are the distinct longitudinal sensory slits or nuchal organs. The dorsal peristomial region is somewhat tumid anteriorly, with a nearly straight profile posteriorly, but strongly arched transversely throughout. From the prostomium it is slightly distinguished by a faint transverse groove which passes laterally into the depression men- tioned above and the groove which continues the latter posteriorly to the end of this segment. The mouth is a large elliptical opening sur- 1906.] NATURAL SCIENCES OF PHILADELPHIA. 243 rounded by a rugous, furrowed lip. Just behind this lip is a shallow transverse ventral groove, while near its posterior end this segment is completely encircled by another faint furrow which, like the groove mentioned, is met by the longitudinal furrow on each side. Besides the pygidium the type has 24 segments, while the only other complete specimen in the collection has 25. The peristomium is achsetous, and somites II to XXII (or XXIII) inclusive are setigerous, leaving two preanal achsetous segments, as in N. lumbricalis. Somite II is as long as the head ; the next seven or eight segments are succes- sively of slightly increased length, the last named being nearly twice as long as the first. Behind IX several segments remain nearly equal in length, then the length diminishes, at first slowly, then rapidly to the last, the last three setigerous segments each equalling III in length and the two achsetous preanal segments together barely exceeding the last setigerous. Somites II, III and IV are of much greater diameter an- teriorly, and slope to the posterior end which is inserted slightly into the next succeeding segment. The next five segments (V to IX) are more nearly cyhndrical, but bear slightly produced glandular collars at their anterior ends. The last four are simple short rings. At the two ends of the body the intersegmental furrows are deep and distinct, but in the middle region they are obscure. The pygidium is funnel-form, but little widened, regular and some- what longer than the last two segments combined. On the type its margin is divided into twenty-one equal and regular triangular teeth each with a minute cirriform tip. The only other anal funnel present has but sixteen marginal teeth and cirri. The anus is central and very large, and is surrounded by about ten regular lobes separated by as many radiating furrows. At about somite VIII the neural groove begins to be distinct between the lateral muscle ridges and continues to XXIV, from which point a faint neural line continues across the pygidium to end in the median marginal tooth which is narrower than the others. The other anal funnel has the three median ventral teeth smaller and the nerve cord terminating between two of them. No distinctly elevated tori or setigerous tubercles exist anterior to XIII, on which the latter are midway between dorsum and venter, And the ridge-like tori extend from just below them through the centre of the broad glandular zone nearly to the neural line. On succeeding segments they are even more prominent, and behind XV have shifted with the glandular zone to the posterior end of the segment. Toward the posterior end as the thin-walled portion of the segment diminishes ih length they become more central, and finally on XXI and XXII the 244 PROCEEDINGS OF THE ACADEMY OF [April. swollen glandular tori form the entire length of the segment and bear the uncini along the middle. Although XXII and XXIV are achsetous they are similarly constituted. The head and first four segments are very smooth, iridescent and of a uniform deep purple color above. The next three segments are smooth and dull yellowish throughout; those of the middle region have pale anterior glandular zones, the rest being yellowish and roughened, while posterior segments are pale throughout and thin-walled, except for the ventral longitudinal bands and posterior glandular zones. The peristomium and two preanal segments are achaetous. Somites II, III and IV have a single stout spine situated at about the end of the anterior one-third of the segment and about twice as far from the dorsal as from the ventral median line. A little above and in front of this is a short vertical series of capillary setae. On the remaining segments the setae form more prominent tufts arranged in vertical series midway on the sides of the body and in the middle of the glandu- lar area which lies on the anterior end of the somites as far as about XV, then for some segments on the posterior end, and finally, with the disappearance of the non-glandular region, occupies the entire seg- ment. Uncini occur on all segments from V to XXII inclusive. They are in all cases arranged in a single series in the same plane as the setae, and extend in a line, for a greater or less distance according to their number, on to the ventral surface. On the anterior segments they are slightly larger and fewer, the number counted in the type being 7 to 9 on V, 16 to 19 on X and 10 to 12 on XXII. The single ventral spines on II, III and IV are stout, fibrous and yellow, tapering to a slightly curved, blunt, hard point. On the same segments the notopodial tufts contain about six longer, stouter setae (fig. 44) with longitudinally striated stems tapering to stiff, straight, acute tips and provided with a broad and extensive wing on one side and a very narrow one on the other; also a corresponding number, arranged in a parallel row, of much shorter, more delicate setae (fig. 43) with nearly wingless stems tapering to a long hair-like and very delicate, flexible tip provided on each side with minute cilia-like hairs. The latter are so minute that they are visible vmder high powers only. Except for a slight increase in number the stouter setae remain unchanged throughout, but the more delicate ones become considerably modi- fied. By about the tenth 'segment they have become more numerous; some of them (fig. 43) are only moderately elongated (about as much or only slightly more than those of the preceding segments), but the lateral processes have become much larger, ensheathing and awn- 1906.] NATURAL SCIENCES OF PHILADELPHIA. 245 like, ancrmay be arranged spirally (fig, 43a) ; others are very greatly elongated, thread-like, openly spiral, tapered very gently to acute tips, and have the lateral processes reduced to minute appressed scales (fig. 36). The latter spring in a group from the ventral side of the bundle. Typical crochets (figs. 38 and 39) have rather slender, strongly curved stems with an asymmetrical nodulus near the middle, the head very little enlarged and provided with a stout, rather long, moderately acute, strongly hooked, and slightly recurved rostrum, above which is a prominent crest consisting of five, or sometimes of four, depressed, overlapping, diminishing teeth. There are no lateral teeth, but the striations end in several groups of conspicuous markings just below the principal teeth. The guard arises close beneath the rostrum, but separated from the latter by a distinct space. There is no distinct guard process or shoulder, but the guard arises as a distinct transverse plate, soon becoming divided into about twelve slender, tapering fila- ments, which spread around the end of the rostrum or overarch it. The stem, neck and posterior part of the head are strongly striated. Posteriorly the stems of the uncini increase in length. On somite V all of the uncini have the form shown in fig. 37. The stem is straighter, beak less hooked and the teeth of the crest fewer and more erect. The guard is rudimentary. The head and succeeding three or four segments are very smooth, iridescent and of a uniform deep purple or reddish-brown color above, not at all spotted or blotched ; the next three segments are smooth and yellowish throughout; those of the middle region have pale anterior glandular zones, the remaining parts being yellowish and roughened; while posterior segments are pale throughout and thin-walled, except in the position of the longitudinal muscles and glandular zones. The tubes are stout, thick-walled, hard but fragile structures, com- posed of fine sand grains, sponge spicules and bits of rock cemented together firmly and lined by a thin mucoid layer. Several tubes are sometimes coherent side by side. Fragments of this species occur in the collections from the Gulf of Georgia, Station 4,197, 31 to 90 fathoms, on a bottom of sticky green mud and fine sand, and Station 4,198, 157 to 230 fathoms, on a soft green mud bottom. The type locality is Station 4,227, in the vicinity of Naha Bay, Behm Canal, 62 to 65 fathoms, dark green mud and fine sand. Nicomache carinata is easily distinguished from A'^. per sonata John- son by the possession of two achaetous preanal segments, while the latter has but one. In this respect it resembles A'', lumbricales Malmg,, 246 PROCEEDINGS OF THE ACADEMY OF [April, but differs from that species in color, the form of the imcini and spines and the greater number of marginal divisions of the pygidium. Lumbriclymene pacifica sp. nov. Plate XII, figs. 40-42. The type and largest complete specimen, which is unduly elongated through maceration of the middle region, is 103 mm. long and 2 mm. in diameter. The head (composed of prostomium and peristomium) is slightly compressed laterally and very short, the length only very slightly exceeding the depth. It totally lacks a cephalic plate and has a nearly straight dorsal profile, meeting the vertical and convex anterior profile at nearly a right angle. There is just the faintest indication of a palpode and of the anterior end of the median ridge. About midway on the side of the head a longitudinal furrow extends from the furrow Yj to a point just over the mouth, and just anterior to the ends of these furrows are the nuchal organs — a pair of small round depres- sions. A few faint short furrows run from near the end of the longi- tudinal furrow toward the mouth, and behind the latter the longitudinal furrow is cut by a transverse furrow which is very strongly developed on the dorsal side but fades out ventrally. From the posterior end of the peristomium a short re-entering half-ring is cut off ventrally by a rather distinct furrow which ends dorsally at the longitudinal furrow. The mouth is a rather conspicuous transverse slit situated about opposite the middle of the head and bounded both anteriorly and posteriorly by rather prominent lips, the former of which is marked by a short but deep longitudinal furrow. Somite II is slightly shorter than the head. Behind it the segments increase regularly in length to IX at least, which is about three times as long as the head. Behind this several segments are probably still longer, but owing to their much softened state the exact length is uncertain. At the posterior end XVI is about as long as VI or VII and succeeding segments diminish rapidly, XIX and XX together only about equalling V in length. The two achsetous preanal segments are again much reduced. Except that at the extreme posterior end a few are provided with prominent tori, all somites appear to be perfectly cylindrical. Somites II, III and IV bear setae about the middle and just behind them is a faint furrow. The next five bear them near the anterior end just behind the glandular girdle. Posterior somites have their large swollen tori situated near their posterior ends and bearing the small tufts of setse above and moderately long lines of crochets below. The two 1906.] NATURAL SCIENCES OF PHILADELPHIA. 247 preanal segments have two tubercles on each side, one above the other, in the position of the tori, but without setse or uncini. In the type the pygidium is very short — less than the two preanal segments — and is terminated squarely by a simple slightly convex plate of broadly elliptical outline. Another specimen has the pygidium more extended and nearly twice as long (on the dorsal side) as the two preceding segments, and the terminal plate is oblique at about 45°. In both specimens the plate is margined by a just evident fold, con- tinuous except on the median ventral region. Just anterior to this margin dorsally is the anus. The color is well preserved at the anterior end in a sharply con- trasted pattern of reddish brown and white. On the head the former color occurs as follows : over the entire frontal surface, a half-ring on the dorsum of the posterior end of the prostomium, and a broad ring which occupies most of the peristomium and which is much more deeply colored dorsal to the longitudinal furrow than below it. The white areas are a broad band occupying the sides and dorsum of the prostomium above the mouth, ahd a very narrow ring on the anterior end of the peristomium. On several succeeding segments the arrange- ment is in a narrow anterior red ring, then a white ring occupying most of the glandular zone, and succeeding this an extensive red area occupying all of the rest of the segment. These colors are best devel- oped on the dorsum, and gradually fade until by about somite VII only a nearly uniform dull yellow prevails. The glandular areas are, however, always whiter and more opaque than the remainder of the segment. The peristomium and two preanal somites are achsetous, II to V bear small tufts of dorsal capillary bristles, and immediately below them two stout straight spines. Remaining somites have dorsal setse tufts, and below them ventral series of crochets occupying from one- eighth to one-sixth of the circumference of the segment. The anterior spines (fig. 40) are deep yellow, with opaque fibrous centres and straight blunt ends. In each tuft the setse number from fourteen to twenty in two ranks. On the proximal part of the exposed portion is a rather wide wing, obliquely striated and often frayed and fringed on the free margin. Just below this the shaft is usually con- stricted, and beyond it tapers to a capillary tip which is short on the anterior and very long and dehcate on the posterior segments, but always, so far as determined, quite devoid of hairs or awns. Crochets are always few in number, never more than nine to fourteen occurring in each torus. They (fig. 41) are similar on all segments and 248 PROCEEDINGS OF THE ACADEMY OF [April, have well-differentiated shoulder, neck and head. The latter (fig. 42) bears a stout, strongly hooked beak, above which are four stout profile teeth of diminishing size, flanked by a few small lateral teeth. The guard arises well below the beak and consists of about twelve coarse fibres which spread in front of and above the apex of the latter. The internal fibrous structure is strongly developed. Several tubes of this species occur in the collection and are interest- ing in structure. They measure 70 to 80 mm. long and 2.5 mm. in diameter, and occur either singly or attached in groups or to foreign bodies. Their walls are thin but hard and very brittle and are com- posed chiefly of small sand grains and sponge spicules, the latter of which are so attached that their pointed ends project freely toward the mouth of the tube. Various kinds of foraminifera and other foreign bodies are attached to the tubes, which are always dark- colored at the distal end and clean and pale elsewhere. Two complete worms and a fragment together with four or five tubes were taken at Station 4,264, off Freshwater Bay, Chatham Strait, 282 to 293 fathoms, bottom of green mud. Sabellaria cementarium sp. nov. Plate XII, figs. 45-51. The fine species which represents the genus Sahellaria along the Pacific coast from Washington to Alaska is represented by a number of specimens, but unfortunately only one of these is complete, the others having lost the posterior end either through an attempt to remove the living worm from their tubes or by maceration in the tubes. The type and only perfect specimen is 81 mm. long, of which the very slender fecal tube contributes 28 mm. The operculum has a diameter of 4.5 mm., the thorax a width of 6 mm., from which thick- ness the abdomen tapers regularly to about 2 mm. at the posterior end and then suddenly contracts to the 1 to 1.5 mm. of the fecal tube. Counting the peristomium there are five thoracic segments exhibit- ing three distinct types of setation, then follow forty ordinary ab- dominal segments and about forty-six segments in the reduced fecal tube; finally the pygidium is a tubular structure 1.5 mm. long and .6 mm. in diameter with its posterior end serrate with about twenty minute teeth. The prostomium is minute and completely concealed beneath the enormously developed peristomium. The small slit-like mouth is enclosed between a pair of closely appressed longitudinal folds, bovmded laterally and somewhat enfolded ventrally by the enlarged palps, which are completely connate with the peristomium. The peristomium is greatly enlarged and in the type measures 6.5 1906.] NATURAL SCIENCES OF PHILADELPHIA, 249 mm. long and 4.5 mm. wide. Dorsally it is completely closed, over- arching and concealing the prostomium and mouth. Together with the palpi which it bears below it forms a broad flat plate, of which the palps form the margins and are rolled inward and nearly touch ven- trally. At its distal end the peristomium bears the operculum — a circular or somewhat elliptical disk directed slightly dorsad and com- posed of three whorls of stout, stiff, hard spines or peristomial setse, so fitted together that they form a flexible and at the same time close- fitting and impenetrable plug to the tube. There are some indications in one specimen that this region may be regenerated when lost. The paleoli of the three rows differ in form and number as indicated below. Just external to and below the operculum is a circle of rather promi- nent conical opercular papillse, which probably represent the much sub- divided dorsal cirri of the peristomium. They are clearly divided in two symmetrical halves like the opercular paleoli, and number 16 to 20 on each side. The palps may be opened from the ventral side and spread, together with the peristomium, as a flattened plate deeply pigmented below and bearing the branchial folds on the lateral thirds. Anteriorly the two palps are conjoined in the operculum, the branchise also meeting in the middle line below and behind the operculum. From 12 to 18 of these l^ranchial folds or ridges occur on each side, the usual number in full-grown specimens being 18 pairs. They are prominent ridges running transversely across the free ventral margin of the palps and diminishing in size from behind forward. Owing to the medial bend- ing of the anterior end of the palps, the anterior six pairs of branchiae are arranged in the form of an arch and several of them lie in a nearly antero-posterior plane. Each gill consists of a stiff plate or ridge, with a serrate free margin behind which the filaments are borne. Except the very last, which usually bears but 10 filaments, the pos- terior gill ridges have about 20, the number decreasing regularly to the most anterior. All gill filaments are very slender and thread-like and the longest posterior ones equal the width of the peristomium. At the base of the branchial region is a pair of short rounded lobes having much the aspect of the branchial plates, but shorter and thicker than they and coming in contact across the mouth slit. Continuing from this dorsally is a sloping ridge bearing a pointed conical cirrus longer than the rounded lobe, and above this again a small tubercle carrying a fan-shaped tuft of slender setse. Somite II is a short, simple ring partly fused with I and III ventrally to form the first ventral gland plate. Laterally it bears three conical 250 PROCEEDINGS OF THE ACADEMY OF [April, cirri, the ventral about as long as the segment, the middle about two- thirds as long, and the dorsal more than twice as long, slender distally and resembling the branchiae, with which it stands as the first of a series. The remaining three thoracic segments are a little longer and IV and V bear distinct ventral glandular plates. Their neuropodial setse are stouter than those on I and II, but have no associated cirri and arise from the anterior margin of the segment on a level about mid- way between the neuropodial tufts of I and II. ' Each of these seg- ments bears a dorsal cirrus or gill similar to that on II and in line with it, leaving a broad, naked area on the dorsum. Unlike II these somites bear no middle cirrus, but in place of it a prominent vertically elongated notopodial tubercle bearing a vertical series of large, coarse, paddle- shaped setse. Abdominal segments are somewhat flattened and only obscurely separated. As indicated above they decrease gradually in both length and diameter to the caudal appendage. Dorsally they present a somewhat arched area between the rows of gills and ventrally are excavated by a deep groove for the fecal tube, which is held in place by the tufts of slender neuropodial setae crossing from side to side beneath it. The body walls are exceedingly thin and delicate in the dorsal median area and permit the intestine to show through distinctly ; the sides and ventral parts are decidedly thicker and more muscular. The abdominal gills have the same general form and position as the thoracic but are more perfect in structure, having larger blood-vessels, thinner walls and more numerous and prominent transverse cihated ridges along their medial aspect. The first five or six are also much longer and have a length equal to the width of the interbranchial area. From this point they decrease in size very gradually and are totally absent from the last five or six segments. On the first abdominal segment the parapodia are very extensively developed and the notopodial uncinigerous tori occupy the entire side of the segment from the branchia nearly to the ventral median line. Ventrally they become higher and terminate in a small free lobe, from beneath which arises a delicate ventral cirrus, and below this again the small neuropodial tubercle with its tuft of setse. The tori decrease in length chiefly from the dorsal end, at first very slowly, then rapidly and at the same time become more sharply defined as distinct, thin, outstanding lobes bearing the uncini on their margins. Toward the posterior end they become merely small prominent pro- 1906.] NATURAL SCIENCES OF PHILADELPHIA. 251 cesses with somewhat constricted stalks standing midway between the branchiae and the neuropodial setae tuft. After the first abdominal segment the neuropodial cirrus quickly becomes reduced and in four or five segments has become quite rudimentary or totally absent. The caudal or fecal tube has already been partly described. It is very delicate and thin-walled and bears no trace whatever of para- podia. Along its ventral side, however, what appear to be nerve ganglia can be distinctly seen and counted through the body wall, and delicate lines running in pairs from their neighborhood dorsad and cephalad have the same metameric arrangement. The type specimen is richly colored. The outer whorl of paleoli are a warm golden, the inner a golden brown. The thoracic region is a rich sienna brown, especially deep on the dorsal interbranchial region of the peristomium as well as of succeeding segments, and on the sides of the latter and the first ventral plate. The ventral post-branchial lobe of I and the dorsal branchiae are also well colored, but other cirri are pale. The gill filaments are pale purple, each with a deep brown basal spot. This specimen is a male, and the abdomen is colored pale cream from the contained sperm. Except for delicate lines formed of minute dots of reddish browm, which begin at the ventral setae tufts and then pass dorsad along the anterior margin of the segment and in most cases continue on to the antero-lateral margin of the gill, there is no pigment in this region. The intestine is filled with a greenish-gray matter that colors the fecal tube. Most of the other specimens are paler, but one has the anterior brown parts of a deep chocolate. The genital products escape by means of a pair of openings through the body walls behind the parapodia of each segment and in several cases from large masses within the tubes. The opercular paleoli vary from bright yellow to the more usual deep golden brown. They are all very stout, hard and rigid, and of peculiar irregular forms difficult of accurate description. All have slender, elongated stems, smooth superficially but striated longi- tudinally at the core, bearing very hea^•y and strong external blades, the great part of whose surface is roughened by numerous fine wavy parallel ridges w^hich are slightly imbricated with their edges directed outward. The markings are not shown in the figures. The outer whorl of normal opercula contains from 39 to 64, according to the size of the worm, between 50 and 60 being the most usual num- ber. They are arranged in two symmetrical groups, though there appears to be a decided tendency for the right half to include one or two more than the left. Exclusive of the stem they (fig. 45o) consist '252 PROCEEDINGS OF THE ACADEMY OF [April, of a massive irregularly twisted base, from which arises a tongue-like piece the end of which is strongly flattened at right angles to the great- •est thickness of the base and terminates in a fringed tip, from the central cleft of which arises a slender, densely hairy process (fig. 46). These paleoli are so arranged that they present a spreading rim, in which their twisted form causes the basal portions to be somewhat imbricated from below dorsalwards. The middle paleoli number from 14 to 25, usually about 20, in sym- metrical halves. They (fig. 45m) also have a thickened base, from which arises a very prominent upright spine, gracefully curved and tapered to an acute tip. The inner circle of paleoli contains from 14 to 22, but in nearly all cases they equal and alternate with those of the middle row. They (fig. 45i) have abruptly widened cleaver-shaped ends bearing a knife-like edge and strengthened on the opposite side by a thickened rim which ends in a short triangular beak. They are dis- posed somewhat obliquely to the middle line and converge ventrally. Somite II bears a small tuft of strongly doubly-fringed, feather-like neuropodial setae (figs. 49 and 50). On III the neuropodial setse are partly of a short, more brush-like form and partly very small and of a slender, slightly fringed form. On the remaining thoracic segments they all become more sparsely fringed. The notopodials (fig. 47) on III, IV and V form a single vertical series of stout setse, shaped like an Indian paddle with the end much split and frayed. At the base of each of these is a minute spatulate hairy setse, with the end entire and somewhat produced in the middle, and the blade bent flatways. Abdominal setse are all very long and slender, but owing to different states of development appear of unequal length. Their shafts (fig. 48) are provided with numerous whorls of hairs united at the base into collars and much produced on one side. The uncini (fig. 51) are very numerous and vary from 100 to 200 or even more on a torus. They are nearly colorless, delicate and incon- spicuous, and consist of narrow elongated bodies provided with from 7 to 9 (seldom the former) very acute, appressed, overlapping teeth which are arranged in one series and diminish in size toward both ends. Each bears a stiff and rather brittle tendon at each end, the upper one being short and expanded distally into a small plate, while the lower bifurcates immediately into a shorter, irregularly thickened, more ven- tral branch and a slender uniform branch greatly elongated to many times the length of the uncini. This species is probably rather common and occurs in the collections from the following stations: 4,220 (type). Admiralty Inlet, near Port 1906.] NATURAL SCIENCES OF PHILADELPHIA. 253 Townsend, Washington, 16-31 fathoms, green mud, sand and broken shells; 4,247, Prince of Wales Island, 89-114 fathoms, green mud with sand and broken shells; 4,274 (cotypes), Kadiak Island, 35-41 fathoms, green mud and fine sand. Samytha bioculata sp. nov. Plate XII, figs. 52, 53. The anterior prostomial lobe is quadrate, broader than long, its anterior border wider than the posterior, slightly concave and its lateral angles somewhat produced. Immediately behind and separ- ated from it by a distinct transverse furrow is a second small quadrate lobe, with a rather prominent eye or close aggregation of several pigment specks at each antero-lateral angle. Behind this lobe the peristomium forms a broad, smooth, slightly convex area reaching to the branchiae. Ventrally there is a very broad truncate under lip with a glandular margin. In the two known specimens only four to SIX very short clavate tentacles exist on each side. The considerably enlarged peristomium projects ventrally as a broad lobe enveloping the lower lip an^l is marked by a narrow trans- verse line of gland cells. Elsewhere it is a simple smooth convex ring. The second somite is about two-thirds as long and marked by a broader glandular half-ring. The third (first setigerous) somite is very short, but succeeding ones increase in length rapidly and by about VIII equal the peristomium. All of the thoracic segments, of which there are nineteen, the last seventeen of which are setigerous, are somewhat muscularly thickened on the ventral half and provided with a narrow presetal half-girdle of glands. The entire dorsum between the setse tufts is thin-walled and smooth throughout. Fourteen segments in the type and larger specimen, and thirteen in the smaller cotype, form the abdomen. In the former this region comprises about one-third of the entire length ; in the cotype not more than one-fourth. Owing to the projection of the parapodia from the angles this region is decidedly quadrate, but somewhat arched dorsally and marked ventrally by a narrow neural groove. The somites dimin- ish regularly in all dimensions from before backward and the entire region tapers to the pygidium, which forms a ring surrounding the large anus and bears a pair of slender, prominently projecting cirri about equal to half the diameter of the body at the posterior end. The four pairs of branchiae are so much crowded antero-posteriorly that they appear to form a single series extending quite across the dorsal area of II and III., Closer study shows that two are anterior and slightly more lateral and two posterior and more median. They are all similar, slightly flattened, slender and elongated, their length. 254 PROCEEDINGS OF THE ACADEMY OF [April, being about three times the diameter of the region of the body from which they spring. Setae begin on III as a pair of minute tufts arising from an elevated position just beneath the gills. The tufts on IV are also small and placed nearly as high. On succeeding somites the setae become larger and more numerous and the tufts gradually assume a low station on the sides of the body as the abdomen is approached. They continue to XIX. Uncinial tori begin on VI, or the fourth setigerous somite. Through- out the thoracic region they are low, but freely projecting, compressed folds which arise immediately ventral and slightly posterior to the setae tufts and end ventrally in a freely hanging lobe or process. On abdominal segments the notopodia are distinctly separated from the neuropodia and project as small papillae from the dorsal angles of the body. The neuropodia are small, compressed lappets constricted at the base and bearing the' uncini along the free margin; apparently they lack cirri. Setae are all of one kind, slender, tapered and narrowly winged on both sides of the free portion nearly to the very acute tip. The uncini form a close single rank in which they are quite numerous (108 on X). From dorsal to ventral end of each series they decrease in length. They (fig. 52) are roughly triangular with a nearly square but sometimes projecting upper ligament process, above which the tooth-bearing margin rises but slightly. The inferior process varies in form, but is usually more or less incurved. Those of the type specimen almost constantly bear four long, slender, acute, overlapping teeth, the ven- tralmost of which is the stoutest, the two middle the longest and the upper the most slender. On the cotype a fifth smaller and sometimes minute tooth exists at the upper end of nearly all uncini (fig. 52a). Abdominal uncini (fig. 53) have the same form, but are much fewer in number. This species is founded on two much macerated specimens, of which the larger and type is 30 mm. long and filled with large ova. The smaller example is 16 mm. long. A nearly complete tube is 65 mm. long. Its basal third is soft, thin-walled and membranous. Beyond this the walls gradually thicken by the accumulation of silty material until they attain a diameter of about 5 mm. This entire region of the tube is supported and protected by siliceous sponge spicules, the ends of which project in all directions and produce a very firm and bristling structure. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 255 Both specimens come from Station 4,197, Gulf of Georgia, 31 to 90 fathoms, sticky green mud and fine sand. Amphicteis scaphobranchiata sp. nov. Plate XII, figs. 54-61. A single well-preserved specimen represents this noteworthy species. The length without branchiae is 32 mm., the branchiae 7 mm., and the maximum '\\ddth in the anterior part of the thorax 2.8 mm. The anterior lobe of the prostomium (fig. 54) is almost completely divided by a longitudinal fiu-row into a pair of somewhat slender tentacle-like divisions which are somewhat divergent anteriorly. A narrow transverse welt passes across the base of this region and is partly concealed by the much larger and more prominent sensory folds which lie behind it. These pass from the lateral borders of the head nearly transversely to the middle line, where they meet in a wide angle. Laterally they are continuous with the lateral lobes or folds which join the frontal lobe beneath. The tentacular membrane lies below the frontal and lateral lobes and is partly embraced by the peristomium. It bears from twelve to fifteen tentacles on each side, the lateral ones being very short and the middle ones as much as two- thirds the width of the thorax. All and especially the shorter ones are more or less clavate. The prostomium passes without any clear demarkation into the peristomium, which reaches as a smooth unbroken convex surface to the branchial segment. Dorsally it is remarkable for its extent, smoothness and absence of furrows. Its length and breadth are each equal to about two-thirds of the width of the branchial segment. In the middle line it reaches to the interbranchial shield, but its postero- lateral angles are cut off by the paleolar tubercles and lateral portions of somite II. Ventrally the peristomium forms a large, broadly trun- cated lower lip, slightly inserted into II and embraced laterally by the paleolar tubercles. The second somite is more than half as long ventrally as the peristo- mium, but except for the narrow strip extending dorsally beneath the branchiae it ends abruptly at the enlarged paleolar tubercles. Somite III is very short and IV slightly longer, the two combined just equalling the length of II. Behind this region the segments again diminish in both length and diameter to the pygidium. With the exception of the last three or four, the thoracic segments are ill defined on the ven- tral and not at all on the dorsal side. Up to the level of the setae tufts, where a lateral shelf -like ridge is formed, the body walls are rather thick and firm, with slight anterior glandular bands and obscure ventral plates. Dorsal to the setae the body for the entire length is 256 PROCEEDINGS OF THE ACADEMY OF [April, perfectly smooth and iinsegmented. The last three or four thoracic segments are transitional in character to the abdominal. The latter are more distinctly differentiated, owing in part to their more promi- nent parapodia and in part to the greater depth of the ventral furrows. The last few segments are separated by obvious furrows even on the dorsal side. Owing to its being crushed the pygidium cannot be described further than to refer to the pair of prominent, slender and stiff cirri which it bears. A distinct neural groove extends through- out the abdominal region and even on to several of the posterior thora- cic segments. A remarkable smoothness and peculiar dull iridescence characterizes the entire cuticle of this worm. If normally formed in this specimen the branchise are highly char- acteristic. All four pairs are large and stout and arise in the usual manner, two pairs anterior and more lateral from II and III and two more posterior and median from IV, those of each side being com- pletely coalesced at the base. A small shield-shaped area separates them in the middle line, but otherwise they cover the entire width of the dorsum. Viewed from above they cover and conceal the entire head and curve downward in front of the prostomium and beneath the tentacles. All are very broad and flat at the base, especially the two inner ones on each side. In the case of three of them the outer one- third is less flattened and tapers to a blunt point. The anterior median of both sides, however, expands into a broad flat plate (fig. 56) some- what lobed on the margins and terminated by a tapering process bent sharply on itself into a hook. This process is complete on one side only, having been broken off of the other. Probably all of the branchise are straight in life, but they are peculiarly stiff and rigid and their cuticular covering is thick and marked by fine transverse strige, but otherwise smooth. Thoracic parapodia consist of simple setigerous tubercles arising from the lateral muscular shelf and short auriculate tori with both dorsal and ventral angles produced somewhat freely. Abdominal parapodia (fig. 57) consist of prominently projecting hatchet-shaped ventral tori, each bearing a minute cirriform process, and sharply bent, clavate dorsal cirri which replace the setigerous tubercles of the thoracic region. Parapodia of successive somites are united by a beaded muscular ridge which passes along the side of the abdomen between the torus and cirrus. The worm consists of thirty-two seg- ments, of which II bears the paleoh. III to XIX ordinary setse, and VI to XXXII uncini. The paleoli (fig. 58) are of a clear, bright glistening straw color and 1906.] NATURAL SCIENCES OF PHILADELPHIA. 257 are arranged in a half tubular figure at the base, from which they diverge and spread distally in a fan-shaped fascicle. In each tuft the shortest one (at the inner end of the posterior limb) is about twice as stout as the longest. Distally they all taper regularly and rather rapidly to very long, slender, acute and stiff tips. Except for these tips they are striated conspicuously in both the longitudinal and transverse direction. The ordinary set* (fig. 59) are few in number and form rather small close tufts. They are apparently similar on all segments, being rather slender with tapering stems and acute, tapering tips, striated some- what obliquely, and provided on one side with a mng of moderate width marked with distinct oblique striae. The uncini are numerous on the thoracic segments (135 on X). They have the form shown in figs. 60 and 60a, being roughly tri- angular, with large bases, a prominent superior ligament process, a projecting rounded lower angle and long tooth-bearing border with five or six strong acute teeth, the lowermost of which is peculiarly bevelled and covers a short inferior ligament process which nearly touches its lower surface. Abdominal uncini (fig. 61) are fewer in number (78 on XXV), much smaller, and formed quite differently. They have small bases and high outstanding tooth-bearing rami, and apparently lack the inferior ligament process. Several thick-walled mud tubes in the same bottle were probably fashioned by this species. Type locality. Queen Charlotte Sound, off Fort Rupert, Vancouver Island, B. C, Station 4,201, 138 to 145 fathoms, soft green mud, sand and broken shells. Chone gracilis sp. nov. Plate XII. figs. 62-66. A very pretty small species resembling Chone duneri Malmgren in proportions but differing from that species in the short, and broadly winged, barbless tips of the branchiae. From C. teres Bush this species is easily distinguished by its much longer branchiae, small number of abdominal segments and the more elongated form of the abdominal uncini. The type specimen has a total length of 32 mm., of which the branchiae make 12 mm. and the thorax 5 mm., the diameter for nearly the entire length being about 2 mm. The branchial bases are simple and very low, being completely con- cealed by the high collar. Ten pairs of delicate branchiae extend equally to a length of more than one-half of the body. The stems are united by a very delicate membrane for three-fifths of their length and 17 258 PROCEEDINGS OF THE ACADEMY OF [April, the remainder is broadly winged on both sides. The barbs are very delicate and are absent from an acute foliaceous tip about 2 mm. long. Eyes are totally wanting. A collar of the form typical of the genus rises to a height of two and one-half times the length of the segment which follows it. The dorsal ends are refolded into the dorsal sinus and a slight notch occurs in a median ventral position ; otherwise it is quite simple. The very small tentacles are shorter than the diameter of the body. Nine segments, 8 of which are setigerous, form the thorax and 51 segments the abdomen. For nearly its entire length the body is cylindrical and of almost uniform diameter, but in the posterior fourth becomes somewhat broader and depressed before tapering rapidly to the caudal end. No distinct glandular ventral plates are developed, but all of the thoracic and the anterior three or four abdominal segments are completely encircled by a thick glandular layer. The thoracic and a few anterior abdominal segments are strongly biannulate, the middle abdominal are simple and half as long as wide, while the posterior are much shorter and more crowded. The fecal groove is distinct on the caudal fourth of the abdomen, and conspicuous on a few anterior abdominal segments, also where it passes obliquely around the right side of the first abdominal segment and on the dorsum of the thorax ; elsewhere it is faint or absent. A few small ova are present in the coelom. Every portion of this worm is white, without a trace of pigment anywhere. All setse tufts are short but rather prominent ; uncinigerous tori are also short and nearly uniform, diminishing only slowly and regularly from the first to the last. The setse and uncini are all delicate and colorless. The collar fascicle consists of a small number of narrowly winged, acute, capillary setse. Remaining thoracic somites have a larger number of partly capillary, partly spatulate setse placed be- tween two small lappet-like processes. The former consist of a dorsal row of acute tapering setae with one moderately wide and one just perceptible wing and a small number of very small and inconspicuous crooked setse (fig. 62) with rudimentary wings found below the spatu- lated setse. The spatulate setse (fig. 64) are arranged in two rows of about ten each; they have delicate, striated, usually symmetrical, obovate blades, and long very slender mucronate processes. The abdominal setse (fig. 63) are also in one series, few in number, rather coarser than the thoracic and with the two wings narrow and about equally developed. Thoracic uncini contain a single series of erect crochets or hooked setae (fig. 65) with striated stems and the slightly enlarged head thrown back, a large coarsely divided crest and stout> 1906.] NATURAL SCIENCES OF PHILADELPHIA. 259 blunt beak. On somite V there are 13 in a torus, on somite VIII 11. The abdominal tori contain uncini of the form shown in fig. 66, with quadrate bases and heads of much the form of the thoracic crochets but strongly recurved on the base. They are more numerous than the thoracic crochets, somite XXI containing 17. The tube is nearly transparent, little tortuous and almost free from sand. The single specimen comes from Station 4,274, Alitak Bay, Kadiak Island, 35 to 41 fathoms, on a bottom of gree» mud and fine sand. A smaller specimen was taken at Station 4,253, Stephens Passage, 131 to 188 fathoms, rocks and broken shells. Explanation of Plates X, XI and XII. Plate X. — Notophyllum imbricatum — figs. 1-3. Fig. 1. — Dorsal view of head of type. X 13. Fig. 2. — Outline of parapodium of middle region, seen from in front. X 13. Fig. 3. — Distal end of a neuropodial seta of average length. X 600. Eulcdia quadrioculata — figs. 4-6. Fig. 4. — Dorsal view of anterior end of type. X 24. Fig. 5. — Posterior xiew of parapodium from somite XXV. x 32. Fig. 6. — End of an average seta from somite X. X 600. Eulcdia longicomuta — figs. 7 and 8. Fig. 7. — Posterior view of parapodium from somite XXV of type. X 32. Fig. 8. — End of a seta from somite X. X 600. Some of the setae have even longer appendages. Pionosyllis magtiifica — figs. 9-11. Fig. 9. — Dorsal view of anterior end of type. X 32. Fig. 10. — Anterior \-iew of parapodium of somite XXV. X 32. Fig. 11. — End of a seta from the middle of a fascicle on XXV. X 600. Stauronereis annulatus — figs. 12 and 13. Fig. 12. — Anterior end of cotype from above. X 82. Fig. 13. — A parapodium from somite XXV, anterior \'iew. X 82. Brada pilosa — figs. 14-17. Fig. 14.— A medium-sized extended papilla from the dorsum of the middle region, showing ring of adhering silt. X 113. Fig. 15. — An extended papilla from the ventral surface. X 113. Fig. 16. — A portion of the middle of a notopodial seta from XV. X 250. Fig. 17. — Exposed portion of neuropodial seta from XV. X 98. a, portion of middle of the same. X 250. Plate XI. — Stauronereis annidatus — figs. 18-22. Fig. 18. — Distal portion of jaw from below. X 130. Fig. 19. — Forked seta from middle of notopodial fascicle of somite XXV ; a, another of the same slightly rotated and foreshortened. X 800. Fig. 20. — Portion of the middle of a capillarv neuropodial seta from somite XXV. X 800. Fig. 21. — A short-bladed compound neuropodial seta from the middle of the fascicle of XXV. X 800. Fig. 22. — A long-bladed compound seta from the same somite. X 800. Travisia pupa — fig. 23. Fig. 23. — A small portion from near the distal end of a notopodial seta from the middle region. X 800. Notomastus giganteus — figs. 24 and 25. Fig. 24. — An entire crochet from the middle of the body. X 333. 260 PROCEEDINGS OF THE ACADEMY OF [April. P'ig. 25. — Profile \'iew, and a face view, of the end of one of the same. X 800. Maldane similis — figs. 26-30. Fig. 26.— Slightly w-inged seta from II. X 333. Fig. 27. — Hispid capillary seta from somite XV. X 480. Fig. 28.— End of crochet from somite III. X 480. Fig. 29.— Entire crochet from X. X 110. Fig. 30.— Distal end of the same. X 480. Maldanella rohusta — figs. 31 and 32. Fig. 31. — Two entire crochets from somite XV. X 110. Fig. 32.— The end of one of the same. X 480. Clymenella tentaculata — figs. 33-35. Fig. 33. — Middle crochet from somite II, entire. X 110. Fig. 34.— End of the same. X 480. Fig. 35.— End of a middle crochet from V. X 480. Nicomache coronata — figs. 36-39. Fig. 36. — Small portion of a fiber seta froin somite X. X 800. Fig. 37. — End of a crochet from somite V. X 480. Fig. 38. — An entire crochet from somite XV. X 110. Fig. 39.— End of the same. X 480. Plate XII. — Lumbrichymene pacifica — figs. 40-42. Fig. 40.— End of a spine from II. X 98. Fig. 41. — An entire crochet from somite X. X 83. Fig. 42.— End of the same. X 440. Nicomache coronata — figs. 43 and 44. Fig. 43. — SmaU seta from somite X. X 360 ; a, small portion of the same. X 600. Fig. 44. — Large seta from somite X. X 360. Sabellaria cementarium — figs. 45-51. Fig. 45. — Three spines forming a middle segment of the operculum, shown in profile in as nearly as possible their natural relations: i inner, m middle, and o outer spines. X 32. Fig. 46. — A face \iew of the end of an outer spine. X 32. Fig. 47. — Middle notopodial seta from somite III. X 32. Fig. 48. — Portion of an ensheathed capillary notopodial seta from somite XX. X 440. Fig. 49. — Bipinniform neuropodial seta from II. X 98. Fig. 50. — A portion of the middle of the same. X 440. Fig. 51. — Middle abdominal uncinus. X 440. Samytha bioculata — figs. 52 and 53. Fig. 52. — Uncinus from somite X of type. X 600; 52, the same from somite X of the cotype, somewhat foreshortened. X 600. Fig. 53.— Uncinus from somite XXV. X 600. Arnphicteis scaphobranchiata — figs. 54-61. Fig. 54. — Dorsal aspect of anterior end of type, with branchiae cut away and separated to show the prostomium. X 9. Fig. 55. — Ventral aspect of the same with the branchise in place. X 9. Fig. 56. — Distal portion of one of the anterior middle branehise. X 9. Fig. 57. — One of the abdominal parapodia. X 24. Fig. 58. — A paleolus from the middle of the fasciculus. X 83. Fig. 59.— A seta from somite X. X 250. Fig. 60. — Two forms of uncini from somite XII, the five-toothed one i somewhat foreshortened. X 600. Fig. 61.— An uncinus from XXV. X 600. Chone gracilis — figs. 62-66. Fig. 62. — A small bent seta from somite V. X 360. Fig. 63.— A winged seta from XXI. X 250. Fig. 64. — A spatulate and mucronate seta from VI. X 360. Fig. 65.— A crochet from VI. X 360. Fig. 66.— An uncinus from XXI. X 360. PROC. ACAD. NAT. SCI. PHILA. 1906. / ■ i i ■ : ! i ■ p ■i'h -a ■'- r i--;-P P -t od pi pe CO STEVENS AND BORING. PLANARIA MORGANI N. SP. PROC. ACAD. NAT. SCI. PHILA. 1906. BURNETT SMITH. PHYLOGENY OF THE RACES OF VOLUTILITHES PETROSUS. PROC. ACAD. NAT. SCI. PHILA. 1906. /,*??•//* Mi^ *^W OGCOCEPHALUS RADIATUS (mitchill). PROC. ACAD. NAT. SCI. PHILA. 1906. MALACLEMMYS LITTORALIS RHIZOPHORARUM FOWLER. PROC. ACAD. NAT. SCI. PHILA. 1906. PLATE V. ..,s,a^SSi&i. ^ J i 2 .5 4 5 ^^^ 6^ r ^b L ^ 1 #i^ ^^^^^^QUH^^^^ ' * m^M 8 1^ 9 Ifi^ ^H P^JM .^^HSBh^ %^ i7 20 ^"^^ PILSBRY AND FERRISS. MOLLUSCA OF SOUTHWESTERN STATES. PROC. ACAD. NAT. SCI. PHILA. 1906. PILSBRY AND FERRISS. MOLLUSCA OF SOUTHWESTERN STATES. PROC. ACAD. NAT. SCI. PHILA. 1906. PILSBRY AND FERRISS. MOLLUSCA OF SOUTHWESTERN STATES. P^OC. ACAD. NAT. SCI. PHILA. 1906. PLATE VIII &^c1 PILSBRY AND FERRISS. MOLLUSCA OF SOUTHWESTERN STATES. PROC. ACAD. NAT. SCI. PHILA. 1906. ^X 11 »1 :) PILSBRY AND FERRISS. MOLLUSCA OF SOUTHWESTERN STATES. PROC. ACAD. NAT. SCI. PHILA. 1906. MOORE. NORTH PACIFIC POLYCH/ETA. PROC. ACAD. NAT. SCI. PHILA. 1900. MOORE. NORTH PACIFIC POLYCH^TA. PROC, ACAD. NAT. SCI. PHILA. 190G. PLATE XII. 55 65 MOORE. NORTH PACIFIC POLYCHA^TA. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 261 May 1. Mr. Arthur Erwin Brown, Vice-President, in the Chair. Forty-six persons present. The PiibHcation Committee reported that a paper entitled "Addi- tional New Species of Polychseta from the North Pacific," by J. Percy Moore, had been offered for publication (April 19, 1906). The deaths of Clarence H. Clark, a member, March 13, and of Lionel S. Beale, a correspondent, March 28, were announced. Dr. Benjamin Sharp made a communication on pearls and pearl fisheries. (No abstract.) May 15. The President, Samuel G. Dixon, M.D., in the Chair. Thirty-four persons present. The Publication Committee reported that a paper entitled ' 'Records and Descriptions of Non-Saltatorial Orthoptera from British Guiana," by James A. G. Rehn, had been offered for pubhcation (May 7, 1906). The death of Carl Y. Yischer, M.D., a member, May 14, was announced. Henry Leffman, M.D., made a communication on the newer views regarding the nature of matter and their bearings on biology. (N abstract.) Mr. Walter F. Herzberg was elected a member. The following were ordered to be printed : 18 262 PROCEEDINGS OF THE ACADEMY OF [May, RECORDS AND DESCRIPTIONS OF NON-SALTATORIAL ORTHOPTERA FROM BRITISH GUIANA. BY JAMES A. G. REHN. The records given in the following pages are all from a collection of Orthoptera made in 1901 at Demerara, British Guiana, by R, J. Crew, and for the privilege of studying which the author is indebted to Prof. Lawrence Bruner. The series contained sixty-one specimens, repre- senting thirty-four species, of which six are new. The types are in Prof. Bruner's collection. BLATTID^. ANAPLECTA Burmeister. Anaplecta pulchella n. sp. Type: 9 ; Demerara, British Guiana, 1901. (R. J. Crew.) Closely allied to A. mexicana Saussure and dohrniana Saussure and Zehntner, but differing from mexicana in the narrowly incised supra- anal plate, the more angulate proximal portion of the appendicular field of the wings, and in some details of the coloration. From dohr- niana it differs in the wings, in the coloration and somewhat in the shape, but agrees in the shape of the supra-anal plate. Size rather large (for the genus); form rather flattened and sub- elliptical; surface glabrous. Head about completely hidden under the pronotum; interspace between the eyes almost twice the depth of one of the eyes; antennse sparsely haired. Pronotum transversely ovate, the cephalic margin regularly arcuate, caudal margin very slightly arcuate, lateral margins rounded obtuse-angulate ; disk somewhat deflected laterad. Tegmina exceeding the apex of the abdomen, sub- lanceolate with the costal margin slightly arcuate except proximad where it is rather strongly arcuate, sutural margin straight, apex acute; basal field small, sub-coriaceous; anal field elongate sub-pyriform in shape and containing four veins; costal veins thirteen in number, the interspaces between the distal ones supplied with short spurious veins which do not coalesce with the discoidal vein; median vein with four rami of which the second (enumerating proximo-distad) is bifurcate. Wings with the appendicular field of medium size, sub-rectangulate with the margins rounded and the proximal angle decidedly obtuse : 1906.] NATURAL SCIENCES OF PHILADELPHIA. 263 clavate costal veins seven in number, the clavation being very large and short, in fact decided inflations, non-clavate costals five in number, distad of the clavate ones; medio-discoidal area with quadrate inter- spaces; axillary vein with one principal branch diverging caudad, which latter has two rami. Supra-anal plate transverse, rounded with a slightly median emargination; subgenital plate not reaching to the apex of the supra-anal plate, inflated and with the caudal section- compressed and narrowly divided, reminding one of the structure of the valves of the Blattinae. Femora with their ventral margins spined. General color tawny above; head ochraceous with a pattern of four transverse and one short longitudinal occipital bars of seal brown, eyes seal brown, antennae ochre yellow proximad and dull brownish distad. Pronotum with the disk seal brown and the lateral portions hyaline, the disk broad caudad narrowing cephalad. Tegmina of the general color with a spot in the anal field, a more prominent one at the base of the discoidal field and another rounded one near the middle of the discoidal field raw umber. Wings very slightly washed with olive, except the sections adjacent to the proximal portion of the costal margin and the caudal section of the inserted angle of the appendicular field, which are washed with ochre yellow. Ventral surface and limbs pale clay color, margined, lined and touched with mummy brown. Measurements. Length of body, 7.6 mm. Length of pronotum, 2.2 " Greatest width of pronotum, 3 " Length of tegmen, 8 '' Greatest width of tegmen, 2.5 '' The type is the only specimen seen. Anaplecta fulgida Saussure. Two males, two females. Two of the above specimens belong to the form of the species with the disk of the pronotum dark, both sexes being represented. The tegmina in these dark specimens are somewhat darker than in the other individuals. This species has been recorded from Tabasco, Mexico, Guatemala and Piedras Negras, Costa Rica, this constituting the first South American record. PSEUDOMOPS Serville. Pseudomops crinicornis (Burmeister). One male, one female. While these specimens would, by their uniform black antennae, fall 264 PROCEEDINGS OF THE ACADEMY OF [May, under luctuosa (Saussure), the original description of crinicornis con- tains nothing at variance with the material examined, no mention being made of the antennae, while the median whitish spots on the tegmina are distinctly specified. It would appear from the description that the form described as crinicornis by Saussure and Zehntner^ is not the crinicornis of Burmeister. Whether lectuosa is distinct from crinicornis is a matter which cannot be settled here, the two having already been synonymized. The specimens studied show that the male is more uniformly colored than the female, the pronotum being without a light caudal margin and the median tegminal spots are absent, while the pale border of the coxae and ventral abdominal segments are much narrower than in the female. PSEUDOPHYLLODROMIA Bmnner. Pseudophyllodromia pavonacea Rehn. 1903. Pseudophyllodromia pavonacea Rehn, Trans. Amer. Ent. Soc, XXIX, p. 262. [Bartica, British Guiana.] One male. This specimen is inseparable from the type with which it has been compared. Pseudophyllodromia fasciatella Saussure. Three males, one female. These specimens agree perfectly with a series of sixteen individuals from Bartica, British Guiana. Pseudophyllodromia obscura Saussure. 1874. Pseudophyllodromia obscura Saussure, Mem. Soc. Phys. et d'Hist. Nat. Gen6ve, XXIII, p. 98. [Brazil.] Four females.. These specimens agree well with the description except that the disk of the pronotum is red-brown overlaid with blackish instead of maroon brown as described. Pseudophyllodromia prona n. sp. Types: cJ^ and 9 ; Demerara, British Guiana. 1901. (R. J. Crew.) Allied to P. peruana and lineolata Saussure, and differing from the former in the strongly lined interocular space, the color of the face and the coloration of the tegmina, the pale portions of the latter being small maculations and not whole areas. From lineolata it differs in the maculation of the pronotum, the general color of the head and the smaller tegminal markings. No other species of the genus appears to be any way as closely related as the two mentioned above. » Biol. Oent.-Amer., Orth.. I, p. 33. . 1906.] NATURAL SCIENCES OF PHILADELPHIA. 265 Size medium; form as in other species of the genus but shghtly broader. Head considerably exserted, the eyes rather prominent, the interspace between them not as wide as the depth of the eye in either sex; antennae fihform, very sparsely haired, in length very con- siderably exceeding the length of the body. Pronotum transverse sub-elliptical, cephalic margin with an extremely slight median angu- lation; caudal margin distinctly but not greatly angulate, the margin laterad of the angulation slightly emarginate; lateral margins arcuate producing slightly lateral cephalic and caudal angles; surface of the disk with two impressed areas, one or each side, extending caudo- laterad from about the caudal fourth of the pronotum ; lateral portion of the pronotum considerably depressed. Tegmina with the venation prominent ; costal margin arcuate on the proximal half, about straight with a hardly perceptible emargination in the distal half, sutural margin with two arcuations, one of the anal field, the other of the remainder of the tegmen; apex narrowly rounded; basal field narrow, limited to about the proximal third; anal field elongate-obovate, the anal sulcus more impressed in the male than in the female; costal veins thirteen to sixteen in number, rami of the median vein all longi- tudinal. Supra-anal plate of the male shorter than the subgenital plate, cerci over twice the length of the subgenital plate, moniliform, tapering from the middle, subgenital plate rather short, transverse, the apex narrow and shallowly emarginate. Supra-anal plate of the female shorter than the subgenital plate, cerci similar to those of the male, subgenital plate flat, the apical margin with a deep, very narrow median incision. Femora spined. General colors clove brown and pale ochre yellow. Head with three median, longitudinal, closely placed, parallel ochre yellow lines on the interocular space, the face with five transverse maculations of the same color, the dorsal one a complete fine line, the second, between the antennae, a broadly V-shaped mark, the third a row of four sub- circular maculations, the fourth, on the clypeal suture, a thick, nar- rowed mesad line, the fifth a simple line across the clypeus ; eyes walnut brown ; antennae strongly inf uscate except a brief ferruginous proximad section. Pronotum with the disk clove brown, a narrow median line and the lateral section pale ochre yellow, a spot on the caudal section of the pale lateral sections clove brown; caudal margin narrowdy pale. Tegmina with the base color deep brownish, the basal field and inter- spaces between the veins of the anal area and between the costal veins and two spots in the proximal section of the discoidal field, one in the very narrow proximal section and the other just distad of it, pale 266 PROCEEDINGS OF THE ACADEMY OF [May, ochre yellow. Limbs ochre yellow marked with clove brown, the caudal broadly lined dorsad, and the tibiae infuscate distad. Measure7nents. Length of body, 9.2 mm. 10 mm. Length of pronotum, 3 " 3 Greatest width of pronotum, 3.9" 4 Length of tegmen, 9.6 " 10.5 Greatest width of tegmen, 3 '' 3 " The types are unique. ISCHNOPTEKA Burmeister. Ischnoptera nyctiboroides n. sp. Type: ? ; Demerara, British Guiana. 1901. (R. J. Crew.) Belonging to the group containing /. azteca, tolteca, nana and parvula, but differing from all in the larger size, and apparently from most of them in the comparatively large intercalary area of the wings. Size rather large; form robust, general form and coloration super- ficially reminding one of a Nyctibora; surface glabrous, but not with a very high polish. Head not hidden by the pronotum, the occipital margin evenly arcuate, interspace between the eyes slightly wider than the depth of one of the eyes; mandibular palpi with the distal joint elongate-elliptical when viewed dorsad, the ventral surface ex- cavate for the entire length. Pronotum of the shape usually found in Ischnoptera, the cephalic margin sub-truncate, the caudal margin sub-truncate with an extremely slight median angulation, lateral margins slightly arcuate, the greatest width caudad; lateral sections somewhat depressed, the disk without depressions. Tegmina extend- ing beyond the tip of the abdomen by about two-thirds the length of the pronotum; costal margin evenly arcuate; sutural margin straight except in the distal fourth, where it is obliquely diverted toward the apex which is in the general curve of the costal margin and rounded acute-angulate in character; basal field not extensive; anal field elongate-obovate, the apex blunt; veins of the anal area seven in num- ber, the costal veins very numerous, discoidal field with the veins distinctly longitudinal. Wing with the greatest width contained slightly more than one and one-half times in the length; costal veins about eleven in number excluding the branches of the mediastine vein, the majority being distinctly but not at all greatly thickened near the margin; discoidal vein with a caudal ramus diverging near the middle anterior ulnar vein with two complete and about two incomplete rami ; 1906.] NATURAL SCIENCES OF PHILADELPHIA. 267 intercalary area rather small but distinct, extending about a sixth the way to the base of the wing, the apical margin of the area ver}^ gently obtuse-angulate. Supra-anal plate transverse, somewhat produced mesad and very narrowly emarginate; cerci about equal to the sub- genital plate in length, tapering in the distal half; subgenital plate somewhat transverse, the distal margin arcuate, not emarginate and not distinctly excavated for the cereal bases. Cephalic femora armed on the ventro-cephalic margin, with about five large spines placed mesad and a number of slightly smaller ones distributed between the larger spines and the distal extremity; median and cephalic femora with the ventral margins armed. General color vandyke brown, narrow lateral areas on the pronotum and the basal areas of the tegmina ochre yellow. Head uniform except for the antenna, the regions of their insertion and the clypeiis which are dull ochraceous. Pronotum with the median area very dark, almost seal brown, the lateral light areas being connected by a thread of the same color, which extends around the cephalic margin. Teg- mina almost ochraceous-rufous when viewed by reflected light, the area of the right tegmen covered by the left being vandyke brown. Wings washed with vandyke brown, the anterior field with the suf- fusion very strong, the color being quite solid around the costal veins Abdomen with the segments margined with seal brown, a broad bar of the same color paralleling the lateral margins and suffusing the subgenital of the preceding plate. Limbs ochraceous, the coxae each with two transverse bars of seal brown connected along the lateral sulcus of the coxa by a longitudinal bar of the same color. Measurements. Length of body, 17.6 mm. Length of pronotum, 4.3 " Greatest width of pronotum, •. . . . 6 " Length of tegmen, 17 " Greatest width of tegmen, 5.2 " The type is the only specimen seen by the author. BLATTELLA CaudeU. Blattella minor (Brunner). 1865. Ph[yllodromia] minor Brunner, Nouv. Syst. Blatt., p. 94. [Brazil.] One male. This specimen fits Brunner's description of this species, except that the bars on the pronotum are fairly well separated, while Brunner says ' ' linea separanti tenuissima. ' ' 268 PROCEEDINGS OF THE ACADEMY OF [May, Blattella inexpectata n. sp. Type: ? ; Demerara, British Guiana. 1901. (R. J. Crew.) In general appearance this form resembles B. parana (Walker) from Para, but it is larger with a richer coloration and with the caudal margin of the pronotum more arcuate. In some respects this species resembles the vitrea-dilatata-pavida group but is more slender, bearing a considerable superficial resemblance to Ischnoptera hlattoides. Size medium; form moderately depressed, elongate-elliptical; surface glabrous. Head almost entirely covered by the pronotum, narrowest portion of the interspace between the eyes slightly less than the depth of one of the eyes; antennae well separated at the bases, in length ex- tending caudad to the distal extremity of the caudal coxae. Pronotum moderately transverse, ovate, narrow, cephalic margin subtruncate, caudal margin moderately arcuate with an extremely slight median angulation, lateral margins arcuate with the greatest width very slightly caudad of the middle; lateral sections moderately depressed. Tegmina slightly exceeding the apex of the abdomen, the greatest width con- tained slightly more than three times in the length; costal margin arcuate, the arcuation of the proximal third much greater than that of the remainder of the margin; sutural margin nearly straight, very obhquely truncate in the apical fourth where it turns to meet the nar- rowly rounded apex ; basal field narrow and covering slightly more than a third of the length of the tegmen; anal field obovate in shape and about equalling the basal field in length ; costal veins about f om'teen in number, median vein with four rami, the distal of which has four branches, some of which has the branching carried two stages farther; anal area with five veins. Wings with the costal veins numbering about nine, the majority of the veins being distinctly incrassate distad; anterior ulnar vein bearing three complete rami. Supra-anal plate trigonal with a rather large median emargination; cerci about equal to the subgenital plate in length, fusiform; subgenital plate moderately inflated, somewhat tectate, apical margin shallowly emarginate, well marked cereal sinuations present. Cephalic femora armed on the ventro-cephalic margin with four large median and one large distal spine, the intervening area filled with fine comb-like spines. Median and caudal femora with the ventral margins spined. General color ochraceous, becoming ochraceous-rufous on the disk of the pronotum and the section of the tegmina which when in repose cover the body. Eyes bistre; lateral sections of the pronotum and costal section of the tegmina translucent ochre yellow, the section of the left tegmen which in repose is covered by the right is very pale, 1906.] NATURAL SCIENCES OF PHILADELPHIA. 269 nearly buff in color. Wings strongly washed with ochraceous; limbs pale ochraceoiLS. Measurements. Length of body, 13 mm. Length of pronotum, 3.5 " Greatest width of pronotum, 4.2 " Length of tegmen, 11.3 " Greatest width of tegmen, 3.8 " The type is unique. NYCTIBORA Burmeister. 1838. Nyctihora Burmeister, Handb. d. Entom., II, Abth. II, pt. I, p. 501. Included N. crassicornis, sericea, holosericea and latipennis Burm. Kirby has selected sericea as the type. Nyctibora tomentosa (Serville)? 1839. Blatta tomentosa Serville, Orthopteres, p. 86. [Surinam.] Two males. These specimens show some few characters at variance with the original description, but agree with the description of a female given by Saussure.2 Walker's tenebrosa from Demerara appears to be a rather different insect. PARATROPES Serville. Paratropes subsericeus Saussure. 1864. Paratropes subsericeus Saussure, Mem. I'Hist. Nat. Mex., 3me Mem., p. 63. [Surinam.] One female. This specimen is referred to subsericeus instead of elegans Bur- meister, of which the former has been considered a synonym. The description of elegans, although very brief, does not fully fit the speci- men in hand, while that of subsericeus fully describes the individual examined. The author does not assert that the two are distinct species, but Saussure's name represents the type examined and Bur- meister's does not. EPILAMPRA Burmeister. Epilampra abortivipennis Rehn. 1903. Epilampra abortivipenna Rehn, Trans. Amer. Ent. Soc, XXIX, p. 273. [Bartica, British Guiana.] One female. The tjrpe of this species was erroneously described as a female, a re-examination showing it to be a male. 2 Miss. Scient. Mex., Orth., p. 69. 270 PROCEEDINGS OF THE ACADEMY OF [May, The female is larger and bulkier than the male, but is otherwise very similar. The supra-anal plate of the female is transverse, rounded and with a moderately deep V-shaped median emargination, the cerci are about equal to the supra-anal plate in length, subequal in the proximal half, tapering in the distal half, supra-anal plate large, moderately inflated, cereal sinuations slight, the apical margin arcuate. Measurements of Female. Length of body, 26 mm. Length of pronotum, 6.5 " Greatest width of pronotum, 8.3 " Length of tegmen, 18.5 " Greatest width of tegmen, 6.5 " Epilampra lucifuga Rehn. 1903. Epilampra lucifuga Rehn, Trans. Amer. Ent. Soc, XXIX, p. 271. [Southern British Guiana.] Three males, one female. The female is slightly larger than the type specimen, but otherwise the series is perfectly typical of the species. The supra-anal plate of the male is rather produced, trigonal and fissate mesacl with the incision a mere slit to very near the apex; cerci about half again the length of the subgenital plate, tapering in the distal two-thirds; subgenital plate not reaching the caudal margin of the supra-anal plate, trigonal, supplied with lateral styles nearly a third the length of the cerci. Measurements of Male. Length of body, 22.5 mm. Length of pronotum, 5.5 '' Greatest width of pronotum, 6.5 '' Length of tegmen, 21.2 " Greatest width of tegmen, 6 " Epilampra maculicollis (Serville). 1839. Blatta maculicollis Serville, Orthopteres, p. 92. [Brazil.] One male, one female. The base color of these specimens is maize yellow, and between the eyes the head bears dorsad a pair of tapering spots of mummy brown, ■ ventrad a blotch of seal brown, which has laterad short bars of the same color extending ventrad between the antennae. The pronotum is very finely sprinkled with, and the delicate pattern traced, in vandyke brown, while the tegmina have the discoidal and anal areas very thickly flecked with large and small spots of burnt umber, the basal and costal regions being comparatively clear with only small spots. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 271 Epilampra fusca Bruimer. 1865. Epilamjyra fusca Brunner, Nouv. Syst. Blatt., p. 170. [^'enezuela.] One female. As far as can be determined from Brimner's description, which was based on a mutilated specimen, this individual appears to represent the species. The color pattern is rather regular and typical of the genus, the maculations of the tegmina being uniformly distributed but larger in the discoidal field, while the whole insect has a strong suffusion of raAv sienna. The head bears two spots of seal brown between the eyes, these spots being connected by a line of the same color which encloses a quadrate area between the antennae. Measurements. Length of body, 41.5 mm. Length of pronotum, 9-2 Greatest width of pronotum, 11.5 " Length of tegmen, 38.5 " . Greatest width of tegmen, 12 PERIPLANETA Burmeister. Periplaneta australasiae (Fabricius). One male, two females. CHOEISONEURA Brunner. Chorisoneura gracilis (Saussure). 1862. Blatta gracilis Saussure, Revue et Magasin de Zoologie, 2e ser., XIV, p. 167. [Brazil.] One damaged specimen. PYCNOSCELUS Scudder. Pycnoscelus surinamensis (Linnteus). Two females. One individual is quite uniform blackish above, with the cephalic margin of the pronotum laterad and the marginal field of the tegmina ochraceous. PANCHLORA Burmeister. Panchlora viridis (Fabricius). 1775. [Blatta] viridis Fabricius, Syst. Entom., p. 272. [America.] One male. This species has previously been recorded from Cayenne. Panchlora peruana Saussure. 1864. P[anchlora] Peruana Saussure, Revue et Magasin de Zoologie, 2e ser., XVI, p. 342. [Peru.] '272 PROCEEDINGS OF THE ACADEMY OF [May, One male, two females. The female individuals differ somewhat in size, but both appear referable to this species. Measurements of Females. Length of body, 20.4 mm. 21.5 mm. Length of pronotum, 5.2 " 6 " Greatest width of pronotum, 6 " 7.3 " Length of tegmen, 19.5 " 22.6 " Greatest width of tegmen, ...... 5.2 " 6.5 " HOLOCOMPSA Burmeister. Holocompsa mtidula (Fabricius). 1781. [Blatta] nitidula Fabricius, Spec. Ins., I, p. 345. [Surinam.] Two females. These specimens have a dark brown patch on the caudal section of the disk, and the caudal margin of the pronotum is narrowly suffused with the same color. This pattern of coloration is evidently present in the type of Fabricius as he states: "Thorax ferrugineus, nitidus macula baseos nigra." The golden hairs which are present on the lighter sections of the pronotum are also distributed over the brown area. Holocompsa cyanea Burmeister. Two males. This species has been recorded from Brazil by Brunner. BLABERUS Serville. Blaberus posticus Erichson. 1848. Blabera postica Erichson, in Schomburgk's Reisen Brit. Guiana, III, p. 580. [British Guiana.] One female. This specimen is the first of a number of individuals of this section of the genus examined by the author, which appears to fit Erichson's very poor description. The pronotal maculation is confluent laterad with the blackish caudal margin. This character appears to be what Erichson refers to in " Der Halsschild mit spitzer gerundeten Seiten, auf dem Riicken mit einigen kleinen, unbestimmten schwarzen Flecken, dagegen der Hinterrand breit schwarz gesaumt." The figm*e of B. thoracicus given by Saussure and Zehntner represents very nearly the same thing, but a specimen from Costa Rica which I had considered thoracicus ^ has these lateral bars lacking. Kirby's synonymizing of 3 Proc. Acad. Nat. Sci. PhiJa., 1905, p. 792. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 27S the two species would appear, from the evidence in hand, to be correct^ The author's B. hiolleyi * differs from posticus in the paler coloration, the more numerous and finer veins of the anal area of the tegmen and the separate and distinctly outlined maculations of the pronotum. MANTID.^. ACONTISTA Saussure. Acontista perspicua (Fabricius). 1787. [Mantis] perspicua Fabricius, Mant. Insect., I, p. 230. [Cayenne.] One male, one female. This species is known only from the Guianas. Angela guianensis n. sp. Types: c? and ? ; Demarara, British Guiana. 1901. (R.J. Crew.) Closely allied to A. quinquemaculata (St oil) from Cayenne, but dif- fering in the coloration of the wings in both sexes, particularly in the absence of yellow in the male, and also in the short tegmina and wings of the female. From brachyptera (Stoll) it can be separated by the unspined margins of the caudal section of the pronotum and the tuber- culate abdomen, as well as the color of the wings of the male. The coloration of the wings of fulgida-yand trifasciata readily separates the new form from these species. Size rather large; form as usual in the genus, male bacilliform with elongate tegmina and wings, female^elongate with quite short tegmina and wings; surface smooth but not polished. (^ . Head strongly transverse, occipital outline straight except for a pair of slight lateral depressions; eyes prominent, ovoid in outline when viewed laterad, rounded and projecting slightly cephalad of the general level of the face but nearly plane caudad; ocelli large, elliptical, placed in a low, broad triangle; facial scutellum arcuate transverse, the arcuation dorsad; antennae as long as the head and pronotum together, moderately serrulate. Pronotum with the coxal dilation slight, subequal cephalad and caudad of the same; lateral margins of the collar distinctly dentate, of the shaft slightly dentate cephalad, smooth caudad; median carina distinct throughout the length of the pronotum but quite delicate on the collar, the shaft compressed trigonal in section. Tegmina slightly longer than the head and pronotum, when in repose reaching nearly to the tip of the third abdominal seg- ment; margins subparallel, apex rounded acute-angulate, the sutural margin being considerably trimmed toward the apex; costal field moderately expanded proximad and very gradually narrowing distad ; * Ibid., p. 792. 274 PROCEEDINGS OF THE ACADEMY OF [MaV, ulnar vein with three almost parallel rami, stigma distinct. Wings exceeding the tegmina by about a third the length of the pronotum, slightly longer than twice the width; costal margin nearly straight, apex rather narrowly rounded. Cephalic coxse slightly less than half the length of pronotum, not armed ; cephalic femora about five-eighths the length of the pronotum, the ventro-caudal margin bearing four large and one small distal spines, the ventro-cephalic margin armed with two grades of spines, the larger size numbering six and the smaller eight, the smaller being placed one each between the larger spines, except that distad of the distal large spine two small ones are present and between the first and second distal spines three are present, dis- coidal spines four in number; cephalic tibiee (exclusive of terminal claw) less than a third the femoral length, the cephalic margin with sixteen spines increasing in size distad, caudal margin with five spines; tarsi about equal to the tibiae in length. Median and caudal limbs very slender but not greatly elongate, the caudal femora not quite reaching the middle of the third abdominal segment. General color Prout's brown. Tegmina hyaline washed narrowly along the sutiu-al margin (distad of the dividing vein) and very broadly along the costal margin with Prout's brown, the greater portion of all the longitudinal veins being marked with numerous fine linear seal brown touches. Wings with proximal half hyaline, the apex of the anterior field yellowish coriaceous, the remainder vandyke brown with very strong violet reflections, this coloration also extending nearly to the base as a varying line along the mediastine vein, the caudal section of the broad transverse bar extending clear to the margin, and broken only by less infuscate areas immediately surrounding the radiate veins and the clear pattern of the transverse veins. ? . Head much as in the male, facial scutellum slightly less arcuate ; antennse very short, hardly exceeding the length of the collar of the pronotum, filiform. Pronotum with the collar nearly a fifth the length of the whole pronotum, coxal dilation slight; collar slightly tapering cephalad, the cephalic angle narrowly rounded and with an apparent elevation caused by a marked depression of the collar immediately cau- dad ; lateral margins of the collar dentate, shaft with the margins un- armed; median carina distinct on the collar and on the cephalic and caudal sections of the shaft, collar trigonal in cross section, shaft rounded dorsad. Tegmina about a fifth the length of the pronotum, reaching in repose to about the middle of the first abdominal segment, margins very slightly approximate distad; apex broadly rounded, stigma distinct. Wings in repose extending beyond the tegmina by not quite a third 1906.] NATURAL SCIENCES OF PHILADELPHIA. 275 the length of the latter, the length but slightly greater than the length of the tegmen. Abdomen with the apex broad, depressed, the three terminal segments with very distinct median longitudinal marginal folds; supra-anal plate about as long as broad, the distal portion rounded; cerci broad, lamellate and truncate as usual in the genus; subgenital plate strongly compressed, rostrate, diverted ven- trad out of the general abdominal plane. Cephalic coxae slightly less than half the length of the pronotum, slightly armed, distal section expanded and compressed, being abruptly trigonal in section, while the proximal three-fom-ths is transverse in section; cephalic femora slightly more than half the length of the pronotum, the margins armed as in the male except that the cephalic margin has one or tw6 additional spines of the second order; tibiae and tarsi as in the male. Median and caudal limbs slender but not greatly elongate. General color tawny-olive finely washed, spotted and sprinkled with Front's brown; eyes marked with seal brown. Pronotum with the two colors showing very distinctly in a pattern which is pardaline in ■character, the shaft with its cephalic fifth inclined toward bistre. Tegraina tawny-olive in the proximal half, becoming brick red in the distal half, the costal section Front's brown, some of the rami of the ulnar vein and the dividing vein with short fine linear touches of bistre, the folded proximal section blackish purple. Wings with the anterior field broccoli brown, yellowish along the costal margin and rich brick red in the distal third ; posterior field with the base color purplish brown with a very strong purplish-blue sheen, bearing two irregular maculations of lemon yellow, limited to about the cephalic half of the field and each bearing distad an area of equal size of the clear base color of the area, the whole wing except these areas and the semi-coriaceous red section of the anterior field having all the transverse veins white or yellowish- white. Abdomen of the general color; limbs of the general color with the pardaline appearance distinctly marked on the median and caudal limbs. Measurements. Length of body, 67.5-1- mm. 88 mm. Length of pronotum, 25 " 36.5 " Greatest width of pronotum, 2.3 " 3 " Length of tegmen, 28.5 '' 13.9 " Length *of wing, 34 " 13.5 " Length^of cephahc femur, 14.8 " 20.3 " The types are the only specimens seen by the author. 276 PROCEEDINGS OF THE ACADEMY OF [May, PHOTINA Burmeister. 1838. Photina Burmeister, Handb. d. Ent., II, Abth. II, pt, I, p. 531. Type, by elimination, P. vitrea Burm. Photina vitrea Burmeister. 1838. M[antis] (Photina) vitrea Burmeister, Handb. d. Entom., II, Abth. II, pt. I, p. 532. [Brazil.] One female. PAKAMTJSONIA Rehn. Paramusonia conspersa (Saussure). 1870. Th[espis] conspersa Saussure, Mitth. Schw. Ent. GeseU., Ill, p. 238. [South America.] « One male, one female. Apparently this is the first record of the species with exact data. This form can be separated from the allied P. injumata (Serville) by the finely denticulate lateral margins of the j^ronotum. MIOPTEEYX Saussure. Miopteryx grenadensis Saussure. Two males. CAILIBIA Stai. 1877. Callibia Stal, Bihang till K. Svenska Vet. Akad. Handl., IV, No. 10, pp. 79, 85. Type. — Harpax pictipennis Serville = Mantis diana Stoll. Callibia diana (Stoll). 1813. [Mantis] Diana Stoll, Natuur. Afbeeld. Besch. Spooken, pp. 74, 78, pi. XXV, fig. 100. ["East Indies."] One male. This species has been recorded from Cayenne. OXYOPS Saussure. 1869. Oxyops Saussure, Mitth. Schw. Ent. GeseU., Ill, pp. 56, 66. Type. — 0. rubicunda (Stoll). Oxyops rubicunda (Stoll). 1813. [Mantis] Rubicunda Stoll, Natuur. Afbeeld. Besch. Spooken, pp. 73, 79, pi. XXV, fig. 96. One male. This specimen has the internal face of the cephalic coxse unicolor as in one of the males examined by Saussm-e. PARASTAGMATOPTERA Saussure. Parastagmatoptera tessellata Saussure and Zehntner. 1894. Parastagmatoptera tessellata Saussure and Zehntner, Biol. Cent.- Amer., Orth., I, p. 188. [Cayenne.] One male. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 277 PHASMID^. CEROYS Serville. 1839. Ceroys Serville, Orthopteres, p. 262. Type. — C. perfoliatus (Gray). Ceroys littiiis n. sp. Type: 9 ; Demerara, British Guiana. 1901. (R. J. Crew.) Not closely allied to any previously known species of the genus, dif- fering in the absence of spines and in the lobation of the median and caudal femora. This species may not be true Ceroys, but it does not belong to any of the present allied genera. Size median; form moderately robust; surface smooth, not polished, unarmed. Head rather long, subequal caudad of the eyes, occiput with a faint median longitudinal impressed line and several very weak ones laterad; eyes subglobose, moderately prominent; antennae filiform, about equal to the head and thorax in length, the proximal joint de- pressed. Pronotum nearly half again as long as broad, the lateral mar- gins somewhat incurved, cephalic margin moderately concave, caudal margin arcuate, transverse depression distinct, arcuate caudad, median longitudinal depression very faint and incomplete. Mesonotum five and a half times the length of the pronotum, a distinct dorsal line present but no carina. Metanotum (including median segment) about five- eighths the length of the mesonotum, the median segment very dis- tinctly longer than broad and exceeding the length of the remainder of the metanotum. Abdomen distinctly exceeding the remainder of the body in length; four proximal joints simple, longitudinal and very slightly increasing in length distad; fifth segment expanded meso- caudad, in length about ec^ual to the fourth segment; sixth segment simple, slightly shorter than the first proximal segment; seventh seg- ment two-thirds the length of the sixth, somewhat expanded caudad; eighth segment slightly shorter than the seventh ; ninth segment shorter than the eighth, compressed, carinate dorsad, narrowly emarginate caudad; supra-anal plate with the angulate tip alone visible; cerci compressed, tapering, blunt, the length about two- thirds that of the ninth dorsal segment; subgenital opercule boat-shaped, carinate in the distal half, lanceolate, acute, not exceeding the tips of the cerci. Limbs multicarinate, robust. Cephalic femora about as long as the mesono- tum, compressed, basal flexure very marked and abrupt, the two prominent dorsal and single ventral carina lamellato-carinate ; cephalic tibise about equal to the femora in length and with three lamellate carinse; tarsi with the proximal joint slightly longer than the remaining joints. Median femora slightly shorter than the metanotum, the two 19 278 PROCEEDINGS OF THE ACADEMY OF [Ma}^, ventral lamellate carinse each with a prominent proximal and distal rounded foliaceous lobe, the dorso-cephalic carina also with a low proximal lobe; median tibiae about as long as the femora, the ventral carinse lamellate, but no lobes present; median tarsi with the proximal joint about equal to the remaining joints. Caudal femora as long as the proximal two and a half abdominal segments, in structure similar to the median but with the dorsal lobe even smaller; caudal tibise slightly longer than the femora, in structure similar to the median tibise but slightly slenderer; caudal tarsi with the proximal joint distinctly but not greatly longer than the remaining joints. General color Prout's brown obscurely marked on the head and limbs with bistre, the femoral lobes and extremities of the median and caudal tibise decidedly bistre ; antennse wood brown sprinkled with the general color. Measurements. Length of body, 101 mm. Length of pronotum, 4.5 " Length of mesonotum, 24.5 " Length of metanotum (including median segment), ... 16 " Length of cephalic femur, 22.5 " Length of median femur, 14 " Length of caudal femur, 17.3 " The type is unique. OLCYPHIDES Griffini. Olcyphides fasciatUB (Gray). 1835. P[hasmd\ fasciatum Gray, Synopsis Phasm., p. 24. [Brazil.] One female (sex ?). This individual is broken, but I presume it is a female. The original description, while very brief, sums up the chief diagnostic characters of the species. 1906.] NATURAL SCIENCES OF PHILADELPHIA, 279 DESCRIPTIONS OF FIVE NEW SPECIES OF ORTHOPTERA FROM TONKIN. BY JAMES A. G. REHN. PHASMID^. Phryganistria grandis n. sp. Type: c? ; Tonkin, Indo-China. [A. N. S. Phila.] Allied to P. sarmentosa Westwood from Sylhet, but differing in the larger size, the greater number of medium-sized spines on the ventral margins of the median and caudal femora, and the presence of a strong distal spine on each of the same margins, as well as the proportionately slenderer abdomen. Size very large; form very slender; surface glabrous. Head about equal to the pronotum in length, somewhat narrowed caudad, and with the caudal section of the occiput with foiu- longitudinal subparallel impressed lines ; eyes rather prominent, globose ; ocelli absent ; antennse when extended caudad reaching nearly to the caudal margin of the second sCbdominal segment, first antennal joint moderately depressed. Pronotum distinctly longer than broad; ceiDhalic margin slightly arcuate-emarginate, caudal margin subtruncate; transverse depression slightly bent caudad in the middle, a faint trace of a longitudinal de- pression in the cephalic section, none in the caudal section. Mesono- tum but little shorter than the metanotum, median segment and first segment of the abdomen nearly seven times the length of the pronotum, subequal except in the caudal fourth where it is slightly and gradually expanded. Metanotum (including median segment) slightly more than two-thirds the length of the mesonotiun ; median segment slightl}^ less than a third the length of the remainder of the metanotum; cephalic half subequal, evenly expanding in the caudal half. Abdo- men exceeding the head and thoracic segments in length by more than the length of the head, as a whole subequal in width, the slight enlargement of the segments at their articulation with con- tiguous segments giving the abdomen the appearance of a sec- tion of bamboo ; fu'st segment very slightly"^ longer than the head and pronotum together, second and third subequal and slightly longer than the first, fourth and fifth subequal and slightly longer than the second and third, sixth segment two-thirds the length of the fifth, 280 PROCEEDINGS OF THE ACADEMY OF [May, Fig. 1. — Phryganistria grandis n. sp. Dorsal view of male type. (One- half natural size.) united ; arolia present. Median seventh segment about half the length of the sixth, compressed and deep, eighth segment com- pressed, as deep as and slightly shorter than the seventh seg- ment, ninth segment with its extreme length about equal to the seventh and eighth united, strongly compressed, tec- tate and carinate dorsad, bear- ing two lateral digitiform pro- cesses which enclose a deep V-shaped emargination, the pro- cesses being rounded distad and bearing a number of short re- curved spines on their inner faces ; cerci slender, simple, strongly curved distad; sub- genital opercule extending to the apex of the eighth dorsal segment, not produced, the dis- tal margin straight, the dorsal margin rounded-angulate, the ventro-caudal angle blunt-rec- tangulate. Cephalic femora slightly longer than the head, pro- and mesonotum, somewhat compressed, multicarinate, basal flexure short but very distinct, dorso-cephalic margin with a number (14 to 18) of rather irregularly distributed spines ; cephalic tibiae considerably ex- ceeding the femora in length, as long as the six proximal ab- dominal segments in length, very slender and with the mar- gins unarmed ; cephalic metatarsi distinctly longer than the re- maining segments of the tarsi femora in length equal to the 1906.] NATURAL SCIENCES OF PHILADELPHIA. 281 mesonotum and half of the pronotum, rather robust, strongly carinate, the dorsal carinse with a few short spines on each but the caudal with a greater number than the cepha- lic, ventral margins with a series of strong spines, 12 to 14 in number, in addition to which is placed distad a very strong acute spine, the genicular lobes are spiniform and the ventral tn^^gral'dTs^i^^.^'Lt- face bears a median longitudinal series of 6 or 7 terai view of the apex fine spines; median tibiae very slightly longer (NarraUiz J^)° "' ' "" " than the femora, slightly curved proximad, all the margins spined but the ventrals with a much greater number than the dorsals; metatarsi about equal to the remain- ing tarsal joints, all more robust than in the cephalic limits. Caudal femora reaching nearly to the apex of the fom-th abdominal segment, dorsal margins with very few spines, ventral margins armed as the median femora, the large distal spine and genicular lobes similar in form, and the lateral series numbering 16 to 17 spines, the median ventral series 9 to 10 in number; caudal tibise similar to the median in •structure but longer; caudal metatarsi distinctly longer than the remaining tarsal joints. Generla color ferruginous, pale on the head and distad on the abdo- men, the greater portion of the mesothorax and the cephalic section of the metathorax wood brown, the former washed with malachite green. Antenna? blackish brown. Ventro-lateral margins of the median and caudal femora dull oil green; spines tipped with black. Measurements. Length of body, 198 mm. Length of head, 7.5 " Length of pronotum, 7 " Length of mesonotum, 46 " Length of metanotum (including median segment), . . . 34.5 " Length of cephalic femur 63.8 " Length of median femur, 48.5 " Length of caudal femur, 58 " Length of caudal tibia, 70.5 " The type only has been examined. TETTiaONID^. Chondrodera ' maxima n. sp. Type: ? ; Tonkin, Lido-China?^ [A. N. S. Phila.] ^ It appears from the Hterature on this genus that Brunner misinterpreted the shape of the mesosternum and metasternum, and should have placed this genus, in'his revision of the Pseudophyllinse, under 4 instead of 4^ of the Phyllomimi. ^ Some doubt is attached to this locality. 282 PROCEEDINGS OF THE ACADEMY OF [Maj^, Allied to C. notatipes and C. suhvitrea Karsch from West Africa, but differing in the greater size, the deeper lateral lobes of the pronotum, the smaller eye, the differently shaped tegmen and the lesser number of transverse veins in the same. From C. borneensis Brunner it is readily separated by the presence of a; longitudinal spurious vein in the anal field of the tegmen, and in the strongly spined and but slightly ciliate femora. Size moderately large; form generally compressed. Head broad, somewhat flattened cephalad; occiput not elevated, straight, rounded Fig. 3. — Chondrodera maxima n. sp. Lateral Adew of type. (X 1^.) transversely; fastigium produced, extending slightly beyond the antenna! scrobes, acute, moderately sulcate distad; eyes rather small, exserted and quite prominent, subglobose; antennae with the margins of the scrobes touching mesad, first joint large, subcylindrical, joints beyond the second poorly defined, in an incomplete condition extending to the distal third of the tegmen. Pronotum some- what saddle-shaped, the lateral lobes moderately flaring ventrad, surface with a number of blunt tu- bercles; cephalic margin obtuse-angulate with the immediate angle truncate, caudal margin arcuate with a slight median emargination ; longitudinal median sulcus distinct, precurrent, bearipg in its de- pression a blunt median carina on the prozona (proper) and metazona, on the latter becoming more conspicuous caudad by the shallower character of the sulcus ; median sulcus flanked through nearly its entire length by a pair of series of rather irregularly placed tubercles ; sulci separating the prozona, mesozona_and metazona very distinct and deep, the meso- FiG. A. — Chon- drodera maxima n. sp. Dorsal view of female pronotum. (X2.) 1906.] NATURAL SCIENCES OF PHILADELPHIA. 283 metazonal one particularly heavy; lateral portions of the caudal mar- gins of the disk somewhat elevated ; lateral lobes slightly longer than deep, caudal margin with a distinct but shallow elytral sinus, ventral margin sinuate obtuse-angulate, the margin supplied with low tubercles which are the terminations of low subobsolete ridges radiating from a point on the ventral section of the lateral lobes. Prosternum unarmed. Mesosternum distinctly but not greatly transverse, the foramina large and well separated. Metasternum strongly transverse, the cephalic margin slightly broader than the mesosternum, caudal section regularly narrowed, the foramina nearer the caudal margin, transverse and con- nected by a rather broad sulcus. Tegmen rather coriaceous in textiu'e, the greatest width (excluding the anal field) contained slightly more than three times in the length; dorsal outline when viewed laterally (excluding the horizontal anal field) nearly straight, costal margin slightly arcuate cephalad, strongly arcuate distad; apex dorsad and rather narrowly rounded, costal field with a distinct rounded proximal lobe; costal field with the mediastine vein short, oblique, anterior radial vein sending nine principal oblique rami to the costal margin, the most of which are bifurcate in the distal portion of their length; radial ramus issuing distinctly but not greatly proximad of the middle of the tegmen, the area between this and the posterior radial vein with six principal transverse veins, which enclose areas roughly quad- rate in shape; anterior ulnar vein undulate, reaching to the apex of the tegmen, the area between the radial ramus and this vein having five principal transverse veins ; area between the two ulnar veins with seven principal transverse veins; anal area narrow but continued to very near the apex of the tegmen, and bearing in addition to one principal precurrent though erratic longitudinal vein several others of similar character in the proximal section, short transverse veins numerous. Wings in repose reaching almost to the tips of the tegmina. Abdomen moderately compressed ; supra-anal plate slightly longitudinal, rounded, the apex emarginate and a shallow longitudinal median depression is present on the proximal half, ciliate; ovipositor about twice as long as the pronotum, heavy, hardly curved, the dorsal margin nearly straight and with a slight basal emargination, ventral margin regularly arcuate, apex acute, dorsal margin and short distal section of the ven- tral margin fiftely serrate; subgenital plate short, transverse trigonal, the apex with a V-shaped emargination. Cephalic femora about as long as the head and pronotum, carinate, the ventral margins wath fom- to six blunt spines; cephalic tibiae slightly longer than the femora, bullae perforate and expanded on both sides, dorsal carinae with one 284 PROCEEDINGS OF THE ACADEMY OF [May, short spine on the caudal, ventral carinse both with a numl^er (6 to 7) of short spines; tarsi with the third joint very broad. Median femora about as long as the cephalic, more compressed, deeper, ventro-cephalic margin strongly carinate and armed with five short spines, ventro- caudal margin armed with a number of very small spines, the margin not strongly carinate like the ventro-cephalic; median tibiae slightly longer than the femora, compressed, carinate, ventral margins armed with seven small spines. Caudal femora not quite half the length of the tegmen, moderately compressed, and considerably inflated in the proximal half, cephalic face bearing two longitudinal depressions, ven- tro-cephalic margin lamellate and armed with ten or eleven flattened spines, which are small proximad and increase to a considerable size distad; caudal tibiae about equal to the femora in length, compressed, carinate, the dorso-cephalic margin armed wdth four short spines, the dorso-caudal with twelve to fourteen larger lamellate spines, the ventral margins with about seven rather small spines on each margin. General color very dull canary yellow, becoming saffron yellow on the limbs. Head touched with saffron yellow; eyes walnut brown; antennae with widely placed annuli of walnut brown. Pronotum with the median longitudinal area faintly washed with lilac. Tegmina with the principal transverse veins bordered and touched with patches and lines of wine purple, the proximal portion of the sutural margin faintly washed with the same color. Ovipositor clay-color, the distal half and the entire ventral margin seal brown. Limbs with the spines either wholly or in part black. Measureinents. Length of body (exclusive of ovipositor), 43 mm. Length of pronotum, 8.5 " Greatest width (ventral) of pronotum, 10 " Length of tegmen, 61 " Greatest width of tegmen, 20 " Length of caudal femur, 27.5 " Length of ovipositor, 19.5 " The type only has been examined. Anabropsis ■' tonkinensis n. sp. Type: ?; Than-Moi, Tonkin, June-July. (Friihstorfer.) [A. N. S. Phila.] Allied' to A. frater (Brunner),^ supposed to be from India, but differ- * Schcenobates of authors, cf. Rehn, Canad. Entom., XXXIII, p. 272, 1901. * Verhandl. K. K. Zool.-bot. Gesell. Wien, XXXVIII, p. 275. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 285 ing in the smaller size and the fewer spines on the margins of the caudal tibiae, ten instead of fourteen in number. Brunner's species is very briefly described and based on a male, but the specimen in hand differs in the lesser length of the body and of the caudal femora and tibiae, as well as the fewer tibial spines. Size rather large (compared with American species of the genus); form somewhat compressed ; surface entirely glabrous, polished. Head short and broad, occiput rounding toward the fastigiiun and bearing a very slight median longitudinal sulcus, which is continued nearly to Fig. 5. — Anabropsis tonkinensis n. .sp. Lateral \-iew of tj'pe. (About X 2.) the extremity of the compressed fastigium, which latter is separated from the vertex by a broad and shallow depression; frontal fastigium touching the fastigium of the vertex; paired ocelli large, placed on the lateral faces of the fastigium, median ocellus placed closer to the apex of the facial fastigium than to the clypeal margin; eyes ovoid, flattened cephalad, slightly prominent; antennal scrobes little elevated, proximal and second antennal joints subcylindrical, the second considerably smaller than the first, remainder of the antennae slender, moniliform. 286 PROCEEDINGS OF THE ACADEMY OF [May, about twice the length of the body, the third joint slightly longer than the proximal and second joint together. ]\Iandibles, clypeus and lab- rum large and produced as is usual in the genus ; palpi large, with the first and second joints subequal, the third slightly longer than the others and distinctly longer than the labrum. Pronotum hardly arched longitudinally, distinctly arched transversely ; cephalic margin truncate, caudal margin arcuato-truncate ; lateral lobes very greatly longer than deep, the greatest depth in the cephalic section ; no median carina pre- sent, but an extremely fine median sulcus extends the whole length of the pronotum; distinct transverse sulci two in number, one near the cephalic margin, the other slightly caudad of the middle and extend- ing cephalo-laterad of the axis of the pronotum and forming oblique broad depressed areas on the lateral lobes. Tegmina very minute, not half as long as the exposed lateral portion of mesonotum, pad-like. Mesonotum, metanotum and abdominal segments provided dorsad with a slight but distinct median carina, laterad the distal section of each segment is provided with a number of short welt-like ridges. Supra-anal plate small, trigonal, depressed with the margins elevated ; ovipositor rather short and weak, slightly arcuate, tapering, apex moderately acute, margins unarmed. Prosternum with a pair of very long spines, nearly erect, the tips needle-like; mesosternum with a pair of similar spines, which, however, are shorter and broader than those on the prosternum ; metasternum with a pair of blunt triangular lobes. Cephalic femora moderately compressed, slightly longer than the pronotum, margins and genicular lobes unarmed; cephalic tibiae slightly longer than the femora, dorsal margins each with a distal^ spine and the cephalic with one median one, ventral margins each with five spines, the proximo-caudal small, as is also the proximo-dorsal; cephalic tarsi very slightly more than half the length of the tibiae, arolia absent. Median femora equal to the cephalic in length and unarmed ; median tibiae very slightly longer than the femora, armed on the dorso-cephalic margin with three spines and on the dorso-caudal with four spines, the ventral margins each armed with five spines. Caudal femora slightly shorter than the length of the body, strongly inflated in the proximal two-thirds, the distal third subequal, pagina sculptured with a regular series of clear-cut oblique lines, caudo- ventral margin with five to seven short spines on the distal section; caudal tibiae about as long as the femora, dorsal margins each armed with^ten fixed spines, ventral margins armed wdth three short mobile spines on the cephalic margin and one on the caudal margin, while three pairs of distal spurs are present, the proximal being very large 1906.] NATURAL SCIENCES OF PHILADELPHIA. 287 on each margin and spiniform; caudal tarsi with the distal joint about equal to the proximal in length. General color dorsad seal brown, ventrad including the limbs pale orange-ochraceous. Head solid color dorsad, ventrad of the dorsal color laid over cream-buff, the facial f astigium clear cream-buff ; mouth parts, except clypeus, of the ventral color; eyes broccoli brown; anten- nse of the ventral color except for the three proximal segments which maculate or suffuse with seal brow^n. Pronotum slightly sprinkled with buffy laterad. Abdomen with the welt-like ridges buffy, as are also the styles and ovipositor. Cephalic and median femora strongly blotched and suffused with seal brown distad ; median and caudal tibiae with proximal section washed with seal brown. Caudal femora with the pagina sprinkled with seal brown, the carina immediately ventrad of the pagina strongly blotched with the same in the proximal two- thirds and the genicular lobes and arches of the same color; caudal tibiae with the genicular region touched with seal brown. All spines and spurs narrowly tipped with seal brown. Measurements. Lengthof body (exclusive of ovipositor), 28.5 mm. Length of antennae, circa 62 " Length of pronotum, 9.5 " Length of caudal femur, 25 " Length of ovipositor, 8.5 " The type alone has been examined. Diestrammena longipes n. sp. Type: d"; Than-Moi, Tonkin. June-July. (H. Friihstorfer.) [A. N. S. Phila.] Allied to D. unicolor Brunner^ from Siberia, China and Tenasserim, but differing in the very much longer limbs and the very marked colora- tion. Size rather large ; body in life probably covered in the greater part with golden scales, of which traces remain. Head short, broad and deep; occiput slightly rounded into the f astigium, which is short, rather broad, deeply sulcate and bifurcate; frontal f astigium low, separated from the f astigiimi of the vertex by a considerable space ; ocelli small ; eye ovate, flattened cephalad, moderately prominent, considerably shorter than the infra-ocular portion of the genae ; antennal scrobes not elevated, proximal joint short subcylindrical, second joint smaller, sub- cylindrical, remainder moniliform, very nearly five and one-half times 5 Verhandl. K. K. Zool.-bot. Gesell. Wien, XXXVIII, p. 299. 288 PROCEEDINGS OF THE ACADEMY OF [May, the length of the body; mandibles, clypeus and labnim moderately prolonged; palpi long and slender, the proximal joint quite short, the second and third longer and siibequal, the distal slender, gradually enlarging, over half again as large as the second. Pronotum arched transversely and with faint indications of lateral shoulders; cephalic margin truncate, caudal margin subtruncate; lateral lobes slightly longer than deep, the cephalic section of the margin oblique, mesad Lateral view of U-pe. (X 2.) with a rounded obtuse angle and caudad with the same, the caudal mar- gin itself nearly straight. Mesbnotum with the caudal margin obtuse- angulate ; metanotum very slightly obtuse-angulate. Abdomen slightly tectate; cerci styliform, slightly shorter than the pronotum. Cephalic femora very slightly longer than the head and thoracic segments, armed on the ventro-cephalic margin with three spines and on the 1906.] NATURAL SCIENCES OF PHILADELPHIA. 289 cephalic genicular lobe with one; cephalic tibiae equal to the femora in length, unarmed dorsad except for a short distal spine on the caudal margin, ventral margins each armed with two mobile spines, the distal ventral margin with one short median spine, spurs two in number, comparatively short, subequal; cephalic tarsi compressed, two-thirds the length of the tibiae, metatarsi slightly longer than the remaining joints, distal joint much slenderer than the others, arolia absent. Median femora very slightly shorter than the head and thoracic segments, ventro-cephalic margin with a single spine in the distal section, ventro-caudal margin unarmed, both genicular lobes with large mobile spines, that of the cephalic lobe slightly shorter than the other; median tibiae somewhat longer than the femora, the ventral margin with four paired spines and one median one on the distal mar- gin, spurs rather short, the caudal the longer and that not more than a third the length of the metatarsus; tarsi similar to the cephalic. Caudal femora elongate, nearly half again as long as the body, consid- erably inflated in the proximal half, slender in the distal half with a slight genicular enlargement, ventro-cephalic face of the proximal inflated portion with several deep longitudinal impressions, the cephalic ventral carina much lower than the caudal, which is moderately lamellate, cephalic armed with ten spines in the distal section, the caudal armed with sixteen spines in the distal two-thirds, genicular lobes each with a very short mesad spine; caudal tibiae slightly longer than the femora compressed, ventral margins imarmed, dorso-cepha- lic margin with thirty-two fixed spines, dorso-caudal with thirty, spurs in three pairs, the dorsal mvich the larger and with the caudal one slightly longer than the cephalic and but little shorter than the metatarsus ; caudal tarsi similar to the others in structure but slightly larger and more robust. General color mars brown, the caudal margins of the thoracic and abdominal segments rather irregularly and broadly marked with seal brown, the median sections of the segments, however, being clear or nearly clear mars brown, while the mesonotum and metanotum bear irregular touches of sulphur yellow. Face ventrad of the eyes and of the antennae touched with vandyke brown ; eyes walnut brown ; anten- nae seal brown, the proximal joints paler and maculate with seal brown, joints beyond the second with two characters of annulations of lemon yellow, a broad annulation and a number of narrow annulations be- tween each broad one, both characters of annulations becoming distant distad. Limbs with the base color straw yellow with a faint touch of lavender on the proximal portions of the femora, all strongly annulate 290 PROCEEDINGS OF THE ACADEMY OF [May, with clove brown, the number of annuU on the caudal limbs being four principal ones on the femur, five on the tibia and two on the tarsus. Measurements. Length of body, 21.5 mm. Length of antenna, circa 118 " Length of distal joint of the palpus, 5.2 " Length of caudal femur, 27.5 " Length of caudal tibia, 29 " The type is unique. Diestrammena palpata n. sp. Type : c? ; Than-Moi, Tonkin. June-July. (H. Frtihstorfer.) [A. N. S. Phila.] Allied to D. unicolor Brunner and D. Jongipes described above, but differing from D. unicolor in the much longer limbs as is the case with D. longipes, from which palpata differs in the very long palpi and the different coloration. Size rather large; form somewhat compressed. Head with the occi- put descending to the fastigium which is short and deeply, but very narrowly, divided ; facial fastigium separated from the fastigium of the vertex by a considerable space; eye elongate subreniform; antennae over five times the length of the body, the proximal joint large, subcylin- drical, second joint small and very short, remaining joints moniliform; palpi very slender, elongate, first and second joints short, third and fourth quite long, subequal, fifth very long, over half again as long as the fourth. Pronotum rounded transversely; cephalic margin sub- truncate, caudal margin slightly projecting, roundly obtuse-angulate, no lateral angles present; lateral lobes slightly longer than deep, the ventral margin bluntly obtuse-angulate. Mesonotum distinctly obtuse- angulate caudad; metanotum with the caudal margin very slightly arcuate. Abdomen compressed. Cephalic femora somewhat longer than the head and thorax, slender, very slightly expanded proximad, but otherwise subequal, ventral margins unarmed, cephalic genicular lobe with a very short fixed spine, caudal genicular lobe with a long mobile spine; cephalic tibiae very slightly longer than the femora, unarmed dorsad, ventral margins with four paired spines and one median one on the distal border, principal spurs considerably shorter than the large genicular spine on the cephalic femora; cephalic tarsi considerably more than half the length of the tibiae, compressed, the metatarsus slightly more than half the tarsal length, arolia absent. Median femora distinctly but not greatly shorter than the cephalic femora, similar in structure but with a large mobile spine on the cepha_ 1906.] NATURAL SCIENCES OF PHILADELPHIA. 291 lie genicular lobe, in length, however, shorter than that on the caudal lobe; median femora and tibiae as in the cephalic limbs but shorter. Caudal femora elongate, slightly more than twice the length of the median, considerably inflated in the proximal half, cephalic ventral margin unarmed, caudal margin with a number of spines, closely Fig. 7. — Diestrammena palpata n. sp. Lateral view of type. (X 2.) placed proximad, sparsely distributed distad; caudal tibiae slightly longer than the femora, dorsal margins with thirty to thirty-fom- fixed spines, dorsal spurs very large, the caudal as long as the metatarsus; metatarsi about as long as the remaining tarsal joint. General color burnt umber, meso-dorsal region suffused with clove brown. Eyes vandyke brown; antennae annulate much as D. longipes, but the color is duller, the larger annulations narrower and the smaller 292 PROCEEDINGS OF THE ACADEMY OF [May, ones more numerous; palpi dark except the distal joint which is pale. Cephalic and median limbs cream-buff, proximal three-fourths of the femora obscurely maculate with mummy brown, a clear light pregen- icular annulus present, genicular region of the femora and tibise seal brown, remainder of tibise of the pale ground Color. Caudal femora cream buff overlaid and mottled with burnt umber proximad and clove brown distad ; caudal tibii3e, except for a certain amount of scattered proximal clove brown maculation, cream buff, spines tipped with dark brown, extreme distal portion of the tibige and caudal margins of the tarsal joints washed with seal brown. Measurements. Length of body, 21 mm. Length of antenna, circa 120 '' Length of distal joint of the palpus, 8 " Length of caudal femur, 29 " Length of caudal tibia, 30.5 " The type is the only specimen examined. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 293 FURTHER KNOWLEDGE OF SOME HETEROGNATHOUS FISHES. PART I. BY HENRY W. FOWLER. In this paper I have attempted to arrange systematically the Heter- ognathi contained in the collections of the Academy of Natviral Sciences of Philadelphia. This group is largely made up of collections formed by John Hauxwell, James Orton and H. H. Smith. Nearly all of Cope's typical specimens, besides a number of comparatively rare or obscurely defined species, are represented. Owing to the increase in knowledge of the vast fresh-water ichthyic fauna of South America, it is believed that further detailed information concerning rare, nominal or species only known from the original specimens will be welcomed. ERYTHRINID^. Hoplias malabaricus (Bloch). Macrodon trahira Cope, Proc. Amer. Philos. Soc. Phila., XI 1SG9-70 (August 19, 1870), p. .5GG. Pebas. John Hauxwell. Cope, Proc. Acad. Nat. Sci. Phila., 1871 (January 16, 1872), p. 257. Ambyiacu River. John Hauxwell. Cope, Proc. Amer. Philos. 8oc. Phila., XVII, 1877-78 (May 17, 1878), p. 694. Peruvian Amazon. Prof. J. Orton. Macrodon tareira Cope, I.e., XXXIII, 1894 (Januarj'^ 5), p. 84. Brazilian province of Rio Grande do Sul. H. H. Smitli. A series of 28 examples represented by those in the Hauxwell and Orton collections from the Ambyiacu and Peruvian Amazons, and the H. H. Smith collection from Rio Grande do Sul, also others from Bahia, Rio das Vehlas and Surinam. They show: Head 2f to 3^; depth 3f to 51; D.ii or iii, 10 1 to 12 1, mostly 111, and only rarely 10 1 or 12 1 ; A. ii or III, 8 I or 9 I, most always 8 i, or only rarely 9 1 ; scales 36 to 42 in lateral line to base of caudal, with 2 to 4 more on latter; 6 scales obliquely back from origin of dorsal to lateral line, only rarely 5 ; between lateral line and base of ventral 5 or 6 scales, usually 5; 15 to 19 scales before dorsal; snout 4 to 4| in head, measured from tip of upper jaw; eye 41 to 8; maxillary If to 2^; interorbital 3^ to 4f; total length of body 2{ to 12^ inches. OPHIOCEPHALOPS subgen. nov. Type Erythrinus unitoeniatus Agassiz. Dorsal fin rounded or posterior rays not elevated. i!'0us). Fig. 6. Curimatus cyprinoides Cope, Proc. Acad. Nat. Sci. Phila., 1871 (1872), p. 258. Ambyiacu River, Equador. John Hauxwell. Of 11 examples 4 before me have the squamation fairly perfect. They show the scales in the lateral line ranging from 45 to 48 when 1906.] NATURAL SCIENCES OF PHILADELPHIA. 301 counted to base of caudal, and 4 or 5 more on the latter. The figure is based on the largest of these. <■■ ■«! Fig. C. — Curimatd cijprinoide.'^ (Liniux'us). Subgenus CURIMATA Cloquet. Postventral region trenchant. Scales small, 51 to 70 in a lateral count, and forming more or less even longitudinal series. In view of the uncertainty of identification of Salmo edentulus Bloch, which is more like the example I shall identify as Curimata schom- hurgkii, I have allowed this subgenus to fall with it and related species. Further, its identification with Salmo cyprinoides Linnaeus may be questioned, and if any credit is attached to Bloch's figure the scales are a little large. Curimata oopei sp. nov. Fig. 7. Head 3; depth 2J; D. iii, 9; A. iii, 9, i; P. ii, 13; V. ii, 9; scales 50? in lateral line to base of caudal (squamation injured), and about 4? more on latter; about 15 scales between origin of dorsal obliquely down to lateral line; about 9 series of scales between lateral line verti- cally to origin of ventral; about 9 scales in a vertical series between lateral line and origin of anal; about 28 scales before dorsal; width of head 2^ in its length; depth of head 2J; snout 3^; eye 3^; width of mouth 3^; interorbital space 2^; base of dorsal 2; least depth of caudal peduncle 2f ; base of anal 2; ventral (damaged) 1^^. 302 PROCEEDINGS OF THE ACADEMY OF [June, Body well compressed, greatest depth at origin of dorsal, and back well elevated. Upper anterior profile gibbous or convex. Edge of back rounded. Caudal peduncle compressed, its length about f in least depth. Head rather constricted below, upper surface convex and with a nearly straight profile to occiput. Nuchal region convex. Head of rather even width. Snout broad and obtuse, especially when viewed Fig. 7. — Curimata copei Fowler. from above, and a little produced beyond tip of mandible. Eye large, its center near first f in length of head, and a trifle longer than deep. Adipose eyelids a little broad. Mouth broad, a little inferior, and symphysis with a little knob fitting in a depression in upper jaw. Mandibular angle broadly obtuse. Tongue thick, rounded, hardly free in front, and not broad. Nostrils adjoining, superior, a little nearer front of eye than tip of snout and posterior larger. Anterior nostril circular, and with a cutaneous margin. Interorbital space broad, a little elevated, and flattened medianly. Postorbital about equal to preorbital or about size of pupil. First infraorbital a little 1906,] NATURAL SCIENCES OF PHILADELPHIA. oO.'^ more than half length of second which is long. Preopercle with several distinct flutings. Gill-flap rather narrow. Opercle smooth. Gill-opening extending forward not quite opposite to posterior margin of pupil. Upper cleft of gill-opening extending forward a little more than at a point equal to last f in length of head. About 15 + 28? small short rudimentary filamentous-like rakers on first arch. Gill-filaments long, longest about 2 in horizontal orbital diameter. Isthmus nearly forming an equilateral triangle. Scales moderately small, adherent, and forming longitudinal series parallel with lateral line. Scales below lateral line, or those on breast and abdominal region, a little larger than others. Scales passing over ventral carina. No narrow median naked strip from occiput to dorsal, scales passing over. Head naked. Scales on chest a little smaller than those on breast. No scales on dorsals, pectorals and ventrals. Base of caudal with scales, and along basal region of anal. Base of ventral inside with a broad flat scaly flap, and another flap, but narrow and at present equal to about f of fin (damaged), placed in axil. No flap in pectoral axilla. Lateral line of simple tubes, continuous, slightly decurved at first and then straight to caudal. Dorsal inserted about midway in vertical between tip of snout and posterior basal margin of adipose fin. Dorsal high, first branched rays evidently highest and others graduated down. Adipose fin well de- veloped and its base about last f in space between dorsal and base of caudal. Anal evidently low and its origin a little nearer base of caudal, in vertical, than origin of ventral. Caudal emarginate (damaged), and rays well branched. Pectoral small and low. Ventral inserted about opposite base of second branched dorsal ray, and reaching about f of distance to anal (damaged). Vent close in front of anal. Color in alcohol brassy-brown, back with a dull olivaceous tinge. Fins all brownish. Iris deep brown. Length 4f inches. Type, No. 8,201, A. N, S. P. Surinam. Smithsonian Institution. One example, the type. This was long ago considered a new species by Cope. It appears to be closely related to Curimata schomburgkii (Giinther), but differs at once in the deep body, the upper profile of which is more gibbous anteriorly. (Named for Prof. Edward D. Cope.) Curimata schomburgkii (Gunther). Fig. 8. Body well compressed. Edge of back rounded. Caudal peduncle compressed. Head compressed, a little constricted inferiorly, upper surface convex. Nuchal region also convex. Head of rather even 304 PROCEEDINGS OF THE ACADEMY OF [June, width. Snout broad, blunt when viewed above. Mouth broad and symphysis with a knob fitting in a cavity of upper jaw. Mandibular angle broadly obtuse. Tongue small, as usual rather far back, flattened, a little thick and only edges free. Interorbital space broad and a little Fig. S. — Curimata schomhurgkii (Giinther). elevated convexly, not especially flattened medianly. Gill-opening forward till not quite opposite posterior margin of pupil. Upper cleft of gill-opening extending forward f in length of head. Gill-rakers 10? + 32?, small short rudimentary-like weak filaments. Gill-filaments about 2 in eye. Isthmus a little broad and triangular. Scales more or less adherent, passing over ventral keel and ridge before dorsal. Scales on chest a little smaller than those on breast. Base of ventral inside with a broad scaly flap. Vent close to anal. Color in alcohol brassy-brown. Length 4f inches (caudal damaged). Two examples from Surinam, larger described above. Dr. Hering. The smaller example is in better preservation. It shows: Head 2^; depth 2f ; D. iii, 8; A. iii, 9, i; scales 54 in lateral line to base of caudal, and 4 more on latter; snout 3f in head; eye Sf; pectoral 1|; ventral 1906.] NATURAL SCIENCES OF PHILADELPHIA. 305 If; upper caudal lobe about 1. Rakers distinct. First branched ray of dorsal longest, and reaching well beyond others, when fin is depressed, or to origin of adipose fin. Adipose fin long, its length along posterior margin but little less than length of its base. Anterior anal rays elevated. Caudal long, deeply forked, and each lobe well pointed. Pectoral long, pointed, and reaching origin of ventral. Ventral long and pointed, and reaching origin of anal. Semitapicis laticeps (Valenciennes). Curimatus altamazonicus Cope, Proc. Amer. Philos. Soc. Phila., XVII, 1877-78 (May 17, 1878), p. 684. Xos. 21,118 (tj-pe) to 21,120, A. N. S. P., cotypes. Peru\dan Amazon. Prof. J. Orton. Coll. of 1873. Head 3i; depth 2f ; D. ii, 9, i; A. in, 12, i; P. i, 16; V. ii, 8; scales about 88 in lateral line to base of caudal (squamation injiu-ed), and 6 more on latter; width of head 2^ in its length; depth of head If ; snout 3f; width of mouth 3f ; interorbital space 2^; base of dorsal 2^ ; base of anal 2 ; least depth of caudal peduncle 3 ; pectoral (damaged) 2 ; ventral (damaged) 1*. No gill-rakers, and filaments H in eye. Color in alcohol brown, paler on lower surface or below lateral line. Head dark on top, like back, and sides and under portions pale. Fins all brownish. Eye dusky. Length (caudal damaged) 7f inches. Type. The other cotypes show: Head 3 and 3,^; depth 2f; D. ii, 10, i; A. II, 13, i; total length of body 6i and 7^ inches respectively (caudals damaged). Psectrogaster ciliatus (MuUer and Troschel). Curimatus rutiloides Cope, Proc. Acad. Nat. Sci. Phila., 1871 (1872), p. 258. Amb\aacu River, Equador. John Hauxwell. (Xot of Kner.) Curimatus cyprinoides Cope, I.e., p. 291. Between the mouth of the Rio Negro and the Peruvian Amazon or Ucaj^ale River. Robert Perkins. (Not of Linna?us.) Head 3; depth 2J; D. iii, 9, i; A. iii, 7, i; scales 48 in lateral fine to l)ase of caudal, and 5 more on latter; 12 scales in an oblique series back from origin of dorsal to lateral line; 9 series of scales obliquely back from origin of ventral to lateral line; 9 series of scales obliquely forward from origin of ventral to lateral line; about 32 scales before dorsal; snout 3f in head; eye 3^; width of mouth 3f ; interorbital space 2^ ; pectoral If ; ventral If ; least depth of caudal peduncle 2|. Rakers none. Gill-filaments about' f of orbit. Postventral carina well developed, scales strongly pectinate which form it. Scales on body all more or less ctenoid. Upper and lower profiles of body more or less evenly convex. Scales on trunk enlarged anteriorly on middle of side, so that longitudinal series are formed which are more or less parallel with course of lateral line above, and below converging as they 306 PROCEEDINGS OF THE ACADEMY OF [June, approach caudal. In alcohol brassy-brown with more or less silvery. Back till about level with occiput a dull olive-brown. Upper surface of head brownish. Body mostly washed with silvery. Fins plain brown, and pale like side. Iris brown. Length 5^ inches. This is the largest example. It is from between the mouth of the Rio Negro and the Peruvian Amazon. Robert Perkins. Also another example with same data. These are labelled, evidently in Cope's handwriting, as "Curimatus cyprinoides v. ajf." Besides the above are 5 examples from the Ambyiacu river in the Hauxwell collection. Cope's label reads ' ' Curimatus rutiloides. ' ' An examination of the air-vessels of two of these examples shows that it persists nearly as far posteriorly till opposite base of penultimate anal ray. It is thus apparent that Cope may have intended all the ex- amples before me to represent his C. rutiloides, but the original labels in his own handwriting would lead to the above allocating of the references. The form called Psedrogaster amazonica by Dr. and Mrs. Eigenmann I cannot distinguish in my examples. Potamorhina pristigaster (Steindachner). Cope, Proc. Amer. Philos. Soc. Phila., XVII, 1877-78 (May 17, 1878), p. 685. Peruvian Amazon, probably from Nauta. Prof. J. Orton. Coll. of 1873. Head 3; depth 2 to 2^, a little larger in smaller specimen; D. iii, 9; A. Ill, 11 to III, 13, i; ventral scutes 22 to 28; scales (squamation damaged) about 90 in lateral line to base of caudal and several more on latter ; snout 3^ to 3f in head ; eye 4|- to 4| ; width of mouth 3^ to 3-^ ; interorbital space 2^ to 2f ; total length of 3 examples 6 to 9 inches (caudal damaged). Anodus elongatus Agassiz. Fig. 9. Anodus steatops Cope, Proc. Amer. Philos. Soc. Phila., XVII, 1877-78 (Mav 17, 1878), p. 683. Nos. 21,498 (type) and 21,499, A. N. S. P., cotypes. Pebas. Prof. J. Orton. Coll. of 1877. Body rather plump. Head broadly convex above, a little restricted below. Snout when viewed above rather rounded. Edges of jaws not sharp. Lips thin. Rami of mandible well elevated inside mouth. Tongue large, elongate, flattened above, rounded in front and free. Inside mouth upper membrane broad. Interorbital space broadly convex. Gill-opening extending forward to front margin of orbit. Rakers 66? + 100?, long, slender, or very fine and numerous and long- est equal to 1|- eye-diameters. On inner edge of first branchial arch also a series shorter in length. Longest filaments nearly equal orbit. Isthmus long, narrow and slender. A triangular naked space extend- ing back on occiput. Edges of body rounded except those of lateral 1906.] NATURAL SCIENCES OF PHILADELPHIA. 307 line. Vent close in front of anal. Color in alcohol dull olive-brown, darker on back, and top of head deep brownish. Fins dull greenish, bases of caudal lobes blackish. A brownish or dusky blotch on several scales about and in lateral line about opposite tip of depressed dorsal. Greenish tints all probably due to the preservative. Iris and adipose eyelid brownish. Length (caudal damaged) 10^ inches. Type. In explanation of Cope's remarks it may be stated that these fishes Fig. 9. — Anodus elongatus Agassiz. agree best with Anodus elongatus Agassiz, a species from which I am unable to separate it. The pectorals and ventrals, although damaged, reach more than half the distance credited. The ventrals are more anterior in position, or originate under the anterior portion of the dorsal, and not opposite its middle. EIGENMANNINA gen. nov. T3-pe Anodus melanopogon Cope. Head large, especially opercular apparatus, and upper profile strongly concave. Mandible well produced beyond tip of upper jaw so that mouth is superior. Eye above middle in depth of head. Body heavy anteriorly, and tapering towards caudal. In other respects more or less allied to Anodus. (Named for Dr. Carl H. Eigenmann, of the Chair of Zoology in Indiana University, a well-known authority on South American fishes.) 308 PROCEEDINGS OF THE ACADEMY OF [June, Eigenmannina melanopogoa (Cope). Fig. 10. Anodus melanopogon Cope, Proc. Amer. Philos. Soc. Pliila., XVII, 1877-78 (May 17, 1878), p. 682. Nos. 21,227 (type) to 21,232, A. N. S. P., cotjijes. Peruvian Amazon. Prof. J. Orton. Coll. of 1873. Width of head 3^ in its length; interorbital space 4^. Body, head and caudal peduncle compressed. Upper surface of head narrow, convex, and lower surface constricted. Snout rounded and convex when seen from above. Jaws more or less flattened or spatulate. Each ramus of mandible well elevated inside of mouth. Lips hardly developed or very thin. Tongue small, narrow, rather thick, fleshy. Fig. 10. — Eigenmanmna melanopogon (Cope). and united with floor of mouth by a median fleshy frenum. Upper buccal membrane rather narrow. Interorl^ital space a little elevated convexly and flattened medianly. Gill-opening large, extending for- ward till a little before front rim of orbit. Vent close in front of anal. Color in alcohol more or less silvery, back brown fading to white on sides and under surface. Top of head brown, sides and lower surface silvered white. Fins pale brownish, dorsal and caudal a trifle darker, and each ray of most fiixs speckled or spotted rather indistinctly with darker brownish. Iris rather brassy. Length 2|- inches (caudal damaged). Type. One example shows 35? + 54? slender rakers, and longest about equals eye or much longer than filaments. With reference to the original description, Cope evidently intended to state that the base of the first dorsal ray was three millimeters nearer the end of the muzzle than the base of the caudal. Both pectorals and ventrals reach m.ore than half way in the spaces stated. Cope evi- dently counts 128 rows of scales from the occipital region, and those on base of caudal. The origin of the ventral is below the bases of 1906.] NATURAL SCIENCES OF PHILADELPHIA. 309 the anterior dorsal rays. Most of rays of the fins are specked with brownish. This species is related to Eigenmannina orinocensis (Steindachner), but differs in the fewer scales, absence of the dark median lateral blotch and the coloration of the caudal. CHILOMYZON subgen. nov. Type Prochilodus steindachneri sp. nov. Scales large, less than 40, usually between 33 and 38 in a lateral count. Mouth disk-like and inferior. Related to Prochilodus Agassiz. (XeVjK^j lip; /x'jCw, to suck.) Prochilodus steindachneri sp. nov. Fig. 11. Head 3^; depth 2f ; D. iii, 9, i; A. iii, 8, i; P. i, 14; V. i, 8; scales 34 in lateral line to base of caudal and 3 more on latter; about 14 scales Prochilodus steindachneri Fowler. before dorsal; 7 scales obliquely back from origin of dorsal to lateral line; 6 scales obliquely forward from origin of ventral to lateral line, and same number in similar count from origin of anal; about 16 scales from isthmus to origin of ventral ; 3 scales obliquely back from origin 21 310 PROCEEDINGS OF THE ACADEMY OF [June, of adipose fin to lateral line; width of head 1| in its length; depth of head 1^; snout 3f ; eye 3f ; width of mouth 2|-; interorbital space 2^0 ; third dorsal ray 1|^; third anal ray 2; length of pectoral If; of ventral 1^; least depth of caudal peduncle 2h. Body deep, compressed, rhomboid in shape, and suggesting certain Cyprinidce. Upper profile more or less evenly convex, back elevated, and greatest depth at origin of dorsal. Lower profile at first straight for a good distance, and becoming convex posteriorly. Caudal peduncle compressed, short, and its least depth about equals its length. Head small, compressed or restricted a little below, and upper surface broadly rounded. Lower surface of head also flattened. Snout broad, obtuse, fleshy, and produced well beyond tip of mandible. Eye circular, a trifle anterior, and above center in depth of head. Eyelid narrow. Mouth broad, and in profile of gape curved downwards. When opened mouth is broad, directed inferiorly, and jaws furnished with broad thick and fleshy lips formed somewhat as a disk. Margin of this disk with a single series of small weak or movable ciliiform teeth. In front of each jaw behind outer series a short second or inner series of similar ones, convex or angular in its course, and with bend or angle directed inwards. Aperture of mouth small. Tongue small, hardly free from floor of mouth. Nostrils close together on side of snout above, and much nearer upper front rim of orbit than tip of upper jaw. Anterior nostril circular, with its posterior cutaneous margin more or less concealing posterior which is thus lunate. Inter- orbital space broad and convex. Infraorbital rim narrow, lowest or most posterior largest. Preorbital a little swollen, with a deep cavity and large thick lip and maxillary more or less filling it when mouth is closed. Opercle striate. Opercular flap broad and rather cutaneous. Gill-opening extending forward till nearly opposite posterior margin of orbit. Rakers none. Filaments of inner series a little longer than those in outer, or about equal to f of orbital diameter. Isthmus broad. Branchiostegals long, broad, subequal, and 4 on each arch. Scales large, of more or less even size, disposed in longitudinal series parallel with lateral line, and each one with several strise. Margin of each scale also a little rough. Small scales on bases of caudal and anal, otherwise fins and head naked. A pointed scaly flap in axil of ven- tral equal to about ^ of length of ventral. Both predorsal and post- dorsal regions with a median keel, former most distinct, and latter also extending on upper surface of caudal peduncle behind adipose fin though still less distinct. Behind first dorsal also an indistinct lateral keel on each side, and below and posterior to adipose fin they are also 1906.] NATURAL SCIENCES OF PHILADELPHIA. 311 evident. Preventral region keeled in similar manner to postdorsal, only median keel most pronounced. Postventral and postanal region keeled, former may be considered almost trenchant. Lateral line continuous, of simple tubes, a little above middle in depth of body and continuous to caudal. Dorsal high, third ray longest, and origin of fin would fall in vertical about midway between tip of snout and base of adipose fin. Anal small, anterior rays longest, margin of fin concave, and its origin nearer base of caudal than origin of ventral. Adipose dorsal small, its base inserted about opposite bases of last anal rays. Caudal long, deeply emarginate and end of each lobe apparently more or less pointed. Pectoral rather long, pointed, and extending beyond origin of dorsal or about f of space to ventral. Ventral inserted nearl}' opposite middle •of base of dorsal or about midway between origin of anal and that of pectoral, and extending about | of distance to former, ^'ent close in front of anal fin. Color in alcohol more or less pale brownish washed with silvery- purplish. Back a little darker than side and lower surface. Body also with about fifteen or rnore indistinct vertical or transverse purplish- dusky bands. Dorsal with about six series of brownish spots on rays. Other fins pale plain brownish. Iris brassy. Length 5 inches. Type, No. 8,207, A. N. S. P. Parahyba, Brazil. jMuseum of Com- parative Zoology, Cambridge, Mass. Only one example, the type. This species is closely related to Prochilodus corimhata (Kner) = Salmo corimhata -Natterer, in Kner = P. nigricans Kner, nee Agassiz = P. oligolepis Giinther. Dr. Steindachner's account of the last does not give the coloration. Other related species are P. hiimeralis Giinther and P. vimboides Kner, both differing in proportions, etc. (Named for Dr. Franz Steindachner.) Subgenus PROCHILODUS Agassiz. Scales small, more than 40 or usually between 40 and 60 in a lateral count. Mouth similar to that of Chilomyzon. Prochilodus ortonianus Cope. Fig. 12. Proc. Amer. Philos. Soc. Phila., XVII, 1S77-7S (May 17, 1878), p. 685. No. 21,267 (type), A. N. S. P., cotype. Nauta, Peru. Prof. J. Ortou. Coll, of 1873. Width of head 1§ in its length ; width of mouth 2f ; interorbital space 2. Body robust, compressed, and caudal peduncle similar. Head broad, robust, convex on upper surface, and lower surface but little restricted 312 PROCEEDINGS OF THE ACADEMY OF [June^ though more evenly convex. Width of head more or less even. Snout broad, convex above and broadly rounded when viewed from above. Interorbital space convex and median narrow fontanel extend- ing from internasal space to occiput. Head with arborescent mucous canals on upper side posteriorly and on suborbital region. Gill-open- ing falling a trifle short of posterior margin of orbit. Rakers none. Gill-filaments about 1^ in eye. Isthmus broad. Branchiostegals 4, large, well developed and subequal. Scales a little rough and each one Fig. 12. — Prochilodus ortonianus Cope. with several striae. Predorsal region slightly keeled. Postdorsal region rounded. Upper and lower surfaces of caudal peduncle a little flattened. Preventral region flattened medianly. Postventral region sharply keeled medianly and an obsolete keel on each side, all three converging towards vent. Color in alcohol at present dull olivaceous or brassy-brown, back with more or less dull metallic shades. Fins all pale brownish, dorsal and caudal each with about 8 series of brownish spots on each ray of former, and about 6 on each lobe of caudal, so that on latter they form more or less transverse bands. Iris brownish. Length 7| inches. Type. Prochilodus cephalotes Cope. Fig. 13. Proc. Amer. Philos. Soc. Phila., XVII, 1877-78 (May 17, 1878), p. 686. No. 21,211, A. N. S. P., type. Peruvian Amazon. Prof. J. Orton. Coll. of 1873. Width of head If in its length; width of mouth 2^; interorbital space 1906.1 NATURAL SCIENCES OF PHILADELPHIA. 313 2. Width of head greatest in postero-supraocular region. Gill-fila- ments about f of orbit, Predorsal region apparently keeled. Pre- ventral region apparently rounded, and postventral region keeled. Color in alcohol more or less pale plmnbeous and somewhat silvered. Back darker or dusky-plumbeous. Upper surface of head like back, Fig. 13. — Prochilodus cephalotes Cope. and lower surface like that of belly. Damaged dorsal with at least 4 series of deep brownish spots on rays. Caudal also with transverse series of spots, other fins plain. Iris brassy. Length 2f inches. Otherwise like the preceding. Frochilodas theraponura sp. nov. Fig. 14. Prochilodus insignis Cope, Proc. Acad. Nat. Sci. Phila., 1871 (January 16, 1872), p. 258. Ambyiacu River, Equador. John Hauxwell. Head 2|; depth 2i; D. iii, 9; A. iii, 8, i; P. ii, 11?; V. ii, 8; scales about 40? (squamation damaged), and 5? more evidently on base of caudal; about 10? scales obliquely back from origin of dorsal to lateral line; about 8? scales obliquely forward from origin of anal to lateral line; 14 scales before dorsal; width of head about 2 in its length; depth of head 1^; snout 3^; eye 3; width of mouth 3; interorbital space 2^; length of fu-st branched dorsal ray 1^; base of dorsal 2^; base of anal 2^; length of upper caudal lobe 1; pectoral If; ventral If; least depth of caudal peduncle 3^. 314 PROCEEDINGS OF THE ACADEMY OF [June, Body robust, compressed, back but little elevated so that upper profile would form an obtuse angle at origin of dorsal, and lower profile more or less evenly convex. Greatest depth at origin of dorsal. Caudal peduncle compressed, and its length about | its least depth. Head compressed, convex on upper surface, and sides constricted below. Greatest width in postero-supraocular region, and upper profile nearly straight or only a trifle concave. Snout short, convex, rather broad, and rounded when viewed from above. Eye circular, Fig. 14. — Prochilodus theraponura Fowler. anterior, and a trifle above middle of depth. Eyelid narrow. Jaws even, a little broad, and together with mouth and nostrils like in pre- ceding species of Prochilodus. Interorbital space convex, and median fontanel extending from internasal region to occiput, broader. Post- orbital largest in suborbital rim. Mucous channels on cranium later- ally not pronounced. Opercle with indistinct curved transverse striae. Gill-flap apparently narrow. Gill-opening extending forward a little in advance of posterior margin of orbit. Rakers none. Gill-filaments equal, about f length of orbit. 1906.] NATURAL SCIENCES OF PHILADELPHIA, 315 Isthmus a little narrowly compressed. Branchiostegals 4, large and subequal. Scales mostly fallen, little rough, apparently disposed in even longi- tudinal series parallel with lateral line, each one with one or more striae, and apparently of more or less even size. Small scales appa- rently along bases of dorsal and anal, and on that of caudal. No trace of ventral flap remains. Predorsal region keeled. Postdorsal region rounded. Upper and lower surfaces of caudal peduncle, preventral and postventral regions keeled. Lateral line (damaged) continuous, on base of caudal, nearly straight, more or less median, and of simple tubes. Origin of dorsal about midway between tip of snout and base of adipose fin, first branched ray longest, and others apparently gradu- ated down. Adipose fin wnth base over those of last anal rays. Origin of anal much nearer base of caudal than origin of ventral, or nearly midway between base of last dorsal ray and that of caudal, and anterior rays apparently longest. Pectoral low, rather long, and reaching ventral. Ventral with origin a trifle in advance of middle of base of dorsal, and reaching nearly | of space to anal. Vent close in front of anal. Color in alcohol pale plumbeous, sides and lower surface more or less silvered, and back dusky-plumbeous. Upper siu-face of head dusky- plumbeous, sides and lower surface silvered. Fins all pale brownish. Dorsal with four well-defined broad deep brown cross-bands. Caudal with two similar colored oblique bands on each lobe, and a median or horizontal one from base of fin to tips of middle rays. Anal with lower anterior tip of fin brownish, and another horizontal transverse band from origin of fin to tips of more posterior rays also of brownish. Pectoral and ventral apparently plain pale brownish. Iris dull brownish. Length 2{ inches. Type, No. 8,033, A. N. S. P. .Ajiibyiacu River, Equador. John Hauxwell. Only one example. Formerly identified by Cope with Prochilodus insignis Kner, this species may at least provisionally be regarded as distinct. This is in view of the identity of Prochilodus insignis Jardine and P. insignis of Kner, and later of Dr. Glinther, not yet having been proved. Schom- burgk's figure, as presented by Jardine, shows each lobe of the caudal with five oblique bands besides the median one, which is also in agree- ment in the description. Dr. Giinther's Amazon examples are said to have only three or four bands across each lobe besides the median one. 316 PROCEEDINGS OF THE ACADEMY OF [June, Prochilodus iheraponura may thus be said to differ from all of the others in the fact that it has but two obUque dark bars on the caudal, aside from the median one. {dtpdr.iov, servant, with reference to the caudal bands, like those of Therapon; obpa^ tail.) Froohilodus amazonensis sp. nov. Fig. 15. Head 3^; depth 2f ; D. iii, 9, i; A. iii, 8, i; P. i, 16; V. i, 8; scales about 43 in lateral line to base of caudal, and 5 more on latter; 11 scales obliquely back from origin of dorsal to lateral line; 8 scales obliquely forward up from origin of anal to lateral line; 10 scales obliquely up from origin of ventral to lateral line; 15 scales before dorsal; wddth of head If in its length; depth of head 1^; snout 3; eye 3^; maxillary 3j^ ; Fig. 15. — Prochilodus amazonensis Fowler. width of mouth 2f ; interorbital space 2; length of base of dorsal 1|; length of base of anal 2|; length of pectoral If-; ventral (damaged) U, least depth of caudal peduncle 2f . Body rather deep, compressed, back but little elevated so that upper profile would form a rather obtuse angle at origin of dorsal, at which 1906.] NATURAL SCIENCES OF PHILADELPHIA, 317 point is also greatest depth. Lower profile more or less evenly convex. Caudal peduncle compressed, and its length equals about f its least depth. Head robust, broad and convex on upper surface with profile nearly straight or only very slightly concave, and lower surface- but little restricted, though more evenly convex. Width of head more or less even. Snout broad, rather short, convex above, and broadly rounded when seen from above. Eye circular, anterior, and a little above middle in depth of head. Eyelid narrow. Jaws broad, rounded, almost even or upper a trifle produced. Lips thick and fleshy, and teeth as in preceding species. Upper buccal flap broad and with a median fleshy tubercle in front. liOwer buccal flap broad. Tongue and nostrils as in preceding species. Interorbital space also similar, and fontanel rather narrow. Postorbitals largest in suborbital chain. Cranium 'and suborbitals with mucous canals, some arborescent. Opercles with traces of very faint radiating strise. Gill-flap narrow. Gill-opening extending forward till nearly opposite posterior margin of pupil. Rakers in form of 4? + 8? or more short inconspicuous fleshy points along outer edge of first branchial arch. Filaments long, series on inner edge of first arch longer, equalling about f of orbital diameter. Isthmus broad. Branchiostegals 4, large, well developed and subequal. Scales rather small, striate, rather smooth, in even longitudinal or horizontal series parallel with lateral line, and of more or less even size. Small scales along bases of dorsal and anal and on that of caudal. Ventral with a rather short pointed scaly flap about ^ length of (dam- aged) fin. Predorsal region with a median keel. Postdorsal region rounded. Upper and lower surfaces of caudal peduncle flattened. Preventral region flattened posteriorly, and with a low or obsolete keel anteriorly. Postventral region trenchant and with an obsolete keel on each side, convergent posteriorly. Lateral line continuous, median on side, extending on base of caudal, decm'ved a little in front, and mostly of simple tubes except those on first 6 scales which are arl^ores- cent. Origin of dorsal falling in vertical about midway between tip of snout and origin of adipose fin, rays long, first branched one apparently highest, and others graduated down so that last one is less than half length of first. Adipose dorsal small, its base over bases of posterior anal rays. Anal inserted nearly midway between base of last dorsal ray and base of caudal, anterior rays longest and edge of fin a little concave. Caudal emarginate, lobes apparently broad. Pectoral low, 318 PROCEEDINGS OF THE ACADEMY OF [Juiie, pointed or upper rays longest, and reaching ventral. Ventral inserted about opposite first third of base of dorsal. Vent close in front of anal. Color in alcohol pale or dull brownish more or less silvered or with brassy reflections. Back with pale or dull purplish reflections becom- ing very dilute greenish on sides. Each scale more or less paler on outer or exposed portion, so that rather pale longitudinal lines are formed on back. Upper surface of head brownish, sides and low^er surface silvered. Fins all pale brownish, dorsal with about eight dusky cross-bands. Caudal with a median blackish band from center of its base to tips of median rays, and each lobe with three oblique bands of similar color, but broader anteriorly. Anal with three horizontal dusky bands, lowest near tips of anterior rays, median beginning at origin of fin, and upper at posterior rays. Pectoral and ventral plain. Iris brownish. Length 4| inches. Type, No. 21,350, A. N. S. P. Lower Amazons. Prof. J. Orton. Coll. of 1874. Prof. E. D. Cope. One example. This species is also apparently closely related to Prochilodus ijisignis Jardine, but differs in the fewer blackish bars on the caudal lobes. It will possibly prove identical with the examples recorded by Dr. Giinther which have three dark bars on each caudal lobe. (Named for the Amazon river of South America, also written Ama- zons and Amazonas. The name Amazon is said to be derived from the Indian word Amassona or ' ' boat-destroyer, ' ' with reference to the destructive tidal phenomenon or proroca.) HEMIODOPSIS subgen. nov. Type Hemiodus microlepis Kner.' This group is distinguished from subgenus Hemiodus Miiller and Troschel by the small scales in a lateral count, at least 100 or more. {'H/ii, half; 68oh?, tooth; o4>t?, appearance, with reference to the subgenus Hemiodus.) Hemiodus microlepis Kner. Cope, Proc. Acad. Nat. Sci. Phila., 1871 (1872), p. 291. Between the mouth of the Rio Negro and the Peruvian Amazon. Robert Perkins. Cope, Proc. Amer. Philos. Soc. Phila., XVII, 1877-78 (May 17, 1878), p. 686. Peruvian Amazon. Prof. J. Orton. Head 4; depth 3y\; D. iii, 9, i; A. iii, 9, i ; scales about 105 (squama- tion injured) in lateral line to base of caudal, and 7 more on latter; about 24? scales obliquely back from origin of dorsal to lateral line; about 12? scales between lateral line and origin of ventral, and about same number between former and origin of anal ; snout 3^ in head ; eye 1906.] NATURAL SCIENCES OF PHILADELPHIA. 319 4f ; interorbital space 3; pectoral H; ventral 1|; least depth of caudal peduncle 2f. Eye midway in depth of head. Gill-opening extend- ing forward till about opposite middle of orbit. Rakers 22? + 38?, compressed, cuneated triangularly, and inner edge of each ciliate. Gill-filaments about equal to i of orbital diameter. Scales above pectoral anteriorly, both above and below lateral line, and on breast, enlarged. Scales on base of caudal also a little large. Scales on post- dorsal region of back formed into more or less convergent series and larger than those just below. Lower lobe of caudal much longer than upper, just the reverse of that indicated by Kner. Length 9 inches (caudal damaged). One example from Robert Perkins, taken between the mouth of the Rio Negro and the Peruvian Amazon. Also two smaller examples from the Peruvian Amazon. Prof. J. Orton. Coll. 1873 and 1877. Prof. E. D. Cope. These both show about 120 scales in lateral line to base of caudal, and 7 or 8 more on latter. Middle of lower lobe of caudal dusky longitudinally, and deepest basally. The Perkins example shows this as faint. CITHARININ.5^. Citharinus citharus (Geoffroy St. Hilaire). Citharinus geoffroii Giinther, Proc. Zool. Soc. London, 1896 (February 4), p. 223. Lake Rudolf. Dr. A. D. Smith. Head 2f ; depth 2^; D. iv, 14, i; A. v, 24, i; scales 83 in lateral line to base of caudal, and 5 more on latter; 21 scales in a vertical series between origin of dorsal and lateral line; 17 scales between origin of anal and lateral line in a vertical series; snout 4^ in head; eye 3^; width of mouth 3; interorbital space 3; base of dorsal 1|; base of anal 1^; least depth of caudal peduncle 3§; pectoral If; ventral H. Back rather elevated, profile forming an angle at origin of dorsal. Lower profile of body more or less evenly convex. Head becoming com- pressed below. Snout short and broad. Eye about circular. Inter- orbital space broad and a trifle convex. Gill-rakers not evident. Color in alcohol, back brownish from a little above lateral line, and lower surface and side silvery- white. Fins pale brownish. Iris pale straw-color. Length 2| inches. Two examples, the other a little smaller. They both differ from the original figure of Geoffroy St. Hilaire in the straight upper anterior profile. PITHECOCHARACIN.C subfam. nom. nov. (Anostomince Auct.) PITHECOCHARAX gen. nov. Type Salmo anostomus Linnseus. Snout narrow and conic, and mouth superior. 320 PROCEEDINGS OF THE ACADEMY OF [June, This name is proposed as Anostomus Klein, in Walbaum, Pet. Arted. Gen. Pise, III, 1792, p. 659 (type Salmo anostomus Linnaeus) = Gray, Cat. Brit. Mus. Fish. Gron., 1854, p. 153, is preoccupied in Orni- thology by Anostomus Bonnaterre, Enc. Meth. Ornith., 1790, p. xciii. (fJiOrjico^j ape, with reference to the short snout or nose; /a/jaf, Charax.) Pithecooharax trimaculatus (Kner). Schizodon trimaculatus Cope, Proc. Amer. Philos. Soc. Phila., XVII, 1877-78 (May 17, 1878), p. 690. Peru\aan Amazon. Prof. James Orton. Coll. of 1877. Head 4; depth 3; D. ii, 10, i; A. in, 8, i; scales 40 in lateral line to base of caudal, and 2 more on latter; about 16 scales before dorsal; 6 scales obliquely back from origin of dorsal to lateral line; 6 scales obliquely up from origin of ventral to lateral line ; 5 scales obliquely up from origin of anal to lateral line; pectoral 1^ in head; ventral If; snout 3 in head, from tip of upper jaw; eye 3f ; interorbital space 2^. Rakers absent. Color in alcohol with each scale marked medianly with a pale area so that longitudinal series are formed, becoming convergent posteriorly. Length 5^ inches (caudal damaged). Also another, 3 inches long, with same data. The opercular blotch, on both of my examples, appears to be superior according to such traces of it as remain, rather than inferior as indicated on Kner's figure. Pithecocharax ucayalensis sp. nov. Fig. 16. Head 3; depth 3^; D. in, 10; A. in, 8, i; P. i, 12?; V. i, 8; scales about 33? in lateral line to base of caudal (squamation injured), and appa- rently several more on latter; about 5? scales between origin of dorsal, obliquely back, and lateral line; about 4? scales between lateral line and origin of anal ; width of head about 2^ in its length ; depth of head , over middle of orbit, about 2; least depth of caudal peduncle 3^; snout 3^ in head, measured from tip of upper jaw; eye 3; interorbital space 3. Body elongate, compressed, back a little elevated or with upper pro- file a little more convex than lower, and greatest depth about middle of predorsal region. Predorsal, postdorsal, and preventral regions (desquamated) apparently rounded. Postventral region possibly with median keel? Caudal peduncle compressed, and least depth about | its length. Head compressed, elongate, and attenuate inferiorly, or with upper profile much more inclined than lower. Snout short, moderately broad, straight in profile and upper surface convex. Eye circular and 1906.1 NATURAL SCIENCES OF PHILADELPHIA. 321 a little anterior. Mouth superior and mandible well protruding in front, cleft nearly vertical. Maxillary small and vertical. Teeth large, sharp, crenulate, uniserial, and two median mandibulars largest and most conspicuous. Tongue rather broad, rounded, and hardly free. Anterior nostril in a fleshy tube about equal to diameter of pupil in length, lateral, and nearly midway in length of snout. Posterior nostril large, a little inclined, slit-like, and close to middle of anterior rim of orbit. Interorbital space broad and nearly flattened. Gill-opening extending about opposite posterior margin of pupil. Fig. 16. — Pithecocharax ucayalensis Fowler. Rakers short weak fleshy protuberances, and moderately numerous. Filaments well developed. Scales mostly fallen, large, evidently in rows parallel with lateral line, and also apparently all of more or less even size. Base of caudal apparently scaly. Lateral line evidently complete, and of simple tubes. Origin of dorsal nearly midway between tip of mandible and base of caudal. Origin of adipose fin placed about last fourth in space between origin of dorsal and base of caudal. Anal inserted well behind dorsal, or near middle of space between end of ventral and base of caudal. Pectoral low, and though' damaged apparently not reaching ventral. Ventral inserted a short distance before origin of dorsal, and though also damaged not reaching perhaps more than half way to anal. Color in alcohol brown, lower surface and fins paler. Body with 322 PROCEEDINGS OF THE ACADEMY OF [June, about seven broad ill-defined dark brown or dusky transverse bands. First three nuchal or predorsal, fourth from base of dorsal, fifth and sixth from postdorsal region, and seventh from base of adipose fin. Fins pale, caudal lighter and with a subbasal transverse blackish line. Iris brownish. Length (caudal damaged) 1^ inches. Type, No. 21,997, A. N. S. P. Peruvian Amazon or Ucayale River. Prof. J. Orton. Coll. Prof. E. D. Cope. This species is distinguished from Pithecocharax anostomus (Linnseus) and P. trimaculatus (Kner) chiefly by the coloration. (Named for the Ucayale or Ucayali River, sometimes called Peru- vian Amazon or Paro.) Sohizodon fasciatus Agassiz. Cope, Proc. Amer. Philos. Soc. Phila., XI, 1869-70 (August 19, 1870), p. 566. Para. De Schulte Buckow. Cope, I.e., XVII, 1877-78 (May 17, 1878), p. 689. Peruvian Amazon. Prof. J. Orton. Anostomus fasciatus Cope, Proc. Acad. Nat. Sci. Phila., 1871 (1872), p. 258. Ambyiacu River, Equador. John Hauxwell. Head 3|; depth 3^; D. ii, 9, i; A. ii, 8, i; scales 37 in lateral line to base of caudal, and 4 more on latter; 11 scales before dorsal; 5 scales obliquely back from dorsal to lateral line; 5 scales obliquely up from origin of ventral to lateral line; 4 scales obliquely up from origin of anal to lateral line; pectoral If in head; ventral If; snout 2-^^ in head, from tip of upper jaw; eye 3f ; interorbital space 2. Rakers none. Coloration faded dull brown. Length 6^ inches (caudal damaged). Para, Brazil. De Schulte Buckow. Prof. E. D. Cope. Three examples from the Peruvian Amazon, the largest about 7^ inches in length, show: Head 3f to 4; depth 3f to 3i; D. ii, 10, i; A. Ill, 8, i; scales 38 to 40 in lateral line to base of caudal, and 4 on latter. The color-pattern is constant. None of my examples show the dark transverse bar on the lower caudal lobe figured by Agassiz. The dark bars on the trunk are not absolutely vertical, the first two at least inclined a little back. They are also placed about equal in space and not extending across the ventral surface. Agassiz also indicates the eye too low in depth of head. One example from the Ambyiacu, 7^ inches long (caudal damaged). John Hauxwell. It agrees with the Orton examples. Laemolyta taeniata (Kner). Cope, Proc. Acad. Nat. Sci. Phila., 1871 (1S?2), p. 259. Ambjiacu River, Equador. John Hauxwell. Head 4; depth 5; D. ii, 10, i; A. iii, 8, i; scales 37 in lateral line to base of caudal, and 5 more on latter; 5 scales obliquely back from 1906. NATURAL SCIENCES OF PHILADELPHIA. 323 origin of dorsal to lateral line; 4 scales obliquely up from origin of ventral to lateral line; 4 scales from origin of anal obliquely up to lateral line; 13 scales before dorsal; pectoral If in head; ventral 1|; snout 3 in head, from tip of upper jaw; eye 3; interorbital space 2^. Rakers none. Length 4^ inches (caudal damaged). Two examples. The median dusky longitudinal band extends from the snout to the caudal. Kner's figure does not indicate it on the side of the snout or on the postocular region. PCECILOSOMATOPS subgen. nov. Type Characidium etheostoma Cope. Scales smaller transversely, about 4 from opposite origin of anal in an oblique series forward to lateral line. (/7oja:£Ao9^ variegated; ^rai^a, body; f^^'', appearance. Poecilosoma was an old name employed by Agassiz for certain Etheostomatincv , to which these fishes bear a certain resemblance.) km Fig. 17. — Characidium etheostunia Copi- Characidium etheostoma Cope. Fig. 17. Proc. Acad. Nat. Sci. Phila., 1871 (1872), p. 259, PI. 13, fio;. 3. No. 8,152 (t}-pe) and 8,153, A. N. S. P., cotj-pes. Ambyiacu River, Equador. John Hauxwell. Scales 11 before dorsal; width of head about 2^ in its length; inter- orbital space 3 J. Body well compressed. Muzzle compressed. Snout rather narrow. Lips thin. Teeth in jaws uniserial, even and fine, none on maxillaries. Interorbital space convex. Gill-opening 324 PROCEEDINGS OF THE ACADEMY OF [June, extending forward opposite posterior margin of pupil. Rakers short, lanceolate and weak, in small number. Filaments well developed. Isthmus a little broad. Each scale with several radiating strige. Vent placed about last f in space between origins of ventrals and anal. Color in alcohol brownish, lower surface paler, and fins still paler. Body with about eleven transverse brownish bands, at first of about equal width with alternate interspaces, but posteriorly interspaces becoming wider. First band nuchal, second and third predorsal, fourth just before origin of dorsal, fifth from middle of base of dorsal, sixth from just behind last dorsal ray, seventh and eighth from post- dorsal region with latter entirely in front of adipose fin, ninth and tenth across caudal peduncle, and eleventh which is dusky on base of caudal. A rather narrow deep brownish band from tip of snout across side of head and embracing lateral line to base of caudal. Dor- sal with at least three brownish longitudinal rather narrow bands or lines, upper ones less distinct. Iris brassy-silvery. Length 1| inches. From Dr. Steindachner's description it would hardly seem probal^le that his Characidium purpuratum is identical with C. etheostoma, in fact it falls in a different subgenus as the species are here, understood if there are but 2^ to 3 scales between the origin of the anal and the lateral line. Subgenus CHARACIDIUM Reinhardt. Scales large transversely, about 3 from opposite origin of anal in an oblique series forward to lateral line. Fig. 18. — Characidium stemdachneri Cope. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 325 Characidium steindachneri Cope. Fig. 18. Proc. Amer. Philos. Soc. Phila., XVII, 1877-7S (Mav 17, 1878). p. 688. No. 21,428, A. N. S. P., type. Peruvian Amazon. 'Prof. J. Orton. Coll. of 1873. Width of head 2h in its length. Gill-opening extending forward about opposite middle of pupil. Vent about last fourth in space between origins of ventral and anal. Color in alcohol brownish with ab&ut 9 transverse distinct dusky bars. Fins all pale brownish. Iris brownish. Length about lyV inches (caudal damaged). Other- wise like the preceding. Characidium tenuis (Cope). Chorimycfenis tenuis Cope, Amer. Nat., XXVII, 1894, p. 67. The upper waters of the Jacuhy River, in the Brazilian State of Rio Grande do Sul. H. H. Smith. Cope, Proc. Amer. Philos. Soc. Phila., XXXIII, 1894 (January 5), p. 86, PI. 5, fig. 3. Head 4f ; depth 6f ; D. in, 9; A. in, 6, i; scales 36 in lateral line to base of caudal and 2 more on latter; 4 scales obliquely back between origin of dorsal and lateral line; 3 scales obliquely back from origin of adipose fin to lateral line; 3 scales from opposite origin of ventral obliquely up and f on\-ard to lateral line ; 3 scales obliquely forward and up from origin of anal to lateral line; 12 scales before dorsal; width of head 2 in its length ; snout about 4^ ; eye 3f ; maxillar}^ 4^ ; interorbital space about 5; least depth of caudal peduncle 2J; length of depressed dorsal about 4; lower caudal lobe (damaged) about 4; pectoral 44-; ventral 5; depressed anal 5f. Edges of trunk rounded. Gill-opening extending about opposite posterior margin of orbit. Rakers about 5 + 7, short, lanceolate and rather well separated. Filaments a little less than vertical diameter of orbit. Isthmus a little broad and tri- angular. Vent about first f in space between origins of ventrals and anals. Color in alcohol brownish, rather pale, especially below, and side with silvery reflections. Each scale of back with darker brown edgQ than shade of body-color. About 8 indistinct brownish trans- verse bars. Fins pale brownish, shaded with darker, anal and ventral a little lighter. Iris pale brownish. Length 2yV inches (caudal damaged). Type. The other example agrees. This species is most closely related to C. steindachneri Cope, but differs chiefly in the slightly larger eye and more slender body. The main character advanced for the nominal genus Chorimycterus, i.e., the presence of biserial mandibular teeth, is entirely fallacious, as both of the cotypes before me have but a single series in the mandible. It is therefore a synonym of Characidium. 22 326 PROCEEDINGS OF THE ACADEMY OF [June, Subgenus RHYTIODUS Kner. Scales small, 80 to 90 or more in a lateral series, Rhytiodus microlepis Kner. Head 4f ; depth 5f ; D. ii, 10, i; A. ii, 8, i; scales 77 in a lateral series to base of caudal, and 8 more on latter; 12 scales obliquely back from origin of dorsal to lateral line; 9 scales obliquely up posteriorly from origin of ventral to lateral line; 8 scales from origin of ventral in a similar way; about 23? (squamation damaged) scales before dorsal; snout 2f in head; eye 5; interorbital space 2; pectoral H; ventral 1^; least depth of caudal peduncle 2f. Small short slender weak gill- rakers developed on outer edge of first arch. Pebas, Equador. Prof. J. Orton. Coll. 1873-77. Prof. E. D. Cope. GARMANINA subgen. nov. Type Rhytiodus argenteo-fuscus Kjier. Restricted to those species of Rhytiodus Kner with large scales, about 50 to 60 in a lateral series. (Named for Prof. Samuel Garman, of Cambridge, Mass., author of many excellent contributions to Ichthyology.) Ehytiodus argenteo-fuscus Kner. Schizodon Sagittarius Cope, Proc. Amer. Philos. Soc. Phila., XVII, 1877-78 (May 17, 1878), p. 689. No. 21,474, A. N. S. P., type. Peruvian Amazon. Prof. J. Orton. Coll. of 1877. Head 4f ; depth 5*-; D. in, 10; A. iii, 8; scales 47 in lateral hne to base of caudal, and 5 more on latter; 7 scales obliquely back from origin of dorsal to lateral line; 4 scales obliquely back from origin of adipose fin to lateral line; 6 scales obliquely up from origin of anal to lateral line; about 16 scales before dorsal; width of head 1|^ in its length; depth of head over middle of orbit 2f ; least depth of caudal peduncle 3; pectoral 1^; ventral H; depressed dorsal 4|; fourth anal ray If ; snout 2| in head measured from tip of upper jaw ; eye about 4^ ; interorbital space 2^. Body tapering evenly back from opposite origins of dorsal and ventral. Predorsal region with an obtuse median keel. Postdorsal region flattened. Preventral region rounded, except region right at bases and just before ventrals, which is flattened. Post- ventral region apparently rather rounded. Least depth of caudal peduncle about half its length. Head depressed above and below anteriorly, and with convex surface. Snout broad, rounded when viewed above. Width of mouth about | of orbit. Teeth broadly expanded or compressed. Maxillary reaching about to posterior nos- tril. Anterior nostril in a short tube. Interorbital space broadly convex. Gill-opening extending forward about opposite last fourth in 1906.] NATURAL SCIENCES OF PHILADELPHIA. 327 head. Rakers short, not numerous and weak. Filaments equal orbit. Isthmus broad and convex. Scales non-striate. Ventral scaly flap f of fin. Pectoral reaching f of space to ventral, and ventral -f- to anal. Vent close in front of anal. Color in alcohol brownish, upper f of body darker and sharply demarcated from lower or whitish surface, and former color extending well below lateral line. Dorsal and caudal brownish, especially middle of each lobe of latter. Pectoral, ventral and anal pale brownish, especially latter. Line of demarcation on sides pronounced by a longitudinal lateral band extending from tip of snout to base of caudal. Iris brownish. Length 6f inches. Prof. Garman's view that this is the young of R. argenteo-fuscus Kner is fully established, as I find that the mandible has 8 teeth rather than 6 as stated by Cope. Leporellus vittatus (Valenciennes). Leporinus vittatus Cope, Proe. Amer. Philos. Soc. Phila., XVII, 1877-78 (May 17, 1878), p. 690. Peruvian Amazon. Prof. J. Orton. Coll. of 1877. Head 3f ; depth 3|; D. iii, 10, i; A. iii, 8; P. ii, 16; V. ii, 8; scales 37? (squamation injured) in lateral line to base of caudal, and 5? more on latter ; 5 scales obliquely back from origin of dorsal to lateral line; 4 scales obliquely up posteriorly from origin of ventral to lateral line; 14? scales (squamation injured) before dorsal; snout 2^ in head; eye 4; interorbital space 2^; least depth of caudal peduncle 3. Gill- rakers short weak fleshy processes. Color in alcohol with back sharply defined from that of band running along lateral line, which is also well defined and dark. A narrow line on flank, well below lateral line, extending from base of pectoral to origin of anal. Top of head with indistinct brownish spots. Dorsal blackish with a broad transverse whitish band below middle. Length 2f inches. Castelnau figures Leporinus vittatus Valenciennes, which will be seen to differ at once in the coloration. It shows each scale of the back marked with a single black spot and the dorsal crossed by a blackish transverse band a little above its middle. The side of the head is spotted, and there are also no traces on the trunk of the dark contrasted color-pattern which my example now shows. Kner's figure of Leporinus pidus agrees better, and though it differs according to the figm-e in having a pale dorsal marked above and in front with a dark blotch, another or basal one is mentioned in the description. The figure also shows no trace of the lower dusky line seen on my example, which extends from the pectoral to the anal. Dr. Steindachner records a large example from Cauca which had both dorsal and anal fins marked with blackish bands. 328 PROCEEDINGS OF THE ACADEMY OF [June, Leporinus fasciatuB (Bloch). Head 3i; depth 2|; D. in, 10, i; A. iii, 8, i; scales 34 in lateral line to base of caudal, and 5 more on latter; 6 scales obliquely back from origin of dorsal to lateral line; 5 scales obliquely up from origin of ventral to lateral line ; 5 scales obliquely forward from origin of anal to lateral line; 12 scales before dorsal; snout 2^ in head; eye 4|-; inter- orbital space 2f; pectoral If ; ventral If; least depth of caudal peduncle 2^. Middle of orbit a little anterior in head. Rakers 5 + 8, denticle- like, compressed and weak. Predorsal and postdorsal regions rounded. Very slight lateral keel on each side of postventral region, and also one on each side of preventral region. Median line of both these areas also with a slight keel, that of latter most pronounced. In coloration it agrees largely with Bloch's figure. The occipital band is broader. First band on trunk forking above, second inclined to base of pectoral, sixth also forked above and extending towards origin of anal, seventh from base of adipose fin to bases of last anal rays, and last or ninth in form of large blotch at base of caudal. An indistinct brownish bar about opposite middle of pectoral and a blotch near its tip or above base of ventral. Traces of two transverse caudal bands. Length 5^ inches (caudal damaged). Rio Parahyba, Brazil. Museum of Com- parative Zoology, Massachusetts. One example. Leporinus friderici (Bloch). Leporinus frederici Cope, Proc. Amer. Philos. Soc. Phila., XI, 1869-70 (August 19, 1870), p. 566. Pebas, Eastern Peru. John Hauxwell. Cope, I.e., XVII, 1877-78 (May 17, 1878), p. 690. Peruvian Amazon. Prof. J. Orton. Head 3^ to 3^; depth 3^ to 3f ; D. ii or in, 10, i; A. in, 8, i; scales 33 to 35 in lateral line to base of caudal, and usually 4 more on latter; 11 to 13 scales before dorsal; 5 scales in an oblique series back from origin of dorsal to lateral line; 5 scales obliquely forward from origin of ventral to lateral line ; 5 scales obliquely forward from origin of anal to lateral line; snout 2f to 3 in head; eye 3f to 4f ; interorbital space 2 to 2^. Gill-rakers slender short denticles usually a little bent distally, and about 20? on first arch. Color in alcohol very dark, base of each scale darker than other portion, and line of demarcation very distinct, so that an imbricated appearance is assumed. In smaller examples pectoral approaches nearer ventral than in adults. Total length of 7 examples 3^ to 9 inches. Peruvian Amazon. Prof. J. Orton. Coll. of 1877. Two examples from Pebas. John Hauxwell. Both show about same number of scales as preceding. The smaller one has the dark lateral blotches more or less confluent posteriorly, as described by Dr. Giinther. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 329 The above specific name is restored in the original, for in all cases Bloch spelled it exactly as above. Leporinus multifasciatus Cope. Fig. 19. Proc, Amer. Philos. Soc. Phila., XVII, 1877-78 (May 17, 1878), p. 690. Peruvian Amazon. Prof. J. Orton. Leporinus megalepis Cope, I.e., XI, 1869-70 (May 19, 1870), p. 566. Para. De Schulte Buckow. Cope, Proc. Acad. Nat. Sci. Pliila., 1871 (1872), p. 259. Ambyiacu River, Equador. John Hauxwell. The cotypes of L. multifasciatus, Nos. 21,448 (type) to 21,450, A. N. S. P., are in very poor condition, though they appear to be identical with the other material. The largest shows: Head 3^; depth 3f ; D. II, 10; A. IV, 8; width of head 2 in its length; depth of head over middle of orbit 2; snout 3 in head from tip of upper jaw; eye 3^; interorbital space 2^; least depth of caudal peduncle 2^. Though dark lateral blotches are not present they may have faded. Length of tj^je 2^ inches. Fig. 19. — Leporinus midtifasciatus Cope. The example from Para has very indistinct bands on the back and a rather long pectoral. The largest of the Hauxwell examples shows : Head 3^; depth 3 ; D. Ill, 10, i; A. Ill, 8, i; scales 33 in lateral Hne to base of caudal, and 4 more on latter; 5 scales obliquely back from origin of dorsal to lateral line; 4 scales obliquely up from origin of anal to lateral line; 10 scales before dorsal; width of head lyf in its length; third dorsal ray 1^; third anal ray If; pectoral 1% ; ventral H; least depth of caudal peduncle 2f ' 330 PROCEEDINGS OF THE ACADEMY OF [June, snout 2f in head from tip of upper jaw; eye 3f ; maxillary 4f; inter- orbital space 2^ ; upper caudal lobe nearly 3 in head and trunk. Pre- dorsal region rounded at first and posteriorly with a median keel ex- tending to dorsal fin. Postdorsal region rounded, also preventral region. Postventral region with a median obscure keel. Head broad, depressed in front, rounded below and sides somewhat compressed. Snout when viewed above broad and triangular with rounded tip. Width of mouth about | of orbit. Maxillary about half way in snout. lips fleshy. Teeth 8 in each jaw, anterior largest, and 2 median mandibulars especially conspicuous. Tongue rather far back, narrow, and not free. Interorbital space broadly convex. Gill-opening ex- tending forward about first third in head. Rakers 10 + 14?, short, lanceolate, and longest about half of longest filaments which are f of orbit. Vent close in front of anal. Color in alcohol faded brownish, a little darker on back and upper surface. About 14 indistinct bars of deeper brownish than body-color across back, and each one also narrower than pale interspace. In some cases they become faded altogether on side, and frequently indistinct traces occur below. Three blotches of deep brownish, larger than eye, along middle of side. First blotch midway in length of trunk, and second midway between it and third, which is on base of caudal. Dorsal, caudal and pectoral pale, and anal and ventral dusky or deep brown. Iris brown. Teeth warm brownish marginally. Length 4|- inches. Identified with L. megalepis Gunther, but that species is said to have the body with large blackish spots arranged in 2 or 3 series. Leporinus holostictus Cope. Fig. 20. Cope, Proc. Amer. Philos. Soc. Phila., XVII, 1877-78 (Mayll7, 1878), p. 690. Nos. 21,467 (type) and 21,468, A. N. S. P., cotj^jes. Peru\-ian Amazon. Prof. J. Orton. Coll. of 1877. Width of head 2^ in its length; interorbital space 2^. Predorsal, postdorsal and preventral regions rounded. Postventral region with an obsolete median keel, and another on each side. Snout broad and triangular seen from above. Lips fleshy. Teeth 8 in each jaw, anterior ones largest, and 2 median mandibulars conspicuous. Tongue rather far back, narrow and not free. Interorbital space broadly convex. Gill-opening extending forward about last foiu-th in length of head. Rakers 5 + 11?, short, lanceolate, and longest about f of longest filaments, which are f of orbit. Scales non-striate. Color in alcohol brownish, and lower sm-face paler or whitish. Body with 10 broad transverse bands as wide as interspaces. First includes or extends over snout, second over interorbital space, third on occiput, 1906.] NATURAL SCIENCES OF PHILADELPHIA. 331 fourth on middle of predorsal region, fifth just before dorsal, sixth vertically across dorsal so that its posterior margin is even with last ray and then down till behind ventral, seventh on postdorsal region anteriorly, eighth on postdorsal region posteriorly till down just in front of anal, ninth behind adipose fin and continued transversely Fig. 20. — Leporinus holostidus Cope. across anal, and tenth at base of caudal. Fins otherwise all more or less unmarked, except posterior margin of caudal which is dusky. Iris brownish. Length 4f inches (caudal damaged). T}^e. The other example shows the depth as 3^. ABRAMITES gen.'nov. Type Leporinus hypselonotus Giinther. Closely related to Leporinus Agassiz, but distinguished by the larger anal basis. {Ahramites, from Ahramis, Bresma and Prasmus, old names applied to Brama hrama or the common European bream. The reference is to the superficial appearance.) Abramites hypselonotus (Gunther). Leporinus hypselonotus Cope, Proc. Amer. Philos. Sec. Pbila., XVII, 18 (May 17 1878), p. 690. Permaan Amazon. Pre " ' '^ 1877. Head 3f ; depth 2f ; D. iii, 10; A. iii, 12, i; scales 34 in lateral fine to base of caudal, and 4 more on latter; snout 3{ in head from tip of upper jaw; eye 3; interorbital space 2f. Length 2|| inches. Dr. Giinther's excellent figure largely agrees, though my example 78 Orton. CoU. of 332 PROCEEDINGS OF THE ACADEMY OF [June, shows traces of an interorbital band and a nuchal one extending down just behind opercle. TETRAGONOPTERIN.E. Cheirodon monodon (Cope). Chirodo7i monodon Cope, Proc. Amer. Philos. Soc. Phila., XXXIII, 1894 (January 5), p. 91, PI. 6, fig. 9. Nos. 21,561 (type) and 21,562, A. N. S. P., cotypes. Brazilian province of Rio Grande do Sul. H. H. Smith. Edge of back rounded, except a httle before spinous dorsal and upper surface of caudal peduncle where it is a little trenchant. Belly rounded. Postventral region and lower surface of caudal peduncle a little trench- ant. Each ramus of mandible elevated a little inside mouth. Tongue rather large, flattened, moderately broad and free in front. Inter- orbital space convex. Gill-opening extending forward nearly opposite anterior margin of orbit. Rakers 12 + 12 on first arch, lanceolate, and longest a little shorter than filaments, which are a little longer than pupil. Isthmus narrowly triangular. Vent close in front of anal. Color in alcohol pale brownish, back a little darker than lower surface, and body everywhere more or less silvered. A pale or dull leaden- brassy streak from above eye back to caudal. Fins all pale or immacu- late brownish. Iris dull silvery. Length 1|| inches. Type. Fig. 21. — Cheirodon pulchcr (Steindachner). Cheirodon pulcher (Steindachner). Fig. 21. Tetragonopterus diaphanus Cope, Proc 1877-78 (May 17, 1878), p. 691. (Part.) Amer. Pliilos. Soc. Phila., XVII, Peruvian Amazon. Prof. J. Orton. 1906. NATURAL SCIENCES OF PHILADELPHIA. 333 Edges of body but little trenchant, or only upper and lower surfaces of caudal peduncle and postventral region. Rami of mandible well elevated inside mouth. Rakers 10 + 14 on first arch, lanceolate, and longest about length of filaments which are f of orbit. Scales non- striate. Color in alcohol brow^nish, back darker and lower surface paler, washed everywhere with more or less silvery. A dull leaden streak, a little high, from below dorsal to caudal and then continued out on its middle rays as dusky. Fins otherwise all dull brownish. Iris dull brassy. Length If inches. Two examples, wTongly identified by Cope with Astijanax diaphanus. Aphyocharax pusillus Giinther. Fig. 22. Cope, Proc. Acad. Nat. Sci. Phila., 1871 (1872), p. 260. Amb\aacu River, Eastern Equador. John Hauxwell.— — ^Cope, Proc. Amer. Pliilos. Soc. Phila., XVII, 1877-78 (May 17, 1878), p. 689. Permian Amazon. Prof. J. Orton. Interorbital space 2|- in head. Snout convex. Rami of mandible but little elevated inside of mouth. Tongue a little elongate, fleshy and free in front. Interorbital space broad and a little convex. Gill- openings extending forward till nearly opposite front margin of pupil. ^^ Fig. 22. — Aphyocliarax pusillus Giinther. Rakers 7 + 8, short, pointed, and much less than filaments. Isthnms narrow and with median groove. Each scale with 2 or 3 striae. Color in alcohol brown, sides paler, and body more or less brassy. Fins brownish, median caudal rays dusky. Iris brownish. Length 2y^6 inches. Peruvian Amazon. Orton. Coll. of 1873. 334 PROCEEDINGS OF THE ACADEMY OF [June, Three examples from the Ambyiacu in the Hauxwell Coll. Head 3f to 3|; depth 3^ to 4; D. ii, 9; A. in or iv, 16, i to 18, i; scales 32 to 34 in lateral line (squamation mostly injured) to base of caudal, and 2 more on latter; total length 2 to 2||- inches. Aphyocharax filigerus Cope. Fig. 23. Proc. Amer. Philos. Soc. Phila., XI, 1869-70 (August 19, 1870), p. 564. Nos. 8,059 (type) and 8,073, A. N. S. P., cotypes. Pebas, Eastern Equador. John Hauxwell. Interorbital space 3 in head. Edges of trunk apparently more or less rounded. Rami of mandible hardly elevated inside mouth. Tongue a little long, narrow and free in front. Interorbital space elevated convexly. Gill-openings carried forward about first f of Fig. 23. — Aphyocharax filigerus Cope. orbit. Rakers 7 + 10, lanceolate, and apparently longer than longest filaments or a little longer than half of orbit. Isthmus narrow. Color in alcohol faded dull brownish, and fins pale, though median caudal rays a little paler than rest of fin. Iris brownish. Length 2^ inches (caudal damaged). Type. Taken from the stomach of a siluroid. Diapoma speculiferum Cope. Amer. Nat., XXVIII, 1894, p. 67. No. 21,580, A. N. S. P., type. Upper waters of the Jacuhy River in Brazihan State of Rio Grande do Sul. H. H. Smith. Cope. Proc. Amer. PWlos. Soc. Pliila., XXXIII, 1894 (January 5), p. 92, PI. 5, fig. 4, from t^^pe. Edge of back rounded. Upper and lower edges of caudal peduncle trenchant. Chest with a median rounded ridge. Postventral region 1906.] NATURAL SCIENCES OF PHILADELPHIA. 335 trenchant. Snout convex. Tongue small, a little back, narrow, and a little free. Interorbital space narrowly convex. Gill-opening car- ried forward nearly opposite anterior margin of orbit. Rakers 6 + 12?, slender, fine, longest a little longer than filaments and also present on lower portion of external arch. Isthmus rather broadly triangular. Scales striate. Vent close in front of anal. Color in alcohol more or less brownish, back more brown than lower surface and body ^-ith silvery wash. A leaden lateral streak from head opposite eye straight to base of caudal. Fins plain pale brownish. Eye brassy. Length 1^^ inches. Cope was entirely mistaken when stating that the inferior limb of the external branchial arches was without rakers. Hemigrammus sclimardae (Steindachner). Head 3; depth 2f ; D. iii, 9; A. iv, 21?; about 30? scales in a lateral covnit from gill-opening to caudal, including those on latter; snout 4^ in head, measured from tip of upper jaw; eye 2^; maxillary 2f ; inter- orbital space 3. Color in alcohol brownish, scales fallen leaving a narrow silvery lateral band extending from shoulder to base of caudal. No blackish blotch at base of caudal, which while a little darker than rest of fin is not darker than peduncle. These conditions may all be due to the preservative. Length If inches. A single small example in poor condition from the Peruvian Amazon. Coll. Prof. J. Orton. Hemigrammus interruptus (Lutken). Hemigramvivs luetkenii Cope, Proc. Amer. Philos. Soc. Phila., XXXIII, 1894 (January 5), p. 91. In the Jacuhy (Brazilian pro\'ince of Rio Grande do Sul). H. H. Smith. The large series from the Jacuhy, Brazilian province of Rio Grande do Sul, and collected by H. H. Smith, shows : Head 3 to 3|; depth 2J to 2f ; D. Ill, 9; A. IV or v, 18, i to 22, i, usually with 21, occasionally 22, sometimes 18, and rarely 20; scales 30 to 33 in lateral line to base of caudal, 2 or 3 more on latter, former count usually 33, frequently 32 or 31, and rarely 30; 11 or 12 scales in a transverse series obliquely back from origin of dorsal; 12 or 13 scales before dorsal; eye 2h to 3 in head, measured from tip of upper jaw; length of specimens If to 3 inches. Gill-rakers about 8 + 14. Hemigrammus robustulus Cope. Fig. 24. Proc. Amer. Pliilos. Soc. Pliila., XI, 1869-70 (August 17, 1870), p. 561- Nos. 8,037 (type) to 8,052, A. N. S. P., cot}T)es. Pebas, Equador. John Hauxwell. Cope, I.e., XVII, 1877-78 (May 17, 1878), p. 690. Peruvian Amazon. Prof. J. Orton. Body strongly compressed. Rami of mandible but little elevated inside mouth. Tongue rather flat, rounded, compressed and a little 336 PEOCEEDINGS OF THE ACADEMY OF [June, free. Interorbital space rather broad and a little convex. Gill-open- ing extending forward to anterior margin of pupil. Rakers about a dozen? on lower outer limb of first arch. Filaments short. Isthmus narrow. Scales striate. Color in alcohol pale or dull brown, abdominal region and lower surface of head paler. Flanks also tinted with a livid grayish or dull slaty hue. A streak of slaty, diffuse and indis- tinct at first, extending along side of caudal peduncle and then con- tinued out on median caudal rays as a blackish band. A dusky diffuse humeral blotch about size of orbit. Fins otherwise mostly uniform dull brownish. Muzzle deep brownish and front of mandible same. Hemigrammus rubustulus Cope. Spots or specks on cheek indistinct. Iris dusky-coppery. I^ength 2 inches (caudal damaged). Type. The other cotypes range as follows : Head 2f to 3^ ; depth 2 to 2f ; D. II, 9; A. Ill, 23 to iii, 29, i; scales 27 to 32 (squamation mostly injured) ; total length H to 2 inches. Other examples in the Hauxwell Coll. Astyanax lepidurus (Kner). Tetragonopterus lepidurus Cope, Proc. Amer. Pliilos. Soc. Pliila., XI, 1869-70 (August 19, 1870), p. 566. Para. Mr. De Schulte Buckow. Head 3^; depth 3|?; D. iii, 9; A. iii, 21 ?; scales (according to pockets) 1906. NATURAL SCIENCES OF PHILADELPHIA. 337 about 33? in a lateral series to base of caudal; 9? scales between dorsal and ventral; snout 4 in head, measured from tip of upper jaw; eye 2f ; maxillary 2^; interorbital space 3|; least depth of caudal peduncle 3. Maxillary toothless and extending posteriorly a little beyond front rim of orbit, not to pupil. Rakers 10 + 16?, long, slender, and longest nearly as long as longest filaments. Scales large, mostl}^ all fallen, and in alcohol leaving a narrow silvery band broadening out on posterior ■| of its course till about equal to diameter of pupil. No evidence of humeral or caudal blotches. Median caudal rays dusky, also tips of upper and lower lobes. Length If inches. This example was received from Cope, and is most likely the one referred to above. Astyanax ipanquianus (Cope). Fig. 25. Tetragonoptenis ipanquianus Cope, Proc. Amer. Philos. Soc. Phila., XVII, 1S77-78 (May 4, 1878). p. 44. Nos. 21,114 (tjiae) and 21,115, A. N. S. P., cotypes. The upper waters of the Urubamba, one of the sources of the Ucayale. Prof. J. Orton. Coll. of 1876-77. Cope, I.e., p. 692. Uru- bamba River; elevation 11,500 feet. Prof. J. Orton. Fig. 25. — Astyanax ipanquianus (Cope). Body compressed, and edges rounded. Snout convex. Rami of mandible not elevated inside mouth. Lips thin. Tongue fleshy, broad, rounded in front and hardly free from floor of mouth. Inter- orbital space convex. Gill-opening extending forward about opposite anterior nostril. Rakers 9 + 11, lanceolate, short, and longest a little less than half length of longest filaments. Filaments f an eye- 338 PROCEEDINGS OF THE ACADEMY OF [June, diameter. Isthmus long and narrowly triangular. Each scale with a number of stria?. Color in alcohol dull brassy-brown with silvery reflections, back slightly more brownish with a leaden tint. A humeral blotch of grayish about equal to orbit in size. Top of head brownish. Fins pale brownish, dorsal and caudal a trifle darker than others. Iris dull brassy-brown. Length (caudal slightly damaged) 5 inches. Type. The other examples, including those later reported by Cope, show: Head B^ to 4; depth 3 to 3f ; D. ii, 8; A. in or iv, 23, i to 27, i; scales about 50? (squamation damaged) to 55 in lateral line to base of caudal, and 2 or 3 on latter; 23 to 25 scales before dorsal; 11 scales in an oblique series back from origin of dorsal to lateral line ; 8 or 9 scales in a vertical series between lateral line and origin of anal ; eye 3j to 4 in head ; total length 2^ to 4f inches. Astyanax phcenioopterus (Cope). Fig. 26. Tetragonopterus phoenicopterus Cope, Proc. Acad. Nat. Sci. Phila., 1871 (1872), p. 261. Nos. 8,093 (type) to 8,128, A. N. S. P., cotypes. Ambyiacu River, Equador. John Hauxwell. Gill-opening extending forward about opposite anterior margin of [styanax phcenicupterus (Cope). pupil. Rakers small and inconspicuous, about 7? on lower part of first arch. Filaments short. Isthmus compressed and narrow. Color in alcohol dull brown, more or less silvery. Side of head silvery. 1906. NATURAL SCIENCES OF PHILADELPHIA. 339 Lower surface of body pale brown, and peritoneum showing through pale. A dusky humeral and caudal spot. Fins and iris pale brownish. Length (caudal damaged) 2^ inches. Type, The other cotypes vary as follows : Head 2^ to 4, usually a little less than 4; depth 3 to 3|; D. iii, 8, rarely iii, 7; A. iv, 21 to 26, and fre- quently I, branched rays frequently 25; scales 30 to 35 (squamation damaged); total length H to 2\ inches. Astyanax diaphanus (Cope). Fig. 27. Tefragonoptenis diaphanus Cope, Proo. Amer. PhiJos. Soc. Phila., XVII, 1S77-7S (May 17, 1878), p. 691. Nos. 21,216 (type) to 21,218, A. N. S. P., cotypes. Peruvian Amazon. Prof. J. Orton. Coll. of 1874. Predorsal region with an obsolete median keel, and postdorsal region rounded. Preventral region rounded and postventral region trench- ant. Rakers 7 + 10, slender, lanceolate, and longest a trifle more than longest filaments which are about equal to diameter of pupil. Isthmus narrowly triangular. Color in alcohol dull brownish, more or less brassy, and especially broad band from eye to base of caudal. Fins all dull brownish. Iris brownish-yellow. Length 2y^g inches. Type. Fig. 27. — Astyanax diaphanus (Cope). The other examples show: Head 3f to 4; depth 3^ to 3^; D. ii, 8; A. Ill, 17, I to 22, i; scales 33 in lateral fine to base of caudal and 3 more on latter; usually 4 scales in a vertical series between origin of anal and lateral line; eye 2| to 2f in head; length about 2 inches with damaged caudals. 340 PROCEEDINGS OF THE ACADEMY OF [June, Aatyanax hauxwellianus (Cope). Fig. 28. Tetragonopterus hauxwellianus Cope, Proc. Amer. Philos. Soc. Phila., XI, 1869-70 (August 19, 1870), p. 560. Nos. 8,137 (type) to 8,142, A. N. S. P., cotypes. Pebas, Equador. John Hauxwell. Cope, I.e., XVII, 1877-78 (May 17, 1878), p. 691. Peruvian Amazon. Prof. J. Orton. Coll. of 1873. Gill-opening extending forward about opposite front of eye. Rakers 10? on lower part of first arch. Filaments a little longer than longest rakers. Isthmus compressed and narrow. Color in alcohol faded to dark brown, lower surface paler. Fins all brownish. A dusky blotch at base of caudal and a similar shade in humeral region. Iris brassy. Length 2^^ inches (caudal damaged). Type. Fig. 2S. — Astyanax hauxwellianus (Cope). flThe other cotypes show: Head 3h to 3§; depth 2 to 2f ; D. usually III, 9, sometimes iii, 8 or iii, 10; A. usually iv, 40?, sometimes iv, 41?; total length 1|| to 2-^-^ inches. The distinct lateral silvery band mentioned by Cope is now hardly e\ddent. Also 59 examples from the Orton Coll. of 1873, Peruvian Amazon, and an example from Robert Perkins, from between the mouth of the Rio Negro and the Peruvian Amazon. 1906. NATURAL SCIENCES OF PHILADELPHIA. 341 Astyanax pectinatus (Cope). Fig. 29. Tetragonopf.erus pectinatus Cope, Proc. Amer. Philos. Soc. Phila., XI, 1869-70 (August 19, 1870), p. 560. No. 8,090, A. N. S. P., type. Pebas, Equador. John HauxweU. Gill-opening extending forward about anterior third of orbit. Rakers about 6 + 9, slender, lanceolate, and less than filaments which are about half of orbit. Isthmus narrow and compressed. Scales Fig. 29. — Astyanax pectinatus (Cope). without striae. Color in alcohol with fins and body nearly uniform brownish, former paler. Body more or less washed with silvery. Iris brassy-brown. Traces of a humeral blotch, though indistinct. Length If inches (caudal damaged). Astyanax longior (Cope). Fig. 30. Tefragonopterus longior Cope, Proc. Amer. Philos. Soc. Phila., XVII, 1877-78 (May 17, 1878), p. 691. Nos. 21,222 (type) and 21,223, A. N. S. P., cotypes. Moyabaraba. Prof. J. Orton. Coll. 1874. Predorsal region with a median obsolete keel and postventral region apparently trenchant, edges of body otherwise rounded. Gill-opening forward till about opposite front of eye. Rakers 9 + 11, lanceolate, and longest about f of filaments, which are f of orbit. Isthmus nar- rowly triangular. Each scale with several radiating striae. Color in alcohol faded dull brownish, and back darker than silvered sides. A broad silver line, separating color of back from that of lower surface, extending from shoulder to base of caudal. Humeral blotch rounded 23 342 PROCEEDINGS OF THE ACADEMY OF [June, and a little longer than pupil. Fins all brownish. Iris brownish with rosy tint. Length 3f inches (caudal damaged). Type. Cope's description would give the depth of the body 4^^ in its length without the caudal, wiiile in the above example it is but 2|-. The head is said to be 4^, while the above example shows it about 3f . He gives Fig. 30. — Astyanax longior (Cope). the eye 3h in the head, though the interorbital width is much less than Ij-^Q in the eye, in fact but little less than the orbit itself. Both examples have Cope's original label. Astyanax maximus (Steindachner). Tetragnnoptems sp. indet. Cope, Proc. Amer. Philos. Soc. Phila., XVII, 1877-78 (May 17, 1878), p. 691. Peruvian Amazon. Prof. J. Orton. Coll. 1873. Head 3f to 4; depth 2f to 2^; D. i, 9; A. iv, 28, i to 29, i; scales 34 to 35 in lateral line to base of caudal and 2? more on latter; 7 scales in an oblique series back between origin of dorsal and lateral line; 7 scales from base of ventral to lateral line; 7 or 8 scales in a vertical series between origin of anal and lateral hne; 17 scales before dorsal; pectoral a trifle less than head; ventral 1^ to 1^; snout 3^ to 4 in head, from tip of upper jaw; eye about 3; interorbital space 2f ; length of larger example with damaged caudal 4f inches. Rakers 10 + 16, lanceolate, slender, longest but little shorter than filaments which are about § 1906. NATURAL SCIENCES OF PHILADELPHIA. 343 diameter of eye. Pectoral reaches well beyond base of anal. Color in alcohol nearly plain brassy-brown, back a little darker, and caudal with a brownish streak from base out to ends of median rays. No evidence of a humeral blotch. Astyanax cauoanus (Steindachner). Head 3f ; depth 2f ; D. ii, 9; A. iv, 42; scales (according to pockets) about 38 in lateral line to base of caudal; pectoral about 1^ in head; ventral If; snout about 4f in head, measiu-ed from tip of upper jaw; eye 3 ; maxillary 3 ; interorbital space 2f ; least depth of caudal peduncle about 3. A few small teeth at base of maxillary inside. Rakers about 7 + 12?, and longest about f of longest filaments. In alcohol brown- ish, fins paler than body-color, and a silvery band, widest on supra- costal region, from shoulder to base of caudal where it terminates in a brownish spot. A brownish blotch on shoulder at origin of silvered lateral streak. Iris silvered brownish. Length 2^^ inches. One example from Paramaribo. Dr. Hering. Astyanax bartlettii (Gunther). Fig. 3L Tetragonopterus bartlettii Cope, Proc. Acad. Nat. Sci. Phila., 1871 (1872), p. 260. Ambyiacu River, Equador. John Hauxwell. Predorsal region with a median keel, rather ol^tuse, and postventral Fig. 31. — Astyanax bartlettii (Gunther) 344 PROCEEDINGS OF THE ACADEMY OF [June, region but little trenchant, edges of body otherwise rounded. Gill- opening extending forward a trifle before front rim of pupil. Rakers about 10 + 14, slender, lanceolate, longest about | of longest fila- ments and apparently rather weak. Filaments about f of orbit. A few strise on each scale. Color in alcohol nearly uniform dull brownish, back a little darker and line of demarcation about level with upper margin of orbit distinct. A grayish elongate humeral blotch, most likely much faded, and in length nearly equal to diameter of eye. Base of caudal brownish and this color also extending out on median caudal rays. Fins otherwise all pale brownish. Iris brassy-brown. Length 4\ inches. Two examples. Astyanax finitimus (Vaillant and Pellegrin). Fig. 32. Predorsal region with a low median keel and one also on postventral, other edges of body rounded. Gill-opening not quite reaching front of orbit. Rakers 8 + 13, lanceolate, longest about | of filaments and Fig. 32. — Astyanax finitimus (Vaillant and Pellegrin). latter about f of orbit. • Isthmus a little broadly triangular. Each scale with several radiating striae. Color in alcohol brownish, back deeper and this color sharply separated from that of lower surfacej^y^a 1906.] NATURAL SCIENCES OF PHILADELPHIA. 345 broad silvery band from shoulder to caudal, where it becomes deep brown or dusky and extends out on median caudal rays. A dusky humeral blotch a little larger than pupil. Lower side and under por- tions of body silvered. Fins pale brownish, dorsal and caudal a trifle dusky. Iris brassy. Length 2i| inches (caudal damaged). Possibly from some part of Central America, most likely Nicaragua? Dr. J. F. Bransford. Five examples. The others show: Head 3i to 3|; depth 2f to 3^; D. in, 9; A. in or IV, 24? to 26, i; scales 36 in lateral hne to base of caudal, and 2 more on latter; 8 scales obliquely back from origin of dorsal to lateral line; 8 scales in a vertical series between lateral line and origin of anal ; eye 2f to 24 in head measured from tip of upper jaw; total length of speci- mens lyV to 3^ inches (damaged caudals). The young examples are more elongate and have larger eyes in proportion, though at all ages the caudal and humeral blotches are evident. Although the name Tetragonoptenis finitimus was originally proposed by Bocourt {Ann. Sci. Nat. Paris, Zool, 5, IX, 1868, p. 62), his account is entirely too brief and imperfect to permit identification. Prof. Vaillant and Dr. Pellegrin next describe examples from Guatemala which they identify with this species and w^hich seem to agree with my own specimens. Astyanax mexioantis (Filippi). Tetragonopterus mexicanus Jordan and Snyder, Bull. U. S. Fish Comm., XIX, 1899 (1901), p. 125. Rio Ixtla at Puente de Ixtla, Morelos. Profs. D. S. Jordan and J. O. Snvder. Tetragonopterus streetsii Cope, Proc. Acad. Nat. Sci. Phila., 1871 (November 28), p. 217. From the headwaters of the Coatzacalcos River among the Cordilleras. Dr. T. Hale Streets. Tetragonopterus argentatus Cope, Proc. Amer. Philos. Soc. Phila., XXII, 1885 (December 10, 1884), p. 1G8. In the city of Monterey. Prof. E. D. Cope. Jordan and Snyder, I.e., Rio Verde near Rascon. Fowler, Proc. Acad. Nat. Sci. Phila., 1904 (April 7), p. 248. Del Rio and De\al's River, Texas. Dr. H. A. Pilsbry. A very large series of examples of most all ages from Rio Ixtla at Puente de Ixtla, Morelos, and Rio Verde near Rascon, Mexico. These collected and presented by Profs. Jordan and Snyder. Also a large series from Monterey, Mexico, from Prof. E. D. Cope, and others from the same evidently from Mexico. Others from Del Rio and Devil's river, Texas, from Dr. H. A. Pilsbry. All show the following: Head 3 to 4; depth 2^ to 3f ; D. in, 9, though sometimes ii or in, 8, and rarely in, 10; A. ni to iv, 17, i to 24, i, usually 21 or 20, frequently 19 or 22, others with 23 and occasionally 24, though rarely 17 or 18; scales in lateral line to base of caudal 32 to 37. and 2 or 3,more on latter, usually 35, frequently 34 or 36, and rarely 32, 33 or 37 ; 7 or 8 scales in an oblique 346 PROCEEDINGS OF THE ACADEMY OF [J line » series back from origin of dorsal to lateral line; 6 scales from base of ventral in a vertical series to lateral line, sometimes 7, rarely 5, and same between origin of anal and lateral line; 13 to 19 scales before dorsal, usually 15 to 16, less frequently 14 or 17, and rarely 13, 18- or 19; eye 2^ to 3^ in head, measured from tip of upper jaw, larger measiu-ements in young; length of specimens ly^g- to 4f inches. Astyanax ceneus (Giinther) is a closely related species differing chiefly in the absence of maxillary teeth, according to Dr. Meek's account. The other characters are seen to be more or less correlated, as several of my examples of A. mexicanus which possess as many as 24 developed or branched anal rays (would be 26 or 27 evidently according to Dr. ]\Ieek's count) also have well-developed basal maxillary teeth. A number of examples, possibly cotypical of Tetragonopterus streetsii Cope, have uncertain data. They are included in the above measure- ments, and are Nos. 32,371 to 32,426, A. N. S. P. Astyanax fasciatus (Cuvier). Tetragonopterus fasciatus Cope, Proc. Amer. Philos. Soc. Phila., XI, 1869-70 (August 19, 1870), p. 566. Para. Mr. De Schulte Buckow. Head 3^; depth 2f ; D. iii, 9; A. iv, 26; scales 34 in lateral Hne to base of caudal, and 3 more on latter; 6 scales in an obliquely vertical series from origin of dorsal to lateral line; 6 scales in a nearly vertical series from origin of ventral to lateral line; 14 scales before dorsal; snout 4 in head, measured from tip of snout ; eye 3 ; interorbital space 2§ ; least depth of caudal peduncle 2J; pectoral 1^ in head, from tip of mandible; ventral If. Several minute maxillary teeth. Rakers about 9 + 15 on first arch, those on ceratobranchial largest, though shorter than filaments. Each scale of body with several striae. Pos- terior distal extremity of maxillary extending back till opposite pupil. Humeral and caudal spots pale. Rather broad pale or silvery band on side of trunk evident. Iris brassy-brown and side of head silvery. Length 2|- inches. One example. I accept the above name, as Dr. Giinther pointed out the inconsist- encies of Cuvier's figure and description, also when compared with other accounts. The examples identified by Prof. Ulrey {Ann. N. Y. Acad. Sci., VIII, 1893-95, p. 284) as Tetragonopterus fasciatus (Cuvier), and credited to Dr. Eigenmann, with the developed anal rays 19 to 25, would seem to indicate a rather unusual range of variation. Astyanax eigenmanniorum (Cope). Tetragonopterus eigenmanniorum Cope, Proc. Amer. Philos. Soc. Phila., XXXIII, 1894 (January 5), p. 89, PI. 6, fig. 8. Nos. 21,598 (type) to 21,602, and 21,627 and 21,628, A. N. S. P., cotypes. BrazUian province of Rio Grande do Sul. H. H. Smith. Predorsal region with median keel. Apparently a low keel on each 1906.] NATURAL SCIENCES OF PHILADELPHIA. 347 side of preventral region. Postventral region trenchant and edges of body otherwise rounded. Gill-opening extending forward a little beyond front margin of pupil. Rakers 8 + 10? (damaged), slender, pointed, and shorter than filaments. Isthmus narrow and compressed- A few strige on each scale. Color in alcohol pale brownish, back darker than lower smi ace and traces of a pale or silvery band from eye to base of caudal. An indistinct brownish humeral blotch and another at base of caudal, also extending out on median rays of latter. Fins pale brownish. Eyes silvered. Length 2^^ inches (caudal damaged). Type. The other examples all show: Head 3^ to 3^; depth 2f to 2f ; D. iii. 9, rarely ii, 8; A. iv, 21, i or iv, 22, i, usually latter; scales 30 to 3-1 in lateral line to base of caudal and 2 or 3 on latter ; 6 scales obliquely back from origin of dorsal to lateral line; 5 or 6 scales from origin of ventral to lateral line in a vertical series; 13 to 15 scales before dorsal, usually 14; eye 2f to 2| in head ; total length of examples 2 to 2f inches. This form has been united with Astyanax fasciatus (Cuvier) by Prof. Ulrey. The above characters will show that it had at least best be regarded as distinct, if only provisionalh'. Further, it may be stated as differing principally from A. fasciatus in the fewer anal radii. All of the above characters are also constant, and it is possible to separate these specimens from those representing Astyanax laticeps, as Cope rightly determined. I shall therefore be obliged to remove the latter from the synonymy of A. fasciatus, where it has also been placed by Prof. Ulrey, as it differs principally in still fewer anal radii. Thus from an examination of the material before me I shall refuse to con- sider A. eigenmanniorum and A. laticeps either as variations of A. fasciatus or in fact either as the same species. Finally Cope's descrip- tion and figure of A. eigenmanniorum may hardly be styled excellent, much less that it leaves little doubt that the fish in question is simply a variation of T. fasciatus. Astyanax iheringii (Boulenger). Tetragonopterus pliodus Cope, Proc. Amer. Philos. Soc. Phila., XXXIII, 1894 (January 5), p. 90, PI. 5, fig. 5. Nos. 21,578 (type) and 21,579, A. N. S. P., cotypes. Brazilian pro\ance of Rio Grande do Sul. H. H. Smith. Predorsal region with median rounded ridge, and postventral trench- ant, edges of body otherwise rounded. Gill-opening reaching front of pupil. Rakers about 10 + 12, short, lanceolate, and longest about i of longest filaments which are about equal to ^ of vertical orbital diameter. Isthmus triangular and rather broad. Each scale with several radiating strise. Color in alcohol brownish, lower portion of 348 PROCEEDINGS OF THE ACADEMY OF [June, body, from level with upper edge of eye, silvery. A lateral silvered streak from eye to caudal. Traces of a faint humeral and caudal blotch. Dorsal and posterior edge of caudal brownish. Iris dull orange. Length 2|-f inches. Type of Tetragoiwpterus pliodus Cope. An example labelled Montevideo, Uruguay, received from Cope. Cope was evidently in error in stating that the humeral and caudal spots were wanting, as even now the former is present though indistinct, and the median caudal rays are darker than the rest of the caudal fin. There is also a dull brownish shade at the base of the caudal. Outer portions of dorsal, caudal and anal brownish, not exactly "finely specked with black" as stated by Dr. Boulenger. Astyanax laticeps (Cope). Telragonopterus laticeps Cope, Proc. Amer. Philos. Soc. Phila., XXXIII. 1894 (January 5), p. 89, PI. 6, fig. 7. Nos. 21,852 (type) to 21,863 and 21,743 to 21,747, A. N. S. P., cotjTjes. Brazilian pro^^nce of Rio Grande doSul. H.H.Smith. Predorsal region with a median keel and postdorsal with an obsolete median ridge with one on each side. Preventral region apparently rounded. Postventral region with median keel. Gill-opening ex- tending a little before front of pupil. Rakers 9 + 12, slender, pointed, and a little less than longest filaments. Isthmus narrow and com- pressed. Color in alcohol pale brownish, back darker than sides and lower surface, and a silvery band from shoulder to base of caudal, its width less than orbit. Body, especially lower surface and side of head, more or less silvered. A rounded dusky humeral blotch, and a dusky dash at base of caudal continued out on median rays of latter. Fins plain pale brownish. Iris brassy with a plain pale brownish ring. Length 2|- inches (caudal damaged). Type. The other examples show: Head 2f to 3^; depth 2^ to 3; D. iii, 9; A. IV, 18, I, rarely iv, 19, i; scales 30 to 35 in lateral line to base of caudal, and usually 3 or 4 more on latter; 14 to 16 scales before dorsal, usually 15; 6 to 8 scales in an oblique series back from origin of dorsal to lateral line, usually 7; 6 or 7 scales from origin of ventral to lateral line, usually 6; eye 3 to 3^ in head, from tip of upper jaw, usually but little over 3; total length Iff to 2f inches. This species has been united with Astyanax fasciatus (Cuvier) by Prof. Ulrey {Ann. N. Y. Acad. Sci., VIII, 1893-95, p. 283). A. laticeps has three distinct small denticles at the base of the maxillary. Astyanax lineatus (Perugia). Tetragonopterus lineatus Cope, Proc. Amer. Philos. Soc. Phila., XXXIII, 1894 (January 5), p. 107. Near Chapada in Matto Grosso from the head- waters of the Paraguay. H. H. Smith. Head 3^; depth 2f ; D. in, 9; A. iv, 25, i; scales 33 in lateral line to 1906.] NATURAL SCIENCES OF PHILADELPHIA. 349 base of caudal, and 3 more on latter; 7 scales in an oblique series back from origin of dorsal to lateral line; 6 scales in a vertical series from base of ventral to lateral line, and same number from origin of anal to latter: about 15 scales before dorsal; pectoral H in head; ventral If; lower caudal lobe 1 ; snout 3f in head, measured from tip of upper jaw ; eye 3; interorbital space 2f ; least depth of caudal peduncle 2t. At least one small denticle at base of maxillary. Posterior distal ex- tremity of maxillary extending till opposite anterior margin of pupil. Rakers about 9 + 15, slender, pointed and longest much shorter than filaments. In alcohol with a pale dusky humeral blotch. A dash of dusky on middle of side of caudal peduncle at base of caudal, fading out over posterior portion of anal, though posterior it extends out on middle caudal rays. About ten distinct longitudinal brownish lines on body, darker on back, and made up of dark brownish specks or dots. Iris brownish with a darker brownish ring. Length 3if inches. Dr. Steindachner's figure in Sitz. Ak. Wiss. Wien, C, 1891, p. 368, PI. 2, fig. 1, does not show the dusky streak on the side of the caudal peduncle. Astyanax chapadae sp. nov. Fig. 33. Head 3f ; depth 3; D. m, 8; A. iv, 18; P. i, 9; Y. i, 7; scales 39 m lateral line to base of caudal, and 3 more on latter; 6 scales from origin of dorsal obliquely back to lateral line, and 4 in same way from origin of adipose fin; 4 scales between origin of ventral and lateral line obliquely forward; 4 scales between origin of anal and lateral line; 16 scales before dorsal; width of head HI in its length; depth of head over middle of orbit, U; snout 3^; eye 3; maxillary 2^; interorbital space 3; mandible 2f ; length of depressed dorsal H; pectoral 1^; ventral 2; upper caudal lobe l^V; least depth of caudal peduncle 2f . Body elongate, compressed, rather ovoid, lower profile a little more convex than upper, and greatest depth at origin of dorsal. Predorsal recrion rounded. Postdorsal region with a median obtuse keel. Pre- veiitral region rounded. Postventral region slightly trenchant. Caudal peduncle compressed, and its least depth about f its length. Head a little long, somewhat attenuated, well compressed, muzzle a little pointed, upper profile a little concave above eye posterioriy, and lower profile rather evenly convex. Snout a little long, convex, and produced a little beyond tip of mandible. Eye circular, a little superior and its center about first f in head. Mouth a little oblique. Maxillary obliquely vertical, its distal expanded extremity reaching about op- posite front margin of pupil and its width about half of diameter of same. Mandible strong. Lips rather fleshy. Teeth in jaws mostly 350 PROCEEDINGS OF THE ACADEMY OF [June, quindeiitate, especially those in mandible anteriorly which are also long and pointed, and five in each outer series above. At least three small serrated teeth along base of maxillary. Tongue elongate, fleshy and not free. Nostrils together near upper front rim of orbit. Interorbital space a little convex. Greatest width of infraorbital rim a little less than orbit or much broader than greatest exposed width of opercle. Gill-opening extending forward opposite front margin of pupil. Rakers about 7 + 10, lanceolate, rather far apart, and longest about f of longest filaments. Filaments about equal to pupil. Isthmus narrow. Branchiostegals 4. Scales large, well striated, largest on flanks, and in longitudinal Fig. 33. — Astyanax chapadce Fowler. series parallel with lateral line. Basal anal sheath low and of small scales. Base of caudal well covered with small scales. Axillary ventral scale about ^ length of fin. Lateral line well decurved and running rather low posteriorly to base of caudal. Tubes simple. Origin of dorsal a little nearer base of caudal than tip of snout, fin low, and when depressed reaching f of space to base of caudal. Adipose fin small, its origin near last third in space between origin of dorsal and base of caudal, or just behind base of last anal ray, and length of fin less than orbit. Origin of anal about opposite base of penultimate dorsal ray, or about midway between that of pectoral and base of caudal. Caudal rather small, emarginate, and lobes rounded. Pectoral falling a little short of ventral. Ventral 1906.] NATURAL SCIENCES OF PHILADELPHIA. 351 inserted nearly midway between tip of snout and tip of last anal ray, or well before dorsal, and reaching about f of space to anal. Vent close in front of anal. Color in alcohol brown, especially back, sides and lower surface silvered and pale. A humeral blotch. A faint trace of a slightly lighter silvered band from shoulder to base of caudal, posteriorly becoming plumbeous though only slightly dusky at base of caudal, and same shade also continued out on median caudal rays to their tips. Base of each caudal lobe with a pale straw-colored blotch about size of pupil, then membranes between rays blackish and fading out into pale dusky posteriorly. Dorsal brownish with an indistinct dusky cross-streak. Other fins all pale and unmarked. Iris brownish. Length 2|- inches. Type, No. 21,828, A. N. S. P. Near Santa Anna da Chapada, in Matto Grosso, Brazil, from the headwaters of the Paraguay. August, 1884. H. H. Smith. Prof. E. D. Cope. Also paratype. No. 21,829, A. N. S. P., with same data. This species is closely related to Astyanax tnoenkhausii (Eigemnann and Kennedy), from which it differs chiefly in the coloration. A. paucidens (Ulrey) and A. jenynsii (Steindachner) both differ in the fewer scales, apparently not more than 33. (Named for Santa Anna da Chapada, province of Matto Grosso, Brazil.) 352 PROCEEDINGS OF THE ACADEMY OF [July, DESCRIPTIONS OF TWO NEW POLYCHJSTA FROM ALASKA. BY J. PERCY MOORE. Among a number of Alaskan Polychseta in the collections of this Academy, the two of which the descriptions follow are new. Syllis quaternaria sp. nov. This sexual free-swimming form of the type for which Malmgren instituted his genus Chcetosyllis is probably the epitokous form of a true Syllis, but it cannot be correlated with any atokous form hitherto described. The length varies from 7 to 9 mm.; the width between the ends of the parapodia is .8 mm. and between the tips of the capillary setsB 4.5 mm., or one-half the length of the body. Syllis quaternaria. — a, dorsal view of prostomium, peristomium and first two somites bearing capillary setse, X 24; 6, parapodium from IX with setse cut off, X 32 ; c, a long and d, a short notopodial seta from X, X 600. The prostomium (a) consists of a pair of prominent hemispherical lobes separated by a deep, narrow, median cleft, and each again divided by a shallow transverse groove. A pair of simple unjointed or obscurely wrinkled cylindrical tentacles arise from the latter grooves and incline decidedly toward the dorsum. They reach a length of 1|- to 1-J times the prostomium. No other cephalic appendages exist, but there are two pairs of dense, black or deep brown, short spindle-shaped eyes, each provided with a small cuticular lens. The dorsal eyes are the more posterior and look vertically upwards ; the large ventral pair look downwards, forwards and outwards. Sometimes 1 or 2 small, rod-like 1906.] NATURAL SCIENCES OF PHILADELPHIA. 353 bodies of pigment occur with the eyes. The mouth is a minute open- ing situated at the very bottom of the vertical prostomial cleft near the ventral surface. The peristomium (a) is but little modified and a segment just like the others, except that the parapodia lack the capillary setae. The sides of the body are nearly parallel almost to the caudal end, where the last 12 to 15 of the total number of 40 to 43 segments taper to the pygidium. Each segment is about ^ as long as wide, or anteriorly rather more than this. The small triangular pygidium bears a pair of large, stout moniliform cirri of about 25 joints, and in the males an additional short, slender, and un jointed median ventral cirrus. Excepting the first pair the parapodia (b) are biramous, and all but the first and the last few bear capillary setse in the notopodium. The basal part is a rather large dome-shaped structure, from which the neuropodium projects rather prominently and ends in a short, slightly curved dorsal acicular process, while a postsetal lamina, stepped at the ends for the rows of setae, forms its more ventral portion. Two or three straight aciculi support the neuropodium. From the middle of the ventral surface of the foot arises a short, slender, tapered and un- jointed ventral cirrus. The notopodium is a very slight, conical tu- bercle which receives the end of a single aciculum strongly curved at the tip and caps a large yellowish setigerous sac, from which the capillary setae spring. Well above and separated from this tubercle arises the dorsal cirrus, which is always prominent and strongly moniliform. With considerable regularity the dorsal cirri are alternately longer and shorter, the even numbers being about equal to f of the width of the body exclusive of the parapodia and consisting of 20 to 25 joints, whereas the odd numbers are about f as long as they and have only 14 to 16 joints. As stated above, all but the peristomium and the last 5 or 6 segments bear capillary notopodial setae which form dense and very conspicuous bundles. They are perfectly smooth and simple in structure, glass- like in appearance, tapered to excessively fine but apparently flattened ends, and are more or less curved and sometimes curled and twisted at the ends — the latter condition being probably the result of preserva- tion. Neuropodial setae are arranged in 5 or 6 horizontal sub-acicular rows. They are relatively short and stout and perfectly colorless and vitreous. The curved stems are deeply cupped, and toothed along the more pro- longed side. The blades end in a single, rather stout, little hooked tooth, and the dorsal or shorter side is provided with a moderately 354 PROCEEDINGS OF THE ACADEMY OF [July, strong fringe best developed on the shorter blades, which also exhibit an aggregation of some of the terminal hairs to simulate an accessory tooth. On the ventral setae the blades seldom exceed a length of twice the diameter of the end of the shaft, while those of the 3 or 4 setas in the dorsalmost row are narrower and 4 or 5 times as long as the distal end of the shaft. Although all of the specimens examined are fully mature and have the body cavity filled with ova or sperm, the alimentary canal is com- plete and well developed, extending continuously from mouth to anus without exhibiting any of that occlusion or degeneration seen in many sexually mature syllids and other annelids. The type and about a score of cotypes are No. 1,091 of the Academy's collections. Taken on the "surface in a lead 4 miles from shore by Mr. E. A. Mcllhenny at Point Barrow, Alaska." Ammotrypane brevis sp. nov. The type of this species is a somewhat contracted specimen 15 mm. long, 1.5 mm. broad and 1.9 mm. high in the middle, being therefore rather robust and tapering to both ends. The general resemblance to A. aulogaster is close, but if the differences exhibited by the single speci- men prove to be constant the two species are readily separated. Ammotrypane 6re)'/s.— Parapodium and gill from somite X, X 56. Instead of being laterally compressed the prostomium is blunt and flattened dorso-ventrally. From it a slight median ridge passes back- ward and gradually merges into the dorsal curvature of the back. Below this ridge on each side are the small sensory pits or nuchal organs. The mouth is a small trifid opening bounded by a pair of small anterior lobes and a single posterior lobe. Although the body seems to be quite complete and exhibits no indications of regeneration, there are only 29 setigerous somites not distinguishable from one another externally except through the presence of the parapodia. No intersegmental furrows whatever are discernible and the segments pass continuously into one another. Ventrally a pair of stout muscular bands, separated by a deep median groove, produce a somewhat sole- like surface. Above this the body is narrower and strongly arched 1906.] NATURAL SCIENCES OF PHILADELPHIA. 355 and marked only by the small transverse muscle bundles, apparently 4 to each segment. If perfect, as it appears to be, the pygidium presents striking char- acteristics. The large median spoon-shaped lobe of A. aulogaster is absent and represented only by a minute slender process. The lateral lobes are much larger, obliquely truncated above, and slightly indented at the end. The parapodia arise from the dorso-lateral surface of the lateral muscle bands and are about equally well developed throughout. They form small but distinct bluntly rounded tubercles produced into a small, conical prominence and apparently unsupported by an aciculum. They bear only a small ventral cirrus and a prominent dorsal branchia, the small dorsal cirrus which has been figured for some species being absent. Branchise are absent from the first pair of parapodia, but are present on all of the others. They are largest on the middle segments, where they exceed a millimeter in length and the parapodium in diameter. They are slender and slightty tapering in form and provided along one side with a strongly developed longitudi- nal muscle, rendering them very contractile. The setae are all simple, colorless, tapering, capillary bristles, longi- tudinally fibrillated and soft in texture. They are arranged in two fascicles, a dorsal of 6 to 10 setse which reach f the length of the branchia, from the base of which they arise, and a ventral which arises from beneath the conical prolongation of the parapodium and contains 5 or 6 setae less than h as long as the dorsal bristles. Along the sides of the body between the parapodia, but nearest to the succeeding one, are small black or dark brown and very conspicuous eye spots, which are hemispherical in shape and indistinctly facetted on the surface. The only specimen, a female filled with eggs, is No. 284 of the Acad- emy's collection, taken by Dr. Benjamin Sharp at Icy Cape, Alaska, on August 17, 1895. 356 PROCEEDINGS OF THE ACADEMY OF [July, NOTE ON THE DUSKY SALAMANDER. BY HENRY W. FOWLER. Near Port Allegany, in McKean county, and in the valleys of the headwaters of the Genesee, Allegheny and Susquehanna, in Potter county. Pa., Mr. T. D. Keim and the writer spent a few days, studying the fauna, in the summers of 1904 and 1906. Most of the hemlock forests have been cleared off, except near Port Allegany, where one still has but a short time left to study the animal life in its primeval conditions. Perhaps the most abundant in individuals among cold-blooded verte- brates was the little dusky salamander, Desmognathus fusca. It was found nearly everywhere in the valleys, about and in streams, and on the comparatively dry mountain-tops, though there less numerous and more solitary. About the headwaters of the Genesee they were even more abundant, so that we had the opportunity of examining multitud&s of living examples. So abundant were they in some localities that as many as four or five would frequently occur under a single small stone. Owing to the nature of the brooks about the Genesee headwaters, which are very swift, cold and sinking out of sight in some places to appear again in others, very favorable habitats are secured in the comparatively damp intercovirses which are only flooded in times of heavy rains. Very great range in color-pattern and color-variation was noted in this species. Young an inch long had the pale area on the back more or less immaculate pink or pale brown. The uppermost of the accom- panying figures, Plate XIII, indicates this type. They varied from these shades into the dusky-black adults. They did not always appear to change from pinkish to brown, as some with pink backs were found equally as large as the larger brownish ones. The pale colors of the back were not always immaculate, many being variously blotched or speckled with darker or brownish. Very dark or blackish examples, not larger than the largest pink or brown ones, also occurred, though these all showed traces to some extent of the dorsal color-pattern. Most all full-grown examples lose every trace of the dorsal color-pattern in the blackish of the back. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 357 A comparison of the types of Desmognathus ochrophcea Cope seems to leave it perfectly clear that it is nothing more than the intermediate brown color-variety of D. fusca, which I have represented in the tw^o lower median accompanying figures. The identity of D. ochrophcea with D. fusca, however, was fii-st suggested by Mr. G. M. Allen in the Proc. Bost. Soc. N. H., 29, 1901, p. 73. In view of the material ex- amined I am now able to fully endorse Mr. Allen's view as final. The characters pointed out by Cope in 1889 are certainly not of specific value as they range more or less through most of my material. The habitat of the ochrophcea stage is not confined entirely to the mountain- tops, as the animals are found under damp stones about streams and in the valleys. I do not agree with Cope's suggestion that Salamandra haldemani Holbrook is "an unusually spotted Spelerpes hilineatus'' as the tail is somewhat short. It agrees with some of my examples of D. fusca to some extent in the color of the belly, which is usually more or less livid in life. It had best be regarded as a synonym of D. fusca. As small and medium-sized examples greatly resemble Plethodon erythronotus, which is rather abundant in the dry forests, a good char- acter for distinguishing the two species at a glance will be found in the pale or whitish streak extending down from the eye behind. This is present on most all examples and may frequently be traced to some extent in the blackish adults. 24 358 PROCEEDINGS OF THE ACADEMY OF [Sept. A CONTRIBUTION TO THE KNOWLEDGE OF THE ORTHOPTERA OF MONTANA, YELLOWSTONE PARK, UTAH AND COLORADO. BY JAMES A. G. REHN AND MORGAN HEBARD. During the month of August, 1904, the junior, author was occupied in collecting Orthoptera at a number of localities extending from eastern Montana to Salt Lake City, and from that locality to east- central Colorado. Collecting at some localities was limited to a few minutes search permitted by the stoppage of the train, at other locali- ties several days were spent and very extensive and interesting collec- tions made. The careful study of the Orthoptera of the Yellowstone Park and of Pike's Peak yielded some interesting and scarce species, as well as two new forms and a considerable amount of information on the habitat and actions of a number of species. The number of specimens in the collection studied is about 2,800, and 121 species are represented. Notes on the habits and other infor- mation has been given under most of the species by the collector, and the chief localities visited have been briefly described, to aid in a proper understanding of their conditions and environmental features as well as to aid future students in this region. A considerable representation of the species treated in this paper has been placed in the Academy's collection by the junior author. Montana. Waco, Yellowstone Co., elevation 2,700 feet. On the sage-brush plains just east of the foothills of the Rockies. Grey Cliff, Sweetgrass Co., elevation 3,900 feet. In a more rugged character of country than Waco but with a similar vegetation. Orthoptera appeared to swarm as the specimens collected were taken in less than two minutes. Billings, Yelloivstone Co., elevation 3,115 fe^t. Specimens taken here were from the dusty main street of the town. Livingston, Park Co., elevation 4,488 feet. Material from this locality was taken among weeds growing in a vacant lot near the station. Emigrant, Daileys, Sphinx and Electric, Park Co., elevation 4,800- 1906.] NATURAL SCIENCES OF PHILADELPHIA. 359 5, 1 00 feet. Dming the brief stops made by the train between Livingston and Gardiner a httle collecting was possible. These stations are situated in the valley of the Yellowstone river between the Snowy Jlountains and the Gallatin Range. The vegetation is chiefly sage- brush and other plants which grow in a semi-arid soil. Orthoptera were very plentiful, and had time allowed a great quantity could easily have been taken. Muir, Bozeman Tunnel, Park Co., elevation 5,49G feet. At this point a few minutes' stop afforded an opportunity to collect in the heavy growth of weeds and grasses along the track. The vegetation was very luxuriant and wholly different from that found in the semi- arid sections. The Orthoptera of regions of this character was always much less abundant than in the sage-brush country. Three Forks, Willotv Creek and Sappington, Gallatin Co., and Jefferson City and Whitehall, Jefferson Co., elevation 4,061-4,529 feet. During the brief stops made at these towns a few specimens were taken. The soil w^as barren and Orthoptera was far less plentiful than in the more eastern portion of the State. Spire Rock, Jefferson Co., elevation 5,217 feet. Among the great boulders on the summit of the mountains near Butte a few specimens were seen and hardly any vegetation found. This locality was by far the most desolate of any at which collecting was done. Yellowstone Park. Mammoth Hot Springs, elevation 6,215-7,000 feet. At Fort Yellow- stone, where the Mammoth Hot Springs are situated, a great variety of collecting grounds are within easy reach. About the fort itself, among the weeds and grasses, Orthoptera were not very plentiful ; but immediately behind the hotel, in the dense sage-brush growing on the sides of the foothills, vast quantities of Orthoptera of many species were to be found. Collecting here was somewhat difficult owing to the steep hillside and the agility of the majority of the species, as well as the difficulty of following more interesting specimens on account of the quantities of more common species which everyw^here swarmed. At a greater elevation on the sides of these hills the sage-brush gave place somew^hat to grasses and other low growing plants, where Orthop- tera were less prevalent but more desirable. On the summits, however, the ground was pebbly and overgrown with sparse and stunted grasses, in which situations many most desirable species were abundant and could be easily taken in the scanty vegetation. In one grass}^ hollow near the top of the foothills, among scattered bushes near a thicket 360 PROCEEDINGS OF THE ACADEMY OF [Sept., of aspen, I found species which did not seem present in the surrounding sections, among which were Hippiscus maculatus, Circotettix rabula, Bradynotes ohesa and Mdanoplus alpinus. At the head of the hot springs I found a plot of marsh grass among the stunted pines which revealed several species of crickets and other Orthopterous insects not found in the surrounding country. On the summits of the higher foothills nearby an even greater variety of the more interesting forms were found. Apollinaris Spring, elevation 7,250 feet. iMaterial collected here was from dense pine woods. Fountain, Lower Geyser Basin, elevation 7,230 feet. Very few specimens could be found here, the pine woods nearby revealing nothing whatever, while but very few individuals were taken on the geyser formation and in the short grass near the hotel. This was the most unfavorable locality found in the Park. Old Faithftd Inn, Upper Geyser Basin, elevation 7,830 feet. Here two distinct collecting localities were accessible in which few though interesting forms were found. Between Old Faithful geyser and the Firehole river a space of short grass contained numerous Arphia pseudonietana, while on the surrounding geyser formations the same species could be found in very small numbers. Back of the Inn, among the short but thickly-growing pines, several species of Melanoplus {M. hruneri and fasciatus) were found which were not met with else- where in the Park. Unlike most of the Yellowstone localities, long and diligent search was required to secure each specimen. Norris Pass, Continental Divide, elevation 8,240-8,320 feet. About the Norris Pass the vegetation was scanty owing to the elevation. In the more open spaces a short grass grew plentifully and several species of Orthoptera could have been taken in numbers had time permitted. The Thumb, Yellowstone Lake. About The Thumb on the shore of Yellowstone Lake the vegetation was heavy, but owing to the few cleared spaces but few Orthoptera were found. Exit of the Yellowstone River, elevation 7,742 feet. At this point above the high abrupt bank on the left side is a large treeless expanse covered by high plants and shrubs. In this location all forms of insect life were very plentiful, and many varieties of boreal Lepidoptera were noticed. The Orthoptera of this section differed materially from that found in the more arid geyser regions. Mud Volcano, elevation 7,740 feet. CoUecting here was done on sandy soil. Near Grand Canon, elevation 8,000 feet. The country about the 1906.] NATURAL SCIENCES OF PHILADELPHIA. 361 Grand Canon much resembled that about Yellowstone Lake, but Orthoptera were less abundant. As in other parts of the Park, very few specimens could be found in the woods, Utah. Vicinity of Salt Lake City, Salt Lake Co., elevation 4,230-6,000 feet. About Salt Lake City I found the vicinity of Eixsign Peak (top 6,000 feet elevation) by far the most productive collecting groimd. There the soil was dry and gravelly, but covered with a somewhat sturdy growth of sage-brush and other desert plants. In this place Orthoptera swarmed and great difficulty was experienced in following scarce specimens among the legions of less desirable ones. About the foot of the hill was the most productive situations, but great numbers were found both on the hillside and at the top. In a dry field of cut alfalfa some miles south of the city (elevation 4,230 feet) collecting was profitable among the stubble, although both there and in the surrounding countiy far fewer species were to be found than w^as expected and nothing of the variety found in the more barren region about Ensign Peak. Colorado. Grand Junction, Mesa Co., elevation 4,573 feet. In the dry sandy country at this locality quite good collecting was found, though recent freshets had washed over much of the groimd, and the only vegetation in most places consisted of tall scraggy bushes. Antlers, Garfield Co., elevation 5,310 feet. An area of cactus and sage at this locality proved very productive in the few minutes spent there. Debeque, ]\Iesa Co., elevation 4,935 feet, and Newcastle, Garfield Co., elevation 5,552 feet. Both localities in the rather arid canon of the Grand River. Vicinity of Colorado Springs, El Paso Co. To the east of the city in the prairie (elevation about 6,000 feet) Orthoptera w'ere present in countless numbers. The true prairie-grass land revealed many species, while the more luxuriant grass about a small stagnant pond produced other species. This locality, while somew^hat dry, was nothing like the parched sage-brush plains in which collecting was done at other places. Austin Bluffs (elevation 6,960 feet) rise abruptlj^ from the prairie land, many dry shrubs growing upon them, plentifully interspersed with scrub pines. Vicinity of Manitou, El Paso Co. The mesa (elevation 6,400 feet) 362 PROCEEDINGS OF THE ACADEMY OF [Sept., at this locality was somewhat like the prairie at Colorado Springs, but not nearly as productive. On the hillside near the Pike's Peak railroad station (elevation 6,700 feet) Orthoptera were plentiful among the high grasses and scrub oak thickets. The Garden of the Gods (eleva- tion 6,300 feet) proved everywhere unproductive except a small field of tall grass near the Manitou street-car line. Many species were to be found here in numbers. Pike's Peak. The' summit of Pike's Peak (elevation 14,147 feet) is a mass of huge boulders with no apparent vegetation. The region about Windy Point (12,000-12,300 feet) is treeless and very steep. A short but sturdy grass grows everywhere that the quantity of huge jagged rocks permit, and in this grass a few species of Orthoptera such as Gomphocerus clavatus, Melanoplus altitudinum, fasciatus and monticola were quite plentiful. Along the timber-line (11,578 feet) in the stunted vegetation Orthop- tera were somewhat more varied, but not as interesting as those at Windy Point. At Middle Hudsonian (elevation 10,500 feet) and above the pump station (elevation 10,100 feet) the forest of quaking aspen was exceedingly unproductive and very few specimens were seen. In Dark Canon (elevation 8,900-9,000 feet) in spite of an abundance of grass few Orthoptera could be found, although other forms of insect life were plentiful. On the densely wooded upper slope of Dark Canon work was attended with less success than in the burned-over forest. Cripple Creek, Teller Co., elevation 9,591 feet. The few specimens taken at Cripple Creek were from the tailings about the mines. Fort Morgan, Morgan Co., elevation 4,338 feet. A rank growth of weeds along the railroad track at this place swarmed with Orthoptera, and in a few minutes a number of interesting forms were taken. MANTID^. LITANEUTRIA Saussure. Litaneutria minor (Scudder). Antlers, Colorado, Aug. 15, 1 immature $ . This specimen is referred to this species with some uncertainty, as it may represent L. horealis Bruner, which, however, is of doubtful validity.^ This individual was taken on dry soil among dense beds of cactus and sage. 1 Vide Caudell, Proc. U. S. Nat. Mus., XXVIII, pp. 463, 464. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 363 AORIDID^. ACRYDIUM Geoffrey. Acrydium granulatum Kirby. Hill at head of springs, Mammoth Hot Springs, Yellowstone Park, Aug. 5, 2 immature specimens. Careful search revealed no mature specimens. Acrydium crassum (Morse). Manitou, Colorado, 6,324 feet, Aug. 16, 3 d', 8 ? ; 6,700 feet, Aug. 23,' 1 c?, 1 9. This species has previously been recorded from localities in northern and north-central Colorado: Fort Collins, Laporte, Steamboat Springs, Poudre river, Platte Canyon and Denver. This species was found among the mountain-loving plants in the immediate vicinity of scrub oak bushes. It appeared to be extremelj^ local, for I found it in but one small area, where, however, it was common. ACRIDINiE. SYRBULA Stai. Syrbula admirabilis (Uhler). St. Louis, Missouri, Aug. 27, 18 d', 12 9 . These specimens are inseparable from New Jersey individuals. A common species in the tall weeds of a vacant lot in the city. ACROLOPHITUS Thomas. Acrolopliitus hirtipes (Say). Mammoth Hot Springs, Yellowstone Park, 6,215 and 6,500 feet, Aug. 5, 2 d^, Colorado Springs, Colorado, Aug. 18, 4 d^, 9 9 . Prairie land, Colorado Springs, Aug. 18, 1 9 . Knob Hill, Colorado Springs, Aug. 17-22,7 J^, 13 9. This series exhibits a great range of color variation, extending from the typical blotched form to the uniform one tentatively named unijormis by Bruner.^ The latter does not appear worthy of recogni- tion, as the extremes and intermediates are present in both the larger series examined. This species was quite scarce on the hillsides about the iMammoth Hot Springs, being found among sage and mountain plants. On the prairie outside of Colorado Springs the species was extremely plentiful and ^ Biol. Cent.-Amer., Orth., II, p. 47. 364 PROCEEDINGS OF THE ACADEMY OF [Sept., easily capkired. The insect is not swift of flight and often relies on its strongly protective coloration to escape detection. ,^:' Fig. 1. — Akentetus unicolor McNeill. Knob Hill, Colorado Springs, Colorado. Lateral -view of male. (X 5.) AKENTETUS McNeill. Akentetus unicolor McNeill. Mannnoth Hot Springs, Fig. 2. — Akentetus unicolor Mc- Neill. Knob HiU, Colorado Springs, Colorado. Dorsal \ae-w of head and pronotum of male. (X 5.) Yellowstone Park, Aug. 5, 1 9 . Knob Hill, Colorado Springs, Colorado, Aug. 17, 2 (^. Akron, Washington Co., Col- orado, Aug. 24, 1 6^. This scarce species, previously re- corded only from the type locality ("Colorado") and Lakin, Kansas, is also represented in the Academy collec- tion by a male from Nebraska and a female from Livermore, Larimer Co., Colorado, the latter taken July 23, 1899. The pattern of coloration is very similar to that of Amphitornus, but the longitudinal bars are narrower. In the Akron and Mammoth Hot Springs specimens the postocular regions are distinctly infuscate. The bars on the caudal femora are distinct in all the specimens examined. The four specimens of this species taken were captured in three decidedly different locations. At Mammoth Hot 1906.] NATURAL SCIENCES OF PHILADELPHIA. 365 Springs one was taken on the hillside among sage-brush. At Ivnob Hill two were captured on the prairie, while at Akron it was taken in a dry situation near the railroad. AMPHITORNUS McNeill. Amphitornus coloradus (Thomas). 1872. S[tenobothrus] bicolor Thomas, Prelim. Rep. U. S. Geol. Sur. INIontana, p. 465.3 1873. S[tenobothrus] coloradus Thomas, SjTiopsis Acrid. N. Amer., p. 82. ]\Iammoth Hot Springs, Yellowstone Park, Aug. 5, 4 9,1 nymph. Summit of foothills near Mammoth Hot Springs, 6,500 feet, Aug. 5, 3 9 . Hill at head of Mammoth Hot Springs, 7,000 feel, Aug. 5, 1 9 . Salt Lake City, Utah, Aug. 13, 1 9 . Hill near Salt Lake City, 4,700 feet, Aug. 13 and 14, 3 cJ^, 2 9 . Knob Hill, Colorado Springs, Colorado, Aug. 17-22, 4 c?, 1 9 . Prairie land, Colorado Springs, Aug. 18, 1 9 . The series examined contains both the typical form and the suffused form called "Var. a" by Thomas. This is apparently the first record of the species from Utah. This species was found in Wyoming, Utah and Colorado in almost the same numbers. Each specimen seen was captured, if that was in any way possible. The insect was found among the sage on hillsides and also among the grasses of the prairie. OPEIA McNeill. Opeia obscura (Thomas). Salt Lake City, Utah, Aug. 13, 1 9 . Colorado Springs, Colorado, Aug. 18, 1 9 . Knob Hill, Colorado Springs, Aug. 17-22, 65 d"', 62 9 . Prairie land, Colorado Springs, Aug. 18, 2 c^, 2 9. Garden of the Gods, Colorado, Aug. 19, 1 &. Specimens of this species were received from Prof. C. P. Gillette from Merino, Lamar, Snyder and Fort Collins, Colorado, labelled Eritettix variahilis with a query. The Salt Lake City specimen is apparently the first recorded west of the Rockies. About Colorado Springs on the prairie this species was to be found in countless numbers. The males are very rapid in their movements, but so common were they that sometimes over a dozen would be taken with one sweep of the net. CORDILLACRIS Rehn. Cordillacris occipitalis (Thomas). Colorado Springs, Colorado, Aug. 18, 1 9 . Knob Hill, Colorado 3 Invalidated by Stenobothnis bicolor (Charpentier) , the combination dating from Selys-Longciiamps, 1868 (Ann. Soc. Entom. Belg., XI, p. 31). 366 PROCEEDINGS OF THE ACADEMY OF [Sept., Springs, Aug. 18-22, 1 6^, 2 9. Austin Bluffs, Colorado Springs, Aug. 18, 8 c?, 6 9 . Prairie land, Colorado Springs, Aug. 18, 1 d^. This was an extremely scarce species on the grassy prairie, but among the low plants growing in a forest of dwarf pines on Austin Bluffs it was quite common. The insects spring about rapidly and are also able to fly well. Cordillacris cinerea (Bruner). Sphinx, Park Co., Montana, Aug. 4, 1 d^. Austin Bluffs, Colorado Springs, Colorado, Aug. 18, 2 9 . This form appears to be hardly separable from the preceding. The male captured was taken in the sage-brush. So rapidly did it jump about that several minutes were occupied in making the cap- ture, as I had no net at the time. Cordillacris crenulata (Bruner). Antlers, Garfield Co., Colorado, Aug. 15, 2 9 . Colorado Springs, Colorado, Aug. 18, 1 9 . Knob Hill, Colorado Springs, Aug. 17-22, 8 (J^, 10 9 . Austin Bluffs, Colorado Springs, Aug. 18, 1 9 . Garden of the Gods, Aug. 17 and 19, 1 d', 3 9 . This species was quite plentiful in the grass of the prairie outside Colorado Springs. It was also taken among cactus and sage, in a damp grassy meadow and among the low scattered plants growing in a forest of dwarf pines. PHLIBOSTROMA Scudder. Phlibostroma quadrimaculatum (Thomas). Colorado Springs, Colorado, Aug. 18, 2 d^, 2 9 . Knob Hill, Colo- rado Springs, Aug. 17-22, 56-6^, 39 9 . Manitou, Colorado, Aug. 23, 1 9 . Garden of the Gods, Colorado, Aug. 17-23, 8 d", 5 9 . Akron, Washington Co., Colorado, Aug. 24, 1 c^, 1 9 . One of the extremely plentiful species found in the prairie. Hun- dreds could have been taken in a short time. ■ Quite a diversity of coloration occurred in the specimens, for among those taken were both brown and green-marked individuals. The females were generally of the green form, while the males were more of the wholly brown type. OEPHULELLA Giglio-Tos. Orphulella pelidna (Burmeister). Knob Hill, Colorado Springs, Aug. 17-22, 5 d^, 5 9. . The separation of 0. pratorum from this form is a task apparently hopeless. For the present at least we prefer to call these pelidna, as pratorum as understood by us, on the basis of material from the south- 1906.] NATURAL SCIENCES OF PHILADELPHIA. 367 eastern States and New Jersey, appears to be slenderer with longer wings. However these differential characters are only averages, numerous specimens which cannot otherwise be separated from the usual pratorum type being intermediate in proportions. The species was found in short marshy grass growing about a shallow pond in the prairie. Considerable search was required to find the ten specimens captured. Orphulella salina Scudder. Grand Junction, Colorado, Aug. 15, 1 6^. This species has previously been recorded from Grand Junction and Montrose, Colorado, by Caudell,* and Grand Junction and Delta by Gillette.^ This specimen was taken in short grass near the Grand river, in a locality which had been recently inundated. No other specimens of Orthoptera were seen about this location. DICHROMORPHA Morse. Dichromorpha viridis (Scudder). Table Rock, Pawnee Co., Nebraska, Aug. 25, 1 c?, 1 9 . St. Louis, Missouri, Aug. 27, 2 9 . In both localities plentiful among thick grass and weeds. CHLOiiALTIS Harris. Chloealtis abdominalis (Thomas). Summit of foothill of Gallatin Range, Mammoth Hot Springs, Yellowstone Park, 7,000 feet, Aug. 5, 4 d", 7 9 . Manitou, Colorado, 6,300-6,900 feet, Aug. 23, 7 d", 4 9 . The male individuals are inseparable from a specimen of the same sex from Pequaming, Michigan, while the females are similar to individuals of that sex from Manitoba and Truchas Peak, New IMexico. The Truchas Peak specimen is slightly smaller with rather slenderer caudal limbs. This species was doubtfully recorded from Colorado by Cockerell ^ as taken at Willow Creek, Cusack Ranch, Custer County, in September. At Mammoth Hot Springs I found this species in a small glade at the top of the foothills. Here in the thick grass growing about a clump of aspen the series listed above was taken with little difficulty. The males of this species are ^-ery quick and wary, but the females are * Proc. U. S. Nat. Mus., XXVI, p. 782. 5 Bull. 94, Colo. Agr. Exp. Sta., p. 26. « Trans. Amer. Ent. Soc, XX, p. 337. 368 PROCEEDINGS OF THE ACADEMY OF [Sept. extremely heavy and slow, and appear to rely a great deal on their protective resemblance to a bit of twig. The stridulation of the j^^ Fig. 3. — ChloeaUis ahdominalis (Thomas). Manitou, Colorado. Lateral Adew of male. (X 4.) ^ '^i. -r^" Fig. 4. — CMoeaUis ahdominalis (Thomas). Mammoth Hot Springs, Yellowstone Park. Lateral view of female. (X 3.) males much resembles that of C. conspersa, but is not nearly so loud. At Manitou, I found this species in about the same numbers among the scrub oaks and mountain-loving plants of a steep hillside. The 1906.1 NATURAL SCIENCES OF PHILADELPHIA. 369 species appeared to be very local in habitat and, if a colony of the insect is not found, long search for the species would very probably be useless, STATJKODERUS Bolivar.^ Stauroderus curtipennis (Harris). Muir, Bozeman Tunnel, Park-Gallatin Co., Montana, Aug. 12, 1 9 . Sappington, Gallatin Co., Montana, Aug. 12, 1 d". Hill at head of Mammoth Hot Springs, Yellowstone Park, Aug. 5, 2 cJ*, 1 ? . Fountain, Lower Geyser Basin, Yellowstone Park, Aug. 6, 1 6^. Near Grand Canon, Yellowstone Park, 8,000 feet, Aug. 10, 2 c?, 2 ? . Yellowstone Lake, Aug. 8, 5 d', 4 9 , Emerald Springs, Yellowstone Park, Aug. 7, 2 c^ , 2 9 . Knob Hill, Colorado Springs, Aug. 17, 1 9 . Manitou, Colorado, Aug. 16 and 23, 5 cJ^,5 9.- These specimens vary considerably in size and somewhat in structure, but such variation is par- alleled in Eastern specimens, from which they cannot be separated. This species was rather boreal in distribution and was found in numbers in tall marshy grass in a large open pasture on the shore of Yellowstone Lake. It was also found plentiful at Manitou, Colorado, in the same locality as Chloealtis ahdominalis. At Muir, Mammoth Hot Springs and Manitou the species was found in dry locations, all the other localities in which it was found being more or less damp. The stridulations of the males resemble sik-sik-sik-sik swiftly repeated, the sound being neither loud nor sharp. Fig. 5. — Chloealtis ah- dominalis (Thomas). Mammoth Hot Springs, Yellowstone Park. Dorsal view of head and prono- tum of female. (X3.) PLATYBOTHRUS Scudder. Platybothrus brunneus (Thomas). ]\Lammoth Hot Springs, Yellowstone Park, Aug. 5, 2 d^. Suimnit of foothills near Mammoth Hot Springs. Aug. 5, 4 d^, 4 9 . Sunmiit of foothill of Gallatin Range, Mammoth Hot Springs, Aug. 5, 1 d^, 5 9 . ' This generic name (Catal. Sinopt. Ortopf. Fauna Iberica, pp. 46, 57, 1898) should be used in this connection in place of Stenobothrus if we recognize Bolivar's di\-isions as genera, as has been done by Burr {Entom. Record and Journ. Variat., XVI, p. 320) . If these groups are recognized as of only subgeneric rank, Fieber's Chorthippus must be used as the generic term in place of Stenobothrus, as has already been done in several previous papers by the senior author. 370 PROCEEDINGS OF THE ACADEMY OF [Sept., On geyser formation, ]\Iammoth Hot Springs, Aug. 5, 1 9 . Top of bare hill opposite Devil's Kitchen, Mammoth Hot Springs, Aug. 5, 1 9 . Meadows near Norris Pass, Continental Divide, Yellowstone Park, Aug. 5, 1 9 . Near Grand Caiion, Yellowstone Park, Aug. 10, 2 c^. These specimens have the caudal tibiae with pale proximal annuli followed by dull glaucous and with the distal third reddish. This species is distinctly boreal and apparently uncommon in collections, the only definite records being from Fort McLeod, southern Alberta, several localities in Idaho, Fort McKinney,. Wyoming, and Estes Park, Colorado. Individuals from all except the Idaho localities have been examined. The specimens taken were all found among the more boreal forms of vegetation in the Park. Nowhere was it possible to take a large series of specimens, in spite of the fact that when found individuals were easily captured. The insects move about quite hurriedlj^ on the ground, and when disturbed can also use their wings advantage- ously. GOMPHOCERTJS Thunberg. Gomphocerus clavatus Thomas. ^lammoth Hot Springs, Yellowstone Park, Aug. 5,6 9 . Summit of foothill of Gallatin Range, Mammoth Hot Springs, Aug. 5, 1 9 . Manitou, Colorado, 6,800-7,000 feet, Aug. 23, 2 d', 5 9. Pike's Peak, 10,200 feet, Aug. 20, 1 9 ; Dark Canon, 8,920-9,000 feet, Aug. 16, 8 c? , 4 9 ; above Windy Point, 12,300 feet, Aug. 20, 16 d' , 8 9 , 6 nymphs ; below Windy Point, 12,000 feet, Aug. 20, 6 c? ; along timber line, 11,578 feet, Aug. 20, 1 d' ; Mountain View, 9,700 feet, Aug. 20, 2 c?, 5 9 . With this series of nearly seventy specimens, and twenty others in the collection of the Academy, we are unable to find any constant character or characters to separate the three so-called American species of this genus. We have attempted to apply the characters given by Bruner^ to distinguish them, but the Windy Point series exhibits enough variation to prove the futility of attempting to recognize several species. The length of the antennje and the angulation of the fastigium vary to a great extent, and the inflation of the cephalic tibiae of the males apparently is as variable as in the European species of the genus. I found this species by no means abundant about the Mammoth Hot Springs, but took it in both the highest and lowest locations. On ' Biol. Cent.-Amer., Orth., II, pp. 92, 93. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 371 3 the hillside above Manitou, Colorado, I found the species in about the same numbers as at the Hot Springs. From Dark Canon to the highest grassy locations on Pike's Peak, however, the species was common in all favorable locations. The insects were found most common in all open grassy spots, especially above the timber-line where short grasses grow everywhere between the scattered boulders. The males are very active and are constantly stridulating — sik-sik-sik- sik — a sound louder but similar to that produced by Stauroderus curtipennis. The females are slow and clumsy in their movements, appearing quite different from the males. PSOLOESSA Scudder. Psoloessa maculipennis Scudder. Knob Hill, Colorado Springs, Colorado, Aug. 17, 1 9 . This specimen is somewhat smaller than Kansas and New Mexican individuals of the species. The previous Colorado record of this species was simply "Southern Colorado." While sweeping the prairie grass, a single specimen of this species was taken from among the hundreds of more common species. So great was the number of common Orthoptera in this locality that it was impossible to search for the scarcer individuals, but this species was not met with again even though much sweeping was done in the same vicinity. AGENEOTETTIX McNeill. Ageneotettix deorum (Scudder). Livingston, Park Co., Montana, Aug. 4, 1 c?. Salt Lake City, Utah, Aug. 13-14, 9 d", 6 ? . Grand Junction, Colorado, Aug. 15, 1 c? . Antlers, Colorado, Aug. 15, 3 cJ', 1 9 . Knob Hill, Colorado Springs, Colorado, Aug. 17-22, 73 c?, 87 9. Austin Bluffs, Colorado Springs, Aug. 18, 2 9. Manitou, Colorado, Aug. 16, 3 d", 2 9. Garden of the Gods, Colorado, Aug. 17 and 19, 3 c^, 3 9 . Roggen, Weld Co., Colorado, Aug. 24, 1 9 . Aki'on, Washington Co., Colorado, Aug. 24, With this extensive series before me, I am unable to separate Bruner's scudder i and occidentalis. The characters given by him in his key of the species of this genus ^ do not appear to be constant, as specimens from the type locality (Garden of the Gods) cannot be separated from others from the supposed range of occidentalis (Salt Lake City, Grand Junction and Antlers), while a sufficient series from any one locality » Bull 94, Colo. Agr. Exp. Sta., p. 58, 1904. 372 PROCEEDINGS OF THE ACADEMY OF [Sept., contains individuals with the characters supposed to be diagnostic of any of the three forms and numerous others which cannot be placed in one or the other of the three ''species." This was one of the most plentiful species encountered. It was found in many of the arid locaUties, but on the prairie it truly swarmed. Each sweep of the net would take a dozen or more specimens in this location, and so great were its numbers that other scarcer specimens, even when seen, would often easily escape in the myriads of this species. AULOCAfiA Scudder. Aulocara elliotti (Thomas). Gray Cliff, Sweet Grass Co., Montana, Aug. 4, 2 d^, 4 9 . Living- ston, Park Co., Montana, Aug, 4, 1 9 . Electric, Park Co., Montana, Aug. 4, 1 ? . Summit of foothills near Mammoth Hot Springs, Yel- lowstone Park, Aug. 5, 5 c?, 6 ? . Top of Ensign Peak, Salt Lake City, Utah, Aug. 13, 1 9 . Knob Hill, Colorado Springs, Colorado, Aug. 17 and 22, 3 c3^ . Prairie land, Colorado Springs, Aug. 18, 3 ? . Garden of the Gods, Aug. 17, 19 and 23, 11 d", 4 ? . The largest series of this species was taken in a meadow of tall grasses and weeds at the western entrance of the Garden of the Gods. It was also taken on the prairie at Colorado Springs, in dry grass and sage at Gray Cliff and in the other localities in hilly sage-brush locations. Aulocara femoratum Scudder. Gray Cliff, Sweet Grass Co., Montana, Aug. 4, 1 cd. Garden of the Gods, Colorado, Aug. 17, 19 and 23, 3 9 . As stated by Gillette ^^ the females of this species are very similar to the same sex of elliotti, and rather difficult to distinguish. The slenderer caudal femora of femoratum are probably one of the best differential characters. The Montana record is the most northerly for the species. The specimen taken in Montana was captm-ed among the dried grasses in prairie land. Had I had more time in this locality I think more specimens could possibly have been taken. In the Garden of the Gods the species was scarce, the only individuals taken being captured in the grass near the western entrance. Aulocara parallelum Scudder. Hillside at Salt Lake City, Utah, Aug. 13, 2 d", 2 9. Top of Ensign Peak, Salt Lake City, Aug. 13, 1 d^, 1 9 . Antlers, Colorado, Aug. 15, 2 d". 1° Bull. 94, Colo. Agr. Exp. Sta., p. 30, 1904. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 373 The color of the caudal tibiae, which were missing in the types, is similar to that of A. elliotti, with the pale proximal annuli more dis- Fig. 6. — Aulocara parallelum Scudder. Salt Lake City, Utah, female. (X 3.) Lateral view of tinct in the females than in the males. The blackish postocular stripe mentioned by Scudder appears to be a variable character and is absent in some specimens and distinct in others. The Antlers specimens are the first recorded out- side of Salt Lake Valley. This species was scarce in both localities in which it was taken. I found it on dry soil overgrown with sage, and owing to the activity of the individuals not many of even the few seen were taken. The caudal limbs of this species are so loosely attached that they snap off at the least strain and perfect specimens were consequently difficult to seciu'e. (EDIPODIN.E. ARPHIA St&l. Arphia pseudonietana (Thomas). Upper Geyser Basin, Yellowstone Park, Aug. 7, 3 6", 1 9. Salt Lake City, Utah, Aug. 13, 1 6". Hill- side at Salt Lake City, Aug. 14, 1 d^, 1 9 . Prairie Fig. 7. — Auloca- ra parallelum Scudder. Salt Lake City. Utah. Dorsal view of head and pronotum of female. (X 3.) 374 PROCEEDINGS OF THE ACADEMY OF [Sept., land, Colorado Springs, Colorado, Aug. 18, 1 d", 1 ? . Akron, Color- ado, Aug. 24, 1 c^. This series presents the variation in the intensity of the overlying blotches of dark brown usual in the species, and which in specimens of a light base color produces a strongly maculate type. The specimens from Upper Geyser Basin all have the sutural margin of the tegmina distinctly and broadly marked with pinkish ochraceous. In the Upper Geyser Basin this species was to be found in small numbers in the small grassy spots between the geyser formations and the Firehole river. In the other localities the species was scarce. Its flight is very vigorous and accompanied by a loud clattering. Arphia arcta Scudder. Hill at head of Mammoth Hot Springs, Yellowstone Park, Aug. 5, 1 o^. This specimen has the wings orange red. The individual collected was the only representative of the species seen. CHORTOPHAGA Saussure. Chortophaga viridifasciata (De Geer). St. Louis, Missouri, Aug. 27, 1 $ . ENCOPTOLOPHUS Scudder. Encoptolophus coloradensis Bruner. Knob Hill, Colorado Springs, Colorado, Aug. 17, 1 d^. The recognition of this distinct species was quite a surprise. Its closest ally is E. sordidus, from which it is separated by numerous Fig. 8.— Encoptolophus coloradensis Bruner. Knob Hill, Colorado Springs, Colorado. Lateral view of male. (X 3:) characters. The only published record is the original description, the type being from Fort Collins, Colorado. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 375 This specimen was beaten from the rank growth of weeds about a small water-filled depression in the prairie. Owing to the immense number of common species in this place no other specimens of this species was noticed, and much beat- ing in the same locality failed to secm-e additional specimens, CAMNULA stai. Camnula pellucida (Scudder). Waco, Montana, Aug. 4, 3 d^, 2 ? . Emigrant, Montana, Aug. 4, 1 9 , Daileys, ^lontana, Aug. 4, 1 9 . Sphinx, Montana, Aug. 4, 1 9 . Electric, Mon- tana, Aug. 4, 1 9 . ]\Iuir, Bozeman Tunnel, Montana, Aug. 12, 1 c?,2 9 . Sappington, Montana, Aug. 12, 1 d^. Spire Rock, Montana, Aug. 12, 1 cJ*. Mammoth Fig. 9.—Encop- Hot Springs, Yellowstone Park : hill at head of springs, SS^ ^Bru- Aug. 5, 1 9 ; on geyser formation at the upper end of ner. Knob springs, Aug. 5, 1 c?. Fountain, Lower Geyser Basin, J?'^l' Colorado Yellowstone Park, Aug. 6, 1 d^. Upper Geyser Basin, orado'! Dorsal Yellowstone Park, Aug. 7, 1 d". The Thumb, Yel- ^'^^^ of If^d ' andpronotum lowstone Park, Aug. 7, 1 9 . Near Grand Canon, of male, (x 3.) Yellowstone Park, Aug. 10, 1 9 . Yellowstone Lake, Aug. 7, 1 9 . Upper slope of Dark Caiion, Pike's Peak, Colorado, Aug. 16, 4 d", 1 9,1 nymph. The specimens from Waco are very interesting as they are exceed- ingly pale in color, some appearing rather uniform dull straw color with the maculations much reduced and quite feeble; others have the maculations distinct but the lighter shades all the same color as in the other pale individuals. The specimen from geyser formation at Mammoth Hot Springs is coated with the mineral deposit. This species was one of the most plentiful of those found in Montana. At Waco the insects fairly swarmed in the dry prairie grass and sage. In this locality the specimens were extremely pale in coloration; exactly the color of the plants on which they were found. I have never seen greater swarms of insects than those of C. pellucida in the sage at the foot of the hills at Mammoth Hot Springs. The species was everywhere in Montana far more plentiful in the low dry locations, but in Colorado I found it only at a considerable elevation in Dark Canon, Pike's Peak. One specimen was found in the Mammoth Hot Springs completely coated by the white formation. In another spring quite a few specimens of this and other species were found dead, killed probably by the fumes. 376 PROCEEDINGS OF THE ACADEMY OF [Sept., HIPPISCUS Saussure. Hippiscus rugosus (Scudder). St. Louis, Missouri, Aug. 27, 1 ? . A single specimen of this species was seen and taken in a vacant city lot overgrown with weeds. Hippiscus conspicuus Scudder. Colorado Springs, Aug. 17, 1 9 . But one specimen of this species was seen. Hippiscus maculatus Scudder. Mammoth Hot Springs, Yellowstone Park: summit of neighboring foothills. Aug. 5, 1 9 ; summit of foothills of Gallatin Range, Aug. 5, 1 9. A little doubt is attached to the identification of these specimens. This species was scarce, but other specimens could have been taken had not the country been so hilly. Specimens would rise from the sage-brush hillside and fly some distance much more swiftly than one could follow. Hippiscus altivolus Scudder. Near Mountain View, Pike's Peak, Colorado, 10,100 feet, Aug. 20, 1 d^ . This is apparently the first additional record of the species since the Fig. 10.— Hippiscus altivolus Scudder. Pike's Peak, Colorado. Lateral \-iew of male. (X 3.) original description (Mt. Lincoln, Colorado, above timber, 11,000- 13,000 feet above sea-level; Como, Colorado, 9,500 feet). This specimen was found on the cog-wheel railroad track, in a spot where on each side almost the only vegetation was a dense thicket of aspen. The insect's flight was erratic and exceedingly swift. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 377 Hippiscus neglectus (Thomas). Mammoth Hot Springs, Yellowstone Park, Aug. 5, 3 9 ; summit of neighboring foothills, Aug. 5, 2 6^, 4 9 ; summit of foothills of Gallatin Range, Aug. 5, 3 9 ; hill at head of springs, Aug. 5, 1 6^ ; top of bare hill opposite Devil's Kitchen, Aug. 5, 1 d^, 3 9 . Upper Geyser Basin, Yellowstone Park, Aug. 7, 1 9 . The Thumb, Yellowstone Park, Aug. 7, 1 d". Knob Hill, Colorado Springs, 7,000 feet, Aug. 17, 1 9 . This species varies considerably in size and greatly in the intensity of the color pattern and the sutural stripe. Individuals from Camas Prairie, Idaho, have been used for comparison. About the Mammoth Hot Springs this species was the most plentiful of the larger Orthoptera. It preferred the hillsides to other locations, and, owing to its somewhat lighter form than other species of the genus found in that locality, it proved swift of flight and not always easy to capture. The single specimen taken at The Thumb was found in a small grassy space in the woods. Only one specimen was seen on the grassy prairie at Knob Hill, Colorado Springs, LEPRUS Saussure. Leprus interior Bruner. Salt Lake City, Utah, Aug. 13, 1 9 . Hillside at Salt Lake City, Aug. 13 and 14, 4 d", 7 9. These individuals agree well with a pair received from Prof. Bruner labelled Salt Lake City, and as a series are smaller and more conspicu- ously colored than a pair of L. cyaneus from Nephista, Colorado. The latter have the dark bars on the tegmina subobsolete, the light sutural stripes and a transverse median light bar being the only distinct mark- ings.- In the Salt Lake specimens the bars are distinct and rather regular, comparatively little variation being exhibited. Several of the latter specimens have the blue of the caudal tibiae weaker than in the other individuals. This series was only taken after long and careful search. There was but one place on the steepest portion of Ensign Peak where the insect could be found, and although the females were easily captured when found, the males flew vigorously. So steep was the hillside that it was bare of everything but scattered grasses and an occasional sage-bush. When at rest the insects were wonderfully protected by their colora- tion. DISSOSTEIRA Scudder. Dissosteira Carolina (Linnaeus). Billings, Yellowstone Co., Montana, Aug. 4, 1 6^, 1 9. Salt Lake 378 PROCEEDINGS OF THE ACADEMY OF [Sept., City, Utah, Aug. 13 and 14, 2 ? . Grand Junction, Colorado, Aug. 15, 1 d", 1 ?. Colorado Springs, Colorado, Aug. 17, 3 d", 2 9. Knob Hill, Colorado Springs, Aug. 17-22, 6 d", 1 9 . Fort Morgan, Morgan Co., Colorado, Aug. 24, 1 9 . Ever3^where common in dusty locations, Dissosteira longipennis (Thomas). Colorado Springs, Colorado, Aug. 17 and 18, 2 c?, 1 9 . Knob Hill, Colorado Springs, Aug. 17 and 18, 4 d^, 1 9 . Specimens have also been examined from Denver, Pueblo, Fort Morgan and La Junta, Colorado, and Sidne}^ Nebraska. A somewhat scarce species and one of the wariest insects collected . I found the specimens on the prairie, and would sometimes be obliged to follow one several hundred yards before an opportunity was offered to capture it. The flight of this insect is swift and high and often of considerable length. Dissosteira spuroata Saussure. Salt Lake City, Utah, Aug. 13 and 14, 13 c?, 5 9 . This series exhibits a great amount of variation in the intensity of the color pattern, some individuals being very dull, almost uniform, while others have a strikingly contrasted pattern. As previously remarked by Saussure, this species bears a considerable superficial resemblance to Spharagemon venustum. These specimens were all taken in a field of alfalfa stubble. Although in this locality the species was abundant, specimens were not taken with ease owing to their extreme wariness and their sudden zigzag flight. Their coloration was of great protective value, as when at rest it was almost impossible to distinguish them from the dried alfalfa. SPHARAGEMON Scudder. Spharagemon aequale (Say). Colorado Springs, Colorado, Aug. 17 and 18, 3 d^, 3 9 . Knob Hill, Colorado Springs, Aug. 18, 1 d^, 1 9 . Prairie land, Colorado Springs, Aug. 18, 1 d', 4 9 . Salt Lake City, Utah, Aug. 13, 2 d', 3 9 . Hill- side at Salt Lake City, Aug. 14, 1 9 . One specimen onlj- is collared, a female from the prairie at Colorado Springs. This species was moderately abundant on the grassy plain near Colorado Springs and was easily captured, as it was by no means as active as other species of the genus. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 379 Spharagemon bolli Scudder. Knob Hill, Colorado Springs, Colorado, 7,000 feet, Aug. 17, 2 'd. Manitou, Colorado, 6,900-7,000 feet, Aug. 23, 5 d", 3 ? . These specimens are not separable from New Jersej^ specimens of the species. Very few specimens of this species were found on the prairie, but it was quite common on a path leading through pine woods on the hills above Manitou. Spharagemon coUare (Scudder). Billings, Yellowstone Co., Montana, Aug. 4, 1 c^. Colorado Springs, Colorado, Aug. 17 and 18, 2 d^, 4 9 . Knob Hill, Colorado Springs, Aug. 17-22, 14 c^, 37 9 . Roggen, Weld Co., Colorado, Aug. 24, 1 d". The specimen from Billings is assigned here with some doubt. One of the most plentiful of the large species found on the prairie. Spharagemon collate angustipenne Morse. Salt Lake City, Utah, Aug. 13, 2 9 . The two individuals recorded here were taken in the stubble field in which Dissosteira spurcata was abundant. This form appeared to be quite scarce. DEROTMEMA Scudder. Derotmema cupidineum Scudder. \\liitehall, Jefferson Co., Montana, Aug. 12, 1 c?. Grand Junction, Mesa Co. , Colorado, Aug. 15, 4 d" , 3 9 . Antlers, Garfield Co. , Colorado , Aug. 15, 9 d", 3 9. These specimens are slightly larger than the original measurements given by Scudder, but otherwise fit the description quite well. The species appears to be readily separable from D. haydeni by the slightly longer and distinctly slenderer caudal femora. The pectinate char- acter of the metazonal rugosities is not characteristic of haydeni, as each of the above series contains specimens in that respect not differ- ing from haydeni, but the femora show them to be cupidineum. The Grand Junction specimens as a series are more suffused than the Antlers individuals. The previous Colorado records of this species are from "\Aliite river, near the Utah boundary, Montrose, Grand Junction and Palisades. The specimens taken at Grand Junction were captured on the re- cently inundated silt along the bank of the Grand river where numer- ous specimens were seen. Those from Antlers were captured on the dry cactus and sage plain. 380 PROCEEDINGS OF THE ACADEMY OF [Sept., Berotmema haydenii (Thomas). Colorado Springs, Colorado, Aug. 17, 8 cJ^, 7 ?. Prairie land, Colorado Springs, Aug. 18, 3 c?, 4 9,1 immature individual. Knob Hill, Colorado Springs, Aug. 17-22, 8 c^, 8 9 . Garden of the Gods, Colorado, Aug. 17 and 23, 4 cJ^, 3 9. Fort Morgan, Morgan Co., Colorado, Aug. 24, 1 d^, 1 9. Both red and yellow winged forms are represented in this series. One of the common prairie species, which I often found in large numbers about damp depressions in the prairie and on the sandy- borders of several shallow ponds. MESTOBKEGMA Scudder." Mestobregma plattei (Thomas). Austin Bluffs, Colorado Springs, Colorado, Aug. 18, 2 9 . Garden of the Gods, Colorado, Aug. 23, 1 9 . But three specimens of this species were seen, although much time was spent searching for additional specimens. The insects were found around stunted pines and bushes growing on cliff-like outcrop- pings in both localities. One specimen had just emerged. The flight of this species is low but swift. Mestobregma kiowa (Thomas). Mammoth Hot Springs, Yellowstone Park, top of bare hill opposite Devil's Etchen, Aug. 5, 1 9 . Prairie land, Colorado Springs, Colo- rado, Aug. 18, 2 d^, 3 9 . Austin Bluffs, Colorado Springs, Aug. 18, 1 9 . Knob Hill, Colorado Springs, Aug. 17-22, 41 c?, 40 9 . Mani- tou, Colorado, Aug. 16 and 23, 3 d^. Garden of the Gods, Colorado, Aug. 17, 15 d", 11 9 . Roggen, Weld Co., Colorado, Aug. 24, 1 9 . This large series exhibits the usual amount of variation in macula- tions seen in this species, chiefly in the intensity of those on the head and pronotmn and in the line between the discoidal and anal areas of the tegmina. One female from Knob Hill has the anal area of the tegmina sub-roseate. " The authors cannot agree with Prof. Brunei's suppression of the genus Mestobregma {Ent. News, XVI, pp. 259-260) on the grounds that the insect on wliich it was erected was not (Edipoda plattei Thomas, as stated by Scudder, but one later described by Saussure as Psinidia (Trachyrhachis) parda- lina. Granting this to be so, the action of the author, and not his intention, shoiild alone be considered, and as Mestobregma was based on (Edipoda plattei Thomas, regardless of material in hand, it should rest on that specific name. The question of the apphcation of the bird genus Ixoreus Bonaparte is a parallel case, for discussion of which see Sclater, Ibis, 1903, p. 142, the final ruUng on which case by the American Ornithologists' Union (Committee on Nomenclature (vide Auk, XXI, p. 424) being similar to the opinion expressed above. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 381 An extremely abundant species wherever found, but swarming in certain grassland locations. Mestobregma thomasi Caudell. St. Louis, Missom-i, Aug. 27, 3 c? , 1 9 . These specimens are not separable from male individuals from Indiana received from Prof. Blatchley. Found in a field overgrown with tall weeds in the city. METATOR McNeill. Metator pardalinus (Saussure). Waco, Montana, Aug. 4, 1 d^, 3 9 . Knob Hill, Colorado Springs, Colorado, Aug. 17 and 22, 1 c^, 2 9 . Prairie land, Colorado Springs, Aug. 18, 6 d", 2 9. Of the series here mentioned nine possess red or reddish wings, five yellow wings and one (Colorado Springs, prairie, Aug. 18, 9 ) has the wings orange. The evidence of this material would thus support the view held by Caudell,^^ w^ho regards Psinidia maculosa Saussure as merely the yellow winged form of pardalinus, an opinion entertained as possible by Saussure in the original description, his words being, 'Annae prgecedentis varietas?" The Waco record is the most northern for true pardalinus, but if maculosa is identical the range is extended as far north as Fort McLeod, Alberta. The specimens taken at Waco were captured among the sage-brush where the species appeared extremely plentiful. I also found this species plentiful at Colorado Springs, in the prairie wherever there was a damp depression. The insect is not swift of flight and may be easily taken with a net. CONOZOA Saussure. Conozoa sulcifrons (Scudder). Grand Junction, Mesa Co., Colorado, Aug. 15, 11 d*, 12 9 . Several specimens of this species were submitted to Mr. A, N. Caudell for comparison with material in the U. S. National Museum, and his remarks are quite interesting and worthy of record. In part he writes : ' ' They are not Conozoa wallula ; they are either C sulcifrons or hehrensi .... I can scarcely tell these species apart. Behrensi is supposed to have shorter wings and elytra, have the median carina of the prozona more deeply cut and have the wing-bands wider. I have 12 Proc. U. S. Nat. Mus., XXVI, p. 790. 382 PROCEEDINGS OF THE ACADEMY OF [Sept., typical specimens of it from California .... these are quite distinct, but a good series before me grades into the sulcifrons form. ' ' Speci- mens of behrensi from San Diego and Claremont, California, appear readily separable from the Grand Junction individuals, but these are probably typical specimens. This species has a rather wide distribution, having been recorded from Nogales and the Huachuca mountains, southern Arizona, in addition to the type locality and Grand Junction, Colorado. It would thus appear to be an Upper and Lower Sonoran species. I took this species on the bare silt along the banks of the Grand river. In this locality it was plentiful and many specimens could easily have been taken as their flight is short and not vigorous. The coloration of the insects is undoubtedly of great protective value. TRIMEROTROPIS St&l. Trimerotropis bilobata n. sp. T3^pes: c^ and 9 ; Antlers, Mesa Co., Colorado, Aug. 15, 1904. Collected by IMorgan Hebard. [Hebard Collection.] Fig. 11. — Trimerotropis bilobata n. SY>. Antlers, Colorado. Lateral \dew of type- (X 3.) This species is one of a number which might with almost equal propriety be placed in either Conozoa or Trimerotropis, but which we have placed here chiefly because McNeill has considered its allied species as a member of Trimerotropis rather than Conozoa. The new form is apparently closely allied to T. rebellis (Saussure) and differs in the absence of a distinct projecting process on the ventro- caudal angle of the pronotum, in the apparently less distinct lateral 190G.] NATURAL SCIENCES OF PHILADELPHIA. 383 angles of the pronotum, the more distinct Conozoa-Yike tegminal bars and the color of the caudal femora. The great elevation of the prozonal section of the median carina of the pronotum immediately separates it from most of the species. No close relationship exists to Conozoa wallula or any of the recently described species of that genus, the whole build being different from that seen in most of the forms of that genus, although C. acuminata approaches it very closely in this respect. Size small; form compressed, slender and elongate; surface finely punctate. Head with the occiput considerably elevated, domed; vertex but slightly lower than the summit of the eyes; fastigium strongly declivent, rather broad in both sexes, greatest width very slightly more than the length, lateral margins distinct and very dis- tinctly converging cephalad, passing into the carinse of the frontal costa without interruption, median carina of the fastigium precurrent, less distinct than the lateral margins of the same; frontal costa rather regularly but not strongly expanding ventrad, the margins sub- parallel for a short distance around the median ocellus, median sections distinct sulcate, less so dorsad, the marginal carinse reaching the clypeal margin ; antennse slightly less than twice the length of the head and pronotum, slightly depressed; eye subovoid, moderately promi- nent particularly in the male, slightly greater than (male) or equal to (female) the length of the infra- ocular sulcus. Pronotum rather narrow; cephalic margin of the dor- sum very obtuse-angulate, caudal margin very slightly obtuse-angulate; median carinae cristate on the prozona and strongly divided into two lobes by the first sulcus, the incision being V-shaped and quite broad in the male and very narrow in the female, the cephalic lobe is subtrigonal in the male and lower and more longitudinally rounded in the female, second lobe equal to the first lobe in height and rounded dorsad in both sexes; carina of the metazona subcristate in both sexes; prozona about two-thirds the length of the metazona; lateral angles of the disk very distinct, and slightly carinate on the cephalic section of the metazona; lateral lobes slightly deeper than long, the ventro-caudal angle rounded w'lih. a very slight indication of a ventral blunt protuberance. Tegmina elongate, narrow, subequal in width, the costal dilation -Trivicrotropis hilohata n. sp. Wing. (X 2.) 384 PROCEEDINGS OF THE ACADEMY OF [Sept., slight and rounded; apex slightly recurved dorsad, rovuided obliquely truncate; intercalary vein distinct, about intermediate in position except dorsad where it regularly approaches and joins the median vein. Wings long and narrow, the width being contained twice in the length ; costal margin evenly and verj^ gently arcuate in the distal half; apex very narrowly rounded; axillary field arcuate apically. Caudal femora quite slender, gently tapering, the width contained about three and a half times in the length, pregenicular constriction not great, pagina rather flat, the sculpturing regular and quite distinct ; caudal tibiae with nine spines on the external margins. General colors ecru drab overlaid with bistre in the male, the lighter color on the head and pronotum hoary white. Head with the facial and costal carinas maculate with blackish brown, a large maculation beneath the base of each antenna and the ventral section of the eye marked with the same color in the male, the ventro-caudal section of the gense being washed with the same tint and a postocular stripe indicated, while the fastigium is marked with pale cinnamon-rufous and a pair of arcuate diverging bars of the same color are present on occiput as well as a blackish interocular bar on the vertex; antennse rather broadly annulate with the two general colors. Female with the head finely sprinkled with blackish spots with a trace of a light postocular bar. Pronotum of the male with a median bistre area which narrows caudad and the lateral lobes of prozona are strongly washed with bistre, forming a contrast Avith the hoary metazona. Pro- notum of the female marked much as in the male, but the whole darker and without light colors, while the lateral angles of the disk and a distinct oblique bar on the lateral lobes of the prozona are darker than any of the other pronotal markings. Tegmina with two complete trans- verse bars, one median, one at the proximal fourth and an incomplete bar just distad of the median transverse one, which is broader mesad than at the costal and sutural margins, the bar at the basal fourth toward the sutural margin approaching the median bar and also extending proximad in the intercalarj^ region; distal third with small irregular maculations which are chiefly along the veins. Wings with the transverse bar bistre and situated in the middle, ciuite narrow as the width is not quite a third and a little more than a fourth of the width of the wing, caudad the band arches very slightly proximad and only slightly suffuses the caudal margin, the stigma broad and heavy and extending about half way to the base ; distal section hyaline with the longitudinal veins broadly suffused with bistre toward the apex; proximal half primrose yellow. Cephalic and median limbs 1906.] NATURAL SCIENCES OF PHILADELPHIA. 385 ecru drab distinctly annulate with bistre. Caudal femora ecru drab with three distinct very dark bars and a solidly colored genicular region, the bars being paler laterad than dorsad, the pregenicular annulus ochraceous bujff; caudal tibise rather dull citron yellow touched with spots of fuscous and with the genicular sections bistre and the tips of the spines blackish. Measurements. Length of body, 17 mm. 24 mm. Length of pronotum, 3.5 " 4.5 " Length of tegmen 18.2 " 26 " Length of caudal femur. 9.2 " 12.5 " A series of one male and two female paratypic specimens have been examined in addition to the types. The paratypic male has the markings slightly more subdued than in the male type. One of the female specimens is marked as in the male type with strongly con- trasted colors, except that the tegmina have the bars somewhat obsolete and the anal area with the veins more strongly lined with dark. The remaining female specimen is colored much the same as the type, except that the bars of the tegmina are slightl}' less con- spicuous. The female type is slightly larger than the paratypic individuals of the same sex, while the additional male is slightly larger than the type. This species was found in the arid valley of the Grand river near Antlers station, where the only vegetation was a heavy growth of low cactus interspersed with occasional sage. Specimens were by no means uncommon and could have been easily taken in numbers had the cactus not interfered so much with collecting, as when alarmed the individuals would invariably seek refuge in the dense beds of cactus. Orthoptera were not present in great numbers at this locality but the few species represented were quite interesting. The species yEoloplus chenopodii was the most abundant and most striking. Trimerotropis gracilis (Thomas). Antlers, Garfield Co., Colorado, Aug. 15, 3 9 . This very distinct species has been recorded from localities ranging from central Idaho to Salt Lake Valley and Colorado. Thomas has recorded it also from the northern boundary (49th parallel) of Mon- tana and Dakota, and Scudder includes Nebraska in its range. Caudell has definitely recorded this species from Durango, Colorado, and Gillette credits it to nineteen localities in the same State, the most eastern 386 PROCEEDINGS OF THE ACADEMY OF [Sept., of which are Greeley and Pueblo, the majority being west of the divide. This species was found among the cactus and sage in the dry bottom lands of the Grand river valley. The insects were by no means plentiful and were also extremely shy, Trimerotropis cincta (Thomas). Manitou, Colorado, 6,700-7,000 feet, Aug. 23, 11 6", 4 ? . The above series shows considerable variation in the strength of the median carina of the fastigium, some individuals having it strongly marked and in others it is hardly perceptible. The color of the caudal tibiae is quite variable in this species, as already noticed by McNeill, and the series in hand exhibits the following : ten with the distal three- fourths of the caudal tibise greenish yellow, four with the same bluish green, and one red. According to Gillette this species is found in the mountains and foot- hills of Colorado at altitudes ranging between 6,000 and 10,000 feet, and on both slopes of the divide. In his series of ninety-five speci- mens the hind tibise were without exception bluish or yellowish on the distal portion. The species has been recorded from Nebraska, in addition to New Mexico and Colorado. The collection of the Academy contains specimens from Salida, Manitou Park and Fort Collins (Horse- tooth Mountain), Colorado. This species was plentiful on a path among stunted pines on the upper slopes of the foothills back of Manitou. The location would have made it particular^ difficult to capture specimens had not the individuals apparently trusted to their protective coloration. A close approach was necessary to cause them to take wing. Trimerotropis ferruginea McNeill. Daileys, Park Co., Montana, Aug. 4, 1 c?. Sphinx, Park Co., Mon- tana, Aug. 4, 3 c?, 2 9 . Spire Rock, Montana, Aug. 12, 1 cJ' ("on bare mountain") . Upper Geyser Basin, Yellowstone Park, Aug. 7, 4 d' . The specimen from Spire Rock is assigned to this species with con- siderable doubt, as it is larger and has a rather different facies, but does not appear to be any closely allied form. The males from Upper Geyser Basin appear rather different from the Daileys and Sphinx specimens, being a little smaller and more like the Spire Rock individual in their somewhat more robust form, though considerably smaller size. This species is probably little more than a form of bifasciata, and hardly worthy of a name in view of the great amount of individual variation found in most of the species of the genus. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 387 This species was found in rather bare mountainous sections and on the formation about the larger Yellowstone geysers. Trimerotropis monticola Saussure. i\Iammoth Hot Springs, Yellowstone Park, Aug. 5, 1 ? . Summit of foothills. Mammoth Hot Springs, Aug. 5, 1 c? . These specimens agree quite well with Nebraska, Colorado and Wyoming specimens in the collection of the Academy. The specimens examined present considerable variation in the length of the tegmina. This appears to be the first record of the species north of Colorado, except Tie-Siding, Wyoming, from which locality Gillette has recorded it. These specimens were taken in the more hilly portions about the Mammoth Hot Springs, where, on account of the character of the country, they were captured with difficulty. The individuals fly swiftly and for long distances when alarmed. Trimerotropis praeolara McNeill. Grand Junction, Mesa Co., Colorado, Aug. 15, 5 cJ*, 1 9 . This species has the fuscous band of the wing very broad, slightly more than one-third the length of the same, and well continued around on the caudal margin, while the spur is almost absent, the disk very pale greenish yellow, and the apex hyaline except for the infuscation of some of the veins. The fight base color is almost bone white in some specimens, and distinctly washed with rufous in others. This species is known only from Salt Lake, Utah, and Sidney, Ne- braska, in addition to the locality here given. Among the bushes growing in the silt deposit along the banks of the Grand river these specimens were secured. The species was by no means as plentiful as T. montana, which was found in the same locality. Trimerotropis laticincta Saussure. Colorado Springs, Colorado, Aug. 17, 6 o^ , 5 ? . Knob Hifi, Colorado Springs, Aug. 18-22, 7 d", 4 9 . This series exhibits considerable variation in the intensity of the bands of the tegmina, and in the shade of the light basic color. In some the latter is decidedly reddish, in others pale yefiowish, and in the majority dull ochraceous. The caudal tibiae are more orange in some specimens than in others, and the pronotum varies from dull ferruginous through warm browns to dull olive, and in one specimen dull blackish-brown. As far as can be determined from the material in hand Scudder's 388 PROCEEDINGS OF THE ACADEMY OF [Sept., T. latifasciata from Utah is the same as this, and his name would have precedence. However, as no Utah material is available I prefer to use Saussiire's name, provisionally at least, for specimens from eastern Colorado. The evidence given by Cockerell ^^ regarding the difference in color characters of individuals of this species taken in coitu is especially interesting in view of the value which has been attached to the char- acters there mentioned, i.e., bands on the inner face of the caudal femora, and the width and curve of the wing band. Not much lati- tude is needed in these characters to connect forms like laticincta and latifasciata and melanoptera. Gillette does not consider latifasciata and laticincta separable from T. citrina.^* This species has been recorded from Fort Collins, Montrose, Grand Junction, Denver, Baileys and Golden, Colorado, by Caudell. One of the more common of the larger prairie land species. The flight of this insect was less vigorous than is usual in species of the genus. Trimerotropis melanoptera McNeill. Colorado Springs, Colorado, Aug. 18, 1 d" . Knob Hill, Colorado Springs, Aug. 17 and 18, 2 d^, 1 ? . Prairie land, Colorado Springs, Aug. 18, 1 d". Two of these specimens agree well in the color of the tegmina with a topotypic (Silver City) male, while three have the bars more prominent and contrasting with the lighter base color. The metazona of the pronotum is paler than the prozona in these three specimens, and in two of them the contrast is as great as in T. monticola. The black areas of the wings appear to be slightly less extensive than the specimens described by McNeill. The senior author's T. snowi ^^ may prove to be but a form of this with a less extensive black area on the wings. From the material then in hand it appeared well worthy of separation, but the additional material shows variability which weakens its status. This species was scarce even on the prairie land which appeared to be its favorite habitat. In habits it much resembled T. laticincta. Trimerotropis strenua McNeill. Salt Lake City, Utah, Aug. 13 and 14, 4 d", 8 9 . This species is very close to T. californica Bruner, and probably '^ Proc. Davenp. Acad. Sci. IX, p. 38. 1^ Bull. 94, Colo. Agr. Exp Sta., pp. 35-36. 15 Trans. Kansas Acad. Sci., XIX, p. 223. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 389 should be considered nothing more than a race of it. The bars of the tegniina are usually somewhat narrower and the caudal tibiae not quite so deep in color. The majority of the specimens have a more or less distinct ferruginous or ochraceous wash suffusing the tegmina, while several have the lighter sections without the same, and clear bone white in color as in the pair of californica used for comparison. A pair of specimens from Salt Lake Valley, received from Prof. Bruner, and labelled by him T. prceclara, belong to this species. Some of the specimens examined have the tegminal bars more distinct and less broken into spots than others, while the width also varies as much as the intensity. A female individual of this species from Salt Lake City, in the collection of the Academy/^ has the tegminal bars much broken into annuli. In nine of the specimens examined a second median bar on the lateral face of the caudal femora is more or less distinctly indicated, while in six it is absent or represented only by a very faint clouding. This beautiful species was quite common in the arid sage-brush region on the outskirts of Salt Lake City and the series was taken with ease. In spite of the insect's bold coloring it was most inconspicuous when at rest. Trimerotropis montana McNeill. Grand Junction, Mesa Co., Colorado, Aug. 15, 10 6^, 4 9 . This species, if such it might be called, is very closely allied to the preceding, as has been stated by McNeill." As far as the material in hand goes it appears to be only a more subdued form, with the dark tegminal bars broader and pale areas more ochraceous and in a measure lacking the strong contrast seen in strenua. The character of the median carina of the fastigium and the proportions of the pronotum do not differ from the series of strenua, while the lateral face of the caudal femora in the greater majority have a median band more or less faintly indicated in addition to the constant subapical one. This species has been previously recorded from Durango and Delta, Colorado, in addition to Grand Junction, as well as Boise City, Idaho^ the type locality. These insects were taken on the banks of the Grand river, on ground that had been recently inundated. The soil was sandy and clusters of tall bushes grew here and there. The specimens were active and not taken without difficulty. " Trans. Amer. Ent. Soc, XXVII, p. 333. Proc. U. S. Nat. Mus., XXIII, p. 433. 26 390 PROCEEDINGS OF THE ACADEMY OF [Sept., Trimerotropis salina McNeill. Colorado Springs, Colorado, Aug. 17, 1 cJ*. Knob Hill, Colorado Springs, Aug. 17, 3 ? . These specimens are slightly larger than the measurements given by McNeill, but we have placed them here chiefly on account of the poorly contrasted tegminal bars and the broad wing band. The lateral face of the caudal femora presents a median bar of varying intensity in several specimens. The coloration {i.e., blackish bars and suffusion) of the ventral sulcus of the caudal femora appears to be too variable to be given the prominence assigned to it by McNeill. The male individual here studied would be placed in another section of the vinculata group on the femoral coloration, while it unquestionably is the same as the three females which have the coloration as described by McNeill. Instances such as this are by no means confined to this species, but can be found in most any large series of a species of the genus. Trimerotropis fratercula McNeill. Daileys, Park Co., Montana, Aug. 4, 13 d^, 6 9 . Sphinx, Park Co., Montana, Aug. 4, 1 9 , Fountain, Lower Geyser Basin, Yellowstone Park, Aug. 6, 6 d^, 6 9. Upper Geyser Basin, Yellowstone Park, Aug. 7, 3 d^, 4 9 . Mud Volcano, Yellowstone Park, Aug. 10, 1 cJ^. Emerald Spring, Yellowstone Park, Aug. 7, 1 9 . Yellowstone Lake, Yellowstone Park, Aug. 7 and 10, 1 d^ , 1 9 . This series is assigned to this species with some little doubt. Some affinity exists with T. inconspicua Bruner, at least as far as published diagnostic characters are concerned. The coloration of the ventral sulcus of the caudal femora varies in individuals from the same locality and not separable by any other character. Some present a solid blackish sulcus with a distinct light pregenicular annulus, others have a well-marked additional light median bar, and a number of specimens have the black interrupted on the sides of the sulcus but undivided in the middle. The evidence of this series of forty-three specimens of unquestionably but one species is that the coloration of the sulcus of the caudal femora is of little or no diagnostic importance. The coloration of the lateral and internal faces of the caudal femora are also unstable, in the authors' opinion, while the intensity of the tegminal bars and the width of the wing bars are by no means as stable as would be supposed from the importance given them as diagnostic characters. The series studied presents some variation in size, the males from Fountain being somewhat larger than males from Daileys (length including tegmina: Fountain 28.5 mm., Daileys 25), while the Mud 1906.] NATURAL SCIENCES OF PHILADELPHIA. 391 Volcano and Upper Geyser Basin males are slightly smaller than the Daileys specimens (24 mm.). The females appear to be rather uniform in size. Three of the four females from Upper Geyser Basin are suffused with rufescent, while the other female and three males from the same locality are without any suffusion. The females from Fountain have the colors more contrasted than the majority of the Daileys females, while the Emerald Spring female has the contrast just as great. The males from Fountain have the pale tints more grayish and less ochraceous than in the Daileys males. The species was so abundant about Daileys that during the iDrief stop of the train nineteen specimens were easily taken. This was also the most abundant species of Orthoptera found about the geyser for- mations of the Park. The insects are wary and rise on the wing with speed. The coloration of the individuals almost invariably blended exactly with the soil on which they were found. Trimerotropis vinoulata Scudder. Electric, Park Co., Montana, Aug. 4, 1 c?. Muir, Bozeman Tunnel, Montana, Aug. 12, 1 9 . Whitehall, Jefferson Co., ^lontana, Aug. 12, 1 c? , 3 9 . Upper Geyser Basin, Yellowstone Park, Aug. 7, 1 9 . Salt Lake City, Utah, Aug. 14, 1 9 . Antlers, Garfield Co., Colorado, Aug. 15, 1 d^. Debeque, Mesa Co., Colorado, Aug. 15, 1 d^. Austin Bluffs, Colorado Springs, Colorado, Aug. 18, 2 9 . Manitou, Colorado, 6,700-7,000 feet, Aug. 23, 4 d", 4 9 . Garden of the Gods, Colorado, Aug. 19, 2 9 . This series represents a great and interesting amount of variation even for this plastic species, ^'ariation in size is considerable, the largest males being from Electric, Debeque and Manitou, the smallest from Antlers ; the largest female from Salt Lake City, the smallest from Austin Bluffs. These differences in size appear to be individual, as the smallest male (Antlers) is from the same general region as the largest (Debeque). The specimens from Muir and Wliitehall, ]\Ion- tana, are large, of rather uniform size, and but little smaller than the large Salt Lake City individual. In coloration we find such variation that a detailed account may be of interest. The Electric specimen has the bands solid and the lighter colors hoary and sharply contrasted; the Muir and Whitehall repre- sentatives are quite dull, the Muir specimen almost blackish, while the bands are only moderately distinct, in one of the Whitehall individuals broken up into annuli. The Muir and Whitehall type has the pro- notum with little contrast, and the caudal femora less variegated than in the Electric individual. The Yellowstone specimen is overcast with 392 PROCEEDINGS OF THE ACADEMY OF [Sept., dull hoary, the bands very weak, narrow and broken into annuli, the whole insect being much duller and more inconspicuous than the usual type. The Salt Lake City female has the bands of about average width, but weak and rather broken. Antlers and Debeque specimens are of the type which might be called normal or typical, with the bands distinct, broad and solid, and the lateral lobes of the pronotum moder- ately variegated. Austin Bluffs specimens are overcast with reddish ochraceous, one with the bands more distinct and solid than the other; Manitou individuals overcast with dull red, the bands variable in intensity and solidity, but not in width, one specimen having the bands quite blackish. The Garden of the Gods specimens are particularly interesting, being entirely washed with brick red, the tegminal bands distinct but narrower than usual. The Garden of the Gods specimens are very interesting as the suffus- ing tint is the predominating color of that region, and in line with this one would expect a uniform very pale type from strongly alkaline regions, which does not hold true of the large amount of material of this species seen. Series from San Diego, California, one of the original localities; Alamogorclo, New Mexico, and Quartzsite, Arizona, are extremely variable among themselves, and each exhibit extremes as different, or nearly so, as those in the material here treated. How- ever, no specimens quite as reddish as the Garden of the Gods individual have been seen from other localities. The band of the wing is quite narrow and weak in some of the Manitou specimens,^* and heavy in Whitehall and Muir individuals. The species was generally distributed and equally abundant in the localities in which it was collected. The specimens taken about the Garden of the Gods showed remarkable protective coloration, as all were overcast with brick red which blended exactly with the character- istic brick red formation of the Garden. This was the only species found on these outer oppings. Trimerotropis coeruleipennis Bruner. Salt Lake City, Utah, Aug. 13, 2 d', 1 9. Top of Ensign Peak, Salt Lake City, Utah, Aug. 13, 1 d", 2 9 . These specimens of this beautiful species have the tegminal bars uniform in pattern and varying but slightly in intensity, except in the female from Salt I^ake City in which they are quite dark. The colora- tion of the ventral sulcus of the caudal femora varies from black with two light bars in the distal half to but one bar in the region, while in '* These may possibly represent one of the other species of the vinculata group - 1906.] NATURAL SCIENCES OF PHILADELPHIA. 393 several specimens the evanescent bar (proximal) is present but incom- plete. The coloration of the caudal tibiae varies from the usual light tint to a dark glaucous blue. The size of the specimens is quite uniform. This species has been recorded from Washington (Yakima), Oregon, (Umatilla), Idaho, Montana, Wyoming, Utah and Los Angeles, California. This species was found in the same locations as T. cyaneipennis, Avhich species it much resembled in habits. Trimerotropis cyaneipennis Bruner. Salt Lake City, Utah, Aug. 13, 4 d", 4 9 . Foothills, Salt Lake City, Utah, Aug. 14, 3 ^^,2 ?. This series is quite uniform in size, while variation in coloration is chiefly due to the intensity of the broken tegminal bars, though one specimen from Salt Lake City is uniforml}^ very pale with most of the markings almost obsolete. This species has been recorded from the Salt Ijake Valley region, and from Grand Canon, Flagstaff and Prescott, Ai-izona. The two males recorded from Prescott are very considerably darker than any of the Salt Lake A' alley specimens examined. This species was somewhat common but inhabited the most inacces- sible portions of Ensign Peak, and owing to the rapidity of its flight specimens were taken Avith no little trouble. CIRCOTETTIX Scudder. Circotettix carlinianus (Thomas). ]\Iammoth Hot Springs, Yellowstone Park, Aug. 5, 1 9 . Circotettix undulatus (Thomas). Manitou, Colorado, 6,400 and 6,700 feet, Aug. 23, 1 c^, 1 9 . pike's Peak, Mountain View, 9,700 feet, Aug. 20, 1 d', 1 9 . Cripple Creek, Teller Co., Colorado, Aug. 19, 4 c^. These specimens have been compared with a series of twenty speci- mens from localities in the mountain regions of New Mexico, and one from Utah. Manitou individuals are more reddish than those from the other localities. This species was nowhere abundant, but several specimens were found on the tailings about Cripple Ci'eek mines. Circotettix rabula^n. sp. Types : (d and 9 ; summit of hill at head of springs. Mammoth Hot *' I.e., a jangling fellow. 394 PROCEEDINGS OF THE ACADEMY OF [Sept., Springs, Yellowstone National Park, Wyoming; altitude 7,000 feet, Aug. 5, 1904. (M. Hebard.) [Hebard Collection.] Allied to C. undulatus (Thomas) and shastanus Bruner, but differing from the former in the more robust form, the broader head, broader heavier and less strongly sulcate frontal costa, larger eyes, slightly less angulate caudal margin of the pronotum and in details of the coloration, as the lighter and usually more variegated tegmina and pronotum and the less distinct fuscous band and spur of the wing. From shastanus it appears to differ in the heavier form and the structure of the first subjacent radial vein. The form of the wing of this species is essentially that of C. undulatus, and its closest relationship appears to be with that species. As shastanus is not available for study, comparison can only be made with descriptions, but the results of such comparisons seem sufficient to separate the specimens in hand, especially when we consider the peculiar structure of the first subjacent radial in the older species, while in the new one here treated the structure of that region does not differ materially from the type seen in undulatus. In coloration a number of specimens of this species bear a striking superficial resemblance to C. carlinianus. Size medium; form quite robust; surface finely rugulose. Head with the occiput gently but distinctly arched, the interocular region no more elevated than the summits of the eyes and in width but little less than the length of the eye; fastigium slightly declivent, forming an obtuse angle with the face, nearly as wide as long, shallow, with the median carina rather faint and the marginal carinse but little more distinct; frontal costa broad, expanded between the antennae and moderately constricted immediately ventrad of the ocellus, margins carinate and nearly (c^) or not (?) reaching the clypeal margin, broadly sulcate ventrad of the ocellus particularly in the male, strongly punctate dorsad with a depressed area at the junction with the fasti- gium; eyes moderately prominent, rather small, very broad ovoid in shape; antennae slightly longer than the head and pronotum together, somewhat depressed in the proximal section. Pronotum deplanate dorsad, broad, distinctly expanded caudad; cephalic margin very broadly and obtusely angulate; caudal margin rectangulate in the male, obtuse-angulate in the female, the angle finely formed and little rounded in either sex; lateral angles rounded and hardly marked on the prozona, distinct and heavy shoulders present on the metazona; median carina fine, little elevated, depressed at the principal sulcus, prozona about half the length of the metazona; lateral lobes slightly 1906. J NATURAL SCIENCES OF PHILADELPHIA. 395 deeper than long, considerably impressed mesad, ventro-caudal angle rounded. Tegmina rather broad, the greatest width contained slightly more than four times in the length ; costal lobe very slight and low, sutiu-al margin straight, apex obliquely truncate toward the sutural margin, the distal third of the costal margin arcuate, making the immediate apex rounded rectangulate; intercalary vein distinct, intermediate in position proximad, close to the median vein distad. Wings moderately falcate, broad, the breadth being contained about once and a half in the length; humeral field with the apex rounded acute-angulate ; axillary field with the margin very slightly arcuate and not lobate, being no more prominent than the next area in the radiate field ; posterior axillary vein with its ramus paralleling closely the anterior axillary vein for a considerable distance, but not coalescing with it; radiate veins moderately thickened. Caudal femora rather slender, the ventral carina distinctly but slightly produced arcuate; caudal tibiae with nine spines on the lateral margins. General color ecru drab, sprinkled, spotted and blotched with drab; eyes burnt umber; antennse of the paler color annulate with bistre; dorsum of the pronotum and the lateral lobes marked with the darker color. Tegmina with traces of three bands of quadrate and ocellate spots, these spots being almost wholly formed by the infuscation veins and cross veins, the bands being proximad, proximo-mesad and disto-mesad, while the distal third is supplied with scattered quadrate ocelli of the darker color. Wings with the transverse band very weak and almost severed, being chiefly an infuscation of veins and without definite limits, extending caudo-mesad and with a broad, moderately distinct stigma which extends half way to the base, the color of the band and stigma being drab ; distal portion hyaline except for inf uscate veins, proximal portion citron yellow. Cephalic and median limbs of the general colors annulate ; caudal femora with a pale pregenicular annulus, and with traces of two transverse bars which are more distinct dorsad ; caudal tibiae pale, infuscate proximad and distad, the spines touched apically with blackish. Measurements. Length of body, 27.5 mm. 29.7 mm. Length of pronotum, 6.4 " 6 " Length of tegmen, 29.5 " 28.7 " Length of caudal femur, 13 " 13.2 " A series of ten specimens in the Hebard Collection have been ex- amined in addition to the types: five, two males and three females, 396 PROCEEDINGS OF THE ACADEMY OF [Sept, from the type locality and date but taken on geyser formation at the upper end of the springs, one male from the immediate vicinity of the Hot Springs, August 5, another male from the summit of the first foot- hill of the Gallatin Range at the type locality, August 5, and one Fig. 13. — Circotettix rabula n. sp. Mammoth Hot Springs, Yellowstone Park. Lateral view of female type. (X 3.) female from Spire Rock, Montana, August 12. These specimens vary slightly in the strength of the tegminal bars, which are partially solid in some specimens, but all other characters appear to agree well with the types. Two males and two females received from Prof. Gillette and labelled C. verruculatus are ref- erable to this new form. They are from Dutch George's and Virginia Dale, Larimer Co., Colorado. This species was by no means plentiful about the Hot Springs and much time was occupied in collecting the ten specimens. The insects are most swift and shy, so that great difficulty was experienced in Fig. 14. — Circotettix rabula n. sp. (X2.) Wing. 1906.] NATURAL SCIENCES OP PHILADELPHIA. 397 approaching them close enough to strike with the net. They may be heard high in the air making for several minutes at a time a noise like the whirring of machinery. The sound is continuous while the insects rise and fall in a most peculiar erratic flight. I have seen specimens of this species fly for over a quarter of a mile making continually this peculiar sound. Of all the Orthoptera met with this was without doubt the most vigorous and elusive species. Circotettix suffusus (Scudder). Summit of foothill of Gallatin Range, ^Mammoth Hot Springs, Yellow- stone Park, Aug. 5, 1 9 . Hill at head of springs, ^Mammoth Hot Springs, Aug. 5, 1 cJ'. Upper Geyser Basin, Yellowstone Park, Aug. 5, 6 6^,2 9, 1 nymph. These specimens are rather uniform in color for representatives of this genus, one being more clear grayish than the others, and several with a faint suggestion of ferruginous in their general tone. About the Hot Springs this species was found in the open sage- brush of the hillsides, while at the Upper Geyser basin it was found in the small grassy openings in the dense pine woods. Circotettix verruculatus (Kirby).=f Muir, Bozeman Tunnel, Gallatin-Park Co., Montana, Aug. 12, 1 &, 1 9. These specimens do not differ in size, the female, however, being slightly more robust. The frontal costa in the male is quite broad and similar to the female, thus differing from a large number of Eastern specimens examined. A series of thirty-one specimens, covering localities extending from the mountain regions of Pennsylvania and Maine west to the northern peninsula of Michigan (Pequaming), shows that while considerable variation in color and size is present in a series, it cannot well be correlated with locality. The two Montana specimens are among the largest seen. These two specimens were taken on a cinder pile beside the railroad. Their crackling was very noticeable whenever they took flight. HADROTETTIX Scudder. Hadrotettix trifasoiatus (Say). Sappington, Gallatin Co., Montana, Aug. 12, 1 d^ . Colorado Springs, ^° The type of this species was taken in "Latitude 57° N., " and in all probabihty was collected by Drummond, who traversed the Athabasca and Peace river region in tliis latitude in western Athabasca and northeastern British Columbia. If collected by Richardson, the other naturalist of the Franklin expedition, it was no doubt taken either in the vicinity of York Factory, Keewatin or on the Athabasca river south of Lake Athabasca. 398 PROCEEDINGS OF THE ACADEMY OF [Sept., Colorado, Aug. 17, 2 6^. Knob Hill, Colorado Springs, Aug. 17 and 18, 4 c?, 6 9 . Prairie land, Colorado Springs, Aug. 18, 7 d", 10 9 . These specimens exhibit considerable variation in the intensity and width of the bands of the tegmina in the female, some having them very weak and broken up into spots, while a few have them as distinct as in the males. The males are quite uniform in the coloration of the tegmina, both in width of bars and intensity. This species was quite common on the prairie. The flight is slow and specimens may be easily captured even without a net. BRACHYSTOLA Scudder. Brachystola magna (Girard). Colorado Springs, Colorado, Aug. 18, 9 d^, 4 9,1 nymph. Knob Hill, Colorado Springs, Aug. 17, 5 d^, 1 nymph. Prairie land, Colorado Springs, Aug. 18 and 19, 2 d', 6 9,1 nymph. This series of thirty specimens clearly shows that the species has two color phases and that the green phase should not be confused with Charpentier's B. virescens, a Mexican species of a quite distinct char- acter, as an examination of his plate will show. Of the series examined eight are of the brown phase and twenty-one of the green, while one is too discolored to determine its true color. The green specimens vary considerably in the shade of the green, but the males appear to have the richer coloration. There is considerable variation in the size of the males. The following color description is made from a female specimen in the green phase from Colorado Springs received alive in Philadelphia and carefully stuffed after death, to-day presenting the same coloration as in life. Predominating color oil green. Head dark oil green above becoming paler ventrad on the face and gense, the caudal margin of the gense, the mandibles and broad band extending from cephalad of the eye ventrad to the clypeal sulcus and the lateral portions of the clypeus pale flesh color; eyes clay color sprinkled with bistre; antennae dusky, margined laterad in the proximal section with pale greenish. Pro- notum with the lateral and median carinse marked with very dark French green, on the metazona extending slightly ventrad and some- what suffusing the lateral lobes; lateral carinse bordered mesad by a line of pinkish vinaceous; recurved caudal and lateral lobe margins chiefly flesh color. Tegmina tawny ochraceous with fair-sized bistre spots well distributed over them. Abdomen suffused with bistre dorsad ; a pair of rather broad distinct longitudinal lines, one on each side of the median carina, and a transverse series of spots on the caudal 1906.] NATURAL SCIENCES OF PHILADELPHIA. 399 margin of each segment, vinaceoiis. Venter bone white. Cephalic and median hmbs bone white washed with greenish. Caudal femora oil green dorsad, whitish ventrad, the ventro-lateral sulcus with quadrate touches of vinaceous-ruf ous, while the internal face is suffused with the same color, a median line of French green is present on the lateral face and one of brick red on the internal face, genicular region bluish green dorsad, with a proximal transverse blackish bar ventrad, the arches chestnut; caudal tibise lavender laterad, vinaceous internally, the genicular region bluish, the spines bone white tipped with black. This species was by no means abundant, but in the prairie region specimens would every now and then be found. The insects were so awkward that they seemed to have almost no control over their move- ments. Their coloration, however, undoubtedly proves a great pro- tection to them. LOCUSTIN-E. PAROPOMALA Scudder. Paropomala wyomingensis (Thomas;. 1871. Mesops wyomingensis Thomas, Proc. Acad. Nat. Sci. Phila., 1871, p. 152. [.\long the east base of the Black Hills, in the ^^cimty of Cotton- wood Creek, Fall River Co., South Dakota, or Converse Co., Wyoming.] 1890. Mesops cyUndncus Bruner, Proc. U. S. Nat. Mus., XII, p. 48. [Valen- tine, Nebraska, along the north side of Keya Paha creek, also on the bluffs south of Chadron, Nebraska.] Roggen, Weld Co., Colorado, Aug. 24, 6 cJ', 1 $ . As far as can be determined from the material" in hand, including specimens determined by Bruner as both wyomingensis and cylindricus, and the literature bearing on the subject, no reason is apparent for separating the two above-mentioned forms. Bruner's cylindricus was based on the brown phase, while in addition to the coloration larger size was given as a differential character. This latter would appear to be purely an individual matter, as green specimens have been examined nearly as large as Bruner's measurements of that sex in cylindricus. All the Roggen males are in the brown phase, and the single female in the green. Material has also been examined from Valentine and Haigler, Nebraska, and Denver and Greeley, Colorado. These specimens were beaten from some very tall weeds near the railroad track, while the train was making a brief stop. SCHISTOCERCA Stll. ScMstocerca lineata Scudder. Brush, Morgan Co., Colorado, Aug. 24, 1904, I <^ . 400 PROCEEDINGS OF THE ACADEMY OP [Sept., This individual is much smaller than the males measm'ed by Scudder. Gillette ^^ has called attention to the variation in size in this species. Beaten from thicket of weeds. HESPEROTETTIX Scudder. Hesperotettix viridis (Thomas). Knob Hill, Colorado Springs, Colorado, Aug. IS, 1 ? . Prairie land, Colorado Springs, Aug. 18, 1 d^, 2 9. Hill near cog-road station, Manitou, Colorado, Aug. 16, 2 9 . These specimens vary considerably in size, the Manitou females being larger than any of the others, and similar to two from Arizona and New Mexico, while seven specimens from five other Colorado localities are in size about equal to the Colorado Springs individuals. The coloration varies very little, and then only in the intensity of the blackish head and pronotal markings. This species was found distributed over a considerable region, but was in no place common. Hesperotettix festivus Scudder. Salt Lake City, Utah, Aug. 13 and 14, 10 c?, 11 9 . Hillside at Salt Lake City, Aug. 14, 1 c?, 1 9. Top of Ensign Peak, Salt Lake City, Aug. 13, 2 d", 2 9. This series as a whole has shorter wings than nine individuals from southern Aiizona, and in consequence of this the specimens appear somewhat more robust. The variation in size in the females is con- siderable, the larger type greatly predominating, while the males vary appreciably. The coloration tends in some specimens toward brownish and in a few is distinctly brown. This is also true of a male Salt Lake City specimen -^ which has been in the Academy collection quite a time. This species was by far the most plentiful of the genus Hesperotettix which I found, and individuals were quite coimnon in the sage growing from the foot to the top of Ensign Peak. Individuals of the species were very active, but relied chiefly for protection on their jumping ability, which was very considerable. Hesperotettix gillettei Bruner.^^ Newcastle, Garfield Co., Colorado, Aug. 15, 1 9 . This recently described species is only known from the Grand river drainage. " Bull. 94, Colo. Agric. Exp. Sta., p. 38. " July 24, 1898. 23 Hesperotettix Gillettei Bruner, Bull. 94, Colo. Agr. Exp. Sta., p. 61, 1904. [Rifle, Glen wood Springs, Delta and Grand Junction^ Colorado.] 1906.] NATURAL SCIENCES OF PHILADELPHIA. 401 During a brief stop of the train at Newcastle, I succeeded in captur- ing a single specimen of this species in the dust beside the railroad track. Hesperotettix speciosus (Scudder). Akron, Washington Co., Colorado, Aug. 24, 1 c^. The median carina of the pronotum of this specimen is deep maroon piu-ple, which depth of coloration is almost equalled in a male from Rocky Ford, Colorado. The specimen was beaten from tall weeds. ^OLOPLirS Scudder. .ffioloplus regalis (Dodge). Fort Morgan, Morgan Co., Colorado, Aug. 24, 1 d^. A single specimen was beaten from low bushy weeds growing densely about a water tank. Vigorous beating for several minutes failed to produce other specimens. .ffiloloplus chenopodii (Bruner). Antlers, Garfield Co., Colorado, Aug. 15, 11 d^, 14 9 . This series varies considerably in the intensity of the coloration, some individuals being more ashy than others. The specimens were easily taken in spite of the cactus, in which they invariably sought refuge, for their movements were slow and they could jump but a short distance. The species was abundant. BRADYNOTES Scudder. Bradynotes obesa (Thomas). Summit of first foothill behind hotel. Mammoth Hot Springs, Yellow- stone Park, Aug. 5, 1 ? . Top of bare hill opposite DeviUs Kitchen, Mammoth Hot Springs, Aug. 5, 1 c?, 2 ? . Although much time was spent searching carefully for specimens of this species, two days' collecting resulted in but four specimens being found. All were taken on the gravelly tops of the foothills where vegetation was almost absent. Their color blended perfectly with the soil and they appeared to rety on this as a means of protection. PODISMA Latreille. Podisma dodgei (Thomas). Pike's Peak, Colorado, at elevations of 9,700 (Mountain View) and 10,200 feet, Aug. 20, 2 d", 1 ? . These specimens and the specimen of Hippiscus altivolus were the only Orthoptera seen in the rather barren tract of stunted aspen 402 PROCEEDINGS OF THE ACADEMY OF [Sept., above the railroad pumping station. This species seemed to be strictly terrestrial in habitat. Podisma oreas-^n. sp. Types : cJ^ and 9 ; summit of foothill of Gallatin Range, Mammoth Hot Springs, Yellowstone National Park, Wyoming. Altitude, 7,000 feet. Aug. 5, 1904. [Coll. Morgan Hebard.] Closely allied to P. ascensor Scudder from American Fork Canon, Utah, agreeing almost perfectly in the form of the abdominal append- ages and other characters, but differing in the vertex being somewhat elevated, the interspace between the eyes in the male over twice as broad as the first antennal joint, the frontal costa more sulcate, the size slightly larger, the antennse reddish and the caudal tibiae differently colored. Size medium (for the genus). Head with the occiput rounded and distinctly elevated above the pronotum in the c?, hardly elevated in the ? ; fastigium moderately declivent, very shallowly and broadly excavated, interspace between the eyes in both sexes over twice as wide as the proximal joint of the antennae; frontal costa broad, hardly (c?) or very slightly (?) narrower than the interspace between the eyes, slightly depressed around the ocellus, not sulcate, biseriate punctate dorsad ; antennae rather short, defective in the c? , distinctly but not considerably longer than the pronotum in the 9 ; eyes hardly prominent, truncate cephalad, distinctly longer than the infraocular sulcus. Pronotum very slightly inflated, very slightly tectate, meta- zona finely punctate; cephalic margin subtruncate, caudal margin very slightly arcuate, no distinct lateral carinae but shoulder angles moder- ately prominent caudad; prozona slightly longer than the metazona in both sexes, quadrate in the c?, slightly transverse in the 9 , principal transverse sulcus very well marked, median carina more distinct on the metazona than cephalad, but not markedly elevated anywhere in its length; lateral lobes distinctly longer than broad. Tegmina slightly less than the pronotum in length, sublanceolate, twice as long as broad in the c^ , half again as long as broad in the ? , apex narrowly rounded, tegmina well separated dorsad. Abdomen with the apex but little elevated and not at all recurved; furcula about a third the length of the supra-anal plate and distinctly longer than the segment from which they spring, simple, parallel; supra-anal plate produced trigonal, considerably inflated mesad, median sulcation and bordering ridges distinct on the proximal half, the sulcation obsolete distad; ^* 'Opeiai, in allusion to its habitat. 1906; NATURAL SCIENCES OF PHILADELPHIA. 403 cerci slender, strongly tapering in the proximal half, very slightly tapering in the distal half, apex blunt and very slightly excavated and falling considerably short of the apex of the supra-anal [plate; sub- Fic i5.—Podisma oreas n. sp. Mammoth Hot Springs, Yellowstone Park. "^ Lateral ^^ew of male type. (X 4.) genital plate with the apical margin slightly elevated, the whole plate evenly tapering to this section, which is narrowly subtruncate and with a slight depressed area immediately caudad of the margin. Prosternal spine erect, slightly retrorse, thick and rather blunt. Inter- space between the mesosternal lobes distinctly broader than long in the c^ . over half again as broad as long in the 9 ; metasternal lobes sub-attingent in the c?, sep- arated by a subquadrate interspace in the 9 . Median and cephalic limbs no more inflated in the d than in the 9 . Caudal limbs moderately robust, femora reach- ing to the apex of the abdomen; caudal tibiffi with eleven spines on each margin. General color bistre in the 6", burnt umber in the 9 . Postocular streak rather faint and weak, seal brown in color; antennse carmine, rather dull distad. Tegmina in the c^ darker ventrad than dorsad, unicolorous in the 9 . Stripes on the pleura distinct. Caudal femora with indications of two dark bars, chevron-shaped on the faces, the light color being an ecru drab, genicular arches and proximal section of the lobes on the internal face blackish, ventral face canary yellow, duller in the male than m the female; caudal tibitB very dull olive green in the 6" and ochraceous m the 9 , except the dorsal face which is gamboge yellow in both sexes; spines black. Fig. 1 6. — Po- dlsma oreas n.sp. Dor- sal view of apex of maleabdo-' men. (X4.j 404 PROCEEDINGS OF THE ACADEMY OF [Sept., Measurements. Length of body, 20 mm. 23 mm. Length of pronotum, 4.8 '' 5.5 " Length of tegmen, A " 5 " Length of caudal femur, 10.5 " 12 " The type specimens are unique. These two specimens were taken on the pebbly summit of the foot- hills back of the Mammoth Hot Springs Hotel. The ground in this region was quite bare, but few plants being noticed. Podisma oregonensis (Thomas). The following localities are represented, all in Yellowstone Park: Mammoth Hot Springs, Aug. 5, 4 cJ^, 1 ? . Summit of first foothill of Gallatin range, Mammoth Hot Springs, Aug. 5, 4 d^, 8 9 . Con- tinental Divide near 8,300 feet elevation (alpine meadow), Aug. 7, 1 d^. ApoUinaris Spring, Aug. 6, 1 9 . Yellowstone Lake, Aug. 7, 8 d", 7 ? . Near Grand Caiion, Aug. 10, 10 d", 4 ? . The specimens from the higher altitudes, as the Continental Divide and Grand Canon representatives, average smaller than the individuals from lower altitudes. One male from Mammoth Hot Springs is partially covered with geyser deposit and is larger than any repre- sentative of the same sex (d*) seen. In coloration some specimens have the paler markings more brilliant than in others, and some few individuals from Grand Canon are suffused with smoky brown. This species appeared to be widely distributed and was found quite plentiful in the short grass of all the elevated open lands in the Park. The species was strictly terrestrial, and individuals were invariably active and possessed of considerable saltatorial power. MELANOPLUS Stil. Melanoplus lakinus (Scudder). Knob Hill, Colorado Springs, Colorado, Aug. 17 to 22, 6 d^, 4 9 . , Prairie land, Colorado Springs, Aug. 18, 1 d^, 1 9. Fort Morgan, Colorado, Aug. 24, 11 d*, 8 9 . Akron, Colorado, Aug. 24, 2 d". Several of the above specimens belong to the richly colored green phase described by Gillette. In all the localities in which I collected on the Colorado prairie where the vegetation was high and rank, I found this beautiful species in moderate numbers. The individuals were very active, but could be easily beaten from the weeds in which they had taken refuge. The green form was noticeable for its brilliancy, Melanoplus occidentalis (Thomas). Emigrant, Montana, Aug. 4, 1 d^, 2 ? . Knob Hill, Colorado 1906.] NATURAL SCIENCES OF PHILADELPHIA. 405 Springs, Colorado. Aug. 17 to 22, 7 d", 11 ? . Austin Bluffs, Colorado Springs, Aug. 18, 1 9 . Prairie land, Colorado Springs, Aug. 18, 4 cJ^, 5 ?. Garden of the Gods, Manitou, Colorado, Aug. 17, 2 cJ", 8 9. Pike's Peak, Colorado, Aug. 20, 1 d at Mountain View, 9,700 feet, 1 9 at 10,100 feet, 1 9 at Middle Hudsonian, 10.500 feet. The specimens here recorded from Emigrant, Montana, are, as far as known, the most northern definite record except one from Bismarck, North Dakota. The species has been recorded by Gillette from locah- ties up to an altitude of 8,000 feet, and in view of the Pike's Peak records here given it will be seen to extend to an elevation of at least 10,500 feet, and to be truly an alpine species. In size these Pike's Peak individuals do not appear appreciably smaller than specimens from the vicinity of Colorado Springs, and are considerably larger than others from Durango, Colorado. From the material collected this species appears to frequent a variety of habitats, from a distinctly arid region at Emigrant to a grassy gully in the Garden of the Gods. On the whole this form frequented the more sterile sections of the regions in which it was collected. Melanoplus flabellifer Scudder. Knob Hill, Colorado Springs, Colorado, Aug. 18, 2 o' . Garden of the Gods, Manitou, Colorado, Aug. 17, 4 c^. These specimens are identical with Scudder's flabellifer in the form of the cerci, but the validity of the species is, in the authors' opinion, questionable. As Gillette has shown ,^5 a large series of specimens will contain forms referable to true occidentalis, flabellifer and cuneatus, and have a number of intermediates between these types as well. A number of the specimens referred in this paper to occidentalis are not typical of it, but were placed there as it is the older species and has priority over flabellifer in case the latter is synonymized. This action appears to us to be inevitable, but the material at hand is not extensive enough to take such action. No specimens of the cuneatus type are in the series. The specimens of this species were taken in the same localities as M. occidentalis. Melanoplus bowditchi Scudder. Grand Junction, Mesa Co., Colorado, Aug. 15, 3 d'. Newcastle Garfield Co., Colorado, Aug. 16, 2 d" . These specimens are quite peculiar in appearance and not at all =>= Bull 94, Colo. Agr. Exp. Sta., pp. 53-54. 27 406 PROCEEDINGS OF THE ACADEMY OF [Sept., similar in coloration to individuals from eastern Colorado and Nebraska. The coloration is decidedly brown, sometimes gray brownish, and little yellow or clear gray is apparent. The head and pronotum are sprinkled with fine umber dots and the tegmina have distinct quadrate macula- tions. The length of the body ranges from 17.5 to 19.5 millimeters. All of the specimens of this species were found in the valley of the Grand river. Those taken at Grand Junction were found on silt near the river, where very little vegetation was to be seen. The insect's coloration was extremely protective and the species relied almost entirely on its wings for locomotory power. Melanoplus flavidus Scudder. Knob Hill, Colorado Springs, Colorado, Aug. 17-22, 21 d^, 10 9 . Prairie land, Colorado Springs, Aug. 18, 4 9 . Garden of the Gods, Manitou, Colorado, Aug. 17 and 19, 1 cJ*, 1 9 . This series shows considerable variation in size and color, the latter chiefly in varying intensity of the pattern. This species was found in the same situations as M. occidentalis, but in much greater numbers. Melanoplus sanguineus Bruner. Garden of the Gods, Colorado, Aug. 17, 1 c?, 1 9 . This striking and beautiful species was recently described from Lamar and Las Animas, in the prairie section of the Arkansas valley, Colorado. This record shows it ascends well into the foothills. The coloration of the caudal femora and tibiae is very striking and serves to readily identify the species. The female specimen in hand has the head, caudal section of the disk of the pronotum and the dorsal face of the caudal femora somewhat suffused with greenish. The specimens were taken in the high grass of a meadow at the western entrance to the Garden of the Gods. Melanoplus bruneri Scudder. Old Faithful Geyser, Upper Geyser Basin, Yellowstone Park, Aug. 7,lc^. This single specimen was taken in a small grassy field in the dense pine woods back of Old Faithful Inn. Melanoplus intermedius Scudder. Sphinx, Park Co., Montana, Aug. 4, 1 c5^. Sappington, Gallatin Co., Montana, Aug. 12, 1 d^. Geyser formation, Mammoth Hot Springs, Yellowstone Park, Aug. 5, 1 cJ*. Salt Lake City, Utah, Aug. 13, 6 d" , 3 9 . Top of Ensign Peak, Salt Lake City, Aug. 1 3, 5 d* , 2 9 . Newcastle. Garfield Co., Colorado, Aug. 16, 1 d^. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 407 The form of the cerci of the male individuals here placed under this species agree fairly well with Scudder's figure of those parts, and not- withstanding his remarks to the contrary the authors believe inter- medius is at the most not more than a form of atlanis, and probably not even worthy of a name, varietal or otherwise. Numbers of specimens have been examined which seem to connect atlanis and intermedius, and all such are here treated under atlanis; only those closely approxi- mating Scudder's figure being referred provisionally to intermedius. Considerable variation in size is noticed, males ranging from 18 to 24 millimeters in the length of body. One male has the caudal tibiae pale reddish, all the others varying shades of glaucous. One of the typical sage-brush species. In some localities it was quite plentiful. Melanoplus atlanis (Riley). Livingston, Park Co., Montana, Aug. 4, 3 c^, 4 9 . Electric, Park Co., Montana, Aug. 4, 1 d^. Jefferson Valley, Montana, Aug. 5, 1 9 . Three Forks, Gallatin Co., Montana, Aug. 12, 1 cJ*. ]\Iammoth Hot Springs, Yellowstone Park, up to 7,050 feet, Aug. 5, 2 c?, 2 $ . Foun- tain, Lower Geyser Basin, Yellowstone Park, Aug. 6,19. Pine woods, Old Faithful, Upper Geyser Basin, Yellowstone Park, Aug. 7, 4 cJ*, 1 9 . Salt Lake City, Utah, Aug. 13, 14 c?, 12 9 . Hillside, 4,700 feet. Salt Lake City, Utah, Aug. 14, 5 d" 1, 9 . Top of Ensign Peak, 4,900 feet, Salt Lake City, Aug. 13, 3 cJ^ . Five miles south of Salt Lake City, in alfalfa field, Aug. 13, 2 c^, 3 9 . Grand Junction, Mesa Co., Colorado, Aug. 15, 1 cJ^, 2 9 . Newcastle, Garfield Co., Colorado, Aug. 16, 1 6". Knob Hill, Colorado Springs, Colorado, Aug. 17 to 22, 7 d^, 2 9 . Prairie land, Colorado Springs, Aug. 18, 3 d^, 2 9 . Manitou, Colorado Aug. 23, 9 d", 1 9 ; same locality at 6,700 feet, Aug. 16, 4 d", 1 9 .' Garden of the Gods, Colorado, Aug. 17, 5 d^, 7 9. Dark Caiion, Pike's Peak, 8,920 feet, Aug. 16, 2 9 . Roggen, Colorado, Aug. 24, 2 9 . This series exhibits a very great amount of variation in size and coloration, and in the form of the cerci many of the males placed here approach intermedius, as mentioned in the remarks under that form. The dullest specimens with little contrasted coloration are from the Garden of the Gods, and the palest ones are from the vicinity of Salt Lake City. However, while the majority of specimens of each series from the two above localities are uniformly dark or light, a few speci- mens from each locality are of what might be called the average type. Quite a number of specimens have the caudal tibite of various shades of glaucous. A common species in both sage-brush and prairie-grass. 408 PROCEEDINGS OF THE ACADEMY OF [Sept., Melanoplus bilituratus (Walker). Salt Lake City, Utah, Aug. 13, 1 c? . Top of Ensign Peak, Salt Lake City, Aug. 13, 1 d". These specimens are perfectly typical of bilituratus, which is probably only a race of atlanis. This is the first record of this species from Utah, and the first one east of Fort Halleck, Elko Co., Nevada, except one from Weeksville, Missoula Co., Montana. These two specimens were both taken on hilly ground covered with Melanoplus spretis (Thomas). Garden of the Gods, Colorado, Aug. 17, 2 $ . ^Mountain View, Pike's Peak, Colorado, Aug. 20, 9,705 feet, 1 c?. The capture of these specimens is of considerable interest as it shows that this very destructive species is still present, either as a migrant or as a native, in regions formerly devastated by it. The absence of recent records of this species from Colorado shows it has become a scarce species. The material examined has been compared with individuals from Nebraska and North Dakota. Two specimens from the Garden of the Gods were taken in a grassy meadow, while the other specimen was taken in an aspen thicket in sparse grass. Melanoplus defectus Scudder. Muir, Bozeman Tunnel, Park-Gallatin Co., Montana, Aug. 12, 2 d^. Mammoth Hot Springs, Yellowstone Park, Aug. 5, 1 c?. Summit of first foothill of Gallatin range. Mammoth Hot Springs, Aug. 5, 2 c?, 3 9 . Top of bare hill opposite Devil's Kitchen, Mammoth Hot Springs, Aug. 5,1 6". This species, previously known only from the type locality. Grand Junction, Colorado, and Nebraska, is quite distinct from the atlanis- spretis type, the form of the subgenital plate, which is considerabh^ elongate, being at a glance sufficient to separate it. Melanoplus dawsoni (Scudder). Muir, Bozeman Tunnel, Gallatin-Park Co., Montana, Aug. 12, 2 c?, 8 9 . Knob Hill, Colorado Springs, Colorado, Aug. 22, 2 9 . Manitou, Colorado, Aug. 23, 1 c^, 7 9 . Hillside near Manitou, Colorado, 6,700 feet,Aug. 16, 5c?,7 9. Material from Fort Collins, Colorado, has been examined in this con- nection. All the material studied belongs to the short-winged type of the species. Taken as a whole the series examined is c{uite uniform in size, and with such color variation as does exist being chiefly in the lighter or dark shade of the general color. 1906.] NATURAL SCIENCES OF PHILADELPHIA. 409 I found this species quite plentiful in a grassy field choked with many bushes. In this locality the species far outnumbered any other Orthopteron. At ]\Ianitou I found it on the hillside, where mountainous plants and scrub oaks formed practically the only vegetation. Melanoplus montanus (Thomas). Mammoth Hot Springs, Yellowstone Park, Aug. 5, 3 c^, 6 9 . Norris Pass, Continental Divide, Yellowstone Park, Aug. 7, 2 9 . This rather conspicuous species has previously been recorded only from Montana. The males in the series examined are quite uniform in size, and fairly so in coloration. The females vary considerably in size, one from Mammoth Hot Springs being distinctly smaller than the others, which latter are slightly larger than the two females from Norris Pass. The dark colors are more blackish and the browns more rufes- cent in the Norris Pass individuals than in any others. A distinctly mountainous species which I found only on the higher hillsides. The insects were powerful but too awkward to escape captvu'e. Melanoplus altitudinum (Scudder). Pike's Peak, Colorado, Dark Canon, 8,920 feet, Aug. 16, 16 d", 9 9 , 1 nymph; upper slope of Dark Canon, 9,000 feet, Aug. 16, 2 c?, 2 ? ; below Windy Point, 12,000 feet, Aug. 20, 3 d", 2 9 . The Dark Canon specimens are quite uniform in size, but the three Windy Point specimens are smaller than any of the former and the males are smaller than any specimens of the species seen. The small