f ^-1^' HARVARD UNIVERSITY Library of the Museum of Comparative Zoology DOES MOT rf!>/-f II A-rr- N. E. WRIGHT CAMBRIDGE 38. MASS l»»TU«N P08TA0E OUARANT.ED BREVIORA MUSEUM OF COMPARATIVE ZOOLOGY AT HARVAED COLLEGE, IN CAMBRIDGE Numbers 121-178 1960-1962 CAAIBRIDGE, MASS., T.S.A. 1963 Edited By Nelda E. Weight CONTENTS BREVIORA Museum of Comparative Zoology Numbers 121-178 1960 No. 121. Three new species of Micrathena ( Araneae. Argiopidae) from South America. By Arthur M. Chickering. 11 pp. March 10. No. 122. Notes on certain species of Micratkcna (Araneae, Argi- opidae) from South America. By Arthur M. Chick- ering. 7 pp. March 11. No. 123. Alepisaurus hrevirostris, a new species of lancetfish from the western North Atlantic. By Robert H. Gibbs, Jr. 14 pp. March 14. No. 124. Anisian ammonoids from Malaya. By Bernhard Kum- mel. 8 pp., 1 pi. March 15. No. 125. The luminous organs of Proctoporus (Sauria, Rep- tilia) — a re-evaluation. By Willard D. Roth. 12 pp. May 27. No. 126. Mid-Scythian ammonites from Iwai formation, Japan. By Bernhard Kummel and Sumio Sakagami. 11 pp., 3 pis. June 3. No. 127. Notes on the cranial anatomy of Necrolcmiir. By E. L. Simons and D. E. Russell. 14 pp. December 19. No. 128. Size of endoceroid cephalopods. By Curt Teichert and Bernhard Kummel. 7 pp. December 20. Xo. 129. Type and type locality of the Gulf Coast spiny softsheli turtle, Trionyx spinifer asper (Agassiz). By Robert G. Webb. 8 pp., 2 pis. December 21. No. 130. The mechanisms of carapacial and plastral hinges in chelonians. By R. V. Shah. 15 pp. December 22. No. 131. A second record of the fossil rodent Palustrinius Wood. By Craig C. Black. 3 pp. December 30. No. 132. The status of Sphaerodactylus pictus, with comments on the distribution of 8. sputator and S. sahanus. By Wayne King. 5 pp. December 30. 1961 No. 133. On the generic limits in the family Pilidae (Proso- branchia: Mollusca). By Edward H. Michelson. 10 pp. February 27. No. 134. Enzymatic constitution of Alsophis saliva and its bio- logical implications. By George Hegeman. 8 pp. February 28. No. 135. Notes on Hispaniolan herpetology. 2. A review of the Anolis semilineatus group with the description of A7iolis cochranae, new species. By Ernest E. Wil- liams and A. Stanley Rand. 11 pp. April 7. No. 136. Notes on Hispaniolan herpetology. 3. The evolution and relationships of the Anolis semilineatus group. By Ernest E. Williams. 8 pp. April 8. No. 137. Notes on Hispaniolan herpetology. 4. Anolis kooprnani, new species, from the southwestern peninsula of Haiti. By A. Stanley Rand. 4 pp. April 10. No. 138. Pfeiffer's unfigured species of Strophocheilus (Megal- ohnlhnus). By T. E. Crowley and T. Pain. 8 pp., 2 pis. June 14. No. 139. A new species of Sphaerodactylus from northern Haiti. By James D. Lazell. 5 pp. June 15. No. 140. A preliminary review of the Nearctic species of Siero- lomorpha (Hymenoptera). By Howard E. Evans. 12 pp. June 27. No. 141. Three new toads from South America: Bufo manicor- cnsis, Bufo spinulosus altiperuvianu^ and Bufo quechua. By Jose M. Gallardo. 8 pp., 3 pis. June 28. No. 142. Australian carabid beetles VI. The tropical and some subtropical species of Pamborus, Mystropomus, and Nurus. By P. J. Darlington, Jr. 13 pp. June 30. No. 143. Miocene lizards from Colombia, South America. By Richard Estes. 11 pp. August 20. No. 144. A large ophiacodont pelyeosaur from the Pennsyl- vanian of the Pittsburgh region. By Alfred Sher- wood Romer. 7 pp. August 21. No. 145. A new species of the cetomimid genus Gyrinomimus from the Gulf of Mexico. By Henry B. Bigelow. 2 pp. September 5. No. 146. New rodents from the early Miocene deposits of Sixty- Six Mountain, Wyoming. By Craig C. Black. 7 pp. December 14. No. 147. Australian carabid beetles VIII. Leiradira, especially the tropical species. By P. J. Darlington, Jr. 12 pp. December 15. No. 148. Australian carabid beetles IX. The tropical Noton- omus. By P. J. Darlington, Jr. 14 pp. December 18. No. 149. A preliminary study of the Silurian ceratiocaridids (Crustacea: Phyllocarida) of Lesmahagow, Scot- land. By W. D. Ian Rolfe. 9 pp. December 19. 1962 No. 150. The genus Bethylus in North America (Hymenoptera : Bethylidae). By Howard B. Evans. 12 pp. Jan- uary 5. No. 151. A new phyllocarid crustacean from the Upper Devon- ian of Ohio. By W. D. Ian Rolfe. 7 pp., 1 pi. January 12. No. 152. New Australian dacetine ants of the genera Meso- struma Brown and Codiomyrmex Wheeler (Hymen- optera-Formieidae). By Robert W. Taylor. 10 pp. January 15. No. 153. Anolis scriptiis Garman 1887, an earlier name for Anolis Icucophaeus Garman 1888. By A. Stanley Rand. 5 pp. February 15. No. 154. Notes on Hispaniolan herpetology. 5. The natural his- tory of three sympatric species of Anolis. By A. S. Rand. 15 pp. April 4. No. 155. Notes on Hispaniolan herpetology. 6. The giant ancles. By Ernest E. Williams. 15 pp. April 12. No. 156. The fossiliferous Triassic deposits of Ischigualasto, Argentina, and preliminary description of Ischigual- astia, a new genus of dicynodont. By Alfred Sher- wood Romer and C. Barry Cox. 9 pp. April 13. No. 157. A rhachitomous amphibian, Spathicephalus, from the Mississippian of Nova Scotia. By Donald Baird. 9 pp., 1 pi. May 28. No. 158. A fossil gerrhosaur from the Miocene of Kenya (Rep- tilia: Cordylidae). By Richard Estes. 10 pp., 1 pi. May 29. No. 159. Age in a small sample of bliiefish {Pomatomus salta- trix (Linnaeus)). By Richard H. Backus. 4 pp. May 31. No. 160. Tvv'O new arthropod carapaces from the Burgess shale (Middle Cambrian) of Canada. By W. D. Ian Rolfe. 9 pp., 1 pi. June 12. No. 161. A comparative study of the respiratory muscles in Chelonia. By R. V. Shah. 16 pp. July 16. No. 162. Australian carabid beetles X. Bemhidion. By P. J. Darlington, Jr. 12 pp. July 25. No. 163. New worm-lizards { An cylo cranium and Amphisbaena) from southeastern Tanganyika Territory. By Arthur Loveridge. 6 pp. July 26. No. 164. Notes on the herpetology of Hispaniola. 7. New ma- terial of two poorly known anoles : Anolis monticola Shreve and Anolis christophei Williams. By Ernest E. Williams. 11 pp. August 22. No. 165. An extinct solenodontid insectivore from Hispaniola. By Brjmn Patterson. 11 pp. August 22. No. 166. Lectotypt's of species of Ogcocephalidae selected from syntypes in the Museum of Comparative Zoology. By Margaret G. Bradbury. 4 pp. September 5. No. 167. BatJiyclupea schroederi, a new bathyclupeid fish from the western tropical Atlantic. By Myvanwy M. Dick. 4 pp. September 5. No. 168. Two new species of fossil talpid insectivores. By Katherine M. Reed. 6 pp., 1 pi. September 7. No. 169. New records of inshore fishes from the Atlantic coast of Panama. By Ira Rubinolf and Roberta W. Rubi- noff. 7 pp. October 15. No. 170. The brain of the emu {Droniacns nouaehollandiae, Lath). I. Gross anatomy of the brain and pineal body. By Stanley Cobb and Tilly Edinger. 14 pp., 4 pis. November 16. No. 171. Notes on amphisbaenids (Ampliisbaenia; Reptilia i , 6. Redescription and range extension of Aniphishaena spurrelli Boulenger. By Carl Cans. 8 pp., 3 pis. December 14. No. 172. A new species of the rodent Pipcstoneomys from the Oligocene of Nebraska. By Raymond Alf. 7 pp. December 14. No. 173. New species of land molluslvs from the Republica Dominicana. By William J. Clench. 5 pp., 1 pi. December 24. No. 174. A new arctocyonid from the Paleocene of Wyoming. By Bryan Patterson and Paul 0. McGrew. 10 pp. December 24. No. 175. A picrodontid insectivore (?) from the Paleocene of Wyoming. By Paul 0. McGrew and Bryan Patter- son. 9 pp. December 24. No. 176. On the races of Kinixys hclliana Gray. By R. F. Laurent. 6 pp. December 27. No. 177. Rhipidistian classification in relation to the origin of the tetrapods. By Keith S. Thomson. H pp., 1 pi. December 27. No. 178. On a new species of the earthworm genus Trigaster Benham 1886 (Octochaetidae). By G. E. Gates. 4 pp. December 27. INDEX OF AUTHORS BREVIORA T MUSEUM OF COMPARATIVE ZOOLOGY Numbers 121-178 1960-62 No. Alf, Raymond 172 Backus, Richard H 159 Baird, Donald 157 BiGELOw, Henry B 145 Black. Craig C 131, 146 Bradbury, Margaret G 166 Chickering. Arthur M 121, 122 Clench, William J 173 Cobb, Stanley 170 Cox, C. Barry 156 Crowley. T. E 138 Darlington, P. J., Jk 142, 147, 148, 162 Dick, Myvanwy M 167 Edinger, Tilly 170 Estes, Richard 143, 158 Evans, Howard E 140, 150 Gallardo, Jose M 141 Gans, Carl 171 Gates, G. E 178 GiBBs, Robert II 123 Hegeman, George 134 King, Wayne 132 Kummel. Bernhard 124, 126, 128 Laurent, R. F 176 Lazell, James D 139 Loveridge, Arthur 163 McGrew. Paul 174, 175 MicHELSON, Edward H 133 Pain, T 138 Patterson, Bryan 165, 174, 175 Rand, A. Stanley 135, 137, 153, 154 Reed, Katherine M 168 RoLFE, W. D. Tan 149, 151, 160 RoMER, Alfred Sherwood 144, 156 Roth, Willard D 125 RuBiNOFF, Ira 169 RuBiNOFF, Roberta W 169 Russell, D. E 127 Sakagami, Sumio 126 Shah, R. V 130, 161 Simons, E. L 127 Taylor, Robert W 152 Teichert, Curt 128 Thomson, Keith S 177 Webb, Robert G 129 Williams, Ernest E 135, 136, 155, 164 BREVIORA Muiseiiiim of Comparative Zoology Cambridge, Mass. March 10, 1960 Number 121 THREE NEW SPECIES OF MICRATHENA (ARANEAE, ARGIOPIDAE) FROM SOUTH AMERICA Bv Arthur M. Chickering Albion College, Albion, Michigan During the summer of 1959 I had the privilege of examining the South American Mierathenae in the Museum of Comparative Zoology at Harvard College. Among the specimens in this col- lection I found representatives of what I am compelled to regard as new species. I am describing these in this brief paper under the names : Micrathena hamata sp. nov. ; M. lata sp. nov. ; and 31. shealsi sp. nov. I have also added a few notes together with figures illustrating features of the epigynum of M. fissispina (C. L. Koch), 1836, with the hope that they will aid somewhat in a more precise identification of this species. The types of the new species are deposited in the Museum of Comparative Zo- ologv. It is with pleasure and a deep sense of gratitude that I again acknowledge my indebtedness to members of the staff of the Mu- seum of Comparative Zoology at Harvard College for their con- tinued encouragement over a period of more than twenty-five years. Persons now active on the staff of the museum and chiefly responsible for this encouragement may be specifically named as follows: Dr. A. S. Romer, Director; Dr. P. J. Darlington, Jr., Curator of Insects; Dr. Herbert W. Levi, Associate Curator of Arachnology; Miss Nelda E. Wright, Editor of Publications. Several other members of the museum staff have also greatly aided me by providing privileges of the laboratories, collections, and library. BREVIORA No. 121 MiCRATHENA FISSISPTNA (C. L. Koch), 1836 (Figures 1-3) Acrosoma fissispina C. L. Koeh, 1836 Plectnna fissispina Walckenaer, 1841 M. fissispina Simon, 1895 M. fissispina Reimoser, 1917 M. fissispina Roewer, 1942 M. fissispina Bonnet, 1957 There are three mature females in the Nathan Banks Collection in the Museum of Comparative Zoology. These are simply la- belled as having come from Brazil with no dates of collection External Anatomy of Micrathena Figures 1-3. M. fissispina (C. L. Koch). Figs. 1-3. Epigynum: from below; profile, right side; posterior surface, respectively. assigned. The abdominal spination is a fairly reliable feature for identification, but the epigynum is also important. Since this organ appears somewhat different than represented in Reimoser's (1917) figures I am offering three more drawings with the hope that they will help others to a more exact identification of this species. The male remains unknown. 1960 NEW SPECIES OF MICRATHENA 3 MiCRATHENA HAMATA sp. nov. (Figures 4-7) The two niature males treated in tliis part of this paper were found filed in the Banks Collection in the Museum of Compara- tive Zoology with females considered to belong to M. petersi (Tacz.). The females are probably immature specimens of M. sc.rspinosa (Hahn). The males appear to be new to science and. consequently, one has been selected as the holotype and is here- with described in accord with my usual procedure. External Anatomy of Micrathena Figures 4-7. M. hamata sp. nov. Fig. 4. Body of male, dorsal view. Fig. 5. Left palpal tarsus and tibia. Fig. 6. Palpal tarsal hook ; nearly retrolateral view. Fig. 7. Ibid., a different view. 4 BREVIORA No. 121 Male holotype. Total length 4.68 mm. Carapace 1.92 mm. lonj;, 1.105 mm. wide opposite second coxae where it is widest; .4r)5 mm. tall ; nearly level from behind PME to beginning of posterior declivity. Median thoracic fovea a shallow longitudinal depres- sion. Dorso-lateral foveae lacking. Eyes. Eight in two rows as usual in the genus; lateral ocular tubercles weakly developed; median ocular tubercles bearing ME quite pronounced. Viewed from above, both rows strongly recurved; viewed from in front, anterior row nearly straight, posterior row moderately procurved; central ocular quadrangle wider behind than in front in ratio of 22 : 17, about as long as wide l)ehind. Ratio of eyes AME : ALE : PME : PLE = 6 : 6.5 : 8 : 5.5 (long diameter used when there are ditferences). AME separated from one another by four-thirds of their diameter, from ALE by nearly five-halves of their diameter. PME sepa- rated from one another by nearly three-halves of their diameter, from PLE by about three-halves of their diameter. Laterals separated from one another by two-thirds the diameter of AME. Clypeus strongly receding; height equal to about twice the di- ameter of AME. Chelicerae, Maxillae, and Lip. Apparently quite normal for males in the genus. Fragility of specimen precludes examination of fang groove for marginal teeth. Sternum. Very rugulose ; with a marked transverse groove between third and fourth coxae ; fourth coxae nearly touching. Legs. 4123. Width of first patella at "knee" .il913 mm., tibial index of first leg 11. Width of fourth patella at "knee" .12996 mm., tibial index of fourth leg 13. (All measurements in millimeters) Femora Patellae Tibiae Metatarsi Tarsi Totals 1. .990 .374 .748 .594 .418 3.124 2. 1.012 .396 .660 .5.50 .396 3.014 3. .704 .264 .396 .330 .352 2.046 4. 1.078 .330 .706 .660 1.450 3.224 Palp .400 .220 .330 2.682 1.632 1 Estimated because of loss of both fourth tarsi. 2 Including tarsal hook. 1960 NEW SPECIES OF MICRATIIENA 5 There is no ventral hook on first coxa nor any corresponding proximal, prolateral femoral ridge and groove on the second feninr. There also seems to be a complete lack of modified spines on the legs. Palp. Essential features shown in Figures 5-7. The tarsal hook and tibia appear to be quite distinctive. Abdomen. General shape as shown in Figure -i. It seems high- ly probable that the female of the species wull be found to have a series of prominent paired spines, remains of which seem to be present in the male. Color in alcohol. Carapace a nearly uniform rich reddish brown. Sternum somewhat lighter. Legs and mouth parts with various shades of brown. Abdomen: Dorsum yellowish along each dorso-lateral margin; broadly and irregularly brownish in the middle with a central lighter irregular stripe and some indi- cation of reduced black spots; venter irregularly brownish, yel- lowish and nearly black. Type locality. Holotype male and one paratype male in the Nathan Banks Collection from Para, Brazil. No date of collection is given. The female is unlaiown. MiCRATHENA LATA Sp. nOV. (Figures 8-12) The specimen described below was filed in the Reimoser Col- lection in the Museum of Comparative Zoology at Harvard Col- lege as M. digitata (C. L. Koch). This is obviously an error and. since I can find no record of it in the literature I am compelled to regard it as new and am describing it as such. Female holotype. Total length from anterior margin of base of ehelieerae to posterior border of abdomen 7.15 mm. Width of abdomen at base of anterior spines 2.925 mm.; width between tips of larger posterior spines 13.325 mm. Carapace smooth; with median thoracic fovea a well defined pit; not markedly raised behind median fovea ; with no dorso-lateral foveae. Eyes. Eight in two rows as usual in the genus. Viewed from above, both rows moderately recurved; viewed from in front, anterior row nearly straight, posterior row gently procurved. Central ocular quadrangle wider behind than in front in ratio of about 3 : 2 ; wider behind than long in ratio of 9 : 7. Ratio BREVIORA No. 121 of eyes AME : ALE : PME : PLE = 8 : 8 : 10 : 8.5. AME separated from one another by five-fourths of their diameter, from ALE by six times their diameter. PME separated from one another by three-halves of their diameter, from PLE by nearly six times their diameter. Laterals separated from one another by three-eig-hths of the diameter of ALE. Height of elypeus equal to about seven-fourths of the diameter of AME. Chelicerae. Normal to genus ; unable to observe teeth along fang groove because of fragility of holotype. External Anatomy of Micrathena Figures 8-10, M. lata sp. nov. Fig. 8. Outline of body of female holotype; dorsal view. Figs. 9, 10. Epigynum; profile, right side and posterior surface, re- spectively. Lip and Maxillae. Normal to genus; details considered unim- portant in the description. Sternum. Extended ; with deep lateral notches and with six exaggerated marginal tubercles; with posterior end extended between fourth coxae as a prominent tubercle nearly one-third as long as entire sternum. Legs and Spines. Only two legs remain entire, hence details not recorded. In general quite normal to genus. True spines 1960 NEW SPECIES OF MICRATHENA rare ; the usual numerous setigerous tubercles moderately well developed. Abdomen. Extraordinarily broadened posteriorly ; with a pair of nearly straight anterior marginal spines extended nearly to PME ; with a pair of very small lateral marginal spines ; tbo posterior end is broadly bifurcated and each fork is subdivided into two spines thus making a total of eight. Epigynum. General features shown in Figures 9, 10, and 12. External Anatomy of ilicraihena Figures 11 and 12, M. Juta sp. uov. Fig. 11. Posterior end of female holotypL'. Fig. 12. Epigynum, from below. Color in alcohol. Light reddish brown with dark streaks and irregular spots. Tyi^e locality. Holotype female from Theresapolis, Brazil, with no date of collection given. There are no paratypes and the male is unknown. BREVIORA No. 121 MiCRATHENA SHEALSI Sp. nOV. (Figures 13-17) Female holotype. Total length from AME to tip of posterior spines 9.23 mm. Carapace 2.79 mm. long ; 2.08 mm. wide opposite second coxae where it is widest; about .98 nnn. tall behind well External Anatomy of Micrathena Figures 13-17, M. sheaJsi sp. nov. Fig. 13. Outline of female holotj'pe, dorsal view. Fig. 14. Eight lateral side of abdunien to show spines on bifurcation. Figs. 15-17. Epigynum : from lielow, riglit lateral side, and from posterior view, respectively. defined median fovea where it is strongly gibbous; with three pairs of dorso-lateral foveae as represented in Figure 13. Eyes. Eight in two rows as usual ; viewed from above, both rows moderately recurved ; viewed from in front, both rows Ii)(i0 NEW SPECIES OF MICRATHENA 9 fjently procurved, measured by centers. Central ocular quad- rangle wider behind than in front in ratio of about 8 : 7, only slio'htly wider behind than long. AME separated from one another by their diameter, from ALE by nearly 3.5 times their diameter. PME separated from one another by their diameter, from PLE by three times their diameter. Ratio of eyes AME : ALE : PME" : PLE = 10 : 10 : 12 : 7 (long diameters used when differences exist) . Laterals separated from one another by nearly the radius of iVLE. Height of elypeus equal to nearly 1.2 times the diameter of AME. Chelirerae. In general, normal to the genus; promargin of fang groove with three teeth, the middle one bidental ; retro- margin with three large teeth. Maxillae. Apparently completely typical of the genus in all observed features. Lip. Wider than long in ratio of about 4:3; transversely grooved in basal third. Sternal suture procurved. Sternum. Elongated scutiform; moderately convex; only with well developed antero-lateral tubercles: not continued between fourth coxae ; sparsely supplied with moderately long stiff bristles. Legs. 4123. Width of first patella at ''knee" .325 mm., tibial index of first leg 11. Width of fourth patella at "knee" .303 nun., tibial index of fourth leg 11. (All measurements in millinieters) Femora Patellae Tibiae Metatarsi Tarsi Totals 1. 2.860 .990 2.100 1.950 .910 8.810 o 2.600 .975 1.820 1.755 .845 7.995 3. 1.625 .650 .980 1.040 .715 5.010 4. 3.250 .780 2.015 2.080 .910 9.035 Spines on legs apparentlj^ unnoteworthy ; many are lost and scars are difficult to locate with certainty. Abdomen. General features essentially as shown in Figures 13 and 14. Moderately flattened. With no anterior marginal spines ; with a pair of lateral dorsal spines of moderate length ; bifurcated posteriorly with each fork subdivided into two spines with the larger below. 10 BREVIORA No. 121 Epigynum. Unlike that seen in any other species; anterior border of modified portion a granulated rim ; features essentially as shown in Figures 15-17. Color in alcohol. Legs and mouth parts a dull reddish brown, lighter beneath. Sternum light brownish. Carapace with a brownish central stripe and a broad dark brown stripe on each side (both represented by stippling in Figure 13) ; remainder of central region yellowish. Abdomen : irregularly yellowish white dorsally with nearly black margins ; venter light yellowish from genital furrow to base and lateral sides of cone surrounding spinnerets ; the cone with a nearly black circular ring ; remainder of venter irregularly dark brown or black with yellowish spots and streaks. Probably some loss of coloration from long preser- vation. Type locality. The holotype was simply labelled : Argentine, Sunchal, with no date of collection given. Apparently Cockerell was the collector. The species is named in honor of Dr. J. G. Sheals, Department of Zoologj^ British Museum (Natural His- tory) . BIBLIOGRAPHY Bonnet, Pierre. 1957. Bibliographia Araneorum. Tome II, 3nie partie. T.ps Artisans de I'Imprimerie Douladoure, Toulouse, France. Cambridge, 0. P. and F. P. Cambridge. 1889-1905. Arachnida-Araneida. Vols. I-II. In: Biologia Centrali- Amerieana. Dulau & Co., Londnn. Kfyserling, Graf Eugen von. 1892. Die Spinnen Ameiikas. Epeiridae. Verlag von Bauer & Raspe, Niirnberg, Germany. Koch, C. L. 1836. Die Arachniden. Niirnberg, vol. Ill, pp. 1-119. Reimoser, Eduard, 1917. Die Spinnengattung Micrathena. Verh. Zool. Bot. Ges. Wien, vol. 67, pp. 73-160, pis. 1-9, figs. 1-31. I960 NEW SPECIES OF MICRATHENA 11 ROEWER, C. Fr. 1942. Katalog der Araneae. Vol. 1. Kommissions Verlaj,' von "Natura." Bremen. Simon, Eugene. 1892-1903. Histoire Naturelle dcs Araignees. Deuxieme Edition. 2 Vols. Librarie Eneyclopedique de Roret, Paris. Walckenaer, Ch. a. 1841. Histoire Naturelle des Insectes. Apteres. Paris, vol. II. / BREVIORA Mmseiiiinti of Comparsitive Zoology Cambridge, Mass. March 11, 1960 Number 122 NOTES ON CERTAIN SPECIES OP MICRATHENA (ARANEAE, ARGIOPIDAE) FROM SOUTH AMERICA By Arthur M. Chickering Albion College, Albion, Michigan During a period of work in the Museum National d'Histoire Naturelle in Paris during the summer of 1958 Dr. Herbert W. Levi, Associate Curator of Arachnology in the Museum of Com- parative Zoology at Harvard College, examined types of nine species of Micrathena Sundevall 1833, all originally described from South America by the great arachnologist, Eugene Simon. All of these species are poorly known and most of them have not appeared in collections since the originals were studied by their author. All were briefly described in 1896 ; four were mentioned in 1895 and accompanied by five simple figures. During the examination of the types mentioned Dr. Levi made free-hand drawings of the dorsal surface of the abdomens to show general form and spination. He also made careful drawings of the exter- nal genitalia when the latter were available ; these were made with the use of a reticule with squares. All of the drawings made by Dr. Levi were turned over to the author to use as he saw fit in connection with his study of the genus. The figures appearing in this paper were made directly from Dr. Levi's original pencil drawings. The outline figures of abdomens are freehand copies with enlargement; the drawings of genitalia were made with tracing paper directly from Dr. Levi's originals. It has seemed worth while to present these data, thus obtained, with the hope that they will be of some help to others who may continue the study of this most interesting genus. BREVIORA No. 122 MiCRATHENA AcicuLATA Simon, 1897 (Fiorure 1) M. acicvlata Petrunkevitch, 1911 M. aciculata Roewer, 1942 M. aciculata Bonnet, 1957 The type is an immature female from Venezuela. Apparently the species has not been reported in collections since the original was taken. The general form of the abdomen with its spination, External Anatomy of Micrathena Figure 1, M. aciculata, abdomen, dorsal view. Figures 2, 3. M. gaujoni, abdomen, dorsal view and epigynum, respectively. Figures 4, 5. M. hamifera, abdomen, dorsal view and epigynum, respectively. seen in dorsal view, is shown in Figure 1. The stippled area is black and the posterolateral corners are white. The male is unknown. Micrathena gaujoni Simon, 1897 (Figures 2, 3) M. gaujoni Petrunkevitch, 1911 M. gaujoni Reimoser, 1917 1960 CERTAIN SPECIES OF MICRATHENA 3 .1/. gaujoni Roewer, 1942 M. gaujoni Bonnet, 1957 Simon stated that the type was 8.7 mm. long and similai* to J/, fissispina (C. Koch), but this hardly seems correct. There are four ])airs of spines, tli*' second pair the longest (Fig. 2). The epigynum has a pair of depressions directed posteriorly be- neath an overhanging rim (Fig. 3). The type is from Ecuador and the male is unknown. MiCRATHENA HAMIFERA SilllOll, 1897 ( Figures 4, 5 ) The length of the holotype was given as 9 mm. The general form of the abdomen with its spination is shown in Figure 4. The dorsal surface is like white enamel in general appearance. Figure 5 shows the form of the epigynum from "slightly be- hind"; just posterior to the curved boundary there is a large ' ' sclerotized knob. ' ' The female is known only from Peru and the male is still unknown. ^MiCRATiTEXA niBELLis Simoii. 1895 ( Figure 6 ) .1/. imbeUis Simon, 1897 M. imhclUs Petrunkeviteh, 1911 M. imhellis Eeimoser, 1917 J/, imhellis Roewer, 1942 M. imhellis Bonnet, 1957 Simon (1895) included a tigure showing the right side of the abdomen with no spines. Reimoser (1917) just mentioned the species and did not include it in his further treatment of the genus. The general appearance of the dorsal surface of the abdomen is shown in Figure 6; the stippled areas in the figure are black in the type. Dr. Levi has determined that the type is an immature female from Venezuela. The male is unknown. MiCRATHENA PERLATA Simon, 1895 (Figures 7, 8) M. perlata Simon, 1897 M. perlata Petrunkeviteh, 1911 M. perlata Reimoser, 1917 BREVIORA No. 122 M. perlata Eoewer, 1942 .1/. perlata Bonnet, 1957 The length of tlie female type is given as 6 mm. Dr. Levi has found that it also is immature. The general appearance of the dorsal surface of the abdomen is shown in Figure 7. The im- mature female is accompanied by a male which may or may not External Anatomy of Micrathena Figure 6. M. imbellis, abdomen, dorsal view. Figures 7, 8. M. perlata, abdomen, dorsal view and male palp, respectively. Figures 9, 10. .1/. pubescons, abdomen, dorsal view and male palp, respec- tively. be properly paired with it. Figure 8 shows certain features of the palpal tarsus. The specimens are simply labelled "ximazon." Simon (1895) stated that the type came from: "Brasilia: S. Paulo de Olwenea (de Mathan)." 1960 CERTAIN SPECIES OF MICRATHENA 5 MiCRATHENA PUBESCENS Simoii, 1895 (Figures 9, 10) M. i)ubescens Siniou, 1897 M. pubt'A-cciis Petrunkevitch, 1911 M. pubescens Reimoser, 1917 M. pubescens Roewer, 1942 M. pubescens Bonnet, 1957 The female type is immature. The abdomen is hairy and sug- gests a close relationship with M. furcula (0. P. Cambridge) from Central America. There are no true spines but the posterior end of the abdomen is somewhat bifurcate. Simon (1895) fur- nished figures to show the right side of the abdomen and the bifurcate posterior end. The immature female is accompanied by a mature male the palpal tarsus of which is shown in one position in Figure 10. Caution must always be exercised in matching the sexes in this genus and this male should be very carefully studied and compared with the growing number of different kinds of known males. The specimens are from Matto Grosso, Brazil. MiCRATHENA PUPA Simon, 1897 (Figures 11, 12) M. pupa Petrunkevitch, 1911 M. pupa Reimoser, 1917 M. pupa Roewer, 1942 il. pupa Bonnet, 1957 The length of the type is given as 8 mm. The form of the abdo- men and its spination are shown in Figure 11 and some of the features of the epigynum in Figure 12. The type female is from Ecuador. Simon apparently had a male associated with the female but Dr. Levi did not find it in the collection. Presumably it is lost and the original description does not give what we now consider to be the important diagnostic male features. MiCRATHENA TOVARENSIS SiuiOU, 1897 (Figures 13, 14) .1/. tucarensisFetrunkevitch, 1911 M. tovarensis Reimoser, 1917 M. tovarensis Roewer, 1942 il. tovarensis Bonnet, 1957 6 BREVIORA No. V2-2 The length of the female type is given as 7.8 mm. The general appearance of the abdomen and its spines, as seen in dorsal view, are shown in Figure 13. The appearance of the epigynum as shown in Dr. Levi 's drawing is given in Figure 14. The type is from \'enezuela. The male is unknown. External Anatomy of Micrathena Figures 11, 12. Al. pupa, abdomen, dorsal view and epigynum, respectively. Figures 13, 14. If. tovarensis, abdomen, dorsal view and epigj^num, respec- tively. Figure 15. M. xantliopyga. abdomen, dorsal view. Micrathena xaxthopyga Simon, 1895 (Figure 15) M. xanthopyga Simon, 1897 M. xanthopyga Petrunkevitch, 1911 1960 CERTAIN SPECIES OF MICRATHENA 7 M. xanthopyga Reimoser, 1917 M. xanthopyga Roewer, 1942 M. xanthopyga Bonnet, 1957 The type is an immature female whose general appearance in dorsal view is shown in Figure 15. Simon (1895) published a figure of the type viewed from the right side. This figure shows four pairs of spines instead of three, as shown in Dr. Levi's drawing. The type is from Venezuela. The male is unknown. BIBLIOGRAPHY Bonnet, Pierre. 1957. Bibliographia Araneorum. Tome II (3nie partie:G-M). Les Artisans de L 'Imprimerie Douladoure, Toulouse. Petrunkevitch, Alexander. 1911. A Synonymic Index-Catalogue of Spiders of North, Central, and South America, etc. Bull. Amer. Mus. Nat. Hist., vol. 29, pp. 1-809. Reimoser, Eduard. 1917. Die Spinnengattung Micrathena. Yerh. Zool. Bot. Ges. Wien, vol. 67, pp. 73-160, pis. 1-9, figs. 1-31. ROEWE}R, C. Fr. 1942. Katalog der Araneae. Yol. 1, Bremen. Simon, Eugene. 1895. Histoire Naturelle des Araignees. Yol. 1, Pts. 3-4. 1897. Descriptions d'Espeees de I'Ordre des Araneae. Annales Societe Entomologique de France, vol. 65, pp. 465-510. BREVIORA MTuseiaim of Comparative Zoology Ca^ibridge, Mass. March 14, 1960 Number 123 ALEPISAURIS BREYIROSTRIS, A NEW SPECIES OF LANCETFISH FROM THE WESTERN NORTH ATLANTIC By Robert H. Gibbs, Jr. Department of Biology, Boston University Exploratory fishing for tunas in the western North Atlantic by the Fish and Wildlife Service vessel DELAWARE has re- vealed the presence of considerable numbers of lancetfishes, oenus Alepisaurus. Through the courtesy of Mr. James L. Squire, I have been privileged to accompany most of these cruises. On the first two, the presence of Alepisaurus was recorded and stomach contents sampled, but only two specimens were saved. During the third cruise, it became apparent that two morphological types were represented, and from then on an attempt was made to measure and make counts on all possible specimens, and to pre- serve a large sample. At first the possibility was entertained that the differences might be due to sexual dimorphism. Gonads were therefore examined on all preserved specimens and on a large number of fresh ones. The appearance was almost exactly the same in all specimens, indicating that sexual dimorphism was probably not a factor. This has been borne out by histological studies of the gonads, which lead me to the rather surprising conclusion that both morphological types, which I am now certain represent valid species, are hermaphroditic. A study of nearly all type specimens and of all original de- scriptions leads to the conclusion that all previously described Atlantic species are conspecific with A. ferox Lowe. The second form is described here. BREVIORA No. 123 Alepisaurus brevirostris, sp. nov. (Figures 1-2) Holotype. U. S. National Museum 186197, 682 mm. in standard length when fresh, 684 mm. preserved ; taken on longiine with 20-fathom buoy lines bv the M/Y DELAWARE at 38° 49' N, 64° 02' W, on September 13, 1957. /,A//^/////^///A. Figure 1. Left lateral view of Alepisaurus hrevirostris (top), and A. ferox ( bottom ). Paratypes have been distributed to the Museum of Compara- tive Zoology, Academy of Natural Sciences of Philadelphia, Cornell University, University of Miami Marine Laboratory, Tulane University, Scripps Institution of Oceanography, Stan- ford Natural Historv Museum, Museum of Zoology L'niversity 1*J60 ALEPISAURUS BKEVIROSTKIS 3 of Michigan, California Academy of Sciences, Britisli ^Museum (Natural History), Museum National d'Histoire Naturelle Paris, and Museu Municipal do Funchal. Diagnosis. A dark-hued species of Alcpisaunis with a grad- ually arcuate dorsal fin profile, the dorsal origin Avell forward of the rearmost margin of the operculum, a short head (6.5 or more in standard length), and a short snout (2.5 or more in head). Description. Dorsal fin high, originating over the middle of the opercle, its longest ray (about number 25-30) about three times the greatest depth of the body ; its rays flexible, easily bent and broken, not branched, joints difficult to observe, though present. Leading dorsal ray thickened, its anterior edge finely serrated. Anal fin highest at the second and third rays, the rays branched distally. Pectoral fin pointed, its middle rays longest ; first ray thickened, serrated anteriorly, unbranched, the rest branched and obviously jointed. Pelvic fin also pointed, the middle rays longest, the first ray thickened, serrated anteriorly, and unbranched, the remaining rays jointed and branched. Caudal fin strongly forked, with eight procurrent rays in each lobe, ten upper principal rays and nine lowers ; uppermost prin- cipal rays considerably elongated in some specimens. Snout inclined downward more sharply than the rear of the head, its length more than 2.5 times in head length. Nostrils a little less than halfway from tip of snout to anterior edge of orbit, the anterior opening round, the posterior crescent-shaped. Upper and lower jaws subequal. Tapper jaw with a row of many small thin teeth on the premaxilla, one or two large fangs on anterior palatine, about three smaller fangs on the rear of the same bone, followed by about 7-10 low, triangular teeth. Lower jaw with an anterior large fang, followed by about 10 small caniniform teeth, one to three large fangs, and about 10-15 low, triangular teeth. No teeth on poorly developed tongue. Two patches of pharyngobranchial teeth. Gill arches with 3-6 upper, 0-1 middle, 17-23 lower groups of low, spinous rakers, totaling 23-28 groups. Branchiostegals mostly 7. Eye about 5 in head length, with vertical adipose eyelids anteriorly and posteriorly. Preopercle smooth. Opercle large, sculptured with lines radiat- ing from its antero-dorsal corner ; subopercle also with striae 4 BREVIORA No. 123 radiating' from its anterior point ; interopercle apparently absent. Interorbital space bonnded by prominent parietal ridg'cs, tlie area forming a flat to slightly concave surface, gradually widen- ing posteriorly. Body elong-ate, its greatest depth, at level of pectoral fins, about 12 in standard length. A low lateral keel occupying most of the rear half of each side. Lateral-line pores opening along the keel and continuing forward l)eyond it ; lateral line on head forming prominent supraorbital, suborbital, and preoperculo- mandibular systems; supratemporal apparently lacking. Anus posterior to pelvic insertion by less than half the length of the pelvic fins. Fin-ray counts and morpliometric data are given in Tables 1 and 2. ' Coloi-afioii. Body iridescent brownish-black dorsally, becom- ing gradually lighter laterally. Region al)Ove lateral line liberal- ly sprinkled with both large and small melanophores, many of the former ocellated. Below the lateral line, and particularly on the belly, many small melanophores present. General coloration decidedly dark in comparison with A. ferox. Lateral keel black. Dorsal fin membrane iridescent black, often with a horizontal row of white spots a short distance above base. Other fins, including adipose dorsal, black. Head dark above, becoming lighter ventrally. Abdominal cavity marked externally l)y alter- nating light and dark bands, the light ones representing strips of muscle, between which the dark peritoneal lining shows through. Visceral Anatomu. Peritoneal lining black. Liver relatively small, covering only the most anterior portions of the stomach and intestine. Stomach black, highly distensible, forming a long, blind sac. Intestine arising at the anterior end of the stomach, continuing straight, without bends, to the anus, divided at al)out one-third of its length into anterior thick-walled and posterior thin-walled portions. Kidneys occupying the entire length of the body cavity, lying retroperitoneally along the ventral side of the vertebral column. Ureters enter a thin-walled urinary bladder which extends about from the level of the pelvic insertion to the anus. Gonads consisting of a prominent pair of elongate, con- tinuous ovaries, which occupy the posterior third of the body 1!)()0 ALEPISAFRT^S BREVIROSTRIS 5 cavity above the intestine, and a pair of thin testes, ahnost in- visible, lying in the dorsal groove formed by the two ovaries. The dncts of the ovaries, and presumably also those of the testes, join the urinary bladder and open by a urogenital pore immediately behind the anus. Swinibladder absent. Related Species. Other than A. hrevirostris, the only recog- nized Atlantic species of the genus Alepisaurus is A. ferox Lowe. The two species are distinguishable by many characters. The most trenchant ones are shown in Table 3, and may be visualized in Figures 1 and 2. In addition to these, many less-perfect ones are demonstrable. Characters associated with head length show significant differences (snout to dorsal origin, snout to pectoral insertion, see Table 2). The pectorals average about one-fifth of the standard length in A. hrevirostris and are slightly shorter in A. ferox. The eye is relatively larger in A. hrevirostris, doubt- less correlated with the shorter snout. Meristic characters show consistent modal differences : dorsal rays most commonly 42-45 in A. hrevirostris, 39-42 in A. ferox; anal raj^s usually 14-15 vs. 15-17 ; pelvic rays 13-14 vs. 14-15. Alepisaurus is common in the Pacific, but at present it is not possible to ascertain the species. I have examined seven Pacific specimens in the U. S. National IMnseum and find them extremely similar to, if not conspeeific with A. ferox (see Table 1). The only disconcerting element was the low number of dorsal rays in four specimens, which indicates at least some degree of dif- ferentiation. I have seen no examples of a form resembling A. hrevirostris from the Pacific. Young Specimens. Five specimens, 85.0 to 190.5 mm. standard length, present a confusing array of characters which defy posi- tive identification. Tables 1 and 2 show for these specimens many characters within or beyond the ranges displayed by adults of both A. hrevirostris and A. ferox. There is obviously a great change between 200 and 500 mm. in the relative proportions, the caudal end in particular increasing proportionally greatly in this time. Presumably the shape of the head also changes, as these five are all extreme, even to A. hrevirostris, in having short heads and snouts. I am inclined to call them all A. ferox, but with considerable doubt. BREVIORA No. 123 be p ™ =t-l K O ^ o be OD O c bJD F^' 1960 ALEPISAURUS BREVIROSTRIS 7 3Iensural Discrepancies. Counts and a few selected measure- ments were made at sea on most of the specimens which came aboard the M/V DELAWARE. A later check has shown that the measurements cannot be used in conjunction with others taken on preserved specimens. Among the 17 preserved speci- mens of each species used in the descriptions here, were 11 A. ferox and 16 A. hrevirostris which were also measured at sea when fresh. In these, the measurements of standard length and head length were consistently less in preserved specimens. In A. ferox, the standard length of one specimen was 0.7 per cent greater, the others 1.0-16.2 per cent less, averaging 6.6 per cent less. The head lengths were 3.4-6.6 per cent less, averaging 5.1 per cent. The corresponding figures for A. hrevirostris w'ere : standard length 0.3 and 1.1 per cent greater in two, others 2.0- 13.1 per cent less, averaging 4.5 per cent less ; head length 0.5-2.8 per cent greater in three, 0.5-9.2 per cent less in 13, averaging 2.7 per cent less. Synonymy. Previous descriptions of Alepisaurus leave much to be desired. Eight nominal species have been described. I be- lieve all these descriptions refer to a single, perhaps polytypic, species. These names are discussed below. Alepisaurus ferox Lowe, 1833. The original description was made from two specimens from Madeira (Lowe, 1835a). In all important characteristics the written description is nearest A. ferox as recognized in the present study, but the accompanying drawing shows an arcuate dorsal fin profile, an absolute character of A. hrevirostris. The matter was further complicated by a description and drawing of a third specimen from Madeira (Lowe, 1835b) which shows a dorsal fin profile characteristic of A. ferox. N. B. Marshall has kindly examined two types in the British Museum. His observations leave no doubt that both are A. ferox as I understand it. In the specimen labeled "TYPES!" (no registered number) the standard length is 1125 mm., head length (to tip of lower jaw, as upper is damaged) 191 mm. (17%), dorsal rays 39 or 40, anal 17, pectorals 14. In the specimen labeled "SYNTYPE" (number 1852.9.13.98) the standard length is 1225 mm., head length 206 mm. (17%) ; snout 90 mm. (44% of head) ; dorsal 41, anal 17, pectorals 15. In both, the 8 BREVIORA No. 123 dorsal origin is over or sliglitly Ix'liiud the posterior edge of the operculum and no large, oeellated melanophores are present; both are larger than any A. hrevirostris I have seen. The name Alepisauriis ferox is thus reasonably established for the long- snouted species. Alcpisaitnis aznreus Valenciennes, 1849. No type is extant, but the following parts of the original description strongly sug- gest A. ferox: ". . . la dorsale est d'egale hauteur jusque vers le trentieme rayon ..." (Cuvier and Valenciennes, 1849 : 531) ; dorsal rays 38, anal 16. The pectoral count of 10 is presumably in error. The length of 5 feet 3 inches is larger than any known A. hrevirostris. The description was based on a specimen from the Canary Islands. Alepisaurus richardsonii Bleeker, 1855. Based on the descrip- tion by Sir -John Richardson (1844) of a head from Van Diemens Land. The drawing of the head shows a snout most like A. ferox. Alepisaurus alfiveJis Poey, 1861. It is difficult to be certain of this description. The anterior rays are all said to be the same height, the posterior ones decreasing rapidly ; dorsal rays 40, anal 17, pectorals 16 ; all most like A. ferox. The pelvic count of 13 is presumably in error. Based on a Cuban specimen. Alepidosaurus horealis Gill, 1863. Based on a head, dorsal, caudal, and pelvic tins from a Pacific specimen. I have examined this specimen and find it close to A. ferox except for a dorsal count of 35. The pelvic count of 13 is apparently an error. The stated snout/head ratio of 2/5 definitely precludes A. hrevi- rostris. Alepidosaurus serra Gill, 1863. Described from a head, caudal, and pelvic fins of a Pacific specimen. I can detect no significant differences between Gill's descriptions of this species and of A. horealis. He places much emphasis on opercular sculpturing, which seems to be quite variable and not a good specific charac- ter. Again the pelvic count of 13 is presumed to be an error. The distance from eye to snout, stated as 2/5 of head length, excludes A. hrevirostris. Alepidosaurus poeiji (iill, 1863. Described on the basis of drawings (which I have not seen) of the second specimen from Cuba mentioned by Poey (1861) in his description of A. altivelis. 1960 ALEPISAL'KUS BREVIKOSTRIS 9 Tlic specinuMi was described as liaviiire la historia natural de la isla do Cuba. Havana. Vol. 2: 1-442. KicHARDSON, Sir John 1844. Ichthyology of the voyage of the HMS Erebus and Terror. Lon- don. Pp. 1-139. Table 1. Pin-ray and gill-raker eonnts of Alepisaurus Dorsal Kays Pelvic Eays 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 8 9 10 A. hrcvirontris 313 5 46111 1193 A.ferox 2 3 10 Id 10 7 4 2 1 2 43 7 Pacific (USNM) 2 2 11 3 3 Small 1 13 1 Anal Kays Pectoral Kays Gill liakers 13 14 15 16 17 18 12 13 14 15 16 23 24 25 26 27 28 29 A. brcvirostris 19 12 1 1 13 10 2 3 4 113 2 A. ferox 1 14 21 15 1 1 5 30 16 14 4 5 2 1 Pacific (USNM) 6 1 4 3 Small 3 2 2 2 1 1960 ALEPISAT'RIS BREVIROSTRIS 13 Table 2. Proportional dimensions of Alcpisaurus expressed as per cent of standard lenoth. Range, with mean in parentheses. Based on seventeen adnlt specimens of each species and five small specimens. Standard li'iiptli, iiiiii. Per (Tilt of standard Iciigtli Snout to anal origin Snout to polvie insertion Snout to pectoral insert Snout to dorsal origin Head length Greatest depth Caudal peduncle depth Pectoral length Pelvic length Anal liase Anal height Per cent of liead length Snout to tleshy orbit Snout to lioiiy oi-\nt Fleshy orbit length Bony orbit length Least interorbital width Snout to anterior nostril Xumlier of gill raker groups (total) brevirostris ferox small 551-894(670.6) 431-1088(777.5) 85.0-190..1 (144.2) 77-83(79.5) 76- 42-50(46.3) 43- 13-16(14.7) 17- 9.4-13(11.4) 16- 12-16(14.5) 16- 7.2-10.4(8.5) 8.0 1.9-3.2(2.4) 2.3 14-20(16.3) 17- 7.2-12.4(9.5) 7.5 8.5-11.4(9.8) 9.2 5.9-7.9(6.9) 7.4 88(80.1) 53(47.5) 23(19.4) 22(17.9) 23(18.6) -12.5(9.5) -4.0(2.8) 24(20.5) -10.4(9.0) -13.8(10.7) 10.3(8.6) 31-37(34.5) 26-31(28.4) 19-23(20.9) 23-33(27.8) 15-20(17.5) 14-16(14.9) 23-28(25.4) 41-46(43.3) 35-42(38.1) 13-20(17.8) 17-28(23.4) 14-18(16.2j 21-24(22.4) 23-29(25.8) 80-81(80.0) 53-58(55.3) 22-25(24.0) 21-25(23.2) 23-30(25.6) 13-16(14.4) 2.7-4.0(3.4) 15-19(17.3) 7.4-8.5(7.9) 9.5-13(11.1) 6.5-9.1(7.8) 36-41(38.4) 27-37(31.4) 25-28(26.2) 31-36(33.4) 15-17(16) 14-18(17.2) 25-27(26) 14 BREVIORA No. 123 Table 3. Principal differential characters of A. hrevirostris and A. ferox A. hrevirostris Coloration dark, more and larger melanophores, many of them ocellated (Fig-. 2) Dorsal fin gradually arcuate, without free anterior rays (Fig. 1) Irregular horizontal row of white spots often present on dorsal membrane Dorsal origin well in advance of rear margin of the opercu- lum Head 6.5 or more in standard length Snout 2.5 or more in head length A. ferox Coloration light, fewer, most- ly small melanophores, few or none ocellated (Fig. 2) Dorsal fin Avith several ante- rior rays elongated, free from membrane; rest of fin about equal in height until sudden posterior drop, or slightly liigher before drop (Fig. 1) No white spots on dorsal mem- brane Dorsal origin about level with the rear margin of the opercu- lum Head less than 6.5 in standard length Snout less than 2.5 in head length BREVIORA Mmseiiiii of Comparative Zoology Cambridge, Mass. Makch lo, i960 Xu.mi'.kk 124 ANI8IAX A:\LM()N()1DS FROM MALAYA By Bernhard Kummel Possiliforous marine strata are sparsely represented or known from Malaya. Even the Triassic system which is one of the better known systems to yield fossils is represented by extremely small faunas consisting- mainly of pelecypods, and only indeterminate ammonoids have been reported by Ne\^i:on (1923, 1925). The first discovery of determinable ammonoids was reported by Savafje (1950) from mudstones near Knala Li]iis, Pahano-. The initial collections were submitted to L. F. Spath of the British Museum (Natural History) who made the following report (in Savage, 1950) -. "Quite a number of common Middle Triassic (Anisian) genera can be recognized in the collection, including Paracerofifes (dominant), ^Sturia, Piijchifes, and Acrochordi- ccras. so that the age of the assemblage is the trinodosus zone. Specific identifications would be more difficult but are unneces- sary; Paraceratifcs trinodosus (Mojsisovics) and such close allies as the Himalayan Ceraiiics fhnilleri (Oppel) and P. winterhot- tomi (Salter) are probably all represented." The rarity of Triassic ammonoids in Southeast Asia, lying as it does at the eastern end of Tethys between the richly fossil- iferous Triassic horizons of the Himalayas and the island of Timor, warrants a more substantial record of these faunas than Spath Avas able to give. Through the kindness of Dr. M. K. Howarth of the British Museum (Natural History) all of the best preserved material from Kuala Lipis, Pahang, was loaned to the writer. Close examination of this small collection (29 specimens) showed that Spath 's conclusions as to genera present and age assignment are correct in spite of the rather poor preservation. 2 BREVIORA No. 124 The fauna contains the following species : Paraceratites trinoclosus (Mojsisovics) Sturia sanftovinii Mojsisovics Acrochordiceras sp. indet. Ptychitcs sp. indet. Data on the geographic and geologic occurrence of this fauna can best be quoted from Savage (1950, p. 76) : "One of the newly recorded areas is some 10.5 miles south-south-Ave.st of Kuala Lipis on one of the branches of the Sungei (= River) Tua, where it flows through Budu Estate (approximately lat. N. 4°02'30", long. E. 102°00'15"; Malayan Survey Department Topographi- cal Sheet No. 2 0/13) . The rocks are mudstones, rarely laminated sufficiently to be called shales. They are fairly homogeneous, slate grey, almost black and carbonaceous ; but some slightly sandier strata (muddy siltstones) weather to a pale buif or brown. The beds show minor flexures but in the main strike 40°-220° and dip to the north-west at angles of about 40°. They are strongly jointed along several directions, the two main joint systems being vertical and striking 60° -240° and 155° -335°. The beds are fossiliferous over a distance of at least 50 yds. in a horizontal direction normal to the strike, equivalent to a strati- graphical thickness of about 100 ft." No data are available with the specimens as to their precise position in the fossiliferous horizon. The species recognized in this small fauna are common forms widely distributed in Tethys and the general cireum-Pacific region. Paraceratites trinoclosus and Sturia sansovi)ui were originally established on Alpine specimens. The specimens as- signed to PtycJiitcs and Acrochordiceras are too poorly preserved to enable specific identification but there is no doubt as to the generic assignment. The two previous records of Triassie annnonoids from Malaya are not as satisfactory. A small fragment of an ammonoid was recorded by Newton (1923, p. 302, pi. 9, fig. 29); it is quite indeterminable. The specimen came from argillaceous sandstone at Mount Faber, Singapore, and was reported to have a "de- pressed whorl with indications of straight ribs and furroAvs con- necting with some well-separated knob-like tubercles situated within a short distance of the inner margin." Newton likewise !!)()() AXISIAN AAi:.ION()lI)S FROM MALAYA 3 (lid not believe the speeiiiieii w;is determinable but thoiiji'lit that it i'es(Miiblo(l the genus B(il(ih>iiil( s. There does not appear to be any .justification for this sugo'estion. The second record of am- monoid i-emains from ^lalaya is also in a paper by Xewton (1925) on a small I'l^per Triassic fanna from the Province of Kedali. The fii-st of the two specimens available to Xewton was cited as Aiiuiioinfis sp. indet. "A" and he sug-gested tliat it may be referred to a form of tlie An-estidae. Tt seems more likely that it is a Jui'drltcs or ])ossibly an A H(it<))iiif( s but the specimen lacks the eharacteristic consti'ictions of these genera (Spath, lOol, p. 10(5. footnote). The second specimen was listed by Xewton as Aninwuifcs sp. indet. "B" and suggested a resemblance to BaJafoiiih s. Spatli (1!'.")1. ]). 15) thought that this specimen might be Ihoinoceras )uisf uri in in (INIojsisovics) . in the adjoining regions of southeast Asia only Indochina has yieUled a large and varied fauna of Triassic anunonoids wliich, however, are generally not well preserved. The literature on the stratigraphy and faunas of this area is very large and need not be reviewed here. A brief summary can be found in Saurin (1956). The only other really new discovery of Triassic am- monoids in southeast Asia has been nmde in Thailand where Anisian and Karnian fannas are now known. The geology of this Triassic region has been described by Pitakpaivan (1955) who includes a preliminary list of the species present, identified by Kummel. Full description of this fanna will be published shortly. In the following description of the species from Kuala Lipis the extensive synonymies for Paraccratiics trinodosus and Sturia saiisovinii have been omitted; essentially complete synonymies can be found in Diener (1915a) and Kutass}' (1933). SYSTEMATIC DESCRIPTIONS Family CERATITIDAE Mojsisovics, 1879 Genus ParacERATITES Hyatt, 1900 Paraceratites trinodosus (Mojsisovics) Plate 1, figures 3-6 The collection contains no less than twenty crushed and in- complete specimens that can be assigned to the well known 4 BREVIORA No. 124 Paraceratites ti'inodosus (Mojsisovics). Allowing for the general faulty preservation, these specimens agree well in most details witli tlie type of this species and with other specimens assigned to it. In his preliminary examination of the fauna, Spath (in Savage, 1950, p. 76) considered that in addition to Paraceratites trinodosus the fauna also contained P. tJivilleri (Oppel) and P. winter!) ottomi (Salter). These are very closely allied forms occurring with P. trinodosns in the Himalayas. However, con- sidering the poor preservation of the specimens, it seems that a more conservative approach is desirable and I am recognizing only the better known and more widely distributed P. trinodosus. This species is particularly widespread in the Alps, Balkans and the Middle East. It is likewise recorded from the Himalayas and Nevada. Identical or closely related species are also known from Japan. This species gives its name to the upper Anisian trinodosns zone. Repository. BMNII — C 55672, C 55673, C .15674, C 55675 (figured specimens) ; C 55653, C 55654, C 55655, C 55657, C 55658, C 55661, C 55662, C 55663, C 55666, C 55667, C 55668, C 55670, C 55671, C 55676, C 55678. Family PTYCHITIDAE Mojsisovics, 1882 Genus PtYCHITITES Mojsisovics, 1875 Ptychites sp. indet. Plate 1, figure 7 The collection contains three crushed and fragmentary speci- mens that without (juestion belong in Ptijchites but identification at the specific level is not possible nor advisable. The most complete specimen is actually only the impression of one side of a conch. The illustration on Plate 1, figure 7 is of a latex cast of this impression. It shows the funnel-shaped umbilicus, broad- ly arched lateral areas, and the radial ribs — all features which are very characteristic of the genus PfijcJiites. Ptychites is known from Middle Triassic strata througliout the world. It likewise includes a very large number of species based largely on differences in shape of the conch, character of ribs, degree of involution, and details of the suture. In the Himalayan 1960 ANISIAN AMMONOIDS FROM MALAYA 5 Musc'hclkalk, PiycJiitcs, of tlie jiTuup of F. nKjifcnts (Oppelj to which these Malayan specimens most likely belong, is one of the most abundant forms present, hciiiu represented by seven species (Diener, 1895, 1907). In southeast Asia the record of Ptychites is very fragmentary and represented mostly by indeterminate species. Even the rich Middle Triassic faunas of Timor appear to have only one species, P. aniarassicus Welter (1915; Arthaber, 1928). The genus is, however, also present in Thailand and Indo- china. It is also known to be present in Japan and New Zealand. Repository: BMNH — C 55659 (figured specimen) ; C 55664, C 55656. Genus Sturia Mojsisovics, 1882 Sturia sansovinii Mojsisovics Plate 1, figure 2 The most easily recognizable species in the collection is this strigate form which is widely distributed throughout Tethys. The specimen consists only of the impression of slightly more than one third volution of one side of a w^horl ; no suture or whorl section is preserved. In spite of this fragmentary preser- vation the ornamentation is so characteristic that there is no reason to doubt its identity with this species. The ornamentation consists of broad, flattened strigations separated by broader, rounded grooves which bear a fine spiral line down the center. The pattern of ornamentation on the Malayan specimen is identi- cal to the fine specimen from the Shalshal Clififs in the Himalayas, figured and described by Diener (1895, pp. 61-62, pi. 15, figs. la, b). In his description of the Himalayan specimen Diener w^as quite emphatic as to the identity of his form with the type from the Alpine Middle Triassic. In this conclusion he had confirma- tion from Mojsisovics who also examined the Himalaj'an speci- men. Sturia sansovinii is known from Anisian and Ladinian strata at many localities in the Mediterranean region. Bibliographic citations to these can be found in Diener (1915a) and Kutassy (1933). The distribution in the region of eastern Tethys and in the circum-Pacific region is not so well known and is of special 6 BREVIORA No. 124 interest here. This species is the only form of Stm-ia from the Himalayas proper in the so-caUed Ilimahiyaii faeies, where Dieiier (1895, 1907) has recorded specimens from tlie upper Musehelkallv at the Shalshal Clift* and at Si)iti. However, in Tibet in some of tlie exotic bloclvs of Alalia -Jobai- near Cliitichun Peak No. 1 (17,74Q ft.), Diener obtained a specimen of Sfuria whicli lie identified as Sturia monogolica (Diener, 1895, p. 113, pi. 29, tig". 4). This form is quite distinct from other species of Sturia in its open umbilicus and the suture, characterized by long', slender, pyramidal saddles. At a later date Diener (1916) erected the genus Fsilusiuria with S. mougolica as the type species. Indeterminate species have been recorded from Middle Triassic horizons in upper Thailand (Kummel, in Pitakpaivan, 1955). These particular forms are small, poorly preserved, and crushed specimens whose relationship to N. sausoriiiii is impossible to determine. Among the inimerous jMiddle Triassic faunas de- scribed from Indochina, Sfuria has as yet not been recorded. Welter (1915) records S. cf. sansovinii from a Ladinian horizon on Timor based on a frag-mentary specimen. Sturia japonica Diener (1915b, pp. 18-20, pi. 6, figs. 1-2) is based on a highly distorted specimen from Middle Triassic forma- tions at Inai, Japan. It is (luite similar to *S^. sansovinii differing in minor features of the suture and character of the strigations. Sturia sansovinii is thus found widely distributed throughout the Tethyan geosyncline where it occurs in strata of Anisian and Ladinian age. Repository: BMNH — C 55669 (figured specimen) . Family ACROCHORDICERATIDAE Arthaber, 1911 Genus AcEOCHORDICEEAS Hyatt, 1877 AcROCHORDicERAS sp. iudet. Plate 1, figure 1 A single, large, crushed and elongated specimen can be as- signed to the genus Acrochordiceras but its poor and incomplete preservation prevents determination of its specific relationship. The whorl sides bear strong radial to slightly curved sharj) ribs. 1900 AXISIAX AM MONOIDS FROM MALAYA 7 Some ul' tlu' ribs bcyiii at tlu' uinbilie-ai .slioulder.s where the}' increase in height forming somewhat of a tubercle beyond which tlu'v bifurcate. Other ribs lack tlie umbilical protuberances and are slightly less prominent. The poor preservation prevents de- tei-mining the pattern of alternation of these two types of ribs. Tlie conch was no doubt slightly evolute but the shape of the wliorl section is not possible to determine nor is any part of the suture preserved. Close comparison of this Malayan specimen with the known species of Acrochordiccras is not very satisfactory but one fea- ture is notable — that is, the rather sharp ribs on the Malayan form. The extent to which these sharp ribs may be due to the deformation of the specimen is hard to determine, however. In most species of Acrochordiceras the ribs tend to be rounded and in some cases broadly rounded. Even though specific comparisons are not possible there is no question of the generic assignment of this form. The genus Acrochordiceras is widely distributed in the Tethy- an belt from the Alps to Timor and is likewise known from a number of localities in the circum-Pacific region. Reposifo)>i. R]\IXII — C 55660 (figured specimen). KEFERENCES .\kthaber, G. V. 1928. Ammonoidea Leio.stiaca aus der obereu Trias von Timor. 2. Nederl. Timor Expedite 1916 oiidcr leiding von Dr. IT. G. Jonker. Uitgegeveu door Dr. H. A. Bromver. IV. Jaarb. Mijnw. Nederl. Ind., vol. LV, no. 2. pp. 1-174, pis. 1-20 (1926). DiENER, Carl 1895. Himalayan Fossils. The Cephalopoda of the Muschelkalk. India Geol. Survey, Palaeont. Indiea, ser. 1.5, pt. 2, pp. 1-118, pis. 1-31. 1907. Fauna of the Ilimalaj'an Muschelkalk. India Geol. Survey, Palaeont. Indiea, ser. 15, pt. 5, pii. l-tlO, pis. 1-17. 1915a. Fossilium Catalogus. I. Pt. 8. Cephalopoda triadiea. 369 pp. Berlin. 19151). Japanisehe Triasfaunen. Denksehr. Akad. Wiss. Wien., vol. 92, pp. 1-30, pis. 1-7. 1916. Einige Bemerkungen zur Nomenklatur der Triaseephalopoden. Centrabl. Min. Geol. Paltiont., pp. 97-105. 8 BREVIORA No. 124 KUTASSY, A. 1933. Fossilium Catalogus. I. Animiilin. Pt. 56, Cephalopoda triadica II., pp. 371-832, Berlin. Newton, R. B. 1923. On marine Triassic shells from Singapore. Ann. Mag. Nat. Hist., ser. 9, vol. 12, pp. 300-321, pi. 9. 1925. On marine Triassic fossils from the Malayan Piovince.s of Kedah and Perak. Geol. Mag., vol. 62, pp. 76-85, pi. 3. PiTAKPAIVAN, KaSET 1955. Occurrences of Triassic Formation at Mae Moh. Royal Dept. Mines, Bangkok, Rept. Investigations No. 1, pp. 1-11, pis. 2-4, map. Saubin, E. 1956. Lexique Stratigraphique International, vol. 3, Asie, Fasc. 6a. Indochine, pp. 1-141. Savage, H. E. F. 1950. Triassic fossils from near Kuala Lipis, Pahang (Malaya). Colonial Geol. and Min. Resources, vol. 1, no. 1, pp. 76-77. Spath, L. F. 1951. Catalogue of the fossil Cephalopoda in the British Museum (Natural History). Part 5, the Ammonoidea of the Trias (D). London, pp. 1-228. Welter, O. A. 1915. Die Annnoniten und Nautiliden der ladinischen und anisischen Trias von Timor. Paliiont. von Timor, vol. V, pp. 71-136, pis. 83-95 Explanation of PLATE The specimens illustrated on this plate are from mudstones of Anisian age from near Kuala Lipis, Pahang, Malaya. They are deposited in the British Museum (Natural History), London. Figure 1. Acrochordiceras sp. indet. BMNH — C 55660. X 0.5. Figure 2. Stiiria sansovinii Mojsisovics. BMNH — C 55669. X 0.5. Figure 3-6. Paraceratites trinodosus (Mojsisovics) BMNH — C 55672 — C 55675. X 1. Figure 7. PtydJiitcs sp. indet. BMNH — C 55659. X 1. \ t:'g. /^ 3«i BREVIORA MuseiLiinii of Comparative Zoology Cambridge, Mass. May 27, 1960 Number 125 THE LUMINOUS ORGANS OF PBOCTOPORUS (SAURIA, REPTILIA) —A RE-EVALUATION By Willard D. Roth Department of Anatomy, Harvard Medical School, Boston, Mass. and Carl Gans Department of Biology, The University of Buffalo, Buffalo, X. Y., and Carnegie Museum, Pittsburgh, Pa. INTRODUCTION The lierpetological literature contains two reports describing the first luminous organs in a terrestrial vertebrate. The two papers discuss identical specimens of the Trinidad lizard Procto- porus shrevei Parker. Sanderson (1939, and observations cited by Parker, same date) claimed that light was emitted by black bordered ocelli on the sides of a male, and Parker (1939) sup- ported this on the basis of his histological examination of the preserved animal. No new observations have been published since that time, but a number of workers have commented upon the original observations. Thus Pope (1955, p. 306) remarked that "other teiids have spots somewhat like those of P. shrevei, so it is highly probable that they, too, can light up." In contrast to this, Harvey (1952, p. 494) in his monograph on bioluminescence stated that he believed "all reports of luminescence in higher vertebrates to be false or spurious due to reflection of light or infection by luminous bacteria." The divergence of opinion on this interesting point prompted a re-examination of this question. The present paper reports a few additional field observations, and includes as well a detailed examination of the histological structure of the ocelli. Since Proctoporus shrevei is very rare, this re-examination had to be carried out on two related and superficially similar forms. 2 BREVIORA No. 125 OBSERVATIONS ON LIVING SPECIMENS Parker (1939, p. 659) mentioned that Sanderson's tield notes contained onh- a brief reference to color pattern involving "five, black spots each containing- a small, vivid Avliite, sometimes Inminons bead." Parker fnrther stated that Sanderson in con- versation informed him "that the animal was kept alive in captivity and could be stimulated to emit light from the lateral spots : the light was of a pale greenish hue, similar to that pro- duced by the hands and figures of a luminous watch. Excitement produced by flashing an intermittent beam of light on the lizard was found to be a very effective stimulus to light production." A more specific first-hand report was given by Sanderson (1939, pp. 41-43) in his popular, considerably amplified and slightly different, account. When initially observed the lizard "turned its liead away from me and both its sides lit up for a few seconds like the ])ortholes of a ship." "After one brilliant display on the night of its arrival in camp it refused to shine with full brightness though the beadlike spots remained plainly discernible in a darkened box when the rest of the animal was invisible." A "loud Avhistle, sudden winds, and flashes of light greatly agitated our lizard, causing it to switch on its "port- holes". . . . The light was much brighter the first time it was switched on after the animal had been quiescent for a period, and more especially after it (the lizard) had previously been sub- jected to intense illumination." We have been able to obtain four sets of further observations on live specimens of the genus Proctoporiis. Julian S. Kenny (V. C. Quesnel, personal communication) some time ago repeated Sanderson's experiments on the original species (P. shrevei) with entirely negative results. Kenny's field notes also indicate that the lizards are diurnal and inhabit rela- tively open spaces on El Tucuche, Trinidad. Dr. Janis Roze (in lift.) states that a specimen of Proctoporiis achlyens Uzzell (M.C.Z. 53128. later used for histological exami- nation reported herein) did not glow when placed in a darkened room after capture. lie adds that exposure to ultraviolet light did cause the spots to shine faintly. The test was carried out in an incompletely darkened room, and the results seem to be open to some question. Harold Heatwole and Owen J. Sexton (personal communica- tion) performed a nuinber of experiments at a field station in Venezuela. They tested one adult male of P. Jucfuosuf; (Peters) 1 !»()() THE LUMINOUS ORGANS OF PROCTOPORUS 3 and two adult males of /'. achlyens for a period of one month. The spots of the first species were yellow and those of the second were red in life, both series of spots bleaching to white after formalin preservation. The specimens were repeatedly moved from light to dark environments and were observed at night. The animals were disturbed. No luminescent effect was ever noted, intraviolet illumination was not attempted. The most extensive series of observations on live animals was made in Ecuador by James A. Peters {in litt.). Specimens of Prionodactijlus vertehralis (O'Shaughnessy) n.\\(\.Neiisticnrus ec- plcopiis Cope were observed while free in the field, during the collecting process, and for several weeks in the laboratory. He reported that neither luminescence nor any other kind of light could be noted in broad sunlight, dim or artificial light, or in the complete absence of light. No reflection could be noted under various types of lighting (sun, fluorescent and incandescent), in quiescent, active or deliberately disturbed animals. The evidence is most valuable because Peters was aware of the lizards' reputa- tion and was deliberately testing the hypothesis of luminescence. SUPERFICIAL APPEARANCE AND PHYLOGENETIC DISTRIBUTION OF THE OCELLI The supposedly luminous ocelli (Fig. 1) are rather similar in the species of Procioporns and Ncusfictirus here discussed. They are ari-anged in a single row along the side of the animal ; each ocellus always shows a light-colored, sharply -defined, black- bordered, circular center. Thev mav be restricted to adult males, with juvenile specimens and females showing onh^ traces. There is usually a size decrease of the ocellar center posteriorly along the series. There may be a marked irregularity in the width of the black border. There is no correlation between the ocellar and the scale patterns. While sharply-defined ocelli are commonly Avell developed only in Boulenger's (]885, p. 332) teiid group II, a check of the more than 130 species of teiid lizards as well as forms of other families represented in the collection of the Museum of Comparative Zoology at Harvard College indicated that patterns with sharply contrasting light and dark rolors are extremely common. A complete morphological series may be demonstrated in the Teiidae. This series ranges from ]iatterns with alternating light and dai'k stri])es. thi-ough those in which the stripes alternately fuse and bi'eak up. to ]iattefiis with well-defined light circles on BREVIORA No. 125 a dark haekgTOund. Such spots may be found over the entire body or may be restricted to the sides. The condition found in the males of Proctoporus and Neusticurus represents only one extreme development of color variation. Similar conditions may also be observed in certain geckonids and iguanids. Here the ocellar pattern may occur all along the side, with the color contrast emphasized around a limited number of spots. Fig. 1. Proctoporus achlyois Uzzell. Lateral view to show shape, size and anangeiuent of ocelli. The scale is graduated in mm. (M.C.Z. 5.3128 — C.G. photo.) H1ST0L0C4ICAL AND HISTOCHEMICAL EXAMINATION OF THE SKIN Methods Two specimens were used for histological examination. These were M.C.Z. 43764, Neusticurus ecpleopus ocellatus Sinitzen from Hacienda Pampayacu, Departamento de Huanuco, Peru, and M.C.Z. 53128, Proctoporus acJiIyens Uzzell from Choroni, Estado de Aragua, Venezuela. The museum specimens were reported to have been fixed in 10 per cent formalin and transferred to 70 per cent alcohol for storage. Sections Avere cut from the second (and largest) ocellus of the left side of each specimen. A sample of faintly pigmented skin from the ventral surface of M.C.Z. 53128 was sectioned for comparison. Small blocks of tissue, including both the center and the black margin, in the case of the ocelli, were excised. These tissue blocks, including the epidermis, dermis, and a small aiiioiiiit of undcrlvino' .skeletal muscle were hvdrated through a 1960 THE LUMINOUS ORGANS OF PEOCTOPORUS 5 descending- alcohol series, post-chromated in saturated aqueous potassium dichromate at room temperature for three days, washed overniglit in running tap water, dehydrated in ethanol, cleared in chloroform, and embedded in tissue mat (56-58°). Five-micron sections were mounted individually so that various staining techniques could be carried out on adjacent sections. The techniques used were : Harris hematoxylin and eosin and the paraldehyde fuchsin method as modified by Halmi (Halmi, 1950) for general morphology, Wilder 's modification of the Bielchowsky silver impregnation method (Romeis, 1948, p. 355) for nerve fibers and endings, Sudan black B with acetone con- trols for lipid compounds, the periodic acid-Schiff technique for 1-2 glycol linkages, the azo dye methods for protein bound sulf- hydryl and disulfide groups (Barrnett and Seligman, 1952, 1954 j, dilute methylene blue at pH's 4 to 9 and dilute light green at pH 's 3 to 8 for a rough approximation of pH signature of proteins (Singer, 1952) and buffered toluidine blue and thio- nine for metachromasia. GENERAL MORPHOLOGY OF THE SKIN AND OCELLUS On the basis of the appearance of nuclei, blood cells, striated muscle fibers, small nerves, and the cells of the epidermis, the fixation was judged to be good. Presumably the external loca- tion of the tissue with the consequent immediate exposure to the formalin had been advantageous. In addition, the tanning with dichromate seems to have been successful in preventing the shrinkage that normally results from paraffin embedding of formalin -fixed tissues. As revealed by hematoxylin and eosin (Figs. 2, 3), the epi- dermis is composed of a very thin stratified squamous epithelium showing a prominent basement membrane. The epithelium con- sists of a single-layered cuboidal stratum germinativum covered by only one or two layers of flattened squamous cells. The surface is covered by dense keratinous scales. The dermis can be subdivided into tw^o layers. Superficially, it is composed of a rather loose fibroelastic tissue which contains a dense accumulation of melanin pigment except in the region of the ocellus. The pigment appears to be contained in chroma- tophores, but this cannot be stated categorically for all of it. Often fine strands of pigment granules extend into the epidermis. In some instances these granules seem to be in processes between the epidermal cells, but in others they seem to occur within the 6 BREVIORA No. 125 e-ytuplasni of tiie epithelial cells. Possibly they occur in both locations. The deep layer of the dermis is composed of typical dense collagenous connective tissue with a rather regular orienta- tion parallel to the surface of the skin. Both layers of the dermis contain an extensive network of elastic and reticular fibers. Fig-. 2. Cross-section through the center of ocellus (l)et\veen the arrows) and surrounding pigmented skin taken from Procfnpnriis (irhhifns Fzzcll. Ilematoxvlin and eosin. .jOX. Fig. 3. Cross-section through the center of the ocellus shown in Fig. 2. Xote the thin epidermis (E), the vacuolated appearance of the superficial layer of the dermis (S), and the dense collagenous deep layer of the dermis (D). Hematoxylin and eosin. 400X. 1 ;)60 I in: LUMINOUS organs of proctoporus The dermis is underlain by typical loose connective tissue con- tainintj- small blood vessels, fat cells, and small peripheral nerves. The white center of the ocellus, which is the prime object of this study, differs histologically from the rest of the skin only ill tJie fact that no melaimi pigment occurs in the superficial layer of the dermis or epidermis. The sections thronoli the ocellus do not show any obvious differences in thickness or arrangement of skin structures as compared with the normal pigmented areas. It should be noted especially that no nerve fillers or specialized nerve endings were recognized in the dermis of either the ocellar or the pigmented regions of the skin, although distinct nerve fibers could be seen in the subcutaneous tissue and in tlie under- Ijang muscle bundles. It should lie noted further that no exten- sive or unusual vascular ai'rangement occurs in the region of the ocellus. THE IIISTOCHEMLSTRY OF TPIE SECTION Ilistochemically, the epidermis shows nothing striking. Its surface gives a moderate reaction for sulfhj'dryl groups and an intense reaction for disulfide linkages, as would be expected if its Fig. 4. Cross-seetioii tlirough tlie same ocellus tis shown in Figures 2, 3. Xote the intense PAS-positive reaction of tlie sn]iei-(icjai layer of the dermis. I'eriodic-acid-Scliiff reaction connterstained witli hematoxylin •400X. 8 IJREVIORA No. 125 scales were keratinized. The epidermis covering the white spot is identical witJi that covering pigmented dermis. The deep layer of the dermis also seems to be identical below the white and pigmented skin. The reactions of both collagenous and elastic tissue present nothing novel. The superficial dermis of the white spot is of special interest, however. In hematoxylin and eosin preparations, it is much more lightly stained than the deep portion of the dermis, contains less collagen and is far more cellular. However, the cytoplasmic limits (and the cell boun- daries) of these cells cannot be made out. The impression, therefore, is of tenuous, presumably branched cytoplasmic proc- esses. Histochemicaliy, this superficial region of the dermis (the cell cytoplasm?) shows a number of characteristics. Thus it gives an intense positive PAS reaction which is diastase resistant (Fig. 4). Further, it shows a moderately strong reaction for sulfhydryl groups which is not greatly intensified Avith the disulfide test. AVith controlled pH staining it shows only a weak staining with light green even at low pH's, but a moderate (pll 5) to heavy (pH 8) staining with methylene blue or thiouine. From these results it seems possible to conclude that the cells of this region are not vacuolated (as might appear from H and E sections), but contain a substance or substances not readily stained by routine methods. The results of the PAS method shoAv that this material contains numerous 1-2 glycol linkages, but is not glycogen. The results of the sulphydryl and the controlled pH methods indicate a moderately basophilic protein which con- tains appreciable cystine or cysteine or both. A complete absence of staining witli Sudan black B indicates that it is not a phospholipid, a glycolipid, or a lipoprotein. Thus these findings suggest the presence of a mucoprotein or mucopolysaccharide. A heavy accumulation of connective tissue ground substance would account for these observations except for the absence of meta- chromasia. While the fixation in this instance is not optimum for a critical evaluation, the ap]iearance of the sections favors an intra- rather than extracellular localization. Furthermore, upon the examination of unstained sections, the region shows no mai'ked granulation under either the light or phase-contrast inicroscope. The reflecting pigment guanine is supposedly insoluble in all of the reagents used for fixation and embedding. Guanine crystals, if present, should therefore be evident in unstained sections, and their absence rules out this pigment as the source of the white appearance. 1960 THE LUMINOUS ORGANS OF PROCTOPORUS 9 DISCUSSION AND CONCLUSIONS Basically there are no important differences between the pres- ent more extensive histological description and that originally fnrnished by Parker (li)39, p. GrtS) . He accented the differences l)et\veen the tissue underlying- the glistening white spot and that below the remainder of the dermis. The points emphasized were tive : (1) a reduction of the epidermis to one-half its normal thickness, (2) absence of the chromatophore layer, (3) presence of a mass of spongy mesenchymatous tissue, (4) large intra- or intercellular spaces in the spongy tissue, and (5) absence of nerve endings. Parker emphasized the poor preservation of the ma- terial. Neither the stains nor the method of preservation were specitied, but the photomicrographs suggest that only H and E or a similar routine staining technique was employed. We differ in failing to find either vacuolated spaces in the s])()ngy tissue or a reduction in thickness of the overlying epi- dermis. It seems clear that the presence of "vacuolated spaces" could be accounted for entirely by the fact that Parker 's material was poorly preserved for histological purposes and that these are presumably fixation or shrinkage artifacts. With regard to tlie variation in thickness of the epidermis. Parker's photographs indicate that such a reduction does not coincide with the absence of melanin. Our sections indicate similar and regional differences in the thickness of the epidermis of some of the pigmented scales. On the basis of the additional evidence reported in this paper, the following statements may be made : Sanderson's observations and Parker's comments thereon con- tain a number of inconsistencies. Sanderson reported that the light is (1) under the control of the animal and can be turned on and off, (2) much brighter the first time it is used after a period of quiescence, and (3) brightest .inst after the spot has been subjected to illumination. In spite of this Parker suggested that the organs are reflectors rather than truly luminous. A reflector might be capable of producing the first of the three effects, but the last two would be characteristic of true lumin- escence. None of the subsequent field observers has reported similar results. Their experiments cover more specimens and a longer period than does Sanderson's report. However, in all but one case the species involved are diff'erent. though externally very similar and probably closely related to the form on which the original report had been based. 10 BREVIORA No. 125 The lateral ocelli of rroctuponm luiglit represent one of four types of structures: independent light-generating organs (either innervated or under liormonal control), receptacles for light-gen- erating organisms, specialized reflecting structures, or simplj^ nonpigmented "white" spots with a hlack margin. The first of these possil)ilities is made unlikely by a number of factors. The center of the ocellus shows no gross difference from the surrounding skin and can be recognized only by the absence of melanin. The ocellus does not, in any manner, resemble pre- viously described luminescent organs. The cells of the white spot do not have an epithelioid appearance and in no way resemble cells previously described for luminescent organs. There is neither special innervation nor vascularization. This is par- ticularly important since Sanderson indicated that the light was turned on (|uite rapidly. The absence of any vacuoles or staining reactions character- istic of bacteria seems to rule out the possibility of storage of luminous microorganisms. There are certain difficulties in distinguishing between re- flecting structures and plain white spots. It seems certain that the white spots do not represent one of the more complicated reflecting systems since specialized epidermal cells and similar structures are lacking. It is, therefore, concluded that the white ajipearance is pro- duced by an inter- or intracellular substance, which lies at the same level of the skin as the dermal melanophores, and which may or may not have sjiecial reflecting properties. The further possibility exists that a local accumulation of connective tissue ground substance, or a specific intracellular mucoproteiu or mucopolysaccharide could be strongly reflective. This conclusion is in good agreement with all reports of obser- vations on live animals but Sanderson's. If any of the species of Proctoporus or related teiids are luminescent, they would seem to glow only under very special circumstances and by a yet un- described mechanism. HoAvever, the inconsistencies of the initial reports by Sanderson and Parker and the completely negative result of the investigations presented here force us to reject, for the present, any interpretation of these "portholes" as l)io- luminescent organs. Our findings, furthermore, suggest a quite different and inter- esting possibility. The location and appearance of the protein- containing cells in the superficial dermis suggests that they could be potential melanophores which have formed no pigment. 1960 THE LUMINOUS ORGANS OF PROCTOPORUS 11 Clearly, proof of this would involve a study of the histochemistry of such "prepiofmented" melanophores. Nevertheless, it seems possible to speculate that tliis color pattern in lizards might be achieved through a precise local control of the chemistry of the melanophore cells and hence, might prove an interesting area for the study of specified control of cellular differentiation. ACKNOWLEDGEMENTS It is a pleasure to acknowledge the original observations kindly contributed by Messrs. H. Heatwole, 0. J. Sexton, J. Kenny and V. C. Quesnel, and Drs. J. A. Peters and J. Roze. We are grateful to A. B. Dawson, D. W. Fawcett, T. S. Parsons, T. Uzzell and E. E. Williams for critical comments on the manu- script. Many of tlie histochemieal slides were prepared by Miss Grethe Aas, and Figures 2 to 4 are from photographs taken by L. Talbert. This study was completed under grant NSF G-9054 of the National Science Foundation. LITEEATUEE CITED BaRRNETT, E. J. AXD A. M. SELIGilAX 1952. Histoeheniit-al demonstrations of proteiu-bound sulphydiyl Sioii])s. Sc-ien-e, n.s., vol. 116, pp. 323-327. 19-ji. Ilistoeliemical demonstration of the siilfhydiyl and disulfide groups of protein. Jour. Nation. Cancer Inst., vol. 14, pp. 769-803. BOULEXGER, G. A. 1885. Catalogue of the lizards in the British Museum (Natural His- tory). 2nd ed., London, vol. 2, xiii + 497 pp. Halmi, N. S. 1950. Two types of basophils in the anterior pituitary of the rat and their respective cytophysiological significance. Endocrinology, vol. 50, pp. 140-142. Harvey, E. N. 1952. Bioluminescence. Academic Press, New York, xvi + 649 pp. Parker., H. W. 1939. Luminous organs in lizards. Jour. Linn. Soc. London, Zool., vol. 40, pp. 658-660. Pope, C. H. 1955. The reptile world. Alfred A. Knopf, New York, xxv + 325 -f xiii pp. 12 BREVIORA No. 125 EOMEIS, B. 1948. Mikroskopisehe Techiiik. Oldenbourg, Munich, xi + 695 pp. Sanderson, I. T. 1939. Caribbean treasure. Viking Press, New York, 292 pp. Singer, M. 1952. Factors which control the staining of tissue sections with acid and basic dyes. Int. Eev. Cyt., vol. 1, pp. 211-255. BREVIORA Mmiseiiim of Coimpsirsitive Zoology Ca:sibridok, Mass. .Itne 3, 19()() NiniBEu 126 MJD-SC'VTIUAX AMMONITES FROM IWAI FORMATION, JAPAN By Bernhard Kummel and Sumio Sakagami Lower Triassic ammonoids are known from only a few locali- ties in Japan. The first fair-sized fauna to be recorded was that from the Taho formation on the island of Shikoku described by Yehara (1928) . The majority of ammonite species from the Taho formation belong in A^iasihirites and Hemiprionites; Yehara had erroneon.sly assigned these ammonites to species of Meekoceras, Kijmaiitcs, Ophiceras, and Xenodiscus. This fauna is clearly representative of the zone of Anasihirites multiformis, which is known from Timor, Kashmir, the provinces of Kiangsu and Hupeli in China, British Columbia, Queen Elizabeth Islands, and western United States. A most imi^ortant contribution to our knowledge of the Lower Triassic of Japan was made by Sumio Sakagami who in 1955, described a small fauna of ammonites from the Iwai formation, Kaizawa Valley, Hinode-mura, Nishitama-gun, Tokyo-to. Saka- gami had specimens from two fossiliferous beds, seventeen meters apart. The lower fauna was concluded to be of early Scythian age and the upper fauna to be mid-Scythian (Meekoceras zone) in age. Correspondence between the authors about this fauna and the studj^ of additional material bear out the conclusion that the faunas of both fossiliferous beds are of Meekoceras zone age. The object of this paper is to further document the species present and the age of the Iwai formation. The Iwai formation is well exposed in the Kaizawa Valley where Sakagami has recognized four members. In ascending ordci- these members are: (a) more than 40 m. of black and bluish sandstone, (b) 10 m. of shale, (c) 10 m. of sandstone, and 2 BREVIORA No. 126 finally (d) about 25 m. of black shale which contains the two fossiliferous units. The lower fossil bed occurs about 2 meters above the base of the upper member, where the fossils occur in lenses of black limestone. The species identified from this horizon are: Dieneroceras iwaiense (Sakagami) Diencroceras sp. indet. Owcnitcs shhnizui (Sakagami) Farmiannites sp. indet. Aspeiiites sp. indet. Juvenites sp. indet. The upper fossil bed lies about 17 meters above the lower fossil bed and consists of marl lenses from which Sakagami obtained a single specimen that has been assigned to Aspenites. The genera and species in both the lower and upper beds are forms very characteristic of the mid-Scythian MccJwccras zone. Faunas of this age are well known from several localities in California, Nevada, Utah, and Idaho (Smitli, 1932; Kummel, 1954) ; from the Queen Elizabeth Islands of Arctic Canada (Tozer, 1958) ; from the Island of Timor (Welter, 1922) ; from Southland, New Zealand (Kummel, 1959) ; from the northern Caucasus Mountains, Russia (Kiparisova, 1958) ; and finally faunas of this age appear to be represented in the Kolyma River region of northeastern Siberia (Popov, 1939) and in Yugoslavia (Petkovic and Mihajlovic, 1935). Of all the forms represented in the Iwai faunas the single specimen of Owenites furnishes the best clue as to their age. Oivcnites shimizui (Sakagami) is an immature form that, how- ever, compares very closely with Oivenites Ti-oencni from the Meclfoceras beds of western United States. Owenites is also known from Timor, New Zealand, and the northern Caucasus Mountains, Russia. Dieneroceras is a longer ranging genus but Dieneroceras iwaiense (Sakagami) is close in its general conch morphology to D. eiieneri from the Meekoccras beds of the west- ern United States. The Iwai specimens j^laced in Dieneroceras sp. indet. are more involute than D. iwaiense and of very different appearance. These specimens lack any sign of a suture and the identification can only be considered as tentative. Juvenites is another genus of wides])read occurrence in the Meekoccras zone of western United States but it does range above and below this zone. The single specimen from the upper fossiliferous bed which Sakagami identified as Aspenites sp. was not available for study. 1960 AMMONITES FROM IWAI FORMATION, JAPAN The specimen is fragmentary but appears from the illustration to be reasonably placed in Aspenites. Another specimen has been uncovered from the lower fossiliferous bed that is of much better preservation and is without any doubt an indeterminate species of Aspenites. Previous assessments of the age of the Iwai formation rested largely on the identification of the most common species in the lower fauna — Diencroceras iwaiense (Sakagami) -as an Ophi- ceras. Shimizu (1932) appears to have been the first to comment on the Iwai ammonites, though Fujimoto (1926) discussed some Fig-. 1. l)iagi;iiimi;iti(.- representation of i\w suture of (a) Oiccnitcs shimizui (Sakagami), holotype (from Sakagami, 1955, pi. '2, tig. 2c) X 7; (b) Owcnites koeneni Hyatt and Smitli, from a paratype of 15 mm. in diameter (from Hyatt and Smith, 1905, pi. 10, tig. 10), X 8; (e) Diencro- ceras iioaiense (Sakagami), partial suture of holotji^e (from Sakagami, 1955, pi. 1, fig. le), X 10. Pseudomonutis, earlier. Forms like Dieneroceras iwaieMse are extremely difficult to place stratigraphically, and it was not until the presence of such genera as Owenites and Aspe)titcs was estab- lished that both the age and generic assignment of the "Ophi- ceras" could be properly evaluated. 4 BREVIORA No. 126 SYSTEMATIC DESCRIPTIONS Family DIEXEROCERATIDAE Kiimmel, 1952 Genus DiEXEROCERAS Spatb, 1934 DiENEROCERAS iWAiENSE (Sakagaiui) Plate 1, fio'ures 3-5 ; Plate 2, figures 7-9 Oplticvran iwaiense Sakaganii, 19.").j, pp. 135-136, pi. 1, figs. 1-9. Ophiceras sp. Sakaganii, 1955, pp. 136-137, pi. 1, figs. 10-1]. Dienerocera'i iwaicnsr, Kumniel. 19.~9, p. 430. The dominant element in the lower ammonoid bed at Iwai is a new species of Dieuerocrros. Sakagami (1955) had eleven speci- mens that he described and illustrated, and there are now four additional specimens in the collections of the Museum of Compar- ative Zoology. The conch is very evolute, each whorl embracing the preceding- one only slightly. The whorls are compressed with broadly arched flanks which converge slightly toward tbe venter. The ventral shoulder is subangular and distinctly marked and tbe venter is a low broad arch. The umbilical shoulders are broadly rounded. The serpenticone coiling of the conch exposes all of the inner whorls whicli are more rounded (and less com- pressed) than the outer volutions. The conch is smooth, except that on some of the specimens there appear to be extremely faint radial folds. The measurements of the better preserved speci- mens are as follows : *MCZ 5282a (Topotype) TUE 5255 (Paratype) TFE 5254 (Paratype) TFE 5252 (Paratype) TUE 5257 (Paratype) TUE 5251 (Holotype) MCZ 5282b (Topotype) TUE 5256 (Paratype) TUE 5253 (Paratype) The suture is faintly and only partially visible on the holotype and on one of the paratypes and consists of a large first lateral lobe, and a much smaller second lateral lobe on the umbilical wall. It is not possible to determine whether or not the lobes are denticulated. ♦MCZ = MusiMiiu (pf ('iiiuiiar.irivc Zooliigy : TUB =: Tokyo I'uiversity of Education D II w T' ( .Mcasurciiicin s in mm.) 22.5 7.7 5.3 11.0 21.0 15.8 10.1 20.7 7.0 10.6 20.5 6.5 9.0 18.0 6.2 8.7 15.5 4.5 3.7 7.5 15.0? 4.5 7.7 15.0 5.0 7.2 9.4 3.0 4.5 l!)(i() AMMONITES P^ROM IWAI FORMATION, JAPAN 5 Rcniorks. Many of the mid- and late Scythian ammonoids, that have on various occasions been assi\. -, li;.;;;. -;[ c. f/'/rr/n'/r.s- shi}ni~ui (Rakagaiiii ), Kuniniel, 1959, p. 43(1. The holotype and only specimen of this si)ecies is a small, juvenile specimen that can with confidence be assigned to Oivenites. The specimen measures 21.0 mm. in diameter, 11.3 mm. for the height of the last whorl, 8.0 mm. for the width of the last whorl, and the umbilicus is 1.5 mm. in diameter. The conch is involute with broadly arched whorl sides that converge, form- ing a sharp acute venter. The only ornamentation consists of radial growth lines. The suture (Fig. la) is ceratitic and typical of that found in species of Owenitcs. It consists of a narrow, denticulated ventral lobe, a large denticulated first lateral lobe, a smaller second lateral lobe and a series of small auxiliary lobes. This suture is almost identical in its basic plan to that of a specimen of 15 mm. in diameter of Owenifes koeneiii Hyatt and Smith (1905, pi. 10, fig. 10) reproduced here on Figure lb. Rcnwrks. Mature specimens of Owcnites show marked excen- truinbilication on the outer whorls producing a deep funnel- shaped umbilicus. Tlie immature volutions (roughly up to 25-30 19(iO AMMONITES FROM IWAI FORMATION, JAPAN 7 mm.) form a tightly involute coiieh and the small nmbilicns shows no tendency toward excentrumbilieation. Owenitcs shitni- zui is almost identical in conch shape and proportions to speci- mens of comparable size of 0. koeneni of the Meekoceras zone of western United States. A paratype of 0. koeneni originally illus- trated by Hyatt and Smith (1905, pi. 10, figs. 7-9) by a poor drawing is illustrated here on Plate 3, figures 5-7. This speci- men measures 15 mm. in diameter and is the specimen from which the suture of Text-figure lb was obtained. The suture is very similar in these two species at about the same diameter, ditfering only in minor details. The resemblance to Kingites in conch shape and suture is more apparent than real. Owenitcs is one of the best mid-Scythian zonal markers in the Circum-Pacific region. In western United States (California, Nevada, Utah, and Idaho) the genus is very common in the zone of Meekoceras gracilitatus. The genus was first established for specimens from the Meekoceras limestone in the Inyo range, Cali- fornia (Hyatt and Smith, 1905, p. 82). Smith (1932) recognized a number of additional species of Owcnifes in western United States but most of these appear to be merely intraspecific variants of 0. koeneni. The Timor Owenites cgrediens Welter (1922) has a narrow, rounded keel-like venter formed by the shell, but the internal cast has a sharp venter. Likewise the Timor species is generally more inflated, producing a broader and deeper umbilical funnel. The suture also differs slighth" in the shape of the lobes and the auxiliary series. Eecently, a specimen of Owenites cf. koeneni Hyatt and Smith, has been described from beds of Pre-Etalian age in western Southland, New Zealand ( Kummel, 1959) . Outside of the Circum-Pacific region. Owetiites has been recorded only from the northern Caucasus Mountains. Occurrence. Lower fossiliferous bed of ui)per iiK'inlR'r of Iwai formation, Kaizawa Valley, Iwai, Hinode-mura, Nishitama-gun, Tokyo-to, Japan. Repositorii. TUE 5262. holotype (PI. 2. figs. 5, 6). Family PARAXAXXITIDAE Spath, 1930 Subfamily PAKAXANXITINAE Spath, 1930 Genus PaRAXANNITES Hyatt and Smith, 1905 Paranannites sp. indet. Plate 2, figures 1, 2 I'vopt yclntvs ;iff. i-osi iilrd ii Irl Spntli. S;ik;if;-;iiiii, ]9.').1, ]iii. l.'iT-liiS, pi. H, figs, la, 1). 8 BREVIORA No. 126 Paranannitcs i\i. iudct., Kuniuu'l, ]!».'!), ji. 4'M). This form i.s represented by a single specimen of only moderate l)reservation. The eonch is involute, compressed, with flattened, l)araHel Avliurl sides and a broadly rounded venter. It measures 33.1 mm. in diameter, 15.3 mm. for the heig'ht of the last whorl, 9.3 mm. for tlie width of the last whorl, and 6.5 mm. for the diameter of the umbilicus. Unfortunately no suture is preserved. Lower Triassic ammonoids of this conch morphology are dif- ficult to identify, and without the suture generally impossible to recognize. The fact that the associated fauna includes species of Owenitis and Juvenites precludes the probability that this specimen could represent a species of Propfychitcs, which is gen- erally an earlier Sc^^thian form. The associated genera indicate a mid-Scythian age for the fauna and, of the ammonites of this age, Paranajuiitcs comes the closest in its conch morphology to this specimen from Iwai, Japan. Paranannitcs aspcncnsis from the Meckoceras zone of western United States has a conch of the same degree of involution and rounded venter which, however, is more inflated — ^the whorl width being just slightly less than the whorl height. Paranannitcs prrteiuiis Smith (1932, p. 99, pi. 31, figs. 13-15) has a laterally compressed conch with flattened sides like the Iwai specimen (PI. 3, figs. 9, 10). This species of Smith is believed to be a synonym of P. aspene7iesis. The tentative placement of tliis specimen in Paranannites is, of course, based on the assum])tion that it is a mature specimen. If, however, it is a juvenile form it is most likely not a Paranannitcs, and then could possibly be the inner whorls of a Flcmingites or Arctoceras, or other such larger ammonoids of mid-Scythian age. Occurrence. Lower fossiliferous bed of upper member of Iwai formation, Kaizawa Valley, Twai, Ilinode-mura, Nishitama-gun, Tokyo-to, Japan. Repository. TUE 5261 (PI. 2, figs. 1, 2). Genus JuVEXITp:s Smith, 1927 Juvenites sp. indet. Plate 1, figure 2 The collection contains a single specimen in which only one side of a half volution is preserved. The specimen measures ap- proximately 16.4 mm. in cliameter, 7.3 mm. for the height of the last whorl, and th(> umbilicus is 5.0 mm. in diameter. The conch is involute Avith l)roa(l, de]iressed Avhorls and a broadly rounded 19(i0 AMMONITES FROM IWAI P^ORMATION, JAPAN 9 vonter tliat <>rade.s imperceptibly onto the lateral areas. The eoiieli bears very eoiispiciioiis forward projecting constrictions. There appear to be six such constrictions on the half volution. A^o indication of a suture is preserved. Remarks. The lack of a suture and the incompleteness of the specimen necessarily make the present identification tentative. Even so, in consideration of the association with Owenitcs and Aspcnitcs which are clearly mid-Scythian in age, the assignment of this specimen to Juvenites appears reasonable. The constric- tions on the Iwai specimen are similar in depth and distinctness to those in Juvenites septentrionalis Smith (1032, pi. 31, figs. 31-32) but in the latter species the constrictions are radial. Strongly projected constrictions somewhat similar to those on the Iwai specimen are present on Juvenites thcrmarum (Smith, 1932, pi. 21, figs. 11-12, 16-17, 19-20). Occurrence. Low^er fossiliferous bed of upper member of Iwai formation, Kaizawa Valley, Iwai, Ilinode-mura, Nishitama-gun, Tokyo-to, Japan. lifprmtoyjj. MCZ 5284 (PL 1, fig. 2). Family IIEDENSTROEMIIDAE Waagen, 1895 Subfamily ASPENITINAE Spath, 1934 Genus ASPENITES Hyatt and Smith, 1905 AsPENiTES sp. indet. Plate 2, figures 3, 4 ; Plate 3, figure 8 A.siK nil( K .s[). S;ik;if;'aiiii, 1935, \>. lo9, pi. '2, ligs. 3a, 1). One of the authors (Kummel) has not had the opportunity to examine the single representative in the collection originally described by Sakagami (1955, p. 139) who states that it agrees with Aspenites of the western United States but the suture is not preserved. As well as one can tell from the illustration, this identification appears to be reasonably correct. Another specimen has since been uncovered from the lower ammonite bed that appears without doul)t to he a juvenile representative of Aspe- nites. This specimen (MCZ 5285) measures only 8.0 mm. in diameter and is a completely involute conch, greatly compressed, with broad arched flanks which converge to a narrow, keeled venter (PI. 3, fig. 8). No suture is preserved. The specimen is almost identical with specimens of comparable size of Aspenites acutus Hyatt and Smith (Smith, 1932, pi. 30, figs. 6-7, 9, 11-12). Aspenites is fairly abundant in tlie Mrekoerrns beds of Avestern United States and in Timor. 10 BBEVIORA No. 126 Occurrence. Botli iipi)er and lower fossiliferous beds of upper member of Iwai formation, Kaizawa Valley Twai. liinode-mura, Nishitama-s'un, Tokyo-to, Japan. Ecpositonf. tup: 5268 (PI. 2, %s. :], 4); MCZ 5285 (PI. 3, fig. 8). KEFERENCEW FUJIMOTO, 11. 1926. A new Locality of " Pseudomonotis." Jour. Geol. Soc. Japan, vol. .S.3, no. 390, p. 113. (In Japanese.) Hyatt, A. and .1. P. Smith 1905. The Triassie Cephalopod (ienei'a of America. U. S. Geol. Survey, Prof. Paper 40, pp. 1-394, pis. 1-85. KU'AKISOVA, L. 1958. Geologic Structure of the USSR, vol. 1, Stratigraphy. Moscow, 588 pp. (In N. A. Belyaevsky, et. al.) KUMMEL, BERNHARD 1952. A Classification of the Triassie Ammonoids. Jour. Paleont., vol. 26, pp. 847-853. 1954. Triassie Stratigraphy of Southeastern Idaho and Adjacent Areas. U. S. Geol. Survey, Piof. Paper 254-H, pp. 165-194, pis. 34-40. 1959. Lower Triassie Annnonoids from Western Southland, New Zea- land. New Zealand Journal Geol. Geog., vol. 2, no. 3, pp. 429-447, figs. 1-7. Pktkovic, K. V. and P. Mihajlovic 1935. La faune des cephalopodes trouvee dans le Trias Inferieur en Montenegro (Yougoslavie) ses caracteristiques et son importance. Ann. geol. Penins. Balkan., vol. 12, pp. 253-269. Popov, G. 1939. New Species of Ammonoids from the Triassie of the Okhotsk- Kolyma Land. Arctic Institute, Leningrad, Problems of the Arctic, no. 12, pp. 72-82. SAKAGAill, S. 1955. Ijower Triassie Ammonites from Iwai, Oguno-Mura, Nishitama- gun, Kwanto Massif, Japan. Science Report, Tokyo Kyoiku Daigaku, sec. C, no. 30, pp. 131-140, pis. 1, 2. . SlIlMIZU, S. 1932. On the Lower Triassie Ammonites of Iwai, Oguno-Mura, Nishi- tama-gun, Tokyo-to. Jour Geogr., Tokyo, vol. 44, p. 97. (In Japanese.) 1960 AMMONITES FROM IWAI FORMATION, JAPAN 11 Smith, J. 1'. 1932. Lower Triassic Ammonoids of North America. U. S. Geol. Sur- vey, Prof. Paper 167, pp. 1-199, pis. 1-81. Spath, L. F. 193-4. Catalogue of the Fossil Cephalopoda in the British Museum (Natural History). Part -t — The Ammoiioidea of the Trias. British Museum Publication, London, 521 pp., 17 pis. TOZEK, E. T. 1958. Triassic Faunas from the Queen Elizalieth Ishnuls, Arctic Canada. Al)stract, Bull. (Jeol. Soc. Amer., vol. (i9, pp. 1653-1654. Welter, O. A. 1922. Die Ammoniten der unteren Trias von Timor. Paiiiont. von Timor, Lief. 11, Alili. 19. i)]i. 82-160, pis. 155-171. Yehaba, Shingo 1928. The Lower Triassic Cephalopod and Bivalve Fauna of Shikoku. Jnp. Jour. Geol. and Geog., vol. 5, p]i. 135-172. EXPLAXATIOX OF PLATE I The specimens illustrnted on tliis jilate ;ne frcmi the lower fossiliferous bed of the Iwai formation, Kaizawa Valley, IIino;le-muia, X'isliitama-gun, Tokyo-to, Japan, and are preserved in the Musevmi of Comparative Zoology. Fig. 1. DUnei-ocerus sp. in;let., MCZ .5283, X 2. Fig. -2. Jiivrniirs sp. indrt., ^K'Z .5284, X 3. Figs. 3-5. Dirurrorrrds iivaioisc (Sakagami"), topotypes, MCZ .5282a,h,c, X :;, PLATE 1 EXPLAXATIOX OF PLATE 2 The speeinienw illustrated on this plate are t'vom the Iwai formation, Kaizawa Valley, Iwai, Ilinode-mnra, Ni.shitania-gun, Tokyo-to and are preserved in the collections of the Geological and ^Nlineralogical Institute, Tokyo University of Education. Figs. 1, -. Parandiinites sp. indet., fi-oni lower fossiliferous bed of Iwai formation, TUE 5261, X 2. Figs. 3, 4. Aspenites sp. indet., fi'om upper fossiliferous bed of Iwai formation, TUE .1263, X 1. Figs. 5, 6. Oivcnites sIiiDihui (tSakagami; from lower fossiliferous bed of Iwai formation, TUE 5262, X 2. Figs. 7-9. Dicncroccras iicaiensc (Sakagami) from lower fossiliferous bed of Iwai formation. Fig. 7, i)aratype, TUE 5254:, X 2; figs. 8, 9, holo- type, TUE 5251, X 2. Fig. 10. Dieneroccrus sp. indet. from lower fossiliferous lied of Iwai formation, TUE 5260, X 2. «r.)P'^5^ EXPLAXATIOX OF PLATE 3 Figs. 1-4. Dieneroceras dieneri (Hyatt and Smith), from MeeTcoceras beds in Wood Canyon, 9 miles east of Soda Springs, Aspen Eidge, Idaho. Figs. 1, 2, holotype, USX:\r 75260, X 1..5; 3, 4, paratype USXM 75260a, X 2. Figs. 5-7. Owenites Icocneni Hyatt and Smith, from Meekoceras beds. Union Wash, Inyo Range, Inyo County, California. Paratype USXM 7526111, X 2. Fig. 8. Aspcnitrs sp. imlet. from lower fossiliferous bed of Iwai forma- tion, Kaizawa Valley, Iwai, Hinode-mura, Xishitama-gun, Tokyo-to, Japan, MCZ 5285, X 4. Figs. 9-10. Faranannitcs pertenuis Smith, from ileeloceras beds in Wood Canyon, 9 miles ea.st of Soda Springs, Aspen Ridge, Idaho. Holotype, USXM 74960, X 1.5. ^«#' V JJs ,■% 1^ 8 10 PLATE 3 BREVIORA Mmseiuijni of Compsirative Zoology Cambridge, Mass. December 19, 1960 Number 127 NOTES ON THE CRANIAL ANATOMY OF NECROLEMUR By E. L. Simons ^ axd D. E. Russell - INTRODUCTION The large number of Avell-preserved skulls of Necrolemur antiquus of the late Eocene Quercy phosphorites of south central France allow for much more detailed study of cranial anatomy in this primate than is possible for most early members of the order. In spite of the fact that cranial osteology can be studied in great detail, views as to the taxonomic position of this primate, and of the allied genera Microchoerus, Nannopitliex, and Pseudoloris show considerable variance. Although not all of the same provenance, little dental varia- bilit}" is evidenced in specimens of Necrolemur antiquus examined by us. In the course of this study, however, a number of differ- ences in position and size of basicranial foramina have been observed, which are in line with mutability of cranial foramina (in individuals of the same species) reported by other authors (see Edinger and Kitts, 1954). Also, in Necrolemur the proba- bility remains that known specimens differ considerably in age (from early Bartonian to late Ludian provincial ages, at least), but locality data are inadequate for precise age determinations. Comparison of upper dentitions in the M.C.Z. and Paris skulls has failed to show anj- dental basis for species distinctions among them. To date, the most detailed studies of the cranium of Necrolemur have been by Stehlin (1916) and by Hiirzeler (1948). Stehlin's thorough and excellent description can scarcely be improved on, ^Zoological Laboratories, University of Pennsylvania. ^ Museum National d'llistoire Naturelle, Paris. 2 BREVIORA No. 127 l)iit copies of this work ai'e not as >ieiierally available as could he "wished. Also, tlie imi^licatioiis of some of his early observa- tions seem to have been neglected in latei- literature. In some rather significant points, recent examination of more and dif- ferent skulls ])ermits comments su|)pleiiientary to his work. ACKXO WLEDi ; e:\ie xts The authors Avould like to take this o|)portunity to thank Drs. J. -P. Lehman, Curator of Fossil Vertebrates at the ^Museum National cVIIistoire Naturelle in Paris, and A. S. Romer of the Museum of Comparative Zoology at Harvard for generously giving permission to pul)lish on the specimens in their respec- tive charges. Preparation of the figures, by ]\Iiss Ellen Cole, was supported by a grant from the AV(Miiiei--(iren Foundation for Anthropological Pesearch. ABBREVIATIONS In the absence of specimen numbers, the Paris Museum skulls of Nccrolcmur have been numberetl 1 through 5 for convenience of reference. Abbreviations used in this paper are as follows : M.C.Z., Museum of Comi)arativ(^ Zoology at Harvard College. Montauban, Natural History Museum (Geological Collection), Montauban, France. Paris, National Museum of Natural History, Paris. CRANIAL CHARACTERS I. AFDITORY REGION Most of the information, published to date, regarding the com- ponents of the auditory bulla in Necrolemur comes from Mon- tauban 9, which has been discussed by both Stehlin and Hiirzeler. Even though this skull was prepared with considerable skill, the crystalline calcite filling was apparently confused with the very similar appearing bone in the region of the epitympanic recess. Consequently, the route of the stapedial artery across the tympanic cavity of the middle ear was lost just anterior to the fenestra ovalis. The extrapetrous portion of the Fallopian aque- duct was lost as well. Preparation of Paris 2 has revealed more details of the epitympanic region (Fig. 1). Exposure of the inside of the bulla in Paris 2 and 5 indicated primarily the lack of a free annular tympanic ring. Moreover, studies by Simons (in press) on a specimen of Necrolemur at the lOfiO CRANIAL ANATOMY OF NECROLEMUR «( 33 o <:<: &K5 1; Rh t ex iD ^ i i ^ i; P rt OQ -. o X ;r o !<; rt 9 ^ -H ft -r '^ cs -^ <» ;h — a -^^ — =4-1 t£ ^ O •n ^- ^ =M CC ^ O ^H ^ • CI ^ b — ^ +- -: ^ ^ * p o 5 .2 - ^ ■> P 1^ *— — * '*^ ■^ a5 ^ 4 BREVIORA No. 127 British IMiiseuni (Natural History) doiiionstrate that the ecto- tympaiiic element is tubular, and mediall}' fused to the ventral bulla wall. Iliirzeler's evidence (1946:353; 1948:28) of the presence of a free ring-, therefore, can no longer be accepted. The bone he identified as such had to be removed during prepa- ration to expose the carotid canal and thus cannot be re-exam- ined. In M.C.Z. 8879 the meatal tubes are reasonal)ly well pre- served but one of the Paris skulls shows an even more complete osseous meatus. Together, these indicate that the ectotyrapanic (external to the bulla) is about as long as the transverse diam- eter of the foramen magnum, curves slightly backward, and may be broadest at the external aperture. None of the fossil or recent Lemuroidea have this sort of meatus. Iliirzeler (1948:27) cites M.C.Z. 8879 as not showing any evidence of a fused tympanic ring. Nevertheless, four and pos- sibly more transverse struts are exposed on the ventrolateral face of the right bulla of the Harvard skull. These bars are su]iports for the internal rim of the tubular ectotympanic. Also of interest is the fact that the anterior route of the ])romontory artery (true entocarotid) is apparently variable. In Montauban 9. it curves sharply anteromedially shortly after leaving the promontory of the petrosal. In Paris 2, this curving is much less accentuated. Some crushing is to be allowed for in the tympanic region of the latter specimen but the amount of curvature illustrated by Hiirzeler (1948, figs. 30 and 31) for Montauban 9 is not indicated in the Paris skull. It should be further noted that this l)ony tube does not lie in a horizontal plane in Montauban 9 and Paris 2, but slopes anterodorsally at an angle of about 45°. The stapedial artery, like the promontory artery, remained enclosed in a bony tube throughout its route within the bulla. Branching from the promontory artery just inside the carotid foramen, the stapedial artery curved dorsally, lying on and following the form of the petrosal promontory. It then passed anterior to the fenestra rotunda to the bottom of the fenestra ovalis. There it diverged laterally, crossing the fenestra ovalis, and continued anteriorly nearly parallel to the promontory artery (Fig. 1). The groove mentioned by Hiirzeler (1948:31) is surely a remnant of the stapedial tube, as he suggested. The exit of the stapedial artery appears to be at the dorsoanterior base of the external auditorv meatus. iy()0 CRANIAL ANATOMY OP NECR0LE:\IUR 5 The extrapetrous portion of the Fallopian a(iiieduct is also enclosed up to, or nearly up to, its exit at the stylomastoid fora- men. Its route lies lateral to the fenestra ovalis, just above the stapedial artery and continues posteriorly under the external semicircular canal. At this point, the tube forms a "T"' giving rise to the small anterolateral opening and a larger posterome- dian branch. The former is a natural foramen. Damage to the latter region makes it impossible to say whether or not the more posterior branch continued as an enclosed tube to the stylo- mastoid foramen. In Paris 2, this foramen is single, in Paris 1, double. Without exposing the interior of the bulla in Paris 1 a possible connection between the branching Fallopian aqueduct and the double stylomastoid opening cannot be confirmed. That the anterolateral foramen of the "T" in Paris 2 could have given passage to the chorda tympani seems likely. Regarding the foramen designated FX, "Foramen von un- liekannter Bedeutung, " by Iliirzeler (1948 :fig. 28), it seems probable that this represents the opening of the inferior petrous sinus, as originally stated by Stehlin (1916:1355). Contrary to the views of a few students, we tind little distinc- tion between Tarsius and Xccrohmur in the major carotid rela- tionships, both inside and outside the l)ulla. The two genera appear to agree in those features of the carotid circulation which distinguish tarsiers from Malagasy lemurs, adapids and lorises (see Le Gros Clark, 1959:151). Location of the internal carotid foramen in Tarsius on the ventral surface of the bulla (instead of on the median wall) constitutes a slight difference from Xco'ohiintr, but it should be stressed that placement of this foramen in the fossil species does approximate the situa- tion in Tdfsius, being more ventral than in most, if not all, other prosimians. Furthermore, Kecrolemnr and Tarsius are alike in having both stapedial and promontory branches within the ])ulla, encased in bony canals or tubes, with the promontorj^ division the larger. In typical Lemuriformes (Malagasy lemurs, adapines and notharctines), the carotid foramen has a quite dif- ferent location at the posteroexternal angle of the bulla and, inside it, the promontory division is A'ery small (Gregory. 1920 :174:-180) . Lorises and the cheirogaleine lemurs differ also, in that the carotid divides outside the entotympanic and the main branch enters the skull through the foramen lacerum medium instead of going through the bulla. A middle lacerate foramen is not present in Tarsius and Nccrohmur. The few differences be- tween these two genera to be observed in the auditory region 6 BREVIORA No. 127 seem best understood with reference to the effects of the anterior shifting of the foramen magnum and greater inflation of the anterointernal part of the bulla in Tarsius. Some primitive fea- tures are also to be seen in the Eocene form. For instance, the canal for the promontory artery in Nccrolemur is only slightly thicker than that for the stapedial. In Tarsius this difference is more pronounced. What is of general significance is that when Necrolemur differs from Tarsius it is usually intermediate be- tween the latter and yet more primitive prosimians. One could hardly expect an Eocene tarsioid to be otherwise. II. BASICRANIUM Stehlin (1916:1351) mentioned that the alisphenoids partici- pated in the composition of the anterior wall of the bulla. How- ever, Hiirzeler (1948:26) has pointed out that, although the l)ulla is overlapped b}^ the alisphenoids, this does not permit the definite statement that the alisphenoids constitute a part of the true bulla wall. In Montauban 9, it is possible to follow the suture between the ])ulla and its neighboring elements from the carotid foramen around the anterior end to the squamosal. The diverticulum D 2 (of Hiirzeler) appears to lie outside this suture. If then, as Hiirzeler suggested, the alisphenoid forms no part of the anterior bulla wall, this diverticulum (D 2) is extra-buUar. The broad overlapping of the external pterygoid plate of the alisphenoid onto the anterolateral bulla wall in Necrolemur (Fig. 2) is a feature of some interest in relating the Quercy form to the modern Tarsius. Cope (1885:467) long ago stressed the distinctiveness of this region of the tarsier basicran- ium when comparing it with the then newly discovered skull of an American Wasatchian prosimian, Tetonius homunculus. In both Necrolemur and Tetonius these external pterygoid plates overlap the bullae, as in Tarsius. Gregory (1920:227) gives the following as a general character of lemuroid Primates : ' ' The elongate pterygoid plates of the alisphenoids extended back to the auditory bullae, whereas in the Anthropoidea they are w^ell separated from them." A further distinction here is possible in that the posterior extremities of the external pterygoid plates in lemurs and lorises, including such fossil forms as SmUodectes, Nofharctus, Adapis, and Frouycticchus, typically (although not in all cases) reach back to the anterior tip of the bulla, but the area of contact is very small and cannot be described as over- lapping. Necrolemur, Tarsius, and Tetonius differ in this respect 1960 CRANIAL AXATOMY OF NECROLEMUR 7 from lemuriform, lorisiforin, platyrrhine, and catarrhine Pri- mates. Stehlin could not find the stylomastoid foramen in his speci- mens: Hiirzeler (1948) shows it in figures 27 and 28 at the posteroexternal angle of the bulla, but does not label it (see Fig. 1). Just anterior to this foramen is a fossette, probalily for reception of the stylohyal, if, as in Tarsius, the tympanohyal was not distinct (van der Klaauw, 1931:239). This foramen and fossette lie in the same depression and have a somewhat variable degree of separation. In M.C.Z. 8879 the external appearence is as a single oblong foramen, while in Paris 1 the two are more distinctly set off (Fig. 2). A specimen at the British Museum is intermediate in this regard. Contrary to Stehlin 's suggestion (1916 :1348) that a true post- glenoid i)rocess does not exist in Nccroloiiur, Paris 1 and 5 ex- hibit a process that can justly be termed postglenoid. Also a l)ostglenoid foramen is present (M.C.Z. 8879, Paris 1, 2 and 5) median to this process and between the posterior limit of the glenoid area and the external auditory tube. Two large foramina, one on either side of the alisphenoid pterygoidal wing, were described hy Stehlin (1916:1353-1354) as the inner and outer openings of the canalis civinninii (or foramen pterygospinosum, Stehlin 1912:1205). He named a smaller opening situated anterodorsally in the same region the foramen ovale. The position of this latter small foramen is vari- alilc. l)ut is always anterior to the glenoid fossa instead of being appi-oximately on a line with it, as is the foramen ovale in Primates generally. In those specimens in which this foramen is relatively large, a groove extends laterally and slightly pos- teriorly from it, which would probably not be the case if it were the foramen rotundum. Removal of matrix from Paris 1 and 2 in the region of the external pterygoid plate or wall has re- vealed a foramen opening directly into the cranial cavity, lying within the wall at the juncture of canals from the three foramina. Given this information, probably not known to Stehlin, we sug- gest that the posteroexternal foramen (outer opening of Stehlin "s canalis civinninii) is the foramen ovale. In a footnote, Stehlin (1916:1354) cited Gregory (1915:430) as confirming his identifications in the region of the foramen ovale. Gregory does this only partially, committing himself no further than to say that the foramen ovale is on the external (as opposed to the internal) side of the external pterygoid Avail. 8 BREVIORA No. 127 FIGURE 2 Necrolemur antiquus x 3.5 1. Foramen for branch of internal 6. maxillary artery 7. 2. Foramen ovale 8. 3. Postg'lenoid foramen 9. 4. Stylomastoid foramen and fos- 10. sette for ? stylohyal 11. 5. Foramen for auricular branch of pneumogastric Hypoglossal foramen Alisphenoid canal Opening of eustachian tube Internal carotid foramen Inferior petrous sinus Posterior lacerate foramen 1960 CRANIAli ANATOMY OF NECBOLEMUR 9 He did not definitely say that Stelilin's small foramen is the foramen ovale. However, he identified as the foramen rotundum the smaller foramen (Stehlin's f. ovale) regarded by us as being for the internal pterygoid branch of the internal maxillary artery. III. ORBIT Along its median Avail the orbit is composed principally of the frontal and the maxillary. Careful search of Paris 1 revealed no OS planum present in the orbital wall (Fig. 3). The lacrymal forms a narrow band within the orbit along the anterior rim. Frontal, parietal and alisphenoid comprise the posterior wall. Xo anterior sutures between the small orbitosphenoid, frontal, and palatine, respectively, could be made out in Paris 1, in which this region is entirely undistorted. Only a small palatine com- ponent is present in the orbit, and this does not separate the frontal from the maxillary. Stehlin (1916:1345) noted that his material was not adequate to allow determination of the maxillo-malar suture. This led him to suggest that the malar might reach the lacrymal. Paris 1 and 2 show that this is not the case ; the maxillary makes up part of the orbital rim. AVhen discussing the Eocene lemuroid Xoih- arctus, Gregory (1920:227) remarked "... the malar if not in actual contact Avith the lacrymal certainly came very close to it, whereas in tarsioids and anthropoids it becomes widely separated from the lacrymal and limited to the outer side of the orbit." Consequently, yccrolcmur resembles the higher Primates in this regard, and not the majority of prosimians other than Tarsius. The absence of an ethmoid component in the rostral orbital wall of Necroleniur has been taken by some students as an indi- cation of a lack of affinity betAveen it and Tarsius, since in the latter the os planum is large. To the Avriters this distinction does not seem to have much significance. A primitiA'e prosimian condition, Avhere the ethmoid has no orbital plate, is retained in such forms as Necroleniur, Pronycticehus, SmUoclectes etc., in Avhich, perhaps, there has not been enough orbital expansion to effect an expression of this bone in the orl)it by impinging on the anterior part of the interorbital septum. In Tarsius and some Lorisif ormes the interorbital septum is A'ery narroAv — evi- dently an accommodation for relatively large eyes, and in both groups an os planum occurs. Moreover, only in Cheirogaleinae, 10 BREVIORA No. 127 among living and fossil Lemuriformes, is the os planum present. In these small lemurs also, the occurrence of an os planum is coupled with large orbits (compared to body size) and a thin interorbital septum. For the Anthropoidea, a similar origin for FIGURE 3 Necrolemur antiquus x 3.5 approx. 1. Sphenopalatine foramen and posterior palatine canal 2. Foramen rotiindum and ante- rior lacerate foramen ( ?coal- esced) 3. Ethmoid foramen Al)breviation,s: as, alisphenoid fr, frontal 7)(.r, maxilla OS, orbitosphenoid pi, palatine pr, parietal this orbital element may be considered. In most Ceboiclea the orbits restrict the interorbital septum to a thin plate on which the OS planum is exposed laterally. Although in Old World Anthropoidea the rostrum between the orbits is occasionally 1960 CRANIAL ANATOMY OF NECROLEMUR 11 rather broad, it is possible to posit that such breadth is secondary and tliat they descend in common from a form in which rela- tively large eyes impinged on the interorbital area enough to induce the appearance of an orbital ethmoid component. This hypothesis is strengthened by observed interorbital narrowness in the only known part of an Oligocene catarrhine skull (Simons, 1959 :8). If the foregoing suggestions apply, then it is not neces- sary to expect the presence of an os planum in the stock from which Tarsins may have arisen. In Nccrolcmur (Paris 1), a small venous foramen can be seen situated near and beneath the median dorsal rim of the orbit. As in Tarsius, but apparently not in other Primates, below this foramen a deep groove curves posteroventrally and (in both) lies at a juncture between the plane of the lateral wall of the rostrum and that of the hack of the orbit. This is another un- usual feature (occurring in both Nccrolcmur and Tarsins) which has to be attributed to independent acquisition, by those who doubt that any known fossil prosimians have a close phyletic relationship to Tarsius. A cranio-orbital foramen exists in Paris 1, l)ut could not be found in Paris 2, 3 or 5. Running anteroposteriorly and slightly above the optic foramen is another groove. A small opening near its anterior end appears to be the ethmoid foramen. The sphen- opalatine foramen and the posterior palatine canal in Paris 1 are combined to open posteriorly through a common large fora- men in the suture between maxillary and palatine. Apparently coalescence of the foramen rotundum and the anterior lacerate foramen occurs in Paris 1 but they are separate in Paris 5. In both cases, however, the foramen rotundum lies within the orbit, as Stehlin noted (1916:1353), and not lateral to the postorbital part of the alisphenoids, as Gregory (1915 :430) suggested. The orbital region of Paris 1 is entirely undistorted and shows that the postorbital opening was small. It is of some interest that neither lemurs nor lorises, nor any other known fossil prosimians of similar size possess a smaller aperture here. As with so many cranial characters the primate showing the most interesting re- semblance to NecroUmur in respect of the structure of the post- orbital region is Tarsius. However, because of the huge flanges that encircle the orbit in mature specimens of Tarsius, and the greater degree of postorl)ital closure seen in such adults, much more revealing comparisons can be made between skulls of 12 BREVIORA No. 127 Nccrolcmnr and those of juvenile tarsiers. In the latter, the C'ircuniorbital flanges are not yet very pronounced and resemble the slight flanges seen in Nccrolcmnr (Fig. 3) . It seems necessary to assume, first, that these flanges in Tarsius are concerned pri- marily with the support of the enormous eyes, and second, be- cause of their very uniqueness, that they were not so developed in Eocene forerunners of the living genus. Nccrolemur clearly has such flanges in an incipient stage. Closure behind the orbit in Tarsius (on the outside) can be seen to proceed in successively older juveniles from three main centers : 1, ventrolaterally, by an upgrowth of the posterior mid- region of the orbital plate of the maxilla; 2, laterally, by an anteroposterior spreading of the middle of the postorbital bar ; and 3, dorsolaterally, by growth of a flange from the frontal, which arises beneath the frontal insertion of the postorbital bar. Ossification proceeds downward as this flange, or plate, grows alongside the postorbital l)ar, and it eventually fuses with the bar, leaving no sutural indication. In most specimens of adult Tarsius, the frontal and jugal components of the postorbital wall can be distinguished by the fact that the region where they fuse is much thinner and consequently more translucent. In Ncc- rolemnr, at least two of these components effecting closure ap- pear to be partially developed. Paris 1 has an uncrushed post- orbital bar which shoAvs an anteroposterior spreading at the middle, much as in Tarsius. The evidence is less clear because of breakage, but the posterior part of the orbital plate of the maxilla also bears a flange in Nccrolcmnr. The third center of closure seen in Tarsius, the frontal element, is indicated in Nccrolemur only by a distinct angulation along the lateral wall of the cranium betwen the orbital and temporal fossae. Although such characters as tlie greatly reduced paraconids and the lo.ss of certain anterior lower teeth eliminate Nccrolemur from the direct ancestry of Tarsius'^, the incipient circumorbital flanges and characteristics of postorbital closure in Necrolemnr are sug- gestive of a stage to be expected in the Tarsius ancestry. Perhaps Pscudoloris or Nanvopifhex are nearer the actual line leading to the modern form, but incompleteness of known specimens pre- vents the sort of comparisons here made with Nccrolcmnr and leaves this possibility insoluble at present. ^Even if this were not so it would be almost irrelevant, in the absence of intermeiliates, to urge an ancestor-desoendent relationship for forms so separate in geographic distribution and in time. 1960 CRANIAL ANATOMY OF NECROLEMUR 13 It may be noted in passing-, that the manner of postorbital closure in Tarsius (insofar as the malar and frontal are con- cerned) is distinct from that seen in catarrhines and platyrrhines. Closure in this area in Tarsius is chiefly effected by an outward and downward growth of a flange of the frontal (with relatively little malar expansion) while in higher Primates the greater part of the dorsolateral area of enclosure is contributed by the de- velopment of an orbital plate of the malar. These differences strongly imply that the partial postorbital closure of Tarsius only parallels that of the Anthropoidea and is not a character of their common inheritance. If it be agreed that some postorbital clo- sure arose at least twice among Primates, the possibility that this feature also was independently acquired in the ancestral platyr- rhine and catarrhine stocks can be more seriously entertained. EXPLANATION OF FIGURES The lateral and ventral views of the slaill of Necrolemnr are based on Paris 1, to which details of missing regions have been added from other specimens, principally M.C.Z. 8879. Mislead- ing stains and fractures in Paris 1 are largely omitted from these illustrations. Certain details of the anterior dentition and mandible are drawn from specimens figured by Stehlin (1916). REFEEENCES Cope, E. D. 1885. The Lemuioidea and the Iiiseetivora of tlie Eocene period of Xortli America. Amer. Naturalist, 19: -l:o7-471. Edinger, T. and D. B. Kitts 1954. The foremen ovale. Evolution, 8: 389-404. Gregory, W. K. 1915. I. On the relationship of the Eocene lemur Notliwctus to the Adapidae and to other Primates. II. On the classification and phylogeny of the Lemuroidea. Bull. Geo. Soc. Amer., 26: 419-446. 1920. On the structure and relations of Notharctus, an American Eocene primate. Mem. Amer. Mus. Xat. Hist., (n.s.) 3: 49-243. HURZELER, J. 1946. Zur Charakteristik, systematischen Stellung, Phylogenese und Yerbreitung der Xeerolemuriden aus dem europiiischen Eocaen. Eclogae geol. Helvetiae, 39: 352-354. 1948. Zur Stammesgeschiehte der Xeerolemuriden. Schweiz. palaont. Abh., 66: 3-46. 14 BREVIORA No. 127 KlAAUW, C. J. VAN DEK 1931. The auditory l)ull;i in sonic fossil iiiamiiials. Bull. Aiiier. Mus. Xat. History., G2 : 1-352 Le Gros Clark, W. E. 1959. The antecedents of man. 374 pp. Edinbiiro-h University Press. Simons, E. L. 1959. An antliropoid frontal bone from the Fayum Oligocene of Egypt: the oldest skull fragment of a higher iirimate. Amer. Mus. Xovit., no. 1976: 1-1(5. 1960. Notes on Eocene tarsioids at the British Museum. Bull. Brit. Mus. Xat. Hist., Geol. Ser., in press. Stkhlin, H. G. 1912. Die Siiugetiere des sehweizerischen Eocaens. Abh. Schweiz. Palaont. Gesell., 38: 1165-1298. 1916. Die Saugetiere des sehweizerischen Eocaens. Ihid., 38: 1299- 1552. BREVIORA Mitaseimim of Coimparative Zoology ('A.Mi'.iui)(ii;, ^Iass. I )k(kmhkr lM), nXil) Ximi'-f.r 12s SIZE OF EXDOCEUOID CEPIIALOPODS i By Curt Teichert- and Bernhard KrM:MEL3 The maximum size of fossil animal groups, ^vhether mammals, reptiles, or invertebrates has always been a fascinating subject of inquiry'', because phyletic size increase is one of the important trends that dominate the evolution of living things. In the case of large animals, tlie evidence is often hard to assemble because their remains are difficult to obtain, to transport, and to store. Squids are the largest living invertebrates and a tradition has been handed down in paleontological literature that the largest fossil invertebrates likewise arc to be found among the cephalo- pods, but few accurate data are to be found in published sources which are now readily available. Among the nautiloid cephalopods, it has long been suspected that the Endoceratida furnished the real giants, but no accurate measurements in support of this statement are available. Clarke (1897) stated that entire shells of Canicroccras pro- tfiformf, 10 to 15 feet long (3 to 5 uK^ers). had been found in the ]\Iiddle Ordovician of Minnesota, hi the same publication, Clarke figured an internal cast of part of a siphuncle, from the base of the body chamber to the adapical end of the spiess, which Avas 3 feet and 3 inches long. Miller and Kunnnel (1944) described and illustrated additional species of these :\Iiddle Ordovician endoceroids from ^Minnesota, which are deposited in the Carnegie Museum. One of their paratypes of Endoccras clarkei measured 750 mm long, is sejitate throughout and is not complete, adapically or adorally. The holotype of Endoiunts gracillimvm Miller and Kunnnel (1944) measured 070 nun in U'ublicatioii authorizea by the Pirectdr, V. S. Geological Survey. -U. S. Geological Survey, Denver, ('olo. = Museum of Comparative Zoology. 2 BREVIORA No. 128 length, again an iiicoiiiplctc six'ciiiicii consisting only of phrag- niocone. These same authors (IcscimIxmI a new species, Endoccras (lccora]i(nsf, on two ])ortions of tlic internal mold of the phrag- nioeone from the Deeorah formation, Winneskiek Connty, Iowa. The larger portion is ahout 62.") mm long and the length of the smaller measures about :>2() mm. They estimated the interval between the two pieees as about 11.1 mm, so tlie total length of this phragmoeone was about 1, ()()() mm. These authors also men- tioned that there is on display in the Chicago Natural History Museum a larger endoceroid that measures 6 feet in length. Teiehert (1927) noted the occurrence, in Middle Ordovician limestones of Estonia, of endoceroids as iiuicli as 5 meters long, but gave no further details. Flower (19.3.")) stated that specimens 12 feet in length had been collected and added that he was "not wholly inclined to discredit a report of an endoceroid found in a ([uarry n(>ar Watertown, New York, which was measured before it was broken up and found to attain a length of .30 feet." As far as we have been able to ascertain, these somewhat vague statements are all that is presently available in the ])ublished rec- ord on the subject of the maximum size of endoceroid cep- halopods. It does not seem to Ix' gcurrally kno\\n that the Museum of Comparative Zoology at Harvard laiivei-sity possesses Avhat a])- pears to Ix' the largest fragment of an eiuloceroid cephalopod on display anywlicrc^ in the world. As Flower (1955) has stated, "the removal of even reasonably coiu])lete sjiecimens involves something vei-y close to ([uarrying oj)erations, storing them is anothei- ])roblem." The specimen in the collections of the Mu- seum of Comparative Zoology is, therefore, pi-obably unique in nniseums of the Avorld. The specimen measures 3,000 nnn in length but is not com- plete, adorally or adapically. In general the preservation is fair, but as a result of weathering and crusliing the full diameter of the conch is preserved only in one jilane, and in the other ])lane the outer shell is removed exposing traces of septa and in places the siphuncle. The tirst recognizable septa are 500 mm from the adoral end but the whole s])ecinien could well ho phragmoeone as this adoral 500 nnn is slightly crushed and weathered and one cannot tell whether septa are present or absent. The adoral diameter of the specimen is 2(S() nnn. The couch ta|)ers at a uniform rate and the adai)ical diameter meas- ures 120 nnn. 1960 SIZE OF EXDOCEROID CEPIIALOPODS Figure 1 — Large endoeeroid on exhibit in the Museum of Comparative Zoology. BREVIORA No. 128 rii I he s('i)ta sl()|)c adapieally at an an^lc of about 45° and in the mid-part of tlic spcciiiicn are spaced 17 to 20 mm apart. Tlic siphunele is visible only on tlie adapieal lialf of tlie specimen. About 1,000 mm from the adoral end of tlie shell the siphunele has a diameter of about 0" mm; at l.ToO mm from the adoral end of the shell the siphunele has a diameter of 75 mm. The first endocones ap])ear 2.000 nnu behind the adoral end of the shell. The spiess measures 510 mm in length. The surface of the shell bears faint annulations that are spaced approximately 10 to 12 mm a]iart. SUMMARY OF MEASUREMENTS Length 3,000 iimi Adoral dianieter 280 mm Diameter 1,00(1 mm from adoral end 220 mm Diameter 1,750 mm from adoral end 170 mm Adapieal diameter 120 mm Diameter of siphunele 1,000 nun from adoral end 95 mm Diameter of siidium-le 1,75(1 mm fiom adoral end 75 mm Spiess length 510 mm A graphical reconstruction of the sludl on the basis of these measurements shows that the entire fossil from its presently pre- served adoral end to the a])(^\ may have measured about 5.S00 nnn. The total length of the body chamb'er is a matter of guesswork. There are tew ])ublished and illustrated records of any straight fossil cephaloiiod shells, complete from apex to aperture, which are more than a foot or so long. In short shells the ratio of body chamber to i)hragmocone may l)e high, even larger than 1:1. With increasing total hmgth of conch, however, ratio of body chamber to ]ihragm()cone is likely to decrease, although no defi- nite figures can be stated, in a specimen of Act{)ioccras beloifoisc (Foerste and Teichert, 11)80, pi. 28), which was 450 mm long, the ratio of body chandicr to phragmocone was about 1 :2. Leith (1942) described a specimen of Ldinlxoccras lamhii (Whiteaves) which was 45.5 in. (1,155 mm) long. lie estimated the total length of the shell at 1,405 mm. The body chamber was almost wholly preserved and not more than 250 mm long. Ratio of body chamber to iihragmocone was thus 1 :4.6 in this specimen. It shotdd be noted, however, that both Acfi)toc€ras beloitensc and Laiiibcoccrds hinthii have body chambers with constricted 1960 SIZE OF ENDOCEROID CEPHALOPODS 5 apertures, whereas no endoceroids Avitli constricted apertures are known. It seems physiolop'ieally plausible that in large straight ccphalopod shells the animal shonld have a Ijetter "grip" on a body chamber with constricted ai)erture than on one with an unconstricted aperture; therefore, in shells which expanded uniformly from the ai)ex to the apertui-c. like the endoceroids, the animal itself, and thus its body chamber, should have been relatively larger. In a juvenile specimen of a straight ammonoid, Baculitcs o vat us, Trueman (1941) determined the ratio of length of body chamber to phragmocone as 1 :0.7, but in adult shells this ratio becomes much smaller. If we assume the ratio of length of body chamber to phragmocone in endoceroids to be more like that of Actinoceras hdoitense we arrive at a length of the body chamber for the Harvard Enduceras of 2,650 mm and for the en- tire shell of 8,150 mm, or 28 feet. This is a conservative esti- mate, yet close to the possible maximum figure of 30 feet men- tioned by Flower. Add to this the length of the tenacles which must have ex- tended a considerable distance in front of the aperture, cer- tainly no less than half the length of the l)ody chamber, and we have an invertebrate animal considerably longer than 30 feet — a truly imposing size. Today's giant squid, Architeuthis, rivals and slightly exceeds in length the largest extinct endocer- oids. Sparck (1928) records specimens of Architeuthis dux from tlie North Atlantic, washed ashore on the Norwegian Coast, that have body lengths of up to 2 meters and tentacles as much as 10 meters long. The largest specimen to our knowledge is that of Aychiteuthis harveyi? recorded by Yerrill (1879, p. 196) which measures 624 inches (17 meters). There is a model in the ]Museum of Comparative Zoology of a specimen of Archi- t(itthls j)rl)u-(ps. which was washed ashor<' in Newfoundland, which measures about 15 meters in length. ^Nlore recently. Lane (1960, pp. 198-227) has critically reviewed a larger number of reports of tinds of and encounters with giant squids. He is in- clined to believe that individuals of Aychit( tffhis or some other genus, as yet undescribed, may reach overall hnigths of some 70 feet. While the Harvard specimen represents by far the largest nautiloid cephalopod on which accurate data are now available, it is interesting to compare it with the largest aiuinonoid on record. This is Pachydiscus seppenradense Landois from the 6 BREVIORA Xo. 128 Upper Cretaceous of western (!ei-iiiaiiy (Landois, 1895, 1898). lu 1895, Landois fii-st described this fossil amiiiouoid whose shell was 1,800 iiini in diameter and in which tlie last camera was 550 mm high. Landois' reconstruction provitled the animal with a body chambei- e(jnivai(Mit to only one-fonrth of a complete whorl. From this he concluded that Uie total diameter of the complete specimen of his ammonoid had been about 2,550 mm. From later studies (Tnieman. 1941) it is. however, likely that Landois' estimate of the len Lower Si!uri;iii ( 'ciili.-ilopocUi of ^liinu'.sota. Geology of Minnesota, vol. o, pt. 2, Paleontology, jip. 7(il-812, pis. 47-60. Flower, li. IT. ]955. Status of eniloccroiil classification. Jour. I'aleoutology, vol. •29, no. ?,, lip. 3119-:;71. FoERSTE, A. F., and Teickert, Curt 1930. The actinoeeroids of east-central Xorth America. Denison Univ. Bull., Sei. Lab. Jour., vol. 25, pp. 201-296, pis. 27-59. Laxdois, H. 1895. Die Riesenannnoniten von Sepjienrade, Pachj/disou!^, Zittel, Sep- penradense H. Landois. Westfal. Provinzial. Ver. Wiss. und Kimst. f. 1894/95, vol. 23, pp. 99-108, 2 pis. 1898. Gewichtsverluiltnisse der Riesen-Anmioniten. Ih'ul., vol. 26, pp. Laxe, F. W. 1957. Kingdom of the Octopus. Javrolds, London. 286 pp. Leitit, E. I. 1942. Xotes on the cephalojiod Ldmbcoccras lamhii from Manitoba. Jour. Paleontology, vol. 16, no. 1, pp. 130-132, 1 text-fig., pi. 22. Miller, A. K., and Ktmmel, Berxiiakd 1944. Rome large straight Ordovician cephalopods from Minnesota. Annals Carnegie Museum, vol. 30, pp. 19-38, 4 pis. Sparck, Ragiiar 1928. Xordens Pyrcverden. Henrik Koppel, Copenhagen, 658 pp. Teichert, Curt l!t27. Der estliindische Glint. Xatur u. ]\ruseum, vol. 57, pp. 264- 272, 7 figs. 1930. Biostratigraphie der Porauiboniten. Xeues Jahrb. f. Mineral, etc., Beil. Bd., Abt. B, pp. 177-246, 8 tigs., 4 pis. True.max, a. E. 1941. The ammonite liody-chaniher, with special reference to the buoyancy and mode of life of the living ammonite. Quart. Jour. Geol. Soc. London, vol. 96, pp. 339-378. Verrill, a. E. 1879. The cephalopods of the northeastern coast of America. Trans. Conn. Acad. Sci., vol. 5, pp. 177-476, pis. 25-56. BREVIORA Miaseiim of Compsirative Zoology Cambridge, Mass. December 21, 1960 Number 129 TYPE AND TYPE LOCALITY OP THE GULF COAST SPINY SOFTSIIELL TURTLE, TRIONYX SPINIFER ASPER (AGASSIZ) By Robert G. AVebb Museum of Natural History, The University of Kansas The currently accepted type locality of Trionyx spinifer asper, Lake Concordia, Louisiana, is in an area of intergradation be- tween three subspecies of Trionyx spinifer. Of the nine available syntypes of asper, none has been designated as a lectotype, and only one of the syntypes of T. s. asper is recognizable as belong- ing to that taxon. Abbreviations of names of museums from which specimens are mentioned are : KU, Museum of Natural History, University of Kansas ; MCZ, Museum of Comparative Zoology, Harvard College ; SM, Strecker Museum, Baylor University ; TU, Tulane University ; and USNM, United States National Museum. Agassiz (1857:406) described Aspidonectes (= Trionyx) asper as having : "... very coarse and large tubercles of the front and hind part of the carapace, which extend, behind, even over the bony shield, and are there supported by prominent warts of the bony plates. These bony warts exist in no other species with which I am acquainted: their form is very irregular, sometimes oblong and sometimes orbicular ; they also project more or less. Another marked peculiarity of this species consists in the greater bluntness of the extremi- ties of the jaws, which are more rounded than in Asp. spini- fer. The jugal arch is also broader. The difference between the males and the females is more striking in this species than in any other, the males being regularly oval, whilst 2 BREVIORA No. 129 the females are almost circular in their outline ... in younger specimens of Asp. asper there are . . . two or three l)lack lines separating the pale rim of the posterior mar- gin Several of the syntypes show "prominent warts of the bony plates" {supra), which appear posteriorly and principally on the seventh pair of pleurals toward the midline on the bony carapace (Plate 1). I have seen bony elevations, which are circular or elongate resembling short ridges, on the carapaces of specimens of Triomjx fcrox (USNM 4373, 55316, 62217) ; a photograph of a fcrox (Stejneger, 1944 :pl. 7) clearly shows bony prominences on the posterior part of the bony carapace. These bony elevations also occur in populations of T. spinifer (SM 2552, 2558, Texas; USNM 54731, Iowa; USNM 100396, 100404. Louisiana) . Bony warts and ridges on the posterior part of the carapace are not diagnostic for T. s. aspcr. To my knowledge, the subspecies of Trionyx spinifcr are not distinguishable by characteristics of the skull, bony carapace or plastron. T. s. aspcr closely resembles the subspecies hartwcgi and spinifcr but differs in usually having two or more blackish lines paralleling the rear margin of the carapace, and usually in having the postocular and postlabial stripes united on the side of the head. Other characters mentioned above liy Agassiz seem not to be of taxonomic worth. Agassiz (op. cif .-.4:05-06) did not designate a type, mention l^rceise localites, or state the number of specimens that formed the basis of his description of Aspidonectes asper. Baur (1893: 220) restricted the type locality of aspcr to Lake Concordia, Louisiana, but did not mention any specimens ; presumably Baur's action was based upon an examination of the bony cara- pace of USNII 012349, which shows the prominent bony ridges described and considered diagnostic by Agassiz and has "Lake Concordia, Louisiana" written in ink on the underside of the carapace. Barbour and Loveridge (1929:225) listed MCZ 1597 and 1622 as cotypes. Stejneger (op. cit. :56-58) discussed some of the syntypes of asper, and also regarded the type locality as "Lake Concordia, La.," designating USNM* 12349 ( = 012349) and MCZ 37173 as "cotypes." The type locality of T. s. asper is currently accepted as Lake Concordia, Louisiana (Schmidt, 1953:109). T. s. asper intergrades w-ith T. s. hartwcgi and T. s. spinifer in the lower Mississippi Valley (Conant and Coin, 1948:11). 1960 TRIONYX SPINIFER ASPER 3 The softshell turtles inhabiting the Mississippi River and its tributaries in Louisiana (including Lake Concordia) and Mis- sissippi represent an intergrading population of spmifer and hortwegi, and, to a lesser extent, asper. Most turtles from the Pearl River drainage and rivers that drain into Lake Ponchar- train adjacent to the east are typical of asper. Lake Concordia is a large oxbow on the west side of the Mississippi River in Concordia Parish, Louisiana. 1 was a member of a field party, from Tulane Uiiiversit}^ which collected three specimens (TU 16524, 16524.1, 16524.2) of Trionyx spinifer from Lake Con- cordia on August 1-3, 1954. Because none of these turtles has the postocular and postlabial stripes united on the side of the head or any indication of more than one marginal line parallel- ing the rear margin of the carapace, none is considered referable to asper. Some of the nine syntypes discussed below were mentioned by Stejneger (op. cit.:51-58). (1) USNM 012349 (Plate 1) is represented by a bony cara- pace and may be considered the present type (lectotype), although never designated as such. ''Trionyx Ferox?, Lake Concordia, Louisiana, BLC Wailes, 1851" is written in ink in the same handwriting on the underside of the bony carapace. Suljsequently, "Ferox?" has been crossed out and "asper" added in pencil, and "012349" inked on tlie second pleural. A gummed label is pasted on the fifth rib, right side, and bears the inscription "asper Ag. (Type)." There is also an attached metal tag bearing the number 22676. The carapace has a maxi- mal length of 14.9 cm. and width of 12.7 cm. ; its size suggests that it is that of a female. On the seventh pair of pleurals and in line with tlie longitudinal sutures of the neurals are two, short, longitudinal, elevated, bony ridges. (2) USNM 01086 is represented by an intact bony carapace (19.8 cm. in length and 16.5 cm. in width), disarticulated parts of the bony plastron (epiplastron and preplastra lacking), and a skull including the lower jaw. There is an attached paper tag labeled "cotype." "Miss." is written in ink on the underside of the carapace. There are small, slightly-elevated bony warts on the sixth pair of pleurals (with few, less conspicuously de- veloped, on the seventh pair), and a semblance of ridging as seen on the carapace of USNM 012349. "\Vashington Lake" (crossed out in pencil), and "1086" are written in ink on the skull, which has a basicranial length (occipital condyle to tip 4 BREVIORA No. 120 of upper jaw) of 62.8 mm. Tlie locality "Washington, Adams Co." is written in pencil on the eard for USNM 01086 in the card file in the USNM. The specimen is certainly a female judging from the size of the skeletal parts. (3) USNM 01084 is represented only by the skull (lacking lower jaw) of a female, which has a basicranial length of 63.1 mm., and the number "1084" and "Wash." inked on it. The skull is presumably from the same locality as USNM 01086. (4) MCZ 46621 is represented only by the skull (lower jaw present) of a female having a basicranial length of 63.0 nnu. ; it was received from B.L.C. Wailes. (5) MCZ 1622 has been photographetl by Stejneger (op. cit.: pi. 16), who mentioned a label in lead pencil, "No. 1622. Ti/pe Amyda asper (Agassiz) Lake St. John, Miss. W. Sargent leg. et don." (which I have not seen) {op. cit. :5S). Stejneger sug- gested that Lake St. Jolm was in Louisiana (op. cit. :footnote 2) ; however, tlie caption for Plate 16 reads Lake John, Florida. Coin (1948:304), commenting on the last -mentioned locality, wrote that MCZ 1622 "... is listed in the museum catalogue as having come from Lake St. John, Mississippi." Lake St. John is a few miles north of Lake Concordia occurring in Concordia and Tensas jiarislies. Louisiana, in the Mississip]ii Uiver drain- age. MCZ 1622 is a young alcoholic specimen, probably a female, having a plastron 5.0 cm. in length, and a carapace 6.8 cm. in length and 6.0 cm. in Avidth. The postlabial and postocular stripes do not join on either side of the head. The undersurface is more heavily pigmented than that which 1 judge to be "nor- mal" for T. .s\ a^pcr. The most significant character is the lack of more than one dark marginal line posteriorly on the carapace ; there are small blackish dots posteriorly but these are Avidely- separated and not closely-set in a linear fashion to suggest a second marginal line. Neill (1951:19-20) believed this specimen to show "... some approach to spinifera.''^ I do not regard this specimen as representative of T. s. asper. Schwartz (1956:15), however, referring to MCZ 1622, stated that it ". . . differs in no way, so far as head and carapace pattern is concerned, from similarly sized specimens of T. s. ogassizi (= asper) from South Carolina and Georgia." It would be surprising to find specimens from South Carolina and Georgia having a pattern on the carapace resembling that of MCZ 1622 : Schwartz's photograph of a juvenile from South Carolina (op. cit. -.14:, pi. 3") is certainly asper. 1960 TRIONYX SPINIFER ASPER 5 (6) MCZ 46633 bears the locality '' Washington, Mississippi" written in ink on an attached paper tag. In a different hand, the data ' " AVashington, Miss. Pres. by B. L. C. Wailes" and the number "9" is written in ink on an unattached, folded piece of stiff, brownish paper that is held in place l)y a metal rod against the plastron. The data suggest that this specimen is from the same locality as USNM 01084 and 01086. MCZ 46633 is a stuffed, adult male having a plastron approxi- mately 11.0 cm. in length and a carapace 16.0 cm. in length. The postlabial and postocular stripes fail to join by a narrow segment on the left side of the head, but appear to be in contact on the right side. Bony, longitudinal welts, resembling those on the bony carapace of USNM 012349, appear posteriorly on the seventh pair of pleurals. The carapace is darkened laterally and sprinkled with widely -spaced, black dots ; there is no indica- tion of a second marginal line. The pattern on the carapace most closely resembles that of T. s. hartwegi. Therefore, I do not consider this specimen to be representative of T. s. asper. (7) MCZ 46615 bears an attached label having the following data written in ink: "Triouyx ferox SOUTHERN SOFT- SHELLED TURTLE Natchez, Miss. H. Wheatland coll'n." Subsequently '"Trionyx ferox" has been crossed out and ''Amyda aspera" written in ink; the name ''spinifera" also has been added in pencil. The reverse side of this label bears the number "8" and the name "B. Chase" written in pencil. A small piece of paper also bearing the number "8" is pasted on the left side on the bony carapace. Presumably this specimen has some relationship to MCZ 46633 that bears the number 9. MCZ 46615 is a large, stuffed adult female having a plastron that measures approximately 33.0 cm. and a carapace 43.0 cm. in length ; the anterior edge of the carapace is studded with conical tubercles. Scattered, oblong, bony elevations adorn the seventh pair of pleurals. There is no pattern evident on the carapace, and the striping on both sides of the head is obscure. This large female might represent any of the subspecies spinifer, hartwegi or asper; there is no character that identifies the speci- men as T. s. asper. (8) MCZ 37173 is the stuffed specimen that Agassiz mentions liaving received from the University of Oxford (at Oxford, Lafayette County), Mississippi (op. cit. AOo) ; the specimen is discussed by Stejneger (op. cit.-.ol). The plastron is approxi- mately 23.5 cm. in length; the length of the carapace, wrinkled 6 BREVIORA No. 129 posteriorly, measures (in a straight line) approximately 29.5 cm. Although the tail extends noticeably beyond the posterior edge of the carapace and is presumably the basis for Stejneger refer- ring to MCZ 37173 as an adult male, its large size indicates that it is a female. A red paper label bearing the writing "A. asper Ag. Cotype" is pasted on the plastron. Elevated bony knobs appear toward the posterior margin of the bony carapace on the seventh pair of pleurals. There are well-defined stripes on the head, but the relationship of the postocular and postlabial stripes is obscure. There is no evidence of more than one dark marginal line paralleling the rear margin of the carapace. Hence the siDccimen is not recognizable as T. s. asper. (9) MCZ 1597, a large alcoholic female, is considered repre- sentative of T. s. asper (Plate 2). A paper label on the left foreleg bears the inscription "Natches, Miss W. Sargent." The carapace, measuring approximately 43.0 cm. in length and 37.0 cm. in width, has more than one dark marginal line and several ocelli. Inner marginal lines in the posterior right and left quadrants are mostly continuous, but are obscured by the wrink- ling and scratching on the posterior part of the carapace. The tubercles on the anterior edge of the carapace are worn and resemble rounded knobs ; more lateral tubercles are equilateral. The seventh pair of pleurals are in contact medially behind the seventh neural. There are indications of raised bony welts on the last pair of pleurals. The head is partly extended, but does not show the relationship of the stripes on the side of the head. The pattern on the snout is obscured, and blackish marks are evident on the dorsal surface of the limbs. The plastral surface lacks dark markings and measures approximately 32.5 cm. in length. MCZ 1597 is discussed by Stejneger, who mistakenly refers to the specimen as a male {op. cit.:58). MCZ 1597 (Plate 2) is herewith formally designated as lectotype of Trionyx spiyii- fer asper (Agassiz), as it alone, of the nine specimens hitherto considered syntypes of Aspidonectes asper Agassiz, is recogniz- able as referable to the subspecies asper. Agassiz (1857:405) mentions having received specimens from Mr. AVinthrop Sargent of Natchez, Mississippi (MCZ 1597, 1622, 46615), Dr. L. Harper of the University of Oxford, Mis- sissippi (MCZ 37173), and Professor B. L. C. Wailes of Wash- ington, Mississippi (MCZ 46621, 46623, USNM 01084, 01086, 012349). These localities suggest places of residence and are not necessarily localities at which the specimens were captured ; at least two syntypes (MCZ 1622, USNM 012349) are from locali- 1960 TRIONYX SPINIFER ASPER 7 ties different from those mentioned immediately above. Al- though forwarded to Agassiz from Natchez, the specific locality from which the lectotype (MCZ 1597) was captured is unknown. Natchez, Oxford and Washington, Mississippi, are in the drain- age basin of the lower Mississippi River, w-hich is inhabited by softshells that are intergrades between T. s. spinifer and T. s. hartwcgi, although few specimens from there are typical of T. s. asper. It is possible, but unlikely, that MCZ 1597 was captured at Natchez. Those syntypes having a discernible pat- tern on the carapace, which is not that of asper, probably came from tributaries in the lower Mississippi River drainage. The lectotype, MCZ 1597, probably came from the Pearl River drainage (adjacent eastward from the Mississippi River drain- age), where most individuals are representative of T. s. asper. Occasional specimens of asjJer from the Pearl River drainage have only one dark line paralleling the rear margin of the carapace, and resemble softshells occurring in the Mississippi River drainage. An adult male given to me by William E. Brode was stated by him to have come from the Pearl River ; this turtle (KU 47120) has only one dark marginal line, and Stej- neger (op. cit. :64) mentions another from the Pearl River drain- age. Schwartz (op. cit. -.16) and Crenshaw and Hopkins (1955: 20) write that some specimens from Georgia have only one solid line at the margin of the carapace. However, most of the softshells inhabiting the Pearl River drainage are typical asper, and the Pearl River drainage is probably the provenance of the lectotype, MCZ 1597. The type loealitj^ of Trionyx spinifer asper (Agassiz), repre- sented by the lectotype, MCZ 1597, is herewith designated as the Pearl River at Columbus, Marion County, Mississippi. The geographic range of T. s. asper (Platypeltis agassizi Baur con- sidered a synonym) is the southeastern United States except peninsular Florida from the Florida parishes of Louisiana east to southern North Carolina ; in streams of the Gulf Coast drain- age including that of Lake Ponchartrain, Louisiana, eastward to the Apalachicola River system, and those of the Atlantic Coast drainage including that of the Altamaha River in Georgia northward to the Pee Dee River drainage in South Carolina. 8 BREVIORA No. 129 LITERATURE CITED Agassiz, L. 1857. Contributions to tlie natural history of the United States of America. A^ol. I. Part II. North American Testudinata. Little, Brown and Co., Boston, pp. 233-452d. Barbour, T. and A. Loveridge 1929. Tyijical reptiles and amphibians. Bull. Mus. Comp. Zoo!., 69(10) :205-360, Baur, G. 1893. Notes on the classification and taxonomy of the Testudinata. III. The genera of the Trionychidae. Proc. Amer. Phil. Soc, 31:213-221. Conant, E. and C. J. GoiN 1948. A new subspecies of soft-shelled turtle from the central United States, with comments on the application of the name Amyda. Occas. Pap. Mus. Zool., Univ. Michigan, 510:1-19, 2 pis. Crenshaw, J. W., Jr. and M. N. Hopkins, Jr. 1955. The relationships of the soft-shelled turtles Trionyz fcrox ferox and Trionyx fcrox aspera. Copeia, 1955, No. 1:13-23, 4 figs. Goin, C. J. 1948. The occurrence of Amyda spinifcra aspera in Florida. Copeia, 1948, No. 4:304. Neill, W. T. 1951. Taxonomy of North American soft-shelled turtles, genus Amyda. Publ. Res. Div. Ross Allen's Rept. Inst., l(2):7-24, 1 fig. Schmidt, K. P. 1953. A check list of North American amphibians and reptiles. Sixth edition. Amer. Soc. Ichth. Herp., LTniv. Chicago Press, pji. viii +280. Schwartz, A. 1956. The relationships and nomenclature of the soft-shelled turtles (genus Trionyx) of the southeastern United States. Charleston Mus. Leaflet, 26:1-21, 1 fig., 3 pis., 2 maps. Stbjneger, L. 1944. Notes on the American soft-shell turtles with special reference to Amyda agassizii. Bull. Mus. Comp. Zool., 94(l):l-75, 30 pis., 10 tabs. Transmuted January 19, 1960. Plate 1 (X appiox. Vs). Bony carapace of Trionyx spinifer, USNM 012349, from Lake Concordia, Louisiana; parts of ribs that extend beyond carapace on right side are detached. Top. — Dorsal view showing elevated prominences posteriorly that Agassiz considered diagnostic for Aspidonectes asper. Bottom. — Ventral view showing inscriptions. Plate 2 (X appiox. Yo). Dorsal view of lectotype of Trionyx spinifer asper, MCZ 1597 ; locality designated as the Pearl River at Columbus, Marion County, Mississippi. BREVIORA Mnaseiam of Compsirative Zoology Cambridge, Mass. December 22, 1960 Number 130 THE MECHANISMS OF CAEAPACIAL AND PLASTRAL HINGES IN CHELONIANS By R. V. Shah Department of Zoology, University of Baroda, Baroda, India Deraniyagala (1930) described the carapace and plastron of Lissctnys and reported that there are two movable corselets in the carapace and four in the plastron. In the carapace the pre- nuchal is looseh^ connected to the anterior border of the nuchal by a thick ligament. This anterior corselet is moved downwards and acts as a valve. The last two marginals form a posterior movable valve in the carapace. In the plastron, the epiplastral lobe acts as an anterior plastral valve, and a small portion of the posterior part of the plastron demarcated by a transverse groove forms the posterior plastral valve. Besides these, there is a pair of cutaneous flaps which are used as valves to cover the posterior limbs after they are retracted under the shell. These flaps are known as the femoral valves. Presence of the cutaneous femoral valves is a distinguishing character of the subfamily Cyclan- orbinae, in which the three genera Lisscmys, Cyclanorhis and Cycloderma are grouped. Hasan (1941) worked out the shell- closing mechanism and described the principal muscles respon- sible for the movements of the six movable parts in Lisscmys. George and Shah (1955), while working on the musculature of Lisscmys, gave pos.able homologies of the muscles described by Hasan. In the present study an attempt is made to describe the closing mechanism of tlie shell in most ^ of the chelonians in which part of the plastron or carapace is movable and acts as a valve. Care- 1 Xo adult was available of Kotochcli/s platynota in which an indistinct trans- verse hingp has been reported. Similarly there is only a juvenile at hand of Pyxis arachnoidcs. Siebenrock has described in this species a mobile anterior plastral lobe in which the entoplastron becomes involved. Since hinges only develop with age, it has not been possible to investigate hinge mechanisms in these species. 2 BREVIORA No. 130 ful dissections of the muscles involved in the movements of the plastral lobes or the carapace parts were made of the following chelonians: Cuora amhoinensis, Emys orbicularis, Emydoiden hlandingi, Terrapene mexicana yucatana, T. Carolina Carolina, T. c. haurii, T. c. triunguis, T. ornata, Testudo graeca, T. hermanni, T. kleimnayini, Kinixys erosa, Kinosternon iaurii, K. cruentatum, K. flavescens spooneri, K. herrcrai, K. leucostomum, K. scorpioides scorpioides, K. suhruhrum, Sternotherus carinatus minor, 8. odoratus, Cycloderma frenatum, Pelusios adansoni, P. carinatus, P. castaneus, P. nanus, P. suhnigcr. All specimens examined are in the collection of the Museum of Comparative Zoology. Only two genera of the chelonians listed above, viz., Cycloderma and Kinixys, show some part of their carapace movable ; the others have some part or parts of their plastra movable. In this regard all the chelonians used for the present study could be classified in four different groups : The first group has only the anterior plastral lobe hinged and capable of valvular movements. The hinge is usually situated at the joint of hyo- and hypoplastral plates. The species of emy- dines examined show this condition. The second group includes those that have only their posterior plastral lobes movable. The posterior plastral lobe is hinged at the level of the joint of the hypo- and xiphiplastral plates. This group is composed of the species of Testudo cited above. The third group includes those which have both anterior and posterior plastral lobes movable. The kinosternines examined belong here. The fourth group consists of the members of the subfamily Cyclanorbinae in which besides the movable anterior and pos- terior plastral lobes there are the two cutaneous femoral valves. MOVEMENTS OF CARAPACE VALVES In Cycloderma and Cyclanorhis, as in Lissemys, the prenuchal is movable. Hasan (1941) has described as the M. nucho- prenuchalis the muscle which is responsible for the movements of the prenuchal. This muscle arises from the under surface of the nuchal and is inserted on the posterior half of the under surface of the prenuchal. When this muscle contracts, the prenuchal is pulled downwards and acts as a small valve to close the shell after the head and the fore limbs are drawn under the shell and the anterior lobe of the plastron is lifted upwards. George and Shah IHGO TURTLE HINGES 3 1,1955) hoinologised this muscle with that part of the spinalis- semispinalis system occurring in this region. In no other chelon- ians are there such movable prenuchals. The last two marginals in Cydoderma and Cyclanorhis, like those in Lissemys, are movable and act as a supracaudal valve. The )na)-cjino-i)ifracaudaIcs are responsible for the downward pull (Hasan 1941). These muscles arise from the undersurface of the marginals and are inserted on the flexible subcaudal valve of the plastron. AVhen these muscles contract the two movable marginals are pulled downwards and at the same time the subcaudal valve jf the plastron is pulled upwards. The joint action of these two valves, i.e. the subcaudal and supracaudal valves, close the shell from behind and protect the tail and surrounding soft parts. These muscles are homologised with part of the flexor caudac supcrficialis by George and Shah (1955). Tn Kinixys erosa an entirely different situation is present. Siebenrock (1916) has described the morphological features of the hinged carapace of this species in great detail. Briefly, how- ever, it may be mentioned that the hinge is at the level of the fourth and fifth costal plates, the posterior part of the carapace being moved up and down like a valve at this point. This move- ment of such a large part of the carapace is unique to this genus. The closing movement is brought about in the following man- ner. First the head and the neck are pulled under the shell as a result of the contraction of the rctrahens capitis coUiqiic, the rectus capitis cervico-plastralis and rectus cervicis muscles. Of these muscles the retrahens capitis coUiqiie is the principal muscle responsible for the withdrawal of the head and the neck. The main part of this muscle in Kinixys erosa arises from the body of the seventh and eighth dorsal vertebrae and the undersurface of the costal plates near these vertebrae. The muscle inserts on the cervical vertebrae and on the base of the skull at its ventral side. After the head and neck are retracted fully, the muscle contracts still further and this additional contraction pulls the hinged posterior part of the carapace downwards and closes the shell from behind. In addition, the pelvic girdle muscles, i.e. the attrahens pelvium, also help in pulling the hinged part of the carapace downward. The attrahens pelvium is very highly devel- oped in this genus and when it contracts the entire pelvic girdle is pulled forward, and since the girdle is firmly attached to the carapace the carapace is pulled with it. Although this pull is not so great as the one exerted by the retrahens capitis collique 4 BREVIORA No. 130 it does contribute to the movement. A few muscles of the legs, i.e. the ilio-tihialis and ilio-femoralis which in part arise from the costal plate, also exert a pull on the carapace after the legs are fully retracted under the shell. Thus a combined pull resulting from the contraction of all these muscles is responsible for bring- ing about the valvular movements of the posterior hinged part of the carapace. MOVEMENTS OF THE PLASTRAL LOBES Anterior plastral valve : The anterior plastral lobe is hinged at the level of the hyo-hypoplastral joint in Emys orbicularis, Emydoidea 'blandingii, Cuora amhoinensis, Terrapene mexicana yucatana, T. Carolina Carolina, T. c. haurii, T. c. triunguis, T. ornata, Pelusios suhniger, P. castaneiis, P. carinatus and P. adansonii. This anterior plastral lobe can be moved upwards like a valve which shuts off the shell in front after the head, neck and the fore limbs are retracted. This upward movement of the an- terior plastral lobe is brought about in the following manner : After the head, neck and the fore limbs are retracted, the clavic- ular part of the deltoideus muscle contracts and pulls the plastral lobe upwards. The anterior part of the 'pecioralis muscle, which has its origin from the plastral lobe, also on contraction pulls the plastral lobe upward. Some part of the scalenus muscle which is inserted on the border of the plastral lobe helps in puULng the lobe upward. The precoracoid cartilage is attached to the ento- plastral plate by a thick ligament and so when the whole girdle is rotated forward with the scapula acting as a fixed point the precoracoid and the coracoid which normally lie in a horizontal plane come now into an inclined plane with the precoracoid lifted upwards. Along with these the plastral lobe also gets pulled upwards and thus the rotation of the girdle helps in lifting the plastral lobe. These combined actions bring about the valvular movement of the anterior plastral lobe. Normally the anterior plastral valve has a straight transverse hinge. This disposition of the hinge is very necessary for free valvular movements of the plastral lobe. But to this normal condition there is an exception in Pelusios adansonii. In this animal the hinge is not a straight transverse one but is V-shaped with the apex directed posteriorly. Because of this type of hinge the valvular movements of the anterior plastral lobe in this animal are very restricted compared to those possible with a straight transversely placed hinged. 1960 TURTLE HINGES 5 Posterior plastral valve : A hinged posterior plastral lobe occurs in Tcsfiido Ideinmanni, Testudo gracca and Testudo her- manni. The hinge is at the level of the hypo-xiphiplastral joint and is also transversely placed. The posterior plastral lobe is pulled upward as a result of the contraction of the retrahens pclvium muscle of the pelvic girdle. This muscle arises from the spine of the pubis and is inserted on the entire inner surface of the xiphiplastral plate. In these animals the muscle is more highly developed than in forms which do not have a hinged posterior plastral lobe. Besides the action of the retrahens pel- vium muscle the skin connecting the legs and the plastron exerts some pull on the plastral lobe which helps in pulling it upward. Anterior and posterior plastral lobes : Kinosternon subrubrum, K. flavescens spooneri, K. leucostomum, K. scorpioides scorpi- oides, K. baurii, K. crucntatum, K. herrerai, Sternotherus carin- atus minor, and Sternotherus odoratus possess anterior as well as posterior plastral lobes which are hinged and are capable of valvular movements. The upward movements of these two plas- tral lobes are brought about in the same w^ay as described sepa- rately in the first two groups which had only one lobe, whether anterior or posterior, movable. In Kinosternon, as far as the anterior plastral lobe is con- cerned, there is an additional factor pulling the plastral lobe upward. There are thick tendons which loosely connect the fourth, fifth and sixth cervical vertebrae wdth the plastral lobe. The point of attachment of these tendons on the plastral lobe is at the border of the epiplastral plates. When the head and the neck are pulled under the shell as a result of the contraction of the retrahens capitis collique (the main retractor of the head and the neck), an indirect pull is exerted on the plastral lobe because the latter is connected by thick tendons with the cervical vertebrae. Such a type of tendinous connection between the cervical vertebrae and the plastron is not met with in any of the other chelonians studied. The Terrapene species examined have been placed in the first group in which only one anterior plastral lobe is hinged and movable like a valve. These species, however, require special at- tention since, although there is no true posterior plastral hinge, the posterior lobe is capable of being pulled dorsally and can close the shell posteriorly as efficiently as in those which have the posterior plastral lobe hinge. Here the bridge between the cara- pace and the plastron is very short and a large portion of the posterior part of the plastron is left unsupported. The retrahens BREVIORA No. 130 pelviuni muscle in this animal is very highly developed and so when it contracts the posterior plastral border is lifted upward. This lifting occurs because the plastral plate bends under the pull exerted by the retrahens pelvium muscle. In this way, although there is no hinge at the level of the hypo- and xiphiplastral joint, the plastral plate closes the shell from behind, after the hind limbs and the tail are retracted within the shell. Four plastral valves : The fourth group of the chelonians in- cludes those which possess four plastral valves capable of valvular movements — the members of the subfamily Cyclanorbinae — Lissemys, Cyclanorhis and Cycloderma. Here the anterior plas- tral lobe is not as well demarcated and hinged as in the animals described above, but since the plastron is soft the movement is effected at the level of the posterior border of the entoplastral plate and the anterior border of the fused hyo-hypoplastral plate. This anterior plastral valve is moved upward by the con- traction of the specially developed muscle — the nucho-plastralis (Hasan, 1941), described as the trapezius by George and Shah (1955). This muscle arises from the nuchal plate, runs along the base of the neck, and is inserted on the epiplastral plate. When the trapezius {—nucho-plastralis) muscles contract the anterior plastral lobe is lifted upward to close the shell. In addi- tion, other muscles are also taking part in the closing mechanism of the anterior plastral lobe ; these are : the clavicular part of the deltoideus, the rectus capitis cervico-plastralis, and the part of the pectoralis muscle which has its origin on the movable part of the plastron. The skin connecting the legs and the plastron also exerts some pull after the legs are retracted under the shell. The trapezius muscle is found only in the Trionychidae. Thus, though the anterior plastral lobe cannot be completely closed in Trionyx gangeticus, to a certain extent it can be lifted up to pro- tect the retracted head, neck and fore limb. In this animal, as in the Cyclanorbinae, the trapezius muscle is present and works in a fashion similar to that in Lissemys (George and Shah 1955). The infracaudal valve in all the members of the subfamily Cyclanorbinae is a small posterior part of the plastron which is demarcated by a transverse groove and which lies behind the xiphiplastral plates. Here also there is no regular hinged con- dition. Part of the flexor caudae superficialis muscle described for the closing mechanism of the supracaudal valve in the carapace of these animals is used for moving the infracaudal valve upward. Thus, when this muscle contracts the supracaudal as well as the 1960 TURTLE HINGES 7 infracaudal valves are moved in opposite directions towards eacli other to close the shell posteriorly. For the protection of the hind limbs in these animals an extra pair of cutaneous flaps is developed. These are known as the femoral valves. Each femoral valve is more or less hinged with the side of the xiphiplastral plates and is capable of valvular movements. The closure of the femoral valves is brought about when the ligament connecting the shank of the leg and the valve is tensed, as the legs are retracted under the shell. The skin connecting the legs and the valves also exerts some pull on the valves when the skin is drawn in as the legs are pulled under the shell. Finally, it may be said that some chelonians have developed certain features like movable plastral or carapace parts by which they can close themselves in a box to protect their soft parts. Besides this, wherever the posterior plastral lobe or the posterior part of the carapace is movable, this capacity for movement also serves to facilitate the egg-laying process by widening the gap between the carapace and the plastron. ACKNOWLEDGMENT I am grateful to Dr. A. S. Romer for giving me all possible facilities to carry out this work at the Museum of Comparative Zoology at Harvard Universit}'. I am also grateful to Dr. E. E. Williams for his valuable suggestions in preparing this paper. This work was done during the tenure of a Fulbright Smith- Mundt scholarship. EEFERENCES Debaniyaqala, p. E. P. 1939. The Tetrapod Reptiles of Ceylon, Vol. 1. Testudinates and Croc- odilians. xxxii-l- 412 pp. Colombo. George, J. C. and E. V. Shah 1955. The myology of the chelonian trunk and tail. J. Anim. Moiph. Physiol., vol. II, no. 2, pp. 49-64. 1957. The myology of the chelonian limb. 1. The forelimb of Lisseviya punctata. Ibid., vol, IV, no. 2, pp. 81-95. 1958. The myology of the chelonian Ihnb. 2. The hind limb nmseula- ture of Lissemys punctata. Ibid., vol. V, no. 1, pp. 21-33. BREVIORA No. 130 Hasan, S. I. 1941. The shell and the mechanism of its closure in the Indian pond terrapin, Lissemys pimctata punclata (Bonaterre). Proe. Ind. Acad. Sci., vol. XIV, sec. B, no. 3, pp. 235-249. Hoffmann, C. K. 1890. Schildkroten in Reptilien. Bronn's Klassen und Ordnungen des Thier-Reichs, Bd. 6, Abt. 3, pp. 1-442. Leipzig. SlEBENROCK, F. 1906. Schildkroten von Ostafrika und Madagaskar, in Voeltzkow, Reise in Ostafrika in den Jahreu 1903-1905, Bd. 2, pp. 1-40. 1916. Schildkroten aus dera nordlichen Seengebiet und von Belgisch- Kongo. Annalen des K. K. Naturhistorischen Hofmuseums, vol. XXX, pp. 1-12. ABBREVIATIONS At. P. Attrahens pelvium At.P.' Area of origin of attrahens pelvium Car. Carapace C.Carp. The position of the posterior hinged part of carapace in Kinixys when shell is completely closed C.Dl. Clavicular part of deltoideus C.DL' Area of origin of clavicular deltoideus C.S.Isc. Cut surface of ischium C.S.Isc' Area of cut surface of ischium where it is fused with f)lastron C.S.P.Isc. Cut surface of pubo-ischial joint which is fused with plastron C.S.P.Isc' Area of pubo-ischial attachment on plastron F.C.S. Part of flexor caudae superfieialis F.C.S.' Area of insertion of F.C.S. on plastron F.V. Femoral valve H.A.P. Hinged joint of the posterior plastral lobe H.A.P.' Place of valvular movement of the anterior plastral lolie in Lissemys H.P.P. Hinged joint of the posterior plastral lobe H.P.P.' Place of valvular movement of the subcaudal valve of plastron in Lissemys J.Car. Hinged joint in carapace in Kinixys P. Pectoralis P.' Area of origin of pectoralis P.C.Ca. Precoracoid cartilage P.G. Pelvic girdle P.G.' Shifted position of pelvic girdle in Kinixys when the shell is completely closed 1960 TURTLE HINGES 9 PI. Plastron R.Carp. The position of the posterior hinged part of the carapace in Kvnixys when the shell is fully opened E.C.C. Retrahens capitis collique E.C.C.P. Rectus capitis cervico-plastralis B.C.C.P/ Area of origin of rectus capitis cervico-plastralis Rt.P. Retrahens pelvium Rt.P.' Area of origin of retrahens pelvium S.Dl. Scapular part of deltoideus Sk.C. Skin connection between the two hinged parts of tin- (ara]i;ue of Kinixys Tu. Tendons connecting cervical vertebrae with plastron Tn.' Place of insertion of the tendons on the plastron Tr. Trapezius Tr.' Area of insertion of trapezius 10 BREVIORA No. 130 Car. J, Cor, R.C.C. Fig. 1. Kinixys erosa: Side view showing extended neck and the position of the muscles, the hinged posterior part of carapace nnd the pelvic girdle when the shell is open. Car, J. Car. R.C.C. RG: At.P Rt.^^^G. Fig. 2. Kinixys erosa : Side view showing the retracted head and neck, the contracted muscles and the shifted position of the pelvic girdle when the shell is closed. I960 TURTLE HINGES 11 / tVVV <^^ -C.S.PISC. C.DI.' R.C.C P' C.S.RIsc' C.S.Isc' Fig. 3. Pelusios suhniger : Left, the superficial ventral muscles in the shell region exposed. Eight, the inner surface of the plastron showing the area of origin of various luuscles. 12 BREVIORA No. 130 H.A.R Fig. 4. Terrapene Carolina Carolina: Left, the superficial ventral muscles in the shell region exposed. Eight, the inner surface of the plastron sho-wing the area of origin of various muscles. 1960 TURTLE HINGES 13 R.C.C.R Fig. 5. Testiido herma>nni: Left, the superficial muscles in the shell region exposed. Eight, the inner surface of the plastron showing the area of origin of the various muscles. 14 BREVIORA No. 130 Fig. 6. Sternotherus carinatus minor : Left, the superficial ventral muscles in the shell region exposed. Eight, the inner surface of the plastron showing the area of origin of various niusuk-s. I960 TURTLE HINGES 15 H.A.P Fig. 7. Lisscmys punctata: Left, superficial ventral muscles in the shell region exposed. Bight, the inner surface of the plastron showing the area of origin of various muscles. BREVIORA Mmsemim of Comparative Zoology Cambridge, Mass. December 30, 1960 Number 131 A SECOND RECORD OF THE FOSSIL RODENT PALUSTRIMUS WOOD By Craig C. Black Carnegie Museum, Pittsburgh 111 July of 1959, while working in the Goshen Hole area in eastern Wj^oming, Laura McC4rew took Sabra B. Black and me to a microfaunal locality about six miles west northwest of Fort Laramie National j\Ionument, just south of the North Platte River. A small surface .sample of isolated teeth and about twenty pounds of fiiie, gray channel sand were collected at that time. The following genera have been recognized in the assemblage to date: Prosciurus, Promylagaulus, Hcliscomys, Palustrimus, and Palaeocastor. The presence of Heliscomys and especially Prosciurus would indicate an earlv Miocene age. Until recently Proscii{7-ns was not known to occur after the late Oligocene, but it has now been recorded from a new basal Miocene forma- tion in South Dakota (Macdonald, personal communication). Its presence in this assemblage together with the definitely early Miocene Promylagaulus and Palustrimus would lead me to be- lieve that the fauna is lowermost Miocene. A detailed report on this fauna is deferred in the hope that a larger sample can be secured in the near future. The purpose of the present note is to call attention to this new early Miocene locality and to record the presence of Palustrimus in the assemblage. Palustrimus was described by AVood in 1935, from a single tooth, LM \ in the Yale Peabody Museum collections from the Upper John Day. The present specimen represents the second record for the genus and corroborates "Wood's original determi- nation. This occurrence is doubly important since the type of Palustrimus lewisi, Y.P.M. No. 10572, cannot be found in the Yale collections. There is a note with the label for the specimen which states that it has been lost since 1950. BREVIORA No. 131 I take this opportunity to tliank Laura McGrew for showing us the locality, and Dr. J. T. Gregory for allowing me free access to the John Day rodent collections then in his care. The drawing is by Mr. James 0. Parley and was made possible by a grant from the Gulf Oil Corporation. Figure 1. Palustrimus sp., LM^, X30. Abbreviations used : Y.P.M. — Peabody Museum of Natural History, Yale Uni- versity M.C.Z. — Museum of Comparative Zoology. Harvard Uni- versity Order RODENTIA Family MURIDAE Palustrimus sp. Referred Specimen. M.C.Z. No. 7353, LM ^ Horizon and Locality. Lower Miocene. S. 15, T. 26 N., R. 65 W., Goshen County, Wyoming. Description. The tooth is composed of three transverse lophs joined lingually by a high ridge. The valleys between the three 1960 THE FOSSIL RODENT PALUSTRIMUS WOOD 3 lophs are extremely deep and pass directly across the tooth to the lingual ridge. Each loph is composed of three cusps, the cen- tral cusp in each case being dominant and the marginal ones lower. The lingual cusps of each loph are joined together by the lingual ridge, which carries several smaller cuspules along its length. The lophs increase in width from front to back, giving the tooth a triangular appearance. The transverse valleys are much deeper than the notches between the cusps. The notch between the buccal and central cusps is quite deep on the an- terior loph, shallower on the central loph and almost completely absent on the posterior loph. There is a small accessory cusp rising from the floor of the posterior transverse valley just in front of the shallow notch between the buccal and central cusp of the last loph. The first and second crests are slightly convex anteriorlv, while the last crest is slightlv concave anteriorlv. The measurements (in mm.) are: anteroposterior, 2.45; width anterior loph, 1.35; median loph, 1.60; posterior loph, 1.77. Discussion. There is little doubt that this specimen is con- generic with Palustrimus. That it is conspecific with P. lewisi seems highly doubtful, however, but the present material does not warrant the erection of a new species. It differs from P. lewisi in the possession of a high, lingual ridge connecting the trans- verse lophs and bearing several accessory cuspules and in having convex rather than concave anterior and median lophs. In other respects the tooth is extremely similar to that of P. lewisi. Both have the three transverse crests composed of three cusps each. Unfortunately, until more material of this peculiar genus is known its taxonomic position eannot be definitely determined. It is like nothing else known among the rodents from the North American Tertiary. Wood's {op. cit.) assignment of the genus to the Muridae would seem to be the most likely one at present, although the possibility of a geomyoid relationship (Wilson. 1949, p. 126) should not be overlooked. REFERENCES Wilson, R. W. 1949. Early Tertiary rodents of North America. Carnegie Inst. Wash- ington, Publ. no. .584: 60-164, figs. 1-13. Wood, A. E. 1935. Two new rodents from the John Day Miocene. Amer. Jour. Sci., (5) 30: 368-372, figs. 1-3. BREVIORA Miisenjnm of Cooiparative Zoology Cambridge, Mass. Dkcembeh '40, 1900 Number 132 THE STATUS OF SFHAERODACTYLUS PICTUS, WITH COMMENTS ON THE DISTRIBUTION OF S. SPUTATOR AND S. SAB ANUS By Wayne King Departiiiont of Biology and Florida State Museum, Universitv of Florida Garmaii (1887:20), in his description of Sphaerodactyliis pictus, lists its range as the island of St. Christopher's (= St. Kitts), West Indies. Barbour, in his monograph of the genus (1921:263) and in his first elieeklist of Antillean amphibians and reptiles (1930:85) lists the range as St. Kitts. In his second (1935:10-1:) and third (1937:115) Antillean checklists, Barbour includes the island of Nevis in the range of picins, and states that it (pictus) is possibly a synonym of S. sputator. In his monograph, Barbour (1921:226) separates jnctus from sputator on the degree to which the dorsal scales are keeled (rather weakly in pictus, and strongly in sputator), and on the number of dorsal scales equal to the distance from the tip of the snout to the center of the eye (9 in pictus, and 10 in sputator). Examination of the types of pictus (MCZ 6071) from St. Kitts, and of a large series of sputator (MCZ 16598—16633, 16635— 16641) from St. Eustatius (= Statia) indicates that the tirst of the two characters used by Barbour to separate these species is extremely subjective and of doubtful value. The second charac- ter, which involves allometric growth of the head and dorsal scales, varies from 7 to 10 scales in sputator. The number of dorsal scales in the "standard distance" of pictus is thus in- cluded. Further examination reveals no character of scutellation which will separate the two forms. 2 BREVIORA No. 132 (laniiau (1887:20) describes the color aiul pattern of pictus as: ''Greyish with tliree or four roAvs of brown spots on each side. On the snout there is a brown band from each eye around the end ; a median band meets these on the rostral. Behind the eyes, on the head, there are six longitudinal bands of brown, four of which join to form two on the occiput, and these meet the laterals on the neck forming two which are continued above the shoulders. A light line across the forehead from one orbit to the other. Two or three light streaks, across the ])ack of the head and neck, appear in some. On a very young one there are five narrow, transverse, dark-edged streaks of white between the eyes and the base of the tail. There are traces of brown blotches on the lower surface. ' ' A series of sputator in the Museum of Comparative Zoology indicates that the juvenile pattern of this species consists of white crossbands, with dark browai edges, on the neck, trunk, and tail. There are 5 to 8 of these crossbands between the level of the eyes and the base of the tail. The crossbands alternate with areas of dull brown ground color. The adult has a pattern of 2 to 8 white crossbands, usually with dark brown edges, on the neck and shoulders. The white markings on the trunk are always edged, at least in part, with dark brown. The pattern may be in the form of crossbands, or may be broken into wavy lines or spots. The white markings may fade to a light tan. In a few individuals the dark brown edge of the dorsal spots tends to form longitudinal rows or stripes on the light ground color. Both juveniles and adults have a dark brown stripe on the canthus i-ostralis, and a white stripe connecting the orbits across the top of the head. From the above description it is evident that the variation which occurs in the color ]iattern of sputator includes the pattern thought to be characteristic of pictus (see Fig. 1). Since their scutellation and patterns are identical, Sphaero- dactylus pictus Garman should be considered a synonym of Sphaerodactylus sputator (Sparrman), and the range of sputa- tor should be extended to include the island of St. Kitts. A confusing factor in establishing the status of S. sputator is Barbour's reference (1923 :2) to its color pattern. lie states that sputator is: "... One of the dichromatic forms, as are so many of the large-scaled species — and perhaps others as yet 19G0 SPIIAERODACTYLUS PICTUS = "C s « I o ^ S5j CO .r ■Id ''H cS -a K a *^^ cS CJ ^ Q -♦- ^ c OC CO «H o o O 1—1 CS ^ ^~, p c (5 O 1= 0} CC ■r T. ^_ o O OS > ^ s S o OQ o p (M -4J CS rH 05 -^- N "3 ■3 rt be ^ '• o o 4 BREVIORA No. 132 little known. Tlie t.y])es are females evidently. The males are much smaller than the females, uniform grey- ish brown through life, or at the most with a few fine scattered dots usually on the head. The females are large, bulky and with a great variety of broken bands, blotches and spots of varying size. ' ' Although Barbour does not list the catalogue numbers of the specimens he refers to in this paper, he does state that they were collected on Statia in 1922 by James L. Peters. This series is MCZ 16598—16633, 1 6635— 16641. At the time of his writing, this series contained both (S". sputator and S. sabanus. This is evident in Barbour's reference to the large ''females" {= sputa- tor) with blotches, bands and spots, and to the small ''males" (= sahanus) of a uniform brown. ^ Examination of the present series of sputator shows little or no sexual dichromatism, and the snout-vent length of the males (31-35 mm., with a mean of 32.8) is slightly greater than that of the females (28-35 mm., with a mean of 31.7) . On 10 July 1958, Dr. Walter Auffenberg and I collected an adult specimen of S. sputator in Basseterre, St. Kitts, thereby confirming the existence of this species at the type locality of the synonym, pictus. On 15 July 1958, we collected two adult sputator on St. Martin, 21/2 miles west, and 14 ™il6 north of Philipsburg, near Devil's Hole. The range of S. sputator, as indicated by the specimens avail- able to me at this time, includes the i-slands of Statia (MCZ 16598—16633, 16635—16641 (29); IJMMZ 57010), St. Kitts (MCZ 6071 (3); UF 10038), and St. Martin (UF 10039 (2); PWII 474A, 606 (5)). Future collecting Avill probably estab- lish its existence on Nevis and Saba, the two remaining islands of the Saba to Nevis chain of islands. The range of 8. sahanus includes all of the islands of the Saba to Nevis chain of islands. It is found on Saba (MCZ 45215 — 45217; USNM 103985—103993, 103995—104003), Statia (MCZ 54010—54015 (158) ), St. Kitts (UF 10041 (9), 10042 (9), 10043 (9), 30044 (9), 10045 (10); PWH 422; ITMMZ 83317), and Nevis (UF 10040 (3); PWII 414; UMMZ 83316 (2): MCZ 1 The Hfihaitiis were later sfjuirateil from th<' series and recatalogued under tlie luiiiio I'lrgantiiliis as MCZ .54010 - .')401.'5. That they are not elegantuhis is now- dear since in addition to other features tliere is no crossbandinff in the juveniles iif this .series. Cochran (IJKiS) pointed out tiiat the juveniles of sgbanus are \iuniarl;ed lilie the adults, while Harbour (1921) descril>ed the erfwsbanded juveniles of flegaiitiiliis. Tliis confusion of elegantuhis and sabanus is not sig- uiticant for the present pai)er ; a study of tiiese and other Lesser AntiUean sphiierodactyls, including a redetinition of all the species, is in preparation and will be presented later. 1960 SPHAERODACTYLUS PICTUS 5 38374). Barbour's record of pictus {= sputator) on Nevis (1935:104 and 1937:115) seems to be based on the Museum of Comparative Zoology specimens of sahanus listed above. The fieldwork during the summer of 1958 was supported by National Science Foundation Grant G-3896 and the Florida State Museum. Specimens collected on this trip have been placed in the University of Florida Collections, Gainesville (= UF). I would like to thank Dr. Ernest Williams (Museum of Com- parative Zoology, Harvard = MCZ), Dr. Doris Cochran (U. S. National Museum, Washington = USNM), Dr. Norman Hartweg (University of Michigan Museum of Zoology, Ann Arbor = UMMZ) and Dr. P. Wagenaar Hummelinck (Der Rijks-Uni- versiteit Zoologisch Laboratorium, Utrecht = PWH ) for the loan of specimens in their care, and Dr. Walter Auffenberg, Dr. William Riemer, Andrew Arata, and Charles Myers for their criticism and interest. LITERATURE CITED Barbour, Thomas 1921. Sphnerodactylus. Mem. Mus. Comp. Zool., vol. 47, no. .3, pp. 217-278. 1923. West Indian investigations of 1922. Occ. Papers Mus. Zool. Univ. Michigan, no. 132, pp. 1-9. 1930. A list of Antillean reptiles and amphibians. Zoologica, vol. 11, no. 4, pp. 61-116. 193."). A second list of Antillean reptiles and amphibians. Zoologica, vol. 19, no. 3, pp. 77-141. 1937. Third list of Antillean reptiles and amphibians. Bull. Mus. Comp. Zool., vol. 82, no. 2, pp. 77-166. Cochran, Doris M. 1938. Reptiles and amphibians from the Lesser Antilles collected by Dr. S. T. Danforth. Proc. Biol. Soc. Washington, vol. .51, pp. 147-156. Garman, Samuel 1887. On West Indian Geckonidae and Anguidae. Bull. Essex Inst., vol. 19, pp. 17-24. BREVIORA Museum of Comparative Zoology Cambridge, Mass. Fkbiu'ary 27, 19()1 Xu.mbek 138 OX THE GENERIC LIMITS IX THE FAMILY PJLIDAE (PROSOBPvAXClIlA: MOLLUSCAji By Edward H. Michelsox Department of Tropical Public Health, Harvard School of Public Health, Boston, Massachusetts Members of the moUuscan family Pilidae have b?eii known to science since pre-Linnaean times. Although the family has been defined, the generic limits — and particularly generic relation- ships — require further clarification. Morphologic investigations have been conducted on representatives of individual species, but only rarely have these studies been of a comparative nature. An attempt has been made in the present paper to review and collate the available information upon which generic limits may be established. Studies on the comparative morphology of the kidney and the ])enial complex are also presented. In the following discussion the family Pilidae will be con- sidered to consist of seven genera. Included in the genera Pila, Lanistes, Afropomns, and Saitha are the Old AVurld species: members of Poniacea, Marisa, and Asolenc constitute the N^ew World species. These genera have been erected primarily upon conchological characteristics based on such criteria as color, size and shape of the shell, types of sculpturing (if present), and the presence or absence of an umbilicus. 11 Attempts have been made by others to divide the family into two major groups, the Old and New World forms, on the pres- ence or absence of a calcareous operculum. Newer knowledge 1 This study was supported (In part) by a research grant (E-513-C) and a training grant (2E-4f)( from the National Institutes of Allergy and Infectious Diseases. National Institutes of Ilealtli, I'ublle Health Service. 2 BREVIORA No. 133 has shown that this criterion is unsound, since only PiJa has a calcareous operculum. However, the phylogenetic significance of the operculum even in Pila is limited since calcification is a secondary process which occurs after the snail hatches (Ranjah, 1942). The longisiphonate or brevisiphonate nature of the respiratory siphon has also been used as a criterion for the separation of Old and New World species, and appears to have some validity. Pila and Lmiistes are brevisiphonate, and the present study in- dicates that the siphon of Afropomus is similar. Saulea can only tentatively be accepted as a valid genus since no description of its anatomy exists. The New AVorld genera, Pomocea and Marisa, are longisiphonate; however, AiioJene has been reported (Scott, 1943) to have an aberrant siphon (brevisiphonate ?), and may be closely related to the Old World genera. It has been suggested that the morphology of the siphon is not of phylogenetic signifi- cance, but reflects an adajitation to ecological conditions (Prashad, 1925). The radulae of the Pilidae are all taenioglossate and have the formula 2 :1 :1 :1 :2. Intra-specific variations, however, reduce the value of radular morphology at levels below the family. The genus Turhinicola w^as erected on the basis of radular morphology (Annandale and Prashad, 1921; Prashad, 1931) ; however, Pils- I)ry and Bequaert (1927) consider this group to be no more than a subgenus of Pila. The eggs of the various species provide several promising and characteristic differences which may aid in arriving at taxonomic limits of the genera. Lipochromes, which color the eggs or egg shells, occur in species of Pomacea, but are absent in the eggs of Pila and Lanistcs (Comfort, 1947). Eggs of Marisa cornnarietis are peculiar in possessing an orange pigment when first depos- ited, but this soon disappears (Michelson, 1956). The presence of an egg-shell is biologically significant and may be of phylogenetic importance. Both Pila and Pomacea produce such eggs, and in both cases the eggs are deposited out of water. Marisa is completely aquatic and its eggs are gelatinous. Eggs of Lanistes were initially reported as membranous by D'Ailly (Pilsbry and Bequaert, 1927), but recent- observations have demonstrated that they are gelatinous and are deposited below the water-line (McMahon et al., 1957). Information concerning the eggs of Asolcne is limited to a report by von Ihering {vide Pilsbry, 1933) in which it is stated that the eggs are gelatinous. The eggs of Saulea and Afropomus have not been described. 1961 GENERIC LIMITS IN THE PILIDAE ITI The soft parts of 8 species of Pilidae (representing the genera Pilo, Lanisies, AfropoinKs, Pomorra, and Mnn'sa) were examined for characteristic anatomical differences. Only two structures, the kidney and the penial complex, appeared promising in this respect. The identity and sources of the material used in this study are presented in Table 1. Specimens of Pila, Lanistcs, Afro- pomus, and the South American Pomacra were obtained from the collections of the Museum of Comparative Zoology, Harvard University, through the courtesy of Dr. William J. Clench. These specimens were fixed either in Bouin's solution or 70% ethyl alcohol and sul)sequently stored in TO*^! alcohol. Speci- mens of Marisa and Pomacra palndosa obtained from laboratory colonies were first relaxed in boiled water and subsequently fixed in Bouin's, Zenker's, or Newcomer's solution. Since not all specimens were fixed in the same way only gross and micro- anatomical features were studied. For micro-anatomical study, tissues were embedded in paraffin, sectioned at 0-12.5;;., and stained with Lillie-Mayer hemalum and eosin. A total of 82 snails were examined including at least 5 specimens of each species. A. The Kidney. The kidney of members of the Pilidae is com- posed of two distinct regions, an anterior and a posterior cham- ber. The anterior chamber is a discrete tubular structure that partially extends into the mantle cavity opening into it through an excretory pore. The posterior chaml)er is embedded in its entirety in the visceral mass. This chamber is not compact, but consists of a large vacuolated area surrounded laterally and ventral ly by a thin, transparent membrane, and bounded dorsally by a large shield-like mass of tissue. The posterior chamber is further limited anteriorly Ijy the pericardial membrane, although access to the pericardium is provided by the renopericardial pore. Descriptions of the kidney of specimens representing the various genera follow : Pila. In P. gloho.sa (Fig. IE) the dorsal surface of the pos- terior chamber is broadly rectangular and measures approxi- mately 1.8-2.0 times the length of the anterior chamber. The dorsal surface is broAvn and blood vessels are not prominent. The anterior chamber is triangular and is so oriented that its main axis is continuous with the axis of the posterior chamber. The kidney in the species from Siam was morphologically similar. & BREVIORA No. 133 Lanistcs. Tn L. bolfancanus (Fig. 1 C) the dorsal surface of the posterior ehaniber ditt'ers from that of PUa in being liroader anteriorly. The ratio of the length of the j)osterior elianiber to the anterior chamber is approximately 0.8 :1 to 1.1. The anterior chamber is considerably longer than in Pila and is triangular in shajie. The main axis of the anterior chamber in Lanistcs is also oriented so that it is continuous with the axis of the ]iosterioi' chamber. Figure 1. Senii-diagraninuitic sketch of the dorsal surface of the kidney in five species of Pilidac: (A) Afropomus halanoideus, (B) Pomacca paludosa, (C) Lanistcs holtaneaniis, (D) Maiisa cornuarieiis, (E) Pila fjlobosa. The posterior chamber (p) and the anterior clianilier («) of each kidney illustrated arc oi'iented as in Figure lA. 1961 OENERIC LIMITS IX THK IMLIDAF, Afropomns. Tlie kiclm-N' in ^1. balanoidciis {Fig. 1 A) differs most radically from those of the other genera. The posterior ehamher is distinctly triangular in shape and its apex is reflected to the right. Furthermore, it is the only form in which blood vessels are prominent on the dorsal surface of the kidney. The anterior chamber is irregularly rectangular in shape, and its axis lies at an obtuse angle to that of the posterior chamber. Pomacea. The morphology of the kidney was similar in the four species examined. In P. paludosa (Fig. 1 B), the posterior chamber is broadly rectangular and has a prominent protuber- ance situated anteriorly on its left margin. The anterior cham- ber is irregular in shape and its long axis lies at an obtuse angle to the axis of the posterior chamber. Marisa. The kidney of M. cornuayidis (Fig. 1 D) is very similar to that found in species of Pomacea, thus further strengthening the suggestion that Marisa, should be considered a subgenus of Pomocea (Baker, 1930; Pain, 1950). The lateral protuberance of the posterior chamber in Marisa is larger and fartlior anterior than in Pouiacca. In addition, the posterior chamber surrounds the anterior chamber more completely in Marisa than in Pomacea B. The Pcnial Complex. The penial complex in the members of the Pilidae arises from the mantle as a finger-like projection. It consists primarily of a large outer penial sheath Avhich en- folds the true penis. Since there is no direct connection betAveen the i^enis and the vas deferens, sperm must be transmitted from the latter organ to the former, and thence to the female. Sach- watkin (19201 first described the presence of an internal sperm canal in the penis of Ampullaria gigas. Prashad (1925), how- ever, found that in Pila glohosa an external sperm canal was present. Our study indicates that both workers were correct and that an internal canal is characteristic of the New World species and an external canal characteristic of the Old AYorld species (Figs. 2-4). TV Although additional studies will be needed to establish the generic limits in the family Pilidae, there appears now to be sufficient information to separate the Old from the New World genera, as shown in Table 2. BREVIORA No. 133 It is apparent from the foreg'oinfi: that sufficient data are not available to permit a critical interpretation of the phylogenetic relationships within the family. Nevertheless, there appears to be an evolutionary trend towards the establishment of the family in the terrestrial biotope. The presence of a respiratory sac, in addition to gills, suggests a mori^hological adaptation for the transition from the aquatic to the terrestrial habitat. Species in two genera {Pila and Poniacca) are highly amphibious and even deposit their egg masses out of water ; the presence of a calcareous egg shell further reduces the dependence of Pila and Pomacea on an aquatic habitat. If the premise is accepted that the family Pilidae is evolving towards a terrestrial mode of life, then we must conclude that both Pila and Pomacea repre- sent evolutionary advances in that direction. Table 1 The Specific Identity, Origin, and Number of Specimens Examined No. Specimens Species Origin Examined Pila gJohnsa Swainson Lanitites bolt an (anus (Eodiiig) Afropomus halanoidcus (Gould) Marisa cornuariclis (Linne) Pomacea paludosa Say Pomacea interrupta Soweiby Pomacea columellaris Eeeve Pomacea nitilia Eeeve Cakntta, India, Bangkok, Thailand Cairo, Egypt Liberia Rio Picdias, Puerto Eico Miami, Florida Chonta anticline, Peru Huanuco, Peru Iluanuco, Peru / 5 10 8 15 12 10 10 5 Table 2A Diagnostic Characteristics of the ]\Iajor Genera of the Family Y 'ilidae i Shape of Eespiratory Sperm Genus Shell Operculum Siphon Canal Pila Dextral; sub-ovate to globose Calcareous Brevisiphonate External Lanistcs Sinistral; sub-ovate, turbinate, or earinate Corneous Brevisiphonate External Afropom us Dextral ; globose Corneous Brevisiphonate External Saulea Dextral; sub ovate Corneous ? ? Pomacea Dextral; sub-ovate, ovate, or globose Corneous Longisiphonate Internal ilarisa Dextral ; secondarily planorboid Corneous Longisiphonatc Internal Asolene Dextral; sub-ovate, Corneous Longisiphonate ? ovate, or neritoid 1 The morphological characteristics ol' the kidney of each genus are presented 1961 GENERIC LIMITS IN THE PILIDAE Table 2B Dia. Cross-section through the penial sheath and penis of a specimen of Mnrisa cornunrictis. The arrow points to the internal sperm canal, iremalnm and eosin, X29. Figure 4. Cross-section through the penial sheath and penis of a speci- men of lAinistcs holtaneanus. The arrow points to the external sperm canal. A similar type of sperm canal was found in the penes of specimens of Pila glohosa and Afropomus halanoideus. Hemalum and eosin, X43. l')61 GENERIC LIMITS IX THE PILIDAE X x. 7} •"5 > ^ ,^t 5i; ]0 BREVIORA No. 133 LITERATURE CITED AriNANDALE, N. niirl B. PRASHAn 1921. Materials for a generic revision of tlic fresh-water gastropod molluscs of the Indian Empire. No. 4. The Indian Ampulla- riidae. Reo. Indian Mus., Oalcutta, 22:7-12. Bakeir, II. B. 1930. The mollusea collected by the University of ^lichigaii-Williani son Expedition in Venezuela. Occ. Pap. Mus. Zool., Univ. Michigan, No. 210, 94 pp. Comfort, A. 1947. Lipochronies in the ova of Fila. Nature, 160:333-334. McMahon, p., T. von Brand, and M. O. Nolan 1957. Observations on tlie polysaccharides of aquatic snails. J. Cell, and Comp. Physiol., 50:219-240. MiCHELSON, E. H. 1956. Studies on the liiology of the genus Crrafodes (Mollusca: Pilidae). Doctoral Dissertation, Harvard Univ., 171 pp. (Pub- lished in part.) Pain, T. 1950. Pomacea (Ampullariidae) of British Guiana. Proc. Malac. Soc London, 28:63-74. PiLSBRY, II. A. 1933. Zoological results of the Matto Grosso expedition to Brazil in 1931, II. Mollusca. Proc. Acad. Nat. Sci., Philadelphia, 85:67- 76. PiLSBRY, H. A. and J. Bequaert 1927. The aquatic niollusks of the Belgian Congo, with a geographical and ecological account of Congo malacology. Bull. Amer. Must Nat. Hist., 53:69-602. Prashad, B. 1925. Anatomy of the couuuon Indian apple snail Pila ylobosa. Mem, Indian Mus., Calcutta, 8:91-151. 1931. Some noteworthy examples of parallel evolution in the molluscan faunas of South-eastern Asia and South America. Proc. Roy. Soc. Edinburgh, 5:42-53. Ranjah, a. R. 1942. The embryology of the Indian apple-snail, Pila globosa (Swain son) Mollusca, Gastropoda. Rec. Indian Mus., Calcutta, 44:217- 322. Sachwatkin, V. 1920. Das urogenitalsystem von Ampullaria gigas Spix. Inaugural Dissertation, Univ. Zurich, Alb. Bonniers Boktryckeri, Stock- holm, 64 pp. ScoTT, M. I. H. 1943. Sobre la organizacion de Ampullaria (Asolene) megastoma Sowerby. Notas del Museo de la Plata, 8, Zoologia, No. 70, pp. 269-280. BREVIORA laseimi of Comparative Zoology Cambridge, Mass. February 28, 1961 Number 134 ENZYMATIC CONSTITUTION OF ALSOPHIS SALIVA AND ITS BIOLOGICAL IMPLICATIONS By George Hegeman University of California Berkeley, California I. Introduction As long as fift^'-eight years ago (Alcock and Rogers, 1902) it was noted that the saliva of certain ostensibly harmless colubrids had toxic properties. Recently, members of the genus Alsophis have been suspected to have a toxic saliva. A. S. Rand (MS.) has detailed accounts of Alsophis portoricensis feeding behavior and of the effects of bites in man. Neill (1954) relates an in- stance of toxic effects in himself from an A. anguilifer bite. The purpose of this paper is two fold: 1) to attempt to pro- vide enzj^matic confirmation for the tentative conclusion that the colubrid Alsophis portoricensis has a truly active salivary secre- tion (i.e. with venomous properties) ; and 2) to characterize the activity found in terms of similar enzymes represented in another colubrid and in a crotalid, thus providing a comparative basis for evaluation of the biological implications of this activity. Natrix sipedon from Florida and Crotalus atrox from Texas provided the other preparations. Commercial crystalline Crota- lus venom was made available by Dr. W. R. Sistrom. The Natrix and Alsophis preparations were collected by a straight- forward method in the laboratory. II. Collection of Saliva A small tablet of washed, fine-pore cellulose sponge, previous- ly moistened with distilled water, was inserted in the reptile's mouth with stainless steel forceps. Alsophis readily grasped and chewed on the sponge, but with Natrix swabbing was necessary. 2 BREVIORA No. 134 When relinquished, the sponge was squeezed and rinsed into a tared weighing bottle. The bottle was placed in a chilled desic- cator and evacuated until dry over potassium hydroxide at 5°C. This procedure was repeated over the course of several weeks until a quantity deemed sufficient for subsequent tests was col- lected. Ten "milkings" coming from two individuals provided 22 milligrams of Alsophis preparation over a period of three weeks. Six Natrix, each milked once on four occasions equally spaced over the same three wrecks, gave a total of 26 milligrams of preparation. III. Estimation of protein content Since all known enzymes are proteins, the first step in gaining a fair comparison of activities depends on the total protein con- tent of the saliva. This was estimated for small redissolved duplicate samples of preparation by the Folin-Low^ry method (Lowry et al., 1951). Bovine serum albumin was adopted as the reference protein standard. Protein by percent weight for the three preparations is given below (Table 1). '.%) Alsophis is here seen to lie between the harmless colubrid and overtly venomous crotalid with respect to potentially active protein content. These results were used to adjust the dissolved preparations used in the later experiments to equivalent protein content. IV. Spreading factor estimation The reports of Neill and Rand (above) indicate the presence of a "spreading factor," seen in many natural venoms (Kella- way, 1939). In the case of snake venoms this is probably a complex mixture of proinvasins and hyaluronidase (Zeller, 1948). These enzymes respectively inhibit destruction of the hydrolytic agent of the venom and destroy the mucoid structural material of the cell lattice by a hydrolytic action. The resultant loosening of the matrix favors the passive transport of the venom components of the tissue. TABLE 1 Preparation Jo protein ( : Alsophis 43.0 Crotalus 99.5 Natrix 16.5 1961 ALSOPHIS VENOM 3 The method followed here has been modified from Fanilli and McClean (1939). It has the disadvantage that specificity is lost and the results are ill-defined. It is well suited to demonstrating low activities however. Two dilutions of the preparation in question, adjusted to equal protein content, were mixed with equal parts of india ink and 0.9 saline. Of each dilution 0.025 ml. was injected intradermally on the shaved back of an etherized rat. Bovine serum albumin of approximately equal protein content was substituted for the saliva preparations in five otherwise similar control injections. Sites of injection were placed within the shaved area so that each was straddled by two controls. The diameter of each weal was read at one-half hour intervals. Since the progress of the spots was a nearly constant percentage of the original diameter for the first hour and one half, the observations below are for that range (Table 2). Rates fell off sharply after that period; the rate for Crotalus fell off sharply after the first half hour. TABLE 2 mgm preparation % increase in diameter protein injected per hour (± 10%) Crotalus 0.125 0.068 80 55 Natrix 0.125 0.063 20 Alsophis 0.130 0.065 50 20 Controls 1 0.000 2 ( ( 14 3 ( ( 4 ( 1 5 •' 17 Again, according to this index, Alsophis shows a definite tendency toward "venomous" aetivit3\ V. Proteinase Another type of enzyme frequently active in venoms (Zeller, 1948) is proteinase. Work on the genus Dromicus, a relative of Alsophis, has demonstrated the activity of this enzj'me type in extracts of the parotid gland (Donoso-Barros and Cardenas, 1959). Proteinases act hydrolyticly to reduce the chain length of proteins and large polypeptides, proAdding necrotic pre- digestion of the food object and destruction of the tissues. The fortuitous release of histamine from damaged and dissolved 4 BREVIORA No. 134 tissue might be expected to enhance auj- neurotoxic activity in the venom too. Proteolytic activity was estimated (Northrop et al., 1948) by measuring the residual material giving the Folin-Lowry reaction in an incubation mixture of enzj^me and bovine serum albumin after precipitation with 0.1 molar trichloroacetic acid. Quanti- ties of 0.25 mgm protein of preparation and 2.5 mgm of bovine serum albumin per milliliter were incubated in M/15 phosphate buffer at pH 7.2 and 37° C. Saliva preparations alone and albumin alone were also assayed for autogenous proteolysis ; the added rates of release in these provided a subtractive correction for the incubation mixtures of preparation and protein. Proteolysis rate is best taken over the first half hour where first order kinetics appear to obtain. The relative activities are given below, the average of two experiments. TABLE 3 'reparation ^Lgm BSA hydrolyzcd/ mgm preparation protein/ hour (±2%) Crotalus 28.3 Alsophis 12.3 Natrix 0.3 \\. Hemolysis The lysis of red l)lood cells is a conspicuous part of the action of snake venoms (Kellaway, 1939). The major part of this effect is due to the action of phospholipase A on tissue phosphatides to produce lysolecithin. This substance renders erythrocytes extremely fragile (Zeller, 1948). The subsequent action of proteoh'tic enzymes results in the disruption of the cell wall (lysis). Other substances in venoms besides lipases promote lysis without first acting on an intermediate, and it is these which were measured here. Following Bernheim (1947), aliquots of washed horse erythro- cytes were incubated with measured quantities of the various preparations in buffered isotonic saline at 37° C. A blank was run to correct for autolysis in the buffered saline, and one aliquot was lysed by repeated freeze-thawing to give 100% lysis. At the end of the incubation period the cells were centrifuged down and the supernatant liquid was immediately read for absorbance at 579 miJ., an absorption peak of hemoglobin. This quantity when related to that of the completely lysed portion gave a 1961 ALSOPHIS VENOM 5 percentage lysis for the cells. Hemolysis rates are given below as percent lysis per ten micrograms preparation protein per milliliter per hour. TABLE 4 Preparation lysis/hour Crotalus 4.42% Alsophis 1.20% Nairix — 0.73% The negative value for the Natrix preparation indicates a lysis rate less than the control, i.e. a protective effect. VII. Cholinesterase The typical elapine venom, noted for neurotoxicity, is ex- tremely high in a unique type of cholinesterase (Zeller, 1947). Presumably, since it is present in high concentrations, the curare-like effect of this venom is due to this enzyme's hydro- lytic action on acetylcholine. This blocks transmission at neuro- muscular junctions, and hence paralyzes the prey. However, inhibitors of this enzyme do not protect animals as fully as might be expected from the effects of cobra venom, including the neurotoxic effects (Zeller, 1948). The situation here is not as clear as might be hoped. The assay method was essentially that of Hestrin (1949). The preparation was incubated with M/lo phosphate buffer at pH 7.2 and 30° C ; acetylcholine concentration was four micro- moles per milliliter. From the known protein content of the preparation and the measured disappearance of acetylcholine the activity of the preparation was established. Under the con- ditions described, accuracy was one tenth of a micromole. This was also the minimum amount of hj'drolysis detectable. TABLE 5 Preparation Units esterase Crotalus 0.0 Alsophis 0.2 Natrix 7.4 The unit of activit}^ is conventionally defined as micromoles acetylcholine hydrolyzed per hour per milligram preparation protein under the conditions of the assay. Cholinesterase is normally low or non-detectable in crotaline venoms. The hio-h activity of the Natrix saliva is surprising. Subsequent tests showed that the Natrix preparation was 50 per cent inhibited fi BREVIORA No. 134 by 10"^ molar liexamethonium (1,6-bis hexaiie trimethylam- moiiium) ; completelj' inhibited by 10"'* and 90 per cent inhibited by 10~*^ molar concentrations of neostigmine (prostigmine), re- spectively. This preparation did not appear to hydrolyze either benzoyl choline or acetyl /3-methyl choline. Characteristically, elapine venom cholinesterase does not act on benzoyl choline, but will hydrolyze acetyl ^-methyl choline. It was unfortunately impossible to treat the Ahophis preparation to this sort of analy- sis due to its low activity and a waning suppl}^ of preparation. VIII. Discussion The statement that a particular animal is or is not "venom- ous" is outwardly a qualitative judgment, but it conceals a quantitative .statement. The demonstrated presence of enzymes characteristic of venom in Ahophis' saliva constitutes necessary but not sufficient proof for the postulated function of the secre- tion in subduing prey. Alsophis' saliva protein content and the relatively greater activity of that protein are superior to those found in the "non- venomous" Notrix. Two large teeth surrounded by a fleshy ridge in the rear of the upper maxilla appear to be modified for administration of the saliva. Although these rear teeth are not grooved, the total picture for Alsophis indicates a degree of specialization trending in the direction of the adaptations of the truly venomous species. However, these morphological features, like the relative potency of the salivary secretion, are only an indication of the possible function. Critical evidence with re- gard to the use and thus the significance of the "venom" will probably come from the careful study of feeding behavior. Animals killed by .snake venom decompose at a much more rapid rate than similar individuals who have died peacefully (Zeller, 1948). Vipers, kept from injecting venom by fang removal or ligature of the duct, take from two to three times the normal time to digest a food object of constant size. Apparently the digestive function of venom is a large part of its adaptive significance, especially in an animal unable to mechanically masticate its prey. It may be that at the level of development seen in Alsophis, the effect of the saliva in subduing prey is secondary and gratuitous to the digestive function. At higher levels of specialization it is almost impossible to disentangle the two roles. The enzymes of snake venom are most easily inter- preted as a digestive complex (Zeller, 1948). 1961 ALSOPIIIS VENOM 7 The cliolinesterase of Natrix saliva, unlike the usual elapid type (Zeller, 1947), will not hydrolyze acetyl y8-methyl choline. This is especially interesting in light of the problem of the rela- tionships between the cokibrids and elapids (Johnson, 1956). An enzyme study of the analogous secretions of these two groups in a larger range of genera could prove valuable. It would, of course, first be necessary to establish the stability of enzvme type as a character wathin a genus or genera for which the rela- tionships are relatively clear. Acknowledgments I would like to acknowledge the guidance and suggestions of Dr. William R. Sistrom, who cheerfully permitted a small plague of snakes in an otherwise orderly bacteriology laboratory. I am also indebted to Mr. A. S. Rand and Dr. E. E. AYilli'ams for suggesting the problem, supplying valuable advice and informa- tion, and seeing the paper through several revisions. Dr. Karl Kofford generously supplied the Alsophis portoricensis used. PAPEES CITED Alcock, a., and L. Eogers 1902. On the toxic properties of the saliva of certain "non-poisonous" colubrines. Proc. Roy. Soc. London, 70:446. Bernheim, a. W. 1947. Comparative kinetics of hemolysis induced by bacterial and other hemolysins. Jour. Gen. Physiol., 30:337. Doxoso-Barros, R. and S. Cardenas 1959. Estudio del veneno de Dromieiis chamissonis (Wiegmann). Inv. Zool. Chilenos, 5:93-95. Fanilli, G., and D. McClean 1939. A spreading factor in certain snake venoms and its relation to their mode of action. Jour. Exptl. Med., 69:81. Hestrin, S. 1949. The reaction of acetylcholine and other carboxylic acid deriva- tives with hydroxylamine, and its analytic application. Jour. Biol. Chem., 180:249. Johnson, R. G. 1956. The origin and evolution of venomous snakes. Evolution, 10: 56. Kellaway, C. H. 1939. Animal poisons. Ann. Rev. Biochem., 8:541. LoWRY, O. H., N. J. Rosenbrough, A. C. Farr and R. J. Randall 1951. Protein measurement with the Folin phenol reagent. Jour. Biol. Chem., 193:265. 8 BREVIORA No. 134 Neill, W. T. 1954. Evidence of venom in snakes of the genera Alsophis and Rhadi- naea. Copeia, 19-i4, no 1:59. Northrop, J. H., M. Kunitz and E. M. Herriot 1948. Crystalline Enzymes. Columbia Univ. Press, New York. Phisalix, M. 1922. Animaux Venimeux et Venins (Vol. 2). Masson et Cie, Paris. ZELr.ER, E. A. 1947. Uber das Vorkommen und die Natur der Cholinesterase der Schlangengifte. Experimentia, 3:375. 1948. Enzymes of snake venoms and their biological significance. Advances in Enzymology, Volume 8, p. 459. F. F. Nord ed., Interscienee Pub., New York. BREVIORA Miaseiuiini of Comparative Zoology Cambridge, Mass. April 7, 1961 No. 135 NOTES ON IIISPANIOLAN HERPETOLOGY 2. A REVIEW OF THE ANOLIS SEMILINEATUS GROUP WITH THE^DESCRIPTION OF ANOLIS COCHRANAE, NEW SPECIES By Ernest E. Williams and A. Stanley Rand Introduction : Recent investigations in Haiti and the Domin- ican Repnblic by expeditions from the Mnseum o± Comparative Zoology have added considerably to our knowledge of the dis- tribution of the so-called grass-anoles of the semilineatus group and have resulted in the discovery of a third member of the series in the Cordillera Central of the Dominican Republic. In addition to describing the new species just discovered we here attempt a summary of the information now available on this group. Acknowledgments: The new species was collected in the summer of 1958 by Clayton E. Ray and A. Staule}- Rand during an expedition partly supported by a grant from the Societj* of Sigma Xi and enjoying the cooperation of the University of Santo Domingo and of the government of the Dominican Re- public. The essential and generous aid of Dr. Eugenio de Jesus Marcano, who accompanied the expedition, and the use of a car and driver furnished by the University of Santo Domingo are very gratefully acknowledged. In August 1959, E. E. Williams and A. S. Rand collected in the vicinity of Port-au- Prince with the support of National Science Foundation Grant NSF G-5634. In 1960 A. S. Rand and J. Lazell collected aroiind Port-au-Prince in northern Haiti and near Aux Cayes on the southwest peninsula, aided by a grant from the American Philo- sophical Society. In both Haitian trips the letters provided by M. Gerard Philippeaux, Minister of Agriculture, and the willing assistance of M. Leonce Bonnefil fils. zoologist in tln' Department 2 BREVIORA No. 135 of Agriculture and Natural Resources, were indispensable ele- ments in the success of the venture. In addition to the specimens collected by these expeditions and those already present in the Museum of Comparative Zoology (MCZ), material was obtained on loan from the United States National Museum (USNM) and the American Museum of Natu- ral History (AMNH). The assistance of the curators of these collections is gratefully acknowledged. Dr. P. S. Humphrey made available for study material collected by him for Yale Peabody Museum (Yale) and the University of Florida (UF). The previously known species : In order to provide a frame of reference for the new species, we first summarize the knowl- edge now at hand for the previously known species : Two species of "grass anoles," A. semilineatus Cope and A. olssoni Schmidt, have long been known in Hispaniola. Both are small {ca. 40 mm snout-vent length), slender-bodied forms with elongate heads, a dorsal zone of about 10 rows of enlarged keeled scales as large as the strongl}^ keeled belly scales, tail only slightly compressed and with no clear demarcation of the breaking zones or verticils. They are thus a morphologically strongly marked group within the Hispaniolan anoles. In habits they are also distinctive, being associated characteristically with grass and low bushes. Morphology: (See also table, below). Structurally the two species differ, very little — they differ in size of scales and slightly in body size, tail length and shape of head but in none of these regards so strikingly that instant identification can be confidently made even by the experienced worker. Body colora- tion differs also but not without some puzzling cases. The dew- lap in males, both as to color and squamation, is the diagnostic difference easiest to employ. The dewlap skin in A. semilineatus is white and the gular scales about the same size as the ventrals. In olssoni the dewlap skin is red or orange (darkly pigmented in alcohol) and the gular scales up to three times as large as the ventrals. In life the iris of semilineatus is steel blue, that of olssoni dark brown. This difference is not determinable in alcoholics. Ecology: There exists a real ecological difference between the two species but again there is overlap and the species do occur side by side. Mertens (1939) describes A. semilineatus as "eurytop" occur- ring in both dry and wet areas. He found it in mangroves (Puerto Plata, Sabana de la Mar), corn fields (Moca), dry open brush (Barahona, San Pedro de Macoris, Ciudad Trujillo), meadows 1961 ANOLIS COCHRANAE 3 iu the lower drier pine woods (Jarabacoa, Monciou), and in damp lush vegetation (Samana, the top of the pass between Santiago and Puerto Plata). He, in fact, states that it is absent only in the cactus- steppe. Rand (1958, field observations) saw it in the Dominican Republic primarily in open situations and along- roadsides and in pastures but found also a single specimen sitting on a rock in a muddy trail through heavy forest (Bejucal), and a dense population living in low vegetation in an area of rather dense bamboo along a stream bank (nr. Sabana de la Mar). Hassler (field notes for November 4, 1929) reports finding this species at Laguna near Samana on "leaves in damp woods and in fields nearby." Near Port-au-Prince, scmilineatus occurs in the hills to the south of towTi and up to the vicinity of Furcy at 4000-5000 ft. It is absent from Port-au-Prince itself and from the Cul de Sac Plain. In one place it has been observed to overlap with olssoni (see below). Mertens has described A. olssoni in contrast to A. semilineatus as "stenotop," confined to open dry areas. It seems, indeed, to be more limited than semilineatus, but in 1959 Williams and Rand found it in the moderately dense vegetation of a Port-au- Prince garden and also in the irrigated areas of the experi- mental farm at Damien, Haiti. It is present in open thorn scrub of the Cul-de-Sac Plain both in Haiti and in the Domini- can Republic. It is known to occur with semilineatus at several localities. Mertens records both forms from the vicinity of Ciudad Trujillo, and at Moncion, Sabana de la Mar and Bara- hona, all localities in the Dominican Republic. Cochran reports both species at San Michel du Nord, Haiti. In 1960 Rand and Lazell obtained both species at Gros Morne in northern Haiti. They found "A. olssoni on grass and low vegetation along sunny roadsides, A. semilineatus on vegetation at the edge of forest." In August 1959 Williams and Rand studied a contact area between semilineatus and olssoni on Bontillier Road which climbs the foothills south of Port-au-Prince. A. olssoni occurred only on the lower reaches of the road, A. semilineatus only on the portion of the road which parallels the crest of the hill. No striking vegetation or habitat difference was evident between the two portions of the road, both of which traverse very dis- turbed, cut over, areas. The distance between the places at which SI inilineatus and olssoni were found closest to one another was a matter of a few A^ertical vards. Rand and Lazell returning 4 BREVIORA No. 135 to the same area in 1960 found the same general pattern but in one instance a semilineatus was taken about 20 feet from an olssoni and at the same level. Habits: The two species differ very little in habits. Both are commonly seen on grass stems and the slender twigs of small bushes and plants. Sometimes they occur on fence posts and stands of barbed wire and occasionally on the ground. None have been seen on the trunks or in the crown of even small trees. Numbers of A. olssoni were observed by Williams and Rand at Damien, Haiti. Individuals of all sizes were found in the tall grass, while on the fence posts most of the animals were large males, and in the short 4-8 inch grass most were juveniles. They are frequently found facing head downward on a vertical perch with the neck bent so that the head is almost horizontal. Both sleep with the hind legs fully extended. Both escape by jumping off their perch into grass cover nearby. In one area near Port- au-Prince a few individuals were chased out of the grass into a rock pile. Here they did not go deep into the rock pile in contrast to juvenile A. cybotes but hid close to the surface and could usually be chased out by poking into the holes with a short stick. A. semilineatus is perhaps shier than A. olssoni and, according to Mertens, it is less pugnacious. Mertens reports that freshly caught olssoni carried out biting battles with each other, raising the nuchal crest and displaying their dewlaps. Very little is known of the reproductive habits of these forms. There is one observation by Rand (field notes, 1958) on A. semilineatus in the Dominican Republic : "In one locality (Rancho La Guardia, San Rafael Province) in a coffee plantation about twenty yards from o])en pasture, a number of eggs of this species were found. These were discovered in three of 22 rotten logs examined in a small area. One log contained 12 eggs, another 7 eggs, and the third 4 eggs. These three logs differed from the others examined in that they contained nests of a large black stinging ant. The eggs were mostly in the loose soil just under the logs but some were in the galleries of the ant nest. When I picked up the first egg an ant stung me and my iuA^oluntary jump sent the egg flying several feet. The other logs in the area contained a variety of invertebrate life but the ones with the lizard eggs had only the black ants and a single centipede nest. The ants provided an effective protection for the lizard eggs and it seems possible that the lizards had sought out these nests in Avhich to lay their eggs. 1961 ANOLIS COCHEANAE 5 The eggs were ovoid with a white flexible skiu. They ranged in size from 12 by 9 nun to 7 by G mm. Collected on August 13 fourteen of the twenty-two eggs hatched, the first on August If) and the last on September 17.'* Distribution : Both species are very widely dispersed in Ilispaniola. Because of their ecological differences the two dis- tributions coincide only in limited and scattered areas, though in broad terms they overlap widely. No olssoni are at present known from the Saniana peninsula and the adjacent areas in the north of the Dondnican Republic north of La Vega and east of Puerto Plata, or from ths southwestern peninsula of Haiti, but none of these areas is so well collected that the absence of olssoni from collections can be taken as a demonstration of a real absence in the field. Except perhaps in the southwestern peninsula, the distribution can be interpreted better in terms of present ecology than in terms of any other factor. We list all the verified localities^ for the two species below": Semilineatus. HAITI: Dcpt. du Nord, Cap Haitien (MCZ, USNM), Citadelle (MCZ), Dondon (MCZ); Dept. de I'Arti- bonite, Gros Morne (MCZ), San Michel (USNM); Dept. de Oucst, Basin Bleu nr Furcy (MCZ), Bontillier Road nr Port- au-Prince (MCZ), 5 km south of Dufort, south of Leogane (MCZ), Furcy (MCZ, AMNII), Obleon nr Furcy (MCZ); De- partment du Sud, Miragoane (MCZ), Les Platons north of Aux Caves (MCZ), Place Negre near Jeremie (MCZ). DOMINICAN REPUBLIC: Prov. Sa7i Rafael, Rancho La Guardia (MCZ); Prov. Barahona, Barahona (AMNH, Senckenberg), Palo (AMNH) ; Prov. Benefactor, 7 km north of Carpintero (MCZ) ; Prov. San- tiago Rodriguez, Moncion (Senckenberg) ; Prov. Santiago, top of pass between Santiago and Puerto Plata (Senckenberg) ; Prov. Puerto Plata, 8 km north of Pena (MCZ), Balneario Colon, Puerto Plata (Senckenberg), Rio Munoz, 7 km from Puerto Plata (Senckenberg) ; Prov. La Vegn, Jarabacoa (Senckenberg); Proi'. EspaiUat, Moca (Senckenberg), Rio San Juan (USNM); Prov. Duarte, Las Bracitas (AMNH); Prov. Trujillo, nr San Cristobal (MCZ) ; Prov. San Pedro de Macoris, San Pedro de Macoris (Senckenberg) ; Prov. Seibo. Boca del Inferno (USNM) ; San Francisco, 6 km east of Hato Mayor (MCZ) : Rio Yabon 1 The record of sonUincatus for the island of Narassa is doubtful. A parutype of A. olssoni was recorded by Schmidt (3 919) as probably from this island. In 1921 Schmidt redetermined the si>ecimen as A. semilineatus and cited the species without qualification as a member of the Navassa fauna. No additional speci- mens of A. semilineatus have been collected, and the record is unconfirmed. 6 BREVIORA No. 135 (MCZ) ; Prov. Samana, Laguna (AMNH), Samana (MCZ, Senckeiiberg), Sanchez (MCZ) ; Distrito de Smito Domingo, Ciudad Trujillo (Seuekenberg). Olssoni. HAITI: Dcpt. da Nord Guest, Bombardopolis (MCZ), Jean Rabel (MCZ, AMNH), Mole St. Nicolas (MCZ); Dept. du Nord, Cap Haitien (USNM) ; Dept. de I'Artibonite, bridge over the Artibonite (MCZ), south end of Etang Bois Neuf (MCZ), Gros Morne (MCZ), St. Marc (USNM, AMNH); Departmeiit de Quest, Boutillier Road nr Port-au-Prince (MCZ), Carrefour (AMNH, Yale, UF), Cabrite Id (AMNH), Damien (MCZ), Delmas (MCZ), Diquiui (MCZ, USNM), Eau Gaillee (Yale, UF), Etang Saumatre (MCZ), Fond Parisien (AMNH), Hatte Latham (MCZ, USNM), Manneville (MCZ), Mon Repos (USNM), Morne Decayette (MCZ), Petionville (Yale, UF), Port-au-Prince (MCZ, USNM, AMNH, Yale, UF), Ste. Philo- mene (USNM), Thomazeau (MCZ), between Thomazeau and Manneville (MCZ), Trou Caiman (USNM), Trou Forban (MCZ), Gonave Id, Anse a Galets (MCZ), Encafe (MCZ, USNM, Yale), Pointe-a-Raquettes (Yale, UF). DOMINICAN REPUBLIC: Prov. Monte Cristi, Monte Cristi (AMNH, Senckenberg) ; Prov. San Rafael, Banica (MCZ); Prov. Independencia, has Baitoas (AMNH), Duverge (AMNH); Prov. Barahona, Barahona (AMNH), Senckenberg), Cabral (MCZ) ; Prov. Santiago Rodri- guez, Monciou (Senckenberg) ; Prov. El Seiho, Sabana de la Mar (Senckenberg); Distrito de Santo Domingo-. Ciudad Tru- jillo (Senckenberg). A THIRD SPECIES DISCOVERED: On September 7 to 8, 1958, col- lecting in the vicinity of Constanza in the high interior of the Dominican Republic, C. E. Ray and A. S. Rand of Harvard University and Sr. Eugenio de Jesus Marcano of the Uni- versidad de Santo Domingo obtained 20 specimens of a new species of "grass anole." The greater number of these speci- mens were collected at night, sleeping on grass stems. A renewed effort to collect the same form the next morning obtained very few individuals, the lizards being then very wary and difficult to see or catch. Examination of these specimens reveals that they differ from the two previously known species in just the ways cited by Doris Cochran (1941, pp. 139-140) for a single specimen from Constanza which she then referred hesitantly to A. olssoni. Her remarks are quoted in full : 1961 ANOLIS COCHRANAE 7 "With some doubt I place with Anolis olssoni a single adult male (USNM No. 62103) collected by Dr. W. L. Abbott in the hills 5 miles south of Constauza. This individual has much smaller scales ou the gular fan than does typical olssoni from San Michel, Haiti. It does not approach, however, semilineatus in fineness of scales. In fact, while the gular scales are finer, the dorsal and ventral scales of the Constanza lizard are actually perceptibly coarser than they are in the San Michel specimens. The color pattern of this Constanza specimen shows none of the definite black markings that so often appear on true olssoni. It is lilac gray above, tinged with china-blue on the supraocular region, the dorsal tone shades into drab above the lateral light stripe, which is very sharply marked anteriorly but less so after it passes the shoulder, back of which it fades out almost com- pletely. The only definite head marking is a black diagonal bar across the temporal region which does not occur in olssoni but is found in every specimen of semilineatus. A series of examples from Constanza will be needed to determine whether these characters are stable and definite, meriting specific separation or whether they represent an aberrant or hybridized olssoni with some of the seniilineatus characters." Dr. Cochran has excellently characterized the new species, which may therefore be appropriately named : Anolis cochranae new species Type. MCZ 57660, an adult $ collected at Constanza, Dom- inican Republic, September 7-8, 1958 by C. E. Ray, A. S. Rand, E. de Jesus Marcano. Paratijpes. MCZ 57661-79, same data as above : USNM 62103, hills 5 miles south of Constanza, collected by Dr. AY. L. Abbott, April 29, 1919. Diagnosis. An Anolis allied to semilineatus and olssoni, dif- fering from the first in the much larger dorsal, ventral and supratemporal scales, from the latter in having a white rather than a red or orange dewlap in the somewhat larger dorsals and in having the gular scales little if at all larger than the ventrals, differing from both in having the ventrals nearly as large as the enlarged dorsals. Description. Head: All head scales multicarinate rather than smooth or singly keeled. Five to eight scales across head between second and third canthals (usually six to seven). Frontal de- pression very shallow, the scales in it nearly or as large as the posterior frontal or anterior supraorbital. 8 BREVIORA No. 135 Supraorbital semicircles in contact (five specimens) or sep- arated by one scale row (fifteen specimens), wholly or partly separated by one scale row from the supraocular discs. Supra- ocular disc consisting' of two to five large keeled scales separated from the elongate supraciliaries by at least two rows of scales. Canthus distinct, canthal scales four (five in one specimen), the second largest diminishing gradually forward. Naris anterior to canthal row\ The anterior nasal scale in contact with rostral. Loreal rows four to five (three on both sides in one specimen). Temporal scales subgranular. Supratemporal scales larger, keeled, grading into the large keeled scales surrounding the inter- parietal. Interparietal larger than ear, separated from the supra- orbital semicircles by one to three scales (usually two). Posterior frontal as large as anterior supraorbital. One scale nearly as large as the posterior frontal between the latter and the canthals. Suboculars in contact with supralabials. One scale intervening between subocular series and canthals. Five to six (six on one side, seven on the other in one specimen) supralabials to the center of the eye. Mentals wider than long, one to two scales inserted between the tips posteriorly. One sublabial on each side in contact with the infralabials. Central throat scales keeled, elongate. Gular fan in males. Trunk : About ten longitudinal rows of much enlarged keeled middorsal scales, broader than long, as large as the ventrals (10 to 12 ill standard distance), grading laterally into the smaller flank scales wliicli in some specimens are keeled imbricate, in others nearly granular. Ventrals in longitudinal rows, keeled, imbricate, mucronate. Postanal plates present in males. Scales of gular fan moderate, not extremely elongate, hardly larger than ventrals, not clearly arranged in lines. Limhs and digits: Hand and foot scales raulticarinate, about 17-19 lamellae under phalanges 2 and 3 of fourth toe, about 26-31 under whole toe. Largest arm and leg scales unicarinate. about as large as ventrals. Tail: Tail subcircular in section, very long, more than 21/2 times snout-vent length ; verticils not apparent. Sisc : Largest 3 41 nnn in snout-vent length ; largest 9 38 mm snout-vent length. 1961 ANOLIS COCHRAN AE Color: Essentially as in semilineatus. The more significant characters of cochranae may be compared with those of semilineatus and olssovi in tabular form: semilineatus skill of tjular fan white flank stripe short ;;ular scales ea. = ventrals 14-17 enlarged dorsal scales in distance snout to middle of eye (standard distance) median rows of enlarged dorsal scales about as broad as long 17-21 ventrals in standard distance olssoni skin of gular fan orange to red dank stripe long gular seales>> ventrals 11-13 enlarged dorsal scales in standard distance median rows of enlarged dorsal scales mostly longer than broad 11-14 ventrals in standard distance cochranae skin of gular fan while flank stripe short gular scales ca. = ventrals 10-12 enlarged dorsal scales in standard distance median jows of enlarged dorsal scales about as broad as long 11-14 ventrals in slandard distance Other differences have been listed by Mertens or Cochran, but they are at best modal differences or they alter markedly with age. These species are indeed close, and females and juveniles are sometimes difficult to distinguish. Discussion: Anolis cochranae combines in new ways charac- ters of ^1. semilineatus and .1. olssoni. It is in no sense an intermediate; its characters are either those of one or the other or are somewhat exaggerated versions of a trend present in one. It is necessary to admit that we know very little about this species beyond its existence. Its distribution would appear, on present evidence, to be extraordinarily limited. It may well be confined to the high interior, but its real range is surely more extensive than known at present. The area from which it comes is remarkable for certain peculiar forms : Celestus darlingtoni and Audantia shrevei in the higher elevations, Anolis aliniger (described as a subspecies of A. chlorocyanus by Mertens in 1939 but in reality a full species) in the vicinity of Constanza itself, Anolis clarlimjtoni both from Constanza and from higher elevations. The region merits extensive and systematic collect- ing. The biological relation of A. cochranae to the other two mem- bers of the semilineatus group is equally unknown. We do not 10 BREVIORA No. 135 know its contacts with either form. Its relationship to A. semilineatus in particular is puzzling. In squamation it differs strongly enough that we have called it, as a matter of judgment, a distinct species. The scale differences from both semilineatus and olssoni are as great or greater than the differences between other closely related sympatric fully valid species. But in other Anolis such, (or lesser) diff'erences are correlated with color and dewlap differences that are evident visual cues to species recog- nition. In color and dewlap A. cochranae exactly resembles one of the neighboring species — semilineatus. If A. cochranae is indeed a full species that is at some point in contact with semi- lineatus, it is necessary to suppose that there is some unlcnown behavioral difference that maintains the distinctness of the population in the absence of color cues. EEFERENCES CITED Cochran, D. M. 1941. The herpetology of Hispaniola. Bull. X. S. Nat. Mus., 177; 1-398. Mertens, R. 1939. Herpetologische Ergebiiissc einer Reise naeh der Insel His- paniola, Westindien. Abh. Senckenberg. Naturf. Ges., 449: 1-84. Schmidt, K. P. 1919. Descriptions of new amphibians and reptiles from Santo Domingo and Navassa. Bull. Amer. Mus. Nat. Hist., 41 : 519-52.3. 1921. The herpetology of Navassa Island. Bull. Amer. Mus. Nat. Hist., 44: 555-559. 1961 ANOLTS COCHRANAE 11 l-H O •a as A .a Pi O Si OB IS O o & o M O o El -S Q BREVIORA Musemnri of Comparative Zoology Cambridge, Mass. April 8, 1961 Number 136 NOTES ON HISPANIOLAN HEKPETOLOGY 3. THE EVOLUTION AND RELATIONSHIPS OF THE ANOLIS SEMILINEATU8 GROUP By Ernest E. Williams The discovery of a third species of the Anolis semilmeatus group, confined apparently to the high interior of the Dominican Republic, poses problems in the distribution, biology and evolu- tion of the group. The distributional data for the semilineatus group has been given in Williams and Rand (1961) and need not be repeated in detail here. A. semilineatus and A. olssoni are both widely distributed north of the Cul cle Sac Plain but occupying eco- logically somewhat different situations and thus with but limited actual contact or overlap ; only A. semilineatus at present is known south of the Cul de Sac Plain in the southwest and Barahona peninsulas. A. cochranae is found in the center of Hispaniola in the Cordillera Central — geographically in the midst of the other two species though its contacts with these others are not known. The biological peculiarity in the relation of A. cochranae to A. semilineatus has also been pointed out in Williams and Rand (1961). Thus, though differing strongly from the closely related A. semilineatus in certain scale characters, A. cochranae is iden- tical in body and dewlap color. This phenomenon is highly unusual in the genus Anolis in which body and dewlap color dif- ferences are important cues in species recognition. (There are, for example, strong body and dewlap color differences between A. semilineatus and A. olssoni.) A. cochranae, if it is in contact with A. semilineaius, as A. semilineatus and A. olssoni are in contact with one another, would seem to be a most anomalous case in which it would be necessary to provide some ad hoc explanation — such as some unknown behavior difference — for the maintenance of the species distinction. 2 BREVIORA No. 136 The problem is thus to provide an explanation of the central geographic position of Anolis cochranae in Hispaniola and of the curious absence in cochranae of the usual anoline species recog- nition characters contra a related species that occurs literally on every side of it. I propose below a suggested history of the semilineatus group that appears to solve this problem. It must be admitted that this proposed history depends upon taking at face value the distributions of the three species as they are known at present. This is patently unsafe, but it provides a useful starting point. On our present knowledge of distribution it is simplest to suppose that the postulated biological problem has not arisen, that cochranae and semilineatus are nowhere in contact. This is at the moment only a brave hypothesis. Anolis cocliranae is known from only two collections; our more extensive knowledge of the distributions of semilineatus and olssoni is by no means good enough to prove contact or absence of contact with coch- ranae. Critical to the proposed history is the supposition — uncon- tradicted by the available evidence — that olssoni is really absent from the southwest and Barahona peninsulas. It does appear to be absent from the moist coastal zone at Aux Cayes (observations by A. S. Rand and J. Lazell in 1960) and Rand did not collect it in the dry area of Oviedo on the Barahona peninsula in 1959. It is not present in Hassler's collections from these two areas. Let us then take the present distributional evidence at face value. Let us assume then that semilineatus is the only grass anole of the southwest and Barahona peninsulas and that olssoni just touches this area at the southern edge of the Cul de Sac Plain. The soutliAvest and Barahona peninsulas taken together are just that portion of the island which was cut off from the mass of Hispaniola by the Pleistocene seaway through what is now the Cul de Sac Plain. Residual salt lakes and coral rocks still testify to this former seaway. The division of Hispaniola into two parts which resulted from this seaway provides two suitable theatres — a main island and a southern counterpart — for the classic pattern of speciation during separation, and intensification of species difference ("character displacement") during renewed contact. On this hypothesis semilinedius is the "autochthonous grass anole of the southern cut-off "portion -of " Hispaniola Mid' alss'om 1961 ANOLIS SEMILINEATUS GROUP 3 and cochranac autoehtlions of the northern main mass of the island. Scmilhieatus has infiltrated the northei-n island all but completely, while olssoni is not known to have invaded the south- ern island. The spread of sonilhicatus through much of the northern island is not too surprising in view of its eurytopic ecology (Mertens, 1939, AVilliams and liand, 1961). Though character- istic of a specialized open habitat, it seems to be sufficiently tolerant of forests that these would be less efficient barriers to its spread than they would to stenotopic olssoni. It is somewhat more surprising" — if it is true — that olssoni has not spread along the dry north coast of the southwest peninsula or the east coast of the Barahona peninsula, but it would be stopped easily by discontinuities in suitable habitat and would for this reason be unlikely to reach localities otherwise quite suitable to it on the southern island. The different coloration in olssoni, including the dewlap color, and the large size of the dewlap scales may Avell have developed after olssoni came into secondary contact with semilineatus dur- ing the latter 's invasion of the northern island fragment. In suggesting this we assume that the features in common of cochranac and semilineatus are primitive and that modification in these features took place exclusively or almost so in olssoni. (Surel}' the lack of enlargement in the gular scales is primitive ill semilineaius and cochranae; this leaves only color in ques- tion.) What, however, about the origin and relationship of cochranae and olssoni? It must first be noticed that there is some plausi- bility in considering these two more closely related to each other than to semilineatus. In body squamation (i.e. scale size), coch- ranac and olssoni are very similar. This is a feature which, unlike the characters of the dewlap or of body pattern, is un- likely to be a matter of intra- or inter-species recognition. AVe do not know that it is per se adaptive: the difference in scale .size between semilineatus, on the one side, and cochranue-olssoni, on the other, is more likely to be the external expression of more fundamental genetic divergencies. No. physiographic barrier, however, will account for the divi- sion of the grass anole population of the northern or main Ilispaniolan island into two species. It is necessary to suppose that the barrier was an area of unsuitable ecology, i.e. moist dense forest. Olssoni may then be supposed to have arisen in 4 BREVIORA No. 136 the arid coastal lowlands while cochranae arose in the open areas of the high pine woods' of the interior valleys of the Cordillera Central. (We note that Wetmore and Swales, 1931, p. 24, describe the natural vegetation of the Valle Constanza as "forests of open pine mingled with areas of dense rain forest.") The knowni habitat of cochranae — Valle Constanza — is a high interior valley of the Cordillera Central. Though the floor of this valley is not very high {ca. 3000 feet) it is surrounded by some of the highest peaks in Hispaniola and ingress to it at moderate elevations is someAvhat narrow and limited. In such an area a grass anole population might indeed enjoy a measure of isolation from other grass-bush populations — the more so if Vv-e suppose that the separation of olssoni and cochranae dates from a period in which the density of the hardwood forest of intermediate elevations was at a maximum. Relationshipy of the semilineatus group. There are no other anoles in Hispaniola which either very much resemble or seem very closely related to the semilineatus group. A search for close relatives and ancestors takes us at once outside Hispaniola. Two Greater Antillean groups of Anolis are structurally simi- lar — tlie alutaceus-clivicolus-cyanopleurus-spectrwni group in Cuba and the krngi-pidchellus-ponccnsis series in Puerto Rico. (None of the anoles of Jamaica or the Bahamas are similar either ecologically or structurally.) Both the Cuban and the Puerto Rican series share with the semilineatus group the middorsal zone of enlarged scales (least developed in krugi of Puerto Rico). All except alutaccus-clivi- colus have keeled ventrals. The Cuban anoles are all forest species, A. alutaceus occurring in rather deep shade, A. spectrum in less deep shade. But, though in this regard they differ from the Hispaniolan species which are fonder of open areas, they are closer to the semilinea- tus group in structure than are the Puerto Rican species. Like the semilineatus group they are small, usually under 40 mm snout-vent length, slender, with large dewlaps and well developed postanal scales in the males. In color, however, they differ in never possessing the flank stripe so characteristic of the semi- lincaius group, tending instead to emphasize the light middorsal stripe. 1 I'ine in Hispaniola. in contrast to e.g. Cuba, is confined to higher elevations. 1961 ANOLIS SEMILINEATl'S GROUP 5 Of the alutaceus series, clivicolus, which may be a subspecies of alutaceus, has the least slender habitus and the least specialized squamation. It is easy to envision this as representing the primi- tive stock of this series. The Puerto Rican series is, on the other hand, more similar to the Ilispaniolan species in habits. Two of the three species — pulchellus and poncensis — are "grass anoles" or at least anoles of open reaches. The third species — krugi — is an anole of denser brush. All are larger than any species of the senii- lincatus group — nearer 50 mm than 40 mm snout- vent length. They are perhaps not as slender as their parallels in Hispaniola (though this is a character difficult to estimate objectively) ; the dewlaps are relatively small and the postanal scales poorlj^ developed. All three have a flank stripe passing forward through the eye more or less well expressed. In both Cuba and Puerto Rico the series exhibit a wider range of structure than do the Ilispaniolan forms. In each series there is a species with the middorsal zone of enlarged keeled scales less developed than in any Ilispaniolan species (in Cuba — clivicolus-alutaceus, in Puerto Rico — krugi) and one Avith this zone much more strongly developed than in any Hispaniolan species (in Cuba — spcctrutn, in Puerto Rico — poncensis). One difference appears in this regard : in all the Cuban forms the width of the zone of enlarged dorsal scales is about the same {ca. 8 scale rows as compared with ca. 10 in Hispaniolan forms), while in the Puerto Rican forms concurrently with increase in the size of the middorsal scale zone, there is an increase in the number of rows enlarged {ca. 4 in krugi, ca. 12 in pulchellus, 1.5+ in poncensis). The evaluation of these resemblances, which are in each case beset with significant differences, is difficult. Parallelism is very probable, and it is especially likely that the Puerto Rican series is an independent radiation witliin Puerto Rico from the same stock that gave rise to A. crisfatellus, A. stratidus, A. guncllachi and A. evermanni. The primitive member of the Puerto Rican series, A. krugi, is not very different from cristatellus and guncllachi and would certainly be classed with them except for its obvious position at the base of a small grass anole radiation on Puerto Rico. The Cuban anoles which display a strong structural affinity in spite of some hahilat difference are more probably close relatives of the Hispaniolan series. There is in fact no substan- tial reason for doubting the relationship. 6 BREVIORA No. 136 It must be pointed out that the squamation pattern with a strongly developed middorsal zone of enlarged keeled scales, smaller laterals, and strongly keeled ventrals as large or larger than the middorsals is common in mainland Anolis, particularly so in Central America. This pattern occurs also in the Greater Antilles in three species which, though certainly anoline, are currently referred to other genera: the Cuban species {ophio- lepis) to Norops, and a species from Navassa (harboiiri) along with one from Hispaniola (wetmorei) to ChamaeUnorops. The mainland forms exhibit a whole spectrum of conditions in regard to the distinctness, number of scale rows, size of scales involved in the dorsal zone, etc. No described form seems close enough to the Hispaniolan or Cuban grass anoles to be worth serious consideration as representing the ancestral stock. Norops opliioh'pis, which occupies the grass anole habitat in Cuba, does not seem related either. It has some features peculiar to itself — the reduction of the canthal ridge scales to two, the small number of scales in the loreal area {ca. 10-12), the very elongate scales betv/een the nostrils, the large mental scales — that are unlike not onh- the semilineatus-alutaceus groups but its supposed congeners on the mainland. The relationships of ophiolcpis are probably with Anolis sagrei and more remotely with the homolechis complex ; there are certainly no grounds for postulating close affinity to the scmilineatiis-ahitaceus set. ChamaeUnorops harhouri and C. wetmorei are even more dis- tinct. The basic pattern of squamation is quite heterogeneous and yet upon this has been imposed a second pattern of enor- mously enlarged keeled dorsals and hugely enlarged keeled ventrals exaggerated beyond that seen in any other forms. This picture, like the apparent radiation of forms on Puerto Rico and the extraordinarily varied array of forms on the mainland, suggests strongly that the pattern — enlarged mid- dorsal zone, enlarged keeled ventrals — is one of several stereo- types that the anoles have again and again produced, that this is one of a limited set of squamation patterns possible to the anolines and therefore produced in parallel fashion in many times and places. It is this parallelism that contributes to the notorious "dif- ficulty" of the genus Anolis. Narrow groups are rather easy to recognize (though the specific and infraspecific structure within 1961 ANOLIS SEMILINEATUS GROUP 7 the group may be puzzling in the extreme) but wider relation- ships (at least when externals only are considered) are problem- atical, becoming- obscurer with each step more distant from the species group. Origin of the semilineatus group The species of the semilineatus group are more uniform than the related Cuban series. They most resemble cyanoplcurus, the middle term in the morphological series of Cuban forms. This species has its range in extreme eastern Oriente and is thus geographically closest to the Hispaniolan group. It therefore seems probable that the semilineatus series on Hispaniola has been rather recently derived from a cyanoplcurus-VikQ Cuban ancestor but has been on Hispaniola long enough to achieve island-wide dispersal and moderate dilferentiation at the specific level. Acknowledgments I have had the advantage of discussions with A. Stanley Rand and with Dr. Hicliard Etheridge. The latter 's osteological evi- dence for species groupings within Anolis (unpublished thesis, University of Michigan) has in part confirmed, in part guided my own thinking on the wider relationships of Anolis species. The map-diagram was prepared by Patricia Grubbs. REFERENCES CITED Mee.tens, R. 1939. Herpetologischer Ergebnisse eiiier Reise nach der Insel His- paniola, Westindien. Abhandl. Senckenberg. naturf. Ges., 449: 1-84. Wetmoee, a. and B. H. Swales 1931. The birds of Haiti and the Dominican Republic. Bull. U.S. Nat. Mus., 155: 1-483. Williams, E. E. and A. S. Rand 1961. Notes on Hispaniolan Herpetology. l'. A review of the Anolis semilineatus group with the description of Anolis rochranae, new species. Breviora, No. 135 : 1-11. BREVIORA No. 136 ^ ^ 3 S --S . o a; o o 03 CO g^ c m Ol a Ol 3 c3 I- o o s 05 ^ o •^ .s CO o 52 OJ o in : ft > ^ .S « -^3 e 05 10 15 J2 S o O ft ^ O d t/) J- "^ o .2 «. V =H 'J^ '■« *^ s o :;: £ c ;« -rl o _ «H ft el 2 "^ S <^ en (D ;-> :§ 1 ? ^ ■73 •-* A en o -^ ft 05 Jh ^ o OP -si o -o .a 2 O 03 '^ ® 0) 3 o ^ ^ §'^ " g =2 *j o *^ O >» 13 '^ S o CO ,^ 5:; S » o 3 fc- BREVIORA MusenaiM of Comparative Zoology Cambridge. Mass. April 10, 19(31 Number 137 NOTES ON HISPANIOLAN HERPETOLOGY 4. ANGUS KOOPMANI, NEW SPECIES, FROM tllE SOUTHWESTERN PENINSULA OF HAITI By a. Stanley Rand Biological Laboratories, Harvard University Among the reptiles and amphibians collected for the Museum of Comparative Zoology in Haiti, with the support of an Ameri- can Philosophical Society grant, during the summer of lfJ60, are seven specimens that appear to represent an undescribed species of Anolis. These lizards are small, moderately proportioned, dull colored in life, as well in preservative, and rather nondescript animals. In life the adult males possess a dark gray gular fan and an orange-pink chin and throat that distinguish them immediately from any otlier Hispaniolan Anolis. When this chin and throat color disappears in alcohol, the combination of scale characters distinguishes the species, but there are no unique characteristics. Dr. Karl Koopman, Assistant Curator of Mammals, Chicago Natural History Museum, provided financial assistance and encouragement that helped to make the collection of these specimens possible. In recognition of his aid the new species is called : Anolis koopmani new species Type. MCZ 62541, adult male. Type locality. Carrefour Canon, 350 m. altitude, near Ducis, N. of Aux Caves, Haiti. Collector. A. S. Rand and J. Lazell, 4 August 1960. Paratypes. Adult males, MCZ 62542-3; adult females, MCZ 62544-5 ; young males, MCZ 62546-7. All from Les Platons, 750 2 BREVIORA No. 137 in. altitude, above Carrefour Cauoii, Haiti, A. S. Hand and J. Lazell, 5 August 1960. Diagnosis. The presence of a zone of much enlarged middorsal scales, keeled head scales, and keeled, imbricate, and pointed ventrals distinguish this species from all the Ilispaniolan Anolis except those of the seynilineatus group {stmilineatus, olssoni, and cochranae) . It is distinguished from the latter in having 6-8 (not 10) enlarged middorsal rows, 5-8 (not 3-5) loreal rows, and 3-5 (not 1-3) scales separating the interparietal from the supraorbi- tal semicircles. It differs also in coloration, the males having pinkish-orange chin and throat. Description. (In the following description variations occurring in the paratypes follow, in parentheses, the condition in the Head. Head scales strongly keeled. Frontal dejiression mod- erate. Scales across snout between second and third canthals 8 (7-9). Nares anterior to canthal ridge; separated from rostral by 1 scale. Canthal ridge distinct, not exaggerated, composed of 4 (4-5) large scales preceded by 2 (1-3) small ones. Second canthal largest, third next in size, first and fourth subequal. Posterior frontal subequal to (slightly smaller than) anterior supraorbital ; separated from canthals by 2 (1-2) scales. Anterior supraorbital separated from canthals b}' 3 (2-3) scales. Supra- orbital semicircles separated by 3 (1-3) scales; separated from supraocular disc by one row of small scales (occasional narrow contact). Supraocular disc of 5-6 (5-7) enlarged keeled scales; separated from superciliarjr by 5-6 rows of granules. A single elongate superciliary. Interparietal scale slightly smaller than ear (lA to slightly smaller) ; separated from supraorbital semi- circles by 4-5 (3-5) scales. Scales in center of supratemporal area, granular, smaller than flank scales, smallest in center. Scales over temporal bar not (very slightly) enlarged. Temporal scales like supratem- poral scales. Suboculars broadly in contact with supralabials, separated from canthals (very narrow contact), not continued behind eye as a series of large scales. Supralabials to center of eye 5 (5-6). Loreal rows 5 (5-8). Loreal scales subequal in size. Mentals broader than long, in contact with 6 (5-6) throat scales posteriorly. Xo series of enlarged sublabials. Central throat scales small, elongate, keeled. (hilar fan: Gular fan small. Scales slightly smaller than (sub- equal to) ventrals, keeled. Trunk: Middorsal scales much larger than flank scales, grad- i:»g: ANOLIS KOOPMANI iug into them. Rows of enlarged middorsals 6-8. Veutral scales larger than middorsals, imbricate, keeled. Limbs and digits: Scales on arms and legs larger tlian ven- ti-als, multiearinate. Hand and foot scales multicarinate dorsally. Lamellae under phalanges 2 and 3 of fourth toe 18 (17-19). Interdigital pads narrow. Tail: Tail round in cross-section. Verticils indistinct. En- larged postanal scales present in males. MEASUREMENTS Snout-vent Total Head Tibia Hvnd ley Sex MCZ # length length length lengtli length male 62541 34 mm 114 mm 9 nmi 11 mm 28 mm (type < 1 62542 39 135 10 12 33 ( 1 62543 38 10 12 31 i i 62546 23 68 6 7 18 ( I 62547 21 6 7 17 female 62544 33 7 9 26 i t 62545 33 109 8 10 26 In life a low nuchal and dorsal crest seems permanently raised (absent in females and young males). Color in life: Male, MCZ 62542, uniform gray-brown above; a whitish line, black edged above, from over shoulder to hind leg, indistinct for the posterior half of its length, below this line the flanks with scattered dark spotting. Belly light brown, chin and throat pale pinkish-orange, with a few scattered black spots, gular fan scales colored like the chin, but the skin dark gray; iris blue. Female, MCZ 62544, plain brown above, a middorsal stripe with a scalloped margin outlined with darker brown, a yellow stripe from below eye to over shoulder, continued faintly to hind leg; venter yellov/ish with faint dark spotting on throat; iris blue. . •..,. , Habitat: The type was taken in a bush along a trail "git the edge of a coffee grove. It was found at dusk among small twigs about three feet above the ground where it probably had climbed to spend the night. Of the six paratypes, three adults were found in heavily shaded coffee groves. All were on the ground among damp leaf litter. One juvenile was found six inches up in low herbaceous vegetation at the edge of the coffee grove. The others were purchased from a small boy for two cents each. Relationships: The relationships of this species are obscure; 4 BREVIORA No. 137 it does not seem to be particularly close to any species now known either from Ilispaniola or elsewhere. The only Anolis outside of Hispaniola that are at all similar to this species are the alutaceus, clivicolus, spectrum, cyanopleu- rus groups of Cuba, and this similarity lies primarily in the presence of a zone of enlarged middorsals and does not extend to other details. I interpret this as parallelism. Within Hispaniola, A. koopmani is superficially most like semilineatus and olssoni. A zone of enlarged middorsals, keeled imbricate, pointed ventrals, keeled head scales, a lateral stripe, narrow digital expansions and small size occur also in .semilineatus and olssoni and suggest a relationship to them. However, most of these characters are not unique to semilifieatus and olssoni even in Hispaniola. The nature of the zone of enlarged middorsals (fewer rows that decrease in size laterally), the strong sexual dimorphism in color, the less attenuate body shape, and generallj' smaller scales all argue against this rela- tionship. Anolis ricordi, clistichus, cybotcs, armouri, shrevei, chloro- cyanuSyCoelestinus and Chamaelinorops wetmorei all have special- izations that seem to exclude them from close relationships with koopmani. The remaining species, Anolis monticoJn, darlingtoni, chr'sto- phei, heridcrsoni, hahanicocnsis and Xiphoccrcns darlingtoni are poorly known. It is possible that some if not all of them are closely related to one another and that koopmani's relationships lie with these. However, until more information is available for this assemblage of species, particularly in regard to color in life, behavior and ecology, this hj-pothesis must remain very tentative. Acknowledgments: I am grateful to Dr. Ernest E. Williams for his advice and to Mr. James Lazell and M. Luc Whiteman for their assistance in the field. I wish also to thank M. Leonce Bonnfil fils and the other members of the Department of Agri- culture of Haiti who helped us in so many ways. BREVIORA Mmseiuiioi of Comparative Zoology ("A:Mi{Hir)(;E, ]Mass. June 14, 19G1 Number 138 PFEIFFER'S UNFIGURED SPECIES OF STROPHOCHEILVS {MEGALOBULIMUS ) By T. E. Crowley and T. Pain This paper is a supplement to J. C. Bequaert's "Moiioy,Taph of the Strophoeheilidae, a Neotropical Family of Terrestrial Mollusks" (Bull. Mus. Comp. Zool., Harvard, 100: 1-210, 1948). When Dr. J. C. Bequaert published his monograph on the Strophoeheilidae he was unable to deal in detail with two species described by Pfeiffer from specimens in the Cumino' collection, now in the British Museum (Natural History), London. Neither of these had ever been figured, and the types are so far the only specimens known. Opportunity has, therefore, been taken to complete Dr. Bequaert's monumental work with figures of Pfeiffer 's almost unknown species. In addition we are describ- ing and figuring S. (M.) capiUaccus (Pfeiffer), so far unfigured, and S. indigcns Fulton. The authors wish to express their grateful thanks to the Brit- ish IMuseum authorities for permission to examine and photo- graph the types for reproduction herein, to Dr. R. Zischka for specimens of S. (M.) nuligens Fulton, to Drs. W. Blume and W. Weyrauch for the loan of material, and to Mr. S. P. Dance and Mr. J. A. AYillson for their generous assistance in photo- graphing specimens. Strophocheilus (Megalobulimus) hector (Pfeiffer) Plate 1, Figure 1 BuUmris hector Pfeiffer 1857, Malak. Blatt., 4, p. 157 (Brazil); 1859, Monosr. Helie. \iv., 4, p. 367; 1868, Op. cit., 6, p. 11; 1876. Op. rit., 8, p. 15; 1877, Op. cit., 8, p. 604. Paetel, 1889, Cat. Conch. Samml., 4tli ed., 2, p. 212. BuUmus (Bonis) hector von Martens 1860, in Albers, Die Heliceen, 2nd 2 BREVIORA No, 138 ed., p. 192; 1876, Xovit. Com-hol., Abt. 1, 5, pts. 50-51, p. 21. Pfeiffer, 1879, Nomeiicl. Helie. Vic, p. 224. Strophocheihis (Thaiimastus) hector Pilsbry 1895, Man. of Conch., (2) 10, p. 50. ThaumasUis hector Pilsbry 1902, Man. of Conch., (2) 14, Classification, p. xxi. Strophocheilits {? MegalohuUmus) hector Pfeiffer, Bequaert 1948, Bull. Mus. Comp. ZooL, Harvard, 100: 118. Original description: "T. subimperforata, elongato-ovata, solidula. sub epidermide decidua, fulvida alba; spira conica, apice rotundata; anfr. 6 convexiusculi, summi conferte capillaceo striati, ultimus spiram vix superans, plicato-striatus et obsolete decussatus ; columella leviter arcuata, nou plicata ; apertura sub- verticalis, acumiuato-ovalis, intus albida, nitida ; perist. album, marginibus callo albo junetis, dextro subiuerassato, brevissime expanso, columellari superue dilatato, adnato. — Long. 71, diam. 35 mill., Ap. 36 mill, longa, 191/2 lata." New Measurements of Adult Holotype Greatest Aperture Aperture Length Width Length Width 71 mm. 39 mm. 35 mm. 26 mm. 6 whorls Specimen examined: Brazil (Miers Coll.), holotype (Brit. Mus., Nat. Hist.). Remarks. Bequaert (11)48, p. 118), who had seen no speci- mens, was inclined to exclude hector from the Strophocheilidae, and Pilsbry (1895, p. 50) placed it in Thaiimastus. Von Mar- tens (1876), however, suggested that it might be related to Stropliocheilns (M.) ohlongus (Miiller). Careful examination of the type has con\dnced us that it is indeed correctly referred to the Strophocheilidae, being by reason of its nepionic sculpture a member of the subgenus Mega- lohulimus. It does not appear, however, to be in any way related to S. (M.) ohlongus, the shell being longer and narrower in proportion, much paler and with a white lip. It is further- more much thinner, the apical sculpture finer, and is covered with a brown periostracum. S. (M.) hector would appear to us to be quite distinct from any other species of Megalohulimus so far known. 1961 STROPHOCHEILUS 3 Strophociieilus (Megalobulimus) cocapatensis (Pfeiffer) Plate 1, figure 2 Bulimus cocapatensis Pfeiffer 1855 (August), Proc. Zool. Soc. London, (for 1855), p. 115 (Coeapata, Bolivia); 1859, Mouogr. Ilelic. Viv., 4, p. 367; 1868, Op. cit., 6, p. 11; 1876, Op. cit., 8, p. 15. Paetel, 1889, Cat. Conch. Samnil., itli ed., 2, p. 209. Bulimus {Bonis) cocapatensis Pfeiffer 1856 (January), Malak. Blatt., 2, (for 1855), p. 1-17. Von Martens, 1860, in Albers, Die Heliceen, 2nd ed., p. 192; 1876, Xovit. Conchol., Abt. 1, 5, pts. 50-51, p. 9. Pfeiffer, 1879, Nomencl. Helic. Viv., p. 224. Strophociieilus (Borus) cocapatensis Pfeiffer, Pilsbry 1895, Man. of Conch., (2) 10, p. 20. Strophocheilus (Borus) cocopatensis Pfeiffer, Pilsbry 1895, Man. of Conch., (2) 10, p. 12; 1902, Op. cit., (2) 14, Classification, p. v. Misspelling of cocapatensis. Bulimus corapatensis Pfeiffer, Paetel 1889, Cat. Conch. Samml., 4th ed., 2, p. 10. Misspelling of cocapatensis. Strophocheilus cocopatensis Pfeiffer, Pilsbry 1930, Proc. Ac. Xat. Sci. Phila- delphia, 82, p. 355. Misspelling of cocapatensis. Strophocheilus (Megalohulimiis) cocapatensis Pfeiffer, Bequaert 1948, Bull. Mus. Conip. Zool., Harvard, 100, no. 1, p. 126. Original description: "B. testa imperforata, ovato-oblonga, solida, minutissime decussata. sub epidermide virenti-fulvida violaeeo-earnea ; spira convexo-euiiiea, apiee obtusa; sutura al- bida, irregulari ; anfr. oYo superis radiatim costatis et minutis- sime granulatis, sequentibus peroblique descendentibus, parum couvexis, ultimo spiram sub-aequante, basi rotundato ; columella recedente, leviter arcuata; apertura subverticali acuminato- ovali, intus margaritacea ; perist. iiicrassato, breviter expanse, marginibus eallo nitido junctis, eolumellari dilatato, adnato. Long. 67, diam. 30 mill." Measurements of Adult Shells Greatest Aperture Aperture Length Width Length Width 67 mm. 33 mm. 31 mm. 21 mm. 5^4 whorls. Holotype 67 33 31 20 514 whorls. Para type 66 33 32 20 5i/> whorls. Para type Specimens examined: Coeapata, Bolivia (Bridges Coll.), 3 types from the Cuming Collection (Brit. Mus. fXat. Hist.]). Reynarlxs. As pointed out by Bequaert (1948. p. 127), the radially ribbed and minutely granulated nepionic whorls, men- tioned in Pfeiffer 's original description, are characteristic BREVIORA No. 138 of Mcgalohulimus, to which subgenus S. cocapateusis un- doubtedly belongs. Pilsbry (1930, Proe. Ac. Nat. Sei. Philadel- phia, 82, p. 355), in describing his S. carrikeri, infers that it may be related to 8. cocapatensis, but we are unable to see any justifi- cation for this assumption. Bequaert suggests that it is not im- possible that cocapatensis may be the same as S. intcrtextus Pilsbry, but comparison of a specimen of the latter with Pfeif- fer's type has convinced us that they are in no way related. The shell of cocapatensis is imperforate, long, thin and deli- cate, brown in color, with a very streaky, pink-fiushed appear- ance. The spire is attenuated, the apex pointed, the mouth long and narrow, the aperture brown within, and the outer lip thin, white, slightly reflected. Columella and callus white. Pfeiffer compared cocapatensis with S. rosaceus, but, as pointed out by Bequaert, the nepionic sculpture is typical of Megalohuli- inus and is not found in CliiUhoi'us, to which subgenus S. rosaceus belongs. Pfeiffer later 1856) placed it between "S. matthewsi" {— leucostoma) and *S'. capillaceiis but, as he does not show it as being- then in his collection, this opinion would seem of little value. Strophocheilus (Megalobulimus) capillaceus (Pfeiffer) Plate 1, fig-ure 3 Bulimus capillaceus Pfeiffer 18.35 (July), Proc. Zool. Soc. London, (for 1855), p. 93. StropJiocheilus (Bonis) capillaceus Pfeiffer, Pilsbry 1895, Man. of Conch., (2) 10, p. 31, PI. 14, fig. 69. StrophocJtcilus {Mcgalohulimus) capillaceus Pfeiffer, Bequaert 1948, Bull. Mus. Conip. Zool., Harvard, 100, p. 120 (full synonymy;, PI. 14, fig. 5. The type of *S'. (M.) capillaceus is in the British Museum (Nat. Hist.), from the Cuming Collection. It consists of three syntypes, of which that figured herein is now chosen as lectotype, no holotype having been selected by Pfeiffer. Measurements of Adult Shells Greatest Aperture Aperture Length Width Length Width 64 mm. 38 mm. 37 mm. 24 mm. 5 whorls. Lectotype 60 39 35 30 5 whorls. Sjnitype S3 33 23 19 5 whorls. Syntype 67 40.5 40.5 22 5% whorls. Huanaco 69 42 41 21.5 5% whorls. Santa Ana 1961 STROPHOCHEILUS Plate 1 Fig. 1. Strop1(0chcihtf; {McgalobuUmm) hector (Pfeiffer). Holotypt". iiat. sizt". Fig. 2. Sirophoclteilits (Mcgalohulimus) cocupatensis (Pfeiffer). Holotyi)o, nat. size. Fig. 3. Strophoeheihis (Mer/alobulimus) capiUaceus (Pfeiffer). Lectotype, nat. size. Fig. 4. Slrophochcilus (ilegaJohuJiinus') maximus indigens Fulton. Apical aspect of immature shell, much enlarged. 6 BREVIORA No. 138 Sprciyucns examined : "Banks of the Kiver Solimoes, " Peru (Cuming Collection). Near Santa Ana, Kio Urubamba, Peru, 3500 ft. (W. Weyrauch Coll., in Pain Collection). Huanaeo, Peru (Pain Collection). Remarks. Bequaert (1948, p. 120) has dealt at length with the probable relationships and position in the subgenus of S. capil- laceus, and gave an excellent figure of it, together with a com- plete synonymy. To this very full account there is nothing which we can profitably add. Strophocheilus (Megalobulimus) maximus indigens Fulton Plate 1, figure 4; Plate 2, figures 5, 6 Bulimus Jcremnoicus d 'Orbigny 1837, Voyage Am6r. Meridion., 5, pt. 3, Moll., p. 300 (in part only: some specimens from Yuraeare, Bolivia, the locality given in Explanation of Plates, p. 695, for fig. 3), PI. 35, fig. 3 only. Not Helix kremnoica d 'Orbigny 1835. Stroplwcheilus (Bonis) maximus ? var. Icremnoicus d 'Orbigny, PUsbry 1895, Man. of Conch., (2) 10, p. 16, PI. 5, fig. 28 (copy of d'Orliigny's fig. 3); 1902, Op. 0(7., (2) 14, Classification, p. iv. Strophocheilus (Borus) indigens Fulton 1914, Proc. Mai. Soc. London, 11, pt. 3, p. 165, fig. (Peru). Strophocheilus (Megalobulimus) iyidigens Fulton, Bequaert 1948, Bull. Mus. Comp. Zool., 100, No. 1, p. 98, PI. 24, fig. 1 (copy of Fulton's 1914 fig.). Strophocheilus indigens Fulton, Blume 1952, Arch. f. Mollusk., Frankfurt, 81, pts. 4-6, p. 105. Strophodheilus (Borus) l-remnoieus subsp. vcstitus Pilsbry 1926, Proc. Ac. Nat. Sci. Philadelphia, 78, p. 6 (Bolivia, probably in Dept. Cochabamba), PI. 2, fig. 7. Strophocheilus (Megalobulimus) maximus vestitus Pilsbry, Bequaert 1948, Bull. Mus. Comp. Zool., Harvard, 100: 94, PI. 19, fig. 4. Original description: "Shell ovate-oblong, yellowish brown, moderately solid; spire about 13 mm. longer than the aperture; whorls 61/2? apex smooth, the second and third whorls with prom- inent oblique plications, last two volutions polished and appar- ently smooth, but under the lens are seen to be finely granulated, the granulations being strong on the middle whorls and gradu- ally becoming weaker towards the aperture ; the lower whorls have also some irregidar and almost obsolete plications ; aperture sub-oval, whitish within ; peristome thickened and very slightly expanded, white, margins joined by a moderately thickened white callus. Alt. 110, Diam. Maj. 47 mm. The nearest species to this is *S'. (Borus) huascari Tschudi, which is broader, has a wider aperture, a rougher and duller surface, and its apical plications are much finer and closer together than in indigens." 1961 strophocheilus Measurements of Adult Shells Length Greatest Aperture Aperture Width Length Width Whorls 135 mm. 57.5 mm. 60 nun. 41 mm. 7 Bolivia: Sacha, Yungas, 800-1500 m. (Bavarian State Mus.) 132 63 59.5 37 7 127 58 56 39 7 128 58 55 31.5 6Y2 118 58 56 24 — 116 48 47 31 6% 101 •16 47 26 51/0 100 49 48 27.5 51/2 Bolivia (W. Weyraudi Collection) 118.5 61 57.5 40.5 6 Bolivia: Chapare, 400 m. (Bavarian State Mus.) Type of vestitus Pilsbry Holotype of indigens, Peru Bolivia (T. Pain Collection) Bolivia (W. Weyrauch Collection) Specimens examined: Peru, type of indigens Fulton (British Museum [Nat. Hist.], No. 1915-1-5-199). Yungas cle Palmar, 1200 m., Bolivia (R. Zischka Coll., in W. Blume, T. Pain, and W. Weyrauch Collections). Remarks. Comparison of the type of indigens, together with the shells from Bolivia, with Bequaert's (1948) figure of vestitus Pilsbr3% leave us in no doubt that they are identical. All shoAV the strong, prominent oblique plications on the second and third whorls, noticeably absent on both the typical maximus and the subspecies huascari. As pointed out by Bequaert (1948, p. 94), indigens { = vestitiis) is of considerable interest in that it bridges the gap between typical maxim its and subspecies huascari in re- spect of its relatively wider spire, narrower body-whorl and smaller mouth. Fulton, who described indigens from a unique holotype, did not apparently connect it with maximus, although he drew at- tention to its close resemblance to huascari. Bequaert (1948, p. 94), dealing with vestitus, makes no men- tion of the prominent sculpture, although this is easily recog- nized in his excellent photograph of the shell he selected as holotype from Pilsbry 's type set (Ac. Nat. Sci. Philadelphia No. 138105). Dr. Weyrauch informs us (in lift.. 1958) that he considers the elongated shell from Oxapampa, Peru, referred by Bequaert to vestitus, to be a typical maximus on account of the aperture being much longer than in ineligens (= vestitus). A similar elongated shell from Peru, without more definite locality, BREVIORA No. 138 kindly sent by Dr. Weyraudi. sliows traces of a dark periostra- cum and, althonjili much worn aliont tlie spire, has the long aper- ture characteristic of typical ))ioximiis. S. [M.) ntn.ri))U(s ijuligois Fulton has not so far been ob- tained in Peru by Dr. Weyrauch, but, from the similarity of the fauna of southeastern Peru and northeastern Bolivia, there can be little doubt that iiidigcns occurs also in Peru. Plate 2 Fig. 5. Strophocheilus (Megalobulimus) maximus indigens Fulton. Hol- otype, nat. size. Fig. 6. Strophocheilus (MegaloiuUmus) maximus indigens Fulton. Yungas de Palmar, Bolivia, nat. size. BREVIORA Mmseium of Comparative Zoology Cambridge, ]Mass. JiiXk 15, 1961 Numbef^ 139 A NEW SPECIES OF SPIIAEEODACTYLVS FKOM NORTHERN HAITI By Jaaies D. Lazell Although the genus Sphaerodactylus on Ilispaniola is suf- ficiently diversified and confused to warrant at least a partial revision, the species here described is so remarkably different from any other form that I am confident in naming it at this time. The new species is named for Mr. Benjamin Shreve of the Museum of Comparative Zoology for his current work on the sphaerodactyls of Hispaniola. Sphaerodactylus shrevei sp. nov. Type: MCZ No. 62548, Mole Saint Nicolas, Haiti. Coll.: J. Lazell and A. S. Rand, 16 July, 1960. Diagnosis: The combination of the following characters serves to distinguish this species from any other found in the Antilles: the presence of a highly convex snout as seen from the side (loreal region also somewhat convex) ; very large keeled dorsal scales beginning at the level of the axilla ; and the paravertebral arrangement of these dorsals, producing a middorsal zone not of granules but of small and large, irregularly placed scales. Description of type. Snout short. Eye nearer tip of snout than ear. Snout, as seen from the side, highly convex. Loreal region also somewhat convex. Rostral large with a partial median cleft. Nostril between rostral, first supralabial, a large supranasal (= internasal) and two small postnasals. A single scale between the supranasals (= internasals), which border the rostral posteriorly. Granular scales on top of snout somewhat larger than interoeular or nape scales. Four supralabials, of 2 BREVIORA No. 139 about equal length, to the center of the eye, followed by two smaller ones. Four infralabials gradually decreasing in size followed by two abruptly smaller ones. Mental short, wider than long, bordered posteriorly by two postmentals. Supraciliary spine small. Squamation of head and neck granular to the level of the shoulder; at that level a moderately rapid change to large, flattened, heavily keeled dorsals. Twenty-five dorsals from the level of the axilla to the posterior level of the hind limb. Five dorsals in the standard distance. A very ill-defined middorsal zone of smaller keeled scales not forming a continuous row but with the large dorsals meeting irregularly along the middorsal line. Middorsals subimbricate or not imbricating, the laterals more distinctly imbricating. Throat scales smooth, granular, juxtaposed. Chest and belly scales larger, smooth, cycloid, broadly imbricate, about nine ventral scales in standard distance. Scales of anterior surfaces of limbs imbricate, cycloid, smooth, somewhat smaller than ventrals. Scales of posterior surfaces of limbs granular, smooth. Digital pads approximately twice as broad as the subdigital lamellae. Ten infradigital lamellae under fourth toe. The type, the only specimen yet col- lected, lacks the tail. It is a female and the escutcheon there- fore cannot be described. Coloration in life. S. shrevei is a dull-colored animal with a pattern composed of three basic hues — each tending to be unique on an individual scale. There are very irregular dark grey-broW'U blotches across the dorsum ; beginning at the back of the head there are three such markings to the slioulders. There are three more crudely "Y" shaped markings on the body, the most anterior of which bifurcates to the right, the remaining ones bifurcating to the left. There are two small blotches on the right side of the rump and one on the left. The second transverse blotch, on the nape, is broken by a light middorsal line that continues down through the fourth marking and then fades out. The ground color of the dorsum is ash grey. There are scattered over the dorsal surface short transverse series of white or partly white scales — from two to four in a row — that appear to have no correlation whatever with the rest of the animal's pattern. The top of the head is ash grey except for a very irregular, dark, grey-brown blotch on the parietal area. Coming back from the eye are two stripes, one of which runs downward across the cheek ; the other nearly connects with the first transverse marking on the back of the head. Xot 1961 SPHAERODACTYLUS SHREVEI 3 including the stripe across the cheek, there are five vertical dark markings across the pale labials, the anterior two of which are connected at the edge of the mandible. There are dark streaks on the lateral edges of the chin and throat, the underside of the hind limbs, and across the venter just anterior to the anus. The ventral surface is white ; a close examination reveals that on each scale there are tiny black dots. This peppering becomes more noticeable laterally and posteriorly. Along each side of the animal is a line of partially connected, small, dark, grey- brown blotches; just ventral to this row is another composed of widely spaced, single, dark, grey-brown scales. All three of the animal's hues: white, grey and grey -brown, are simply variations in the intensity of speckling on each scale with tiny black or brown dots. The pattern of the animal bears no resemblance to that of the young or females in the species to which it has been com- pared, or to any other Sphaerodactylus I have seen. Habitat. The type specimen was taken from a large circular rock pile about two-and-one-half feet deep ; this sort of rock pile is the result of removing the debris from a heap of charcoal after burning, and is composed of rocks that vary in size from that of a golf ball to nearly the size of a football. This particu- lar heap was of some age, for even in the arid terrain of Mole Saint Nicolas several fair-sized thornbushes had sprouted up in it. Collecting was very difficult, for any animal uncovered could generally manage to dart back into the pile before the collector could safely ascertain that it was not a scorpion or some other unpleasant handful. In order to get best results we excavated areas through the pile, dividing it up into more manao-eable smaller piles; this system netted Cdcstns, TypJihrps. and Tropidophis, as well as the type of Sphaerodactylus shrevei. Another specimen of apparently the same species escaped, leav- ing only its tail behind. Due to the relative inaccessibility of peninsular north-western Haiti it may be some time before additional specimens can be obtained. Comparisons. From S. copei, the only other comparably large- scaled Hispaniolan form, S. shrevei differs in the following characters: (1) Snout seen from the side highly convex: (2^ No middorsal zone of gramdes; (3) Dorsal scales flatter, not swollen, apt not to imbricate, especially in the middorsal area ; (4) Pattern a series of irregular dark dorsal blotches with 4 BREVIORA No. 139 a line of often connected smaller blotches along each side ; dorsal blotches broken by a light middorsal line anteriorly. From S. scaber of Cuba, S. shrevei differs in all the men- tioned characters and in snout length, which averages slightly shorter in the Cuban form. S. samanensis Cochran of the Dominican Republic resembles ^■. shrevei somewhat in squamation but the dorsals are smaller and the ventrals larger. There are no smaller middorsal scales and the pattern is very different. >S*. shrevei differs from S. becki of Navassa again in the absence of a middorsal zone of granular scales and in the flatness of the dorsal scales, which are rounded, swollen, and rather tubercular in 8. becki. Two Jamaican forms, S. richardsoni and 8. parkeri, occa- sionally possess dorsal squamation similar to that of 8. shrevei in that while there is no middorsal zone of granules there may be small scales irregularly scattered along the middorsal line, but m general the arrangement of the scales is much more regular and not of a paravertebral nature. From both of these species 6'. shrevei differs in the following characters: (1) Snout seen from the side highly convex; (2) Head granules extending posteriorly to level of axilla instead of just onto nape; (3) No enlarged, clearly defined canthal scale; (4) Pattern and colora- tion. From 8. parkeri it differs also in having only a single small scale between the internasals. There is a ^'aguer resemblance to 8. shrevei in some Lesser Antillean forms. 8. vincenti and aS'. microlepis, for example, show a tendency towards reduction in size of the middorsal scales, but in these forms the pattern of squamation tends to be very regular and the scales are much smaller. The new species has been compared with these forms largely because there are apparently no closer ones, although it bears little resemblance to any of them. Its relationships at the moment are not at all clear. Acknowledgments. My thanks are due to Mr. A. S. Rand for help in collecting the specimen described, to Mr. B. Shreve for checking the description and com.parisons made and to Dr. E. E. Williams for the loan of comparative material and for readin'v the manuscript. A grant from the American Pliiht- sophical Society supported the expedition which resulted in the discoverv of 8. shrevei. 1961 SPHAERODACTYLUS SHKEVEI 5 EEPERENCES Barbour, T. 1921. Sphaerodactylus. Mem. Mus. Comp. Zool., 47: 217-277. Cochran, D. M. 1941. The herpetology of Hispaniola. Bull. U. S. Nat. Mus., 177: 1-398. BUIBAL, R. 19j9. a new Sphaerodactylus from Oriente, Cuba. Herpetologica, 15: 89-93. Table 1 Comparisen of the snout length ratios of six species of SpTiaerodactylus. The ratio is obtained by dividing the standard distance (i.e., the distance from the center of the eye to the tip of the snout) into the distance from the center of the eye to the ear opening. The higher the figure obtained the shorter the snout. Specimens Mean Value Range S. richardsoni 5 .72 .67- .81 S. parlceri 3 .76 .73- .78 S. copei 10 .76 .70- .80 S. shrevei 1 .80 - S. scaber 6 .82 .71- .92 S. becki 4 .90 .78-1.00 BREVIORA Miiseiuim of Compsirative Zoology Cambridge, Mass. June 27, 1961 Number 140 A PRELIMINAKY REVIEW OF THE NEARCTIC SPECIES OF SIEROLOMORPHA (HYMENOPTERA) By Howard E. Evans In the course of collecting Bethyliclae and examining material in various museums, I liave encountered a good many specimens of the curious bethylid-like genus Sierolomorpha Ashmead. Some of these specimens are so different from the type, ambigua, for a long time the only known species, that it seems desirable to provide names for them. Krombein (1951, U. S. Dept. Agri. Monogr. 2, p. 748) has already pointed out that there appear to be several species in North America. The present paper does not pretend to be an exhaustive treatment of this genus. Names are provided for several of the more distinctive species, but before a definitive revision is possible more collecting must be done and more detailed studies made of structural details and their vari- ation. Actually, Sierolomorpha is not a bethylid or even especially close to the Bethylidae. Schuster (1949, Ent. Amer., 29: 124) is correct in pointing out its close similarity to certain Mutillidae and Tiphiidae, and the arrangement in the Synoptic Catalog (Krombein, 1951, op. cif.), in which the genus is placed in a monogeneric family between the Tiphiidae and Mutillidae, is probably the best that can be achieved at present. Sierolomorpha is related to the most primitive Scolioidea, and may not be far from the stock which gave rise to the Bethylidae. However, in virtually all structural details it stands very much closer to the Tiphiidae than to the Bethylidae. Only one species of the genus, hospes Perkins, has been de- scribed from outside the Nearctic region (from Hawaii). Krom- bein has suggested that this species may have been introduced from North America. In the collection of the U. S. National Museum there is a single male of an undescribed species from 2 BRRVIORA No. 140 Panama. I have seen no other specimens from outside the United States and Canada. The most useful structures for the separation of species in this genus appear to be the antennae, especially the tyloides of the male (Figs. 1-6), the propodeum, and the first two abdominal tergites. There seems to be little variation within the genus with respect to the mandibles, elypeus, and most details of the thorax including the legs and wing venation. I have therefore made little mention of these structures in my descriptions. I have studied the male terminal ia of selected specimens and found some minor variation, particularly in the volsellar cuspis. However, the differences are so slight and subtle that further study to determine the extent of individual variation did not seem war- ranted. I have therefore made no mention of characters of the terminalia in the keys and descriptions. The following abbreviations are used for the museums and private individuals that supplied material for this study: AMNII, American Museum of Natural History, New York; CIS, Cali- fornia Insect Survey, Berkeley; CNC, Canadian National Col- lections, Ottawa; HKT, Henry K. Townes, Ann Arbor, Michi- gan ; MCZ, Museum of Comparative Zoology, Cambridge, Mass. ; UCD, UniverNity of California at Davis; USNM, U. S. National Museum. Key to Species Females ^ . Abdomen with the constriction between the first two tergites strong, the first tergite with a weak to strong apical trans- verse depression which is longitudinally striate ; propodeum with a moderately wide median groove which is irregularly margined by carinae 2 Abdomen with the constriction between the first two tergites weak, first tergite smooth, not depressed or striate apically ; propodeum with median groove absent or linear and not margined as above 3 . Legs beyond the coxae In-ight yellowish-brown ; basal segments of antennae suffused with light yellowish-brown ; notauli strongly diverging anteriorly (Florida and Arizona to Alberta, Ontario, and Massachusetts) canadensis (Provaneher) 1 The females of two species, apache and hrevicornis, are unknown. 1961 XEARCTIC SPECIES OF SIEROLOMORPHA 3 Legs brown except front tibiae and tarsi yellowish-brown; antennae brown, sometimes weakly suffused with paler brown on sides of basal segments ; mesoscutum rather flat, notauli weakly diverging anteriorly (California, Arizona, and Colorado to Saskatchewan and Yukon) nigrescens n. sp 3. Front, pronotum, and mesopleura wnth strong punctures pro- and mesonota bright rufocastaneous (Arizona) . . hicolor n. sp Front, pronotum, and mesopleura with only minute, widely spaced punctures; thorax entirely black (Georgia to Con- necticut and Kansas) similis n. sp. 3Iales 1. Abdomen with the constriction between the first two tergites strong, the first tergite with a weak to strong apical trans- verse depression which is longitudinally striate ; tyloides present on antennal segments seven through ten, short and rather prominently projecting (Figs. 1, 2) ; propodeum as in female (see couplet 1 of key to females) .... 2 Abdomen with the constriction between the first two tergites weak or absent, the first tergite smooth, not depressed or striate apically ; tyloides not present on antennal segment seven or, if so, very long and low on segments eight and nine 3 2. Tibiae and tarsi mostly or entirely light yellowish-brown in most specimens ; temples rather short, as seen from above much shorter than eyes (Florida and Arizona to Alberta, Ontario, and Massachusetts) . . . ca7ia densis CProyancher) Legs brown except front tibiae (sometimes also front tarsi and middle tibiae) light yellowish-brown; temples strongly developed, as seen from above nearly as wnde as the eyes (California, Arizona, and Colorado to Saskatchewan and Yukon) nigrescens n. sp. 3. Antennae extremely short and compact, segment eleven about 1.2 X as long as thick, segment thirteen about 1.5 X as long as thick; tyloides present on antennal segment eleven (Fig. 3) ; head extremely broad, about 1.25 X as wide as high (South Carolina) hrevicornis n. sp. Antennae elongate, segment eleven about twice as long as thick, segment thirteen much more than twice as long as thick ; tyloides not present beyond segment ten ; head 1.1-1.2 X as wide as high 4 4 BREVIORA No. 140 4. Ocelli distinctly enlarged (diameter of anterior ocelhis about .22 X minimum width of front) ; tyloides present on antennal segments seven through ten (Fig. 6) ; front with a strong median groove in front of anterior ocellus (Arizona) apache n. sp. Ocelli not enlarged (diameter of anterior ocellus less than .2 X minimum width of front) ; tyloides present on seg- ments eight through ten (Figs. 4, 5) ; front with a weak median impression if any 5 5. Punctures of front moderately strong ; tyloides rather short (Fig. 5) ; middle and hind tibiae and tarsi dull brown (Arizona) ? hicolor n. sp. Punctures of front very weak; tyloides elongate (Fig. 4) ; all tibiae and tarsi light yellowish-brown (Georgia to Connecticut and Kansas) similis n. sp. SiEROLOMORPiiA BicoLOR new species Holotype. — 9 , ARIZONA: Southwestern Research Station, 5 mi. W. of Portal, Cochise Co., 9 August 1959, 5400 feet eleva- tion (H. E.Evans) [MCZ]. Description of type female. — Length 5.8 mm., length of fore wing 4.3 mm. Head black ; pronotum, mesoscutum, and scutellum w^holly bright ruf o-castaneous ; remainder of thorax and pro- podeum black ; abdomen very dark brown, approaching black basally; mandibles light brown, darker basally and apieally; clypeus suffused with reddish-brown apieally; scape black, flag- ellum dark brown, outer side of apical segments paler ; legs dark brown except front and middle tibiae and all tarsi light brown ; fore wing lightly, uniformly infuscated, veins and stigma dark brown. First four antennal segments in a ratio of about 12:5:7:9, segment three 1.4 X as long as thick. Front strongly polished, punctures small but rather strong ; spacing of punctures rather irregular, those on the sides being mostly rather close, often not much more than their own diameters apart, those on the middle of the front rather sparse ; temples also with distinct punctures. Head subcircular in anterior view, vertex evenly rounded off a short distance above eye tops; inner orbits sub- parallel, minimum width of front about .9 the eye height. Ocelli small, in a broad triangle, the front angle greater than a right angle ; postocellar line slightly exceeding oeello-ocular line. Thoracic dorsum strongly polished, non-alutaceous ; pro- notum with strong, widely spaced punctures; mesoscutum with 1961 NEARCTIC SPECIES OF SIEROLOMORPHA 5 a very few punctures on the sides, impunctate medially. Pro- podeum strongly polished medio-basaliy, elsewhere slightly roughened by obscure punctures ; median line weakly impressed. Mesopleurum polished and with well-defined punctures. First abdominal tergite without a subapical depressed and striate baud, completel}^ smooth ; constriction between first and second tergites weak ; abdomen rather strongly hirsute beyond segment three. Male (assigned here tentatively). — ARIZONA: Cochise Stronghold, Dragoon Mts., 4850 feet elevation, oak-juniper zone, 2 July 1947 (Werner & Nutting) [Univ. Arizona]. Description of male. — Length 5 mm., length of fore wing 4.8 mm. Head and thorax entirelj^ black, abdomen dark brown ; apical half of mandibles light brown; antennae wholly dark brown; legs wholly dark brown except front tibiae and tarsi bright yellowish-brown; wings subhyaline. First four antennal segments in a ratio of about 20 :8 :13 :20, segment three 1.5 X as long as thick, segment four 2.2 X as long as thick, this segment typical of the remaining segments except the last ; segments eight through ten each with a short, rather weak longitudinal polished ridge (Fig. 5). Front polished, wholly covered with small but well-defined punctures which are separated by scarcely more than their own diameters ; vertex and temples more weakly punctate; front somewhat impressed along the inner orbits and beside the posterior ocelli, faintly impressed medially. Minimum width of front approximately equal to eye height; ocelli in a broad, flat triangle, postocellar and oeello-ocular lines subequal ; ocelli of moderate size, diameter of anterior ocellus .18 X mini- mum width of front. Pronotum short, shining and w^th dense, rather weak punctures. Mesonotura shining, rather sparsely and weakly punctate. Propodeum polished, impunctate, and non- alutaceous over most of its surface. Mesopleurum polished, densely but weakly punctate. First abdominal tergite smooth, without a subapical depressed and striate band. Remarks. — The type female was taken on the ground beneath oak trees in dry, open forest. The male associated with it tenta- tively was apparently taken in the same type of forest and at nearly the same altitude. SiEROLOMORPHA APACHE new spccies Holotype. — $ ARIZONA: 3-5 mi. SW of Apache, Cochise Co., 8 August 1959, about 4300 feet elevation (H. E. Evans, on Baccharis glutinosa) [MCZ]. 6 BREVIORA No. 140 Description of type male. — Length 4.5 mm., length of fore wing 4.0 mm. Entire body dark brown, the head nearly black ; mandibles light brown ; antennae uniformly light brown ; legs dark brown except tibiae and tarsi light brown, the front tibiae a rather bright yellowish-brown ; wings hyaline. First four an- tennal segments in a ratio of about 15:9:13:17, segment three 1.6 X as long as thick, segment four 2.0 X as long as thick, segment eleven 2.1 X as long as thick ; segments seven through ten each with a weak longitudinal polished ridge (Fig. 6). Front polished, with weak, scattered punctures, with a very strong median groove extending downward from the anterior ocellus ; eyes large and prominent, inner orbits subparallel below; head about 1.2 X as wide as high. Minimum width of front 1.07 X eye height ; ocelli large, diameter of anterior ocellus .22 X mini- mum width of front; postocellar line 1.15 X ocello-ocular line. Pronotum short, shining, weakly punctate. Mesonotum strongly shining, obscurely punctate. Propodeum in large part strongly polished, median area without ridges or other sculpturing except for a simple carina on the posterior third. Mesopleurum shining, obscurely punctate. Abdomen with scarcely any indication of a constriction lietween the first tAvo segments either dorsally or vcntrally; first tergite without an apical striate depression. Remarks. — • This striking specimen was taken on vegetation in tlie daytime, although the large ocelli suggest that the species may be nocturnal or crepuscular. Tlie locality was a dry wash in an area of desert grassland. SiEROLOMORPHA siMiLis new spccies Holotypc. — S . MARYLAND: Takoma Park, 22 Sept. 1945 (II. & M. Townes) [HKT]. Description of type male. — Length 4.4 mm., length of fore wing 4.1 mm. Body dark brown, head and thorax almost black; mandibles light brown ; antennae uniformly dark brown ; coxae dark brown, femora medium brown, paler apically, tibiae and tarsi yellowish-brown ; wings lightly tinged with fuscous. First four antennal segments in a ratio of about 19 :10 :13 :18, segment three almost twice as long as thick, segments four and eleven about 2.1 X as long as thick ; tyloides in the form of low carinae on segments eight through ten, the carina on eight extending for much of the length of the segment, the others slightly shorter (Fig. 4). Front strongly polished, with small, rather evenly distributed punctures which are separated by 1-2 X their own 1961 NEAKCTIC SPECIES OF SIEROLOMORPHA / diameters; median line of front weakly impressed. Head 1.12 X as wide as high, subcircular in anterior view, temples only moderately developed, contracted immediately behind eyes. Inner orbits convergent below, minimum width of front subequal to eye height ; ocelli of moderate size, diameter of anterior ocellus .17 X minimum width of front; postocellar line 1.1 X ocello-oeular line. Pronotum strongly shining, with a great many minute punctures. Mesoscutum strongly shining, obscurely punctate, notauli strong, nearly reaching anterior margin. Dorsal surface of propodeum shining, with a linear median groove. Mesopleurum strongly polished and nearly impunctate. First abdominal tergite polished, smooth, with no evidence of a transverse apical de- pression; second tergite with a narrow transverse basal impres- sion, so that there is vague evidence of a constriction between the first two tergites. Alloiype.— 2, Kearny, New Jersey, 9 Sept. 1935 (C. W. Funaro) [AMNH]. Description of allotype female. — Length 5.1 mm.; length of fore wing 3.6 mm. Head and thorax dark brown, abdomen medium brown ; mandibles and apical half of clypeus light yellowish-brown ; antennae dark brown except scape and pedicel suffused with yellowish-brown and flagellum light brown be- neatli ; coxae brownish but legs otherwise wholly bright yellow- ish-brown ; wings lightly tinged with fuscous. First four antennal segments in a ratio of about 23:10:11:15, segment three 1.5 X as long as thick, segment eleven 1.7 X as long as thick. Front strongly polished, punctures small and widely separated Head subcircular in anterior view, about as wide as high ; inner orbits subparallel, minimum width of front subequal to eye height. Ocelli small, in a broad triangle; postocellar line subequal to ocello-oeular line. Pronotum polished, with scattered small punctures. Mesonotum impunctate, notauli strong on posterior .8 of mesoscutum, diverging and attenuate anteriorly. Pro- podeum mostly smooth and shining, with a very thin median groove which posteriorly is paralleled by some irregular carinae. Mesopleurum convex, smooth and shining. Abdomen fusiform, depressed ; first tergite smooth and polished, with no evidence of a transverse apical impression; second tergite barely depressed basally. Paratypes. — CONNECTICUT: 1 9 , Bank of Conn. Kiver, East Hartford, 2 Sept. 1947 (H. E. Evans) [MCZ] : NEW YORK: 1 9 , Sea Cliff, Long Island [MCZ] ; MARYLAND: 3 $ $ , Takoma Park, same data as type [MCZ, HKT] : WEST 8 BREVIORA No. 140 VIRGINIA: 1 $, Shaver's Fork, Tucker Co., Oct. 1938 (G. E. Wallace) [Carnegie Mus.] ; SOUTH CAROLINA: 1 $, Green- ville, Oct. 1952 (L. & G. ToAvnes) [HKT] ; GEORGIA: 1 $, Macon, 1 Dee. 1923 (T. II. Ilubbell) [USNM] ; KANSAS: 3 $ $ , Manhattan, Sept., Nov. (D. A. Wilbur, T. F. Winburn) [Kansas State Univ.]. Variation. — The two female paratypes are very similar to the allotype. The Connecticut specimen has the basal two antennal segments bright amber-colored, contrasting strongly to the flagel- lum ; the Long Island specimen is without a head. The males show a small amount of size variation ; the smallest specimen (Manhattan, Kansas) has a fore w4ng measuring 3.6 mm., the largest (Takoma Park, Md.) has a fore wing measuring 4.4 mm. The Kansas specimens tend to haA^e the body (especially the abdomen) slightly paler in color, but otherwise little variation in color or body sculpture can be noted. In most specimens the postocellar line is subequal to the ocello-ocular line, and in one specimen it is somewhat shorter. SlEROLOMORPIIA BREVICORNIS UCW SpCCieS Holotype. — S , SOUTH CAROLINA: Greenville, 21 Sept. 1952 (L. & G. Townes) [HKT]. Description of type male. — Length 3.4 mm., length of fore wing 2.7 mm. Body dark brown, head almost black ; mandibles light brown on apical half; antennae dark brown, very slightly paler beneath ; legs dark brown except front tibiae and tarsi light yellowish-brown ; wings hyaline. First four antennal seg- ments in a ratio of about 12 :6 :7 :8, segment three 1.1 X as long as thick, apical segment unusually short and thick, about 1.5 X as long as thick; tyloides present on segments nine through twelve, but rather small (Fig. 3). Front strongly polished, punctures minute, shallow; median line strongly impressed in front of anterior ocellus. Head very broad, 1.25 X as Avide as high ; front broad, its minimum width .61 X width of head, 1.18 X height of eye ; ocelli small, diameter of anterior ocellus .14 X minimum Avidth of front; postocellar line very slightly greater than ocello-ocular line. Vertex forming a broad, even arc above the eye tops; temples moderately developed, in dorsal vicAV about two-thirds as wide as eye. Pro- and mesonota shining, obscurely punctate; notauli strong, complete, diverging ant rior- ly. Propodeum mostly smooth and polished, med'an line Aveakly grooved, ecarinate. Mesopleurum strongly shining, obscurely 1961 NEARCTIC SPECIES OF SIEROLOMORPHA 9 punctate. Venation differing from that of other species only in having the second recurrent and second transverse cubital veins very weakly indicated and the margin cell somewhat more rounded apicalh'. First two abdominal segments without a con- striction between them, first tergite smooth and polished, without a striate depression along its apical margin. SlEROLOMORPHA NIGRESCENS new SpecicS Holotype.— $, WASHINGTON: Olympia [USNM]. Description of type male. — Length 4.4 mm., length of fore wing 3.7 mm. Body dark brown, almost black; mandibles light brown ; antennae uniformly dark brown ; legs dark brown except front tibiae bright yellowish-brown ; wings lightly tinged with fuscous. First four antennal segments in a ratio of about 20:8:11:16, segment three only 1.3 X as long as thick, segment four about 1.7 X as long as thick; segments seven through ten each with a short but rather strong longitudinal ridge (Fig. 1). Front strongly polished, with minute punctures which are rather close together below, much more widely scattered above ; median line of front weakly impressed. Head about 1.15 X as wide as high, rather thick, seen from above with the temples nearly as thick as the eyes, the head across the temples nearly as wide as across the eyes. Inner orbits subparallel below; minimum width of front 1.13 X height of eye ; ocelli small, diameter of anterior ocellus .14 X minimum width of front; postocellar line subequal to ocello-ocular line. Pronotum strongly shining, obscurely punctate. Mesoscutum polished and nearly impunctate, with very strong notauli which diverge slightly anteriorly and do not quite reach the anterior margin. Propodeum with two median carinae between which it is somewhat grooved, disc otherwise with weak and irregular sculpturing, somewhat shining. Meso- pleurum shining and with very small punctures. Abdomen with a strong constriction between the first two segments ; first tergite depressed along the posterior margin, second tergite along its anterior mars-in, both depressions with longitudinal striations. Allotype.— 9, V/ASHINGTON: Seattle [USNM]. Description of allotype female. — Length 4.5 mm., length of fore wing 3 mm. Head and thorax dark brown, abdomen medium bro"v\Ti, somewhat darker apically; mandibles and apical half of clypeus light brov\-n ; antennae dark brown, suffused with lighter brown on the sides of the basal flagellar segments ; legs dark brown except front tibiae bright yellowish-brown ; wings lightly 10 BREVIORA No. 140 tinged with fuscous. First four antennal segments in a ratio of about 22 :9 :9 :13, segment three 1.2 X as long as thick. Front strongly polished, punctures minute and widely scattered. Head subcircular in anterior view, very slightly wider than high ; inner orbits subparallel, minimum width of front 1.16 X height of eye. Ocelli small, in a broad triangle; postocellar line 1.1 X ocello-ocular line. Pro- and mesonota polished, obscurely punc- tate ; notauli deeply impressed, diverging only slightly anterior- ly, terminating well short of anterior margin of mesoseutum. Propodeum and mesopleurum as described for male. Abdomen fusiform, shining ; first tergite with a transverse apical depres- sion which is strongly longitudinally striate. Paraiypcs. — WASHINGTON: 1 $ , same data as tvpe [USNM] ; 1 9 , Almota, 24 June 1911 [MCZ] ; IDAHO: l' $ , Harvard, 24 June 1935 (J. M. Beck) [USNM] ; 2 5 rtical diameter of tympanum, 3 mm. Tymi)anum on an outward sloping plane. Paratoids subtriangu- lar with indistinct borders, with their long axis oblicpie medio- laterally, with flat dorsal granules. One subgular vocal sac. Dorsum with flat granules. Belly with the larger granules on abdominal region. Limbs with dorsal conical granules. Without interdigital membrane in the hand ; 1st and 2nd fingers sub- equal ; subarticular tubercles double on fingers 2 and 3 ; two carpal tubercles, the inner one smaller, elli])tic and somewhat salient, the outer larger and rounded. Subarticular tubercle double on toe 4 ; interdigital membrane in the toes to near the tip of the digits, but in toe 4 basal and prolonged a.s a serrated cutaneous fringe; two metatarsal tubercles small but elongati\ the inner more salient ; tarsal fringe absent. Dorsal coloration light brown, with some darker spots, not well marked. Belly b'ght. Dimensions. Head and body 50 mm. Head length 12. .5 mm. Head width 17 mm. Head height 5.5 mm. Interorbital space 2.5 mm. Elbow to the third linger 21 mm. Femur 18 mm. Tibia 16 mm. Heel to the fourth t(^e 25.5 mm. Foot 17 nun. Diagnostic fcafurrs. Biifo niaHicornis's dilfers from all other Neotropical toads ; the shape of the head and cephalic crests recall certain Asiatic forms. In the Neotropical area the nearest species is Bufo int('rni''diHS Giinther, but this is distinguished by the well nmrked ])arietals and preurbitals, the interorbital space narrower than the upper eyelid, the distinct tympanum, and by the paratoids wlrch are not elli])tical and are separated from the eye by the orbitotympanic crests. The new species is not as close to Bufo rolliceps Wieginann and differs from that species in having less prominent cephalic crests, the interorbital space narrower, double subarticular tubercles on the foot and no lateral granules in a row continuing the paratoids posteriorly. Material studied. BMNH 1898, 3.10.1 (1 specimen), Mani- eore, Rio Madeira, Brasil, B. Pift'ard. Bufo spixulosus alttperuvianus subsp. nov. Type. AMNH 14418, adult female, Challapata, Department of Oruro, Bolivia. Description. Head very short and wide. Loreal region sloping 1961 NEW TOADS FROM SOUTH AMERICA 3 outAvard. Rostrum vertical. No cephalic crests, excepting maxil- laries; cantlius rostralis thick. Interorbital space granular. Tym- panum sloping out. Paratoids well marked and rounded, con- tinued laterally by large granules (each granule with many horny points). Two types of dorsal granules: the larger with one central horny point and many others around it, the smaller with only one horny point. Larger granules on abdominal region. First finger longer than second ; subarticular tubercles on the fingers, double or semidivided ; palmar outer tubercle larger and rounded, inner smaller and elongate. Interdigital membrane of the foot basal but prolonged as a fringe on the toes; subarticular tubercles on the toes, simple or sometimes double ; two metatarsal tubercles, the inner more salient ; a thick tarsal fringe. Dimensions. Head and body 80 mm. Head length 17 mm. Head width 30 mm. Head height 11 mm. Interorbital space 6 nun. Ujijier eyelid Avidth 6 mm. Eye 7 mm. Tympanum 3.5 mm. Paratoid 9 mm. by 8.5 mm. Elbow to the third finger 38 mm. Femur 36 mm. Til)ia 30 mm. Heel to the fourth toe 50 mm. Foot 35 mm. Parafype. AMNH 1-1417, Choro, Bolivia, adult female 82 mm. Dlstrihution. The two localities of the material studied, Chal- lapata and Choro, are in the Department of Oruro, Bolivia ; the type locality is at 3700 metres altitude. Diagnostic features. According to Vellard (1959), there are six subspecies of Bufo spinulosus -. B. s. spinulosus, B. s. are- quipensis, B. s. limensis, B. s. trifolium, B. s. flavopictus and B. s. orientalis. B. s. altipernvianiis adds a seventh. Bufo s. altiperuvianus differs from B. s. spinulosus, the Bol- ivian subspecies structurally and geographically closest (De- partment of La Paz) in having the head shorter, not so distinct from the body ; the loreal region sloping more laterally ; tym- painim larger; paratoids larger and more rounded. Capurro (1950: 11) has cited B. spinulosus from Tarapaca Province, Chile (west of Oruro Department, on the other side of the Cordillera Occidental), but specimens of this provenance that I examined at the Chicago Natural History Museum are dif- ferent from the form here described. Remarks. I name this subspecies after the old Spanish name, Alto Peru, of the region from which it derives. Material stueVed. AIMNLI 14418 (1 specimen) Challapata, Bo- livia. AMNH 14417 (1 specimen) Choro, Bolivia. 4 BREVIORA No. 141 BUPO QUECHUA sp. IIOV. Type. CM 4225, adult t'eniale, liK-aohaca, 2500 m., Department of Cochabamba, Bolivia. Description. Head triaiioular, widest at the angle of the mouth ; loreal region sloping outward. Maxillary border marked ; canthus rosti-alis thick; supraorbital crest absent; parietal crest visible ; orbitotympanie crest thick ; one rostral-internasal crest more or less marked. Tympanum not visil)le. Paratoids approximately elliptic, dorsally smooth. Body dorsally with sparse large granules, but with abundant small granules. One lateral row of granules, continuing the paratoids, each granule with a large central papilla and smaller papillae around it ; below the row, lateral granules of the same type. Belly with abundant conical and simple granules. Elongate limbs, with conical granules dorsally. First finger longer than second ; fingers free, borders with small conical granules ; subarticular tubercles generally simple, but double on the third finger ; outer palmar tubercle large and rounded, inner one smaller and elongate. Tarsal fringe absent ; two metatarsal tubercles elon- gate and approximately of the same size, the outer one more salient ; interdigital membrane near the toe tips, but on the fourth only a little more than half its length and prolonged as a cutaneous fringe ; subarticular tubercles on toes, small and simple. Dorsum light l)rown with three large darker triangular spots not well marked ; one interocular with base to the front and two others on the body with the base to the rear ; a vertebral light line divided the last two triangles. Limbs dorsally with transverse wide dark bands. Belly yellowish with dark spots shaped very irregularly. Dimensions. Head and body 50 mm. Head length 12 nnn. Head width 17 mm. Head height 7 mm. Eye 5 mm. Upper eyelid width 4.5 mm. Interorbital space 5 mm. Paratoid length 7.5 mm. Elbow to the third finger 22 mm. Femur 20 mm. Tibia 17 mm. Heel to the fourth toe 28 mm. Foot 20 mm. Parafupcs. CM 4223, 4224, 4226, lucachaca, Dept. Cocha- bamba, Bolivia. Head and body: 62 mm., 41.5 mm., 37 mm., respectively. Diagnostic features. Four other species related to Bufo que- chua have been previously described : B. ockendeni Boulenger, B. inca Stejneger, B. leptoscelis Boulenger and B. fissipes Bou- leiiqer. In the table below their diiferential characters and the altitude at which they are found are shown. 1961 NEW TOADS FROM SOUTH AMERICA 00 « -3 K5 5 o^-e ^^ ^«g Qj ^ W <1 S HO 2 o h «fc +^ 2--— - — ..._; Pq to -i-io^ iH-Scur' 3 60 be o :;3 SJ. ^ ^ -^ ^M ^ -M o '^ ^ '5 ■*! GQ ^ - ^ *^ bD o TS s cS a> t:! »3 « o '^ G > • fH > ft p 05 ,2 "3 =H a2 '« • ^ ^ tc O o s ^ cS r^ a 03 o 4^ O ft 'a3 o lO 71 Ti >■ ^ ^ a o L. f; ^ «l-l 5- O -5 o bJD 13 be o d o _ p s CS G o o 33 C3 c 'S: rt -4— .^ ft rt 1 i^ ^ ft « ^^ ^ *"* 1 ^ C^ o ''j c; Cw O H Ph H t— ( o oo ?D o cq o m f~i t-^ C>] C^l ^ -^ — "yn -*-)i t^ sa, S ^ t:: t: Q> a> ;z; K r2 ^ , , 3 a> rt S •^ P ^ rt 3 be -M be -^j l> < 6 BREVIORA No. 141 From this comparison of t-liaracters it may be deduced that Biifo Jissipes, described from Santo Domingo, Province of Cara- Laya, Peru, is the nearest relative of B. quechua. But B. fissipes differs in tlie absence of parietal crests and the rudimentary interdigital membrane. Comparing the altitudes of the five species, it is noticed that B. quechua is seen to live at the high- est altitudes; in ."ome localities, such as Yungas de Chapare, Department of Cochabamba, Bolivia, it coincides with B. odiendeni but this last lives also at lower altitudes. It is inter- esting to see that B. ockendeni is, in this group of five species, the one that lives at the lowest altitude and has the widest known distribution, from Central and S.E. Peru to the depart- ments of Cochabamba and Santa Cruz in Bolivia. The other three previously described species are restricted to S.E. Peru: B. in^a (departments of Ayacucho and Cusco), />. leptoscclis and B. fssipes (Department of Puno, but at different altitudes). B. quechua occurs in the Yungas of the Department of Cocha- bamba, Bolivia. Material studied (including comparative material). Bufo quechua: CM 4223-26 (4 specimens) Incachaca. Department of Cochabama. Bolivia, 2500 m.. J. Steinbaeh. USN]\I 118704 (1 specimen) Socotal, Yungas del Chapare, Department of Cochabamba, Bolivia, J. Steinbaeh, 11-1929. MZUM 89414 (1 specimen) Yungas del Chapare, Department of Cochabamba, Bolivia. MZUM 76075 (1 specimen) Yungas de Cochabamba, 2200 m., Department of Cochabaml^a, Bolivia. MZUM 68163 (3 specimens) Yungas de Cochabamba, 2200 m.. Department of Cochabamba, Bolivia. MZUM 68166 (1 specimen) Cochabamba Valley, 2600 m.. De- partment of Cochabaml)a, Bolivia. Bufo fissipes. AMNH 6105 (1 specimen) Juliaca, Peril, H. II. Keays. Chicago Natural History Museum 64879, 64850 (2 speci- mens) Puno, Peru. Bufo iuea. USNM 107648 (1 specimen) 1 mile above San Miijnel, Avacueho, Peru, 0. F. Cook. V-27-1915. MCZ 4758 (1 specimen) Idma. I^ubamba Valley, 6000 ft.. Peril, E. Heller, X-1915. B}(fo ockeudeni. ]\ICZ 15425 (1 specimen) Chaijuimayo, Peru. Chicago Natural History IMuseum 3581-82 (2 specimens) Chaqui- mayo, S.E. Peru, H. C. Watkins. 1961 NEV." TOADS FROM SOUTH AMERICA 7 AxMXll (mi-lT (16 specimens) Juliaca, Peru, H. H, Keays. MZUM 68153 (1 specimen) Yungas del Chapare, Department of Cochabamba, Bolivia. .MZTM 68152 (2 specimens) Tarata. 1900 m.. liolivia. C^l 3806b, 4515 (2 specimens) Cerru Husanu, west of Santa Cruz, 1400 m., J. Steinbaeh. Acknowledgments 1 must thank the Consejo Nacional de Investigaciones Cien- tificas y Tecnicas de la Kepul)lica Argentina for the fellowship given me for investigations on Neotropical amphibians ; I am grateful also to Dr. A. is. Komer, Director of the Museum of Comparative Zoologj' at Harvard I'niversity and to Dr. E. E. Williams. Curator of Reptiles and Amphibians of this Museum, for facilities aiforded me during 1959-60 ; to Dr. A, G. C. Grandi- son, Dr. D. Cochran, Mr. C. :^.I. Bogert, Dr. R. Inger, Mr. N. Richmond and C. Walker for sending me material from their respective museums; and to Dr. \\ Vanzolini for the photo- graphs illustrating this ]iaper. Bibliography Barbour, T. and G. K. Noble 1920. Amphibians and reptiles from southern Peru, collected by the Peruvian Expedition of 191-1-1915 under the auspices of Yale University and National Geographic Society. Proe. U. 8. Nat. Mus., 58: 609-620. BOULENGEB, G. A. 1902. Descriptions of new liatraeliians and reptiles from the Andes of Peru and Bolivia. Ann. Mag. Nat. Hist., (7; 10: 39-4-402. 1903. Descriptions of new batrachians in the British Museum. Ann. Mag. Nat. Hist., (7; 12: 552-557. 1912. Descriptions of new l)atrachians from the Andes of South America, preserved in the British Museum. Ann. Mag. Nat. Hist., (8) 10: 185-191. Captirro, L. F. 1950. Batracios de Tarapacii. Invest. Zool. Chilenas, 1(1): 9-12. GUNTHER, A. 1858. Catalogue of the Batrachia Salientia in the collection of the British Museum, London: i-xvi -f 1-160, pis. I-XII. 8 BREVIOBA No. 141 Mertens, E. 1952. Amphibieii und Eeptilien in Titschack, Beitr. zur Fauna Perus, 3: 257-266. Stejisteger, L. 1913. Results of the Yale Peruvian Expedition of 1911. Batrachians and reptiles. Proc. U. S. Nat. Mus., 45: 541-.547. Vellakd, J. 1959. Estudios sobre batracios andinos. V. El genero Bufo. Mem. Museo Hist. Nat. Javier Prado, 8: 3-48, pis. I-XIV. Plate 1. Bufo manicorensis, new species. Type: BMNH 1898.3.10.1. o p^ ia§;%^^^0 ill m- tf^m^ "^mm^ H. S^fiSistt 1i ^, Xi»:; Plate 3. Bufo quechua, new species. Type: CM 4225. BREVIORA Musemnii of Comparative Zoology Cambridge, Mass. Jl'xe 30, 1961 Number 142 AUSTRALIAN CARABID BEETLES VL THE TROPICAL AND SOME SUBTROPICAL SPECIES OF PAMB0RU8, MYSTROPOMVS, AND NURU8 By p. J. Darlington, Jr. This is the first of four papers describing new flightless tropical (and a few related subtropical) Australian rain forest Carabidae of zoogeographic importance. However, these special papers will be treated as parts VI to IX of my general series on Australian carabid beetles. Some of the species now de- scribed have been referred to (but not by name) in the preced- ing paper of the series, on transition of wet forest carabid faunas from New Guinea to Tasmania (1961b). My localities are listed, mapped, and briefly described in No. IV of this series (1961a). The holotypes of new species described from my ma- terial are placed, at least temporarily, in the Museum of Com- parative Zoology. Paratypes will be deposited with the Com- monwealth Scientific and Industrial Research Organization at Canberra (where they can be compared with the Sloane Collec- tion) and in most cases in the Queensland Museum too. In the descriptions, proportions are usually given as simple fractions (%, %, etc.) but are based on actual measurements made under the microscope. Pamborus Banninger (1940) has correctly distinguished the real species of Pamhorus known to him, after examining some of the older types that Sloane (1904) was unable to see. What I have to say now is mostly concerned with tropical forms unknown to Banninger. Three species-groups of Pamhorus occur in tropical North Queensland. Two of them consist of single, very distinct species, elegans SI. and punctntus n. sp., respectively. The third, which 2 BREVIORA No. 142 I call the tropicus group, consists of a series of slightly dif- ferentiated, allopatric forms extending from South Queensland to the base of the Cape York peninsula and probably derived from a common ancestor that dispersed rather recently. Charac- ters given in the following key to distinguish the four species ( ? or subspecies ) of the tropicus group should be supplemented by comparison of descriptions and of specimens if possible. Key to tropical species of Pamborus (with some subtropical species in parentheses) 1. Each elytron with 6 to 8 costae (which may be raised or nearly flat) separated by narrower crenulate intervals 2 — Elytron with 15 or more nearly equal costae 7 2. Tip of aedeagiis dentate near apex; neck constriction deep; prothorax subcordate (wet forests of New South Wales and South Queensland) (alternans) — Tip of aedeagus not dentate; neck constriction shallow; prothorax variable in shape 3 3. Form more convex; sides of ijrothorax not or slightly sinuate; 7th and 8th elytral costae usually strongly developed and not much in- terrupted (drier woodlands of New South Wales and South Queens- land) (viridis) — Lfess convex; sides of prothorax often more sinuate (tropicus group) 4 4. Elytra with costae relatively wide and weakly convex, 5th and 6th nearly as wide as 4tli on disc, 7th and 8th usually distinct, but variable 5 — Elytra with costae usually narrower and more convex, 7th and 8th more often interrupted or disintegrated 6 5. Intervals between costae slightly vdder and more strongly crenulate; larger (31-36 mm.) (South Queensland) {suhtro-picus) — • Intervals between costae narrower (1st reduced to a fine irregular impressed line) and less strongly crenulate; smaller (26-30 mm.) (Eungella Range) transitus 6. Shining (Mt. Spec to Atherton Tableland, etc.) tropicus — Duller (northern Atherton Tableland to Mossman-Daintree area) opacus 7. Pronotum with basal impressions, not punctate; elytral costae not much interrupted; elytral margins green (Herberton — ? to v. Cook- town) elegans — Pronotum without basal impressions, entire surface densely coarsely punctate; all elytral costae much interrupted; bluish black (Ather- ton Tableland, etc.) punctatus 1961 AUSTRALIAN CARABID BEETLES 3 Pa.MBORUS SUBTROPICUS 11. sp. Form of alternans but often broader, somewhat variable, sliglitly depressed; rather shining black, pronotum with mar- ginal channels (narrowly) and posterior impressions bluish or greenish, elytra with crenulate intervals and margins green. Read: neck constriction weak. P yoihorax obovii y^ (Mt. Jacob) or 1/^ (Kenilworth) wider than long at middle; base slightly wider than apex ; sides broadly rounded anteriorly, broadly but not strongly sinuate posteriorly ; posterior angle moderately produced backward; linear basal impressions uniting with mar- ginal channels in moderate impressions, with convex areas not or vaguely reaching posterior margin; each lateral margin with 1, 2, or 3 (number variable, sometimes asymmetrical) seta-bear- ing punctures near and before middle. Elytra with margins not serrate; each with 8 slightly elevated costae wide on disc, narrower externally and apically, separated by strongly crenu- late intervals of which the 1st is narrowest but plainly crenu- late; 7th and 8th costae usually distinct at least to near middle of length but not strongly raised, slightly interrupted, 8th some- times disintegrated. Aedeagus not dentate. Length 31-33 (36) ; width 12 (13) mm. (^ figures in parentheses show probable size of an individual of which I have only elytra). Holotype i (M. C. Z. Type No. 30,346) and 1 9 paratype from Mt. Jacob, c. 45 miles south of Gladstone, South Queens- land, March 1958; and 1 6 paratype from Kenilwortli. west of Blackall Range. South Queensland, May 1958 ; all taken by nivself in or on the borders of rain forest. 1 have also elvtra of this species from Mapleton, on the north end of the Blackall Range. See key for distinguishing characters of this species. PaMBORUS TRAXSITUS 11. sp. Form almost of rather broad alternans, slightly depressed; moderately shining black, marginal channels and basal impres- sions of pronotum without or with only slight metallic color, margins and crenulate intervals of elytra green or greenish. Head: neck constriction weak. Prothorax between % and % wider than long at middle ; Ijase slightly wider than apex ; sides broadly rounded anteriorly, sometimes vaguely angulate at middle, broadly but weakly sinuate posteriorly; posterior angles 4 BREVIORA No. 142 moderately produced backward; linear basal impressions unit- ing with marginal channels in moderate impressions, with con- vex areas sometimes reaching or nearly reaching posterior mar- gins; each lateral margin with 1 to 4 (number variable, often asymmetrical) seta-bearing punctures near and before middle. Elytra with margins not serrate ; each eh^tron with 8 slightly elevated costae very wide on disc, narrower laterally and apic- ally, 5th and 6th not or not much narrower than 4th on disc, 7th and 8th narrower, usually distinct but not strong, often somewhat interrupted but rarely disintegrated ; crenulate in- tervals very narrow on disc, less strongly crenulate than in related forms, interval between 1st and 2nd costae reduced to a fine line almost without crenulations anteriorly. Aedeagus not dentate. Length 26-30; width c. 10.5-11.5 mm. Holotype S (M. C. Z. Type No. 30,347) and 67 paratypes all from the Eungella Range, c. 40 miles west of Mackay, Queens- land, c. 2000-3000 ft. altitude, Nov. 1957, taken l)y my wife, my son, and myself, in rain forest. For characters and relationships of this geographically iso- lated form, see preceding discussion and key. Pambortts tropicus n. sp. Form almost of alternans, slightly depressed; rather shining black, marginal channels and posterior impressions of pronotum and margins and crenulate intervals of elytra green. Head : neck constriction weak. Prothorax appearing as long as wide but by measurement nearly % wader than long at middle ; base not or slightly wider than apex: sides broadly rounded an- teriorly, sometimes vaguely angulate at middle, broadly sinuate posteriorly; posterior angles projecting backward as usual in group ; basal impressions uniting with marginal channels in moderate impressions, sometimes with vague convex areas reach- ing or nearly reaching base ; each lateral margin with 1 to 4 seta-bearing punctures near and before middle (number vari- able, often asymmetrical). Elytra with margins not serrate; each elytron wnth 8 costae usually slightly narrower and more convex than in preceding forms especially externally, but 7th and 8th costae variable, sometimes distinct (but not strong), sometimes much interrupted or disintegrated; crenulate inter- vals narrow on disc, broader externally, strongly crenulate. Aedeagus not dentate. Length (types) 28-31 ; width 10.5-11.5 mm. 1961 AUSTRALIAN CARABID BEETLES 5 Holotype 6 (M. C. Z. Type No. 30,348) and 30 paratypes all from Mt. Spec plateau (Paluma Kange), about 40 miles north of Townsville, 2000-3000 ft. altitude, Nov. -Dec. 1957, taken by the Darlingtons, in or on the edges of rain forest. Additional speci- mens, not types: 5, Kirrama Range, 2000-3000 ft., Dec. 1957; 3, Millaa Milfaa, April 1932; 2, Longlands Gap, Sept. 1952 (J. G. Brooks) ; 6, mountains above (SW of) Atherton, Dec. 1957 and Feb. 1958 (this and the 2 preceding localities are on the south- central Atherton Tableland) ; 6, Mt. Bartle Frere, west slope, 2000-3500 and 3000-5000 ft., Dec. 1957; and 1, Davies Creek Road, northern Atherton Tableland, May 1958; all specimens except Brooks' collected by the Darlingtons, in rain forest. This species (or subspecies) is more like suhtropicus of South Queensland than like transitus of the Eungella Range. It dif- fers from suhtropicus in having elytral costae usually slightly narrower and more convex especially externally, with 7th and 8th costae more often interrupted or disintegrated. Pamborus opacus Gehin Sloane (1904, p. 702) and Biinninger (1940) have applied the name opacus to this species. It differs from tropicus in being obviously duller, and it is also slightly more slender, with slightly narrower and more convex elytral costae, and on the average it has more marginal pronotal punctures, although these vary in number and position. Although this species and tropicus are not very different structurally, I think they probably are real species, for they overlap geographically. Of opacus, I have 6 specimens from mountains north of Kairi, on the Atherton Tableland between the Atherton area and Davies Creek, taken in rain forest at close to 4000 ft. altitude, and 18 specimens from Mt. Lewis, near Mossman, taken at about 3000 ft. altitude in rain forest. My northernmost locality for tropicus is between these two places but at a somewhat lower altitude, near the southern end of the Davies Creek forestry road. Pamborus elegans Sloane Sloane (1915) described this species from "scrub" (rain forest) east of Herberton on the Atherton Tableland. It should occur in the rain forests on the mountains between Atherton and Herberton, but I did not find it. Banninger (1940) records 6 BREVIOEA No. 142 it from ''Mac Ivor Kiver, " which may be the Melvor Kiver north of Cooktown. Pamborus punctatus n. sp. Form as figured (Fig. 1) ; entirely dull bluish or purplish black. Head quadrate ; eyes small ; antennae short ; neck con- striction very deep. Proihorax % or more wider than long at middle, widest behind middle, narrowed in front and behind Fig. 1 . Pambokus punctatus n. sp. Fig. 2. NURUS REX n. sp. 1961 AUSTRALIAN CARABID BEETLES 7 but base about % wider than apex ; sides weakly rounded except strongly rounded or vaguely angulate behind middle, slightly sinuate near base; basal angles only slightly produced back- ward; side margins thickened but marginal channels and basal impressions obsolete ; dorsal surface almost evenly but not strongly convex, with middle line almost obsolete except an- teriorly, and transverse impressions obsolete ; whole surface coarsely, irregularly punctate with longitudinal, deep punctures. Elytra oval; humeri not serrate; each elytron with 15 distinct and 2 additional partial costae, all much interrupted, the outer- most reduced to tubercles. Abdomen extensively but not uni- formly punctate. Tip of last ventral segment subtruncate in male, broadly rounded in female. Aedeagus not dentate. Length 17-19.5; width 7.3-8.0 mm. Holotype ? (M. C. Z. Type No. 30,349) and 2 paratyi^es from mountains above (8W of) Atherton. 3000-4000 ft. alti- tude, Dec. 1957 and Feb. 1958; and additional paratypes as follows: 1, south of Ravenshoe, c. 3000 ft., Feb. 1958; 2, east side Mt. Bellenden Ker, 3000-4500 ft., Dec. 1957. All speci- mens taken by myself in rain forest. This species differs from all previously known Famborus in obliteration of all pronotal impressions and in heavy punc- tation of pronotum. The small size and numerous elytral costae suggest a distant relationship with P. guerini Gory of South Queensland etc., but (in addition to the other differ- ences) guerini has serrate humeri and pimctatus has not. In life this small Pamhorus strikingly resembles the heavily catenulate species of Notonomus (especially masculinus Darl. 1953) which occur in the same rain forest areas. I suspect this is a case of mimetic convergence. Mystropomus Two species of this genus occur in tropical eastern Austra lia. One, with alternate elytral intervals raised, occurs in rain forest on the Eungella Range. It seems to be at most a poorly defined subspecies of subcosfaiiis Chd. of South Queensland and New South Wales. I do not think it is worth naming. The other tropical species, with elytral intervals equal, is N. regu- laris Banninger (1940), which occurs probably throughout the main (base-of -peninsular) rain forest system of North Queens- land. I now have 113 specimens of it from localities north to 8 BREVIORA No. 142 Thonitou Peak aiiil south to the ]\It. Spec plateau. There is some geographical variation, but specimens from most localities can be referred to the typical subspecies. However, the form on the Mt. Spec plateau at the southern extreme of the species' range seoms worth distinguishing as follows. ^MysTKOPO-^IU? REOrLAUIS LAEVIS U. Subsp. Similar to large specimens of typical Mystropomus regularis Bann. but almost lacking the weak granular elytral costae of typical regularis. In the latter each elytron has 6 or 7 dis- tinguishable (though scarcely elevated) costae or stripes that are more shining than the intervening spaces. In the new sub- species only about the 4 inner stripes are indicated at all, and they are much less distinct than in the typical regidaris. The rest of the eh'tral surface, including the lateral and anterior declivities, is virtually undifferentiated, dull, and finely granu- lar. As compared with typical regularis, laevis also has slightly Avider and more reflexed prothoracic margins and a somewhat duller pronotum. Length c. 17; width c. 6.5 mm. Holotype S (M. C. Z. Type No. 30,350) and 1 i paratype both from ^It. Spec plateau (Paluma Range), c. 40 miles north of Townsville, North Queensland, 2000-3000 ft. altitude, Feb. 1958, taken by myself in or on the edge of rain forest. NURUS This is a group of very large, stout Pterostichini that I can- not separate from Trichosternus by any single constant char- acter except the relatively heavy build. Nurus atlas Cast, has each $ front tarsus slightly dilated, with only 2 segments squamulose, and most other Nurus have 6 tarsi narrow and without squamae, but both these conditions exist in certain Trichosternus. Tschitscherine (1902) was therefore wrong in using form of $ tarsi as a primary generic character and Sloane (1894) was riglit in not using it. The known species of Nurus form several groups that have been called subgenera and that are named in the following key, although 1 am doubtful of the value of these subgenera. Nurus sensu stricto includes about 4 species in northern New South Wales, south at least to near Ebor, and an additional 1961 AUSTRALIAN CARABID BEETLES 9 species on Mt. Tamborine in southeastern Queensland.' The majority of the species inhabit rain forest, but at least one ex- tends into savannah woodland. The two species of Pachymelas apparently inhabit savannah woodland and perhaps coastal habitats in tropical Queensland; they do not seem to enter rain forest. The two previously described species of Nuridius are localized in South Queensland, probably in rain forest, al- though their habitats are not specified. The three new species described below, one tentatively associated with Nuridnis and the others not assigned to subgenera, are all rain forest species, but they are not directly related among themselves and prob- ably represent three separate invasions of isolated rain forests by different stocks of Nurus. The genus is apparently absent in the main (base-of -peninsular) rain forests of North Queens- land. Key to subgenera and some species of Xurus 1. Mesosternum not setose {Nurus sensu stricto) c. 5 species — Mesosternum setose anteriorly 2 2. Humeri without teeth {Pachymelas) S species — Humeri with margin toothed or thickened (but sometimes only slightly so) 3 3. Elytra without basal margin {Nuridius) 4 — Elytra with basal margin 5 4. Margins of pronotum less strongly reflexed; size smaller (31 & 39 mm.) fortis, grandis — Margins of pronotum strongly reflexed, size larger (41-45 mm.) . . .rex 5. Wholly black; stout, prothorax transverse, not much narrowed behind nox — Greenish black ; more slender, prothorax more narrowed behind .... mediiis Nurus rex n. sp. Form as figured (Fig. 2) ; large; black, head and pronotum shining, elytra (except marginal intervals) dull, lower surface moderately shining. Head rather small (in genus), c. 2/3 width prothorax; mandibles long, straight basally, strongly 1 The type locality of N. imperialis (Slonno 18!>4) is given as "North Queensland," but the species really Inhabits South Queensland. I have Ke<>n it unlv from Mt. Tamborine south of Brisbane. The type locality of \. crami- formin (Sloane 1S99, p. .570) is piven as "Cairns." but the specimen was receive'()('ene (Vdar ("r(M'k beds of northeastern Colorado. Wood {up. cit., p. 25-1-255). in describin:enerically distinct from the species of Eiinij/s but am not prepai'cd to say whether or not it should be referred to Lcidyntys or to anotlun- genus.'" I'nder tiie circumstances, it seems propei- to place Eumns cxiguus in Scottinnis and to refer L(if}i/)i)iis rifiis to the svnonvmv of this species. This brings tlie numhci- of recognized species of Scottimns to three; semble more closely those of //. t( luiicr ps ((lalbreath. 194S). Tlie molars of both agree in having anterior and posterior cingula that rise rather steeply to join the pi-otostyle and (Mitostyh'. These cusps are closely appressed with no gap between them, a further point of resemblance to //. f< iiiiia ps. The median va]le\' is straight on both teeth, slanting somewhat ]iosteriorly, not sinuous as in II. (/rrf/m-iil and //. IkiIcIk )i. The ant(M'ior cingnlum is stronger than the jjosterior on both AC and M-. Diifciissioii : Although it is somewhal larger and shows a few minoi' ditt'erences, Heliscnintis svhhiikjt ri is extremely close to //. hiniictps. Roth sjx'cies lia\'e an nndivded internal eingulum 2 XaiiU'd for K. M. SdilaiUJcT in rccdLrnilinn (if his extensive work in the (ioslicn Hole ai-ea. 6 BBEVIORA No. 146 Mild i\ sti'ai . ^^^" ^^Mt.'B.F ^^^ \ 3 .K\^^/ Ravenshoe Known distribution of Leiiadira, eudoniic suljgt'nus llctadira, in North Queensland. The finely dotted line is the approximate eastern edge of high land (the Atherton Tableland etc.). No. 1, aurifer ; 2, alticola; 3, alternans ; ■i,f-> '3 a; a CO 10 oo 00 T— 1 a; ft ft 55. fl s TS c3 t» ^ fc •a "3 V 02 "u -^ g t-3 ft CS o -^3 (m* .r-( 3 T— 1 bi 13 3 ^tH o oo «c 3 o o • r-l a o I— 1 be 0) 3 bO a CO • r-l > 'El 00 00 ^ -2 a; 3 s^ o o3 tl 1—1 Q 4> II a; a 5 so 5^ c: c3 s la be CS 4. 2 00 3 .^ -^- b£ 1 c f-H 00 cS g s UJ 1 00 o o o > 0^ -♦- "ci CO ■2 C3 CO ^-1 a '-^ 3 ■73 1-^ CO CO o be £ O t-, ^ a. 1-5 ^ 00 3 'zj > CS ^ %4 «> «^ ^ ^ 13 3 ft ft OJ X CO J3 3 be Kl cS «-i CS 5*H o 00 p. +^ V5 c« !2; 00 to "be s CJ & s £- s 3 1 00 irt *r^ 1h be -3 CS CO -73 EC <4-l 5^ • i-i Id s 'S 'a be be 3 «<-i ^ 3 C^ W CO o -t^ 1 s +-' K cc O '3 !3 .s 01 g ft CO "3 be CO Q T-l 33 t3 .S «3 X bi c5 SO o 6 CO to 0" rH -4-' C» _g S 3 "ft 'E 8 BREVIORA No. 149 tion shows that this specimen lies within the 95 percent confidence limits for group 3. The 'primary' types of C. stygia are unknown but one of the "good specimens of C. stygia" figured by Jones and Woodward (1885, pi. 10, fig. 2; 1888b, pi. 12, fig. 2) BM 45154, is similarly within the 95 percent confidence limits for group 5. C. stygia is thus not separable from C. papilio on statistical or indeed any other present evidence, and the former junior synonym should be suppressed. The morphospecies C. papilio and C. stygia form the one biospecies (or transient spe- cies) C papilio. It is worth noting that the specimen of C pap- ilio figured by Jones and Woodward (1885, pi. 10, fig. 1; 1888b, pi. 12, fig. 1) BM 58669 is 'abnormal,' as can be seen from its position on Figure 1. No significant differences in P or Q can be detected in the stratigraphically separated material of the Ritchie and Jennings collections. There is a notable lack of clustering in terms of the size factor P corresponding to the mean size of successive moult stages or instars. Attempts to divide the Ritchie or Jennings specimens alone into instars were also unsuccessful. Intra-instar variation may obscure the limits of successive instars if sufficiently great, but not all crustaceans obey Brooks' Law (Needham, 1950, pp. 10-11), and C. papilio may be another exception. At least part of this difficulty is due to sampling and preservation. Thus small individuals are only rarely collected (see Fig. 1) and all large individuals found have been incomplete. Relatively gigantic specimens up to two feet in total length occur both in the Les- mahagow and liagsliaw Hills inliers, but they can only be recon- structed from fragments and hence do not appear on Figure 1. A further complicating feature of the few large specimens avail- able is their distinctive dendritic carapace ornament. It can be .ijrgued, however, that this ornament is characteristic of adult aistars (Rolfe, in press). Such giant individuals must have had younger growth stages coincident in size with the specimens of Figure 1, and no such dendritic ornament has been observed in that size range. It seems preferable to extend the name C. papilio to include these large individuals, at the risk of 'lumping,' until better sampling has been made. 1061 PRELIMINARY STUDY OF CERATIOCARIDIDS 9 SUMMARY Of ten species of Ceratiocaris recorded from the Lesmahagow inlier, only C. pnpilio and C. stygia are sufficiently well founded to demand preliminary investigation. Both Salter's diagnosis of these two species and Jones and Woodward's sul)sequent de- finitions are artificial. Analysis of 128 carapaces in museum col- lections shows that C. papilio is indistinguishable from G. stygia, and the latter should be suppressed as a junior synonym. Carapaces of C. papilio show isometric growth except in the smallest individuals. The material cannot be resolved into a series of distinct instars. EEFEEEXCES Only those references not listed by Van Rtraelen and Sehmitz, 1934, are given here. Xeedham, a. E. 1950. Growth and regeneration rates in relation to age in the Crus- tacea with special reference to the isopod, Asellxfi aqvai\c\ts (Linn.). Jour. Gerontology, 5: 5-16. Ritchie, A. 1960. A new interpretation of Janioufiiis lenroorTi White. Xature, 188: 647-649. EOLFE, W. D. I. (In press) Grosser morphology of the Scottish Silurian phylloearid crus- tacean, Ceratiocaris inipiUo Salter. Jour. Paleont. Simpson, G. G. 1961. Principles of animal taxonomy. Columltia Univ. Press, Xew York. SnrpsoN, G. G., A. Eoe and E. C. Lewontin 1960. Quantitative Zoology. Eevised ed. Hareourt, Brace and Co., Xew York. ST0RMER, L. 1935. Dictyocaris, Salter, a large crustacean from the Upper Silurian and Downtonian. Xorsk geol. tiddskr., 15: 267-298. Teissier, G. 1960. Eelative growth. In The Physiology of Crustacea, Waterman, T. H. ed., vol. 1, pp. 537-560. Academic Press, XeAv York and London. Van Straelen, V., and G. Schmitz 1934. Crustacea Phyllocarida (=:Archaeostraca). Fossilium Catalogus; pars 64, Berlin. BREVIORA Museum of Compsirative Zoology Cambridge, Mass. January 5, 1962 Number 15U THE GENUS BETHYLUS IN NORTH AMERICA (HYMENOPTERA: BETHYLIDAE) By Howard E. Evans Bethylid wasps are predominantly tropical and subtropical in distribution, with only a few species of diverse genera penetrat- ing temperate regions and virtually none entering arctic or sub- arctic regions. The sole exception to this statement, illogicallv. is the type genus of the family, Beiliylus Latreille. This rather liighly evolved genus is circumpolar in distribution. In North America, specimens have been taken close to the Arctic Circle, but none have been taken south of New York, Illinois, Colorado, and central California. In the Old World there are several species of northerly distribution and several others from the Mediterranean region. The geiuis is not known from the South- ern Hemisphere. In North America, four species have been described in the genus; these are: casianeus Kieffer, amocnus Fonts, hrachypit lus \Vhittaker. and flauicoynis Whittaker. A fifth species, decipiens Provancher, has recently been transferred to the genus by Krombein (1958, U.S. Dept. Agri., Monogr. no. 2, first suppl., p. 98). Examination of types in the U.S. National Museum re- veals that two additional species, Arysepyris californicus Brid- well and Perisetnus oregonensis Ashmead, properly belong in the genus. Since the latter species is the tj^pe of the genus Digoniozus Kieffer (1905, In Andre, Spec. Hymen. Eur. Alger., v. 9, p. 245), this name can be added to the synonymy of Bethylus. One of these seven names can be removed from further con- sideration here. I have recently had an opportunity to study the type and only known specimen of Bethylus castaneus Kieffer (1907, Berlin. Ent. Zeitschr., 51: 295). The wings of this speci- men are in poor condition, but enough remains to be sure that this species belongs not to Bethylus but to the related genus Goniozus (new combination). 2 BREVIORA No. 150 Thus there are six specific names aA'ailable for the North American Bethylns, three of them ncAvly assigned to the genus. The question naturally arises as to how much synonymy is in- volved and how many species, in fact, are there? The present paper is an attempt to answer that question. ANALYSIS OF THE PROBLEM Specimens of this genus are not common in collections, but by borrowing material from many sources I was able to obtain about 80 specimens. One's first impression, on scanning this material, is the remarkable uniformity of the specimens in size, color, and structure. The only notable color differences are sex- ual : the males have yellow mandibles and wholly yellow an- tennae, while the females have dark mandibles and the antennae more or less infuscated apically. There are no noticeable differ- ences in the structure of the mandibles and clypeus, in the sculp- turing of the head or thorax, or in the male genitalia. There is, however, one character which varies strikingly, and that is wing length. The wings vary all the way from small pads scarcely larger than the tegulae to wings of normal size. This is not unusual in the genus, as several brachypterous species have been described and the European fuscicornis is known to exhibit much variation in wing length. In the case of the North American Bethylus, it was of interest to know whether wing length varied in a continuous spectrum or whether there were certain wing-length types which might represent different species. Following 0. W. Richards (1939, Trans. R. Ent. Soc. London, 89: 185-344) in his revision of the British species, I first deter- mined the relative wing length of each specimen by dividing the length of the fore wing by the length of the hind tibia (which is much easier to measure accurately than total body length). I then plotted the number of individuals exhibiting a given relative wing length (Fig. 1). The males fell into two distinct groups. Those of the first group (Type A) might be termed subapterous, since the wings are exceedingly small, barely surpassing the anterior margin of the propodeum. Males of the second group (Type B) might be termed micropterous, since the wings are still very small, extending about to the beginning of the pro- podeal clecliAdty. When one plots the females on this same scale he obtains a somewhat different picture (lower half of Fig. 1). The subapterous forms (Type A) tend to have slightly longer wings, the micropterous forms (Type B) slightly shorter wings, 19()2 NOUTH AMERICAN [-.irrn VM'S CD -ia oS f-l M^ 1^ (> ' 0^ cpoo . OS in in 1 o» o lO'^lOOl-'^'^'-' ^ ^ Ji 0^ o^o\ ijhj if o> O<0 .— k5 Kj •^ ^?^ '_ 10 K) V- 1 Ov =^ sO-fi c i6 «) > -*— r— i oS -*^ -> i-i Op f^^ 0) •-I ,~^ ex ^ZO *r" — t ■-) if • in S o P o •—I .—1 +J ■_ r^ n ^^ u. ^ '^ -« r-t _J ^ +- "x,^^ _j c ' O « Jo c; X r-t r-i u. >■ O) r-l —1 _1 s O o ■ r* "^ -♦— I <^ t:! ■<■ ■«- fj =4-4 • rH o p^ 1 OS ^ «i s <— » ^^ "H »^ »-1 ""^ 5 ^ o o d o Z E be 4 BREVIORA No. 1 30 SO tliat the two curves overlap slightly. Furthermore, there is a class of individuals with wings of moderate length, reaching about to the posterior margin of the first abdominal tergite, which might be termed brachypterous (Type C), as well as a few in- dividuals with wings of normal length (macropterous. Type D). The fact that wing length varies discontinuously suggests the possibility that several species may be involved, each exhibiting a different winff length. Presumablv there would be four such species, with males of the two less common ones still to be dis- covered. However, in the absence of other characters one cannot rule out the possibility of polymorphism. I once again turned to Richards' study of the British species, and discovered that the most useful character for separating the three forms occurring in Britain is the ratio between the distance separating the hind ocelli and the distance separating the hind ocelli from the occiput. I determined this ratio for all speci- mens availalile to me but obtained a unimodal curve, with the mean 1.7, the range of variation from 1.3 to 2.2 (close to the range for the European juscicornis) . Thus these measurements failed to support the po.ssibility of more than one species. How- ever, in the course of making the measurements I found myself able to recognize "long-headed" and "short -headed" individ- uals. The difference was slight, but sufficient to induce me to measure the heads and determine the width/length ratio. In the case of the males I again obtained two separate curves (Fig. 2. top). For the females I obtained a bimodal curve (Fig. 2, bot- tom). It was at once apparent that all the subapterous individ- uals (Type A) were "long-headed" (left hand curves in Fig. 2), all the micropterous individuals (Type B) "short -headed" (right hand curves in Fig. 2). The brachypterous females (Type C) were all "short -headed", while the fully winged females (Type D) were of both types. Here was a suggestion that two species might be involved, with both species being polymorphic for wing length in the female sex. Upon sorting the specimens into two lots represent- ing probable species, several other differences previously over- looked or discounted in importance were discovered. The most important of these involved the sculpture of the propodeum, the shape of the male subgenital plate (Fig. 3), and the wing vena- tion of the few available fully winged females. Thus I am now convinced that two polymorphic species are involved. The name (innit lilts Fonts is applicable to the "long-headed" species, while 1962 NORTH AMERICAN BETHYLUS dccipiens (Provancher) is the earliest name for the "short- headed" species. The two species are widely sympatric east of the Rockies, but amoenus is not known to occur west of the Rockies. The characters separating the two species are summar- ized below, as are their synonymy and distribution. N 0. of Males IZ 11 10 9 a 7 5 4 3 Z 1 .76 79 80 81 .62 &h .54 .85 .£6 .6 7 .es .89 wh/lh No. of Females .78 19 .60 62 &Z .63 64 65 .66 .87 .68 .39 .90 .91 .92 wh/lh Fig. 2. Number.s of individuals (ordinate) exhibiting given relative head lengths (width of head [WH] divided by length of head [LH]), males at top, females at bottom. The only good series of decipicns from one locality is the series of 16 females and 4 males from Chilliwack, British Colum- bia, on which Whittaker based his descriptions of hraehypterus and ftavicornis. As discussed further below, I have studied or oi)tained the necessary information on this entire series. Twelve of the females are micropterous (Type B), three are brachyp- terous (Type C), and one is maeropterous (Type D). This 12 :3 :1 ratio, obtained in a series from one locality, is approached rather closely by the ratio for the species throughout its range. 6 BREVIORA No. 150 which is 41 :9 :4. Unfortunately, only one reared series of this species is available. That is a series of five females and three males in the U.S. National Museum reared from Vicia angusti- folia at the Lummi Indian Reservation, Washington. All indi- viduals in this series are micropterous. Unfortunately, no good series of am.oenus is available; the longest series consists of three females and a male taken on different dates at Bar Harbor, Maine. All of these individuals are subapterous, and in fact only one fully winged individual of this species is known. This is a female taken by 0. W. Richards on the window of an automobile at Buffalo, N. Y., 19 Sept. 1928. The ratio of subaptery : braehyptery : macropter^' in the females of this species is 23:0:1. It is, of course, entirely possible that brachypterous individuals of amoenus may some day be discov- ered. It is also quite possible that polymorphism for wing length may occur in the male sex. At present only seven males of ainocnus are known, only thirteen of dccipiens. Clearly any hy- potheses on the genetics of polymorphism in these wasps will have to await the day when much more material has accumulated in museums. At present it appears that only the females are polymorphic for wing length and that the polymorphism arises from a very simple genetic mechanism. The Nearctic dccipiciiH is undoubtedly closely related to the Palaearctic fuscicornis and may well be derived from it. Not only are the oeellar measurements similar, as noted earlier, but the sculpturing of the propodeum is similar and the male sub- genital plate virtually identical. However, there is no doubt in my mind that they are specifically distinct. The antennae of fusci- cornis are shorter and the scape is black at the base, yellowish apically (the scape is wholly yellowish-brown in both Nearctic species). Richards has found that the frequency distribution of relative wing length in fuscicorni.'i is more or less trimodal or ((uadrimodal, but less distinctly so than in dccipiens and with a much larger proportion of longer-winged individuals. Further- more, the males of fuscicornis are typically macropterous rather than micropterous as in dccipiens. TAXONOMIC TREATMENT Key to North American Species of Bethylus Propodeum with a median polished ridge, remainder of disc con- trastingly alutaceous; head rather short (width/length ratio 1962 NORTH AMERICAN BETHYLUS 7 .85-. 91 ill female, .89-. 92 in male) ; wings of micropterous in- dividuals reaching at least nearly to middle of propodeal disc (relative wing length .57-.91) ; fully winged individuals with radial vein curved upward sharply apically, vein arising from basal vein barely indicated ; fore tibiae clear yellow ; male subgenital plate strongly emarginate, but the side-pieces relatively broad and blunt (Fig. 3, a) . .decipiens (Provancher) Propodeum somewhat convex dorsally but without a median ridge which is set off from the remainder of the disc ; head slightly longer (width/length ratio .79-. 84 in female, .85-.86 in male) ; wings of most individuals extremely small, reaching barely beyond anterior margin of propodeum (relative wing lenglli .26-. 54) ; fully winged individuals with radial vein not curved upvrard sharply at apex, vein arising from basal vein nearly as long as transverse median vein; fore tibiae of female usually at least weakly sutfused with brownish ; male subgenital plate with a strong emargination, the side-pieces reduced to slender, acuminate processes (Fig. 3, b) amoenus Fonts Fig. 3. Subgenital plates of (a) Bethylus decipiens and (h) BetJiylus amoenus. Bethylus decipiens (Provancher) Gonatopus decipiens Provancher, 1887, Add. Corr. Faune Eiit. Canada, Hj-men., p. 179 [Type: 9, Cap liouge, Quebec (Que. Prov. Mus., yellow label no. 1382) ] . — Muesebeek and Walkley, 1951, U. S. Dept. Agri. Monogr. 2, p. 1038. rtrL^icniHs oregonerisis Ashmead, 1893, Bull. V. S. Nat. Mus., 45: 70 [Type: 5, Portland, Oregon (U. S. Nat. Mus. no. 40422)]. New synonymy. 8 BREVIORA No. 150 Digoniozus ore(joncnsis Kieffer, 1905, Spec. Hymen. Eur. Alger., 9: 245 [Made type of new genus Digoniozus]. — Muesebeek and Walkley, 1951, U. S. Dept. Agri. Monogr. 2, p. 732. Arysepyris californicns Bridwell, 1919, Proc. Hawaiian Ent. Soc, 4: 34 [Type: 9, Parkside, San Francisco Co., Calif. (U. S. Nat. Mus. no. 64124)]. New synonymy. Bethyltis bracliypterus Whittaker, 1929, Trans. R. Ent. Soc. Lon- don, 76 : 385 [Type : 2 (not 6 as stated), Chilliwack, Br. Col. (British Museum) ]. — Muesebeck and Walkley, 1951, U. S. Dept. Agri. Monogr. 2, p. 732. New synonymy. Bethylns fiavicornis Whittaker, 1929, Trans. R. Ent. Soc. London, 76: 386 [Type: 6% Chilliwack, Br. Col. (British Museum)]. — Muesebeck and Walkley, 1951, U. S. Dept. Agri. Monogr. 2, p. 732. New synonymy. Glenosema calif ornicus Muesebeck and Walkley, 1951, U. S. Dept. Agri. Monogr. 2, p. 727. Bcthylus decipiens Krombein, 1958, U. S. Dept. Agri. Monogr. 2, First Suppl., p. 98. Remarks on types. — Provancher's decipiens was transferred to Bethylns by Krombein upon his examination of the type. Dr. Krombein has kindly placed his notes at my disposal, and they leave no doubt that Provancher's name applies to this species. The propodeum is alutaceous but with a median polished ridge, and the wings extend almost to the posterior slope of the propod- eum. The type is in good condition. The type of Ashmead's oregonensis is also in good condition and is a fully winged female of this species. The type of Brid- welLs calif ornicus is unfortunately in poor condition, the head, abdomen, and legs all being missing. However, the wings and propodeum are typical of the micropterous form of decipiens. A topotypic female in the collection of the California Academy of Sciences is very similar to the type and is in good condition. Whittaker 's two names require special discussion. The types and most of the paratypes are in the British Museum and I have not seen them. However, Mr. G. E. J. Nixon has been good enough to examine tliese specimens and send me the critical in- formation on them. I have studied one paratype of hrachypterus in the collection of Cornell University as well as two of this species and one of fiavicornis in the collection of Roliert M. Fonts of Laredo, Texas. The characters AVhittaker used for separating the two species are color characters which happen to be those which separate the sexes, and it happens that all the specimens 1962 NORTH AMERICAN BETHYLUS 9 of braclnjpfo'iis are females and all of Jlavicornis are males — Whittaker's statements to the contrary notwithstanding. The entire series is from Chilliwack, British Columbia; the type of hrachypterus is a fully winged female, that of flavicornis a mieropterous male. As indicated earlier, three of the paratypes of brack ijpte)-us are lirachyjiterous, the remaining twelve mierop- terous. Specimens cxantincd. — 40 9 9, 10 £ i. ALASKA: 1 9, Fair- banks, 25 June ]948 [I^SXMJ ; 1 9, Circle, 2 July 1958 (C. Lindroth) [CXC] ; 1 9, Xenana, 17 June 195:3 (K. I. Sailer) [USXM] ; 1 9, 2 <3 6, Mile 1476, Alaska Highway (C. Lindroth) [CXC]. BRITISH COLUMBIA: 1 9, Mile 290, Alaska High- way, 19 June 1951 (W. Mason) [CXC] ; 1 9, Smithers, 12 June 1958 (C. Lindroth) [CXC]; 1 9, Cranbrook, 12 May 1922 (C. Garrett) [CXC]; 1 9, Victoria, 28 Aug. 1923 (K. F. Auden) [CXC] ; 3 9 9,1 S, Chilliwack, May-June, Sept. 1927 (0. Whit- taker) [CU, Coll. R. M. Fonts]; 1 S, Galiano, 2 Aug. 1929 [Coll. Fonts]; 1 9, Kaslo (A. X. Caudell) [USXM]; 2 9 9, Terrace [MCZ]. WASHIXGTOX : 1 9, Olympia [USXM]; 5 9 9,4 S S, Red River Rd., Lummi Ind. Res., 1 Aug. 1944 (Vicia angusHfolia, W. W. Baker) [USXM]. OREGOX : 1 9, Portland [USXM] ; 1 9, Forest Grove, 1 Apr. 1919 (A. C. Bur- rill) [USXM] ; 1 9, Ashland Loop, Siskiyou Mts., Jackson Co., 6 Aug. 1950 (Malkin & Thatcher) [CAS]. CALIFORXIA : 1 9, Land's End, San Francisco, 11 July 1922 (F. X. Williams) [CAS] ; 1 9, Parkside, San Francisco Co., 8 Sept. 1910 (J. C. Bridwell) [USXTVE]. UTAH : 1 9, Logan [MCZ]. COLORADO: 2 9 9, 1 $, Fort Collins, June, Sept. 1895 (C. F. Baker) [USXM]. IDAHO: 1 9, Coeur d^Alene (H. J. Rust) [USXM]. ALBERTA: 1 9, Edmonton, June 1917 [USXM]; 1 $. Elk- water Lake, 19 July 1956 (0. Peck) [CXC]. OXTARIO : 1 9, Sudbury, 1892 [CXC]. QUEBEC: 1 9 , Anticosti Island, 9 Sept. [MCZ]. XEW BRUXSWICK: 1 9, Penobsquis, Dec. 1927 (C. A. Frost) [MCZ]. XOVA SCOTIA: 1 9, Portapique, 23 July 1929 (C. A. Frost) [MCZ]. MAIXE: 4 9 9, Bar Harbor, July- Oct. (A. E. Brower) [USXM]. XEW YORK: 1 9, Grand Island, 11 Oct 1922 [USXM]; 1 9, Xorth Fairhaven, 1 Sept. 1918 [CU].i 1 The following abbreviations have been employed for the museums involved : CAS, California Academy of Sciences. San Francisco ; CXC. Canadian National Collections, Ottawa : CU, Cornell University. Ithaca : MCZ, Museum of Compara- tive Zoology, Cambridge; USNM, U. S. Nati()nal Museum, Washingttm. 10 BREVIORA No. 150 Map showing distribution of North American Bethyhis. Solid triangles : Bethylus amoenus, subapterous form ; hollow triangle : macropterous form of amoenus. Solid circles: B. decipicns, mieropterous form; half -solid circles: braehj'pterous form of decipiens ; hollow circles: macropterous form of decipiens. Bethylus amoenus Fonts Bethylus amoenus Fonts, 1928, Proc. Ent. Soc. Wash., 80: 127 [Type: 9, Slaterville-Caroline, Tompkins Co., N.Y., 14 Jniie 1904 (Cornell Univ. no. 934) ]. — Mnesebeck and Walkley, 1951, U. S. Dept. Agri. Monogr. 2, p. 732. Remarks on types. — The type is in good condition. I have also studied a male allotype, bearing the same data, in the collec- tion of Robert M. Fonts. Specimens exammed. — 24. 5 5, 7 $$. NORTHWEST TER- RITORIES: 2 9 9, Norman Wells, 3-13 July 1949 (W. Mason) [CNC]. ALBERTA: 1 9 , Aspen Beach, 23 Aug. 1944 (0. Peck) [CNC]; 1 9, Elkwater Lake, 19 July 1956 (0. Peck) [CNC]. SASKATCHEWAN: 1 S, White Fox, 10 July 1944 (0. Peck) [CNC] ; 1 c^, Holdfast, June 1946 (W. A. Nelson) [CNC] ; 1 9, 1902 XORTH AMERICAN BETHYLUS 11 Assiniboia, June 1955 (J. R. Vockeroth) [CNC] ; 1 i, Saska- toon, 15 Sept. 1924 (K. M. King) [CNC]. MINNESOTA: 1 ?, Eaglesnest, 26 An^. 1959 (W. V. Baldnf) [TTSNM]. WISCON- SIN: 1 9, Cranmoor, 20 May 1910 (C. AV. Hooker) [USNM]. ILLINOIS: 1 2, Palos Park, 17 March 1933 (Frison & Mohr) [111. Nat. Hist. Survey]. MICHIGAN: 1 9, AVexford Co., 4 July 1952 (R. R. Dreisbach) [Coll. Dreisbach] ; 1 9, Presque Isle Co., 28 July 1952 (P. B. Kannowski) [Coll. Dreisbach] ; 1 9, Midland Co., 20 June 1945 (R. R. Dreisbach) [Coll. Dreisbach]. ONTARIO: 1 9, Jordan, 25 Sept. 1916 (W. A. Ross) [CNC] : 1 9, Rondeau Park, Kent Co., 28 June 1936 (G. Steyskal) [Coll. Dreisbach] ; 1 9, Prince Edward Co., 10 July 1950 (J. F. Brim- ley) [CNC] ; 1 9, Belleville, 2 Oct. 1956 (J. M. Smith) [CNC]. NEW YORK: 1 9, Buffalo, 19 Aug. 1928 (0. AV. Richards) [MCZ]; 2 9 9, Ithaca, 28 May, 23 June (Babiy, Evans) [CU, MCZ] ; 1 9,1 $, Slaterville-Caroline, 14 June 1904 [CI^ Coll. R. M. Fonts] ; 1 <5, Caroline-Harford, Tompkins Co., 15 June 1904 [CU]; 1 S, Gannett Hill, 30 Aug. 1925, 2000 feet [CU]. MAINE: 3 9 9, 1 5 , Bar Harbor, July, Sept., Oct. (A. E. Brower) [USNM] ; 1 9, Southwest Harbor, 6 Sept. 1922 [CU]. NOA'A SCOTIA: 1 9, Portapique, 22 July 1929 (C. A. Frost) [MCZ]. BIOLOGY OF THE GENUS The only specimen of this genus which I have collected was taken walking over the ground in a small sand pit. Several speci- mens in collections are labeled as having been taken sweeping, one while "sweeping Carex," another "while beating for ants." Several specimens of both species were taken by A. E. Brower at Bar Harbor, Maine, on "Great Heath," one of the female decipiens "on flowers of Ilex verticillata.'" A female amoenus from Palos Park, Illinois, is labeled "in wet peat sample," while a series of decipiens from the Lummi Indian Reservation, AYashington, is labeled "Rd Vicia angusti folia." Apparently these insects occur in a variety of situations. B. amoenus has been collected in every month from March to October, decipiens from April to October and also in December. Two specimens of amoenus bear host data. One is the female listed above from Cranmoor, AVisconsin, which is indicated as a probable parasite of Eudemis vocciniana. This name is now re- garded as a synonym of Rhopohata naevana (Hbn.), an oleu- threutid moth known as the black-headed tireworm. The other 12 BREVIORA No. 150 specimen is the female listed from Belleville, Ontario, which is labeled as a parasite of Brachypterohis pulicarius L. This is a nitidulid beetle introduced from Europe to the United States about 1918. If the latter record is correct, it is the only kno^\^l instance of a Bcthylus attacking a beetle. The European cephalofcs Forster and fuscicornis (Jurine) attack various caterpillars, chiefly Microlepidoptera but occasionally Noctuidae. The fe- male wasps sting and malaxate their rather large prey, then drag it to a place of concealment, such as a hollow^ stem. Several eggs are laid on the prey and several larvae develop on a single host. Further details regarding the biology of these two species may be found in the papers of Richards (1932, Trans. Ent. Soc. So. England, 8 : 35-40 ; 1939, Trans. R. Ent. Soc. London, 89 : 185- 344). BREVIORA Miiseiuim of Compsirative Zoology Cambridge, Mass. January 12, 1962 Xr. miser 151 A NEW PHYLLOCARID CRUSTACEAN FROM THE UPPER DEVONIAN OF OHIO By W. D. Ian Rolfe INTRODUCTION Wlieii c'lirating- the collections of non-trilobite arthropods in the Musenni of Comparative Zoology, the writer recently found a fossil crustacean which had been sent to Professor P. E. Ray- mond for determination. The specimen was received from the Cleveland Museum of Natural History through Dr. D. H. Dunkle in January 1944 ; the late Professor Raymond published no description of the specimen and left no manuscript notes with it. Recent collecting in the same area by Mr. G. Lammers of the Cleveland Museum failed to find further specimens, although a fragment of a second specimen from a different locality was recently donated to the Museum of Comparative Zoology by Mr. R. Pritschan of Cleveland. This specimen will be referred to as the MCZ specimen to distinguish it from the original Cleveland ^Museum specimen. During the preparation of this description Mr. Lammers called the writer's attention to the fact that H. K. Brooks of the Uni- versity of Florida had collected and studied the echinocaridids of this region. In correspondence, Mr. Brooks informed the writer that he had photographed this specimen some years ago, but was kind enough to allow the writer to submit this account for publication. The writer is indebted to Dr. G. A. Cooper, Professor A. La Rocciue and Professor F. G. Stehli for searching through the collections at the U. S. National Museum, the Ohio State Univer- sity and Western Reserve University, Cleveland, for additional material, and to 'Mv. W. E. Scheele. Director of the Cleveland Museum of Natural History, for allowing the specimen to be retained for description. Professor H. B. Whittington kindly took the photograph for Plate 1 and offered helpful criticism of the manuscript. 2 BREVIORA No. 151 SYSTEMATIC DESCRIPTION Subclass MALACOSTRACA Latreille, 1806 Siiperorder PHYLLOCARIDA Packard, 1879 Order ARCHAEOSTRACA Claiis, 1888 Suborder RHINOCARINA Clarke in Zittel-Eastman, 1900 Family OHIOCARIDIDAE fani. nov. Diagnosis. Carapace valves deep, with anterodorsal-medio- ventral fold and broad median dorsal plate. Rostral plate and number of thoracic and abdominal segments unknown. Remarks. The family Rhinocarididae comprises five genera which form a compact group characterised by elongate carapace valves and a narrow median dorsal })late. The present genus is so distinct from the previously described Rhinocarina as to Avar- rant the erection of a second family. Genus OhiOCAKIS gen. nov. Type species. Ohiocaris wycoffi sp. nov. Diagnosis. As for the family. Ohiocaris wycoffi sp. nov. Plate 1 ; Figure 1 Descripfion. The Cleveland specimen is exposed with the dorsal surface of the carapace uppermost in one half of a concretion. It is preserved as a very thin film of golden brown fcuticular material, but this has been destroyed over much of the .specimen so that an internal mould is revealed. The concretion has been split apart so that the two carapace valves and median dorsal plate are separated by matrix from two complete abdominal seg- ments, a fragment of a third, and stylet fragments. The where- abouts of the counterpart of the concretion are unknown. As may be seen from Figure 1 and Plate 1, the carapace valves are deep ; a well-defined fold, semicircular in cross section, runs from the anterodorsal region of each carapace valve, immedi- ately posterior of the strong carapace horn, and dies out ventrad of the centre of the valve. The ventral or free margin of each valve is bordered by a narrow reflexed rim except in the mid- ventral region, where it continues as a marginal ridge inside the ventral margin. 1962 NEW DEVONIAN CRUSTACEAN FROM OHIO The broad median dorsal plate is separated anteriorly from the cephalic region of the carapace by a shallow transverse groove. The grooves laterally separating the plate from the main area of the carapace valves are narrower and deeper than the anterior groove, and are confluent posteriorly with the carapace mandibles anterior carapace horn oblique fold left valve of carapace pereiopods p osier o median groove marginal ridge m e dian dors a I plate last abdominal segment (? = 7th) ventral process of style head stylet Figure 1. Outline drawing of Ohiocaris u-ycoffi gen. et sp. nov. showing structures visible on Plate 1. Cleveland Museum of Natural History 33241. X 1.1. rim and a groove marking the posterior edge of the median dorsal plate. These grooves doubtless mark the position of marginal rims analogous in structure to those at the ventral edge of the carapace valves. The plate bears a faint posteromedian groove which extends anteriorly for 1.4 mm. The carapace valves and median dorsal plate are crazed by 4 BREVIORA No. 151 veinlets of a brown mineral ( ?collophane), whereas the abdomi- nal segments and the matrix are unaffected, indicating differen- tial chemical desiccation. The fragments of test preserved are smooth and free from ornament and the few wrinkles present are clearly secondary. The abdominal segments are inverted relative to the carapace and thus the style and stylets are exposed ventral side upper- most. The last segment ( ?7th) is 1.3 times the length of the preceding segment and has a concave posteroventral margin. After the photograph for Plate 1 was taken, the fragmentary st^'le and stylets were broken from the matrix. The dorsal head of the style thus exposed was found to be of the echinocaridid type illustrated by Echinocaris snblevis Whitfield, 1880, figure 6 (=Hall and Clarke, 1888, pi. 29, fig. 13). Only the bases of style and stylets are preserved but the former is triangular in cross section and much shorter tlian the stylets. As in all the known archaeostracans, the head of tlie style embraces the proxi- mal portions of the stylets laterally and ventrally. A small denticle, 0.2 mm. long by 0.4 mm. broad, projects posterolaterally from the right lateral edge of the ventral style process or plat- form. The cuticle of the style head, stylets and abdominal segments lacks ornament. Only the inflated coxal parts of the mandibles are j)reserved. and excavation has failed to reveal the toothed gnathal lobes which were probably broken off at burial. The mandibles have been impressed tli rough the anterodorsal region of the carapace valves, and the plane of section show^s the left mandible to have had a wall thickness of 0.7 mm. At least four recurved ridges on the anteroventral region of the carapace fairly certainly mark the position of simple pereio- pods. As they are only seen as impressions through the thin cuti- cle of the carapace no detail of their structure can be discerned. The MOZ specimen is a fragment sliowing the median dorsal plate only. Dimensions, in millimeters Oleveland ^luseuni of Natural History 33241 Maximum length of undistorted right carapace valve 34.0 Maximum height of undistorted right carapace valve, to right edge of groove bordering median dorsal plate 25.0 1962 NEW DEVONIAN CRUSTACEAN FROM OHIO 5 Length of median dorsal i)late, along mid-line 21.0 Maximum width of median dorsal plate, at a point 7 ram. posterior from transverse groove 7.3 Length of penultimate abdominal segment 4.3 Length of last abdominal segment 5.5 Maximum dorsal width of style head 4.5 Length of style head to base of style 3.3 Width of style at base 1.1 Cross-sectional diameter of stylet 1.6 Museum of Comparative Zoology 6556 Maxinnim length of median dorsal plate ca. 21 Maximum width of median dorsal plate 7.2 Holotype. Cleveland Museum of Natural History 33241. Col- lected by Dale Wycoff, 25th May, 1934, from the Chagrin Shale, Upper Devonian. Locality — shore of Lake Erie at mouth of Porter Creek, 12 miles west of Cleveland, Cuyahoga County, Ohio. Other material. Museum of Comparative Zoology 6556. Col- lected by Raymond Pritschan and donated to the Museum via G. Lammers, July 31, 1961 ; found as float from Chagrin Shale. Locality — Painesville, 25 miles northeast of Cleveland, Lake County, Ohio ( ? Whitfleld's 1880, p. 37, Leroy locality). Remarks. The spread out carapace valves of the Cleveland specimen recall the condition in Dithyrocaris, which possibly lived with the valves in this attitude. The marginal rim fore- shadows the well-developed submarginal wall and doublure struc- ture of Dithyrocaris, and is similar to the condition in the ceratio- caridids Caryocaris curvilata (Gurley) and CaUizoe hohemica Barrande. The anterodorsal-medioventral fold is more anteriorly situated than that of Pephricaris horripilata Clarke, whereas the "Schragrippe" and "Schnabelfurche" of Silesicaris nasuta Ciiirich (1929, p. 29) run closer to the ventral margin. As men- tioned above, the style resembles that in Echinocaris rather than that in any rhinocaridid, but the long smooth last abdominal segment is different from that of every species of that genus. The posteromedian groove on the median dorsal plate may prove to be of phylogenetic significance as a vestige of the non- rhinocaridid simple dorsal hinge, and homologous with the me- dian fold of other members of the Khinocarina. Ohiocaris shows the greatest development of the median dorsal plate and suggests a derivation from the Middle and early LTpper Devonian rhino- caridids. Thus the plate width/carapace width ratio is 0.146 in 6 BREVIORA No. 151 Ohiucaris, Ijut only O.UI)(i in the specimen of Elyinocaris slliqua figured by Beeeher (1902, pi. 19, fig. 8; Yale Peabody Museum 22410). Hall and Clarke's reconstruction of Mcsothyra oceani (1888, pi. 32, fig. 1) is inaccurate in showing the hypothetical median dorsal plate too broad. Measurement of the specimen upon which this reconstruction was based (New York State Museum 4576) shows that the ratio is 0.073, not 0.150 as figured. Two other specimens of M. oceani (NYSM 4577, 4581) give com- parable ratios of 0.093 and 0.082, and in the Rhinocaris columhina figured by Clarke (1893. fig. 4. NYSM 4786) the ratio is only 0.074. The concretion in which the Cleveland specimen occurs has the characteristic orange colour of oxidised Chagrin material described by Cushing, Leverett and Van Horn (1931, p. 34). Fossils do not seem to have been recorded previously from the Chagrin west of Cleveland (Cushing ef al., 1931, p. 35). The MCZ specimen forms part of a small collection comprising many of the same sj^eeies of brachiopods and pelecypods as those listed by Cushing et al. (1931, p. 35), and in addition the crustaceans Echinocaris multinodosa Whitfield, E. suMevis Whitf. and PalacopahKnuni mwlxryiii AYliitf.. EEFERENCES Beecher, C. E. 1902. Revision of tlie Phyllocarida from tlie Chemung and Waverly groups of Pennsylvania. Quart. J. Geol. Soc. London, vol. 58, pp. 441-449. Clarke, J. M. 189.3. On the structure of the carapace in the Devonian crustacean Bhinocaris ; and the relation of the genus to Mesothyra and the Phyllocarida. Amer. Nat., vol. 27, pp. 793-801. Cushing, H. P., Leverett, F., and F. R. Van Horn 1931. Geology and mineral resources of the Cleveland district, Ohio. Bull. U. S. Geol. Surv., 818. GURICH, G. 1929. Silesicaris von Leipe und die Phyllokariden iiberhaupt. Mitt. Min.-Geol. Staatsinst. Hamburg, Heft 11, pp. 21-90. Hall, J. and J. M. Clarke 1888. Trilobites and other Crustacea. Natural History of New York, Palaeontology, vol. 7. Whitfield, R. P. 1880. Notice of new forms of fossil crustaceans from the Upper Devon- ian rocks of Ohio, with descriptions of new genera and species. Anier. J. Sci., ser. 3, vol. 19, pp. 33-42. 1962 NEW DEVONIAN CRUSTACEAN FROM OHIO r *%.. .■•\ # r ^r %•• Plate 1. Ohiocaris icycoffi gen. et sp. nov., Cleveland Museum of Natural History 33241, x 1.7. BREVIORA useiuinn of Connparative Zoology Cambridge, Mass. January 15, 1962 Xtmber 152 NEW AUSTRALIAN DACETINE AXTS OF THE GENERA MES08TRUMA BROWN AND CODIOMYRMEX WHEELER (Hymeuoptera — Formicidae) By Robert W. Taylor Biological Laboratories, Harvard University The two new species described below are of considerable in- terest as members of tlie Australian ant fauna. The rare genus Mcsostruma, to which Mesostruma hrowni n. sp. is added, in- cludes two previously described species which have recently been revised by W. L. Brown, Jr. (1952). Mesostruma is known only from eastern Australia and occupies an important phylogenetic position in the subtribe Epopostrunuti, being almost exactly in- termediate in character between the major Australian genera Epopostruma Forel and Colohostruma Wheeler. Codiomyrmex flagellatus n. sp. is the second member of its genus to be de- scribed from northern Queensland, and as such is the second representative of the important short-mandibulate stock of the subtribe Strumigeniti to be recorded from Australia. The de- ficiency of this element in the Australian fauna is of considerable zoogeographic interest. Its historical absence or scarcity on the continent has perhaps been important in allowing adaptive radia- tion of the short-mandibulate Epopostrumiti of the genus Colo- hostruma (Brown, 1952, 1959; Brown and Wilson, 1959). INIesostruma browni new species (Figs. 1-5) Holotype worker. Synthetic aggregate length (TL)^ 3.6; head length (HL) 0.78; head width (HW) 0.71; mandibular exten- sion (ML) 0.37; Weber's length of alitrunk (WL) 0.8-t; cephalic lAU measurements stated in the present paper are given in miUimeters with indices in units. The conventions foUowed in measurements are those estabiisntu for the Dacetiui by Brown (195;}a, lySSb). An ocular scale with units of O.OKj.s mm. was used for measuring, with correction to the nearest unit, the maximum error being estimated as ±0.01 mm. BREVIORA No. 152 index (CI) 91 ; mandibulo-cephalic index (MI) 47. Correspond- ing precisely with the generic characteristics cited by Brown (1948, 1952) for head shape, mandibular and antennal structure, l^etiole and postpetiole form, and body sculpturing. Head shape as in Figure 1, similar to large M. laevigata workers, but clypeus and anterior part of head somewhat more transverse. Eyes small, strongly convex as in M. turneri. Humeral angles rounded ; dorsum of alitrunk in profile strongly and evenly convex. Pro- podeal lamellae as in Figure 2, more extensive than in either previously described species. Petiolar node as in Figure 2, similar to that of M. turneri, but more massive and less acute above in side view. Mesostruma irowni new species, Figs. 1 and 2, worker (Holotype). Fig. 1. Full-face view of head. Fig. 2, AUtrunk, nodes and base of gaster in side view. Figs. 3-5, male (Allotype). Fig. 3. Full-face view of head. Fig. 4. Alitrunk, nodes and base of gaster in side view. Fig. 5. Forewing. Pilosity and sculpture omitted from Figs. 2-4. Scale line: 1 mm. 1962 NEW AUSTRALIAN DACETINE ANTS 3 Sculpture of head, alitrunk and petiole consisting of large circular umbilicate foveae, more widely spaced than in the other described species, rarely separated by distances less than their maximum diameter ; the surfaces between them smooth and strongly shining. Dorsum of alitrunk less densely sculptured than head, with a median longitudinal area almost devoid of foveae. Posterior parts of sides of alitrunk opaque, the foveae indistinct and mixed with coarse punctures. Foveae of petiolar node smaller than those of head and alitrunk ; those of postpetiole indistinct, postpetiolar dorsum finely and irregularly sculptured and feebly shining. Dorsum of first gastric segment smooth and strongly shining, with no trace of longitudinal striae. Color rich golden brown ; petiole, base and apex of gaster, legs, mandibles and antennae lighter. Type locality. Two miles east of Berry, New South Wales (B. B. Lowery). Worker variation. Thirty-two paranidotype ivorkers, collected with the holotype, have the following dimensions : TL 3.2-3.9 ; HL 0.72-0.83 (mean 0.77) ; HW 0.66-0.74 (mean 0.71) ; ML 0.32-0.38 (mean 0.35) ; WL 0.77-0.92; CI 89-96; MI 42-49. No significant structural variation is indicated in the series. The workers are monomorphic with no bimodality in the frequency distributions of the dimensions listed, and no perceptible allometric differentia- tion between the head and mandibular dimensions, within the sample. The measurements and indices of ten paratype specimens from Riverview College, Sydney (B. B. Lowery), fall within the above ranges and have almost identical means, but an eleventh specimen of the same series is much smaller: TL 3.1 ; HL 0.69 ; HW 0.62 ; ML 0.34; WL 0.74; CI 90; MI 49. The HL of this individual is 2.80 standard deviation units smaller than the overall mean HL of all 45 specimens examined (76.7 ± SD 2.75), its HW is 3.0 standard deviation units smaller than the overall mean HW (70.4 ± SD 2.80). The specimen is thus of extremely small size when compared with workers from mature colonies, and is perhaps an old nanitic which had survived into the mature colony with which it was collected. A single paratype worker from Barrington Tops, New^ South Wales (T. E. Woodward), has the following dimensions: TL 3.2; HL 0.72 ; HW 0.66 ; ML 0.32 ; WL 0.77 ; CI 92 : MI 44. Paratype queens. The first series of dimensions are those of an alate from Burns Bay, Sydney (B. B. Lowery) ; the second 4 BREVIORA No. 152 series those of a dealate from Pymble, New South Wales (C. Mer- covich). TL 4.0, 4.3; IIL 0.80, 0.86; HW 0.76, 0.82; ML 0.37, 0.40; WL 1.07, 1.16; CI 95, 95; MI 46, 46. Differing from the workers in the usual characters of full sexuality: larger size, presence of ocelli and wings, and unreduced structure of ali- trunk. Coloration as in worker, the ocellar area dark brown. Wings clear with pale yellow veins, venational pattern similar to male. The same diagnostic features as those of the worker serve to distinguish queens from those of M. turneri. Allotype male. TL 3.0; HL 0.54; HAV including compound eyes 0.63 ; WL 0.88 ; forewing length ca. 2.2 mm. Head as in Figure 3. Compound eyes large, elliptical, strongly convex, their longest diameters about 0.28 mm. Mandibles slender, acute, probably not opposable. Antennae robust, thirteen segmented. Maxillary and labial palpi well developed; palpal formula ap- parently maxillary 5, labial 3, as in the worker (the mouthparts have not been dissected from the unique specimen). Body profile as in Figure 4. Mesonotum with w^ell developed notauli, their posterior portion, forming the stem of the "Y," indistinct. Propodeal lamellae smaller than in the female castes. Petiole sub-clavate, the node low, sloping gradually back from the anterior peduncle ; antero-ventral tooth obsolete. Lateral edges of post-petiole lacking aliform appendages, but each with a distinct, low, obtuse, longitudinal carina. Gaster broad, some- what flattened basally ; the basal edges of its first segment feebly carinate longitudinally, as in many epopostrumite workers. Geni- talia exposed, enfolded by the parameres. The latter similar to those of Oreciognathus (Brown, 1953b) : broad in dorsal view, with convex lateral outlines and strongly concave inner faces, the apices rounded with their tips turned inwards and opposed mesally. Apex of sub-genital plate acute. Cerci short and stout. The penis valves and volsellae have not been dissected from the specimen. Forewing venation (Fig. 5) of the Solenopsis type, as in Orectognathus, the apical elements (Rsf 5, Mf 4 and Cu-A) feebly developed, and the radial cell open. Hindwing narrow, with a broad posterior fringe of microtrichiae, and four well developed subapical hamuli; venation as in Orectognathus. Head and most of alitrunk coarsely and closely punctate. Punctures of the pronotal dorsum and prescutellar area similar to the foveae of the workers, but almost contiguous. Sides of alitrunk Avith quite extensive shining areas, especially on the 1962 NEW AUSTRALIAN DACETINE ANTS 5 median parts of the larger sclerites. Petiole and postpetiole with eoarse punctures, those on the latter shallow and irregular. First gastric tergite semi-opaque, with very irregular and shallow large, flat, piligerous punctures. Color blackish-brown ; antennae, mandibles, under-mouthparts, and legs yellowish-brown. Wings clear, their veins pale yellow. Described from a unique specimen collected with the holotype and its associated paranidotype worker series. Material examined. Northeastern New South Wales : 2 miles east of Berry (type locality) December 28, 1959, holotype and 32 paratype workers, allotype male (B. B. Lowery). Burns Bay, Sydney, February 2, 1959, ex leaf litter, a single alate queen (paratype) (B. B. Lowery). Riverview College, Burns Bay, Syd- ney, April 19, 1959, eleven paratype workers ( B. B. Lowery). Pymble, Sydney, March 18, 1956, a single dealate queen (para- type) (C. Mercovich). Barrington Tops, ex leaf mould (Berlese funnel sample), a single paratj'pe worker (T. E. Woodward). The holotype, with paratypes, has been returned to Father Lowery for eventual deposition in the Commonwealth Scientific and Industrial Research Organization collection at Canberra; the allotype, with paratypes, is in the Museum of Comparative Zoology, Harvard University ; the remaining paratypes are in the Queensland Museum. Biology. The following information regarding the biology and ecology of 31. hrowni has been provided by Father Lowery. The type locality is about two miles inland from Seven Mile Beach, in low hill country behind scrubby alluvial coastal flats. The collection was made in a grassed clearing in a heavily tim- bered area, with Eucalyptus and turpentine growing in black non-sandy soil. The Riverview College locality overlooks Burns Bay, Lance River Cove, Sydney. The colony taken there was found nesting in damp yellow sand beneath a cover of moss and a little grass. The site was in a clearing in low scrub about 15 to 20 feet high, with Eucalyptus corymhosa, Grevillea, Lantana and Leptosper- nuon. The soil near the nest contained a few small termite gal- leries and a large nest of the locally dominant ant Acropyga australis. The alate paratype queen was collected nearer the Burns Bay foreshore, wandering on leaf litter in warm sunshine, during the late afternoon. Father Lowery has collected three further colonies of M. lyrowni within 200 meters of the type nest site. One of these 6 BREVIORA No. 152 colonies was nesting about four inches below the surface of coarse black sandy soil, under a small rock, but probably not in direct contact with it. A maze of termite galleries was located immediately beneath the rock, and permeated the surrounding soil. The new species is dedicated to Dr. W. L. Brown, Jr. of Cornell University, a leading authority in ant taxonomy who has worked particularly with the Dacetiui and has devoted much study to the Australian ants in general. The addition of M. hrowni to Mesostruma requires no change in the basic concept of the genus, as formulated by Brown (1952). Indeed Mesostruma retains its appearance as a compact and distinctive genus, with its species abundantly distinct from each other. Considering the characters of the worker and queen, M. hrowni seems to be most closely related to M. turneri Forel, which it resembles in the structure of the alitrunk, propodeal lamellae and petiole. It has a proportionately narrower head, however (see Brown and AYilson 1959, fig. 7), and lacks the longitudinal striation of the basal gastric segment seen in turneri. The ab- sence of humeral denticles, and the general form of the alitrunk, propodeal lamellae and petiole distinguish M. hrowni from M. laevigata Brown. The new species differs from both the previ- ously described species in its less dense sculpturing and overall glossiness, and its more extensive propodeal lamellae and more massive petiolar node. The three known species of Mesostruma may be separated by the following key (based on the workers). 1. Ilunieri rounded; dorsum of alitrunk strongly and evenly convex in profile; eyes protruding and very eonvex '1 Humeri acutely subdentate; dorsum of alitrunk not so markedly convex in profile; eyes less convex and protruding only slightly. (Victorian mallee) ^1/. laevigata Browai 2. Head broad, CI 98-100; head capsule opaque, the foveae almost con- tiguous ; gaster finelj' longitudinally striate over basal half or more of segment I (vicinity of Cairns, Queensland) M. turneri Forel Head narrovrer, CI 88-96 ; head capsule strongly shining, the foveae separated by smooth areas at least as wide as their maximum diameter; basal segment of gaster smooth and strongly shining (northern New South Wales) M. hrowni n. sp. In the discussion above I have not considered the enigmatic species f Mesostruma, monstrosa (Viehmeyer), 1925 (Brown, 1948). The unfortunate circumstances surrounding the original 1962 NEW AUSTRALIAN DACETINE ANTS 7 selection of this species, which was based on an apparently ab- normal specimen, have been discussed by Brown (1952). Dr. Brown now believes (personal communication) that Viehmeyer's species was most likely based on a defective Epopostruma speci- men. In anj^ case monstrosa seems best ignored, pending the location and competent re-examination of the type. CODIOMYRMEX FLAGELLATUS UCW SpCcicS (Figs. 6-9) Holotype worker. TL 1.9 ; HL 0.48 ; HW 0.32 ; scape length (SL) 0.23; ML 0.08; WL 0.47; CI 68; MI 19. General form much as in C. semicomptus Brown (1959) (Fig. 9), but smaller and more lightly constructed. Shape of head as in Figure 6. Dorsal surface of cranium convex, sloping towards occiput and clypeus as in C. semicomptus, the convexity less pronounced, however, and the occipital lobes more broadly rounded when viewed from the side (c/. Figs. 7, 9). Mandibles strongly convex, rising above the anterior clypeal border ; with five strong, acute, conical teeth, each slightly smaller than the one posterior to it. Basal lamella normally hidden at full mandibular closure, set at a slightly lower level than the teeth and oblique to them; its shape as in Figure 8 — roughly right -triangular, the posterior edge almost perpendicular to the mandibular axis, and the anterior edge diagonal, forming the hypotenuse. The anterior lamellar edge rises almost immediately from the base of the proximal mandibular tooth so that the diastema is very brief. Clypeus almost perfectly plane. Body profile somewhat as in C. semicomptus, with which it is compared in Figures 7 and 9. Alitrunk narrow, its maximum width 0.59 X the HW; its dorsum almost perfectly plane, with- out sutures or a median longitudinal carinula. Alitrunkal dor- sum in side view evenly arched between the pronotum and the propodeal teeth. In dorsal view the sides of the pronotal disc are evenly rounded and those of the remaining alitrunkal dorsum almost parallel, with the distance between them, at the base of the propodeal teeth, slightly more than half the pronotal width. Dorsum of alitrunk margined with a fine carina, less distinct than in C. semicomptus, enclosing the pronotal disc anteriorly, and continuous with the upper edges of the propodeal spines posteriorly. The latter with their infradental lamellae similar to those of semicomptus. Propodeal spiracle minute, circular, its margin not appreciably expanded ; it contrasts with that of C. 8 BREVIORA No. 152 semicomptus, which is larger and has a wide and very conspicu- ous rini-like margin (Fig. 9). Profile of petiolar node as in Figure 7, shorter than that of semicomptus, only about as long as high in side view ; seen from above the node is slightly longer than broad with rounded sides and a truncate anterior border. A full complement of areolate spongiform appendages is de- veloped, distributed normally as in Figure 7. Gaster depressed ; basigastric costulae reduced to about five feeble lines on either side of segment one, the two groups of costulae separated by a wide median shining area; the costulae extend back about y^ the length of the segment. Mandibles shining, with a few scattered punctures. Head capsule with a close cover of large, flat, irregular, shallow punc- tures, which are almost effaced in a small region in the center of the frons. Antennae very finely punctate, their scrobes punc- tate-granulose. Alitrunk, both nodes, gaster, legs and anterior parts of frons and clypeiis smooth and strongly shining. Petiolar l^eduncle coarsely granulate. Coflionv/rmi .r ffncjcUatits new species, Figs. 6-8, liolotype -workoT. Fig. C^. Full-face view of head. Fig. 7. Alitnuik nodes and liase of gaster in side \iew. Fig. 8. ilandiblc. Codiomyrmcx semicomptus Brovrn, paratope worker. Fig. 9. Alitrunk, nodes and base of gaster in side view. Scale line : O.IG mm., for Fig. 8; 0.25 mm., for Figs. 6, 7 and 9. 1962 NEW AUSTRALIAN DACETINE ANTS 9 Occiput with a number of very fine long (0.05-0.08 mm.) hairs with clavate tips, somewhat finer than in C. semiconiptus. Hairs of clypeus shorter and more distinctly clavate. A very few^ simi- lar hairs are present on the anterior part of the pronotal disc and the dorsal surfaces of both nodes. The pilosity otherwise consists of exceedingly long (0.13-0.36 mm.), fine, tapering hairs, which are distributed symmetrically. On the head, five pairs in the following positions (see Fig. 6) : on the occipital border, about halfway between its midpoint and its lateral extremity on either side ; on either side and a little anterior to the median part of the f rons ; at the edge of the cephalic dorsum just anterior to its widest point; at the edge of the cephalic dorsum at about mid head length ; on either side of the posterior extension of the cly- peus, above the antennal insertions. On the body, six pairs : two pronotal ; one on each node ; two at the base of the gaster — distributed as shown in Figure 7. On the forelegs : a single flagellum on the outer side of the tibia, near its apex. On the middle and hind legs: single hairs on the outer sides of the limbs near the bases of the tibiae and basitarsi. These long flagella curve upward and inward on the head capsule ; those of the body are erect with their apices turned posteriorly ; the anterior pronotal and postpetiolar pairs are more tangential, and inclined laterally. Type locality. Clump Point (near Mourilyan) Queensland, June 3, 1953 (T. E. Woodward), collected from a Berlese fun- nel sample of leaf mould. Worker paratype variation. Seventeen worker paratypes, col- lected with the holotype, show no significant variation in size or structure. The specimens have at some time been subjected to drying and fungal attack while in alcohol storage, with the result that some are fragmentary. A few fungal hyphae are still attached to several of the specimens, and a number have had the pubescence, particularly the elongate flagella, damaged dur- ing the consequent cleaning. Among the known Codiomyrmex, C. flagellatus is most closely related to C. semiconiptus Brown, the only other known Aus- tralian species. The general resemblances between these forms are indicated in the accompanying figures. The two species may be readily separated by the characters discussed in the above comparative dsecription; the difi^erences in size, pilosity, and propodeal spiracular structure are especially characteristic. The adaptive significane of the peculiarly sparse and elongate body pilosity of C. flagellatus is not understood, and deserves examination should live material become available for future study. 10 BREVIORA No. 152 The holotype, with paratypes, has been placed in the collection of the Queensland Museum. Paratypes are deposited in the Museum of Comparative Zoology, Harvard University, and the Commonwealth Scientific and Industrial Research Organization Collection, Canberra. I wish to acknowledge the generous assistance of Father B. B. Lowery of Sydney, and Dr. T. E. Woodward, of the University of Queensland, who collected the bulk of the material cited above and made it available to me for study. I wish also to thank Dr. E. 0. Wilson, of Harvard University, for his assistance and advice during the preparation of this paper. REFEEENCES Brown, W. L., Jr. 1948. A preliminary generic revision of the higher Dacetini (HjTiien- optera: Formicidae). Trans. Amer. Ent. Soc, 74:101-129. 1952. The claeetine ant genus Mesostruma Brown. Trans. Roy. Soc. S. Aust., 75:9-13. 1953a. Revisionary studies in the ant tribe Dacetini. Amer. Midi. Nat., 50:1-137. 19o3b. A revision of the dacetine ant genus Ortciognaihus. !ilem. Queensland Mus., 13:84-104. 1959. Some new species of dacetine ants. Breviora Mus. Comp. Zool. Harvard, 108:1-11. Brown, W. L., Jr. and E. O. Wilson 1959. The evolution of the dacetine ants. Quart. Rev. Biol., 34(4): 278-294. BREVIORA Museum of Comparsitive Zoology Cambridge, Mass. February 15, 1962 Number 153 AN0LI8 SCRIPTUS GARMAN 1887, AN EARLIER NAME FOR ANOLIS LEUCOPHAEUS GARMAN 1888 By A. Stanley Rand Garman in 1887 described Anolis scripfus on the basis of five specimens in the Museum of Comparative Zoology, giving the type locality as "Silver and Lena Keys, Fla." Barbour in 1914 re-examined Garman 's type series and decided that they were identical with A^wlis cristatellus from Puerto Rico and the Vir- gin Islands and therefore placed A. script us in the synonymy of A. cristatellus. In the course of an examination of the Museum of Comparative Zoology anoles referred to A. cristatellus I had occasion to study the type series of A. scriptus. I find that the series is mixed and none is cristatellus. One, a juvenile, is A. homolechis quadrocelifer of Cuba; the other four are conspecific with the form from the southeastern Bahamas described by Gar- man (1888) as Anolis leucophacus, and apparently subspecifically identical with the form from the Turks and Caicos Islands now called alhipalpebralis Barbour 1916. Clearly the name scriptus can no longer be kept as a synonym of Anolis cristatellus, but correction of its status raises certain problems. Since the series is mixed a lectotype must be selected to fix the name. The type series, three adult males and two juveniles, are all somewhat faded from their long period of preservation. One of the juveniles possesses the scale characters of Anolis JiomoJechis and the color pattern, dark spots over the shoulders, is still suf- ficiently evident to identify it as Anolis homolechis quadrocelifer. This is the specimen labeled as coming from Lena Key, which thus would appear to be Cayos de la Lena, near Cabo San An- tonio, Cuba.^ I have arbitrarily excluded this juvenile from the 1 For further information on this form see Rnibal and Williams (1961). 2 BREVIORA No. 153 concept of A. scriptus and it therefore needs no further dis- cussion. The remaining four specimens seem to belong to a single species and I herewith designate M.C.Z. No. 65950 as the lectotype of Anolis scriptus Garman. The labels accompanying these specimens say "Silver Key Florida." I, like Barbour, have been unable to locate a Silver Key anywhere in the West Indies. There is a Silver Bank near the islands from which the types must have come in the south- eastern Bahamas but it is completely submerged. These specimens are very like cristatellus as both Garman and Barbour agreed. Garman distinguished them from cristatellus on the basis of the greater size of the two paravertebral scale rows. Barbour (1914, p. 274) said, "I can not see, however, that these are at all enlarged ; and there is no other character in which they vary from true A. cristatcllns." An examination of the type series helps to explain this contradiction. Two of the males have the two paravertebral scale rows enlarged more than is usual in cristatellus, but the third male has the paravertebral scale rows scarcely enlarged at all and it is undoubtedly this specimen that Barbour examined. However, a close comparison shows certain other and more constant differences between the type series of scriptus and the many specimens of cristatellus examined. In scriptus the dorsal scales lateral to the paravertebral rows are larger than they are in specimens of cristatellus of similar size. In cristatellus also, the frontal ridges are higher and sharper and the frontal depres- sion correspondingly deeper than in the type series of scriptus. Finally, in cristatellus, there are only 1-3 scales behind the inter- parietal and these are abruptly larger than the very small dorsal scales. In scriptus there are many more rows of enlarged scales in this position and they grade more gradually into the dorsal scales. In all of these characters the type series of scriptus differ from cristntellus and agree with specimens of the species now called leucophaeus. So far as I can find, the types of scriptus do not show any scale differences from leucophaeus, nor does leucophaeus show any additional differences from cristatellus. From this it appears that scriptus and leucophaeus are synon- ymous and scriptus as the older name must be substituted for leucophaeus. 1962 ANOLIS SCRIPTUS 3 The species "leucophaeus" is quite widely distributed in the southeastern Bahamas and has been divided into four subspecies. These races have been described primarily on the basis of color pattern, and they are all very similar in scalation. They are diagnosed in Table I. The types of script us lack the many dark spots characteristic of leiicophaeus and the lectotype has a well-developed tail crest which is lacking in sularum. Thus the name scriptus definitely does not apply to the populations called leucophaeus and sularum. Distinguishing between albipalpehralis and mariguanae is more difficult. The diagnostic difference between them is the pres- ence of a broad dark lateral band in mariguanae. This is absent in the type series of scriptus but it is also absent in many of the adult males of mariguanae and best developed only in the juve- niles and females. Even the small ''type" of scriptus lacks this band but this specimen is so faded that one cannot be positive that the band was never present. Many of the females of albipal- pehralis have dark middorsal blotches which are lacking in the small "type" of scriptus but, since they are absent in man}' alhi- palpebralis, this is not conclusive. The male scriptus have a com- plex mottling along the sides in addition to a light narrow lateral line. The light lateral line is found in both albipalpcbralis and mariguanae but the mottling in the types of scriptus is most like that found in albipalpebralis. Finally, the lectotype of scriptus has a dark line running posteriorly from the eye onto the neck. This marking is found in some of the males of albipalpebralis but in none of the mariguanae examined. So far as can be deter- mined there are no useful scale differences between mariguanae and albipalpebralis. From this it appears that the "type" series of scriptus, while not indisputably assignable to either of these races, is most like albipalpebralis and the lectotype most clearly so. For this reason it seems necessary to replace the name albi- palpebralis by the name scriptus. In accordance with this change the type locality of scriptus is restricted from "Silver and Lena Keys" to "Silver Key," Turks and Caicos Islands. Further restriction seems pointless at this time. The correct names for the races of this species now stand as follows: Anolis scriptus scriptus Garman 1887 = Anolis albipalpebralis Barbour 1916 Anolis scriptus leucophaeus Garman 1888 Anolis scriptus mariguanae Cochran 1931 Anolis scriptus sularum Barbour and Shreve 1935 BREVIORA No. 153 ■3 op -- »-. ^ c ft ft O) =0 c « s ci 3 •^^ 3 be Si. tx c3 •«* t ^ =0 ^ a ■2 r«» o s ^ OQ «l-l o o cc ^ 01 15 o a> tH 11 ^ o '^ ft ■r. _ft C3 03 en to .f- "rt 5 K 13 H 1— ( ^ H S3 bD ^ -S HI > rM o o rt >-i cS ■^ -^^ ns iw J5 a> c: o ;« o O) o >5 d 'M ,^ ^ p, o g 3 0^ •« __ § • s- S, r-^ ^ t:? g rt ■;; S o ^ c? ^H ft .^ a* be r^ 1— • bt o "« .5 IH c ^ o 5 Is m ft fii rt c4 OQ 3 I— 3 be es 7? « j^ S ^ O c: 1—1 K ft O o 3 o 3 Jh • p- *" c 1IJG2 ANOLIS SCRIPTUS 5 EEFERENCES CITED Cochran, D. 1931. New Bahaman reptiles. Jour. Washington Acad. Sci., 21: 39-40. Barbour, T. 1914. A contribution to the zoogeography of the West Indies, with especial reference to amphibians and reptiles. Mem. Mus. Comp. ZooL, 44: 209-359. 1916. Additional notes on West Indian reptiles and amphibians. Proc. Biol. Soc. Washington, 29: 215-220. Barbour, T. and B. Shrkve 1935. Concerning some Bahaman reptiles, with notes on the fauna. Proc. Boston Soc. Nat. Hist., 40: 347-366. Garman, S. 1887. On West Indian Iguanidae and West Indian Scincidae in the collection of the Museum of Comparative Zoology, Cambridge, Mass., U.S.A. Bull. Essex Inst., 19: 25-53. 1888. Reptiles and batrachians from the Caymans and Bahamas. Bull. Essex Inst., 20: 103-116. RuiBAL, R. AND E. E. Williams 1961. The taxonomy of the Anolis homoleohis complex of Cuba. Bull. Mus. Comp. Zool., 125: 211-246. BREVIORA Mmseium of Compsirsitive Zoology Cambridge, Mass. April 4, 19G2 Number 154 NOTES ON HISPANIOLAN HERPBTOLOGY 5. THE NATURAL HISTORY OF THREE SYMPATRTC SPECIES OF ANOLIS By A. S. Rand The lizard genus Anolis is abundant in species and individuals in the Greater Antilles. Quite a number of the common species occur together over wide areas. There is thus a special oppor- tunity to study the ecological relations of sympatric species of a single genus under conditions genuinely favorable for field observation. Several papers have now been published which have begun to exploit this fortunate opportunity: Ruibal (1961) and Collette (1961) working on the anoles of Cuba; Oliver (1948) on those of Biniini in the Bahamas; Williams and Rand (1961) on the scmUincatus group in liispaniola. These and earlier workers — Grant (1940) for Jamaica; Stejneger (1904) and Schmidt (1928) for Puerto Rico and especially Mertens (1939) for His- paniola — have established that in each well studied case there are small but definite differences in microhabitat among the sympatric Anolis. In this paper I document the same point for the three com- monest species of Anolis in Hispaniola. I describe also a number of behavioral differences which seem to be associated with the ecological differences and attempt to assess the adaptive sig- nificance of these behavior patterns. This study was made during a two-month field trip to the Dominican Republic and supple- mented with observations on captive lizards kept in the lab- oratory in Cambridge. ACKNOWLEDGMENTS I wish to thank the government of the Dominican Republic that, through the Universidad de Santo Domingo, provided us 2 BRE\'TORA No. 154 with transportation and other assistance in that country ; Dr. Engenio de Jesus Marcano who helped us in the field and who sent me live specimens on our return. I wish also to thank Mr. Clayton Ray for inviting me to accompany him on this trip, and for his assistance in the field; Dr. E. E. Williams for helping to arrange the trip and for his advice and assistance throughout, and Dr. A. L. Rand for critically reading the manuscript. Also I wish to thank Sigma Xi for providing some of the funds that made this trip possible. MATERIAL AND METHODS The field observations during the summer of 1958 were made incidentally to the main project, a survey of Dominican caves for vertebrate fossils. Consequently, they include no lengthy ob- servations in any one locality but are a synthesis of data col- lected from widely separated places. Though I noticed no geo- graphical variation in behavior, obviously this is a possible source of error that runs throughout the present study. This is par- ticularly true as observations are lumped for different subspecies in two cases: Anolis clilorocyanus chlorocyanus and A. cliloro- cyanus cyanostictus ; and Anolis distichus ignigularis and A. distich Its dominice^isis. The main areas, where observations were made, were : Santo Domingo and vicinity, Santiago and the nearby Sierra Septen- trional, Padre Las Casas, Sabana de la Mar and the mountains south of there, and the Sierra de Neiba. These field observations are least complete on A. cJdorocyonus and uneven on the other two species. For example, I found eggs only of cyhotes, while I observed copulation only in distichus. The laboratory observations made at the Harvard Biological Laboratories involved only two species, A. cyhotes and A. chloro- cyanus chlorocyanus. These animals, ten of each species, all adults and about half of them males, were sent to me by Pro- fessor Eugenio de Jesus Marcano. The lizards were released in a room 12 x 20 ft. and 9 ft. high. The room was a constant 80° F. and lighted by a south window. Eight small tree trunks 7 feet tall, mounted on bases so that they stood upright, an 8 foot rub- ber tree, a potted pine bush, and some other potted plants w^ere arranged around the room. The li'-ards were fed meal worms on the floor and in dishes taped at various heights above the floor, on the tree trunks. Fruit flies were also released in the room. The whole room was watered daily. ] 962 THREE IlISPANIOLAN ANGLES 3 For a period of about three weeks, daily observations of vary- iiiw lengths were made. The lizards quickly came to ignore me as I sat quietly in one corner of the room and watched them with low power binoculars. This was obviously not a natural situation but the observations that could be checked against the field notes show a gratifyingly close agreement. This method seems to be a useful adjunct to field observations though obviously no substitute for it. All three of these species are small to medium sized lizards possessing the characteristic specializations of the genus : en- larged subdigital lamellae and a throat fan or dewlap that is best developed in the male. In each species the males grow to a larger size than do the females. A)iolis disticluis varies in dorsal coloration from green to gray to brown, frequently with mottling on the back. The dewlap is yellow in A. d. doniinicensis, and red with a yellow border in A. d. ignigidaris. The males measure 51 mm in snout-vent length (Cochran 1941). Anolis cydotes is dorsally gray or brown, sometimes reddish, frequently with two indistinct greenish lateral stripes. The dew- lap is whitish or yellowish in color. The males measure 67 mm in snout-vent length (Cochran 1941). Anolis chlorocyanus is usually bright green with the ability to change to brown. A. c. cyanostictus also has a rust-red spot in front of the shoulder and one on the head. A. c. chlorocyanus has a dewlap that is light blue anteriorly and dark, almost black posteriorly. A. c. cyanosiicius has the dewlap proximally cad- mium yellow, distally sky-blue. The males measure 71 mm. in snout-vent length (Cochran 1941). OBSERVATIONS The observations from both field and laboratory have been combined and arranged according to topics. Under each topic heading all three species are compared. Geographical Distrihutio7i. All three species are widespread in the Dominican Republic. Apparently they avoid only the high, wet pine forests of the Cordillera Central and are less common in the dry regions on the south coast near Asua and in the northwest near Puerto Plata. However, A. chlorocyanus is replaced by A. coelestinus in the Barahona Peninsula. Geo- graphical variation in each of these species has been described byMertens (1939). 4 BREVIORA No. 154 Hahitat. I found these three species at almost every locality I visited in the Dominican Republic. Mertens called distichus and cybotes eurytopic forms and chlorocyanus only slightly less so. He records both distichus and cybotes from the mangrove areas along the coast to the lower pine forests around Constanza at some 1200 meters altitude. The most significant environmental factor for all three species seems to be some sort of vertical perch, a tree trunk, fence post or cliff face, but A. chlorocyanus seems to avoid the most open areas. Anolis distichus lives primarily on isolated trees and fence posts and along the edges of woods and trails and in open woods. A. cybotes lives in these situations but also occurs in the deeply shaded interiors of densely wooded areas that distichus avoids. A. clilorocyanus certainly occurs in the edge situations that dis- tichus prefers and probably in the heavier woods as well. Though there are differences such as these in the extremes tolerated, these animals live in the same habitats over most of the Dominican Republic and part, at least, of Haiti. Mertens notes that in separated localities he observed individuals of all three species living in the same tree. My observations agree with this. Microhabitat. These three species, though living in the same habitat, have differences in their microhabitat preferences. A. distichus lives almost exclusively on exposed (i.e. not closely surrounded by vegetation) tree trunks, fence posts and similar structures, within 10 to 15 feet of the ground. It is seldom seen on the ground or in the smaller branches of bushes or trees. Though it must descend to the ground to reach isolated trees, it does not spend much time on the ground. A. distichus is often very common on the palisade fences in small villages. A. cybotes is also primarily an animal of exposed tree trunks and fence posts within 10 feet of the ground. However, it also frequents rocks and fallen logs and smaller individuals are frequently seen on the ground. It also avoids, during the day at least (see Sleeping), small twigs and foliage. These field ob- servations are confirmed by the laboratory data which show cybotes spending most of its waking time on perches less than five feet from the floor. A. chlorocyanus lives also on tree trunks and fences but unlike the other two, frequently ranges high in the trees and out among the smaller branches. I saw a female fall from the crown of a 30 foot palm tree. I suspect that this species is one that lives primarily up among the branches and ranges down the trunk 1962 THREE HISPANIOLAN ANGLES 5 rather than the reverse. However, individuals were seen also on fences and once I saw a dozen individuals on a pole frame- work nine feet high that with a few bits of palm thatch was all that was left of a shed in the middle of a treeless pasture. Like distichus, chlorocyanus must occasionally descend to the ground to reach isolated trees. In the laboratory, chlorocyanus spent most of its time high on the perches and ventured out among the leaves of the rubber plant much more than did the cybotes. Territoriality. As Mertens noted, all three species appear to be territorial. He says that for A. cyhotes and chlorocyanus he found only a single adult male and one or more females of each species on any one tree. When he placed an additional male on one of the occupied trees it was immediately attacked. A. dis- tichus, he saj^s, defends a smaller territory and the large trees may have several males, each with its own territorj^, as well as a number of females. These observations agree with mine, although it was only in areas where distichus were extremely common that I found more than one adult male distichus on the same tree, and then only on large trees. Certainly in the laboratory no male of either cyhotes or chloro- cyanus would tolerate another male of the same species in his immediate vicinity for long, and even females were chased away if they approached too closely. Occasionally a male of one species would display to an individual of the other species. In addition to this horizontal effect a vertical stratification on individual trees or posts is evident in at least two of the species, distichus and cyhotes. On the smaller trees that were occupied by two individuals of distichus, the lizards were usually of opposite sex and the male was usually closer to the ground than was the female. On one tree there were four lizards of this species : a large male near the base, a female above him, a slightly smaller female above her and, highest of all, a juvenile who was about 7 feet above the ground. During the hour that I watched this tree all of the lizards moved a number of times but this stratification remained the same for it was actively enforced. When one of the higher individuals moved in sight of and within 2 or 3 feet of a lower one, the lower animal immediately reacted to it. The male did this by bobbing his head and pumping his dewlap, the female by a short charge in the direction of the intruder. In every case the intruder retreated immediately, around the tree and up. 6 BREVIORA No. 154 In cyhotes almost all of the individuals on the trees were large and most of them were males. Tlie individuals that were seen on the ground were almost always juveniles or females; the females are much smaller than the males in this species. This spatial distribution seems to be a real phenomenon but I have no field evidence to support the hypothesis that it is the result of territorial defense. However, there are certain laboratory observations that support this contention. In the laboratory I saw no vertical stratification but I also saw no evidence that the females tended to spend more time on the ground than did the males. This may be because the perches were small and nu- merous enough so that each lizard could occupy its own, as they usually did. However, there was a definite stratification horizontally with the large males on the perches near the window and the smaller males and females on those farther away. AVhen a smaller individual invaded a perch near the window, and this happened quite frequently, the resident male would display to it and the intruder would retreat. Thus, though there is no direct evidence that the distribution seen in the field is due to territorial defense, there is evidence that defense could be at least a contributing factor : the males taking the most de- sirable positions, the elevated perches, and ciiasing away any smaller individual that attempted to move in, with the result that the juveniles and females would spend most of their time on the ground. It is of interest to note that this vertical stratification in cyhotes and disiichus results in greater difference in size between the individuals that occur together. Adult cyhotes are larger, and the adult males much larger, than adult distichus. This size difference could be reflected in the size of the prey items taken and so reduce the competition for food between the two species. The young of cyhotes are of course no bigger than adult dis- tichus but these are the individuals that live primarily on the ground and so do not occur on the tree trunks with the distichus. The importance of this in reducing competition for food, if it acts in this way at all, is of course unknown. It would be inter- esting to have an analysis of the stomach contents of various sized individuals of each species. In chlorocyanus, I have no clear evidence for vertical strati- fication either in the field or in the laboratory. However, the fact that most of the individuals caught on fences and tree trunks were adult males suggests that the females and the juveniles may 1962 THREE HISPANIOLAN ANGLES 7 Stay higher in the trees. This phenomenon may also give maxi- mal emphasis to the size difference between this species and the distichus with which it comes in contact for it is the large indi- viduals of chlorocyanus that descend the tree trunks and they meet first and presumably would compete most for food with the smallest distich us. Posture. Seen in silhouette these three species are usually im- mediately recognizable. This is partly because of their different proportions but even more because of their very dift'erent postures. A. distichus rests facing either up or down or angling across the tree with its head and at least the anterior part of its body well off' the substrate and with its neck bent so that its head is parallel to the substrate but further away from it than are its shoulders (Cf. figure 4, plate 2, Mertens). A. cyhotes typically rests facing down the tree with the fore part of the body off the substrate and the neck bent dorsally so that the head is nearly parallel to the ground. In one individual seen resting on the underside of a log which slanted at about 45°, the neck was bent back well over 90°. This posture is true in the laboratory as well as in the field. On the ground the posture is much like that of distichus with the neck bent so that the head is parallel to the ground and raised above it. A. chlurocyanus usually rests with both its head and body quite close to the substrate and its neck bent only a little if at all, both in the field and in the laboratory. While these postures are typical of the normal resting posi- tion, both cyhotes and disticltus when mildly alarmed flatten against the substrate, and cklorocyanus, when about to display, raises itself up on its legs. I believe these differences in posture can be correlated with feeding behavior as discussed below. Mertens says that most Anolis rest with their heads pointing toward the ground. I noticed this most commonly in cyhotes and less so in the other species. Activity. A. distichus appears to be a much more restless lizard than either of the other tw'o species. Like them it spends most of its time resting quietly but an individual seldom remains in one place for more than a few minutes. It then moves quickly a few inches away on its tree trunk and rests quietly again for another few minutes. A. cyhotes, on the other hand, seems to spend much longer periods of time resting in one spot. Again, when a change is 8 BREVIORA No. 154 made, it is made quickly. Tn the laboratory when a cyhotes moved from one tree to another, the lizard frequently ran down to the base of the tree, stopped for a few moments, left the tree with a jump and ran part way across the floor, paused, ran the rest of the way and with a jump started up the new tree, usually pausing again before settling down. A. chlorocyaniis, though also spending periods of time im- mobile, moved quite frequently, going slowly and deliberately about in the trees, in the laboratory as well as in the field. In the laboratory, moving from one tree to another was a single process. The lizard moved slowly down to the base of the tree, jumped off, dashed across the floor and with a jump started up the new tree. As little time as possible was spent on the floor. Climhing. I have no observations on the climbing ability of distichus. In the laboratory A. cyhotes did not appear to be as sure a climber as A. chlorocyanus. I saw several of the former fall to the ground from the smooth leaves of the rubber plant while I noted that only one of the latter did so even though chlorocyanus spent much more time on these leaves than did cyhotes. While both these species started their climbs in the lab with a jump up on to the vertical surface, and I saw a large male cyhotes make a 6 or 7 inch vertical jump to reach a hanging branch, chloro- cyanus made a great many more horizontal jumps. Particular!}' common was one used to cross the 10 inches that separated the two closest perches. A. cyhotes is apparently a quickly moving lizard, quite at home on the ground but not so much so in the more treacherous footing of smooth green leaves and small twigs. A. chlorocyanus, on the other hand, is shy of the ground, moving across it only occa- sionally and then with as much speed as it can manage, while in the less secure footing of more arboreal situations, its de- liberate movements help keep it from falling. Feeding. On several occasions a distichus was seen to move from its resting position on a tree trunk or fence post and snap up something small from the bark. Twice I saw a male interrupt his displaying to do this. Once I saw a male move up to an ant about an inch away, follow it up the tree a couple of inches and then apparently lose interest and turn away. Thus there seems to be some selection of food and not everything small that moves within range is eaten. 1962 THREE HISPANIOLAN ANGLES 9 A. cyhofes was never seen to catch anything in the field, though on two occasions I saw an individual struggling with a large dragonfly ; each time the lizard had the head and thorax in his mouth and the wings and abdomen still protruding. In the lab- oratory, cyhotes came willingly to the floor to take meal worms. Sometimes a lizard would return to a tree to eat the meal worm but usually it remained for several additional minutes before returning either to its old tree or a new one. I never saw a cyhotes moving around on the floor looking for food. On one occasion a male left his perch and ran about 10 feet across the floor to make an unsuccessful attempt to capture a 2-inch cock- roach. I have no data on chlorocyavus feeding in the field. In the laboratory several individuals were seen snapping at small ob- jects on the leaves and twigs. This species only rarely came down to the floor to capture meal worms and each time that one did so it moved slowly to the base of the tree, then rushed out, seized the meal worm, ran back and climbed up the tree before stopping to eat its captured prey. On the basis of this limited evidence, some tentative generali- zations about the relations of feeding, posture and movement can be made. A. distichus seems to feed primarily on smaller insects that it catches on the tree trunk. The posture with the head held high above the substrate would enable it to see more of the tree trunk than it could if the head was held low. The frequent move- ments are necessary if the lizard is to take advantage of the in- sects that happen to be on the opposite side of the tree trunk. The head is held parallel to the surface in which the lizard is most interested. A. cyhotes males get at least part of their food from the ground after sighting it from their elevated perch. The posture of this species keeps the head roughly parallel to the surface in which the lizard is most interested, in this case the ground. Since the lizard can see a large sector of the ground around him at all times, he need not change his position frequently to maintain a careful scrutiny of a considerable area. A. cJilorocyanus almost certainly gets most of its food from the trees in which it usually lives and its slow movements sug- gest that at least part of the time, unlike the other two species, it goes looking for it instead of lying in wait. The head is held parallel to the surface in which it is most interested. 10 BREVIORA No. 154 We need more observations on all of these points and par- ticularly on the feeding of A. cJilorocyanus. But it is interesting that the two species, distichiis and cyhotes, that occur in the same microhabitat, on the tree trunks, seem to differ so markedly in feeding behavior. Mertens remarks, in passing, that most Ayiolis get their food on the ground and among the roots of the trees and that their usual posture, as he records it, oriented toward the ground, may be related to this. I believe that this is true of cyhotes but not of the other species. Escape. Each of the arboreal species has a noticeably differ- ent method for evading herpetologists and presumably other predators. A. disiichus, when approached, quickly moves around to the other side of its tree trunk, usually moving up or down at the same time. If followed, the lizard may continue this maneuver- ing to keep the tree trunk between itself and the pursuer for some minutes. Soon, however, the lizard will either run up the tree out of reach or run down it to the base where it is concealed by the surrounding vegetation. Only when very hard pressed will one leave his tree or post and run out into the grass. Occasionally, and this was noted especially early in the morning, distich us would hide under a bit of bark or in a hole in the tree. Mertens says that distichiis, when approached, runs around to the other side of its tree or post and then down to its base to hide. He does not mention any of the other behavior described here. A. cyhotes showed the same tactics in the field and in the lab- oratory. When approached, an individual would remain still until I came very close, then suddenly it would dart around to the other side of the tree ; it might stop there but only until I moved into view again, then, instead of employing evasive ac- tion on the tree, it would usually run down and, unlike distichus, frequently leave the tree completely. In the laboratory the lizard usually ran a few feet away, where it might remain on the floor for some moments or might immediately climb a new tree. Mer- tens records these same flight reactions for this species. An A. cyhotes, discovered on the ground or on a fallen log, frequently hid under whatever cover was available. Sometimes a large male, when first approached, would display his dewlap to me before fleeing. This was observed occasionally in the wild and became very common with one male in the laboratory after he became accustomed to me. ]962 THREE IIISPANIOLAN ANGLES 11 Usually in the field, cJilorocyanns, when approached, imme- diately began to climb up the tree. This action was slower and more deliberate than that of the other two lizards but, since it was started sooner, the lizard was usually carried safely out of my reach. In the laboratory and in the field, when on a fence, where the lizard could not climb out of reach, it usually climbed as high as it could get and then dodged about there. Only very occasionally did one run down to the ground, though frequently one jumped to a nearby perch, if available, or ran out among the foliage and twigs at the ends of the branches. Mertens also notes that cJilorocyanns usually climbs up its tree and conceals itself in the crown. The escape behavior of the three species closely parallels the feeding behavior. A. distich us conducts its evasive behavior on the tree trunk. A. cyhotes willingly leaves its perch to escape on the ground, and A. chlorocyanus retreats whenever possible up into the top of the tree. Daily Activity. Since I did not spend long periods of time watching any one lizard or groups of lizards, I have no detailed information on the variations in activity during the day. All three species seem to be strictly diurnal. They w^re seen sitting in the sun more frequently in the early morning than at other times of the day, presumably to raise the body tempera- ture to the preferred level. A. (listicJius, at least, seems to feed more actively in the morning than at any other time. On several occasions I captured one with food in its mouth and three times I had one snap at the knot of the thread noose with which I was trying to snare it. Sleeping. I do not know where A. clisticlnis spends the night but, since individuals sometimes use holes in trees as hiding places in the early mornings, they may possibly use these during the night. Two individuals of A. cyhotes Avere found asleep at night, in the field. Both of them were males and were asleep on top of the foliage of the outermost twigs of small bushes. They were plainly visible from my vantage point, though perhaps not to a predator, such as a snake, climbing the bush. In the laboratory one evening during an examination of the room, I located seven cyhotes asleep. Five of them were in the needles at the tips of the branches of the pine bush, one on a small vine where it stretched away from the tree trunk, and the last on top of the topmost leaf of a large-leaved potted plant. Shaking a branch on 12 BREVIORA No. 154 which a cyhotes was sleeping woke the lizard but it did not move until the branch was shaken vigorously or the lizard touched. Then it jumped immediately to the floor and remained there un- moving. A. chlorocyanus was seen asleep only in the laboratory where during the period mentioned above I located six animals. Three of them were sleeping between large leaves of the rubber plant and the wall, one was between a board and the wall, one inside a cold radiator, and only one exposed, on top of its perch. When disturbed, the chlorocyanus immediately sought new hiding places. As Mertens notes, A. distichns and A. cyhotes sleep with their liind legs partly flexed, while chlorocyanus usually sleeps with them extended along the tail. I do not know enough about the real and potential nocturnal predators of Anolis to speculate on the adaptive significance of these sleeping places. But it is striking that they are so different from the situations in which the animals spend the day. Male cyhotes, which in the daytime live on substantial vertical surfaces, during the night sleep on the small flexible foliage and twigs of bushes and vines. A. chlorocyanus during the day is usually exposed to view and during the night sleeps under or behind some sort of cover. Ue'productwn. Since copulation was observed only in distichus and eggs found only of cyhotes, it is not possible to compare these species with respect to reproductive behavior. I am including these data in the hope that eventually comparative data will be available. Three times pairs of distichus were seen in copulation in early to mid-afternoon on tree trunks from four to six feet above the ground. In no case was the preceding courtship observed. In one case tlie lizards were oriented diagonally up a large tree six feet from the ground. The larger male was on the higher side and on top of the smaller female. One of his front legs was across her shoulders and holding on to the tree in front of her front leg; one of his hind legs Avas across the base of her tail, the toes resting on her thigh and the trunk in front of it. His other legs were spread out holding the tree. His tail was bent under hers and his head Avas resting on but not biting her neck. Her position was that of any resting lizard except that her tail was strongly arched. The position observed in the other cases was virtually identical. 1962 THREE niSPANIOLAN ANGLES 13 SUMMAKY AND CONCLUSIONS These observations on the behavior and ecology of these three species of lizards, Anolis distichus, A. cyhotes and A. chloro- cyanus in the Dominican Republic are obviously incomplete. However, certain tentative conclusions can be advanced. These species are sympatric and occupy the same macrohabi- tats over much of their ranges. The species differ somewhat in their microhabitats but they are not clearly separated spatially in this way; members of each species occur on the same tree within feet and sometimes within inches of each other. They all have the same basic body form, though differing somewhat in size, in proportions and in morphological adaptations such as development of the enlarged subdigital pads. They are all in- sectivorous in diet. However, in the details of their behavior they are very different. These details can be fitted together to form a picture of three morphologically similar species of the same genus living together in the same habitat but living three very different lives. Reviewing these very briefly, Anolis distichus is strictly an animal of the lower tree trunks. Its territorial pattern results in .spacing out the individuals living on the same tree with the largest at the bottom and the smallest higher up. This species feeds on small insects on the trunk, utilizing a posture and pat- ern of activity that enables it to forage effectively in this sort of place. \Yhen frightened it takes evasive action on the tree trunk, and copulation takes place there. A. cyhotes, though partially an animal of the tree trunks, is also closely associated with the ground below. Its territorial behavior results in a spatial distribution with the large indi- viduals on the trees and the smaller ones on the ground. Even for the indi\dduals on the trees the perches there seem to be pri- marily lookouts from which to survey the ground. The postures and patterns of activity seem fitted best for this and much less so for watching the tree on which the lizards sit. Certainly they do go to the ground to capture food spotted from these perches. They run to the ground when frightened and they bury their eggs in the ground. A. chlorocyanus in its behavior seems as closely related to the tree tops as cyhotes is to the ground. It seems to be mainly the large males that descend the tree trunks. Some feeding un- doubtedly occurs high in the trees even in these large males and 14 BREVTORA No. 154 chlorocyanus retreats upwards "when frightened. Its slow de- liberate movements seem adapted to the more precarious arboreal footing among the leaves and twigs. These are not just three similar animals doing the same things in the same way in slightly different places but three similar animals, each with a unicpie set of complex behavior patterns which interlock functionally so that each species has its own way of life within the same habitat. This situation can be an example of the operation of the Gause- Volterra hypothesis that closely related species can live together only if they differ in ecology. Since many of these differences can be correlated with feeding behavior they can be interpreted further as serving to reduce interspecific competition for food. However, the demonstration that these conclusions are valid must await the collection of further data, particularly data on just what environmental factors act to limit the population densi- ties of each species. LITEEATURE CITED Cochran, D. M. 194]. Tlie heipetology of Hispaniola. Bull. U. S. Nat. Miis., 177: 1-.398. COLLETTE, B. 1961. Correlations between ecology and morphologj- in anoline lizards from Havana, Cuba and southern Florida. Bull. Mus. Comp. ZooL, 125: 137-162. Grant, C. 1940. II. The Reptiles. In Lynn, W. G. and C. Grant, The Hcrpetology of Jamaica. Bull. Inst. Jamaica, Sci. Ser., 1 : 61-148. Ingek, R. F. 1959. Temperature responses and ecological relationships of two Bor- nean lizards. Ecology, 40 (1): 127-136. Meetens, R. 1939. Herpetologishe Ergebnisse einer Reise nach der Insel Hispaniola, Westindien. Abb. Senckenb. naturf . Ges., 449 : 1-84. Oliver, J. A. 1948. The anoline lizards of Bimini. Anier. Mus. Novitates, no. 1383: 1-36. RUIBAL, R. 1961. Thermal relations of five species of tropical lizards. Evolution, 15: 98-111. Schmidt, K. P. 1928. Amphibians and land reptiles of Porto Eico, with a list of those reported from the Virgin Islands. Sci. Survey Porto Rico and the Virgin Islands, 10, part 1: 1-535. 1962 THREE HISPANIOLAN ANGLES 15 Stejneger, L. 1904. The herpetology of Porto Rico. Rep. T. S. Nat. Mus., 1902: 553-724. Williams, E. E. and A. S. Rand 1961. Notes on Hispaniolan herpetology. 2. A review of the Anolis semilineatus group with the description of Anolis cochranae, new species. Breviora, no. 135: 1-11. BREVIORA Mmseiiam of Connparsitive Zoology Cambridge, Mass. April 12, 1962 Number 155 NOTES ON HISPANIOLAN HERPETOLOGY 6. THE GIANT ANGLES By Ernest E. Williams INTRODUCTION Mertens (1939) has called attention to the existence of geo- graphic variation in the giant anoles of Hispaniola and has distinguished a typical western race, Anolis ricordii ricordii Dumeril and Bibron, and an eastern race, A. r. haleafus Cope. The distinction between these two forms is sharp and un- equivocal ; the situation is, however, more complicated than Mertens' limited sample (16 specimens) led him to believe. Study of the unreported series of Hispaniolan giant anoles in the American Museum of Natural History (AMNH) plus the specimens in the Museum of Comparative Zoology (MCZ) and the United States National Museum (USNM) (91 specimens in all) makes it clear that at least three vicariant forms are recog- nizable. The form occurring from Port-au-Prince north to Cap Haitien and Port-de-Paix is the one to which Mertens has shown that the name ricordii Dumeril and Bibron must be attached. Another, occurring in the north and east of the Dominican Republic, may be called, following Mertens, by Cope's name haleatus (Eupristis taleatus Cope 1864, Proc. Acad. Nat. Sci. Phila., p. 168, type locality "St. Domingo"). A third unnamed population occurs on the Barahona peninsula. A fourth popu- lation inhabiting the southwest peninsula of Haiti may be dis- tinct. I list below the distinguishing characters of the three well-marked forms. BREVIORA No. 155 ricordii Very low nuchal and dorsal crests Nuchal crest scales as long as or longer than high, and not or but little higher than the very low dorsal crest Head scales small, numerous (7-9 across snout at level of second canthal 2) In S S deep hlacTc spots above shoulder, sometimes also on occi- put ; no other evident pattern Table 1 Barahona population Low nuchal and dorsal crests Nuchal crest scales higher than long but small, not higher than the weakly developed dorsal crest Head scales larger, fewer (4-6 across snout at level of second canthal) Whole of body in both sexes with very irregu- lar small Matches and mottlmg haleafvs ^ A prominent nuchal crest, a variable but lower dorsal crest Nuchal crest scales much higher than long, always much higher than scales of dorsal crest Head scales large, few (2-5 across snout at level of second canthal) Both sexes with no evident pattern or transverse handing or reticulation CHARACTER ANALYSTS The scales on the snout are swollen, bosslike in all Hispaniolan giant anoles. The differences are solely in the size of these boss- like scales. The contrast is extreme in this regard between the giant anoles of northern Haiti {ricordii) and those of the north- ern Dominican Republic (haleatus) . The animals of the Bara- hona and southwestern peninsulas, however, are intermediate, overlapping in this regard the north Dominican (haleatus) and approaching the north Haitian populations. On this character alone it is not possible to separate every specimen of the north Dominican and the southern populations though there is a well- marked average difference. Number of scales across the snout is a measure of a more general feature — general scale size — - which is somewhat greater in haleatus than in ricordii. Mertens has cited a number of key regions, i.e. loreal region, base of the tail, etc., which show this. 1 A. Salle, who collected the type of Eupristis haleatus worked both in the northern Dominican Republic and at the base of the Barahona peninsula (map of travels compiled by W. J. Clench). It was therefore necessary to confirm the application of the name by checking the characters of the British Museum type. Miss A. G. C. Grandison has courteously confirmed that the nuchal crest soales of the type are indeed significantly higher than the dorsal crest scales. 2 I count the canthals forward from the anterior border of the orbit (see Fig. 1). The reverse count — from the anteriormost canthal back — is sometimes used, e.g. Oliver (1948). This, however, has the disadvantage that the scales here are small and variable and do not provide a stable starting point. 1962 HISPANIOLAN GIANT ANGLES 3 The snout, however, provides the clearest expression of this general feature and one which can be reduced to a simple count with an adequate numerical range (2-9). Mertens has used a similar transverse count but makes it directly in front of the eyes. I have chosen a transverse count somewhat further for- ward because this transverse line, anterior to the supraorbital semicircles seems to me more suitable as a place for a standard count utilizable for all species of Anolis. N,S- BflLEflTUS RICORDII Fig. 1. The two extremes of anterior head squamation in Anolis ricordii subspecies. The arrows indicate the place at which the count across the snout is to be made. The pattern of geographic variation of nuchal crest scales is not so simple. Again, the populations of Port-au-Prince and northern Haiti and northern Dominican Republic show the extremes, the crests of Port-au-Prince and of north Haiti popu- lations very low with a long anteroposterior base, while those of north Dominican Republic populations are tapering, high, short based anteroposteriorly. The southern populations could again be described as intermediate but on careful examination this is not quite accurate. Both the north Haitian and southern populations have a reduced nuchal crest as compared with the north Dominican population but the nature of the reduction is different. The nuchal crest scales of north Haitian animals are long-based, low, rounded scales, not at all tapering; they are 4 BREVIORA No. 155 sometimes shorter but usually slightly higher than the very low, long dorsal erest scales. In the giant anoles of the Bara- hona peninsula tlie nuchal scales are, in contrast, still short anteroposteriorly, relatively high, tapered, but very small, not appreciably higher than the tallest dorsal crest scales and no- tably smaller in area than the largest dorsal crest scales. The few specimens from the southwestern peninsula do not permit adequate analysis of the populations of this important geographic area. The four adults examined all come from the vicinity of Fond des Negres halfway along the peninsula. In squamation, these are very like the animals of Port-au-Prince and north Haiti. The one remaining specimen is a juvenile (78 mm snout-vent length) from the foothills of the Massif de la Hotte. Although it is less than half-grown, its dorsal and nuchal scales are appreciably higher and less broad based than those of the adults from Fond des Negres and indeed are very comparable to those of the Barahona population ; in addition, in this individual these scales are double and even triple even on the nape. It thus seems very possible that the populations of the middle of the southwest peninsula and those of its extreme end differ significantly. (See further discussion below.) The nuchal crest scales seem in all populations to be some- what variable. HoAvever, within each adequately-sampled popu- lation the range of variation is very characteristic and does not blur the distinctions tabulated above. Despite statements by Boulenger (1885) and by Mertens (1939), I do not find a clear correlation with sex ; females, for example, of haleatus do not consistently have lower nuchal crest scales than male haleatus of the same size. There is, however, clear ontogenetic change ; the smallest individual of haleatus at hand (AMNH 28651 from San Juan Bay, Samana, 41 mm snout-vent length) has the nuchal crest scales only incipiently enlarged and could not be recog- nized as a member of the haleatus population on that character. However, in haleatus the characteristic, tapering, spinelike nu- chal scales develop very early and even specimens little more than half grown (e.g. MCZ 57719, Santiago, S , 83 mm snout- vent length ; MCZ 5445, Samana Peninsula, S , 88 mm snout- vent length) are very readily recognizable. Dorsal crest scales are more variable than nuchal crest scales, and variable in a peculiar fashion : there is sometimes a regular alternation of relatively high triangular single scales and pairs 1962 HISPANIOLAN GIANT ANGLES 5 of miicli lower, more quadrangular scales.^ However, such a regular alternation occurs — if present at all — on only a por- tion of the back and there is present between nuchal crest and dorsal crest, on the one hand, and dorsal and tail crests, on the other, various irregular conditions with double and single crests erratically intermingled and scale types somewhat inter- mediate and rather irregularly so. In ricordii itself with a very low dorsal crest the double condition with all low scales is most frequent ; in haleatus the situation is individually very variable ; in the Barahona and the southwest peninsulas the double condi- tion of the dorsal crest predominates. The body color of live specimens and well-preserved alcoholics is probably useful. The usual specimen, however, requires much interpretation and its evidence must be received with some skepticism. In tabulating pattern, above, I have ignored all ill-defined discolorations and have tried to assess pattern on the basis of real aggregations of pigment rather than fortuitous darkening of random areas by formalin. Even with this qual- ification the problem is not simple : these are Anolis and have the power of color change ; an adequate discussion of their pat- tern would be possible only if the whole repertoire of color changes were known. I record my information below by popu- lation. Ricordii: Females from Port-au-Prince and north Haiti ap- pear to be plain green above with no markings of any sort. Males from Port-au-Prince are also mainly green (as preserved they may show an obscure, very fine brown reticulation) but are distinguished by a large, very characteristic patch of black above the shoulder but of ill-defined shape and varying extent. There may also be irregular black patches extending onto the back of the head. There may also be a white patch at the corner of the mouth. A Cap Haitien male (Senckenberg 10445) is quite similar to Port-au-Prince specimens. Of three males newly re- ceived from Ti Guinin a little to the east of Cap Haitien (MCZ 66147-9) two have almost no black at the shoulder (only small and inconspicuous spots) and none at all on the head. The other male is as devoid of black spotting as any female. All show fine brown reticulation quite like that of similarly preserved Port- au-Prince males. 1 In the tabulation above I have compared the nuchal crest scales with the highest crest scales at midbody. 6 BREVIORA No. 155 Balcatus: Schmidt (1921, p. 10) records specimens from Sanchez and Villa Rivas as "usually green with dark-edged transverse bands of light greenish yellow. ' ' Preserved specimens which have any definable pattern show this transverse banding which is present on dorsum, limbs and tail; it appears to be most persistent on the tail. It is very conspicuous in a specimen near hatchling size (AMNH 28651). There are frequently num- bers of small spots on the sides of the belly and on the venter. The MCZ Santiago series, though preserved in a light phase, show very irregular light banding much invaded by dark retic- ulations, but the bands are very prominent on limbs and tail. Barahona Population: I have unfortunately been unable to find any description of Barahona giant anoles in Hassler's notes. I must therefore rely entirely on the preserved specimens, which, however, are unusually consistent, though from several localities and with very different collection dates. Collected by Hassler, they differ strikingly from the taleatus specimens, most of which were also collected by Hassler. All Barahona specimens, and most clearly those from Valle de Polo, show irregular small blotches, very variable in tint and extent, scattered over the entire dorsum. The closest approach to this condition is seen in those haleatus which show reticulation. There are never in the Barahona specimens the large shoulder spots of ricordii, sensu stricto. One specimen (AMNH 51241 from Enriquillo) does show transverse banding, but more obscurely than in haleatus. The small blotches are also very obscure in this specimen, which therefore could not in this phase be recognized as a member of the Barahona population on its color. The nuchal and dorsal crest scales are, however, typical harahonae, and the hypothesis which I have adopted is that the blotching so characteristic of most preserved specimens is characteristic only of one of the phases of the color repertoire possible to this population. Population of the Southwest Peninsula of Haiti: Again the material is inadequate for proper analysis, the more evidently so since the few available specimens suggest that this is not a unit sample. The juvenile from the Massif de la Hotte is, as preserved, reddish or purplish brown with indistinct broken dark longitudinal lines on head and nape and more evident narrow brown transverse bands on the back, six between shoulder and groin, in pairs with very slightly lighter reddish-brown be- tween each pair. The tail is obscurely annulate. 1962 HISPANIOLAN GIANT ANGLES 7 111 contrast, the three adult females from the vicinity of Fond des Negres are essentially patternless (MCZ 66016 is dark and obscurely vermiculate, and USNM 72631, 72633 are plain light green except for a subocular half ring of scales that is conspicu- ously white). The single male is similarly nearly patternless but has also the subocular half ring of white and, in addition, a series of small black spots above the shoulder, very like a vestige of the large black shoulder spot of Port-au-Prince ani- mals and very like those of the two Ti Guinin males. TAXONOMIC EVALUATION There appear to be at least three distinctive populations of giant anoles, all readily separable by nuchal crest development and body color, less sharply separable by the size of the scales on the snout. These populations are allopatric and thus may be species or subspecies. The differences between typical haleatus and typical ricordii are such that they could well imply specific distinction. How- ever, all the populations south of the Cul de Sac Plain, both the rather distinctive Barahona animals and the very poorly known populations of the southwest peninsula, are to some de- gree intermediate between the two northern extremes although presenting some features that are their ovm. In thus bridging the morphological gap between the extremes, they strongly sug- gest that the Hispaniolan giant anoles belong to a single species. The giant anoles are scarce and local. It is not to be expected that intergradation between the several populations will be easy to demonstrate ; specimens from many of the critical intermedi- ate areas are conspicuously lacking. The Fond des Negres and Ti Guinin specimens, on the other hand, may be intergrade populations. The vestigial shoulder spots present in the two Ti Guinin males and the single Fond des Negres male suggest this conclusion, as does, in the latter case, the combination of nuchal and dorsal squamation most like typical ricordii with somewhat lower scale counts across the snout. No giant anoles are at hand from the northwest Domini- can Republic. This is the area in which intergradation between ricordii and haleatiis is to be expected. It is thus not at all sur- prising to find the first suggestion of loss of characters of the typical race just to the east of Cap Haitien. 8 BREVIORA No. 155 On the southwest peninsula, the Fond des Negres area is one in which such intermediate populations are to be expected, as recent collections from the area show. Thus, Dromicus parvi- frons parvifrons and D. parvifrons protenus appear to meet in this area and several anole races (to be reported later in this series) show intergradation at just this point. I thus interpret the Fond des Negres giant anoles as inter- grades between typical ricordii and a population to the west occupying the tip of the southwest peninsula. This western popu- lation I infer to be represented at present by the single juvenile specimen collected by P. J. Darlington in the foothills of the Massif de la Hotte. The wide-banded specimen from Enriquillo in Barahona is reminiscent of the banded color phase in typical haleatiis, but in squamation it is not intermediate and it is geographically quite unsuitable as a member of an intergradient population. Nevertheless, the presence of this color phase hints at a closer relation with ialeafus than otherwise could be inferred. The absence of other evidence or hint of intergradation is easil}^ accounted for by the gaps in the distributional record, and, granted the desirability of further evidence, recognition of subspecies status seems justified for typical ricordii, for baleatus, for the Barahona population and probably also for a population at the west end of the southwest peninsula of Haiti. The latter two populations are currently nameless. That from the southwest peninsula is at the moment very insufficiently known, and it would not now be appropriate to describe it. The Barahona population, on the other hand, is well recorded and may be formally named: Anolis ricordii barahonae new subspecies Type: MCZ 43819, Polo, Valle de Polo, Barahona, Dominican Republic, an adult female collected by W. G. Hassler, Septem- ber 1932, J. C. Armstrong donor. ParaUjpes: Valle de Polo, MCZ 56141, AMNH 51036, 51235-6; Herman's Finca near Paradis, AMNH 51231-3; Barahona, AMNH 50255-6, 50261 ; Halfway between Trujin and Enriquillo, AMNH 51230; Enriquillo, AMNH 51241. Diagnosis: A subspecies of ricordii distinguished from the typical form and from baleatus Cope by the nature of the nuchal crest (formed by small but slender tapering scales not or very 1962 HISPANIOLAN GIANT ANGLES 9 little higher than the highest scales of the dorsal crest), by the size of the scales of the snout (4-6 across snout between second canthals), and by a characteristic phase of coloration in both sexes in which small blotches are present, irregular in shape and of varying intensity. List of Specimens Examined: A. r. harahonae (17 specimens). DOMINICAN KEPUBLIC. Polo, Yalle de Polo: ]\ICZ 43819; Valle de Polo: AMNH 51036, 51235-7, MCZ 56141; Barahona: AMNH 50255-6, 50261; Herman's Finca near Paradis: AMNH 51231-3 ; Halfway between Trujin and Enriquillo : AMNH 51230 ; Enriquillo: AMNH 51241; locality uncertain: AMNH 51229, 51234, 51240. A. r. ricordii (12 specimens). HAITI. Port-au-Prince : AMNH 49501; Diquini: MCZ 8619, USNM 118902, 123988; Source Le- clerc, Morne Decay ette, near Port-au-Prince: MCZ 65729-31; PetionviUe: MCZ 60013-4; 3Iarcaco: USNM 69437; Port-de- Paix: MCZ 63338; Ti Guinin near Cap Haitien: MCZ 66147-9. [Records by Mertens : Port-au-Prince, 4 specimens; Cap Haitien, 1 specimen.] A. r. haleatus (57 specimens). DOMINICAN REPUBLIC: Pena, near Santiago: MCZ 57713-9; Sayitiago: MCZ 7831; Las Bracitas: AMNH 41465-6; El Rio: AMNH 41294, USNM 62104- 5; Rio San Juan: USNM 74940-1; Samaria Peninsula (various localities) : MCZ 5445, AMNH 28651, 39807-15, 39817-23, 39825- 9, 39837, 40224-30, 40387-90, 42285, 42775, 44841-4; La Romana: MCZ 16321. [Records by Mertens: Santiago, 5 specimens; Moca, 4 specimens; Finca Arhol Gordo near Villa Altogracia, 1 speci- men.] A. r. subsp. nov. (1 specimen). HAITI: Foothills, Massif de La Hottc: MCZ 38217. A. r. subsp. nov. x r. ricordii. Pemel near Fond des Negres: MCZ 66015-6 ; Fond des Negres: USNM 72631, 72633 ; also 72632, skeletonized. BEHAVIOR The giant AnoUs of Hispaniola is infrequently seen by collec- tors and consequently our information on its ecology and habits is limited. Such information as is available is summarized below by A. S. Rand. Quotation marks refer to Rand's personal ob- servations. 10 BREVIORA No. 155 A. ricordii ricordii, Morne cle Cayette, near Port-au-Prince, Haiti (A. S. Eand and J. Lazell). " A large individual was seen about 15 feet up on a branch of a 30-foot high tree on the edge of a patch of brush on the hillside. It remained sitting quietly in sight for about half an hour until we climbed the tree to attempt to capture it. It then climbed out among the small branches where, desi3ite an intensive search, we were un- able to locate it again." Port-de-Paix, Haiti (A. S. Rand and J. Lazell). "A female was seen sitting head down on the trunk of a large tree about 4 feet from the ground. The tree was one of a series forming a fence row. The lizard allowed us to approach and noose it." A. ricordii haleatus, Pena near Santiago, Dominican Repub- lic (C. E. Ray and A. S. Rand). "Local people who brought in a series at this locality reported that they lived among the upper branches of the larger shade trees in the coffee plantations." Finca Arbol Gordo, Dominican Republic. Mertens reports that the single specimen collected by him jumped to the ground from the trunk of a palm tree. Mertens kept this animal in cap- tivity for some time and wrote that it was active only in the late afternoon. When approached it squirreled around to the other side of its perch. It slept exposed on small branches with its hind legs flexed. A. ricordii subsp., Camp Perrin, Haiti (A. S. Rand and J. Lazell). "A small boy guided Rand to a stand of coffee where a crowd of Haitians had treed a large individual. The lizard was about 20 feet up in a small tree. It was too high to noose and the tree too small to climb so we attempted to scare the lizard to a more favorable situation by poking it with a long pole. After considerable fuss the lizard ran out along a small branch to the next tree, down that to the ground, across the ground to hide among the roots of a large Cieba tree. After considerable prodding among the roots we dislodged the lizard which avoided the herd of assistants successfully and climbed the Cieba out of both sight and reach." Above Maceline, near Camp Perrin (A. S. Rand and J. Lazell). " In a small stand of coffee we saw a single animal about 15 feet up in one of the large trees on a large branch. When Lazell climbed the tree it squirreled about the branch concealing all of itself except one eye. It then climbed up and out among the small branches and disappeared from ^•iew." 1962 HISPANIOLAN GIANT ANGLES 11 Mertens speculates that the giant anoles of Hispaniola were originally inhabitanis of the lowland rain forest and have adapted to cultivated areas by living in the crown of the taller trees. The observations of Kand and Lazell tend to support this conclusion. INTRAISLAND ZOOGEOGRAPHY The portion of Hispaniola south of the Cul de Sac Plain, in- cluding the southwest and Barahona peninsulas and the La Selle and La Hotte ranges, has, as Mertens has emphasized, a number of endemic species and even genera. I have recently (Williams, 19(jl) stressed the importance of this area, which I have called the "southern island," in the initial ditferentiation of forms now widespread in the portion of the island north of the Cul de Sac Plain. Barahonae as an isolate in "southern island" easily fits this pattern. Baleatus and typical ricordii do not. This is merely the first of many examples to be discussed in this series which will show that not all the patterns of differ- entiation in Hispaniola can readily be explained by the simple hypothesis of "southern" and "main island" isolates. As in the present case so in a considerable number of others, there appear to be additional loci of differentiation, e.g. northern Haiti, which are not so evidently marked off' by present or recently past barriers to distribution. These cases of subsidiary loci of differentiation are not at present well analyzed. Hispaniola has not been really well col- lected, and it is now evident that the differentiation of local popu- lations that occurs within it is finer grained than the sporadic collecting of the past could reveal. For many of the instances of minor loci of differentiation, data are just now being gathered. RELATIOXSHIPS OF EICORDII AND THE OTPIER WEST INDIAN GIANT ANOLES Anolis ricordii clearly belongs to the series of giant anoles that includes equestris of Cuba and cnvicri and rooscvclti of the Puerto Rican-Virgin Island complex. Aiiolis garmani of Jamaica seems, even on externals, to show no features, except size, which specially link it to this group, and Richard Etheridge in an unpublished thesis (University of Michigan) has shown that osteologically it belongs to quite a dift'erent section of the genus. 12 BREVIORA No. 155 No full study has ever been made of the Greater Antillean giant anoles as a group, though they clearly merit such a study. ^ I do not intend to attempt this here, but in order to place the races of ricordii within a frame of reference which might add to our understanding of them and of their origin I have compared all of the species for a number of mostly qualitative characters which were simple to determine and to evaluate. Table 2 presents the results of this comparison. It is sufficiently clear even from the limited evidence presented by Table 2 that we are dealing with four good species. The Hispaniolan populations are a unit as compared with the forms of the neighboring islands and, as in the case of the other poly- typic giant anole, A. equestris, the intraisland differentiation has involved primarily scale size and color pattern. There is one unusual zoogeographical feature in the distribu- tion of the giant anoles — the occurrence of A. cuvieri and A. roosevelfi on the same island l)ank, cuvieri on the mainland of Puerto Rico, roosevelti on Culebra in the Virgin Islands. These two are very distinct species. Their differences are at another level than those between ricordii and haleatus, and their occur- rence on what must at various times during the Pleistocene have been a single land mass provides a special problem in accounting for their origin and dispersal. ACKNOWLEDGMENTS This paper was in part made possible by collections made under National Science Foundation Grants NSF G-5634 and G-16066 and grants from the American Philosophical Society and the Society of Sigma Xi. Assistance given by the govern- ments and people of the Haitian and Dominican Republics is also gratefully acknowledged. I am indebted also to Charles M. Bogert and Doris M. Cochran for loans from the collections under their care. Dr. Norman Ilartweg permitted examination of the second known specimen of Anolis roosevelti Grant. Dr. W. J. Clench provided information on the areas visited by A. Salle in the Dominican Republic. 1 For further information on the non-Hispaniolan species see Stejneger (1904), Grant (1931), Schwartz (1958). 1962 HISPANIOLAN GIANT ANGLES 13 S X 13 3 o i^ CO o •■o AS > I ^ CC -G GQ O) a lO ^ ^ CO o o o < — I © 32 rs .9 cS rt ii O *o 60 c 00 « o 05 oj c: a ^ oj 1> ^ i: -S 02 '^ p: m « .s a S m CO CO a a CO u ^^ '-; o3 (M "3 o «o 02 o 02 -*— S ^ -M fl -4^ 3 S bJO "T? 02 O Ol TiH 3 02 3 o o i CO 9 o o CD CD S > 0) CO a a CO a? 03 IE o ^ O ^ -a rS M QJ O ■3 3 ^ 03 CD a l-H ;h « S5 S O • rH g 3 OQ ^ 05 03 =H 03 73 03 p -3 -4^ 03 ;-i 1— H 3 'rt T! 3 a 2 a cS cS 3 -^ 02 a iJ 02 ci 14 BREVIORA No. 155 REFEEENCES CITED BOULENGER, G. A. 1885 Catalogue of the lizards in the British Museum (Natural His- tory). London, vol. 2, i-xiii + 1-497. Grant, C. 1931 A new species and two new subspecies of the genus Anolis. Jour. Dept. Agr. Puerto Rico, 15: 219-222. Mertbns, R. 1939 Herpetologische Ergebnisse einer Reise nach der Insel His- paniola, Westindien. Abhandl. Senckenb. Naturf, Ges., 449: 1 84. Oliver, James A. 1948 The anoline lizards of Bimini, Bahamas. Amer. Mus. Novitates, no. 1383: 1-36. Schmidt, K. P. 1921 Notes on the herpetology of Santo Domingo. Bull. Amer. Mus. Nat. Hist., 44: 7-20. Schwartz, A. 1958 A new subspecies of Anolis equestris from Eastern Cuba. Herpe- tologica, 14: 1-7. Stejneger, L. 1904 The herpetology of Puerto Rico. Rep. U.S. Nat. Mus., 1902: 549-724. Williams, E. E. 1961 Notes on Hispaniolan herpetology. 3. The evolution and rela- tionships of the Anolis semUineatus group. Breviora, No. 136 : 1-8. 1962 HISPANIOLAN GIANT ANGLES 15 5D 03 10 ^i) o C> ^ n, v en <) ;^ ^^ VK- 6 ai ^ 73 ^ P3 O N- O Tl s- MO +J C3 o g O lO J3 — ' b Tl MJ o r* C! t. o .^ o 8 03 !^ fO 1-1 a D O fi ►^ 12; BREVIORA seum of Comparaitive Zoology Cambridge, Mass. April 13, 1962 Xumber 156 THE FOSSILIFEROUS TRIASSIC DEPOSITS OF ISCHIGUALASTO, ARGENTINA, AND PRELIMINARY DESCRIPTION OF ISCHIGUALASTIA, A NEW GENUS OF DICYNODONT By Alfred Sherwood Romer and C. Barry Cox THE FOSSILIFEROUS TRIASSIC DEPOSITS OF ISCHIGUALASTO, ARGENTINA By Alfred Sherwood Romer Early in 1958, a joint expedition of the Museo Argentino de Cieneias Naturales of Buenos Aires and the Museum of Com- parative Zoology, Harvard University, was organized to explore continental deposits in the western Argentine in search of fossil vertebrates. The personnel included, for the Buenos Aires Mu- seum, Dr. Guillermo del Corro, geologist, and Mr. Orlando Gutierrez, assistant ; from Harvard, Professor Bryan Patterson, preparators Arnold C. Lewis and James Jensen, Mrs. Romer and the writer. The earlier part of the trip was spent in the general region of Mendoza. In late April it was decided to move our base northward and explore the valley of Ischigualasto. This region lies some 300 miles to the northeast of Mendoza, in eastern San Juan Province, close to the border of La Rioja, in the department of Yalle Fertil, at a latitude and longitude of approximately 68°W, 30°5'S. No adequate maps of this area exist; Ischigualasto is shown on the San Juan sheet (no. 37) of the 1/500,000 map of Argentina, but the details are highly inaccurate. Frenguelli (1948, fig. 1, pis. 1, 2) gives sketch maps showing the general topography of the valley and its relation to adjacent regions. The name derives from an Indian village once 2 BREVIORA No. 156 present at the marofin of the valley. This, however, disappeared long since and the valley is oompletely uninhabited. In 1874 Stelzner (1885, pp. xii, 74-75) rapidly traversed the valley on mnle-back, presumably following a trail which runs from Jachal, in San Juan Province, through the valley and on eastward to Patquia in La Rioja. Bodenbender (1911, pp. 94-113) similarly crossed by this trail several decades later. A small coal seam had been discovered in the hills west of the valley at a locality named los Rastros because of dinosaur-like footprints discovered there; Huene (1931) was taken, in 1927, on a hasty trip to see these footprints ; he traversed the valley but was unable to stop there. Until recent years nothing was known of the geology of the region except for such observations as Stelzner and Boden- bender could make as they crossed it, and nothing of fauna and flora except the footprints just mentioned and a few plants col- lected by Bodenbender. The reason for this former paucity of knowledge is obvious. The valley is extremely arid, and in the days of mule travel, a stay there was impossible because of the almost complete lack of water and fodder.^ With the coming of automobile transporta- tion and, particularly, of vehicles with four-wheel drive, the situation has changed radically, and in 1948 Frenguelli was able to publish a rather comprehensive account of the stratigraphy of Ischigualasto as the result of exploration of the area in 1943, primarily in the interests of paleobotany. During the course of his work there, three fragments of cyno- dont skulls and jaws were recovered, and were described by Cabrera (1943). These were of interest as indicating the pres- ence in Argentina of Triassic beds similar in fauna and presumed age to those of southern Brasil. But since so little reptile ma- terial was collected during Frenguelli 's survey of the region, it seemed unlikely that work there would be profitable for the ver- tebrate paleontologist. My attention was first called to Ischigualasto by Professor Huene before the results of Frenguelli 's work became known. As noted above, Huene had crossed the valley on mule-back during his inspection of the los Rastros footprint site and, al- though unable to stop and prospect, had been struck by the 1 Bodenbender (1911, p. 96) breaks off in tbe middle of his description of the region to insert a paragraph of emphatic and italicized warning: "NOT A — A los futuros exploradores hago presente que no se puede contar . . . con sufl- ciente pasta para los anUnales," etc. 1962 TRIASSIC DEPOSITS OF ISCHIGUALASTO 3 seemingly favorable nature of the beds for vertebrate explora- tion. Like our Argentinian colleaj^ues, I had failed to be im- pressed as to prospects by the few materials obtained by Frenguelli. A different light, however, was shed upon the possi- bilities by a publication by Heim (1949). He had been commis- sioned in 1944 to report on the los Rastros coal mine, but in crossing the Ischigualasto valley had become so interested in the region that a considerable part of his report is devoted to the Ischigualasto beds and, incidentally, to the fossils found there. He photographed a cynodont skull in situ (pi. VIII, fig. 2) and says (p. 22) : "El senor de la Vega me llamo la atencion sobre los restos de vertebrados que se han hallado frecuentemente en el suelo de las arcillas, especialmente en la parte media de la formacion. Se presentan como acumulaciones de fragmentos de huesos, de color pardo oscuro. A veces aim se distingue la forma del craneo y de la dentadura con restos de dientes negros, pero sin que se pueda restaurar esqueletos." This report indicated that investigation of the Ischigualasto region might prove profitable, and our expedition moved to that region. A road of passable quality runs to Valle Fertil, a village south of the valley, which serves as departmental capital. Be- yond this, settlement is sparse and presently ceases altogether, the roads deteriorate into a confusing series of trails, and it was necessary to hire a local guide to reach our destination.^ At Cerro Morado, about 50 km. north of Valle Fertil, there is reached the southern edge of the valley which extends some 20 km. to the north-northwest. At its southern end its width is about 7 km.; it narrows to 21/2-3 km. at the north. Its western boundary is a range of rugged sandstone hills, the eastern boun- dary an unbroken cliff of red sandstone, "los Colorados," strik- ing in appearance, with a height, for much of its length, of 150 to 200 meters. The valley itself, almost bare of vegetation except for scattered thorn bushes, exhibits an expanse of shales of varied pastel colors and occasional sandstones. It is drained by a series of broad sandy arroyos which converge at about the center of its western margin where, at Agua de la Pena, there is a small stream of water (non-potable, according to our guide). From this point the deep gorge of the Rio de la Peiia runs west- ward through the hills to the Rio Bermejo. Northwestward from 2 Since our work there, a new highway has heen constructed that links Patquia with Pogancilla and Chilecito. This passes a little to the northeast of the Ischi- gualasto basin, rendering it comparatively accessible. 4 BREVIORA No. 156 the main valley extends the narrower valley of the Rio de la Chilca, which turns westward to join tlu> Bermejo. We made camp near Agua de la Pena, began exploration of the valley — and were immediately astounded by the abundance of vertebrate remains which it contained. Every vertebrate paleontologist dreams of finding, someday, a virgin territory strewn with fossil skulls and skeletons. Almost never does this dream come true. To our amazement and delight, it did come true for ns at Tschigualasto. All about us, in the clays of the valley, were exposed specimen after specimen of fossil reptiles. In most instances the greater part of the time of a field party is taken up in prospecting for specimens. Here little of this sort of w^ork needed to be done, and our energies could be de- voted to excavating the better specimens chosen from the wealth of materials readily available. Approximately six weeks were spent here. Our work was hampered by the shortness of the day- light hours at this season of the year, and by the fact that a number of trips to Valle Fertil, each taking a full day, were necessary to carry out fossil blocks and obtain food supplies, gasoline and water. Nevertheless, well over 100 specimens, mainly plaster blocks containing skeletons, partial skeletons or skulls, were recovered. The geology of the region has been described by Frenguelli and by Heim, and discussed by Groeber and Stipanicic (1952, pp. 87-99). Apart from faults of a minor nature, the geologic structure is simple, the sediments uniformly dipping gently to the east-northeast with, in consequence, a regular succession of beds from west to east. Toward the Bermejo, in the western hills (an area not visited by us), are beds thought to be of Car- boniferous and Permian age — the Paganzo beds of Bodenbender — followed by the Estratos de Ischichuca.^ Farther east, the hills bordering the Tschigualasto valley are formed by the Estra- tos de los Rastros, dominantly sandstones, which conformably overlie the Ischichuca and include the coal seam and footprint locality mentioned. The shales occupying the valley constitute the Estratos de Ischigualasto, with a thickness estimated at 400 to 500 meters; it is in this formation that nearly all the verte- brate fossils of the region have been found. The steep cliffs at the east side of the valley constitute the lower part of the Estratos de Gualo of Frenguelli, for which Groeber and Stipanicic prefer the designation of Estratos de los Colorados. The four upper 3 The beds beneath Cerro Morado may be the Isehichuca. 1962 TRIASSIC DKPOSTTS OF TS( HIGUALASTO 5 formations, at least, of the series — the Estratos de Isehiehuea, de Ids Rastros, de Ischiguahisto, de los Colorados — are obviously parts of a sinole sedimentary cycle, Avithout evidence of any dis- conformity. In the region of Agua de la Pefia a local fault (Frenguelli, 1948, pi. II. profile A) obscures the transition be- tween los Rastros and Isehignalasto formations. But elsewhere (as Frenguelli 's profiles B-D, pi. IV) deposition can be seen to have been uninterrupted, and (apart from a conglomerate bed which is taken as the upper boundary of the los Rastros forma- tion) the transition is marked mainly by a diminution in im- portance of the sandstones which are so prominent in the los Rastros. The uppermost beds of the los Rastros yielded frag- mentary verteV)rate remains which appear comparable to those of the Tschigualasto formation. The typical beds of the Estratos de los Colorados contrast strongly with those of the Estratos de Ischigualasto in color, predominance of sandstones, and re- sistance to erosion ; but as Frenguelli notes, a zone of transition is apparent at the base of the cliffs. Stelzner "lumped" the entire series of beds found in this re- gion as "Rhaetic" (in the broad sense in which that term has frequently been used in Argentinian geology). Bodenbender believed the "Gualo" beds to be Cretaceous, the Ischigualasto Jurassic, the los Rastros "Rhaetic." Frenguelli, more correctly, considers the series as a whole to be older, the "Gualo" to be Rhaetic or lower Liassic, the Ischigualasto to be upper Keuper, the los Rastros to be lower Keuper. Groeber and Stipanicic (table I) believe the spread in time of deposition of these beds to be rather narrower ; the los Colorados are assigned with doubt to the Rbaetic, and l)oth Tschigualasto and los Rastros to the upper Norian — i.e., the uppermost Keuper. The vertebrate re- mains suggest a lower position (Romer 1960a, pp. 1291-1292; 1960b, pp. 86-87) although full discussion should be postponed until the fauna has been more thoroughly studied. The Norian stage is one in which is found the typical Upper Triassic dino- saurian fauna ; the Ischigualasto fauna is, on the contrary, one in whicli there is little evidence of dinosaurs and in which gomi^hodont cynodonts and rhynchosaurs are dominant. The Ischigualasto formation is essentially comparable to the Santa Maria beds of Brasil and the Manda beds of Tanganyika. It is surely pre-Norian and not improbably pre-Carnian ; the gompho- dont-rhynchosaur faunas would appear to be essentially Middle Triassic in age. 6 BREVIORA No. 156 By agreement, the entire collection was shipped to Cambridge, where it is being prepared. All types and representative speci- mens of all forms found will be deposited in the National Museum in Buenos Aires. Complete preparation, however, will be a lengthy process, not only because of the considerable quantity of material but also because of the refractory nature of the mat- rix enclosing a large proportion of the specimens. In general, publication of the scientific results will not appear for some time, since we do not wish to publish before preparation has pro- ceeded to the point at which all material of a given form has become available for study. In the case of the dicynodonts, how- ever, the material included only a few specimens of a single large form; these have been prepared, and have been studied by Dr. Barry Cox, of King's College, University of London. A prelimi- nary description is appended ; it is hoped that a full account may be published within the year. It is gratifying to us to have been instrumental in opening up a new Argentinian area for exploration by vertebrate work- ers. Since our trip, several further expeditions to the region have been made by the University of Tucuman, under the direc- tion of Dr. Oswaldo Reig, with successful results ; certain ma- terials collected on these later expeditions have already been described (Reig 1958, Casamiguela 1960). There are vast areas of late Paleozoic and early Mesozoic deposits in the western Ar- gentine which have never been visited by vertebrate paleontolo- gists. Although the chances of finding beds as unusually productive as those of Ischigualasto are not great, it is highly probable that other faunas which will aid in rounding out the early history of vertebrates in Argentina await discovery. I have elsewhere (Romer 1960a) expressed our thanks to a number of friends who aided us in the Mendoza region. We are further deeply grateful to various other persons who aided the general work of the expedition and our exploration of Ischigua- lasto. The cordial cooperation of Dr. Adolfo D. Holmberg, then Interventor, and members of the staff of the Buenos Aires Mu- seum, was much appreciated. Professor Rosendo Pascual of the La Plata Museum accompanied us during the early portion of the trip. Much valuable scientific information was given us by the geologists of the Yacimientos Petroliferos Fiscales, Comision Nacional de Energia Atomica and the Direccion Nacional de Mineria, including, among others, Drs. Pablo Groeber, Pedro N. Stipanicic, Martinez Cal, Hector de la Mota, Luis A. Barrio- nuevo, and Vicente Ferreiro. The Comision Nacional de Energia 1962 TRIASSIC DEPOSITS OF ISCHIGUALASTO 7 Atomica aided us in the difficult matter of water and gasoline supply at Ischigualasto. Dr. Mario E. Terrugi aided us greatly in many regards, and the Harvard members of the expedition appreciate very much the hospitality extended to them by Dr. and Senora Terrugi and by the late Dr. Bernhard Dawson and Senora Dawson of La Plata. Our expedition was made possible by grants from the National Science Foundation and Life magazine. REFERENCES CITED BODENBENDER, G. 1911. Con8tituci6n geol6gica de la parte meridional de La Rioja y re- giones limitrofes (Republica Argentina). Bol. Acad. Nac. Cienc. C6rdoba, 19:1-220. Cabrera, A. 1943. El primer hallazgo de terapsidos en la Argentina. Notas Museo La Plata, 8, Paleont. no. 55:317-331. Casamiquela, R. M. 1960. Noticia preliminar sobre dos nuevos estagonolepoideos argen- tinos. Ameghiniana, 2:3-10. Frenguelli, J. 1948. Estratigrafia y edad del llamado Retieo en la Argentina. Gaea, 8:159-309. Groeber, p., and p. N. Stipanioic 1952. Triasico, pp. 13-141 in Geografia de la Republica Argentina. Buenos Aires, Soc. Argentina Estudios geog. Gaea, 2 (1):1-541. Heim, a. 1949. Estudio geologico del Carbon "Retieo" y del Valle de la Pena (Provincias de San Juan y la Rioja). Ministerio de la Industria y Comercio, Bol. Dir. General de Industria Minera, 89:1-31. HUENE, P. 1931. Die fossilen Fahrten un Rhat von Ischigualasto in Nordwest- Argentinien. Palaeobiologica, 4:99-112. Reig, O. a. 1958. Primeros datos descriptivos sobre nuevos reptiles arcosaurios del Triassic de Ischigualasto (San Juan, Argentina). Rev. Assoc. Geol. Argentina, 13:257-270. ROMER, A. S. 1960a. Vertebrate-bearing continental Triassic strata in Mendoza region, Argentina. Bull. Geol. Soc. America, 71:1279-1294. 1960b. Explosive evolution. Zool. Jahrb., Syst. Bd., 88:79-90. Stelzner, a. 1885. Beitrage zur Geologie und Palaeoutologie der Argentinischen Republik. Geologiseher Theil. Capel & Berlin, xxix -f 329 pp. 8 BREVIORA No. 156 PRELIMINARY DIAGNOSIS OP LSCHIGUALASTIA, A NEW GENUS OF DICYNODONT PROM ARGENTINA By C. Barry Cox Department of Zoology, King's College, University of London Above is given a brief account of a joint Buenos Aires-Har- vard expedition to the Valley of Ischigualasto, San Juan Province, Argentina. Among the remains collected from the Ischi- gualasto formation, presumably of Middle Triassic age, were several skulls and parts of posteranial skeletons belonging to a new genus of large dieynodont. A preliminary diagnosis of this new genus follows below; it is named Ischigualastia jenseni after Mr. James Jensen, who was responsible for the extremely pains- taking collection and preparation of this material. Ischigualastia jenseni, gen. et sp. nov. Holotype of I. jenseni: Number 18.055, Museo Argentino de Ciencias Naturales, consisting of skull and partial skeleton. Geological Horizon and Locality: Ischigualasto formation (Triassic), approximately 100 m. above the base of the forma- tion ; about 2 km. north of Agua de la Peiia, Ischigualasto Valley, Department of Valle Pertil, San Juan Province, Argentina. Genotype: Ischigualastia jenseni Cox. Generic and Specific Diagnosis: Large dicjaiodont (type skull 55 cms. long, 46 cms. broad). No teeth in upper or lower jaws. Skull triangular in dorsal view, greatest width across occiput. Very wide interorbital region, very narrow intertemporal re- gion. Tapering snout, without nasal ridges or bosses. No pineal boss, but a slight mound in front of pineal foramen. No post- frontal bone. Preparietal bone probably present. Interparietal forms whole of posterior half of intertemporal bar, widely sepa- rating squamosals from postorbitals. No sharp median inter- temporal ridge. Zygomatic arches bowed outward. Sharp transition between dorsal and occipital surfaces. Occiput almost semicircular in outline. No tabular bone visible. Stapes lacks stapedial foramen. Short interpterygoid vacuity. No ectoptery- goid bone. Pterygoid broadly meets maxilla. Palatine and pre- maxilla meet, excluding maxilla from internal nares. Palatal surface of premaxilla bears pair of anterior ridges. Premaxilla 1962 TRIASSIC DEPOSITS OF ISCHIGUALASTO 9 extends some way anterior to maxilla. Aseendinji: portion of epipterygoid slender, not expanded to form part of lateral wall of braincase. No lateral winji: on dentary. Stout retro-articular process. Five sacral ribs. Acromion })rocess of scapula absent or ves- tigial. Coracoid foramen between precoracoid and scapula. Sternum constricted halfway along its length; dorsal surface bears bosses for attachment of ribs. Ulna has large olecranon process, with cartilaginous epiphysial union with rest of bone. Femur with well-developed head set off from rest of bone. A more extensive and illustrated account of Ischigualastia will appear later. BREVIORA Mmseiuim of Comparative Zoology Cambridgk, Mass. May 2S, 1962 Number 157 A RHACIIITOMorS AMPHIBIAN, SPATHICEPHALVS, FROM THE MISSISSIPPI AN OF NOVA SCOTIA By Donald Baird Department of Geology, Piiiiceton University, Princeton, New Jersey In the past few years field parties from the Museum of Com- parative Zoology at Harvard College, supported in part by funds from the National Science Foundation, have discovered several deposits of fossil amphibians which date from the earlier half of the Carboniferous period. Most of the material recovered comes from redbeds of the I'pper Mississippian Mauch Chunk group in West Virginia (Romer, 1941) and Pennsylvania; more recently, amphibians have been found in beds of equivalent age at two sites in Nova Scotia. In continuation of the Harvard collecting program, field par- ties sponsored jointly by Princeton University and the Nova Scotia Museum of Science have collected additional bones and trackways of amphibians in Pennsylvanian as well as Mississip- pian beds. One of these finds is noteworthy because no member of the order Temnospondyli from the Mississippian system has hitherto been described. This specimen, the right half of a skull table preserved as a natural mold, comes from the same horizon and locality as the embolomerous lower jaw described as PhoUdcrpeton{1) hrcton- cnsc Romer (1958). The matrix is a flaggy calcareous siltstone characterized by well-sorted quartz grains and silvery muscovite mica. 2 BREVIORA No. 157 Su])erorder Labyrinthodontia Order Temnospondyli Suborder RHACHITOMI^ Family LOXOMMIDAE Genus SpATHICEPHALUS Watson 1!)29 [Spathiocephalus Romer 1945, errore] Emended Diagiwsis. A loxomniid extremely specialized in its expanded cheek and snout, constricted interorbital area and shortened skull table, and having numerous small, chisel-shaped maro'inal teeth. e^ SpATHICEPHALUS PEREGER, U. Sp. Diagyiosis. Differs from N. minis Watson, the only other known species, in its relatively narrower skull table. Type. PU 17182, Princeton University Geological Museum. Occurreyice. Point Edward formation, Canso group, Upper Mississippian (probably early Xamurian). Beach of cove be- tween Point Edward and Keating C'ove, 4 miles northwest of Sydney, Cape Breton County, Nova Scotia. Collected by Donald Baird, William F. Take and Jane McN. Take, 1960. DESCRIPTION The dorsal surface of the skull table bears a deep, coarsely reticular sculpture rather similar to that of Megaloceplialus lineolatus (Cope) but with narrower ridges. Aside from the separation of prefrontal from postfrontal, the contacts between skull elements are normal for a loxommid. The narrow frontal l)0iie is Ijounded medially by a deep, straight suture which is ridged rather like the edge of a file. Anteriorly the frontal is beveled to receive the end of the nasal ; anterolaterally it bears a long, striated facet for the articulation of the prefrontal. As in other loxommids the prefrontal must have swelled laterally to form the antorbital process which dif- ^ AssigiiiiU'iit of the T.oxoiiiinidno to the Rliac-hitoiiii is now (•(iiifiiiued 1iy the association of rhat-hitonious vertebrae with a skull of Mc/jaloeepliahis lineolatus (Cope) from Linton, Ohio (cf. Baird, 19.")7). 1962 NEW SPATHICEPIIALUS ferentiates the orbit proper from its anterior extension, the lacri- mal fenestra, and frives the loxommid eye-socket its characteristic keyhole shape. Most of the lateral edge of the frontal forms the thick orbital rim. At its waist the frontal is 4.3 mm. wide, mak- ing- the interorbital distance a mere 8.6 mm. — extraordinarily narrow for a skull with an estimated width of 185 mm. The short parietal extends laterally into a square lappet ; there is no indication that this lappet represents a former intertem- Fig. 1. Skull of Spaihicepltalus pereger, n. sp., x %. Eestoration in flashed liiU'S based on . 489) do not mention any characters of taxononiii- value, anil Filhol's illustration (ibid., fig. 423) is vague and diagranimatie. Iloffstetter (1944, p. 553) considers it possibly a skink, but notes that vertebrae similar to those of Cordylus {sensu lain) occur in the same deposit. Later, he indi- cates (1955, p. 621) that these fossils "rappellent les pieces homolog'i(|ues de Tactucl Cordiihi^," but the assignment is still tentative. The fossil described here is referable to the subfamih^ Gerrho- saurinae on the basis of the large, broadly imbricate, rounded or slightly squared throat scales, contrasting with the much smaller, non-imbrieate, diamond-shaped throat scales of the cor- dylines. Chamaesaura has larger throat scales than other cordy- lines, but they are anteroposteriorly elongate and mucronate, rather than smooth and transversely widened as in gerrhosau- rines. Fig. 2. Gerrhosaurus cf. G. major, Lower Miocene, Kenya, left lateral view, X 1.5. Though the published generic characters of Gerrhosaurus are in soft anatomy and scale details not preserved here (Loveridge, 1942, p. 488; FitzSimons, 1943, p. 268), similarity between this fossil and Recent members of the genus seems to indicate that it belongs here. The throat scales, what remains of the dorsal scales, shape of the ear and lateral fold all agree closely with these characters of the Recent genus. Moreover, a number of resemblances discussed below strongly suggest reference to the Recent species G. major. 6 BREVIORA No. 158 1. Tli(^ fossil has the sizo and o-oneral proportions of a large atlult individual of (i. major. The latter is the lar<>est of the species of (icrrhosaurus ; most other species are considerably smaller. 2. The slia]ie of the opening' of the external eai- in 0. major may be rounded or sliohtly angular dorsally. Tn the available specimens, G. m. major most frequently shows the rounded con- dition, but a. m. (iraiidis usually has a more angular dorsal edge, as in the fossil. 6'. m. hottegoi resembles G. m. major in this character. 3. Tn oerrhosaurs. the shape of the tympanic shield is con- sidered taxonomically significant. This scale lies on the anterior border of the outer ear, and in all species of Gcrrhomurns except (t. major, is thin, flattened, and often expanded to cover and ])rotect the cavity of the outer ear (see e.g. FitzSimons, 1948, figs. 150, 156). Loveridge (1942, pp. 515, 518) states that tym- panic shields of G. fiavigularis are also narrow and band-like as in G. major. This is grossly true, but in detail the two can be distinguish(^d easily. G. flavujnlaris has a narrowly crescent- shaj)ed tympanic shield (see e.g. FitzSimons, ihid., fig. 154; cf. fig. 164 of (r. m. (jra)idh) which is thin and flattened, w4iile that of G. major is strap-shaped, and thickened. 4. Another similarity to G. m. grandis is the presence, on the posterior border of the outer ear opening, of a small anterior row of scales, flanked by a larger posterior row. Tn G. m. major and G. m. holfcgoi these scales tend to be subequal. This charac- ter varies somewhat, and in any case the time separation as well as lack of further preserved characters precludes reference of this fossil 1o one of the living subspecies. However, this charac- ter and that given as number 2 above seem to suggest a closer relationship to (r. m. grandis and G. m. hotfegoi than to G. m. major. The other subspecies, G. m. zechi, is known from only a few specimens, none of which were available to me. Tt is very closely related to G. m. hotiegoi and its status is not clear at this time. Disti'ibntion of Tvecent GrrrJio.'iairnis major G. major occurs today in principally arid savanna along the eastern coast of Africa, north to Eritrea and south to Zululand. G. m. grandis, the most southern subspecies, is found from Zulu- land north to Morogoro, Tauganyika. G. m. f)WJor is a coastal 1062 MIOCENE GERRIIOSAUR FROM KENYA subspecies, principally in Taiifianyika, but reachinfr as far north as Kenya. G. m. botUijoi ranjies from central Tang-anyika noi'th throup-h central Kenya and reaches north to coastal Eritrea, far- tlici- noi-th than any otlicr gci'i'liosanr. and is the only subspecies of a. nutjor found today in tlic Ivaviroiido (iulf I'eg'ion of Lake Victoria, tlie same region as tiie occurrence of the fossil. The l)rol)lematical (}. m. zcchi has, so far as known, a disjunct distri- bution limited to the northern Belgian Congo and Togo. CONCLUSIONS The fossil described here is closely related to, and perhaps conspeciiie with, the Recent species GcrrJiosaurus major. Thus it is extremely probable that the habitat of the lizards repre- sented by the fossil was semi-arid or arid savanna, like that of the modern species. The presence of a mammalian fauna of Fig. 3. (ierrho.saurus cf. -71, 4 figs., L'l pis. De Stefano, G. 1903. I sauri del Queicy appartenenti alia collezione Rossignol. Atti fioc. Ital. Sci. Nat., 42:382-418, 2 jils. FiLlIOL, II. 1877. Reclieiches sur les pliosplioriles du Quercy. Etude des fossiles qu'on y rencontre et specialement des maniiniferes. Ann. ScL Geol., 8:1-561, 55 pis. FlTZSlMONS, V. F. 1943. The lizards of South Africa. Transvaal Mus., Mem. no. 1, xv -\- 528 pp., 384 figs., 24 pis. IIOPFSTETTER, R. 1944. Sur les Scineidae fossiles. I. Formes europeennes et nord- americaines. Bull. Mus. Nat. Hist. Nat. Paris, 16:547-553, 2 figs. 1955. Squamates du type moderne. In: Piveteau, Traite de Paleon- tologie, 5:606-662, 26 figs. Leakey, L. S. B. 1952. Lower Miocene invertebrates from Kenya. Nature, 169:624- 625, 2 figs. LeGros Clark, W. E., and Leakey, L. S. B. 1951. The Miocene Honiinoidea of East Africa. Brit. Mus. Nat. Hist., Fossil Mammals of Africa, no. 1:1-117, 28 figs., 9 pis. LOVERIDGE, A. 1942. Revision of the African lizards of the family Gerrhosauridae. Bull. Mus. Conip. Zool., Harvard Univ., 89:484-543. McDowell, S. B., Jr., and Bogert, C. M. 1954. The systematic position of Lanthaiiotu.s and the affinities of the anguinoniorphan lizards. Bull. Amer. Mus. Nat. Hist., 105:1- 142, 43 figs., 16 pis. Moreau, R. E. 1951. Africa since the Mesozoic, with particular reference to certain biological problems. Proc. Zool. Soc. London: 121:869-913, 4 tables. ROMER, A. S. 1956. Osteology of the Reptiles. Univ. Chicago Press, xxi + 772 pp., 248 figs. Whitworth, T. 1953. A contribution to the geology of Rusinga Island, Kenya. Quart. Jour. Geol. Soc. London, 109:75-96, 2 pis. 19.")8. Miocene ruminants of East Africa. Brit. Mus. Nat. Hist., Fossil Mammals of Africa, no. 15:1 50, 18 figs., 12 tables. 10 BREVIORA No. 158 - i) i^i S M s 03 Ul h^ a> S ^ "Z O -^j o rt -/] 'A o o ^ .LJ F— 1 -4—' !-• _ o 73 ^ ^. A >i *w -r V^ § ,__. ;2 — • -H ^ ^ -^- • ^ M ^ i^ ? — ' f-i ^ •^ ^ F^ ^ ^ o Ph 2 X ^ 'T O >» GO -a >■ ^ -^ ^* K be C3 »-M o S "5 O-i o ^ ^ _cS p_^ ^ -kJ , — • ^4-^ o oT =1- be o ^ 3 p< cS bl k< .s be o o M -4^ ^^ o "~" •z ^ >: o •f; — -4-" e ^ M B R E V I O R A Miaseiunti of Compsirsitive Zoology Cambridge, Mass. May 31, U)(iL' Nimber 159 AGE IN A SMALL SAMPLE OF BLUEFISH {P03IAT03IUS SALTATRIX (LINNAEUS) )i By Richard H. Backus Woods Hole Oceanographic Institution and Museum of Comparative Zoology Little is known about the life-history- of the bluefish (Poma- tonius saltafrix) even though this lono'-known species is widely distributed in warm and temperate seas and is valuable for food and sport. Not only have the eggs and larval young of this interesting fish not been positively identified, l)ut nothing exact is known of its rate of growth although the "annuli" on its scales prove to be quite easily read. Perhaps it is the great population fluctuations of this fish and its erratic a]:»pearance in many waters which have discouraged naturalists from study- ing it. Bigelow and Schroeder (1953) give a summary of what is known of bluefish life history. Little has been added since their account was written. To learn something of the growth of the species, advantage was taken of the abundance of bluefish around Woods Hole dur- ing the autumn of 1961 and of the angling expertness of Asa Wing, Henry Cain, and Carl Grant, Jr. These men saved heads and scale samples together with records of fork length from the 34 fish that they caught between October 9 and 19 in Great Harbor at Woods Hole, in Woods Hole passage, and along the east shore of Buzzards Bay north of AVoods Hole. The saccular otolith of the bluefish has been figured hj Le Gall (1934) and by Sanz Echeverria (1950). Le Gall {op. cit.) also figured the scales, and in another paper (1935) he said that his examination of the scales in North African specimens suggests that "young individuals attain their adult size and their first sexual maturity at the age of 4 or 5 years." On the other hand, 1 Woods Hole Oceanographic Institution Contribution No. 1234 2 BREViORA No. 159 Boreea (1986) says of Black Sea bluefish : "During the second year they reach dimensions of 14-20 cm and can attain the first sexual maturity. ' ' Table 1 Fork length and age in a small sample of hlucfish. Fork Length Age Pork Length Age (inches) (annull) (inches) (annuli) 13.9 1+ 15.2 1 + 14.0 1+ 15.2 1 + 14.0 1+ 15.5 1 + 14.0 1+ 15.5 1 + 14.0 ? (1 + or 11+) 15.5 1 + 14.1 ? (1 + or 11+) 15.5 11 + 14.1 1+ 15.8 1 + 14.2 1+ 16.0 1 + 14.2 1+ 16.2 11 + 14.4 1+ 16.5 1 + 14.5 1+ 17.0 11 + 14.5 1+ 17.5 11 + 14.8 1+ 18.0 TI+ 15.0 1+ 19.0 11+ 15.0 11+ 19.0 11+ 15.0 ? (1 + or 11+) 20.1 11+ 15.1 1+ 24.0 III + An otolith extracted from a specimen in our sample 14.4 inches in fork length measured about 0.45 by 0.15 inches. The otolith is much sculptured and does not seem suitable for use in age de- termination by simple visual inspection though it might be so suited if x-ray methods were used. Since the scales of this fish Hi'c large, and those examined l)y us seem to show the annuli rather clearly (Figure 1), we have I'elied on the scah^s alone for the ages reported here. Fork length and age for the specimens in our sample are given in Table 1. Fork lengths were taken to the nearest eighth of an inch, were converted to decimals and rounded to the nearest tenth. Ages (expressed in annuli) were determined by examin- ing at least 10 legibh* scales from each specimen. In three cases consistent results could not be obtained. Thus, age has been deter- mined with some confidence in 31 instances. Of these, 21 speci- mens show an age of I+, nine an age of II+, and one an age of 1962 AGE IN BLUEFISH 3 III+. Specimens of age 1+ range from I'-iA) to 16.5 inches in fork length and have a mean length of 14.85 inches. Specimens of age 11+ range from 15.0 to 20.1 inches and have a mean length of 17.48 inches. The sole 111+ specimen measures 24.0 inches in fork length. Because of the small size of our sample these data can do little more than show that snapjiers (as small bluefish are called) of four to nine inches, seen in the autumn (Bigelow and Schroe- der, 1958), are indeed young of the year and that such fish ap- proximately double their length in the succeeding 12 months. Bluefish left the "Woods Hole area during the course of a five- day northeast storm which commenced on the evening of October 19 and lasted until the early hours of October 25. The last fish in our sample was caught on October 19. None were caught after the passage of the storm although the fishing effort continued and produced good catches of striped bass (Roccus saxatiUs). Surface temperature records for Great Harbor, Woods Hole, show fluctuations between 64.9 and 64.5^F for the period October 6 to 12, a decline from 64.5 to 60.8°F from October 12 to October 19, the day the storm began, a decline from 60.8 to 55.0°F during the storm, and small fluctuations about 55°F for the remainder of the month. One may suppose that at about 60°F the bluefish were near the minimum temperature that they can tolerate at this stage of the life cycle and that the catastrophe of the storm, with the accompanying further drop in tempera- ture, was enough to start them on the journey to their winter haunts. LITEEATURE CITED Bigelow, Hexry B. and William C. Schroeder 1953. Fishes of the Gulf of Maine. U. S. Fish and Wildlife Service Fishery Bulletin, 74 (Volume 53) : 1-577. BORCEA, I. 1936. Xotes sur la Biologie du Poniatome (Lufar; dc la Aler Noire. C. E. Aead. Sei. Eoumanie, 1 : 222-223. Le Gall, Jeax 1934. Le Tassergall ou Bluefish (Pomatovius saltatrix Laeepede = Temnodon saltator Linne). Eev. Trav. Office Peches Maritimes, 7: 27-85. 1935. Le Tassergall ou Blue Fish. Bull. Soc. Sci. Xat. Maroc, 15: 232-233. SaNZ EcHEVERRfA, JOSEFA 1950. Notas sobre otolitos de peces procedentes dc las costas del Sahara. Bol. Inst. Espanol Ocean., 27: 1-14. BREVIORA No. 159 m »S»f "^ ' '' ANNULUS I ANNULUS H Fig. 1. Scale from a bluefish {Pamatomits saltatrix) 20.1 inches in fork length taken in Woods Hole passage on October 17, 1961, showing two annuli. BREVIORA Museiimi of Comparative Zoology Cambridge, Mass. June 12, 1962 Number 160 TWO NEW ARTHROPOD CARAPACES FROM THE BURGESS SHALE (MIDDLE CAMBRIAN) OF CANADA By W. D. Ian Rolfe Three carapaces of an undescribed artliropod were found by the writer when curating the large ^Museinn of Comparative Zoology collection of Burgess Shale arthropods, collected by P. E. Raymond, H. C. Stetson, W. E. Schevill and C. H. Burgess in August, 1930. Subsequent search through the material in the U. S. National Museum, Washington, D. C, yielded eleven further specimens and two specimens of another new form, which had been set aside for description Ijy C. E. Resser. The writer is indebted to Dr. G. A. Cooper for permission to borrow and describe the U.S.N.M. material and to Dr. II. B. Whitting- ton for the photographs and for criticism of the manuscript. The USN^M specimens came from Walcott's quarry at locality 35k near Field, British Columbia (Walcott, IDll, pp. 51-52; Resser, 1929, p. 2 : = locality Sllf of Rasetti, 1951. pp. 37-38, 103, 129). The MCZ specimens were also probably collected from this quarry, although it is impossible to be certain of this since the 1930 expedition also collected from "a second layer . . . very fossiliferous . . . some seventy feet further up the mountainside" (Raymond, 1930, p. 32;' 1935, p. 205). This second locality possibly corresponds with Rasetti 's Sllg (1951. pp. 38, 104, 130) and details of the stratigraphy of these two horizons are given in that work. Letters a and b following a specimen number indicate that part and counterpart are present. Carapace shape alone is insufficient to determine the affinities of any arthropod, as Roger has pointed out (1946, p. 59), and hence it seems better to group such isolated carapaces together as follows. 2 BREVIORA No. 160 TRILOBITOIDEA or CRUSTACEA incertae sedis PeOBOSCICARIS gen. nov. Type species. Prohoscicaris agnosia sp. nov. Diagnosis. Carapace valves only known, large, with antero- dorsal region produced into a spatulate beak. Proboscicaris agnosta sp. nov. Plate 1, figures 1, 2; Text-figure 1 Diagnosis. Anterior beak prominent ; length of posterior mar- gin U.48 to 0.82 of greatest depth of carapace ; posterior margin indented at or near midpoint. Description. Valves elongate, ranging from shallow to deep. Since the orientation of the valves is unknown, the straight to slightly concave margin will be treated as dorsal and the pro- duced region as anterior. Postero-dorsal angle rounded and obtuse ; posterior margin slightly indented at or near its mid- point. Ventral margin moderately convex in posterior half to three-quarters of carapace length ; strongly concave in anterior region and separating off a tongue-shaped anterior beak. This beak has been lost from USNM 139866 (see Text-fig. 1) and it seems likely that the distinct outlines shown by USNM 139869 and 139873 (Text-fig. 1) are only due to tlie loss of this region. The absence of a recognizable rim or doublure pre- vents certainty on this point. The valves were doubtless joined by a membranous hinge along the dorsal border as in Canadaspis. Only one specimen shows evidence of two valves preserved, and this is shown on Text-figure 1, USNM 139872. Only the anterior beak of the right valve is preserved and this is displaced anteriorly rela- tive to the almost complete left valve. A reticulate pattern may be seen in patchy areas on USNM 139867 and 139873. The wrinkles subparallel to the ventral margin of USNM 139866 (Text-fig. 1) are clearly due to flat- tening of the originally convex test. Specimens USNM 139867, 139870, 139871, 139873, MCZ 5979/2 and MCZ 5979/3 are blotched by the alga Morania parasitica Walcott, previously recorded on Canadaspis and figured by Walcott (1919, p. 232, pi. 50, fig. 1) on a carapace of Hurdia victoria AValcott. 1962 NEW BURGESS SHALE ARTHROPODS =5 o =0 o o "^ ■-4-I a S o -1 o o 4 BREVIORA No. 160 Remarks. The form of the carapace is so distinct that little confusion is possible with previously described species. Such a small beaked specimen as USNM 139874 approaches Canadaspis perfccia (Walcott) in outline, but the posterior and antero- ventral embayments readily distinguish the new form. Some specimens of Hurdia victoria in the Museum of Comparative Zoology show an indentation of the ?posterior margin similar to that in Prohoscicaris, and in addition show an identical retic- ulation of the carapace surface, so that fragments of the pos- terior ends of the two forms might prove difficult to distinguish. Large-mesh reticulation also occurs in Tuzoia and Carnarvonia (Walcott, 1912, pp. 157-158, 165, 187, 189) so that this char- acter is of little value for suggesting relationships. Small-mesh reticulation is visible in sjiecies of CaryocariH, Dictijocaris and Conca vicar is (as well as in the olenellid trilobites: Raw, 1936; Moore, 1958, fig. 5.22), and in Ceratiocaris and Monfecaris such reticulation can be shown to arise from differential corrosion of the cutieular prisms (Rolfe, 1962a, pp. 45-47). The cuticle of the Burgess Shale specimens is too poorly preserved to ascertain whether this reticulation is sculptural or structural. Holotype. USNM 139871. Plate 1, figure 2 and Text-figure 1. Other material. The twelve specimens shown on Text-figure 1: USNM 139866a/b. 139867a/b, 139868-139870. 139872-139875. MCZ 5979/1, 5979/2a/b. 5979/3a/b. Another specimen, USNM 139876, is a fragment of the anterior end only. Dimensions of holotype. Maximum length parallel to hinge line : 98 mm. Maximum depth perpendicular to hinge line : 52 mm. PrOBOSCICARIS IXGEXS sp. nov. Plate 1, figure 3 ; Text -figure 2 Diagnosis. Carapace valves only known ; anterior beak rela- tively small ; length of posterior margin ca. 0.28 of greatest depth of carapace ; posterior margin sigmoidal. Description. The ventral margin is more of a simple skewed curve than the convex and concave outline of P. agnosta. The posterior margin is shorter and hence is situated more dorsally than in P. agnosta and in addition this margin is sigmoidal rather than indented. The carapace margin is curled under, except along the dorsal border, suggesting the marginal rim common in the later phyllocarids. 1962 NEW BURGESS SHALE ARTHROPODS 5 The surface of the carapace is smooth but little is preserved of the original test save blotches of filmy black material. Occa- sional circular areas of silver sheen on the holotype may repre- sent Morania parasitica. Fig. 2. Outline drawing of the holotype of Proboscicaris ingens sp. nov. — USX:\I 139865b. Postulated anterior to left. Remarks. Again the carapace shape is distinctive, though the asymmetrical outline recalls that of Isoxys, which, however, is smader and has the anterior and posterior dorsal extremities acuminate rather than truncate. This species is the largest of the known Burgess Shale arthro- pods in terms of surface area of the carapace and was doubtless the one which Walcott had in mind when he wrote, "there are also fragments of the carapace of a very large form that possibly may be related to Hurdia victoria" (1912, p. 183). Individuals of H. victoria may be longer but they are also slenderer. Such large carapaces are of particular interest in this fauna since it is among them that a suitable adult for the hypothetically larval Waptia jieldcnsis might be sought (Fedotov, 1925, pp. 386, 389; Heldt, 1954, p. 180: Tiegs and Manton, 1958. pp. 292, 314; cf. Henriksen, 1928. p. 14; Stormer, 1944. p. 100). In this connection is seems worth recalling the striking resemblance of Marria, from this same deposit, to a crustacean nauplius (Ruedemann, 1931, p. 8) or metanauplius. A comparable Upper Ordovician form, Paramarria, occurs in association with an arch- aeostracan carapace, Galenocaris (Wells, 1944). If Paramarria is a naupliar stage, and the larval stage of Galenocaris, and its 6 BREVIORA No. 160 aspect is not merely due to convergence for a planktonic exist- ence (as that of Minietastcr and Bostricliopus seems to be), it would contrast with Kecent Leptostraca. In the latter group, development is direct, the young hatching at a late stage. Simi- larly, Naraoia might l)e regarded as a larval merostomoid. It is possible that J', ingens is simply an older instar of P. agnosta. However, such radical changes in shape are not com- mon except in early ontogeny, and it seems better to distinguisli this form as a separate species. Holotype. USNM 139865a/l) — part and counterpart. Dimensions of holotype. Maximum length parallel to hinge line: 156 mm. Maximum depth ])er]^endicular to hinge line: 75 mm. Other material: USXM 139890 — a fragment of the posterior of a carapace valve, which must have exceeded 150 mm. long by 95 nun. deep when complete. DISCUSSION The laek of limbs or body si'gments precludes any discussion of the affinities of this new genus and the problem of classifica- tion of these early crustacean-trilobitoid forms has been sum- marized elsewhere (Rolfe, 1962b). It seems worthless to classify such isolated carapaces above the generic level in view of the limited number of characters available. Many of the genera attributed to phj'llocarid families, or made the types of new families such as the Isoxyidae (junior synonym of Tuzoiidae Raymond, 1935) and Pseudoarctolepididae of Brooks and Caster (1956, p. 13), will need to be brought together under the inccrtae sedis category shown above. Some idea of the relative abundance of Proboscicaris agnosta in the Burgess Shale fauna may be gained from the following list of numbers of individuals of non-trilobite arthropods col- lected by the 1930 MCZ expedition, and recently curated by the writer : TRILOBITOIDEA Burgessia hella Walcott 54 Emeraldella or fMolaria spp. indet. 3 Leanchoilia superlata Walc. 12 Marrclla splcndens Walc. 202 Naraoia compacta Walc. 3 1962 NEW BURGESS SHALE ARTHROPODS 7 Opahinia regalis Wale. 1 iSidneyia inexpectans Wale. 9 fYohoia plena AValc. 1 TRILOBITOIDEA or CRITSTACEA inceriae scdis Anomalocaris canadensis Whiteaves ca. 22 Canadaspis ohliqua (Wale) 7 C. ovalis (Wale.) 1 C. perfect a (Wale.) 76 C. sp. indet. 12 Fieldia laticeolata Wale. 1 Hurdia triangiilata Wale. 1 H. victoria Wale. 16 Isoxys acutangulus Wale. 10 Prohoscicaris agnosta sp. iiov. 3 Protocaris ef. pretiosa Resser 1 Tuzoia retifera- Wale. 1 T. sp. indet. 2 EEFERENCES Brooks, H. K. and K. E. Caster 1956. Pseudoarctolepis sharpi, n. gen., n. sp. (Phyllocarida) from the Wheeler Shale (Middle Cambrian) of Utah. Jour. Paleont., 30:9-14. Fedotov, D. 1925. On the relations between the Crustacea, Trilobita, Merostomata and Arachnida. Akad. Nauk S.S.S.R., Leningrad, Bull., YI Serie, 18:383-408. Heldt, J. H. 1954. Waptia fiehienain Walcott et les stades larvaires des pen6ides. Bull. Soc. Sci. nat. Tunis.. 6: 177-180. Henriksen, K. L. 1928. Critical notes upon some Cambrian arthropods described by Charles D. Walcott. Vidensk. Medd. fra Dansk naturhist. Fore- ning i Kjerbenhavn, 86:1-20. Moore, R. C. 1958. Introduction to Historical Geology. 2nd ed. McGraw-Hill, New York, Toronto, London, 7 -|- 656 pp. Rasetti, Franco 1951. Middle Cambrian stratigraphy and faunas of the Canadian Rocky Mountains. Smithsonian Misc. Coll., 116, No. 5: 5 + 277 pp. 8 BREVIORA No. 160 Raw, Frank 1936. Mesonacidae of Coniluy in iShrupsliiie, with a discussion of cdassification within the family. Quart. Jour. geol. Soc. London, 92:236-293. Raymond, P. E. 1930. Report on invertebrate paleontology. Ann. Rep. Mus. Conip. Zool. for 1929-1930:31-33. 1935. Leandhoilia and other Mid-Cambrian Artluopoda. Bull. Miis. Comp. Zool., 76:20.5-230. Resser, C. E. 1929. New Lower and Middle Cambrian ('iusta6). describing the musculature of Emys europea, describes the diaphrngmnticus as formed of three parts originating' from the carapace; two parts insert on the wall of the Itmg while the third one inserts on the plastron. This description compares well witli the account given here, except that the two parts of the muscle described by Owen as inserting on the lung are here regarded as tne muscuhiria strUitiim ptihuonnle, and only the third part which inserts on the plastron is the triie diaphragmaticus. BREVIORA No. 161 about expiration and inspiration. He interprets the throat move- ments in Chelonia as functioning in olfaction and not in respira- tion. George and Shah (1954) have studied the respiratory mechanism in Lisseniys and have confirmed McCutcheon's view that the abdominal muscles are effective for respiratory move- ments and that the throat movements are only for olfaction. In addition, they have also described the presence of an extra pair of muscles which cover the lungs completely in Lissemys. These muscles are composed of striated muscle fibres. On con- traction of these muscles the pulmonary air is pushed out of the lungs, and on their relaxation the atmospheric air is taken in. Thus these lung muscles, the muscularis striatum pulmonale, aid the action of the abdominal muscles in bringing about the expiration and inspiration. M.S.P. Fig. 1. Diagrammatic sketch of the disposition of the respiratory muscles in Cyclaxiorbinae where the lungs are completely covered by the muscularis striatum pulmonale. George and Shah (1955, 1958 and 1959) made a compara- tive study of the abdominal muscles and of the lung muscles in some additional chelonians: Lissemys punctata (all three 1962 RESPIRATORY MUSCLES IN CHELONIA subspecies), Geormyda trijuga, Trionyx gangeticus, Testudo ele- gans, Malacoclicrsus iorneri and Eretmochelys imhricata. Ac- cording to their observations the lung muscle, the muscularis striatum pulmonale, covers the lungs completely in Lissemys punctata, partially in Geoemyda, while the muscle is totally absent in the rest of the forms they studied. Of the flank cavity muscles the diaphragmaticus and the transversus: abdominis are O.A. Fig. 2. Diagrtuiimatie sketch of the disposition of the respiratory muscles iu Trionyehinae. well developed in Lissevnys and Trionyx where they join with each other to form a continuous muscle sheath covering the visceral organs including lungs. The diaphragmaticus in Geo- emyda trijuga and Eretmochelys imhricata does not join with the transversus ahdominis to form a continuous muscle sheath, but there is a bridge of connective tissue between them. In Testudo elegans and Malacochersus torneri the diaphragmaticus muscle is totally absent leaving only a thin membranous sheath of connective tissue in its place. The other flank cavity muscles, Adz. the serratus magnus and the ohliquiis abdominis are present with slight variation in all the animals they studied. In the light of these observations on the respiratory muscles in a very few chelonians, it was thought desirable to examine BREVIORA No. 161 n.s.p.L. T^.S.P.M. Fig. 3. Diagrammatic sketch of the disposition of the respiratory muscles in Malacleviys terrainn terrapin. Fig. 4. Diagrammatic sketch of the disposition of the respiratory muscles ill Pseudemys floridana and Fscudemys texana. 1962 RESPIRATORY MUSCLES IN CHELONIA more forms representing, as far as possible, almost all the major groups of the order Chelonia, and make a comprehensive com- parative study of these muscles to get an overall idea of the morphological features of the respiratory mechanism adapted by the animals of this order. For this study some fifty different cryptodiran and nine pleurodiran forms were selected. A list of the animals chosen is given below. This work was carried out at the Museum of Comparative Zoology at Harvard University, Cambridge, Massachusetts, U.S.A. I am thankful to Dr. A. S. Romer, then Director of the Museum, and Dr. E. E. Williams, Curator of Herpetology, for all the facilities given and for their constant help and en- couragement during the course of the study. tvA.D. M.S.P.L.- C.T r\.s.T.T^- Fig. 5. Diagrammatic sketch of the disposition of the respiratory muscles in Emydinae forms in which the diapliragmaticus muscle is absent. MATERIAL STITDIED All the animals selected for the study were alcohol preserved and were found in excellent state of preservation. Careful dis- sections of the flank cavity muscles, the diaphragmaticus, the f)-ansv(rsus abdominis, the serrafus magmis and the obliqnus BREVIOBA No. 161 abdominis and the lung muscle, the miiscularis striatum pulmon- ale, were done on these animals and followino- is the report of the comparative study. List of chelonians selected for the present study: CRYPTODIRA TESTUDINOIDEA Testudinidae Eniydinae Chineniys reevesii Chrysem.ys picta dorsalis Chrysemys picta marginata Chrj'semys picta picta Clemniys caspica caspica Clemmys caspica lepiosa Clemmys guttata Cleininys mutica Cuora aniboiuensis Deiroclielys reticularia Emydoidea blandingii Em^'s orbicularis Geoemyda luanni Geoemyda puuctularia funerea Geoemyda spinosa Geoemyda trijuga Graptemys kohni Graptemys pseudogeograpliica Kachuga tectum tectum Malaclemys terrapin terrapin Malayemys subtrijuga Oeadia sinensis Pseudemys floridana Pseudemys texana Terrapene yucatana TesUidininae CJeoclielone pardalis Pyxis arachnoides Testudo graeea Testudo hermaiini Testudo horsfieldii Testudo kleinmanni Chelydridae Kinosterninae Sternotherus carinatus minor Sternotherus odoratus Chelydrinae Chelydra serpentina TRIONYCHOIDEA Trionychidae C yclanorMnae Cyclanorbis sp. Cycloderma frenatuin Lissemys punctata (all three subspecies) Trionychinae Dogania subplana Trionyx gangeticus Trionyx sinensis Trionyx triunguis CHELONOIDEA Chelonidae Caretta caretta Chelonia mydas Eretraochelys imbricata Lepidochelys olivacea DERMOCHELYOIDEA Dermochelyidae Dermochelys coriaeea PLEURODIRA Pelomedusidae Pelomedusa subrufa subrufa Pelusios subniger Podocnemis expansa Podocnemis lewyana Podocnemis unifilis Chelidae Chelodina longicollis Emydura krefti Hydromedusa niaximiliani Plnynops geoffroana Platemys platycephala 1962 RESPIRATORY MUSCLES IN CHELONIA DESCRIPTION OP THE MUSCLES Of all the respiratory muscles mentioned above, the lung muscle, the muscularis striatum pulmonale, shows the most ex- treme variation in Chelonia. It is so well developed in the forms belonging to the subfamily Cyclanorbinae that it covers the lung completely while in the other subfamily, Trionj'chinae, the muscle is totally absent. In Cyclanorhis sp. and Cyclodcrma frenaium (Fig. 1) the muscle arises from the carapace in the vicinity of the second and third thoracic vertebrae and also from the lateral side of these vertebrae. The fibres arising from the carapace run over the entire dorsal surface of the lung and when they M.T). M.5.P.M- Fig. 6. Diagrammatic sketch of the disposition of the respiratory muscles in Kachnga tectum. reach the outer, anterior, and the posterior limits of the lung they turn onto the ventral side and continue to run towards the entrance of the bronchus. The fibres arising from the lateral side of the thoracic vertebrae run over the medial side of the lung and then come onto the ventral side and reach the entrance of the bronchus. All the fibres of the muscle closely adhere to the wall of the lung. From the place of the origin of the fibres, the muscle could be arbitrarily divided into two parts, a lateral 8 BREVIOBA No. 161 part which arises from the carapace and a medial part which arises from the lateral side of the thoracic vertebrae. The nuis- cularis striatum pulmo'nale in Lisseinys punctata (Fig. 1) (George and Shah, 1954) differs slightly from the one in Cycla- noi'Ms and Cycloderma; the muscle is otherwise very similar in its course and insertion in all the three genera of Cyclanorbinae. In Lissemys the muscle arises entirely from the carapace and does not have its lateral part arising from the side of the verte- brae. The muscle in all the forms of the group Cyclanorbinae is C T.B Fig. 7. Diagrammatic sketch of the disposition of the respiratory muscles in marine turtles. innervated by the branches of the intercostal nerves. As said before, the muscle on its contraction pushes out the pulmonary air of the lungs, and on its relaxation the atmospheric air rushes in. In Trionyx gangeticus, Trionyx sinensis, Trionyx triunguis and Dogania suhplana (Fig. 2) belonging to Trionychinae, the niuscnlaris striatum pulmonale muscles is totally absent. All the forms belonging to the Emydinae have a niuscnlaris striatum pulmonale which partly covers the lung. In this group the muscle shows great variation in different species. 1962 RESPIRATORY MUSCLES IN CHELONIA n.D. Fig. 8. Diagrammatic sketch of the disposition of the respiratory muscles in Testudininae, except Pyxis arachnoides. XA H.SV.L. Fig. 9. Diagrammatic sketch of tlie disposition of the respiratory muscles in Pyxis arachnoides. 10 BREVIORA No. 161 In Malaclemys terrapin terrapin, Clemmys caspica caspica, Clemmys guttata, Grapteniys pseudogeographica, Graptemys kohni, Ciiora amhoinensis, Pscudemys fioridana and Pseudemys texana, the muscidaris striatum pidmonule, though only partly covering the lung, is well developed compared to other Emydinae. In these emydines the medial part of the muscle arising from the side of the thoracic vertebrae is well developed and quite extensive, while the lateral part of the muscle with its origin from the carapace, though well developed, is comparatively small M.D. M.S. P.M. T.A Fig. 10. Diagrammatic sketch of the disposition of the respiratory muscles in Pleurodira forms in which the muscuJaris siriatum pulmonale is partial and the diaphragmaticus muscle is absent. in extent. However, in Malaclemys terrapin terrapin, Graptemys pseudogeographica and Graptemys kohni the lateral part of the muscle is comparatively more developed than in the other emy- dines listed above. The place of oriain of the lateral part of the muscle in these three forms is parallel to that of the medial part of the muscle (Fig. 3). In Pscudemys fioridana and Pseu- demys texana the place of origin of the lateral part of the muscle is perpendicular to that of the medial part of the muscle (Fig. 4). 1962 RESPIRATORY MUSCLES IN CHELONIA 11 111 Emys orhicularis, Kachuga tectum tectum, Ocadia sinensis, Vhrijsemys picta picta, Chnjsemys picta dorsalis, Chrysemys picta marginata, Deirochelys rcticularia, Emydoidca hlandingii, Clenimys mutica, Chinemys rcevesii, Malaycuiys siibtrijuga, Geo- emyda punctularia funerea (Figs. 5 and 6), Geoeinyda manni, Geoemyda spinosa, and Geoemyda trijuga (Fig. 12) the medial part of the muscular is striatum pulmonale is very poorly devel- oped and only covers a very small portion of the anterior medial side of the lung. The lateral part of the muscle is also less developed compared to that of Malaclemys and others and shows variations in its extent, never covering more than a small portion of the anterior and anterolateral side of the lung. M.D M. S. P. C.T. T.A O.A; Fig. 11. Diagrammatic sketch of the di-spositiou of the respiratory muscles in Podocnemis. Six species belonging to the Testudiniuae, Testudo hcrmamii, Testudo graeca, Geochelone pardalis, Testudo horsficldii, Testudo Ideinnianni (Fig. 8) and Pyxis arachnoides (Fig. 9) have been examined. In all these except Pyxis arachnoides the muscularis striatum pulmonale is totally absent and a thin sheet of con- nective tissue is present in its place. In Pyxis arachnoides (Fig. 9) there is a poorly developed lateral part of the muscularis striatum pulmonale present, covering only a very small part of the anterior region of the lung. The presence of part of the 12 BREVIORA No. 161 muscle in Pyxis appears to be a case of an intermediate stage between the typical condition of the Testudininae, on one hand, where the muscle is absent, and that of Bmydinae, on the other, in which it is better developed. In Stcrnotheriis odoratus and Sternotherus carinatus yyiinor, belonging to Kinosterninae, the muscularis striatum pulmonale is similar in its origin, course and insertion to that described for the Malaclemys terrapin terrapin (Fig. 3). In Chelydra ser- pentina of the Chelydrinae the muscle is completely absent and instead a thin layer of connective tissue is present in its place. In Chelonia mydas, Caretta caretta, Lepidochelys olivacea and Eretmochelys imhricata (Fig. 7), which all belong to the family Chelonidae, the muscularis striatum pulmonale is totally absent. Even in Dermochelys coriacea (Fig. 7), of the Dermochelyoidea, OA. C.T.BH— ^ T, A. Fig. 12. Diagrammatic sketch of the disposition of the respiratory muscles in sucli Emydinae as Geoemyda spinosa, etc. the muscle is absent. Thus, it appears that none of the marine forms possess any lung muscle, and that unlike the land forms or the fresh water ones there is no variation in this regard. All the members belonging to the suborder Pleurodira possess a partial muscularis striatum pulmonale rather similar to that seen in the Emydinae (Fig. 5). The members of the genus 1962 RESPIRATORY MUSCLES IN CHELONIA 13 Podocneniis show an unusual extension of the muscularis stria- tum pulmonale. In Poclocnemis nnifilis and the other two species of Poclocnemis, the medial part of the muscle, after its usual origin, runs forward adhering to the medial wall of the lung and (unlike the normal condition where the fibres terminate on the lung near the entrance of the bronchus) after reaching the anterior limit of the lung continues forward and closely adheres to the dorsolateral side of the pericardial membrane. Finally, these fibres insert on the membrane at the level of the anterior side of the auricles (Shah, in press). Such a pericardial extension of the m,uscularis striatum pulmonale is not found in any of the other pleurodirans that were studied. No trace of such an extension is present in any of the Crypto- dira, nor does there appear to be any previous record of the presence of a striated muscle layer on the pericardium in any vertebrate. FLANK CAVITY MUSCLES The diaphragmaticus and the transversus abdominis form the expiratory set of the flauk cavity muscles in Chelonia, while the serratus magnus and the ohliquus ahdominis form the in- spiratory set (McCutcheon, 1943 ; George and Shah, 1954, 1958 and 1959). In all Chelonia the transversus abdominis muscle is well de- veloped. In the Cryptodira studied, the transversus ahdominis muscle is the most highly developed in the Trionychoidea, and in this group it joins with the anteriorly placed diaph ragmaticus muscle of the same side to form a continuous muscular sheath which envelops the visceral organs including the lungs. The muscle arises in all the chelonians from the posterior half of the carapace, but the place of origin is not constant in all forms since great variations in its extent are seen in different individ- uals. In Pseudemys floridana and Pseudemys tcxana the trans- versus abdominis muscle extends almost up to the level of the apex of the heart on the ventral side (Fig. 4). In no Emydinae, whether the diaphragmaticus muscle is present or not, does the transversus abdominis muscle have the extensive spread seen in all the Trionychoidea. In those Emydinae where the dia- phragmaticus muscle is present, there is a bridge of connective tissue between it and the transversus abdominis of the same side. Such a bridge of connective tissue between the diaphrag- maticus and the transversus abdominis muscle is present in all the species of the genus Geoemyda and all the marine chelonians. 14 BBEVIORA No. 161 The diaphragmaticus muscle, in all the cheloiiians in which it is present, arises from the nndersurface of the second or third costal plates of the carapace. Its place of origin is oriented transversely with respect to the vertebral column. In all the forms of Trionychoidea (Figs. 1 and 2) the muscle is very highly developed and, as mentioned above, it joins the trans- versiis ahdo)iiinis muscle of its side to form a continuous mus- cular sheath to envelop the viscera. In some Emydinae the muscle is present (Fig. 12) ; in others it is absent (Figs. 3, 4, 5, and 6). No members of the Testudiuae have the diaphrag- maiicus (Fig. 8) ; there is a thin layer of connective tissue in its place. In all the chelonian studies the inspiratory muscles, the .s'^r- ratus magnus and the ohliquus ahdominis (Figs. 1 to 11), are present with such slight variation that these are not worth detailed discussion. On contraction of these muscles the volume of the body cavity is increased and thus a negative pressure is created in this cavity and so the lungs expand. On expan- sion of the lungs the atmospheric air rushes in and in this way inspiration is brought about. DISCUSSION From the present stud}^ of the respiratory muscles in Chelonia it is evident that there is a great deal of noticeable variation in two muscles, the diaphragmaticus and the muscularis striatum pulmonale. The variations in these two muscles range from a highly developed condition to a total absence, with all inter- mediate stages. The other respiratory muscles are always pres- ent, and although slight variations in different forms are seen, these are very minor ones. The presence of the muscularis striatum pulmonale in all the Cyclanorbinae, where the muscle covers the lung completely, is regarded as a primitive character which is retained in these forms. The early ancestral chelonians presumably developed these muscles as a substitute for the intercostal muscles lost when their body w-all was covered by the rigid shell and could not have the normal movements which are the main component of the respiratory mechanism in all other amniotes. The muscu- laris striatum pulmonale must thus have been of survival value to the early ancestral chelonians, and it is retained fully in all the Cyclanorbinae but shows a gradual trend toward total dis- appearance in other chelonians. Some chelonians, the Triony- chinae, Testudiuae and all the marine forms, have totally lost 1962 RESPIRATORY MUSCLES IK CHELONIA 15 this muscle. It is quite obvious that the presence of the muscle covering the lung is a hindrance to full expansion of the lungs; it must therefore have developed as a stop gap arrangement to tide over the loss of the body wall movements until some better physiological adaptation for respiration was achieved. Unpub- lished Avork by the author on the blood of some cheloniaus shoAvs some interesting results. The oxiphoric capacity of the blood of Lissemys punctata Avhere the muscularis striatum pulmonale covers the lungs completely is much less than that of tlie blood of Trionyx or Tcstudo elegans where the muscle is totally absent. The oxiphoric capacity of the blood of Gcoemyda trijnga, where the muscularis striatum pulmonale muscle is incompletely de- veloped (Fig. 12), shows intermediate values. Thus from the study of blood some light is thrown on the new physiological adaptations which have taken place, substituting for some of the morphological adaptations of the primitive forms. More work on the physiology of respiration in different chelonians will be necessary for a better understanding of the problem of respiratory mechanism in this order. Some aspects of this are being worked on at present in my laboratory at the University of Baroda. ABBEEVIATIONS USED IX FIGUEES Br. Bronchus CT. Connective tissue O.T.B. Bridge of connective tissues between diaphragmnticus and the iransrersus abdoviinis muscles. D. Diaphragmatwus muscle H. Heart Lu. Lung M.D. Connective tissue in place of the diaphra gmuticus muscle. M.S. P. Muscularis striatum pulmonale muscle covering the left lung M.S.P.i Muscularis striatum pulmonale muscle cut horizontally and the right lung removed so as to show the place of origin of the muscle M.S.P.L. Lateral part of the muscularis striatum pulmonale M.S.P.M. Medial part of the muscularis striatum pulmonale O.A. Obliquiis abdomiiiis muscle P.M. Pericardial membrane P. Ext. Pericardial extension of muscularis striatum pulmonale. S.M. Serratus magnus muscle T.A. Transversus abdominis muscle 16 BREVIORA No. 161 EEFERENCES George, J. C. and R. V. Shah 1954. The occurrence of n striated outer muscular sheath in the lungs of Lissemys punciata granosa Schoepff. J. Anim. Morph. Phys- iol., 1: 13-16. 1955. Respiratory mechanism in Chelonia. J. Anim. Morph. Physiol., 1: 30-32. 1958. The structural basis of the evolution of the respiratory mech- anism in Chelonia. Proc. XVth Int. Congress of Zool. London. Papers read by title: 24: 1-2. 1959. The structural basis of the evolution of the respiratory mech- anism in Chelonia. J. Anim. Morph. Physiol., 1: 1-9. McCUTCHEON, F. H. 1943. The respiratory mechanism in turtles. Physiol. Zool., 16: 255- 269. Owen, R. 1866. Anatomy of vertebrates. Vol. I. London, xxxvii + 650 pp. Williams, E. E. and S. B. McDowell 1952. The plastron of the soft shelled turtles (Testudinata, Triony- chidae). A new interpretation. J. Morph., 90: 263-275. BREVIORA Museum of Comparative Zoology Cambridge, Mass. Ji^ly 25, 1962 Number 162 AUSTRALIAN CARABID BEETLES X. BEMBIDION By p. J. Darlington, Jr. Museum of Comparative Zoology, Cambridge, Mass. This is the tenth in a series of papers on Australian Carabidae. Some earlier parts, including a list of localities at which I col- lected in 1956-1958 and a discussion of transition of wet forest carabid faunas from New Guinea to Tasmania, are given under References. The present paper deals with the Australian species of Bem- hidion {sensu lato). This is a zoogeographically important genus, which tends to be bi-zonal in distribution, occurring mainly in the north and south temperate zones of the world (Darlington 1959, pp. 332-333). The distribution, ecology, relationships, and possible history of the Australian forms are therefore note- worthy and will be summarized after discussion of the separate species. I am indebted to Prof. Carl H. Lindroth for dissecting males of all the Australian species and telling me how he thinks they are related to European and North American forms. I could have made the dissections myself, but I am not familiar with the genitalic characters of northern Bemhidion and could not have interpreted the characters of the Australian species. However, neither Prof. Lindroth nor. I have investigated most of the Asiatic species or those of New Zealand or southern South America. This paper is therefore only a limited contribution to the zoogeography of Bemhidion. At the time of Sloane's last study of Australian Bembidiini (1921), he knew five Australian species of Bemhidion and two o/ Cillenus. Two supposed "Bemhidion" of Blackburn's that Sloane did not know {hoharti and watt sense) are in fact not Bemhidion but Tachys. I plan to treat them in mj' next paper. I have series of all five real Bemhidion known to Sloane and 2 BREVIORA No. 162 have seen no other native species, and it may be that these five species of the gemis (excluding Cillenns) are all that are native in Australia, although it is too soon to be sure about this. References and synonymy of the species are given by Sloane (op. cit.) and will usually not be repeated here. The species should be identifiable by the following key, which is based partly on Sloane 's key (1921, p. 193). All the species are winged and presumably able to fly, except that the wings are dimorphic in proprhim (q.v.). Key to Australian Species of Bembidion 1. Large (c. 5.2-6.5 mm.) ; dull bronze, elytra with 2 incomplete transverse pale fasciae ; clypeus with several fine converging grooves on each side (introduced from South America) brullei — Smaller; not marked as above; (native") 2 2. Frontal sulci long, impressed and converging on clypeus ; upper surface of insect dull or shining ; elytron with 6 or 7 dorsal striae 3 — Frontal sulci shallow and rather short, not crossing clypeus; upper surface dull ; elytron with 7 dorsal striae 5 3. Dull brown with vague paler elytral markings; whole upper surface microreticulate ; elytron with (5 dorsal striae (stria 7 absent or faint); length c. 4.2-4.8 mm errans — Shining, upper surface not microreticulate; elytron 7-striate; size smaller 4 4. Elytron with 2 seta-bearing punctures on 3rd interval, none on yth ; color black, elytral apices and sometimes lateral subapical spots slightly paler; length c. 3.3-3.9 mm hlackburni — Elytral intervals 3 and 5 each with several inconspicuous seta-bearing punctures; color irregular dark brown; length c. 2.9-3.4 mm. proprium 5. Prothorax transverse, sides not sinuate posteriorly; color much like following species except apical testaceous area of each elytron broken into subapical and apical marks which are often narrowly connected along outer elytral margin and sometimes connected near suture too ; length c. 3.0-4.0 mm. jacksoniense — • Prothorax cordate, sides sinuate posteriorly ; color greenish or bronzed with elytral apices conspicuously testaceous, the testaceous areas broadly lunate; length c. 4.0-4.7 mm opulentum Bembidion (Notaphus) brullei (Gemminger and Harold) variegatum Brulle 1843, p. 44 (not Say). Bemhieidium hrvllei Gemminger and Harold 1868, p. 409. Form as figured, broader and less convex than usual in Australian species of genus; bronze, appendages irregularly testaceous and fuscous, elytra with complex but variable pale 1962 AUSTRALIAN CARABID BEETLES marks which usually involve epipleuri, parts of elytral bases inside humeri, marginal channels, apices (rather vaguely), an irregular interrupted fascia before middle often including very elongate pale areas on intervals 7 and 8 and isolated spots on 3rd intervals, and an irregular incomplete post-median fascia; sides of abdomen also pale; upper surface dull, with close reticulate microsculpture, isodiametric on head and pronotum, vaguely Betnbidion hrullei (Gemminger and Harold), from between Murray Bridge and Meningie, South Australia. transverse (but nearly isodiametric) on elytra. Head .76 and .76 width prothorax (in S 9 measured) ; eyes large and prom- inent; antennae of moderate length, middle segments c. 2I/2X long as wide ; front slightly convex ; frontal grooves subparallel (slightly curved) and poorly defined on disc, converging and more sharply impressed across clypeus, and latter with addi- tional fine longitudinal (converging) grooves or wrinkles; men- turn with entire tooth. Prothorax transversely subcordate with rather broad base; width/length 1.45 and 1.44 (in measured 6 9); base/apex 1.19 and 1.17; base/head 1.07 and 1.06; sides broadly arcuate for much of length, moderately sinuate before basal angles ; latter well defined and nearly right or slightly 4 BREVIORA No. 162 obtuse; lateral margins moderate, each with usual 2 setae; disc convex, with anterior transverse impression poorly defined, mid- dle line distinct but slightly abbreviated at both ends, basal transverse impression poorly defined; baso-lateral impressions deepest inwardly, bottoms irregular or slightly convex, slightly wrinkled but not much punctate, limited externally by longi- tudinal carinae distinct from prothoracic margins. Elytra c. Ys wider than prothorax (E/P 1.36 and 1.35) ; humeri prom- inent but rounded ; sides subparallel for much of length ; mar- gins moderate (wider than in opulentum) , ending anteriorly opposite ends 6th striae (opposite 5th in opidentum) , forming translucent shelves before subapical sinuations ; striation entire, striae slightly impressed and rather strongly punctate especially on disc ; intervals slightly convex, 3rd with 2 seta-bearing punc- tures about % from base and less than % from apex. Inner wings fully developed. Lower surface microreticulate but mostly not distinctly punctate ; anterior process of raetasternum strongly margined between middle coxae, the margin strongly rounded- angulate at middle. Legs without obvious distinctive characters. Secondary sexual characters: $ with 2 segments each front tarsus moderately dilated and squamulose below; S with 1, $ 2 setae each side last ventral segment. Length c. 5.2-6.5 ; width c. 2.0-2.6 mm. Known in Australia only from six specimens taken by myself beside the road between Murray Bridge and Meningie, South Australia, September 1957, probably beside one of the series of more or less saline lakes near the mouth of the Murray River. However, my collecting there was done under difficulty, in the face of light rain driven by strong wind, and I did not dis- tinguish the species in the field. Specimens of Benil)idio7i errans and proprium were taken on the same occasion. Superficially, this species looks rather like Bemhidion (Nota- pkus) dentellum Thunberg and related species of the Northern Hemisphere, and Prof. Lindroth says its genitalic characters are those of a true Notaphus: "The armature of the internal sac, even in details, comes very close to that of approximatum Lee. and coloradense Hayw." of North America. The sculpture of the clypeus of brullei is distinctive but, I think, not of sub- generic value. I was on the point of describing this as a new Australian species when I discovered specimens of it in the Museum of 1962 AUSTRALIAN CARABID BEETLES 5 Comparative Zoology from South America, from several locali- ties in northern Argentina. It was described more than a hun- dred years ago from specimens taken on the seashore at Monte- video, and it is apparently common in the La Plata region and at Cordoba, Argentina. It has presumably been introduced into southern Australia by shipping. Bembidion (Ananotaphus) errans Blackburn Netolitzky (1931b, pp. 181-182) made this the type of sub- genus Ananotaphus, which may be considered to include also the two following species: propriur)i Blackburn and hlackburni Csiki. The three species in question are somewhat alike in form and agree in having deep frontal sulci, and the male genitalia "show a certain, though not very evident, agreement in the internal sac" (Lindroth). However, the three species differ con- siderably among themselves, and their relationship to other sub- genera is doubtful, so far as the genitalic characters are con- cerned. Blackburn and also Sloane (1921) thought that errans oc- curred only near the coast, and I agree, although it is not con- fined to obviously saline habitats. It has been previously recorded from southern "Western Australia, South Australia, and Victoria. I have a series from southern Western Australia, from several localities including the vicinity of Perth and Pemberton, collected by H. Demarz, and I took two specimens between Murray Bridge and Meningie, South Australia, and four near Hobart, Tasmania. In Western Australia Sloane found it "on the muddy margin of the Vasse River within the tidal influence," and some of Demarz 's specimens were taken at Mandura salt lake. My South Australian specimens were probably taken by slightly saline ponds near (east of) the mouth of the Murray River (see under preceding species). My Tasmanian specimens were on the grassy-muddy bank of a small pond in flat, lowland country north of Hobart. This was essentially a fresh-water habitat, but it was only a few miles from the coast and there may have been a trace of salt there. A series of B. proprium and one specimen of hlackhurni were taken at the same place. Bembidion (Ananotaphus) proprium Blackburn Sloane (1921) knew this species too only from localities on or near the coast of southern Australia, including South 6 BREVIORA No. 162 Australia, Victoria, and southern New South Wales (Wollon- gong). One of his specimens was "beside a little rivulet near where it entered the sea" (presumably at Wollongong). My mainland specimens too are from coastal localities, from between Murray Bridge and Meningie (see under B. Irullei) and between Meningie and Kingston, South Australia (taken by myself), and from Seaford, Victoria (taken "under beach drift" by W. L. Brown). However, I took a series in Tasmania, north of Hobart, beside fresh water, though still near the coast (see under errans). The species thus seems to have about the same ecological distribution as errans. My seven mainland specimens are all fully winged. Most of the sixteen from Tasmania have the inner wings somewhat re- duced, about as long as the elytra, slightly folded or crumpled at apex, and evidently unfit for flight, but one of the Tasmanian specimens is fully winged or nearly so. Bembidion (Ananotaphus) blackburni Csiki Csiki 1928, p. 159. (Jubiinn Blackburn 1888 (not Heer, not Wollaston). Sloane (1921) records this species from South Australia, Victoria, and southern New South Wales, and notes that it occurs beside fresh water, in some cases much farther from the coast than the preceding species. I took it at Winchelsea, Vic- toria, and found it common in Tasmania. It occurred in wet places at low altitudes near Hobart (see under errans) and near Ellendale, but it was commoner on the mountains, near Lake St. Clair (over 2400 ft. altitude) and Great Lake (c. 3400 ft.), for example. On the mountains in Tasmania hhckhurni occurs not only beside standing water and in other wet places but sometimes also on wet, open heaths near and above tree line. Here it runs in sunlight in and on the dense mats of moss and other vegetation that cover much of the ground in open places. This is a true subantarctic habitat. Benihiclion hlackhurni is the only Australian species of the genus that reaches a subantarctic habitat. On the wet mountain heaths, B. blackburni is a member of a small association of superficially similar species of small black Carabidae including Cyphotrechocles gibbipennis (Blackburn), Amblystomus nigrinus Csiki {niger Blackburn), and Euthcnarus nigelhis Sloane. These species look so much alike, superficially, that it is not easy to distinguish them in the field with the naked 1962 AUSTRALIAN CARABID BEETLES 7 eye. They occur together on warm days on wet moss pads, etc., on the mountain heaths as well as in other wet places. This is a striking example of convergence of species of unrelated carabid genera. I do not know its ecological significance. I should add that, although these species occur together under some circum- stances, they do not have identical ecological limits elsewhere. Bembidion (Philochthus) jacksoniense Guerin [Subgenus SloanephUa Netolitzky is here declared a synonym of Philochthus, for reasons given below (new synonymj')-] See Sloane (1921, p. 193) for sj'-nonyms and references, and see Netolitzky (1931b, p. 182) for subgenus SloanephUa, based on this species. However, Netolitzky himself notes the great simi- larity in most characters of jacksoniense and species of the northern (Europe, western Asia, North Africa, etc.) subgenus Philochthus, and Prof. Lindroth finds ' ' general agreement in the arrangement of the internal sac of the male genitalia in jack- soniense and subg. Philochthus . . . ," and it seems to me there is more to gain by recognizing the relationship and putting jacksoniense in Philochthus than by stressing the differences. Netolitzky says the principal difference is that the metasternal process between the middle coxae is margined in the northern Philochthus, not in jacksoniense. Netolitzky adds that in this character and in frontal sculpture jacksoniense closely resembles B. " Notapliojuimus" {=Notaphocampa) opulentum (below). I think these facts may indicate (1) a close relationship between jacksoniense and the northern Philochthus and (2) a less close one between Philochthus and Notaphocampa, with opule7itum perhaps in some ways a connecting link. Sloane says jacksoniense is found over the whole continent of Australia, beside fresh water, and this is probably essentially true. However, it may be absent in small areas in the east and perhaps in larger ones in the north (I am not sure about this) and it has not yet been found in Tasmania, and it certainly occurs in brackish and perhaps alkaline habitats (inland) as well as by strictly fresh water. It is veiy common in Western Au- stralia but less so in the east, although I have specimens from several eastern localities ranging from Townsville in tropical Queensland south to the Blue Mountains of New South Wales and Mt. Kosciusko. In some localities it occurs with the follow- ing species, opulentum. In light-trap material from Cooper Crossing, Lake Eyre, it is represented by a few specimens among many opulentum. 8 BREVIORA No. 162 Bembidion (Notaphocampa) opulentum Nietner [Subgenus Notaphomimus Netolitzky is here declared a synonym of Nota- phocampa, for reasons given helow (new synonymy).] sobrinum auet. (not Boheman). Sloane (1921, p. 193) gives Australian synonyms and earlier references to this species. Netolitzky (1931b, pp. 175-178) makes an (I think) unnecessary special subgenus, Notaphomimus, for it. Except for the difference in frontal sculpture, which I think is over-stressed by Netolitzky, opulentum seems very close to Bemhidion (Notaphocampa) niloticum Dejean, and the two species are closely allied in geuitalic characters (Lindroth). I have discussed the distribution of opulentum {'"sohrinum") recently (1959, p. 339), but my statement that the species ranges from Africa through tropical Asia, etc., was apparently wrong, although made on good authority (Netolitzky, Andrewes). There seem to be two slightly different species: foveolatum Dejean {sohrinum Boheman) of Africa, Madagascar, etc., and opulentum Nietner of the Oriental-Australian region (see Jeannel 1946, pp. 370-371, and Basilewsky 1952, p. 177). B. opulentum occurs through tropical Asia, part (but not all) of the Malay Archi- pelago, part of tropical and south temporate eastern Australia, and apparently also in New Caledonia. It exhibits some geographical variation (Netolitzky, loc. cit.) : the Australian form can be called subspecies riverinae Sloane. In Australia it is widely distributed in the eastern part of the continent at least from the latitude of Townsville south to Tasmania and west to South Australia, but I do not think it has yet been found in Western Australia. (Sloane knew it only from Queensland, New South Wales, and Victoria.) It usually occurs near the margins of standing or slowly moving fresh water. Six specimens that I took behind the bank of the Burdekin River near Charters Towers, Queensland, in March 1958, were running on wet mud and beside a stagnant pool of flood water. A specimen from near Waratah, Tasmania, was taken on a log floating in water at the side of a small pond. And three from west of Renmark, South Australia, September 1957 were, I think, taken on mud behind the bank of the Murray River. However, the species apparently occurs also in inland saline areas, for I have a long series taken at light at Cooper Crossing, Lake Eyre, South Australia, in February 1956 (collected by Dr. G. F. Gross). That this species is so widely distributed in Asia but has not yet spread through the whole of Australia suggests that it has reached Australia 1962 AUSTRALIAN CARABID BEETLES 9 comparatively recently, possibly by way of the Lesser Sunda Islands and Timor, for it does not seem to occur in New Guinea or Cape York. Bembidion, subgenus Cillenus Cillcnus is commonly considered a subgenus of Bcmhidion. Its gross range is from Europe and China to Australia, Tasmania, and New Zealand, but it is strictly coastal. Different species of it occur by running water near the coast, or on ocean beaches, or actually between the tide lines. I have discussed its distribu- tion and possible history elsewhere (1953; 1959, pp. 333-334). Two species of Cillenus are known from (eastern) Australia: a larger, duller one (albovirens Sloane) from tropical Queensland (from Townsville, not Cairns as incorrectly stated by Sloane in 1921), and a smaller, more slender, more shining one (mastersi Sloane) from Sydney and the north coast of Tasmania. Both species probably live on the ocean beach. Both are (irregularly) testaceous, as beach-living Carabidae often are. Both Australian species are fully winged and one or both probably fly. My single specimen of albovirens probably flew to light at Townsville, al- though I did not actually see it do so. Some other Cillenus in other regions including New Zealand are flightless. SUMMARY AND DISCUSSION The five native Australian species of Bembidion (excluding Cillenus, which has evidently had an independent history) are distributed as follows. The three species of subgenus Ananota- jyhus occur in the southern part of Australia, and all of them extend to Tasmania too. Of these three, errans and proprium apparently occur only at low altitudes near the coast, sometimes in saline but also sometimes in fresh or nearly fresh habitats beside water, while blackburni is more widely distributed. It sometimes occurs at low altitudes with the two other species, but it has also entered the subantarctic zone on Tasmanian moun- tains. B. (Philochthus) jacJcsoniense occurs almost throughout Australia (but not Tasmania) in both fresh and saline (interior) habitats, usually beside standing or slowly moving water. And B. (Notaphocampa) opulentum is widely distributed in the Orient and in tropical and south temperate eastern Australia. It reaches Tasmania but apparently not Western Australia. It occurs in about the same variety of habitats as jacksoniense. 10 BREVIORA No. 162 These five native Australian species of BemhkUon are appar- ently descended from tln-ee successive invaders. Tlie first was the ancestor of the endemic subgenus Ananoiaphus, which has difl'erentiated in Australia and formed three verj^ distinct species. All three are now confined to the southern edge of Australia plus Tasmania; one of the three (hlackhurni) extends into sub- antarctic habitats on mountains in Tasmania; and one {prop- rium) is undergoing- wing atrophy. This looks like an early stage in penetration of the southern cold temperate zone by an orig- inally tropical or northern winged group of Carabidae, and evolution of a derivative flightless stock adapted to southern cold temperate habitats. The ancestor (s) of the Australian and Tasmanian "Trechns" may have gone through a stage like this. The second invader was the ancestor of Bemhid/ion (Philoch- thus) jacksoniense. The latter has spread throughout Australia (but not Tasmania) and is now a thoroughly distinct species, although its relationship to the northern species of the sub- genus is still clear. In southern Australia it overlaps the edge of the range of the species of Ananoiaphus but (in my experi- ence) it does not often actually occur Avith them and perhaps does not compete with them very much noAV, although it may limit their distribution northward. The third invader is Bemhidion (Notaphocampa) opulentum. The fact that this species is now widely distributed in southern Asia (with a very close relative in Africa) but has apparently not yet spread through the whole of Australia, and the fact that the Australian population is only subspecifically diiiferentiated, suggest that the species has reached Australia rather recently. In eastern Australia it occurs with jacksoniense, in the same habitats. The tAvo species may compete, and opulentum may be replacing jacksoniense, which seems to be much less common in eastern Australia, where opulentum occurs, than in the west, where opulentum does not occur. This history of three successive invasions, the later invaders perhaps competing with and modifying the distributions of earlier ones, accounts very well for the present distribution of Bemhidion in Australia. The question then is, what has been the history of the ancestral forms outside Australia? Bemhidion is now dominant in the north temperate zone but is very poorly represented in the tropics, where the genus is replaced by swarms of Tachys. The obvious, but not necessarily the correct, guess, therefore, is that the three Bemhidion that have invaded Australia have somehow crossed the tropics from 1962 AUSTRALIAN CARABID BEETLES 11 the north temperate zone. This would be consistent with the distribution of Philochthus, which is now Avholly north temperate except for jacksonicnse isolated in Australia. However, another, perhaps more probable history can be sug- gested. The endemic Australian subgenus Ananotaphus vaguely resembles Notaphocampa in form. Ananotaphns has deep frontal sulci and Notaphocampa opulentum has shallow ones, but opu- lentum is very close to Notaphocampa niloticmn Dejean in most characters including male genitalia (Lindroth), and niloticum has deep frontal sulci. Ananotaphus and Notaphocampa are in this way linked by niloticum, which, incidentally, ranges from North Africa across tropical Asia north to temperate Japan. Similarities between Notaphocampa opulentum and Philochthus jacksoniense are noted under the latter in the preceding pages, and iacksoniense is clearly related to northern species of Philochthus. These are hints that Ananotaphus, Notaphocampa, and Philoch- thus, which include all the Bcmhiclion (except Cillenus) native in Australia, may belong to one group which has had a rather complex evolutionary and geographical history, and the group may have been primarily tropical. B. opulentum and niloticum are widespread in the tropics now ; all the Australian Bemhidion are salt-tolerant, which is characteristic of tropical Bemhidion (Darlington 1953, p. 14), and at least one European Philochthus {aeneum Germar) is salt-tolerant too (Lindroth) ; and evolution in the tropics would facilitate the successive invasions of Au- stralia that have occurred. I suggest, then, that subgenera Ananotaphus, Philochthus, and Notaphocampa may all be derived from one salt-tolerant group of Bemhidion that has evolved primarily in the Old World tropics, has invaded Australia three times, and has invaded the north temperate zone at least twice, once as Philochthus in Eu- rope, etc., and once as Notaphocampa niloticum in Japan. This hypothetical history cannot be critically tested until the phylo- geny of the whole genus Bemhidion is better understood. Geni- talic characters do not clearly confirm it. If the forms in ques- tion have evolved as suggested, their history has probably been complex and may have involved additional groups of Bemhidion that have not reached Australia. Cillenus is another, independent subgenus of Bemhidion (per- haps it should be considered a related genus) that has crossed or evolved in the tropics, in saline habitats, and reached Australia (Darlington 1953: 1959, p. 333). 12 BREVIORA No. 162 It is noteworthy that, although several stocks of Bemhidion have reached the western part of the Malay Archipelago from Asia apparently by "mountain hopping," by somehow dispersing from mountain to mountain and from island to island at con- siderable altitudes (Darlington 1959), none has reached Aus- tralia or New Guinea in this way. No Bemhidion has been found on mountains in New Guinea, and the Australian species are primarily lowland forms and seem to be descended from salt- tolerant ancestors that crossed or evolved in the tropics at low- altitudes. EEFERENOES Blackbuun, T. 1888. [Australian Benibidiini.] Trans. R. Soc. South Australia, 10: 38-45. 1901. [Australian Benibidiini.] Trans. R. Soc. South Australia, 25: 120-124. Basilewsky, p. 1952. [Types of African Bemhidion, etc.] Bull. Ann. Soc-. Ent. Belgique, 88: 173-194. Brulle, a. 1843. (In) d'Orbigny, Voyage dans I'Amerique meridionale, 6, Part 2. Insectes. CsiKi, E. 1928. Junk-Sehenkling Coleopterorum Catalogus, Carabidae, Harpa- linae I. Darlington, P. J., Jr. 1953. [Cniemis.] Coleopterists ' Bull., 7: 12-16. 1959. Bemhidion and Trechus of the Malay Archipelago. Pacific Insects, 1: 331-345. 1961a. Australian carabid beetles IV. List of localities, 1956-1958. Psyche, 87: 111-126. 1961b. Australian carabid beetles V. Transition of wet forest faunas from New Guinea to Tasmania. Psyche, 68: 1-24. Gemminger, M. and B. de Harold 1868. Catalogus Coleopterorum, 1. Jeannel, R. 1946. Coleopteres carabiques de la region malgache, part 1. Netolitzky, F. 1931a. [Palearctic Bembidion.] Koleopterologisehe Rundschau, 28: 28/1-124/96, D/97-70/166. 1931b. Uberpriifungen afrikanischer und australischer Benibidiini. Wiener Ent. Zeitung, 47: 169-183. Sloane, T. G. 1921. [Australian Bembidiini.] Proc. Linn. Soc. New South Wales, 46: 192-208. BREVIORA MiuseiLim of Compairative Zoology Cambridge, Mass. Jilv 26. 1962 Number 163 NEW W0RM-LIZAKD8 (ANCYLOCRAXirM AND A3IPHISBAEi\A) FROM SOUTHEASTERN TANGANYIKA TERRITORY By Arthur Loveridge 1. ANCYLOCRAxn':\r On the 27tli May. lf).")0. Mr. C. J. P. lonides collected at Newala an Anciilocrauium whicli, later that same year, he do- nated to the British Museum (Nat. Hist.). AYhile expressing the liopc that more material would be forthcoming'. I tentatively identified it with A. harkeri Loveridge (1946, Proe. Biol. Soc. \Yashington, vol. 59. p. 78, figs.), a species then, as now. known only from the 6 holotype (M.C.Z. 48950). Last year Mr. lonides was successful in securing six addi- tional Ancylocranium from NcAvala, the most southerly record known for any member of this ]ieculiar genus. I find that all six differ from ^1. harleri of Lindi District in precisely the same characters as did the first, as recorded in my manuscript notes. Consequently, I propose to designate these Newala Anci/locrn})- iiim as a southerly form of harkeri . viz. Ancylocranium barkeri newalae subsp. nov. Holotype. Museum of Comparative Zoology No. 67001 (lonides No. 9356), a presumed 6, from Newala, Southern Province, Tanganyika Territory. Collected by C. J. P. lonides, Esq., on 12th June, 1961. Paratypcs. lonides Nos. 8844, 9351-3, 9355. 9359, of which three are now in the Museum of Comparative Zoology (M.C.Z. 67002-4) and three in the British Museum Nat. Hist.) (BM 1959.1.5.18. 1962.177 and 1962.178). 2 BREVIORA No. 163 Diagnosis. Head shields substantially the same as figured for .4. harkrri from which. howeA-er, it may be distiiifriiished as follows : Median ventrals about six times as broad as their fellows; complete caudal annuli five or less. 27 CIS + 9) or 29 flS + 11) sej^ments in a midbody annulus; 205-215 annuli on body, normally 3 on underside of tail from post-anal rinarietal). 20 (10 + 10) segments in a midbody annulus; 227-247 annuli Oil liody. 28-28 on tail: Kondo Plateau jviulocnsis rostciioi- leiiij)oi'al i)resent immediately below parietal. 20 (10 + 10) seiiiitents in a midbody annulus; 239-255 annuli on body. 22-24 on tail: ^lakonde Plateau ik iroloensis 22 ! 10 + 12) sepmenis in a midliody ainiulus; 252-277 ainiuli on l)ody. 22-27 on tail: Nangiu-nwe, Newala )iangun(Wfnsis 24 (12 + 12). rery rarely 22 (10 + 12) segments in midbody annulus; 264-280 annuli on body, 25-28 on tail: Mbanja and Lindi, Lindi District ewerhecl-i Df scfiption. Due to fusing of the surviving head shields in the subgenus Ciinlsca. their complicated nomenclature is some- what difficuli to follow. The student is therefore referred to figures 22 (( ini-hccki) and 23 ifoiidocnsls) on pages 393-394 of my Pevision oF the Afi-ican Lizards of the Family Amphis- baenidae (1941. P.nll. Mus. Comp. Zool.. vol. 87, no. 5). The lalei'al view of civcrhrcki there shown, will be found to corre- spond with the head of )icic(!lacnsis in displaying behind the temporal a scale (]iosl('rior temjioran immediately below the large ])arietal. This a])idies to all ten Newala lizai-ds, as it did to all fifty fv'crbvcki that 1 collected at Mbanja on the Lindi coast. \'iewevl from above, the ])ari('tal suture of both n-ed. the immediate flanking' pair of laterals trans- versely divided (or entire in some paratyjies). Color. In alcohol. Grayish white ( ' flesh-]unk ui life), uni- form except for the e.rfrrme tip of tail which is purplish. Sizr. Total length of holotype 6 . 144 (127 + 17) mm., largest 6 in the series; largest 9 (I 9114), 152 (13(i + 16) mm. In the six i 6 the tail is included in tlie total length 8.4 to M.3 times: in the three undamaged 9 9 from 9.5 to 10.4 times. AmPHISBAEXA XAXGI'Rl'WENt^Is; sp. UOV. Holotype. Museum of Comparative Zoology Xo. 67010 (lonides Xo. 9249). a 3 from X'^anguruwe. ca. 1600 feet, 8 miles south of Xewala. Xewala District, Southern Province. Tanganyika Territory. Collected by C. J. P. lonides, on 16 May. 1961. Paratypes. :\LC.Z. 67011-19 (lonides Xos. 9250-9355). being 37 ci o [aU used for statistics) and 69 9 9 (of which 62 were measured but only 42 had annuli counted i. AVith same data as type but collected 12-16 May. 1961. Diagnosis. ISee Diagnosis for ^4. niwalaensi^, where the pre- liminary jiaragrajih and key are e(|ually applicable to nanguru- irt nsis. Description. xVgain, see remarks under Description of ihe preceding species, for the head shields of nangurnu'ensis arc substantially similar to those of ewerhecki. Segments in a midbody annulus 22 (10 + 12), but counts made on only a score of paratypes: annuli on body 263 (252-277 in paratypes. viz. 252-2G!-* in 6 6. 261-277 in 99) of which 6 BREVIORA No. 163 the median pair of ventrals are only about l^/o times as broad as the adjacent ones; annnli on underside of tail from post-anal ring to conical tip 27 (22-26 in 37 paratype $ i , 22-25 in 42 9 9); anals 6 (4 in four paratypes, due to fusion), the median pair much enlarged, the immediate flanking pair of laterals entire, transversely divided, or reduced to a tiny wedge Color. In alcohol. Grayish white (? flesh-pink in life), uni- form except that much of the tail, both above and below, may be purplish. Size. Total length of holotype $ , 160 (143 + 17) mm., largest of the 38 S $ ; largest 9 (T 9250), 170 (153 + 17) mm. Total length of smallest 6 , 124 (111 + 13) mm., of smallest 9 , 100 (90 + 10) mm. In the 38 (J (5 the tail is included in the total length from 8.7 to 10.1 times, average 9.4 times ; in the 62 9 9 from 9.4 to 10.9 times, average 10.1 times. /^^ BREVIORA MniseiiirTti of Coimparsitive Zoology Cambridge, Mass. Aigust 22, 1962 Number 164 NOTES ON THE HERPETOLOGY OF HISPANIOLA. 7. NEW MATERIAL OF TWO POORLY KNOWN ANOLES ANOLIS MONTICOLA SHREVE AND AX LIS CHRISTOPHEI WILLIAMS. By Ernest E. Williams Recent expeditions in Haiti have obtained material of two ancles, one of which {A. inonficula Shreve) had been previously recorded on the basis of the unique type specimen, and the other (A. christophei Williams) was known only from type and para- type. Examination of the new material sugg-ests that the two species, though in a number of respects strikingly different, may yet be related. It is, therefore, appropriate to discuss these two species jointly. Comparison is made also with Anolis darlingtoni Coch- ran = .4. ctheridgei new name,^ regarded by Cochran (1939, 1941) as allied to monticola. Anolis monticola A. monticola Shreve 1936 was described from a single male (lacking most of the tail and darkened by formaldehyde) col- lected by P. J. Darlington in "tlie northern and eastern foothills. Massif de La Hotte, 1000-4000 feet, Haiti." No other specimens have been reported since the original description. However, specimens had been collected for the American Museum of Natural History by W. G. Hassler in 1935 in the vicinity of Aux Cayes and Camp Perrin. These, like the type, are in a dark phase and show only the faintest trace of pattern. In Hassler 's notebook, on the other hand, there 1 Etheridge (unpublished thesis, University of Michigan) has shown that the genus Xiphocercus cannot be retained. Thus. Xiphocercux diirlingtoni Cochran l'.i.';."i joiiKs the ijcnus AjioHk. am; Aiwli-y (liirliiuitoni Cdchran liC'.it iiuiHt in conse- tiuenee be renamed. I propose, with Miss ('(ichran's kind consent, tliat tlie later named species be called Anolis etheridgci. 2 BREVIORA No. 164 is an excellent description of colors in life. The American Mu- seum specimens (as well as Hassler's notes) were examined by Max Hecht some years ago and the material tentatively referred to monticola Shreve. In 1960, A. S. Rand and J. Lazell (collecting in Haiti with the aid of a grant from the American Philosophical Society) re- turned with the first well-preserved specimens of this very beauti- ful small lizard. A. S. Rand reports the circumstances under which this mate- rial was collected as follows : "The monticola were found along the trail one-half day's ride above Camp Perrin. The trail there ran along a hillside mostly cut over and grown up to dense bushes and grass. Along this trail were many hendersoni subsp. in the bushes, particularly where they were shaded by occasional trees. I remember also seeing cyhotes and one ricordii. ' ' At one point we came to a small patch of rocks . . . boulders of various sizes, some very large, heaped one on top of another a bit like but less extreme than the boulder heaps in the Panduras Mts. in Puerto Rico. This area had not been cleared and growing up from among these boulders were sizeable trees and much smaller woody vegetation. The result was a dense, heavily shaded patch of bush isolated in an open sunny area. It was in this spot on the trees, bushes, fallen branches among these large boulders that we found the monticola." The available sample of Anolis monticola now consists of the following specimens:^ Haiti. Department du Sud. Norther n and eastern foothills, Massif de La Hotte, 1000-4000 ft.: MCZ 38296 (type) . Mountains 25 miles north of Aiix Cayes on Jeremie Road: AMNH 49818, 49845, 50108-9, MCZ 56139 (formerly AMNH 50110). In moun- tains on road to Jeremie about 8 miles from Camp Perrin, 2000- 3000 ft.: MCZ 56140 (formerly AMNH 50097). Tomheau Cheval between Camp Perri^i and Beaumont : MCZ 62998-6300'. Grande Cayemite: MCZ 58026. The last specimen requires special discussion. It is the Eyerdam specimen (formerly MCZ 25483B) discussed by Miss Cochran (1941, p. 179) as a juvenile coelestinus with keeled ventral scales. Though like many Eyerdam specimens poorly 1 Museums from which specimens are cited in this paper are alilireviateil as follows : AMNH, American Museum of Natural History : MCZ, Museum of Com- parative Zoology ; UMMZ, University of Michigan Museum of Zoology. 1962 TWO POORLY KNOWN ANOLES 3 preserved and dried, it seems a typical ynonticola with the char- acteristic swollen middorsals as well as keeled ventrals. It is important in extending the range of monticola to this small island off the north coast of Haiti's southwestern peninsula. Hassler's and the more recent collection were all made in a relatively small area in the southeastern foothills of the IMassif de La Hotte. However, the type, like the Grande Cayemite speci- men, came from north of the Massif de La Hotte. (Recent evi- dence, still unpublished, indicates that there may be striking differences between the forms to the north and to the south of this ridge.) The new specimens show that the species reaches at least 46 mm snout -vent length. Color notes and sketches are available for A. monficola both from W. G. Hassler and from the Rand-Lazell expedition. Hassler's color notes are for specific specimens: 1. 3Iale (Field no. 74, now AMNH 49845). General dorsal color Hooker's Green. Saddles brown green, three in number, narrowest middorsally, one across shoulder, two between fore and hind legs. A light crescent in the temporal region. Throat and belly dark olive green. Legs barred. Tail barred. Eyes Antwerp Blue, sometimes changing to greenish. Edge of orbit yellowish brown. Skin of fan (which is relatively small) blue, scales light and dark green. Occurring also in a dark phase almost without pattern. 2. Female (Field no. 70, AMNH 50097 = MCZ 56140). Back brownish, bounded laterally by a wavy reddish line edged with a dark line. Sides below the dark line olive shading to very light yellowish green. Sides of neck yellowish green. Tail nearly plain brown. Legs reddish brown posteriorly. Belly nearly white. Upper lips bluish. Throat yellowish with some green. I have examined also by courtesy of Mr. Hassler (now of the Fort Worth Children's Museum, Fort Worth 7, Texas) a color sketch, made by Melville P. Cummin, of the head and neck of a live male from "mountains on Jeremie road about 30 miles from Cayes, 2000-3000 ft. alt., August 28-29, 1935." This shows very well the blue of the chin and the yellow green ground color of occiput and nape as well as the black white-centered ocelli which occur laterally on the occiput and nape of some male specimens. Hassler's photographs of a live male and a female contrast the ocelli and the strong banding of the body in the male with the almost patternless female. 4 BREVIORA No. 164 The observations of Eand and Lazell are in excellent agree- ment with those of Ilassler. A pencil and crayon sketch from life is included in their field book and Lazell 's description of the same specimen, now MCZ 63004, may be paraphrased as follows : Ground color yellow green, top of head darker. Body banded with velvety black. Upper and lower jaws blue, this color con- tinued as a stripe fading to whitish above shoulder. An area behind the eye and a V-shaped band across the neck ochre yellow. In the black mark on the neck behind the yellow is a long, thin sky-blue spot. Another such blue spot in the black area on the occiput and anterior to it, still within the black area, a larger white spot with a pale pink center. Legs banded, with a brown- ish wash. Venter and fan bright yellowish green, but base of fan with a bluish wash on both scales and skin. Lazell comments (in agreement with Hassler) that the males can turn very dark after capture, obscuring the markings to an extreme degree, such that only the markings on head and neck are still visible, but the jaws always retain their bluish color and the fan its bright yellow green with a blue wash. (The type was apparently preserved in this phase ; no markings at all are visible, but the blue of the chin remains evident.) All the males in the Rand and Lazell collection have two prominent pairs of ocelli, one on the occiput, one on the nape. In the figured specimen the light spots in the interior ocelli are divided. In the preserved specimens the light zone between the two pairs of ocelli may be lighter than the ground color or not. (This is one of the areas described by Lazell as ochre yellow in life.) Banding on body and limbs is always conspicuous; on the flanks each dark band is divided by a lighter, narrow, vertical streak. The two females obtained by Rand and Lazell show the lighter mid-dorsal zone bounded by a wavy dark line mentioned by Hassler. Rand, who has described these specimens, cites the color of the middorsal zone as greenish brown. The inward curves of the dark line are gentle, the convex portions of the line are produced into points by larger or smaller spots of color even lighter than the broad middorsal zone. There is blue on the sides of the chin, and the throat is green and the belly whitish in life. Anolis christophei The species A. christophei Williams 1960 was based on two female specimens obtained in the vicinity of La Citadelle of 1962 TWO POORLY KNOWN AXOLES 5 Kino- Christophe, one collected by W. J. Eyerdam for the Mu- seum of Comi:)aratiye Zoology in 1927, and one collected by W. G. Hassler for the American Museum of Natural History in 1935. A series of this species, including the previously unknown males, has now been secured again in the vicinity of the Cita- delle by Luc and George Whiteman collecting for the Museum of Comparative Zoology as part of a collecting program in Haiti partly supported by N.S.F. grant G 16066. Unfortunately these are without notes on color in life or any record of dewlap color. They are recorded as occurring on walls and on the ground. The type (I\ICZ 25485) and paratype (AMNH 49736) are now supplemented by the following specimens (all from La Cita- delle) : MCZ 66900-19, UMMZ 122818(5). The color as ascertained from preserved specimens is not very different from that described by Cochran (1941) for the type, as quoted in the original description. Male. Snout mottled. Two darlv supraorbital stripes, neither sharply defined, the posterior stripe coalescent with an arc of dark pigment surrounding a light area which encloses the inter- parietal. A scalloped, crescent-shaped transverse dark mark on the occiput. A butterfly-shaped mark middorsally in front of shoulder, light-edged behind. Three similar marks between the fore and hind limbs and above the sacrum, but in each of these the anterior margins are ill defined. Tail indistinctly banded. On each side a light line from the ear arching downward to join a dark-edged light line just above root of arm, continuing halfway along the body. Flanks above and below light line strongly mottled with darker. Limbs mottled light and darlc. Labials above and below darkly pigmented. Venter washed and powdered, Avith throat and undersides of limbs and tail darker, with some obscure light spotting, least evident on the tliroat. Dewlap scales white, skin grey. Female. Pattern as far as discernible the same but much more obscure. Throat and limbs more distinctly spotted with white. Even in the male the pattern, while complex and distinct, is not prominent. Near hatchlings show the same pattern especially well-defined, but even in these the contrast of dark and light areas is not really bold or striking. In all cases specimens must be fully immersed in liquid before the pattern becomes at all evident. BREVIORA No. 164 Comparison op A. christophei, A. monticola and A. ETHERrooEi A. monticola, A. christophei and A. etheridgei are all small anoles (44-48 mm snout-vent length) of, as far as known, quite local distribution. (See map.) Only A. etheridgei is known to be montane, reaching at least 6000 feet elevation. A. monticola, despite its name, is not known above 3000 feet, and A. christophei is known only from the Cita- delle which is not more than 2000 feet high. All three belong to the set of Hispaniolan anoles that have the ventral scales arranged in transverse rows. They differ from all other Hispaniolan anoles (except the very different giant anole A. ricordii) in having the subocular scales separated from the supralabials by a row of intervening scales. In the other Hispaniolan anoles the suboculars and supralabials are in contact. The three species share other minor details of squamation : 10-14 scales across the snout ; 6-7 loreal rows ; 6-7 supralabials to center of eye ; temporals and supratemporals finely granular ; interparietal smaller than ear : only one sublabial in contact with the inf ralabials ; lamellae under phalanges ii and iii of fourth toe 14-19. They are somewhat similar also in basic color pattern. All have the dorsum transversely marked. In monticola the dorsal Anolis monticola AAnotis chrislophe Anolis etheridgei Figure 1. Map of the distribution of Anolis monticola, A. christophei and A. etheridgei on Hispaniola. 1962 TWO POORLY KNOWN ANGLES 7 marks extend on to the flanks as transverse bands ; in christopliei and etheridgci they are restricted to the dorsum as saddle-like marking's and in etheridgci rather broken up into mottling. ^ Monticola is strongly sexually dimorphic in pattern; etheridgei weakly so; chrisfophci hardly so at all. In spite of all resemblances, however, we deal with three strongly marked species, almost as distinct from each other as from other Hispaniolan anoles. Table I compares the differences in squamation between the three taxa. As Cochran suggested in the description of darlingtoni ( = etheridgei), this species seems the closest relative of monticola. Christ ophei in its lack of the specializations of the other two forms — in particular, in its possession of a well-developed dew- lap — • seems the most primitive of the trio. The bright bold patern of male A. monticola would seem prima facie an obvious compensation for the extreme reduction of the dewlap, which is in most species a striking species-recognition mark. We do not have a description of colors in life of A. etheridgei; it would be reasonable to expect a more vivid pattern than appears on the available specimens. We also lack ade- quate information on the color in life of christophei. In regard to the latter species I can only state that the material recently re- ceived and in a state of preservation that normally retains well any vivid pattern seemed nondescript and obscurely patterned until very closely examined. Whether there are any more members of this small sub-group of Hispaniolan anoles will have to be determined by more thor- ough search of the island. A. etheridgei is still known only from four localities in the Cordillera Central of the Dominican Repub- lic (Loma Vieja, Loma Rucilla, Valle Nuevo, Constanza), A. monticola from four localities in the foothills of the Massif de La Hotte and on the island of Grand Cayemite, and A. christophei only from the single locality, the vicinity of La Citadelle. These are widely separated areas. If related species are equally local in distribution, they may well have been missed. Of course, we lack at the moment any information which would permit sig- nitieant hypotheses on the history of these forms. There is indeed evidence that suggests that these three forms are one branch of an extremely interesting intra-Hispaniolan radiation ; this con- ception, however, will be documented in another paper. 1 1 rely on Miss Cochran's description of A. darlingtoni = etheridgei for evi- dence of transverse, dorsal, saddle-like markings in this form. At the time of writing twenty years later they are not at all evident on the specimens. BREVIORA No. 164 ■t.3 s o Ci X t •e* m 5^ M O (M e o 1 « «»• tc s o Q h S w o -a « SQ o rr* '« 1> n Ci 'S h «i n r« QJ '3 •\ "3 C3 ti bi p Q ^ =w -> e« ft o 00 05 -2 iH '3 m .^ 1:3 .a .rH .-• c 03 O) 13 CO a" 0) <~— « 02 -*-i a; 'rS i-H CCI o OS 03 ^^ CO ft 3 >i to T3 J ri h b CO f ft ^ <» 03 en a C 05 M Zj 03 ^ •rH • r- '« . t— 1 -!.:> ft g X2 C« •40 s ut 00 oj ;h ;3 o3 ft t< fi C3 t4 J= w +- •+^ ^ p _o 03 ^1 > f^ 2' f~ -c> a Oj '2 tH -*•- G •rH CO Ih O) ■^r co ft b aj c -^ CS g ,C c« -(•^ ft t; O; CO CO ft ■►^ B c3 03 ■^ cu Lh ^ « c3 +H (5 s a ^ CO o »H o o cS Ih ft o e! ft CO cS «H aj N • l-< 05 ;^ e l-i ,_H r5 cS -♦•^ "3 c ^ 03 <4H Vr ft tH 3 _o Ol .2 CO '{•M O) ft +H CO ''I 1^ CO « N • fH 00 ?f^ !;' c t^ 03 03 c3 ft Ih S CO fH tH 05 ';h « ft +H 1962 TWO POORLY KNOWN ANOLES 9 be 3 VI o ,^ ai ^ 0) § &; -*j -i-* OJ o U> o m r^ o > CB 01 [» "« CO -M ^_, •1 « M o D ' — 3^ c M f^ L^ 3 O [» r- sT a p 3 o o ^ CO b •* V ^ > C8 X ri 3 '4-' _2 13 tf c; ;I3 eS >» ^ 1 r^ SJ tn -*- ^ "x S ^ 1— 1 -^ cS o X X > X X n < o X O o It o > H X ;^ ^ -^- X o § X S ■? X S T3 15 •^ rj: "^ O '$■ o3 s > o ft 1^ X X X 3 o3 & X -1-5 a £ 3 0) 3 'fcH a7 > X +3 O o a p 3 S r^ X X )5 3 3 fO o X 13 5 > X 3 3 X 13 X p _7r ■^ s: 3 s 15 p +-• 1-1 O ■^ ■^- -w 3 u v 3 O) 3 o ^ ■f^ >» bj3 l-< o o ^ +J i X o p< k> X 4J Ol TS 13 ■? X _x ^ « ■*! '1 13 _3 3 • r-t ^ a fe »— 1 I— < X c« X o <;] a> X M » P» X M 3 ;h aj -M -^ 3 3 3 • fh O t^ TJ a -3 01 3 ^ aT .4^ ^ 3 3 o >-4 3 ^ m J2 X e ■PH Is X 3J o 3 5 t>l 3 -2 X ^^ i ;2 3 'o t^ «H 3 TJ '.^ 3 X be 3 aj > ^ 'P a M »-*i 3 -*^ 13 nd 00 X 10 BEEVIORA No. 164 LITERATTJEE CITED Cochran, D. M. 1935. New reptiles and amphibians collected in Haiti by P. J. Darling- ton. Proe. Boston Soc. Nat. Hist., 40: 367-376. 1939. Diagnoses of three new lizards and a frog from the Dominican Republic. Proc. New England Zool. Club, 18: 1-3. 1941. The herpetology of Hispaniola. Bull. U. S. Nat. Mus., 177: 1-398. Shreve, B. 1936. A new Anolis and new Amphibia from Haiti. Proc. New Eng- land Zool. Club, 15: 93-99. Williams, E. E. 1960. Notes on Hispaniolan herpetology. 1. Anolis christophei, new species, from the Citadel of King Christophe, Haiti. Breviora, No. 117: 1-7. 1962 TWO POORLY KNOWN ANOLES 11 o o CO to N Q Si o BREVIORA MiaseiLiiiri of Comparative Zoology Cambridge, Mass. August 22, 1962 Number 165 AN EXTINCT SOLENODONTID INSECTIVORE FROM HISPANIOLA By Bryan Patterson During the summer of 1958, Drs. Clayton E. Ray and A. Stanley Rand carried on field work^ for this museum in Puerto Rico and the Dominican Republic. Particular attention was paid to cave deposits, and a number of previously unexplored caves were examined. The cave containing the material re- ported on here is in the Sierra de Neiba near Rancho La Guardia in the Province of San Rafael, Dominican Republic; it has no local name. Situated in a limestone cliff and accessible only by a nearly vertical climb of some 30 feet, the cave con- tains a number of chambers connected by narrow^ passages. Bones were encountered in the deposits on the floor of the antechamber and of the first chamber beyond it. In the ante- chamber two layers are present, an upper, dark grey to black one some six inches thick, and a lower, reddish-brown one eight to ten inches in depth. The upper layer is either lacking or negligible in depth in the first chamber, the lower there attaining a thickness of six to twelve inches. Scanty remains from the upper dark layer include Rattvs, which clearly indicates post- Columbian age. No introduced forms were encountered in the reddish-brown layer, which contains indigenous rodents and the three species of Nesopliontcs- known to have inhabited the island, together with scanty remains of bats, birds, lizards, and solenodontids. In addition to the bones of the extinct form de- scribed below, one specimen of Solenodon paradoxus was ob- tained. This is the left horizontal ramus of a young individual 1 Partly supported by a Sigma Xi-RESA grant-in-aid. -N. paramicnis, luipomicrus and zamicrus of Miller (1929). A', '•panimicru.s," the largest of the three, agrees in size with the Cuban 2f. micrus, from which Miller separated it on molar characters. I have examined several hundred speci- mens of the Hisiianiolan form and compared them with Cuban material. The supposed differences are not constant and I have so far been unable to find others that might validate Miller's species. 2 BREVIORA No. 165 with lo in process of eruption and alveoli of the other teeth (M.C.Z. no. 7260) ; as far as I am aware this find constitutes the first fossil (or subfossil) record of the species. Judging from the good preservation and completeness of most of the bones from the reddish-broAvn layer, Dr. Ray (MS field notes) is inclined to doubt that their occurrence is a result of owl-roost accumulation. Most of the material, including all the solenodontid remains, was found in the first chamber of the cave. As regards the age of the layer, all that can presently be said is that it is almost surely pre-Columbian. During their stay in the Dominican Republic, Drs. Ray and Rand were accompanied by Professor Eugenio de Jesus Marcano F., Universidad de Santo Domingo, whose aid was invaluable in all phases of the work. It is a pleasure to name the extinct solenodontid in his honor. Dr. Ray will discuss the rodents from this and other localities in the course of his comprehensive studies on the Antillean Rodentia. I am indebted to Miss Linda Loring for the cleaning, sorting and cataloguing of the other groups represented in the collection. The drawings are by Mrs. Dorothy Marsh. T have been fortunate indeed as regards com- parative material, thanks chiefly to the excellent series of Solenodon in the Mammal Department of the museum. Thirty- six mandibles, fifteen humeri and six ulnae of S. paradoxus and three mandibles of ^. cuhanus have been available. I have not seen limb bones of S. cuhanus, but published figures (e.g. Peters 1863) indicate that these do not differ appreciably, either in structure or in proportions, from those of the larger species. INSECTIVOEA SOLENODONTIDAE Antillogale^ gen. nov. Type: — A. marcanoi sp. nov. Distribution: — Quarternary, Hispaniola. Diagnosis: — Differing from Solenodon as follows: P4 and lower M smaller relative to size of jaw, lower M with lingual cleft between paraconids and metaconids deeper, paraconid wings directed more anteriorly, paraconids consequently farther from metaconids, low ridge between bases of paraconids and metaconids isolating slight valleys at bases of trigonid basins, anterior ] The Antilles, plus 70X77 , -weasel. 1962 EXTINCT SOLENODONTID INSECTIVORE 3 ciispules below bases of paraconids very weakly developed, heels less broadly shelf-like labially. Post-dental portion of ramus larger relative to anterior portion. Ilnmeriis and ulna shorter, mueh wider relative to length, ulna with pit proximo-medial to sigmoid notch. ^e' Antillogale marcanoi sp. nov. Type: — M.C.Z. no. 7261, incomplete right ramus of mandible with P3-M0 and alveoli of other teeth. Hypodigm: — Type and the following specimens: M.C.Z. nos. 7262, incomplete left ramus with P4-M2 and alveoli of Po, M3 ; 7266, posterior portion of left ramus with alveoli of Mo, juvenile ; 7263, right humerus lacking ectepicondyle ; 7264 ; left humerus lacking proximal epiphysis; 7265, right ulna. Horizon and locality: — Late Pleistocene or Recent; unnamed cave 2 kilometers SE of Rancho La Guardia, Municipio de Hondo Yalle, Provineia de San Rafael, Republica Dominieana. Diagnosis: — As for the genus. Description: — Knowledge of the lower incisors, canine and anterior premolar derives solely from alveoli. On this evidence, the relative sizes of the incisors and the degree of enlargement of I2 were essentially as in Solenodon, as was the relative size of the single-rooted canine. The anterior premolar would appear to have been larger than P^, to about the degree seen in S. para- doxus. P3 is small relative to P4, nearly as much so as in ^. paradoxus, and much smaller, both actually and relatively, than in S. cnhanus. In structure it is very similar to the corresponding tooth of S. paradoxus, and rather different from the larger, more globular one of S. cuhanus. The paraconid of P4 is a larger cusp than in the majority of specimens of S. paradoxus and much larger than in *S^. cuhanus. The essential characters of the molars have been given in the diagnosis and can be seen in the figures : I may add that, as far as these teeth are concerned, Antillogale could only with difficulty be distinguished generically from Apternodus. The horizontal ramus is shorter and more slightly built than in Solcnodon and, instead of the two, or even three, mental foramina almost invariably present in that form, there is only a single large one, situated beneath P3. The postdental portion of the mandible is relatively large and robust. "Whereas the horizontal ramus is rather shorter than in S. cuhanus the pos- terior part is rather larger and longer, being intermediate in BREVIORA No. 165 1 rfl a Figurfi 1. a) AntiUogale marcanoi, type, M.C.Z. no. 7261, incomplete right ramus, dorsal view, b) Solenodon paradoxus, M.CZ. no. 12384, crown view of right P3-M2. X 3. these I'espects between cuhanus and paradoxus. The coronoid procesis, so far as can be judged from the juvenile M.C.Z. 7266 and the incomplete M.C.Z. no. 7262, appears to have been less tapering than in the living species. The base of the anterior border of the process is convex, as in S. cuhanus, and not ex- cavated as it is in S. paradoxus. The masseteric fossa is small, 1962 EXTINCT SOLENODONTID INSECTIVORE relatively and actually, oval in outline, and not very sharply defined. The marginal process, for the insertion of M. digastricus, is comparable to those of the living forms. The angle is not complete in any of the specimens but enough is preserved in M.C.Z. no. 7262 to demonstrate that, in contrast to S. pa7'adoxiLs and in agreement with S. cuhanus, the ventral border is not a Figure 2. AntillogaJe marcanoi. a) M.C.Z. no. 7261, type, incomplete right ramus, lateral view, b) M.C.Z. no 7266, incomplete left ramus, lateral view, X 2. downcurved and that, consequently, there is no lunate notch between angle and marginal process. The limb bones differ markedly in proportions from those of Solenodon. The humerus of Antillogale may be summed up as a bone having the width but not the length of that of S. para- doxus, and hence massive. The relative lengths of humerus and ulna appear to be essentially the same in both. The ulna, like 6 BREVIORA No. 165 the humerus, is shorter than but equally as wide as that of the living species; other differences of note between the two are the presence in the fossil of a very distinct pit proximo-medial to the sigmoid notch and of a rugose interosseous border. Figure 3. Lateral views of left rami of a) Solenodon panidoxus, M.C.Z. no. 34828, L) Antillogale marcanoi, M.C.Z. no. 7262, and e) Solcnodon cubanus, Y.P.M. no. 1203. In b) areas in solid outline are restored from M.C.Z. nos. 7261 and 7266. X 3/2. 1962 EXTINCT SOLENODONTID INSECTIVORE ^•^f% Figure -4. Anterior vie^vs of humeri (a-e) and ulnae (d, e). Antillogale marcanoi, a) M.C.Z. no. 7264, b) M.C.Z. no. 7263, (!) ]\[.C.Z. no. 7265. Sohnodon pdmdoxus, c) and e) ^M.C.Z. no. 12416. X 4/3. 8 BREVIORA No. 165 Measurements in mm. M.C.Z. nos. 7261 7262 Ps, length 2.7 — width 2.0 — P4, length 3.1 3.2 width 2.3 2.5 Ml, length 3.0 3.4 width 2.9 — M2, length 3.3 3.4 width 3.0 3.3 Depth of ramus beneath Mi 6.5 — 7263 7264 Humerus, total length 40.0 — a.-p. diameter of proximal end 9.9 — tr. diameter of proximal end 10.9 — a.-p. diameter at center of shaft 7.1 6.6 tr. diameter at center of shaft 7.3 6.5 a.-p. diameter of distal end 5.1 4.9 tr. diameter of distal end — 17.7 tr. diameter of trochlea — 9.2 7265 Ulna, total length 45.6 a.-p. diameter of olecranon 4.8 tr. diameter of olecranon 7.4 tr. diameter of proximal articular surface 6.4 a. p. diameter at center of shaft 4.3 tr. diameter at center of shaft 2.3 a.-p. diameter at distal end 3.9 tr. diameter at distal end 2.8 Discussion: — Tlie Solenodontidae have thus far been known from one genus with two species. Splitting of these taxa has of course been attempted. Soletwdon cubamis has been made the type of Aiopogale by Cabrera (1925), and Barbour (1944) has described a second Cuban species, S. poeyanus, based on pelage characters. Barbour explicitly stated that he could detect no cranial or dental distinctions. Aguayo (1950) and Koopman and Ruibal (1955) have regarded this form as a subspecies of cuhanns. The latter authors noted that some fragmentary remains (M.C.Z. no. 7054) from a cave in Camaguey, which they referred to the living species, were somewhat larger than Recent specimens of cubanus examined by them. In recent years, members of the Sociedad Espeleologica de Cuba have obtained further material, including some skulls, from the Province of Havana. Thanks 1962 EXTINCT SOLENODONTID INSECTIVORE 9 to Sr. Oscar Arredondo, I have seen most of these specimens ; they also represent individuals rather larger than, although not other- wise different from, any in the small sample (3) of cuhanus available to me. It is of some interest that the type of poeyamcs (M.C.Z. no. 6957) is also rather larger than this sample, a point not mentioned by Barbour; in fact there is agreement in this respect between the type and these fossils (or subfossils). To conclude from this that poeyanus is distinguishable from cuhanus on the basis of size and that the fossils so far found are referable to the former would, I think, be premature. The available series are too small to rule out accidents of sampling. The size differ- ence is slight and if the size range of cul)mius is comparable to that shown for paradoxus by the adequate series at hand we could be dealing simply with segments of a normal distribution. Reinforcing caution is the fact that one of the two mandibles from Maisi, Oriente (M.C.Z. no. 10065), mentioned by Allen (1918) is smaller than the other fossils. Whatever the solution of this minor problem may prove to be, I agree entirely with Aguayo and with Koopman and Ruibal that subspecies, at most, are involved. There is at present no good evidence of more than one species of Solenodon in Cuba. Atopogale is recognized, either as a genus or as a subgenus, by some authors. There can be no question that paradoxus and cuhanus are clear-cut taxa but they seem to me, as to others, to merit no more than specific rank. AntiUogale marcanoi helps to clarify matters here ; it is sharply distinct from either of the living species and the latter share most of the characters that differentiate them from it. AntiUogale is quite evidently a member of the Solenodontidae. The structure, although not the proportions, of the known limb bones ; the general structure of the mandible, and especially the presence of a marginal process; the small Ij and the degree of enlargement and mode of implantation of I2 (as revealed by the alveoli) — all these characters combine to place this con- elusion beyond reasonable doubt. Only in lower molar structure is there a closer resemblance to members of another family, the Apternodontidae. The difference here I believe to be the result of specialization in the Solenodon phylum. The molars of the two living species give the impression of having undergone anteroposterior compression of the trigonids. A majority of the characters in which they differ structurally from those of Aniillogale could be directly correlated with such a change. Aside from molar structure, the differences between the extinct and the living species are chiefly of a proportional nature, as 10 BREVIORA No. 165 pointed out above. AntillogaJe evidently had a somewhat shorter facial region and much shorter, more heavily built fore limbs than Solcnodon. It may well have been more fossorial in habits. Solowdon with its tAvo species, one on each of the larger West Indian islands, has always been one of the most isolated of mam- mals, zoogeographically — and even taxonomically — speaking. As knowledge of the past history of mammals has improved, this isolation has become, if anything, even more apparent. In earlier days it was at least possible to assume a common ancestry with if not membership in the Tenrecidae. There is now no real evidence for such an assumption — tenrecids, as far as the record goes, appear always to have been African and Madagascan in distribution — and we must look elsewhere for close relatives and possible ancestors. The only group that seems to me to come close to fulfilling the requirements is the Apternodontidae, a family which at least had the merit of inhabiting North America, the only probable source area.^ The described forms, the Oligocene Aptcrnodus and Oligorycfes, were certainly not ancestral to the solenodontids. but apternodontids are now knoAvn in North Amer- ica at least as far back as the Bridgerian Eocene." The possibility, I would go so far as to say probability, exists that solenodontids were derived from relatively unspecialized apternodontids that inhabited the Central American peninsula during the earlier Tertiary. Rafting of the ancestral stock to the Antilles, for rafting was certainly involved, may have taken place in the later part of the Eocene, at roughly the same time as the rafting of the ancestors of caviomorph rodents and platyrrhine primates to South America. This event, if it occurred at the time suggested, would have insured for the solenodontids a very long residence in the West Indies. There has, I feel, been some reluctance to accept such a possibility, a reluctance based, consciously or un- consciously, on the fact that Solenodon alone has hitherto repre- sented the family and on the assumption that we now have an adequate idea of the Pleistocene, or at least pre-human, faunas of the archipelago. I strongly doubt if we do have a good 1 Allen's belief (1918) that solenodontids were derived from nesophontids. a vifw recently supported liy McDowell (]ft.")S). is not at all convincing'. McDowell's attemnt to ri^id Aiitenifiiiiiit out fit the Insectivorji is ntti'rlv unrealistic, being in large part based on misinterpretation of the fossils he examined. 2 I am indel)ted to Dr. Craig' C. lUaclv for the opportunit.v of examining a speci- men from Tabernacle Bntte. Wyoming. This is tlic earliest kr.own entberiaii with zalambdodont molars. Contrary to earlier statements (including one of mine made on the b;isis of the literature), the molars of the I'aleocene I'dliirn' riirtvft are not zalandidodont in structure, nor are they even pre-zalaiubdoatross" during 1891, Lieut. -Commander Z. L. Tanner, U.S.N., commanding. XXVI. The Fishes. Mem. Mus. Comp. Zool. Harvard Coll., 24: 1-431, pis. 1-88, A-N. HuBBs, Carl L. 1958. Ogcocephalus darwini, a new batfish endemic at the Galapagos Islands. Copeia, 1945 (3) : 161-170, pis. 1-5. BREVIORA Mnasemm of Cooipsirative Zoology Cambridge, Mass. September 5, 19B2 Number 167 Bathyclupea schrocderi, a New Bathyclupeid Fish from the Weslerii Tropical Atlantic By Myvanwy M. Dick Museum of Comparative Zoology, Harvard University The species described below is based on specimens collected by the Atlantis during the Harvard-Havana Expedition to Cuba, and by the Oregon (exploratory vessel of the U.S. Bureau of Com- mercial Fisheries) during her work in the Gulf of Mexico. To Mr. Harvey R. Bullis, Jr., director of the Oregon's program, go my thanks for representatives of both western Atlantic species of Bathyclupea. I am also indebted to Mr. Loren P. AVoods of the Chicago Natural History Museum, and to Dr. Leonard P. Schultz, U.S. National Museum, for the loan of comparative specimens of related species and for additional representation of the new form. Bathyclupea schroederi, new species . . Holotype: A specimen 140 mm. in standard length, taken at Atlantis Station 2987, 23°23'N., 79°39'W., 280 fathoms. MCZ 41498. Paratypes: MCZ 39416, 4 specimens, Atlantis Station 2987, 23°23'N., 79°39'W., 280 fms. ; MCZ 39380, 1 specimen, Atlantis St. 2987a, 23°22'N. ; 79°39'W., 260 fms. ; MCZ 40600, 1 specimen, Oregon St. 2635, 17°37'N., 63°28'W., 220-235 fms.; CNHM 65145, 1 specimen, Oregon St. 1872, 16°41'N., 82°20'W., 300 fms.; CNHM 65147, 1 specimen, Oregon St. 1888, 16°41'N., 81°02'W., 250 fms.; CNHM 65146, 1 specimen, Oregon St. 1886, 16°55'N., 81°12'W., 275 fms. Diagnosis: The following species of Bathyclupea have been described : Bathyclupea hoskynii Alcock, 1891, from the Andaman Sea (the type species) ; Bathyclupea argentea Goode and Bean, 2 BREVIORA No. 167 1895, from Nevis, West Indies; Bathychipra ynnylayana Weber, 1934, from the Flores Sea, Indonesia ; Bathyclupea megaceps Fowler, 1937, from off Mindanao, Philippines ; and Bathyclupea gracilis Fowler, 1937, from off Makyan Island, Moluccas. Bathyclupea schroederi differs from all of these by havino- a spine and 37 to 39 rays in the anal fin (cf. 33 or fewer). It can also be distinguished from B. hoskynii by its less deep body (3.9 to 4.4 in standard length; cf. 3.3), and from B. argent ea and B. maylayana by the placement of the anal fin, which originates in the anterior half rather than in the posterior half. B. schroederi can be distinguished from B. megaceps by the relative length of the base of its anal fin, which is longer than the head in the new form but shorter in B. megaceps. The length of the head in B. schroederi, 3.3 to 3.4 in standard length, is shorter than that of B. gracilis (3.0 in length). Description: The description which follows is based on the holotype and paratypes, 120 to 144 mm. in standard length, listed above. Counts and proportional dimensions are provided in Table 1. Body laterally compressed, its greatest depth at origin of anal fin, 3.9 to 4.4 in standard length. The dorsal profile is nearly straight. There is a well defined, rather straight lateral line bear- ing 38 pored scales which are somewhat more adherent than the scales elsewhere on the body. Scales cycloid, deciduous, thin and relatively large. Head naked with large mucous cavities. Verte- brae 10-20-1, total 31. Head 3.3 to 3.4 in standard length, interorbital flat, 16.3 to 17.1 in standard length. Eye mainly in the posterior half of the head. Nostrils small, contiguous, almost superior. The mouth is nearly vertical. Minute conical teeth on the lower jaw, a single row of minute teeth on the vomer and premaxillary. The anterior end of the vomer protrudes into the mouth, frequently extend- ing forward to beneath the tip of its snout. There are four gill arches. Moderately long gill rakers, with .slight protuberences, on the first and second arches ; rakers much reduced and stublike on the third and fourth arches. Gill rakers on the first arch 3 + 16. The gill opening is wide, the branchi- ostegal membranes free from one another and from the isthmus. Pseudobranchiae present. The pectoral fin is large, pointed, the upper rays the longest, about 3.5 in standard length. The ventral fins are anterior to the pectorals, short, very slightly separated. The rays are rather 1962 NEW BATHYCLUPEID FISH like the ribs of a fan, fitting into a shallow groove when con- tracted. The dorsal fin originates in the postmedian half of the body, the anal fin in the anterior half. Color in alcohol: Opercle and abdominal wall dusky. In life the general color was silvery. It is a pleasure to name this species for William C. Sehroeder of the Department of Fishes, Museum of Comparative Zoology, who collected the holotype and several of the paratypes. His many publications, the collecting he has done, and his work in the department have greatly enriched the field of ichthyology. TABLE 1 Counts, and proportional dimensions expressed in percent of standard length, of the fiolotype and paratypes of Batfiyclupea schroederi. M 5 1 rvj-o CO ^ CT- en 1^ S in §2 Standard lengtfi (mm.) 144 144 140 139 137 137 130 120 116 Greatest deptfi of body (percent) 27.8 26.4 25.0 26.4 25.5 27.0 24.6 25.0 23.3 Greatest depth of body at midpoint 26.4 23.6 25.0 25.0 24.1 24.1 23.1 25.0 23.3 Least depth of caudal peduncle 10.4 9.7 9.3 9.5 9.5 8.7 9.2 9.2 8.6 Greatest width of body 10.4 9.7 7.9 9.5 9.5 8.7 7.7 9.2 7.8 Greatest width of head 12.5 11.8 12.2 12.2 12.4 11.7 11.5 12.5 11.2 Snout to origin of dorsal fin 54.2 55.6 55.0 53.6 54.7 53.2 54.6 50.8 51.7 Snout to origin of anal fin 43.8 43.1 42.8 42.8 45.2 43.8 43.1 47.5 44.8 Snout to insertion of ventral fin 27.8 27.8 28.6 28.6 31.4 26.9 29.2 29.3 Snout to insertion of pectoral fin, ventral angle 28.5 27.8 27.8 29.3 29.9 29.9 26.9 28.0 29.3 Length of base of dorsal fin 10.4 10.4 10.7 10.6 10.9 10.9 10.8 10.0 10.3 Length of base of anal fin 51.4 49.4 47.1 49.3 48.2 51.1 48.4 48.3 50.9 Distance between insertion of ventral fin and origin of anal fin 20.7 20.2 17.8 17.8 20.4 17.7 16.7 17.2 Length of head 28.5 28.5 30.0 29.3 29.2 30.6 27.7 30.0 32.0 Length of snout 8.3 8.3 8.6 8.6 8.7 8.7 8.5 9.2 7.8 Greatest diameter of eye 11.8 11.1 11.4 11.4 11.7 11.7 10.8 11.7 11.2 Width of interorbitai space 5.6 5.6 5.0 5.0 5.1 5.1 5.4 5.8 6.4 Length of upper jaw 13.2 13.2 12.2 12.2 12.4 13.1 12.3 12.5 12.9 Counts: Dorsal fin 1-9 1-9 1-8 1-9 1-8 1-9 1-9 1-8 1-9 Anal fin 1-37 1-38 1-39 1-37 1-38 1-39 1-38 1-39 1-37 Pectoral fin 1-28 1-28 1-29 1-28 1-28 1-28 1-29 1-28 1-29 BREVIORA No. 167 X:: o M en 00 Ci T-l N O > o Hi K5 D O BREVIORA Museium of Coimpsirative Zoology Cambridge, Mass. Sp:i'temher 7, 19()2 Number KiS TWO NEW SPECIES OF FOSSIL TALPID IXSECTIVORES By Katherine M. Reed During work on the talpid siilifamily Proscalopiiiae, two new sealopine moles came to my attention and are described in this note. I am grateful to ^Ir. Richard Tedford, University of Cali- fornia Museum of Paleontology, and Dr. C. W. Hibliard, Uni- versity of Michigan Museum of Paleontology, for the loan of the material. The work was carried out at the ]\Iuseum of Compara- tive Zoology. I am obliged to Professor Bryan Patterson for critical comment and to ^liss Barbara Lawrence for access to the collections of Recent inseetivores. I also appreciate the com- ments and assistance of Dr. J. R. Macdonald and Dr. Mary Dawson. The illustrations were made by Mr. Richard Stafford. The following abl)reviations are used : UCMP — University of California ^luseum of Paleontology UMMP — University of IMichigan ^Museum of Paleontology 1., w., trig., tal. — length, width, trigonid, talonid TALPIDAE Subfamily Scalopinae DOMXIXOIDES Green 1956 DOMNINOIDES VALENTINENSIS n. Sp. Type: UCMP 33152, right ramus with Pj, P;.-4. Mo-. Hrjpodigm: Type and UCMP nos. 36150-36157, including iso- lated teeth, partial rami and some liml) bones. UCMP 29215 and 29215-A refer to limb material marked "float." Horizon and locality: Late ^Miocene, Valentine formation, from the quarries near the middle of the exposed Valentine formation 2 BREVIORA No. 168 at the Port Niol)rara locality on the quarter section line between the NW and SW (|iiarters of Sec. 24, T 34 N, R 26 W, Cherry County, Nebraska, UCMP locality Y3218 (See IMacdonald 1947). The specimens marked "float" are, according to 'Sir. Tedford, "nndonl^tedly Barstovian, but come from an undetermined horizon within the Valentine." Diagnosis: No diastema between P^ and P4 ; slightly larger and more robust tlian DonniiuoirJes ripareiisis; metastylid defi- nite on Mo.' Description : The only uj)per tootli in tlie material, and tlie first known to belong to this genus, is a broken, isolated P^. This tooth has a blade-shaped paracone. A very narrow anterior cingulum widens to a shelf-like lingual cusp and extends up the jiosterior side of the tooth to join the paracone crest posteriorly. The lingual cusp is Avidest opposite the stoutest jiortion of the paracone. The tooth has at least two, jiossibly three roots, two labial, one lingual. In the lower dentition, P^ is small, conical and single rooted. P2 is not represented in the material but is double rooted. P.^ is consideraldy larger than P^, double rooted as in D. rijxirnisis, and with a slight heel. It is situated very close to P.,, with no diastema between these teeth. P4 is double rooted with a very small anterior cuspule. It has a larger heel than Po., the heel sloping downwards labially to a small cingulum which connects the heel to the anterior cuspule. The main cusp is conical. Ml is not represented in the material studied; thus no com- jiarison with D. riparcnsis is ])ossible in this resj^ect. Relative root sizes are in accord with this species. In ^2 the paraconid is smaller than the metaconid, which is the highest lingual cusp. The entoconid is the stoutest lingual cusp. A crest, the crista obliqua. runs from the hypoconid to the metastylid, as in D. ripa)'e)isis. The protoconid is higher than the hypoconid and is slightly labial to the metaconid. There is a large anterior cin- gulum, wide labially and with an irregular border, that extends around the paraconid to the opening of the trigonid valley. The posterior cingulum is small and does not reach to the lingual' face of the tooth. The talonid has a narrow opening. In M.-., tbe size relations of the cusps are as in M^. All exami)les of this tooth iDr. J. R. Macdoiiald has kiiullj- sent iiip a eopv of a forthcoming nianuscript in which lie clpseril)ps a new species of D':i)uiinoi(l' x. D. rnlriitiiirtisis is ilistiii- guished from this species by lacking any trace of cingnla on the lingiinl face of the molars and l>y the crista obli(|iia nnining to tlie metastylid rather than to the metaconid. 10()2 NEW FOSSILS TALPIDS 3 are worn, hut there are stroiifr supgestioiis of a metastylid and of a relatively wide anterior ein<»:uhnn. Xo posterior einfrulmn is present. The talonid is relatively longer than in ^Iv. There are usually two mental foramina which vary in posi- tion. In UCMP 36151, they are below the anterior roots of P4 and between P4 and Mi ; in Ur^IP 33152, between P2-.S and the roots of P4. In IK'^IP 36152. there is only one foramen, l)elow the anterior root of P4. In D. riparensis, the foramina ai-e below P;{ and between the roots of P4. The ramus is mueh like that of D. riparc7isis and is in general similar to that of Macdonald's species, although slightly stouter than in either. Some comparison of the limb material of the fossil has been made with modern species. It must be noted that the majority of tile limb material is "float" and association with the teeth could ])erhaps be doul)ted: the extremely talpid-like nature of both does, however, strongly suggest association. The humeri of the fossil are very similar to but slightly smaller than those of Talpa europaea. They are larger than in Parascalops hreweri and wider than in Condylura. Discussion: Macdonald has suggested (j^ers. comm.) that the jaws of his new species may represent the lower dentition of Proscalops sccundiis. This now seems extremely unlikely, first because of the discovery of P^ of Domninoides valentinensis. which is very unlike that of the Proscalopinae, and second, the greater development of the anterior cingula on lower molars than I would expect on the basis of the known trends in the Proscalopinae (See Reed 1961). Wilson (1960) has recently de- scribed lower molars which he assigns to Proscalops sp. cf. P. secundus; these are much nearer to the proscalopine type of molar than to that of Domninoides. It is evident that Doiiini)ioides is a talpid, not a soricid as Green (1956) originally described it. ]\Ieasurements [^CMP 33152 M. UCMP 36152 .AIo UCMP 33152 M3 1. 3.1mm 1. 3.2 mm 1. (approx.) 2.4 mm w. trig. 2.5 w. trig. 2.2 w. trig. 1.9 w. tal. 2.7 w. tal. 2.7 w. tal. 1.6 depth of jaw below :\Ii, UCMP 33152 : 3.7 mm. 4 BREVIORA No. 168 HesPEROSCALOPS Hibbard 1941 Hesperoscalops sewardensis n. sp. Hesperoscalops rexroadi Hibbard 1953, p. 23, fig. ID. Type: UMMP 27276, partial right ramus with M2-3, fragment of a left ramus, anterior part of a left ramus with worn Mj, ulna, part of a scapula, and a femur. Iliipodigm : Type only. Horizon and localitij: Late Pliocene or early Pleistocene, from area of Saw Rock local fauna, NE 1/4 Sec. 35, T 34 S, R 31 W, Seward County, Kansas. Diagnosis: Distinctly larger than H. rcrroadi and with greater develo])ment of basal accessory cusi)ules. Description and discussion: The difference in size between the new species and specimens of H. rexroadi in which the teeth are in nearly the same state of wear indicates that these cannot be wear differences in the same species and that the two are distinct. Although the cusp pattern is very similar, as Hibbard states (1953, p. 23), the anterior cingular cuspule on Mo is larger and better developed than in H. rexroadi, as is the anterior basal ac- cessory cuspule on ^lo. If Hesperoscalops rexroadi is ancestral to Scalopus, this new, but closely related species must represent an extinct side line, for //. seicardensis is considerably larger than Sealojjus aquaticiis and has much better developed basal cuspules and cingula. ^Measurements H. sewardensis UMMP 27276 M. UMMP 27276 M3 1. 3.0 mm w. trig. 2.5 w. tal. 2.6 depth of jaw below Mo on lingual face : 4.5 mm. length of Mo -3 : 5.8 mm. H. rexroadi UMMP 27278 Mo UMMP 27278 M3 1. 2.3 mm 1. 2.2 mm w. trig. 1.9 w. trig. 1.6 w. tal. 2.1 w. tal. 1.5 depth of jaw below jMo on lingual face : 3.2 nun. length of M2-3 : 5.0 mm. 1. 2.8 mm w trig. 2.4 w tal. 1.7 1962 NEW FOSSILS TALPIDS 5 REFERENCES Green, M. 1956. The Lower Pliocene Ogallala-Wolf Creek vertebrate fauna, South Dakota. Jour. Paleontology, vol. 30, no. 1, pp. 146-169. HiBBARD, C. W. 1953. The insectivores of the Rexroad fauna, Upper Pliocene of Kan- sas. Jour. Paleontology, vol. 27, no. 1, pp. 21-32. Macdonald, J. E. 194:7. A new shrew from the Niobrara River, Upper Miocene of Ne- braska. Am. Jour. Sci., vol. 245, pp. 123-126. 1962. The Miocene fauna from the Wounded Knee area of western South Dakota. Bull. Am. Mus. Nat. Hist, (in press). Reed, K. M. 1961. The Prosealopinae, new sul)family of talpid insectivores. Bull. Mus. Comp. Zool., vol. 125, no. 14, pp. 473-494. Wilson, R. W. 1960. Early Miocene rodents and insectivores from northeastern Colo- rado. Vertebrata, Univ. of Kansas Paleontologieal Contribu- tions, article 7, pp. 1-92. BREVIOBA No. 168 , X 'A , CO ^< CO . lO X ft CI co' >n a o . t>-_ o • r-i CO •rH a o o o & 03 o • r— 1 Qj o cu P( > C3 9 05 o o o ai M ■r. <1^ k; >o lO l- 1^ Cl I-H 1— 1 co' ' « " > « ^ -M ^ be rq CO & c^ u) > .s « ;_i f—^ 1— 1 ^ -4— '^-J ^4h ? •/: ■■K O O p p _bc ^ ^ o c s rH Cj o .. .*" r^ 2 -t- P v£ K* >■ -J—' •"^ ■^ )-^ t~ ^-A -f- w o 9 PLI -a "3 < to" i-H CO 10 10 ?— 1 ^ g oi tH CO rH t^ CO Cl CO Ph CO TO CO ■2" O Oh o Ph Hi CO '^O ^ s •S ■5 to "32 4j to o ^ SC C) '^ ?^ ^ =0 •^ ^ ■^ . cj ci =c "TS o ••!* 10 la •^ 1 o ^ h^ s bil ^ i--' p s o s bi bi T— 1 S s bi P^ . ■■'■' be Ph BREVIORA Miiseiuim of Coeiparsitive Zoology Cambridge, Mass. October 15, 1962 Number 169 NEW RECORDS OF INSHORE FISHES FROM THE ATLANTIC COAST OF PANAMA By Ira Rubinoff^ and Roberta W. Rubinoff During February, March, and April, 1961, a collection of in- shore fishes from both coasts of the Isthmus of Panama was made in connection with a studj' of the effects of geographical isola- tion on fish speciation. This collection consisted of 2095 speci- mens of 136 species. A total of 907 specimens of 47 species was taken from the Pacific coast and 1188 specimens of 89 species were collected on the Atlantic coast. Fourteen species in the latter collection Avere found to represent extensions of known ranges and some of these are sufficiently abundant to be consid- ered resident members of the Panamanian fauna. The purpose of this paper is to report the new records from this area. Notes are also included on three species represented in previous collec- tions from Panama by only one or two specimens each. All the specimens discussed herein are in the collections of the Museum of Comparative Zoology. Rotenone poisoning of coral pools and lagoons was the princi- pal method of collecting, although occasionally sandy beaches were seined. All of the new records are from coral reef areas, particularly from the reefs bordering the Galeta Point Naval Station near Coco Solo in the Canal Zone. Most of the smaller eels were collected when the incoming tide overflowed poisoned coral pools onto the exposed surfaces of the reefs and into small crevices and burrows in the coral. When the poison reaches these burrows the inhabitants are driven onto the flat of the reef and are easily collected. Many young speci- mens of Muraenidae as well as most of the specimens of Morin- guidae and Echelidae were collected in this manner. iThis research was supported in part by a Public Health Service Training Grant to the Department of BioloRy at Harvard University. Cambridge, Massa- chusetts, and in part by a fellowship from the Woods Hole Oceauographic Insti- tution during the summer of 1061. Contribution No. 12.'5,'? of the Woods Hole Oceauographic Institution 2 ]!REVIORA No. 169 In their classic work on the marine fishes of Panama, Meek and Hildebrand (1923, 1925, 1928) collected 23G species from the Atlantic coast, of which only five were eels. In other collec- tions from this area, the apodes are one of the poorest repre- sented groups. Numerous species are known to range along the Atlantic coast north and south of Panama but have yet to be recorded there. Of the nine species of apodes in our collection seven represent new records for this area. Specimen lengths are standard lengths to the nearest milli- meter in all cases except the eels, for which total lengths are given. The authors wish to thank the following people for making this part of our study possible : Dr. Sydney Galler and Mrs. Helen Hayes of the Biology Branch of the Office of Naval Re- search, Dr. L. P. Schultz of the United States National Museum, and Dr. Martin Moynihan and Mrs. Adela Gomez of the Canal Zone Biological Area. Dr. James E. Bohlke of the Philadelphia Academy of Sciences most kindly identified some of the eels. SPECIES NOT PREVIOUSLY RECORDED FROM THE ATLANTIC COAST OF PANAMA XENOCONGRIDAE Kaupichthys atlanticus Bohlke MCZ 41455-58 Five specimens, 50-205 mm. long, were taken from coral reefs at Galeta Point. These specimens agree with the description of Bohlke (1956). Range : Tropical western Atlantic from Bermuda to south of Jamaica. MORINGUIDAE Aphthalmichthys mayeri (Silvester) MCZ 41459-61 One specimen, 275 mm. in length, was taken from the edge of a reef southwest of Las Palmas mountain about two -thirds of the distance from ]\Iaria Chifpiita to Porto Bello. This fish had a dis- tinctly pink head and pink-orange body. This color faded to a dull yellow two days after preservation. Three specimens, 150- 330 mm. long, were taken from the exposed surface of Galeta Point Reef. Range : Bermuda, Florida, Puerto Rico. 1962 NEW RECORDS OF PANAMANIAN FISHES 3 ECHELIDAE Myrophis egmontis Jordan MCZ 41462-64 Twelve specimens, 81-231 mm. in leno-th, were obtained from Galeta Point. These specimens may be differentiated from Eche- lidae previously recorded from Panama by the following char- acteristics : origin of dorsal fin behind the vent, and absence of teeth on the vomer. These eels were found in situations similar to those of Aphthalmichthys mayeri. Range: Florida, Bahamas, West Indies. MURAENIDAE Enchelycore nigricans (Bonnaterre) MCZ 41465-67 Seventeen specimens of this eel. 81-545 mm. in lenp'th, were taken from the reefs at Galeta Point. The slitlike posterior nostril diagnostic of Enchelycore is an unreliable character in separating young specimens from other Muraenidae. Separation of the young of E. nigricans from Gymnothorax moringa which it closely resembles, was facilitated b}' two characters. The anterior nostrils of Enchelycore are shorter than those of G. moringa and the upper jaw of Enchely- core has a series of 4-6 long canine teeth medial to the outer row of teeth on both sides. All Gymnothorax moringa which we ex- amined have 1-3 teeth in this series. Range : Bermuda, West Indies. Gymnothorax viciniis (Castelnan) MCZ 41468,69 Four specimens, 145-330 mm. in length, were taken at the Galeta Point reefs. Range: Bermuda, West Indies to Brazil, Cape Verde Islands and Africa. Gymnoth&rax moringa (Cuvier) MCZ 41470-73 The collection contains nine specimens, 64-185 mm. in total length, taken from the reefs at Galeta Point. Range: Atlantic coast of America from Florida to Brazil, Ber- muda, Bahamas, West Indies, St. Helena. Uropterygius Bohlke (n. sp.. in manuscript) MCZ 41475 4 BREVIORA No. 169 One specimen 190 mm. long was collected at Galeta Point. Our specimen was identified by Dr. James Bohlke as a species he is currently describing and it will be designated as a paratype. AULOSTOMIDAE Aulostomus maculatiis Valenciennes MCZ 41474 One specimen, 185 mm. long, was taken at Galeta Point. Range : Bermuda, Florida, Gulf of Mexico, Central American coast of Caribbean, Bahamas, West Indies. HOLOCENTRIDAE Holocentrus coruscus Poey MCZ 41476 One specimen, 45 mm. long, from Galeta Point, has the folloAv- ing meristics: D-XI, 12; A-IV, 8; gill rakers 9. Color of fresh specimen : dorsal surface of head red ; body red and white lateral stripes ; caudal fin red ; anal and second dorsal dark red at distal portions ; interspinous membranes of dorsal peppermint striped, black spot distally between first three dorsal spines. With the exception of the number of gill rakers this specimen fits the de- scription of H. coruscus by AVoods (1955) in his revision of the Western Atlantic species of IIoloce7itrus. Range : Bermuda, Florida, Bahamas, West Indies. r fVPOGONIDAE Apogonichthys stellatus Cope MCZ 41477 Two specimens, 12 mm. and 29 mm. in length, were taken from the edge of a reef southwest of Las Palmas mountain about two- thirds of the distance from Maria Chiquita to Porto Bello. Range : Bermuda, Florida, Bahamas, West Indies. LABRIDAE Thalassoma hifasciatum (Bloch) MCZ 41478-83 Fifty-eight specimens 19-83 mm. in length, were taken at Galeta Point. Although this species has not been previousl^v recorded from Panama it is one of the most abundant representatives of the reef pool fauna. Many more examples were seen than were collected. Range : Bermuda, Florida, Bahamas. West Indies. Honduras. 1962 NEW RECORDS OF PANAMANIAN FISHES 5 CANTHIGASTERIDAE Canthigaster rostratus (Bloch) MCZ 41484, 85 We collected eleven specimens, 19-42 mm. in length, at Galeta Point. Ten of these specimens have the fin formula D-10, A-9 ; one specimen has D-9, A-9. These specimens agree with the de- scription of Breder (1927). Jordan and Evermann (1898), Evermann and Marsh (1900), and Nichols (1930) report speci- mens with the fin formula D-6, A-8. For a partial explanation of this discrepancy see Breder (1927). Range : Bermuda, Florida, West Indies, Venezuela and Madeira. CLINIDAE Lahrisomus I'alislierae (Jordan) MCZ 41486, 87 Six specimens, 25-68 mm. in length, were found in Galeta Point reef pools. One 56 mm. specimen taken in the first week of April possessed enlarged ovaries from which ova 0.5 mm. in diameter were obtained. A 54 mm. male also taken at this locality had enlarged testes. Range : Florida to Brazil. ^tr-*^ Ldhrisomus nigricinctns Rivero MCZ 41488 One male specimen 45 mm. long was taken at Galeta Point. It was compared with the holotype, MCZ 34150, and with the de- scription given by Springer (1958). Springer (1959) reports the range extension of L. bucciferus and L. guppyi to the Atlantic coast of Panama. L. nuchipinms reported by Meek and Hildebrand (1928) and our specimens of L. kalisherae and L. nigricinctns bring the total recorded number of Atlantic Panamanian Lahrisomus species to five. SPECIES RARELY RECORDED FROM THE ATLANTIC COAST OF PANAMA With the exception of Dinematichthys cayorum these species were not collected by Meek and Hildebrand (1923, 1925, 1928). BLENXIIDAE Rupiscartes atlanticus (Cuvier and Valenciennes) MCZ 41489-91 6 BREVIORA No. 169 Eight specimens, 39-74 mm. in length, from the coral reefs at Galeta Point are in the collection. The previous record of this species from Panama was a single specimen taken by Fowler (1916) at Toro Point. Range : Bermuda, Atlantic and Pacific coasts of tropical Amer- ica, West Indies. Salarichthys textalis (Quoy and Gaimard) MCZ 41492 Three specimens, 24-42 mm. in length, were taken at Galeta Point. One 31 mm. specimen was previously collected at Cale- donia Bay, Panama, by Breder (1925). Range : Bermuda, Florida, AVest Indies, Brazil. BROTULIDAE Dinematichthys cay or urn (Evermann and Kendall) MCZ 41493-96 Twelve specimens of this species (Ogilbia cayorum of Meek and Hildebrand), 26-51 mm. in length, were collected at Galeta Point. On April 6 some adults were found to contain well developed embryos which could be seen through the body wall. These were extruded in gelatinous strings when a slight pressure was exerted on the abdomens of the females. They were about 5-7 mm. in length and had small yolk sacs. When placed in a bucket of sea- water the embryos were free swimming although apparently pre- mature. Fowler (1916) and Meek and Hildebrand (1928) each found only one example of this species. Range : Bermuda, Florida, Bahamas. LITEEATUEE CITED BOHLKE, J. E. 1956. A synopsis of the eels of the family Xenocongridae (including the Clilopsidae and Chilorhinidae). Proc. Acad. Nat. Sci. Phila- delphia, 108: 61-95. Breder, C. M. 1925. Notes on fishes from three Panama localities: Gatun spillway, Eio Tapia and Caledonia Bay. Zoologica, 4 (4) : 137-158. 1927. Scientific results of the first oceanographic expedition of the "Pawnee" 1925, Fishes. Bull. Bingham Ocean. Coll., 1 (1): 1-90. Evermann, B. W., and M. C. Marsh 1900. The fishes of Porto Eico. U. S. Fish Conim. Bull, for 1900: 49-350. 1962 NEW RECORDS OF PANAMANIAN FISHES 7 Fowler, H. W. 1916. Cold-blooded vertebrates from Costa Eiea and the Canal Zone. Proc. Acad. Nat. Sci. Philadelphia, 68: 389-414. JoRDAK, D. S., and B. W. Evermann 189(5, 1898, 1900. The fishes of North and Middle America. Bull. U. S. Natl. Mus., 47 (1-4): 1-3313. Meek, S. E., and S. F. Hildebrand 1923, 1925, 1928. The marine fishes of Panama. Field Mus. Publ. Zool. Ser. Chicago, 15 (1-3): 1-1045. Nichols, J. T. 1929, 1930. The fishes of Porto Eico and the Virgin Islands. (New York Academy of Sciences) Scientific survey of Porto Eico and the Virgin Islands, 10 (2, 3): 161-399. Springer, V. G. 1958. Systematics and zoogeography of the clinid fishes of the sub- tribe Labrisomini Hubbs. Publ. Inst. Mar. Sci., 5: 417-492. 1959. A new species of Labrisomus from the Caribbean Sea, with notes on other fishes of the subtribe Labrisomini. Copeia, 1959 (4) : 289-292. Woods, L. P. 1955. Western Atlantic species of the genus Eolocentrus. Fieldiana : Zool., 37: 91-119. BREVIORA Mmseiuim of Comparative Zoology Cambridge, Mass. Xovkmhkk lo, \\)ii'2 Number 170 THE BRAIN OF THE EMU (DROMAEUS NOVAEHOLLANDIAE, LATH )' 1. Gross Anatomy op^ the Braix and Pineal Body' B>- Stanley Cobb and Tii.LY Edinger The histology of the cerebral hemisphere of the einii has been extensively studied by Craigie (1935a, 1935b, 1940) and three diagrams of the hemisphere have been published. Drawings of the whole brain have also appeared in the literature (Strong, 1911; Kiienzi, 1918), and a photograph was published hy Anthony (1928). No description of the whole brain, however, is to be found. Since the emu is, next to the ostrich, our largest living bird, and since it belongs to a taxonomically controversial group, it seem>s of value to describe the brain and compare it with the brains of other birds. Moreover, the emu is considered, by Py- craft (1900) and many others, to be one of the most primitive of birds. The concept of "primitiveness" will be considered in the discussion at the end of this paper. MATERIAL Three specimens of Droiiiacus novachollandiae were collected by S. J. J. Davies in November 1960 in Western Australia for Professor Ernst Mayr, Director of the Harvard Museum of Com- parative Zoology. Two of them were kindly given to us by Pro- fessor Mayr for neurological study. The lieads had been cut off 1 'I'liis spellirifT of Dronidciis is not the (iiic iU'CPiitpd by soJiii' newer cliecklists. but beciUise Dromiccius (an alternative si)elling) is the jierpetuation of a grapho- logical error (Newton, 1896) and because Dromaius is a less proper Latinization, it seems better to vise the older form. 2 From the Museum of Comparative Zoology and the Department of Neurology and Psychiatry, Harvard University, and the Laboratory for Psychiatric Research, Massachusetts General Hospital. This investigation has been aided by grants from the Foundations Fund for Psychiatry and the National Institute of Neurological Disease and Blindness, grant #03429-02. 2 BREVIORA No. 170 and skinned, the eyes liad been removed, and the speeiniens had been fixed in 10 per cent formalin solution in the field. After about two months the heads were packed in moist condition and shipped in cellophane bags to the Fnited States. Here the brains were removed from the skulls after making photographs of vari- ous stages of the dissections. The brains were then fixed in fresh neutral formalin solution (10 per cent) for a month. One brain (that of Emu #85) was divided into its com])onent ])arts for weighing and special histological studies. The other brain (Emu 4^104) was removed, photographed, fixed for a month in 10 per cent formalin as above, and embedded whole in celloidin for serial sectioning. Both are brains of adult males. The first (#85) appears somewhat larger and weighed 27.7 grams ; the second (#104) weighed 25.1 grams. DESCRIPTION The position of an avian brain within the skull is determined by many developmental factors. The most obvious are the shape of the bill, the size and position of the eyes, the habitual posture of the bird, and the size and shajie of the brain itself. Starck (1955) has given an excellent discussion of these relationships and emphasizes the importance of the size of the eye and the position of the orbit. One way of describing the position of the brain is to measure the angle between the cerebral axis and the axis of the bill (Cobb, 1959). In tlie emu this angle is about 27° (see Fig. 1), an angle somewhat smallei- than that of the gull {Larui< ar(/( itfafus: 84°) and the grouse {Boiiasa kdiIxHus: 36°), but distinctly greater than that of the cormorant {Phalacrocorax ourifu^: 15°) which has the sti-aightest (most extended) type of skull and an exceptionally small hraiii-bill angle. Besides showing the relation of the bi-ain to the skull. Figure 1 shows the olfactory bulb and memln-anous sac of the olfactory nasal chamber ; the bulb s(>ems to be in direct contact with the chamber, but closer scrutiny shows that there is a space bridged by the short olfactory nerves. When the light, dijiloic bone of the bill is removed, the sac which forms the lining of the olfactory chamber is revealed. It is a fairly tough structure containing blood vessels and many nerve fibers. It is crossed anteriorly by a branch of the first division of the trigeminal nerve. The main nerve trunk of this division is seen passing through the orbit, close to the ()])tic nerve and u]) to a ])oint just below the olfactory bulb. 1^62 THE BRAIN OK THE EMU 3 This is the main sensory nerve from the bill, innervating' tiie skin and vibrissas Its large size suggests tliat taetile sense in the bill is acute and important. On opening the olfactory chamber, the most posterior of the three nasal chambers, a well-developed turbinal mound (superior or olfactory concha) is seen on the lateral Avail. It is covered with a soft, yellowish epithelium, which becomes thinner and less yellow as it spreads out over tlie dorsal and mesial aspects of the chamber. A vertical section through the nasal chambers of the bill at this level (Fig. 3) reveals that the concha is raised to, a height of about 5 mm. and is slightly constricted at its base, but is not folded into a spiral like the conchae of some vultures and albatrosses (Bang, I960). A specimen for microscopic examina- tion was taken from the dorsal surface of the olfactory chamber ; it shows cells and cilia typical of olfactory epithelium. About 5 cm. anterior to the olfactory concha there is a large nostril (Fig. 1), which is the external opening of the anterior nasal chamber. The emu has large eyes and the orbits are spacious. As one sees in Figure 1 the brain lies mostly behind the orbit with the olfactory chambers in the bony structures just in front. The optic nerve enters the chiasm and passes directly to the optic lobe of the opposite side of the midbrain (Fig. 2C). The large fascicles of nerve fibers can be seen as they cross. The optic lobe is a large and conspicuous structure (Figs. 1. 2B and 2C'). Tn the lateral view only about one-fifth of it is covered by the overlying hemisphere. In Figure 2B (in which the parts of the brain are slightly separated) the relation of the optic lobe to hindbrain and forebrain is emphasized. It is clearly a part of the midl)rain. In fact, the optic lobes are homologues of the corpora bigemina of reptiles, and of the anterior corpora quadrigemina of mam- mals. They have taken a A-entrolateral position in birds, perhaps because it was easier there to make room for the extraordinary tectal development in this class of vertebrates. The emu brain when viewed from above (Figs. lA and 2A) impresses one by its triangular shape, with cerebral hemispheres broad posteriorly and narrow anteriorly. The olfactory bulbs protrude, forming the anterior pole of the hemisphere. On the vertex the two sagittal elevations of hyperstriatum stand out con- spicuously and are separated from the lateral parts of the hemis- pheres by a distinct sulcus, tlie vallc^cula (Portmann and Stinge- lin, 1961). 4 BREVIORA No. 170 The cerebellum is larger in comparison to the forebrain than in passerine birds. It has a greater diameter dorsoventrally than laterally (Figs. 1 and 2) although the auricles ]n-otrude laterally on each side. These lobes, composed of flocculus and nodule, are the only ones that complicate the simple conformation of the cerebellum, the corpus cerebelli being largely a mid-line organ corresponding to the vermis of mammals. Between the anterior surface of the cerebellum (culmen and declive) and the posterior poles of the cerebral hemispheres there is ample space for the pineal stalk and gland. The lateral view of the brain (Fig. 2B ) shows the relative sizes of the main subdivisions. For this photograph the forebrain, midbrain, and hindbrain were slightly pulled a])art. The hemi- spheres of the forebrain are well developed and extend backwards covering parts of the optic lobe and of the cerebellum. The great- est diameter of the hemisphere is 36 mm. and the greatest diam- eter of the olfactory bulb is !) nun., giving a ratio of 4 to 1 or 25 per cent. This places the emu among those birds that have large olfactory bulbs (the Gruiformes, Caprinudgiformes, Procel- lariiformes, Podicepidiformes, and Apterygiformes). In a list of 47 different species of birds, arranged according to the relative size of the olfactory bulb, the largest at the top. the kiwi would come first and the emu seventh (Cobb, 1960). The anterior end of the hyperstriatum accessorium (sagittal elevation or AVulst) is close to the olfactory bulb, and the posterior end shades off into the neostriatum before reaching the occipital pole of the hem- isphere. Thus the emu has a large Wulst that reaches well back towards the occipital pole (Figs. 1 and 2) and well forward to a point close to the olfactory- bulb. A comparison of the external configuration of the brain of the emu with that of other birds shows that it resembles most sonu' herons and ducks. Tn comparing it with Stingelin's (1958) photographs, it is seen to be strikingly similar to the brain of Ixobrychus minutus (see his fig. 21 "Zwergreiher"). Seen from below (Fig. 2C ) the conspieuous characteristics of the emu brain are ; ( 1) the large, separated olfactory bulbs, form- ing the anterior pole; (2) the flatness of the ventral aspects of the lateral parts of the two cerebral hemispheres; and (3) the pair of big optic lobes shaped like flasks with their necks joined in the optic chiasm. The cerebellum is so narrow that it is almost hidden by the medulla oblongata, only the flocculi .showing on each side. The roots of the third, seventh, eighth, ninth, and tenth cranial nerves show in this view. 1962 THE BRAIN OF THE EMU 5 The brain of emu ^85 (after formalin fixation) weighed 27.7 grams; liis body weight was 34 kg. The brain of emn #104 weiglied 25.1 grams (also after formalin fixation ) ; body weight 31 kg. This gives a ratio of brain weight to body weight in emu #85 of 1/1227 and in emu #104 a ratio of 1/1235. Little sig- nificance. h()\v('^•el■, should be given to these ratios because it is known that a living emu may vary 30 to 40 per cent in weight during a year due to conditions of food, climate and water supply. The first brain was separated into 8 pieces, for weighing, as follows : Olfactory bulb (right) (injured) Olfactory bulb (left) 0.12 grams Cerebral hemisphere (right) 8.85 grams Cerebral hemisphere (left) 8.8 grams Optic lobe (right) 7.3 grams Optic lobe (left) 7.3 grams Cerebellum 4.6 grams Brainstem 3.9 grams p?' The brainstem (defined by Portmann. 1I)4(). and named "Stammrest") is the basal mass of nerve tissue made i\p of thalamus, midbrain (with optic lobes removed) and hindbrain (with cerebellum removed). Portmann 's purpose was to choose as his common denominator that part of the brain \vhich varies least in its size relative to the size of the whole bird. That part is obviously the brainstem. He then compares its size to other parts of the brain and, by dividing the weight or volume of the stem into the eoresponding value for another part, he obtains his index. This ''index of cerebralization" he finds for an emu to be 4.18, obtained by dividing the weight of the "Stammrest" into the combined weight of the two hemispheres. In our emu #85 this index is 17.6/3.9 = 4.5. According to Portmann 's list the figures 4.18 and 4.5 both place the emu far below parrots and ravens, but above loons, grebes, and quail. He believes that this quotient gives an expression of the "level of integration" of the brain for each species. The Pineal Body In the dissection of enni #104, a large })art of the post-central area of the calvarium was left intact and carefully lifted off the brain. The pineal stalk was thus torn away at its attachment to the diencephalon. It is 10 imn. in length and remained attached 6 BREVIORA No. 170 to the pineal body (Fig. 4). The body itself is embedded in the dura and lies in a dciiression of the cranial roof between the an- terior and posterior fossae. The dorsal position of the epiphysis is thus clearly demonstrated ; it lies between cerebrum and cere- bellum at the level of their dorsal surfaces. The stalk leaves the brain at a point just rostral to where forebrain joins midbrain. The pineal body is round and firm, slightly flattened dorso- ventrally. It is yellowish in contrast to the white skull. The fibrous envelope is continuous with dura which has strong bands spreading laterally and anteriorly. Removed from the mem- branes, the pineal body is roughly triangular, 7 mm. long on each side. With stalk attached, it Aveighs 0.1 gm. after formalin fix- ation. DISCUSSION The description of the gross anatomy of the brain of the emu brings up five points for discussion : 1 ) the size of the brain, 2) the question of primitiveness, 3) the general shape of the brain in relation to the base of the skull, 4) the size and position of the Wulst, and 5) the topographic relations of the pineal body. The size of the brain in relation to body size and "intelligence" has been the subject of much study and many pronouncements. Suffice it to say here that in our opinion the relation of brain weight to body weight (so called cephalization) is a ratio too simple to give information of much significance. Portmann's (1952) pioneer work in describing an index of encephalization is an advance in the right direction. Body weight in birds is too grossly variable to be used in com])arison to the much more stable brain weight. Small birds may show rapid and marked change in weight. There is good evidence that some birds may lose from 30 to 50 per cent of their body weight in 24 honi-s during a mi- gratory flight (Odum et al., 1961 ; Helms and Drury, 1960). The emu, being flightless, lives in a fairly uniform environment and does not go through the prolonged exertion of migratory flights. Its ratio of brain weight to body weight might, therefore, be relatively stable. Actually, in Dromaeus novaeJiollandiae this ratio is approximately 1/1230 (see p. 5). From the weights s'iven by Crile and Quiring (1940) we deduce that the ratio for an ostrich (Strnthio canielus massaicKs) is 1/2929; for a sparrow (Passer domesUcus) it is 1/23 ; and for a hummingbird {AmazUia tzacatl) it is 1/24. This does not mean that the hummingbird has a "better" brain than the emu. It merely indicates that the 1962 THE BRAIN OF THE EMU 7 body controlled by the brain of the hummingbird is just as com- plex as the body of the emu, thouo'h much smaller. The question as to which brain is "better," or more highly evolved, is mean- ingless unless one asks. ' ' Better for what ? ' ' Obviously, the hum- mino;bird's brain is better for Hij>ht and the emu's better for running. Another factor relative to brain size must be considered. It has been pointed out by Sholl (1956) that small brains are in general more closely packed with nerve cell bodies than large brains which have more glial structures between neurons. Man has 10.5 nerve cell bodies per cubic micron ; a mouse has 142.5. In short, the need is to learn what parts of the brain, control- ling what organs, are larger or smaller in each family of birds. With more investigation into the quantitative anatomj' of the brain, some of these questions may be answered. Fritz (1949) has estimated the volume of four parts of the striatum in four different species of birds ; he found significant differences, but no correlation with Portmann 's cerebral index. Many authors have spoken of the emu, and in fact all ratite birds, as primitive, but their concept of primitiveness is not clear. Some seem to call these birds primitive because they are flightless and have no keel on the sternum (Leach, 1923), others because they have a straight type of skull base (Streckschadel) (Mari- nelli, 1928, p. 156). Stingelin (1958) considers those birds, with a small ^\^ulst which lies neither far forward nor far back, to be the less evolved type. The point would seem to be that one must not apply the term primitive in a general way to the emu (or probably any other bird). One should specify in what respect a given type or family is less evolved ("primitive") and in what respect it is more evolved (specialized). Even then, the gaps in our phylogenetic knowledge do not allow us to say whether the ratite sternum is due to a devolution from carinate ancestors or an evolution from cursorial reptiles. The presence of feathers and the avian type of brain suggest strongly a descent from flying ancestors. In respect to running and adaptation to life in open plains one feels confident in saying that the emu is highly evolved. Much work has been done on the development of the avian skull. Pertinent to an understanding of the shape of the emu's brain are three recent lines of investigation. Duym (1951) described the bending of the base of the skull in different birds and specified four types — the stretched or extended type of 8 BREVIORA No. 170 skull and three de<>-rees of bendiiifj'. Dullemeijer (1960) has re- lated the shape and size of the pi-ineipal parts of the brain to the amount of bendin9U animals. Ohio J. Sci. 40: 219-259. DULLEMEI.JER, P. 1960. Shape and size of the biain ])aits as architectonic factors in tlie skull of birds. Acta nioiphol. neerl.-scandin. 4: 96. Dl'YM, M. 1951. On the head posture in birds and its lelation to some anatomical features. Proc. Kon. Xed. Akad. Wetensch. (C) 54: 202-211, 260-271. Edinger, L., A. Wallenberg and G. Holmes 1903. Untersuehungen iiber das Vorderhirn der Viigel. Abh. sencken- berg. naturf. Gis. 20: 341-426. Edinger, T. 1961. Fossil brains reflect specialized behavior. World Xeurol. 2: 934-941. Fritz, Walter 1949. Vergleichende Studien iilier den Anteil von Striatumteilen am llemisphiirenvolunien des Vogelhirns. Eev. suisse Zool. 56: 461-491. Helms, C. W., and W. H. Drury, Jr. 1960. Winter and migratory weight and fat field studies on some North American buntings. Bird Banding 31: 1-40. HuENE, F. v. 1911. Ulier Erijtlirosueliii.'f. Yertreter der neuen Eeptil-Ordnung Pely- cosimia. Geol. Pal. Abh. 14: 1-6(1. Jaekel, O. 1910. Ulier einen neuen Belodonten aus dem Buntsandstein von Bern- burg. Sber. Ges. naturf. Freunde Berlin: 197-229. Krabbe, K. H. 1961. La glande pineale. Worl. riiotogiaph, life size, of dorsal view of same dissert ion 1962 THE BRAIX OF THE EMU 17 FIG 2^^ DORSAL FIG 2^ LATERAL FIG 2^ VENTRAL Figure 2. Three views of the brain of the emu ( Drnmaeus novaehollan- diae) #104. Life size. F, forebrain hemisphere, E, hindbrain, il, mid- brain, showing optic lobe (T) and optic- chiasm (C), V. vallecula, TT, Wulst or hyperstriatum aceessorium. 18 BREVIORA No. 170 FIG 3 FIG 4 FIG B Figure 3. Section tlirough the nasal chambers cut in frontal vertical showing the olfactory (or posterior) chamber (P) into which protrudes the olfactory condha (C) covered with yellow olfactory epithelium. Below is seen part of the middle nasal chamber (M). The two chambers are di^aded by the sepfiiin. They connect anteriorly witli the anterior chamber and the external nostril. Life size. Figure 4A. Photograph, life size, of pineal body (PB) lying in shallow cavity of the ealvarium, posteroanterior view. The stalk (S) protrudes downward. Figure 4B. Ventral view, looking upwards at under surface of ealvarium. The stalk (S) is bent backwards. BREVIORA Mmseiiioi of Comparative Zoology Cambkidgk, Mass. 1)kcembp:r 14, 11H)2 ^'^MB^:R 171 NOTES OX AMPHLSBAEXIDS ( AMPHI8BAEXIA : REP- TILIA). 6. REDESCRTPTIOX AXD RANGE EXTEXSION OF AMPHISBAENA SFURRELLI BOULENGER. By Carl Gans Department of Biology, The University of Buffalo, Buffalo 14, Xew York 111 1915 Boulenofer (p. 659) described the new species of Amphishaeua spurrdJi, characterized primarily by the presence of tubercular or subcoiiical segments on the dorsal surface of the tail. The two syiitypes were collected at "Anda Goya [Colombia], at the junction of the R. Condoto and San Juan." The only subsequent record of the species (Burt and Burt, 1931. p. 40) is the citation of a single specimen (A.M.X'.H. 18261) from the neighborino: locality of ''Boca de la Raspadura, " Colombia, without su])p](MU(Mitary description. The present note is based upon a re-examination of these three and of two additional specimens, one of which extends the range of A. spnrrcUi into Panama. The original description has been amended and rewritten according to the standard pattern (Gans and Alexander. 1962). Simple, non-idealized illustrations are included in the present paper. It is a pleasure to acknowledge the support of the X'ational Science Foundation (XSF G-9054. G-21819). Examination of the types was made possible by assistance from the estate of Leo Leeser. Specimens Avere examined through courtesy of C. M. Bogert. The American Museum of N^'atural History (A.M.X.H.), Miss Alice G. C. Grandison. British Museum (Xatural History) (B.M. ). R. F. Inger and H. Marx. Chicago Natural History Museum (C.X.H.M.), and E. E. Williams, Museum of Comparative Zoology (M.C.Z.). I am particularly grateful to Dr. Federico Medem who made the C.X.H.M. speci- men available, and to Miss C. Rhodes for technical assistance. BREVIORA No. 171 Amphisbaena spurrelli Boulenger, 1915. Amphubaena spHrrelli Boulenger, 191.5, p. (359. Terra typit-a : ' * Anda Goya, at the jiiiK'tion of the R. Coiidoto and San .luaii," Colonihia. LECTO- TYPE: B.M. 19ir3.10.lM.9 (liy present designation). J'ARATYPE: B.M. 1915.10.21.8. Diagnosis: A form of Ampliisbacna without fusions of head scah^s; with 4 oval [not round] prech)aea] pores; and witli the dorsal and lateral surfaces of the eaudal tip covered with eonieal or tubercular segments. Specimens have 218 to 222 body annuli ; 18 to 20 caudal annuli ; 16 to 18 dorsal and 16 to 18 ventral seg- ments per midbody annulus; and one row of postgenial and one row of postmalar chin shields. There is no visible autotomy con- stricton of the tail. Autotomy takes place after the seventh annulus. Notes 0)1 tin types: Boulenger (191."), p. (i59) illustrated the smaller of his syntypes, which has here been chosen as a lecto- type. The types, still extant and in good condition, suggest that his illustrations were idealized, and s(>veral of his counts [shown 78 Fig. 1. Amphisbarna .^purrrlli. Map showing lo'-alities mentioned in te.xt. Anda Goya and Boea de la Rasi>adura are actually cdoser together than can be indicated on a map ilrawii to this scale. 1962 AMPHISBAENA SPURRELLI 3 ill brackets in the table] erroneous. Tbe erroi-s do not affect the validity of the species. I)( scvipiiou : ^leristic charactcM's are listed in the table. Fiynre ."] shows the head scalation, l-''ij:ure 4 the sef):nientatioii of cloaca and tail, P^igures 5 to 8 ])hotographs of head, niidbody pattern and tail. Presei'A'ed specimens are various [faded] shades of brown dorsall}', somewhat lighter vent rally. The darker dorsal color is in part produced by a darkening of the rectangular center of each segment, the contrast with the lighter segmental margins giving the impression of dark spots. The fully dark dorsal spots descend the sides to approximately the third ventral below the lateral groove on each side. Ventrad from this the dark center shrinks drastically or may fade out entirely. The anterior fifth (M.C'.Z. 8f)784), or tlir head alone, lacks the dark colored seg- ments. The head scalation shows some variability and no major fu- sions. An azygous rostral tiarely visible in dorsal view is fol- lowed by three pairs of enlarged cephalic shields in contact along the dorsal midline. The nostrils pierce the first pair (nasals). The second pair (prefrontals) are the largest segments of the head. There are three supra- and two and a half infra- labials, as the third infralabial extends considerably beyond the angulus oris. The suj^ralabials are large, the second much the largest. The CX.H.^I. s))ecimen has this segment subdivided differently on both sides. The second infralabials are the largest segments on the lower .jaw. Small segments lie beyond the angulus oris in the position of fourth supralabials. The mental is shaped like a truncated wedge with a posteriorly con- vex tip. The postmental is hexagonal and elongate. It lies in lateral contact with the medial edges of the second infralabials. as well as the anterior poi-tion of the medial edges of the rela- tively short, wedge-shaped malars. There are one to two rows of postgenial segments, follo\\ed by a single postmalar row, the lateral segments of which are slightly enlarged. The head is relatively blunt, flattened slightly dorsoventrally and oval in cross-section. The lower jaw is but slightly shorter than the upper. The sides of the head would, if extended, inter- sect some distance anterior to the rostral tip, even in adults BREVIORA No. 171 E E c 0) 27 25 23 21 19 17 o» 20 22 24 26 28 30 Snout - Vent Length - cm Fig. '2. Amphishama spiirrclli. Scatter diagram showing plot of tail length versus snout-vent length for the available specimens. The lectotype is shown as a hollow circle. in which the bulge of the temporal musculature changes the out- line. The attachment of the skin to the crest of the skull pro- duces a concave dishing of the interfrontal suture, particularly in adult specimens. The first body annulus curves forward to contact the frontals. Its dorsalmost segments may be somewhat enlarged, and one of the specimens has an intercalated dorsal half-annulus. The second through fifth annuli are narrowed, and the fourth marks the level of the head joint or the point at which the bulge of the temporal musculature returns to normal. 1962 AMI'HISBAENA SPURRELLI Fig. 3. AmpJiisbaena .fpurreUi. Dorsal, lateral and ventral views of the head of A.M.X.H. 18261 from Boca de la Raspadura, Colombia. The line equals 1 mm to scale. (V. Cummings, del.). BREVIORA No. 171 Fig. 4. AmpliLsbacua spiimlli. noisnl, lateral ami ventral virws of tlu' cloaea aiul tail of A.M.X.H. ]8l^(il from Boi-a de la Raspaduia, Cdlomliia. Note the extent of cone formation on the tail ami the Hattened, .slit-sha]itMl precloacal pores. Tlie line equals 1 mm to scale. (V. C'umniinf»H, del.). The dorsal gTOOve is only indicated on the head. The ventral groove is indicated mainly as a gap between aligned segments. The lateral grooves start about one and a half head lengths be- hind the head joint, and are well defined by a double row of triangular segmental fragments. The middorsal segments are almost twice as long as wide, the midventral segments are almost twice as wide as long. 1962 A.Ml'lilSHAEXA SPURRELLI 7 Tlip oval pi'ccloacal pores lie in a single uninterrupted row ol" normal sized seg-ments antcrioi- to the predoaeal shield. The preeloaeals ai-c cliaraeterized by a eentral group of four some- what elongate segments. The posteloaeals, slightly gi-eater in number, eliaraeteristieally have a set of two to four midventral and enlarged segments and, tiaidving these, several split segments entering the doaeal sides. The cloaca may be entii'ely prolapsed. The tail becomes gradually wider posterior to the eloacal slit and somewhat higher as well. The ventral surface appears plane and an ext(Misi())i of the pi-ecloacal region. The terminal third of the tail shows reduction, with the tip about twice as high as wide. The segments of the dorsal and lateral surfaces are strongly t nbcrculate or cone-shaped. This character facilitates diagnosis of specimens with intact tails. Caudal autotomy takes place behind the seventh ])ostcloacal annulus (cf. Vanzolini, 1951, p. 2.3). Range: Lowland ri\er valleys of northwestei'n South America, from extreme northern Colombia (Choco) to southern Panama. Distrihulioit rcconh: COLOMBIA: Choco Province: Anda Goya, mouth of Rio Condoto ( l)()ulenger, 1915); B.M. 19i5.10.21.8 (PAIJATYPE). 1915.10.21.9 (LECTOTYPE ) ; C.N.H.M. 18098S [E. K. Dunn.' leg. per F. Medem|. Boca de la Raspadura (Burt and Burt, VX^\ ) ; A.M.X.H. 18261. PANAMA: Tucuti branch. Tuira River [H. C. Clark, leg.] M.C.Z. 39784. LiTi:K.\'r["'Ri': cited Boi'LENGER, George Alhert 19in. Descriptions of a new AmpJiishaiiui nnd a new snake discovered l^y Dr. 11. Or. F. Spuriell in soutliern ("olninl)i;i. I'roc. Zool. Soc. London, (1915), i)p 659-61. BiRT, Charles E. and .\[ay Daxiieim Dirt 1931. Sonth American lizards in tlie collection of the American Mu- senm of Natural History. Bull. Amer. Mus. Xat. Hist., vol. 61, art. 7, PI). 227-395. Gans, Carl Axn Alexander Allan Alexander 1962a. Studies on amphisbaenids (Amphisbaenia, Reptilia), 2. On the amphisl)aenids of the Antilles. Bull. Mus. Comp. Zool., vol. 128, no. 3, pp. 65-158. N'anzolixi, Paulo Emilu) 1951. Anipliisbaena fiiliginnm. Contribution to the knowledge of the Brasilian lizards of the family Amphisbaenidae Gray, 1825. 6. On the geographical distribution and differentiation of .1 iiiiiliishaena fuliginosa Linne. Bull. Mus. Comp. Zool., vol. lUti, no. 1 , i)p. l-()7. 8 BREVIORA No. 171 Fig. f). Amphisboena fipiin-fUi. Lateral view of the liead of the topotype, C.N.H.M. 180988, from Aiida (ioya, rohiiiiliia. Note tlie irresuhir sulj- division of the second suprahibial. 1962 AMPHISBAENA SPURRELLI 9 Fig. (i. .1 iiiiihislKii'iKi siHirnlli. Dorsal ( k'ft ) and ventral (riuiit) viuw.s of A.M.X.H. spec-inien at midbody. Note the darkening of the dorsal seg- mental centers. Fig. 7. Amphisbaena spurreUi. View of caudal tip erossdighted to empha- size the knolilied nature of the tenninal segments. 10 BREVIOKA No. 171 Fig. 8. Amphishaena spurrelli. Ventral view of the cloaea and tail of the A.M.N. H. .specimen. Note the onset of autotoniy at annulus .seven, also the contrast between the plane ventral and eonieal laterodor.sal i-andal segments. 1962 AMPHISBAENA SPURRELLI 11 Si. =0 ■5 ^-1 o ^ ^i' H > x — W "& "71 + + 11 + 11 + o •t t^ o or. ^ ci o; 11 11 11 CI n 11 11 M CC I'l Z: 3C 00 QC m + + n o 11 (M oc -^ CO 11 •^ tC OC ■£ -^ in or -t 11 f-H 11 HI 11 11 11 11 L-^ Lt X ^^ — ^ r— < Cl '"'' o; X "~ -~ ^— _« _- ^ OO . 1 • 1 — ~ ^' ~ o 'jH ^- ' — ii -— S: ^ X t^ — ' _ i; i.» r" * OS «rt CO ~, 3 S M N3 d <;' ;^ M d S BREVIORA Mmseiinii of Comparative Zoology Cambridge, Mass. December 14, 196 2 Number 172 A NEW SPECIES OF THE RODENT PIPE.STONEOMYS FROM THE OLIGOCENE OF NEBRASKA By Raymond Alf Webb School of California, Claremont In 1956 John C. Donohoe described from the Pipestone Springs formation of Chadronian age in the Montana Oligocene two jaw fragments of a new rodent, Fipestoneomys bisulcatus, that he placed in the family Aplodontidae. The purpose of this paper is to record the occurrence of Pipe- sioneomys in the Chadron formation of Nebraska, to suggest that it be placed tentatively with the castorids rather than with the aplodontids, and to propose a new species of the genus. North of Crawford, Nebraska, well known exposures of Oligo- cene sediments occur along the Pine Ridge escarpment. For several years, through the kindness of Frank Arner of Crawford, Nebraska, the Webb School of California has had the privilege of collecting fossils on the Arner ranch. In Sec. 26, T. 33N., R. 53W., Sioux County, Nebraska, are several harvester ant mounds in the Chadron formation, and from these have been collected a rich microfauna including Pipestoncomys. I wish to thank W. D. Turnbull of the Chicago Natural History Museum for the loan of the type specimen of P. hisnlcatua fur study, and Bryan Patterson of Harvard University, and Dr. Mary Dawson of the National Science Foundation for advice. The new species is named in honor of Professor Bryan Patterson. Tlic drawings are by Nick Strekalovsky. The abbreviations refer to the following institutions or collections: C.N.H.M., Chicago Na- tural History Museum; M.C.Z., Museum of Comparative Zoology; R.A.M., Raymond Alf Museum, Webb School, Claremont, Cali- fornia. 2 BREVIORA No. 172 Family CASTORIDAE ? PiPESTOXEOMYS PATTERSOXI 11. Sp. Type: M.C.Z. no. 7113, fragment of left maxilla with Mi-2. mjpodigm: The type and M.C.Z. nos. 7110, RP* ; 7106, LP-*; 7108, LP4 ; 7102, LMi ; 7111, RM2 ; 7104, LP4 ; 7107, RP4 ; 7103, RMi ; 7109, RMj ; 7112, RMi ; 7101, LM^ ; and R.A.M. nos. 3072, RM2 ; 1683, LM2 ; 671, LM2 ; 1283, RP4 ; 2105, RP4 ; 2104, RMj. Horizon : The base of the Chadron formation along Pine Ridge is marked by a basal conglomerate resting on the weathered sur- face of the Pierre shale. The top is characterized by the upper purple-white layer, a continuous purple-tinted white limestone designated by Schultz and Stout (1938) as the boundary be- tween the Chadron and the Brule formations of the White River group in this area. About half way between this and the basal conglomerate there is a second limestone lens, which is referred to as the lower purple-white layer. The specimens come from well below the lower purple-white layer. Localtiy: Sec. 26, T. 33N., R. 53W., Sioux County, Nebraska. Diagnosis: Differing from P..hisulcatus as follows: paracone and metacone of Mi"2 not rounded, paracone and anteroloph not sharply separated, anteroexternal and anterointernal lakes sepa- rate, relatively small ; Mi smaller, more rectangular in outline. Description : The paracone and metacone of M^'^^ though dis- tinct and higher than the remainder of the crown, are not the rounded cusps that they are in P. hisulcatus. In P. hisulcatus the paracone is separated from the anteroloph by a deep groove com- municating with the anteroexternal lake, whereas in P. pattersoni the separation is very faint. In P. hisulcatus the anteroexternal and anterointernal lakes communicate by a slight groove. In P. pattersoni the corresponding lakes are completely separate and are also relatively smaller and are more delicately outlined. There is no trace of a mesostyle. P^ is larger than either M^ or M^. A deep groove separates the anteroloph from the rest of the tooth and communicates openly with a deep valley that separates the paracone-metacone loph from the hypocone. On the anteroloph there is a lake below the anterocone and an embayment projecting inward to the protocone. As on the upper molars, a mesostyle is lacking. The lower first molar of P. pattersoni differs in shape from the corresponding tooth in P. hisulcatus, being much more rectangu- lar. It is also smaller and more delicate. P4 is larger than Mj. 19G2 NEW OLIGOCENE RODENT Fig. 1. Pipestone&mys pattersani. A, M.C.Z. no. 7113, left maxillary frag- ment bearing Mi-2, type. B, M.C.Z. no. 7110, RP^. C, M.C.Z. no. 7103, RMi. D, M.C.Z. no. 7107, RP4. X 25. BREVIORA No. 172 The anterolophid, metalophid, and posteroloi)hid are well devel- oped and separated by deep grooves that terminate at the meta- conid and entoconid. There is a lake posterior to the metaconid and one on the posterolophid adjacent to the hypocones. Discussion: Pipestoncomys, especially in the light of the new evidence reported here, does not fit well into the Aplodontidae. The upper Eocene Eohaplomys, the earliest known aplodontid, 1 '^ 3 has the dental formula (Stock, 1935; McGrew, 1941). _L U J. o This formula is retained to the present. The presence or absence of P^ is indeterminable in the holotype of P. hisulcatus, but a 2. mm » Fig. 2. Pipestoiieomys bisulcatus. Right M^"^, crown view, C.N.H.M. no. UM 409, type. small fragment of the maxilla anterointernal to a P"^ of P. pattcr- soni (M.C.Z. no. 7106) shows no sign of the presence of P^. P| are large relative to the molars. The anterior margin of the mas- seteric fossa in P. bisidcatus extends to beneath the middle part of P4, and the ramus is relatively thick and deep. These charac- ters are not typical of the early aplodontids, but are quite like those occurring in castorids. The cheek tooth structure seems to me to accord better with that of castorids than with that of aplo- dontids. A lower molar of P. pattersoni (M.C.Z. no. 7112) was ground down to show the pattern of a well worn tooth. The median external groove (hypoflexid) persists but its medial portion be- comes transformed into a lake, forming a pattern similar in gen- eral to that found in, e.g., Paleocastor (cf. Fig. 3 and Stirton 1935, Fig. 30). The incisor enamel of Recent Castor, Miocene Paleo- castor, and Oligocene Agnotocastor is characterized by an orange 1962 NEW OLIGOCENE RODENT color (Wilson 1949), that of aplodontids is not pigmented. Among the many incisors picked out of the ant mounds are a few with orange color, and those of appropriate size could well pattersoni. belong to P Fig. 3. Pipestoneomys pattersoni. M.C.Z. no. 7112, EMi ground down to show pattern near base of crown. X 25. The past and present distribution of aplodontids is also of some interest in this connection. Not only is the range of Recent Aplo- dontia limited to a strip along the west coast of North America, but no unquestioned fossil form has been found east of the Great Basin. 1 If this picture of geographic distribution is adequate, we may have another item of evidence that argues against aplodontid affinities for Pipestoneomys. Pipestoneomys seems to me to fit more reasonably into the Cas- toridae than into the Aplodontidae, and I know of no other family of appropriate range to which it might be referred. If this tentative assignment is correct, the genus is the earliest beaver so far known, although its position in castorid phylogeny is un- certain, EEFERENCES DONOHOE, J. C. 1956. New aplodontid rodent from Montana Oligocene. Jour. Mam- mal., 37: 264-268. McGrew, p. O. 1941. The Aplodontoidea. Geo!. Ser., Field Mus. Xat. Hist., 9: 1-30. iShotwell. in his interesting studj of the aplodontids (I'J-jS), mentions Pipes- toneomys only in passing, noting that P. bisulaatits occurred well to the east of thf other known fossihs. 6 BREVIORA No. 172 ScHULTZ, C. B. and T. M. Stout 1938. Classification of Oligocene sediments in Nebraska. Bull. Univ. Nebr. State Mus., 4: 15-52. Shotwell, J. A. 1958. Evolution and biogeography of the aplodontid and mylagaulid rodents. Evolution, 12: 451-484. Stibton, E. a. 1935. A review of the Tertiary beavers. Univ. Calif. Publ. Bull. Dept. Geol. Sci., 23: 391-458. Stock, C. 1935. New genus of rodent from the Sespe Eocene. Bull. Geol. Soc. Amer., 46: 61-68. Wilson-, E. W. 1949. Early Tertiary rodents of North America. Carnegie Tnst. Wash- ington, Publ. 584: 71-164. Table 1 Pipestoneomys. Upper Dentition Measurements in millimeters P. pattersoni . P. bisuleatus C.N.H.M. UM 409, type Crown length Ml M.C.Z. 7102 1.62 1.78 M.C.Z. 7113 1.47 M2 M.C.Z. 7111 1.48 1.6 M.C.Z. 7113 1.23 E.A.M. 671 1.48 E.A.M. 1683 1.62 E.A.M. 3072 1.44 P4 M.C.Z. 7106 2.16 M.C.Z. 7108 2.07 M.C.Z. 7110 1.98 Greatest width Ml M.C.Z. 7102 1.71 1.95 M.C.Z. 7113 1.67 M2 E.A.M. 1683 1.62 1.81 E.A.M. 671 1.44 M.C.Z. 7111 1.35 M.C.Z. 7113 1.67 E.A.M. 3072 1.35 P-t M.C.Z. 7106 1.98 M.C.Z. 7108 2.09 M.C.Z. 7110 1.94 1962 NEW OLIGOCENE RODENT Table 2 Pipestoneomys. Lower Dentition Measurements in millimeters P. pan ersoni Crown length Ml M.C.Z. 7101 1.44 M.C.Z. 7103 1.48 M.C.Z. 7109 1.74 M.C.Z. 7112 1.89 E.A.M. 2104 1.62 P4 M.C.Z. 7104 1.62 M.C.Z. 7107 1.51 E.A.M. 1283 1.80 E.A.M. 2105 1.71 Greatest width Ml M.C.Z. 7101 1.51 M.C.Z. 7109 1.62 M.C.Z. 7112 1.62 M.C.Z. 7103 1.35 E.A.M. 2104 1.53 P4 M.C.Z. 7104 1.35 M.C.Z. 7107 1.42 E.A.M. 1283 1.48 E.A.M. 2105 1.62 P. bis III cut US C.N.H.M. UM 408 dP4 1.78 1.69 dP4 1.18 BREVIORA MiuiseiLim of Comparative Zoology Cambridge, Ma>s. I)k( f.mhkh 24, 19(iL' Xu.mbei; l?,'! NEW (SPECIES UF LAND MOJ.HSKS VUijM THE REPUBLICA DOMINICANA By "WiLLiA.M J. Clench We are deeply indebted to Drs. Staidey Rand, Clayton Kay and Juan Rivero for two important collections of land moUusks collected in the Republica Dominicana. Both collections were made in areas hitherto unknown as far as the land inollusks were concerned. Of the foni- species herein described, two represent a subfamily and a and strongly convex. Outei- ]i]i simple, jmrietal wall with a thin glaze. Suture slightly indented. Columella long, sinuous and somewhat oblique. Sculpture consisting of very fine axial growth lines. Height Width 65.0 mm. 67.0 nun. Holotype 67.0 mm. 67.5 nnn. Paratype Typ( s. The holotype is in the Museum of Com])arative Zool- ogy, no. 230508, from Cueva de San Fi-ancisco, Cerros de San Francisco, Mun. Pedro Santana, San Rafael, Republica Domini- eana. Collected by Stanley Rand and Clayton Ray, Augast 1958. There is single paratype from the same locality. Renunl.s. This species is placed pi-ovisionally in the genus ZapJiyseind, a genus heretofore known only from Jamaica and Xavassa islands. It is about three times as large as the lai-gest Jamaican species, Zaptnisrnid niacniurrayi C. B. Adams. Both specimens were dead when found in tlu' cave and appear to be ((uite old as they are white and completely devoid of perios- tracum. This s])ecies is as large as the largest PoUidontes r/if/antea Scapoli from llisjiainola and Zaelu'i/sia petitiana d'Orbigny from the Triindad Mountains of Cuba. Its large size and thinner shell makes it moi'e capacious tlian eithei- of these two species. 1962 LAND MOLU'SKS FROM HISPANIOLA 5 This is by far the largest species in the family Sagdidae, a family which occurs in the West Indies, southern Ignited States,^ and south through Central America and northern South Amer- ica. Its greatest generic development is centered in the island of Jamaica. REFERENCES Bartsch, Paul 1946. The operculate Iniul luolliisks of the family Aiinulaiiidae of the Ishxiul of Hispaniohi and the Bahama Archipelago. Bull. U.S. Nat. Mus. 192: 1-3. Crosse, H. 1891. Fauiie nialac-ologi(|iu' tenestre et fluviatile de I'lle de Saint Dominge. Jour. Coneiiyl. 39: (59-95. Turner, R. D. 1960. Land shells of Xavassa Island. Bull. Mus. Comj). Zool., 122: 233-2-14. lA single jrenns. Microphusiila. with two spcci^-s. iiigersolli lihuid and cookci Pilsbry extends mirth Ihrougli tlic K(iel( \V.\ uniiiii;. 2 During the 1{»61 season a concentration similar to Imt less ricli lliaii tliat of the bone befl was found some 00 feet low<'r in the member. 2 BREVIORA No. 174 few frafinients of a medium-sized pantodont. We propose that the assemblage be known as the Shotgun local fauna, from the name of the member in which it occurs. The age, so far as can be deter- mined at present, appears to be early Titfanian. At Shotgun Butte, the type locality, where the member has a thickness of 2,830 feet (Keefer, 1961), l'li< nacodHs s]). and Plesiadapis sp. cf. P. cookci, have been found in the uj)per i)art of the sequence (Keefer and Troyer, 19r)6). We are greatly obliged to l)v. William R. Keefer for his help in the field. For the opportunity to examine comparative material we are indebted to Dr. C. Lewis Gazin, U. S. National Museum, Dr. Malcolm ('. McKenna, The American Museum of Natural His- tory, and Dr. Craig C. Black, Carnegie Museum. During the field season of 1960 we were assisted by James A. Jensen, Lee A. Wooder.son, John Zamecnik, Floyd Andrews, Clyde T. Williams. Richard P. Timmerme.yer, Charles P. Lyman, Jr., and Robert F. Wallin. The ]ihotographs arc the work of Miss Linda Loring. The following aI)l)reviations are used: A. M.N. II., American ^luseum of Natural History; M.C.Z., Museum of Comparative Zoology: I'.AV.. Fniversity of Wyoming. AHCTOCVONIDAE 1 .Muiiay OXYCLAENINAE Matthew OOLPOCLAEX rs -gen. nov. Ttjpc sijccies: C. keeferi sp. nov. Known distribution : Paleocene, vavW Titfanian, Wyoming. I)i(i(jnosis: Enamel of molars strongly wrinkled, all crests cren- ulated, accessory cuspules numerous. Prine-ijjal cusjis of upper .\[ high, massive, tightly grouped; central basin small; protocone nearly central in position with long lingual slojx' ; conules large, blunt, separated by grooves from principal cus]is; hypocone small on ^P-, i-udimentary on 'SI'': cingnla sti'ong, not continnoiis around jtrotocones ; external cingulum intcrruptetl by labial con- linuation of sharp cleft between paracone and metaconc; ni)per M wide relative to length. Trigonids of lowei- M high, short ; i)ai-a- conid internal in position, closely ap])ressed to metaconid, i)ai-a- conid crest well dcx'cloped : postci'ior ci'cst connecting protoconid 1 Out- <(1' us (li.I'.l proiPdscs to liarislCr tlic A rcl myiniiil.-ii' Irniii the ( 'rciKlonlM to the Coiid.vliirthni. 'I'liis traiist'cr .iiid tiic \:jiiois (|iicsli(iiis tliiit it raises wiU l(c discussed in ;i later jiaiier. -Kidpcis, a fold and cluviius; in alliisinii to I lie \vrinl>- XKW i'Ai,i;()(i:\i-: AKc'rocvnxii) ;; and metaeonid present: protoeonids and metaeonids with slioii, centrally sitnated erests rnnning lino-nally and labially. respee- tively. forming a third, transverse trigonid crest : deejjest jxjrtion of talonid basin near lingnal side : hypoconnlid of ^I.-j long, broad, higli. cuspidate. COLPOCLAKXrS KEEFERI ' s]l. nov. Type: M.G.Z. no. 8355, LM.,. Hypodigm: Type and F.W. nos. li)31. LM... ; ll»3l2. KM.-., ; 1933, LMi; 1934. RM3; 1935. L.M^. : :\I.r.Z. nos.' 8356. KM,, "(much worm ; 8357. LM2 ; 8358. ini.. Horizon and locality: Shotgun local fauna. Shotgun member of the Fort Union formation; XEI.4 SEf^., sec. 31. T. (i X.. R. 3 E.. % mile SW of the more iiortlici-ly of the Twin Rnttes. Fremont Comity. AVyoming. J)in(/)iosis : Sole known species of th(^ genns diagnosed above. 1) F SCRIPT lOX Colpoclaenus kcrferi is notable for the massive, high, princij^al cusps of the upper molars, the elevated trigonids and the extreme- ly rugose enamel. The latter feature disappears after very little wear, as is demonstrated by a moderately worn M2 (M.C.Z. no. 8358). There can be. we believe, no doubt that the various lower molars represent the same species, and their association with the n]i])ers would appear to be demonstrated by excellent occlusion between .AI.C.Z. no. 8357, LM^, and U.W. no. 1935, L.M.. as well as by the general structure. The upper molars are considerably wider than long. ^V - being e.ssentially quadrangular, ^l^ suboval in outline. The i)rotocone is the largest of the three principal cusps; its apex is a little lin- gual to the center of each tooth of the series, and its lingual face is very long and gently sloping, rather shorter in ]\F" than in M^"-. The paracone is more labial in position than the metaeone and is slightly larger and higher, being approximately the same height as the protocone. Paracone and metaeone are grooved on their basinward faces and are separated by a well marked cleft. The large blunt proto- and metaconules are separated from the primary cusps by grooves and are united by crests to the anterior and posterior cingulum respectively; the protoconule is smallei- 1 For Dr. W. R. Keefer who discovered the Shotgun bone bed in the (■llur^se (if his worl; on the Fort Union of the Winn River Basin. BREVIORA No. 174 Figure 1. Colpoclaenus keeferi gen. et. sp. nov. Upper molars: a, LM-, M.C.Z. no. 8357; fe, Li\[l, U.W. no. 1933; c, EM^, U.W. no. 1934. X3. Stereoscopic views. 1962 NEW PALEOCENE ARCTOCYONID .") than the metaconnle on ^l^-, al)ont eqnal to it in size on "Sl'^. The ratiier shallow basin enclosed by these cusi)s and conules is situ- ated in the labial half of the tooth and bears various i)Oorly de- fined cuspules on its floor. A hypoeone is differentiated on Mi"-, on which it is considerably smaller than the coiiulcs. I)iit is bai-ely distinguishable on M^. Apart from an interruption on the exter- nal face, where the cleft Ix'tween paraeone and metacone con- tinues on to the labial margin, a strong', crenulated cinguluiii runs from the hypoeone around the tooth to the antero-internal corner, at Avhicli point the cingulum is broken in AP. M- shows no j)i'()to- stylar enlargement here and M"^ a very slight one. With the ex- ception of a slight rise in the cingulum at the parastylar site on ^r"^, there is no indication of external styles. Anterior and pos- terior cingula extend farther lingually on M^ than on the jjreced- ing teeth and the two are almost united by a median lingual cuspule. Several cuspiiles are present on the posterior face of M^ above the cingulum. ^Fo is slightly constricted at the junction of trigonid and talo- nid but is otherwise nearly quadrangular in outline ; the trigonid and talonid are approximately e(iual in length and width. The trigonid is somewhat higher than the talonid and narrows apically. The protoeonid is a large massive cusp that forms nearly half of the trigonid. The heavy, cuspidate paraconid crest runs from the apex of the protoeonid to the lingual side of the tooth, where the paraconid is poorly ditferentiated. The blunt metaconid is but little smaller than the protoeonid and equal to it in height ; the apices of the two cusps are connected by a papillate crest that is more lightly built than the paraconid crest. Blunt, Avrinkled crests run directly lingually and labially down the facing slopes of the protoeonid and metaconid. These al)ut (U.W. no. l!):^.')) or fuse (M.C.Z. no. 8358 ) to form a third, central transverse crest. A sinuous, antero-posterior cleft partially separates the trigonid apex into labial and lingual halves. All this complicated structure is shallow and would rapidly be ol)literated by weai-. The pos- terior face of the trigonid is sloping and bears various small, papillate crests; that behind the metaconid is the most prominent but there is no trace on it of a metastylid. The very large, blunt hypoconid nearly equals the protoeonid in size and makes up nearly half of the talonid. The cuspidate crista obliqua is low and fades away toward tlie base of tlie talonid. Two minor crests run forward from it to the trigonid base, isolating a vei-y sitiall fossette. The hypoconulid is low but relatively large, and in the BREVIORA No. 174 Figin'o 1". Colpocliit mis l-crfrri gen. et .sp. iiov. ^li h. M.C.Z. no. S3;")8. X'^. Stt'rt'oscu]ji(' views. I. r.W. 710. 193; unw()i-ii r.W. no. 19.'^.") is ti-iciispidatc and set off fi-oiii tlic .'ulja- (M'lit (Mis|)s by sliallow ' tooth of Z . paleocena, differino- only in a few points. The tooth is slightly smaller rela- tive to M^ the postero-internal angle is sonunvhat more scinared, and the antero-external style is i-athei- more ])i-ominent and con- nected by a low inconspicuous crest to the tip of the paracone. The adjacent styles of AP- abut to form an apex fi-oiii which the outer margin of the cheek tooth series falls aw;iy anteriorly and posteriorly. The tirst lower molar diffei's only in detail from Ihat of /'. sill)erli)\(ji. The ])araconid is usually higher and slightly more independent and is invariably larger than the metaconid. Tn several teeth the latter cusp has almost or (piite lost its identity in the oblique ti-igoiud crest. Ti-igonid and talonid are con- sistently better ch'mareated externally. The crest that forms the external border of the talonid cui-ves inward at its anterior extremity to reach the center of the [jostcrior face of the trigonid a short distance below the protoconid rathei- than abutting against the base of the trigonid externally, as is the rule in P. silberlin(/i. A narrow, shallow gi-oove lies between the trigonid face and the incurving position of the crest. The main iX)rtion of the talonid crest, that forming the lateral and posterior borders of the basin, appears to be more cuspidate than in the Lebo sample; the cuspules number from four to idne rather than from two to three. The two cuspules on the lingual side of the talonid are extremely variable in the Shotgun sample, amount- ing in some specimens to little more than irregularities on the talonid margin. An interradicular crest is present. Mo resembles the corresponding tooth of P. silb( rlliu/i more closely than does -M, : Simpson's description (1937, ]). 138) applies practically verbatim to our single specimen. 1 The iiostHi-ior ciii;^!!]!!!)! of >[i of Zaiii/ctiri.s iialeoccna is not as distinct from llif ilcntriil Icisiii lis Siin|isI Jjength Ml Width Mi; 2. -10 -2.85 2.6G ±.051 .135 ±.03(5 5.07 ±1.35 1.00 -1.40 I.IG ±.038(3 M.C.Z. 8423 1.90 i.lo .122 ±.027 10.5 ±2.34 1 Nor. a fortiori, for such forms a.s soriculs and various inicrochiropterans that are exoedafuoiloiit to varyin.^ d»';rre('S. - As measured liy .Simpsou ( l!»,'i7. p. 13S). 1962 A PALEOCENE PICRODONTID DISCUSSION JMatthew, in his description of Zanycteris paleocena, took no account of Picrodus silh€)-Ii)tyi. and indeed there was nothing- in Douo'lass' description and figures to invite comparison between tlie two forms. It remained for Simpson, with new material of P. silhcrlingi and tir.st-kaml knowh^dge of Z. paleocena, to show that they were closely related. As he pointed out (1937, p. 136), the species were distinct, l)ut in the absence of comparable parts it could not certainly l)e stated that this was also true of the genera. We now have lower molars indistinguishable generically from Picrodus found with upper molars clearly distinct from those of Zanycteris. On the face of it, then, the two previously described species would appear to belong to distinct genera. Howevei' i)robable, this cannot yet be considered as certain. The highly specialized lower molars may i)erhaps prove to be rather stereotyped within the family and hence unreliable for generic discrimination. Although quite unlikely, it is thus still conceiv- able that P. .siJberlingi and Z. paleocena may represent one genus, and P. sp. another. Finds of associated upper and lower teeth of the previously described forms will l)e required before all doubt on this score can be set at rest. As regards the broader question of picrodontid affinites, we are in complete agreement with Simpson (1937) that the family cannot be referred to the Chiroptera. The enlarged lower incisor, small upper canine, small i)remolars, long and slender muzzle, and origin of the anterior root of the zygoma opposite M^'^ are definitely non-ehiropteran characters. Since ^latthew's work, it has always been stated that the molars resembled those of the specialized Phyllostomatidae, members of the subfamilies Stur- nirinae, Phyllonycterinae and Stenoderminae. Thanks to the excellent collection of bats in the ^luseum of Comparative Zool- ogy, we have been able to make comparisons with nearly all members of these three groups. AVe are quite unimpressed by the resemblances between their molars and those of the picro- dontids. Essentially these are limited to wrinkled enamel and pointed trigonids, and are more than offset by numerous dif- ferences. Sturnira and the Phyllonycterinae have smooth enamel. In Sturnira, the upper molars have deep, antero-posteriorly run- ning central valleys and no traces of lophs; the lowers have the trigonid and talonid basins continent, the metaconid widely sep- arated from the protoconid, and the trigonid of Mj neither com- pressed nor elevated. In the phyllonycterines, the upper molars 8 BREVIORA No. 175 are triangular and lack lophs and a stylar area ; the lowers have the trig'onids scarcely elevated above the talonids, and all cusps, save the prt)toeonid of M^, incorporated in the rims of tlie talo- nid basins. The Stenoderniinae have wrinkled, often strongly wrinkled, enamel, l)ut the molar structure is very different from that of the picrodontids. The upper molars ai'c short in com- parison with their widths, 'SI- is large (except in I'ygodcrma), frequently larger tluiii AP, and the i)aracones and metaeones are high and trenchant (in Pygoderma the paracones only), standing well above the flattened lingual portions of the teeth. The trigonids are elevated and compressed to points, hut the compression does not involve the whole trigonid as it does in the picrodontids. The metaconid, when not subordinated in the crest running to the apex of the ])r()t()coni(l, is a distinct element situated low on the crown. The jiointed ])rot()c()nitls of both molars and premolars give every appearance of being involved in the same morphogenetic gradient. The few resemblances in molar structure between picrodontids and specialized j^hyllostonuitids scmmu clearly to ])e of the sort brought about by convergenee. The Alicrochiroptera can be elim- inated as possible relatives of the Picrodontidae, and we can see in the latter nothing suggestive of the JMegachiroptera. On the positive side, we can offer no really useful suggestion as to the ordinal affinities of the ])icrodontids. Simpson has reniarked that "reference to the l*rimates is merely a possibility, with no positive evidence to c(unmend it." The possibility, of course, exists, and such ])icrodontid characters as the small size of the trigonid comjjared with that of the talonid and the tight group- ing of the trigonid cusps could be cited as resemblances, how- ever vague, to early members of that order. Kesemblances of this sort scarcely constitute positive evidence, however. Kefer- ence to the Insectivora is e(|ually unsatisfactory. Our (|ueried placement of them there is strictly faute de mieux, due largely to reluctance to use the Primates as a scrap-basket order. REFERENCES Bell, W. G. MS. The geology of the southeastern flank of tlie Wind River Moim- [19.15]. tains, Fremont County, Wyoming. Pp. 1-204. On deposit in the library of tlie Deiiartment of Geology, The University of Wyoming. 1962 A PALEOCEXE I'K IK iD' )X'1II) 9 (;\zix, C. T,. 1SI')I'. I'ak'ocoiU' ii!;iii:iii;ili;iii f';iiii:;is of tlio Bison Basin in south t-putral Wyoming. Sniitlison. Misc. Coll., 131, no. 6: i-iv, 1 ;")". lIKil. Occiiirences of Palcocene Maniiiialia in Tertiary basins of Wy- (imiiit;. Wyo. Geol. Assoc. KUli Ann. Field Conf. 47-52. 1! i'liZKLKK, .1. 11144. lieitriige ziir Kcnntnis dcr Oimylidae. Alili. S = 18.50 b) mertensi (N = 21) 0-13.8 m= 8.95 v = 41.87 c) "schoensis" (N = 7) 11.6-15.3 m = 12.86 v = 13.83 Student's t is significant between helliana and mertensi (9.48), and between "schoensis" and mertensi (6.65), dubiously signifi- cant between helliana and "schoensis" (3.11). The same relation between 'belliana and "schoensis" becomes, however, significant if analyzed by regression lines (t = 4.26). 2. The ratio between the width of the plastron at level of the lateral ends of the humero-peetoral suture and at the abdominal suture : a) helliana (iV = 86) 110-200% )u = 145.45 i; = 11.30 b) mertensi {N = 21) 103-147% m = 114.52 t; = 31.54 Student's t (8.05) is highly significant. As a result of these computations, mertensi appears definitely valid, but "schoensis" seems to be so only for two ratios: 1. the ratio of the sum of humeral and gular sutures to the breadth of the plastron at the level of the humero-peetoral suture and 2. the regression line for the correlation between the length of the pec- toral suture and the maximum breadth of the carapace. From a taxonomic point of view, it does not seem advisable to revive "schoensis" on such slender evidence — -more espe- cially as the helliana sample is geographically highly heteroge- neous. A complex elinal variation from Rhodesia to Ethiopia is possible, since we have no data from the populations between Kenj-a and Ethiopia or Eritrea. Some records from Eritrea (Sordelli 1901, Calabresi 1927, Scortecci 1928) disclose a short- ening of the pectoral suture, as in mertensi. On the other hand, no east-we.st cline exists between mertensi and helliana: the British Museum specimens from Entebbe and Mount Elgon are clearly mcHensi. The two (luestions propounded have thus the following an- swers : (1) A race "schoensis" cannot now be revived for the north- eastern populations of Kinixys helliana. 1962 KINIXYS BELLIANA GRAY 5 (2) The distribution of Kinixys helliana mcrtensi includes Uganda, but its northern limit remains unknown. These conclusions are still provisional. More material could prove not only that "schoensis" is valid but that other subspe- cies can be recognized, or on the contrary that even mertensi merges in helliana through Sudanese and Abyssinian popula- tions. Material examined Kinixys hclliaria helliana Gray MUSEUM OF COMPARATIVE ZOOLOGY: Sudan: Torit (1). Kenya: Golbanti (1), Ithanga Hills (1), Kibwezi (1), Voi (5). Tanganyika: Amboni (1), Kilosa (1), Kiponda to Mitungu (1), Kitaya (1), Mikindani (4), Morogoro (1), Simo near Tabora (2), Turiani (1), Ujiji (2). Zanzihar Island: Zanzibar (1). Nyasaland: Cholo Mtn. (1), Mtimbuka (5). Northern Rhodesia: Isoka (1). Southern Rhodesia: Birchenough Bridge (1), Bula- wayo (1), Hot Springs (2), Lumani (1), Selinda Mtn. (3), Umtali (1). Trawsraai; Naauwpoort (1). Katanga: Kaipiri (1), Lukafu (1). BRITISH MUSEUM (NATURAL HISTORY) : Sudan: Ka- dugli (1). Ethiopia (1), Anseba (2). Somaliland: Berbera (1). near Berbera (1). MUSEE ROYAL de l'AFRIQUE CENTRALE (Tervuren, Belgium). Kenya: Kibwezi (1), Voi (1). Northern Rhodesia: Abercorn (3). Katanga: Kahamhaie (l),Kabinda (1), Kakanda (3), Kansenia (12), Kapiri (8), Lofoi sources (1), Lukafu (2), Lukonzolwa (4), Mwera (1), Ste. Walburge (2). Kivu: Ma- kunga (2). Kinixys helliana mertensi Laurent MUSEUM OF COMPARATIVE ZOOLOGY: Ituri: Mahagi- Port (1). BRITISH MUSEUM (NATURAL HISTORY) : Ucjanda: En- > tebbe (1), Mt. Elgon (2). MUSEE ROYAL de l'AFRIQUE CENTRALE (Tervuren, Bel- gium). Cow(/o, without locality (4). Uele: Dika, (3),Mauda (1), Niangara (1). Ituri (1) : Abimva (1), Gangala na Bodio (3), Mahagi-Port (3). Stanleyville District: Avakubi (1). 6 BREVIORA No. 176 Kinixys helliana nogueyi Lataste. MUSEUM OF COMPARATIVE ZOOLOGY: Dahomey: Bassila (1). Togo: Tohoun (1). Sierra Leone: Kabala (1). Two specimens from Lukolela (western Congo: M.A.C. 4648)^ and from the Kwango District (southwestern Congo : M.A.C. 10736) suggest that a differentiated population exists in the lower Congo region, which somewhat resembles mertcnsi in hav- ing a short pectoral suture and a long abdominal suture. Other specimens are, of course, needed. Key to the races of Kinixys belliana A. Forelimb with 4 claws — Range : Western and northern Cameroon, west to Senegal K. b. nogueyi Lataste B. Forelimb with 5 claws (occasional specimens have 4 claws) 1 — Eatio between the median gular suture + the median humeral suture and the breadth of the plastron at the level of the lateral ends of the humero-pectoral sutures: 75 to 94% (less than 75 in juveniles). Ea- tio between the pectoral suture and the sum of the gular and hu- meral sutures: 0-31% Range: Northeastern Congo and Uganda (presumably also Eepub- lique Centre Africaine and eastern Cameroon). K. b. mertens^i Laurent 2— These ratios, respectively, 49 to 83% and 25 to 69%. Range: Sudan east to Eritrea and Somaliland, south to Natal, north- west to Angola and southern Congo. K. b. belliana Gray BIBLIOGRAPHY Calabresi, E. 1927. Anfibi e rettili raccolti nella Somalia dai Proff. G. Stefanini N. Puccioni. Atti. Soc. Ital. Sci. Nat. Milano, 66: 14-60, pi. I. Laurent, R. F. 1956. Contribution a 1 'herpetologie de la region des Grands Lacs de I'Afrique Centrale. I. Generalites. II. Cheloniens. III. Ophi- diens. Ann. Mus. Congo, Zool., 48: 1-390, pis. I-XXXI. SCORTECCI, G. 1928. Rettili dell 'Eritrea esistenti nelle collezioni del 'Museo Civico di Milano. Atti Soc. Ital. Sci. Nat. Milano, 67: 290-339, figs. 1-8, pis. VII-IX. SORDELLI, F. 1901. Anomalia in una testuggine (Cinixys belliana) del Sudan Orien- tale. Atti Soc. Ital. Sci. Nat. Milano, 39: 111-114, figs. 1-2. Wermuth H. and R. Mertens 1961. Schildkroten. Krokodile. Briickenechsen. G. Fischer, Jena. 1 M.A.C. : Mus6e Royal de I'Afrique Centrale (Tervuren). BREVIORA Mmseium of Contiparsitive Zoology Cambridge, Mass. December 27, 1962 Number 177 RHIPIDISTIAX CLASSTFICATTOX TX RELATION TO THE ORIGIX OF THE TETRAPODS By Keith S. Thomson Department of Biology and Museum of Comparative Zoology, Harvard University IXTRODUCTIOX 111 an extensive study of the nasal region of the lower piiatho- stomes, Jarvik (19-1:2) eoneluded that the four families of rhipi- distiau Crossopterygii represent two distinct stocks (superorders according to Lehman, 1959) — the 'Porolepiformes' (Porolep- idae and Iloloptyehidae ) and the 'Osteolepiformes' (Osteolep- idae and Rhizodontidae). Of major interest to students of vertebrate evolution was his conclusion that the structures seen in the snout of 'Porolepiformes' were closely comparable to those seen in the recent LTrodela, and those of 'Osteolepiformes' were comparable to those of the recent Anura, and that here was evidence of a biphyletic origin of the tetrapods. His work was based upon detailed studies of two genera, PoroJcpis and Eu.sthcnoptcron (a rhizodontid). Jarvik 's material was unusual in that it showed details of the endocranial part of the snout which it is not usually possible to study in fossil crossopterygi- ans ; in recent years, however, more such material has been described (Vorobjeva, 1959, 1960a, 1960b; Kulcyzcki, 1960) and 0rvig (1957) has published a full treatment of the inter- relations of the Rhipidistia on the basis of the structure of the scales. For the last year or so I have been engaged in a study of the ethmoid region of the osteolepids Megalichthys and Ectosteorhachis, continuing the work started by Romer (1937, 1941). Briefly, the newly available evidence does not support all of Jarvik's original conclusions. Because of the relation of this 2 BREVIUKA No. 177 \\(»i'k to the problem of tlic oi-ijiin of tlic tctrapods and particu- larlv to the currently disinitcd tlieories of the ancestry of the recent Amphibia, I haye decided to publish this short reyiew of the more ]iertinent points in adyance of a more thorough treat- ment. This may also be useful since the work of Vorobjeya, be- ing in Russian, may not be readily ayailable to eyeryone. This paper forms })art of the work to be submitted to the De- partment of Biology, Haryard Uniyersity, in fulfillment of the requirements for the degree of Doctor of Philosojihy ; it is my pleasure to acknowledge the constant help and guidance of Pro- fessor A. S. Romer during all this study; he and Dr. E. E. Wil- liams and Professor B. Patterson read and criticized the manu- script. I am also grateful for the friendly encouragement of Drs. Jaryik and ^ryig in Stockholm ; the former has allowed me to see an as yet unjiublished manuscript on this subject (Jaryik, in press). My friend ]\Ir. Simon Karlinsky provide*! the translations from the Russian. My studies haye been sup- ported by the awai-d of X.A.T.O. Science Studentship 3/60/955 by H. M. Department of Scientific and Industrial Research. London, and by tlie .T(^ffries Wyman Scholarship at Harvard Uniyersity. MATERIAL It is necessary to establish the taxonomir status of certain of tlie rhipidistians concerned in this study; these are: McgalicIifJii/s Agassiz 1841- (including Ecio^fi orhavhis Cope 1880). Carboniferous and Lower Permian of Europe and North America. PlatycephaIicJit]iys Vorobjeva 1959. Upper Devonian of U.S.S.R. Panderichthys Gross 194U Upper Devonian of U.S.S.R. Porolepis {P. ex grege posiKntiensis [Kade 1858]). Lower Devonian of Poland. Mfgalichthys-Ectosirorhach is. Ecfosfcorhochis was originally described by Cope (1880) from material from the Lower Per- mian of Texas; it w-as later (1891) referred by him to the Car- boniferous genus Megalichihiis. In recent years it has been sug- gested several times (see, for example, Romer, 1941) that the two genera are, in fact, distinct and that the Lower Permian material will have to be referred back to EctostcorJwchis. No formal diagnosis of this has been given ; however, in advance of 1962 RIIIPIDISTIAX KELATI0X8I Ill's 3 siR-h a diagnosis T shall follow precedent and refci- to tlic Per- mian fish, for {•onvenience, as ' EctostcorJiachis.' Mcgalichfhys and Eciostrorhachis are generally acknowledged to be members of the Osteolepidae. Berg (lf)58) niiites them with Owen's Parahatmchn.'i to form the separate family Para- batrachidae; however, available evidence, e.g. Bystrow (1950), shows that Megalichfhys is closely related to Osicoh pis itself. PoroJepis. Knlcyzcki's (1960) material of ForoJi pis is with- out doubt correctly referred to that genus. Platijcephalichihys and Pandcrichthys. These genera are in- volved in a long series of taxonomic shufflings concerning the genera Cricodus, Dendrodus and Polycodus; they may be con- sidered most conveniently together. The best way to review the situation seems to be to start with the description by Rohoii (1889) of several specimens from the Upper Devonian of Ru.ssia, some of Avhich he assigned to Dendrodus hiporcafus Owen (Ro- hon, 1889, plate 1, figures 1, 5, 9, a skull, and 2, 7 8, a tooth and two scales), while others he named Cricodus wenjucovi (Rohon, 1889, plate 1. figures 4, 6, a skull from the River Ojatj, and 3, 11. a skull and lower jaw from the River Sjass). Since these specimens are not given numbers by Rohon, I shall refer to them, for convenience, by the number of the figure by which they are illustrated. Gross (1933) placed tho.se specimens figured in plate 1, figures 1, 3, 4, 5. 6, 9, together with new material from the Baltic Old Red Sandstone, in the genus Pclyplocodus Pander 1860, as the new combination Polyplocodus wenjucovi. Jarvik (1937), how- ever, put "figures 1, 3, 4, 5, 6. 9, 11" in Eusfhenopteron as £". wenjucovi. He also concluded that the names Cricodus, Den- drodus, and Prjlyplocodus should not be assigned to any fresh material because of the fragmentary and enigmatic nature of the originals. Hence Gross (1941) named a new genus Pan- derichthys in order to reassign material named by him (1930) Polyplocodus (Cricodus) rJwmholepis (see also Gross 1933. 1936). Now there enters a possible source of confusion because Gross (1941) named a second new species — PanderichfJiys bysfrovi — for the material from the Baltic Old Red which he had named Polyplocodus wenjucovi in 1933 (see above), while leaving the Rohon material in Eusfhenopteron tvenjucovi. Xov- objeva (1960a) named a third species — Panderichthys stol- bovi. 4 BREVIORA No. 177 Til 1959 Vorobjeva described a new genus Platycephalichthys, with the type riaf)/C( pJialichfhifn hischoffi, based on recently collected material from the Tjiper Devonian of Russia, and in- cluded in this genus the Rohon specimen from River Ojatj ("fig- ures 4 and 6") as a new species Plafyccphalichthys rohoni. Later (1960b) she referred the remaining Rohon material to Eusthenodon Jarvik 1955 as Eusihenodon wenjucovi. Thus the two genera now comprise : Panderichfhys P. rhomholepis Gross 1941 P. hystrovi Gross 1941 P. stolhovi Vorobjeva 1960 Plafycephalich th ys P. hischoffi Vorobjeva 1959 P. rohoni Vorobjeva 1959 From the structure of the scales and teeth, Plafyccphalichthys is a member of the Rhizodontidae. Panderichfhys was assigned by Gross (1941) to the Rhizodontidae but there seems little doubt (Vorobjeva 1960a, and Orvig 1957) that it should be placed with the Osteolepidae. SUBDIVISION OF THE RIIIPIDISTIA Jarvik (1942, p. 489) lists a series of dift'erences between Porolepis and Eusfhcnoptcron which he considers to be repre- sentative of a basic split within the Rhipidistia. He also (1942, pp. 417, 495) discusses the connections between these fishes and the recent Ami^hibia. In general, the same characters are in- volved in the two arguments. In the next few pages I shall review recent findings which indicate that certain of these char- acters no longer support Jarvik 's distinctions. These are all characters which may readily be determined in the fossils; this is in contrast with some of Jarvik 's points involving the passage of nerves and vessels, the refutation of which could be as diffi- cult as their interpretation. Some of Kulcyzcki's (1960) state- ments may be disputed, as I shall mention later, on the grounds that his material was insufficiently well preserved. This criti- cism cannot apply to Vorobjeva 's work, particularly since the most important of the characters she has described are, as already mentioned, easily ascertained. The recognition, in certain anatomical characters, of resem- blances between specific rhipidistians and recent amphibians, without the supporting evidence of a fossil lineage (lacking in 1962 RHIPIDISTIAN RELATIONSHIPS 5 the ease of the Urodela and Anura) becomes, to a certain ex- tent, a matter of subjective judgment. In the following pages I shall consider primarily the evidence for a basic distinction between 'Osteolepiformes' and 'Porolepiformes, ' for it is upon the validity of this supposed dichotomy that all subsequent phylogenetic hypothesis must rest. The nature of the anterior palatal fenestrae. According to Jarvik (1942, p. 489, etc.), an important difference between the "Osteolepiformes' and the 'Porolepiformes' lies in the nature of the paired palatal recesses present between the anterior edges of the vomers and the posterior rim of the premaxillae. In Eusthenopteron this region has the form of a shallow, partially sub-divided groove, the 'prenasal groove,' limited posteriorly by the edges of the vomers; in Porolepis there is a pair of '])its' extending backwards between and separating the vomers. These palatal recesses which are described by three names — 'fossae apicales, ' 'anterior palatal fenestrae,' and 'pre-nasal pits' — were assumed by Jarvik to have contained glandular organs homologous with the various intermaxillary glands of recent Amphibia. It was further proposed that the condition in Eus- thenopteron foreshadowed that of recent Anura, and the condi- tion in Porolepis tliat of Urodela. It had already been suggested, however, that these recesses served solely to receive the points of large tusk-like teeth set in the tips of the lower jaws (Holmgren and Stensio, 1936; Romer, 1937; and now also Kulcyzcki. 1960). The material at my disposal shows quite clearly that this latter explanation is the true one for the Osteolepidae. Plate I shows the tusk fitting into the palatal recess, leaving no room for any glandular struc- ture. That this is also the case in other 'Osteolepiformes' may be deduced from the similar large tusks of the lower jaws — for example, in Panderiehthys (Gross, 1941). In 'Porolepiformes' the situation is very similar, but in tliis case, as Jarvik has .shown, the teeth concerned are a pair of tooth whorls. This extremely interesting discovery provides, incidentally, positive indication that the familiar Onychodiis, long known from such tooth whorls, is, in fact, a 'porolepiform' rhipidistian (Jarvik, in press). The presence of the paired tooth wiiorls speaks, indeed, for tile unity of the 'Porolepiformes,' but the morphological rela- tionship between the type of palatal fenestra and the lower jaw dentition completely precludes any phylogenetic relationship 6 BREVIOBA No. 177 between the fenestrae in Rhipidistia and the jilandnlar organs of recent Amphibia. Evidence from, the cnniiaJ cavity. A point of resemblance, apparently possible of interin-etation as evidence of relationship, between Porolepis and the Hrodela is that in both "... the in- ternasal wall is broad and does not form any nasal septnm. It lodges the ethmoid part of the cranial cavity" (Jarvik, 1942, p. 417). This, moreover, is supposed to be in direct contrast with the 'Osteolepiformes' (and Anura) — by extrapolation from the situation in Eusthenopteron in Avhicli the internasal wall is relatively narrow and also solid. Kulcyzcki (1960) noted that his material of Porolepis did not show any forward extension of the cranial cavity between the nasal sacs; his material was perliaps not as good as might be desired fully to substantiate this view, since it consisted entirely of natural casts and not true bony remains. At the same time that Kulcyzcki 's work was published there appeared the work of Vorobjeva. Her material consisted of several portions of the skulls of Plotycephalichthys and Panderichthys, both of which are undoubtedly 'osteolepiform' and both of which, she states, have an ethmoid ext(»nsion of the cranial cavity. Her descrip- tion of PlatyeephalicJttJiys is interesting; ". . . a wide inter- nasal portion (of the braincase) with a cavity stretching forward almost to the front edge of the skull" (Vorobjeva, 1960a, trans.). In the face of evidence that a rhizodontid and an osteolepid have a 'pars ethmoidalis cranialis' and at least one species of Porolepis may not, one is forced to conclude that this character is of no significance in any attempt to distinguish supra-familial groupings within the Rhipidistia. One may further ])ear in mind that the extent of the cranial cavity need have no relation to the extent of the brain contained therein. A dramatic demon- stration of this is afforded by the coelacanth Latimcria (Mi]U)t and Anthony, 1958), in which a large cranial cavity contains but a small brain. Presence or absence of the pars ethmoidalis cranialis is thus a rather labile character in the Rhipidistia and. having no great anatomical or functional basis, no phylogenetic speculation may reasonably be drawn from it. Nasal apertures aud the )iasal cavity. All known 'Porolepi- formes' are characterized by the presence of two external nares and a choana. Further, the endocranial opening for the poste- rior naris is confluent with that for the choana. In Evsthenop- teron, on the other hand, there is but one external naris and this 1962 RHIPIDISTIAX RELATIUXSHIPS 7 is separated from the choana by the lamina iiario-choanalis of the endoskeletal nasal capsule. T'ntil the work of X'orobjeva this was believed to hold true for all 'Osteolepiformes" but 7^^//;- dcrichihijs is described fVorobjeva, 1960a) as liavinures it is not possible to determine Avhether the described contiuenee of the posterior naris with tiie fenestra endochoanalis is (liic iiici'dy to a defect in the preservation oi- not. This is indeed a strano'e situation for a fish which is in all major respects to be consid- ei'ed 'osteolepiform.' Jarvik has other speculations based on the detailed configura- tion of the nasal cavity itself, but these seem to be more open to dispute in connection with phylogenies spanning 300 million years. Ectosteorhachis certainly lacks most of the ridges, grooves and depressions described in Eusth< nopfi ran and supposed to be typical of all 'Osteolepiformes.' Jarvik described in Eustlienopteron a foramen in the post- nasal wall which he states is the posterior endonarinal fenestra (Jarvik 's terminology). This is also present in Ectosteorhachis, but the presence of an anterior naris in Panderichthys may pos- sibly indicate that the single external naris of 'Osteolepiformes' is homologous with the posterior naris of Porolepis. In this case the foramen in the post-nasal wall, which Jarvik considers to be the forerunner of the tetrapod naso-lachrymal duct, may ])oh- sibly have to be interpreted in some other fashion, but consider- ably more evidence is needed to settle this point. Other structures. There are several other points, noted by Jarvik as indicative of a division within the Rhiiudistia, which are contradicted by the anatomy of Ectosteorhachis: 1) Jarvik (1942, p. -t92) states that a difference between 'Osteolepiformes' and 'Porolepiformes' is that in the former the vomers are in mesial contact, while in the latter they are separate from each other. Although in tiu^ sectioned material at my dis- posal the vomers are slightly displaced, it seems that these bones are not in mesial contact. Further, they lack the pos- terior extension passing back on either side of the tooth-bearing part of the parasphenoid, which has also been stated to be typi- cal of 'Osteolepiformes.' The vomers are thus much more simi- lar to those of Porolepis than to tbose of Eusthenopteran : they lie entirely anterior to the tooth-bearing part of the parasphe- noid. 8 BREVIORA Xo. 177 2) Jarvik states (1942, p. 492) that in 'Osteolepiformes' the parasphenoid is narrow and in 'Porolepiformes' it is broad. Bnt in Ectosteorhachis (an osteolepid) only the tooth-bearing part is narrow — the parasphenoid is continued forward and later- ally as a broad, if thin, film of bone fused to the ventral surface of the endocranium (cf. Romer, 1937, p. 19). 3) Jarvik states that lack of the external parietal foramen is tyi)ieal only of 'Porolepiformes'; both MegalichfJnis and Ecto- steorliachis lack it, however. Kuleyzeki further criticizes Jarvik 's description of the snout anatomy of Porolepis, especially the detailed description of the nasal capsule and the canals for nerves and vessels (see espe- cially Kuleyzeki, 1960, pp. 81-94). On the other hand, Jarvik states (in press) that in most details the new material he has of Glyptolepis fully bears out his description of Porolrjns. DISCUSSION AND CONCLUSIONS The nature of the fossil remains, by which the Ehipidistia are known, is extremely variable, ranoiny from the excellent and fairly plentiful material preserved *in the round' of Eus- thenopteron, Ectosteorhachis and Glyptolepis (Jarvik, in press, and in preparation), to the fragments of skulls, isolated teeth, and scales typical of most genera. It is not surprising, there- fore, that most phylogenies and taxonomic studies have been based on the histology of teeth and scales. The latter approach is, of course, open to some doubts and reservations, but the re- cent work of 0rvig seems particularly important (see 0rvig, 1957, p. 409 for phylogeny). It is especially interesting to note the positions assigned by 0rvig to Pandcriclithys and Platycephalichthys. According to Jarvik 's interpretations of crossopterygian anatomy, the pres- ence of the pars ethmoidalis cranialis in both genera and the two external nares in Panderichthys would ally them with the Poro- lepis lineage. All other evidence, however (J^rvig and Vorob- jeva), opposes this view and maintains their 'osteolepiform' status. One is forced to conclude that these two characters are more labile than was formerly supposed ; they fail to show cor- relation with other, diagnostic, characters. Schmalhausen (1959) rejected Jarvik 's theories on the grounds that the two 'types' of skull merely reflected the relative pro- portions of the skull, Porolepis being broad-snouted and Eus- tJienopteron narrow-snouted. Schmalhausen was referring par- ticularly to the difference between the two types of anterior 1962 RHIPIDISTIAN RELATIONSHIPS palatal fenestra, the nature of whieh we have already seen to depend on the type of lower jaw dentition. Although Sehmal- hausen's idea is an attraetive one, it does not explain the dis- crepancy in the oeenrrenee of the pars ethnioidalis ci'aiiialis. The 'Osteolepiformes' contain both broad and narrow snouted forms, but the presence of the pars ethnioidalis cranialis is in- dependent of this factor : Pars ethnioidalis Osteolepidae Rhizodontidae Snout cranialis Broad Absent Narrow Present Narrow Absent Broad Present j Ectostf orliachis I Pandcrichfln/s j Eusfhenopfcron [ Platyccphaiichthys This character is not even related to the relative width of the internasal wall, which seems to be dependent on the relative size of the nasal capsules rather than on the external proportions or intrinsic 'osteolepiform'/'porolepiform' nature of the snout. I can think of no explanation of the apparently random oc- currence of these characters except that the Rhipidistia form a fairly close-knit group within which comparable variations can occur in all families. There is no doubt that a distinction of some sort can he ili-awii between a porolepiform and an osteolepiform assemblage (al- though the former is a much more compact group than the lat- ter). The controversy lies in the status to be assigned to eacli. The two groups have been interpreted by Jarvik (1942. I!).!."). 1960) to represent a fundamental split within the Rhipidistia and, by extrapolation, within the recent Amphibia also. There are two major objections to the arguments which are presented to substantiate this hypothesis. Firstly, only one genus of 'Osteo- lepiformes' (Ei(sthenopfcrou) was available for consideration in any detail, and secondly, all the characters dividing the two groups are chosen with reference to the proposed relation to the Urodela or Aiiura and not at all with reference to the Rhipidis- tia in general. Taking into consideration the evidence presented above and bearing in mind other more detailed jioints such as are dis]nited by Kulcyzcki and to which 1 hope to return in a future work, it seems that the two groui)s of Rhipidistia arc more closely related than has been stated. Furthermore, those characters which might reflect a relationship with particular 10 BREVIORA No. 177 recent Amphibia are present in both croups of rhipidistians and are not, therefore, justifiably so considered. From my own studies of the Rhipidistia, I personally favour the view propounded by Berg (1958) that there are three <>roups of Klii])idistia : the Porolepiformes, the Osteolepiformes and the Rhizodontiformes. I would allow each group no more than super-familial rank. Literature cited Berg, L. S. 1958. System der rezenten iind fossileii Fisehartigeii uiid Fisehe. 31(t pji. Deutscher Verlag, Berlin. Bystrow, a. p. 1939. Zahnstniktur der Crossopterygier. Acta. Zool., 20: 283-338. 1950. Mieioscopic structure of tlie bones and teeth of the Carbonifer- ous crossopterygian tish MepaUclitliys (family Osteolepidae). Dokl. Akad. Xauk. U.S.S.B., 34: 119-121. (In Russian.) Cope, E. D. 1880. Second contrilnition to the histuiy of the Vertel)rata of the Permian formation of Texas. Proc. Am. Phil. Soc, 19: 38-58. 1891. On the character of some Palaeozoic fishes. Proc. U.S. Nat. Mus., 14: -447-463. Gross, W. 193(1. Die Fische des mittleren Ohl Red Sud-Livhmds. Geol. Paliiont. Abh., N.F., 18: 121-ir)(i. 1933. Die Fische des Intltischen Devons. Palaeontographica, 79, Alit. A: 1-72. 1936. Beitrage zur Osteologie lialtis(diei- und rlieinischer Devon-Cros- sopterygier. Palaont. Z., 18: 129-155. 1941. Uber den Unterkiefei ciniger devonistdier Crossopterygier. Abh. Preuss. Akad. Wiss., umth.-nat. Kl., no. 7: 1-51. Holmgren, N. and E. A. Stensio 1936, Kranium und Visceralskelett der Akranier, Cyclostomen und Fisclie. In L. Bolk ei alio, Handbucli der Vergleichenden Ana- tomie der Wirbeltiere, vol. 4: 345-353. Jarvik, E. 1937. On the species of Eusthenopteron found in Russia and the Baltic States. Bull. Uppsala Geol. Inst., 27: 63-127. 1942. On the structure of the snout of crossopterygians and lower gnathostomcs in general. Zool. Bidrag, 21: 237-675. 1955. The oldi>st tetrapods and their forerunners. Sci. Monthly, 30: 141-154. 1960. Theories de 1 'evolution des vertebres. 104 pp. Masson et Cie, Paris. 1!)62 RHIPIDISTIAN RELATIONSHIPS 11 (ill l>i't'ss I. On tlu' Porolepiformes and the origin of tlic iiidiji'les. (Unpub- lished, read before tbe Intei-national ('olloiniiutn oil Kvobition, Paris, May, l!t(jl.j KlLCYZCKI, J. 1 !•<)(». Pnrolcpis (Crossopterygii) from the Lower Devonian of tlie Holy Cross Mountains. Aeta Palaeont. Polonica, 5 (1) : 65-106. Lehman, J. P. 1959. L 'evolution des vertebres inferieures. 188 pp. Dunod, Paris. MiLi>()T, J. and J. Anthony 1958. Anatomie de Latimeria ehalumnae. Vol. T. Squelette, Museles et Formations de soutien. 122 pp. Centre National de la Rech- erche Scientifique. Paris. Orvig, T. 1957. Remarks on the vertebrate fauna of the lower Upper Devonian of Escuminae Bay, P.Q., Canada, with especial reference to the porolepiform crossopterygians. Arkiv. Zool., 10 (6): 367-126. ROHON, J. V. 1889. Die Dendrodonten des devonischen Systems in Russland. Mem. Acad. Imp. 8ci. St. Peterslnirg, (7) 36 (14): 1-52. ROMER, A. S. 1937. The braincase of the Carlioniferous crossopterygian Megalich- ihys nitidus. Bull. Mus. Conip. Zool., 82 (1): 1-67. l9-tL Notes on the crossopterygian hyoniandilnilar and braincase. J. Morph., 69 (1,): 141-160. SCHMALHAUSEN, I. I. 1959. Concerning monophyletism and i)olyphyletism in relation to the problem of the origin of the land vertebrates. Bull. Mosk. Obshch. Isp. Pri. Biol., 64 (4): 15-33. (In Russian.) VOROB.JEVA, E. I. 1959. A new genus of crossopterygian fish — Plaii/cephalichihjis from the Upper Devonian of Lovat. Palaont. J. Akad. Nauk. U.S.S.R., 1959, no. 3: 95-106. (In Russian.) 1960a. New facts aljout the genus Paiiderichthi/s from tlic Devonian of the U.S.S.R. Palaont. J. Akad. Nauk. U.S.S.R., 1960, no. 1: 87-96. (In Russian.) 1960b. Concerning the .systematic position of Eusfhenopteron wenju- covi (Rohon). Palaont. J. Akad. Nauk. U.S.S.R., 1960, no. 2: 121-129. (In Russian.) Xo. 177 Plate 1. Megalichtlni.s fidiu tlie Scottish Carboniferous. M.C.Z. 8941. Tip of the ?noi:t in anterior view, showing iiremaxillary teetli, dentary tusks, and the anterior pahital fenestrae exposed by a natural break in the specimen: A. Photographed immersed in water. B. Semi-diagrammatie sketch of A, emphasizing the pertinent features. V / V-t-.^ BREVIORA Museium of Comparsitive Zoology Cambridge, Mass. December 27, 1962 Number 178 ON A NEW SPECIES OP THE EARTHWORM CENTTS TBIGASTEE BENIIAM 1886 (OCTOCHAETIDAE)i By G. E. Gates University of Maine, Orono, Me. The presence, in Cuba, of an apparently endemic Trigaster is of such zoogeographical importance as to contra-indicate further and perhaps indefinite delay in publishing' avail;i1 Mexican one. The Porto Rican species seems to be closer to those of St. Thomas. Earthworms (of course after exclusion of the transported forms), because of their v^-ontinement to, breeding in, and slow migration through the soil, long have been believed to be second in paleogeographical importance to no other group of animals. Trigaster, perhaps more than any otiier North American genus, now seems likely to provide especially interesting data if the native faunas of the Mexico-Virgin axis can be studied before they are too profoundly affected I)y human interference with their habitats. 2 Male funnels of x, in other octoehaetid genera, frequently are retained after the testes of that segment ceased to mature sperm. 3 Little is known of the oligochaete faunas of the West Indies, including His- paniola. Indeed, nothing- at all is known of the oligochaetes of Santo Domingo- I Date Due Mff^*M99V Harvard MCZ Libra 3 2044 066 302 720 nocc MOT /"•if^/^i II A ■*•!