et) trrrestectetts Pasi, Pauly Tahcuay PR@GEEDINGS OF The Academy of Natural Sciences OF PELL A Diner itl A VOLUME LXVI 1914 PHILADELPHIA : THE ACADEMY OF NATURAL SCIENCES LOGAN SQUARE | 1915 G b i l F | \3 Pal Tue ACADEMY oF NATURAL SCIENCES OF PHILADELPHIA. JaNuARY 30, 1915. I hereby certify that printed copies of the Procreprnes for 1914 were mailed as follows:— Pacesy © IRASL. J 5, sterscetesstevaconstct ass vo cescuescecensotaherds cote cae eee March 31, 1914. « 49-192.. 11, 1914. 193-224. 21, 1914 «225-290, 3, 1914 “ 291-434., % 25, 1914 6 ABS -B 38] Sactis sweet eek cm sacccdbseas re sesesen00sSdeartee eietenancnes betstaeee PTS 5, 1914. Ea ToAl ots TRS Rime fee RRP NG ic SEPEREEN Phar erat c August 25, 1914. §47-554.. .. September 19, 1914. “555-602. ...November 24, 1914, “603-618. ..December 9, 1914. es Wa Kt rele hepreretnsrooreras-reer a ere osdeachaa scasiaebb sb Gola aves eee January 23, 1915. EDWARD J. NOLAN, Recording Secretary. PUBLICATION COMMITTEE: Henry Skinner, M.D., Sc.D., WITMER StonE, A.M., Sc.D., Henry A. Piussry, Sc.D., Wiuram J. Fox, Epwarp J. Nouan, M.D. The President, SAMUEL Gipson Dixon, M.D., LL.D., ex-officio. EDITOR: Epwarp J. Nouan, M.D. CONTENTS. For Announcements, Reports, etc., see General Index. ALEXANDER, CHartes P. New or little-known craneflies from the United States and Canada. Tipulid, Diptera (Plates XX V—X XVII) inch Ee eau ere ANDREWS, Roy CuapmMan. Notice of a rare ziphioid whale, Mesoplodon densirostris, on the New Jersey Coast (Plates XVI-XVIII)..... a Banks, NatHan. New ane eis. nde ead pote (GEA Hem 5G U0), ae Ro RO BARRINGER, Danie Moreau. Further notes on Meteor Crater, Arizona (Plates XXI-X XIII)... Boyer, Cuarues 8. A new diatom (Plate X).. ones Brown, Amos P., and Henry A. Prussry. Brean aan sii lusks of the Oligocene of Antigua (Plate IX)................ CocKERELL, T. D. A. Miocene fossil insects... CrawLey, Howarp. Two new Sarcosporidia .... The evolution of Sarcocystis muris in the fetesaoate cells of the mouse (Plate XY) Fow er, Henry W. Fishes fron ins Rupimuni Rive er, ‘British Guiaman enc. atts Description of a new Blenne frond Slew fee Ww sain “gue on other fishes from the Middle Atlantic States Fishes collected by the Peary Relief Expedition of 1899 Fox, Henry. Data on the orthopteran faunistics of eastern Pennsylvania and southern New Jersey Hearn. Haroup. Certain features of Soledaenenne ee Alaa ment. eens KEELEY, FRANK do Makes on some igneous sods ae Oompa. Maine, and Pigeon Cove, Mass McInpoo, N. E. The olfactory sense of Hy meontera (Pl ates SNOT Xai ee The Speneeducing organs ae ‘he honey hee ‘(Plates DOIEXE XOX)... PAGE lv CONTENTS Pitspry, Henry A. Description of a new echinoderm (Plate WOLD) 5. eek Rind eine ee ae PILsBRY, HENRY A., and Amos P. Brown. The method of progression in Grane (PIBBCANSIING eters cc sncece trae List of land and fresh-water mollusks of Antigua : Poutton, Epwarp B. Mimicry in North American butter- flies. A reply (Plate V).... en Reese, Aubert M. The eee ean a the faiende Alligator (Plate SXaLDD) Peco eear tear terre rascaesciteiipene Renn, James A. G., and Morecan Heparp. A study of the species of the genus ee (Orthoptera, Tetti- gonide).. On the Orthoptera founda on ane Florida em aad in extreme southern Florida, IL... SmitH, Burnerr. Morphologic sequences in ans ‘eanaliculite fulgurs (Plate XXIV)... : SpaETH, Reynotp A. The distribation of the ¢ genus Gy noes in the vicinity of Haverford, Pennsylvania (Plates Stone, Witmer. On a collection of mammals from Ecuador TuHompson, JoserpH C. Contributions to the anatomy of the Ilysiidse PORE in See Vanatra, E. G. Land and fresh-water shells from eastern Canada..... Ras Montana shells... os ee as Warpie, H. NEWELL. Descrtntiontal of a Tsantsa in the ethnological collection of the Academy, with notes on another specimen (Plates VI, VII) 197 s eS PROGBEDINGS OF THE ACADEMY OF NATURAL SCIENCES a OF PHILADELPHIA. 1944. JANUARY 20. Mr. Cuarues Morais in the Chair. Nineteen persons present. The Publication Committee reported that papers under the following titles had been presented for publication in the PRocEEp- INGS: “Notes on some igneous rocks at Ogunquit, Maine, and Pigeon Cove, Mass.,”” by Frank J. Keeley (December 30, 1913). “Conspicuous flowers rarely visited by insects,” by John H. Lowell (January 2). “Mimiery in North American butterflies: A reply,” by Edward B. Moulton (January 9). The death of Silas Weir Mitchell, M.D., a member, January 4, was announced. The Council reported the following appointments: CoMMITTEE ON Finance.—John Cadwalader, A.M., Edwin S. Dixon, Effingham B. Morris, James D. Winsor, and the Treasurer. On Liprary.—Thomas Biddle, M.D., George Vaux, Jr., Henry Tucker, M.D., Frank J. Keeley, and Witmer Stone, A.M., Sc.D. On Pusuiications.—Henry Skinner, M.D., Witmer Stone, A.M., Se.D., Henry A. Pilsbry, Se.D., William J. Fox, and Edward J. Nolan, M.D. 2 PROCEEDINGS OF THE ACADEMY OF [Jan., On Instruction AND Lecrures.—Henry A. Pilsbry, Sc.D., Charles Morris, Henry Tucker, M.D., George 8. Morris, and Stew- ardson Brown. SoLiciror or THE AcaDEMy.—George Vaux, Jr. Curator oF THE WituiamM S. Vaux Coiections.—Frank J. Keeley. CusTopIAN OF THE Isaac Lea CoLuEections.—Joseph Willcox. The President of the Academy is ea officio a member of all Com- mittees. Puitie P. Catvert, Pu.D., made an illustrated communication on epiphytic Bromeliads of Costa Rica and their animal inhabitants. (No abstract.) Mr. Arthur Howell Napier was elected a member. The following were ordered to be printed: 1914.] NATURAL SCIENCES OF PHILADELPHIA. 33 NOTES ON SOME IGNEOUS ROCKS AT OGUNQUIT, MAINE, AND PIGEON COVE, MASS. BY FRANK J. KEELEY. It would probably be difficult to find a more remarkable display of igneous rocks than that along the coast of Maine south of Ogunquit. Here for a couple of miles the shale, dipping nearly vertically, is penetrated by almost innumerable dikes, varying from a few inches to over fifty feet in thickness and showing great variety in color and texture. The shale itself, as the result of these numerous intrusions, has been metamorphosed and indurated until it is frequently as hard as the igneous dikes. Fresh fractures are usually gray with faint indications of differently constituted lamina, but on the weathered surfaces the various layers assume different colors, often producing a decidedly striped appearance resembling banded jasper, becoming particularly noticeable in the rounded pebbles occasionally lining the shore. Numerous ramifying veins of white and yellow quartz further characterize the shales, and the extremely rugged character of the coast line, with several coves and an overhanging cliff exceeding fifty feet in height, together with the almost unlimited variation in color due to weathering of the shale and its igneous intrusives, has resulted in this section becoming a favorite haunt of artists. Prom early times it has likewise attracted the attention of geologists, and in the first geological survey of Maine, published in 1838, Charles T. Jackson gives considerable space to the description of the features of this district and calls attention to the manner in which some of the dikes intersect each other, as indicating that the igneous intru- sions can be referred to at least three periods. During the past summer, with the view of becoming better acquainted with the petrographical character of these rocks, I collected a number of specimens, from which I have since prepared sections and studied them microscopically. The locality receiving particular attention was a small cove on Israel’s Head, between the mouth of the Ogunquit River and Lobster Point. Here a patch of sand beach, used by the guests of the Ontio and Lookout Hotels as a bathing place, is surrounded by the usual shales of the region, 4 PROCEEDINGS OF THE ACADEMY OF [Jan., penetrated by several dikes and intrusions of igneous rocks. A series of these rocks was collected and this particular place selected for the purpose, not only because the intrusions seemed to include the principal types of igneous rocks, but also for the reason that they are located where they can be readily identified from the description by anyone interested. Commencing with the rocky point which extends out to low-water mark on the north of the bathing beach, this is penetrated by a dike about twelve feet thick of diabase porphyrite with phenocrysts of plagioclase too much zoizitised for specific identification, in a matrix of diabasic texture, composed of augite, biotite, plagioclase, and chlorite, the latter apparently altered pyroxene; also as accessory constituents, titanite, apatite, and secondary calcite. Dr. F. Bascom, who kindly looked over these sections with me, suggests that the reason much of the pyroxene is entirely fresh or in part altered to hornblende, while in other cases it is completely replaced by chlorite, is probably that there may have been two distinct varieties of pyrox- ene originally present, one more readily altered than the other. Near contact with the shale, this dike becomes basaltic in texture, a fine-grained mixture of feldspar, biotite, magnetite, and brown horn- blende, the latter no doubt replacing primary pyroxene, with pheno- erysts having the outlines of pyroxene, almost invariably completely altered to chlorite. A short distance toward the south, in the rocky wall back of the beach, is a twelve-inch dike of diabase with a small branch dike forking from it. Except that it contains a few small vesicles filled with secondary calcite, this is a typical diabase, fine and uniformly grained. Beyond it is a dike of basalt, four to eight inches thick. At the contact it is glassy, with lath-shaped feldspars oriented parallel to the wall. The interior is more completely crystalline, with phenocrysts of pyroxene altered to chlorite and many small, rounded patches of calcite, apparently filling vesicles. Further south is an irregular angular intrusion of trachyte. It consists almost exclusively of intermeshed rods of feldspar, apparently orthoclase somewhat kaolinized, with scattered patches of ferru- gious material slightly translucent and dark red in color when sufficiently thin, also generally red by reflected light. This rock corresponds in texture to the dyke rocks which have received the name of bostonite, but in the absence of any microscopical evidence of the presence of anorthoclase, a chemical analysis would probably be necessary to determine whether it should be so classed. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 5 Next comes another dike of diabase, characterized by the presence of considerable pyrite, which occurs in rounded aggregates, filling the interstices between crystals of augite and plagioclase. It also contains vesicular cavities averaging about a millimeter in diameter, filled with calcite and a little quartz, margined by acicular secondary hornblende. Somewhat south of the beach is a large intrusion having an irregu- larly rhomboidal outline, consisting of trachyte porphyry with large orthoclase phenocrysts in a felsitic matrix containing some horn- blende and a little quartz. Four sections were made from specimens of the shale associated with dikes mentioned, one broadly striped, another showing finer laminations, and the remaining two of rather uniform texture. All are highly silicious, including quartz grains up to a half millimeter in diameter scattered among finer grains of quartz and some secondary minerals, chiefly micas, sometimes biotite, and in one sample, a bright green mica. The extent of metamorphosis is indicated by apparent metosomatic penetration of the secondary minerals into some of the primary quartz grains. Striping, when present, is due to the concen- tration of such secondary minerals in layers, which in the original sediment were probably less purely silicious than the rest. Beyond a gully south of Lobster Point is a very noticeable dike about five feet in diameter, transected at an acute angle by another of same size. The first may be classed as a diabase porphyrite and contains vesicles about a millimeter in diameter such as characterize so many of the dikes here, but in this case there is about as much quartz as calcite in the cavities, while generally the filling is entirely of calcite. This rock also contains much pyrite in the form of isolated grains in the interior of the dike, but in clouds of minute particles several millimeters across, in the basaltic textured rock near the contact. The other dike is an olivine diabase, notable for numerous large idiomorphie phenocrysts of olivine now completely altered to serpentine of unusually high double refraction, apparently consisting, in part at least, of chrysotile, showing development along irregular cracks, so characteristic of the alteration process in olivine. Possibly a mile further south, beyond Perkin’s Cove, there is exposed on the shore an extensive intrusion of diabase, under which there is a water-worn cave between tide levels, locally known as the Devil’s Kitchen. This rock is a rather coarse-grained diabase with some primary biotite, in which the augite is perfectly fresh, but another ferro-magnesian constituent originally present has been 6 PROCEEDINGS OF THE ACADEMY OF [Jan., completely altered to a brown serpentine-like material sometimes apparently mixed with felted masses of biotite and chlorite. This may have been an orthorhombic pyroxene, as slight traces of it remaining in the heart of a couple of the brown areas showed parallel extinction, and it has none of the characteristics of olivine. In addition to the usual magnetite, apatite is present as an accessory mineral, but not at all plentifully. While the indurated shale is continuous along the sea coast for a couple of miles south of Ogunquit, granite outcrops at a number of places not far back from the shore, as at Pine Hill and further to the west at Mt. Agamenticus. It is hornblendic at the outcrops noticed, but I did not collect or further study any specimens. Some additional collections of dyke rocks were, however, made just north of Ogunquit on the road to Portland. Here, in widening the road, several outcrops have been cut away, leaving fresh exposures. One such is located on east side of road about one-eighth mile north of the car barn and shows three different igneous rocks penetrating or in contact with each other. Toward the south, there is first a gray, medium fine- grained diabase, then a compact black basalt. A section of the contact» demonstrates that the basalt was a later flow than the diabase. Next to it comes a coarse diabase porphyrite with feldspar phenocrysts, sometimes exceeding an inch in length, and beyond this another fine-grained diabase, and then indurated shale similar to that described from the shore. On the west side of the road, one-eighth mile further toward the north, is another good exposure of diabase porphyrite, in which the phenocrysts are developed to an extent that they appear to make up more than half the rock, in contact with basalt of later origin. For comparison, I give the following brief description of igneous rocks at Pigeon Cove, Mass., where years ago I collected and studied specimens from the dikes along a similar short section of the shore. The end of Cape Ann consists of light gray hornblende granite, quarried extensively for commercial purposes. Its feldspar is almost exclusively microcline and the hornblende is generally accom- panied by biotite. This granite is penetrated by many igneous dikes, although these are not so numerous or varied in character as those at Ogunquit. Near the extreme point of the cape, known as Andrew’s Point, below an unfinished square stone tower, is a dike of solvsbergite, a uniformly crystalline mixture of plagioclase with hornblende showing pleochroism from olive to indigo-blue, much finer grained in 1914.] NATURAL SCIENCES OF PHILADELPHIA. a an offshoot which extends into a parallel crack in the granite. A block of granite which has been picked up by the molten dike rock © is exposed in the interior of the latter. East of this is an extensive intrusion of quartz porphyry, so classed from the general characteristics of the whole mass, rather than from the microscopical examination of individual sections, some of which would otherwise rank as fine-grained granites, while others show a few phenocrysts and patches of micro-pegmatite. Within this intrusion are segregations containing crystals of hornblende several -inches long and large masses of blue quartz. It has been injected with diorite, but as it does not split in straight lines like the granite, no regular dike is exposed. The diorite has forced its way irregularly among the fragments of the older rock, some of which are included in it. The diorite consists of a fine-grained mixture of hornblende, biotite, and triclinic feldspar, with a few phenocrysts of zoizitised plagioclase and occasionally a small one of light colored pyroxene. It is intersected by numerous small white veins, no doubt of secondary origin, and consisting in one section examined of feldspar, both orthoclase and plagioclase, and light colored pyroxene. To the south, the dike of solvsbergite, which crosses the point, again appears, and further on a sharply defined dike of quartz porphyry several feet thick. Still further south are three small dikes of diorite, differmg from that at the point in several minor respects. There is but little biotite, and the hornblende is of a bluish- green color. No veins were noted, and the smallest dike, which is but a few inches thick, is very fine-grained and free from phenocrysts. They are probably all derived from the same source. Beyond them comes another series of dikes, all no doubt of similar origin. They are, respectively, two to three inches, twenty-eight inches, sixteen inches, and eighteen feet in thickness, the latter just below the Ocean View Hotel, while further on is still another nearly as large. The larger dikes are typical fully crystallized diabases, coarser or finer grained according to size of dike, with unaltered constituents and basaltic texture near the contacts. The two- to three-inch dike is basaltic throughout. It passes close to a swimming pool blasted out of the rocks, and is visible over the sloping shore for a couple of hundred feet, occupying a crack in the granite as straight and sharply defined as if cut with a knife. The sixteen-inch dike is admirably adapted for illustrating the effect of quick or slow cooling on an igneous rock, as it has an offshoot or branch, three-eighths to two inches thick, extending into the 8 PROCEEDINGS OF THE ACADEMY OF [Jan., granite. Where three-eighths to one-half inch thick, chips of the rock are procurable which permit of sections being made showing the granite penetrated by the small dike. Here the matrix is an almost opaque glass with plagioclase rods and phenocrysts of augite. When it becomes three-fourths inch thick, a slight tendency toward ecrystal- lization of the matrix is noticeable in the centre, and so on until the middle of the main dike is reached, where but little trace of the basaltic texture remains and the rock is a characteristic diabase. A similar series of sections can of course be made by starting from the contact in one of the larger dikes, but the transition from basalt to diabase is much more sudden. Two other rocks occurring nearby, but not appearing on the shore line just considered, are worthy of mention. One is a highly por- phyritic andesite with phenocrysts sometimes two inches long, indicated by their extinction angles to be oligoclase, in a matrix consisting of uralite, biotite, and plagioclase. This rock is not well exposed at Pigeon Cove. I have noted an outcrop in a door yard near centre of village and another in a hollow west of what is known as Sunset Rock, but across Sandy Bay it appears as a sharply defined dike in the granite, on the shore between Rockport and Straits- mouth. As the granite here seems capable of cleaving in a straight line for an indefinite distance and the three exposures are approxi- mately in line, although widely separated, they may all pertain to the same dike. The other rock referred to oecurs in a cut leading from the shore to the Rockport Quarry, near the archway under main road. It is a light brown crystalline rock which proved to consist entirely of micro-pegmatite, the best example I have seen of this intergrowth of quartz and microcline. 1914.] NATURAL SCIENCES OF PHILADELPHIA. +7 ON A COLLECTION OF MAMMALS FROM ECUADOR. BY WITMER STONE. Mr. Samuel N. Rhoads made a collecting trip to Ecuador, February- July, 1911, and secured a valuable series of vertebrates. His entire colle¢tion was purchased by the Academy of Natural Sciences of Philadelphia, and reports on the fishes and reptiles have already appeared in the Procrrpincs.!| The mammals, comprising sixty- eight specimens, referable to nineteen species, were obtained for the most part on the paramo and the region immediately below, on Mt. Pichincha, 10,000-13,000 feet, while a few additional specimens were obtained from the mountains above Chambo, from the Pagma forest near Chunchi, 7,000 feet, and from Bucay, province of Guayas, 975 feet. As the Academy previously possessed no mammals whatever from the Andes, the satisfactory identification of much of Mr. Rhoads’ material was rendered impossible until such specimens could be secured for comparison. Upon his return-from Peru, Mr. Wilfred H. Osgood, being anxious to make comparisons with certain Eeuador species, generously offered to compare Mr. Rhoads’ specimens with the series in the Field Museum in return for their use in the identifica- tion of his Peruvian mammals. This he has done and has given me his opinion as to their relationships. The American Museum has, through Dr. J. A. Allen, curator of mammals, loaned me specimens of Blarina thomasi and B. squami- pes for purposes of comparison and the U. 8. National Museum a series of Sciurus hoffmanni. For this aid I would extend my sincere thanks especially to Mr. Osgood, without whose co-operation this paper could not have been prepared. Mr. Rhoads has kindly furnished me with some field notes on Cenolestes and other interesting species which are duly credited. 1. Ichthyomys soderstromi de Winton. Ichthyomys séderstr6mi de Winton, Proc. Zool. Soc. London, 1896, p. 507. Two specimens of this interesting fish-eating rodent were obtained from Mr. Séderstrém, of Quito, who collected: the type specimen on 11911, p. 493; 1913, p. 153. 10 PROCEEDINGS OF THE ACADEMY OF [Jan., the Rio Machangara, Ecuador, February, 1895. One of the speci- mens before me was obtained March 16, 1904, but neither has an exact locality. 2. Epimys rattus (Linn). Mus rattus Linnzeus, Syst. Nat., ed. 10, vol. 1, p. 61, 1758. (Sweden.) One skin and three skulls in the collection, obtained at Hacienda Jalaneay, Chunchi, Chimbo, and at Bucay, Guayas. 3. Epimys norvegicus (Erxleben). Mus norvegicus Erxleben, Syst. Regni Anim., vol. 1, p.381,1777. (Norway.) One skull, from specimen caught in a house at Bucay, June 15, 1911. 4. Mus musculus (Linn). Mus musculus Linneus, Syst. Nat., ed. 10, vol. 1, p. 62, 1758. (Sweden.) Five specimens obtained at Riobamba and Hacienda Garzon at the southern foot of Mt. Pichincha, 11,000 feet. 5. Oryzomys minutus (Tomes). Hesperomys minulus Tomes, Proc. Zool. Soe. London, 1860, p. 215. (Eeuador.) The Hesperomys minutus of Tomes was based upon an immature specimen obtained by Fraser, but without definite locality, although it was supposed to be from Pallatanga. Oldfield Thomas, in describing a new species of this genus from Peru (Ann. and Mag. Nat. Hist., 1894 (XIV), p. 357), identified with Tomes’ type an adult specimen in the British Museum which was also colleeted by Fraser at Palla- tanga, December, 1858. Later (op. eit., 1898 (II), p. 267) he de- scribed this specimen as new under the name Oryzomys dryas. His explanation of this action is that specimens received from Mr. Séderstrém (locality not given) were obviously identical with Tomes’ type and different from the Pallatanga skin. Mr. Rhoads’ five specimens were all obtained in the vicinity of Hacienda Garzon (or Rosario) at the southern foot of Mt. Pichincha, some of them “in meadows and swamps below the house, 10,150 feet, ”’ others higher up near the paramo, 12,000 feet. Hind Length. Tail. foot. Par. 3, May 12, 1911 180 100 23 i has oc’, May 12, 1911 . 183 103 23 12 o', May 13, 1911 . 178 98 23 13 oc’, May 15, 1911 . 180 100 23 13 9, May 12, 1911 . 190 103 22 12 The Séderstr6m specimens mentioned by Thomas doubtless came from Pichincha, which is of easy access from Quito, and probably 1914.] NATURAL SCIENCES OF PHILADELPHIA. 11 Tomes’ type was also obtained there. While the latter is probably not absolutely identifiable after this lapse of time, it seems best to retain his name minutus for this form. The animal isa little smaller than Mus musculus and almost exactly the same color below, while above it is very much more rusty with a clearer line of demarcation on the side. The ground color above, at the base of the tail where it is purest and brightest, is ‘‘tawny ochraceous” of Ridgway’s Color Standards 1912, but is duller and browner on the back and head and thickly mixed everywhere with black hairs. 6. Reithrodontomys soderstromi Thomas. Reithrodontomys Séderstrémi Thomas, Ann. and Mag. Nat. Hist., 1898 (1), p. 451. (Quito.) Four specimens obtained from Mr. Séderstr6m, who collected them at or near Quito, and one obtained in Quito by Mr. Rhoads, April 25, ad. 2, length 185 mm., tail 105, hind foot 20, ear 14.5. 7. Phyllotis haggardi Thomas. Phyllotis Haggardi Thomas, Ann. and Mag. Nat. Hist., 1898 (II), p. 270- (Mt. Pichincha.) One immature female (12,697, Coll. A. N.S. Phila.) obtained at Hacienda Garzon at the southern foot of Pichincha, at 12,000 feet, nearly up to the paramo, May 12, 1911. It agrees very well with Thomas’ description, excepting that the tail is only 50 mm. in length, whereas that of the type measured 86 mm. 8. Hpeomys vulcani (Thomas). Aipeomys vulcani Thomas, Ann. and Mag. Nat. Hist., 1898 (1), p. 452. (Mt. Pichincha, 12,000 feet.) A partially mummified skin (12,698, Coll. A. N.S. Phila.) obtained from Mr. Séderstrém, collected on the west side of Mt. Pichincha. Thomas seems to have made a mistake in citing the tail as “barely as long as the head without the body,” since a few lines below he gives length of ‘head and body 111 mm. and tail 84 mm.” The measurements are probably correct as given, since our dried specimen is 100 mm. long exclusive of the tail which measures about 72 mm. 9. Thomasomys paramorum Thomas. Thomasomys paramorum Thomas, Ann. and Mag. Nat. Hist., 1898 (1), p. 453. (Paramo, south of Chimborazo.) Ten specimens obtained on the paramo of Mt. Pichincha are apparently referable to this species, obtained originally in a similar region near Mt. Chimborazo. They are almost exactly like the much smaller Oryzomys minutus in color, both above and below. 12 PROCEEDINGS. OF THE ACADEMY OF [Jan., Mr. Rhoads’ measurements are as follows: Hind Length. Tail. foot. Ear. @, Hacienda Garzon, 10,500 itis; May Aad ccs cc . 201 108 23 14 oF juv., Hacienda Garzon, 10,500 1a Wenig TNS): erates LT! 95 23 14 2, Hacienda 'G arzon, 10,500 $iftiys UY Deohieed Sy ee Seas 215 118 24 12 ils Hacienda Garzon, 10, 500 oe ela yl Oneal P2LG 118 24 16 Cus Paramo, 13,000 ft., May ve 205 100 24 15.5 2, Paramo, 13,000 ft., May 7. 205 125 25 16 Q, Paramo, 13,000 ft., May 7. 204 112 24 14 2, Paramo, 13,000 ft. , May 7. 195 102 24 14 o, Paramo, 13,000 ft. ’ May 210 115 25 15 10. Thomasomys rhoadsi sp. nov. Mr. Rhoads secured a series of seven specimens of another Thoma- somys on the paramo of Mt. Pichincha apparently allied to T. cinereus Thomas, from Cutervo, Peru. As no such animal seems to have been described from Ecuador, I propose to name it Thoma- somys rhoadsi, in honor of Mr. Samuel N. Rhoads whose expedition to Ecuador has brought to light so mamy interesting species of verte- brates. Type No. 12,709, Collection Academy of Natural Sciences of Philadelphia, @. May 15, 1911, Hacienda Garzon, Mt. Pichincha, 10,500 feet. Coll. by Samuel N. Rhoads. Skull similar to that of 7’. cinereus Thomas, but a little larger with much wider interpterygoid fossa. Fur long and soft. Brown tints of upper parts nearly bistre of Ridgway’s ‘‘ Color Standards,”’ but so mingled with black hairs that the general appearance is very much darker, nearly black on the back. The fur of the under parts is gray with buffy tips, not whitish. The upper side of the hind feet is clothed with dusky hairs down to the base of the toes, while the latter have scattered white hairs, notably at their extremities. The fore feet are but scantily haired, while the hairs on the tail in no way conceal the scales. The ears are well haired. Length 250 mm., tail 110, hind foot 31.5, ear 16. Skull measure- ments.2. Total length 35 mm., greatest breadth 18, molar series 7, incisors to first molar 10, breadth of constriction between orbits 6 lower jaw (bone only) 20. 2 As used by Thomas in description of T. cinereus. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 13 The series of skins measures as follows: Hind Length. Tail. foot. Kar. o&, Hacienda Garzon, 10,500 : ft., April 28 5, eo) 115 32 17 2, Hacienda Garzon, 10,500 ft., April 28 wes 212 103 29 16 2, juv., Hacienda Garzon, 10,500 ft., April 28 . 210 100 28.5 15.5 o&, Hacienda Garzon, 10,500 ft., April 28 » PBS 115 30.5 1525 o&, Hacienda Garzon, 10,500 ft., May 12 ee 30 15 2, Hacienda Garzon, 10,500 ft., May 15... _ 250 110 Bil 5) 16 Mr. Osgood informs me that so far as he is aware all of the de- scribed species approximating this in size have relatively shorter tails, while the blackish back distinguishes it from all those species available for comparison. “Several specimens of this mouse were secured above the Casa Garzon, along the trail from that Hacienda to the summit of Mount Pichincha. The elevation was a few hundred feet above the valley, where the marsupial Cwnolestes was secured, approximately 10,500 feet and probably this is their lowest range, as much trapping was done 200 to 300 feet farther down without securing any. Their habitat was on the rocky wooded slopes, where they had burrows similar to those of our North American Microtine rodents, from one to three inches below the surface of the soil and débris. “T was interested to notice, when skinning these mice, that all of them, or possibly only the males, were supplied with a remarkable prolongation of the anus, that organ being extended, or rather, extensible, beyond the thighs for half an inch or more, as indicated in the memoranda on the labels of the specimens. This prolongation was not an internal extension, but external, being hairy throughout.” (S. N. Rhoads.) 11. Akodon mollis altorum Thomas. Akodon mollis altorum Thomas Ann. and Mag. Nat. Hist., 1913 (II), p. 404. (Canar.) Nine specimens of this mouse were obtained on the paramo of Pichincha or near Hacienda Garzon, a little lower down. Externally they are almost exactly like Thomasomys paramorum, averaging perhaps a little darker or duller, but are easily recognized by the shorter tail. 14 PROCEEDINGS OF THE ACADEMY OF [Jan., - The measurements of the series are as follows: Hind Length. Tail. foot. Ear. 3, Hacienda Garzon, 10,500 Hite AOU 2S yes Eee 185 73 24.5 14 Q , Paramo, 13,000 ft., May 7. 105 2034) eee 14 o, Paramo, 13,000 ft., May 7. 169 69 23 14 2, Paramo, 13,000 ft., May 7. 160 64 23 12 @, juv., Paramo, 13,500 ft., IMiatyeS: sets cates Balo5 60 23 10 o, Paramo, 13,500 ft., May 8. 187 80 23.5 11 9, Hacienda Garzon, 10,500 ft., May22...:......... mali, 72 23 13.5 9, Hacienda Garzon, 10,500 ttre Mlayel Gece aaa eed 68 2303 15 2, Cumboya, N.S. of Quito, IM ya29 8 cae ee ee 52 60 23 13.5 12. Sciurus irroratus (Gray). ” Macroxus irroratus Gray, Ann. and Mag. Nat. Hist., 1867, XX, p. 431. (Upper Ucayali River, Brazil.) : One specimen (12,725, Coll. A. N.S. Phila.), male, procured in the Pagma forest, July 11, 1911, is perhaps referable to this species, though no suitable material is available for comparison. Length 330, tail 152, hind foot 50, ear 20. “Squirrels were reported by the natives to be in the forests about Huigra from 4,000 feet and upwards, but we saw none until we penetrated the Pagma forest above Hacienda Jalancay, 6,000 to 7,000 feet. They were exceedingly rare, however, even in this forest. I saw one, after the specimen secured was taken, in an orange grove near the Casa, 1,500 feet lower down. I can state nothing about their habits, not having personally observed them in life. No other species of squirrels were observed in our wanderings.” (S. N. Rhoads.) 13. Sciurus hoffmanni soderstromi subsp. noy. One specimen (No. 12,726, Collection Academy of Natural Sciences of Philadelphia), from Mt. Pichincha, November, 1903, collected by L. Séderstrém. While a member of the S. hoffmanni group, this specimen is much more rusty-red especially across the shoulders and on the fore legs than any specimens I have seen from Costa Rica or any in a con- siderable series with which Mr. Osgood has compared it in the Field Museum. In other respects it does not seem to differ. I find no name applicable to this form and would propose that it be called Sciurus hoffmanni séderstrémi, in honor of its collector, 1914.] NATURAL SCIENCES OF PHILADELPHIA. 15: who has done so much in developing our knowledge of the birds and mammals of Ecuador. 14. Sylvilagus andinus (Thomas). Lepus andinus Thomas, Ann. Nat. Hist., XX, 1897, p.551. (Mt. Cayambi.) Three specimens obtained on the mountains above Chambo, 10,000-10,400 ft. Hind Length. Tail. foot. Kar- OPe April Oke eoes., ence GSU) 17 73 57 Gin calyoval 1), see ceeneee wee 840 7 72 59: o', juv., April 17...... . 280 15 60 55 These specimens no doubt represent S. a. chimbanus Cabrera (Trans. Mus. Cien. Nat. Madrid; Zool. Series, No. 9, 1913), but as I am unable to appreciate the difference between the two forms and have no typical material of either for comparison, I prefer to refer the specimens to Thomas’s S. andinus. “These cottontails were found, as we rose from the Chimbo River valley up the slopes, 500 to 1,000 feet above the town of Chambo. They frequented the brush-grown pastures in similar situations to those frequented by our cottontails of the United States, only they kept more closely to the bushes and were not found in the open pastures. They seemed to range from that point no lower, but to reach up the slopes into the paramo, at an elevation of 12,000 to 13,000 feet, but we secured none in the paramo region, abundant as they evidently were by their tracks and droppings in the thick tussock grass. We also saw several on the paramo of Mount Pichincha, above Hacienda Garzon, but were fated to secure no specimens. None were seen in this locality. below the paramo, viz., 11,000 feet, approximately. It is quite possible that the paramo rabbit is distinct from the animal of the templada, at least subspecifically, or that there are two species, not distinguishable at a distance by a field observer.’ (S. N. Rhoads.) 15. Mazama americana (Prxl). Moschus americanus Erxleben, Syst. Regni. Anine, vol. I, p. 1777. One female specimen (12,730, Coll. A. N. S. Phila.) obtained at the junction of the Chanchan and Chaguancay Rivers on the lower western slope of the Andes, February 27, 1911. Length 1,050 mm., height at shoulder 600, height at rump 715, tail 160, ear 115. Mr. W. H. Osgood has revived Erxleben’s name for M. nemo- rwagus (Field Musewm, Nat. Hist. Publ., No. 155, vol. X, p. 48, 1912) 16 PROCEEDINGS OF THE ACADEMY OF [Jan., but Oldfield Thomas (Ann. and Mag. Nat. Hist., XI, 1913, p. 585) considers that it applies rather to MW. rufus of authors, and I so use it. 16. Mustela aureoventris Gray. Mustela aureoventris Gray, Proc. Zool. Soc. London, 1864, p. 55. (Eeuador.) One specimen (12,731, Coll. A. N. 8. Phila.), male, procured in the Pagma forest, July 11. Measurements: Length 420 mm., tail 165, hind foot 48, ear 12. While there is an earlier Mustela auriventer Hodgs, 1841, which in my opinion would invalidate Gray’s name, yet it does not seem desirable to propose a substitute until the relationship of the Ecuador species to those described from Colombia and Peru is definitely settled. At present I am unable to secure any of the material necessary for comparison. “The only weasel seen on the expedition was shot by Mr. Lemmon, my assistant, in the Pagma forest from a tree at a low elevation, and, until picked up, was thought to be a squirrel.” (S. N. Rhoads.) 17. Blarina osgoodi sp. nov. Eight specimens of a Blarina were obtained at Hacienda Garzon on Mt. Pichincha at an altitude of 10,500 ft. and on the paramo 1,500 ft. higher. Four of these were prepared as skins and the others preserved in spirits. So far as I am aware, no Blarina has previously been found south of Colombia (B. thomasi Merriam and B. squamipes Allen) and Merida, Venezuela (M. meridensis Thomas); and as the Ecuador specimens differ from all of these, I propose to"name them Blarina osgoodi for Mr. W. H. Osgood, of the Field Museum of Natural History, whose expeditions to South America have done so much to enrich our knowledge of its mammals and birds. Type from Hacienda Garzon, Mt. Pichincha, 10,500 ft. altitude, May 12, 1911, female, collected by Samuel N. Rhoads. No. 12,732, Collection A. N.S. Phila. Similar in size and coloration to B. thomasi Merriam from Bogota, but not quite so brown. Anterior unicuspid teeth less distinctly angulate on the inner side than in B. thomasi and the third and fourth unicuspids quite different. The third unicuspid in B. thomas? is apparently as long as broad, while in B. osgoodi it is very much broader than long, and both it and the fourth unicuspid are smaller in every way. Total length 105 mm., tail 30 mm., hind foot 14 mm. Skull measurements: total length 22 mm., greatest breadth 10 mm. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 17 Measurements of the other skins: Hind Length Tail foot 2, Hacienda Garzon, 10,500 ft., May 5... 113 30 14 9, Paramo, 13,000 ft., May 7.................... 105 26 14 o', Paramo, 13,000 ft., May 7.................... 106 26 14 I am not prepared to say how good a character the squamation of the feet may prove to be, but B. osgoodi exhibits scaly plates on the hind feet similar to, but smaller than, those of B. squamipes. They are not apparent on the fore feet nor on any of the feet of B. thomasi which has the feet much more thickly haired than either B. osgoodi or B. squamipes. In general size, color, and length of tail B. meridensis and B. squamipes appear to be very similar, and it would be interesting to know whether the former exhibits the marked squamation. “These shrews were first taken at the Hacienda Garzon on the southern slope of Mount Pichincha, about on the level with the Casa. They were trapped in runways along the banks of a deeply cut arti- ficial ditch, the kind used in that country in lieu of a fence against cattle. The location was wooded clearing and brush land, at that point deeply shaded upland. The runways in location and character were precisely like those made and used by Blarinas in the United States.” (S. N. Rhoads.) 18. Didelphis marsupialis Linn. Didelphis marsupialis Linneus, Syst. Nat., ed: X., 1758, p. 54. One specimen obtained at Quito collected on the east side of Cayambe. Also a ramus of a lower jaw found at Bueay, June 20. Mr. Osgood has compared the Cayambe specimen with true marsupialis of Guiana and Venezuela and finds but little difference except that it is somewhat larger. It may possibly be referable to D. m. colombica Allen, type locality Santa Marta. 19. Cenolestes fuliginosus (Tomes). Hyracodon fuliginosus Tomes, Proc. Zool. Soc. London, 1863, p. 51. (Ecuador.) : Two female specimens of this little known marsupial were obtained by Mr. Rhoads at Hacienda Garzon, near the paramo of Mt. Pichincha. May 12, 1911, about three months after Mr. Osgood had rediscovered the very closely allied C. obscwrus on the Paramo de Tama on the borderland of Colombia and Venezuela. So far as I am aware, no specimen of this interesting animal has been obtained in Ecuador since the type was taken by Fraser about 2 “a 18 PROCEEDINGS OF THE ACADEMY OF [Jan., 1859 (see Tomes, P. Z. S., 1860, p. 213). Fraser’s specimen was sup- posed to have come from Pallatanga, but this was not certain as the collection had been mixed up, and it is quite as likely, in view of Mr. Rhoads’ discovery, that he got it on Mt. Pichincha. As given by Tomes in describing the animal in 1863, the measure- ments of the type (reduced to millimeters) are: head and body 97 mm., tail 97, head 31. In his description of C. obscurus Thomas gives the measurements as head and body 151, tail 144, hind feot 23, ear 12x 11.5, and bases the species as distinct from C. fuliginosus on the fact that it is “double the size.” Mr. Rhoads’ measurements of his two specimens are: Hind Length. Tail. foot. Har. 2, Hacienda Garzon, 10,500 ts, Misiy. di Diet eee te . 198 103 22 ail @, Hacienda Garzon, 10,500 ft., May 12... f . 217 110 22.5 11.5 Mr. Osgood’s series of five females of C. obscwrus averaged as follows: Total length 223 mm., head and body 107.6, tail vertebree 115.4, hind foot 22.5. It will be readily seen, therefore, that his specimens and those of Mr. Rhoads are practically identical in measurements, while a com- parison of the skins made by Mr. Osgood shows ‘“‘no appreciable difference.” Unless it is maintained that there are two species in Neuador, it looks very much as if C. obscurus Thomas might become a synonym of C. fuliginosus Tomes. Such a view, involving the assumption that Tomes’ measurements were quite erroneous or that his specimen was a young one, seems to me much more rational than to suppose that two species of quite different size occurin Ecuador. “The two specimens were secured in swampy ground, the edge of a large pasture on the Hacienda Garzon, within a few feet of a swiftly flowing stream of considerable size. They were caught in small cyclone mouse traps set in underground runways among the thick grass, these runways being about on the level with the waterline of the swamp. They were caught on the same day, soon after placing the traps in that locality, but although I continued to trap there for a week longer, having as many as 40 or 50 traps in that place, I secured no more specimens there, nor in any other similar localities where trapping was done. The stream alluded to runs over a bed strewn with volcanic rocks and boulders and is in 1914.] NATURAL SCIENCES OF PHILADELPHIA. 19 an open cultivated valley-head, draining the south slopes of Mount Pichincha, about 8 miles south of Quito and at an elevation of about 10,500 feet, the valley at this point being about half a mile wide and extending to even greater widths as far as one can see, in a southerly direction.” (S. N. Rhoads.) 20 PROCEEDINGS OF THE ACADEMY OF [Jan., THE DISTRIBUTION OF THE GENUS CYCLOPS IN THE VICINITY OF HAVERFORD, PENNSYLVANIA. BY REYNOLD A. SPAETH. The original purpose of this paper was, after the notes and obser- vations of two years had been collected, to prove that by a careful and regular study of the water from a single locality, many of the forms of the Cyclopide which are now considered rare would be found to be quite abundant, at least at certain seasons of the year. Owing to unforeseen circumstances, this purpose had to be abandoned in part, and although all the forms recorded in this paper have been taken from a single small pond in the vicinity of Haverford College, the records of monthly abundance or rarity of the different species are not complete. Certain species, notably C. varicans, C. fimbriatus var. poppei, and C. phaleratus, appear far more abundantly in the spring from the middle of March to the end of May. C. prasinus I have found in the greatest numbers in September and early October collections. (C. albidus, C. viridis var. insectus, and C. serrulatus do not vary so noticeably in the collections, while C. fuscus is the most unvarying species of the genus, a few being found in the water taken the year round. In his paper of ’97, E. B. Forbes has made a most excellent revis- ional study of the North American Cyclopide. His work has sim- plified investigation for all future workers in this field, and his very careful and excellent observations can hardly be overestimated. I have followed his system of subgenera to avoid confusion. Cragin’s paper of ’83 has scarcely received due credit from the investigators who have followed him. His drawings show con- siderable accuracy of detail and very few of the important structural features have escaped his notice. Marsh’s work has been rather more of a plankton study than one of individual structure. Brewer’s paper of ’98 lays considerable emphasis on feet-armature, which is now generally considered a variable feature and not a reliable character for specific distinction. Miss Byrnes’ recent paper of ’09 is somewhat confusing. Rather unfortunately she has reverted to Herrick’s varietal names of C. signatus, has credited Forbes with both C. bicuspidatus Claus and C. serrulatus Fischer in her list of species studied, and has confused 1914.] NATURAL SCIENCES OF PHILADELPHIA. 21 C. bicolor Sars with C. varicans Sars. Many of her drawings, however, are excellent, and she has furthermore presented all figures on the* same scale, which is of value in a comparative study of the species. Her chief source of information has apparently been Herrick, whose work, while remarkable, considering the number of species described, is, owing to its large field, often misleading and at times quite incor- rect. She has neglected both Schmeil and Forbes—the most com- plete modern works on the subject. Her most careful study has been onthe armatures of the four pairs of swimming feet, which, while interesting in showing the variations that occur, are not sufficiently constant characters to warrant a deduction of general conclusions. : EK. B. Forbes was the first American investigator to lay much stress on the importance of the shape of the receptaculum seminis. He was convinced by Schmeil’s work of 792 that it was the most important character for specific distinction. Its shape, while often somewhat obscure, varies very little, and the same general outline is preserved in all members of the same species. This outline is effected largely by the number of spermatozoa contained in the receptaculum. Other important distinguishing characters are the number of female antennal segments, the length of the first female antennz, hyaline plates and sensory hairs and clubs, armature of the stylets (variable in some species), and the shape and armature of the fifth foot. The number of female antennal segments is usually quite constant. Of the specimens from this locality, C. phaleratus and C. varicans were the only exceptions to the rule. The former may have either ten or eleven joints and the latter eleven or twelve. In the plate of C. varicans I have shown only the eleven-jomted form of the first antenne, as this seems to be a winter transitional stage, though such individuals were all sexually mature. The twelve-jointed form did not appear until the April collections. The length of the female first antennze varies remarkably in C. serrulatus, but im all species having antennz of less than twelve segments, it is quite constant. Hyaline plates vary very slightly as to their edges. Minute serra- tions occasionally appear or are absent, notably in C. prasinus. Sense-clubs and hairs are constant features. The armature of the stylets and their proportions are constant in some species, as C. modestus, C. varicans, and C. phaleratus. In others, as C. bicuspidatus and especially in C. serrulatus, the variation is very great. 22 PROCEEDINGS OF THE ACADEMY OF [Jan., Forbes considers the armature of the swimming feet ‘‘of con- siderable value in certain cases, and constant as a rule;” that “the general character with regard to strength, etc., may usually be relied upon; but I have often seen in a single specimen all the gradations between spines and sete, and it would be impossible from this character to say which of the two names should be applied.” The more I have worked out the armature of the swimming feet, the more IT am convinced that a constant standard cannot be obtained for such an armature for all members of the same species. The following table is an illustration of the confusing results obtained in attempting to establish a standard swimming-feet armature for C. phaleratus. Case NUMBER l. First pair—outer ramus, three spines, five sete; inner ramus, one spine, four sete. Second pair—outer ramus, four spines, four sete; imner ramus, one seta, one spine, four sete. : Third pair—outer ramus, four spines, five sete; inner ramus, one seta, one spine, four sete. Fourth pair—outer ramus, three spines, five sets; inner ramus, one seta, two spines, two sete. Case NUMBER 2. First pair—outer ramus, three spines, five sete; inner ramus, one seta, one spine, four sete. Second pair—outer ramus, four spines, five set#e; inner ramus, one seta, one spine, four sete. Third pair—outer ramus, four spines, five sete; inner ramus, one seta, one spine, four set. Fourth pair—outer ramus, three spines, five setz; imner ramus, one seta, two spines, two sete. CasE NUMBER 3. First pair—outer ramus, four spines, five sete; inner ramus, one spine, four sete. Second pair—outer ramus, four spines, five sete; imner ramus, one seta, one spine, four sete. Third pair—outer ramus, four spines, five sete; inner ramus, one seta, one spine, four sete. Fourth pair—(right) three spines, five sete; inner ramus, one seta, two spines, two sete. Fourth pair—(left) four spines, four sets; inner ramus, one seta, two spines, two sete. ; The above cases show a variation in both outer and inner ramus of the first pair, in the outer ramus of the second pair, and the outer 1914.] NATURAL SCIENCES OF PHILADELPHIA. 23 ramus of the fourth pair. The case of a spine becoming a seta as in the outer ramus of the fourth foot of Number 3 right and Number 3 left is not as unusual as the appearance of an extra spine, as on the outer ramus of the first pair of Number 3 compared with the cor- responding ramus of 1 and 2. In some species the armature is considerably more constant. In C. modestus, for example, it rarely varies at all, specimens from widely separated localities showing an identical arrangement of both spines and sete. After the examina- tion of 'a great many specimens of a number of species, I have con- cluded that unless there are very distinct differences in some of the other important distinguishing characters, a slight variation in swimming-feet armature is not sufficient proof for establishing a new species or even a variety. The fifth foot, both in armature and shape, is constant as a rule. There are slight variations occasionally in the shape of the segments and comparative lengths of spines and sete. In a single case I have found a mature female of C. viridis var. insectus with an extra com- pletely developed seta on the distal segments of both fifth feet. This very unusual form is now in the collection of C. D. Marsh. The receptaculum seminis has already been mentioned. In pre- serving specimens it is advisable to use a 1 per cent. solution of formalin to avoid contraction and distortion of this organ. All of the plates have been drawn from life. The movements of the living animals may be readily overcome by the use of a 1-1000 parts solution of chlorotone. This I have found to be very con- venient and all danger of flattening and distorting the outline of specimens may be thus avoided. By this method the same individual may be repeatedly used, as the animal recovers a few moments after having been replaced in fresh water. Care must be taken that the solution of chlorotone does not become concentrated by evaporation, in which case the animals are killed by its too violent effects. The work in this paper covers a period of three years. All of the species described herein have been taken from a single pond of less than one acre area. The Copepoda having been largely neglected in this part of the country, it is interesting to note the presence of some of the more unusual forms, as C. varicans, C. fimbriatus var. poppei, and several others. I am greatly indebted to Dr. C. D. Marsh for his identification of specimens and his general interest and assistance throughout the period of study. My thanks are also due Mr. E. B. Forbes and Dr. H. S.. Pratt. The latter, through his unfailing personal interest and encouragement and by placing the 24 PROCEEDINGS OF THE ACADEMY OF ‘Jans very best of the laboratory equipment at my disposal, is largely responsible for the existence of this paper. I gladly take this oppor- tunity to acknowledge also the assistance of Mr. J. Ashbrook in collecting material. Genus CYCLOPS O. F. Miiller. Subgenus CYCLOPS Claus s, str. Cyclops bicuspidatus Claus. PI. II, figs. 1-5. Cyclops bicuspidatus Claus, 757, p. 209. Cyclops pulchellus Sars, ’63, pp. 246, 247, pl. XI, figs. 6 and 7. Cyclops navus Herrick, ’82a, p. 229, pl. V, figs. 6-13, 15-17. Cyclops thomas* Forbes, ’82a, p. 649. Cyclops bicuspidatus Schmeil, ’92, pp. 75-87, pl. II, figs. 1-3. Cyclops minnilus Forbes, 93, p. 247. Cyclops serratus Forbes, ’93, pp. 247, 248. Cyclops forbest Herrick and Turner, 795, p. 104. Cyclops navus Brewer, ’98, p. 133. Cyclops pulchellus Brewer, °98, pp. 133, 134. Cyclops bicuspidatus Lilljeborg, ’01, pp. 11-14, pl. I, figs. 12-17, pl. II, fig. 1 Cyclops pulchellus Byrnes, pp. 24, 25, pl. X. Cyclops bicuspidatus Byrnes, pp. 25, 26, pl. X. Synonymy and Distribution —In his discussion of the synonymy of this very variable and widely distributed species, Forbes has cleared up the question of C. thomasi Forbes, C. navus Herrick, _C. minnilus Forbes, and C. serratus Forbes=C. forbesi Herrick. He finds, after a careful comparative study, that they should all be considered as slight variations of the type C. bicuspidatus, but the differences are not sufficient to warrant the varietal names. Schmeil has also discussed the question at great length. He does not consider C. thomasi Forbes nor C. navus Herrick of specific value, since the latter is only a variety of C. thomasi Forbes. Brewer has described, as C. navus Herrick and C. pulchellus Koch, two species of cyclops from the vicinity of Lincoln, Nebraska. His descriptions of the rudimentary and swimming feet show that he was dealing with slightly different specimens of C. bicuspidatus Claus. As C. pul- chellus Herrick and C. bicuspidatus Forbes, Miss Byrnes has described examples of C. bicuspidatus Claus from Long Island. Her descrip- tions of the swimming feet and her drawings of the receptaculum seminis show very conclusively that the two forms both belong under this specific name. The variation of the single seta on the fifth foot is frequently encountered in the species. The form of the fifth foot in what she describes as C. bicuspidatus Forbes is very interesting, for Forbes has found it but once and on that occasion from Woods Hole, Mass. It corresponds exactly with the European forms and with the representatives of this species from this locality. C. bicuspidatus Claus is very widely distributed over the United 1914.] NATURAL SCIENCES OF PHILADELPHIA. Day: States. Forbes states that “it has been found in Massachusetts and Wyoming and in all the intervening territory,’ and further that “it is the commonest Cyclops in the Great Lakes.”” Miss Byrnes has recently found it to be a common form in the Long Island waters. I have noted that it occurs very abundantly in one of the large streams in this vicinity, but in the small pond from which I have. taken every other species described in this paper I have found but a single specimen.! Specific Description.—In the specimens of C. bicuspidatus Claus from this vicinity, the lateral angles of the cephalothorax are very prominent, as in the “‘specimens from the far west”? (Forbes). The first thoracic segment is a little more than half the length of the entire cephalothorax (PI. II, fig. 1): The posterior borders of all the thoracic segments are smooth. The first abdominal segment (PI. II, figs. 1 and 5) is unusually expanded, laterally, on the anterior side of the suture. It is about as long as the remaining three abdominal segments. The posterior margins of the first three abdominal segments are finely serrated (Pl. II, fig. 5). The fourth segment has the usual row of spinules on its posterior margin (PI. II, figs. 1 and 2). The stylets (Pl. I, fig. 2) are more than twice as long as the last two abdominal segments. Their length is about six times their width and they are frequently slightly out-curved. The above proportions hold good only in the cases of the specimens examined. They vary considerably in different localities. The lateral spine is inserted at the beginning of the posterior third of the ramus, and at a distance of about one-quarter of the length of the stylet from its insertion point there is always present a minute lateral comb of spinules (PI. II, fig. 2). Of the apical sete, the outermost is heavy, finely plumose, and about as long as the delicate inner seta. The longer of the two prominently developed setz is about equal in length to the abdomen and twice that of the shorter seta. They ‘are both delicately plumose. : The first antennze of the female (Pl. II, fig. 1) are seventeen- jointed and terminate at the end of the first thoracic segment. At the distal end of the twelfth segment there is borne an unusually long, spear-shaped sense-club (PI. II, fig. 3). The setze are all quite short and plumose for the most part. The terminal segments bear 1Jn late March and April, 1910, collections in the vicinity of Cambridge, Mass., this was by far the most abundant form. 26 PROCEEDINGS OF THE ACADEMY OF [Jan., no hyaline plates and the last three gradually increase in length to the end. The armature of the swimming feet is as follows: First pair—outer ramus, two spines, four sete; inner ramus, one seta, one spine, four sete. Second pair—outer ramus, three spines, four sets; inner ramus, one seta, one spine, four sete. Third pair—like second. Fourth pair—outer ramus, three spines, four sete; inner ramus, one seta, two spines, two sete. The fifth foot (Pl. II, fig. 4) is two-segmented, the basal segment being about as long as broad and bearing a plumose seta on its outer distal angle. The distal segment is cylindrical, about twice as long as wide, and bears a long plumose seta and a short, thick spine. For this spine is often substituted a longer seta-like form, though that is not the regular armature in the local specimens. The receptaculum seminis (Pl. II, fig. 5) consists of two divisions. The anterior portion extends as a low arch across the segment. The posterior division is bag-shaped and reaches a point half-way to the posterior margin of the first abdominal segment. Its anterior border branches out abruptly on either side along the suture. The porus is situated on the median line, between the lateral angles on the suture. The egg-sacs are unusually large and are carried at a considerable angle from the body. Forbes gives the size of this species as 1-1.4 mm. Schmeil gives 1.3-2 mm. for the European forms. The specimens from this locality average about 1.5 mm. The color of C. bicuspidatus is generally a very pale shade of yellow. Often individuals appear to be quite colorless. The dorsal surface of the thorax, especially the anterior portion, usually has a peculiarly shiny appearance. The last characters for the distinction of this species are its slim form, the shape of the fifth feet and the receptaculum seminis, and especially the small lateral combs of spinules on the outer sides of the stylets. Subgenus MARCOCYCLOPS Claus. Cyclops fuscus Jurine. Pl. 1. Monoculus quadricornis fuscus Jurine, ’20, pp. 47, 48, Taf. II, fig. 2. Cyclops signatus Koch, ’38, Heft 21, Nr. 8. Cyclops coronatus Claus, '63, pp. 97-99, Taf. II, fig. 16 and Taf. X, fig. 1. Cyclops signatus var. coronatus Herrick and Turner, '95, p. 106, pl. XV, figs. 1-4. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 27 Cuniaes fuscus Schmeil, ’92, pp. 123-127, 136-140, pl. I, figs. 1-76; pl. IV, . 16. Cyclops fuscus Jurine, Marsh, ’95, pp. 16, 17, pl. VI, figs. 5, 7, and 11. Cyclops signatus var. coronatus Herrick, Brewer, 98, pp. 129, 130. Cyclops fuscus Jurine, Lilljeborg, ’01, pl. Ill, figs. 12-15. Cyclops signatus var. coronatus Herrick, Byrnes, 06, pp. 193-200, pl. VII, figs. 1-6, pl. VIII, figs. 1-3. Cyclops signatus var. coronatus Byrnes, ’09, pp. 9, 10. Synonymy and Distribution —In his discussion of the differences between this species and the closely related Cyclops albidus Jurine, Dr. Schmeil has proven not only that we are dealing with two distinct species, but also that Jurine was the first investigator to distinguish these two. It is only right, then, that Jurine’s names should stand and that Koch’s Cyclops signatus give way to (1) C. fuscus and (2).C. albidus Jurine. In spite of Schmeil’s careful proof, several of our American investigators have clung to Herrick’s varieties tenuicornis and coronatus of C. signatus Koch. C.D. Marsh accepted Schmeil’s proof in his paper of ’95. Even after the publication of Forbes’ paper of ’97, in which he abandoned Herrick’s terms for those of Jurine, Brewer in ’98 and Miss Byrnes in ’06 and again in 09, have reverted to Herrick’s varietal names. In speaking of the ‘‘two varieties,” coronatus and tenuwicornis, Brewer states that ‘‘the real differences between them are confined ito the seventeenth joint of the antenna and the caudal sete.’ He then continues: ‘The difference between their first cephalothoracic segments and their furce is hardly distinguishable.” On p. 136 of Schmeil’s monograph there is a table of “‘the most important dis- tinguishing characters’? of the two species in question. Of the eleven “important characters’? mentioned, Brewer has noted four, and no mention is made of such important features as the receptacu- lum seminis and the sense-club (present or absent) on the twelfth antennal segment. It is quite evident that Miss Byrnes has overlooked the more important distinguishing differences between the two species. In her recent paper, The Fresh Water Cyclops of Long Island (09), in spite of the carefully tabulated proofs of Schmeil in Germany and the acceptance of his views by both Marsh (’95) and Forbes (797), Miss Byrnes has clung to the obsolete name of C. signatus Koch and attempts to revive Herrick’s varietal name coronatus and Richard’s(?) annulicornis. In a foot-note (on p. 9) she states: “T have used the more recent name C. signatus var. coronatus instead of the older name C. fuscus Jurine, because it expresses more clearly the evidently close relationship that exists between C. signatus var. 28 PROCEEDINGS OF THE ACADEMY OF [Jan., coronatus and the allied form C. signatus var. annulicornis, called C. albidus by Jurine.”’ It is quite evident that Miss Byrnes has not had Dr. Schmeil’s exhaustive work on the Cyclopide of Germany at hand. He has shown conclusively (129, 130 and 137-140) that by right of priority Jurine’s names should stand. No matter how “close” the “relation- ship” between the two forms in question, Miss Byrnes is hardly justified in using ‘‘the more recent’? name. In her description of Cyclops signatus var. coronatus (p. 9) she states that this form has ‘serrations in the hyaline plate on the two distal segments of the antenne’”’ and, furthermore, that “‘the notches in the hyaline plate of the antenne form gradually and may or may not be present. In fig. 4 they are seen on the last segment only.”” The first statement I am-unable to verify, nor do any of the investigators mention a serration of the “two distal segments of the antenne.’’ They are always present in the hyaline plate of the distal segment in adult forms. In the immature forms the plate on the distal segment of the antenn is often exactly as in C. albidus Jurine. The serrations do not “form gradually.” I have several times observed a young specimen of C. fuscus just before the last eedysis. In such cases there was a finely serrated hyaline plate on the last segment, but the coarsely serrated plate of the adult form could be distinctly observed below the transparent chitin folded flat down along the segment. In every case the serrations of the coming plate were complete. In her conclusion Miss Byrnes states that “there are wide ranges of variability in the reduced seta on the mner ramus of the fourth foot of annulicornis and in the hyaline plate of the antenne of both varieties—in short, in the most important differential characters of the two varieties.” It is not at all surprising that Miss Byrnes considers the ‘two varieties’? so closely related when we find that half of (to her) the “‘most important differential characters’”’ are concerned with a single seta on the inner ramus of the fourth foot. In her description of this species, as well as all the others described by her, Miss Byrnes obviously neither considers the receptaculum seminis a most important feature, nor does she mention the presence of sense-club or hair in a single species. Miss Byrnes has added very little to the evidence of the close rela- tion that may exist between these two species. She gives but one- quarter of the observations in Herrick’s ‘“diagnosis,’’ which is incomplete even in its full form, and “then draws her conelusion mainly from her own notes on the difference in the single seta of the 1914.] NATURAL SCIENCES OF PHILADELPHIA. 29 inner ramus of the fourth foot, already mentioned. I quote from her paper once more. ‘He (Herrick) states that the two varieties have similar armature of the swimming feet, but that tenwicornis differs from coronatus in the absence of serrations on the antennal lamellze and in the divarication of the ovisacs.’’ Here follows the original description of Herrick’s “‘Cyclops signatus var. tenwicornis. First segment of thorax shorter, its length to that of the entire thorax as 1:1.9. Last segment of antennze with unserrated lamella. Caudal stylets longer, length to width as 2.1 :1, inner aspect not ciliated. Outer apical seta half as long as inner. Second segment of antennules longer. Color variable, but always banded or splotched. It is also generally true that the ovisacs in the present variety are more strongly divaricate:than in the variety coronatus.”’ Miss Byrnes now concludes that since “the antennal lamella of annulicornis sometimes bears serrations and coronatus sometimes bears its egg masses in a divaricate position, as I have found in attempting to distinguish the two forms by this character, ‘ consequently Herrick’s diagnosis is untenable.” This gives Herrick absolutely no credit for six of his eight points of difference noted. Herrick’s diagnosis should not be considered untenable, but merely incomplete. This species, while not as common as Cyclops albidus, appears to be quite generally distributed over the United States. Forbes reports it from the ‘‘ ponds and lakes of Wisconsin, Michigan, Illinois, and Massachusetts,” where it ‘‘oecurs sparingly.’ Brewer found it with C. albidus Jurine (= C. signatus var. lenwicornis Herrick) in the vicinity of Lincoln, Nebraska, but “always in small numbers.” Miss Byrnes has studied the species at Cold Spring Harbor, Long Island. Kofoid does not mention it in his studies of the “ Plankton of the Illinois River.” I have found it in this locality; rather more abundantly in September, October, and April than during the winter months. I have never found it in very great numbers. During the summer and fall of 1909, it occurred sparingly in collections from Lake Winnepesaukee, N. H., and in a small pond adjoining “Fresh Pond” at Cambridge, Mass. Specific Description.—The cephalothorax (PI. I, fig. 1) in this spe- cies is a little more than twice as long as the abdomen. They are to each other as 21:10. The first segment is to the entire cephalothorax as 7:11. The length of the thorax to its width is as 11:6. In the living animal none of the lateral angles of the thoracic segments are prominent. The posterior borders of the thoracic segments are all e 30 PROCEEDINGS OF THE ACADEMY OF [Jan., unserrated. The fifth segment has a row of minute chitinous teeth extending transversely across the ventral side between the fifth feet. This same segment has on its lateral sides a row of fine spinules as well as a minute fringe of hairs, as in C. prasinus. The abdomen tapers but little towards its posterior end. The first segment is about as long as the three following ones. The pos- terior edges of the first three abdominal segments are smooth. Occasionally there are very slight and uneven serrations present (Pl. I, fig. 3). The ventral posterior border of the fourth segment has a prominent fringe of spines which do not extend to the edge of the anal opening, as is the case in C. albidus. I find no mention of this characteristic in any descriptions of C. fuscus that I have at hand. The stylets (Pl. I, fig. 3) are short; the branches often slightly out-curved. The length is three times the width. Schmeil states that the inner border is ‘‘densely”’ set with hairs. In the specimens from this locality these hairs are often very irregular, rarely ‘‘dense,”’ but always present. The apical sete are well developed and densely plumose. The outer is to the inner as 4:7. The longest is to the next in length as 7 : 5. The first antennz of the female (PI. I, figs. 1 and 9) reach to the anterior border of the first abdominal segment. They are seventeen- jointed and bear a minute sensory hair upon the twelfth segment in place of the sense-club found in C. albidus (PI. I, fig. 9). On their anterior edges at the point of juncture with the following seg- ments, the eighth, ninth, tenth, twelfth, thirteenth, and fourteenth segments are ornamented with a row of prominent, obliquely set spinules. These rows of spinules form almost a quarter circle on the eighth, ninth, and tenth segments. On the twelfth and thirteenth segments they are not so closely set and are fewer in number. It is interesting to note that where Schmeil found six spinules on these segments in European forms, I have found but four or five, and on the fourteenth segment where he records four, I have noted five in every case. I have never seen more than seven of these spinules on the eighth, ninth and tenth segments, and the eighth and tenth usually have but five. Besides these regular rows of spinules, there are smaller transverse and longitudinal rows and irregular groups of very minute spinules (PI. I, fig. 9) on the ventral side of the first fourteen antennal segments. The longitudinal rows mark the boundary between the smooth portion of the segment and that on which the spinules occur. The three terminal segments bear transparent hyaline plates. cc 1914.] NATURAL SCIENCES OF PHILADELPHIA. 3h These plates are smooth on the fifteenth and sixteenth segments, but on the seventeenth (PI. I, fig. 7) the plate is coarsely serrate from the base of the segment to the insertion point of the middle seta; from thence to the end minutely serrated as in C. albidus. This plate projects somewhat beyond the end of the seventeenth segment. The second antennz (PI. I, fig. 5) are unusually long. The third segment is the longest of the four—a distinguishing difference between this and the following form. The swimming feet are armed as follows: First pair—outer ramus, four spines, four sete; inner ramus, one seta, one spine, four sete. Second pair—outer ramus, four spines, five setae; inner ramus, one seta, one spine, four sete. : Third pair—like second. Fourth pair—outer ramus, three spines, five sets; inner ramus, three spines, two sete. Marsh has noted that “the larger of the two terminal spines of the endopodite of the fourth foot, instead of being serrated on its edges as is customary in all the spines of the swimming feet, is beset on its inner margin with long, rather irregular teeth.”’ I find this character present in the local specimens, though the ‘Grregular teeth’? do not extend to the tip of the spine on its inner margin. There are but five or six of these long serrations at the middle of the inner margin, thence to the tip of the spine the serrations are normal. Schmeil notes that the lamella which connects the basal segments of the fourth pair of swimming feet is ornamented by a fringe of long hairs. This is found also in C. albidus, and is therefore hardly a distinctive character of C. fuscus. In all the specimens that I have examined, these “hairs” are very coarse, resembling rather long serrations. There is’a very noticeable characteristic in the lamella between the basal joints of the third pair of swimming feet. There are two very minute rows of blunt spinules extending transversely across the lamella. The upper row is broken in the middle. In the following species this character is very different. The fifth foot is practically identical with that of the following species. Miss Byrnes states that the basal segment of the rudiment- ary fifth foot in “coronatus” (= C. fuscus) is conspicuously short. Schmeil finds no such difference. Among all the representatives of C. fuscus that I have compared with C. albidus, I have found only very slight differences in the lengths of the basal segments of the fifth feet. The apparent difference in length may often be 32 PROCEEDINGS OF THE ACADEMY OF [Jan., accounted for by the strongly arched cephalothorax of C. albidus, which enables the animal to place the fifth feet close upon the first abdominal segment. In C. fuscus the shorter basal segment is frequently only a foreshortening because of the angle at which the feet are held. The arrangement of the spinules, the spines and the setz are exactly as in C. albidus (PI. I, fig. 13). The receptaculum seminis (Pl. I, fig. 11) consists of two main divisions. The anterior portion is wide, shaped very much like the corresponding portion of the receptaculum in C. albidus, but with a very distinct. indentation on its anterior border. The posterior part appears as a pair of slightly elongated reniform divisions sepa- rated by a median line and fusing at the porus. The color of the receptaculum is always a deep reddish-brown which aecteensly makes it difficult to distinguish its outline. The egg-sacs are carried very close to the abdomen. The eggs are dark; in the living animal they look quite black, but in reality they are a deep shade of brown. This applies only to freshly de- posited eggs, as all Cyclops eggs from which the young are about to emerge show a characteristic salmon tinge due to the color of the bodies of the nauplius. The length of the female varies but little. The following five measurements give an approximate average for C. fuscus from. this locality. All measurements include the caudal sete. No. 1 3.75 mm No. 2 oO. TOM No. 3. ‘ 8-4 Mmm No. 4 3.4 mm No. 5 : 3.58 mm Average sissies ee otitis 20) DOM. Schmeil gives 3.4mm. Brewer’s figures are much smaller-—1.4 mm.— 1.8 mm. The males are often only half as long as the females. An average length is 1.75 mm. The color of the first four thoracic segments and the abdomen from the posterior half of the first segment to the furca is usually a dark green. The fifth thoracic segment, the stylets, and the fourth, fifth, twelfth, thirteenth, and fourteenth segments of the first antennze are blue. The anterior half of the first abdominal segment is reddish- brown. Occasionally there are irregular blotches and streaks of blue on the posterior borders of the first four thoracic segments. The remaining segments are a dirty yellow color, deeper on the anterior border. 1914.| NATURAL SCIENCES OF PHILADELPHIA. 33 C. fuscus may be readily distinguished from all other members of the genus by its size, dark color, and closely lying dark egg-sacs. The absence of the sense-club of the twelfth antennal segment, the coarsely serrated hyaline plate of the seventeenth antennal segment, the form of the receptaculum seminis, and the ciliated inner border of the stylets are the most easily distinguishable characteristics of this species. In order to simplify the comparison of C. fuscus with C. albidus, I have arranged the following tabulated form for the characters of the two species: Cyclops fuscus. Cyclops albidus. (1) Inner borders of the stylets ciliated. smooth. (2) Third segment of second long. short, pear-shaped. antennze (3) Receptaculum seminis (Pl. I, fig. 3.) (Pl. I, fig. 4.) (4) Twelfth segment of first sense-hair. 'sense-club. antennz bears (5) Hyaline plate of seventeenth coarsely serrate. finely serrate. antenne segment (6) Egg-sacs carried close to abdomen. widely divergent. (7) Length of outer fureal seta 4 : 7. ese to that of inner (8) Fureal sete densely plumose. lightly plumose. (9) Posterior border of fourth smooth. finely serrate laterally. thoracic segment (10) Serration of posterior border do not extend to anal extend to the anus. of fourth abdominal seg- opening. ment (11) Lamella of third pair of minute spinules. ‘coarse spinules. swimming feet | (12) Color dark (green and blue). light with occasional | black portions. Schmeil notes differences in the form of the spermatozoa and their position in the spermatophore. The twelve points of difference noted above should enable anyone to distinguish the two. forms positively. 34 PROCEEDINGS OF THE ACADEMY OF [Jan., Subgenus MARCOCYCLOPS Claus. Cyclops albidus Jurine. Pi. I. Monoculus quadricornis var. albidus Jurine, ’20, p. 44, pl. II, figs. 10 and 11. Cyclops signatus var. tenuicornis Herrick and ‘Turner, ’95, pp. 106, 106, pl. XV, figs. 5-7; pl. XX, figs. 1-7; pl. XXXII, figs. 1, 2. Cyclops albidus Schmeil, ’92, pp. 128-132, pl. I, figs. 8-14b; pl. IV, fig. 15. Cyclops albidus Forbes, ’97, pp. 47-49, pl. XIII. Cyclops albidus Lilljeborg, ’01, pp. 49-51, pl. IL, figs. 21 and 22. Cyclops albidus v. Daday, ’06, p. 184. Cyclops signatus annulicornis Byrnes, ’09, pp. 10-13, pl. IV. Specific Description—The first segment of the strongly arched elliptical cephalothorax (Pl. I, fig. 2), is in the proportion of 3 :4 compared with the entire length of the cephalothorax. The lateral angles of the segments are not prominent. The fifth segment is rarely visible from above, owing to the arched form of the cephalo- thorax. The first three segments are smooth along their posterior borders. I find in all of the specimens examined from this locality that the fourth segment has on its posterior border, laterally and not extending to the median line, a row of very minute chitinous serrations. These are only visible when the animal is turned on its side. I find no mention of these serrations in any of the descriptions of this species to which I have access. The fifth segment bears, dorsally situated, three transverse rows of spinules. The last of these is the only complete one and borders the segment posteriorly. The second row does not extend to the median line. The first row is quite short, lateral in position, and the spinules are considerably larger than in the other two. Schmeil states in his note number three, p. 130, that, with two exceptions, these rows of spinules have “never been observed.’”’ In Cragin’s paper of ’83 in his description of Cyclops tenuicornis Claus (= Cyclops albidus Jurine), he says: “Wither side of the fifth thoracic segment is furnished with three transverse rows of serrulations, of which the posterior one is mar- ginal.” Furthermore, he illustrates the point in his Pl. II, fig. 13. His drawing is inconsistent, however, with his description, as it does not show the marginal row extending the entire width of the segment. Forbes, very properly, does not mention the row of blunt spinules between the insertion points of the fifth feet. This is not a “char- acteristic,” as it is found in Cyclops fuscus Jurine. The width of the cephalothorax is to its length as 1:2. Its length to that of the abdomen is as 7:4. (Schmeil reverses these figures and gives abd. : ceph. : : 9 : 5, obviously an oversight). The abdomen is heavy; the first segment tapers only slightly, but the enlargement at the anterior end extends beyond the width 1914.| NATURAL SCIENCES OF PHILADELPHIA. 35 of the last thoracic segment. The second and third segments are cylindrical and their posterior margins are very slightly and unevenly serrated, largely on the under side. The fourth segment has the usual fringe of spines on its posterior edge, and tapers suddenly to the insertion point of the stylets. The stylets (Pl. I, fig. 4) are short, but slightly divergent and smooth on their inner sides. This is an important character which Forbes does not note in his deserip- tion. The proportion of the length to the breadth of each ramus is 3:1. There are four well-developed apical bristles. The longest is to the second in length as 7:5; the outer to the inner as 1 : 3. They are all plumose, but not as densely so as in Cyclops fuscus Jurine. The first antenne vary very little in length. In the female they usually reach to the middle or posterior border of the last thoracic segment. They are seventeen-jointed, the terminal joints attenuated, the last three being each armed with an hyaline plate. The edges of these plates on the fifteenth and sixteenth segments are for the most part entire, but I have repeatedly seen them, especially at the base of the fifteenth segmental plate, minutely serrated. More rarely these serrations extend along the entire edges of all three plates. The plate of the last segment is always finely serrated on its distal half. The point where these serrations. cease and the smooth edge begins is sharply defined by a much deeper notch or indentation (see Pl. I, fig. 8). The twelfth segment bears a well- developed sense-club (Pl. I, fig. 10). Its length is about equal to that of the thirteenth segment. All of the segments, except the three terminal ones, bear an irregularly broken, longitudinal row of minute spinules on their under side. The eighth, ninth, tenth, twelfth and thirteenth segments have each a short row of small cone-shaped serrations at the poimt of juncture with the following segments, as in Cyclops fuscus. The twelfth segment (PI. I, fig. 10) has, in addition, several (usually two) rows of smaller spinules extending parallel to the marginal semicircular row. ; Marsh (95) failed to find these ‘‘ crowns of spines” on the antennz of ‘a large-number of mature females” of C. albidws which he “examined with great care.’”’ He concludes that this peculiar character ‘“‘seems to be rarely true in our forms.” Forbes has found it in the specimens examined by him from many parts of the country. I have never failed to find it in the local specimens. The third segment of the second antenne (PI. I, fig. 6) is short and somewhat pear-shaped. The armature of the swimming feet is as follows: 36 PROCEEDINGS OF THE ACADEMY OF {[Jan., First pair—outer ramus, four spines, four sete; inner ramus, one seta, one spine, four sete. Second pair—outer ramus, four spines, five sete; inner ramus, one seta, one spine, four sete. Third pair—like second. Fourth pair—outer ramus, three spines, five sete; inner ramus, one seta, two spines (inner smooth-edged), two sets (distal one reduced) (see Pl. I, fig. 14). The lower row of spinules on the lamella connecting the basal segments of the third pair of swimming feet is very large and well developed (about twelve coarse spinules). The fifth foot (Pl. I, fig. 13) is essentially like that of C. fuscus Jurine. Lilljeborg, in his Pl. IIT, fig. 21, shows the fifth foot differing from that of C. fuscus (fig. 13) on the same plate by the absence of the rows of spinules on the two segments. The fifth foot consists of two segments. The basal segment is slightly convex on its outer margin, while the inner margin is correspondingly concave (see Pl. I, fig. 13). Toward the inner lateral surface of this same segment there are several, usually three, rows of well-developed spinules. At the outer distal corner it bears a long seta, plumose on its distal half. The distal segment is set well towards the inner side of the lower segment and at the point of juncture is ornamented with a ; circle of small spines. It bears on its tip two heavy spines and a slender seta. The inner spine is slightly longer than the outer. At its base there is a semicircle of quite prominent spinules. Both of these spines are more densely plumose on their inner edges. Be- tween them and borne at the end of a truncated cone-shaped projec- tion is the long middle seta. It is only slightly plumose at its distal end, and these hairs (8-10 on each side) are placed at regularly diminishing intervals. The shape of the receptaculum seminis (PI. I, fig. 12) can be readily noted. The anterior division is almost elliptical when fully dis- tended. The posterior portion is two-lobed and has the form of a low, widely spread letter w. This organ is practically colorless, and for this reason its form can be unmistakably observed. The egg-sacs (Pl. I, fig. 2), carried at a considerable angle from the abdomen, are nearly as long as the abdomen. Ina @ in which the abdomen measured .6 mm. the egg-sacs were .57 mm. long. In discussing the size of this species, Forbes states that “the usual length of the female in America is from 1.26-1.4 mm., but it seems to be much greater (2.5 mm.) in the European representatives 1914.] NATURAL SCIENCES OF PHILADELPHIA. 37 of this species.”’ It is interesting to note the measurement of the following five females taken at random: ING ool 2.5 mm. INGOs an ents eee on 2.51 mm. INIGy Shscakaneeee =e 2.21 mm. INOS Aces. dak . 2.50 mm. No. 5. vs ; 2.47 mm. This gives an average measurement of 2.43 mm. for representatives of the species from this locality—less than .1 mm. difference in size from European forms. Taking Forbes’ figures and the above together, we get an average measurement of 1.88 mm. The color, a blue-gray, occasionally nearly black, is most prominent in the stylets and last abdominal segment, the second, third, first half of the fourth, ninth and tenth segments of the first antennee (@) and in irregular transverse bands, usually four, on the cephalo- thorax. The remaining portions of the animal are nearly colorless or a pale shade of yellow, though I have.frequently noted individuals in which they were a bright blue-green. The species seems to be one of the commonest not only in America, but all over the world. Forbes found it “‘in all localities examined.” In nearly all the lists of crustaceans to which I have access it is recorded as a common but rarely abundant variety. Kofoid records it as ‘‘numerically the least important of the dominant members of the genus in our plankton” (Illinois River). I have found it to be one of the commonest of the Cyclopide. Miss Walker has studied the species in Todd’s Pond, Oregon, where she found it to be, with Cyclops serrulatus, second in abundance after Cyclops prasinus. In collections made during the summer of 1909, at Lake Winne- pesaukee and in the vicinity of Cambridge, Mass.. in March and April, 1910, C. albidus was a common but not an abundant form. Subgenus ORTHOCYCLOPS Forbes. Cyclops modestus Herrick. PI. IJ, figs. 6-11. Cyclops modestus Herrick, ’83a, p. 500. Pa Cyclops modestus Heryick and Turner, ’95, pp. 108, 109, pl. XXT, figs. 1—5- Cyclops modestus Marsh, ’93, pp. 213, 214, pl. V, figs. 10-13. Cyclops capilliferus Forbes, '93, pp. 248, 249, pl. XL, figs. 14-17; pl. XLI, fig. 18. Cyclops modestus Forbes, ’97, pp. 51-53, pl. XV, fig. 4; pl. XVI, figs. 1-3. Cyclops modestus Byrnes, ’09, pp. 26, 27, pl. XI, figs. 4 and 5. Specific Description.—The shape of the cephalothorax is very characteristic in this species (PJ. IT, fig. 6). The first segment reaches 38 PROCEEDINGS OF THE ACADEMY OF {Jan., its maximum width considerably in front of the posterior border, and the anterior border line is somewhat straighter than usual, resembling C. prasinus in this respect. This segment is to the entire thorax as 5:8. The fourth segment is regularly, semicircu- larly excavate on its posterior margin. Forbes finds the posterior edges of the first three segments irregularly notched, but the fourth smooth. I am unable to verify this observation; all the thoracic segments of specimens examined by me had smooth posterior edges. The thorax is about twice as long as wide and one-third as long again as the abdomen. In the male the cephalothorax (Pl. II, fig. 7) tapers only a very little. The first segment is distinctly concave on its anterior border. It expands abruptly in its anterior third, but its middle does not attain the width of the preceding thoracic segment. The first three segments have smooth or faintly uneven posterior margins. The last segment Forbes notes to be ‘‘ peculiar in lacking the usual fringe of spines on the posterior edge.’’ I have noted on either side of the anal opening peculiar, out-curving, hook-like projections of the chitin (PI. II, fig. 8) on the posterior border of the fourth abdominal segment. Otherwise its edge is smooth. It is interesting to note that half-way between the anterior border of the anal opening and the posterior margin of the third abdominal segment there are present in this, as in all the Cyclopide, two peculiar button-like projections whose function may be sensory (Pl. II, fig. 8). The stylets (Pl. II, fig. 8) are rather slim and about twice as long as the fourth abdominal segment. Their length is four times their width. The lateral spine is inserted slightly beyond the middle of each ramus and from this point half-way to the posterior margin of the stylets, they are in the female peculiarly excavate. Along the curving line that marks this character there is a very minute row of spinules. The outer apical seta is rather short and finely plumose. The other three apical sete are all well developed. The outer is to the inner as 4:3. The middle is the longest and is to the next in length as 3:2. Though Forbes has shown the inner borders of the stylets to be densely -and coarsely plumose in his fig. 4 of Pl. XV, he fails to mention this character in his des¢ription. I have never seen a specimen of C. modestus with the stylets as densely plumose as he shows them to be. All of the specimens examined from this locality showed very fine and often unevenly distributed hairs on the inner surfaces of the stylets (Pl. II, fig. 8). They are only visible under a high power. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 39 The female first antenne consist of sixteen segments. Herrick notes a similar form with seventeen joints, but this again I am unable to verify. They extend a little beyond the posterior margin of the first thoracic segment. The second, third, tenth, and thirteenth segments bear unusually long, heavy, plumose setz. In the males these spines, which correspond to those of the third segment in the female antenne, are very conspicuous (PI II, fig. 7). When the first antenne are folded under the body they protrude as shown in the figure. This also occurs when the female antenue are folded under the cephalothorax. The female antenne show an unusual change of direction between the third and fourth segments, giving the animal a characteristic appearance and making the females resemble the males to the naked eye. ‘On the fourteenth and fifteenth segments of the first female antennze occur several characters that may be sensory in function. The fourteenth segment bears a minute sensory hair and a very inconspicuous, transparent, hyaline plate that contains about twelve coarse serrations (Pl. II, fig. 11). The fifteenth segment bears a smooth-edged hyaline plate of which the distal end is club-shaped and protrudes considerably beyond the anterior end of the segment. Forbes (’97) indicated the position of this plate in his fig. 1, Pl. X VI, of the first female antenne, but he seems to have overlooked the character of the organ. In the figure, the positions of the large setze of the fourteenth and fifteenth segments are indicated. The armature of the three segmented swimming feet is very constant. The sets: and the apical spines are long and slender. Forbes notes that “‘the distal segments of the third and fourth pairs of legs turn inward in a way peculiar to this species.’”’ The inner border of the distal segment of the inner ramus of all four pairs is finely plumose. The feet armature is as follows: First pair—outer ramus, four spines, four sete; imner ramus, six sete. Second pair—outer ramus, four spines, five sete; inner ramus, six sete. Third pair—outer ramus, three spines, five sete; inner ramus, six sete. Fourth pair—outer ramus, three spines, five sete; inner ramus, one seta, two spines, two sete. The lamellze connecting the swimming feet have a peculiar form. Pl. II, fig. 10, shows this feature in the third pair. In the fourth pair the place of the seta on the basal joint is taken by a short chitinous 40) PROCEEDINGS OF THE ACADEMY OF [Jan., point. The lamella of the second pair is similar to that of the third, while in the first pair the chitin points have become quite rounded and smooth. The fifth foot (Pl. II, fig. 9) is generally considered as having three joints. Forbes states that it “has three freely movable segments, though the basal one is small.” Miss Byrnes simply states that “the fifth foot is very distinctly three-jointed, and not ‘obscurely’ as Herrick observed.”” The basal segment is certainly very short, though its presence is no longer a matter of conjecture. Whether - it is “freely movable’? I am unable to say. The second segment is almost square and bears a short seta on its outer side. They are both plumose and the larger is borne at the end of a slight projection of the segment. The shorter of these setze is usually folded under the longer, as shown in the figure. The rudimentary feet in this species are unusually large. The receptaculum seminis is very much like that of C. bicuspidatus Claus. The anterior division is very low, extending but a little beyond the porus. The posterior, bag-shaped portion reaches half- way to the posterior margin of the abdominal segment. The egg-saes are narrow and extend a little beyond the ends of the stylets. They usually contain from 10-12 dark ove. Forbes gives 1.2 mm. for the length of C. modestus. I find it slightly larger, 1.3 mm. being an average length for females. The color of this species is most beautiful, violet and lavender shades predominating. It is evenly distributed in the chitin and persists in preserved material. There aré usually a number of large, orange-colored globules below the chitin. These are specially numerous in the cephalothorax and the swimming feet. C. modestus can be readily distinguished from all other species by the sixteen-jointed first antenne, the three-jointed fifth foot, the very characteristic stylets, and the receptaculum seminis. While nowhere an abundant species, C. modestus appears to be very: generally distributed over the United States. Forbes reports it from several localities in Illinois and from Grebe Lake in the Yellowstone Park. Marsh found it in Rush Lake, Wisconsin, and Herrick in Cullman County, Alabama. Miss Byrnes has recently studied the species from the Long Island waters. I have found it, always in small numbers, one of the rarer species of the genus from this locality. In a collection from ‘Fresh Pond,’’ Cambridge, Mass., made in the fall of 1909, this form outnumbered all other species. It occurs 1914.| NATURAL SCIENCES OF PHILADELPHIA. AT also in collections from Lake Winnepesaukee made during July and August, 1909. Subgenus MICROCYCLOPS Claus. Cyclops varicans Sars. PI. IL, figs. 6-10. Cyclops varicans Sars, 62, pp. 252, 253. Cyclops varicans Schmeil, ’91, pp. 33, 34; ’92, pp. 116-118, pl. VI, figs. 1 Cyclops varicans Forbes, ’97, pp. 63 and 64. Cyclops varicans Lilljeborg, '01, pl. IV, figs. 23, 24, pp. Cyclops rubellus Lilljeborg, ’01, pl. IV, figs. 25 and 26, pp. Cyclops bicolor Byrnes, ’09, pl. XIII, pp. 29-31. as fo-l. Synonymy and Distribution—Though I am unable to translate Lilljeborg’s Swedish description of his new species C. rubellus, I am convinced from the Latin synopsis as well as his drawings of the receptaculum seminis and abdomen that this species is synonymous with C. varicans Sars. His drawing of the receptaculwm seminis shows this organ to have the form of that pictured by Schmeil (Taf. VI, fig. 3) with a slightly greater expansion of the “wings” of the anterior division. By a careful study of this organ in a number of specimens, I have concluded that what Lilljeborg pictures as the receptaculum seminis of the type C. varicans (see Taf. IV, fig. 24, of his paper of ’01) is the extreme form of narrow anterior portion, Schmeil’s drawing showing a slightly greater expansion of this same- division. Lilljeborg’s drawing of the receptaculum of C. rubellus and my own (PI. III, fig. 10) of this organ in C. varicans are examples of the extremely wide and wing-like form. The proportional lengths of the caudal sete, as shown by Lilljeborg, are identical with those- of the typical varicans (Pl. ILI, figs. 6 and 7). Miss Byrnes in her recent paper (March, ’09) has described a species of Cyclops under the name of C. bicolor. Her description of the- antenn, which she states ‘‘contain each twelve segments,” at once suggests C. varicans, for nowhere do I find a record of C. bicolor with more than eleven joints in the first female antenne. Schmeil gives eleven, Lilljeborg 10-11, and Marsh has noted a form with ten segmented antenne, though he finds the usual number is eleven. Herrick, with whose description she seems to have compared her own, also gives eleven as the number of female antennal segments. Miss Byrnes’ formula for the swimming feet agrees exactly with that of specimens examined from this locality. In speaking of the variation of this species, she says: ‘“‘The species C. bicolor is usually placed (e.g., by Marsh) among Cyclops having ten or eleven segments. The occurrence, therefore, of a twelve-jointed antenna shows that considerable variation may occur in this organ. Except in this / 42 PROCEEDINGS OF THE ACADEMY OF [Jan., respect, the form from Cold Spring Harbor agrees with Professor Marsh’s description.” The above extract needs no comment. Herrick assigns C. varicans a place among the twelve-jointed-antennze forms having a two-segmented fifth foot. This may account for Miss Byrnes’ error, for Herrick’s drawing of the fifth foot of C. varicans is quite incorrect. C. varicans is one of the rarer species of the genus. Schmeil found it sparingly in the vicinity of Halle. Lilljeborg reports it and describes it from Sweden. Herrick found it but once, and Miss Byrnes also collected it in a single instance from a pond on Long Island, describing it as a twelve-jointed variation of C. bicolor. In his summary of species Forbes records it as “‘a fairly common species throughout the range of Cyclops in North America.’’ I have found it in small numbers in the March and April, 1909, collections in this locality and sparingly in the vicinity of Cambridge, Mass., and from Lake Winnepesaukee collections taken in July and August, 1909. Specific Description—The first segment of the almost elliptical cephalothorax (Pl. III, fig. 6) is about as long as wide and a little over half as long as the entire thorax (3:5). The lateral angles of the third, fourth, and especially the fifth thoracic segments are prominent. The proportion of cephalothorax and abdomen is as 10:7. The posterior borders of all the thoracie segments are smooth. The fifth segment is somewhat flattened and extended laterally. Its lateral edges protrude considerably beyond the first abdominal segment. At the ends.of these wing-like projections of the fifth thoracic segment are inserted long, curving, plumose sete, which are usually considered homologous to the outer sete of the basal segments of the two-jointed rudimentary fifth feet. The fifth feet are inserted at the inner corners of the lateral projections of the fifth thoracic segment on its posterior edge (PI. III, fig. 9). The first abdominal segment is considerably expanded in its anterior half (Pl. III, figs. 6 and 10). The entire abdomen, which is rather slim, tapers gradually to the furea. The posterior borders of the first three segments are smooth. On the posterior edge of the fourth abdominal segment, ventrally placed and rarely extending half way around the circumference of the segment, there is present a short row of long serrations. The stylets (Pl. III, fig. 7) taper slightly and are carried very close together. They are slightly shorter than the last two abdominal segments taken together. The outer apical bristle is rather heavy, sparsely plumose, and nearly as long as the delicate inner one. Of 1914.] NATURAL SCIENCES OF PHILADELPHIA. 43 the two developed apical setz, the inner longer onc is to the outer as 10:7. Schmeil shows both of these setee to be evenly plumose on their distal seven-eighths. This character is rarely present in the specimens from this vicinity. The anterior plumose portion is replaced by a short row of delicate spines on either side of the sete (PI. ILI, fig. 7). These do not merge into the hair-like forms gradually, but end abruptly at the beginning of the plumes as in the long median caudal seta of C. phaleratus. The lateral spine is inserted two-thirds of the length of the stylets from their anterior border. It is usually delicately plumose, although the bare form is not rare. The female first antenne may have either eleven or twelve seg- ments. The number given by Schmeil is twelve, but I have frequently found eleven-jointed sexually mature forms, bearing eggs. Such forms occur more frequently in late winter and the twelve-jointed- - antenn forms in April and May (PI. ILI, figs. 6 and 8). The antenne are a little over two-thirds as long as the first thoracic segment. The division from eleven to twelve joints takes place in the third joint. The fourth, fifth and sixth joints are often narrower than the seventh and eighth. Most of the antennal bristles are not plumose. Schmeil notes the presence of a well-developed, closely lying sense-club on the ninth segment. This I cannot discover on the specimens that have come under my observation. In its place there is a minute sensory (?) hair. The first antenn of the male are peculiar in the unusual develop- ment of the sensory structures of the first division. These are larger than in any other species of the genus and may be readily noted even under a slight magnification as long blue, semi-transparent, narrow bag-shaped structures on the posterior side of the antenne. The swimming feet are all two-jointed. The outer ramus of the fourth pair is generally visible from above in life. This is due to the unusually large lamella connecting these feet which makes them protrude at a greater lateral angle. Schmeil notes that the fourth pair of swimming feet is less fully developed than the other three pairs. In fig. 11 of Pl. III I have shown a foot of the third pair. It is interesting to note an indication of the third segment; a row of short hairs at the middle of the distal segment of the outer ramus and a group of longer hairs on the corresponding segment of the inner ramus. This character is always present in the first three pairs of swimming feet. The armature of the swimming feet is as follows: First pair—outer ramus, three spines, five sete; inner ramus, one seta, one spine, four sete. 44 PROCEEDINGS OF THE ACADEMY OF [Jan., Second pair—outer ramus, four spines, five setee; inner ramus, one seta, one spine, five sete. Third pair—outer ramus, four spines, five setae; inner ramus, one seta, one spine, five sete. Fourth pair—outer ramus, three spines, five setae; inner ramus, one seta, two spines, three sete. The above armature is quite constant. Herrick states the fifth foot is two-jointed. Since he has only collected the species in a single instance, this must be considered an error. There is often a very inconspicuous indication of a former large basal segment (Pl. III, fig. 9), but the movable fifth foot consists of a single cylin- drical joint. At the middle of its distal end there is pore a long, delicately plumose seta (Pl. III, fig. 9). The shape of the receptaculum seminis has already been discussed. Fig. 5 of Pl. III shows what is probably the extreme ‘“‘wing-like”’ formation of the anterior division. The posterior portion is about as long as the anterior and has the form of a short bag. The porus is situated immediately between the two portions on the narrow transverse division that extends entirely across the first abdominal segment. The egg-sacs contain from ten to twelve ove and are carried at a slight angle from the abdomen. The following measurements of six females taken at different times give an average length of .867 mm. INO:8L.. At ys Total le ngth = .895 mm. No. 2 = = .841 mm. No. 3 ne < -= 9662. No. 4 “ “ = 296). No. 5 . = )-60 nine No. 6 = 380) ome The females average .21 mm.’in width. The males are somewhat smaller than the females—.69 mm. being an average oe Schmeil gives .8-.92 mm. for the size of the females and “about” .7 mm. for the males. Sars’ figures are somewhat greater—l mm. " ‘Herrick gives .8 mm. The color of C. varicans is ordinarily very pale. A faint shade of yellow is noticeable throughout the body, making the animal un- usually inconspicuous. It may be readily distinguished from all other species of the genus by its short, twelve-jointed antenne, the very characteristic receptaculum seminis, and the two-jointed swim- ming feet. 1914.| NATURAL SCIENCES OF PHILADELPHIA. 45 Subgenus EUCYCLOPS Claus. Cyclops prasinus Fischer. PI. III, figs. 1-5, Cyclops prasinus Fischer, '60, pp. 652-654, pl. XX, figs. 19-26a. Cyclops fluviatilis Herrick, ’82, p. 231, pl. VII, figs. 1-9. Cyclops magnoctavus Cragin, ’83, pp. 70, 71, pl. ITI, figs. 14-28. Cyclops prasinus Schmeil, ’92, pp. 150-156, pl. V, figs. 1-5. Cyclops fluviatilis Herrick and Turner, ’95, pp. 114, 115, pl. XXVI, figs. 1-8; pl. XXX, fig. 1. Cyclops prasinus Forbes, 97, pp. 57-59, pl. XUX, figs. 1 and 2; pl. XX, figs. 1 and 2. Cyclops fluviatilis Brewer, ’98, pp. 135, 136. Cyclops prasinus v. Daday, ’06, p. 180. Cyclops fluviatilis Byrnes, 09, pp. 28, 29, pl. XV, figs. 1 and 2. Specific Description—The form of the cephalothorax in this, the smallest species found in this locality, is that of an ellipse slightly flattened at the ends. The first segnient is to the total length of the cephalothorax as 5:7. Its length is to its width as 5:4. The lateral angles of all the thoracic segments are obscure. Their posterior borders are unserrated. The lateral edges of the last segment bear each a fringe of very minute hairs (PI. ITT, fig. 5). The abdomen, which is to the cephalothorax as 5:9, is rather slender and tapers only slightly towards its posterior end. The first segment is enlarged at its anterior end and about as wide as the last cephalothoracic segment. The posterior borders of all the abdominal segments are unevenly and minutely serrated. The short stylets (Pl. III, figs. 1 and 2) stand well apart in the living animal. The lateral spines are situated 3 of the length of the stylets from their anterior end. Of the apical sete only two are well developed. The inner and outer bristles are small. Of these the outer is much heavier and is slightly shorter than the inner. Neither attains the length of the stylets. Of the large middle pair the inner is to the outer as 5:3. Brewer gives 4:5 “or equal.” The larger of the two is four times, the shorter about three times, as long as the stylets. Both are delicately plumose. The first pair of (twelve-jointed) antennze (PI. III, fig. 1) reach in the female to the posterior border of the third thoracic segment. Forbes finds them often reaching “quite to the first abdominal segment.”’ The eighth joint is the longest, exceeding that of the two preceding segments. Dr. Schmeil finds on the ninth segment a ‘well-developed sense-club,’’ and, in his foot-note, states that Richard ‘“‘even denied the presence of a sense-club.” All of my specimens agree with Forbes’ description in the absence of the sense-club, but “minute sensory bristle” on the tenth segment. This seems, then, to be undoubtedly a characteristic point of differ- 46 PROCEEDINGS OF THE ACADEMY OF [Jan., ence from the European forms. The three terminal joints are slightly curved and each bears a well-developed hyaline plate. The edges of these plates, which Forbes found to be entire, seem to vary. I have frequently found them finely serrated on the distal third of the terminal segmental plate. Again I have noted a very distinct notch or indentation at the posterior end of the serrations. This resembles somewhat the characteristic ‘‘notch”’ in the corresponding plate of Cyclops leuckarti Claus, though it is not so deep nor con- spicuous. Occasionally I have observed a specimen in which all three hyaline plates were finely serrated. The commonest form has two of the plates with smooth edges, though the plate of the last segment invariably has the slight notch mentioned above (PI. III, fig. 3). The very long and strong set on the first and fourth antennal segments and the change of direction of the remaining segments beyond the fourth, as well as the short caudal stylets, suggest a superficial resemblance to Cyclops modestus Herrick.. The four pairs of swimming feet are armed as follows: First pair—outer ramus, three spines, five setee; inner ramus, one seta, one spine, four sete. Second pair—outer ramus, four spines, five sete; inner ramus, six sete. Third pair—lke second. Fourth pair—outer ramus, three spines, five sete; inner ramus, one seta, one spine, three sete. Both the spines and sets of these feet are peculiarly long and slender. - The rudimentary fifth foot (Pl. III, fig. 4) consists of a single seg- ment, armed with a spine and two sete. The spine is inserted immediately above the characteristic bulge of the minutely plumose inner side. The longer of the two sete is borne at the end of a cone- shaped projection of the distal end of the segment. The remaining seta is inserted on the outer side of the segment at the base of the cone-like process. Its length is two-thirds that of the apical seta and slightly greater than that of the spine. Both the sete are delicately plumose; the spine is more coarsely so. Brewer notes that the “three set” are bare in his specimens. The form of the receptaculum seminis (Pl. ILI, fig. 5) is the most characteristic feature of this species, though it is frequently quite difficult to distinguish owing to the density of the pigment matter in the first abdominal segment. It consists of two very distinct 1914.] NATURAL SCIENCES OF PHILADELPHIA. 47 divisions. The anterior portion has a central arm extending half- way to the upper edge of the first abdominal segment, which branches into two wing-like projections at right angles to the central arm and extending transversely across the abdomen on either side of the median line. The outer ends of these projections are often slightly enlarged. Forbes and Schmeil both find that these side arms are “S-shaped.”’ Among all the specimens that I have examined from this locality, I have noted the ‘‘S shape” in only one instance. In many cases the outer ends of the wing-forms tend to turn up slightly, but the “‘S-shaped”’ canals are certainly the exception and not the rule among the local representatives of this species. Consequently, in my drawing (PI. III, fig. 5) I have shown what I consider a rather more characteristic form of the receptaculum seminis for C. prasinus. This somewhat insignificant detail shows that the receptaculum seminis may be a variable character. The portion of the receptaculum behind the suture consists of two lateral sacs connected by a narrow transverse canal lying close under the suture. In the middle of the anterior edge of this lower con- necting arm is the porus. The anterior division of the receptaculum fuses with the posterior at this same point. The contents of the upper and lower portions differ in appearance only. Dr. Schmeil has proved that the spermatozoa in the upper portion are simply more densely packed together than in the lower. The egg-sacs contain only a few ova. I have never noted more than ten; seven is an average number. They adhere closely to the abdomen, often covering two-thirds of its dorsal area. An average length for the female from this locality is .82 mm. Forbes gives .48—.7 mm. Some of the European measurements are as follows: Vosseler............ a0) mme Richard.............. : .9 mm. (after Schmeil). Vernet... .88 mm. Schmeil..... .8-.9 mm. The males average .62 mm. long and .14 wide. They have a very long sete at the distal end of the first division of the first antenne. The color seems to be quite constant. The first thoracic segment is a pale yellowish-brown with irregular patches of blue-green along its posterior border. The remaining thoracic segments are a deep shade of blue-green. The first abdominal segment is brown, more dense in the anterior half. The remaining segments are irregularly blotched with the thoracic shade of green which becomes solid in 48 PROCEEDINGS OF THE ACADEMY OF : [Jan., the stylets. The first pair of antennz are a brownish-yellow; their anterior border is often tinged with green. Forbes has ‘“‘seen both blue and pink individuals.’’ Herrick found that the color varies “from deep indigo to greenish-brown.’’ Cragin states: “Animal dirty blue-green, antenne lighter. Dark green pigment masses are scattered beneath the integument in various places, particularly along the anterior side of the first antennz.”’ This species is quite generally distributed over the eastern and central United States. Herrick found it in Lake Minnetonka, Minn. Marsh reports the species from Lake Erie, Lake Michigan, and some of the smaller lakes of Michigan and Wisconsin. Forbes has found it in collections from Sister Lake, Fla., and many localities in the State of Illinois. As C. flwiatilis Herrick, Brewer reports it from the vicinity of Lincoln, Nebraska, and Miss Byrnes from Long Island. Miss E. R. Walker reports it as the most abundant form from Todd’s Pond, Oregon. It is one of the most abundant forms in September and October dredgings, but I have noted only a few individuals during the winter months. It is easily distinguished by its small size, its habit of swimming near the surface, its dark color (it frequently appears to be black to the naked eye), and, under the microscope, by the very charac- teristic form of the recepiaculum seminis and short caudal sete. Subgenus PARACYCLOPS Claus. ‘Cyclops phaleratus Koch. PI. IV, figs. 1-4. Cyclops phaleratus Koch, ’35—’41, Heft 21, pp. 8, 9, pl. IX. Cyclops perarmatus Cragin, ’83, pp. 72, 73, pl. I, figs. 9-18. Cyclops phaleratus Schmeil, ’92, pp. 170-178, pl. VIII, figs. 1-11. Cyclops phaleratus Herrick and Turner, ’95, pp. 120, 121, pl. XVII, figs. 1-7; pl. XVII, figs. 2-2d; pl. XIX, fig. 1; pl. XXI, figs. 6-10. Cyclops phaleratus Marsh, °95, pp. 19, 20. a ‘Cyclops phaleratus Forbes, ’97, pp. 59-62, pl. XX, fig. 3. Cyclops phaleratus Lilljeborg, ’01, pp. 105-109, pl. VI, figs. 20, 21. Cyclops phaleratus Byrnes, ’09, pp. 31-33, pl. XIV, figs. 1-9. Specific Description —The first segment of the unusually broad cephalothorax is a little longer than the other four thoracic segments. Its width is slightly greater than its length. The posterior borders of the first three thoracic segments are smooth. The fourth segment is ornamented with a minute fringe of short serrations. The chitinous covering of the fifth thoracic segment is composed of a continuous cylinder like the abdominal segments, and not of a dorsal and ventral plate, as is the case in the corresponding segment of other species of this genus. This segment is armed ventrally along its posterior margin by a row of heavy, cone-shaped teeth (Pl. IV, fig. 4) which ee 1914.] NATURAL SCIENCES OF PHILADELPHIA. 49 extend at a slight angle and not parallel to the body. This row is, furthermore, not continuous, but generally interrupted in the middle for a short distance on either side of the median line. Occasionally, however, I have noted specimens in which this row of spinules is practically continuous, though there is always a slight break. The length of the entire cephalothorax to that of the abdomen is as 7 : 5. The first abdominal segment is very slightly smaller than the fifth thoracic segment, and the whole abdomen tapers but little. How small this taper is may be readily seen from the following pro- portional width of each segment at its posterior border. BTS DRSE STIG Ups et cee lana (ct hk ES nO EDT, Second segment... op tings, acta een see PAG Third segment ee : oles rah edt ee 24 Fourth segment.......... Je cn Meat Nie pee SL The proportional lengths of the four abdominal segments beginning with first are 11:7:6:2. All the segments are cylindrical. The posterior borders of the first, second, and third segments are minutely serrated (Pl. IV, fig. 2). The last abdominal segment is less than one- fifth as long as the first segment and bears on its posterior border a fringe of unusually long and heavy spines (PI. IV, fig. 2). The stylets are short (Pl. IV, fig. 2). Their length and width are to each other as 4 :3—a proportion that shows them to be very wide. Below the point of insertion of the short lateral spine, which may or may not be plumose, the stylets taper rapidly. In addition to this lateral spine, there is a row of slightly shorter spmules (usually from- four to six) extending from the lateral spine, ventrally, slightly below the middle of the side of each stylet. Forbes mentions ‘‘a row of long spinules on the ventral side of each ramus, extending from the middle line of the anterior border to the point of insertion of the lateral spine.”” I have never noted such a row of spinules in any of my specimens, though this entire minutely spinose armament of the stylets must be considered a variable feature. There are two or three rows of minute hairs extending obliquely from the middle of the anterior border of each stylet towards the posterior border. The interior border of the stylets is often plumose (Schmeil found an “unbehaarten Innenrande”’) and the inner and posterior dorsal surfaces are armed with irregular groups and rows of short, blunt spines (Pl. IV, fig. 2). Forbes and Schmeil both note that the outermost apical bristle ‘‘is placed high up on the side of the stylet.”’ In his fig. 1 on pl. VIII, Schmeil has shown this to be the case, but in fig. 2 of the same plate he shows this spine inserted directly beside 4 50 PROCEEDINGS OF THE ACADEMY OF [Jan., the shorter of the two developed caudal sete. I have found it only as he shows in fig. 2 (Pl. IV, fig. 2). This spine is often quite blunt and always densely p:umose on both sides. At the point of insertion there are several long, dorsally placed serrations extending a third of the distance about the base of the spine. The innermost bristle is very delicate, about as long as the outer, and plumose on its outer side only. Of the two well-developed median bristles, the inner is from two and one-half to three times as long as the outer. Its first third is bare, the middle portion fringed with small spines, and the last third finely plumose. It is longer than the abdomen and un- usually wide at its insertion point. The smaller of the two developed caudal sete is bare for one-quarter of its length, thence to its tip it is fringed with a row of small spines on the outside and delicately plumose on the inner. Between the insertion point and the beginning of the outside row of spines there is often a minute fringe of hairs. The dorsal median seta is quite slender, a little longer than the outer apical spine, and not plumose. The first antennie of the female (PI. IV, fig. 3) are eleven-jointed in most cases, though the ten-jointed form appears occasionally. Schmeil records only the ten-jointed form from Germany, and Lilljeborg the same from Sweden in his paper of 1901. Forbes states that they “may be either ten- or eleven-segmented,”’ and Miss Byrnes has noted a single individual in which the left antenna had eleven, the right but ten segments. I have noted a similar individual in a single case. Herrick found that ‘‘the antenna is usually ten-jointed, but frequently is eleven-jointed (?), and is much shorter than the first thoracic segment.’’ The question-mark seems to indicate some doubt as to the existence of the eleven-jointed form, though subsequent workers have verified his observation in the case of the American representatives of this species. It is interesting to note here that the eleven-jointed form has, as far as I am able to ascertain, been recorded only by the American investigators. In the eleven-jointed form there is borne at the distal end of the eighth segment, in place of a sense-club, a minute sensory bristle. The antennz taper but little in the first nine joints, the last two being considerably narrower than the others. The last joint bears an unusually strongly developed seta. The armature of the other antennal joints is quite uniform. The length of the female first antenne is a little more tham half that of the first cephalothoracic segment. Schmeil states the first antenns of the male are normal. I have 1914.| NATURAL SCIENCES OF PHILADELPHIA. 51 noted that there is on the last division an unusually large spine, set well towards the distal end. There is also a small blunt spine witli a large comb-like fringe on its outer margin, placed about in the middle of the second large division of the male antennze. The re- maining spines are rather smaller than in other species. On the joints of the first division there are present, on the under side, the usual long, blunt, finely plumose sensory structures. The second antennz are very short and broad. The long curving seta of the basal segment I have never found to be evenly plumose as Schmeil shows it. It is usually armed with a row of short spmules on each side of its distal three quarters, while the first quarter bears, only on its upper side, an uneven row of longer spinules. The second segment bears on its outer side a double row of delicate spines, and on its upper side a fringe of spinules and near its distal end a peculiar spine. The distal half of this spine is curved and bears on its upper side a comb-like fringe of fine spinules. The smaller spine on the distal end of the third segment has this same peculiar curve and fringe. The curved sete at the top of the fourth segment are short and very broad. The four pairs of swimming feet are all three-segmented, both spines and sete are well developed, and the outer edges of the first and second segments of each ramus bear rows of heavy spinules. The armature is most variable. For example, I have found on the outer ramus of the first pair either three or four spines and five sete. Again in the second pair on the corresponding ramus I have found four spines with either four or five sete. In another case I have found on the outer ramus of the fourth foot of the right side, three spines and five setze and on the same ramus of the opposite side four spines and four sets. Furthermore, the armature of the inner ramus of the second and third pairs of swimming feet differs consistently from that given by Forbes, in the presence of an extra seta on the inner side. This seta is very obscure and may bé overlooked on account of the long spinules on the same side of the ramus, but its existence can be positively identified by its position and by the fact that, though it is-often not much larger than the accompanying spinules, it is the only one that is plumose. I give this rather lengthy dis- cussion of the swimming feet merely as an example of the armature not being constant nor reliable as a point of differentiation of species. (See Introduction.) The following is an average armature for the local specimens: : First pair—outer ramus, three spines, five setae; inner ramus, one seta, one spine, four sete. s 52 PROCEEDINGS OF THE ACADEMY OF {Jan., Second pair—outer ramus, four spines, five sete; inner ramus, one seta, one spine, four sete. Third pair—outer ramus, four spines, five sete; Inner ramus, one seta, one spine, four sete. Fourth pair—outer ramus, three spines, five sete; inner ramus, one seta, two spines, two sete. The fifth feet (Pl. IV, fig. 4) are merely flange-like projections of the fifth thoracic segment. They are moré lateral than ventral, and there is ordinarily no line marking the position of a former segment. Occasionally, however, I have noted a faint indication of a possible segment, extending from the base of the outer spine obliquely to a point below the inner spine where the large row.of serrations con- necting the two fifth feet ceases (see description of cephalothorax and fig. 4, Pl. IV). The fifth foot is armed with three nearly equal spines, of which the inner is coarsely plumose, the middle one more finely plumose, and the outer bare or finely and sparingly plumose on its outer side. Schmeil’s criticism of Brady’s “otherwise excellent drawing,” in which he shows all three spines to be plumose, is hardly justifiable (see note 2, p. 176, of Schmeil’s monograph). In addition to the three large spines there are several small serrations about the base of the mner spine and an uneven mass of similar small serrations laterally below the outer spine. The receptaculum seminis consists of two nearly equal divisions extending as narrow bands transversely across the entire first abdominal segment. The porus is situated in the middle of the short, median, common portion. Owing to the dark color of this species, the structure of the receptaculum is extremely difficult to observe. The egg-sacs are borne close to the abdomen and frequently extend quite beyond the end of the stylets. They usually contain from ten to twenty large dark eggs. The oviducts are of unusual interest in this species. While in all other species they terminate within the cephalo- thorax, in C. phaleratus they extend as a blind duct to the anterior border of the fourth abdominal segment (PI. IV, fig. 1). The following measurements of six females covering collections from the same spot during two years, give an average length of 1.73 mm. No. 1 : ; Total length = 1.6 No. 2 : . Fes So BO ee he) Ss * = 1.95 NO. o.. ER Ren i! Sg okie AE eos : zs cs = 1.88 No. 4 ate ce rh ga = 1.78 NO: See: Pa? _ + i = 71-61 No. 6 : ee te ee eee ee ee “e 4 = 1:6 . 1914.] NATURAL SCIENCES OF PHILADELPHIA. 53 The males are only a little smaller—1.6 mm. being an average length for local specimens. The coloring of C. phaleratus is most brilliant. The ground-color; which in this species is directly in the chitin, is a brick-red. The second thoracic segment, the last abdominal segment, the stylets and the caudal sete and spines, the swimming feet, the mouth parts, second antenne, and the last segment of the first antennze are sky-blue, which varies in intensity. Often these parts appear quite colorless. The egg-saes are dark blue or lavender in the first stages of development. C. phaleratus is a pelagic species. In aquaria it may often be found a little above the water line, where it sometimes crawls even beyond the upper margin of the meniscus line. Its swimming motion is a rapidly darting one. The easiest way of distinguishing it from all other members of the genus is by its superficial resemblance to the genus Canthocamptus. The short, eleven-jointed antenn, and form of the rudimentary fifth feet, are certain microscopic points of identification. This species seems to be very widely distributed in America, though nowhere is it particularly abundant. Forbes reports it from several localities in Illinois and Wisconsin and from Portage La Prairie, Manitoba. Marsh has found it in several of the Michigan lakes and Cragin reports it from Cambridge, Mass., as Cyclops perarmatus. Miss Byrnes has studied the species on Long Island, where she found it in ‘shallow, fresh-water ponds.’’ I have noted it rather more abundantly in the spring collections, but never in great numbers from the vicinity of Haverford, Pa., and from a small spring-water pond near Gillette, Wyoming, as well as in collections from Lake Winne- pesaukee and the vicinity of Cambridge, Mass. Subgenus PARACYCLOPS Claus. Cyclops fimbriatus var. poppei Rehberg. PI. IV, figs. 5-11. Cyclops poppet Rehberg, ’80, p. 550, Taf. VI, figs. 9-11. Cyclops fimbriatus Schmeil, ’91, pp. 35, 36. Cyclops fimbriatus var. poppet Schmeil, ’92, pp. 168-170, Taf. VII, figs. 14-16. Cyclops fimbriatus Herrick, ’95, pp. 121, 122, pl. XVII, figs. 8, 9; pl. XXI, fig. 11; pl. XXV, figs. 9-14. Cyclops fimbriatus var. poppet Forbes, ’97, pp. 63 and 65. Cyclops fimbriatus Byrnes, ’09, p. 33, pl. XV. Synonymy and Distribution —In his Beitrdge zur Kenntniss, etc., of 791, Schmeil considered Cyclops poppet Rehberg, synonymous with the typical C. fimbriatus. The following year, however, after a more careful study of the species, though he still claimed the differences to be too few to warrant a new species, he granted that 54 PROCEEDINGS OF THE ACADEMY OF [Jan., there should be a variety of C. fimbriatus, which he named C. fimbriatus var. poppet Rehberg. All of Herrick’s drawings of the stylets of what he describes as C. fimbriatus Fischer show that his specimens belonged to the var. poppet, and not to the typical form. Forbes recognizes the variety poppei, but gives no description or figures in connection with his note of the occurrence of the species. Miss Byrnes, following Herrick’s mistake, has described this variety as the typical C. fim- briatus from Long Island. Her fig. 5, pl. XV, of the stylets with their characteristic armature of a longitudinal row of spines proves that the animal examined by her was not fimbriatus, but the variety poppet of Rehberg. There has been not a little confusion of the above-mentioned two forms. I cannot find a single description of | the typical C. fimbriatus Fischer by any American investigator. The species seems to be represented in this country by the variety only. This has been described at least twice under the name of the type form. Cyclops fimbriatus var. poppei seems to be one of the rarer species of the genus. Several of the investigators have failed to find it. Kofoid states in his Plankton of the Illinois River, ““E. B. Forbes (797) records in May, September, 1896, C. varicans Sars as common, and C. fimbriatus var. poppet Rehberg and C. bicolor Sars as rare.”’ Forbes states that this is ‘‘a rare species in Manitoba, Alabama, and the north central States.’”’ Brewer does not record it from the waters about Lincoln, Nebraska. Marsh, in 95, names “ fimbriatus”’ in his Key to Species of Cyclops, but gives no description. Miss Byrnes has studied the species from Long Island waters, where she states that “Cyclops fimbriatus has been taken in great numbers— especially in the collections made in the early spring.”” This obser- vation agrees very closely with my own regarding the variety poppei in this vicinity. In a collection made in the spring (March) of 1907 it was the most abundant form. I have never recorded it from September to January. In February, 1909, I found a few egg-bearing females. I consider it one of the rarest of the members of the genus found in this locality. Specific Description—The somewhat slim cephalothorax (Pl. IV, fig. 5) tapers only slightly posteriorly. The first segment is half as long as the entire thorax. The dorso-ventral diameter is short in proportion to the length of the animal. In a specimen measuring 1.13 mm. it was but .14 mm. The width of the three posterior thoracic segments diminishes but a little. The lateral angles of the 1914.] NATURAL SCIENCES OF PHILADELPHIA. 55 thoracic segments are not prominent. Both Herrick and Miss Byrnes have obviously drawn their specimens under pressure, when, as I have repeatedly observed, the thoracic outline is entirely changed. Fig. 5, Pl. IV, shows the form of cephalothorax in life. The third segment is finely serrated along its entire posterior margin (fig. 10, Pl]. 1V). These serrations often are slightly larger at the end of the marginal row. The fourth thoracic segment bears, laterally, a short row of minute but rather coarse “hairs” (fig. 11, Pl. 1V). On the corresponding portions of the fifth segment there are present similar short rows of coarse hairs, but on this segment they are much larger (fig. 8, Pl. IV). The entire cephalothorax is to the abdomen as 7 : 5. Its length to its width is as 2: 1. ~The abdomen is wide, the first segment being but a little narrower than the fifth segment of the thorax. In my drawing of the living animal (fig. 5, Pl. IV) the first abdominal segment is foreshortened owing to the curve of the entire dorsal surface. This fact, taken together with the short first antenne and the method of locomotion, suggests a strong resemblance to members of the genus Cantho- camptus. The first three abdominal segments are finely serrate on their posterior margins. These serrations in the fourth segment, instead of stopping at the sides of the anal opening, turn anteriorly and extend a short distance along either side of this opening. There are, furthermore, between the upturned portion of the posterior row of serrations and the side of the anal opening, two very minute rows of what Schmeil ealls: ‘“button-like projections” (fig. 6, Pl. IV). Neither Herrick’s nor Miss Byrnes’ drawings give an accurate view of this armature, though the former evidently noticed this detail from his fig. 11, Pl. 21, of his report of ’95. Miss Byrnes’ drawing shows the marginal row of serrations of the fourth abdominal segment extending across the anal opening. All the abdominal segments -bear transverse rows of minute indentations of tne form of the marginal serrations, but not projections of the cuticula. This may easily be proved by turning the animal on its side, when the dorsal and ventral lines of the abdominal segments will appear as unbroken lines. I find this character quite constant in specimens of the variety poppei, though Schmeil notes that it is often missing in the type fimbriatus. The stylets (fig. 6, Pl. [V) differ from those of the typical fimbriatus. They are only as long as the last two abdominal segments. Their length is three times their width and their inner margins almost meet at the point of insertion in the abdomen. One of the main differences 56 PROCEEDINGS OF THE ACADEMY OF [Jan., between this form and the type lies in the characteristic armature of the dorsal surface of the stylets. The lateral spine is inserted well towards the median line. On the outer lateral surfaces and a little below the position usually occupied by the lateral spine there is a row of prominent serrations. The position and course of this row of serrations may be seen in fig. 6 of Pl. 1V. There are two well- developed apical setze which are carried prommently divaricate. Of these two, the outer is a little over half as long as the mner and is ornamented on its outer surface with minute spines, its inner surface bearing the usual hairs. Herrick’s statement, ‘“‘imner two-thirds as long as the outer,’’ is undoubtedly another case of reversed propor- tions. The longer seta is plumose. The delicate mmermost sete are as long as the outer and generally curved as in fig. 6, Pl. LV. The outermost setz are very heavy, rather blunt-and delicately plumose on their inner surface only. About the base of each there is a ¢ circle of long serrations. The first antenne of the female (fig. 7, Pl. IV)-are eight-jointed. They are only half as long as the first thoracic segment and are carried at right angles to the median line, as is the case with practically all of the Cyclopide. The segments taper rapidly, the distal one being but one-quarter as wide as the first at its line of juncture with the second. The fourth segment is the longest. The two distal segments are slightly bent forward in life (fig. 5, Pl. IV). Herrick states: ‘The basal joint with a small semicircular series of fine bristles.”” This is not a characteristic of this species, but of the entire genus. Miss Byrnes’ description is this: ‘‘The antenne contain but eight segments; they are short and are characterized by two well-developed sete.’’ The fifth segment bears a well- developed sense-club (fig. 7, Pl. IV). This is somewhat different from Schmeil’s figure of this organ for the type form. In the speci- mens that I have examined I find it to be rather more slim and spear-shaped than ‘“‘club-shaped”’ as in the type. It suggests the form of the corresponding organ in C. bicuspidatus Claus (PI. II, fig. 5). Many of the sete of the first seven segments are delicately plumose, but those of the terminal segment are, as far as I have been able to observe, quite bare. Along the line of the distal third of the third segment (fig. 7, Pl. [V) I have repeatedly noted what appears to be the beginning of the segmentation of another antennal joint. However, I have never observed a specimen of ‘var. poppet” with nine-jointed first antennw. The semicircle of fine bristles at the base of the first joint is unusually prominent. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 57 The antennules are characterized by the two very short terminal segments. The armature of the three-jointed swimming feet is as follows: First pair—outer ramus, three spines, five*setee; inner ramus, one seta, one spine, four sete. Second pair—outer ramus, four spines, five setze; Imner ramus, one seta, one spine, four sete. Third pair—like the second pair. Fourth pair—outer ramus, three spines, five sete; immer ramus, one seta, two spines, two sete. The inner margins of all the branches of the swimming feet are ornamented with rows of bristles. This is especially the case in the second, third, and fourth pairs. In her description of the variations of this species, Miss Byrnes says: ‘‘ The most striking variation in C. fimbriatus is a tendency toward a reduction in the armature of the inner ramus of the first swimming feet, where in the apical position, for example, a single large spine replaced two sete or a spine and a seta, which is the typical armature of the organ.’’ She does not give the ‘‘typical armature” in her table, and the small seta on the inner side of the inner ramus of the first pair of swimming feet she has probably overlooked. The fifth foot (fig. 8, Pl. IV), which is one-jointed, I find bearing the following armature: On the inner side of the distal end of the single segment there is a heavy spine, serrate on its ner surface and finely but not densely plumose on its outer margin. The row of minute serrations which Schmeil shows to be present in the type form at the base of the corresponding spine (see his Taf. VII, fig. 12) I do not find at all after a careful examination of a number of speci- mens. There is a single spine-like projection at the base of the inner spine, but no indication of a row. On the outer distal corner of the fifth foot there is a very coarse spine-like seta that is densely plumose on its inner and outer side. These two “spines”’ (the outer is more nearly a spine than a seta) are of equal length. Between them and borne at the end of a button-like process is the middle seta, slightly longer than the two spines and plumose on its distal half. Rehberg noted that “das rudimentire Fiisschen ist mit zwei gleichlangen Dornen und einem kiirzeren Haar besetzt, wahrend. sich bei Cycl. fimbriatus nur ein kurzer Dorn und zwei lange Haaren befinden.”” Schmeil admits that Rehberg’s description and drawing. is quite accurate, states, however, that the fifth feet in type and. variety are identical. I am inclined to agree with Rehberg that 58 PROCEEDINGS OF THE ACADEMY OF [Jan., the outer seta of the type form is replaced in his variety by a spine. There has been much confusion over this point. Herrick states that the fifth foot has ‘‘three spines.”” Miss Byrnes agrees in her description with Schmeil, for she finds “a coarse inner spine and two setze.”’ This is undoubtedly one of those details of structure which may vary considerably, and I merely give the observations of the investigators to show their opinions on this point. The receptaculum seminis (fig. 4, Pl. [V) is of a very characteristic form. It consists of two portions. These are in the shape of long, closely lying narrow ellipses. The upper portion has a slight inden- tation in the middle of its anterior margin and the ends are slightly upturned. The porus is situated in the middle of the short common portion which connects the two divisions of this organ. The struc- ture of the receptaculum seminis has never before been observed in the American representatives of this variety. The egg-sacs contain but a few dark ove, 7-10, and are carried close to the abdomen in life. I have found that 1.17 mm. is an average length for females of the variety poppei. Schmeil gives .86 mm. and Herrick .8 mm. The former states that “die Varietiit ist etwas kleiner als die typische Form.” For the ‘‘typische Form’? Schmeil gives for the female “092-1 mm.”’ The smallest mature female that I have measured was 1.13 mm. long—longer than the typical European form. Schmeil’s observations on the size of this species does not seem to hold good, then, for the American.forms. In fact, the reverse is true of specimens from this vicinity. Further study of the species will undoubtedly show that type and variety are of a size. The variety poppet is practically colorless. Irregular chains (fig. 5, Pl. IV) of rose-colored globules are scattered occasionally throughout both cephalothorax and abdomen, and the ovaries, when distended, appear a deep shade of lavender. The most readily distinguished characters of this variety are the short eight-jointed antennze (which show that it belongs either to the type or variety) and the very characteristic armature of the dorsal surfaces of the caudal stylets. The following table illustrates the main points of difference between type and variety: 1914.] NATURAL SCIENCES OF PHILADELPHIA. 59 | c : . Y. fimbri 5 C. fimbriatus Fischer. C. fimbriatus var | (1) Fifth foot with a spine and two sete, with two spines and a seta, the spine circled at its) one spine plumose, and base by a row of minute) no row of spinules at the spinules. | base of either. (2) Receptaculum semi-‘“‘may reach the anterior narrow, low lying, ellipse nis | border of first abdominal] (fig. 9, Pl. IV), with | segment” (Schmeil). slight indentation in mid- dle of anterior border. (3) Fourth abdominalserrations stop at side of serrations turn up (fig. 6, segment anal opening. DAI (4) Transverse row ofjform semicircle about sty- extend longitudinally (fig. serrations on sty-| lets stopping above lat- 6, Pl. IV). lets eral spine. (5) Stylets narrow, set far apart,shorter, inner margins | almost equal to last three almost — meet, length | abdominal segments. | equal to last two abdomi- | nal segments. Schmeil notes further minute differences in the structure of the male first antenne. N. B.—Since the compilation of the above table I have had occasion to examine collections from Lake Winnepesaukee, N. H., in which a form occurs which bridges over the gap between type and variety as regards character No. 4. In these forms the transverse row of serrations on stylets is exactly as in the type form as figured by Schmeil, but the proportional length of stylets agrees with var. poppet and not with the type form. List or CycLopipm RECORDED FROM A SINGLE SMALL POND IN THE VIcINIrY OF HAVERFORD, Pa. (The above small pond is on the estate of Mr. McFadden, directly on the south side of the Philadelphia and Western Railroad tracks and to the right of the bridge that crosses these tracks at Haver- ford station.) Genus CYCLOPS. I. Subgenus CYCLOPS Claus s. str. 1. Cyclops viridis var. insectus Forbes. Not described in this paper, but an abundant form. ‘2. Cyclops bicuspidatus Claus. II. Subgenus MACROCYCLOPS Claus. 8. Cyclops fusous Jurine. 4, Cyclops albidus Jurine. 60 PROCEEDINGS OF THE ACADEMY OF {Jan..,. IV. Subgenus ORTHOCYCLOPS Forbes. 5. Cyclops modestus Herrick. V. Subgenus MICROCYCLOPS Claus. 6. Cyolops varicans Sars. VI. Subgenus EUCYCLOPS Claus. 7. Cyclops serrulatus Fischer. Not described in this paper, but the most abundant form. 8. Cyclops prasinus Fischer. VII. Subgenus PARACYCLOPS Claus. 9. Cyclops phaleratus Koch. 10. Cyclops fimbriatus var. poppei Rehberg. In the above list, subgenus III is omitted since it only contains C. ater Herrick, a form not found thus far in this locality. This arrangement is taken directly from Forbes’ paper of ’97. Description of C. viridis var. insectus Forbes and C. serrulatus Fischer are not given in this paper. The latter form is perhaps the commonest one of the genus and will not be mistaken for any other since no other American form approaches it in morphological details. Forbes has discussed the synonymy in his paper of ’97, but since that time several European investigators have established new species and varieties all intimately related with the type form. A careful comparative study of these new forms will be necessary before the description of C. serrulatus Fischer can be brought up to date. It is known to be a most variable form. Size and stylet proportions are rarely constant in individuals taken at the same locality and even in the same collection. BIBLIOGRAPHY. Brewer, A. D. ’98. A Study of the Copepoda found in the Vicinity of Lin- coln, Nebraska. Studies from the Zool. Lab. of the University of Nebraska, Lincoln, Nebraska, No. 29, Article XIII, pp. 119-138, pl. VII. Byrnes, E. F. ’03. Heterogeny and Variation in some of the Copepoda of Long Island. Reprinted from the Biol. Bulletin, Vol. V, no. 3, August, 03. pp. 152-168, 5 figs. —— ’06. Two Transitional Stages in the Development of Cyclops signatus var. coronatus. Reprinted from Biol. Bulletin, Vol. X, no. 5, April, "06. —— 09. The Fresh Water Cyclops of Long Island, Cold Spring Harbor Mono- graphs, No. VII, March, 09. Pub. by the Brooklyn Inst. of Arts and Sciences. Fifteen plates, 43 pp. CuicuKorr, G. ’06. Copépodes d’eau douce de Bulgarie, Zool. Anzeig., Band 31, pp. 78-82. Criaus, GC. ’63. Die freilebenden Copepoden, mit besonderer Beriicksichtigung der Fauna Deutschlands, der Nordsee und des Mittelmeeres. 230 pp., 37 pls. Leipzig. Craaty, F. W. ’83. A Contribution to the History of the Fresh-water Copepoda. Trans. Kansas Acad. Sci., Vol. VIII, pp. 66-80, pls. I-IV. 1914.) NATURAL SCIENCES OF PHILADELPHIA. 61 Dapay, E. v. ’97. Resultate der wissenschaftlichen Erforschung des Bala- tonsees. IX Section—Crustaceen, 31 pp., 40 drawings. Budapest, ’97. —— 03. Mikroskopische Siisswasserthiere aus der Umgebung des Balaton. Zool. Jahrbiich., Abt. f.’Syst. ete., vol. XTX, Heft I. —— '06. Untersuchungen iiber die Copepoden-fauna von Hinterindien, Sumatra und Java, nebst einem Beitrag zur Copepodenkenntnis der Hawaii-Inseln. Pp. 175-206. 3 Taf., Zool. Jahrbiich., Abt. f. Syst. etc., Bd. 24, Heft 3, Jena. Douwsr, C. v. 799. Zur Morphologie des rudimegtiiren Copepoden-fusses. Zool. Anzeig., Bd. 22, pp. 447-450. —— 03. Zur Kenntnis der freilebenden Siisswasser Copepoden Deutschlands —Cyclops crassicaudis Sars. Zool. Anzeig., Bd. 26, pp. 463-465. —— 07. Zur Copepodenfauna von Java und Sumatra. Zool. Anzeig., Bd, 32, pp. 357-364. Forses, EB. B. ’97. A Contribution to a Knowledge of North American Fresh- water Cyclopide. Ill. State Lab. of Nat. Hist., Vol. V, Article II, pp. 27-82, pls. VIII-XX. Fores, 8. A. ’82. On some Entomostraca of Lake Michigan and Adjacen Waters. Amer. Nat., vol. XVI, pp. 537-543, 640-650, pls. VIII, IX. ; Herrick, C. L. ’83. Heterogenesis in Copepod Crustacea. Am. Nat., Vol. XVII, pp. 208-212. — ’g3a. Heterogenetic Development in Diaptomus. Am. Nat., Vol. XVII, pp. 381-389, 499-505, pls. V-VII. Herrick, C. L., and Turner, C. H. ’95. Synopsis of the Entomostraca of Minnesota. Geol. and Nat. Hist. Surv. Minn., Zodl., Series II, 525 pp. SI plates. Koror, C. A. ’08. The Plankton of the Hlinois River, Part IT. Constituent Organisms and their Seasonal Distribution. Bull. Ill. State Lab: of Nat. Hist., Vol. VIII, art. 1. Leumann, H. 03. Variations in Form and Size of Cyclops brevispinosus Herrick and Cyclops americanus Marsh. Trans. Wise. Acad. of Sci. Arts and Letters, Vol. XIV, part I, pp. 279-298, pls. XXX—XXXIII. ; LitisesorG, W. 701. Synopsis Specierum hue usque in Suecia Observatorum Generis Cyclopis, sive Bidrag Till en Ofversigt af de Inom Sverige Iakttagna Arterna af Slaktet Cyclops. 118 pp., VI Tab. Stockholm, ’01. Marsu, C. D. 95. On the Cyclopidee and Calanide of Lake St. Clair, Lake Michigan and Certain of the Inland Lakes of Michigan. Bulletin of the Michigan Fish Commission, No. 5, 24 pp., 9 plates. —— °03. The Plankton of Lake Winnebago and Green Lake. Wisc. Geol. and Nat. Hist. Survey, Bull. No. XII, Series No. 3. Ranpotrpn, H. ’00. Chloretone .... an Anesthetic and Macerating Agent for Lower Animals. Reprinted from Zool. Anzeig., Bd. XXIII, No. 621. Rarasun, M. J. ’05. Fauna of New England, 5. List of Crustacea. Occa- sional Papers of the Boston Society of Natural History, VII. Boston. Say, T. 718. An Account of the Crustacea of the United States. Journ. Acad. Nat. Sci. of Phila., Vol. I, pp. 421-458. Scuuei, O. ’91. Beitrige zur Kenntnis der Siisswasser Copepoden Deutsch- lands, mit besonderer Beriicksichtigung der Cyclopiden. Zeitschr. f. Natur- wiss., Halle, Bd. 64, pp. 1-40. 99.’ Deutschlands freilebende Siisswasser-Copepoden. I. Teil. Cyclopide. Bibliotheca Zoologica, Heft II, 191 pp., 8 Tafeln. —— °93. Copepoden des Rhiitikon-Gebirges, 40 pp.,4 pls. Abhandl.d. Naturf. Ges. zu Halle, Bd. XIX. Waker, E. R. ’08. Observations on the Microfauna of an Oregon Pond. Transactions of Amer. Microsc. Soc., Vol. XXVIII, pp. 76-84, pl. VI. EXPLANATION OF PLATES. Prare 1.—Cyclops fuscus Jurine. Cyclops albidus Jurine. _ Fig. 1—C. fuscus, adult female. Oc. 2, obj. 1. (Ocular and objective numbers refer to Leitz lenses unless otherwise noted. All drawings were made with the help of the camera lucida.) a ae lel, 62 PROCEEDINGS OF THE ACADEMY OF (Jan., Fig. 2.—C. albidus, adult female, showing the characteristic divaricate position of the egg-sacs. Oc! 1, obj. 1 (Bausch & Lomb). Fig. oe aspect of stylets and last abdominal segment of C. fuscus- a Oc. 3, obj. 3 Fig. 4. Stylets and last abdominal segment of C. albidus. Oc. 1, obj. 3 (B. & L.) Fig. 5.—Three proximal joints of the second female antenna, C. fuscus. Oc. 2, obj. 3. Fig. 6.—Third and fourth joints of the second female antenna, C. albidus. Oc. 1, obj. 3 (B. & L.). Fig. 7.—Terminal segment of first female antenna, C. fuscus, showing the typical deep serrations of the hyaline plate. Oc. 2, obj. 3. Fig. 8 —Terminal segment of first female antenna, C. albidus. The finely serrated plate is characteristic. Oc. 3, obj. 5. Fig. 9—Twelfth segment of the first female antenna of C. fuscus, shoWie the armature of minute serrations and the sense-hair. Oc. 2, obj. Fig. 10.—Twelfth segment of the first antenna of a female C. albidus Late rows of ‘‘thorns”’ and sense-club. Oc. 2, obj. 5. Fig. 11.~—Receptaculum seminis of C. fuscus. Oc. 3, obj. 3. Fig. 12.—Receptaculum seminis of C. cael Oe. 3, obj. 3 Fig. 13.—Fifth foot of C. albidus. Oc. 2, obj. 5. Fig. 14.—Distal segment of the inner ramus of the fourth pair of swimming feet of C’. albidus, showing a group of small hairs in place of the second seta of the inner side. Oc. 0, obj. 5 Puate II.—Figs. 1-5, Cyclops pendants Claus. Figs. 6-11 Cyclops modestus, 4 Herrick. : Fig. 1—An adult female (slightly extended by pressure). Oc. 0, obj. 3. Fig. 2 r stylet showing the minute serrations near the anterior end. Oc. 3, obj. 5. Fig. 3.—The sense-club of the twelfth segment of the first female antennz.- Oc. 3, obj. 7. Fig. 4.—A foot of the fifth pair. Oc. 3, obj. 7. Fig. 5—The receptaculum seminis. Oc. 3, obj. 3. Fig. 6.—An adult female. Oc. 0, obj. 3. Fig. 7.—An adult male with first antenn folded under the cephalothorax. Oc. 0, obj. 3. Fig. 8. —Stylets and last-abdominal segment. Oc. 2, obj. 5. Fig. 9.—A foot of the fifth pair. Oc. 1 (B. & L.), obj. 7. | Fig. 10.—Lamella of the third pair of swimming feet. Oc. 1 (B. & L.), | obj. 5. Fig. 11.—The three terminal joints of the first female antennwe showing | hyaline plates and sense-hair. Oc. 1 (B. & L.), obj. 5. Puate IiI.—Figs. 1-5, Cyclops prasinus Fischer. Figs. 6-11, Cyclops varicans Sars. Fig. 1.—An adult female. Oc. 2, obj. 3. Fig. 2.—The stylets and last abdominal segment. Oc. 3, obj. 5. Fig. 3.—The three terminal joints of the female first antenna showing the characteristic hyaline plates. Oc. 2, obj. 7. Fig. 4.—The fifth foot. Oc. 2, obj. 7. Fig. 5.—The receptaculum seminis. Oc. 3, obj. 5 Fig. 6.—A mature female of C. varicans with eleven-jointed first antenn : a winter transitional form. Oc. 2, obj. 3. Fig. 7.—The stylets and last abdominal segment. Oc. 2, obj. 5. Fig. 8.—An eleven-jointed antenna of the first pair in the female. Oc. 3, eee obj. 5 Fig. §.—"The ae thoracic segment showing the rudimentary fifth feet. Oc. 3, obj. 5 Fig. 10.—The Peceminion seminis. Oc. 0, obj. 5. ‘ a Fig. 11.—A_two-jointed swimming foot of the third pair. The division of the third joint is always mdicated by minute rows of hairs. Oc. 2, obj. 5. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 63 Prats 1V.—Figs. 1-4, Cyclops phaleratus Koch. Figs. 5-11, Cyclops fimbriatus var. poppet Rehberg. Fig. Fig. Fig. Fig. 1—An adult female. Oc. 0, obj. 3. 2.—The stylets and last abdominal segment. Oc. 0, obj. 5. 3.—The female first antenna. Oc. 2, obj. 5. 4.—The fifth foot. Oc. 2, obj. 5. Fig. 5.—An adult female; the first cephalothoracic segment appears some- what foreshortened. Fig. Fig. Fig. Fig. Fig. Fig. 6.—The stylets and last abdominal segment. 7.—The female first antenna. 8.—A foot of the fifth pair. 9.—The receptaculum seminis. 10.—Posterior serrated margin of the third thoracic segment. 11.—Posterior margin of the fourth thoracic segment. 64 PROCEEDINGS OF THE ACADEMY OF [Jan., A STUDY OF THE SPECIES OF THE GENUS DICHOPETALA (ORTHOPTERA: TETTIGONIID#). BY JAMES A. G. REHN AND MORGAN HEBARD. The possession of the extensive series of this genus secured by us in the southwestern United States in the summers of 1910 and 1912, with the acquisition by the junior author of the very important representation of the same group contained in the Bruner Collec- tion, prompted us to make a detailed study of this interesting but previously little-known genus of long-horned grasshoppers. The scope of our work became so extended that practically all the material of the genus in American collections was finally examimed. The few types contained in European collections were relatively unimpor- tant. Our work has required the description of a number of new forms and the synonymizing of several old ones. The color descriptions have been based on Ridgway’s recent set of color standards.1 DICHOPETALA Brunner. 1878. Dichopetala Brunner, Monogr. der Phaneropt., p. 77. 1891. Dichopetala Brunner, Verhandl. K.-K. Zool.-bot. Gesell., Wien, XLI, p. 4. 1897. Dichopetala Saussure and Pictet, Biol. Cent.-Amer., Orth., I, p. 315. 1900. Dichopetala Scudder, Proc. Davenp. Acad. Nat. Sei., VIII, p. 67. 1900. Dichopetala Rehn, Trans. Amér.-Entom. Soc., X XVII, p. 88. 1901. Dichopetala Rehn, Entom. News, XII, p. 207. 1902. Dichopetala Rehn, Trans. Amer. Entom. Soc., X XVII, p. 335. 1902. Dichopetala Scudder and Cockerell, Proc. Davenp. Acad. Sci., LX, p. 51. 1902. Dichopetala Morse, Psyche, IX, p. 381. 1906. Dichopetala Kirby, Synon. Catal. Orth., II, p. 388. 1907. Dichopetala Rehn, Proc. Acad. Nat. Sci. Phila., 1907, p. 56. 1909. Dichopetala Rehn and Hebard, ibid., 1909, p. 167. 1912. Dichopetala Hunter, Pratt and Mitehell, Bull. 113, Bureau of Entom. U.S. Dept. of Agric., p. 50. This genus was based on two species—mexicana and emarginata Brunner. GenotyPE: Dichopetata mexicana Brunner (selected by Kirby, 1906). The genus is a member of the Phaneropterine and of the group Odonture, constituting with the genera Odontura Rambur, Pseudi- sotima Schulthess, Epiphlebus Karsch, Atlasacris Rehn, Peropyrrhicia ‘Color Standards and Color Nomenclature. By Robert Ridgway. Wash- ington, D. C., 1912. 1914.| NATURAL SCIENCES OF PHILADELPHIA. 2 65 and Angara Brunner a section of the group. Of these genera all are exclusively Old World except Angara, which is Brazilian. Generic Description—Fastigium of vertex more or less compressed, short, not or distinctly sublamellate, not at all suleate or finely suleate proximad, more or less in contact with facial fastigium. Antennz subcrassate proximad, from two to five times the length of the body. Pronotum not at all, or more or less constricted mesad, dorsum more or less arcuate in transverse section; caudal margins of lateral lobes more or less arcuate or subtruncate. Tegmina in male abbreviate; anal field extending nearly the entire length of tegmen; sutural margin at apex of stridulating vein obtuse-angulate to rectangulate produced. Tegmina in female very short, not reaching or distinctly surpassing the caudal margin of the metanotum, overlapping, subcontiguous, or more or less decidedly remote from one another; distal margin of female tegmina arcuate to truncate. Abdomen more or less dilated; disto-dorsal abdominal segment with distal margin emarginate, bisinuate, truncate or arcuate, supra- anal plate simple or (in o@ of tauwriformis) bearing a dorsal erect T-shaped structure. Cerci of male incurved, acute, falciform, simple, with dorsal margm rarely serrato-dentate or with median tooth or lobe on dorsal or external face, occasionally with an accessory digitiform lobe from base. Subgenital plate of male broad and short or produced, more or less narrowed distad, free lateral margins coneave, subparallel or converging, distal margin truncate or more or less deeply and completely V- or obomegoid emarginate, unicari- nate or tricarinate ventrad. Ovipositor from one and one-half to three times the length of pronotal disk, more or less arcuate, apex more or less acuminate and with its margins serrato-dentate. Sub- genital plate of female emarginato-truncate, arcuato-emarginate, or more or less completely divided into two halves, these more or less acute distad. Limbs more or less elongate. Cephalic femora from one and one-half to three times as long as the disk of the pronotum in the male, one and one-third to two and one-half times in the female. Caudal femora from four to nearly seven times the length of the pronotal disk in the male, from four to five and three-fourths times in the female. Classification —From a systematic standpoint, the characters of greatest value in the differentiation of the species are: in both sexes, general form of the body and shape of the eyes; -in the male, form of the pronotum, form of the tegmina, form of the cerci and subgenital plate; in the female, form of the pronotum, form of the tegmina, 3) 66 PROCEEDINGS OF THE ACADEMY OF [Jan., relative size and width of space between the tegmina, form of the ovipositor and that of the subgenital plate.. The general form is much more robust in some species than in others, the females almost always more robust than the males, in falcata and pollicifera less different in this respect than in the other forms. The outline of the eye is, in a few cases, of assistance in distinguishing females of closely allied forms, as castanea and brevihastata. The pronotum ranges from not at all constricted, to decidedly constricted mesad in both sexes. In the male sex the tegmina show modifications in the form of the margins and the width of the fields, the promimence of the stridulating vein and the projection of the sutural margin at the apex of the same vein. The tegmina of the female are as diagnostic as the more complex appendages of the male, their relative position and the interspace between the same, as well as the form of their margins, being of importance. The characters of the genitalia of the two sexes are discussed in detail below. Morphological Notes on Male Genitalia.—The variation in struc- tural form in the cerci of the male covers a number of types which show six different lines of development, relatively as follows: mexicana falcata | —— durangensis | castanea brevihastata | ( gladiator \ emarginata oreaca catinata caudelli r— tridactyla E—tauriformis C—pollicifera B—-serrifera The position of durangensis is more or less problematical, as we have only nymphal males. 1914.| NATURAL SCIENCES OF PHILADELPHIA. 67 The extremes in structural variation in the form of the cerci are a simple incurved falciform type, found in the group A, and one with a median tooth and an accessory lobe from the base, found in group F. The general cercal structure of the various groups can be pre- sented best in tabular form. mexicana | _. : Group A | faleata j Simple, falciform. Group B (serrifera).—Simple, dorsal margin serrato-dentate. Group C (pollicifera).—With an external median tooth. Incertze sedis (dwrangensis)—With evidence of a dorsal median tooth. {castanea brevihastata | With a dorsal median fork, developing gladiator from a simple tooth to a large flattened Group D ; : A emarginata | lobe covering the greater portion of the orececa distal section of the cercal shaft. catinata Group E (tawriformis) —With a greatly developed lobiform median tooth arising from the external margin of the shaft, the distal portion of the latter peculiarly modified. A transverse proximal lamella present dorsad on the shaft. tridactyla | With a dorsal median tooth and an accessory Group F | caudelli | digitiform lobe from the base of the shaft. The species in group D exhibit a regular development in the form of the dorsal median fork from a simple median tooth which becomes depressed and flattened, spreading laterad until it is as wide as the proximal portion of the shaft, to the other extreme which has it modified into a great inverted spoon-like plate covering the greater portion of the cereal shaft. The peculiar digitiform accessory appendage of tridactyla and caudelli springs from a proximal trans- verse ridge, which is apparently homologous with the more decided transverse lamella found in the same region in tawriformis. From the evidence of eighteen immature males, belonging to six species (durangensis, brevihastata, gladiator, oreeca, catinata, and pollicifera), it is evident that the separation of the median fork of the cercus is never accomplished before the mature condition. Of brevihastata and pollicifera we have material representing two con- secutive instars, one preceding the mature condition, the other species being represented by this stage alone. In the forms of which we have two stages no indication of the lobe is apparent in 68 PROCEEDINGS OF THE ACADEMY OF [Jan., the earlier instar, while in the other stage, in all of the species repre- sented in this condition, there is a more or less distinct indication of an incipient lobe or tooth, this being most pronounced in dwrangensis and catinata. In the latter this embryonic lobe is more definitely formed than in durangensis, consisting of an ovate vertical area of relatively large size. In the closely related oreaca, the incipient lobe is not vertical, but horizontal in position. The male subgenital plate is very varied in form, the distal margin ranging from truncate with lateral styliform processes to obomegoid emarginate; the general form broad with the distal portion little produced and narrowed, the lateral angles more or less blunted, to an opposite extreme, elongate, narrow, concave laterad with the lateral angles acute, between which extremes are a number of modifi- cations of one or the other. Quite curiously, there exists no correla- tion between certain forms of cerci and certain forms of the subgenital plate, forms nearly related in cereal structure, as orewca and catinata, having very different subgenital plates. Morphological Notes on Female Genitalia—The ovipositor ranges in general form from the elongate, very slender, decidedly arcuate type seen in gladiator, and the elongate robust type with a more or less straight ventral margin as found in a number of species, to a short, moderately arcuate form seen in castanea and brevihastata. It is evident that there is considerable individual variation in the depth of the ovipositor, this being very apparent in those species represented by considerable series, so much so that the extremes have different facies, but the major portions of such series always bridge the apparent gaps. In ovipositor length there is marked variation in gladiator and brevihastata, this being most apparent in the former species, the extremes of which are quite different in appearance. We have before us ten female nymphs which we can positively refer to five species (dwrangensis, brevihastata, gladiator, oreewca, and polli- cifera). Of durangensis we have represented the second instar preceding maturity, of brevihastata the two preceding maturity, and of the other three species the instar preceding maturity. From this material it is evident that the development of the ovipositor is very rapid, but in no case do the external margins acquire distal teeth until the mature condition is reached. In one specimen which is appar- ently on the eve of the last ecdysis (the type of levis) the teeth of the enclosed ovipositor can be seen through the sheath when it is held to the light. The subgenital plate of the female presents great diversity in 1914.] NATURAL SCIENCES OF PHILADELPHIA. 68 development, which in its details are not always correlative with apparent affinities. These diversities can be placed in two categories, one (mexicana, falcata, durangensis, castanea, and brevihastata) with the plate entire, the other (comprising the remaining species) with it divided completely in two, at least as far as the chitinous portion is concerned. In the first section we have from an extreme which is very broad and short, with the distal margin emarginato-truncate, to one of a similar general form with the margin bisinuate to arcuato- emarginate. In the second section we have even greater diversity, the paired lobes varying from broad to very narrow, blunted to aciculate, the general form of the margins differing to a lesser degree. In the forms with an entire subgenital plate, the distal margin has a different appearance when the plate is flat or when it is compressed; which factor should always be considered in determining the character of this margin. For the sake of uniformity, we have endeavored to give the tharacter of this margin from the plate were it flattened out. Notes on Tegminal Structure—In the male the tegmina are more ample in oreeca and more reduced in size in tridactyla than in the other species. The angle of the sutural margin is very greatly pro- duced in fridactyla and on the other hand almost imperceptible in catinata. The stridulating vein is apparent in all the forms of the genus, but variable in strength and curvature, while the tympanum is also of variable form and definition. In the female the considerable variation in form and position indicated in the generic description is not correlated with the general relationship of the forms, as certain species with overlapping quadrate tegmina and others with nearly contiguous similarly shaped tegmina occur in sections of the genus which on sum total of characters are well removed from one another. The reduction of the female tegmina has proceeded further in emarginata than in any other form of the genus, as there they are decidedly lateral and very small, while the development of the tegmina in the same sex is most marked in falcata, where they are overlapping, covering all of the metanotum and the greater portion of the proximal dorsal abdominal segment. The venation in the female tegmina is always generalized, being more complex in falcata than in any of the other forms. Color Pattern.—The color pattern of all of the forms of this genus is similar in several respects; first, in the possession of pale paired lines extending from the eye caudad to the apex of the abdomen and, second, in the general uniformity of the lateral and ventral color. In the majority of the forms the color of the dorsum between the pale 70 PROCEEDINGS OF THE ACADEMY OF [Jan., lines is more or less uniform and, for convenience in describing the extent and character of the pattern, we have referred to this as the dorsal color, the ventral and lateral tones as the lateral color, and the pale paired lines and their developments as the pale pattern. The range of tone in all three of these principal components of the colora- tion is very considerable, the extreme on one hand having the pattern intense, the contrasts decided and the tones darker and richer, while in the other extreme the pattern is dilute, the contrasts poor and the tones paler and weaker. To facilitate reference to these extremes we have termed them the intensive and recessive extremes. In the recessive condition the pale pattern is frequently much restricted as well as weakened, while the dorsal color is often but little, in part only, or not at all different from the lateral color. Distribution Extending from north-central Texas (Dallas), southern New Mexico (Dry Canyon and Mesilla Valley) and central southern Arizona (Tumamoc Hill and Syeamore Canyon), south to the upper Rio Balsas Valley m1 Guerrero, Mexico, on the west reaching Tepic and on the east the vicinity of the coast at Corpus Christi and Brownsville, Texas, and Tamos, Vera Cruz, Mexico. Vertically the genus ranges up to at least 6500 feet (in the Davis Mountains, Texas). It reaches its greatest diversity in southern’ Texas and the northern and central parts of the Mexican tableland. History.—In 1878, Brunner? erected the genus for two species then described, viz., mexicana (from Mexico) and emarginata (from Texas). In 1880, Bormans’ described a species from Schoa, Abyssinia, as Dichopetala massaiew, which has since been placed in the genus Peropyrrhicia, which is exclusively African. Seudder, in 1900, described* a Dichopetala brevicauda from California, which we now know to be an Arethea and not at all related to Dichopetala. In 1901, Rehn® described a new form from Mexico as D. pulchra, basing it on material which he had previously recorded as mexicana. Scud- der, in 1902, in Scudder and Cockerell’s list of New Mexican Orthop- tera® described as new a species of the genus from New Mexico, calling it Dichopetala brevicauda, but as that name was preoccupied, Morse, at Scudder’s suggestion, renamed the species D. brevihastata.’ In 1907, Rehn described a species from Arizona as D. levis.$ ® Monogr. der Phaneropt., p. 76. * Ann. Mus. Civ. Stor. Nat., Genova, XVI, p. 218, fig. ‘Canad. Entom., XXXII, p. 331. § Entom. News, XII, p. 207. © Proc. Davenp. Acad. Sci., UX, p. 51. 7 Psyche, IX, p. 381. § Proc. Acap. Nar. Sci. Puiva., 1907, p. 56. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 71 Material.—In the preparation of the present paper the types of the following species have been before us: (Dichopetala pulchra Rehn, synonym of D. mexicana Brunner.) Dichopetala falcata n. sp. Dichopetala serrifera n. sp. Dichopetala durangensis n. sp. Dichopetala pollicifera n. sp. Dichopetala tauriformis n. sp. Dichopetala. castanea n. sp. Dichopetala brevihastata Morse. (Dichopetala levis Rehn, synonym of D. brevihastata Morse.) Dichopetala gladiator n. sp. Dichopetala oreeca n. sp. Dichopetala catinata-n. sp. Dichopetala tridactyla n. sp. Dichopetala caudelli n. sp. ’ The entire series of the genus examined by us numbers 362 speci- mens. The great majority of these (239) were taken by the authors on recent trips and are located in the Hebard Collection and that of The Academy of Natural Sciences of Philadelphia. Of the remainder of the representation we have had before us, 37 specimens were from the Hebard Collection ex Brunner; 27, comprising the entire series of the genus in the United States National Museum, were examined through the kindness of Mr. A. N. Caudell; 50, forming the entire series in the Scudder Collection, were either loaned or made accessible to us by Dr. Samuel Henshaw, of the Museum of Comparative Zoélogy, and a few specimens each were loaned by the authorities of the Field Museum and the American Museum of Natural History. To the above-mentioned gentlemen and the authorities of these museums we wish to express our hearty thanks for their assistance in the work. We have, with their co- operation, been able to examine almost all of the material in America on which the records of the genus were based. Aside from the typical material of the two original species of the genus, no recorded specimens of the group exist in other collections. KEY TO THE SPECIES. Males. A.—Cercus subfalciform, non-fureate. B.—Cercus non-serrate. C.—Length of pronotum equal to one-third that of cephalic femur. Subgenital plate produced into lobes.............. mexicana Brunner. 72 PROCEEDINGS OF THE ACADEMY OF [Jan., CC.—Length of pronotum equal to two-thirds that of cephalic femur. Subgenital plate not produced into lobes... falcata n. sp. BB.—Cercus with the dorsal margin serrate. (Subgenital plate broadly subtruneate with decided lateral substyliform PPCM GALES.) ese cin Peace ta na rene serrifera Nn. sp. AA.—Cercus with a median fork (either tooth or lobe). (No access- ory digitiform lobe from base of cercus.) B.—Subgenital plate little produced. (Cercus with median tooth simple. Hyes OVate.) wcrc Castaneda n. sp. BB.—Subgenital plate distinctly produced meso-caudad. C.—Subgenital plate greatly produced. Cereus with the tooth very long, as long as the remainder of the shaft, and needle-like distad......... tauriformis n. sp. CC.—Subgenital plate moderately produced. Cereus with the tooth not as long as the remainder of the shaft and not needle-like distad. *“D.—Cereus with the median tooth blunt and simple. (Byes elliptical.)..... cn ..Drevihastata Morse. DD.—Cercus with the median tooth depressed and lamellate. E.—Cerecus with the median lobe (7.e., tooth) acute, not rounded when seen from the dorsum, - placed on the external margin of the cercus. Pronotum little constricted mesad. General coloration green..... coe pollicifera n. sp. EE.—Cercus with the median lobe generally rounded when seen from the dorsum, placed on the dorsal face of the cercus. Pronotum mod- erately constricted mesad. Coloration varie- gated. ; F.—Subgenital plate with the distal margin weakly emarginate and the lateral angles ilunted’s...4) eae gladiator n. sp. FF.—Subgenital plate with the distal margin decidedly emarginate and the lateral — acute (variable in degree) (?) durangensis n. sp. G.—Median lobe of cereus decidedly shorter than the proximal half of the cercal Sha toocccccccceneenn emarginata Brunner. GG.—Median lobe of cercus at least as long as the proximal half of the cereal shaft, spoon-like in shape and inverted over the shaft. H.—Margins of the cereal lobe converging distad, apex hardly truncate, ven- tral margin of the lobe decidedly cingulate vee OP CCCH Ni. Sp. 1914.| NATURAL SCIENCES OF PHILADELPHIA. 73 HH.—Margins of the cereal lobe hardly converging distad, the apex sub- truncate, ventral margin of the lobe weakly cingulate ........catinata n. sp. AAA-—Cercus with a dorsal median tooth and an accessory digiti- form lobe attached at the dorsal base. B.—Tegmina shorter than the pronotum, the portion of the anal field of former distad of stridulating vein very brief, sutural margin at apex of this vein decidedly produced. Cereus with the median tooth proportion- ately longerv.......... ee OC OCLY Omen SDs BB.—Tegmina longer than the pronotum, the portion of the anal field of former distad of stridulating vein normal, sutural margin at apex of this vein slightly produced. Cercus with the median, tooth proportionately shorter, caudelli n. sp- Females. A.—Ovipositor very decidedly longer than the head and pronotum together.® B.—Tegmina slightly overlapping mesad. C.—Ovipositor hardly or not at all longer than half the length of the caudal femora. Subgenital plate not produced laterad into large trigonal lobes. D.—Size large (body 21.5 mm., pronotum 6.9, ovipositor 14.5). Ovipositor slenderer................ falcata n. sp. DD.—Size medium (body 15.5-18.2 mm., pronotum 4.1-4.3, ovipositor 10.2-10.5). Ovipositor more TODUS bes cece ceantnrnmnciianninnmn OUTONgensts D. ‘Sp. CC.—Ovipositor distinctly longer than half the length of the caudal femora. Subgenital plate produced laterad into large trigonal lobes. (Size medium; pronotum NOt Sellate.) ..n.ccsweecernneenenneernennennn nd Uriformies D. Sp. BB.—Tegmina not attingent or subattingent mesad. C.—Subgenital plate compressed, truncate, shallowly arcuato-emarginate or biconvexo-emarzinate distad. : mexicana Brunner. CC.—Subgenital plate with distal margin very profoundly triangularly emarginate or broadly divided to the base. D.—Form slender, subcompressed. Pronotum narrow, elongate (of the type usual in the genus). Limbs and ovipositor proportionately longer than in the opposite category. E.—Subgenital plate with lateral apices very acute, more or less spiniform. 2 Occasional specimens of D. brevihastata, which belong to the opposite cate- gory, have the ovipositor appreciably though not decidedly longer than the head and pronotum. These specimens are exceptional and do not: represent the average condition of the species. 74 PROCEEDINGS OF THE ACADEMY OF [Jan., F.—Ovipositor with ventral margin nearly straight except at distal extremity. Teg- mina small................emarginata Brunner. FF.—Ovipositor with ventral margin more (gladia- tor) or less (orewca) arcuate. Tegmina medium sized. G.—Size large. Ovipositor slender. Limbs very elongate.............. oow..gladiator n. sp. GG.—Size medium. Ovipositor more robust. Limbs moderately elongate..ore@ca n. sp. EE.—Subgenital plate with lateral portions trigonal, but apices blunted, non-spimiform. (Ventral margin of lateral lobes of pronotum distinctly sinuate. Eyes relatively prominent.)............... catinata n. sp. DD.—Form compact and robust. Pronotum not elon- gate, short. Limbs and ovipositor proportion- ately shorter, the ovipositor robust. E.— Ovipositor not quite half as long as the caudal femora; subgenital bigs with lobes rotundato- angulate.. ovum tridactyla n. sp. EE.—Ovipositor equal to or slightly more than half as long as the caudal femora; subgenital plate with lobes quite acute....caudelli n. sp. \A.—Ovipositor slightly or not at all longer than the head and pronotum together. B.—Tegmina separated hy pack! or quite their own width. (Eyes ellipticuls)ssemeanare rane brevihastataM. orse. BB.—Tegmina separated by muc ch less than their own width. C.—Form more robust. “Ovipositor elongate, slender pollicifera n. sp. CC.—Form slenderer. Ovipositor quite robust. castanean. sp. Dichopetala mexicana Brunner. 1878. D{ichopetala] mexicana Brunner, Monogr. der Phaneropt., p. 77, pl. I, fig. 6. [Cuernavaca, Morelos, Mexico.] 1897. Dichopetala mexicana Saussure and Pictet, Biol. Cent.-Amer., Orth., ip. old: 1900. Dichopetala mexicana Rehn, Trans. Amer. Entom. Soc., X XVII, p. 88. [Rio Cocula, Guerrero, Mexico.] ; 1901. Dichopetala pulchra Rehn, Entom. News, XII, p. 207. [Rio Cocula, Guerrero, Mexico.} This species needs comparison only with D. falcata (vide infra), from which the male can be immediately separated by having the subgenital plate exserted in lobes, the female by having the tegmina not overlapping and both sexes by the proportionately more elongate limbs. Types: o& and 2; Cuernavaca, Morelos, Mexico.” [Brunner Collection.] 1914.] NATURAL SCIENCES OF PHILADELPHIA. , 75 We here describe a topotypic female from the Hebard Collection. (data: Cuernavaca, Mexico; January 4, 1899). Description—Size medium; form subecompressed. Head with the occiput subglobose, strongly descending to the antennal fosse; fastigium compressed, short, subcultriform-lamellate dorsad, apex rounded when seen from the side and not projecting cephalad of the Fig. 1.—Dichopetala mexicana Brunner. Lateral outline of topo- typic female. (X 2.) antennal scrobes, almost touching the facial fastigium; face, gene, clypeus and labrum glabrous; eyes ovate in basal outline, moderately prominent; antennze incomplete. Pronotum with the greatest dorsal length subequal to the greatest ventral width (across lobes) of same; dorsum of pronotum with the impressed transverse sulcus distinct, broadly V-shaped, placed mesad, the dorsum slightly constricted at the same point; cephalic and caudal margins of disk subtruncate; lateral lobes of pronotum with the greatest depth contained one and two-thirds times in the greatest dorsal length of the same, ventral margin Nala subtrunecate, cephalic and caudal angles of the same _. ; margin rounded, transverse sulcus marked only on the : ae dorsal portion of the lobes and there descending ob- ™exicana liquely ventro-cephalad. Tegmina very short, reaching en as the caudal margin of the metanotum, much broader outline _ of than long, the greatest length contained twice in the aupeen sal greatest width, distal margin broadly arcuate, disto- topotypic costal and disto-sutural angles broadly rounded, sutural ae pe margins narrowly separated. Abdomen heavy; cerci very short, crassate, substyliform, the apex rather sharply attenu- ate; ovipositor heavy, robust, the length half that of the caudal femora, dorsal margin moderately arcuate, ventral margin straight for the median three-fourths, strongly arcuate proximad and distad, 76 PROCEEDINGS OF THE ACADEMY OF [Jan.,. the proximal half of the whole ovipositor in consequence tapering to the middle, thence subequal in width to the subacute apex, dorsal margin strongly serrato-dentate for a third of its length from the: apex, with nine to ten teeth, ventral margin armed in a similar fashion for a fourth of its distal length, with seven to eight teeth;. subgenital plate moderately transverse, lateral margins broadly con- vergent, distal margin broadly and shallowly arcuato-emarginate, the lateral angles forming very short rectangulate lobes. Cephalic femora with the length slightly greater than the dorsal length of the ovipos- itor, slightly less than the ventral length of the ovipositor; cephalic tibiz slightly longer than the femora, spined on all the margins, for- amina elliptical. Median femora very slightly longer than the cephalic pair. Caudal femora about two and one-third times the length of the cephalic femora, moderately inflated proximad, ventral margins unspined, genicular lobes spined; caudal tibize surpassing the length of the femora by about the length of the pronotum, dorsal margins more heavily spined than the ventral ones. Description of the Male Type (from Brunner).—Tegmina of male with the internal margin having a considerably produced angle. Cerci of male robust at base, not far from base horizontally incurved at a right angle, attenuate, compressed, acute acuminate. Sub- genital plate of male very much flattened, attenuate in the middle, exserted caudad in two lobes. Measurements (in millimeters). Cuernavaca, Mex. Rio Cocula, = == Mex. fof i] 2 te) (Type, ex (TYPE, ex {Hebard (Type of Brunner). Brunner). Coll.] —_ pulchra.). Length of body (exclusive of [A.N.S.P.] ovipositor) poltsye 15. 17.5 20. Length of pronotum 5) 4. De 4.5 Greatest caudal width of disk of pronotum Ss ie 3.7 3.2 “Length of tegmen ; ee 136 L. Greatest width of tegmen 2.6 PA: Length of cephalic femur........ 10. 9. 10. ll. Length of median femur Mas 11.4 12. Length of caudal femur 24. 23. 25.6 24.5 Length of ovipositor : 12. 12.3 11.2 Color Notes.—The original color characters given by Brunner are as follows (paraphrased): Green. Occiput rufous, with a fine median line of sulphur. Pronotum rufous, disk fuscous, marked. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 77 with longitudinal lines of sulphur, approximating mesad. Tegmina of male green, disk fuseco-maculate, external margin albo-vittate. All of the femora rufescent at their bases, toward the apices green, apex of the caudal femora, as well as the base of the caudal tibis, infuscate. Abdomen of male fuscous dorsad, marked with a black line and a broad vitta of sulphur, distal segment of male rufous. Cerci of male nigro-lineate. Subgenital plate of male light green, lobes nigro-marginate internally at termination. Abdomen of female rufo-punctate. Ovipositor with both margins rufescent at apex. The two females before us show considerable color variation which the following description covers. General color of the face, genx, lateral lobes of the pronotum, pleura, and greater portion of the lateral aspect of the abdomen honey yellow to dull green yellow, ventral aspect of the body similar but inclining toward wax yellow. Dorsal aspect of pronotum, fastigium, occiput, dorsum of abdomen, and dorsal portion of lateral aspects of the latter brick red to claret red. Eyes buckthorn brown to raw sienna, crossed obliquely by a fine seal brown line; antenne with the two proximal joints largely morocco red to claret brown, remaining joints of dorsal color; ventral portion of infra-ocular region, at base of mandibles, with a more or less distinct blotch of claret brown. Pronotum with a pair of narrow discal lines varying from honey yellow to light viridine yellow, these bordered more or less distinctly on one or both (dorsal and lateral) margins by blackish lines, the pale lines regularly diverging cephalad and caudad from the middle of the pronotum; cephalic and caudal margins of the disk more or less distinctly beaded with blackish. Tegmina with the discoidal section of the color of the dorsum of the abdomen, occasionally washed in part with blackish; marginal field of lateral color; distal margin in one specimen edged narrowly with the green of the limbs, sutural region approaching the lateral color in the other specimen. Abdomen with traces of a seal brown pattern bordering the lateral section of the dorsal color, the pattern sometimes enclosing areas of the clear dorsal color and again merely a line of seal brown ; marginal beading of segments distinct dorsad, more or less distinct laterad. Ovi- positor of lateral color, more or less washed with parrot green, the rufous margining extending to the base on the dorsal margin and little proximad of the armed section on the ventral margin, dorso- proximal section with the suleus more or less blackish. Limbs absinthe green to claret brown, proximal portion of the cephalic and median femora washed with chestnut when the general color of the 78 PROCEEDINGS OF THE ACADEMY OF [Jan., femora is green; apices of all the femora, or only the caudal femora, and the adjacent portion of the tibie blackish; caudal femora proxi- mad with a reticulate pattern of blackish brown on the lateral face, a central line of which pattern is more or less decidedly indicated; - distal portion of the caudal tibiz and tarsi blackish. Distribuiion.—The present species is known only from two localities on the slopes of the upper Rio Balsas Valley in the states of Morelos and Guerrero, Mexico. The species apparently has a range in vertical distribution extending from about 1900 to 5440 feet, from the evidence of the twa localities from which it is now known. Morphological Notes——The ovipositor, as is usual in species of the genus, varies somewhat in length. The subgenital plate of the female shows considerable variation in the form of the distal margin, ranging from the truncate type originally described, through the moderately arcuato-emarginate condition found in our Cuernavaca topotype, to the moderately decided and distinct obtuse-angulate emarginate condition found in the Rio Cocula specimen. The tegmina of the female have a certain amount of variation in the distal margin of the same, this being more truncate in one of our specimens than in the other. There is also some little variation in the width of the interspace between the tegmina in the same sex. Synonymy.—tThe senior author is responsible for the only synonym of the present species—D. pulchra. Tje female specimen on which that synonym was based was first recorded correctly by him as D. mexicana, but later differences in the subgenital plate were noticed which seemed of specific value, and the individual was separated as D. pulchra, its closest relationship being supposed to be with D. emarginata. The apparent difference on which this separation was made we now know to be untrustworthy, as the amount of com- pression of the plate produces a different form in the margins of the same. A certain amount of individual variation in the emargination of this plate is also evident from the form of it in the three known individuals of that sex. Remarks.—Owing to our lack of male individuals of this species, the type of that sex being unique-as far as known, we have placed the species in the male key from the evidence of the original deserip- tion. Specimens Examined.—2; 2 females. Cuernavaca, Morelos, Mexico, January 4, 1899, 1 2. [{Hebard Coll.] Rio Cocula, Guerrero, Mexico, May 12, 1898, (Otis W. Barrett), 19. Type of pulchra. [A. N.S. P.] 1914.] NATURAL SCIENCES OF PHILADELPHIA. 79 Dichopetala falcata n. sp. This species belongs to the same section of the genus as mexicana, but differs in the greater general size, the shorter limbs of the male, in the tegmina being overlapping in the female, the male subgenital plate being without exserted lobes, in the same portion in the female being distinctly emarginate and in the greater length of the ovipositor. TypE: &; Tepic, Mexico. (Hisen.) [Hebard Collection.] Description of Type—sSize quite large (for the genus); form comparatively robust. Head with its greatest width contained about one and one-half times in the depth; occiput well inflated, steeply declivent to the very short, compressed, and sublamellate fastigium, the apex of which is very bluntly recurved, very slightly separated from the apex of the facial fastigium; eyes small, elliptical in outline, Fig. 3—Dichopetala falcata n. sp. Lateral outline of type. (X 2.) the length about equal to half that of the infra-ocular portion of the gene; antenne elongate, proximal joint subdepressed. Pronotum with the greatest dorsal length subequal to the greatest width across the ventral portion of the lateral lobes, dorsal line when seen from the side straight, the whole dorsum slightly constricted mesad when viewed from the dorsal aspect; cephalic margin of disk moderately arcuato- emarginate, caudal margin truncate; but a single complete transverse sulcus present, this entering the disk laterad at the middle and on the middle of the disk strongly arcuate caudad; lateral lobes distinctly longer than deep, the greatest depth contained one and one-half times in the length of the same, ventro-cephalic-angle very narrowly rotundato-rectangulate, caudal margin obliquely rotundato-truncate, 80 PROCEEDINGS OF THE ACADEMY OF [Jan., ventro-caudal angle very broadly rotundate, ventral margin slightly arcuato-emarginate. Tegmina with their exposed length about three-fourths that of the dorsum of the pronotum, broad, the greatest width slightly exceeding the length of the same, the general form subquadrate ; marginal field regularly narrowing distad, distal extremity broadly arcuato-truncate, disto-costal angle narrowly rounded, sutural margin with the proximal angle distinct, subrect, thence the margin is obliquely biundulate to the distal margin; texture of the tegmina coriaceous, the principal veins poorly indicated, interstices obscurely and irregularly reticulate, stridulating vein strongly indicated. Disto-dorsal abdominal segment transverse, a rectangular depressed area present mesad; supra-anal plate trans- verse trigonal, the apex distinctly produced in a short slightly upturned acute process; cerci simple, depressed, falciform, the proximal third moderately broad, slightly lamellate laterad, dis- tinctly tapering from the proximal third, the internal margin regularly arcuate, the external one with a rounded angle where the proximal oF Pig. 4.—Dichopetala falcata n. sp. Fig. 5.—Dichopetala_falicata n. sp. Outline of apex of abdomen of type Outline of ovipositor of allotype. seen from the dorsum. (X 3.) (X 2.) lamellation disappears, the form of the whole cercus appearing bent-arcuate in consequence, apex acute; subgenital plate broad, short, narrowing distad, the apex very narrowly subtruncate. Cephalic femora about one and one-half times the length of the dorsum of the pronotum. Median femora nearly twice the length of the pronotum. Caudal femora with their length not greatly inferior to that of the body, distinctly but not greatly inflated proximad, ventral margins unarmed, genicular lobes very weakly or not at all spined; caudal tibie distinctly but not greatly exceeding the femora in length, dorsal spines more numerous than the ventral ones. AuuotyPE: 2; Tepic, Mexico. (Hisen.) [Hebard Collection.] Description of Allotype.—Differing from the type in the following characters: Pronotum with the dorsal length slightly greater than the greatest ventral width across the lobes, median constriction extremely slight, hardly evident, cephalic margin emarginato- truncate, caudal margin truncate. Tegmina shorter, their exposed length no greater than half of the pronotal length, decidedly broader 1914.| NATURAL SCIENCES OF PHILADELPHIA. 81 than long, margins as in the male, the sutural margins overlapping for the greater portion of their length, with their form much the same as in the male. Disto-dorsal abdominal segment and the supra-anal plate similar in form to that found in the male, but the terminal tubercle of the latter is broader and less evident; cerci very short, crassate, tapering, apex acute; ovipositor with the length about one and one-half times that of the cephalic femora, moderately heavy, the dorsal margin regularly and distinctly arcuato-concave, ventral margin straight except at the extreme proximal and distal extremities, at the latter well arcuate dorsad to the subacute apex, for the distal third of the dorsal and a fourth of the ventral margins strongly serrato-dentate; subgenital plate transverse, distal margin bis- arcuate emarginate, produced into ‘brief trigonal lobes laterad. Cephalic femora slightly less than one and one-half times the length of the disk of the pronotum. Median femora about one and two- thirds times the length of the pronotum. Caudal femora with their length distinctly exceeding that of the body (exclusive of that of the ovipositor). Measurements (in millimeters). Tepic, Mex. 2 (TYPE.) (Allotype.) Length of body.......... mraaceie OON Dales cy Length of pronotum Nardone tin 55) 6.9 Greatest ventral width of pronotum... o.7 6.3 Length of tegmen.. ees Gee ne one Length of cephalic femur Saheces 8. 9.2 Length of median femur ee? 10. ihe Length of caudal femur 21.5 DAS). Length of caudal tibia 23. 26.5 Length of ovipositor. 14.5 Color Notes.—Both specimens of this species seen by us have been at some time immersed in a liquid preservative which has completely removed their original color, leaving them in general a pale ochraceous. Fortunately, however, sufficient of the pattern remains to enable us to give a few notes on the same. Caudal portion of the occiput, which area is usually covered by the pronotum, seal brown, a very fine postocular line of the same and sometimes a similar weak medio-longitudinal line on the occiput present, the post-ocular continued ventro-cephalad across the eye; antenn irregularly but very closely and strikingly annulate with seal brown. Pronotum with the disk margined laterad with fine continuations of the post- 6 82 PROCEEDINGS OF THE ACADEMY OF [Jan., ocular lines of the head, subparallel caudad to the transverse sulcus and thence distinctly but not greatly diverging (male), or regularly but very gently diverging throughout their entire length (female). Tegmina with the dorsum darker than the marginal field, the venation of the pale general color on a darker, nearly wine-colored, background, humeral trunk of the darker color. Abdomen with the dorsum of the proximal segments in the male narrowly edged cephalad with seal brown, this portion like that similarly colored on the oeciput probably normally concealed; the dorsum of the abdomen separated from the sides by a more or less distinct line, which in position is continuous with the postocular line of the head and pronotum; caudal margin of the dorsal segments in the female more or less distinctly and broadly edged with darker color. Limbs more or less decidedly washed with madder brown, a slight edging of the same color on the dorsal margin of the ovipositor, the termina! teeth of the same tipped with seal brown. Distribution —This very striking species is only known from the type locality, the territory of Tepic, western Mexico. Specimens Examined.—2; 1 male, 1 female. Tepic, Mexico, (Eisen), 1 1,1 2. Type and allotype. |Hebard Collection. | Dichopetala serrifera n. sp. On account of the peculiarly serrate cerci of the male, this species occupies a unique position, and comparison with other forms is not necessary. Tyre: o; Barranéa,; twelve kilometers north of Guadalajara, state of Jalisco, Mexico. Altitude not less than 3,500 feet. Septem- ber 13, 1903. (W. L. Tower.) [American Museum of Natural History.| Description of Type.—Size medium; form subcompressed. Head with greatest width contained about one and one-half times in greatest depth; occiput moderately declivent to fastigium and antennal scrobes; fastigium low, acuminate, faintly suleate dorsad, ventrad subattingent with frontal fastigium; eyes moderately prominent, reniform in basal outline, depth about two-thirds that of the infra-ocular portion of the gene; antennz incomplete. Pro- notum moderately sellate, greatest ventral width about five-sixths that of the dorsal length of the pronotum, greatest caudal width of disk about two-thirds length of same; cephalic margin of disk very broadly and shallowly obtuse-angulate emarginate; lateral margins of disk of pronotum (as indicated by color pattern) slightly con- 1914.| NATURAL SCIENCES OF PHILADELPHIA. 83 verging caudad from the cephalic margin to the middle, thence decidedly diverging to the caudal margin; transverse sulcus rather weak except at median line, crossing margins of disk mesad, broadly V-shaped on disk; lateral lobes with the greatest depth contained one and two-thirds times in the greatest dorsal length of lobes, cephalic margin straight, ventro-cephalic angle obtuse, ventral margin obliquely sinuato-truncate, ventro-caudal angle more or less Fig. 6.—Dichopetala serrifera n. sp. Lateral outline of type. (X 3.) broadly rounded, caudal margin obliquely subtruncate except for a short dorsal section which is truncate with the caudal margin of the disk. Tegmina subequal to four-fifths the length of the pronotal disk, width of discoidal and anal fields subequal to the caudal width of pronotal disk; marginal field broad, costal margin gently arcuate, strongly arcuate distad, distal extremity of whole tegmen obliquely truncate, sutural margin strongly obtuse-angulate produced at the apex of the stridulating vein, distad of this pro- jection straight and rounding into the distal margin; stridulating vein decided, straight, distal portion of stridulating field with anastomosing short cross veins. Disto-dorsal abdominal segment with main portion of same truncate distad, a broad triangular impressed area indicated; supra- ob er ag anal plate trigonal with the apex briefly and Outline of apex of narrowly fissate; cerci with the proximal half abdgmien. of type . rom aorsum. robust proximad, thence decidedly tapering, at the (x 3.) middle the shaft is bent rather sharply meso- dorsad, subdepressed and slightly expanded at the apex, the margins proper unarmed, dorsal face with an elevated ridge bearing ten to eleven teeth of unequal width but subequal length, the distal extremity of 84 PROCEEDINGS OF THE ACADEMY OF [Jan., the cereus proper rectangulate, distal extremity of the toothed ridge with the terminal tooth distinctly projecting, immediately proximad of which, on what is properly the lateral margin of the shaft, is placed an extra adpressed tooth; subgenital plate very ample, moderately produced, lateral margins subparallel, distal margin arcuato-truncate, lateral angles produced into considerable styliform appendages, which in length are about equal to one-half the distances between their bases. Cephalic femora slightly more than one and-one-half times the length of the dorsum of the pronotum; cephalic tibize with foramina elliptical. Median femora one-third again as long as the cephalic femora. Caudal limbs damaged. Measurements (in millimeters)—Typrn: length of body, 16.5; length of pronotum, 5; greatest dorsal width of pronotum, 3.7; length of tegmen, 4; width of discoidal and anal fields of tegmen, 3.3; length of cephalic femur, 8.2; length of median femur, 10. Color Notes—General color cinnamon buff. Dorsum of proximal portion of occiput, extending cephalo-laterad as far as the eyes, disk of pronotum, proximal portion of anal field of tegmina, inter- marginal section of proximal third of the sutural margin of the same, greater portion of discoidal field of same and dorsum of abdomen, sharply delimited laterad, black. Medio-longitudinal region of pronotum and abdomen with a bar varying from burnt sienna to clay color, this area narrow cephalad on the pronotum, somewhat expanded caudad on same, very poorly defined on abdomen and there broad mesad. Eyes cinnamon brown flecked with blackish brown; antennze with the two proximal joints touched with claret brown laterad, remaining joints and ventral surface of the two proximal ones black, the simpler joints narrowly annulate with the general color distad. Lateral lobes of the pronotum sparsely and weakly punctulate with bone brown. Tegmina with region of humeral trunk burnt sienna; edge of proximal third of sutural margin of general color. Lateral aspect of abdomen rather heavily punctulate with bone brown; margins of all segments more or less beaded light and dark; a pale unmarked area present on each side of abdomen in the position usually occupied, in species of the genus, by pale bands; disto-dorsal abdominal segment with the black of the dorsum limited to proximo-laterad trigonal areas. Limbs more or less weakly washed with victoria lake, the femora considerably and tibiz less decidedly lined and speckled in linear fashion with black; tarsi black. Distribution —The species is only known from the type locality. 1914.| NATURAL SCIENCES OF PHILADELPHIA. 85 Remarks.—The type of this remarkable species is unique. Specimens-Examined.—1; 1 male. Barranca, twelve kilometers north of Guadalajara, state of Jalisco, ore elevation about 3,500 feet, September 13, 1903 (W. L. Tower), 1o. Type. [{Amer. Mus. Nat. Hist.] Dichopetala durangensis n. sp. Related, as shown by the female sex, to D. falcata, on which it differs in the much smaller size, the more sellate pronotum, the rela- tively more prominent and larger eyes, the shorter and more abbrevi- ate tegmina, the more decidedly trigonal extremity of the disto-dorsal abdominal segment and in the much more robust ovipositor. The available males are not mature, but they show conclusively that the species has a median tooth or lobe on the cercus, while falcata has the same simple, aside from which the form of the subgenital plate is characteristic. As the females of all of the species except D. serrifera are known, we have no hesitation in describing the species without Fig. 8.—Dichopetala durangensis n. sp. Lateral outline of type. (X 2.) adult males, as the possibility of the present form being the female of serrifera is exceedingly remote. Type: @; Durango, Mexico. (Palmer.) [Scudder Coll.] Description of Type—Size medium; form rather robust. Head with the occiput sharply declivent to the fastigium, strongly arcuate in transverse section; fastigium little elevated, slightly recurved at the apex, elongate, but little compressed, shallowly sulcate dorsad, ventrad touching the fastigium of the face; eyes moderately promi- nent, ovate, the depth of same at least two-thirds that of the infra- ocular portion of.the gene; antenne incomplete. Pronotum weakly sellate, broad, the greatest ventral width but slightly surpassing the greatest dorsal length of disk; disk of pronotum with the lateral margins, which are weakly indicated structurally by calloused lines and strongly by color pattern, parallel to the transverse sulcus, which 86 PROCEEDINGS OF THE ACADEMY OF [Jan., severs the same mesad, thence slightly divergent caudad; cephalic margin of disk weakly arcuato-emarginate, caudal margin of same truncate, width of disk caudad contained about one and one-third times in the length of the same; transverse suleus forming a broad V-shaped pattern mesad on the disk; lateral lobes of the pronotum with the greatest depth contained one and two-thirds times in the dorsal length of the same, cephalic Fig. 9.— Dichopetala Margin of same very faintly emarginate dorsad, durangensis n. sp. yentro-cephalic angle rotundato-rectangulate, Dorsal outline of . 5 head, pronotum ventral margin moderately sinuato-truncate, ven- and tegmina. tro-caudal angle broadly rounded, caudal margin (<2) obliquely arcuato-truncate. Tegmina with the exposed portion about two-fifths the length of the dorsum of the pronotum, transverse, greatest width about twice the apparent length, considerably overlapping mesad; costal margin obliquely arcuate, sutural margin subtruncate, distal margin subtruncate, disto-sutural angle narrowly rounded; marginal field comprising about two-fifths the entire tegminal width. Supra-anal plate rotundato-trigonal, the distal margin of same slightly thickened and recurved; cerci very short, conical, apex slightly incurved; ovipositor about one-half the length of the caudal femora, moderately falciform, median depth about one-sixth of the length, dorsal margin con- siderably and regularly arcuate, ventral margin for about three-fourths of the length subtruncate, the distal fourth of the ventral margin strongly arcuate, dorsal margin with distal two-fifths armed with six to seven decided teeth which are well spaced and increasing in length distad, ventral margin armed on distal fourth with nine spines, which increase in length distad and are slightly recurved at the same end of series; subgenital plate small, broadly emarginato-truncate mesad, laterad with short trigonal lobes at’ the angles. Cephalic femora slightly shorter than length of head, pronotum and tegmina combined, very faintly clavate distad; tibise distinctly exceeding the femora in length, tympanum small, elliptical. Median femora one- third longer than the cephalic femora. Caudal femora moderately elongate, proximal dilation moderate, regularly tapering to the narrow subequal distal portion; caudal tibize exceeding the femora by about one-half the length of the pronotal disk. Notes on Male Sex.—As all the specimens (two in number) of this sex are immature, we can give only a few notes on the genitalia as there found. The cereci are provided mesad on the dorsal surface 1914.| NATURAL SCIENCES OF PHILADELPHIA. 87 with the beginning of what is unquestionably in the adult a very decided lobe, the distal margin of which in the nymphal condition is nearly rectangulate, the whole being considerably elevated dorsad of the shaft of the cercus. The distal portion of the cercus is little curved, robust, slightly depressed, the apex acute. Subgenital plate moderately produced, subequal in width, the distal margin deeply rotundato-emarginate, the lateral angles acute, slightly recurved toward the median line. Paratypic Series.—We have before us two paratypic adult females, one of which is measured below. Measurements (in millimeters). Durango, Mex. (TyeeE.) (Paratype.) g g Length of body (exclusive of ovipositor) sl) 18.2% When cGy Ole ROMO UU rere estates 4.3 4.1 Greatest dorsal width of pronotum......... 3.2 2.9 Length of tegmen.. 2. 2. Greatest width of tegmen 3.2 2.9 Length of cephalic femur 7.3 7.5 Length of median femur 8.9 9. Length of caudal femur 20.2 oe Length of ovipositor... Bos 10.5 10.2 Color Notes.!—General shade ranging from old gold to oil green, on the pronotum paling (in the old-gold individual) to light viridine green. A pair of narrow lines of blackish are more or less distinctly indicated, extending from the dorsal margin of the eye caudad over the sides of occiput and along the lateral angles of the pronotal disk, margined laterad by a band of empire yellow of varying width and definition. Eyes argus brown, blotched with blackish; antenne with the two proximal joints ranging from claret brown to burnt sienna, remaining joints blackish with narrow dull apricot yellow distal annuli; occiput more or less washed with very dull weak maroon. Dorsum of pronotum with cephalic half of disk very weakly washed with morocco red, a medio-longitudinal line of 10 Ip both of these females the abdomen has been bent ventro-cephalad and in consequence the length given above is not the real length of the insect, but only the shortest distance between the point of the fastigium and the base of the ovipositor. It is not possible to get a true measurement of length from the present material. : Jn the present notes only the two well-preserved adults have been used, a few notes on the nymphs being placed at the end. 88 PROCEEDINGS OF THE ACADEMY OF [Jan., seal brown present, the latter finely divided cephalo-caudad by a, thread of morocco red; cephalic and caudal margins of disk with more or less regular and decided beading of blackish, the whole of the pronotum with a more or less decided sprinkling of bay colored points which vary in intensity with the blackish lateral lines. Teg- mina varying from apricot orange to hazel, humeral trunk hay’s russet. Abdomen with the proximo-dorsal portion of each segment (these areas hidden when the abdomen is not stretched) blackish, distal margin of segments more or less beaded with blackish or prout’s brown, the entire surface dorsad and laterad more or less sprinkled with stipples of the latter color. In line with the post- ocular line and tegminal humeral trunk there is continued to the apex of the abdomen a pattern, which is indicated by a limiting to the region between the same of the decided marginal beadings and the proximal blackish markings of the segments, or a pale line similar in color to the pale portion of the postocular line. Mesad on the abdomen is indicated more or less distinctly a line similar in color to the above-mentioned yellowish lateral ones. Limbs more or less washed with madder brown, lined and dotted in linear fashion with black; caudal femora proximad of the color of the lateral lebes of the pronotum triply lined with blackish; all tibize lined dorsad and laterad with blackish. Ovipositor more yellowish than the general tone, with the dorsal margin more or less maroon. The type has the general color old gold with the abdominal segments blackish proximad. One of the nymphs is nearly uniform parrot green, another ochraceous- tawny, both with pale postocular lines, and the third with the two colors combined, the first cephalad and the second caudad, the limbs and whole dorsum strongly punctulate with black and the lined pattern very decided blackish and whitish, the medio-longitudinal line continued on the head. Distribution.~The species is only known from Durango, Mexico. Remarks.—Of the material known belonging to this most interest- ing and beautiful species, one adult female has been badly damaged in the past by insect pests, so it has not been considered in the previous description. The nymphs’ are not perfect, but there can be no question of their identity with the adult females. Specimens Examined—7; 4 females, 2 male nymphs, 1 female nymph. Durango, Mexico (Palmer), 3 2 (Typr, paratypes), 2 co nymphs, 1 2 nymph. [Seudder Coll.] Durango, state of Durango, Mexico, November 27, 1909, (F. C. Bishop), 1 9. [U. 8S. N. M.] © 1914.| NATURAL SCIENCES OF PHILADELPHIA. §9 Dichopetala pollicifera n. sp. This very distinet species requires comparison with none of the other forms of the genus, the general build, nearly uniform greenish coloration, depressed external tooth on male cercus and the form of the tegmina readily distinguishing it. Tyree: o'; Brownsville, Cameron County, Texas. July 31— August 5, 1912. (Hebard.) [Hebard Collection.] Description of Type—Size medium; form quite robust (for the sex); surface smooth, unpolished. Head not at all elevated dorsad of the level of the disk of the pronotum, occiput roundly but rather Fig. 10.—Dichopetala pollicifera n. sp. Lateral view of type. (xX 2.) decidedly declivent cephalad; fastigium narrow, very acute lanceo- late, low, dorsum subsuleate, caudad from the base of the fastigium. extends for a short distance a faint elevated line, ventral portion of the fastigium of the vertex touching the fastigium of the front; eyes very tw prominent, ovate, the greatest width contained about one and one-half times in the depth of the Fig. 11—Dichope- : tala pollicifera eye, the depth of the eye contained about one and yn. sp. Outline of one-half times in that of the infra-ocular portion Pex of abdomen , A of type from dor- of the gene; antenne filiform, rather heavy,in sum. (x 4.) perfect specimens (the type has the antenni broken) about three times as long as the body. Pronotum with the greatest (ventral) width but little less than the length of the dorsum of the pronotum, caudal width of the dorsum con- tained about one and one-third times in the length of the same, dorsum straight cephalo-caudad when viewed from the lateral aspect, appreciably arcuate in transverse section, cephalic margin of the disk emarginato-truncate, caudal margin faintly arcuate, slightly flattened with the faintest possible sinuation mesad, lateral margins of the disk indicated by the usual callous lines, 90 PROCEEDINGS OF THE ACADEMY OF {[Jan., severed once and then by the transverse sulcus, which is placed very slightly cephalad of the middle, subparallel caudad to this sulcus, thence slightly diverging to the caudal margin; transverse sulcus of a V-shaped form mesad; lateral lobes with the greatest depth contained about twice in the greatest (dorsal) length of the same, cephalic margin nearly straight, ventro-cephalic angle nearly rectan- gulate, ventral margin arcuato-sinuate cephalad and arcuately expanded caudad, rounding into the oblique, slightly arcuate caudal margin, transverse sulcus impressed dorsad. Tegmina with their length subequal to that of the dorsum A of the pronotum, their greatest width i (not flattened) subequal to their length; yY\ costal margin straight with the disto- EN ' ) costal angle well rounded, distal margin bluntly arcuate, passing broadly without any Wipa, 12 audi epee sign of angulation into the strongly ob- fala pollicifera n. sp. lique and gently arcuate distal portion of the Dorsal outline of head, sytural margin, proximal portion of the pronotum, and tegmina Z 2 of male type (12) and sutural margin obliquely truncate, at the roaa allotype (15). extremity of the stridulating vein slightly aoe produced, rotundato-rectangulate; marginal field very narrow; discoidal field with a fine, irregular network of veins; stridulating vein slightly oblique, slightly arcuate; tympanum very faintly outlined, trigonal, the sharpest angle directed meso- proximad; proximal portion of the anal field (7.e., proximad of the stridulating vein) closely but rather cdarsely areolate. Abdomen quite plump; disto-dorsal abdominal segment strongly arcuato- emarginate laterad by the bases of the cerci; supra-anal plate forming a transverse lappet which has its greatest length contained about three times in its greatest width, distal margin of same truncate, broadly arcuate laterad; cerci simple, heavy, slightly depressed, distal portion of the cereus and median tooth decidedly depressed, the distal portion of the cereus tapering, acute, the immediate apex very fine, clawlike and slightly hooked, median tooth placed on the external face of the cercus, broad, bluntly lanceolate, reaching about half way from the point of its origin to the apex of the cercus, closely apposed to the shaft of the cercus for the greater portion of its length; subgenital plate moderately elongate, V-shaped in section, distal half with the margins moderately and regularly convergent, distal extremity deeply V-emarginate, the emargination extending slightly more than a fourth the way to the base of the plate, laterad of the S 1914.] NATURAL SCIENCES OF PHILADELPHIA. Ol median emargination the converging lateral margins cause the lateral angles to appear as acute trigonal projections with their immediate angles blunted. Cephalic femora nearly twice as long as the dorsum of the pronotum, slender; cephalic tibize appreciably exceeding the femora in length, foramina elliptical. Median femora subequal to the head, pronotum and tegmina in length, similar in build to the cephalic femora; median_tibie exceeding the femora by about the same proportion as in the cephalic limbs. Caudal femora one and one-third times as long as the body, very moderately inflated in the proximal half, slightly compressed, genicular lobes not dis- tinetly spined, but with a minute point (sometimes absent) at the apex of each lobe; caudal tibie surpassing the femoral length by about four-fifths the length of the dorsum of the pronotum. Allotype: 2; Brownsville, Cameron County, Texas. July 31- August 5, 1912. (Hebard.) [Hebard Collection. | Description of Allotype——The following characters are solely those cof difference from the above description of the type, features not mentioned are essentially as in the male sex. ; Size large (for the genus); form robust. Head noticeably broad, the greatest width nearly equal to the depth of the head as far as the clypeal suture; fastigium very brief, slightly elevated at the extremity; eyes moderately prominent, slightly compressed, ovoid, their depth contained slightly more than one and one-half times in the depth of the infra-ocular portion of the gene; antenne about one and one-half times the length of the body. Pronotum heavy, the dorsum more arcuate in section than in the opposite sex, caudal Q : : iD width of the disk contained one and \Y Ne Ly one-third times in the length of the Weise De same, greatest ventral width of the Sei pronotum equal to about five-sixths Fig. 14. — Dichopetala pollicifera of the length of the dorsum of the Bie Oe guaposttor et same; cephalic margin of the disk emarginato-truncate, caudal margin of the same moderately arcuate, no appreciable callous bounding lines present on the disk, which rounds into the lateral lobes; transverse sulcus weakly indicated, placed mesad and on the middle of the disk impressed in a broad, V-shaped figure; lateral lobes with the greatest depth contained one and two-thirds times in the dorsal length of the same, margins as in the males but with the ventro-cephalic angle blunter. Tegmina with their apparent length about one-fourth that of the pronotum, their greatest width nearly two and one-half times their visible 92 PROCEEDINGS OF THE ACADEMY OF [Jan., length, interspace between the tegmina very slightly less than half the apparent length of the tegmina; distal extremity of the tegmina arcuato-truncate, broadly rounding to the costal margin and more narrowly to the sutural margin. Supra-anal plate very broad trigonal, apex very blunt; cerci short, conical, distal portion slightly elongate; ovipositor subequal to the combined length of the head, pronotum and tegmina, the greatest proximal depth contained slightly more than three times in the length of the same, moderately tapering in the proximal two-thirds, moderately arcuate, dorsal margin moderately arcuate, ventral margin with slightly more than the median half of its length nearly straight, considerably arcuate proximad, strongly arcuate distad, general angle of the margins distad slightly more acute than a rectangle, distal fourth of the dorsal margin with seven to eight teeth, low proximad and increasing in length distad, erect, directed disto-dorsad, ventral margin with the distal fourth armed with eight to nine spines, slightly increasing in length distad, the extreme distal ones slightly hooked; subgenital plate small, V-emarginate distad for about half of its length, the portions of the plate laterad of the median emargination present as acute trigonal lobes, the apex of which is slightly beyond the general form of the lobe. Cephalic femora about one and one-half times the length of the dorsum of the pronotum, more robust than in the male sex. Median femora about a third again as long as the cephalic femora. Caudal femora slightly more robust than in the male, but of similar proportions. s Paratypic Series.—A series of fourteen males and eight females bearing exactly the same data as the type and allotype have been selected as paratypes. Measurements (in millimeters). Brownsville, Tex. (TyYPE.) (Paratypes.) Length of body 16. AF 15. Length of pronotum 4.2 4.4 3.8 Greatest dorsal width of pronotum 3.2 3.2 3.2 Length of tegmen 4.2 4. 4.2 Greatest width of discoidal and anal fields of tegmen 3.8 3.8 5 Length of cephalic femur 8. 6a fiers: Length of median femur 9.5 10.6 9 Length of caudal femur 21.5 22.3 19.4 2 Regenerated limb. The other cephalic limb is lacking. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 93 2 Brownsville, Tex. (Allotype.) (Paratypes.) Length of body (exclusive of ovipositor).. 18. 21. 16. Length of pronotum.. Seared. 5.6 5.5 ‘Greatest width of dorsum of pronotum's 4.2 4.3 4. Apparent length of tegmen... 1.4 ho) 1.4 Greatest width of entire tegmen 3s oF 2.5 Length of cephalic femur... 8.4 8.1 8.9 Length of median femur......... 10.2 10. 10.4 Length of caudal femur... . 24. 23.8 23.5 Length of ovipositor 9.5 9.3 9: In size the type represents what is the average of the entire male ‘paratypic series, practically none but the extremes measured above varying appreciably from the more general size. The allotype is in most measurements the maximum extreme for the female sex, from which the series grades rather evenly to the minimum measurements given for that sex. The length of the body is an uncertain measure- méht, as it depends so largely on the stuffing of the abdomen, and, while the present series was stuffed in its entirety by the authors, a certain amount of variation in bulk is impossible to prevent. The length of the cephalic femora, particularly in the female sex, shows appreciable variation—in females of approximately the same bulk measuring 7.8 and 9.2 mm. This variation is also noticed in the median femora, while the tibiz of the respective limbs vary propor- tionately. Such variation in the length of the caudal femora as is evident is of a far less degree and not disproportionate to the general bulk variation. The tegmina of the male show some variation in length, never being shorter than the pronotal length, but occasionally surpassing the same. Color Notes—The general pattern of the male sex consists of a dorsal shade covering the dorsum of the head, disk of the pronotum; another covering the tegmina (aside from the marginal field and region of the principal veins); a paler lateral color involving the face, sides of head, lateral lobes of the pronotum, and sides of the abdomen; an area covering the dorsal surface of the abdomen either concolorous with or darker than the dorsal shade of head, pronotum and tegmina, and a pair of pale, narrow postocular lines extending along the angle 8 Owing to the poor definition of the dorsum of the pronotum in the female, this measurement is of less value than in the male, but it is here given to cover relatively the same portion as that measured in the other sex. 94 PROCEEDINGS OF THE ACADEMY OF [Jan., of the tegmina, involving the whole marginal field of the tegmina and dividing the dorsal and lateral colors of the abdomen. The female sex is so unicolorous that it has no pattern distinctive enough to describe. The color of the dorsum of the head, pronotum, discoidal and anal fields of the tegmina and the limbs of the male ranges from yellowish oil green to cosse green,'! occasionally lined along the internal margin of the pale lateral pronotal lines with maroon, and in all thickly and more or less regularly sprinkled with very minute points or stipples of the same color. Discoidal and anal fields of the male tegmina varying from lettuce green to serpentine green; !” area of the principal veins (7.e., humeral trunk) more or less broadly and strongly lined with a shade varying from morocco red to maroon. Paired pale lines in the male (also covering the marginal field of the tegmina) varying from cream-white to light green-yellow, less prominent on the abdomen in some specimens than in others. Dorsal color of the abdomen sometimes the same as the dorsal color of the pronotum, again as dark as the major portion of the tegmina, and in a fair proportion ranging through pompeian red to madder brown, such brownish tones being due to a great increase in number of the overlying stipple points of those colors, similar to the condition found on the dorsum of the pronotum. These points are also present in individuals having an apparently uniform greenish dorsal tone on the abdomen, but they are so few that they do not affect the general shade. Lateral color in the male ranging from javel green to cosse green, sulphine yellow in a single individual.'® General color of female varying from uniform lettuce green to snuff brown dorsad, all finely and more or less thickly stippled with garnet brown to maroon (one specimen), paling into court gray on the sides of head, pronotum, and pleura; tegmina of the dorsal color with marginal field pale and a line of maroon present on the principal veins. Pale lines indicated but weakly in a few specimens on head and pronotum, not present on abdomen. Limbs unicolorous with the dorsum of the pronotum in both sexes. Eyes in both sexes varying from russet brown to hazel, crossed obliquely dorsad of the middle by a fine line of chestnut brown, which touches the caudal margin of the eye at the ventral margin of the pale postocular line. Antenne in both ‘Frequently the pronotum is in part paler than these shades, but this is apparently due to the stuffing and drying, so that no importance is here attached to such fluctuation in the color of the dorsum of the pronotum. 16 The latter in but one specimen, the remainder between lettuce green and’ spinach green. . 16 The latter shade may be due to drying, as it is found in but asingle specimen 1914.| NATURAL SCIENCES OF PHILADELPHIA. 95 sexes ranging from olive ochre to lime green, the proximal joints more or less speckled with maroon. Tibial spines tipped with black. Ovipositor of the general color tone and usually of the dorsal shade, finely stippled with garnet brown and dorsal margin more or less washed with same. ; Distribution.—The present species is known only from three locali- ties in the arid tropical Tamaulipan region of southern Texas, Brownsville, Piper Plantation (along the Rio Grande about ten miles southeast by east of Brownsville), and Lyford (in the same county about forty-seven miles north of Brownsville). It doubtless ranges over an adjacent section of Mexico. Biological Notes——At Brownsville and Piper Plantation the present species was scarce and local, occurring in tangles of Clematis (probably C. reticulata) growing over the ground and on low mesquite and huisache. Individuals, when disturbed, endeavored to hop, crawl, or drop into recesses of these vines, where they are so well protected by their coloration that beating was the best method of securing them. At Lyford the single specimen was taken with D. gladiator in a weedy field which had a low cover of sand spur (Cenchrus sp.) and grasses. This species was found to be by far the least active of any of the forms of the genus taken by the authors. Morphological Notes—In the female sex the interspace between the tegmina varies from one extreme, in which the sutural margins are touching, to one in which the space separating them is nearly half of the width of a single tegmen. Remarks.—The possibility of confounding this very peculiar species with any other form of the genus is very remote. It is interesting that in a region which has been examined as often as the Brownsville section, as striking a species as this should have been overlooked, for which the character of its habitat is probably responsible. Specimens Examined.—43; 17 males, 9 females, 17 nymphs. Brownsville, Cameron County, Tex.; July 31—-August 5, 1912; (H.); 16 &, 9 2 (Typn, allotype, paratypes), 6 co’ nymphs, 9 2 nymphs. Piper Plantation, near Brownsville, Cameron County, Tex.; August 3, 1912; (R. & H.); 2 9 nymphs. Lyford, Cameron County, Tex.; August 6-7, 1912; (R. & H.);1 7. Dichopetala castanea n. sp. 1912. Dichopetala brevihastata Hunter, Pratt and Mitchell (not of Morse), Bull., 113, Bureau of Entom., U.S. Dept. of Agr.,p. 50. (Part) [Corpus Christi and Maverick County, Texas.] D. brevihastata Morse—in This species differs from its nearest ally 96 PROCEEDINGS OF THE ACADEMY OF {Jan., the broad, hardly produced subgenital plate of the male, in the more spiniform and strongly incurved distal portion of the male cercus, in the slightly deeper lateral lobes of the pronotum in both sexes, in the broader and more approximate tegmina of the female, in the less tapering ovipositor, which has the ventral margin straighter, and in the broader, more ovate, and less elliptical eye. Type: co; Laguna del Gato, three miles west of Sam Fordyce, Hidalgo County, Texas. Elevation 175-200 feet. August 6, 1912. (Rehn and Hebard.) [Hebard Collection.] Description of Type—Size medium; form moderately slender; surface subglabrous. Head with the occiput not elevated dorsad of the level of the pronotum, gently arcuate; fastigium compressed, Pig. 15.—Dichopetala castanea n. sp. Lateral view of type. (X 2.) lamellate, barely touching the frontal fastigium; eyes prominent, ovate, the greatest width contained less than one and one-half times in the length, the length of the eye contained about one and one-third times in the infra-ocular length of the gen; face moderately flat- tened; antenne over three and one-half times as long as the body, filiform. Pronotum with the dorsum subdeplanate, strongly nar- rowed mesad, the margins of the same regularly converging caudad from the cephalic margin and from the middle somewhat more strongly diverging toward the caudal margin, the median width not more than half that at the caudal margin, the width of the latter equal to slightly more than half of the length of the disk, cephalic and caudal margins of the disk subtruncate, transverse broad V-shaped sulcus | [ placed slightly caudad of the middle, \ Wks caudad of which sulcus there is oe PG NS laterad on the disk slight indication Figs. 16 and 17—Outline of left of another sulcus; lateral lobes with cereus of males (types) of Dichope- the greatest depth contained nearly tala castanea (16) and D. brevihas- c : tata (17). (X 10.) twice in the dorsal length of the 1914.] NATURAL SCIENCES OF PHILADELPHIA. 97 same, ventro-cephalic angle subrectangulate, ventral margin moder- ately sinuato-emarginate cephalad, ventro-caudal angle broadly rounded, caudal margin oblique-truncate, transverse sulcus well im- pressed dorsad, lateral shoulders slightly indicated in the humeral region. Tegmina very slightly shorter than the dorsal length of the pronotum, the distal extremity slightly surpassing the margindf the proximal abdominal segment, costal margin moderately arcuate, distal margin truncato-arcuate, the lateral angle rounded, sutural margin rotundato-rectangulate at the apex of the stridulating vein, the margin obliquely subtruncate thence to the distal extremity; mar- ginal field narrow, discoidal field expanding from the middle of the tegmen, with a number of irregular areas, stridulating vein nearly transverse, gently arcuate caudad, speculum proper with the greatest length exceeding the greatest width. Abdomen subfusiform; disto- dorsal abdominal segment with the median impressed area transverse and arcuate; supra-anal plate trigonal with a broad median V-shaped emargination; cerci with the proximal portion straight, robust, cylindrical, distal portion tapering, gently curving mesad, the distal fourth straight, spiniform and at a right angle to the thickened - proximal portion, tooth placed at the junction of the proximal portion and the tapering section, on the dorsal section of the cercus toward the external face, moderately acute, subdepressed, little divergent dorsad from the general plane of the cercus, not more than a third the length of the distal half of | the main cercal shaft; subgenital plate broad, USGS) the eB Bae Figs. 18 and 19.—Outline of subgen- est width considerably exceeding ital plate of males (types) of Dichope- the length of the plate, lateral Gone: ee) and D. brevihastata margins straight convergent in ibe F the proximal two-thirds of their length, thence parallel for a very short distance, the width of the subequal section about half that of the greatest width of the proximal section, distal margin with a \-shaped emargination mesad, this occupying slightly more than a quarter of the median length of the plate, laterad of this emargination the distal margin is nearly truncate, thus forming lateral angles slightly more acute than a rectangle, a moderately distinct median carina present on the ventral surface of the plate. Cephalic femora subequal in length to the head, pronotum and tegmina, very slender; cephalic tibie distinctly exceeding the femora in length, foramina small, elliptical. Median femora slightly longer than the cephalic femora; ‘ 98 PROCEEDINGS OF THE ACADEMY OF {Jan., median tibie nearly half again as long as the femora. Caudal femora about a third again as long as the body, moderately inflated in the proximal three-fifths, genicular lobes unispinose; caudal tibize surpassing the femora by about the length of the pronotum. Allotype: 2; Laguna del Gato, three miles west of Sam Fordyce, Hidalgo County, Texas. Elevation 175-200 feet. August 6, 1912. (Rehn and Hebard.) [Hebard Collection.] Description of Allotype-—Characters not specifically mentioned are not markedly different from the male sex. Size medium; form moderately robust. Pronotum much less compressed, the median width of the disk of the pronotum contained less than three times in the length of the same. Tegmina much broader than the apparent length, apices hardly surpassing the margin of the metanotum, the distal margin rotundato-truncate, the interspace between the tegmina not more than half the width of a single tegmen; venation irregular. Cerci very short, conoid; ovipositor nearly equal to the length of the head and pronotum together, tapering in ~the proximal two-thirds, the proximal depth slightly more than a third of the length, moderately arcuate, ventral margin appreci- ably flattened for a portion of its length, dorsal margin more arcuate distad than proximad, apical margins proper slightly more acute than a _ rectangle, armed. on the distal third of the Figs. 20 and 21.—Outline of ovipositor dorsal margin with seven erect of females (allotypes) of Dichopetala teeth, increasing in length distad, are (20) and D. brevihastata (21). ventral margin strongly arcuate distad, armed with nine spines, the proximal several of which are smaller than the others, the distal ones distinctly recurved; subgenital plate brief, strongly transverse, distal margin obtusely brace-shaped (—-—) emarginate, Cephalic femora slightly exceeding the combined length of the head, pronotum and tegmina, slightly more robust than in the male. Median femora as robust as the cephalic pair. Caudal femora a fourth again as long as the body (exclusive of the-ovipositor), appre- ciably more robust than in the male, the proximal dilation slightly more extensive. Paratypic Series.—A paratypic series of five males and eight females from Laguna del Gato has been selected, the measurements of the same being given below. 1914.| NATURAL SCIENCES OF PHILADELPHIA. 99 Measurements (in millimeters). A A lomo} Laguna del Gato, Tex. ford 1 SS) (TYPE.) (Paratypes.) Length of body 2 Pal hays 20 18. ate Length of pronotum 4.2 4.5 4.5 4.3 Greatest width of dorsum of pronotum........... 2.6 3 2.3 2.9 Length of tegmen 3.9 4 4. 4. Greatest width of discoidal and anal fields of tegmen 3.2 Bas! 348" 3. Length of cephalic femur 9.2 9.7 9.5 8. Length of median femur 9.8 10.8 10. 9. Length of caudal femur 20. 22. 22 19. 22 Laguna del Gato, Tex. (Allotype.) (Paratypes.) Length of body (exclusive of ovipositor) Pale 16.8 18.5 19. Length of pronotum 5D. 4.8 Heal 5.2 Greatest width of dorsum of pronotum........... 3.6 3. 3. 3.2 Apparent length of tegmen 128} 1.6 1.4 156 Greatest width of tegmen Pst 2.9 2.4 Boe Length of cephalic femur 8.6 8. 8.2 8.2 Length of median femur 9.2 8.9 9. 9. Length of caudal femur 2125 20. Pale 21.2 Length of ovipositor 7.2 7.5 ail U8 Measurements of extreme individuals. fou foil g Q Southwest Victoria, Southwest Tamo , Texas. Taman as, Texas. Vera (Schaupp) Mex. (Schaupp) Cruz, {Hebard [Hebard {Hebard Mex. Coll.] Coll.] Coll.} (Bishop) [U.S.N.M.] Length of body 5 23.5 16.8 21. Length of pronotum 4 4.7 4.9 6.1 Greatest caudal width of dorsum of pronotum 226 3.2 4.1 Length of tegmen 3 3.4 2 2a Greatest width of discoidal and anal fields of (co), or entire widthof (9)tegmen. 3. | 3.8 22 2.8 Length of cephalic femur (ee 9. iG) fit. Length of median femur 8. 10. 8.3 12.9 Length of caudal femur. 18. 21.5 20. 29. Length of ovipositor ie peeves 6.7 10. 100 PROCEEDINGS OF THE ACADEMY OF {Jan., The variation in size appears to be purely individual aside from the possible presence of a larger race at the southern extreme of the. range of the species. The individuals from Victoria, Tamaulipas, are of peculiar interest in this connection, the pronotum in both sexes being shorter and broader than in the other specimens of the species, while the legs are somewhat thicker and more robust than in by far the greater portion of the Texas series, but their length pro- portions are matched in other individuals. The tegmina in the Victoria male are shorter and broader than in any other individual of the same sex seen, the distal portion being much less produced with the margin decidedly truncate. The size of the tegmina is seen to vary somewhat in the female sex, the width of the imterspace __ between the two also fluctuating, but, the latter is never wide enough to cause any difficulty in determining the species. The Tamos female is unique in its great dimensions, aside from which it is per- fectly typical of the species. We doubtless have here parallelism to the condition of sporadic giantism found in the next species, under which the matter is discussed. It is possible that in this ease the variation is geographic, but our material from Mexico is too meager to make any definite assertions. Color Notes.—The following color notes are all based on stuffed specimens, all of which, with the exception of two individuals, were prepared by the authors and have retained in practically its entirety the original coloration.” The general pattern-of the Male has the dorsum of the head, pronotum, and abdomen dark and generally uniform, paired pale lines extending from the caudal margin of the eye, marking the boundaries of the disk of the pronotum, involving the whole marginal field of the tegmina and present as broader lateral bars on the abdo- men. The abdominal segments have pale beaded margins caudad, these rarely outlined proximad with blackish. In the male the general color of the venter and of the lateral aspects of the pronotum, pleura, abdomen and proximal portions of the femora vary from honey yellow to parrot green in shade and also considerably in intensity. The color of the dorsum of the head, pronotum, and abdomen in the same sex ranges from clear mahogany red through argus brown to nearly clear pyrite yellow. This color is usually ‘7 Three males and eight females listed in the summary of specimens have not been stuffed and are disregarded in the remarks given above. They are all much browner than any of the freshly prepared individuals collected by the authors and in three cases have the color pattern much intensified on the abdomen. « 1914.] NATURAL SCIENCES OF PHILADELPHIA. » 101 purest on the disk of the pronotum, on the dorsum of the abdomen being solid, bounded by the pale marginal lines on the segments, frequently bisected by an adventitious pale medio-longitudinal line _or divided by a broad median area of the lateral color and in the extreme condition only present laterad in the neighborhood of the pale lines. Head of the male with the cephalic and lateral aspects slightly paler than the lateral color, varying from baryta yellow to chalcedony yellow, an infra-ocular bar of indian red frequently present, rarely observed, and then but faintly, in the paler specimens; eyes varying from light buff to seal brown, this apparently due in part to the drying of the insect; antennz varying from raw sienna to kaiser brown, the proximal joint in greater part of the color of the face. Pronotum of male with the pale lateral lines varying from nearly pure white to buff yellow in one extreme and viridine yellow in the other, always broader cephalad and caudad than mesad, generally severed once and sometimes twice at the point of greatest constriction; lateral lobes of pronotum with a trace of the dorsal color present cephalad and caudad contiguous to the pale bars. Tegmina of the male with the underlying color varying from bay to maroon, the venation and marginal field similar to the pale lateral bars of the pronotum in color; base of the narrow discoidal field entirely suffused with weak mahogany red or bright chestnut, the pale veins of the distal portion of the same field sharply contrasted. Disto-dorsal abdominal segment of the male washed with yellow, varying from pale cadmium to mars yellow, the cerci wholly of the same color. Cephalic and median femora of the male with more or less of their length ochraceous-orange, occasionally green without any of the former color; cephalic and median tibize with green the underlying color, more or less suffused with ochraceous-orange or occasionally (Victoria, Tamaulipas, specimens only) almost wholly blackish, the region of the tympanum almost invariably (but a single exception) touched dorsad with blackish. Caudal femora-of male with the proximal three-fifths parrot green'®’ with a medio-lateral stippled pattern of blackish; distal extremity of femora blackish brown of variable intensity, rarely very weak, remainder of distal two-fifths of the femora varying, with the general tone, from mustard yellow to zinc orange. Caudal tibie of male with their ventral color agreeing with the pega distal pec of the femora, oe dorsad 18 Even in the specimens with honey-yellow lateral fo (Victoria, T; amauli- pas) the femora are parrot green, which would lead one to the conclusion that green was the natural coloration of the greater portion of the whole insect. 102 _ PROCEEDINGS OF THE ACADEMY OF [Jan., more or less washed with blackish brown, very strongly and decidedly so in the Victoria, Tamaulipas, pair. The type is of a well-contrasted color form only surpassed in intensity by the Victoria individuals. The stuffed Victoria, Tamaulipas, female is practically a color duplicate of the male from that locality, the following comments being based on the remainder of the stuffed females. The general pattern of coloration of the females differs from the pattern of the male only in that the pale lateral lines are not indicated on the tegmina and the beaded character of the marginal coloration of the dorsal abdominal segments is less frequent. The lateral and ventral color of the females is, in the stuffed individuals, always green, varying from light yellow-green to parrot green. The dorsal color as indicated in the male is rarely developed at all in the female, and then mostly indicated only on part of the pronotum and extreme lateral sections of the dorsum of the abdomen. A narrow medio- longitudinal line varying from victoria lake to english red is indicated more or less distinctly on a part or all of the pronotum in a portion of the females, while in‘one specimen the dorsum of the pronotum is weakly suffused with the last-mentioned color. Head of the female colored as in the opposite sex, but antenne ranging from oil yellow to mars yellow. Pronotum of the female with the lateral lobes colored as in the male; the usual pale lateral lines of the pronotum are subobsolete in about half of the specimens, being indicated on the head in several which have them almost lacking on the pronotum. Tegmina of the female varying from cream white to baryta yellow, the interstices of the dorsal section occasionally slate color, the median section proximad with a variable but always weak narrow line of hay’s russet. Abdomen of the female with the lateral pale lines washed with the lateral color as in the male, occasionally very weak, sometimes relatively broad and frequently with the portion on each individual segment roughly elongate trigonal. Ovipositor passing from courge green at the base to ivy green at the apex, washed along the dorsal margin with the color of the median line of the dorsum of the pronotum when that line is present; teeth of the ovipositor pitch brown. Limbs of the female varying from a type like that of the majority of the males to one in which they are practically course or light bice green, the pattern of the external face of the caudal femora always, although often very faintly, indicated. The presence of the blackish near the tympanum of the cephalic tibiz is exceptional and not the rule in the female. The allotype is about midway between the extremes here de- 1914.] NATURAL SCIENCES OF PHILADELPHIA. 103 scribed, having a short median line on the pronotum, very weak incomplete pale pronotal lines and practically uniform limbs, without blackish near the tympanum of the cephalic tibiz. The Victoria, Tamaulipas, individuals stand apart from the other specimens, having an intensified pattern surpassing in contrast anything else belonging to the species seen by us. Certain structural peculiarities may compel the separation of these, as a distinct race, at a later date when more Mexican material is accessible, so that the following comments do not include-them. The female from Maverick County has a very intensive coloration, with the dorsum of the pronotum and lateral portions of the dorsum of the abdomen suffused with garnet brown. The Laguna del Gato series is as a rule weaker in color contrasts than Uvalde, Del Rio and Mission individuals of both sexes, the type alone excepted. San Antonio specimens stand about intermediate in color intensity. That color is of little geo- graphie significance is shown by the fact that the two lots taken closest together (about eighteen miles apart), 7.e., Laguna del Gato and Mission, are nearly as widely divergent as any examined. Distribution—The range of this species extends from an undeter- mined point on the Pecos River, probably near the New Mexican line, east to the vicinity of Corpus Christi and south as far as Tamos, Vera Cruz, Mexico. Aside from the uncertain Pecos record and that from Tamos, the range of the species is approximately co-extensive with the area called the Rio Grande Plain by Bray.” While the Pecos locality is more elevated, the highest points at which we have noted the species (Del Rio and Uvalde) are at an elevation of 1,100 feet, while the Corpus Christi and Tamos individuals were taken almost at sea-level. It is probable that the Victoria, Tamauli- pas, specimens were taken at a higher elevation than 1,100 feet, but we have no definite information to this effect, the general region, however, being near the 500 meter (approximately 1,640 feet) con- tour. At Mission the species was taken just below the line of gravel hills, while the Laguna del Gato series was taken in these hills. . At Del Rio and Uvalde it occurred on the rolling plateau country, while 19 The data with this specimen is ‘‘Pecos, Aug. 18.’’ As the specimen came to Seudder through Uhler, we can, judging by analogy with other specimens similarly credited by Scudder, probably consider it one of C: apt. Pope’s collecting. Capt. Pope’s camps along the Pecos reached from above the New Mexican line to considerably below the same, but the dates were all in March. The specimen is in poor shape, having been dried from a liquid preservative. It is possible, however, that the specimen was taken much later near the present town of Pecos, Texas. 20 Botan. Gazette, XXXII, p. 116, fig. 6. 104 - PROCEEDINGS OF THE ACADEMY OF [Jan., at San Antonio it was taken in the hilly country ainemate ly north of the city. Tamos, the southern limit of the species, is near Tampico, on the Rio Panuco, in the extreme northern part of Vera Cruz. Biological Notes—At Laguna del Gato the present species was taken on a low, very green rhamnaceous shrub (probably Condalia obovata), where it was common locally. At Mission five individuals were heard at night in bushes, several as much as five feet from the ground. At San Antonio the species was very local, not at all common and hard to find in a low, stout, green rhamnaceous bush. On the hill slopes at Uvalde it occurred on the low Acacia berlan- landieri, which there replaced the ubiquitous mesquite, while at the foot of the hills it was taken on tall rank green weeds. On the Del Rio hills, which were clothed with low Acacia, numerous other thorny bushes, occasional arborescent yuccas, and several species of Opuntia, the present species was secured in catclaw (Mimosa sp.) and other thorny bushes. The specimens taken by Pratt at San Antonio were from Opuntia lindheimeri, while at Corpus Christi and in Maverick County it occurred on plants of the same genus. _ Synonymy.—The erroneous determination of the species as brevi- hastata by Hunter, Pratt and Mitchell, we have corrected by an ex- amination of the original material. Remarks.—The most striking diagnostic characters of this species are, the abbreviate subgenital plate of the male and the combination of a short ovipositor and the narrow interspace between the tegmina in the female. The latter space never equals the width, and rarely as much as half the width, of a single tegmen. The male subgenital plate is seen, when compared with that of brevihastata, to be broader, not at all produced mesad, except that the plate is in general narrowed in that direction, the distal margin is V-emarginate mesad with slight subtruncate sections laterad, which are flanked at the angles with very short blunt subobsolete tubercles. The cercus has the extremity more acute than in brevihastata, also more elongate, with the character of the median tooth slightly different. The ovipositor has the ventral margin straighter, the proximal depth less in propor- tion to that of the apex and the apex slightly more acute. The more ovate eye is immediately apparent on comparison. Specimens Examined.—54; 21 males, 33 females. Pecos. August 18, 1 9, [Scudder Coll.]. Del Rio, Valverde Co., Texas, elev. 900—-1,100 feet, August 22-23, 1912, (R. and H.), 3 o',, 1 9. Uvalde, Uvalde Co., Texas, elev. 1,000-1,100 feet, August 21-22, 1912, (R. and H.), 3 o; last week of July, 1 &, [Seudder Coll.]. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 105: San Antonio, Bexar Co., Texas, August 15-16, 1912, (R. and H.), 2,5 2; October 29, 1905, and June 16, 1908, (F. C. Pratt; on Opuntia lindheimeri), 3 2, [U. 8. N. M.]. Southwest Texas, November, 1884, (Schaupp), 1 o7, 3 92, [Hebard Coll.]. Maverick Co., Texas, May 15, 1906, (J. D. Mitchell; on Opuntia), 1 9,(U.S. N. M_]. Carrizo Springs, Dimmit Co., Texas, June, 1885, May, 1886. (A. Wadgymar), 1 &, 1 9, [Hebard Coll.]. Corpus Christi, Nueces Co., Texas, October 20, 1905, (F. C. Pratt), otele? AUS. Ne ME]. Ringgold Barracks (now Rio Grande City), Starr Co., Texas, (Schott), 1 o, [Seudder Coll.]. Laguna del Gato, Hidalgo Co., Texas, elev. 175-200 feet, August 6, 1912, (R. and H.),6 #,9 2. Tyen, allotype and paratypes. Mission, Hidalgo Co., Texas, elev. 138 feet, August 5-6, 1912, (Ri and! H:); 2c", 3 2. Victoria, Tamaulipas, Mexico, July, 1 o&, 1 2, [Hebard Coll.]; December 10, 1909, (F. C. Bishop), 2 2, [U.S. N. M.]. Tamos, Vera Cruz, Mexico, December 7, 1909, (F. C. Bishop), 12 SUES: IN. Mi: Dichopetala brevihastata Morse. 1902. Dichopetala emarginata Scudder and Cockerell (not of Brunner, 1878), Proc. Davenp. Acad. Sci., LX, p. 51. [Mesilla Park, New Mexico.|} 1902. Dichopetala brevicauda Scudder (not Dichopetala brevicauda Scudder, 1900), bid., p. 51, pl. IV, fig. 1. [Riley’s Ranch, Mesilla Valley, New Mexico; Mesilla Park, New Mexico; Mexico.] 1902. Dichopetala brevihastata Morse, Psyche, IX, p. 381. (To replace D. brevicauda Seudder, 1902.) 1907. Dichopetala brevihastata Rehn, Proc. Acad. Nat. Sci. Phila., 1907, p. 56. [Carr Canyon, Huachuca Mts., Arizona.| 1907. Dichopetala levis Rehn, Ibid., p. 56, fig. 10. [Carr Canyon, Huachuca Mts., Arizona.] 1909. Dichopetala brevihastata Rehn and Hebard, Jbid., 1909, p. 167. {Mouth of Dry Canyon, Sacramento Mts., New Mexico.| This species needs comparison with no form of the genus except castanea, under which species the important differential characters are given. Types: 1,2 2. Riley’s Ranch, Mesilla Valley, New Mexico, August 16 (Cockerell); College campus, Mesilla Park, New Mexico, on Atriplex canescens, August 2 (nec 7), (Cockerell); Mesilla Park, New Mexico, September 11 (Cockerell). Single Type here Designated: 3; College campus, Mesilla Park, Donna Ana County, New Mexico. August 2,1898. (Cockerell; on Atriplex canescens.) (Scudder Collection.] 106 PROCEEDINGS OF THE ACADEMY OF [Jan., Description of Type—Size medium;?! form moderately slender; surface subglabrous. Head with the greatest width of the eye contained more than-one and one-half times in the,length of the same; antenne* (in perfect individuals) five times the length of the body. Pronotum with the caudal section of the same equal in width to about two-thirds the length of the disk, transverse V-shaped sulcus placed on the middle of the disk; lateral lobes of the pronotum with the greatest depth contained slightly more than one and one-half times in the dorsal length of the same, cephalic margin of the lateral lobes slightly arcuate, ventro-caudal angle and caudal margin of the lobes moderately arcuate. Tegmina slightly longer than the dorsal length of the pronotum, distal margin obliquely arcuato-truncate, the lateral angle moderately rounded; stridulating vein slightly oblique. Disto-dorsal abdominal segment with a recurved trigonal production which is very deeply V-shaped emar- Fig. 22.—Dichopetala brevihastata Morse. Lateral view of type. (X 2.) ginate mesad, laterad of which this production is rounded; supra-anal plate trigonal with the apex narrowly truncate; cerci as in castanea in the form of the proximal portion and general curve of the distal section, median tooth slightly less divergent dorsad from the main body of the cercus, distal extremity of the cerci moderately produced, caniniform, but not spiniform; subgenital plate longer than the proxi- mal width, lateral margins moderately rect-convergent for the greater portion of their length, thence briefly subparallel, the width of the subequal portion about half of the broad proximal portion, distal margin wholly V-emarginate, only narrowly rounded lobes re- maining laterad, a very weak medio-longitudinal carina present ven- trad. Cephalic femora proportionately as in castanea, but slightly slenderer. Medianfemora very slender. Caudal limbs as in castanea. Allotype here Selected: 9; Riley’s Ranch, Mesilla Valley, Donna "(On account of the close relationship of this species to castanea, only characters showing some difference from those of the latter species are here mentioned. 2 Imperfect in type. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 107 Ana County, New Mexico. August 16, 1898. (Cockerell.) [Seud- der Collection. | Description of Allotype-—Characters not specifically mentioned are not markedly different from those of the male sex or of the female of the closely allied D. castanea. Pronotum with the whole form slightly more compressed than in castanea. Tegmina small, little broader than the apparent length of the same, reaching the caudal margin of the metanotum, distal margin broadly rotundato-rectan- gulate, the interspace between the tegmina subequal to the width of a single tegmen. Ovipositor slightly surpassing the length of the head and pronotum together, considerably arcuate, proximal depth distinetly less than a third of the length of the ovipositor, ventral margin moderately arcuate, not at all flattened, dorsal margin very slightly more arcuate distad than proximad, apical margin proper acute-angulate, armed on the distal third of the dorsal margin with seven teeth, ventral margin considerably arcuate distad, armed with nine spines, those of both margins similar to those found in castanea : subgenital plate with the lateral angles produced into distinct spini- form lobes equal in length to the main depth of the plate, median emargination of the plate much as in castanea. Cephalic femora slightly inferior to the combined length of the head, pronotum and tegmen. Caudal femora slightly longer than the body, appreciably weaker than in castanea. Paratypic Series.—We have examined all three specimens on which Scudder based the species, one being selected as the single typé, another as the allotype, the third (Mesilla Park, September 11, Cockerell) remaining as a paratype. Measurements (in millimeters). al Mesilla Park, Srenmiore Canyon, Average of three; N.M. (Type). Pima Co., Ariz. Tumamoce Hill, Pima Co., Ariz. Length of body ...* 1S EY 14.8 13. (12.5-13.6) Length of pronotum.... 4.3 4. 3.9( 3.8 4. ) Greatest caudal width of disk of pronotum 5 OE agit 2.9: 222= 3. ) Length of tegmen....... 4.4 4.2 3.8 ( 3.5— 4.1) Greatest width of discoidal and anal fields of teg- men Fee ee oars 5 Boe 3. Il 2EOiG249= 3.) Length of cephalic femur 9.2 10. 8.3.( 7.7= 9.2) Length of median femur... 10.3 10.8 a We (Gi 0) )) Length of caudal femur 21.3 22.8 19.1 (17. —21.3) *s Scudder’s original measurements of this specimen are: body, 14; pronotum, 3.7; cephalic femur, 11; caudal femur, 21. The discrepancy in body length is 108 PROCEEDINGS OF THE ACADEMY OF [Jan.,. * opto Carr Canyon, Average of six; Average of six; Cochise Co., Marathon, Brews- Laguna del Gato, Ariz. terCo., Tex. . Hidalgo Co., Tex. Length of body.................138. 15.8(18.7-17.7) 16.33.8-18. )- Length of pronotum............ 3.9 3.8(3.6- 4.) 4. (3.9 4.2): Greatest caudal width of disk of pronotum.............-2.6 2:5( 2.5— 2.6) 2.5( 2.3- 2-7)- Length of tegmen........... 4. 4.1(3.9- 4.4) 4. ( 3.9- 4.5) Greatest width of discoi- - dal and anal fields of tegmen..... : Orly 3 (22,9132) ur 2). Bi a2 oe) Length of cephalic femur. 9. 8.9( 8.5- 9.5) 9.5( 9.1- 9.9) Length of median femur.. 9.8 9.4( 9. -10.1) 9.9( 9.1-10.3) Length of caudal femur...20.3 19.9 (19.3-20.5) 21. (19.9-22.2) oios Average of three; Montelovez, Coahuila,. Mex. Length of body... . 14.8 (12.8-16.3) Length of pronotum ; ; 4.3( 4. -— 4.7) Greatest caudal width of disk of pronotum 2.8 ( 2.7— 2.9) Length of tegmen.. 3.7( 3.4 4.2) Greatest width of discoidal and anal fields of tegmen........... ico sla k = oem Length of cephalic femur . 9.6( 8.9-10. ) Length of median femur . 10.6( 9.8-11. ) Length of caudal femur 22.5 (21. —23.5) . : 29 Riley’s Ranch, Mesilla Park, New Mex. New Mex. (Allotype.) (Paratype.) Length of body (exclusive of ov pare sath Alas) 18.3 Length of pronotum... Sat: 5s 6.4 Greatest dorsal width of pronotum. 3. 3.9 Apparent length of tegmen’ WA 1.4 Greatest width of tegmen me Pat Length of cephalic femur 8. 9. Length of median femur... 8.5 10.5 Length of caudal femur 7A 24. Length of ovipositor 8.2 8.8 probably due to the same being taken from different points, that of the cephalic femur on account of his measurement including the coxa, but the pronotal difference is incomprehensible to us. The only explanation which seems at all likely is that the length was taken along the lateral angles of the disk instead of along the median line, as we are accustomed to take the latter measurement to- get the maximum. 1914.] Average of three; Sycamore Canyon, Uvalde, Uvalde Co., Tex. Length of body » 29 Length of pronotum 6. ‘Greatest dorsal width of pro- notum.... Te Apparent length of tegmen 2 ‘Greatest width of tegmen... 2.5 Length of cephalic femur 10.4 Length of median femur FA Ds Length of caudal femur ...........: 26.7 Length of ovipositor 9. Average of six; Laguna del Gato, Hidalgo Co., Tex. NATURAL SCIENCES OF PHILADELPHIA. ae Average of three; Tumamoc Hill, Beeville, Gregory, San Laredo, 109 Pima Co., Ariz. Pima Co., Ariz. Length of body xc e of ovipositor)....... / 205720. —2-5) 21.4 (20.7-22.7) Length of pronotum D22)/ (5: = 5.5) 5.3 ( 5.1— 5.5) ‘Greatest dorsal width of pro- notum.... olen (asa oe) Bole Gss = 382) Apparent length of tegmen... eee On( eae ea)) 12)@ Ie = 156) ‘Greatest width of tegmen..... 2a (Gale See) 2.2( 2.1- 2.6) Length of cephalic femur........... 8.7( 8.5- 8.9) 9. ( 8.8 9.3) Length of median femur... 9.9( 9.7-10. ) 9.6( 9.2-10. ) Length of caudal femur............ ose (21 . 9-23 .7) 23.3 (23. —24.1) Length of ovipositor............. . 8.8( 8.7— 9. ) 8.6( 8.5- 8.9) ) 9} Average of six; Marathon, Brewster Co., Tex. Length of body (exclusive of peotton) » 2002107 58-22. ) Length of pronotum.... 5.38( 5. — 5.8) ‘Greatest dorsal width of pr onotum... 3.2( 3.1- 3.5) Apparent length of tegmen 0.000... 12) Wels. 5) ‘Greatest width of tegmen................... Qel (62222) Length of cephalic femut.......... 8:9( 8. = 9.7) Length of median femur . 8.9( 8. - 9.7) Length of caudal femur. 22.8 (21.9-24. ) Length of ovipositor.... 8.7( 7.8 9. ) one Bee Co., Patricio Co., W aap Go., Tex. Tex. Tex. 19. 20.2 18.7 5.4 eT oy 3. 33415) Be) ard ie 1.4 Pye Ae 1.9 9. 9.1 8.2 10. 10.3 9.2 Doe 22.8 PAP. 8.5 8.5 7.6 Pe Montelovez, Coahuila, Mex. Average of three; Length of body . 20.5 (18.5-23. ) 18.1 (15.3-21. ) Length of pronotum........0..... oO (Geo Neo. Weer (a= (sar) Greatest dorsal width of pro- HOWE e eee ee Ae o= 3-0) Beale 3e2= 928) 110 PROCEEDINGS OF THE ACADEMY OF [Jam:, 2 Average of six; : Average of three; Laguna del Gato, Montelovez, Hidalgo Co., Tex. Coahuila, Mex. Apparent length of tegmen... 1.2( 1.1— 1.5) IAs = 28> 158) Greatest width of tegmen Qe 2a Deo) 25422. — 227) Length of cephalic femur... 9. ( 8.7— 9.3) 9.4( 8.8-10. ) Length of median femur... 10.2( 9.8-11. ) 10.3 (8. 7-112) Length of caudal femur.......... 23.3 (22.4-24.7) 24.6 (23.3-26.5), Length of ovipositor 8.2( 7.7- 8.7) 9.1( 8.6— 9.8) oS Average of four; Jaral, Coahuila, Mex. ; Length of body ........ : 20. iP 4) —? bo fo) | bo —_ or ~~ Length of pronotum 5.3( 5.2- 5.8) Greatest dorsal width of pronotum 3.0 (d02-e050) Apparent length of tegmen Pag SS) Greatest width of tegmen 223) (Bie eeO) Length of cephalic femur 8.4( 8. —9. ) Length of median femur Lor 3Cor =10s8) Length of caudal femur. . 21.1 (20.6-22.2) Length of ovipositor.. 8.1 (¢ 8:= 823) From the above measurements it is evident that considerable variation, both geographic and individual, is present in this species. The geographic evidence shows that material from southern Arizona and the elevated portions of western Texas averages smaller than the series from southern Texas (Laguna del Gato) and Coahuila,. Mexico, particularly in the femoral length. The Jaral, Coahuila, females are not, however, of the same general proportions as the Montelovez, Coahuila, specimens of that sex, and, although we are unable to locate the latter locality, possibly altitude may be respon- sible, Jaral being near the four-thousand-foot contour line. It is possible that the character of cover and richness of same may be a factor in determining the amount of geographic variation. The individual variation is considerable in all of the series, but the only really puzzling feature is the occurrence of large females with heavy pronoti. In the Montelovez series these are more numerous (three out of four) than elsewhere, but the paratypic female from Mesilla Park and to a lesser degree the Uvalde specimen are of this type- This point is treated in greater detail under Remarks. Color Notes.24—The series of this species exhibits a very considerable 24 The remarks here set forth on color variation have been made entirely from material which has been stuffed or which is considered by the authors to have retained in a great measure the color tones of the living insect. 1914.| NATURAL SCIENCES OF PHILADELPHIA. 111 amount of variation in intensity of color pattern and color tone. The pattern is considerably recessive and intensive in the male, but in the female the recessive extremé is even greater than in the male, although the intensive is not as decided. The normal pattern as found in both sexes is as follows: a dorsal color involving the occiput, fastigium, dorsum of pronotum, and dorsal surface of the abdomen: pale postocular bars extending to the dorsal base of the cerci; a lateral and ventral color involving face, gene, lateral lobes of prono- tum, pleura, and lateral and ventral aspects of abdomen. In intensive individuals the dorsal color is decidedly darker than the lateral and ventral color, in average individuals in part at least so, in recessive individuals nearly (co) or quite (9) similar in tone. Strongly intensive individuals generally have the distal extremity of the caudal femora infuscate, but this is not a rule, as occasionally average and rarely moderately recessive individuals have this marking present. Male. Dorsal color ranging from ochraceous tawny and sudan brown to claret brown and maroon, solid and uniform on head and pronotum, generally restricted to the lateral sections of the dorsum of the abdomen and in intensive individuals blackish next to the pale lateral lines, the distal margins of the abdominal segments edged with the pale color of the lateral bars, this edging subobsolete in the recessive specimens and narrowing mesad in all, a more or less distinet dark beading characterizing the same margins. Pale lateral bars ranging from cream white to very pale orange yellow always indicated in the male. Lateral and ventral color ranging from cream buff and pinkish buff to brownish olive, the green complimentary phase ranging to civette green. Face in the extreme green condition with broad paired vertical bars of the pale color of the lateral bars placed ventrad of the eyes and antennal bases, the median pair weakly continued to the clypeus; eyes varying from cameo brown to bay; antenne varying from raw sienna and mars yellow to madder brown, weaker distad. Lateral lobes of the pronotum with a broad margin of the pale lateral color on the ventral and greater portion of the caudal margin, the dorsal section of the lateral lobes washed more or less with the dorsal color; pale paired lines more or less severed at the median sulcus. Tegmina with the marginal field of the color of the pale lines; region of the humeral trunk varying from orange rufous to morocco red; remainder of tegmina with the base color blackish brown, the venation of the same tones as the pale lines. Limbs of the lateral color more or less washed with the dorsal color, in intensive individuals decidedly so, the blackish-brown infuscation 112 PROCEEDINGS OF THE ACADEMY OF {Jan., of the distal extremity of the caudal femora (when present) being generally correlated with a similar infuscation of the dorsal face of the caudal tibize, the caudal femora, which are infuscate distad, having the adjacent section of the distal half more or less inclined toward tawny or yellow ochre. Pattern of caudal femora always present. Female. Dorsal surface varying from being concolorous with the lateral color through buffy citrine to russet, the intensive type having the abdominal coloration largely produced by stipplings. The segments of the abdomen in these contrasted specimens are marked as in the male, but with decided blackish lateral patches which extend more or less toward the median line proximad on each segment. Color of the pale lines (when present) varying from eream white to pale lumiere green and warm buff. Lateral color varying from clay color (in this specimen pale chalcedony yellow on head and pronotum) through course green to hellebore green, the extreme recessive green condition being without distinct pale margins to the lateral lobes and having the whole coloration uniform except for a darkening of the distal margins of the ovipositor. Eyes and antenne as in the male. Limbs varying as in the male, but in the recessive green type uniform with the general coloration and with the paginal pattern weak. Tegmina varying from a type nearly uniform with the lateral color, to one with the costal and distal margins of the color of the pale lines, humeral trunk claret brown and remainder blackish brown with pale venation. Ovipositor varying from uniform with mass color*(recessive green and brown types) with margins edged distad with clove brown to blackish brown to elm green washed dorsad with hazel and teeth blackish (extreme intensive type). Geographically considered, the coloration of the species shows plasticity in some localities and constancy in others. The Arizona and New Mexico individuals are all more or less recessive, the large Laguna del Gato series decidedly so, while the Marathon series is about evenly divided (recessive, intermediate, and intensive). The five Garden Springs adults are chiefly intensive, as are four of the five .Kent individuals. The Uvalde and Laredo specimens are recessive, while the Beeville and Gregory representatives are average. The Montelovez and Jaral individuals are almost all intensively colored. From the basis of the Arizona, Marathon, Kent, Garden Springs, and Laguna del Gato series it seems possible that direct and reflected 1914.| NATURAL SCIENCES OF PHILADELPHIA. 113 light may be a factor in determining the intensity of the color pat- tern, the percentage of intensive specimens being greatest from those localities known to us where the cover is densest, more light resistant and the surface conditions less favorable for reflecting light. Distribution—This species has the widest distribution of any in the genus, its range extending from southern New Mexico (Sacra- mento Mts.), south to southern Coahuila, Mexico (Jaral), west to the Baboquivari Mts., central southern Arizona, and east to the Texas coast at Gregory (San Patricio Co.). Its vertical range is from practically sea-level at the last-mentioned locality to as high as 4,800 feet in the Sacramento Mountains. Its zonal range appears to be entirely Lower Sonoran. As far as known, it does not extend into the region of the Edwards Plateau in central Texas and does not occur in the mountains of Trans-Pecos, Texas. Scudder in his original description stated that he had specimens from Mexico, this probably referring to the Montelovez specimens examined by us, as these were contained in his collection. Biological Notes.—The present species was fairly numerous on ereosote-bush (Covillea tridentata) at Dry Canyon, Sacramento Mountains, New Mexico, and occurred on the same plant on Tumamoc Hill, Arizona, while at the latter locality it was also found on the ground in short, dry, yellow grass. On the slopes of Sycamore Canyon, Baboquivari Mountains, Arizona, it also occurred in grasses and was taken from Acacia sp. At Marathon, Texas, it was generally common in various low bushes and grasses, the males, particularly, often sprawled out in a loose manner somewhat reminding one of phalangids or harvest-men, while at Kent and Garden Springs it occurred in similar situations. At Beeville and Uvalde it was taken from green weedy plants, at Gregory it was found in the green tangle about a mesquite clump, while it was beaten from a low bush on a sandy slope at Laredo. At Laguna del Gato it was taken rather commonly with D. castanea on a low, very green rhamnaceous shrub (probably Condalia obovata). A correlation of the dates on the present series brings out some very interesting points on the time of maturity of the species. The earliest dates on which adults were secured are July 28 at Beeville, July 30 at Gregory, August 6 at Laguna del Gato, August 7 at Mesilla, August 10-12 at Laredo, and August 16 at Mesilla. At Dry Canyon, New Mexico (elevation 4,800 feet), on July 13 nymphs not more than half grown were not uncommon. while at Marathon and Garden Springs on August 26-27 and September 2, respectively, nothing but tS) 114 PROCEEDINGS OF THE ACADEMY OF [Jan., nymphs were seen, while on September 11 both nymphs and adults were taken at Garden Springs, and adults outnumbered the nymphs at Marathon on September 12-13. From this it is apparent that in the low, warmer Rio Grande Plain and interior valleys (Mesilla) the species matures at least a month earlier than in the higher regions of the plateau. The latest date is November 1-3 at Jaral, Coahuila. Morphological Notes.—The subgenital plate of the male varies in the depth of the emargination of the distal margin and also in the degree of acuteness of the flanking angles of the same, in some exam- ples these latter being quite acute and in others appreciably rounded. The ovipositor varies slightly, almost inappreciably, in general curvature of the margins and little in length, but the relative depth varies very decidedly, particularly in all of the Mexican specimens, which, however, are almost or quite equalled in this respect by individuals of the sex from Kent and Garden Springs. The lateral angles of the subgenital plate of the female vary in the degree of angulation, in one extreme being practically rectangulate, in the other with subspiniform. extremities. While this latter variation is frequently correlated with that in the depth of the ovipositor, the rectangulate type with the deeper ovipositor, the more spiniform type with the narrower ovipositor, this relationship is not at all absolute. Synonymy.—By an unfortunate lapse, Scudder, when originally describing this species, applied to it the same name (Dichopetala brevicauda) that he had given two years previously to a species now known to belong to the genus Aretha, as explained by Morse who renamed the present form. The name brevihastata, proposed by Morse to replace the preoccupied brevicauda, cannot, in our opinion, be credited to Scudder, notwithstanding Morse’s statement that the name was suggested by him, as the note is entirely by Morse without a direct quotation from Scudder. The naming of a species by proxy does not seem possible under present-day rules. In the same paper in which the last D. brevicauda was described, Scudder and Cockerell recorded D. emarginata from Mesilla Park on Atriplex, a locality and situation from which they, a few lines below, described brevicauda. There are no specimens in the Scudder Collection labelled emarginata from Mesilla Park or in the National Museum, and Prof. Cockerell can give me no additional information. In view of these facts and also that true emarginata is not found within hundreds of miles of that locality, as far as known only brevihastata occurring in that region, it seems perfectly logical to assume that the determination 1914.] NATURAL SCIENCES OF PHILADELPHIA. 115 of Mesilla material as belonging to two species 1s due to a compilation of determinations made at different times, one before the recognition of brevihastata, and that both records relate to the same species. The species D. levis, erected by the senior author on a single female, is a synonym of the present species. The peculiarities of the ovipositor of the type of /evis, the unarmed margins of which sug- gested the specific name, we now know are due to the immaturity of the individual. The latter is in the stage immediately preceding maturity and its proper relationship to the other material now in hand is very evident. The more robust character of the limbs and smaller size of the type of levis are similarly explained. Remarks.—The variation in the length of the pronotum as found in certain females has been touched upon above under the measure- ments, this being the most decided fluctuation from the more general type found in the species. After considerable study and considera- tion from different view points, we have concluded that this phase eannot be separated from the more typical one of brevihastata, that it is a fluctuation occurring anywhere in the range of the species, although more numerous in certain regions than in others, and is approached by a few specimens not typical of the same and again not exactly similar to the type of brevihastata. There is, of course, a possibility that future work may show the advisability of recognizing the larger pronotum type as distinct, but we do not feel that the necessary evidence will be forthcoming. Specimens Examined.—103; 47 males, 46 females, 4 male nymphs, 6 female nymphs. Syeamore Canyon, Baboquivari Mts., Pima Co., Arizona, elev. 3,700-4,700 feet, October 6-9, 1910, (R. and H.),2 47,3 9. Tumamoc Hill, Tucson Mts., Pima Co., Arizona, elev. 2,400-3,092 feet,-October 3-4, 1910, (R. and H.), 3 o',3 Q. Carr Canyon, Huachuca Mts., Cochise Co., Arizona, August, 1905, (H. Skinner), 1 o, 1 9 nymph. ‘Type (2 n.) of D. levis, [A. N.S. P.]. Riley’s Ranch, Mesilla Valley, New Mexico, August 16, (Cockerell), 1. Typs, [Scudder Coll.]. Mesilla Park, New Mexico, August 7 and September 11, (Cock- erell), 2 2. Allotype and paratype, [Scudder Coll.]. Dry Canyon, Sacramento Mts., Otero Co., New Mexico, elev. 4,800 feet, July 13, 1907, (R. and H.), 1 2 nymph. Marathon, Brewster Co., Texas, elev. 3,900-4,160 feet, August 26-27 and September 12-13, 1912, (R. and H.), 16 o', 14 9,2 f and 2 29 nymphs. 116 PROCEEDINGS OF THE ACADEMY OF [Jan., Garden Spring, Brewster Co., Texas, September 2 and 11, 1912, (R. and H.), 3 #, 2 9,2 2 nymphs. 5 Kent, Culberson Co., Texas, elev. 3,900-4,200 feet, September 17-18, 1912, (R. and H.),3 0’, 2 @. Uvalde, Uvalde Co., Texas, elev. 1,000-1,100 feet, August 21-22, 1912, (R. and H.), 1 @. : Beeville, Bee Co., Texas, July 28, 1912, (R. and H.), 1 2. Laredo, Webb Co., Texas, elev. 500-550 feet, August 10 and 12, 1912, (R. and H.), 1 re gy kes Gregory, San Patricio Co., Texas, July 30, 1912, CEs) sie O"7 Laguna del Gato, three miles west of Sam Fordyce, Hidalgo Co., Texas, elev. 175-200 feet, August 6, 1912, (R. and H.), 11 #7, 8 9, 1 #@ and 1 Q nymph. Montelovez, Coahuila, Mexico, September 20, (Palmer), 6 7,4 92, [Scudder Coll. and U. 8. N. M.]. Jaral, Coahuila, Mexico, November 1-3, 1909, (J. Friesser), 4 9, [Field Museum of Natural History]. Dichopetala gladiator n. sp. 1901. Dichopetala emarginata Rehn (not of Brunner), Trans. Amer. Entom. Soc., X XVII, p. 335. [Texas.] 1912. ” Dichopetala emarginata Hunter, Pratt and Mitchell (not of Brunner), Bull. Hes Bureau of Entom., U. 8. Dept. of Agric., p. 50. [Hebbronville, Texas 1912. Dichopetala brevihastata Hunter, Pratt and Mitchell (not of Morse), Tbid., p. 50. (Part) [Alice, Texas.] The present form is closely related to the following species, D. emarginata Brunner, from which, however, it can easily be separated by the more ample tegmina of the male, the more longitudinal lateral lobes of the pronotum of the same sex, the less crassate tooth to the male cereus, the less deeply and sharply angulato-emarginate apex of the subgenital plate of the same sex, the proportionately longer limbs of both sexes and the much longer, slenderer, and more arcuate ovipositor. Type: o; Lyford, Cameron County, Texas. August 6-7, 1912. (Rehn and Hebard.) [Hebard Collection.] Description of Type.—Size rather large; form moderately elongate. Head with the occiput but little declivent to the fastigium and antennal scrobes; fastigium moderately compressed, sublamellate at the apex, subsuleate proximad, extremity partially in contact with facial fastigium; eyes prominent, ovate, their depth contained one and one-third times in that of the infra-ocular portion of the gene; antenne with the proximal joints rounded, very long, over four times as long as the body. Pronotum subsellate, the dorsum 1914.] NATURAL SCIENCES OF PHILADELPHIA. 117 bisinuate when seen from the lateral aspect; disk of the pronotum very narrow mesad, the median width about half the caudal width of the same; lateral margins of the disk regularly converging to the middle of the same, thence regularly diverging caudad, indicated caudad by a rounded angle, but elsewhere by color only, in section the cephalic portion of the disk is arcuate, the caudal section deplan- ate; cephalic margin of disk truncate, caudal margin of same gently arcuate, distinctly elevated; transverse suleuS V-shaped, obliquely Fig. 23.—Dichopetala gladiator n. sp. Lateral view of type. (X 2.) severing the margins of the disk mesad, on the lateral lobes extending considerably ventrad; lateral lobes with the greatest depth contained twice in the dorsal length of the same, cephalic margin subtruncate, ventro-cephalic angle rotundato-rectangulate, ventral margin arcuato- sinuate, slightly emarginate cephalad, ventro-caudal angle and caudal margin gently arcuate, dorsad rounding into the caudal mar- gin of the disk, the surface of the lobes strongly drawn in to the Figs. 24 and 25.—Outline of left cercus of topotype of Dichopetala emarginata (24) and type of D. gladiator (25). (x 10.) lateral margins of the disk. Tegmina slightly longer than the disk of the pronotum, greatest width of the discoidal and anal fields of the tegmina subequal to the length of the anal field; costal margin moderately arcuate, distal margin obliquely arcuato-truncate, disto- caudal (literally apical) angle well rounded, sutural margin arcuate- obtuse-angulate, slightly sinuate distad of the extremity of the stridulating vein; marginal field moderately broad, discoidal field 118 PROCEEDINGS OF THE ACADEMY OF [Jan., not particularly broad, expanding distad, anal vein moderately arcuate, stridulating vein strongly arcuate proximad, straight distad, tympanum proper rather large, subtrigonal. Abdomen with the disto-dorsal abdominal segment having the margin sinuately arcuato- truncate, moderately arcuato-emarginate laterad; supra-anal plate tongue-shaped; cerci regularly arcuate inwards, the proximal portion of the shaft rather heavy, somewhat tapering, median lobe placed on the dorsal surface, short, depressed, slightly concave ventrad, when seen from the dorsum with the external margin arcuate, internal margin straighter, distal extremity of the lobe weakly subtruncate, the lobe narrowing along the same lines as the proximal portion of the shaft of the cereus, distal portion of shaft more sharply arcuate, subdepressed, acute subaciculate, the distal portion (distad of tooth) subequal in length to the proximal portion; subgenital plate moderately produced, elongate, lateral margins moderately converging on median half, subparallel on distal fourth, distal margin with a V-shaped median emargination which occupies not more than Figs. 26 and 27.—Ventral outline of subgenital plate of topotype of Dichopetala emarginata (26) and type of D. gladiator (27). (X 4.) one-half of the margin, laterad of the emargination subtruncate, distal portion of the ventral surface of plate tricarinate, mesad and laterad, the lateral carine following the lines of the distal fourth of the lateral margins, the oblique portion of the lateral margins thick- ened and subcarinate. Cephalic femora about two and one-half times the length of the disk of the pronotum; cephalic tibiz with the foramina elliptical. Median femora but slightly less than twice the length of the head and pronotum. Caudal femora nearly one and one-half times the length of the body, moderately inflated in proximal half. Allotype: 2; Same data as the type. Description of Allotype-—The following characters are those of difference from the type. Size large; form moderately robust. Head with eyes slightly less prominent and more elliptical than in the male, their depth contained one and one-half times in that of the infra-ocular portion of the gene. Pronotum with the disk less deplanate than in the male and nearly straight when seen from the 1914.] NATURAL SCIENCES OF PHILADELPHIA. 119 lateral aspect, form of disk similar to that of male but less constricted mesad; cephalic margin emarginato-truncate, caudal margin moder- ately arcuate; lateral lobes with their greatest depth contained slightly less than twice in the dorsal length of the same, margins of the lateral lobes as in the male. Tegmina semi-ovate, reaching the caudal margin of the metazona, the greatest width distinctly greater than the apparent length, interspace between the tegmina about Figs. 28 and 29.—Dorsal outline of head, pronotum, and tegmina of topotype of Dichopetala emarginata (28) and allotype of D. gladiator (29). (X 2.) two-thirds that of a single tegmen. Cerci brief, conical, acute; Ovipositor subequal to the median femora in length, arcuate in general form, moderately slender, tapering in proximal two-thirds, ventral margin less arcuate than the dorsal one aside from a decidedly arcuate distal portion, distal third of dorsal margin with eleven to twelve teeth increasing in size distad, ventral margin with nine to ten teeth on distal fifth, the extreme distal one slightly recurved; sub- Figs. 30 and 31.—Outline of ovipositor of topotype of Dichopetala emarginata (30) and allotype of D. gladiator (31). (x 4.) genital plate almost divided to the base by a V-shaped emargination, the lateral sections developed into very acute trigonal lobes. Cephalie femora slightly less than twice as long as the disk of the pronotum. Median femora twice the length of the disk of the pronotum. Caudal femora almost two and one-half times the length of the cephalic femora. 120 PROCEEDINGS OF THE ACADEMY OF [Jan., Paratypic Series.—We have before us a paratypic series of twenty- three males and twenty females from the type locality—Lyford, Cameron County, Texas. Measurements (in millimeters). Average and extremes Lyford. of six Lyford spec. (Type) (Type and Paratypes) Length of body... ay ee tee ea) 16.7 (15.9-17.8) Length of pronotum... ee “ 4.5 4.4( 4.2- 4.6) Greatest dorsal width of disk of pronotum 2.6 2.7. (225-35) Length of tegmen.. 4.6 4.4( 4.3- 4.8) Greatest width of discoidal and anal fields Ole PIVEN sever cerca one ‘ ( a Length of cephalic femur... atin ne tee ee teal (O03) 10.4 (10. —10.9) Length of median femur... a perter nl Kael 11.1 (10.8-11.4) Length of caudal femur concn LOO 23.3 (22.8-23 .8) oskon Carrizo Springs Ce Wades. {(Hebard (U.S.N.M.] [U.S.N.M.] Coll.) Length of body. ener : Ve iWfe 16.2 Length of pronotum 4.8 4.7 4.7 Greatest dorsal width of disk of pro- notum hah Dees eRe, 2.4 2.9 Length of tegmen... 4.1 4. 4.4 Greatest width of discoidal and fae fields of tegMeN .... ecco He enti eke ee 3.4 3.2 Length of cephalic femur. . 10.2 9.2 9.7 Length of median femut......... LE 10. 10.3 Length of caudal femur........... D3 ie2, 22.5 22.5 C8 Average and extremes of six Lyford spec. ‘ Lyford. (Allotype and (Allotype) Paratypes) Length of body (exclusive of ovipositor).... 22. _ 21.9 (20.2-22.6) Length of pronotum........ 520 5.9 ( 5.6— 6.3) Greatest dorsal width of disk of pronotum. 3.1 3.3( 3.1- 3.6) Apparent length of tegmen 1.4 1.4( 1.2— 1.7) Greatest width of tegmen Le ae te 2.2;(-2.1= 259) Length of cephalic femur...... 10.5 10.8 (10.3-11.1) Length of median femur es 12. (11.3-12.7) Length of caudal femur ; 7 fe 27.4 (25. 5-28 .5) Length of ovipositor......... 11.2 11. (10. -11.8) 1914] NATURAL SCIENCES OF PHILADELPHIA. = 121 2 Carrizo Springs. {Hebard Coll.] Length of body Gace of oviposi- tor)... A dpa ste eee OD 18. 20.8 Length of pronotum... 6.8 6.2 6.5 Greatest dorsal width of disk of pro- TIVO ULLTEN Se eet ccs av 3.4 Jel Apparent length of tezmen SL 1.6 le Greatest width of tegmen Sele 2.4 220 Length of cephalic femur... supe ulate} 10.3 Length of median femur... peed 11.6 Length of caudal femur................. 28. 26.2 Path Meniti Of OVAPOSILOL.. 3 sh Satnan LOT 14.5 14.5 Males from Wades, Cotulla, and Carrizo Springs have the pronotum slightly longer, tegmina slightly shorter, and femora very slightly shorter than Lyford males. Females from Carrizo Springs have the general size slightly less, the pronotum appreciably longer and ovipositor distinctly longer than in Lyford individuals of the same sex. The most striking variation in measurements is in the length of the ovipositor, which varies geographically more than individually, _ the Carrizo Springs females having this actually and proportionately far exceeding the measurement of the Lyford specimens. Color Notes——The following notes have been based wholly on material which was stuffed in the field or which we have every reason to believe has retained its original coloration. The recessive and intensive extremes are considerably different—-in fact, decidedly different in the male sex. Instead of describing a dorsal and lateral color, it seems best in the present species to speak of the tones as a general color, a pronotal wash, a pale pattern, and an abdominal infuscation. General color of male varying from cinnamon buff to yellow ochre, passing in certain individuals to apple green. The pronotal wash varies from uniform with the general color through sanford’s brown to claret brown, the area covered by the same consisting of the occiput, disk of the pronotum, more or less of the dorso-cephalie and dorso-caudal portions of the lateral lobes of the same, the tegminal humeral trunk and more or less of the discoidal field and vicinity of the anal vein of the tegmina. The pale pattern consists of the usual postocular bars outlining the disk of the pronotum, the greater portion of the margins of the lateral lobes, the marginal field of the tegmina, paired lateral bars on the abdomen, transverse edgings on 122 ‘ PROCEEDINGS OF THE ACADEMY OF [Jan., the segments of the same, and an adventitious medio-longitudinal abdominal thread. The tone of this pattern varies from barita yellow (in the recessive extreme where it is very poorly contrasted) to buff yellow, in the generally greenish individuals running to whitish with the transverse edgings of the abdominal segments cendre green. The abdominal infuscation varies in intensity with the general condition of the coloration, in the extreme recessive condition being absent and in the other extreme covering the entire dorsum of the abdomen (except for the medio-longitudinal thread) and the dorsal portion of the lateral faces of the same, between which types are regular graduations in the dorsal width of this shade, the lateral patches decreasing in size toward the average and recessive condition. The tone of this infuscation is always blackish. Eyes varying from terra cotta to vandyke brown; antenne lightly more intense than the general color. Pronotum with the pale borders of the lateral lobes somewhat variable in width. Tegmina with the greater portion of the anal field and much of the discoidal field with the base color varying from wood brown to seal brown, the venation and a large proximal patch on the anal field varying from sulphine yellow to raw sienna, frequently more or less washed with greenish. Abdomen with a dorsal medio-longitudinal bar of from sulphine yellow to antique brown in those specimens approaching the intensive extreme which have the dorsum of the abdomen not solidly infumate, the medio-longitudinal thread of the pale pattern of course dividing this bar; segments with the pale pattern edging narrowing mesad, broadest where they bisect the lateral bars of the same pattern, in the intensive extreme a suggestion of beading of the same margin is due to the breaking of the pigment into regular though subcon- tiguous patches; disto-dorsal abdominal segment varying from deep chrome to sanford’s brown; cerci and subgenital plate ranging from deep chrome to orange rufous. Cephalic and median limbs of the general color, more or less suffused with orange rufous distad on the femora and all of the tibie, the latter in the intensive condition becoming blackish brown distad. - In one of the recessive specimens the cephalic and median femora are dusky olive green distad. Caudal femora of the general color, pattern always distinctly indicated, the distal extremity always with a blackish area occupying about one-seventh the femoral length, the distal half otherwise varying from light orange yellow to mars yellow, passing into the color of the proximal half; caudal tibia more or less deeply and almost wholly suffused with blackish brown, with a brief genicular section of the general color. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 123 Female with the general color varying from sulphine yellow to cosse green, the dorsum of the abdomen in intensive individuals ranging to dilute raw sienna. The pronotal wash is only indicated in the extreme intensive condition, never as extensive as in the male and only solid on the occiput and the cephalic half of the pronotal disk. Pale pattern in the extreme recessive type but faintly indicated on the head and pronotum, in the extreme intensive condition much as in the male, but the transverse edgings to the abdominal segments are narrower and decidedly beaded by blackish intervals. Marginal field of tegmina always solidly whitish. Abdominal infuscation represented only in intensive specimens by blackish areas of variable size placed dorso-laterad at the bases of the segments. Antenne varying from apricot yellow to ochraceous orange. Ovipositor of the general color, strongly olive green distad, edged there with blackish brown, the dorsal margin of proximal half washed with raw sienna. Limbs of general color, in the intensive extreme ap- proaching viridian green proximad on caudal femora, distad on same sulphine yellow, distal infuseation usually present only in the inten- sive individuals; tibize of the general color, distal extremity of the same and tarsi touched with buffy brown. The type is an average male, while the allotype is an extremely intensive female. In the Lyford series we have both extremes in both sexes and every intermediate, so it is evident that the color variation has no geographic significance. Distribution —This large species is known from seven localities in the Rio Grande Plain of Texas (vide Bray), its range extending east to Corpus Christi, south to Lyford, north to Wades and Cotulla, and northwest to Carrizo Springs. Its distribution probably extends south of the Rio Grande into Mexico, but we have no material from that country. Its vertical distribution is limited, extending from or near sea-level at Corpus Christi and Lyford to about seven hundred and fifty feet at Carrizo Springs. Biological Notes—The present species, from data on the specimens, ‘was taken on cotton at San Diego (nymphal individuals) and on prickly pear (Opuntia) at Hebbronville and Alice. At Lyford we found the species fairly common but somewhat local in fields of high weeds, which had a low cover of sand spur (Cenchrus) and grasses. Its stridulation, which was heard at night and with the aid of which specimens were taken, is very faint. *5 One exception, a recessive female from Lyford has them indicated. 124 PROCEEDINGS OF THE ACADEMY OF [Jan.,. The material from San Diego taken April 30 is all mymphal, while at Carrizo Springs nymphs were taken in May and adults only in. June. At Wades and Cotulla adults were taken May 21 and 12, respectively, while at Lyford, August 6-7, but a few nymphs were secured, although the adults were numerous. At Hebbronville the: species was pairing August 29. Morphological Notes.—In the male there is some variation in. the: relative size of the tegmina and some slight differences in the character’ of the margin of the median lobe of the cerci, but the variation in the genitalia of that sex, including the distal margin of the subgenital plate, is extremely slight. The really noteworthy feature in fhe form variation is that of the ovipositor, as it varies some in form as well as dimensions. The even curve of the latter is appreciably flattened mesad in certain individuals. This latter condition, how- ever, is never decided enough to cause the ventral margin to appear wholly or in part straight. The caudal margin of the disk of the pronotum varies in the female from truncate to slightly but dis- tinetly arcuate. Synonymy.—The misidentifications of this species, first as D. emarginata by Rehn in 1901, and second by Hunter, Pratt and Mitchell in 1912 as D. emarginata and in. part as brevihastata, we are able to correct, having the material before us. The first error can be explained by the absence of any material for comparison of the closely allied emarginata, while the same reason was doubtless respon- sible for its determination. as brevihastata, only female individuals having been at hand, aside from an alcoholic male from Hebbronville. Remarks.—There exists a possibility that the acquisition of more material from the western part of the range of the species may make desirable the separation of a western race based on the more elongate ovipositor, but our present representation is too limited to convince us of the desirability of that action. This form is complemental to D. emarginata, living in the main in a different region although in much the same situations. Specimens Examined.—65; 29 males, 29 females, 3 male nymphs, 4 female nymphs. Lyford, Cameron Co,, Texas, August 6-7, 1912, (R. and H.), 24 @, 21 9,3 2 nymphs. Type, allotype and paratypes. Corpus Christi, Nueces Co., Texas, July 29, 1912, (H.), 1 2 nymph; October 20, 1905, (F.C. Pratt), 2 9, (U.S. N. M.]. Wades, Nueces Co., Texas, May 21, (E. A. Schwarz), 1 o, (U.S. N. M.]. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 125 Alice, Nueces Co., Texas, August 28, 1908, (J. D. Mitchell; on Opuntia), 1 2,[U. 8S. N. M.]. + San Diego, Duval Co., Texas, April 30, (E. A. Schwarz; on ‘cotton), 2 co’ nymphs, [U. 8. N. M_]. Hebbronville, Duval Co., Texas, August 29, 1908; on Opuntia, Wit lkse> (WSs Newel: Cotulla, La Salle Co., Texas, May 12, 1906, F. Cy Pratt) i ot, [U. S.N. M_]- Carrizo Springs, Dimmit Co., Texas, May and June, 1885, (A. Wadgymar), 2 o, 3 2, 1 & nymph, [Hebard Collection]. Texas; 1°9), [A. N.S: P:1. Dichopetala emarginata Brunner. 1878. Dfichopetala| emarginata Brunner, Monogr. der Phaneropt., p. (Texas.]} 1897. Dichopetala emarginata Saussure and Pictet, Biol. Cent.-Amer., Orth., I, p.315. [Texas.] : This species constitutes with D. gladiator (vide supra) a section of the genus readily recognized by genital characters of both sexes. These characters are emphasized in the keys, and in the foregoing description of gladiator the differential features of the two forms are also given. Types: o& and @; Texas. [Brunner Collection and Geneva Museum.” ‘ Description of Male (Dallas, Texas; U.S. N. M.).—Size medium; form as usual in the genus. Head with occiput moderately declivent; fastigium moderately compressed, low, subcontiguous with the facial fastigium; eyes moderately prominent, ovate, greatest depth contained one and one-third times in the greatest depth of the infra- ocular portion of the gene; antennz with proximal joints not depressed. Pronotum not sellate, dorsum subdeplanate, on cephalic ‘half slightly rounding laterad; disk with lateral margins slightly indieated by angles caudad, by color for their whole length, decided and regularly narrowed mesad, the median width about one-half that of the caudal margin of the disk; cephalic margin of disk trun- ‘cate, caudal margin of same considerably arcuate, slightly flattened mesad; transverse sulcus bisecting the lateral margins of the disk mesad, on the disk forming an obomegoid figure caudad of the middle; lateral lobes of the pronotum with the greatest depth contained *° From the information given by Saussure and Pictet we learn that this material, or at least the portion of it in the Geneva Museum, was collected by Boll. We have examined fourteen specimens collected by Boll at D: allas, so we consider Dallas material typical. 126 PROCEEDINGS OF THE ACADEMY OF [Jarie about twice in the dorsal length of the same, cephalic margin of lobes straight, ventro-cephalic angles narrowly rotundato-subrectangulate.,. ventral margin very slightly oblique, sinuato-truncate, ventro-caudal angle roundly obtuse-angulate, caudal margin obliquely arcuato- truncate, the dorsal portion of the margin passing into the arcuation of the caudal margin of the disk. Tegmina over three-fourths the length of the disk, the greatest width of the discoidal and anal fields little inferior to the tegminal length; costal margins moderately arcuate, rounding into the slightly oblique subtruncate distal margin, sutural margin subrectangulate, the angle at the extremity of the stridulating vein broadly rounded, the distal section of the same margin slightly sinuate; marginal field narrow, discoidal field regularly expanding distad, anal vein arcuate, anal field with its greatest length subequal to its greatest width, stridulating vein: decidedly arcuate, tympanum proper poorly developed. Disto- Fig. 32.—Dichopetala emarginata Brunner. Lateral view of male topotype. (x 2.) dorsal abdominal segment with the margin truncate distad; supra- anal plate trigonal, acute, the lateral margins of same slightly arcuate ; cerci with the proximal portion moderately robust, slightly flattened dorsad, median lobe placed on dorsal surface, depressed, broad, when viewed from the dorsum the margin of the lobe is broadly rounded on the external and bluntly angulate on the internal side, shaft proper arcuate inwards from proximad of middle, tapering, acute, subaciculate, slightly curved dorsad; subgenital plate with proximal width slightly less than greatest length, moderately produced, distal half with margins converging for the greater portion of their length, then subparallel, distal margin rectangulate emarginate, the lateral angles very faintly blunted, ventral surface weakly tricarinate distad. Cephalic femora somewhat shorter than the combined length of the head, pronotum and tegmina; foramina of cephalic tibie elongate elliptical. Median femora slightly exceeding the combined length of head, pronotum, and tegmina. Caudal femora slightly more than 1914.] NATURAL SCIENCES OF PHILADELPHIA, 127 twice the length of the median femora, considerably robust in the proximal three-fifths. Description of Female (Dallas, Texas; U.S. N. M.).—The following points are those of difference from the male. Size rather large. Eyes slightly more elliptical than in the male; proximal antennal joints slightly deplanate. Pronotum with the lateral margins of disk marked nowhere by angles and only indicated by color, shape of color margins as in male, but median width slightly more than half the caudal width of the disk; lateral lobes with greatest depth contained less than twice in the greatest length of same. Tegmina distinctly laterad, separated by more than their own width, in form semi-ovate, not reaching to caudal margin of the metanotum, humeral trunk indicated. Disto-dorsal abdominal segment subtruncate; supra-anal plate slightly broader than long, rounded subtrigonal. Ovipositor elongate, more than twice the length of the pronotal disk, moderately robust, dorsal margin moderately arcuate, slightly flattened mesad, ventral margin straight except for a short and decided arcuation distad, dorsal margin armed on distal fifth with six spines, ventral margin armed on same section with six to eight spines, the extreme ones of latter series slightly recurved; subgenital plate with chitinous portion divided in two, these present as lateral subpyriform lobes with very acute apices. Cephalic femora slightly and median femora decidedly exceeding the combined length of head, pronotum, and tegmina. Caudal femora slightly less than twice the length of median femora. Measurements (in millimeters). oMon Dallas, Texas. (Described Average spec.; (Scudder of three U.S. N. M.) Coll.) (U.S.N.M.) specimens. Length of body pallor, Boden 13:2 14.2 Length of pronotum ant pO 3.6 4. 3.9 Greatest caudal width of disk of pronotum.... er RO Oe 2.5 ei Pa Length of tegmen . cpr ias HOS TA 3.5 3.4 3.5 Greatest width of discoidal and anal fields of tegmen..... 3.2 2.5 3.1 2.9, Length of cephalic femur... 8.9 er et, Length of medianfemur... 9.9 es 9.4 9.6 Length of caudalfemur.....__ 20 ills. 19.7 18.2 27 This specimen is so badly shrunken that we have not considered the body length. 128 PROCEEDINGS OF THE ACADEMY OF {Jan., Texas. (TYPE meas., Gregory, Beeville, Uvalde, ex Brunner) Tex. Tex. Tex. Mengthon Odiawemare nance alas. 1726 15.7 16.7 Length of pronotum ....... 4.5 4.1 4 3.9 Greatest caudal width of disk of pronotum ae ZDa,tf 220 297 Length of tegmen........ ee 4. 3.6 3.6 Greatest width of discoidal and anal fields of tegmen.. irs oem 3) 3. Length of cephalic femur... 9.5 10. 9.4 10. Length of median femur... 0... 11-2 10. 10.7 Length of caudal femur 20. 23. 21.8 22.5 fe Dallas, Texas. (Described spec., (Seudder — (Scudder U.S. EN. M.) ) Coll.) Coll.) Length of body . 18.6 ifs); 16. Length of pronotum malOe 5. 6. Greatest caudal width of disk of pro- notum Bo 3. 3.4 Apparent length of tegme n 1.4 um iF Width of entire tegmen 2. 2-1 1.6 Length of cephalic femur O53) 4 oe Sinan, ee Length of median femur 10.2 Py Length of caudal femur 20.8 23.8 Length of ovipositor 10.3 inte 12.7 + O59 ey } Dallas, Texas. Texas. Average (TYPE of four meas., ex (U.S.N.M.) specimens. Brunner.) Length of body 17.2 16.1 Gy Length of pronotum aya: 5.6 be Greatest caudal width of disk of prono- tum. 3. on Apparent length of tegme n 1:2 ie Width of entire tegmen Di. 1.9 Length of cephalic femur 8.4 8.8 9. Length of median femur 9. 9.6 Length of caudal femur . al. 21.8 22. Length of ovipositor . 11.3 11.3 its. The Dallas male from the Scudder Collection is the smallest of seven of that sex from the same locality, while the two females from the same series represent the extremes of six females contained 1914.] NATURAL SCIENCES OF PHILADELPHIA. 5 ol29 therein. It is quite evident that considerable individual variation is present in the species, this being pronounced in the Dallas repre- sentation, which is the only series of any size. Too little material is available to consider possible geographic variation in size. Color Notes.—But two of the specimens before us have been stuffed, these (both males) forming the basis of most of the following notes. As usual there is a dorsal color, a lateral color, and paired pale lateral lines in the male, while in the female the dorsal and lateral colors are nearly or quite uniform and the pale lines hardly or but weakly indicated. Lateral and ventral color of male ranging from lumiere green to apple green, in unstuffed specimens ranging to old gold. Dorsal color consisting of an overlying tint running from chestnut through burnt sienna to orange rufous (in unstuffed specimens), covering the lateral portions of occiput, all or nearly all of disk of pronotum and lateral portions of dorsum of abdomen. Pale paired lines in male extending from eye to base of cercus, ranging from pale viridine yellow to creamy white. Head with fastigium and occiput bearing a narrow medio-longitudinal line of the color of the pale lines, finely bordered with lines of the dorsal color; eyes ochraceous tawny to liver brown, crossed obliquely by an irregular slightly darker line. Prono- tum in some specimens with the color dilute caudo-laterad, in tone approaching the lateral color, the pure dorsal color being restricted to a median band; lateral lobes more or less suffused dorso-caudad with the dorsal color; ventral margins edged with the color of the pale lateral lines. Tegmina with the marginal field of the color of the pale lateral lines, humeral trunk chestnut, discoidal field and large part of the anal field blackish brown, the venation varying from buff yellow to neva green, the latter tone only present on the discoidal field, the venation of the proximal half of anal field in all specimens approaching buff yellow. Dorsum of abdomen with the median section very close to the lateral color, the margins of all the segments finely and closely beaded with the dorsal color on an edging of the tone of the pale lateral bars. Limbs with the femora of the lateral color, distad becoming infuscate with ochraceous-orange, the tibize entirely of the latter color; cephalic tibize with the vicinity of the foramina narrowly lined with blackish, a disto-genicular area on the median tibie more or less marked with the same, tarsi of the same limbs more or less clouded with bone brown. Caudal femora with distal eighth more or less solidly blackish brown, pattern on external face more or less distinct in all, medio-longitudinal - in 9 130 PROCEEDINGS OF THE ACADEMY OF ans position; caudal tibie more or less completely infuscate with bone brown. General color of female (unstuffed specimens) olive ochre to honey yellow, finely punctulate with maroon, these punctulations thickest on the dorsum of the abdomen and practically absent from the face, gene, and lateral lobes of the pronotum. Limbs varying from uniform apple green to the general color, more or less suffused with vandyke brown. Pale lateral bars hardly indicated or weak in the female, tegmina almost wholly of the same tone. Ovipositor of the general color, suffused distad with sepia to brownish black. Distribution —The range of this species covers a considerable area of Texas south and east of the Edwards Plateau and plateau plains, being known from four localities, viz., Dallas, Gregory, Beeville, and Uvalde. Dallas constitutes the northern and eastern limit of its known range, Gregory the southern, and Uvalde the western. The vertical range of the species is from practically sea-level at Gregory to eleven hundred feet elevation at Uvalde. Biological Notes—All we know regarding the habits, ete., ean be taken from our own notes, based on the capture of three specimens. At Gregory we obtained the species from a green tangle about a mesquite clump, where D. brevihastata was also secured; at Beeville it occurred in weeds near a tangle of low vine-covered bushes, while at Uvalde it occurred with D. castanea on Acacia berlandieri growing on the low hill slopes. . Morphological Notes——The male cerci seem to be very constant in form, but the distal margin of the male subgenital plate shows considerable variation, in some specimens (Dallas) considerably approximating D. gladiator in this respect, from which species, however, cercal and other characters readily separate them. This variation is due to a certain amount of plasticity in the shape of this margin, which ranges from distinctly rectangulate emarginate (as it is in the majority of specimens) to a type which has the angulation obtuse with the lateral angles much more rounded than in the typical form. In the female the ovipositor shows some variation in the straight- ness of the ventral margin, this being slightly arcuate in three Dallas individuals, but this arcuation is not as decided as in gladiator, the general form and robustness of the ovipositor being different from that found in the latter. In two Dallas females the tegmina are hardly visible beyond the pronotum, but the specimens are unques- tionably adult. The tegmina project slightly caudad of the adjacent 1914.| NATURAL SCIENCES OF PHILADELPHIA. 131 (lateral) margins, but not (or hardly) caudad of the dorsal margin of the pronotum (caudal margin of disk). Remarks.—There can be little doubt of the correctness of associ- ating Brunner’s name with this species, as apparent discrepancies between the description and present material seem to be entirely color differences or else due to a different conception of terms and the relativity of such. The evidence we have of the probable source of the original material and locality of the same assists one considerably in locating the species, as but one form of the genus is, so far as known, found in the Dallas region. Specimens Hxamined.—21; 12 males, 9 females. Dallas, Texas, (Boll), 7 o', 6 2, [Scudder Collection]; 1 @,[U.5S. N. M.]. Dallas, Texas, 1! co, 2 9, (U.S. N. M.]. Texas, (Belfrage), 1 2, [Scudder Collection]. Gregory, San Patricio Co., Texas, July 30, 1912, (H.), 1. Beeville, Bee Co., Texas, July 28, 1912, (H.), 1 o. Uvalde, Uvalde Co., Texas, elev. 1,000-1,100 feet, August 21-22, 1912, (R. and H.), 1 @. Dichopetala orewca* n. sp. Closely related to but a single species—D. catinata (vide infra)— from which it can immediately be separated by the lobe of the male cercus having the margins converging distad, by the ventral margin of the same with a distinctly indicated longitudinal cingulum, the narrower subgenital plate of the male, which has the lateral margins subparallel distad, and by the more ample tegmina in the same sex, while in the female sex the ventral margin of the lateral lobes are straighter in, the present species, the ovipositor is more elongate, with the dorsal margin more regularly arcuate and the distal teeth of same more numerous, and the apices of the subgenital plate spiniform, while in both sexes the eyes are less prominent in orewca. Type: o; Canyon behind Pulliam Bluff, Chisos Mountains, Brewster Co., Texas. Elev. 4,600—5,000 feet. September 7, 1912. (Rehn and Hebard.) [Hebard Collection.] . Description of Type.—Size medium; form moderately elongate. Head with the occiput full, but little declivent to the fastigium and antennal scrobes; fastigium compressed, lamellate, in contact with the fastigium of the face; eyes prominent, ovate in outline, infra- ocular portion of the gene but little longer than the eye; antenne 8 ‘Opévovkoc, Mountain-dwelling. 132 PROCEEDINGS OF THE ACADEMY OF [Jan., about four and a half times as long as the body, proximal joints cylindrical. Pronotum very faintly sellate, dorsal lme weakly as- cending caudad when seen from the side; form of disk as usual in the genus, the median width hardly more than half that of the caudal margin of the same, regularly diverging cephalad and caudad, more sharply so cephalad; lateral margins of disk indicated on metazona by rounded angles, by color alone on prozona; cephalic margin of disk subtruncate, caudal margin of disk almost imperceptibly Fig. 33.—Dichopetala oreeca n. sp. Lateral view of type. (xX 2.) arcuate; transverse sulcus broadly V-shaped mesad on disk, severing the lateral margins of the disk mesad; lateral lobes of the pronotum with the greatest depth contained one and one-half times in the dorsal length of the same, cephalic margin sinuato-truncate, ventro- cephalic angle blunt, nearly rectangulate, ventral margin arcuato- truncate, gently rounding into the oblique arcuato-truncate caudal margin. Tegmina in length nearly equal to that of the combined ios \ ea 2 Figs. 34 and 35.—Outline of cercus of males (types) of Dichopetala orewca (34) and D. catitana (35). ( X 10.) head and pronotum; costal margin gently arcuate, disto®caudal angle quadrantiform, distal margin obliquely passing into the distal portion of the sutural margin, the latter roundly obtuse-angulate at the extremity of the stridulating vein, obliquely arcuato-sinuate distad of the same; marginal field moderately wide, discoidal field regularly widening in the distal two-thirds, anal field with the greatest length little more than the greatest width, stridulating vein gently 1914.] NATURAL SCIENCES OF PHILADELPHIA. 133 arcuate, tympanum unequally trigonal, anal vein moderately arcuate. Abdomen with the disto-dorsal segment having the distal margin proper sinuato-truncate; supra-anal plate broad, rotundato-trigonal, with folded rounded lateral flaps; cerci with proximal portion moder- ately robust, cylindrical, lobe diverging proximad of the middle of the shaft, expanding into a broad convex structure like an inverted spoon, the dorsal margin arcuato-truncate, ventral margin considerably arcuate and VK i with a distinct marginal longitudinal cin- | gulum, apex rather narrowly rounded, whole lobe directed dorso-mesad, re- Figs. 36 and 37—Ventral out- mainder of shaft acute, tapering, acicu- live .3 PERO mares late, triquetrous in section, arcuate, lying —catinata (36) and D. orewca under the lobe and following the same pe C. 5 3) general curve; subgenital plate rather ample, distal half with lateral margins moderately converging, thence straight and subparallel to the tips, distal margin nearly rectangulate emarginate, the lateral angles moderately acute, the ventral surface with a distal medio-longitudinal carina. Cephalic femora about one and two-thirds times as long as the length of the head and pronotum together; cephalic tibie with the foramina elongate elliptical. Median femora subequal to the median pair in length. Caudal femora slightly more than twice the length of the cephalic femora, moderately inflated proximad. Allotype: 2; Moss Well, Chisos Mountains, Brewster Co., Texas. Elev. 4,700 feet. September 5-8, 1912. (Rehn and Heb- ard.) [Hebard Collection.] Description of Allotype-—The following characters are those of difference from the male sex. Form rather robust. Antenn:e about twice the length of the body. Pronotum not at all sellate, the dorsal line nearly straight when seen from the side: median width of the disk proportionately greater than in the male; lateral margins of the disk indicated almost wholly by color; cephalic and caudal margins of the disk as in the male; lateral lobes as in the male except that the ventral margin is subsinuate. Tegmina lateral, very small, rotundate, distal margin slightly flattened, inter- Figs. 38 and 39—Dorsal outline *P¢e between the tegmina subequal to of head, pronotum, and teg- Wwidthofasingle tegmen. Disto-dorsal mina of females (allotypes) of — ahdominal segment with the distal Dichopetala oreeca (38) and 5 5 D. catinata (39). (xX 2.) margin subarcuate, arcuato-emarginate 134 PROCEEDINGS OF THE ACADEMY OF [Jan., laterad; supra-anal plate broadly tongue-shaped; ovipositor about twice the length of. the disk of the pronotum, moderately robust and arcuate, tapering in proximal two-thirds, dorsal margin very considerably and regularly arcuate to the very apex, ventral margin very faintly arcuate except for the distal fourth which forms a quadrant, dorsal margin with seven to eight distal teeth, increasing in length distad, ventral margin with nine to ten distal teeth increasing in length distad and with the apical ones somewhat recurved; sub- oS genital plate almost divided ee in two (chitinous portion completely divided), lateral sections elongate acute, the ee tips aciculate, slightly Seer curved toward the median line of the body. Cephalic Figs. 40 and 41.—Outline | of ovipositor of femora about one and two- oe ENTE orewca (40) and thirds times the length of the head and pronotum to- gether. Median femora slightly longer than the cephalic femora. Caudal femora about two and one-half times the length of the cephalic femora. Paratypic Series —We have selected as paratypic eight males: two from canyon behind Pulliam Bluff (locality of Type) and the re- mainder from Moss Well-(locality of Allotype). Measurements (in millimeters). Canyon behind ao Pulliam Bluff, Average of six Average of four Chisos specimens from specimens from Mts.,Tex. Chisos Mts., Tex. Davis Mts., Tex. (Tyre) (Type and Paratypes) (Paratypes) Length of body 15.8 15.8 (14. -16.9) 15.1 (14.3-16. ) Length of pronotum 3.9 3.7( 3.5- 3.9). 3.9( 3.8 4. ) Greatest caudal width of disk of pronotum 2:5. 2-6)( 2.5-)2-9)' 2: 8'(Ge6=3a) Length of tegmen 5.38 4.8(4.3-'5.38) 47 ( 4:68. Greatest width of discoi- dal and anal fields of tegmen 3.2 ~3:3(-3.2= 5:8), 3.40 3s Length of cephalic femur.10.2. 9.7( 8.9-10.3) 8.8( 8.4- 9.5) Length of median femur.10.1 10.1(9.2-11. ) 9.5 ( 9.2-10.1) Length of caudal femur..21.5 21.1 (19 6-22.5) 20.3 (19.5-21.8) 1914.] NATURAL SCIENCES OF PHILADELPHIA. 135 - Monte- Moss Well, ~ Average of four _ lovez, Chisos Marathon, specimens from + Coahuila. Mts., Tex. Tex. Davis Mts., Tex. {Seudder ~ (Allotype) (Paratype) (Paratypes) Coll.} Length of body (exclu- sive of ovipositor)....16.3 21.5 19.9 (19. —21.3) tee Length of pronotum..... 5. 5: Sie = a3) 4.4 Greatest caudal width of disk of pronotum. 3.1 3.9 Bee omen) 3. Apparent length of ff tegmen..... 1.1 ie 1 (Ge 9= als 5) Greatest width of teg- roaYe) desea 2° 2.2 1.9( 1.9- 2. ) 12 Length of cephalic fe- : ATNUUD Serene 9.7 10.2 9.3( 9. — 9.5) Length ‘of median fe- mur pels iil 3} 10.2 (10. —10.5) 8.2 Length of caudal fe- TAU 2a 24.6 23.5 (23.2-24. ) 20. Length of ovipositor... 10.4 10.9 10.8 (10.5-11.6) 9. From the very small size of the Montelovez female it would seem that at the southern end of its known range the species is quite under the proportions of Texan specimens, although it is best to make such a statement guardedly, as it would appear from the evidence of the Texan material that size variation is, in large part at least, individual. In the Texan series our individual lots are not of sufficient size to be really comparable, although the Marathon female appreciably exceeds individuals of the same sex taken at higher elevations, while in the male sex the measurements so overlap in the Chisos and Davis series that the differences appear to be purely individual. Probably a series from Marathon would show as much size variation as similar representations from other localities. Color Notes.—The intensive and recessive extremes of this species are considerably different, the latter being more decided in the female than in the male. The components of the pattern are the usual dorsal latero-ventral, and pale pattern colors, the first two being wholly or in large part indistinguishable in the recessive females, the pale pattern almost completely lacking in the same and weaker than usual in recessive males. In the more or less intensive males and females the pale pattern is very broad, in fact broader than in any of the other forms of the genus. Dorsal color in recessive males limited to the sides of the dorsum of the head, cephalic two-thirds of the disk of the pronotum, humeral 136 PROCEEPINGS OF THE ACADEMY OF [Jan., trunk, vicinity of the anal vein and part of the anal field of the tegmina, indicated on the abdomen only by a lineation margining the pale lines dorsad; in the intensive males coloring the occiput, the dorsum of the pronotum, greater portion of the tegmina and gener- ally the dorsum of the abdomen except mesad. The tone of the dorsal color varies from claret brown to maroon, in the intensive extreme blackish laterad on the abdomen. Lateral color in the male varying from lettuce green to oil green, in the recessive extreme coloring the middle of the occiput and the greater portion of the dorsum of the abdomen with oil green, in the intensive extreme the latter is repre- sented by a median section of oil yellow, occasional individuals having this mars yellow and antique brown. The pale pattern varies in tone from flat white to pale orange-yellow (on the abdomen only), occasionally tinged with greenish, the pattern coloring the usual areas and in the intensive individuals almost as wide on the pronotum as on the tegmina, the ventral margin of the lateral lobes edged with the same in intensive specimens. Head with a medio- i longitudinal occipito-fastigial thread of the pale color, faintly tinged with the encompassing color; broad vertical imfra-ocular and infra- antennal bars of the pale pattern rather strongly (intensive) or weakly (recessive) contrasted; antenne of the dorsal color (intensive) or orange (recessive); eyes varying from auburn to bay. Tegmina with the ground color of the discoidal field and much of, the anal field blackish brown, the overlying venation and solid paler section of the anal field of the lateral color (recessive) or mars yellow (inten- sive). Abdomen with the dorsal section of the segments more or less decidedly edged distad with the pale pattern; disto-dorsal abdominal segment largely of the dorsal color; cerei varying from greenish proximad and weak mahogany red distad to entirely orange rufous. Limbs of the lateral color, not at all (recessive) or more or less (intensive) washed with burnt sienna on the tibie and the distal extremities of the femora; genicular region of the caudal femora blackish in intensive individuals; pattern of the pagina of the caudal femora restricted, decided in intensive and weak in recessive indi- viduals, ventro-lateral face of caudal femora flat white in intensive specimens; foramina of cephalic tibiz whitish with a seal brown figure. The recessive females are nearly uniform old gold to biscay green, passing into civette green on the limbs, the head pale green yellow with no markings except a faint postocular pale bar and an edging of claret brown dorsad to the same; pronotum more or less parrot 1914.] NATURAL SCIENCES OF PHILADELPHIA. 137 green caudad, the pale lines weak; tegmina claret brown mesad, marginal field of the pale pattern; abdomen with weak narrow paired pale lines, more or less distinctly edged dorsad by a line of claret . brown; ovipositor touched with pinkish proximad, the teeth black tipped. Intensive females with the dorsal color covering most of the occiput (not mesad), the cephalic and at least part of the caudal section of the disk of the pronotum, in tone varying from claret brown to mahogany red. The lateral color varies, in intensive individuals (stuffed specimens) from olive green to ochraceous-tawny, passing into variscite green on the pleura and coxze of ochraceous- tawny specimens, the lateral color covering the dorsum of the abdo- men and limbs as well as the lateral and ventral aspects. Pale pattern in intensive specimens broad, very broad on abdomen. Head in intensive specimens with the vertical bars described in the male rather weakly indicated, otherwise as in that sex. Pronotum with the color of the caudal portion of the disk passing from the dorsal color into that of the lateral regions, the pale bars outlined dorsad more or less distinetly with blackish; lateral lobes occasionally washed with hoary white. Tegmina of intensive females with the base color of the discoidal field blackish. Abdomen with the lateral coloration more or less sprinkled with claret brown stipplings; lateral pale bars more or less washed with flesh pink to rose pink, sharply outlined dorsad on each segment by semi-lunate edgings of black, which form continuous series conforming in arcuation to the form of the abdomen; ovipositor in intensive specimens strongly garnet brown to victoria lake on proximal two-thirds of dorsal margin. All limbs with the genicular region more or less strongly and sharply suffused with claret brown; all tarsi blackish. Caudal limbs with the pattern as in intensive males, in one individual the dorsal section of the proximal half of the femora is largely whitish. Both the type and allotype are intensive individuals. With a single exception, all of the nymphs seen are in or approaching the intensive condition, the exception being about midway between the two extremes. * Distribution —The present species has a very limited range, being found so far as known only at certain elevations in western Texas and at an unlocated point in Coahuila, Mexico. Aside from Mara- thon, Texas, the species is known only in that State from the Davis and Chisos Mountains, the former range beginning about forty miles northwest of Marathon, the latter lying seventy-five miles due south from the same point. At Marathon (where it was very infrequent 138 PROCEEDINGS OF THE ACADEMY OF [Jan., and occurred with D. brevihastata) it was taken between 3,900 and +,160 feet, while in the Davis Mountains it occurred in Lower Limpia ‘Canyon at 4,900 feet, at Maguires Ranch in Upper Limpia Canyon at 5,600 feet and on the slopes of Pine Mountain at 6,500 feet. In the Chisos Mountains it was secured at Moss Well at 4,500-5,300 ~ feet, in the canyon behind Pulliam Bluff at 4,600—5,000 feet and on the slopes of Lost Mine Peak at 6,000 feet. The vertical range of the species is thus seen, at least in Texas, to extend from about 3,900 to 6,500 feet. Biological Notes.—This peculiar species has been found in a number of situations, on bare rock, in grasses and weeds, in a number of species of shrubby plants and in low trees, once in a nogal or walnut tree (Juglans rwpestris). In such places they climb gingerly about, at night giving occasionally a very faint lisping stridulation, of a tinkling, sibilant character, which can be represented by zip-a zip-a Zip-a Zip-ip-ip-ip, the last portion being given infrequently and then very rapidly. This note can scarcely be heard at.a distance greater than six feet. Morphological Notes—In the male the greatest morphological variation appears to be that in the angulation of the distal margin of the subgenital plate, this being more broadly obtuse-angulate in many specimens than in the type, while the bottom of the emargina- tion is nearly rounded in one individual. The male tegmina vary somewhat in bulk, this causing the disto-sutural margin to appear nearly straight in those having the longest tegmina and more or less arcuate (or subangulate) at the apex of the anal vein in those with shorter tegmina. The caudal margin of the disk of the pronotum is truncate in some and feebly emarginate in other specimens, but weakly arcuate (as in the type) in the majority. The female tegmina vary considerably in proportionate size, and the interspace between the same consequently shows an equal amount of variation, ranging from but little over half to that of a whole tegmen width. The oviposi- tor exhibits similar variation in depth to that seen in certain other species of the genus, in the majority of specimens the distal half being subequal in depth and tapering only in the proximai half, although the form of the margins remains practically the same. The spines at the distal extremity of the ovipositor vary in number from seven to nine dorsad and seven to ten ventrad. Remarks.—The peculiar characters of the male of this form imme- diately separate it from all of its congeners except catinata, from which it can readily be distinguished by the lobe of the cereus not being 1914.] NATURAL SCIENCES OF PHILADELPHIA. 139 subtruncate at the apex and having the margins of the same con- verging distad. The female is not so readily separated, but it is hardly likely to be confused with anything but catinata, the charac- ters of difference from which are given in the diagnosis. This is peculiarly a mountain form, the Marathon locality being very close to the foot of mountains in conditions not at all desert-like. The last-mentioned locality was the only place in Texas where it was found associated with another species of the genus (D. brevihastata), which there far outnumbered the present form. The large tegmina of the male are quite characteristic. of orewca, which in the male sex and in the intensively colored female is remarkably pretty. Specimens Examined.—30; 13 males, 7 females, 1 male nymph, 9 female nymphs. Pine Mountain (slopes), Davis Mountains, Jeff Davis Co., Texas, elev. 6,500 feet, August 29, 1912, (R. and H.), 1 &. Maguires Ranch, Upper Limpia Canyon, Davis Mountains, Jeff Davis Co., Texas, elev. 5,600 feet, August 29, 30, 1912, (R. and H.), Spartan Oke Lower Limpia Canyon, Davis Mountains, Jeff Davis Co., Texas, elev. 4,900 feet, August 31, 1912, (R. and H.), 1 Marathon, Brewster Co., Texas, elev. 3,900-4,160 feet, September 12, 13, 1912, (R. and H.),1 9. Moss Well, Chisos Mountains, Brewster Co., Texas, elev. 4,500—- 5,300 feet, September 5-8, 1912, (R. and H.), 6 o, 1 2 (paratypes and allotype), 5 9 nymphs. - Canyon behind Pulliam Bluff, Chisos Mountains, Brewster Co., Texas, elev. 4,600-5,000 feet, September 7, 1912, (R. and H.),3 ¢ (Typ and paratypes), 4 2 nymphs. Lost Mine Peak, Chisos Mountains, Brewster Co., Texas, elev. 6,000 feet, September 6, 1912, (R. and H.), 1 @ nymph. Montelovez, Coahuila, Mexico, September 20, 1 @ , [Seudder Coll.]. Dichopetala catinata n. sp. Closely related to only D. oreeca (vide BER under which the differential diagnostic characters are set forth. Tyre: o'; Brownsville, Cameron Co., Texas. July 31, 1912. (Hebard.) [Hebard Collection.] Description of Type—It seems necessary only to state characters not fully in accord with those of orewca. Size moderate. Eyes very prominent, ovate, their depth contained one and one-third times in .that of the infra-ocular portion of the gens. Pronotum hardly sellate; disk of pronotum with median width very slightly more than 140 PROCEEDINGS OF THE ACADEMY OF [Jant,. half that of the caudal margin of the same; lateral margins of the disk regularly diverging cephalad and caudad; transverse sulcus with. an impressed subobomegoid figure; lateral margins of the disk with a more or less distinct angle everywhere except mesad; cephalic and caudal margins of the disk arcuato-truncate; lateral lobes of the- pronotum with the ventral margin distinctly smuate and the caudal margin less oblique. Tegmina not longer than the pronotal disk, general form as in oreeca, but with the sutural margin rotundato- rectangulate at the apex of the stridulating vein; marginal field narrow, discoidal field regularly expanding for nearly its whole length, anal field with its greatest width about two-thirds of its length, stridulating vein arcuate, slightly bent near the proximal third, tympanum poorly defined, but with the general form much as Fig. 42.—Dichopetala catinata n. sp. Lateral view of type. (X 2.) in oreeca. Disto-dorsal abdominal segment with the distal margin slightly emarginato-truncate, subrectangulate laterad of the same and deeply and sharply arcuato-emarginate at the bases of the cerci; supra-anal plate quadrate with rectangulate angles; cerci in general much as in oreeca, but the lobe is larger, the margins subparallel, and the apex arcuato-truncate, while but a*trace of the ventral cingulum is present; subgenital plate ample, produced, arcuate in transverse section, lateral margins concavely emarginate, distal margin arcuate V-emarginate, lateral angles moderately acute, diverging, ventral surface with a low median carina distad. Cephalic femora about one and two-thirds times as long as the head and pronotum together. Median femora slightly longer than the cephalic femora. Caudal femora about twice the length of the median femora. Allotype: 2; Brownsville, Cameron Co., Texas. August 1, 1912. (Rehn and Hebard.) [{Hebard Collection.] Description of Allotype——The following characters are those of difference from the female of oreaca. Pronotum with the disk broad mesad, at least three-fourths the caudal width of the same; cephalic 1914.] NATURAL SCIENCES OF PHILADELPHIA. 141 and caudal margins of disk subtruncate; lateral lobes of the pronotum with the ventral margin sinuato-emarginate dorsad of the coxe. Tegmina more dorsal than in orewca, semi-ovate, separated by an interval less than half the width of a single tegmen. Disto-dorsal -abdominal segment with the distal margin subtruncate; supra-anal plate transverse, rounded; ovipositor slightly less than twice the length of the pronotal disk, moderately robust, margins as in oreeca, the extremity of the dorsal margin with six, that of the ventral margin with seven spines; subgenital plate with the chitin divided in two, the lateral sections elongate, sublanceolate, the immediate apex blunted. Cephalic femora about twice the length of the disk of the pronotum. Median femora slightly longer than the cephalic femora. Paratypic one an imperfect adult male, the other an immature male, both taken at Brownsville, July 31—August 5. Measurements (in millimeters). Brownsville, Texas. Co ofl 9 (Allo- (TYPE) (Paratype) type) Length of body 17.6 15.4 16.2 Length of pronotum ...... 4. 4. 4.5 ‘Greatest caudal width of disk of pro- notum...... eed. 2.0 3. Length of tegmen oS 4. 3.7 1.5 Greatest width of discoidal and anal fields of tegmen (co) or of entire teg- men (9 ).... 2.8 Dell Dell Length of cephalic femur 9.8 10.5 8.8 Length of median femur . 10.8 iil. 9.5 Length of caudal femur eee 22. Length of ovipositor 8.9 Color Notes.—As but the type and allotype’ of catinata have fully retained their original coloration, the following notes are based en- tirely on them. Male (Typr). General pattern consisting of a dorsal color, a latero-ventral color and a pale pattern, the first covering the occiput, the dorsum of the pronotum, humeral trunk, discoidal and greater portion of anal fields of the tegmina and greater portion of the dorsum of the abdomen. The tone of this color is between burnt sienna and chestnut, that of the latero-ventral color cosse green, while the pale pattern runs from creamy on the head to white on the tegmina and 142 PROCEEDINGS OF THE ACADEMY OF [Jan., pinard yellow on the abdomen, the latter pattern limited to a medio- longitudinal occipital and fastigial thread, and paired bars extending caudad from the eyes along the lateral margins of the disk and margining laterad the dorsal color of the abdomen. Head with the face, mouth-parts,; gene, and much of the postocular region maize yellow; eyes bay; antennz of the dorsal color proximad, passing into antique brown with a few well-spaced moderately broad annuli of seal brown. Pronotum with the pale bars slightly tinged with greenish, in the vicinity of the angle the caudal margin of the disk and of the lobes is blackish. Tegmina with the marginal field wholly of the pale color, the distal portion of the discoidal field with the base color blackish brown and the vein pattern of the general dorsal color, anal field with the vicinity of the proximal two-thirds of the anal vein broadly blackish brown, the proximal portion of the sutural margin edged with same, median section of the anal field washed with barita yellow. Dorsum of the abdomen with the median portion of the segments having the base color weak and each segment with a proximal area of pinard yellow (most decided proximad), laterad the dorsal color is outlined with blackish, this latter oblique and independent on each segment, the contiguous yellowish portion of the pale pattern broken up in consequence. Disto-dorsal abdomi- nal segment of the dorsal color, cerci mars yellow. Cephalic limbs mars yellow, passing into the latero-ventral color proximad, the tarsi clove brown. Median femora largely of the lateral color passing into mars yellow, tarsi clove brown. Caudal femora of the lateral color, passing distad into mars yellow with a decided genicular area of blackish, pattern of pagina pronounced, but not extensive; caudal tibiz seal brown dorsad, honey yellow ventrad, caudal tarsi seal brown. Female (Allotype). Nearly uniform light yellowish olive (probably more vivid in life), passing into light bice green on the limbs, ap- proaching forest green on the medio-longitudinal portion of the caudal femora, the ventral carina of the same whitish. Pale lines feebly indicated on the head and the cephalic portion of the pronotum; caudal margin of the pronotum with blackish as in the male; disk of the pronotum with a faint medio-longitudinal thread of auburn, which is intersected by a black spot at the crossing of the transverse sulcus; eyes auburn; antennz aniline yellow passing into pyrite yellow distad with a few scattered weak annulations. Tegmina weakly suffused with antique brown mesad. Abdomen with the tegmina covering a blackish blotch. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 143: Distribution —The present species is only known from the vicinity of Brownsville in the arid tropical Tamaulipan section of the lower - Rio Grande Valley, Texas. The range of the species unquestionably extends south into Mexico. Biological Notes.—This form was numerous in vine-covered hedges and tangles near old Fort Brown, Brownsville, where they were heard stridulating at numerous points about dusk, but they were _ extremely difficult to secure, owing to their surroundings, as they always sought refuge within the tangled hedges when approached. The stridulation is a faint ts’kh, repeated at intervals of about twice the length of the note. Morphological Notes—From the evidence of the two males, the median width of the disk of the pronotum is seen to vary somewhat, in the paratype this being as much as two-thirds the caudal width of the disk. Specimens Examined.—4; 2°, 1 9,1 o nymph. Brownsville, Cameron Co., Texas, July 31—August 3, 1912, (R. and H.), 2 #, 1 2,1 nymph. Typr, allotype and paratypes. Dichopetala tauriformis n. sp. This is a véry peculiar and distinct species having no close relation- ship to any other form in the genus, in the female sex showing some affinity to falcata and in the male sex approximating pollicifera more nearly than anything else. The peculiar appendage of the supra-anal plate of the male, the anomalous cerci, the strongly depressed median section of which, together with the elongate aciculate tooth which is peculiarly curved, and the unusual structure of the distal section of the shaft, as well as the very decided peculiarities of the subgenital ‘ plate at once distinguish the male sex, while in the female the ovi- positor is proportionately the longest and heaviest in the genus, the subgenital plate with its lateral trigonal lobes also being quite different from that found in falcata. Type: oc’; Mountains twelve leagues east of San Luis Potosi, Mexico. (Palmer.) [Scudder Collection] Description of Type.—Size above the average for the genus; form moderately slender. Head with the occiput rather strongly declivent to the fastigium and antennal scrobes; fastigium low, slightly com- pressed, weakly sulcate dorsad, not touching the frontal fastigium; eyes prominent, elongate ovoid, their length two-thirds that of the infra-ocular portion of the gene. Pronotum weakly sellate, the 144 PROCEEDINGS OF THE ACADEMY OF [Jan., dorsum moderately depianate in transverse section; disk with the lateral margins indicated by a weak angle as well as by color, angle weakest mesad, arcuate inbowed, most approximate at the transverse sulcus which is slightly cephalad of the middle; greatest caudal width of dorsum about two-thirds the greatest length of same; cephalic margin subtruncate, very weakly emarginate mesad, caudal margin very gently arcuato-emarginate; transverse sulcus impressed in an obomegoid figure mesad, faint cephalic and caudal traces of a Vig. 43.—Dichopetala tauriformis n. sp. Lateral outline of type. (xX 2.) longitudinal sulcus; lateral lobes of the pronotum with greatest. depth contained one and two-thirds times in the greatest dorsal length of same, cephalic margin straight, ventro-cephalic angle narrowly rounded rectangulate, ventral margin gently arcuato- emarginate, the greatest depth of the lobes caudal, ventro-caudal angle and caudal margin modefately arcuate except the dorsal por- tion of the latter which is truncate. Tegmina somewhat inferior to pronotum in length; costal margin straight, disto-costal angle well rounded, distal margin slightly oblique, truncato-arcuate, sutural margin rect- angularky produced at the extremity of the stridulating vein, distal portion of the same margin obliquely sinuato- truncate; marginal field rather nar- Figs. 44 and 45.— Dichopetala yow, discoidal field regularly expand- tauriformis n.sp. Dorsalout- . ; 3 4 lineof head, pronotum andteg- ing from the proximal third, anal vein roa of male (type : 44) and straight and not arcuate, anal field sare ni Yeas oi) Pag aS with the greatest length but little more than greatest width, stridulating vein slightly arcuate, tympanum proper poorly defined. Abdomen with lateral margins subparallel, disto-dorsal abdominal segment strongly transverse, the greatest length of the same not more than one-fifth its greatest width, - 1914.] ; NATURAL SCIENCES OF PHILADELPHIA, 145 caudal margin of the segment arcuato-sinuate, slightly produced mesad into a low truncate lobe, which is the hinge of the supra-anal plate, the latter with the length subequal to the greatest proximal “width, lateral margins approximating distad, distal margin broadly V-emarginate, lateral Fig. 46.—Dichopetala angles slightly acute, from the dorsal surface of the ee mis D. Sp. z : 5 orsal outline of supra-anal plate immediately proximad of the apex of abdomen apex is erected a structure like the Greek letter Y, ” is ate (type). but with the cross bar slightly straighter; cerci very complex, having first a semicircular transverse lamellate ridge proximad, distad of which the whole cercus is depressed, obliquely and strongly so toward the internal margin, that which we consider the shaft proper directed meso-caudad, narrowing, strongly depressed, the distal extremity bent inward at a right angle, apex acute, tooth developed from the external margin but little distad of the base, depressed proximad, there lamellate, becoming aciculate distad, curving dorsad and mesad, as long as the shaft; subgenital plate greatly produced, reaching nearly to the tips of the cerci, lateral mar- } gins regularly arcuato-concave, the distal ex- tremity distinctly broader than the median width, distal margin with a decided median Fig. 47.—Dichopetala quadrate emargination, laterad of which the tauriformisn.sp. Ven- . Cries cu : tral outline of sub- Margin is obliquely truncate, angles acute genital plate of male with the immediate angle blunted. Cephalic ype) tS) femora subequal to the length of the head, . pronotum, and tegmina; cephalic tibize with elliptical foramina. Median femora half the length of the caudal femora. Caudal femora longer than the body, moderately inflated, very gradually tapering distad. Allotype: 9°; same data as the type. Description of Allotype—The following points are those of differ- ence from the type. Head with the occiput more roundly declivent. Pronotum with the dorsum broader, the lateral margins of the disk (which are indicated almost wholly by = color) nearly parallel to the transverse CS 8 ange sulcus, thence moderately diverging; Cb ee cephalic margin with the emargination more decided, that of caudal margin less Fig. eae ae pe oe aurea e decided; disk with almost no traces of (<8) pen 10 146 PROCEEDINGS OF THE ACADEMY OF [Jan., pattern of transverse sulcus; lateral lobesasin male. Tegminavery short, broad, sutural margins decidedly overlapping, distal margin somewhat oblique, arcuato-truncate. Supra-anal plate rotundato-trigonal; cerci short, conic, apices slightly Le elongate, acute; ovipositor very robust, elongate, about two-thirds as long as the Fig. 49.—Dichopetala tauriformis n.sp._ Ven- eaydal femora. dorsal mar- tral outline of subgenital plate of female 4 = 7 (allotype). (x 4.) gin considerably arcuate, ; more sharply so distad, ven- tral margin straight except distad where it is decidedly arcuate, dorsal margin with eight to nine teeth on the distal third, ventral margin with seven to nine teeth on the distal fourth, those on the latter faintly recurved distad; subgenital plate with the chitinous portion completely divided, the lateral sections developed as acute trigonal lobes slightly longer than broad. Cephalic femora very slightly longer than the head and pronotum together. Median femora subequal to the length of the head, pronotum, and tegmina to- gether. Caudal femora subequal to the length of the body, mod- erately robust (for the genus). Paratypic Series.—We consider all of the material before us, other ‘than the type and allotype (three males and seven females), para- typie. Measurements (in-milli meters). oe Twelve leagues east of San Luis Sierra de San Potosi, Mex. Miguelito, Mex. Average (Paratypes.) of three (Tyrs.) (Paratype.) ————_-—_—_——__._ paratypes. Length of body 16.5 15o2 16. 5} 16.1 Length of pronotum.. 3.9 3.7 4. 3.7 3.8 Greatest caudal width of disk of pronotum. 3. Bi 2.6 2.5 ot Length of tegmen 3.5 33-5) 3.3 Bee 3.3 Greatest width of dis- coidal and anal fields of tegmen 3.2 3.1 2.8 2.6 2.8 Length of cephalic fe- mur 7.9 7.3 (fe 726 7.3 Length of median fe- mur... 9.9 9.5 8.7 9 9.1 Length of caudal fe- mur 19.7 18.5 itd 19. 18.4 1914] NATURAL SCIENCES OF PHILADELPHIA. 147 co} Twelve leagues east of San Luis Potosi, Mex. (Allotype.) (Paratypes.) Length of bodvy................. x alle 19.2 Die o Length of pronotum ... 4. 4.6 4.7 Greatest caudal width of disk of prono- GUT esse tera). terrae Ae Pera at See ae RO Sri 4. Length of tegmen................ Stina atl 1.5 12 Greatest width of tegmen... Nan ee AeA oi 3.2 Be ' Length of cephalic femur... ee 6. Tall 8. Length of median femur... Pani hee: 8.6 9.5 hen eGhvotacaucdal engin Makaha antl lle 18.2 20.6 eno LHOtmovdpOsltOlere an meets LORS 12 12.8 ee Sierra de San Miguelito, Mex. Average Alvarez, Mex. (Paratypes.) of five (Paratype.) —————— paratypes. Length of body... 20.4 5 }5) 16.8 18.7 Length of pronotum...... 4.5 4.3 4.4 4.5 Greatest caudal width of disk of pronotum..... ee ORO 3.3 3.4 3.6 Length of tegmen. ace o dla sd 1.2 1.4 Greatest width of tegmen... 3c 26 2.9 2.9 Length of cephalic femur Galt oe 7.5 7.4 Length of median femur... 8.2 8.4 8.6 Length of caudal femur.............. 17.2 19.2 18.5 18.7 Length of ovipositor eos LAS, te eS 11.8 Color Notes.—This species has the usual intensive and recessive extremes, the former of which has a dorsal color much darker than the lateral one, in the recessive extreme there being almost no differ- ence in tone between the lateral and dorsal colors. As far as present material goes, the extremes are almost equally marked in the two sexes and the tones are very similar in both. We here give the colors as found in the material, but as none of it has been stuffed there is a strong probability that the greens, at least, have lost much of their intensity. Dorsal color varying from sulphine yellow (extreme in the females alone) to dull maroon (intensive of both sexes), traces of the latter being present in the recessive males,” while this color is solid and pure on the head and disk of the pronotum of intensive individuals of both sexes. On the dorsum of the abdomen * Possibly the recessive condition in the male is wholly due to ance ation, the original dorsal color being ‘left in patches. We, however, do not feel con- vinced that this is the case, as the general tonal correlation of what we consider the recessive male is essentially the same as in the undoubtedly recessive females. ¥- 148 PROCEEDINGS OF THE ACADEMY OF [Jan., this intensive color is only pure laterad, mesad the tone being aniline yellow to sulphine yellow, thickly and closely stippled with maroon. The lateral maroon bordering lines are narrowly present in even the recessive males as well as faintly indicated in the recessive females. Lateral color varying from sulphine yellow to buffy citrine, the males being almost all sulphine yellow, particularly pure in the intensive males. Pale lateral lines very narrow, more or less indicated in all, varying from creamy white to maize yellow, rarely touched with orange pink, extending from the caudal margin of the eye to the base of the cercus. Eyes varying from chamois (recessive male) and buffy citrine (recessive female) to old gold (intensive male) and cinnamon brown (intensive female). Antenne varying much the same as the dorsal color. Tegmina of male largely oil green, the proximal portion of the humeral trunk blackish, large portion of anal field washed with warm sepia, marginal field shell pink; temina of female with discoidal field oil green, anal field similar, occasionally (intensive female [allotype]) washed with maroon, marginal field shell pink to ochre red. Ovipositor varying with the dorsal color. Limbs varying from pois green to grape green, occasionally washed with purplish vinaceous on median and cephalic pair in recessive specimens, of the same greatly suffused, lined and spotted with maroon in intensive individuals. The latter condition is very decided in its extreme, the femora having nearly solid pregenicular patches dorsad, while the distal extremities of the tibize and all of the tarsi are suffused with maroon. The type and allotype are in the extreme intensive con- dition, which is shared or approximated by several other specimens. Distribution.—The present species is known only from three locali- ties in the state of San Luis Potosi in the east-central portion of the Mexican tableland: Sierra de San Miguelito, mountains twelve leagues east of San Luis Potosi city, and mountains at Alvarez. The first-mentioned locality we are unable to locate, so its altitude cannot be given, but it probably has much the same elevation as the other localities, which range between five and six thousand feet. Alvarez is on the upper course of the Rio Verde, a head tributary of the Rio Panuco, east of the city of San Luis Potosi. Morphological Notes.—The tegmina of the male show variation in the angulation of the sutural margin and in the character of the distal margin. The latter is more arcuate in one specimen than in the type and in one paratypic male it is more truncate. The curve of the stridulating vein also varies somewhat. The stalked process on the male subgenital plate in one paratype is similar to that of 1914.] NATURAL SCIENCES OF PHILADELPHIA. 149 the type, while in the others the head of the process is more or less expanded with the distal margin arcuate. The anomalous cerci seem to vary little or not at all, while the subgenital plate varies only in that the quadrate emargination of the distal margin is re- placed by a V-shaped emargination in one paratype. The female shows variation chiefly in the robustness of the ovipositor, although this is not as pronounced as in some other species of the genus. Remarks.—The structure of the apex of the abdomen in the male of this species and the very heavy ovipositor of the female are char- acters which serve to easily distinguish the present peculiar form. There is no approach to the genital structure of the male in any of the other forms of the genus, except that the tooth springs from the external margin of the shaft of the cercus in this and in pollicifera, which similarity is somewhat augmented by the general form of the pronotum and tegmina, but there the analogy ceases, as the details of the abdominal appendages and of the tegmina are quite different. The female sex, however, shows no close affinity to pollicifera, while it does have much in common with falcata, to which the male sex shows no affinity. Specimens Examined.—12;. 4 males, 8 females. Mountains twelve leagues east of San Luis Potosi, Mexico, (Palmer), 2 co, 4 2. Typx, allotype, and paratypes. [Scudder Collection. | Sierra de San Miguelito, state of San Luis Potosi, Mexico, (Palmer), 2%, 3 2. Paratypes. [Scudder Collection.] Mountains at Alvarez, state of San Luis Potosi, Mexico, (Palmer), 192. Paratype. [Scudder Collection.] Dichopetala tridactyla n. sp. This species can be immediately separated in the male sex from all of the species of the genus, except D. caudelli, by the peculiar appen- diculate character of the cercus, while from caudelli it can be separated in the male sex by the shorter tegmina, the very brief distal portion of the anal field of the same, by the sutural margin of the tegmina being strongly produced at the apex of the stridulating vein and by the more elongate median tooth of the cereus. In the female sex tridactyla can be separated from caudelli by the shorter ovipositor and blunter apices to the lobes of the subgenital plate. TypE: co’; Camacho, Zacatecas, Mexico. November, 1877. (Law- rence Bruner.) [Hebard Collection.| 3 Description of Type—Size small. Head with the occiput well * 150 PROCEEDINGS OF THE ACADEMY OF [Jan., rounded, regularly descending to the fastigium and antennal scrobes; fastigium slightly elevated, compressed, linear, rounded at the apex when seen from the side, not touching the frontal fastigium; eyes very prominent, subglobose, depth about one and one-half times that of the infra-ocular portion of the gene; antennze with the proximal joints large, slightly depressed. Pronotum sellate, dorsal length little greater than caudal width of dorsum of same and distinctly less than greatest ventral width of pronotum across lateral lobes; cephalic margin subtruncate, caudal margin very slightly arcuato- emarginate; lateral margins of disk slightly marked caudad by rounded angles, elsewhere by color only, the general form of same considerably narrowed mesad; transverse sulcus severing lateral margins of disk mesad, represented on the disk by a median trans- verse impression placed at the caudal third, but not connected with the sulci severing the lateral margins of the disk; lateral lobes with greatest depth contained about one and one-half times in the dorsal length of the same, cephalic margin of the lobes arcuato-emarginate, Fig. 50.—Dichopetala tridactyla n. sp. Lateral outline of type (male). (X 3.) ventro-cephalic angle narrowly rotundato-rectangulate, ventral margin slightly sinuato-truncate, ventro-caudal angle and caudal margin very broadly arcuate. Tegmina slightly shorter than the pronotum, broad, the width of the discoidal and anal fields subequal to the tegminal length; costal margin slightly arcuate, disto-costal angle rounded, distal margin moderately arcuate, passing into the sutural margin, latter strongly rotundato-rectangularly produced at the apex of the stridulating vein, distal portion of the sutural margin strongly oblique; marginal field rather narrow, discoidal field short, strongly expanding distad, anal field very broad. Abdomen with lateral margins of segments subparallel, proximal segments sub- tectate; disto-dorsal abdominal segment with the distal margin arcuato-truneate, considerably arcuato-emarginate at the dorsal 1914.| NATURAL SCIENCES OF PHILADELPHIA. 151 base of the cerci; cerci trifid, from the dorsal base projects an arcuate subequal blunt digitiform lobe, which in general follows the curve of the shaft of the cercus when seen from the dorsum and when viewed from the lateral aspect is subparallel with the same, not quite Fee teat ~ reaching the apex of the median tooth, Heenan acne and latter diverging proximad of the middle, —_ tegmina of males (types) of é . ae m Dichopetala tridactyla (51) moderately acute, tapering, slightly de- and D.caudelli(52). (X 3.) pressed, diverging moderately disto- dorsad, subequal to half the length of the shaft of the cercus distad of the tooth, shaft very robust proximad of the divergence of the tooth, falciform, strongly depressed, triquetrous, margins sharp, apex acute, tapering for a short distance proximad of apex; subgenital plate large, cymbiform, moder- ately produced, distal margin broadly and rather deeply V-emarginate, lateral angles Figs. 53 and 54. Outline of left cercus HALEN pa CapEn@ of males (types) of Dichopetala tri. ™ora slightly less than half the dactyla (53) and D. caudelli (54). length of the caudal femora; ese) cephale tibie with tympanum elliptical. Median femora subequal to one-half the length of the caudal femora. Caudal femora about one and one-half times the length of the body, moderately inflated proximad. Allotype: 92 ; data the same as the type. Description of Allotype.—Size medium; form robust (for the genus). Head broad, form of _. Seiies = 4 fastio; ere h Figs. 55 and 56.—Ventral outline of occiput and fastigium as in the subgenital plate of males (types) of male, the latter, however, not as Dichopetala tridactyla (55) and D. ‘ judelli (56). 8. compressed; eyes prominent, Ee!) OSE) niore ovate than in the male, depth of eye contained about one and one-half times in that of the imfra-ocular portion of the gene. Pronotum in general form similar to that of the male, but less sellate, non-depressed mesad; caudal margin of disk subtruncate; lateral margins of disk hardly indicated even caudad;_ trans- verse sulcus as in male, the median discal remnant of same less distinct and V-shaped; lateral lobes with greatest depth contained 152 PROCEEDINGS OF THE ACADEMY OF [Jan., nearly one and one-half times in greatest dorsal length of same, margins of lobes as in male, but the cephalic truncate and not emarginate. Tegmina very short, : J) over twice as wide as the apparent Le length, costal margin arcuate with the distal portion more or less Figs. 57 and 58.—Ventral outline truncate; interspace between teg- of subgenital plate of females 3 : (allotypes) of Dichopetala tridac- ™ina slightly more than half the tyla (57) and D. caudelli (58). width of a single tegmen. Abdo- OoRd men somewhat compressed, proximal segments tectate dorsad; supra-anal plate moderately produced, rounded; cerci very short, conic; ovipositor slightly surpassing the length of the median femora, moderately arcuate, more sharply so distad, robust, dorsal mar- gin faintly flattened mesad, SS ee distal third of same margin Se. Ree eee armed with six distinct teeth, increasing in length distad, ventral margin very weakly Nae arcuate except in the distal : Sere pee third where the arcuation is pe ee decided, there armed with six Figs. 59 and 60.—Outline of ovipositor of to seven short slightly re- females (allotypes) of Dichopetala tridac- curved teeth; subgenital plate tyla (59) and D. caudelli (60). (X 4.) divided into two parts, as far as fhe chitinous structure is concerned, these connected mesad for a third of their length by soft integument, the lateral halves of the plate subovoid-trigonal, the apices bluntly angulate. Cephalic femora about one and two-thirds times the length of the disk of the pronotum. Median femora slightly less than half the length of the caudal femora. Caudal femora robust (for the genus), considerably inflated. Paratypic Series —AI\l of the material belonging to this species now before us, in addition to the type and allotype, is considered paratypic—four males, fifteen females. Measurements (in millimeters). Camacho, Mexico. Average of four (TYPE) paratypic Length of body aare LORS 11.3 (11.2-11.7) Length of pronotum 2.8 3. (2.9- 3.1) Greatest dorsal width of pronotum 221 2.5( 2.5- 2.6) 1914.] NATURAL SCIENCES OF PHILADELPHIA. 153 Camacho, Mexico. Average of four (TYPE) paratypic os foilfos Length of tegmen... 2.6 2.6 ( 2.5— 2.6) Greatest width of discoidal and ‘anal fields of tegmen.............. 2.3 2.6( 2.4 2.7) Length of cephalic femur Tes (geile (elect, 8) Length of median femur Shee 8. (7.4 8.4) Length of caudal femur. 9 alia). 1673: (15. 2-17. 1) Camacho, Mexico. Average of four ( aya pe) paratypic Qe Length of body (exclusive of ovipositor) 15.5 15.3 (14.7-17. ) Length of pronotum......0.0...... 4.3 3.8( 3.4 4. ) Greatest dorsal width of pronotum.. peta 71 3.1(.2.9- 3.2) Apparent length of tegmen 9 I G29 1b) Greatest width of tegmen.................. Dyali 1.9( 1.8- 2. ) Length of cephalic femur... area G.(6.5= 7.) Length of median femut..... a HB, TeGiGiel— 8.3) Length of caudal femur ...... , 18.2 17.1 (6.3=18.3) Length of ovipositor. 8.3 8.1( 7.8- 8.8) From these measurements it is evident that the type is distinctly under the average in size, while the allotype is considerably over the average for the female sex, both showing in certain proportions the minimum and maximum proportions, respectively, for their sexes. The selection of the type and allotype was based solely on the con- dition of the specimens, so no size factors were considered in the matter. It is apparent from the above evidence that there is con- . siderable individual variation in size in the species. Color Notes.—As none of the material belonging to this species has been stuffed, we are compelled to take the colors found at their face value, although doubtless some have altered very materially. The pattern of both sexes consists of a more or less uniform dorsal color involving a variable portion of the occiput, dorsum of the pronotum, dorsal portion of the lateral lobes of the pronotum and dorsum of the abdomen, and a pale lateral color which involves the face, genx, ventral portion of the lateral lobes, and lateral aspect of the abdomen, the latter color always (o”) or frequently (@) modi- fied in extent and tint. Male. Dorsal color varying from prout’s brown .to clove brown, most decided near its lateral borders on the abdomen. Lateral color varying from buckthorn brown to dresden brown, the ventral hal 154 PROCEEDINGS OF THE ACADEMY OF [Jan., of the lateral lobes of the abdomen and the marginal field of the tegmina creamy white in intensive specimens; paired lateral lines on the abdomen (in the usual position of pale lines in the genus) contrasted with a wash of the dorsal color on the proximo-ventral portion of the abdomen. Head with narrow postocular lines and a medio-longitudinal thread on the occiput and. fastigium creamy white, in intensive individuals the dorsal color covering much of the postocular portion of the gene; eyes varying from dresden brown to cinnamon brown; antenne ferruginous dorsad, ventral surface mahogany red to chestnut. Pronotum with the caudal section of the disk washed with auburn, continuations of the postocular lines, converging to the transverse sulcus and diverging caudad of the same, subobsolete near the caudal margin, varying from creamy white to buff yellow; dorsal color more or less strongly clouding an obliquely delimited dorsal section of the lateral lobes, ventral section of same creamy white. Tegmina with the discoidal and anal fields with a blackish-brown base, over which the veins are outlined in ochraceous orange, the greater portion of the.sutural margin washed with hay’s russet. Distal half of the appendiculate lobe of the cerci infuscate with the dorsal color in intensive individuals. Limbs varying from old gold to olive lake, more or less generally infuscate with chestnut brown, most decided ventrad and there linear in pattern; caudal femora with a pair of fine blackish lines on the ventral portion of the lateral face; tibize more or less lime green. Female. Dorsal color ranging from argus brown to dark mummy brown; lateral color ranging from creamy white to dresden brown, the latter in recessive individuals and there very poorly separated from the dorsal shade. Head with pale lines less distinctly indicated _ than in the male sex, the extreme intensive individual having the greater portion of the head opaline green. Inthe average individuals the abdomen has no decidedly indicated lateral bars dorsad margining the dorsal color, but in the intensive specimens these bars are decided creamy white, of variable width and the lateral base of the abdomen is contrastingly washed with the dorsal color. Pronotum as in the male, but nearly uniform in recessive individuals. Tegmina varying from nearly uniform with the lateral color to blackish brown, with the venation of the lateral color, in the intensive extreme having the costal portion of the latter color. Ovipositor varying from citrine to orange-citrine, distal portion infuscate in some specimens. Limbs varying from viridine green (in this the femora pale bluish white proximad) to cosse green, marked much as in the male, but with the 1914.] NATURAL SCIENCES OF PHILADELPHIA. 155 infuscation more or less tessellate or marmorate, linear and punctate in character. Distribution.—The present species is only known from two localities in the central portion of the Mexican tableland—Camacho, Zacatecas, and Jimuleo, Coahuila. Its vertical distribution is apparently from somewhat below five thousand to about six thousand feet. Morphological Notes——In the male sex the only morphological variation worth noting is that of the degree of arcuation of the margins of the distal excision of the subgenital plate. In the type these margins are straighter than in the others of the sex, but in all they are more or less arcuate toward the angles. In the female sex the ovipositor varies appreciably in robustness without correlation with the general size. . Remarks.—The present species and D. caudelli constitute a very distimet section of the genus, having no close relationship to any of the other forms. Specimens Hxamined.—21; 5 males, 16 females. Camacho, Zacatecas, Mexico, November, 1887, (Lawrence Bruner), 5 o&@, 14 @. Typk, allotype, and paratypes. [Hebard Collection. | Jimuico, Coahuila, Mexico, November, (Lawrence Bruner), 2 @. Paratypes. [Hebard Collection.] Dichopetala caudelli n. sp. This species is close to D. tridactyla, but can be readily separated in the male sex by the distinctly longer tegmina, the more normal distal portion of the anal field of the same, the sutural margin of which is but little produced at the apex of the stridulating vein and by the shorter median tooth of the cercus, while in the female the slightly longer ovipositor and acute apices to the lobes of the sub- genital plate enable one to distinguish the present form. The species is similar to ¢ridactyla in the majority of the characters, so we have made our description in large part comparative® When not mentioned specifically, the structure is understood to be similar to that in tridactyla. Type: o'; San Luis Potosi, state of San Luis Potosi, Mexico. (Palmer.) [Seudder Collection.] Description of Type.—Size and form as in tridactyla. Head as in tridactyla. Pronotum with the caudal margin of the disk very slightly arcuate, disk itself (indicated by color) broader mesad than in tridac- tyla, the lateral borders of the same less sharply diverging cephalad and caudad; transverse sulcus severing the lateral borders of the 156 PROCEEDINGS OF THE ACADEMY OF {Jan.,. disk mesad, forming a V-shaped figure near the caudal third of the disk; caudal margin of the lateral lobes of the pronotum obliquely subtruncate, ventro-caudal angles of lobes rounded. Tegmina appreciably longer than the dorsum of the pronotum, width of dis- coidal and anal fields slightly less than the length of same; costal margin considerably arcuate, distal margin obliquely arcuato-truncate, sutural margin obtuse-angulate at the extremity of the stridulating vein, appreciably sinuate distad of the same; discoidal field less sharply expanded than in tridactyla. Cerci with the appendicular lobe straighter than in ¢tridactyla, slightly expanded distad, median tooth short, depressed, when seen from the dorsum with the margin rounded and not acute, acute distal extremity of the shaft of the cercus slightly shorter and more regularly tapering than in tridactyla ; Fig. 61.—Dichopetala caudelli n. sp. Lateral outline of type (male). (xX 3.) subgenital plate with distal margin more deeply V-emarginate than in tridactyla, the margins of the excision slightly arcuate, the lateral angles quite acute. Limbs as in tridactyla, but cephalic and median femora very slightly slenderer. Allotype: 2°; Mountains twelve leagues east of San Luis Potosi, state of San Luis Potosi, Mexico. (Palmer.) [Seudder Collection.| Description of Allotype.—Differing from the allotype of tridactyla in the characters here described. Form and size as in ¢tridactyla. Head as in tridactyla. Pronotum as in tridactyla, but with the disk broader mesad, the margins (indicated almost wholly by color) very slightly narrowing to the transverse sulcus,. thence moderately diverg- ing caudad; caudal margin of disk gently arcuate; transverse suleus more continuous than in tridactyla, but weak mesad; lateral lobes of the pronotum shallower than in tridactyla, the greatest depth contained nearly twice in the greatest dorsal length of the same, margins similar. Tegmina similar to those of tridactyla. Abdomen very similar to that of tridactyla; ovipositor slightly more robust 1914.] NATURAL SCIENCES OF PHILADELPHIA. 157 and slightly more elongate, teeth of distal portion slightly longer; subgenital plate completely divided to the base, lateral halves nar- rower than in tridactyla and more acute distad. Cephalic femora almost twice the length of the disk of the pronotum. Median femora slightly more than twice the length of the pronotal disk. Caudal femora very slightly more than twice the length of the median femora. Paratypic Series—We have designated as paratypes two males and two females from the type locality, and one male and four female from the mountains at Alvarez, San Luis Potosi, Mexico. Measurements (in millimeters). San Luis Potosi, Mex. Average of three (Paratypes.) Alvarez, Mex. para- (TYPE.) ——— (Paratype.) types. Length of body if 11.6 12.3 14. 12.6 Length of pronotum 2.9 3. 2.9 3. 3. Greatest caudal width of disk of pronotum...... 2.2 Pet) 2.2 2.) 2.4 Length of tegmen 3.2 3.4 oS 3.3 3.3 Greatest width of discoidal and anal fields of tegmen 2.7 3 200 oe 2.8 Length of cephalic femur 8. . Length of median femur. 8.9 : 8.9 oF 8.9 Length of caudal femur 16.2 16.3 18.3 18.4 17.6 Twelve mas San Luis eastofSan Luis Potosi, Potosi, Mex. Mex. (Allotype.) (Paratype.) Length of body (exclusive of ovipositor) 000... 15. 16.5 Length of pronotum............... ia eee ee eee 4.1 ‘Greatest,caudal width of pronotum. Bee on Length of tegmen .8 122 Greatest width of tegmen 125 De Length of cephalic femur 7.7 Length of median femur... 9. 8.4 Length of caudal femur... . 19. 19 Length of ovipositor 8.5 8.5 ete, 2 Alvarez, Mex. Average of (Paratypes.) four para- types. Length of body (exclusive of ovipositor)...0........ resco lays 16. Lome 15.6 Length of pronotum............... 4.1 4.2 4.3 4.2 158 PROCEEDINGS OF THE ACADEMY OF [Jan:; oe Alvarez, Mex. Average of (Paratypes.) four para- — types. Greatest caudal width of pro- notum™= Ache ee St A 3.4 3.2 3.2 Length of tegmen......... ile ile eal 1. Greatest width of tegmen 1.8 1.8 2. 1.9 Length of cephalic femur 7.4 7.6 7.4 7.4 Length of median femur 8.5 9. 8.9 8.7 Length of caudal femur 18.5 18.8 18.1 18.6. Length of ovipositor 9. 9.2 9. $.9 Color Notes—In recessive individuals this species is unicolorous, while in intensive individuals the dorsal color is distinctly darker than the lateral, and pale paired bars are more or less decidedly indi- cated in all but the extreme recessive specimens. The extreme recessive condition is represented by one female, closely approached by another of the same sex and less closely by a male. The extreme intensive condition is represented by one male and one female and approached by three other females. The type is nearly intermediate between the extremes, but slightly nearer the intensive condition; the allotype approaches the intensive condition, but is not typical of it. - The extreme recessive condition ( @ ) is uniform light brownish olive except for the tegmina. The extreme intensive condition in the male has the dorsal color maroon along the lateral margins of the area on the head, disk.of pronotum, and abdomen, paling to cedar green and weak buff yellow (on abdomen only) mesad, while in the female the general tone of the dorsum is tawny, washed with claret brown caudad on each abdominal segment and margined laterad by the same. The suffusing color of the dorsum is largely produced by stippling and the extreme margins of the pronotum and abdominal segments are of the color of the pale lateral lines beaded with the suffusing tone. The dorsal color of the male ranges through old gold with weakly indicated blackish lateral margins, of the female through kildare green finely sprinkled with maroon. Pale lateral lines varying from chalk white to weak orange pink (in part only and in the intensive male), narrow in the female and broad in the male, cover the entire marginal field of the tegmina in both sexes. In the male these lines are crenulate on the abdomen, while in the female they are more or less obliquely offset on each segment. Lateral color of male varying from yellowish olive green (intensive extreme) to chamois (recessive extreme), in the female from the recessive extreme with it uniform with the dorsum to the intensive extreme 1914,] NATURAL SCIENCES OF PHILADELPHIA. 159 which has it sayal brown, the segments of the abdomen stippled and beaded as on the dorsum, the lateral lobes of the pronotum touched with mignonette green in this type. Head with the eyes varying from cinnamon buff to snuff brown. Pronotum with the lateral lobes edged with the color of the pale bars in all but the recessive females, in one of the intensive females there being a considerable area of chalky white on the ventral portion of the lobes. Tegmina of male with the humeral trunk claret brown to bay, discoidal field varying from yellowish olive green to cosse green, anal field sharing the same tone. but more or less oil green mesad with the proximal portion more or less mahogany red. Tegmina of the female divided between the dorsal and lateral color with the region of the humeral trunk claret brown to bay. Limbs almost wholly of the lateral color, occasionally more greenish in tone, in intensive individuals more or less washed, lined and stippled (particularly on cephalic and median pair) with claret brown. Dorsal aspect of the cerci of male washed with claret brown. Ovipositor with teeth blackish. Distribution —This species, as far as known, has a limited range in the east-central portion of the Mexican tableland, occurring at three localities in the state of San Luis Potosi: San Luis Potosi City and hills near the same, mountains twelve leagues east of San Luis Potosi and mountains at Alvarez. The latter locality is south of the upper -course of the Rio Verde, a tributary of the Rio Panuco which empties into the Gulf of Mexico near Tampico. As far as we are able to determine from several topographic maps, the localities are situated between five thousand and six thousand two hundred feet elevation. ‘ Morphological Notes.—In the female sex there is some little varia- tion in the shape of the caudal margin of the disk of the pronotum, this ranging from gently arcuate to sinuato-truncate. The inter- space between the tegmina also varies considerably in width in the same sex, in the greatest extreme being subequal in width to a single tegmen. The ovipositor varies slightly but appreciably in depth and in the number of teeth on the dorsal margin (6 to 8). Remarks.—We take pleasure in dedicating this species to Mr. A. N. Caudell, of the United States National Museum, who called our attention to the peculiar cerci of the male sex. Specimens Examined—12; 4 males, 8 females. ~ San Luis Potosi, Mexico, (Palmer), 3 o™, Type and paratypes, 1 9. [Scudder Collection and U. 8. N. M.] Hills near San Luis Potosi, Mexico, October 15, (Palmer), 2 9° . Paratypes. [Scudder Collection.] 1600 PROCEEDINGS OF THE ACADEMY OF [Jan., Mountains twelve leagues east of San Luis Potosi, Mexico, (Palmer), 1 9. Allotype. [Scudder Collection.] Mountains at Alvarez, San Luis Potosi, Mexico, (Palmer), 1 o%, 4 2. Paratypes. [Scudder Collection and U. 8. N. M.] 1914.] NATURAL SCIENCES OF PHILADELPHIA. _ 161 MIMICRY IN NORTH AMERICAN BUTTERFLIES: A REPLY. BY EDWARD B. POULTON, D.SC., M.A. OXON. CONTENTS. PAGE partrod ue toners cee ree nea seen Sco» lltont 1. The Attacks of Birds on Butterflies and the Theories of Mimicry. Sela 2. Haase’s Name “Pharmacophagus” and his Hypothesis that Mimicked Butterflies (Models) derive Nauseous Qualities from the Larval Food-plants .. 162 3. Indirect Evidence that Phar macophagus philenor i is a Model possessing Distasteful Qualities : 165 4. The attempt to explain Mimetic Resemblance as due to Affinity between Model and Mimice.. 167 5. Sexual Dimorphism (Antigeny of Scudder) and Mimiery 168 6. The Female of Neophasia terlooti, another North American Mimic of Danaida plexippus 172 7. The Colored Pigments of the Pierine as illustrated by Neophasia 176 8. The Restriction of Sex-limited Inheritance to the Mimetic Pattern of Neophasia terlooti.. 177 9. The Evolution of Limenitis (Basilarchia) archippus from an Ancestor with a Pattern like that of L. (B.) arthemis....000.0....0cccccccceees 178 10. The Relation of the Pattern of Limenitis obsoleta (hulsti) to that of archippus, arthemis and weidemeyeri? .... 180 11. The Male Genital Armature of the North American forms of Limenitis 190 12. Similar Environmental Conditions versus Mimicry as an Interpreta- tion of Color Resemblances....................... ; 192 Certain criticisms of the theories of mimicry and warning colors have recently appeared in the publications of The Academy of Natural Sciences of Philadelphia, and it is, I think, a convenience that the controversy should be continued in the same channel. The oceasion also enables me to contribute in, I hope, an appropriate way to the publications of the great and learned society with which I have the honor and pleasure of being specially associated. In the present paper I propose to deal with the friendly criticisms contained in Dr. Henry Skinner’s paper (32). It will be most convenient, I think, to consider the author’s arguments under separate heads, which I have arranged as far as possible in the same order as that adopted in his memoir. 1. THe Arracks or Birps ON BUTTERFLIES AND THE THEORIES OF Mrnicry. The believers in these theories, both Batesian and Miillerian, will cordially agree with Dr. Skinner as to the paramount importance 11 162 _ PROCEEDINGS OF THE ACADEMY OF [Jan., of showing ‘“‘that birds are in the habit of eating butterflies and that some butterflies are poisonous or nauseous to them and others not.”’ (32, p. 121.) It must be admitted also that we require vastly more evidence than we at present possess. But evidence is accumulating steadily, and some of the best has been forthcoming in recent years. I may refer especially to Mr. 8S. A. Neave’s observation (36) on January 12, 1912, of a Wagtail devourmg Lycenid and Pierine butterflies, but rejecting an Acraea, in the bed of a forest stream near Entebbe, Uganda. - Dr. Skinner, in a more recent paper (34, p. 25) refers to the fact that the Biological Survey of the United States examined fifty thousand bird stomachs and only found butterflies in five of them. Mr. C. F. M. Swynnerton has quite lately thrown much light on this method of investigation (33). He is convinced, as the result of recent work at Chirinda, Gazaland, southeast Rhodesia, ‘that conclusions based on stomach-examination are likely to be fallacious, unless that examination has been so thorough and minute that even such small objects as the scales of Lepidoptera must have been detected if present, even in small numbers, in either stomach or intestines, unless a very large series has been so examined for each species, and unless, finally, a note had been made at the time of the shooting of each specimen as to the probable proportions in which insects of various kinds were present at the moment.’’ Mr. Swyn- nerton’s paper was especially intended as a reply to Mr. G. L. Bates (25), whose statements are quoted by Dr. Skinner (32, p. 122). I have treated this subject very briefly and inadequately because I hope to return to it in a later paper dealing with the attacks made by Mr. W. L. McAtee in a memoir (28) written in a very different spirit from that of Dr. Skinner. 2. Haase’s NAME “‘PHARMACOPHAGUS”’ AND HIS HyPpoTHESIS THAT Mimickep Burrerrires (MopgELs) DERIVE NAUSEOUS QUALITIES FROM THE LARVAL Foop-PLANtTs. Dr. Skinner, influenced by my use of Haase’s term ‘‘ Pharmaco- phagus,” is apparently under tle impression that I am a convinced follower of his hypothesis. This is by no means the case. In a review (14) of Haase’s work (13) I expressed the opinion, to which I still adhere, that the hypothesis is probably true—although as yet quite unproved—for some distasteful species, but that it is certainly not true of others. Rothschild and Jordan (20, 433-4), following Horsfield (1) and Haase (9), have shown that the Papilionine are 1914.] NATURAL SCIENCES OF PHILADELPHIA. 163 divisible into three well-marked sections differing in larval, pupal and imaginal characters. They give descriptive titles to each of the sections, but do not suggest names which can be used at any rate provisionally as genera. In the meantime, it is highly inconvenient to include in the genus Papilio the species of all three groups. For this reason, and for this reason alone, I provisionally adopted Haase’s Pharmacophagus for the “Aristolochia swallowtails,”” his Papilio for the “Fluted swallowtails,’’ and his Cosmodesmus for the “Kite swallowtails.”’ I am quite ready to abandon any or all of these when it is proved that the three groups may be referred to by other names with a prior claim, and, under any circumstances, Haase’s terms cannot, with their present meaning, permanently stand for genera, because, as I learn from Dr. Jordan, each of the three sections is a much larger group which must itself be split up into genera. Furthermore, I do not, as Dr. Skinner states on p. 124, accept any conclusions or use any argument based on the meaning of the word when I pro- visionally employ “ Pharmacophagus” as the name of a genus of the Papilios, and I do not think that any words of mine can be quoted which will bear out Dr. Skinner’s interpretation. Inasmuch as Haase’s hypothesis occupies so large a place in Dr. Skinner’s memoir, I venture to offer a few remarks upon the idea itself as well as upon some of the author’s criticisms. The great majority of the pigments possessed by plant-eating insects are built up in the laboratory of the living organism, in spite of the fact that the larval food is rich in chlorophyll. Nevertheless, this color exists ready-made, and certain insects have been specially adapted to avail themselves of it and thus to gain certain pigments. I proved this many years ago by spectroscopic examination as well as in other ways (3, 4), including experiments in which larvze were fed upon parts of leaves devoid of chlorophyll (10)—experiments recently repeated with confirmatory results upon different species by Prof. W. Garstang (24). I think it probable that nauseous or poisonous substances, when they exist in a plant or in a group of allied plants, may be employed by certain species which are restricted to it or them; but as yet the proof is wanting. Among the most probable instances, and those which should first be tested by chemical means, are the Danaine feeding on the Asclepiads and the “ Pharmacophagus”’ swallowtails feeding on Aristolochia and its allies. I may here remark that Dr. Skinner is mistaken in supposing that Haase in his hypothesis drew any distinction between the Danaine and the 164 PROCEEDINGS OF THE ACADEMY OF [Jan., section of the Papilios to which he gave the name Pharmacophagus. He maintained that both of them, and the Ithomiine and Acrewine as well, derived their distasteful qualities direct from the larval food- plants. I refer to the following passage in which Dr. Skinner is speaking of Danaida plexippus (32, p. 126): “The protective idea in this case is the same as in the so-called pharmacophagus butterfly, the imago of plexippus which is said to be repugnant to birds but the repugance is not based on the idea of the butterfly feeding on a poisonous plant (Asclepias) in the larval stage.” As regards the specially protected and much-mimicked group of the Acreine, the recent hitherto unpublished researches of my friend Mr. W. A. Lamborn upon their larvee in the Lagos district strongly suggest that the butterflies do not derive the nauseous qualities, which they undoubtedly possess, in the manner assumed by Haase; for the food-plants belong to varied groups. In a letter written July 16, 1913, and received as I am preparing this paper, Mr. Lamborn states: “By far the most common Acrea here [the neighborhood of Ibadan, 8. Nigeria, W. Africa] is lerpsichore. Its larve abound, and seem, like so many other distasteful caterpillars, to have a wide range of food-plants.”’ The facts brought forward by Dr. Skinner do not appear to me to affeet the probability of Haase’s hypothesis. It is well known that insects feeding on a great variety of plants commonly include among these species with pdisonous qualities. Haase’s hypothesis only refers to certain insects confined to poisonous or acrid food-plants. I say “certain” insects, for the power of utilizing the poisonous quality, if it exist at all, is undoubtedly a special adaptation by no means necessarily present in any larva which feeds on the plant possessing the quality. The other fact alluded to by the author, that the acrid principle may be present in very small quantity, is, I think, equally devoid of bearing on the hypothesis. If the adapta- tion exist at all, we should expect small quantities to be stored up and concentrated. The percentage of lime in a leaf is very small, yet the larva of Clisiocampa neustria reserves enough to render its cocoon opaque with minute crystals of the carbonate in the form of aragonite (5) and Eriogaster lanestris enough to make its eggshell-like cocoon out of the oxalate (8). Haase’s hypothesis cannot be proved or disproved by discussion. It is the work of the chemist that is needed. The most appropriate field in the world for this work is North America with its hundreds 1914.] NATURAL SCIENCES OF PHILADELPHIA. 165 of skilled cheraists and its well-equipped laboratories, and with two abundant species—Danaida (Anosia) plexippus feeding on Asclepiads and Pharmacophagus (Papilio) philenor feeding on Aristolochias— by which to test the validity of Haase’s hypothesis. 3. InprREcT EvIDENCE THAT PHARMACOPHAGUS PHILENOR IS A MODEL POSSESSING DISTASTEFUL QUALITIES. I have myself only seen this insect alive on one or two occasions, and have certainly never had the opportunity of observing it in relation to its natural enemies, nor do I know of any such observa- tions. Scudder states (6, 1248-9) that the larve are gregarious when young and semigregarious in later life, that the perfect insect is very tenacious of life, and he quotes Edwards for the observation that it has a strong and disagreeable scent. These qualities, espe- cially the two latter, are generally characteristic of distasteful species; but Skinner states (32, p. 124) that later specially directed observa- tions have failed to confirm Edwards. Skinner also records (p. 125) the fact that the larve are attacked by parasites, but this is commonly true of Danaine, Acreine, and other distasteful much-mimicked groups. Haase is, so far as I am aware, the only writer on the subject who has supposed that the immunity of models is complete, and probably in all cases protection from insectivorous vertebrates is to a large extent balanced by exceptional exposure to the attacks of parasites and certain other insect enemies, such as Asilid flies and Hemiptera (19). I quite agree with Skinner (p. 125) that the principal attacks are made during the earlier stages of an insect’s life—and think of the elaborate protective adaptations which are common in these stages— yet I do not doubt that the imago is subject to severe persecution from enemies of many kinds. Furthermore, it must be remembered that each imago, the heir of all the other stages, and especially each female, is of far greater value to the species than a single pupa or larva and often hundreds of times as valuable as an ovum. Although I must admit that there is no direct evidence to prove that P. philenor is nauseous to birds (p. 123), I believe that much might be learned if American naturalists would offer large numbers of this swallowtail to many species of insectivorous birds in confine- ment, offering at the same time other butterflies with a procryptic under-strface, such as Vanessa milberti or species of Grapta (Hugonia). The North American Danaine models might be tested at the same time. Although the records of field observations are greatly to be 166 PROCEEDINGS OF THE ACADEMY OF [Jan., preferred to this method of experiment, yet in the absence of such observation much may be learned by comparing the behavior of the same individual bird with different species of insects. The indirect evidence that Ph. philenor acts as a model and pos- sesses the qualities of a model seems to me extremely strong. On this hypothesis many facts receive their interpretation; without it they are unexplained and meaningless. Philenor is one of the “Aristolochia swallowtails,’ a section which is abundantly repre- sented in tropical America and in the Oriental Region, but, with the exception of Ph. antenor in Madagascar, absent from the Ethiopian Region. The mimicry we observe in North America is not only repeated in both Regions where these swallowtails are abundant, but repeated in a more convincing manner, because the patterns are often far more elaborate, and because an ‘‘ Aristolochia swallow- tail’ may break up into numerous geographical races with distinctly different patterns which are mimicked in each locality by correspond- ing races of the “Fluted swallowtails”’ and, in the Neotropical Region, of the “Kite swallowtails.”” A good example is the Oriental Ph. aristolochie with its subspecies mimicked by the females of Pap. polyles. Furthermore, there is in this case experimental evidence that aristolochie is distasteful, and its slower, more flaunting flight has often been remarked upon. In the Oriental Region species of: Pharmacophagus are also sometimes mimicked by day-flying moths, and, in the Neotropical Region, not only by these, but by ‘ Kite swallowtails” (Cosmodesmus) and Pierines. Throughout the whole range, as in North America, the mimicking ‘Fluted swallowtails”’ are as a rule females, while on the other hand the ‘‘ Kite swallowtails”’ are mimetic in both sexes (23). Just as the other much-mimicked groups—the Danaine, Ithomiine, Heliconine, and Acreine—are themselves specially subject to mimicry—the genera or sections of the same subfamily superficially resembling each other and also resembling those of the other subfamilies—so is it in both respects with the South American “Aristolochia swallowtails.’’ In every way these butterflies behave like the great distasteful groups supply- ing the best known models for mimicry. If we had no experimental or other evidence that the Danainw are unpalatable, the indirect evidence is strong enough to warrant at any rate a provisional accept- ance of the hypothesis that they possess some peculiar means of defence which renders them specially advantageous as models. For wherever they are indigenous in the Old World they are mimicked by butterflies of other groups, and even in North America, where 1914.] NATURAL SCIENCES OF PHILADELPHIA. 167 there are only three forms, each one of them is mimicked. It is not as if the models for mimicry were distributed indiscriminately among the butterflies. They are furnished by a few genera here and there among the Nymphaline, Pierine, etc., but the vast majority of them are concentrated in the four Sab iemilicn mentioned above and in the “Aristolochia swallowtails.”’ Until these remarkable and very numerous facts are explained by some other hypothesis or until something stronger than negative evidence is forthcoming, we are justified in accepting the hypothesis of advantageous resem- blanee to a specially defended model. I should be the last to rest content with indirect evidence, however strong, and for many years I have urged naturalists, and especially those in the tropics, té make observations and to undertake experiments. As a result of much work, a considerable body of direct evidence, which cannot be ignored by any fair-minded opponent, has been steadily accumulating, especially from Africa; but I freely admit that more is greatly needed, and I shall continue to urge my friends to seek for it. 4. THe ATTEMPT TO EXPLAIN Mimetic RESEMBLANCE AS DUE TO AFFINITY BETWEEN MopeEt ANp Mimic. Dr. Skinner appears to adopt the above interpretation of the SRE between the Papilios and Pharmacophagus when he says “ The three species, glaucus, asterius, and troilus, do bear a ee to philenor but this happens in any aggregation of species in a genus.’ (32, p. 125.) This interpretation does not bear inspection. In the first place, the butterflies do not in any real sense belong to the same genus, and it is for this very reason that I have provisionally adopted Haase’s Pharmacophagus for philenor. In the second place, the three mimetic species are placed by Rothschild and Jordan in three different groups of the section “Papilio” (‘Fluted swallowtails”’). In the third place, it is clear that the true affinity is shown by the non- mimetic patterns rather than by the mimetic ones—by the upper surface of the male asterius and by the males and glaucus females of glaucus. Darwin suggested that mimicry began “long ago between forms not widely dissimilar in color,” and Scudder adopts the same hypothe- sis in the following passage: “The process has been a long one, so that... . , we may readily presume far less difference between mocker aod mocked when the “mimicry between them first began, than now exists between the mocked and the normal relatives of the mocker.” (6, Deel) Los 168 PROCEEDINGS OF THE ACADEMY OF [Jan., obvious that this interpretation of the resemblances borne by other insects to the stinging Hymenoptera cannot be thus explained, and, within the Lepidoptera themselves, the study of detail has often furnished a refutation. Thus Prof. Gowland Hopkins (12, p. 680) writes: “The mimicking Pierid retains the characteristic pigments of its group, while those of the mimicked Heliconid are quite distinct. This would seem wholly to refute the argument that in such cases the likeness may spring from a real affinity between the two insects.” (See p. 176.) 5. SexuAL DIMORPHISM (ANTIGENY OF SCUDDER) AND Mumicry. The mimetic butterflies of North America, as in other parts of the world, are in large part mimetic in the female sex only, forming a special subsection of the far wider group of sexually dimorphic or antigenetic species. Dr. Skinner seeks to explain the special sub- section and the inclusive group by an appeal to the same general law. Thus, speaking of the mimetic females of North American Papilios, he says on p. 125: ‘These differences [between the sexes] occur in numerous species and it seems logical to consider that they are governed by a general law rather than that a few of them are caused by protective resemblance.’’ He uses the same argument concerning the female Argynnis diana, which Scudder maintains in the most positive terms to be a mimic of Limenitis (B) astyanax. (6, I, pp. 266, 287, 718; III, p. 1802). Comparing this Argynnid with five other sexually dimorphic species of the genus in North America, Skinner says on p. 126: “It does not seem consistent to pick out one species (diana) and say that its antigeny is due to tertiary mimicry. How can the dimorphism of the other species be explained?” But the female diana is, according to two eminent North American entomologists, Scudder and Edwards, picked out by nature and distinguished among the other antigenetic females by the fact that it resembles a species of a very different Nymphaline genus. I agree with them—although my opinion is worth very little as compared with theirs, for I have never seen the species alive—and I was seeking to place a resemblance which puzzled Seudder, in its true position among the mimetic butterflies of the Region. The far wider ques- tion of sexual dimorphism in general did not fall within the scope of my paper. Again, referring to the mimetic female Papilio, I do not know why it is specially logical to seek to explain by the same general law two very different categories, viz., the sexually dimorphic females that closely resemble other species and.those that bear no 1914.] NATURAL SCIENCES OF PHILADELPHIA. 169: such resemblance. I doubt whether Dr. Skinner would venture to apply the same argument to the polymorphic mimetic females of the Ethiopian Papilio dardanus or to many other examples that could be cited. The North American females are not nearly so striking as these, but their patterns are explained by the theories of mimicry and by no other theories as yet suggested. There are doubtless certain general principles which underlie the whole phenomenon of sexual dimorphism. One of these is obvious— the linking of color, pattern and structure (as we see in the shape of the wings or in the forefeet of so many butterflies) with sex a linking which is so apt to occur in insects as well as in several other groups, and is so specially conspicuous in the Lepidoptera Rhopalo- cera. To this principle I think another may be added, at any rate so far as the butterflies are concerned—the greater variability of sex-limited patterns in the female (23). But these general principles do not explain the different categories of antigenetic females, although they may, and I think do, explain the fact that there is material out of which these categories have been built by selection. They would also, of course, account for any antigenetic characters, if such there be, that have not been subject to selection. They are the nearest approach to a general law governing antigeny as a whole that can be offered in the present state of our knowledge. Beyond these principles we have, I submit, to look for special explanations rather than for general laws. (1) The mimetic females are probably to be explained, as Wallace suggested (2, p. 22), by the special needs and special habits of the sex, but also by the fact that the difference in pattern variability may be such that the evolution of mimicry is initiated in one sex and prevented in the other (23, p. 132). (2) A second class of female patterns is procryptic, meeting the special needs of the sex by promoting concealment. (3) Ina third class the whole or a certain proportion of the females of a species retains ancestral patterns (or structures like the fore feet mentioned above) which have been lost or become more degener- ate in the males. (4) Finally the fact that males are so often distinguished from their females by brilliant tints which are pigmentary in some species and structural in others and by scent-producing organs of many kinds strongly suggests anfimportant fourth class due to the operation of sexual selection. The summary briefly set forth in the last paragraphs will, I think, 170 PROCEEDINGS OF THE ACADEMY OF [Jan., show the hopelessness of any attempt to bring all the examples of sexual dimorphism under any single law except one which expresses the two principles explained on p. 169. The complexity of the subject is still further increased by the fact that different elements in the pattern of a species will often fall into more than one class. Thus Dixey has maintained that the female of Argynnis diana belongs to the third of the above classes except as regards “the large expanse of blue ground colour,’’ which is mimetic and belongs to the first class (7, p. 106, footnote). Tn his later paper (34) Dr. Skinner has still further developed his objections to any special interpretation of the various classes of sexual dimorphism in butterflies. He speaks of velvety patches on the fore wings of male Satyrin@ and brands on the wings of male Hesperide. The researches of Fritz Miiller (29) show that these structures are scent-producing organs, and there is no doubt that they are of use in courtship, or epigamic. The law that would be so comprehensive as to explain at once an epigamic scent-patch, the more rudimentary anterior foot of a male Nymphalid and the mimetic pattern of its female, would be so very general that it would not carry us any distance in the attempt to understand each of these different facts. Concerning Papilio glawcus glaweus and its dark turnus female (I adopt Rothschild and Jordan’s synonymy, 20, p. 582), which some naturalists at least regard as mimetic of Ph. philenor, Skinner says (34, p. 25) in criticism of Edwards: “There is also an assumption ‘to which I take exception. Does anyone know which onee|the dark or the male-like female] appeared first and why?” With regard to the last word “why,” Edwards had expressly disclaimed know- ledge, for he speaks of ‘‘some unknown influence” causing the appear- ance of the black female, and we can say no more than this to-day. With regard to the other part of the question, I think it may be shown that Edwards took the reasonable view in supposing that the dark female appeared later than the male-like one. The male pattern is shown to be ancestral, because it bears an intimate relation- ship to the pattern of other allied Papilios. This is the argument used by Scudder (6, p. 534) in the following passage: “In Jasoniades glaucus, where we sometimes have a black female, it is more difficult to decide what should be considered the normal color, owing to diversity of view upon the relationship of many of the swallowtails; but, to judge only from those agreed by all to be most nearly allied to it, there can be no question whatever that the striped character prevails.’’ 1914] NATURAL SCIENCES OF PHILADELPHIA. 171 The turnus female is a partially melanic variety, but the lines of the male pattern can be detected beneath the overspreading pigment. It also exhibits many features in its pattern which have received no interpretation except that they are mimetic of philenor or secondarily mimetic of the other Papilionine mimics of philenor (21, 467-471). No doubt there are examples in which it is probable that melanic females preserve something of an ancestral pattern, as in Argynnis diana or the valesina form of our British A. paphia (7, 103-5, 119-21), but I do not think that anyone has maintained that this is true of the melanic females of Papilios. It is, I submit, unreasonable to suppose that the male-like pattern first appeared almost hidden under the melanism of the twrnus female, and that the full pattern became evident by the clearing up of the dark pigment; whereas the opposite view, that the partial melanism appeared later, obscuring but not completely hiding a pre-existing male-like pattern, seems to me entirely probable. Such partial melanism, in my opinion, provided the foundation on which the details of the mimetic resem- blance were gradually built. As regard this same species, Dr. Skinner’s final conclusions (34, p- 26) are comprised in the following statement: “The evidence in favor of glaucus bemg brought about by mimicry is almost nil, while the evidence against it is very considerable. The species swarms in countless thousands in the north where glaucus does not exist.”’ When we add to these last words the fact that the model P. philenor is also non-existent in the north, Dr. Skinner’s argument seems to support the view he is attacking. P. philenor only enters New England and Southern Canada as a strageler and barely overlaps the range of the northern subspecies of P. glawews glawews, which Rothschild and Jordan distinguish under the name of P. glaucus canadensis (20, p. 586). As regards the closely allied P. rutulus, the same great authorities give reasons for considering it a distinct species. The whole range of glaucus glaucus—Florida to New England and westward to the- Mississippi basin—lies within that of P. philenor, and over this whole range the dark turnus female occurs intermingled with male-like females—the latter preponderating in the north, the former in the south. The evidence based on geo- graphical distribution seems to me strongly to support Edwards’ conclusions. And we may add that there are, as I have already said, details in the pattern of the dark females which ‘are not explained by any other hypothesis. Objections based on the great abundance of the non-mimetic ancestor are considered on pages 178, 179. Le, PROCEEDINGS OF THE ACADEMY OF {Jan..,. 6. THe FemaLe or NEOPHASIA TERLOOTI, ANOTHER NORTH AMERICAN Mimic oF DANAIDA PLEXIPPUS. Dr. Skinner remarks (34, p. 27): “What is the cause of the ex- traordinary antigeny seen in Neophasia terlooti? The male in this species is white and the female orange. The female of the species. was once sent to me as a ‘little Danais’ and it really looks like one. Here would be a good opportunity to build up a mimicry theory.’’! At the time when I read these words I had never seen the species, but Dr. Skinner has now very kindly sent me a male and female from Reef, Arizona (Noy. 2, 1903: Biederman). There can be no. doubt that the female 7s a mimic of D. plexippus. The comparison between the yellow of the under surface exposed in the position of rest and the orange of the upper surface, the blackening of the veins on the upper surface of the hind wing and other details to be de- scribed below are quite inexplicable on any other hypothesis. The mimicry is rather rough and there is no approximation in the shape of the wings. In both respects this female stands in about the same position as the females of the Neotropical Perrhybris (‘‘Mylothris”’). I am greatly indebted to my friend for this opportunity of examining and writing on what is to me an entirely new example of butterfly mimicry in North America—another result of its invasion by the Old World genus Danaida. My friend Commander J. J. Walker, who has had an intimate experience of the allied Neophasia menapia in Vancouver Island, tells me that during flight the latter is one of the feeblest of Pierines and that it suddenty appears upon the wing in immense numbers. He has kindly permitted me to make use of the following unpublished extracts from his journal, on H. M. 3. “ King- fisher” at Esquimalt, Vancouver Island: 1882, August 7: “Day still, hot, and cloudless. During the forenoon I was agreeably surprised by the appearance of a good number of specimens of a very pretty ‘‘White’’ butterfly... . It seems to come very near to Leucophasia, by the elongate shape and delicate texture of its wings, as well as by the rather short antenne and hairy palpi... . They were flying sluggishly in the sunshine over the water, and the signalman and I caught 15 on the poop in a very short time [the ship being about 300 yards off shore]. . Landed at 4 P. M.; the Leucophasia? was still on the wing, and I 1 Dr. F. A. Dixey remarked of N. terlooti in 1905 (Proceedings of the Entomo- logical Society of London, p. xx) : ‘“This latter butterfly is especially interesting as possessing a female which closely resembles some of the mimetic forms of Euterpe.” 1914.] NATURAL SCIENCES OF PHILADELPHIA. 173 caught 15 or 16 (at flowers of Matricaria), all in the most exquisite condition, like those taken on board ship. They all appeared to have emerged from the pupa on that day, as I had been on the lookout for some days past, and certainly did not see one on the wing yesterday.” August 8:. “Went on shore this forenoon at 11.30, to get a few more of the Leucophasia |Neophasia| while it remained in good condition. .... I had no difficulty in getting as many as I wanted - +... a day, however, had made a perceptible difference in its condition, as a good many were getting somewhat worn and chipped. They were very easy to catch, flitting from flower to flower in the open places [among the pines] and of very weak and sluggish flight.” August 14: The first 9 was taken on this date. “TI beat it out of a fir-tree.” c The fact that the only Pierine mimic in the Nearctic Region belongs to a genus with the characteristics described by J. J. Walker Suggests an interpretation on the lines of Fritz Miiller’s hypothesis. I now propose to institute a detailed comparison between the colored pigments of Neophasia menapia and terlooti. THE FEMALE OF NEOPHASIA MENAPIA.—Under surface of hind wing.—A colored spot, roughly triangular in shape, is found in the black marginal band of areas 2,3,4,5,and6. The spots, as well as the other markings described below, were orange in 4 females, orange-red in 2, and a rather pale vermilion in one. The tint in some individuals tends to deepen towards the base of the wing—especially along the costa. Beyond vein 7, viz., in areas 7 and 8, the pigment is continued at first as a narrow marginal line, which filling area 8 except at its very base, broadens with it toward the base of the wing. In the opposite direction, beyond vein 2, area le bears two spots, of which the upper is sometimes roughly diamond-shaped. These spots are placed one on each side of the dark line, representing a lost vein, which divides the area longitudinally into two sections. Below vein 1b a narrow marginal orange line extends over about 4 of the breadth of area 1b. In addition to these marginal orange markings, there is also an internervular development of the same pigment starting from the base of the wing, especially distinct in the lower or inner marginal section of area lc, which in favorable examples is highly colored over more than half its length starting from the base. In strongly marked females a few scattered orange scales are also seen in area 7 and in the upper section of area 1c, and they could probably be found in other areas of certain individuals. 174 PROCEEDINGS OF THE ACADEMY OF [Jan., Upper surface of hind wing. Most of the above-described mar- ginal features appear, but far more faintly, on the upper surface. The other orange marks are not represented on the upper surface of those females that I have examined, nor did they appear anywhere upon either surface of the fore wing. Tue Mae or NEOPHASIA MENAPIA.—Under surface of hind wing.— Sixteen examples were examined and of these about half had a comparatively few dull orange or sometimes yellow scales in one or more of areas 6, 7, and 8. When present they are precisely in the position of the marginal markings of the female. Since writing the above I have had the opportunity of examining 6 beautiful specimens from Esquimalt, in Commander J. J. Walker's collection. Well-developed marginal markings appeared on all females: om (1) a beautiful cinnabar red; on (2) a pale cinnabar red, a little deepened at the anal angle, apex, and costa; on (5) orange, becoming orange-red in the same positions. Of the 3 males, two possessed pale cinnabar scales at the apex and along the costa.. one of them bearing a few at the margin of the upper section of area le and still fewer—only 2 or 3 scales—in the lower section. The third male had pale yellow marginal scales at the apex and costa, a few becoming faintly reddish, especially at the apex. Commander Walker tells me that these butterflies were all “set” immediately after capture, and that they have never been ‘‘relaxed”’ and ‘‘reset.”’ . Inasmuch as Prof. Gowland Hopkins has shown (14, 12) that the pigments of Pierine are soluble in water, it is probable that Walker’s specimens more truly represent the colors of the living insect than do any of the others here described, for all of these have been ‘“‘relaxed”’ at least once. Tue Mae or Neopuasta TERLOOTI.— Under surface of hind wing.— The marginal markings of the female menapia are represented on the male of terlooti, smaller indeed, but with a far richer color, being of a bright, rich vermilion tint. In the single specimen I have had the opportunity of examining these markings are solely marginal. They are wanting from area 4 and so slightly developed in all areas except 6,7, and 8 (where they are purely linear and do not fill the last-named area as in the Q menapia) that it would be easy to count the con- stituent scales with a lens. In the specimen before me there are only 3 vermilion scales in area 5 on the left side and only 5 on the right, but they are more numerous and usually far more numerous in all the other markings. Although the dark pigment is com-- paratively weakly developed in the male, area le is divided very 1914.] NATURAL SCIENCES OF PHILADELPHIA. 175: distinctly by a strongly marked linear streak, and the 2 orange spots of the female menapia are represented by 2 marginal groups of vermilion scales, one in each section of the area. Vermilion scales occur nowhere else on the specimen, although those described above can be distinctly seen through the translucent scales of the upper surface. THe FemaLe or N&opHASIA TERLOOTI.—Under surface of hind wing—The vermilion markings are developed almost precisely in the positions of the orange markings of the female menapia—more strongly at the margin and the extreme base of the wing, but much less so elsewhere. The lower section of area le is, however, richly marked with vermilion for 4 of its length from the base. The rest of the colored markings are light yellow of an ochreish tint, rather distinctly different from that seen elsewhere on the wings. Under surface of fore wing.—The marginal markings and the marginal part of the chief orange patch are also light yellow, but of a lemon tinge. The orange of the chief marking and of scattered scales forming a linear mark in the cell is very rich and deep in tint: the mark in the cell is in fact better described as orange-red. The two marks at the end of the cell, in areas 5 and 6, are transitional in tint between the yellow marginal and the more central orange markings, and there is transition to be observed between the yellow margin in areas la and 1b and the rest of the chief orange marking. These changes in color are effected by a gradual increase in the number of orange scales and not by any real transition between the yellow and orange pigments, although if we study the wings as a whole we find several tints of orange and yellow. Upper surface of hind wing.—The vermilion markings are repre- sented by comparatively few scales. Within these markings the submarginal spots and the ground-color of the rest of the wing are deep orange, but of a duller tint than that of the fore wing. The submarginal spots of the outer (hind) margin are slightly less deep in tint, while along the costa, where the surface is concealed beneath the fore wing, the orange scales are gradually replaced by yellow, and again, at the extreme margin, by black, with perhaps a trace of the vermilion which is so distinct on the opposite surface. The vermilion scales could not be properly investigated because of the overlap of the wings. Upper surface of the fore wing.—The colors are nearly as on the under surface, but, except at the apex, the submarginal spots and the margin of the principal marking are distinctly less pale and 176 PROCEEDINGS OF THE ACADEMY OF [Jan., therefore much nearer to the tint of the orange ground-color of the rest of the wing. The transition here does not appear to be effected by a gradually increasing number of deep orange scales, but by a gradual increase in the depth of the tint. The two marks at the end of the cell are nearly as rich an orange as in the expanse be- low the cell, and the transition towards yellow is, on the upper surface as compared with the lower, shifted towards the costal margin, occurring in the two spots of the same series placed above the end of the cell in areas 10 and 11. The linear spot in area 11 is yellow with thinly scattered orange scales, which are far more thickly placed on the spot in area 10. 7. THE CoLoRED PIGMENTS OF THE PIERINAS’ AS ILLUSTRATED BY NEOPHASIA. Professor F. Gowland Hopkins has shown (11, 12) that the white pigment of Pierines is an impure uric acid, and that the yellow orange and probably the red pigments are a derivative of uric acid which he calls ‘“‘lepidotie acid.”” No pigments of simiiar constitution were found in any other butterflies. Therefore, when a Pierine mimics an Ithomiine or, as in N. terlooti, a Danaine, the resemblance is effected by the production of an entirely different coloring matter. Gowland Hopkins believes that the yellow, orange, and red Pierine pigments are chemically nearly allied and may pass one into the other by slight changes perhaps in the degree of oxidation. He observed that one tint was represented by another in corresponding markings of opposite sexes or allied species. Thus he remarks (12, p. 678): “Tt is interesting to note, by comparing various allied species of Delias, that the red marginal spot may become more yellow, while the yellow area usually found at the root of the wing may become more red, till both may exhibit a uniform orange colour, or the change may go farther and red and yellow change places without the general color-plan of the wing being altered.” These conclusions are strongly supported by a careful study of Neophasia, where it has been shown that in different individuals of the same sex of menapia the same markings may be either orange, orange-red, or pale vermilion, while in the opposite sex they may be absent or feebly represented in dull orange or yellow. Again in the allied terlooti the corresponding markings are a rich deep vermilion in both sexes. We are led to realize that it is very easy for Neophasia to produce any shade between a pale lemon-yellow and a rich ver- milion. The colored markings of menapia cannot be regarded as - 1914.| NATURAL SCIENCES OF PHILADELPHIA, 177 mimetic, and, if Danaida plexippus had never entered America, it is highly improbable that anything more than the corresponding colored markings would have been evolved on the wings of the female terlooti. The range of tints in the markings common to menapia and terlooti gives an indication of the variational material out of which selection built up the mimetic pattern. The peculiar shade of yellow of the under surface of the hind wing, the rich orange of the central parts of the upper surface, the paler tints of the marginal markings, especially at the apex of the fore wing, the emphasis by black pigment of the veins of the hind wing upper surface, are all elements in producing the result—a somewhat rough but at a distance almost certainly a deceptive mimetic likeness to D. plexippus. The same considerations help us to understand the prevalence of Pierine mimicry in tropical America as compared with other parts of the world—because of the predominant Ithomiine and Danaine with warning patterns made up of reds, yellows, whites, and blacks. Such patterns are mimicked by the Pierine genera Dismorphia (in the broad sense), Perrhybris (‘‘Mylothris”), Archonias, Hes- perocharis, and we can now add the North American Neophasia. 8. THE RESTRICTION OF SEX-LIMITED INHERITANCE TO THE Mimetice PATTERN OF NEOPHASIA TERLOOTI. The older colored markings common to the females of menapia and terlooti are only partially sex-limited, being inherited in a very reduced form by some of the males of the former species and probably by all of the latter. The more modern mimetic pattern of the female terlooti is strictly sex-limited. The facts harmonize with the hypothesis that female mimicry is largely due to the great variability of this sex in Lepidoptera and the freedom with which it offers to selection a wide range of sex-limited colors and markings, but that when a pattern has been long established it tends to be transferred to the opposite sex. : The older non-mimetic marginal markings suggest that the trans- formation of uric into lepidotie acid is especially easy in this part of the hind wing and invite comparison with the number of mimetic Neotropical Pierines in which marginal or submarginal reds have been developed in the same position, viz., on the under surface of the hind wing—a study that would carry me too far from the subject of the present paper. 12 17s PROCEEDINGS OF 'THE ACADEMY OF [Jan., 9 THe EvoLuTion oF LimeniTiIs (BASILARCHIA) ARCHIPPUS FROM AN ANCESTOR WITH A PATTERN LIKE THAT OF L. (B.) ARTHEMIS. The origin of archippus, suggested in the title of this section, is due to Seudder (6, 277-8, 714). All I have done is to support the published views of this distinguished naturalist by making a careful analysis of the markmgs of archippus and arthemis, by this means demonstrating that the details of the mimetic pattern are accounted for on his hypothesis. I am sorry to find that neither Scudder’s hypothesis nor the results of my analysis carry any conviction to Dr. Skinner, who uses the following words: ‘‘Arthemis and weide- meyeri [with a very similar pattern] have flourished prosperously in the struggle for existence, and it is difficult to understand why archippus should be so specially favored. The statements attempt- ing to prove the evolution of archippus from an ancestral form (arthemis) seem to me very inconclusive” (32, p. 127). Dr. Skinner makes no alternative suggestion as to the origin of the mimetic species. The doctrine of evolution—for it is hardly necessary to discuss the ancient belief which would assume that archippus was originally created in its present form—leaves us only two hypotheses. Either archippus was evolved from some form of Limenitis which has entirely disappeared or from one which is more or less closely repre- sented by a species still in existence. The former alternative aban- dons the problem as insoluble, and abandoned it must be if there is no sufficient evidence that the ancestor can be reconstructed from any existing form. I agree with Scudder in preferring the counsel of hope to the counsel of despair.. L. (B.) arthemis and weidemeyeri present us with an ancestral pattern wide-spread in the genus and found not only in North America, but also with little change in the Old World section of the temperate cireumpolar zone. Archippus is so closely related to arthemis that the larval and pupal stages are almost identical, and although the imaginal patterns are so different, Scudder indicated, and I have attempted to trace in detail, the manner in which one pattern may be derived from the other. I really think that if Dr. Skinner, with specimens of archippus and arthemis before him, will verify the details of the account in my earlier paper (21, pp. 456-459), he will find that many minute features on the wings of the mimetic species are interpreted and correlated in a satisfactory manner. And a hypothesis that interprets stands, until replaced by another that interprets better. . With regard to Skinner’s inference that inasmuch as arthemis 1914.] NATURAL SCIENCES OF PHILADELPHIA. 179 flourishes prosperously, it is unlikely that a mimetic form would arise from it, this is an objection which at once arises when mimicry is studied in the original monograph of its founder, published long before Fritz Miiller had thought of his hypothesis. According to Bates, mimicry was a refuge for the destitute, a last means of escape for a hard-pressed and dying species. It was this very conclusion which was Miiller’s stumbling-block; for the majority of the mimics in southeast Brazil where he lived were clearly successful and abundant species, and the same is true of the majority of mimicking species wherever they are thoroughly known. Nor is there any reason to suppose that these successful forms originally arose from rare and hard-pressed non-mimetic ancestors. Want of space prevents the dis- cussion of more than a single example. I refer to Tirumala (Melinda) formosa, an Oriental invader into the Ethiopian Region (18, 31). This species, abundant east of the Victoria Nyanza, near Nairobi, is there beautifully mimicked by the Ethiopian Papilio rex. The invading Danaine has transformed an indigenous species just as in North America. West of the great lake 7. formosa is represented by an equally flourishing daughter species, 7. mercedonia, with a pattern darker than its parent and one much further removed from the allied Oriental Danaine. Pa pilio rex west of the lake becomes P. mimeticus, as beautiful a mimic of 7. mercedonia as rex is of formosa. The two Danaine models are now distinct species, but their Papilionine mimics, connected by intermediates (P. commixtus) in the interme- diate geographical area northeast of the Victoria Nyanza, are certainly a single interbreeding community. Similarly, in North America Danaida plexippus is a very distinct species from D. berenice and D. strigosa, although these latter may be geographical races of one species. The three forms of Limenitis are, on the other hand, all probably mimetie modifications of a single species, although L. obsoleta is probably distinet from archippus and floridensis. To con- tinue the history of the African invading Danaines: Further westward the flourishing and prosperous 7’. mercedonia has given rise to a still darker species, 7. morgeni, which has altogether lost the appearance of an Oriental Tirwmala and has become the most perfect mimic of the African Danaine genus Amauris. Here, then, we have a species so dominant that it is mimicked by a butterfly of a different family. It gives rise to another species and the mimic undergoes corresponding changes. Finally, in spite of these evidences of prosperity, it becomes itself a singularly perfect mimic. All these changes are far less abrupt than that from arthemis 180 PROCEEDINGS OF THE ACADEMY OF .[Jan., to archippus, and I do not think that any naturalist who recognized the traces of the pattern of mercedonia still lingering almost invisible on the surface of morgeni or concealed by the overlap of the wings would doubt that the former is the ancestor of the latter and that the | model has become itself a mimic. Finally, it must be remembered that L. archippus has a far wider range than orthemis, and it is reasonable to suppose that this advan- tage has been conferred by its mimetic pattern. Arthemis is confined to Canada east of the Rockies and to the northeastern States, while archippus is “found over very nearly the same area as Anosia plex- ippus”’ (6, 278). 10. THE RELATION OF THE PATTERN OF LIMENITIS OBSOLETA (HULSTI) TO THAT OF ARCHIPPUS, ARTHEMIS AND WEIDEMEYERI. When I wrote the paper criticised by Dr. Skinner (22), as well as an earlier paper, in some respects more detailed (21), I had never been given the opportunity of examining a series of the Arizona and Utah mimic, Limenitis obsoleta (hulsti), and my brief account was founded on the excellent fig. 5 on plate VII of Dr. W. J. Holland’s work (417.) In January, 1909, when I had the honor of repre- senting my country at the Darwin centenary in America, my friend Dr. F. A. Lucas, Director of the American Museum, Central Park, New York, showed me a series of obsoleta together with its model, Danaida strigosa. The specimens were in the Brooklyn Museum, of which Dr. Lucas was then Curator. I saw at once that’ the form was very variable and that my work required the study and com- parison of a long series of individuals. Dr. Lucas very kindly obtained a few specimens of the model and mimic for me and put me in communication with Dr. R. E. Kunzé, of Phoenix, Arizona, who has generously provided me with a fine mass of material. The following account has been drawn up from the study of 24 males and 9 females from Phoenix and 2 males and 1 female from Tueson. Thirty-three specimens bear the precise date of capture, one the month and year, one a month of which the interpretation is uncertain, and one for which the month is not recorded. Omitting these last two, the dates of capture are given in the following table. The three 1896 specimens were captured at Tucson (2,400 feet) in southern Arizona, the remaining 31 at Phoenix (1,100 feet) in the valley of the Salt River, southern Arizona. Apr: 9, 1896 nnciuansore 1 Q © Stine: 16,1896 22 hae Apr. 10, 1896... . 1c Apr. 22, 1897...... -1¢ 1914.| “NATURAL SCIENCHS OF PHILADELPHIA. 181 Apr. 17, 1909 eden ol OS ditiby als lO lo Sept. 21, 1909 on ee Othe oat O10), taee = aa terest Sept. 30, 1909 Peer One OC a O10k & add ork @ Oct. 27, 1909 Le lee Apr TOM! e Ael2@ Oct. ......, 1909... etn ollinopl es alyoyr, <1 SURO TE ee Selec Mar. 26, 1910 efile Soe Seales 2o6' Mar. 27, 1910 en leche Sepia oe Lolin a eS Mar. 29, 1910 pert lO) Sepie bie 0911 mv te Apr. 1, 1910 eee leche sepia 2is.1 it ios Apr. 4, 1910 peel ot wept. 24es19i'T1 5 EG Apr. 6, 1910 elect Oke 4. 191d 12 The existence of two broods, one emerging between the end of March and the end of April, the other in September and October, are clearly shown. The two specimens in June and July, respectively, were probably representatives of a third brood. The model, Danaida strigosa, appears to be much rarer than its mimic at Phoenix—at any rate, in the localities where Dr. Kunzé collected. From this place I have only received 2 males, captured July 2 and 6, 1912; from Tucson—1 female May 26, 1 male June 7, 1 female June 9, 1 male August 19, all in 1896; from Prescott (5,400 feet), in western Arizona—2 males and 1 female July 15, 1912. Dr. R. E. Kunzé, of Phoenix, Arizona, who has had a long and intimate experience of the butterfly fauna of the State, kindly informs me that, in the Phoenix (1,100 feet) and Tucson (2,400 feet) districts and between them, L. obsoleta is almost exclusively found in the valleys, along the river-bottoms, and by the canals, where its larval food-plant, a willow, grows It is commoner in the river-bottoms, especially near the streams, than by the canals. Danaida strigosa flies with it in these situations and is indeed commoner there than elsewhere, but, unlike the mimic, it is also found in other places. It is impossible to state the relative proportions of Danaine and Limenitis, but by the rivers and canals the mimic is the commoner in the ratio of about twelve or fifteen to one. The proportions at Tucson and Phoenix seem to be the same. Danaida plexippus occurs, but is scarce in the Salt River valley at Phoenix. Dr. Kunzé estimates that it may exist in the ratio of one to fifteen of D. strigosa, but in some seasons he does not meet with * The armatures of two of these males were studied by Dr. Eltringham (p. 190). 3 Dr. Kunzé adds in his letter of August 5, 1913: “TI should say that obsoleta has here [Phoenix] from 3-4 broods in a season, from April Ist up to November Ist, in a mild autumn, of course. I think the last brood oviposits on cotton- wood, our Populus fremonti and other species, because its leaves keep green fill latter part of December, whereas willow drops leaves earlier.” 182 PROCEEDINGS OF THE ACADEMY OF [Jan., it at all. At Prescott, Arizona (5,350 feet), 135 miles north of Phoenix, strigosa flies in the company of plexippus from July to September, the latter being the commoner of the two. D. strigosa extends as far south as Galveston, Texas, and may also occur in some parts of Mexico, near the northern boundary. Limenitis obsoleta does not occur at Prescott. The fine series of L. obsoleta (hulsti) tabulated on p. 180 at once made clear to me that the Arizona form is not, like floridensis (eros) in Florida, a local race of L. archippus transformed by mimicry of the dominant local Danaine, but the bearer of an ancestral pattern which preserves features lost by the two other mimetic races. I therefore desire to correct my former conclusicn, founded -on the figure of asingle specimen, that obsoleta is a modified form of archippus (21, p. 460, 22, pp. 171-2). At the same time I remarked in the latter paper (p. 172): “I have not yet had the opportunity of ascer- taining whether this hypothesis is supported by evidence derived from a careful study of the pattern.” The hind wing.—The most prominent ancestral features of obsoleta are the ‘traces of the white discal band derived from an ancestor with a pattern like that of arthemis or weidemeyeri. In archippus and floridensis a trace of the white band is found on the under side of the hind wing in some specimens, but so far as my experience goes never on the upper surface. In obsoleta some trace of it is always present on both surfaces, but when, as in the majority of specimens, there is a difference in the degree of development, it is stronger upon the under side. It is more strongly developed in the females than the males, and this is the general rule with the ancestral features of the species, as it appears to be in archippus, of which a certain proportion of the males in the Albany district, but no females, have entirely lost the black discal stripe from the upper surface of the hind wing (recorded by Mr. John H. Cook, 22, pp. 211-212). Thus the white stripe, together with its black outer border, is evanescent on the upper surface of the hind wing of 2 female obsoleta from Phoenix and small in the female from Tucson, whereas the same feature is evanescent in half the males from Phoenix and but slightly developed in others. The evanescent feature in both males and females is more strongly represented, generally far more strongly, on the under surface. The degree of development of the black band is generally related to that of the white, the two being usually evanescent together or well developed together, but the range of variation is much greater in the white than in the black, corresponding with the entire disappearance 1914.| NATURAL SCIENCES OF PHILADELPHIA. 183 of the former but not of the latter from the upper surface of the allied archippus. On the other hand, the development on the under as compared with the upper surface is greater in the black than the white. In both sexes there is a tendency, as in archippus, to throw the white spots on the under surface of the hind wing into relief by an inner edging—a darkened shade of the ground-color in areas 2, 3, 4, and 5, still darker and often black in areas 6 and 7. This feature probably represents the black inner border of the white discal band in the non-mimetic ancestor. When the 32 specimens, omitting the 2 taken in June and July, recorded in the table on p. 180, are arranged according to their two broods—the 15 March and April specimens together and the 17 September and October together—it is seen that there is a small but distinct seasonal difference in the development of the trace of the white discal band of the hind wing and its black outer border. The spring brood is in this respect distinctly the more ancestral, bearing on the average stronger traces of the pattern of weidemeyeri and arthemis. ‘This is true of the females as well as the males, as may be inferred from the following statement: Females (spring brood = 4, autumn brood = 5).—The only 2 specimens with evanescent band and border bear the dates Sept. 11, 1911, and Oct. 10, 1910. The most reduced band of the spring brood is seen in the Tucson specimen, April 9, 1896. In all the remaining 4 spring females, the band and, in all but one, the border is distinctly stronger than in either of the 2 remaining autumn females. Males (spring brood = 10, autumn brood = 18).—It is extremely difficult to classify the degree of development of the band and border —there is a complete and gradual transition. There is, however, a marked difference at both ends of the scale between the two broods. The most evanescent white bands are seen in 6 autumn males. In all these the feature is more reduced than in any spring male. Very small and reduced bands are found in 3 males of each brood. Beyond these there is the most gradual transition to the highest degree. of development found in the sex, and among these we find by far the highest in a specimen captured April 22, 1911, while 2 other spring males are rather beyond any of the autumn brood. Considering the black border separately, the difference is even more marked, for this feature is evanescent in 4 of the autumn brood and none of the spring, while the next 4 are equalled and on the whole slightly exceeded by the 4 spring specimens in which the feature is least de- veloped. The black border is more highly developed in 4 of the spring 184 PROCEEDINGS OF THE ACADEMY OF [Jan., brood than in any of the autumn. This detailed comparison has been extraordinarily difficult to make, because of the perfect transi- tion and the minute shades of difference. When the attempt was made to express the difference, the specimens grouped themselves into fours in an irritating and unnatural manner. It might perhaps have been wiser to attempt no analysis of so transitional a feature, but to be contented with the statement that a distinct difference exists at both ends of the scale, the band and border of the most strongly marked specimens being decidedly more developed in the spring brood, while the reduction of these features in the least strongly marked specimens was carried distinctly further in the autumn brood. I cannot but think, however, that my attempts at an analytical comparison, whatever faults there may be in the details, are a truer expression of the facts. An interesting difference between the upper surface of obsoleta and that of archippus is common to both fore and hind wings, viz., the far more heavily blackened veins gained by the latter in mimicry of D. plexippus. Floridensis here shows its origin from archippus, for it retains the darkening along the veins, although out of place in a mimic of D. berenice. Nosuch evidence of having passed through an archippus stage is to be seen on the upper surface of obsoleta. The veins are heavily blackened on the under surface of the hind wing in all three mimics, in evident likeness to their respective models, although obsoleta in this respect is less darkened and a less perfect mimic than the other tyo. - In certain specimens of obsoleta there is to be seen on the hind wing under surface two largish rich brown sharply outlined patches, one in the cell and one near the base of area 7. On the basal side of each patch is a white spot and a white suffusion commonly surrounds the projection of the precostal into area 8. These elements tend to become evanescent together and distinct together, acting like a single feature. Slight traces of these markings can probably be found on every fresh specimen. They were remarkably pronounced in the female taken Sept. 5, 1911 (p. 181). These vestiges, except in one respect, resemble the well-known basal pattern of arthemis far more closely than that of weidemeyeri. The pale elements are, however, for the most part blue in arthemis, but nearly white in weidemeyeri, and therefore in this respect nearer to obsoleta. Archippus has advanced further from the ancestral forms than obsoleta, for ‘‘the basal red patches have vanished, but the pale blue marks in and on the costal side [area 7] of the base of the cell are retained, and, 1914.] NATURAL SCIENCES OF PHILADELPHIA. 185 lightened in tint, represent the two more conspicuous white spots occupying nearly the same position in Anosia [Danaida]”’ (21, 456-7). Now that I have had the advantage of studying obsoleta, and have re-examined archippus m the light of the new experience, I find that a few examples do possess a very faint trace of the reddish patches of arthemis. In these vestiges as in so many other features in the pattern we are led to conclude that obsoleta represents an older stage in the evolution of archippus. The fore wing—The inner edge of the angulated black outer border of the white band of arthemis and weidemeyeri runs from the costa to the inner margin of the wing, near but well within the pos- terior angle, although it is broadened so far that its outer edge enters this angle; in most specimens of archippus it runs to the junction of the middle and posterior third of the outer (hind) margin (21, p- 457). Some females, however, approach the condition of obsoleta, which is generally far nearer in this respect to the pattern of arthemis and weidemeyeri. In obsoleta the direction of the vestigial black outer border, which, except near the costa, is evanescent on the upper surface, can be easily traced by fixing the attention on the outer ends of the four prominent white spots in areas 3, 4, 5, and 6. With this as guide, the eye is led on to an evanescent white spot nearly always present in area 2, and in certain individuals to the faint continuation of the black line towards the posterior angle. The angle made with the costa is very different from that of archippus. The black line is usually far more distinct on the under surface, and here it may be seen in many specimens that the direction changes abruptly in area 1b, becoming parallel with the outer margin and leading to a termination on the inner margin within, and often well within the posterior angle. In well-marked specimens, especially in the females, the black line is seen to lead to the outer end of a white linear mark close to the inner margin in area la (see p. 186). Faint vestiges of the former white band can even be made out in 1b on the under surface of a few individuals. There is great variation in the position of the black line in 1b. In most males it unites with and continues as a broadening of the black margin. The white spots which represent the costal half of the white band of arthemis and weidemeyeri are far better developed in obsoleta than. in archippus. In the latter the spots are 2 to 4 in number, the last being very small. In obsoleta there are always 4 large and distinct spots, especially well developed in the female, while a minute 5th spot, already mentioned as placed in area 2, is nearly always present * 186 PROCEEDINGS OF THE ACADEMY OF [Jan., and often more strongly marked on the under surface. A trace of it ‘could be made out on the upper surface of all the females and on 17 of the males; from one or both sides of the remaining males it was absent, but it is likely that when these were fresh examination with a lens would have led to the detection of a few white scales. It is clear that the trace of the original discal band is more shortened in archippus than in the Arizona form, and that the 4th spot in area 3, -or In other individuals the 3rd in area 4, is now in the position of the minute trace of a 5th spot in area 2 of obsoleta. Furthermore, the black discal marking retains in obsoleta more of its original appearance as an outer edging to the white band than in archippus—an appear- ance still more fully sustained upon the hind wing. In the fore wing of archippus it is obviously much developed, especially at the costal end, in mimicry of the model plexippus. The trace, on the costa itself, of the anterior end of the white band of the fore wing, already described as generally to be found in archip- pus (21, p- 457), was present in all the females and 19 males of obsoleta, but in some of these it was barely visible. This feature is apparently more often wanting altogether from archippus, but the two forms have reached nearly the same level, and I think that in both examination with a lens would reveal the presence of some trace of the marking in most or perhaps all fresh specimens. I have already incidentally mentioned on p. 185 the most inter- esting ancestral feature in the fore-wing pattern of obsoleta, and one entirely wanting from archippus, viz., a distinct trace in area la of the inner marginal end of a white dis¢al band like that of arthemis or weidemeyeri. This linear mark was present, varying in the degree of its development, in all the females and 23 males, and traces might probably have been found on all when fresh. The mark is also to be found on the under surface where the fore wing is overlapped by the hind, but for this reason it was only examined in a few specimens; in these it did not stand out on the paler ground- color as conspicuously as on the upper surface. It has been already pointed out on p. 185 that the outer end of the mark coincides with the point on the inner margin indicated by the direction of the vestige of the black outer border in some individuals, viz., a point well within the posterior angle of the fore wing. In a single female (Apr. 17, 1909), unfortunately rather worn, the mark in la appar- ently extends to the black margin at the posterior angle. The same relationship to a mark stopping short of the angle is also ndicated, especially in fresh specimens of the female, by a distinctly 1914.] NATURAL SCIENCES OF PHILADELPHIA. 187 paler shade of the ground-color outside the discal black stripe on both surfaces of both wings. The change of shade follows the suggested direction of the black line to the inner margin of the fore wing, although near this border it is not sharply demarcated as on the rest of the wing. Such an abrupt change in the depth of the color is very rarely to be seen on the upper surface of archippus. Scudder has looked on the reddish spots of arthemis, occupying the very position of this paler shade in obsoleta, as the foundation from which the mimetic form arose (6, p. 714), and I have followed him (21, 22). If we are right, and the transformation occurred first in this area and only later in the area inside the white discal stripe, it is easy to understand why there should be a difference in the shade of the ground-color for natural selection to seize upon. The Arizona Danaida strigosa is also paler on the outer than it is on the inner part of the wings, although the transition is gradual and not sharp as in obsoleta. On the under surface of the fore wing archippus is, in this very respect, more strikingly ancestral than obsoleta, the pattern of the model having been such as to emphasize the feature. Archippus is also commonly ancestral as compared with obsoleta in the distinct indication by a reddish-brown tint of the red submarginal spots on the under surface of both wings (21. p. 456). The white mark in area la of the fore wing has this further interest, that it mdicates the point at which the outer edge of the discal band of the hind wing met that of the fore, reconstructing for us a pattern like that of weidemeyeri and arthemis in which the band of the hind wing is placed much further from the outer margin than it is in the other wing. The evolution of the marginal pattern of both surfaces of both wings of obsoleta from a condition like that of arthemis appears to have been the same as in archippus (21, pp. 456-459) and to have reached nearly the same result. The slight differences cor- respond.with those between the respective models and are doubtless due to mimicry. : The two white spots in the fore wing cell on the under side were present in all the males of obsoleta. The females showed greater variability, the basal spot beg sometimes absent, but generally much larger than in the males. On the upper surface of the same wing the distal spot was large, for this feature, in 6 females, small in 3, minute in 1. In 14 males it was sharp and distinct, though small, and it could be detected in 8 of the others. In the remainder the triangular black mark in which the white spot lies could be made out by looking carefully for it. White scales were probably origin- 188 PROCEBDINGS OF THE ACADEMY OF [Jam., ally present on this mark in some of the worn specimens that do not now possess them. This white spot can be far more frequently detected on the upper surface of obsoleta and archippus than on that of arthemis and probably more often than in wetdermeyeri, although it may attain great relative size in this species (21, Pl. X XV, fig. 1). Its freyuent appearance in the two mimics points to an origin from an ancestor of the existing North American species that was in this respect nearer in pattern to L. lorqwini, in which the spot is almost invariably well developed (21, 479, 480, Pl. XX, figs. 6-8). At the same time the redevelopment of an ancestral feature by means of mimicry must not be lost sight of as a probable interpretation. The pattern of D. strigosa is such that the spot in the fore-wing cell of obsoleta probably adds to the likeness, at any rate during flight- The strong development of the feature in the females—in this species the more ancestral sex—favors the former hypothesis. As regards the traces of the Limenitis pattern persisting in the fore-wing cell on the under surface and their transference to the upper surface, obsoleta and archippus have reached nearly the same stage. The most strongly marked individuals of the former are, however, more ancestral, in that the white spot on the upper surface and the two spots below are larger and more conspicuous than in any examples of archippus. The seasonal differences on the fore wing were not so well marked as on the hind. Furthermore, the relationship was reversed, the autumn brood being more ancestral than the spring. The difference, however, was barely detectable except In one feature where it was very distinet—the minute white spot in the fore-wing cell. This was sharp and distinct in 11 out of 13 autumn males and only 2 out of 10 spring males. It was also on the whole better developed in the autumn females. Temperature experiments on the pup and, if possible, on the ova and larve would be well worth trying on this form as well as on archippus and floridensis. Considering what has been done by Dorfmeister, Weismann, Merrifield, and Standfuss, remembering also that Lamborn has recently brought evidence which suggests, although it does not prove, that vestiges of ‘‘tails’’ can be brought back to the hind wings of the tailless mimetic females of Papilio dardanus (26), it is quite probable that some increase in the pattern derived from a non-mimetic ancestor might be induced by the shock of heat or cold applied to the pupal or both larval and pupal stages. And the fact that there are certainly some seasonal differences in the 1914.] NATURAL SCIENCES OF PHILADELPHIA. 189 ancestral elements of L. obsoleta renders such experiments especially hopeful. An experiment made by Edwards and quoted by Scudder (6, p. 278) is also encouraging. The black band of the hind wing of archippus was widened in two butterflies which emerged from pup subjected to cold, being in one specimen, a female, nearly three times the normal width. It is necessary, in conclusion, to point out in a few words some special effects of the Danaine model, P. strigosa. Most prominent among these is the peculiar shade of the ground-color of obsoleta, so different from that of archippus and floridensis and so strikingly like that of the model. The triangular shape of the discal spots of the fore wing, especially pronounced in those of areas 3 and 4, has evidently been produced in mimicry of the characteristic-looking triangular and diamond-shaped spots of the model. The direction -of the line of these spots in obsoleta which has been shown on p. 185 to be more ancestral, viz., more like that of arthemis and weide- meyeri, than i archippus, has doubtless been stereotyped by the model, in which four of the most conspicuous white spots in areas 1b, 2, 3, and 4 are parallel with the outer margin of the fore wing. It is also probable, as suggested in a former paper (21, p. 460), that the retention of the white spots representing the discal band on the hind wing upper surface, and it may be added the linear mark in area la of the fore wing, has been aided by ‘“‘a general likeness ”’ [during flight] “‘to the pale-streaked hind-wings of strigosa.”’ Here, too, the relative development of the feature in the female favors a different interpretation; for, as already pointed out (p. 182), the female is slightly the more ancestral and the male the more advanced mimic in this species. The fact that the traces of the black border of the white discal band, which undoubtedly interfere with the mimetic resemblance, on the whole follow the white spots in the degree of development (p. 182) is also in favor of the sup- position that the entire marking is an ancestral feature which has not yet been got rid of. In order to prove that obsoleta is, as its pattern strongly suggests, ancestral as compared with archippus—that it stands in a position intermediate between the latter form ahd the non-mimetic species -of Limenitis—arthemis and weidemeyeri—it is necessary to seek for another line of evidence. 190 PROCEEDINGS OF THE ACADEMY OF [Jan.,. 11. Toe Mate Genirat ARMATURE OF THE NortH AMERICAN Forms oF LIMENITIS. In former years I have felt, with many other naturalists, some- suspicion of the conclusions based on a study of the male genitalia of Lepidoptera. The organs are so complex and in parts so thin- walled, so liable to be deformed by twisting and pressure, that it seemed unlikely that they could escape alteration in the processes of manipulation and mounting. Their shapes are such that a slight difference in the angle at which a drawing is made or a photograph taken makes all the difference to the result. I have, however, been converted by my experience of the work of my friends Dr. Karl Jordan and Dr. H. Eltringham. I have seen the latter naturalist preparing and studying the same parts in different individuals again and again until he was able to determine with complete certainty the actual form that is characteristic of the species or race. I there- fore asked him if he would kindly help me by preparing and drawing the genitalia of the North American forms of Limenitis. In asking this favor, I was, all unconsciously, making ready for a most valuable test of the validity of the method and its results. At the time when Eltringham made his drawings we had no copy of Scudder’s great work (6) available, but, when they were finished, I borrowed the volumes from the library of the Entomological Society of London. I turned at once to Plate 33, representing the genitalia of the Canadian and eastern North American species of Limenitis, and found that the four figures (9, 11, 12, 15), prepared by Edward Burgess for Scudder, might almost have been copied from Eltringham’s drawings or the drawings from the figures! Two careful pieces of work carried out independently have led to precisely the same result. It will therefore be admitted that we may safely accept the six figures on the accompanying Plate V as the expression of the true structural relationships in the different species. Figures 4, 5, and 6 on the right side of Plate V represent the male genital armatures of species also figured by Seudder, save that his L. astyanazx (fig. 15) represents the eastern race and Eltring- ham’s (fig. 4) that from Arizona. But the form of the genitalia is nearly the same, as may be seen by comparing the figures, allowing of course for the difference in magnification. Eltringham’s figures also show with Scudder’s the close resemblance between astyanaxr and arthemis (fig. 5, Scudder’s fig. 9). The two representations of L. archippus are almost identical, save that Scudder (figs. 11 and 1914.] NATURAL SCIENCES OF PHILADELPHIA. 19t 12) represents the end of the terminal hook as obliquely truncated, Eltringham (fig. 6) as a simple point. Figures 1, 2, and 3 on the left of the plate represent forms of Limenitis from an area outside the limits of Seudder’s monograph. The claspers of L. lorquini (fig. 1) are seen to differ markedly from those of all the other forms. Weidemeyeri (fig. 2), on the other hand, closely resembles arthemis and astyanax, although it is of a stouter build. The main interest of the series of figures is, however, concentrated in obsoleta (fig. 3). Just as the pattern of this species was seen to be intermediate in many details between that of archippus on the one hand and arthemis and weidemeyeri on the other, so is it with the form of its claspers. To make sure that the appearance represented in fig. 3 was not an individual peculiarity, Dr. Eltring- ham made a second preparation, but with precisely the same results. The comparison between figs. 2 and 3 suggests that the mimetic form arose from an ancestral species with claspers more like those of weidemeyert than arthemis. Looking at these figures, some natural- ists may be inclined to suppose that obsoleta sprang from weidemeyeri in the southwest, while archippus developed independently from arthemis in the east and north. Such a conclusion seems to me improbable. It is unlikely that independent lines of evolution would have led to structures with the essential similarity that is to be recognized between the forms shown in figs. 3 and 6—I refer especially to the hook below and the strong teeth above the end of the organ—and still more improbable that such independent eyolu- tion would have led to the resemblances in minute detail that have been shown to exist between the patterns of obsoleta and archippus. Remembering that these conclusions are founded on small differ- ences between organs that are themselves very variable, Dr. Eltring- ham has confirmed his results by making preparations from 3 indi- viduals of archippus, 2 of obsoleta, 2 of weidemeyeri, and 2 of astyanax arizonensis. He finds that the fine points or teeth are not only variable in different individuals, but that they vary on the two sides of the same individual. This he has shown by the careful drawings reproduced on Plate V, where this want of symmetry is apparent in nearly all the figures. The second specimen of weidemeyeri has rather fewer teeth than the one figured. In a single specimen of archippus floridensis (eros) the organs were somewhat larger than in archippus and the clasper points were a little less acute. In spite of great individual variability and the want of symmetry, the claspers of the individuals shown in Plate V exhibit recognizable characters 192 PROCEEDINGS OF THE ACADEMY OF [Jan., common to other individuals of each species examined by Eltringham and, as regards three of them, by Burgess. Knowing my own want of experience in the comparative study of these male abdominal appendages, I submitted Dr. Eltringham’s drawings to my friend Dr. Jordan, who wrote, Aug. 15,1913: “ Archip- pus appears to be a later modification of obsoleta, as yousay. Astyanax arizonensis, weidemeyeri, and arthemis are also closely related to one another.’* Dr. Eltrimgham also agrees that the comparative study of the armatures supports the conclusions arrived at from a study of the patterns. Considering together pattern and the structure of the claspers, there are strong reasons for believing that the mimetic forms arose from a North American Limenitis with the pattern of arthemis and weidemeyeri, but including a white spot in the fore-wing cell upper side now seen most commonly in lorquini among North American species, and with claspers like those of weidemeyeri and arthemis, but probably nearer to the former. : -I trust that Dr. Skinner will consider that this evolutionary history, if not convincing before, has been rendered so by the fresh evidence now produced. 12. SimmuaAR ENVIRONMENTAL CONDITIONS VERSUS MIMICRY AS AN INTERPRETATION OF COLOR RESEMBLANCES. With regard to the resemblance of Limenitis (Basilarchia) flori- densis to Danaida berenice in Florida and of L. (B.) obsoleta (hulsti) to D. strigosa in Arizona, Skinner suggests (32, p. 127) that “similar environmental conditions explain these color resemblances better ‘The remainder of Dr. Karl Jordan’s letter contained an interesting and suggestive criticism of Seudder’s conclusion that proserpina is a hybrid between arthemis and aslyanaz. “The differences in the genitalia between astyanax and arlhemis might render copulation a little difficult, but are too insignificant to prevent it. According to Scudder, proserpina is the hybrid between astyanaz and arthemis. If that is the case, the genitalia should be intermediate. As they are identical (teste Scudder) with those of the northern insect, I do not believe that proserpina is a hybrid. The offspring of a @ proserpina were partly proserpina, partly arthemis. This also points in the direction that astyanax has no part in the production of proserpina. Scudder appears to rely particularly on this point—proserpina inclines towards astyanax where the latter prevails, and towards arthemis in the places where this insect is abundant. But such an agreement in coloration may simply be due to the two occurring side by side. It is not necessarily evidence for hybridization. I have only looked at Scudder’s book, not at the specimens; my opinion is therefore worth very little, but I incline to the belief that arthemis assumes the pattern of astyanaz where it comes into contact with the latter, ze. that proserpina is a southern modification of arthemis, not a hybrid. It would be advisable, however, to examine the genitalia of a series of specimens of all three insects.”’ he ery dint Ah he a A i i a 1914.] NATURAL SCIENCES OF PHILADELPHIA. 193 b than the hypothesis of mimicry.”’ He does not venture to suggest this interpretation for the resemblance of L. (B.) archippus to Danaida plexippus; for the great environmental changes endured by both model and mimic in their extensive north and south range make any such suggestion untenable. With regard to the detailed likeness of three forms of Limenitis to three Danaine butterflies in North America, I may fairly retaliate on my friend and point out in his own words, mutatis mutandis, that “it seems logical to consider that they are governed by a general law rather than that two of them, but not the third, are caused by similar environmental conditions.” I have already many years ago dealt with this supposed interpretation of mimetic resemblance by an appeal to the forces of the environment, and the arguments then brought forward (15) have, so far as I am aware, never been met. Dr. Skinner does not attempt to meet them, nor does he even allude to the peculiarly strong evidence furnished by these very North American mimics against the hypothe- sis of environmental conditions. Although this evidence is clearly set forth in the paper which Dr. Skinner was discussing (22), as well as in earlier publications of mine (16, 21), I will repeat the substance of it on the present occasion. The three Danaines of North America are modern invaders from the Old World, quite isolated and out of place in the New, while the genus Limenitis is an ancestral element in the North American fauna. My own experience of insect systematics is very limited, and I could not with any confidence or authority attempt to weigh the value of characters which have been described as generic. Know- ing these limitations only too well, I applied to my friend Dr. K. Jordan, and he, after making fresh investigations into the male genitalia and carefully studying Moore’s generic characters, came to the conclusion that the Old World Limnas and Salatura and the New World Anosia and Tasitia could not be sustained as separate genera, but that all four were to be properly included in the single genus Danaida. ‘his genus is nearly related to several much-mimicked groups of Danaine in the Old World, but the two species from which the few American geographical forms have been derived are aliens in the New World. Dr. G. B. Longstaff has recently shown that in the gregarious instinct, as manifested by hanging in festoons and clusters from trees, the Old World Danaida genutia (plexippus) resembles its New World representative D. plexippus (27, pp. 75, 76), in which the same habit has often been observed (6, pp. 730, 734-7) 13 194 PROCEEDINGS OF THE ACADEMY OF [Jan., Even in pattern there is but little difference between the most nearly allied Asiatic and American species of Danaida, and if, as Dr. Skinner believes, color and pattern are the expression of environ- mental conditions, then they are the expression of an Old World, and not of a New World environment. On Dr. Skinner’s view, the Old World invader, when it became exposed to the new environment, should have come to resemble the New World resident. Instead of this, the resident has come to resemble the invader. In concluding the present paper I may quote an opinion expressed to me by Professor Svante Arrhenius. A few years ago I asked my friend whether he thought it possible to explain by the incidence of physico-chemical forces, such as those of the environment, the super- ficial resemblance of one form to another when that resemblance required, as in the development of a complex pattern, the co-opera- tion of many different factors. He replied, as I expected, that he did not consider the explanation possible; for the building up of sucha likeness was inconceivable except by the aid of selection. This was the argument I advanced in 1898 (45), after an analysis which showed that mimetic resemblance often requires the co-operation of many different factors; and it was a great satisfaction to find the conclusion con- firmed by an authority with Professor Arrhenius’ broad outlook on the sciences in their relation to one another and to mathematics. BIBLIOGRAPHY. 1. Horsrietp, T. 1857.” In Horsfield and, Moore’s Catalogue of Lepidop- terous Insects, Museum.of the East India Company, London. 2. Wavuace, ALFRED R. 1865. Transactions of the Linnean Society of London, XXV (1866), Pt. I (1865), p. 1. 3. Pourron, Epwarp B. 1885. Proceedings of the Royal Society of London, XXXVIII, pp. 269-315. 4. —— 1886. Proceedings of the Royal Society of London, XL, pp. 135-173. 5 1889. Proceedings of the Entomological Society of London, pp. XXXVII-XXXIX. . 6. ScupperR, SamueL H. 1885. “Butterflies of the Eastern United States and Canada,”’ Cambridge, Mass. 7. Drxey, F. A. 1890. Transactions of the Entomological Society of London, pp. 89-129. if 8. Poutron Epwarp B. 1891. Proceedings of the Entomological Society of London, pp. xv, xvi. 9. Haasp, E. 1893. Untersuchungen tiber die Mimicry. 10. Pouuron, Epwarp B. 1893. Proceedings of the Royal Society of London, LIV, pp. 417-430. 11. Horxrys, F. Gownanp. 1894. Proceedings of the Royal Society of London, LVII, 1894, p. 5. 12. —— 1895. Philosophical Transactions of the Royal Society of London, B., Vol. 186, Pt. II, pp. 661-682. 13. Haase, E. 1896. ‘Researches on Mimicry,” Pt. Il, Stuttgart. English translation. ee ee 1914.] NATURAL SCIENCES OF PHILADELPHIA. 195 14. 15. 16. Poutton, Epwarp B. 1897. Nature, LVII, pp. 1-4, 25, 26. 1898. Journal of the Linnean Society of London, Zoology, X XVI, pp. 558-312. Also published in ‘‘Essays on Evolution,’’ Oxford, 1908, pp. 220-270. 1902. Verhandlungen of the V. International Zoological Congress, Berlin, pp. 171-179. . Houianp, Witiram J. 1903. “The Butterfly Book,’’ New York. 8. Neave,S.A. 1906. Transactions of the Entomological Society of London, pp. 207-222. . Poutton, Epwarp B. 1906. Transactions of the Entomological Society of London, pp. 323-409. . Roruscnitp and JorpAN. 1906. “Novitates Zoologice.’’ Tring., XIII, pp. 411-753. . Pounton, Epwarp B. 1908. Transactions of the Entomological Society of London, pp. 447-488. 1909. Annals of the Entomological Society of America II, pp. 203-242. Also published in ‘‘ Darwin and the Origin,’’ 1909, pp. 144-212. 1909. Mimicry and Sex, in ‘‘ Darwin and Modern Science,’’? Cam- bridge, 1909, pp. 292-295. Also published in ‘Darwin and the Origin,” pp. 1382-139. . Garstane, W. 1910. Nature, LXXXIV, pp. 549, 550. . Bares, G. L. 1911. Ibis, London, pp. 630, 631. . LampBorn, W. A. 1912. Proceedings of the Entomological Society of London, pp. ¢xxxi-cxxxiv. . Lonestarr, G. B. 1912. ‘“Butterfly-hunting in Many Lands,” London. . McAter, W. L. 1912. Proceedings of The Academy of Natural Sciences of Philadelphia, pp. 281-364. . Méurer, Frirz. 1912. Collected papers translated by E. A. Elliott in Appendix to ‘‘Butterfly-hunting in Many Lands,” by G. B. Longstaff, pp. 601-666, plates A, B, C, D, E, F, G, H, J . Neave, 8. A. .1912. Proceedings of the Entomological Society of London, p. lv. . Poutron, Epwarp B. 1912. Proceedings of the Entomological Society of London, pp. xxxi—xxxii. . SkinneR, Henry. 1912. Journal of The Academy of Natural Sciences of Philadelphia, Second Series, XV, pp. 121-127. . SwynnerTon, C.F. M. 1912. Ibis, London, October, pp. 635-640. . Skinner, Henry. 1913. Entomological News, XXIV, pp. 23-27. EXPLANATION OF PLATE V. Genital armatures of male North American Limenitis (Basilarchia). Figures drawn by H. Eltringham. All the figures are magnified about fourteen diameters. Fig. 1—Limenitis lorquini. Fig. 2.—L. weidemeyeri. Fig. 3.—L. obsoleta. Fig. 4.—L. astyanax arizonensis. Fig. 5.—L. arthemis. Fig. 6.—L. archippus. 196 PROCEEDINGS OF THE ACADEMY OF [Feb., FEBRUARY 17. Dr. BENJAMIN SHARP in the chair. Fourteen persons present. The Publication Committee reported the reception of the following contributions to the PROCEEDINGS : “On a collection of mammals from Ecuador,’”’ by Witmer Stone (January 24). “The olfactory sense of Hymenoptera,’’ by N. E. MeIndoo, Ph.D. (January 27). “Description of a Tsantsa in the ethnological collection of the Academy with notes upon another specimen,’’ by H. Newell Wardle (January 30). “Description of a new echinoderm,” by Henry A. Pilsbry (January 31). R. A. F. Penrose, Jr., Amos P. Brown, Frederick Prime, Edgar T. Wherry, and Charles D. Walcott were appointed*to constitute the Committee on the Hayden Memorial Geological Award. Joseph McFarland, M.D., was elected a member. The following were ordered to be printed: 1914.] NATURAL SCIENCES OF PHILADELPHIA. 197 DESCRIPTION OF A TSANTSA IN THE ETHNOLOGICAL COLLECTION OF THE ACADEMY, WITH NOTES ON ANOTHER SPECIMEN. BY H. NEWELL WARDLE. The little mummified human heads, known by their native name of tsantsa, and made by the Jibaro tribes dwelling in the eastern Andean valleys around the head waters of the Amazon, have been known to science for half a century,! yet the specimens are still sufficiently rare for each to merit a full description. The Academy has recently received a fine tsantsa, as a gift from Dr. Thomas Biddle (Plate VI). It was formerly in the possession of the Museum Umlauff of Hamburg, and bears the tag of that insti- tution with the scription, he INO: No. 826 23182 Equador~ A = ae), Jivaros. The head is in fine condition, the flesh being firm and hard, though apparently not brittle. The skin is devoid of wrinkles, despite the excessive shrinkage it has undergone. It shows that peculiar .chocolate-brown tone characteristic of the tsantsa, the portions in relief, such as lips, nostrils, and zygoma, being highly polished and lighter in color, with a distinctly reddish tinge. There are no traces of tattooing observ- able. The measurements follow: m. Eon onbsliy Cir CUMIELEM CO sae eda cesctsateen cassie essed eastern . .240 Transverse supra-auricular curve. pene es LOA. Maximum antero-post. diatneter (inion-glabella)..... Rea 3095: Maximum /OCcipito-taltall Giameters access camera ncianeeencn LILO Maximum) CraNSVerse CIAMetO Reo crtcnaccntmanciacatnickesesemaennie 048 Total length of face................... pee ete eh ee tt.) O80 Maximum bizygomatic diameter... a er ela SNP aE sameees ell The above measurements show that the Academy’s specimen 1 Dr. Moreno-Maiz, Téte d’ Indien jivaro (Pérou oriental) conservée et momifiée par un procedé particulier, avee quelques renseignements sur les Jivaros. Bulle- tins de la Société d’anthropologie de Paris, t. 111, p. 185, 1862. 198 PROCEEDINGS OF THE ACADEMY OF {Feb., possesses the characteristic, narrow, elongated form, with strongly marked prognathism and considerable asymmetry. __ Behind the temples is the usual deep depression, which, especially on the right side, almost amounts to a breaking in of the surface, and gives to the forehead a somewhat conical form. The whole facial region also is narrow and flattened laterally. The cheeks, while not sunken, are depressed below the level of the zygomatic arch and of the muscles of the mouth. The eyebrows are apparently asymmetrical in the arrangement of the hairs—the right supercilium having its inner corner close to the glabella, while that of the left is well round toward the side. The head would seem to have suffered loss at these points during the process of preparation. The hairs are comparatively short— 4 to 5 mm. in length. Considering the great contraction of the piliferous tissue, the growth is not heavy. No eyelashes are visible. The eyelids have been inverted and stitched together, the stitches trom the right eye being continued to close a gash on the right surface of the root of the nose. The line of the crack is traceable across the root of the nose to a corresponding break on the left side which is not sewed. This is found in other tsantsas and doubtless results from the method of preparation which forces the nose forward and upward. The naso-buceal region is strongly projected, which, together with the slightly receding chin, gives to the lower face a snout-like aspect. ¥ : The nostrils, which are rudely circular and distended, are directed almost straight to the front, in a manner suggestive of a double- barreled gun. ‘The orifices show a sufficient supply of vibriss:e. The septum has been broken away in the interior of the nasal passage, which extends straight backward into the cavity of the head. The contour of the nose shows a moderate convexity between root and tip. The lips are proportionately heavy and are pressed forward in a horizontal curve and held together by +hree vertical stitches of vegetable cord. The holes through which these cords pass are circular, as if made with an awl, or kept open during the shrinkage process by round skewers, such as appear on the Murato tsantsa of Colini,? which apparently was in process of preparation when collected. The lip decoration of pendant cords attains a length of 2 Dr. G. A. Colini, Osservazioni etnografiche sui Givari. Real Accademia dei Lincei, anno eelxxx, 1882-1883. Memorie della Classe di scienze morali, storiche, e filologiche, vol. XI, tav. 1, fig. 1. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 199 30 centimeters. It is composed of three two-strand cords in the natural color, which, doubled in the middle, is passed downward through one of the piercings, whereupon its projecting loop is traversed by the free ends after crossing the lips. This is the simplest form of loop knot. The three cords form individual entities at the lips, and, unlike the labial ornaments figured by Dr. Rivet,’ there are no connecting loops with pendant cords. Twenty-one and a_ half centimetres below the lips, one of the ends of the left-hand cord has been broken off. All the remaining strands are gathered together at a point 30 cm. below the lips and, treated as a single strand, tied in a simple knot. Below this they resume their individuality. Both ends of the central cord and the remaining end of that on the left are formed into a kind of uncut tassel by turning back the ends and tying them in a single knot. The other three cord-ends were possibly united by a similar knot, though at present one of the strands is looped back and knotted upon itself. These cords are generally believed to be more than mere decoration, having a mne- monic value, after the manner of the Peruvian quipu (fig. 1). ty, N) “Cords of the lip decoration: Fig. 1. The Academy’s tsantsa; Fig. 2. The Castner tsantsa. The chin is rounded and slightly receding, though the latter appearance is considerably enhanced by the artificial protraction of the lips. Two incisions have been made in it—one on the left side extending almost vertically downward and following the curve from the edge of the lower lip to a point corresponding to the gnathion or slightly posterior to it, the other extending forward from the severed edge of the neck to a point below the right angle of the mouth, where the gash turns upward. These cuts were doubtless made to facilitate the extraction of the lower maxillary bone and were then sewed up with twisted fibre. In the lower seam, the stitches remain in place, but those of the downward incision, with 3Dr. Rivet, Les Indiens Jibaro; etude geographique, historique et ethno- graphique, L’ Anthropologie, t. XIX, p. 79. 200 PROCEEDINGS OF THE ACADEMY OF [Feb., the exception of a single stitch, have evidently been cut away from the surface, leaving two parallel rows of stub ends of fibre visible on the opposite sides of the crack. The ears, though reduced to a length of 3 em., are apparently in nowise distorted. The lobule of each is pierced with a round hole, through which passes the soft cord of two twisted strands, which ties on the pendant ear ornament. ‘These are of unequal length, that of the right ear being only 22 em., while the left one measures 42 em. from its fibre proximal end to the tip of the hair tuft in which it terminates. These cylindrical ear ornaments are formed of a heavy rope of twisted fibre, coarser and more woody than the cords through the ears and lips, and tightly bound with fibre at both ends. To the distal end has been attached, partly by gumming and partly by fibre wrapping, first a heavy tuft of hair, apparently human, 10 cm. long on the longer, and 5 em. on the shorter pendant. Above this, with more gum and more encircling strands, was laid a circle of small yellowish-green feathers, 4 em. long. Above these come the overlapping, iridescent scales which have entirely covered the foundation and consist of the green elytra of a tropic beetle, Mallaspis antennatus, each sewed by a single horizontal stitch of very fine twisted fibre, passing through the two perforations in its proximal end. Occasionally there is but one perforation, but even in that case the stitch is usually horizontal, passing over the opposite edge.* The head has been severed at a point level with the lower line of the chin. The orifice has an oval form, being compressed laterally, in conformity with the whole head. At the section, the thickness of the flesh varies from 3 mm. to 9 mm. By ordinary day light, the hair is of an ebony-black, but in sun light shows considerable iridescence with a marked reddish cast in the shorter hair of the fore part of the head. In character, it is moderately fine, wavy, and rather stiff. Much of it, upon the top and sides of the head, is short; the longest at the back reaches a length of 40 em. Probably much has been removed for the sealp-belt— a trophy the Jibaro esteems only second to the tsantsa. The sealp is seamed from what was, before distorting in desiccation, the posterior median line of the neck to a point midway to the vertex. At the crown of the head, there is a single circular perforation, 4 Dr. Colini (opera citu, tav. LI, figs. 10 and 10 bis) figures ear pendants of this type and ascribes them to the Muratos, a division of the Jibaros. 1914.] NATPRAL SCIENCES OF PHILADELPHIA. 201 through which passes the doubled suspension cord. The free ends of this cord are knotted within the head around a small stick laid in anterior-posterior position beneath the vertex. Some 8 cm. above the point of issuance, the doubled cord is drawn into an incom- plete knot, 7.e., the end is not pulled through, but doubled back upon itself. A centimeter farther and the doubled cord is again knotted. From this point it continues without further interruption, 335 mm., the loop thus formed of 671 mm. being of sufficient size to permit of passage over the head of the former Jibaro owner, when the tsantsa was worn suspended around the neck. This suspension cord is not twisted, but woven, or rather plaited, with a fairly uniform width of 4mm. The cut end within the: head shows ten strands, and the technique is unquestionably that of the five-loop plaiting described by Dr. Roth.° It possesses the attractive arrangement of strands and the flat under-surface, with slightly convex upper face characteristic of this peculiar process, and a series of experiments in ten-strand cord plaiting failed to reproduce it exactly, until the Warrau five-loop plaiting was tried. The result was more than satisfactory, for rather rapid tvork with this method gave all the peculiarities seen in the Jibaro cord—the occasional overlapping of one of the strands of the loop by its mate, thus con- cealing the lower, the consequent thickening and narrowing of the cord with the obscuring of the pattern—points which do not appear in Dr. Roth’s beautifully regular drawing. It is of interest to note the occurrence of this technique—which would seem to be unre- corded elsewhere—in two such widely separated localities as the Amazonian slopes of the Andes in Ecuador and the Pomeroon Dis- trict of British Guiana; employed, in the one case, by the Jibaros, a tribal group of as yet undetermined affinities,® and, in the other, by the Warraus, whose relationships also remain to be fixed; and the question arises as to whether this five-loop plait is made also by the people of the far-flung Carib stock. While engaged in the study of the Academy’s tsantsa, another of these little mummified heads came under the writer’s notice, and it was deemed advisable to include a brief description of it (Plate 5 Dr. Walter E. Roth, Some Technological Notes from the Pomeroon District, British Guiana. Journal of the R. A nilapolosia Institute of Great Britain and Treland, vol. XL, p. 27, Plate VI, figs Dr: ’Rivet, Journ. citu, t. XVII, p. 338, footnote, promises a detailed study of the language of the Jibaros, based on vocabularies in kis possession. Pre- viously available evidence of its affinities was not sufficient to permit of assigning the tribal group to any stock. 202 PROCEEDINGS OF THE ACADEMY OF [Feb., VII, a and b). This trophy is owned by Mr. Samuel Castner, Jr., of Philadelphia, and was obtained by him in 1903 at a sale of the collection of Arthur H. Little, where it was wrongly ascribed to Oceanica. It is a typical Jibaro tsantsa, in excellent condition. The skin is: of the same chocolate-brown color, but without the lighter tones which distinguish the promimences in the one already described. It is smooth and shows no traces of tattooing. The measurements follow: m. Horizontal circumference... te Heer Enh Oe ee, .260 TANS Verse SUpra-aUcieUlaL CULE) ses none nce nnen ocr ee 160 Maximum antero-post. diameter (inion-glabella)... ee BOSE Maximum occipito-labial diameter... cccccsccscsesseeessseeseneesseeseseeeceern . .105 Maximum ‘transverse diameter ticcat.cc.c.ntcton coined eee OGD Total length of face... ee er mee re EU Maximum bi-zy gomatic * Giadneben «thors, <5 eee ee Comparing the two specimens, it is evident that the gain in both the horizontal circumference and in the transverse supra-auricular curve, of the Castner piece, is due to the greater width of the head, both the longitudinal diameters being actually shorter. The face also is broader and shorter, with very slight prognathism. The head is not noticeably asymmetrical and possesses the char- acteristic deep depressions behind the temples. The cheeks are rounded out, so that the muscles of the mouth and the position of the zygomatic arch are not brought into relief. The eyebrows are symmetrical and fairly heavy, with individual hairs reaching a length of 10 mm. No eyelashes are visible, the eyelids being inverted but not stitched. There is no break across the root of the nose, but a deep crease, due to the protraction of the bucco-nasal region. The nasal contour is marked by a moderate curve from tip to root. The nostrils are directed horizontally forward and show numerous vibrisse. Neither septum nor allze have been perforated. The lips, which are proportionately heavy, show three vertical piercings, each occupied by a short twisted cord which traverses both lips and is knotted in front, leaving short pendant ends. To these cords, just above the point of issuance from the perforation in the lower lip, a horizontal cord is attached, which, in turn, bears the characteristic lip decoration of long, pendant cords—twenty-two in number and separated into two groups by the knotting of the horizontal sustainer around the central vertical cord. The long twisted cords of both groups fall straight to a length of m. .367 (fig. 2), 1914.| NATURAL SCIENCES OF PHILADELPHIA. 203 The chin is rounded, but not receding; it falls into a vertical line with the lower forehead. No incisions are visible, only a deep furrow from the neck to behind the position of the maxilla on the left side. In this, as in other details, the Castner tsantsa shows the work of a more skilful preparator than was the Jibaro from whose hand the Academy’s specimen came. The ears are considerably distorted so that an.accurate measure- ment is not possible. Both have been pierced through the lobule, though the right one alone bears an ornament—a section >f bird (?) bone, hung by a doubled cord of twisted fibre drawn, through it, the knotted loop end being pulled back within the hollow bone. The head has been severed by a diagonal cut, which passed close to the head on the right, but left a portion of the neck on the left side. The flesh at the section varies from 4 to 2 mm. in thickness. The hair is of a beautiful ebony-black, fine and wavy, and reaches a length of 56 em. The seaming of the scalp extends from the neck in the posterior median line almost to the vertex, where the single perforation occurs through which the suspension cord is passed. This latter has a length of only 50 cm. from the point of issuance from the perforation to its re-entry therein. The width of the cord is 3 mm., and the technique is evidently the same five-loop plaiting noted in the suspension cord of the Academy’s tsantsa. Comparison with the table of measurements of the eleven mum- mied heads studied by Dr. Rivet,’ shows that in both the tsantsas here described the horizontal circumference, the transverse supra- auricular curve, and the maximum antero-posterior diameter rise above the average, though not reaching the maximum measurements. In maximum transverse diameter, the Academy’s piece falls within 3 mm. of the minimum, while the Castner head is above the average; in total height of face, the Academy’s approaches the maximum, with a bi-zygomatic diameter below the minimum, while the height of the Castner specimen is below the minimum and its width at the zygoma somewhat below the mean. Reference should be made to the preparation of these trophy heads, because of its relation to the structure and condition of the fished product. Three methods have been described by reliable travellers, and it is probable that all are, in the main, correct, the differences being due to local variation of practice among the Jibaro tribes. 7 Dr. Rivet, Journ. cilu, t. XIX, p. 76. 204 PROCEEDINGS OF THE ACADEMY OF [Feb., Dr. Rivet,’ following Lubbock and other writers, gives the fol- lowing procedure. After the extraction of the cranium through the posterior incision, the skin with adherent flesh is boiled in an herb decoction. Withdrawn from this, it is placed around a spherical “stone, superheated, and, after shrmkage, upon a smaller stone, and then upon a third yet smaller. Meanwhile, another hot stone is passed back and forth over the surface, thus facilitating the shrinking and drying of the tissue. The lips, and sometimes the eyelids also, had previously been carefully sewed to prevent the retraction in desiccation, causing them to gape. According to the engineer Von Hassel,’ after the substitution of the hot stone for the cranium, the head is hung in the smoke of a palm-root fire, but there is no mention of boiling. The lips are ‘‘deformed—by means of a cord and a little piece of chonta’’ (wood). The third description of the method pursued, which was given Lieut. Safford by Sefor Tirado”—an eye-witness—is an interesting blending of the two preceding. According to this statement, imme- diately after the extraction of the skull, the scalp is sewed up, and the hole in the vertex pierced and supplied with its cord. Afterwards the head is dipped in the hot infusion of herbs, “care being taken not to allow the roots of the hair to enter,’’ though how this latter precaution is possible is not readily conceivable. Dried by the introduction of hot stones, it is then smoked over the cooking-fire, the hair being wrapped in leaves for protection. After three or four months of curing in the smoke, the lips are pierced and the decorations added. None of these descriptions makes mention of any lashing or means of holding the cranial envelope in position during the curing process. Yet the Murato tsantsa of Colini," which is evidently a head obtained before the finishing touches had been added, shows a slender spike of wood passed backward through the nostrils and out through the perforation at the vertex. A cord is lashed around the ends of this stick and over the forehead, thus forcing the nostrils forward and 8 Dr. Rivet, Journ. citu, t. XIX, p. 71; alsoSir John Lubbock, Note on the Macas Indians. Journalof the Anthropological Institute of Great Britain and Ireland, vol. III, p. 30. Sir John, however, states that the bones were removed through the neck a/ter the boiling. ® Jorge M. von Hassel, Las Tribus salvajes de la regién amazénica del Pert. Boletin de la Sociedad Geogrdfica de Lima, XVII, 1905, pp. 56-57. 1 Dr. Walter Hough, Prepared Human Head. American Anthrapologist, vol. XIV, p. 406. E Dr. G. A. Colini, opera citu, p. 362 et seq., tav. 1, fig. 1; also Dr. Rivet, Journ. cilu, t. XIX, p. 82, Pl. 1, fig. 3. 1914.| NATURAL SCIENCES OF PHILADELPHIA. 205 upward, and causing the deep bend at the root of the nose, which is characteristic of all genuine tsantsas, and, in the Academy’s example, has resulted in an actual breaking of the integument. The three piercings of the lips are also occupied by skewers lashed in place; another stick is thrust into the auditory meatus and the néck is secured to a wooden ring by stitching. The lips, then, in this unfinished trophy, are already pierced and — held together by skewers, which fix them rigidly in the desired position. Together with the spike through nostril and vertex, they form an essential part of the taxidermy and account for the char- acteristic protraction of the lips. Certainly in both specimens herein described, the perforations of the lips, like those of the lobules, were accomplished while the flesh was comparatively soft. On the other hand, the stick thrust into the auditory meatus would seem not to have been a constant feature of the preparation, as the small hole made by it is found only occasionally. Some sort of device must have been in use for keeping open during desiccation the flaccid skin of the neck, and the wooden ciccle of Colini’s tsantsa would have admirably served this purpose. The Academy’s specimen shows at the section of the neck certain fine grooves, running from the outer to the inner surface of the flesh, as if it had been traversed by stitches and had later been cut or broken away along the plane of the piercings, leaving the grooves exposed. It seems to be certain that, whether the preparation occupied one day, one week, or several months, whether it was by means of alter- nate steaming and drying or by smoking, or by a combination of both methods—as seems probable—the seaming of the scalp, the piercing of the vertex and of the lips, and the insertion of the wand from nostril to vertex must have been done while the flesh had not yet hardened; but, apparently, when the last of these operations took place, the trophy was already reduced practically to its final dimensions and, in the case of the Academy’s tsantsa, had lost some of its flexibility. EXPLANATION OF Puatess VI, VII. Prater VI.—Tsantsa, mummified human head, prepared by the Jibaro Indians of Ecuador. A. N.S. P., No. 15,048. Pirate VII.—Tsantsa: a, front view; 6, lateral view. Owned by Mr. Samuel Castner, Jr. 206 PROCEEDINGS OF THE ACADEMY OF [Feb., DESCRIPTION OF A NEW ECHINODERM. BY HENRY A. PILSBRY. Several years ago the Academy received from Mr. Clarence Bloomfield Moore specimens of a fossil sea-urchin, which had been obtained from material dredged from near the mouth of the Withla- coochee River, Florida, in an area mapped as Pliocene. The speci- mens represent an undescribed species, which may be called— Eupatagus mooreanus n.sp. Plate VIII. The outline is broadly ovate, slightly emarginate in front, tapering a little, and abruptly, vertically truncate behind. The upper surface is moderately convex, highest at the posterior third of the length; lower surface nearly flat, slightly concave at the sides and anteriorly close to the peristome, The ambulacral centre is at the anterior six-tenths of the length. Unpaired ambulacral area indistinct, without large pores. Paired ambulacral areas petaloid, nearly straight-sided, closed at the distal ends. Those of the posterior pair are a little longer than the anterior. They form an acute angle with one another and right angles with the areas of the anterior pair.~ The poriferous zones are depressed, with equal, circular pores in pairs connected by furrows. The four genital pores are equal, round, the anterior pair much closer than the posterior pair. The peripetalous fasciole is distinct, not sinuous. The tubercles in the interambulacral areas are very unequal. The larger ones stand in about five waved concentric series and occupy slight depressions. The posterior interambulacral area is convex and not distinctly tuberculate, having a quite distinct median suture. The peristome is in form of a transverse oval, flattened posteriorly, It is situated at the anterior third of the length. The periproct opens at the summit of the flattened posterior area, is shortly oval, somewhat higher than wide, and a little angular at the ends. The base is closely and strongly tuberculate, the tubercles becoming much smaller at the periphery. There is a smooth raised segment from the peristome to the posterior end. Length 61, transverse diameter 51, alt. 29 mm. This species is about the size of E£. clevei Cotteau, from which it 1914.] NATURAL SCIENCES OF PHILADELPHIA. 207 differs by the narrower ambulacral petals, the vertically truncate posterior end, wider peristome, equal genital pores, and the different arrangement of the tubercles. The largest specimen is 64 mm. long. The type is No. 1147 ANS: P3 EXPLANATION OF PuiatEe VIII. Upper, lateral and basal views of Eupatagus mooreanus Pils. 208 PROCEEDINGS OF THE ACADEMY OF [Mar., Marcu 17. Mr. Cuarutes Morris in the Chair. Eleven persons present. The Publication Committee reported the reception of the following papers as contributions to the PrRocEEDINGs: “Fishes collected by the Princeton Expedition to North Green- land in 1899,” by Henry W. Fowler (February 14). “Description of a new Blenny from New Jersey, with notes on other fishes from the Middle Atlantic States,’”” by Henry W. Fowler (February 24.) “Presh-water mollusks of the Oligocene of Antigua,’’ by Amos B, Brown and Henry A. -Pilsbry (March 9). The deaths of the following members were announced: Thomas Wistar, M.D., September 27, 1913. N. Roe Bradner, M.D., February 6, 1914. Charles 5. Welles, February 24, 1914. Walter Rogers Furness, February 27, 1914. Edwin J. Houston, March 1, 1914. Stuart Wood, March 2, 1914. The following were 6rdered to be printed: 1914.] NATURAL SCIENCES OF PHILADELPHIA. 209 FRESH-WATER MOLLUSKS OF THE OLIGOCENE OF ANTIGUA. BY AMOS P. BROWN AND HENRY A. PILSBRY. In a recent paper by one of us! reference has been made to the deposits carrying these fresh-water shells, which were first noted by Nugent.? His collection of Antigua fossils, including, no doubt, speci- mens of these fresh-water mollusks, is still preserved in the collections of the Geological Society of London, now in the British Museum. These deposits were later described by Purves* as his division E, under the name of the “Lacustrine or fresh-water chert.”’ These beds are mapped by Purves as extending completely across the island, in the central plain from Corbizon Point and Dry Hill in the northwest to near Willoughby Bay and Falmouth Harbor in the southeast. His observations on the fossils appear to have been made at Dry Hill and at Corbizon Point. M. Purves records the following genera as occurring in these cherts: Melania, Zonites, Nematura or Amnicola, Planorbis, Melampus, Neritina, Truncatélla, Pomatias. He also states that the specific descriptions of these shells will be published later, but this seems never to have been done. The species described in this paper were collected from the sea cliffs at Dry Hill, where these flinty beds, carrying fresh-water species, outcrop on the seashore and where they have weathered out by the action of the rains and the salt water dissolving away the calcareous material and leaving the silicified shells intact in a remarkably good state of preservation. When these beds were seen inland at several points, the weathered surfaces of the layers exposing the shells were not so well preserved as at Dry Hill or at Corbizon Point, only sections being found in most cases, as the shells were imbedded in the compact flint. This was, of course, the case at the two localities above noted, likewise; the hard, compact flint layers, varying in thickness from one to four inches, being frequently crowded with these fresh-water shells that showed only in sections oe the frac- 1 Brown, Notes on fine (ee of me Taio of SRE a oc. A. N.S. P., 1913, p. 584-616. See also p. 596 of the same paper. 2 Nugent, A Sketch of the Geology of the Island of Antigua, Trans. Geol. Soc. London, ser. 1, Vol. V, (1821), pp. 459-475. 3 Esquisse géologique de lle d’Antigoa, Bull. Mus. Roy. Hist. Nat. Belg., Vol. II, 1884-85, pp. 273-318. 14 210 PROCEEDINGS OF THE ACADEMY OF [Mar., tured surfaces. Where the sea had been eating into the cliff, and detaching slabs of the hard, flinty layers, the surfaces of these flint slabs were completely covered with the flint pseudomorphs of the shells, these in most cases having the finest sculpture perfectly preserved The relation of these beds to the other members of Brown's division 3 as given in his Notes on the Geology of the Island of Antigua is shown in the section of the rocks at Dry Hill, on page 595,4 and there it will be seen that the lacustrine or fresh-water chert layers immediately overlie the Volcanic Sands which form the lower 23+ feet of the section. The same sequence of deposits is also seen at Corbizon Point, where the fresh-water chert layers with silicified wood occur along the shore immediately above these same Volcanic Sands that are found at Dry Hill. Near the Botanic Station, just east of St. John’s, the flinty layers with fresh-water shells are found, but the shells are imbedded in the flint and only show in sections in the hard, compact rock. The reddish beds of the Volcanic Sands are absent at this locality east of the Botanic Station or are only represented by sandy nodules in the white tuffs. No silicified wood was seen at this place. The species described below belong to the genera Hemisinus, Bythinella, and Planorbis. The genus Hemisinus is undoubtedly what Purves has called Melania and indeed he mentions Hemisinus as being found living in Cuba. These species of Hemisinus are described below. The Bythinella is probably what Purves referred to Nematura or Amnicola; we-have described one species. The one Planorbis which we. describe is the only representative of this genus in the collection. We found no representatives of Melampus, Trun- catella, Pomatias, Zonites, and Neritina, mentioned by Purves. Hemisinus antiguensis n. sp. Pl. IX, figs. 1, 3, 5, 6. The shell is slender, diameter contained nearly three times in the length; whorls rather numerous, probably at least fifteen in a perfect shell, as a young one 12.5 mm. long has twelve whorls, the upper part of the spire being very slender. Whoris convex, sculptured with many rounded ribs, as wide as their intervals, somewhat curved, the concavity forward an]j somewhat protractive. There are about 25 ribs on awhorl. Above the lower suture of each whorl there are two or three spiral cords, the lower.one strongest. On the last whorl the ribs extend to the periphery where they disappear, the peripheral region anl the base having numerous spiral cords. The aperture ‘ Brown, loc. cit. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 211 is but rarely preserved, but in the best examples the peristome seems to be somewhat effuse at the base of the columella. Length 16.5, diam. 6 mm., 8 whorls remaining. “ec 19.5, “ee 6.4 “ 8 “ec “ The sculpture of longitudinal ribs with basal spirals is charac- teristic. The same type of sculpture occurs in various South Ameri- can species of Hemisinus. It could readily be matched also in Melania and related forms or in the Pleuroceratide. There seems to be variation in the development of the spirals. Many specimens show weak traces of impressed spirals over the ribs throughout, and this seems to be the normal condition; but in some examples the ribs appear to be smooth except near their lower ends. ; This species, like the associated forms, has the basal sinus or notch obsolete, as in part of the recent species. Hemisinus siliceus n.sp. Pl. IX, fig. 2. The shell is Melaniiform, regularly tapering, the diameter of last whorl contained about 2! times in the total length. The whorls are convex, and apparently without any sculpture except growth- lines. The last whorl has fine, reversed sigmoid growth striz, which retract somewhat below the suture, then advance, as in H. cubaniana. In the type specimen a former peristome, indicating a period of growth arrest, appears as a sigmoid varix on the last whorl. This indicates a more strongly sigmoid outer lip than in the recent Antil- lean species. Length 26 mm., about 6 whorls remaining, the summit lost; diam. 10.8 mm. No entirely perfect aperture was found on the slabs collected, but so far as we can judge, it seems to be much like that of Hemisinus cubanianus (Orb.). It is not unlikely that H. siliceus is ancestral, or at least a collateral species not far removed from the ancestral stock of the smooth Antillean species of Hemisinus. There is, of course, a possibility that this Antigua species belongs to the genus Pachycheilus, which is represented in the recent fauna of Cuba by P. conicus (Orb.) and P. violaceus Prest.; but the straighter columella does not, in our opinion, favor this view. The type has lost the shell from the spire by conversion into flint, but the surface has been preserved in perfection on the last whorl. Hemisinus latus n. sp. PI. IX, fig. 4. This form is represented by somewhat numerous internal casts, of which the largest has been selected for illustration. It differs 212 PROCEEDINGS OF THE ACADEMY OF [Mar., from the associated species by its decidedly broader figure. The diameter of last whorl is contained about 2% times in the estimated total length. The whorls, of which somewhat over 5 are preserved in the type, are shorter and broader than in H. silicews. Very little of the shell is preserved and the sculpture is unknown. It is proba- bly smooth. The aperture is largely concealed by another shell, the thin, arcuate, outer lip alone remaining visible. Length of broken specimen 17 mm.; diameter 8.5 mm. Bythinella antiguensis n.sp. Text fig. 1. The shell is oblong, pupiform, smooth; outlines of the spire convex, the apex conspicuously obtuse. Whorls 4, very convex, aperture vertical, shortly ovate, its length contained 24 times in that of the shell; peristome in one plane, thin. Length 1.8, diam. 1.1, length of aperture 0.7 mm. | This very minute form is not rare. It has the very a obtuse summit and the pupiform shape of the species usually referred to Bythinella, rather than the shape ( ce of Paludestrina, if, indeed, the two groups are distinct. Of course, any generic reference of a minute fossil Amnicoloid shell of this sort is purely provisional, unless it is from a region where the recent fauna and its antecedents are well known. Planorbis siliceus n.sp. Pl. IX, figs. la, 3a, 5a, 6a. This is a species of the section Tropicorbis.2 The shell is rather thick, with the periphery rounded, more convex towards the right side. The two sides are about equal in width of the concavity, but that on the right side penetrates more deeply, being vortex shaped. The last whorl is rounded on this side. On the left side the last whorl is subangular and the cavity is less infundibuliform. Diameter 3.5, greatest alt. 1.7 mm. és 5 rf cae: Monee This species belongs to a widely spread group of the modern tropical American fauna. PALEOGEOGRAPHIC RESULTS. The species of Hemisinus, as of related genera, are river snails. They do not inhabit intermittent streams, ponds, or lakes, except as the latter may form part of a stream system; and they are equally unknown in brackish water. The occurrence of several well-differen- tiated species therefore implies the presence of rivers or permanent « 8 Tropicorbis n. section, type P. liebmanni Dkr. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 213 streams. Melanians are not likely to be distributed by adventitious means from stream to stream, as Physa, Anodonta, and some other fresh-water mollusks are. Their egg capsules are not gelatinous or likely to adhere to the feet of water fowl, but are firmly fixed to stones, shells, or the like. Wherever their distribution has been studied in detail, it has been found remarkably consistent and explicable by actual stream connections or such as have probably existed within the life of the species or group in question. As the relationships of the Antiguan species are with South America (H. antiguensis) and with Cuba (H. siliceus), it seems probable that (1) at the stage of the Oligocene when this bed was formed, a consider- able land surface existed in the Antiguan area, and (2) that this area was, or had been, connected with the South American main. It seems likely that the present Hemisinus species of Cuba and Jamaica are descendants of the same South American stock. There is in Cuba, however, another totally distinct genus of Melanians, Pachycheilus, represented by P. nigrata (Poey) and P. violaceus Prest. of the recent fauna, which are apparently traceable to a Central American connection. The other fresh-water snails of the silex bed are not significant. The Planorbis belongs to a group widely spread in the recent Neotropi- eal fauna and the ‘‘ Bythinella”’ is an ambiguous form of unknown relationships.® EXPLANATION OF PLATE IX. Figures 1, 3, 5, 6—Slabs strewn with Hemisinus antiguensisn.sp. At the posi- tions marked a are seen specimens of Planorbis siliceus n. sp., the type being 6 a. Fig. 2.—Hemisinus siliceus n. sp. Fig. 4.—Hemisinus latus n. sp. 6 The generic characters of the fresh-water Rissoids are cften not expressed in the shell. A fossil form of simple structure cannot be located generically with any degree of certainty, unless in a region where the characteristics and affinities of the associated fauna have been thoroughly worked out. 214 PROCEEDINGS OF THE ACADEMY OF [Mar., TWO NEW SARCOSPORIDIA. BY HOWARD CRAWLEY.” Sarcocystis leporum sp. n. ! The material on which the present study is based consists of the arm and shoulder of a very old male rabbit shot at Bowie, Md., on December 13, 1913. The presence of Sarcosporidia in rabbits has been recorded from time to time in the literature, and there are four. specimens of this parasite in the collection of the Zoological Division of the Bureau of Animal Industry, the localities of which were Maryland, Pennsylvania, New York, and Illinois. No descrip- tion of the organism has ever been published, nor has it received a name. With regard to this latter point certain authors have of late assumed that the Sarcosporidia, like other parasites, are in the case of each species capable of dwelling in any one of several hosts, and there is a certain amount of experimental evidence that this is true. Nevertheless, in the event of a duplication of names, it is very easy to relegate one of them to synonymy, whereas it is extremely awkward . to be obliged to refer to a parasite as the sarcosporidian found by a certain author in a certain animal from a certain locality on a certain date. Hence it seems best to make a new species, and I propose to call this parasite Sarcocystis leporum: In the fresh tissue the parasitic cysts were visible as short, delicate threads or rods lying in the muscles. They were about two milli- meters long, and the diameter measured in paraffin sections was from 200 to 250 microns. It may incidentally be mentioned, how- ever, that the size of a sarcosporidian cyst is of no diagnostic value, since it is wholly a function of the age, and the cysts of this particular specimen were probably much under the possible maximum size. Compared with the infestations seen in rats, mice, and ducks, that of the rabbit here under consideration was very slight, and a casual glance at the flesh would probably not have revealed anything amiss. Data as to how heavily rabbits may become infested are, however, wholly lacking. The eysts, in paraffin sections, showed nothing noteworthy. The cyst wall was from 5 to 6 microns thick, and seen under low powers presented the typical striated appearance. Under high magnifica- 1914.] NATURAL SCIENCES OF PHILADELPHIA. 215 tion, however (500 to 1000), it was easy to see that the wall was composed of a great number of papilliform processes, standing closely packed together upon a sort of basement membrane and with their outer ends wholly free. That is, the cyst wall, at least in this case, does not consist of a membrane pierced by pores, nor of a congeries of rods bounded both without and within by a membrane, but of rods or papille projecting freely from a basement membrane. Furthermore, in this case, there was nothing to show that any part of the cyst wall was derived from the surrounding host tissue. As already stated, the papille rested upon a basement membrane with which they were apparently contmuous. Within there was to be seen the structure usual for sarcosporidian cysts; that is, a division of the central space into compartments, the walls of which were a continuation of the membrane inclosing the cyst. Finally, in the central portions of the cysts there was a small area free of spores, and here the coarsely alveolar structure of the frame work could readily be seen. This is in itself an indication that these cysts were young rather than old, since it is a matter of common obser- vation that in old cysts there is always present a central space of considerable extent in which there are no spores. As is usual, the cysts were closely packed with spores which showed a certain disposition to be arranged in files, meCION from the centre to the periphery. A study of the spores themselves revealed data of considerably greater interest. Several mounts were prepared by smearing out the contents of the cysts on slides, drying, fixing in absolute alcohol and staining in Giemsa. As thus prepared, the spores measured about 13v long by 5 wide, the precise figures for the average of 20 specimens measured being 13.14” long by 5.16% broad. The longest spore measured was 16 long, the broadest 6” wide. The figure given for the length, however, is a trifle too small, smce the measurement was taken in a straight line, no allowance being made for the curvature. Furthermore, it is not unlikely that the fixed spores are too wide, since they are quite large enough to be flattened in the process of fixation. The spores, although possessing the typical banana shape, are not quite symmetrical, it being generally possible to distinguish between a narrower, more pointed and a broader, more rounded end. This narrower end, which may be regarded as anterior, is occupied by a very solid mass of homogeneous cytoplasm, which has but little affinity for the stain and contrasts very sharply with the deeply 216 PROCEEDINGS OF THE ACADEMY OF [Mar., staining cytoplasm of the remainder of the spore. Thus, when viewed with powers of only 200 to 300, the spores show two very clear-cut oval areas, the nucleus in the posterior half and the differ- entiated area in front, and this latter is sometimes so faintly stained that the complete outline of the spore cannot be followed. The clear region is sometimes oval, sometimes truncated behind, as shown in the figure. It may be spoken of as the rostrum of the spore. Behind it, the cytoplasm abruptly assumes the character which it possesses in the remaining portion of the spore where it is densely staining and conspicuously alveolar. It. is to be noted, how- Spores of Sarcocystis leporum. X 3500. ever, that the cytoplasm nearest the rostrum shows the coarsest alveoli, while backward the alveoli become smaller and smaller, so that in many cases the cytoplasm in the posterior half of the spore becomes very dense, on account of the excessive minuteness of the alveoli. In other cases, however, the alveoli are distinctly visible throughout the entire extent of the dense spongioplasm of the cell. At times, also, the spongioplasm encroaches somewhat upon the homogeneous cytoplasm of the rostrum, there being here visible one or more alveoli or one or more strands of spongioplasm. There is, finally, often to be seen one or two clearer regions in the cytoplasm between the rostrum and nucleus, but it is not believed that these represent morphological entities. Following the rule for the spores of Sarcosporidia, there is no mem- brane, the spores being naked masses of protoplasm. The nucleus may occupy nearly any position in the cell, but it is usually placed near the posterior end. Although, as already stated, 1914,] NATURAL SCIENCES OF PHILADELPHIA. 217 it stands out very clearly when the spores are viewed with rather low powers, it seems for the most part to be no more than a clear space in the cytoplasm, and it is only occasionally that a definitive nuclear membrane can be demonstrated. Within it is normally provided with a number of chromatin granules. These granules vary a good deal in size. At times they are quite minute and occur in clusters and-chains. More frequently, however, they are quite large, round, or elongated, and appear to be wholly free in the nuclear sap. At times, also, there is a more or less typical chro natin net. Whereas these differences may have some significance, it is perhaps best to regard them merely as variants of some fundamental plan. The spores of certain species of the Sarcosporidia which attack mammals are described and figured as being liberally provided with rather densely staining granules, concerning which there has been a good deal of theoretical discussion. Frequently, also, such spores have been described as showing a differentiated area at one end, and attempts have been made to correlate this area with the polar cap- sules of the spores of the Myzxosporidia. In the case in hand, the spores of Sarcocystis leporum, there is at least no question about the structure, which is remarkably clear cut and perfectly obvious. And the rostrum of this spore, whatever may be its homologies, is clearly the analogue of the rostra of the sporozoites and merozoites of Coccidia. Moreover, it seems an entirely safe assumption that its function is to enable the spore to drill its way into the intes- tinal epithelium of its host, without prejudice as to whether this host be an invertebrate, another rabbit, or some carnivorous mammal or bird which preys upon the rabbit. Sarcocystis setophage sp. n. Stiles (1895¢)! notes that Dr. Hassall, of the Bureau of Animal Industry, discovered a sarcosporidian in the muscles of a redstart (Setophaga ruticilla), and a description of this parasite was promised at the time. This description, however, was never published, and the material, consisting of two cysts embedded in paraffin, was recently given to me by Dr. Hassall. Based upon the number of sections yielded by the cysts, their length was about 2.5 mm., while the largest cross sections measured about 1 mm. in diameter. Hence the cysts are thick in proportion to their length, which seems characteristic for Sarcosporidia of birds. The cysts were divided into the usual compartments. 1 Stiles. 1895 e—New American finds of Sarcosporidia (Notes on parasites. 28.) Vet. Mag., Phila., v. 1 (11) (Nov., 1894), Jan. 17, pp. 728, 729. 218 PROCEEDINGS OF THE ACADEMY OF [Mar., Unfortunately, the material was not in good condition for cytological study, and very little could be made out as to the structure of the spores. As well, however, as could be determined, the form was that shown by the spore of Sarcocystis rileyi, that is, one end was rounded, the other pointed. The spores were sometimes straight, sometimes curved, but in the latter case the curvature was slight, which also seems characteristic for avian Sarcosporidia. Within, in a few of the better preserved specimens, it was possible to make out a vacuole in which was a chromatin granule. This structure is probably to be interpreted as a vesicular nucleus. The spores were small, measuring from 4 to 5 microns long by .75 to 1.00 micron broad. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 219 A NEW DIATOM. BY CHARLES S. BOYER, A.M. Chetoceros elmorei n. sp. Plate X. Prof. C. J. Elmore, of Grand Island College, Nebraska, sent me recently a slide containing a form of Chetoceros. On first examina- tion, the species appeared to resemble C. wighamii Br. On receipt, however, of material which was subjected to a closer examination, I have concluded that the species is new. The material had been dried upon blotting paper and it was necessary to soak it for a long while to secure filaments of.the proper size, it being impossible, of course, to resort to the proper methods of cleaning. The character of the chromatophores is not known. I have, however, succeeded in mounting a number of slides which show quite clearly the structure of the valves and spores. All species of the genus Chetoceros have heretofore been considered as marine, and have not been found inland except in the Caspian Sea. The species about to be described is found in Devils Lake, North Dakota. It may be of interest. to give a brief description of the locality from the Second and the Sixth Biennial Reports of the State Geological Survey of North Dakota, of the years 1903 and 1912, respectively. In the Report for 1903, from an article by Mr. E. J. Babcock, Water Resources of the Devils Lake Region (p. 208), and also from an article by Mr. Howard E. Simpson, The Physiography of the Devils-Stump Lake Region, in the Report for 1912 (p. 105), the following information is obtained: Devils Lake is in Lat. 48° N., Long. 99° W. It is a glacial lake and “occupies a basin formed largely by morainie ridges.” Its length is “about twenty-four miles, and the width averages, perhaps, between four and seven miles.” It lies at an elevation of about fourteen hundred feet, and its greatest depth is not more than twenty-nine feet. No streams of any size enter the lake, its chief source of supply being the annual rainfall from the surrounding ridges, and it has no outlet. Although originally -a large fresh- water lake, it is now much reduced in size, and its waters ‘“may be termed alkaline and brackish, since they show a salinity of about 220 -PROCEEDINGS OF THE ACADEMY OF [Mar., one per cent., of which magnesium and sodium salts constitute a considerable portion.”’ The following is the diagnosis of the form: Filaments straight, 234 wide. Cells rectangular with sharp angles; valves slightly convex; foramina narrowly linear, irregular, bipartite. Sete straight, hollow, approaching each other at an acute angle and crossing at a right angle near the corners of the valves, about ten times the length of the valve. In valve view they diverge from each other at an angle of about 80°. Terminal sete shorter than the others, somewhat curved in the direction of the filament. Spores with the primary valve arcuat2, secondary valve produced into a subconical frustum. Rarely in free spores the primary valve is covered with minute spines. The valves are joined together near one side by a tubular com- missure, from six-tenths to eight-tenths of a micron in thickness, situated near the edge of the valve, usually at unequal distances from each end, and, so far as noticed, is found near the margin on the same side of all valves in the filament. The presence of the connecting tube and the unusual locality combine to render this form a unique species, which I take pleasure in naming after Prof. Elmore. An examination of fig. II, 7 and k, in Plate III, in Schiitt’s article - (Ueber die Diatomeengattung Chetoceros, Bot. Zeitung, 1888) offers an interesting suggestion as to the origin of the commissure. In the genus Thalassiosira the cells are connected by mucilaginous threads which are central. In the present species of Chetoceros the commissure is eccentric and. appears to be tubular. In C. simile the valves touch each other, and in several species, such as compactum, the centre of the valve is produced or considerably elevated, but in no previously described form has the union of the frustules been consummated, except by the interlacing of the awns. The other diatoms sparingly found in the waters of the lake include species of Fragilaria, Gomphonema, Epithemia, and Surirella. Owing to the presence of great quantities of small crustacea, the material is mounted with difficulty. F EXPLANATION OF PLATE X. Fig. 1—A short filament consisting of seven cells. The filaments usually contain twenty or more cells. In filaments showing the spores, the width is quite constant, but in vegetative cells the width varies considerably, many of them being much wider. 1914,] NATURAL SCIENCES OF PHILADELPHIA. 221 Fig. Fig. Fig. Fig. Fig. Fig. 2 shows the accidental separation of two adjacent cells and the division of the commissure. ¥ 3 is a diagrammatic representation of the position of the commissure near one side of the cell. The valves are frequently in close contact in the middle, making the foramen bipartite. 4.—Represents the usual position of the commissure, although the dis- tance from the edge of the valve is somewhat variable. Sometimes the com- missure is in the middle of the side, but more frequently nearer one end. 5 is a valve view of the secondary valve as seen in the cell. 6 shows a form rarely found of a free spore much more developed than the others and having the surface of the primary valve partly covered with spines, somewhat as in C. wighamii Br. Specimens of the latter in my collection, however, show spores which are smaller, more circular, and with the surface more evenly covered with spines. 7 represents two adjacent cells, as frequently seen, containing spores with their secondary valves opposed. The figures represent a magnification of 1,200 diameters. 1T am indebted to Mr. F. J. Keeley for sketches from which some of the draw- ings are made. 222 PROCEEDINGS OF THE ACADEMY OF {Mar., LAND AND FRESH-WATER SHELLS FROM EASTERN CANADA. BY E. G. VANATTA. The following species of shells were taken by Mr: Bayard Long while collecting plants in the Magdalen Islands and Prince Edward Island in the Gulf of St. Lawrence. The snail fauna seems to be the usual northern type with the addition of the European Helix hortensis Miill., Hygromia hispida L., and a new Succinea related to western American forms. Suocinea bayardin.sp. Figs. 1, 2, 3. Shell rather small, oval, globose, thin, polished, apex red, obtuse, body whorl translucent amber colored, with a few growth strie, whorls about 23, convex, rapidly increasing; suture impressed. Aperture more than half the altitude, very broadly oval, parietal M @ 2 callous, thin, outer and basal lips thin and evenly arched, columella very narrow below, expanded above into a translucent white fold. Alt. 5.7, diam. 4.3, aperture alt. 3.8, diam. 2.8 mm. Locality —Indian River, Kensington, Prince Edward Island. Col- lected on August 29, 1912. The types are No.°106,651 in the col- lection of the Academy of Natural Sciences of Philadelphia. This shell is very closely related to Succinea oregonensis Lea, but has a lower and more obtuse spire and is a smaller species. It differs from Succinea chrysis West by lacking the opaque streaks, is not green or reddish, and is smaller. I take pleasure in naming this shell after Mr. Bayard Long, the botanist, who collected it. Helix hortensis Mull. Nineteen specimens were collected at Basin Island near Coffin Id., Magdalen Islands, of which ten were the form 12345, two (12)345, one (123)(45), two (123)45, one 1(2345), one (12345), one 10305, and one 00000; at Grindstone, Grindstone Island, M. I., nine speci- 1914.] NATURAL SCIENCES OF PHILADELPHIA. 223 mens were taken representing the following forms, four 12345, two 1(23)45, two (12)345, (123)45; four specimens from Alright Island, M. I., all are the form 12345. This species was collected at three locations on Prince Edward Island as follows, one form 12345 from Douglas, four 00000 from Souris and five 00000 from Bloomfield. Hygromia hispida L. Eight specimens were collected at Charlottetown, Prince Edward Island. Pupilla musoorum L. One specimen taken at Basin Island near Coffin Island, M. I. Bifidaria pentodon Say. Collected in alder thickets at Grindstone, M. I., and Tignish, Prince Edward Island. Vertigo ovata Say. Taken along a rill at Summerside and on knolls in alder thickets at Tignish, Prince Edward Island. Vertigo ventricosa Morse. * Collected in woods on Basin Island off Coffin Island, M. I.; and in alder thickets at Tignish, Prince Edward Island. Vertigo gouldi Binn. Taken in the woods at Grindstone, Grindstone Id., M. I., and at Mt. Stewart, Prince Edward Island. Columella edentula Drap. Five specimens taken on Basin Id., M. I. Acanthinula harpa Say. One specimen found near Campbell's Pond, Darnley, Prince Edward Island. Vallonia pulchella Miill. One specimen taken in coniferous woods at Charlottetown, Prince Edward Island. Cechlicopa lubrica Mill. Plentiful in the woods on Basin Id., M. I., one specimen being very tall; also found in alder thickets at Grindstone, M. I., and at Tignish, Prince Edward Island. Vitrina limpida Gld. In a low wet calcareous woods on Basin Id., M. I. Vitrea hammonis Strém. Collected in a meadow and in alder thickets on Grosse Isle, M. I.; in the woods on Basin Id.; in the thickets and woods at 224 PROCEEDINGS OF THE ACADEMY OF [Mar., * Grindstone, M. I.; and in the woods and thickets at the fol- lowing places on Prince Edward Island: Bloomfield, near Camp- bell’s Pond at Darnley, along the Indian River at Kensington, Mt. Stewart, Tignish, and around Lake Verde. One specimen from a wet birch woods at Lake Verde is so deficient in lime that the shell collapsed on drying. Vitrea binneyana Morse. Several specimens were taken in a rich wet thicket along a brook at Grindstone, Grindstone Id., M. I. Striatura milium Morse. Collected at Bloomfield, near Campbell’s Pond at Darnley, and around Lake Verde, Prince Edward Island. Striatura exiguum Stimp. Collected on Basin Id., in the woods and thickets at Grindstone, M. I., and at Bloomfield and near Campbell’s Pond at Darnley, Prince Edward Island. Euconulus fulvus Mill. Taken on Basin Id. and at Grindstone, M. I. Zonitoides arborea Say. Collected on Grosse Isle, Basin Id., at Grindstone, M. I., and on Prince Edward Island at Tignish, near Campbell’s Pond at Darnley, and in-Fullerton’s Marsh at Bunbury. Agriolimax agrestis L. Collected at Bloomfield, on- Brackley Point Road near Charlotte- town, Charlottetown, and near Campbell’s Pond at Darnley, Prince Edward Island. Arion circumscriptus Johns, Taken at Charlottetown, Prince Edward Island. Pyramidula cronkhitei anthonyi Pils. Collected on Grosse Isle; Basin Id.; East Cape, Coffin Id.; at Grindstone, Grindestone Id., M. I.; at Bloomfield, near Camp- bell’s Pond at Darnley, along Indian River at Kensington, and Tignish, Prince Edward. Island: ; Pyramidula (Planogyra) asteriscus Morse. Several specimens taken in thickets at Grindstone, M. I., and one at Bloomfield, Prince Edward Island. Helicodiscus parallelus Say. In the woods and thickets on Basin Id., at Grindstone, M. I., and at Bloomfield and around Lake Verde, Prince Edward Island. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 225 Punctum pygmeum Drap. Collected in thickets at Grindstone, M. I., and along Indian River, at Kensington, Prince Edward Island. Succinea ovalis Say. Was collected at Basin Island, M. I.; Etang du Nord on Grind- stone Id., M. I.; between Brackley Point and Charlottetown, also at Bloomfield on Prince Edward Island. Succinea retusa Lea. Was taken in a larch swamp on Grindstone Island, M. I.; and at North Lake and Tignish on Prince Edward Island. Sucoinea avara Say. One specimen found in a larch swamp on Grindstone Island, M. I. Succinea bayardi n. sp. On grass stems in a salt marsh just above the water along Indian River, Kensington, Prince Edward Island. Planorbis trivolvis Say. East Point Ponds and ponds east of East Cape, Coffin Island; Grand Tracadie and Dundee, Prince Edward Island. Planorbis antrosus Conr. . Was taken at Moncton, New Brunswick, Canada. Planorbis exacutus Say. From a brook in a meadow, Grindstone Island, M. I. Planorbis deflectus Say. Was taken at East Point Ponds and ponds east of East Cape, Coffin Id.; Etang du Nord and in two brooks on Grindstone IIo M. I.; North Lake, Black Pond, Lower Sea Cow Pond, Tignish, stream between Brackley Point and Charlottetown, and in a brook at Charlottetown, Prince Edward Island. Planorbis parvus Say. Was collected at Etang du Nord, Grindstone Id.; pools near the Narrows, Alright Id., M. I.; Lake Verde and in a stream between Brackley Point and Charlottetown, Prince Edward Island. Lymnea palustris Mill. Was collected in ponds east of East Cape, Coffin Id.; Etang du Nord, Hospital Pond and nine other pools on Grindstone Island, M. I.; North Lake, Grand Tracadie. Bloomfield, Fullerton’s Marsh at Bunbury, Dundee, Lower Sea Cow Pond at Tignish, in a stream between. Brackley Point and ( tharlottetown, Hillsborough River at 15 226 PROCEEDINGS OF THE ACADEMY OF {Mar., St. Andrews, and below the mill pond at Southport, Prince Edward Island. Lymneza humilis modicella Say. Was taken at Etang du Nord, Grindstone Island, M. I. Physa gyrina Say. E : ( East Point Ponds and ponds east of East Cape, Coffin Island; Etang du Nord and several small pools on Grindstone Island, M. L.; Black Pond, Tignish, East Lake at Bothwell, Dundee, stream between Brackley Point and Charlottetown, m a brook at Charlottetown, below the mill pond at Southport, Prince Edward Island; Moncton, New Brunswick. Carychium exiguum Say. In a swampy alder thicket at Grindstone and in a larch swamp on Grindstone Island, M. I.; on knolls in an alder thicket at Tig- nish, Prince Edward Island. Pisidium abditum Hald. Was collected on Grosse Isle; in ponds east of East Cape, Coffin Id.; Etang du Nord, Grindstone, Grindstone Id., M. I.; and at Charlottetown, Prince Edward Island. Pisidium variabile Prime. Was found in pools near The Narrows, Alright Id., M.*I., and in a pond at Tignish, Prince Edward Island. Anodonta cataracta Say. Was collected in the stream above Adams’ Pond at Darnley, Prince Edward Island. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 227 APRIL 21. Mr. CHarztes Morris in the Chair. Eleven persons present. The Publication Committee reported the receipt of contributions to the ProcerprnGs under the following titles: ‘Two new Sarcosporidia,”’ by Howard Crawley (March 18, 1914). “A new diatom,’’ by Charles 8. Boyer, A.M. (March 21). “Land and fresh-water shells from eastern Canada,” by E. G. Vanatta (March 21). “Montana shells,” by E. G. Vanatta (March 25). “The vascular system of the Florida alligator,” by Albert M. Reese (March 27). “The method of progression in Truncatella,”’ by Henry A. Pilsbry and Amos P. Brown (April 11). “List of land and fresh-water mollusks of Antigua,” by Henry A. Pilsbry and Amos P. Brown. The deaths of the following members were announced: Charles Sumner Williamson, March 23, 1914. George E. Kirkpatrick, March 26, 1914. Curwen Stoddart, Jr., April 1, 1914. Ernest Comly Dereum, April 10, 1914. The Council reported that Messrs. John Cadwalader, Charles B. Penrose, Witmer Stone, and J. Percy Moore had been appointed to constitute the Standing Committee on By-Laws. On the unanimous nomination of the Committee on the Hayden Memorial Geological Award, the gold Hayden Memorial Medal was conferred on Henry Fairfield Osborn, Se.D., LL.D., in recog- nition of his distinguished work in vertebrate palzeontology. Henry Farrrrmetp Osporn was born at Fairfield, Conn., August 8, 1857. He was educated at Princeton College (1877, 1880), Cambridge University under Balfour (1879), and the Royal College of Science under Huxley (1879-80). He has received the following degrees: Princeton A.B. 1877, Se.D. 1880; Hon. LL.D. Trinity College 1901, Princeton 1902, Columbia 1907; Hon. D.Sc. Cambridge 1904; Hon. Ph.D. Christiania 1911, Upsala 1913. Dr. Osborn began his paleontological explorations in Wyoming in 1877, and has continued them to the present time in various parts of the United States, in Europe and Egypt. From 1877 to 1890 he was connected with the Princeton University expeditions and the Museum of Geology, and between 1891 and 1914 he has directed the American Museum explorations in vertebrate palxontology. His own researches and those of his students fill five volumes of papers from the American Museum Memoir and Bulletin. His publications and papers in 16 228 PROCEEDINGS OF THE ACADEMY OF [Apr., vertebrate paleontology embrace 197 titles, including Evolution of Mammalian Molar Teeth, published in 1907, and The Age of Mammals, published in 1910, beside the American Museum volumes above referred to. They cover contribu- tions to the history of the extinct fishes, reptiles, and mammals. Dr. Osborn succeeded Edward D. Cope as Vertebrate Paleontologist of the Geological Survey of Canada between the years 1900-1904. In 1900 he was appointed successor to Othniel C. Marsh as Vertebrate Paleontologist of the United States Geological Survey, and is still engaged on the series of paleontological mono- graphs which were left unfinished by Marsh, especially those on the titanotheres and the Sauropoda. Among the most important of his investigations are the following: the evolu- tion of the ungulate foot, the evolution of the types of molar teeth of the mam- mals, the evolution of the perissodactyl ungulates, especially the rhinoceroses, horses, and titanotheres, the correlation of the Tertiary geological horizons of Europe and North America, the principles of the evolution of the skull in mam- mals, and the laws of evolution as observed in paleontology. Dr. Osborn’s administration of the Department of Vertebrate Paleontology in the American Museum of Natural History has been instrumental in assembling the most extensive collection of vertebrate fossils in existence, including the complete vertebrate fauna of more than half of the Tertiary horizons of the West and the complete phyletic evolution of many of the most important types of mammals. ‘This administration has been notable also in establishing the American Museum as a training school in vertebrate palzontology from which have issued many of the leading vertebrate paleontologists of the younger generation in this country-and abroad. There have been corresponding advances in paleontological technique and the educational methods of palzontology, so that the fossil collections in the American Museum have become a standard for the same work in other institutions. In addition to these paleontological and geological activities Dr. Osborn has held many educational and administrative positions, including the professorship of comparative anatomy at Princeton (1883-1900) and the professorship of zoology at Columbia University (1891-1914). Since 1890 he has been instru- mental in the development of the American Museum of Natural History and the founding and development of the New York Zoological Park. His contribu- tions on the administration of the Museum and the Zoological Park number twenty-six titles. His total contributions to sciencé cover a broad field, embracing in the 390 titles beside vertebrate paleontology researches and addresses on zodlogy, embryology, neurology, psychology, odontology, zoogeography, geology, biology, antropology, biography, and education. The Chair announced the following elections: MEMBERS: William J. Davis. Arthur W. Sheaffer. John 8. Sharp. CORRESPONDENTS: Shibasaburo Kitasato, M.D., of Tokyo, Japan. Charles T. Ramsden, of Guantanamo, Cuba. Marie Curie, of Paris. N. Charles Rothschild, of London. Gerritt S. Miller, of Washington, D. C. Sdmund Heller, of Washington, D. C. Charles W. Richmond, of Washington, D. C. Frank M. Chapman, of New York. Edgar A. Mearns, of Washington, D. C. The following were ordered to be printed: bo bo les’ 1914.] NATURAL SCIENCES OF PHILADELPHIA. FISHES FROM THE RUPUNUNI RIVER, BRITISH GUIANA. BY HENRY W. FOWLER. In the fall of 1912 the Academy received a collection of fresh- water fishes from the Rupununi River,.in the highlands of British Guiana. The specimens were purchased from Mr. J. Ogilvie, who collected them during the same year and in 1911. Mr. Ogilvie informs me, in lieu of the name of any settlement or town, they were approximately secured in North Latitude 2° to 3°, and West Longi- tude 50° 20’. A number are apparently new or undescribed, while others are not only new records for the Rupununi, but also for Guiana as well. The figures are all drawn to scale, each number over the accompanying line signifying millimeters. SELACHII. DASYATIDZA. Potamotrygon hystrix (Miiller and Troschel). One young foetal example. Color pale uniform brown. . Six pa- pill on floor of mouth. Body mostly smooth, without prickles or warts. Length 145 mm., disk width 60 mm. TELEOSTOMI. OSTEOGLOSSID 45. Osteoglossum bicirrhosum Agassiz. Fig. 1 (young). One 273 mm. Also three young with yolk-saes still adherent. According to Mr. Ogilvie, this fish carries its young in its mouth until they wholly absorb the yolk and they are able to fend for themselves. CHARACIDZ. -CURIMATIN®. Curimatus cyprinoides (Linnzus). ' One example, which agrees with my Ambyiacu River examples in the absence of gill-rakers, and with my figure.t Dr. Eigenmann rightly corrects my use of Curimata® to the present 1 Proc. Acad. Nat. Sci. Phila., 1906, p. 301, fig. 6. 2 Amer. Nat., XLI, 1907, p. 768. 230 PROCEEDINGS OF THE ACADEMY OF [Apr., form, as I overlooked Oken, though he wrongly identifies? the exam- ples I recorded as Psectrogaster ciliatus within the genus Cwrimatus. They are undoubtedly members of the genus Psectrogaster, as their spinescent scales show, and cannot be identified with the diagnosis he gives for the species Curimatus ciliatus, and this after he had examined the type of Anodus ciliatus Miller and Troschel in Berlin. If this species is found identical with Cwrimatus cyprinoides (Lin- nus), this latter name will, of course, supersede. Contrary to my arrangement in 1906, I now believe my Ambyiacu and Upper Amazon specimens to belong to Psectrogaster amazonicus Eigenmann and Eigenmann. Subgenus CYPHOCHARAX Fowler. Back not elevated, but upper profile nearly straight from above eye to near dorsal. Scales large, 36 or less in lateral line, and in nearly even longitudinal series. Fig. 1.—Osteoglossum bicirrhosum Agassiz. (Young.) This diagnosis is recast, as the species I identified with the desig- nated type is certainly different. Curimatus spilurus Ginther. One example 93 mm. CHILODIN #. Chilodus labyrinthicus rupununi subsp. nov. Fig. 2. Head 32; depth 31; D. rv, 9,1; A.1v, 7,1; P,1, 14; V,1, 8; scales 27 in |. 1. to caudal base, and 4 more on latter; 5 scales above 1. 1.; 4 scales below |. |. to ventral origin; 3 scales below 1. 1. to anal origin; about 8 predorsal scales; head width 1} its length; head depth at occiput 14; snout 3; eye 3}; maxillary 34; interorbital 2}; first 3 Mem. Carnegie Mus., V, 1912, p. 269. 1914.| NATURAL SCIENCES OF PHILADELPHIA. 231 branched -dorsal ray trifle longer than head; first branched anal ray 1{ in head; pectoral 14; ventral 12. Body elongate, robust, moderately compressed, and deepest at dorsal origin. Predorsal with slight or obsolete keel just before dorsal origin, other edges of body all convex, and that of preventral broad and flattened, or only very obsolete keel or ridge along outer boundaries. Lower profile evenly convex. Caudal peduncle com- pressed, small, about long as deep. Head robust, somewhat pyramidal, upper surface slightly de- pressed, and lower broadly convex, convex sides not especially converging below. Upper profile slightly convex, less inclined than —_> = === Fig. 2.—Chilodus labyrinthicus rupununi Fowler. (Type.) lower. Snout broadly triangular as seen from above, surface convex, and length about half its width. Eye large, circular, high, and a little anterior m head. Adipose-eyelid broad, exposes most of very broad pupil. Mouth small, terminally inferior, and seen below transversely crescentic. Upper jaw with single series of small weak filament-like teeth, none in lower jaw. Maxillary small, upper edge slips below preorbital, and slightly expanded distal end reaches opposite posterior nostril. Mandible small, shallow, each ramus well elevated inside mouth. Tongue well back, large, depressed, free. Nostrils well developed, superiorly lateral, together, in last third of snout length, anterior with well-developed cutaneous rim 232 PROCEEDINGS OF THE ACADEMY OF {Apr., and posterior exposed as crescent. Interorbital broad, flattened. Infraorbital large, covers most of cheek, surface with radiating strie. Opercle striate, and subopercle projects moderately pos- teriorly, both with striate surfaces. Occipital fontanel well devel- oped, extends forward nearly to front of eyes. Gill-opening extends forward for last third in head. Gill-rakers about 10 + 20 weak flexible filaments, about 4 length of gill-filaments, and outer series separated by high cutaneous or cartilaginous parti- tion. Gill-filaments about 2 im eye. Isthmus broad. Branchioste- gals with inner shortest and outer longest. Scales large, firm, well exposed, in regular horizontal series, entire, and more or less uniform except smaller ones on caudal base. Surface of exposure of each scale pitted or with variously shaped shallow cavities, not interfering with general smoothness to touch. Long pointed free scaly flap in axil of ventral. Lateral line complete, midway along side, of simple tubes and each opening by pore at middle of scale exposure. Dorsal origin midway between that of adipose fin and snout tip, first branched ray longest with fourth simple but slightly shorter, and fin depressed slightly more than half way to caudal base. Adi- pose fin inserted about midway between depressed dorsal tip and caudal base, fin small. Anal inserted slightly before adipose fin, first branched ray longest, lower edge of fin slightly emarginated, and when depressed reaches caudal base. Caudal well forked, lobes broad, apparently (damaged) equal. Pectoral low, pointed, reaches about ;°, to ventral, and latter inserted behind second branched dorsal ray base, reaches } to anal. Vent close before anal. Color in alcohol largely pale brownish, ground color mostly uniform. Above lateral line four lengthwise deeper brownish bands; upper three, together with additional median dorsal band, much broader than lower ones, and each extending over median portions of scales. L. 1. in a broad underlaid pale slaty to dusky band, not quite equal to pupil diameter in width, and extending posteriorly out on median rays of caudal to their tips. Each scale in dark longitudinal bands, besides those in |. |. and for most of extent of two lengthwise series of scales below 1. |., with deep dusky-brown spot formed at its base. Dark median lateral band of trunk continued on head across opercle and in front to snout tip. Head brown above, pale like belly below. Eye slaty. Above pectoral in course of dark lateral band two ill-defined dusky spots, subequal, and each about long as pupil. Fins all pale or whitish in general tint, though dorsal 1914.| NATURAL SCIENCES OF PHILADELPHIA. 233 with broad vertical dusky band, slightly narrowing below, and upper posterior edge of fin slightly dusky. Caudal, except as stated above, pale and uniform. Adipose fin pale brownish. Lower fins all pale or whitish. Length 136 mm. (caudal tips damaged). Type, No. 39,306, A. N.S. P. Rupununi River, British Guiana. J. Ogilvie. Only the type known. This differs from Chilodus labyrinthicus (Kner)* in coloration, that species having but one blackish spot above the depressed pectoral fin. (Named for the Rupununi River.) HEMIODONTIN ©. Hemiodus semiteniatus Kner. One example 83 mm. long (caudal tips damaged), agrees with Kner’s account. The depth (43) is probably due to age, as all the other characters are in accord. The scales (according to the pockets) are 55 in |. |. to caudal base and 3 more on latter, besides 9 above 1. |. Eigenmann identifies examples from Konawaruk and Gluck Island with this species,* though they are likely wrongly so called, as they show the depth 4 to 45 and the scales 44 or 45 in 1. 1., with only 7 above. Hemiodus quadrimaculatus Pellegrin. Two examples, one 115 mm. and the other 32 mm. Anisitsia notata (Schomburgk). One 147 mm. long and another 157 mm. PYRRHULININ#. Pyrrhulina filamentosa Valenciennes. One 70 mm. Eigenmann mentions “adipose brick-red,’’ certainly an error if referring to an adipose fin.® NANNOSTOMIN/. Characidium blennioides Eigenmann. One example 47 mm. (caudal damaged). Characidium fasciadorsale sp.nov. Fig. 3. Head 32; depth 54; D.im, 8; A. 1m, 7; P. 1m, 8; V.1, 8; scales 34 in lateral line to caudal base and 2 more on latter; 5 scales above 4 Microdus labyrinthicus Kner, Denk. Ak. Wiss. Wien, XVII, 1859, p. 149, Pl. 3, fig. 5. Rio Branco and Barra do Rio Negro. 5 Mem. Carnegie Mus... V, 1912, p. 276, PL 36, fig. 3. § L.c., p- 279- 234 PROCEEDINGS OF THE ACADEMY OF [Apr., 1. 1.; 3 scales below 1. 1. to ventral origin; 3 scales below lI. |. to anal origin; 9 scales before dorsal; head width 2+ in its length; head depth at occiput 13; snout 4; eye 32; maxillary 42; interorbital 4%; length of dorsal 1; least depth of caudal peduncle 24; upper caudal lobe 1;; pectoral 1; ventral 14. Body elongate, slender, moderately compressed, greatest depth at dorsal origin, profiles similar and edges rounded. Caudal peduncle well compressed, least depth about 1? its length. Head attenuated, compressed, flattened sides but scarcely con- verging below, upper profile convex and lower concave. Snout conic, protruding slightly, compressed, long as broad. Eye rounded, little longer than deep, high and slightly anterior. Mouth small, transverse. Teeth small, invested with membrane, conic, simple, pointed. Maxillary small, free, well inclined, reaches midway in RR RR DP REMY Aw | RP Pe SS Ss ay ‘ if Hissss Q) — = CS: \\ = SN yy 3 Fig. 3.—Characidium fasciadorsale Fowler. (Type.) £ J yp space between front and hind nostrils. Tongue depressed, free in front. Mandible shallow in front, and rami well elevated inside mouth. Nostrils well separated, anterior about midway in snout length, and posterior close before front eye edge, both simple pores. Preorbital triangular, length about ? of eye. Interorbital broadly convex. Infraorbitals narrow. Opercle moderate, surface smooth. Subopercle extended posteriorly and with broad cutaneous flap. Gill-opening extends forward about opposite hind edge of pupil. Gill-rakers about 4+ 10? short weak points, about 4 length of filaments and latter about 2} in eye. Isthmus moderately broad. Branchiostegals moderate. Seales mostly uniform in size, in even longitudinal series parallel with 1. 1., well exposed, and each showing about six horizontal striw on exposures. Ventral axilla with rather short pointed free scaly 1914.] NATURAL SCIENCES OF PHILADELPHIA. 235: flap. I. 1. complete, midway along side, and of simple tubes which: at first extend half way over exposure of scale and posteriorly about three-quarters. Dorsal origin about midway between snout tip and end of adipose fin, first branched ray highest, extends almost as far posteriorly as tip of last, and depressed fin reaches but slightly less than half way to caudal base. Adipose fin small, inserted about midway between end of depressed dorsal and caudal base. Anal inserted close behind depressed dorsal tip, and fin extends about two-thirds to caudal base. Caudal well forked, sharply pointed lobes about equal. Pectoral pomted, with median rays longest, and fin reaches almost to ventral origi. Ventral inserted about opposite middle of dorsal base, fin reaches four-fifths to anal. Vent close before anal. Color in alcohol with very pale brown ground color, scarcely paler below. A narrow brownish lateral band, about equal to. diameter of pupil in width on head, though on trunk reduced to narrow line, extends from snout tip to caudal base where it ends in small detached dusky spot. Head pale brownish above, whitish or paler beneath. Lips and front half of maxillary dusky, though posterior half of maxillary contrasted pale or whitish like lower surface of head. Eye slaty. Trunk with ten transverse deep brownish bands, a little broader than interspaces, and within area of each edge of each scale deeper brown. Dorsal and caudal dull grayish, with at least three transverse or horizontal dusky streaks across former. Pectoral, ventral and anal pale and uniform. No dark median dorsal streak. Length 47 mm. Type, No. 39,307, A. N. 8. P. Rupununi River, British Guiana. J. Ogilvie. Also Nos. 39,308 and 39,309, same data, paratypes. These show: Head 32 and 34; depth 5 and 53; D. m1, 9 and im, 7; scales 32 ? (injured) in |. |. to caudal base and about 3 more on latter; 5 scales above |. l.; 3 scales below |. |. to anal origin; 9 predorsal scales; snout 42 and 4; eye 3} and 33; maxillary 4 and 4}; interorbital 41 and 4; length 35 and 42 mm., respectively. The smaller example differs from the others in having the pectoral extending slightly beyond the ventral origin and the latter fin reaching the anal origin. This species is related to Characidium etheostoma Cope’ in colora- tion, but is much more slender, thus approaching Characidiwm 7 Proc. Acad. Nat. Sci. Phila., 1906, p. 323, fig. 17. Type. 236 PROCEEDINGS OF THE ACADEMY OF [Apr., catenatum Eigenmann. From the latter it differs chiefly in the ‘dark horizontal cross-bands on the dorsal fin, that species bemg figured and described as having a uniform dorsal. (Fascia, streak; dorsalis, of the back; with reference to the streaked dorsal fin.) : ANOSTOMIN-®. Leporellus vittatus (Valenciennes). Two examples, one 195 mm. and the other 126 mm. Compared with the example from the Peruvian Amazon,’ I find they differ only according to age. I cannot find that the species has been recorded from Guiana previously. The following notes may be of value: Head 33; depth 44; D. m, 9, 1 or m1,-10, 1; A. m, 8, 1; scales 37 to 39 in |. |. to caudal base and 4 or 5 more on latter; 6 scales above |. 1.; 5 scales below |. |. to anal origin; 11 predorsal scales; snout 2} in head; eye 5 to 6; maxillary 3 to 33. In color these examples show a dark or dusky spot on each scale of sides and back. Dark spots on head large and irregular on front and snout, small and less numerous on cheek and opercle. Dorsal with large broad distal blackish blotch, and another blackish streak transversely subbasally. Broad dark lateral band includes |. |. and extends out on median caudal rays to their tips. Each caudal lobe with two blackish transverse horizontal bars. Adipose fin pale, lower edge blackish. Other fins pale, ventrals and anal with median dusky shade. Anostomus anostomus Channaaneh Three examples, 81 to 85 mm. (caudals damaged). Leporinus nigroteniatus (Schomburgk). Five examples, 142 to 190 mm. Leporinus friderici (Bloch). One example 210 mm. in length, and a smaller one 120 mm. long."” Leporinus alternus Eigenmann. Two examples, 70 to 73 mm. in length. The only point at variance in the original account is the statement ‘“‘four graduated, obliquely- pointed teeth in each jaw,’’ my examples, however, showing eight teeth in each jaw. 5 Mem. Carnegie Mus., V, 1912, p. 293, Pls. 38, figs. 5-6. Warraputa, Rock- stone and Crab Falls. ® Proc. Acad. Nat. Sci. Phila., 1906, p. 327. '0° The Parahyba example which I identified with Salmo fasciatus Bloch, in’ Proc. Acad. Nat. Sci. Phila., 1906, p. 328, I feel certain is identical with Lepo- rinus conirostris Steindachner, Sitz. Ak. Wiss. Wien, LXXI, I, 1875, p. 233, PI. 4. 1914.] NATURAL SCIENCES OF PHILADELPHIA. 237 Leporinus paralternus sp. nov. Fig. 4, Head 33; depth 4!; D. rity LOTS AS mt, 8; P> 1, 16; V.1;8: scales 36 in lateral line to caudal base, and 3 more on latter; 6 scales above 1. 1.; 5 scales below 1. 1. to ventral origin; 5 scales below 1. |. to anal origin; 13 predorsal scales: head width 2 in its length; head depth at occiput 13; snout 2?; eye 43; maxillary 4; interorbital 3; first branched dorsal ray 13; first branched anal ray 12; upper caudal lobe 1; least depth of caudal peduncle 22; pectoral 14; ventral 13. Body moderately long and slender, compressed, deepest. at dorsal origin, and edges all convexly rounded. Caudal peduncle well compressed, about long as deep. Head rather conic, compressed, sides flattened slightly though about evenly curving over above and below. Snout elongate, conic, Fig. 4.—Leporinus paralternus Fowler. (Type.) long as broad. Eye rounded, a little high, centre about midway in head. Adipose-eyelid very slight. Mouth small, nearly terminal, or but slightly inferior, so that upper jaw only protrudes very little. Maxillary moderately inclined, small, extends a little beyond anterior nostril, though not to posterior. Lips thick, fleshy, entire. Eight teeth in each jaw, graduated forward to median which are longest, and all ending in a point on cutting-edge. Inner buccal membrane above quite broad. Tongue little distinct. Mandible shallow in front, short rami well elevated in mouth. Anterior nostril lateral, in small cutaneous tube reaching beck nearly to posterior, which latter simple and about last 3 in snout length, also lateral. Inter- - orbital evenly convex. Suborbitals of moderate width. Preopercle * 238 PROCEEDINGS OF THE ACADEMY OF [Apr... edge inclined forward. Opercle moderately large, with a few radi- ating striz on its lower edge. Gill-openings lateral, extend forward about opposite hind pre- opercle edge. Gill-rakers about 11 +9 points, about 3 in gill- filaments, and latter 13 in eye. Isthmus broad. Branchiostegals large, subequal. Scales large, well exposed in longitudinal series parallel with |. 1., more or less uniform in size, though a little smaller on breast and caudal base. Scales ensheath bases of dorsal and anal, though not. extending on the fins. Ventral with free pointed axillary scale trifle less than half length of fin. Lateral line complete, midway along side, extends on caudal base, and formed of simple tubes extending well over scales. Dorsal origin a little nearer hind edge of adipose fin than snout tip, first branched ray longest, and depressed fin extends 3 to caudal base. Anal inserted well behind depressed dorsal, first branched ray longest and reaches caudal base. Caudal well forked, pointed lobes equal. Adipose fin about # of eye, inserted about midway between caudal base and hind edge of depressed dorsal fin. Pectoral small, low, extends + to ventral origin. Ventral inserted behind second branched dorsal ray base, fin extending trifle more than half way to anal origin. Vent at last third in space between depressed ventral tips and anal origin. Color in alcohol largely pale uniform brownish above, below, or on belly and abdomen, paler and whitish. Head brownish above, sides and below paler. Lips pale. Iris slaty. Trunk with: nine dusky to blackish transverse cross bands, of which first or that just after occiput, fourth or that below dorsal, seventh or that just before adipose fin, and last at caudal base, broader or much more expanded than others. Second intermediate, also fifth intermediate transverse streaks entirely dorsal and only extend down each side half way to lateral line. Fins all pale or uniform, dorsal and caudal slightly darker, and adipose fin with a little grayish. Length 120 mm. Type, No. 39,320, A. N.S. P. Rupununi River, British Guiana. J. Ogilvie. Nos. 39,321 to 39,323, paratypes, same data. Head 33 to 33; depth 4 to 44; D. m1, 10,1; A. 1, 8, 1; scales 36 or 37 in I. 1. to caudal base and 3 or 4 more on latter; 6 scales above |. 1.; 5 scales below |. l.; 13 predorsal scales; snout 22 to 2} in head; eye 4 to 4}; maxillary 41 to 42; interorbital 3; length 75 to 82 mm. These: 1914] NATURAL SCIENCES OF PHILADELPHIA. 239 examples all agree with the type in the constancy of their color patterns. The present species is distinguished from its ally, the preceding, by the presence of an extra black streak before the dorsal. (Para, near; alternus, alternate, the related species.) MYOCHARAX subgen. nov. Type Leporinus desmotes sp. nov. Differs from the subgenus Anostomus in the dentition, that of the mandible being composed of two long slender curved rodent-like incisors, and also two somewhat similar though much shorter ones externally and well back. Upper jaw with six tricuspid incisor-like teeth. Anterior nostrils in rather long cutaneous tubes. One species, described below. (vs, mouse; 74945, Charax; with reference to the rodent-like teeth.) Leporinus desmotes sp. noy. Fig. 5. Head 4-depthe42-— Dy m1, 10Sns Avir28; 15 2b tG) V1, 9: seales 35 in lateral line to caudal base and 6 more on latter; 6 scales above |. l.; 5 scales below |. |. to ventral origin; 5 scales below 1. 1. to anal origin; 13 predorsal scales; head width 2 in its length; head depth at occiput 13; snout 23; eye 4; maxillary 4; inter- orbital 23; first branched dorsal ray 1; first branched anal ray 17; least depth of caudal peduncle 2$; upper caudal lobe 14?; pectoral 12; ventral 14. , ; Body elongate, compressed, tapers somewhat posteriorly, deepest at dorsal origin, anterior upper profile rather evenly convex, and edges all convexly rounded. Caudal peduncle compressed, about long as deep. Head rather small, compressed, somewhat conic, surfaces rather conic, and profiles similarly straight and inclined. Snout conic, broad, length # its width, and as seen from above rather acuminate. Eye rounded, lateral, and centre falls trifle behind middle in head length. Adipose-eyelid narrow, best developed anteriorly. Mouth small, inferiorly terminal. Maxillary with thick integument, well inclined and reaches trifle over half way to eye or about opposite front edge of posterior nostril. Lips thick and fleshy. Upper lip broader, with two series of fleshy lamelle, of which outer show much deeper clefts, are longer, though more vaguely defined’ exter- nally. Lower lip peculiar, with deep-cleft narrow fleshy lobe each side leaving broad symphyseal surface, this crowned with seven or 240 PROCEEDINGS OF THE ACADEMY OF [Apr., more series of lamellw, each deeply cleft and therefore distinct, besides each lamella beg broken up transversely into as many small apical papille. Teeth of upper jaw quite different from those in lower, 6 in number, though each rather indistinctly tricuspid median cusp much better defined, and outer tooth of each side a little smaller than subequal inner ones. Lower teeth developed as two long slender upward-curved pointed teeth, and basally though well back on each side another similar, much shorter and inconspicuous tooth, apparently less firm im its socket. Mouth with very broad upper buccal membrane, its surface finely papillose. Tongue not deter- mined. Mandible short, rami not especially high. Anterior nostril lateral, in conspicuous and somewhat bell-shaped cutaneous tube Fig. 5.—Leporinus desmotes Fowler. (Type.) equal to half of eye-diameter in length, and protruded beyond edge of upper lip. Posterior nostril lateral oblique slit, formed about last 2 in snout length about opposite eye centre. Interorbital broadly convex. Suborbitals moderate in width. Hind edge of preopercle slopes forward. Opercle moderate, smooth. Gill-opening restricted, extends forward only about last sixth in head. Gill-rakers 8 + 11 rather short broad firm points, trifle over 4 in gill-filaments and latter slightly exceeding half an eye- diameter. Isthmus rather broad and convex. Branchiostegals 4, rather broad, moderate, subequal. : Seales large, well exposed, slightly smaller on breast and predorsal region, though much-more so on caudal base, otherwise uniform, 1914.] NATURAL SCIENCES OF PHILADELPHIA. 241 and disposed in longitudinal series parallel with |. 1. Sealy sheaths at dorsal and anal bases low, allowmg movements of fins, though not extending on their surfaces. Ventral with free pomted axillary sealy flap, nearly 3 in length of fin. L. 1. complete, nearly straight, and tubes simple, extending well over first half in scale exposures. Dorsal origin midway between snout tip and hind basal edge of adipose fin, first branched ray longest or extends back well beyond tips of last, and fin reaches but trifle less than half way to caudal base. Adipose fin inserted little nearer depressed dorsal tip than caudal base, and length about equals eye. Anal inserted trifle before adipose fin, first branched ray longest, and fin reaches trifle beyond caudal base. Caudal large, well forked, broad pointed lobes about equal. Pectoral low, pointed, reaches about ¢ to ventral. Latter inserted behind first branched dorsal ray base, fin large, and reaches 13 to anal origin. Vent nearly at first third in space between depressed ventral and anal origin. Color in alcohol generally pale brownish or whitish, marked by ten very conspicuous well-defined slaty-black transverse bars, those on trunk meeting their opposite sides and sloping slightly posteriorly. On head lower surface pale and dark bars reflected slightly across. Lips pale or whitish. Iris slaty. First dark bar includes upper edge of snout, extends down along maxillaries and then across lower surface of mandible posterior to symphyseal papille, and its width much less than any of the other bars. Second dark bar close behind, and of moderate width, bridges lores over upper surface of snout. Third dark bar. extends over interorbitals and also reflected on infraorbital. Fourth dark bar extends from occiput, includes opercles and pectoral root. Fifth dark bar slightly forked above and below, midway in predorsal. Sixth dark bar includes dorsal base anteriorly and ventrals. Seventh dark bar postdorsal and postventral in position. Eighth dark bar wholly before adipose fin, extends to front half of anal. Ninth dark bar includes adipose fin and caudal peduncle. Tenth dark bar includes caudal base. Except for such intrusion as noted, all fins of whitish color generally. Length 170 mm. (caudal tips damaged). Type, No. 39,324, A. N.S. P. Rupununi River, British Guiana. J. Ogilvie. : Nos. 39,325 to 39,327, paratypes, same data. Head 32 to 33; depth 32 to 4; D. m1, 10,1; A. m, 8,1; scales 33 to 35 in 1. 1. to caudal base and 3 or 4 more on latter; 6 scales above |. |.; 5 scales 242 PROCEEDINGS OF THE ACADEMY OF [Apr., below |. I.; 12 or 13 predorsal scales; snout 2? to 23 in head; eye 32 to 44; maxillary 44 to 5; interorbital 22 to 23; length 105 to 149 mm. Only comparatively slight mdividual variation is noticed in these examples. One example also has an additional lesser external mandibular tooth on the right side, only very inconspicuous. (despvw-ys, prisoner, with reference to the dark cross-bands sug- gestive of convict garb.) Schizodon fasciatus Agassiz. One example 170 mm. TETRAGONOPTERIN 2%. Tetragonopterus argenteus Cuvier. One example 78 mm. Tetragonopterus chalceus Agassiz. Two examples, one 80 mm. and the other 97 mm. Astyanax rupununi sp. noy. Fig. 6, Head 4; depth 22; Dim;/85 07 1A. mm, 28, 1;— Pr, 12> Vie ee scales 35 in lateral line to caudal base and 3 more on latter; 8 scales above |. 1.; 6 scales below |. |. to ventral origin; 7 scales below 1. 1. to anal origin; 13 predorsal scales; head width 1,°; in its length; head depth 1; first branched dorsal ray 1; first branched anal ray 12; least depth of caudal peduncle 24; pectoral 14; ventral 13; snout 4 in head measured from upper jaw tip; eye 24; maxillary 23; interorbital 23. Body elongately ovoid, well compressed, deepest at dorsal origin, and edges all rounded convexly except median predorsal ridge, which slightly trenchant. Caudal peduncle compressed, length about equals least depth. Head small, compressed, lower profile little more inclined than upper, and flattened sides not converging above or below. Snout convex, length about 2 its width. Eye rounded, placed about first in head. Mouth very slightly inclined, transverse, broad, terminal. Maxillary inclined vertically, free, extends back to front eye edge and greatest expansion 3} in eye. Lips firm, rather thin. Upper jaw teeth biserial, outer series smaller and tricuspid, and inner series of larger mostly quincuspid, in all cases median cusp much largest. Mandibular teeth quincuspid, large, powerful and uniserial. At least one of teeth extending-from inner upper series on inner base of maxillary, though tooth quite small and obsolete. Mandible very slightly protrudes, strong, and rami scarcely elevated inside mouth. Tongue depressed, rounded, and free in front, rather 1914.] NATURAL SCIENCES: OF PHILADELPHIA. 243 broad. Inner buccal folds broad. Nostrils together, anterior simple pore with hind cutaneous rim exposing larger posterior aper- ture in crescent, and anterior falls about last third in snout length. Interorbital evenly convex. Infraorbital broad, covers cheek, with few radiating strie, and its width 2 of eye-diameter. Hind pre- opercle edge nearly vertical. Opercle narrow, with a few obsolete striz, and its width trifle less than. 4 its length. Gill-opening forward to front eye edge... Gill-rakers about 10 + 13, slender, tapering, pointed, and slightly shorter than gill-filaments, which latter about half of eye. Isthmus narrow, constricted, surface rounded though with median. groove in. front. Branchiostegals moderate, subequal. Fig. 6.—Astyanax rupununi Fowler. (Type.) Seales disposed in. even. longitudinal series parallel with |. L., mostly uniform in size except those variably smaller along predorsal, breast, preventral,. postventral, caudal base and along anal base. Ventral with free scaly pointed axillary flap about 2 length of fin. L. |. complete, decurved slightly below median axis, and extending up a little low along side of caudal peduncle at first. Tubes simple, and better exposed in. posterior course of |. |. Dorsal origin midway between. snout tip and caudal base, first branched ray longest, extends back well beyond tips of last, and depressed fin reaches half way to. caudal base. Adipose fin inserted little nearer depressed dorsal tip than caudal base, and its length V7 244 PROCEEDINGS OF THE ACADEMY OF -[Apr., about 2 of eye. Anal with long base, inserted below last branched dorsal ray base, first branched ray longest equals half length of base, and lower edge nearly straight. Caudal well forked, pointed lobes about equal. Pectoral low, pointed, reaches slightly beyond ventral origin, though not back till opposite dorsal origin. Ventral inserted midway between pectoral and anal origins, reaches back { to anal. Vent at tip of depressed ventral. Color in alcohol largely dull brownish on back and upper surface of head. Sides of head and trunk paler than back, and becoming still more so, or whitish, on ventral region. All scales of back and sides with darker brownish edges, made up of dark dots. Sides of head, opercle, postorbital and muzzle sprinkled with larger dusky dots. Iris slaty. Slightly above level of eye centre, just above 1. l., and about opposite middle in length of pectoral, a blackish ellipsoid blotch, its length about equal to eye. From its upper front end and its entire hind end a pale area extends transversely as two ill-defined pale or whitish spots, posterior much greater in extent. Extending back along vertebral axis a dusky line begins behind pale area and continues back along caudal peduncle side, where it widens, to expand still more at caudal base into large dusky or blackish blotch, and also reflected out on median caudal rays to their tips. Fins all whitish, dorsals and caudal tinged grayish. Distal edge of anal its whole extent shghtly tinged grayish. Length 64 mm. Type, No. 39,228, A.N.S.P. Rupununi River, British Guiana. 1912. J. Ogilvie. No. 39,329, paratype, same data. Head 3%; depth 23; D. m1, 9; A. rv, 26,1; scales 36 in |. 1. to caudal base and 3 ? more on latter; 7 scales above 1. l.; 7 scales below |. 1. to ventral origin; 8 scales below |. |. to anal origin; 14 predorsal scales; snout 4 in head; eye 3; maxillary 2}; interorbital 24; length 62 mm. _ This species appears to be related to Astyanax wappi (Valen- ciennes), as described from the type by Eigenmann." It differs in the smaller head, deeper body, and larger eye. There are, how- ever, no “traces of longitudinal streaks between the rows of scales.” (Named for the Rupununi River.) Menkhausia chrysargyrea leucopomis subsp. nov. Fig. 7. Head 34; depth 1,5; D. m, 9; A. 1v, 26,1; P. 1,13; V. 1, 7; scales 34 in lateral line to caudal base and 2 ? more on latter (squama- " Mem. Carnegie Mus., V. 1912, p. 355, Pl. 52; fic. 1. 1914. NATURAL SCIENCES OF PHILADELPHIA. 245 tion injured); 8 scales above 1. 1.; 7 scales below 1. 1. to ventral origin; 8 seales below |. |. to anal origin; 9 scales before dorsal to posterior end of occipital process; head width 2 in its length; head depth at occiput 1; snout 32; eye 3; maxillary 22; interorbital 23; first branched anal ray 12; least depth of caudal peduncle 2}; pectoral 14; ventral 13. Body deep, well compressed, rather ovoid, predorsal and post- dorsal with slight median ridge, and other edges convex, except slight ridge each side along abdomen before-ventral, most distinct Fig. 7—Menkhausia chrysargyrea leucopomis Fowler. (Type.) just before latter. Greatest depth at dorsal origin. Caudal peduncle well compressed, and its length about $ its least depth. Head deep, well compressed, lower profile more inclined than upper, which latter nearly straight from snout front to occiput. Flattened head sides very slightly converging below. Snout convex over surface and in profile, broadly convex as seen from above and length about half its width. Eye rounded, placed about first 2 in head. Adipose-eyelids slightly developed in front and behind. Mouth broad, terminal, about level with upper rim of pupil. Lips thick, fleshy, firm. Jaws about even when closed, mandible scarcely 246 . PROCEEDINGS OF THE ACADEMY OF [Apr., projecting. Maxillary vertically inclined, extends back slightly beyond front eye edge, and its greatest expansion a little less than 3 in eye. Teeth in upper jaw biserial, quincuspid, though those in inner series larger and apparently continued on inner base of each maxillary as a small single tooth. Mandible with uniserial quin- cuspid teeth, median largest and lateral ones very small posteriorly. Tongue broad, depressed, rounded and free in front. Inner buccal folds broad. Mandible strong, convex over surface; and rami not elevated inside mouth. Nostrils together, anterior simple pore with cutaneous flap behind exposing posterior in crescent, and frenum would fall about last fourth in snout length. Interorbital evenly convex. Posterior infraorbital broad, covers all of cheek except lower narrow strip, rather obscurely striate, and its greatest width 3 of eye. Postorbital narrow. Hind preopercle edge nearly vertical, sloping slightly forward below. Opercle deep and narrow, and surface nearly smooth. A long occipital fontanel begins oppo- site front pupil rfm and extends up within occipital process well towards its tip. Gill-opening forward opposite front edge of eye. Gill-rakers about 9 + 12, slender, lanceolate, about 3 length of gill-filaments, and latter 12 in eye. Isthmus narrowly constricted, surface convex, and with slight groove in front. Branchiostegals moderate, sub- equal. Seales mostly large and well exposed, disposed in longitudinal series parallel with 1. 1.; and each one with a number of radiating strie, about 8 usually exposed. Scales become a little smaller along anal base, and two or three series extend over latter, at least on anterior part of fin. Seales of small size over most of caudal, at least its greater basal portion. Free gxillary pointed ventral scaly flap, its length 2 of fin. L. 1. complete, slightly decurved, running a little low along side of caudal peduncle at first, and simple tubes extend about first } over exposures of scales. Dorsal inserted about midway between hind edge of adipose fin and snout tip, first branched ray-longest, and depressed fin extends 13 to caudal base. Adipose fin inserted much nearer depressed dorsal tip than caudal base, its length about equal to eye. Caudal well forked, pointed lobes about equal. Anal inserted opposite last dorsal ray base, with slight elevated lobe in front. Pectoral slender, pointed, low, reaches slightly beyond ventral origin. Ven- tral inserted slightly before dorsal insertion, and depressed fin about reaching to anal origin. Vent at last third in space between ventral and anal origins. ee 1914.] NATURAL SCIENCES OF PHILADELPHIA. 247 Color in alcohol largely pale brownish, upper or dorsal region very slightly darker, and lower surface paler. Head brownish above, paler below. Lips pale brown. Eye slaty. Level with upper part of eye, or on vertebral axis, a blackish humeral blotch, a little deeper than wide, and distant from head 5 seales. All about humeral blotch a pale area, best understood on examining figure. Several scales posterior a narrow dusky streak begins and runs along vertebral axis, and though gradually enlarging to caudal base, not forming spot on latter. Fins all pale or whitish, dorsal and caudal tinged with grayish. Length 95 mm. Type, No. 39,330, A. N.S. P. Rupununi River, British Guiana. J. Ogilvie. No. 39,331, paratype, same data. Head 32; depth 2; D. m1, 9; A. Iv, 28, 1; scales 33 in 1. |. to caudal base, and 2 more on latter 8 scales above |. 1.; 6 scales below |. |. to ventral origin; 7 scales below 1. 1. to anal origin; 10 predorsal scales; snout 4 in head; eye 375; maxillary 22; interorbital 2%; length 95 mm. Apparently a form of Menkhausia chrysargyrea (Giinther),” though it is incompletely described. No mention is made of the pale area surrounding the humeral blotch, and the dark posterior lateral vertebral streak. That this color pattern should have resulted from preservation originally in formaline may be likely in the first case, though less so in the last. (Aevzes, white; éxwy/s, shoulder.) Gymnocorymbus nemopterus sp. nov. Fig. 8. Headge.sdepin (4. Dy rv. 90 1 AN ny 32, 1 Peal" Ve r 7: scales 33 in |. |. to caudal base and 2 more on latter; 10 scales above l. 1.; 10 seales below 1. 1. to anal origin; 12 predorsal scales; head width 12 its length; head depth at occiput 1; snout 4 in head, meas- ured from upper jaw tip; eye 22; maxillary 27; interorbital 23; least depth of caudal peduncle 2; first branched anal ray 13; pec- toral 1; ventral 13. Body deeply ovoid, compressed, upper anterior profile double coneave, and lower profile much more bulging. Body edges con- stricted, though not trenchant, and greatest depth at dorsal origin. Caudal peduncle well compressed, its length about half its least depth. 12 Tetragonopterus chrysargyreus Giinther, Cat. Fish. Brit. Mus., VII, 1868, p. 328. Hssequibo. 248 PROCEEDINGS OF THE ACADEMY OF [Apr., Head small, compressed, upper profile slightly concave from snout to occiput, lower profile a little more inclined, and flattened sides but very slightly constricted below. Snout convex over surface and in profile, broadly convex as viewed from above, and its length about half its width. Eye rounded, placed about first 2 in head. Adipose-eyelid little developed, only extends on eye a little in front and behind. Mouth broad, shallowly cleft, and latter falling on level with upper edge of pupil. Maxillary nearly vertical, reaches opposite front of eye, slender, and greatest expansion about 4 of eye. =: an eases) 3 cas ee Sy MS -> IBS -