/ • HARVARD UNIVERSITY Library of the Museum of Comparative Zoology DO£S NOT CIRCUJLAII BREVIORA ^^ ^^ MUSEUM OF COMPARATIVE ZOOLOGY Harvard University Numbers 265-295 1967-1968 CAMBRIDGE, MASS. U.S.A. 1969 Edited by Nelda E. Wright CONTENTS BREVIORA Museum of Comparative Zoology Numbers 265-295 1967 No. 265. New land-locked fish species of the genus Galaxias from North Auckland, New Zealand. By R. M. McDowall 1 1 pp. April 6. No. 266. A new species of Vesicomya from the Caribbean Sea (Mollusca: Bivalvia: Vesicomyidae). By Kenneth Jay Boss. 6 pp. April 6. No. 267. A new generic arrangement for some Australian scincid lizards. By Allen E. Greer. 19 pp. June 29. No. 268 The ecology and behavior of two sympatric Lyeodactylus geckos. By Allen E.Greer. 1 9 pp. June 29.' No. 269. Two new species of Aniphitarsus (Ophiuroidea) from the Western North Atlantic. By Amy Schoener 9 pp June 29. ' ^^' No. 270. Additions to the unionid fauna of the Gulf drainage of Alabama, Georgia, and Florida (Mollusca: Bivalvia) By Richard I. Johnson. 21 pp. June 29. No. 271. Chelonia depressa Garman re-investigated. By Ernest E. Williams, Alice G. C. Grandiso^n. and Archie F. Carr, Jr. 15 pp. November 17. No. 272. The ecological distribution of the anoline lizards around Kingston. Jamaica. By A. Stanley Rand 18 pp November 17. No. 273. Studies on neotropical Pompilidae (Hymenoptera). III. Additional notes on Epipompilus Kohl. By Howard E.Evans. 15 pp. November 17. No. 274. The genus Nops (Araneae, Caponiidae) in Panama and the West Indies. By Arthur M. Chickering. 19 pp. November 17. No. 275. A new scincid lizard from the northern Solomon Islands. By Allen E. Greer and Fred Parker. 20 pp. Novem- ber 17. No. 276. The generic relationships of the African scincid genus Eumecia. By Allen E. Greer. 9 pp. November 24. No. 277. Lestidiuni bigelowi, a new species of paralepidid fish with photophores. By Michael J. F. Graae. 10 pp. No- vember 24. No. 278. Notes on the emberizine sparrow Rhynchospiza stolz- inanni. By Raymond A. Paynter, Jr. 6 pp. Novem- ber 24. No. 279. A second skink with fragmented head scales from Bou- gainville, Solomon Islands. By Allen E. Greer and Fred Parker. 12 pp. November 24. 1968 No. 280. The notatus group of Sphaerodactyhis (Sauria, Gekkoni- dae) in Hispaniola. By Benjamin Shreve. 28 pp. March 19. No. 28 1 . Additional notes on batoid fishes from the Western Adan- tic. By Henry B. Bigelow and William C. Schroeder. 23 pp. April 5. No. 282. The extinct baboon, Porapapio jonesi, in the early Pleisto- cene of northwestern Kenya. By Bryan Patterson. 4 pp. April 5. No. 283. Scythian ammonoids from Timor. By Bernhard Kum- mel. 21 pp. April 5. No. 284. The relationships of Arwlis of the roquet species group (Sauria: Iguanidae) — III. Comparative study of display behavior. By George C. Gorman. 31 pp. April 5. No. 285. The genera of puffbirds (Bucconidae). By G. W. Cot- trell. 5 pp. April 5. No. 286. The status of Nesogalaxias neocaledonicus (Weber and de Beaufort) (Pisces. Galaxiidae). By R. M. McDowall. 8 pp. April 5. No. 287. The sequence of genera in the Estrildidae (Aves). By Ernst Mayr. 14 pp. April 5. No. 288. The epidermal glands of Lygodactylus (Gekkonidae, Lacertilia). By P. F. A. Maderson. 35 pp. May 29. No. 289. The genus Miagrammopes (Araneae, Uloboridae) in Panama and the West Indies. By Arthur M. Chicker- ing. 28 pp. May 29. No. 290. Geographic Variation in the Hispaniolan frog Eleuthero- dactylus wetmorei Cochran. By Albert Schwartz. 13 pp. May 29. No. 291. A new species of Tribolonotiis (Lacertilia: Scincidae) from Bougainville and Buka, Solomon Islands, with comments on the biology of the genus. By Allen E. Greer and Fred Parker. 23 pp. May 29. No. 292. Herpetogeography of Puerto Rico. V. Description of a new species of Sphaerodactylus from Desecheo Island. By Harold Heatwole. 6 pp. May 29. No. 293. Intra- and interspecific chromosome variation in the lizard Anolis cristatellus and its closest relatives. By George C. Gorman. Richard Thomas, and Leonard Atkins. 13 pp. May 29. No. 294. Distribution and biology of the opisthoproctid fish Win- teria telescopa Brauer 1901. By Richard L. Haedrich and James E. Craddock. 1 1 pp. May 29. No. 295. The Chaiiares (Argentina) Triassic reptile fauna. IV. The dicynodont fauna. By C. Barry Cox. 27 pp. May 29. INDEX OF AUTHORS BREVIORA Museum of Comparative Zoology Numbers 265-295 1967-1968 No. Atkins, Leonard 293 Bigelow, Henry B 28 1 Boss, Kenneth Jay 266 Carr, Archie F., Jr 271 Chickering, Arthur M 274, 289 Cottrell, G. W 285 Cox, C. Barry 295 Craddock, James E 294 Evans, Howard E 273 Gorman, George C 284, 293 Graae, Michael J. F 277 Grandison, Alice G. C 27 1 Greer, Allen E 267, 268, 275, 276. 279. 291 Haedrich, Richard L 294 Heatwole, Harcld 292 Johnson, Rich \rd 1 2 73 No. KUMMEL, Bernhard 283 Maderson, p. F. a 288 Mayr, Ernst 287 McDowell, R. M 265, 286 Parker, Fred 275, 279, 291 Patterson, Bryan 282 Paynter, Raymond A.. Jr 278 Rand, A. Stanley 272 ScHOENER, Amy 269 schroeder, william c 281 Schwartz, Albert 290 Shreve, Benjamin 280 Thomas, Richard 293 Williams, Ernest E 271 BREVIORA Mmseuim of Coimparsitive Zoology Cambridge, Mass. 6 April, 1967 Number 265 NEW LAND-LOCKED FISH SPECIES OF THE GENUS GALAXIAS FROM NORTH AUCKLAND, NEW ZEALAND' R. M. McDowall Museum of Comparative Zoology, Harvard University Abstract: Two species of Galaxias Cuvier (Galaxiidae) are described from small land-locked coastal dune lakes in North Auckland, New Zealand. They are small shoaling species similar to and apparently derived from the sea-going Galaxias maculatus attenuatiis (Jenyns). INTRODUCTION In the New Zealand Galaxiidae, speciation appears to be taking place by the land-locking of sea-going species. This is a pattern familiar in the Northern Hemisphere Salmonidae, and also in the Southern Hemisphere in the Retropinnidae (McDowall, 1965). Stokell (1964) described a new species of Galaxias (G. parrishi), which he considered to be a land-locked derivative of G. maculatus ignotus, and noted another land-locked form from Victoria, Aus- tralia. The writer (McDowall, 1966) noted a similar form col- lected from another Victorian lake (see Pollard, 1964), and the occurrence of further such derivatives from G. maculatus attenu- atus, in New Zealand coastal lakes. About 15 years ago, the New Zealand Marine Department con- ducted a survey of the chain of small lakes on the west coast of the North Island of New Zealand, to determine their ecology in relation to the introduction of suitable game fishes. In their report on this survey, Cunningham et al. (1953) described the fish faunas of these lakes as follows: "Various species of Galaxias, Gobio- morphus, and Retropinna were recorded," and they listed in a table the lakes from which these genera were taken. In 1963, the writer examined a sample of G. maculatus attenuatus-Vike fishes from one of these lakes, which, on close examination, appeared to 1 Fisheries Research Publication No. 93. BREVIORA No. 265 be distinct from G. maculatus attenuatus in a number of details. Because this sample was rather old, it was felt desirable to re- collect the fish, before description, and this was done in March 1965. Large samples of Galaxias were collected from Lake Waiparera, a few miles north of Kaitaia, and from Upper Lake Rototuna, on the peninsula west of the northern arm of Kaipara Harbor (Fig. 1). Inasmuch as these lakes are land-locked, it was to be expected that the Galaxias populations would differ mor- phologically and behaviorally from Galaxias maculatus attenuatus. L WAIPARERA- Figure 1. Locality map of North Island, New Zealand, showing location of: 1. Lake Waiparera, 2. Lake Rototuna. Being land-locked, their life cycles are also necessarily modified from that of G. maculatus attenuatus which has marine juveniles. Examination of the fresh samples from these two lakes, showed them to be distinct from G. maculatus attenuatus and from each other. They do not resemble any other New Zealand species of Galaxias. Characters suggesting relationship to G. maculatus at- tenuatus, apart from their general appearance and coloration, in- clude shoaling behavior as adults, forked caudal fins, similar denti- tion, lack of pyloric caeca, and the long, many-rayed anal fin. 1967 NEW SPECIES OF GALAXIAS 3 The isolation of sea-going species in small lakes, as appears to have taken place in these instances, poses a problem in species delimitation. The alternatives lie between describing each isolate as a distinct species or subspecies, or including all the populations in a single, widely variable species, derived from the parent sea- going species. Since there has been a fundamental change in the life history pattern of the land-locked species, it seems likely that the parent and derivative species have acquired reproductive iso- lating mechanisms. This conclusion is supported by the fact that the derivative populations have developed characteristic morpho- logical differences from the parent population. The specific dis- tinctness of the two derivative populations under examination is another problem. Ideally, breeding experiments and behavioral studies of the two populations would be desirable to provide the answers, but these are outside the scope of the present study. On examining the morphological differences between the two popula- tions, it seemed likely that they were independently derived from the parent stock and therefore more closely related to the parent than to each other. The population in Lake Rototuna is, from both morphological and geological considerations, probably an earlier derivative of G. maculatus attenuatus than is the Lake Waiparera population. Lake Rototuna occurs in well stabilized rolling sand dunes, once covered in bush, at an altitude of 300 ft. (91 m). whilst Lake Waiparera is in still shifting sand dunes, about a hun- dred feet (36 m) above sea level. Corresponding to their apparent difference in age, and perhaps to differing ecological conditions in the two lakes, the degree to which these species have diverged from their common parent species differs greatly. The differences are most obvious in vertebral count, a character which has proved useful in separating Galaxias species in New Zealand. Assuming relationship to G. maculatus attenuatus, the Lake Rototuna popu- lation has traversed the full range of vertebral number in the New Zealand Galaxiidae, from the maximum in G. maculatus attenuatus of 59-64, to a minimum of 47-50. The vertebral number in the Lake Waiparera population is also reduced and almost completely disjunct from both G. maculatus attenuatus and the Lake Rototuna population, viz. 54-59 (Fig. 2). The direction of divergence from G. maculatus attenuatus, seen in the two derivatives, varies; sometimes the counted structures are fewer in number, as in vertebrae and most of the fin-ray counts (Fig. 3). In the case of gill rakers (Fig. 2), in contrast, G. macu- latus attenuatus has a total count on the first arch of 13-17, the Waiparera fish 15-17. and the Rototuna fish 18-23. These data BREVIORA No. 265 C. CBACmS 11 Jk C ATTENUATUS iiL g. USITATUS h jL ll iL Jll jlll 4 S 6 7 8 13 14 1^ 16 17 18 19 SO 21 22 23 an ANCHrOSTECALS ClLl RAKERS 47 48 49 50 SI 52 51 54 S5 S6 57 58 59 60 61 62 63 k4 VERTEBRAE Figure 2. Frequency distribution for counts of branchiostegals, gill rak- ers, and vertebrae in Galaxias gracilis, G. maculatiis attenuatiis, and G. usitatus. DOR5 Al f I N RAYS ANAL FIN RAYS PEL\-1C FIN RAYS Figure 3. Frequency distribution for counts of dorsal, anal, and pelvic fin rays in Galaxias gracilis, G. maculatiis atteniiatHS, and G. usitatus. 1967 NEW SPECIES OF GALAXIAS 5 suggest that the meristic distinctions may not be simply a case of temperature related differences. Differences were also found to occur in body proportions. In some cases both the derivatives diverged from G. maculatus attenuatus in the same direction (Fig. 4-2, 4-5), the measured structure being proportionately either (D A B C © 8 9 10 11 12 STANDARD LENGTH / LENGTH CAUDAL PEDUNCLI A B C (D 1.7 1.8 1.9 2.0 2.1 STANDARD LENGTH / PRE-PELVIC LENGTH A B C ® 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 STANDARD LENGTH / PELVIC - ANAL LENGTH A B C (D 9 10 II 12 13 STANDARD LENGTH / LENGTH DORSAL FIN BASE A B C © 3.8 4.0 4.2 4.4 4.6 4.8 5.0 5.2 5.4 5.6 STANDARD LENGTH / HEAD LENGTH A B c 2.3 2.4 2.S 2.6 2.7 2.8 2.9 3.0 3.1 3.2 3.3 3.4 3.5 HEAD LENGTH / 1 NT E R - ORB IT A L V/IDTH Figure 4. Ranges for proportional measurements in Galaxies gracilis (A), C. maculatus attenuatus (B), and G. usitatus (C). 6 BREVIORA No. 265 larger or smaller in both derivatives than in G. maculatus attenua- tus. In other cases, the divergence has gone in opposite directions, and the range for G. maculatus attenuatus was found to lie be- tween the ranges for the derivative populations (Figs. 4-1, 4-4). These data indicate that each of the two populations should be regarded as a full species, distinct from G. maculatus attenuatus and from each other. SYSTEMATIC DESCRIPTION Galaxias gracilis^ n.sp. Figure 5; Table la. Holotype: Deposited in collection of Fisheries Research Di- vision, New Zealand Marine Department. Paratypes: Dominion Museum, New Zealand. Reg. No. 4499; Museum of Comparative Zoology, Harvard University, No. 45053; Smithsonian Institution Fish Collection No. USNM 201224. Type locality: Upper Lake Rototuna, Kaipara Harbour, North Auckland, New Zealand. Figure 5. Galaxias gracilis n.sp. Length to caudal fork, 59 mm. Description: A small, slender, shoaling species with long head, low vertebral count and rearward placement of the pelvic fins. Maximum known fork length 62.5 mm. D i-iii, 7-10; C 16; A i-iii, 12-16; Pel 6-8; Pec 12-14; Branchiostegals 4-6; Vertebrae 47-50; Gill-rakers 5-7 (upper), 13-17 (lower limb of first arch), total 18-23. Trunk cylindrical, slender, somewhat depressed on head, later- ally compressed on caudal peduncle, deeper than broad. Caudal peduncle slender and short. Head small, slender, a little broader than deep, lower jaw protruding a little, or equal in length to upper; lips thin, cleft of mouth slightly oblique, extending to a 1 Named on account of its slender form, from the Latin gracilis, slender. 1967 NEW SPECIES OF GALAXIAS 7 line through anterior eye margin. Eye large, at upper head pro- file, with inter-orbital flat. Mandibular and premaxillary teeth without canines, lingual teeth strong, entopterygoidal teeth moderately developed; gill-rak- ers strongly developed; pyloric caeca lacking. Fins membranous and short, except anal which is long based. Distal margin of dorsal fin rounded to straight, anterior rays long- est. Anal origin more or less below dorsal origin, distal margin of fin straight, anterior rays longest, maximum fin length little longer than base length. Pelvic fins inserted relatively far back, pelvic-anal interval short, fin relatively short in pelvic-anal length. Pectoral fin disposed laterally, fin short. Caudal fin forked, fin tips rounded, depth about equal to body depth. Coloration: Trunk covered with a dense peppering of large melanophores, which intensify on the head and dorsum of the trunk, failing lateroventrally, and ventrally. Fresh material is silver in these latter areas. Meristic variation: Dorsal rays (segmented), 7(1), 8(21), 9(23), 10(5); Caudal 15(4), 16(42), 17(4); Anal 12(1), 13(5), 14(30), 15(13), 16(1); Pelvic 6(7), 7(40), 8(3); Pec- toral 12(8), 13(28), 14(14); Vertebrae 47(1), 48(16), 49(24), 50(9); Branchiostegals 4(2), 5(27), 6(21); Gill-rakers (total count) 18(2), 19(3), 20(13), 21(7), 22(3), 23(2). Differs from G. maculatns attennatus (Jenyns) in having a much lower vertebral count, fewer dorsal rays, a somewhat lower bran- chiostegal count and more gill rakers; also in a longer pre-pelvic length, longer head, narrower inter-orbital, larger eye, short dorsal fin base, smaller maximum size reached, and the lack of a marine whitebait juvenile stage. Differs from G. usitatiis n.sp. in having fewer vertebrae, fewer branchiostegals, a greater number of gill rakers, a longer pelvic base-anal origin length, shorter dorsal fin base, lack of serrations on the opercular membrane and the smaller maximum size attained. Galaxias usitatus^ n.sp. Figure 6; Table lb. Holotype: Deposited in collection of Fisheries Research Divi- sion, New Zealand Marine Department. 1 Named on account of its lack of highly distinctive characters, from the Latin iisitatiis, ordinary. 8 BREVIORA No. 265 Paratypes: Dominion Museum, New Zealand Reg. No. 4500; Museum of Comparative Zoology, Harvard University, No. 45054; Smithsonian Institution Fish Collection No. USNM 201223. Figure 6. Galaxias iisitatus n.sp. Length to caudal fork, 72 mm. Type locality: Lake Waiparera, North Auckland, New Zealand. Description: A slender, lacustrine, shoaling species, similar to Galaxias maculatus attenuatus, with long head, a short pelvic-anal interval, with many of the fishes examined having six rather than the more usual seven pelvic fin rays. Distinctive in the presence of fine but definite serrations on the free margin of the opercular membrane. Maximum known fork length 81.5 mm. D i-iii, 9-12; C 14-18; A ii-v, 12-16; Pel 5-7; Pec 11-15; Branchiostegals 5-7; Vertebrae 54-59; Gill-rakers 4-5, 11-13, total 15-17. Trunk cylindrical, slender, somewhat depressed on head, lat- erally compressed on caudal peduncle and somewhat deeper than broad. Caudal peduncle slender and short. Head long and nar- row, slender, about as deep as broad; jaws equal, prominent, cleft slightly oblique reaching to about anterior eye margin, gape rather narrow. Eye large, placed close to upper head profile, inter-orbital more or less flat, broad relative to head width. Mandibular and maxillary teeth without canines, lingual teeth strong, entopterygoi- dal teeth well developed; pyloric caeca lacking; gill rakers well developed; free margin of operculum finely serrate. Fins membranous and short, except anal which is long based. Dorsal fin set well back, fin with moderately short base, distal margin of fin somewhat rounded. Anal fin origin more or less below dorsal origin; anal fin long based but greatest fin length little greater than base length, distal margin of fin about straight and inclined to body axis. Pelvic fin inserted at about mid-point of standard length, pelvic-anal length short, pelvic fin moderate in pelvic-anal interval, thus short. Pectoral fin base disposed laterally, fin short. Caudal fin forked, fin tips rounded, fin depth about equal to body depth. 1967 NEW SPECIES OF GALAXIAS 9 Coloration: Trunk a dusky gray-brown and covered with irregu- lar dark blotches similar to G. maculatus attenuatus. When alive, the lateroventral and ventral abdomen is silvery, in preserved ma- terial, largely unpigmented. Meristic variation: Dorsal rays (segmented) 9(4), 10(28), 11(8), 12(l);Caudal 14(1), 15(4), 16(34), 17(1), 18(1); Anal 12(2), 13(10), 14(13), 15(15), 16(1), Pelvic 6(30), 7(11); Pectoral 11(1), 12(5), 13(21), 14(12), 15(2); Vertebrae 54(2), 55(5), 56(10), 57(12), 58(9), 59(3); Branchiostegals 5(1), 6(25), 7(15); Gill-rakers 15(2), 16(11), 17(7). Differs from G. maculatus attenuatus in having a lower vertebral count, fewer pelvic fin rays, longer head, somewhat larger eye, shorter pelvice-anal interval; differs from G. gracilis in details dis- cussed under the diagnosis of that species. It differs from both G. maculatus attenuatus and G. gracilis by the presence of serra- tions on the opercular margin. DISCUSSION The two new species described above are interpreted as land- locked derivatives of one of the New Zealand sea-going species, G. maculatus attenuatus. In Lake Okataina, thermal lakes district. North Island, and in one of the Kaihoka Lakes, near Cape Fare- well, there are two land-locked populations of the sea-going G. fasciatus Gray. The upland-alpine lake species G. koaro Phillipps and G. lynx Hutton are also envisaged, by the writer, as fresh- water derivatives of G. brevipinnis GUnther, a species with marine stages. It was suggested (McDowall, 1965). that Retropinna ab- breviata McDowall and R. lacustris Stokell had speciated under similar circumstances — in isolation, in lakes cut off from the sea with a migratory population captured in the impounded water body. Looked at from this perspective, the distribution patterns of some of the apparently younger species in New Zealand's fresh- water fish fauna make greater sense, e.g. G. lynx, present in alpine lakes all along the eastern side of the main divide of the South Island, to the west of the divide in the Nelson Lakes district, with, according to the arrangement of Stokell (1949) a highly disjunct population in Lake Waikaremoana, near East Cape in the North Island. This pattern appears best explained by derivation of the freshwater lacustrine form, probably several times, from the par- ent species G. brevipinnis. G. koaro probably has the same mode of derivation, from G. brevipinnis, one or several times, in the lakes of the volcanic plateau of the North Island, and further study may well suggest that G. lynx and G. koaro are conspecific. 10 BREVIORA No. 265 This view of speciation in the Galaxiidae probably suppHes the key to understanding much of the fauna, both phylogenetically and zoogeographically. ACKNOWLEDGEMENT The writer is grateful to Dr. Giles W. Mead for his advice and for reading the manuscript, and to Messrs. K. F. Maynard and W. Skrzynski for assistance in collecting the specimens. The finan- cial support of a New Zealand National Research fellowship dur- ing the study is gratefully acknowledged. LITERATURE CITED Cunningham, B. T., et al. 1953. A survey of the western coastal dune lakes of the North Island, New Zealand. Aust. J. Mar. Freshw. Res. 4 (2): 343-386, 2 pis. McDowALL, R. M. 1965. A new species of Retropinna from Lake Omapere, North Auck- land. Rec. Dom. Mus. N.Z. 5 (13): 89-91, 1 fig. 1966. Further observations on Galaxias whitebait and their relation to the distribution of the Galaxiidae. Tuatara 14(1): 12-18, 2 figs. Pollard, D. A. 1964. Studies on the reproductive cycle of Galaxias attemiatus. Aust. Soc. Limn. Newslett. 3(1): 10. Stokell, G. 1949. The systematic arrangement of the New Zealand Galaxiidae. IL Specific classification. Trans. Roy. Soc. N.Z. 77 (4): 472-496, 10 figs. 1964. A new species of Galaxias from Victoria, Australia. Rec. Dom. Mus. N.Z. 5 (6): 45-48, 1 fig. (Received 10 October, 1966.) 1967 NEW SPECIES OF GALAXIAS 11 TABLE 1. Morphometric variation in Galaxias gracilis and G. usitatus. Ratio a. G. gracilis b. G. usitatus Standard length/Body depth at vent Standard length/Caudal peduncle length Caudal peduncle length/Peduncle depth Pre-dorsal length/Standard length Standard length/Dorsal fin base length Dorsal fin length/Fin base length Standard length/Anal fin base length Anal fin length/Fin base length Pre-pelvic length/Standard length Pectoral-pelvic length/Standard length Pectoral fin length/Pectoral-pelvic length Pelvic-anal length/Standard length Pelvic fin length/Pelvic anal length Standard length/Head length Head length/Head width Head length/Inter-orbital width Head length/Eye diameter Head length/Gape width 8.25-9.50 7.70-10.08 7.46-9.65 9.14-11.60 1.54-2.08 1.31-1.69 0.73-0.78 0.74-0.79 10.90-13.86 8.00-11.78 1.38-1.71 1.29-1.64 6.44-7.68 6.02-8.20 1.11-1.33 1.15-1.33 0.53-0.56 0.49-0.56 0.28-0.33 0.28-0.33 0.32-0.42 0.36-0.43 0.20-0.24 0.22-0.27 0.41-0.50 0.36-0.43 3.92-4.39 4.06-4.54 1.87-2.14 1.87-2.29 3.01-3.50 2.67-3.25 3.67-4.50 3.72-4.58 3.29-4.00 2.90-3.80 BREVIORA Mmseimm of Comparsitive Zoology Cambridge, Mass. 6 April, 1967 Number 266 A NEW SPECIES OF VESICOMYA FROM THE CARIBBEAN SEA (MOLLUSCA: BIVALVIA: VESICOMYIDAE) By Kenneth Jay Boss Museum of Comparative Zoology Harvard University During Cruise 104 (fail 1965) of the M /V OREGON of the Bureau of Commercial Fisheries Pascagoula Fishery Station. Pas- cagoula, Mississippi, an unusually large bivalve was trawled, alive, from a depth of over two hundred fathoms (366 m) off the coast of Colombia in the Caribbean Sea. Upon examination and com- parison with material in the United States National Museum, the specimen was found to be referable to the rare and enigmatic family Vesicomyidae. The unique specimen is not conspecific with any known species of this family and is herein described as new. Vesicomya (Callogonia) caribbea sp. nov. Figures 1-6 Holotype: Division of Mollusks, U. S. National Museum, cata- log number 674573. ^Type locality: M/V OREGON Station 5692, 12 = 3 1'N; 70 58'W; 205 fathoms (375 m); Caribbean Sea, off Cabo la Vela, Peninsula de Guajira, Colombia; bottom temperature 55T (12.8°C); 10 October 1965. The specimen was collected alive by a 40' shrimp trawl. Description: Shell 96 mm long, 67 mm high and 54 mm wide, irregularly elliptical, inequilateral, equivalve, solid, frangible, in- flated with both valves of equal convexity, without posterior gape, but with slight undulate flexure to right posteriorly. Umbos an- terior, conspicuous, involute, pointed, prosocline, slightly elevated 2 BREVIORA No. 266 and inflated; umbonal cavity deep. Anterior margin not extensive, very broadly rounded and weakly convex; ventral margin ex- tensive, weakly convex, rising gently posteriorly; anterior dorsal margin short, semisigmoid, concave umbonally, rounded and con- vex distally; posterior dorsal margin convexly arcuate and coexten- sive with posterior margin, forming broad blunt, oblique trunca- tion; outline weakly pointed posteroventrally. Sculpture closely set, fine concentric laminae, grouped between larger, more conspic- uous, irregular growth rings; radial sculpture not evident. Posterior ridge, semisulcated dorsally, radiating from umbo toward postero- ventral region, becoming weaker distally. Ligament strong, pro- tuberant, black, elongate, lanceolate, slightly inset basally, sup- ported by thickened hinge line, and forming broad elongate nymph. Lunule not developed or circumscribed by impressed isocardio- form line. Hinge dentition invested with periostracum. Right valve with two cardinal teeth beneath umbo; ventral tooth stronger and protuberant; dorsal tooth with anterior laminar portion and posterior thickened portion. Left valve with single central cardinal tooth developed, strong and protuberant; anterior laminate portion broken and obsolete; secondary laminate tooth beneath umbo broken. No true internal radial rib, but rib-like thickening present on posterodorsal margin of anterior adductor muscle scar. Muscle scars moderately impressed, large, irregularly ovate, anterior smaller than posterior. Pallial line weakly convex; pallial sinus irregular, narrow; pallial line not directly or obviously contiguous with posterior adductor muscle scar. Internal surface of valve with radial vermiculations. Periostracum dehiscent, olivaceous, wrinkled, heavier peripherally, also thickened over cardinal and anterior portions of hinge internally. Shell basically white, ex- ternally grayish to olivaceous, discolored centrally with russet band; earthy and chalky in texture; internally shining, with dark- ened blotches and pale salmon centrally. Remarks: The family Vesicomyidae (originally spelled Vesi- comyacidae) was established by Dall and Simpson (1901). It has been reviewed by Lamy ( 1920), and recently Odhner (1960) has listed the species referable to Vesicomya. The status of the family has been questioned: Thiele (1935). Odhner (1960) and Clarke ( 1962) included Vesicomya in the Kellyellidae. Jukes- Browne (1913) even mentioned it in relation to the Veneridae. On the other hand, Dall (1908), Thiele and Jaeckel (1931), Prashad (1932), Taki and Habe (1955), and Newell (1965) have maintained the Vesicomyidae separately. 1967 NEW SPECIES OF VESICOMYA 3 There is also a question as to what genera should be included in the Vesicomyidae. For example, Woodring (1938) placed Calyptogena Dall. along with his newly described genera Ectena- gena and Phreagena, in the Vesicomyidae. An historical survey of the family and a critical revision of at least the western Atlantic species is currently being prepared. No western Atlantic Vesicomya approaches the size of V . carib- bea. V. leeana (DalK 1889, p. 440; 1890, pi. 10, figs. 6-9) is the only western Atlantic species that is closely related to and which might be confused with V. caribbea. The latter is easily distinguished by being much larger and having a heavier shell with a distinct posterior ridge. In addition, the hinged teeth of V. carib- bea are stronger and more blunt. The eastern Pacific species. V. gigas (Dall, r896, pi. 18; 1908, pi. 16, fig. 9), attains the size of V. caribbea but it is usually less heavily shelled and its teeth are more distinct. It may be diflferentiated from V. caribbea by its much shorter ligament and shorter, more distinct, nymphal callos- ity. Further, V. gigas lacks a posterior ridge and is generally of longer, narrower proportions. Also, V. gigas is mesially constricted ventrally while V. caribbea is not. The two species most closely related to V. caribbea are known only from widely separated localities and are also rare in collec- tions. The most closely related is Vesico)n\a cliuni described by Thiele and Jaeckel (1931, p. 244, pi. 4. fig. 100) from VAL- DIVIA Station 63, 2 N; 8'4.3'E, from a depth of 2492 m in the Gulf of Guinea, eastern Atlantic Ocean. Unfortunately, the au- thors did not figure an internal view of the shell and no specimens of this species were available for comparison. The outline of the valves of V . caribbea and V . cliuni are quite similar; however, the semisulcated posterior ridge of V. caribbea appears to be a dis- tinguishing character. Further, according to Thiele and JaeckeFs description, V. chuni is relatively thin shelled, whereas V. caribbea is thick shelled. The other species closely resembling V. caribbea is from the East Indian area. Vesicomya winckwortJii, described by Prashad (1932, p. 153, pi. 5, figs. 7-8) from SIBOGA Station 17, 7°28.5'S; 1 15'^28'E, in 1060 m depth, north of Java, is founded on a single right valve which, fortunately, was well illustrated externally and internally. In contrast to V. winckworthi, V . caribbea is apparently more inflated, its anterior margin more narrowly rounded, and its nymph stronger and thicker. Vesicomya caribbea is placed in the subgenus Callogonia Dall 1889 (type-species, by monotypy, Callocardia [Vesicomya] leeana 4 BREVIORA No. 266 Dall 1889), as its pallial sinus is, though indistinct, somewhat angular. Thiele and Jaeckel (1931) placed V. chuni, the species most closely related to V. caribbea, in Callogonia. ACKNOWLEDGMENTS Mr. Harvey BuUis, Jr., Director of the Bureau of Commercial Fisheries Pascagoula Fishery Station made it possible for me to participate in Cruise 104 of the M/V OREGON and gener- ously transferred specimens into my care. Drs. J. Rosewater of the Division of Mollusks of the U. S. National Museum, D. Cohen of the Bureau of Commercial Fisheries Ichthyological Labora- tory, and W. J. Clench and R. D. Turner of the Museum of Com- parative Zoology, Harvard University, critically read the manu- script. The drawings were inked by Mrs. Mildred Carrington. Dur- ing the preparation of this report the author was on the staff of the Ichthyological Laboratory, Bureau of Commercial Fisheries, Washington, D. C. LITERATURE CITED Clarke, A. H., Jr. 1962. Annotated list and bibliography of the abyssal marine mollusks of the world. Bull. Nat. Mus. Canada, No. 181, 114 pp. Dall, W. H. 1889. Report on the Mollusca [BLAKE]. Part II. Gastropoda and Scaphopoda. Bull. Mus. Comp. Zool., 18: 1-492, pis. 1-40. 1890. Preliminary report on the collection of Mollusca and Brachio- poda obtained in 1887-'88 [ALBATROSS]. Proc. U.S. Natl. Mus., 12(773): 219-362, pis. 5-14. 1896. Diagnoses of new species of mollusks from the west coast of America. Proc. U.S. Natl. Mus., 18(1034): 7-20. 1908 The Mollusca and Brachiopoda [ALBATROSS]. Bull. Mus. Comp. Zool.. 43(6): 205-487. pis. 1-22. Dall. W. H. and C. T. Simpson 1901. The Mollusca of Porto Rico. Bull. U.S. Fish Comm., 20(1): 351-524. pis. 53-58. Jukes-Browne, A. J. 1913. On Collistu, Amiuntis, and Pitaria. Proc. Malac. Soc. London, 10: 335-347. Lamy, E. 1920. Revision des Cypricardiacea et des Isocardiacea vivants. Jour. de Conch., 64: 259-307. 1967 NEW SPECIES OF VESICOMYA 5 Newell, N. D. 1965. Classification of the Bivalvia. Amer. Mus. Novit., No. 2206, 25 pp., 3 text figs. Odhner, N. H. 1960. Mollusca. Kept. Swed. Deep-Sea Exped., 2(Zool. 22): 365-400, 2 pis., 12 text figs. Prashad, B. 1932. The LamelHbranchia of the SIBOGA Expedition. Systematic Part, Pelecypoda 11 (exclusive of the Pectinidae). SIBOGA- Expeditie, Monographie 53 c, 353 pp., 9 pis. Taki, I., and T. Habe 1955. [In] Kuroda, T., Vesicomyidae. Illus. Cat. Jap. Shells, 2(1): 6, text-figs. 1-4. Thiele, J. 1935. Handbuch der systematischen Weichtierkunde. Jena, Gustav Fischer, vol. 2, pp. 779-1022, 110 figs. Thiele. J., and S. Jaeckel 1931. Muscheln der deutschen Tiefsee-Expedition. Wiss Ergeb. Deutsch. Tiefsee Exped., 21(1): 161-268 (1-110), pis. 6-10 (1-5). WOODRING, W. P. 1938. Lower Pliocene mollusks and echinoids from the Los Angeles Basin, California. U.S. Dept. Interior, Geol. Surv. Prof. Papers, 190, 67 pp., 9 pis. (Received 20 October, 1966.) BREVIORA No. 266 VesicoDiMi iCallogonia) carihbea n.sp.. Holotype. USNM 674573. Fig. 1. Internal view of left valve. Fig. 2. Internal view of right valve. Fig 3. External view of left valve. Fig. 4. External view of right valve. Fig. 5. Dorsal view of valves, showing posterior ridge. Fig. 6. Anterior view, showing involute umbos. BREVIORA MiiseiLiim of Com para tive Zoology Cambridge, Mass. 29 June 1967 Number 267 A NEW GENERIC ARRANGEAAENT FOR SOME AUSTRALIAN SCINCID LIZARDS By Allen E. Greer In Boulenger's (1887, vol. 3) classification of skinks, the genus Lygosoma was by far the largest and most heterogeneous group that defied further spHtting on the basis of the characters used. Unable to recognize "natural" distinctions which would war- rant full generic rank, Boulenger designated 1 1 "natural" species groups of Lygosoma as subgenera and thereby established a rank- ing system that then, as now, was little used in lizard taxonomy. The relationships of one of Boulenger's Lygosoma subgenera, the subgenus Rhodona, and part of the genus Ablepharus are re- examined in the present paper. Boulenger's (1887:223) subgenus Rhodona was based on an unusually large suite of external characters, being defined as fol- lows: "Limbs short or rudimentary. Lower eyelid with a trans- parent disk. Ear distinct, minute. No supranasals. Praefrontals very small and widely separated, or absent. Frontal broader than the supraocular region." To this list could have been added the following: nasals enlarged so that they just fail to touch or do touch medially, 2-5 pairs of nuchals, and a single pair of enlarged preanals. In addition, it might have been noted that the group is found only in Australia and Tasmania. The widespread Old World genus Ablepharus, on the other hand, was based essentially on a single character, Boulenger (1887: 344) including in this genus all those skinks lacking "movable eyelids" and with a "transparent disk covering the eye." Obviously, Boulenger in this instance put great confidence in a single character, the "ablepharine" eye. This single character approach should not be thought of as being too incongruous with Boulenger's general philosophy. In most other reptile groups with the "ablepharine" eye (non-eublepharine geckos, pygopodids. xantusids, and snakes), the character was 2 BREVIORA No. 267 then and may still be rightly regarded as the hallmark of natural taxa. The "ablepharine" skinks, however, are an exception. Boulenger's basic conception of Rhodona and Ablepharus was accepted by the authors of the two subsequent revisions of the 'lygosomine" skinks (Smith, 1937; Mittleman, 1952). Malcolm Smith (1937) gave Rhodona full generic rank and included within it three southeast Asian (aiiguinoides, roulei, and gyldenstolpei) and one Australian (australis) species that in 1935 he had con- sidered the sole representatives of the genus Ophioscincus} This arrangement, however, was suspect on two counts. First, the species of Ophioscincus differ from those of Rhodona in lacking both fore and hind limbs and external ear openings, and in having no or only one pair of nuchals, and second, such an arrangement involves an improbable geographic distribution. All skinks with an immovable transparent disc covering the eye were retained in Ablepharus by Smith ( 1937). In this he followed Boulenger, although he noted in 1935 that the group was almost certainly polyphyletic, probably having arisen from Leiolopisma and Emoia stocks. Mittleman (1952), taking his cue from Smith's reasonable sup- position of the polyphyletic nature of Ablepharus, cut the Gordian knot by separating out all Ablepharus that had the frontoparietals and interparietal fused into a single scale and placing them in the genus Cryptoblepharus. It is interesting to note that the "able- pharine" eye was still sacrosanct; no "ablepharine" skink was lumped with a non-"ablepharine" skink. Similarly, Mittleman split Boulenger's Rhodona into two genera: Rhodona with the frontoparietals and interparietal fused into a single scale (indicated as 1 in Table 2), and N odor ha in which the frontoparietals are distinct ( 2/ 1 ) or fused ( 1 / 1 ) but always distinct from the interparietal. Mittleman removed the four species of Ophioscincus from the Rhodona-Nodorha complex and was certainly right in doing so. However, in reverting to Smith's 1935 treatment of the genus Ophioscincus, he has re-emphasized the anomalous geographic distribution of that genus. The disjunct Australian, southeast Asian distribution, once again, makes the "naturalness" of the genus suspect. Although Mittleman's arrangement provides groupings which are of value in identifying and cataloging skinks, it fails to demonstrate 1 In 1887 the only known species of Ophioscincus Peters, 1873 was australis, and this was put in the subgenus Lygosoma by Boulenger. 1967 AUSTRALIAN SKINKS: LERISTA 3 any greater awareness for the evolution of higher taxa than did earUer pre-BouIengerian classifications. This has been an a priori criticism of Mittleman's lygosomine classification since its publi- cation, and perhaps explains why his system has been passed over in favor of a Boulenger-Smith arrangement by most present day authors. Even these two earlier authors must have been aware of many of Mittleman's "generic characters" but chose not to use them. In my own research on the higher taxa of the skinks I have attempted to correlate characters of external morphology and skull morphology. It is my contention that when a suite of external and skull characters can be found in only a certain group of species, there is strong evidence that the group is "natural," i.e., mono- phyletic. The geographic distribution of the taxon is used as a further check of the validity of the arrangement. If the distribution of the taxon is consonant with the general geographic barriers in- fluencing the distribtuions of other taxa, then the argument for the validity of the new taxon is strengthened. A consideration of the external morphology has, of course, been the standard approach in skink taxonomy, although, like Boulenger (1887:131), I believe I have found new characters which are important. Little has been done with skull morphology in relation to skink classification. Boulenger examined the relationships of the bones of the palate as seen through the epidermis of the mouth and utilized what he saw at the generic level. This method led to some serious misinterpretations of the anatomical facts and these have been perpetuated (see Smith, 1935, and Mittleman, 1952, for example) until the present time. By removing the epidermis over- lying the palate, Waite (1929) and Mitchell (1950) made a tech- nical advance, but failed to take full advantage of it in their own taxonomic work. My own approach (outlined above) has provided evidence which convinces me of the previously only suspected polyphyletic nature of many groupings in all three earlier classifications (Boulenger, 1887; Smith, 1937; and Mittleman, 1952). Many of the single characters which were utilized exclusively or heavily in diagnosing skink taxa simply do not correlate with any other characters (inter- nal or external), as one would expect them to do in a natural group. Other correlations utilizing skull and external characters are possible, however, and will form the basis of a new classification of skinks. This brief study delimiting a newly recognized taxon and giving it generic rank is an example of the new approach and 4 BREVIORA No. 267 is a first contribution toward the distant goal of a complete generic revision of the skinks. The taxon to be discussed here consists of all the species of Rhodona (sensu Boulenger) and those species of Australian Able- pliarus that conform to the descriptions of Rhodona except for the "ablepharine" eye (Table 2). The earliest named species of the group {Ablephanis liueata) was described by Bell in 1833 (see below) as the type species of Lehsta. The new taxon should therefore be properly known as Lerista, with the following primary generic synonymy. Lerista Bell Lerista Bell, 1883, Proc. Zcol. Soc. London, p. 99 (Type species. Lerista lineata Bell, 1833, by monotypy) Rhodona Gray, 1839, Ann. Mag. Nat. Hist., vol. 2, p. 335 (Type species, Rhodona punctata Gray, 1839, by monotypy). Soridia Gray, 1839, Ann. Mag. Nat. Hist., vol. 2, p. 335 (Type species, Soridia lineata Gray, 1839, by monotypy). Brachystopus Dumeril and Bibron, 1839, Erpetologie generale, vol. 5, p. 778 (Type species, Brachystopus lineopunctulotus Dumeril and Bibron, 1839 = Rhodona punctata Gray, 1839, by monotypy). Praepeditus Dumeril and Bibron, 1839, Erpetologie generale, vol. 5, p. 787 (Type species, Soridia lineata Gray, 1839, by monotypy). Ronia Gray, 1841, Ann. Mag. Nat. Hist., vol. 7, p. 87 (Type species, Ronia catenulata Gray, 1841 = Rhodona punctata Gray, 1839, by monotypy). Leptosoma Fitzinger, 1843, System reptilium, p. 23 (Type species, Riopa bougainviUii Dumeril and Bibron, 1839, by monotypy). Ophiopsis Fitzinger, 1843, Systema reptilium, p. 23 (Type species, Lerista lineata Bell, 1833, by monotypy). Miculia Gray, 1845, Catalogue of the specimens of lizards in the collection of the British Museum, p. 66 (Type species, Miculia elegans Gray, 1845, by monotypy). Pholeophilus A. Smith, 1849, Illustrations of the zoology of South Africa, p. 15 of Appendix (Type species, Pholeophilus capensis Smith, 1849 = Soridia lineata Gray. 1839, by monotypy). Phaneropus Fischer, 1881, Arch. Naturg. Jhrg. 47, p. 236 (Type species, Phaneropus inuelleri Fischer, 1881, by monotypy). Nodorha Mittleman, 1952, Smithsonian Misc. Coll., vol. 17, p. 14 (Type species, Riopa bougainviUii Dumeril and Bibron, 1839, by original designation). DIAGNOSIS OF THE GENUS LERISTA There is no single diagnostic character that will distinguish the species of Lerista from all other skink taxa. This is, however, not 1967 AUSTRALIAN SKINKS: LERISTA 5 to say that once one is familiar with the diversity of skinks that a species of Lerista cannot be identified as such at a glance. On the basis of external characters, the taxon may be diagnosed as follows: no supranasals; nasals enlarged so that they just fail to touch or do touch medially; prefrontals reduced and widely sepa- rated or absent; frontal generally broader than supraocular region; frontoparietal/interparietal configuration = 2/1, 1/1, or 1; 2-5 pairs of nuchals; lower eyelid with a transparent disc and either movable or fused to form a permanent spectacle; no row of scales separating the small scales of the eyelid from the subocular supra- labial; ear opening small to minute, but always present; a single pair of enlarged preanals; body form attenuate; limbs diminutive or rudimentary, but never totally lacking (digital formula from 5-5 to 0-1). DESCRIPTION AND RELATIONSHIPS OF THE GENUS lERiSJA The following description of the taxon Lerista is based on an examination of whole skulls, alcoholic specimens (see Materials section of paper), and also of type descriptions when species have not been available. The skull is considered first, as skull morphology has generally proved more useful in determining the broader relationships of skinks than has external morphology. Skull morphology: The skull characters of the species are re- markably uniform, although the species adapted to a more burrow- ing way of life (e.g., L. bipes and L. praepeditu) show certain modifications in the skull that are often seen in burrowing skinks. To facilitate comparisons with its near relatives, a brief descrip- tion of the skull characters of the genus is given. The whole skull tends to be somewhat conical in shape. The palatines are approximately 2-2 '/a times as long as broad and meet along their entire medial edge. The palatal rami of the pterygoids are "squared-off'" and in some species slightly emarginate posteri- orly, and they also meet along their medial edge (Fig. 1). The palatines and palatal rami of the pterygoids thus form an extensive "secondary palate." Among lizards the tendency toward the forma- tion of a bony secondary palate is unique to skinks, and the extensive palatine-pterygoid palate of Lerista is, in turn, a rather "advanced" development of the palate in skinks. An ectopterygoid process extends forward along the lateral edge of the palatal ramus of the pterygoid, past the palatine-pterygoid suture and onto the posterolateral edge of the palatine, thereby BREVIORA No. 267 Figure 1. Ventral view of the palate of A, Lerista elegans (FMNH 11319), Lake Toolburnup (= Tool Brumup), Western Australia; B, Spheno- morphus pardalis (MCZ 35413), Coen, Cape York, Queensland; C, Lerista boiigainvilli (MCZ 61403), Tallarock, Victoria; D, "Ahlepliarus" lineo- ocellotiis (MCZ 33144), Margaret River, Western Australia. Abbreviations for this figure: e, ectopterygoid; p, palatine: pt, pterygoid, A, C, and D drawn to scale. 1967 AUSTRALIAN SKINKS: LERISTA 7 excluding the pterygoid from a position on the infraorbital fenestra (Fig. 1). This anteriorly projecting ectopterygoid process is also unique to skinks, among lizards, although it is not found in all skinks. The premaxillae and nasals are paired; the vomer, frontal, and parietal are single. There is a parietal foramen in the parietal. The frontal and maxilla form a surface suture or the frontal, pre- fontal, maxilla, and nasal meet at a point. The temporal arch and the postorbital bar (formed by the jugal) are present. A small supratemporal fenestra may be pres- ent but is more frequently covered by the posterior extension of the postfrontal. A postorbital is lacking. The bones of the skull and mandible generally lack pigment. Teeth are present on the maxillae, premaxillae, and dentaries. There are usually 7 (5 in L. bipes and 6 in L. lineopimctulata) teeth on the premaxillae and 9-13 (usually 11-12) on each maxilla. Meckel's groove may be present or partially overlapped by the dentary. Relationships as inferred from skull morphology. On the basis of the relationships of the bones of the palate, Lerista is most similar to Hemiergis and one group of Sphenomorphus species (Fig. 1 ) . For convenience, this latter group may be referred to as the solomonis group; it encompasses so far as is known the following species: solomonis, emi grans, pardalis, australe, antimorum, crassicauda, tenuis, and tryoni. The species of all three groups, that is Lerista, Hemiergis, and the solomonis group, have the palatines and the pterygoids meeting along the midline, and, except for three species of Hemiergis and S. australe, the species of all three groups possess an ectopterygoid process which extends forward to the palatine to exclude the pterygoid from the infraorbital vacuity. In Hemiergis peroni, H. tridactylus, and H. initiale the anteriorly projecting ectopterygoid process has been secondarily reduced to a small nubbin which fails to exclude the palatal ramus of the pterygoid from a position on the infraorbital vacuity. The process is better developed in 5. australe where it extends anteriorly about half way along that outer edge of the palatal ramus that would, in the absence of the process, border the infraorbital fenestra. The species of the genus Hemiergis and the solomonis group, however, differ from the Lerista skinks in possessing more teeth on the premaxillae (8-9) and in having a long thin postorbital.^ 1 The number of teeth on the premaxillae and the absence or presence and shape of the postorbital are of greater taxonomic significance in skinks than in most other lizard families. 8 BREVIORA No. 267 The relationship of the bones of the palate described above for Lerista and its near relatives is not found in any Ablepharus other than in the "ablepharine" Lerista. In all other Ablepharus the palatal rami of the pterygoids are separated (contrary to Boulen- ger's 1887 diagnosis of the genus), and there is no ectopterygoid process (Fig. 1). Furthermore, in all Ablepharus other than the "ablepharine" Lerista there are usually 9 (or more) teeth on the premaxillae, and the prefrontal often forms a broad suture with the nasal. External morphology: In gross aspect the species of Lerista are attenuate skinks in which the tail comprises fully half the total length. The appendages are diminutive and widely separated when adpressed to the body. The head tends to be somewhat conical to wedge shaped. Upon closer inspection it can be seen that the nasals are en- larged in a dorsoventral aspect so that they meet mid-dorsally behind the rostral or just fail to do so. There are no supranasals and the frontonasal is wider than long. The prefrontals are small and widely separated or, in some species, they are absent. The frontal tends to be wider than the supraocular region and is in contact with 1 or 2 (usually 2) of the 2 to 4 (usually 3) supraoculars. The frontoparietals and interparietal display the three possible degrees of fusion. The degree of fusion is indicated in Table 2 by a simple notation in which the number of frontoparietals is given on the left side of the slash mark (/) and the interparietal is indi- cated on the right side. The numeral one means that all three scales are fused into a single scale. It can be seen from the data in Table 2 that there is a very approximate correlation between the state of fusion of the fronto- parietals and interparietal and the reduction of the appendages and digits. Reduction of the appendages and digits is in turn correlated with burrowing habits. It would appear, therefore, that in species of more burrowing habits there is a greater tendency for the fronto- parietals and interparietal to fuse. This is just one of several evolu- tionary trends (see below) in the group and, to me, it seems pointless to break the concept of this trend by splitting the spec- trum into two parts as Mittleman (1952) did when he set up Nodorha as distinct from Rhodona. The parietals meet behind the interparietal, and there are 2-4 pairs of nuchals. There is a well defined transparent area or "window" in the lower 1967 AUSTRALIAN SKINKS: LERISTA 9 eyelid. In most species the lower eyelid is movable, but in a few it has become fused to the upper resulting in an "ablepharine" eye (Table 2). In all species of Lerista a supralabial contacts the small scales of the lower eyelid below the transparent area and thereby acts as a subocular. There are 2-4 (usually 3) supralabials anterior to the subocular (= a supralabial). The external ear opening is small, that is, it is smaller than the transparent area in the lower eyelid, but although the opening is minute in several species, it is never completely covered over by the epidermis. The scales around the midbody are smooth and number 16-24. There is a pair of very large preanals. The three midventral series of subcaudals are subequal. The digital formula (fingers-toes) ranges from 5-5 to 0-1 (Table 2). Only two species have 5-5 while all others show a reduced number of digits, although no species is limbless. The digits are covered above by a single row of scales (Group I of Brongersma, 1942). Color patterns in the genus are variations on a striped motif (Table 2). A solid, dark, dorsolateral stripe is common and often occurs in conjunction with either a third mid-dorsal dark stripe or with a pair of mid-dorsal broken dashed stripes on the two dor- salmost scale rows. A few species, on the other hand, lack broad dorsolateral or mid-dorsal stripes and simply have broken dashed stripes on each dorsal scale row. There seems to be little intra- specific variation in regard to pattern type. In addition to the osteological differences mentioned above, two features of the external morphology serve to further separate the "ablepharine" Lerista from other Ablepharus and unite them with Rhodona in a common genus. These are ( 1 ) the enlarged nasals and (2) the reduced number of digits. Of all other Ablepharus only A. greyi and A. burnetti have a reduced number of digits (4-5). These two Australian species are, however, only distantly related to Lerista and will be discussed at a later date. It might also be mentioned here that the species of Lerista differ from those attenuate, burrowing skinks of the subgenus Lygosoma which occur in Australia in having transparent eyelids and external ear openings rather than having scaly eyelids and completely cov- ered ear openings. The differences in skull morphology between Lerista, Hemiergis and the solomonis species group have been mentioned above. 10 BREVIORA No. 267 Table 1 summarizes the external differences between the three groups. Geographic distribution of Lerista and its relatives: It is signifi- cant that the genus Lerista is found only in Australia and Tasmania. If we discount, for reasons given above, any close affinity with Ablepharus and Ophioscincus, then it is clear that there are no close relatives of Lerista in either Africa or Eurasia. The closest relatives of Lerista, namely Hemiergis and the solomonis assemblage, are quite obviously a product of the Aus- tralian Region. The species of Lerista represent an exclusively Australian adaptive radiation into the "surface litter and subsur- face burrowing niche." They are certainly not the only lizards to have entered this niche — Hemiergis and some of the solomonis group species mentioned above also occupy it — but Lerista is one of the most successful groups to exploit it, if diversity and abun- dance are criteria for such success. Furthermore, it is the most specialized species of this genus that have exploited the relatively new (geologically speaking) arid areas of Australia. EVOLUTIONARY TRENDS WITHIN THE GENUS lERlSlA The species of the genus as it is understood here can be arranged according to certain structural sequences that are undoubtedly indicative of the evolutionary trends that have occurred within the group. One of these trends has been a reduction in the size of the limbs and the number of digits on the fore and hind feet (Table 2). In this regard it should be noted that the number of digits on the forefront are equal to or less than, but never greater than, the number of digits on the hindfoot. This pattern of digit reduction is common among lizards, the most noteworthy exceptions being the amphisbaenid Bipes biporus (digital formula = 5-0) and the scincid genus Opiiiomorus. In the latter genus the number of fingers is equal to or greater than, but never less than the number of toes (Anderson and Leviton, 1966). The reduction of the digits in Lerista is not unexpectedly cor- related with other structural changes indicative of greater adapta- tion to burrowing habits (Mitchell, 1958); e.g., the size of the eye is reduced somewhat, the lower jaw becomes more counter- sunk into the upper, and the ear openings become more minute but never disappear. Although there is a tendency toward smaller external ear open- ings in the species of Lerista more closely adapted to a burrowing 1967 AUSTRALIAN SKINKS: LERISTA 11 existence (e.g., L. bipes and L. praepedita), the quadrate bone undergoes no striking changes in shape from that found in most Hzards with external ear openings. The bone is excavated posteri- orly, and the tympanum attaches to the lateral rim. By way of com- parison, the five species of Hemiergis are less closely adapted to a fossorial life than many of the Lerista skinks but have developed a simple rod-shaped quadrate concomitant with the loss of the external ear opening. Several characteristic changes in the skull are found in the ex- treme burrowers of the group (L. bipes and L. praepedita) . The skull posterior to the orbits is proportionately elongated, and the supra- and post-temporal fenestrae are often completely closed over, although no elements in the arches are lost. The prefrontal also tends to be reduced in its anterior extension, and posteriorly may touch the postfrontal to exclude the frontal from the orbit. In L. bipes and L. praepedita the premaxillae project forward well beyond the level of the premaxillary teeth, a modification which is undoubtedly an advantage to these species which employ the snout as a burrowing wedge. The very approximate correlation between the loss of digits and the fusion of the frontoparietals and the interparietal (Table 2) has already been mentioned. Most of the species of Lerista that have followed these trends to the extreme are among Australia's most closely adapted arid area dwellers. In this regard they are to arid Australia what Ophiomorus and Scincus are to arid southwest Asia and North Africa. The similarities are, of course, entirely convergent, for Ophiomorus and Scincus, which are rather closely related, are fairly distantly re- lated to Lerista. MODE OF REPRODUCTION IN lEMSlA Nothing has been published concerning the breeding habits of these skinks, although from the meager data presented below, it is evident that there are both live-bearing and egg-laying species in the senus. A single Lerista microtis (MCZ 24577) with a snout-vent length of 51 mm contains three young in which scales and color pattern are evident, indicating that the young would probably be born. L. timida, on the other hand, is apparently oviparous. A gravid female (snout-vent length = 40 mm) in the Field Museum of Natural History (97739) contains two oblong, whitish, leathery shelled eggs. An untagged female (snout-vent length = 46 mm) in the MCZ collection is likewise gravid with two shelled eggs. 12 BREVIORA No. 267 Single females of L. bougainvilli (MCZ 6751; snout-vent length ■= 60 mm) and L. pimctatovittata (FMNH 911 Al; snout-vent length = 85 mm) each contain three spherical, yolky oviducal eggs. These eggs were not far enough along in their development to allow speculation as to the mode of reproduction in either species. THE SPECIES OF LERISTA Although a thorough revision of the species in the genus is sorely needed, the following table (Table 2) may be of some use to those faced with the tedious curatorial duties of identifying and organiz- ing collections. The table makes no pretense of being a summary of a generic revision and is based solely on the species as they are currently recognized in the literature. Two characters important in species diagnoses, namely the fusion or independence of the frontoparietals and the number of digits, are in the greatest need of re-evaluation on the intraspecific level. The number of digits is, in fact, already known to vary some- what within a single species. Boulenger (1887:335) recorded R. gerrardi as having either mono- or didactyl forelimbs, and King- horn (1924:180) reported on a specimen of R. pimctatovittata with the "forelimb distinctly didactyl" rather than monodactyl. Loveridge (1934:372) also found a "rudimentary stump of a sec- ond toe" in R. miopa, a species which had previously been consid- ered as having only a single toe on the hind foot. Mitchell (1955:402) studied the variation in the tarsal (carpal) and metatarsal (metacarpal) bones in nine species of Rhodona and came to the conclusion that "among specimens of the same species . . . from diflferent localities ... the loss of digits and digital bones may not be always indicative of specific variation. The genus may be found to contain polytypic species within which the number of digits varies." ACKNOWLEDGMENTS Broad revisionary studies within the Scincidae would be im- possible without the interest and cooperation of the curators of many institutions. The present study has been most generously supported through loans and encouragement from the following people and their respective institutions. Miss A. G. C. Grandison, British Museum (Natural History) (BM); Drs. R. G. Zweifel and C. M. Bogert, American Museum of Natural History (AMNH); Dr. R. F. Inger and Mr. H. Marx, Field Museum of Natural History (FMNH); Drs. A. E. Leviton 1967 AUSTRALIAN SKINKS: LERISTA 13 and Steve C. Anderson, California Academy of Sciences (CAS); Dr. C. M. Walker, University of Michigan Museum of Zoology; Dr. G. S. Myers, Division of Systematic Biology, Stanford Uni- versity. A special word of thanks goes to Dr. Ernest E. Williams of the Museum of Comparative Zoology for his enduring interest in this continuing research on the relationships of skink taxa. His generosity with the reptile collection under his charge and his indefatigable assistance in obtaining specimens from which skulls could be prepared have, in fact, made this research possible. Both Dr. Williams and my wife Carol have read and criticized the manuscript in many helpful ways. MATERIALS Complete skulls of the following specimens of Lerista have been examined: hipes (AMNH 86089); boiigainvilli (MCZ 61403); elegans (FMNH 11319; fragilis (CAS 77190; MCZ 42988); gerrardi (MCZ 33255); lineo- pimctiikita (BM 1902.7.30.5); miopa (MCZ 33260); mueUcri (MCZ 86699); planiventnile (BM 1954.1-2.21); praepcdita (MCZ 33265); punc- tatovitlaui (MCZ 79494); telrudaciyla (BM 1902.7.30.6); timida (MCZ X- 13246). In addition, the relationship of the palatine, palatal ramus of the ptery- goid, and the ectopterygoid were examined in the following specimens of Lerista: hipes (MCZ 3^3251, 35349, 35350. 79570); boiigainvilli (MCZ 67149): elegans (MCZ 33126, 33136); tetradactyla (MCZ 51893); timida (MCZ 33152, 33153). Nineteen of the 28 species of Lerista were available for examination as alcoholic specimens: allanae (2); hipes (20); hougairirilli (20); distin- guenda (1); elegans (17); fragilis (many); gerrardi (12); lineata (2); lineopunctulata (1); microtis (2); miopa (3); muelleri (18); nichollsi (1); planiventrale (2); praepcdita (6); piinctatovittata (24); stylis (1); tetra- dactyla (6); timida (20). For comparative purposes the following skulls of Lerista's near relatives were examined: Hemiergis decresiensis (MCZ 49173); H. initiate (MCZ 33210); H. peroni (MCZ 24648, 24652); H. quadrilineatum (MCZ 33210); H. tridactyliim (MCZ 24595). Sphenomorphus antimoriim (MCZ 25374); S. australe (MCZ 24568); S. crassicauda (AMNH 82606); S. emigrans (MCZ 27043); 5. pardalis (MCZ 35413); S. tenuis (MCZ 35398); S. tryoni (MCZ 35387; 35388); 5. solomonis (MCZ 72618, 72626, 72664, 72665, 77373, 77374). 14 BREVIORA No. 267 LITERATURE CITED Anderson, S. C. and A. E. Leviton 1966. A review of the genus Ophiomorns (Sauria: Scincidae), with descriptions of three new forms. Proc. Calif. Acad. Sci., 33(16) : 499-534. BOULENGER, G. A. 1887. Catalogue of the lizards in the British Museum (Natural His- tory). 2nd. ed., vol. 3. xii + 575 pp., British Museum, London. Brongersma, L. D. 1942. On the arrangement of the scales on the dorsal surface of the digits in Lygosoma and allied genera. Zoologische Mededeelin- gen, 24(1-2): 153-158. KiNGHORN, J. R. 1924. Reptiles and batrachians from south and southwest Australia. Rec. Australian Mus., 14(3): 163-183. LOVERIDGE, A. 1934. Australian reptiles in the Museum of Comparative Zoology, Cambridge, Massachusetts. Bull. Mus. Comp. Zool., 77(6): 243-383. Mitchell, F. J. 1950. The scincid genera Egernia and Tiliqiia (Lacertilia). Rec. South Australia Mus., 9: 275-308. 1955. Preliminary account of the Reptilia and Amphibia collected by the National Geographic Society — Commonwealth Govern- ment — Smithsonian Institution Expedition to Arnhem Land. Rec. South Australia Mus., 11: 373-408. 1958. Adaptive convergence in Australian reptiles. Australian Mus. Mag., 12: 314-317. MiTTLEMAN, M. B. 1952. A generic synopsis of the lizards of the subfamily Lygosominae. Smithsonian Misc. Coll., 117(17): 1-35. Smith, M. A. 1935. The fauna of British India. Reptilia and Amphibia. Vol. II — Sauria., xii + 440 pp., Taylor and Francis, London. 1937. A review of the genus Lygosoma (Scincidae: Reptilia) and its allies. Rec. Indian Mus., 39(3): 213-234. Waite, E. R. 1929. The reptiles and amphibians of South Australia. 270 pp., Harrison Weir, Govt. Printer, Adelaide. 1967 AUSTRALIAN SKINKS: LERISTA 15 /-*s c« re CO ^ E ■*~t '0 CO _S O g^ CJ CO 'n to -a 00 ri ■a Lh 0 ipper p al bo 3 0 c E Lh 0 ^^ >^ .13 i ■a 0 1-5 «J 'zj ;^ CO ^ & qJ 03 K c/3 0 "sT 4-1 S 2 t»-H , 0 to 03 CJ CO -4—' o c 1 0 d wo 0 2 O- < (a! 0 CO < x: ^ -^ •~ .2 CO 0 -3 X) 4J -4-J '3 0 i _CJ CJ to •a ■+-H 0 0 T3 C c ;cept in ipper pi labial f the e\ 13 sCr x: t- x: -' ca 0 UJ -J OQ < d CO 0 Z '5 CO '0 u Dh CO 0 1 Absent, ind sion. x: •4—1 to < a, X 4-H to x: '3 Dh CO "C to ^— ( 0 13 0 Dh CO % 1 r4 fN 'u CO CO 0 e -»-> X (U •1-H ^ 1/3 'S 0 4) 0) ^,^^ tH 0 "3 Dh 0 C/5 ' 0 "3 S Lh .'S CO u •4.^ .— "« 03 4-» 0 to o a> -a c a .2 < . d "o U E 2 S o ■H 'S t/3 Uh 'C c« « t^ D- c o O u OJ >-. C Ui O c c3 u. ' "■ £ Ph S "« -S 3 ■*-' 00 'So 03 Q 1-^ c % .2 V S •3 ^ 1>J c (4-1 ■§ o O o 03 C/5 c/) •S -£ o o GO C/5 1) ^ O f3 0) X) ■^^ ^ 2. — . c 3 < 3 O C/3 fN 1) 5 CN I 05 00 o o •a c 4J H O 3 z < ^ ^ 3 y5 — S ;* c c« > o n u "*^ ^2 ^ "" o ^ u 7X "^ S O tN I OS ON r'", 00 ^■^ .22 ^ O •^ O u^ - (U — . cfl N^X '■^^ ;^ •^^ ■^ c >.x: o c c _o •3 c o o • — ^ & i "2 « c ■ ^ * x: ry) x: c > (U >. X) ' >% ca (U x: O 2 & E re t/5 _0 CO u o +-» ^ (U g c ^ O ii o 3 ■- ^ 3 2 "« j;i u 2 C H < ii E 3 ^ t: (zi 7= o = -^ ■55 ^ « > 2 = -3 - — ■^ XU C3 ^ ^ 3 < £ o c O O fc- ° .. « O CS C <-> 13 C C3 C (U ^L/ P- -i^ t-i t-l _ f^ -^ _C3 u VI 3 < C OJ 3 O OS JS ^ "3 O (» ;h O T3 m (U C JO "S ■« c o ^ o I I CO .3 a. ■a c 03 03 Q 73 C 03 XI 15 lU +-* 03 "o o ■a CQ Oh 03 •a u u MH CO CQ o u. O "o o f en u< l^ o o O CJ T3 •a c ^ M o pa 03 CO Q Q o c ■a c 03 X5 3 o o o 2 " o « XI — o ;- X i- a Q (S T-H M T-1 «s r4 (S CM rj cs CM Pi Pi Pi fN o o\ '— .-V C/3 ^—^ vo ^-s o ^ r~- '-^ ^~^ ON 00 00 00 00 00 .2 C to rri .^ X c :3 ■X- -Si OJ x: o j C/1 > ? 03 o 3 hi. 19: 5d O ^ E T3 jj s < m ^ o c^a\ ,^v 2 " r<-, r<-i •^ u 00 =o ^ C3 +-• VI . u< ^ 3 O -a c .5 CO 3 O CO .2 lo -4-* x: 0 C/5 0 c aj en k- CO M Si CO c m -4—1 E en -a C CO X) 'S •a <4-l en 0 k- X ■? E (U o CO ^ = 00 O X ^ 3 en k. 'cii 3 e« k-c _C -*-* iH -. c ca fN i-H < 1— s a 2 1 1 1 1 ' — s -a 3 JO 3 13 'So Z3 E I fS >— ^ r^i ^- «J Q ^ ^ ■4-^ c "i _o % So •3 c Pi « p< Pi Pi Pi Pi c o o 'ii U 'o a> Oh I/) H 00 k> u ao c ^ t^ ^ 2 00 \D 00 .^ kT t— < ;^ OJ ->:- l-^ ^ "3 >^ ■*^ ^ ;^ CO ^0 ^ ^ 0 eij ^ ^ Co Ci- ON 00 c o » CQ ^ S 13 Tf ^ c <3^ -2 — 5: vu g Q ON (U Ml T3 2 o 7^ »-J 00 X S :3 1967 AUSTRALIAN SKINKS: LERISTA 19 9^ ^. P-il -S^ o- -sba-t= B^ B^ §1 :^: 1 ^ ^ ^ x:„ i § O t ^ C OB cj x: El SS >.o l^-Slcgg-d-g ""5: ""S -cZ . 8i=^?^^o^ S « r-->^'-— r-iL"]^ i-;j::c« oC.E PtD |.^S >.^c ^2= "i3.2 S| I :r^l3 P Ih 0-5-= i—iJ -pi:-^ ^^ss ^re :s;=:j: ^j gSS gS^ iSc 11^ g-s gSc g wo< ^oa z \ / \ -y^ )ly \ KENYA i \ 'lU ^"""""-^ ~ " V INDIAN OCEAN L ^-^ ' TANZANIA "^^ / 5°S- 35° E \^ Ao^E 45° E 1 Figure 1. Map showing the distribution of Lxi^iuUictxhis pictiinitiis keniensis (area bounded by the dashed hne), and locahties outhning the distribution of L. sonutUcits (black dots). The westernmost locaHty for L. somaliciis corresponds to the study area where the data for the paper were collected. L. piitiiratiis (subspp.) occurs throughout the area shown in the map. except on the Arabian Peninsula. The distribution of L. p. keniensis is adapted from Pasteur (1964:77, fig. 20). The abbreviation F.S. stands for French Somaliland. 1 I follow Loveridge's (1947) and Pasteur's (1960) taxonomic treat- ment of the genus. By simply calling L. pictiiratus a superspecies and raising subspecies to species rank, Pasteur (1964) has not substantially improved our understanding of the situation. 1967 I.YGODACTYLUS GECKOS 3 species in the genus found on continental Africa. L. p. keniensis is a subspecies occurring in tiie northern two-thirds of Kenya and the bordering edges of the neighboring countries. L. sonuiUcus is an endemic of the arid Somali region. L. s. battersbyi ranges from the vicinity of the Somaliland-Ethiopian border south into northern Kenya (Fig. 1 ). Turkana is a district in the extreme northwestern corner of Kenya, lying between Lake Rudolf on the east and the Uganda- Kenya border on the west. Both geographically and ecologically the area is a western outlier of the arid Somali region to the east (Schmidt. 1923; Parker. 1932. 1936, 1942). As various parts of Turkana have been described in some detail elsewhere (von Hohnel, 1894; Worthington. 1932; Fuchs, 1935. 1939; Buxton, 1937), no more than a brief description of the area need be given here. South Turkana. where these observations were made (Figs. 2-3), is extremely arid and consists of numerous rugged plateaus and broken hills. Between these extend vast lava and quartzite pebble sheets and barren windswept sand flats. Goats and drought have kept vegetation to a minimum except along the larger water courses, which never carry a permanent supply of water. Acacia is by far the most predominant plant group, and it is in the Acacia trees that one finds the Lxi^odactyliis geckos. The local distribution of the two species in south Turkana was a bit peculiar. Both species were quite common in the study area except along the Kerio River at Lokori (Fig. 3). This area was visited several times during the course of my two-month stay, but only three or four individuals of each species were ever seen. This scarcity was not too unusual in the case of L. s. battersbyi as it was always found less frequently in the large Acacia tortilis which is dominant along the larger water courses such as the Kerio, Kakurio. and Kalabata. L. p. keniensis, however, was usually very common in these acacias, except along the Kerio — an exception that merits special consideration below (see Competition section). Food: Movement of prey organisms seems to be a prerequisite for eliciting a gecko's attention. Food items of L. p. keniensis, as determined by field observations and an analysis of the contents of ten stomachs, included a wide variety of invertebrates ranging in size from Diptera less than three mm long to mantids as long as the snout-vent length of the gecko itself. Most food items, how- ever, are small (5 mm or less in length). Perhaps the most common arthropod in the geckos' microhabi- tat and yet one that is never considered as food is an ant species 4 BREVIORA No. 268 of the genus Crematogastcr (species near paolii) — large numbers of which occur on every Acacia. The ants of this genus are well known for their fetid secretions (W. L. Brown and E. O. Wilson, personal communication), which may account for their being gen- erally avoided by geckos. However, the geckos often "ambush" columns of ants returning from a raid on a tree termite nest and pluck the largest morsel from the clutches of a particularly suc- cessful raider. At other times, the ants are actively avoided, and if, on being wounded and disabled, the gecko adheres to the tree, he is soon covered with the swarming ants. The termites of the Acacia always cover their paths on the sur- face of the tree with thin, dirt arches. When one of these arches is accidentally broken open, the workers immediately begin to repair it. During this process, a worker must expose himself to some degree for a few seconds while at the breach, and there is rarely a breach without a gecko patiently partaking of the feast. There is usually intense competition for the single position at the breach and once the dominant animal is removed, his place is quickly taken. Both species possess great visual acuity, individuals often being attracted by a movement of a small (1-2 cm) object 2-2 '/2 meters distant. Both species of geckos spent a good deal of time licking the extremely viscous, amber colored exudate of the Acacia trees, al- though the reasons for such actions are unknown. Like Kastle's ( 1964:489) Phelsuma, the Lygodactylus geckos under my obser- vation also licked a variety of jams and jellies, apricot being their favorite flavor. Water economy with these geckos, as with all reptiles of the area, must be of crucial importance. No dew was ever formed during the two months of field work, and we encountered rain only twice (1, 14 August); both times the amount was insufficient to cover the ground. One day a L. s. battersbyi was found licking the moisture from near the mouth of a canvas water bag that was customarily hung from the tree it inhabited. The gecko appeared practically oblivious to movement around him and continued his licking for approximately 15 minutes. Activity temperatures and periods: Cloacal temperatures of three active L. p. keniensis taken during the morning were 32.9, 34.5 and 35. TC. Air temperatures at the positions from which these geckos were taken were 30.8, 32.9 and 35.5'C. respectively. Cloacal temperatures for six geckos on the initial loss of the right- ing reflex ranged from 43.3-44.3'C. (average 43.7°C.). 1967 LYGODACTYLUS GECKOS 5 During the preliminary stages of heat distress L. p. keniensis behaves in a manner characteristic of many lizards, i.e., the body is raised oflf the ground on straightened legs, the mouth is held open, and the respiration rate increases. Both L. /;. keniensis and L. s. battersbyi are most active during the morning and late afternoon. During the mid-day they usually remain stationary, sometimes remaining in the same spot for an hour or more. I rarely saw a Lygodactylus on the sunny side of a branch. They usually chose to move along the underside of branches although, if this side were exposed to the sunlight, they would move along the upper, shaded side. Approximately 15 to 30 minutes after the sun has set, foraging individuals of L. /;. keniensis suddenly begin to ascend the branches of their trees. Some seek out a cavity in which to pass the night, but the majority of them climb to the peripheral branches and, turning so as to face the quickest avenue of escape, settle down for the night. In the morning, about 15 to 30 minutes before sunrise, they suddenly "come to life" again and quickly scurry down the branches to the main part of the tree to begin the day's foraging. L. s. battersbyi was also seen ascending and de- scending the terminal branches at approximately the same time as L. p. keniensis, although no individual of this species was ever observed assuming his sleeping position. These were by far the longest activity periods of any of the diurnal lizards in the area. The lacertids and agamids rarely ap- peared until about an hour and a half after sunrise and retired shortly before sunset. It thus seems as if the diurnal Lxgochictyliis geckos have retained enough of their family's nocturnal heritage to enjoy the best of both worlds. External niorpliology: Before proceeding with a discussion of the behavior of the two species, it will be helpful to describe briefly some of the more important aspects of color pattern and external morphology (Table 1 ). Hatchling L. p. keniensis are 13-14 mm in snout-vent length. The dorsum is grayish brown with whitish spots which are bor- dered with black along their anterior edges. These whitish spots extend onto the dorsal surface of the tail where they become wider than long but are still bordered on the anterior edge by black. The result is a generally mottled and cryptic appearance. The tail is brownish orange above and orange-red below. This bright ventral color extends forward along the midventral line to the pectoral region. The dark chevron markings of the throat are faintly evident. 6 BREVIORA No. 268 The change to the adult pattern entails the loss of the bright color on the tail, the change from an orange stripe to a yellow one along the ventral midline, the development of roughly longitudinal light and dark markings on the head and shoulders, and an intensification of the black on the throat. Adult males are usually slightly larger and stockier than females. The largest male of our specimens has a snout-vent length of 35 mm; that of the largest female is 34 mm. The black on the throat of the males also tends to be slightly more extensive than in females. The light and dark chevrons and white. O-shaped postmental spot, said to be characteristic of this subspecies (Loveridge, 1947: 198, 228), are lacking in some adults of both sexes that have almost solid black throats. In the majority of the specimens, however, the "typical"' pattern can be seen. Preanal pores and escutcheon scales appear in males at a snout- vent length of about 26-27 mm. There are usually 6-8 preanal pores arranged in an anteriorly projecting, obtuse chevron. The grayish escutcheon scales are distributed in four separate areas: the underside of the thighs, directly anterior to the preanal pores and between the preanal pores and vent. Females lack these secondary sexual characters. Both sexes have a light yellow midventral stripe on the body bordered laterally by a light gray area which extends to the dark brown dorsal color of the sides. When the animal is cold, how- ever, the pale midventral yellow becomes a very bright mustard yellow and the dark chocolate brown color of the dorsum extends ventrad to the edge of the bright yellow stripe. Hatchling L. .v. battershyi are about 1 1 .5 mm in snout-vent length. The dorsum is a cryptic light brown with scattered white and dark brown flecks. The tail is faintly tinged with reddish orange, and the entire venter is immaculate. Adults show little or no sexual dichromatism. Both sexes are uniformly brown above with dark brown spots on each shoulder and a few less well-defined dark brown spots on the sides. The white flecking and reddish tinge to the tail of juveniles are lack- ing. There is a thin dark brown horizontal stripe through the eye onto the side of the head. Adult females average slightly larger than adult males. The largest male and female in our sample each measures 27 mm. The smallest male with preanal pores and escutcheon scales measures 21 mm in snout-vent length. There are usually 6 pre- anal pores (two specimens have 5 pores) arranged in an anter- iorly projecting obtuse chevron. Only the undersides of the 1967 LYGODACTYLUS GECKOS 7 thighs bear escutcheon scales. Females lack these secondary sexual characters. Both L. p. keniensis and L. s. battersb\i have black pigment in the bones of the cranium and a jet black peritoneum. These characters have been tentatively associated with diurnal habits among geckos (Kiuge, 1967). although experimental work at- tempting to elucidate the adaptive significance of this pigment distribution has been inconclusive (Hunsaker and Johnson, 1939). Adults and hatchlings alike of both species have the tip of the tail very slightly spatulate and provided ventrally with a double row of lamellae. In general appearance this structure closely re- sembles the ventral surface of the expanded digital tip. The tip of a regenerative tail has the differentiated ventral "scansorial" scales of an original tail, but it lacks the ordered arrangement of a double row. Loveridge (1947:195), Kiistle (1964:487-488). Mertens (1964), and Pasteur (1964:12.76) discuss this structure on the underside of the tip of the tail in terms of its "adhesive" or "holding" advantage for these arboreal lizards. Such a function is suggested by its gross similarity to the digital pads, the fact that resting animals often have only the extreme terminal end of the tail in contact with the substrate (see Kastle. 1964. fig. 2a) but move about with the subterminal area dragging on the substrate and the terminal area curved slightly upward, and the fact that dead animals have been seen hanging by the tail. It is difficult to watch these two species of geckos actively forag- ing, especially during the late afternoon when they are most active, and not see another possible function of this terminal subcaudal expansion. During the times when an individual is very active, the tail is frequently raised in a deep arch and the tip of the tail is quickly touched to the substrate, withdrawn, and touched down again several times. In this manner the gecko quickly makes con- tact with parts of the surface behind him through almost 180". In this action it seems as if the tail functions less as a grasping organ than as a tactile sensory organ. Such a tactile sensory function would, of course, be an added advantage in an arboreal life. Display: From Kastle's ( 1964:494-495) description of the intra- specific display of L. p. picturalus it would appear that the dis- play of L. p. keniensis is very similar. In "full display" an adult gecko simultaneously raises the body slightly on straightened legs, arches the back (exposing the yellow midventral stripe?), bows the neck with the snout pointing slightly down and distends the 8 BREVIORA No. 268 black throat. The display may be given with the body in a head- on or broadside position relative to the individual eliciting the response. Often both body positions are assumed alternately or successively, with the displaying gecko moving closer to the antag- onist in the head-on position. It is also in the head-on position that the head is sometimes quickly jerked from side to side through a small arc. Among adults this display may be given by males to males in fighting, by males to females in courtship activities, by females to females in fighting, and by females to males in repulsing courtship overtures. Except in courtship, the displays never lasted more than a few seconds and in only two instances except for courtship did they lead to actual physical contact. One of these cases was brought about by introducing a foreign male into a small scrub Acacia tree occupied by two adult females and a single male. The resident male approached the quiescent foreign male while displaying head-on and broadside until, after a close broadside display, he suddenly attacked the foreign male with such force that both geckos fell out of the tree. The resident male immediately regained his tree, but the introduced male made his way back to the tree slowly and stayed around the base for some time. Several days later, however, he was still in the same tree, but he always elicited displays from the original male when- ever the two met. The only component of the display commonly encountered sep- arately was the distended black throat. This was often given in "mild distress" situations. Another, more common, means of aggression was a simple attack or rush at an individual. This be- havior was seen in juveniles as well as adults, especially when there was a large size difference — in favor of the attacker. During the 15 to 30 minutes after sunset that the geckos re- mained active, the intensity of the day's aggression between indi- viduals declined or diminished altogether. In one instance in which two females and one male were occupying a single, small Acacia tree, the larger female had been chasing the smaller female and male from a favorite food supply at the base of the tree all during the afternoon. Shortly after sunset, however, and until they retired, all three adults were feeding within centimeters of each other at the site. It is interesting to note that L. .v. battersbyi, which is so different in color pattern from L. p. keniensis, displays in the same way, i.e., the body is slightly raised on straightened legs, the back arched. 1967 LYGODACTYLUS GECKOS 9 [he neck bowed with his head pointing downward, and the immacu- late throat distended. The body may assume the head-on or broadside position alternately, and head jerking can be a part of the head-on approach. Although both species display in the same manner and some- times occur in the same tree, the display was never used inter- specifically. Confrontations did occur, however, and they always resulted in the L. s. battersbyi simply being chased from the scene by a larger or more bulky L. p. kcniensis. The hatchlings and juveniles of both species were always com- pletely ignored by the adults, although the adults often ate moving objects larger than the little geckos. The smaller individuals showed little fear of the adults, but they always moved aside when they found themselves in the path of an adult. Foreign objects never elicited a display response from cither species. Courtship and mating: Courtship and mating were observed only once in L. p. keniensis, but the behavior was sufficiently dif- ferent from Kasde's description (1964:497-498) to warrant a complete description based on my observations. The pair was first seen about 12:35 PM on the dark vertical trunk of an Acacia tort His along a large watercourse. The male was posturing in the typical display, slightly broadside to the fe- male. She moved toward his flanks and nipped at the basal third of his tail. The male immediately completed the circle by coming around to her tail which she would then twitch back and forth across his face while moving away with him following. The court- ship was interrupted several times at this point, but it was always initiated again by the female. After separating 70 to 100 cm. it appeared as if the female began searching for the male. She would approach his tail, and the male would stiffen slightly and give a few quick sideways jerks of the head. The female then nipped at his tail and he gave a full display. The female would again ap- proach his tail, and the male, completing the circle, would come up behind her and follow closely behind as the female moved slowly away wagging her tail across his face. Mating took place when the male continued creeping forward over the female after she had stopped, and grasped the female about the body approximately 2 mm posterior to the axilla with his forefeet. Holding the skin of the right side of the female's nape in his mouth, the male twisted his tail up and under hers from the right side, bringing their vents together with the male's tail on the female's left side. The left hind foot of the male was held on 10 BREVIORA No. 268 the dorsal side of the female's left thigh at its base, and his right hind foot was on the tree trunk. After a firm union was obtained, the male dropped his forefeet from around the body of the female onto the surface of the tree. Matins lasted for about 40 minutes, after which time the female flicked her tail sideways several times and the male disengaged himself, arching his tail upward in a deep bow. Each individual proceeded to lick the area around his vent. The cyclic turning of the female on the tail of the male and being turned upon by the male, the wagging of the tail in the face of the following male and the recurrent initiation of courtship activities by the female are elements in the courtship and mating of L. p. keniensis not observed by Kastle in his L. /;. picturatiis. Males were seen to initiate courtship activity on several occa- sions, but were quickly repulsed by a display from the female. Only a single mated pair of the more elusive L. s. battcrsbyi was observed. The pair was found about 10:30 AM in a shadow on the side of a scrub Acacia. The male had grasped the skin of the right side of the female's nape in his jaws and had brought his vent next to hers from the left side. The male's back was slightly hunched and his tail trailed behind on the left side. The male's forefeet were employed in grasping the female around the pectoral region behind the shoul- ders and were not placed on the trunk. His right hind foot rested on the dorsal surface of her right thigh at its base, and his left foot was in contact with the trunk. The female seemed to be made quite nervous by my presence. She wagged her tail from side to side, often touching the male's back. The female stayed at the same locus but frequently turned about 180\ always carrying the male with her and keeping the pair oriented along the longitudinal axis of the branch. After breaking off copulation, the female quickly disappeared, but the male began licking his vent. Reproduction: Both L. p. keniensis and L. .v. battersbyi lay two white, hard-shelled eggs in cavities of the tree which they inhabit. Usually the termites have excavated cavities in one or more branches, and, when these are abandoned, they become favorite sites for egg deposition (Fig. 4). Eggs will be laid in a particu- larly favored spot year after year, the newest pair of eggs being deposited on top of the pile. In this way as many as 40 egg shells can accumulate in a single cavity. L. s. battersbyi lays slightly smaller eggs than L. p. keniensis and the eggs of both species can be found in the same cavity. 1967 LYGODACTYLUS GECKOS 11 Whole egg shells were frequently found with a small circular hole through the shell and the yolky contents completely removed. The most likely explanation for this loss is predation by the ants which invariably inhabit the Acacia trees. Another possible predator of Lygodacrylus eggs or young is Homopholis fasciata, a gecko which inhabits, almost exclusively, these same cavities in which the Lygodactylus species lay their eggs. Developing eggs, hatchlings and gravid females of both species were encountered throughout July and August. The smallest fe- males of L. p. keniensis and L. s. battersbyi with large, shelled oviducal eggs nearly ready for deposition measured 28 and 24Vi mm, respectively, in snout to vent length (Table 1 ). In all females carrying two large oviducal eggs, the egg in the right oviduct was placed more anteriorly in the body cavity than the egg in the left oviduct. Popiilaiion structure: Almost any scrub Acacia over a meter in height was likely to have one or more Lygodactylus inhabiting it. In an attempt to get some idea of the population structure of the L. p. keniensis in these trees, several of them were held under observation and the adults collected as they came to light until it was felt that the entire adult complement had been collected. In one instance, a large Acacia tortilis was also "collected out." The results of such collecting (Table 2) showed that females out- number males 2 to 1 . Whether the 2 to 1 sex ratio is a reflection of a disproportion- ate birth ratio or the result of heavy selection pressure on the males cannot be decided at present. L. s. battersbyi is so secretive that it is impossible to satisfy one- self that the entire adult population of any one tree has been col- lected. The general impression, as indicated by the sex ratio of the specimens collected (18 males. 26 females), is that females may also outnumber males. On the open expanses, where the scrub Acacia predominates, the population inhabiting a single tree usually consisted of a single male, one to three females and a few juveniles. This uni- formity of structure is undoubtedly a reflection of the uniformity in size and form of the Acacia scrub. Very few scrub Acacia were large enough to support two or more males, although this was the rule in the much larger A. tortilis. In the few instances in which more than one male inhabited a tree small enough to allow continuous observation, it was evident that males were highly territorial. Females in trees with two or more males generally restricted their foraging areas to coincide approxi- mately with the territories of the males and were rarely seen far 12 BREVIORA No. 268 outside this area. No territorial display of females was ever ob- served, although a female will display to another female in certain instances, such as confrontation over a food source or a cavity for egg deposition. Confrontations over a favorite food source of limited size, such as a partially opened termite nest, indicated that dominance is largely a matter of size and not sex: a large female for instance will dominate the position at a breach in a termite archway (see above) to the exclusion of the male within whose territory the breach may be. Kastle (1964:491) reported that his L. pictitratus were never observed to come to the ground. Observations on L. p. kcniensis and L. .v. battersbyi showed that, although both species are highly arboreal, they will come to the ground when attracted by some small movement. If the tail of an individual were accidentally broken ofT during capture, another gecko in the tree would usually come to the ground for the thrashing tail. Geckos could also be enticed two or three feet away from the base of the tree by scatter- ing wood termites on the ground. During the late afternoon or early evening the geckos would often stay near the ground in the base of the tree and make forays onto the ground after a small, passing invertebrate. Individuals that were marked artificially (with paint spots) or naturally (lost tail, scars, etc.) were kept under observation for as long as three weeks, and in no case was there emigration from the "home tree." Strong arboreal tendencies and indications of territoriality make it seem likely that most individuals, once they have become established in a tree, rarely leave the tree for the rest of their lives. Competition: As these two species obviously occupy similar niches, the ways in which competition may be reduced are of in- terest. Although it was by no means unusual to find both species occu- pying the same tree and foraging within inches of each other, the general situation was one species to a tree. In this regard there seemed to be some correlation between the bark color of the Acacia inhabited and the occupant species, although no systematic analysis was undertaken to quantify this impression. In the relatively dense stands of the dark-barked A. tortilis along the larger watercourses, L. p. keniensis, the darker species, oc- curred almost to the exclusion of L. s. battersbyi. Even in the few A. tortilis found along the smaller watercourses, L. p. keniensis greatly outnumbered L. .v. battersbvi which might also be found in the same tree. 1967 LYGODACTYLUS GECKOS 13 On the more open flats, the scrub A. etbiaca is the dominant tree. The bark of this species is smooth and ranges in color from light gray to dark reddish brown. In pure stands of the gray- barked trees, L. s. battersbyi, the lighter species, was found to the exclusion of L. p. keniensis. Between these two extremes, however, the correlation became less exact and more difficult to establish. Whether this correlation, to the extent that it does exist, is due to habitat selection or random dispersal and natural selection for body color on the given substrate, or both, cannot be decided on the basis of the data at hand. Whenever the two species were found together in the same tree, it was more frequently the case that a L. s. battersbyi would be found in a "L. p. keniensis tree" than vice versa. In such situations, the smaller L. s. battersbyi tended to occupy the more peripheral branches, while the larger L. p. keniensis would be found on the larger central trunks and branches. Adult L. s. battersbyi also inhabited smaller trees and shrubs than adult L. p. keniensis. Both species were often encountered separately on single trees that were quite isolated by stretches (100-200 meters) of lava or quartz pebble sheets. In one instance a male and female L. p. keniensis and developing eggs were found on a windswept flat in a dark-barked Acacia that some time previously had been stripped of its foliage leaving only the trunks and main branches (Fig. 5). During the afternoon that I worked in this area the tree was hit by three large dust devils. Beyond these differences, however, there was little indication of any other ecological or behavioral differences between the two species. This is not too disturbing as many of these differences, which are expected on the basis of theory, are hinted at by dif- ferences in size and color patterns of the two species. For ex- ample, the size difference may be reflected by average differ- ences in the size and kind of prey organisms taken by the two species. It remains to explain the absence of L. p. keniensis from the Kerio River at Lokori (Fig. 3), while it is so common along all other large river courses investigated. It seems to me that this most peculiar local distribution may perhaps be due to the exclu- sion of L. /;. keniensis by the somewhat larger Hemidactyliis brooki angulatus. This widespread nocturnal gecko was very abundant in numerous situations along the Kerio (under the exfoliating bark of A. tortilis as well as in earth crevices) but was extremely rare beyond the confines of the Kerio except in the chimneys of the ubiquitous termite nests. 14 BREVIORA No. 268 The Kerio has its headwaters far to the south in the high central plateau of Kenya and is undoubtedly the most nearly permanent river in this region of south Turkana. It may be that some en- vironmental factor such as humidity is more nearly optimal for H. b. angulatus in south Turkana only along the Kerio and in the termite nests. Under such special conditions in an otherwise hostile environment the species may be able to occur to the exclusion of L. p. keniensis. ACKNOWLEDGEMENTS My thanks go first and foremost to Professor Bryan Patterson of the Museum of Comparative Zoology for permitting me to accom- pany his paleontological expedition to Turkana. Professor Patter- son and the other members of the expedition, Messrs. Arnold D. Lewis, Muteti Nthale, Daniel Ngumi, William D. Sill, Clyde T. Williams, and Roger C. Wood, aided in making the collections and observations. Without the encouragement and assistance of Dr. E. E. Williams of the Museum of Comparative Zoology, the idea of accompanying the Patterson expedition would never have come to fruition. Dr. Williams has also supplied many worthwhile suggestions during the course of his many readings of the manu- script. Dr. W. L. Brown of Cornell University very kindly identified the single species of ant which I collected, and both he and Dr. E. O. Wilson of the Biological Laboratories, Harvard University, supplied me with information on the biology of the genus Crema- togaster. Mr. Thomas Schoener of the Biological Laboratories at Har- vard made a thorough examination of the contents of 10 stomachs of L. p. keniensis. Mr. J. B. Gillet of the East African Herbarium identified the plants which I obtained in the Turkana country. Dr. A. S. Rand, Mr. George Gorman and my wife Carol have read the manuscript and offered several helpful suggestions. On behalf of the whole expedition I would like to express our appreciation and thanks for the many kindnesses extended to us by Dr. and Mrs. R. J. Drysdale-Anderson and their mission staff at Lokori. My field work in Turkana was supported by a grant from the Evolutionary Biology Fund, which is administered for the National Science Foundation by the Biological Laboratories at Harvard Uni- versity. Professor Patterson's expedition was supported by NSF Grant G. P. 1188. 1967 LYGODACTYLUS GECKOS 15 LITERATURE CITED Buxton, D. R. 1937. A natural history of the Turkana fauna. J. East Africa and Uganda Nat. Hist. Soc, 13 (3-4) . -85-104. FucHS, V. E. 1935. The Lake Rudolf Rift Valley Expedition. 1934. Geograph. J.. 86 (2):114-142. 1939. The geological history of the Lake Rudolf Basin, Kenya Colony. Phil. Trans. Roy. Soc. London, Ser. B. 229 (560): 219-274. HOHNEL, L. VON 1894. Discovery of Lakes Rudolf and Stefanie. Longmans, Green and Co., London, 2 vols. Translated by Nancy Bell. HuNSAKER, D. and C. Johnson 1959. Internal pigmentation and ultraviolet transmission of the in- tegument in amphibians and reptiles. Copeia, 1959, No. 4: 311-315. Kastle, W. 1964. Verhaltensstudien an Taggeckonen der Gattungen Lyxodcuuylii.s und Plielstima. Zeitschrift flir Tierpsychologie. 21 (4):486-507. Kluge, a. G. 1967. Higher taxonomic categories of gekkonid lizards and their evolution. Bull. Amer. Mus. Nat. Hist.. 13S (1 ):l-65. LOVERIDGE, A. 1947. Revision of the African lizards of the family Gekkonidae. Bull. Mus. Comp. Zool.. 98 { 1) : 1-469. Mertens, R. 1964. Der Eidechsenschwanz als Haftorgan. Senck. Biol., 45 (2): 1 17-122. Parker, H. W. 1932. Scientific results of the Cambridge Expedition to the East African Lakes, 1930-1931. 5. Reptiles and amphibians. J. Linn. Soc. London, Zool., 38 (258) :2 13-229. 1936. Reptiles and amphibians collected by the Lake Rudolf Rift Valley Expedition, 1934. Ann. Mag. Nat. Hist., (10) 18:594- 609. 1942. The lizards of British Somaliland. Bull. Mus. Comp. Zool., 91 (1 ):1-101. Pasteur, G. 1960. Notes preliminaires sur les lygodactyles (Gekkonides). L — Remarques sur les sous-especes de Lyi^iutdcfyln.s pictiiratiis. Bull. LF.A.N., Ser. A, 22 (4) : 1441-1452. 1964. Recherches sur revolution des lygodactyles, lezards .Afro- Malagaches actuels. Trav. Inst. Scient. Cherif., Ser. Zool., No. 29:1-^32. 16 BREVIORA No. 268 SCHMIDI, K. P. 1923. Contributions to the herpetology of the Belgian Congo based on the collection of the American Museum Congo Expedition. 1909-1915. Part II.— Snakes. Bull. Amer. Mus. Nat. Hist.. 49:1-146. Underwood, G. 1954. On the classification and evolution of geckos. Proc. Zool. .Soc. London. 124 (3):469-492. WORTHINGTON, E. B. 1932. The lakes of Kenya and Uganda, investigated by the Cambridge Expedition, 1930-1. Geograph. J., 79 (4):275-297. TABLE 1 Summary of the size and color pattern differences between Lygodactylus pictwatus keuicnsis and L. somalicus battersbyi pictiiniliis Snout-vent length (in mm.) of: Hatchling Largest adult male Largest adult female Smallest male with escutcheon scales Smallest female with oviducal eggs 13-14 (N = 10) 35 (N = 53) 34 (N = 79) 26-27 (N = 53) 28 (N=12) sonuiliiiis 111/2 (N=3) 27 (N=17) 27 (,N = 26) 21 (N=18) 241/2 (N=3) Color pattern of: Juvenile Adult Throat with faint dark chevrons; venter with orange-red mid- ventral stripe. Longitudinal light and dark markings on head and shoulders: dark black throat chevrons; venter with yellow mid- ventral stripe. Throat and venter immaculate. Head and shoulders uniform — color of dorsum; throat and midventer immaculate. 1967 LYGODACTYLUS GECKOS 17 TABLE 2 Summary of data on sex ratios of total adult complement of Lygodactylus pictwatiis keniensis inhabiting single trees Number adult lua of les Number of adult females Number of juveniles 0 6 8 3 1 3 2 3 2 2 2 3 1 or more 1 0 0 2 1 or more 2 0 2 4 or more 15 0 21 ( 2 2 1 0 + 47?) several 0 1 0 4 or more 0 30 1 64 0 20 + Ratio males/f emales = 1/2 1 18 BREVIORA No. 268 Figure 2. Typical view of the open flats in the study area. Figure 3. The dry bed of the Kerio River near Lokori. 1967 LYGODACTVLUS GECKOS 19 Figure 4. Two pairs of Ly^odacyliis puiiiniiiis kciiiciisis eggs in the termite-excavated cavity of a scrub Aciuiu. Part of the dead branch has been broken away to expose the eggs. Figure 5. Defoliated Acacia, on a wind blown fiat, harboring a male and female Lxt^ocldcixli/.s picturniiis kenieasis and developing eggs. BREVIORA Mmsemm of Coimiparative Zoology Cambridge, Mass. 29 June 1967 Number 269 TWO NEW SPECIES OF AMPHITARSUS (OPHIUROIDEA) FROM THE WESTERN NORTH ATLANTIC Amy Schoener INTRODUCTION The brittlestars discussed in this paper were recently obtained from the Western North Atlantic, at depths of between 200 and 535 m. Two new species are described and both are referred to the previously monotypic genus Amphitarsus H. L. Clark (1941). The type species of Amphitarsus is A. mirabilis (by orig- inal designation), remarkable in possessing peculiar winglike structures on the oral surface of the disk; these structures are in reality lateral arm plates which have become greatly expanded in the genital region. The two new species have similar structures (Figs. 1, 2, 3, B). Both species are apparently rare, since only seven specimens in all have been obtained in collections which included several thousand brittlestars. In 1941 H. L. Clark described the monotypic ophiuroid genus. Amphitarsus, from the Caribbean region. The familial relationship of the genus was ambiguous, since the mouth papillae, which are extremely important diagnostic characters, varied considerably, and Clark observed that "no two of the jaws look exactly alike." Clark tentatively placed it in the Amphiuridae. No further contri- bution towards the solution of this problem has since been made. Recent collections obtained by Drs. Robert Hessler and Howard Sanders (Woods Hole Oceanographic Institution), as part of a detailed study of benthic assemblages of the Western North Atlan- tic (Sanders, Hessler and Hampson, 1965), have yielded two new species of this genus, one of which apparently sheds light on its familial relationship. 2 BREVIORA No. 269 Amphitarsus mirabilis Clark Figure 1 Amphitarsus mirabilis H. L. Clark, 1941, Mem. Soc. Cubana Hist. Nat., 15 (1): 83-85, pi. 8. Holotype, MCZ 6232 from off Cayo Coco, Camaguey Province, Cuba, from 230 fathoms (420m). Two paratypes, MCZ 6398, 6399, from off Santa Clara Province, Cuba, from 175-235 fathoms (320- 430m). Description: Disk: Diameter ca. 6 mm. Aboral disk surface (Fig. 1, A) covered with fine granulation; very short radial areas alternate with much longer and concave interradial areas; radial shields small and narrow. Oral interradial surface (Fig. 1, B) densely covered with minute spinules or pointed granules. Oral shields rounded, pentagonal, slightly swollen, wider than long, and not granulated. Oral papillae in two groups: inner oral papillae typically two; short, swollen, and located on each side of the jaw angle, the innermost forming the pair characteristic of the Amphi- uridae, but on several of the jaws two of the papillae have fused laterally into a single wide papilla; outer oral papillae consist of a group guarding the large outer oral tentacle. Genital plate appears to be fused with four basal winglike arm plates. Arms: Reaching a length of 50-60 mm. Arm spines, 9. Ven- tral arm plates rather poorly defined, more or less pentagonal. Tentacle scales: 1 very large nearly round scale present through- out most of the arm; 1-3 smaller scales on basal arm joints (Fig. 1,C). Remarks: Clark's original description was accompanied by photographs of the holotype but no detailed illustrations were included. Amphitarsus nike' new species Figure 2 Holotype: MCZ 6797, about 140 miles N. of Surinam; Latitude 1° 53.5' N; Longitude 54° 33.3' W. Depth: 535 m. Woods Hole Oceanographic Institution CHAIN cruise 35, Dredge Station No. 33. April 25, 1963. Paratypes: From the same locality, MCZ 6799, 6800. Description: Disk: Diameter of 7.4 mm measured from outer radial shield edge to opposite edge of disk. Disk subpentagonal, flattened, and covered with fine granulation. Primary and other ^ The specific name, a nominative in apposition, refers to the Winged Victory (Nike of Samothrace). 1967 NEW SPECIES OF AMPHITARSUS .. • •'. < — ^v.. '•r-:"t-.'-'":'Uly(^ .■'■.■;. / radial shield V'.-'.'" '" i^---- ''"'■'! dorsal arm plare lateral arm plate 0,5nim .arm spme J.v';.-;,^ latere I arm plate u_arm spine tentacle scale ventral arm plate Figure 1. Ainphiiursiis mirabilis H. L. Clark, 1941. A, Aboral view of arm and part of disk. g.Oral view of two arms and part of disk. C, Lateral aspect of arm. 4 BREVIORA No. 269 C ^ 6 mm arm spme oral papilla oral shield genital area lateral arm plate tentacle scale ventral arm plate Imm Figure 2. Amphitarsus nike n.sp. A, Aboral view of arm and part ol disk. B, Oral view of arm and part of disk. C, Lateral aspect of arm. 1967 NEW SPECIES OF AMPHITARSUS 5 plates not visible, and only the tips of the radial shields exposed. Midline between pairs of radial shields barely distinguishable. Oral interradial surface of disk covered with fine granulation; oral shields small, subtriangular with rounded angles and a slightly de- pressed convex base; length about 0.9 mm. and width 0.8 mm. Adoral shields narrow proximally, becoming much wider at base between second oral tentacle pore and oral shields. A single papilla generally present at the jaw apex, but in other cases it is difficult to distinguish whether one, two, or none of the papillae are occupying this position. Inner oral papillae: four to five, conical, becoming broader and oval towards the outer edge of the jaw. A vertical series of teeth occurs at the jaw apex. Oral tentacle pores partially visible, opening into the mouth angle. Genital clefts elongate, flaring from the ventral arm plate and narrowing at the margin of the disk; they extend from the adoral shield to the margin of the disk. These clefts are superficially partitioned by five over- lapping winglike lateral arm plates; these plates begin at the second ventral arm plate and are attached to the lateral arm plate and not directly to the disk itself. Arms: Arms incomplete, at least 36 mm beyond the disk on 7.5 mm specimen. At the arm base, the fifth (outermost) over- lapping winglike lateral plate is divided into two or three flat, finger-like projections which are visible from the aboral side of the disk. Dorsal arm plates angular, broadened ovals, i.e. with obtuse angles inward in the proximal segments, becoming widelv oval after the first six segments. Narrow lateral arm plates meet at the midline on both oral and aboral surfaces, thereby separating the ventral and dorsal arm plates, respectively. Arm spines: three occur on most of the lateral arm plates; four are present on those joints which carry the winglike lateral plates. Beyond the disk, upper arm spine ca. 0.8-1.0 arm joints in length; lower two spines increasing slightly in length, the most ventral spine being ca. 1.2 arm joints in length. Ventral arm plates indistinct, somewhat broadened, hexagonal, with narrower proximal portions and wider, distal, convex portions beyond the disk. Tentacle pore large, and conspicuous, round-oval; on the basal pores there are two to three tentacle scales of reduced size. Thereafter, there is a single, flat, tentacle scale covering the entire pore. Color in life: Unknown; color in alcohol and dry specimens: whitish. Specimens examined: Holotype and paratypes. 6 BREVIORA No. 269 Amphitarsus spinifer^ new species Figure 3 Holotype: MCZ 6798, about 200 miles east of New York City; Latitude 40 01.8' N; Longitude 70 42' W. Depth: 200 m. Woods Hole Oceanographic Institution ATLANTIS cruise No. 283; WHOI Benthic Slope Station No. 2. August 28, 1962. Paratypes: From the same locality, MCZ 6801. Description: Disk: Disk diameter 3.8 mm from the outer edge of the radial shields to the opposite disk margin. Short radial areas alternate with longer interradial areas; disk subpentagonal, gen- erally flat, covered with multitudinous somewhat poorly-defined scales with isolated spines present on many of them; primary plates clearly visible, much larger than the surrounding scales, irregular in shape and separated from each other by the smaller disk scales. Radial shields clearly exposed, closely paired but not touching, and less than half the disk radius in length; length about 0.6 mm and width 0.2 mm. Oral disk interradial surface covered with scales and isolated spines similar to those on the aboral sur- face; the spines are largest on the centro-dorsal disk surface {ca. 0.4 mm long), becoming slightly smaller and narrower on the disk aboral margin {ca. 0.3 mm long), and decreasing to 0.25 mm on the oral surface. Genital clefts elongate, flaring from the first lateral arm plate, narrowing at the disk margin, extending from the oral shield to the margin of the disk. Clefts partitioned super- ficially by two overlapping plates which begin at the second ventral arm plate, to whose surface these winglike lateral plates are joined. Oral shields small, length ca. 0.3 mm and width ca. 0.5 mm, rounded triangular, with a slightly convex distal margin which buds off from the triangle base thereby creating two laterally pro- truding bulges just proximal to it. Adoral plates long, narrow, slightly widened where they meet on the midline. Two conical oral papillae present, on each side of the jaw, the terminal one form- ing one side of the pair characteristic of the Amphiuridae. The second papilla is separated by a diastema and is closest to the papillae encircling the second tentacle pore. Jaw character, nearly constant in specimens examined. Arms: In a smaller specimen (of 3.4 mm disk diameter), the arms, though broken, extend 20 mm beyond the disk edge; in the specimen described, the break is 6 mm from disk edge. At the 1 The specific name is intended to call attention to the spiny character of the disk. ■ 1967 NEW SPECIES OF AMPHITARSUS 0.4inm B — — radial shield 'i^^— arm spine dorsal arm plate arm spme oral papilla -s^oral shield disk spine 5^ -^ — lafcral arm plate arm spme 0.5mm Figure 3. Amphitarsus spinifer n.sp. A, Aboral view of arm and part of disk. B. Oral view of two arms and part of disk. C. Lateral aspect of arm. 8 BREVIORA No. 269 arm base, second (outermost) winglike lateral plate, visible from the aboral side, is broken into five, flattened, finger-like projections. Dorsal arm plates rounded hexagonals, length exceeding width. Arm spines: six at arm base, five to four further out on arms; two spines on joints carrying winglike lateral plates. Beyond disk, arm spines ca. 0.4 mm, equalling approximately the length of an arm joint; dorsalmost spine is shorter and flat. Tentacle scales: two on proximal joints; one externally placed, rounded, leaflike scale; one internally placed erect spine, attached to the median edge of the ventral arm plate (present only in first six-seven segments). Ventral arm plates basically hexagonal, with deeply emarginated lateral areas where one tentacle scale attaches. Color in life: Unknown; color when dry: whitish. Remarks: Three of the four paratypes were fragmented, but it was clear that they were not from the same individual, since each bore an almost complete set of primary plates on the aboral disk surface. The fourth was badly mangled. Specimens examined: Holotype and paratypes. DISCUSSION The three species of the genus Amphitarsus illustrated and dis- cussed above possess the singularly striking character of winglike lateral arm plates in the genital region. Although variable in num- ber, these plates distinguish the genus from all others. If the main characteristic relating these species is taken to be these unusual winglike plates, then the genus could be placed in the family Am- phiuridae, on the basis of the jaw structure of A. spinifer. The presence of two papillae at each jaw angle in A. spinifer is quite constant, even though the species is represented by smaller indi- viduals than either of the other two species; perhaps the younger stages of A. spinifer differ less from the adult condition, or it may not be as immature as is suspected on the basis of size alone. The variation in the oral papillae evidenced in A . mirabilis and A . nike may be due to the immaturity of these specimens. All individuals of this genus so far as known are fairly small, although one arm fragment possibly belonging to A. spinifer suggests that a larger size is attained. Thus it is possible that these are young individuals, and if this is the case, they might belong to the Ophiacanthidae or even to the Ophiocomidae (Fell, personal communication). At present the genus is known only from the Western North Atlantic. Whether or not it is truly confined to this region, only continued investigations of other areas of the ocean will tell. 1967 NEW SPECIES OF AMPHITARSUS 9 This work was carried out at the Museum of Comparative Zool- ogy during the tenure of fellowships from the International Wom- en's Fishing Association and from Harvard University; research facilities were provided under NSF grant GB-3532, and speci- mens were collected by the Woods Hole Oceanographic Institution under NSF grants GB-15838 and GB-563. I wish to thank Dr. H. B. Fell of this department for his guidance and advice, Dr. R. Turner for reading the manuscript, and Drs. R. Hessler and H. Sanders for allowing me to study their collections. BIBLIOGRAPHY Clark, H. L. 1941. Report of scientific investigations of the Atlantis expedition to the West Indies. Mem. Soc. Cubana Hist. Nat. 15 (1):83-85. Fell, H. B. 1960. Synoptic keys to the genera of Ophiiiroidea. Zool. Piibl. Vic- toria Univ., Wellington, No. 26:1-44. Sanders, H. L., R. R. Hessler, and G. R. Hampson 1965. An introduction to the study of deep-sea benthic faunal assemblages along the Gay-Head — Bermuda transect. Deep Sea Res. 12: 845-867. BREVIORA MiLiseiiaiM of Comparsitive Zoology Cambridge. Mass. 29 June 1967 Number 270 ADDITIONS TO THE UNIONID FAUNA OF THE GULF DRAINAGE OF ALABAMA, GEORGIA AND FLORIDA (MOLLUSCA: BIVALVIA) Richard I. Johnson More than a decade has passed since Clench and Turner ( 1956) published their pioneering work, "Freshwater MoUusks of Ala- bama, Georgia, and Florida from the Escambia to the Suwannee River." It is not surprising that in the time since the publication of that paper certain additions and corrections have been, and can be, made concerning the Unionidae. Athearn (1964) added two new species, Lcunpsilis haddleton'i and Villosa choctcnvcu.sis. both from the Choctawhatchee River, Alabama. Johnson (1965) described Auodonta peggyac which ranges from the Choctawhatchee River, Alabama, to the Hills- borough River, Florida. Athearn (1964) convincingly differenti- ated Lcunpsilis jonesi Vander Schalie from L. australis Simpson, both from the Choctawhatchee River. Alabama, and regarded as synonymous by Clench and Turner. In 1964 Athearn collected a live specimen of Lcunpsilis bimym- inatiis Simpson in the upper Flint River drainage of Georgia, which confirmed its distinctness from Lcunpsilis cxccivaius (Lea) and vali- dated the earlier records thought to be in error by Clench and Turner. An examination of the type, and of subsequently col- lected specimens, has clearly shown that Alasmidonta wrii^/uicuui (Walker) of the Ochlockonee River, Florida, is distinct from A. iricuigulata (Lea) of the Apalachicola River system of Georgia and Florida. Under Amulontoides radicitus (Conrad), referred to as A. elliotti (Lea) by Clench and Turner, are included records of Strophints subvexus (Conrad) from the Chipola River, Florida. Obovarici mtulota (Wright) from the Escambia River, Florida, was overlooked and has recently been rediscovered. This paper is intended to be supplemental to the work of Clench and Turner. All of the species mentioned here are figured except 2 BREVIORA No. 270 Anodonta peggyac Johnson which I described and figured in 1965. Strop/lit us subvexus (Conrad), Anodontoides radiatus (Conrad), Lampsilis binominatus Simpson, and L. jonesi Van- der SchaHc are fully described. Remarks are given on Alasmidonta wrightiana (Walker), Obovaria rotulata (B. H. Wright), Lampsilis haddk'toni Athearn, and Villosa choctawcnsis Athearn. All of the extant types of the various named forms mentioned here have been examined and photographed. I wish to thank Mr. Herbert D. Athearn for allowing me to examine his specimen of Lampsilis binominatus Simpson. Thanks are also extended to Drs. K. J. Boss, W. J. Clench, and R. D. Turner for reading the manu- script and suggesting improvements. The following abbreviations have been used. ANSP — Academy of Natural Sciences of Philadelphia, Penn- sylvania. MCZ — Museum of Comparative Zoology, Harvard Uni- versity, Cambridge, Massachusetts. MZUM — Museum of Zoology, University of Michigan, Ann Arbor, Michigan. USNM — United States National Museum, Washington, Dis- trict of Columbia. All specimens mentioned in the text are in the Museum of Com- parative Zoology unless otherwise noted. Strophitus SUBVEXUS (Conrad) Plate 1, fioures 1-3 o Anodonta siihvexa Conrad 1834, .Amer. Jour. Sci.. 25: 341, pi. !. fig. 12 (Black Warrior River [Alabama] type, ANSP [lost] ). Mciri^arifdna connasangaensis Lea 1858. Proc. Acad. Nat. Sci. Phila., 10: 138 (Connasaiiga River, one of the headwaters of the Alabama River, Gilmer Co.. Georgia). Lea, 1859, Jour. Acad. Nat. Sci. Phila.. (2)4: 229, pi. 32. fig. 113; figured holotype USNM 86277. Lea 1859, Obs. Unio. 7: 47. Margcirilana spilUnunii Lea 1858, Proc. Acad. Nat. Sci. Phila., 10: 138 ([Magbys Creek, written on type, not located on modern county map] Tombigbee River, near Columbus [Lowndes Co.[ Mississippi). Lea, 1862, J^our. Acad. Nat. Sci. Phila., (2) 5: 105, pi. 17, fig. 252; figured holotype USNM 86278. Lea, 1862, Obs. Unio, 8: 109. Muri^aritana tombecbeensis [sic\ Lea 1858, Proc. Acad. Nat. Sci. Phila., 10: 138 (Tombecbee River, Columbus [Lowndes Co.] Mississippi). Margcirilana tombigbeensis Lea 1862, Jour. Acad. Nat. Sci. Phila. (2) 5: 107, pi. 18, fig. 255; figured holotype USNM 86253. Lea, 1862, Obs. Unio, 8: 111. 1967 GULF DRAINAGE UNIONIDS 3 Margaritana gesnerii Lea 1858, Proc. Acad. Nat. Sci. Phila., 10: 138 ( Up- haupee Creek [Macon Co.] Alabama; Chattahoochee River, below Columbia [sic] [Columbus, Muscogee Co.] Georgia). Lea, 1862. Jour. Acad. Nat. Sci. Phila. (2 ) 5: 211. pi. 32, fig. 280; figured holo- type USNM 86212 from Uphaupee Creek. Lea. 1863. Obs. Unio. 9: 33. Morgaritaiia aJahamensis Lea 1861, Proc. Acad. Nat. Sci. Phila.. 13: 41 (Talladega Creek [Talladega Co.] Alabama). Lea. 1862. Jour. Acad. Nat. Sci. Phila.. (2)5: 104. pi. 16. fig. 249; figured holotype USNM 86262. Lea, 1862. Obs. Unio, 8: 108. Margaritana coliimheiisis Lea 1867, Proc. Acad. Nat. Sci. Phila.. 19: 81 (Tombigbee River, near Columbus [Lowndes Co.] Mississippi, type not in USNM [lost]). Strophitus conasaiigaensis [sic] (Lea). Ortmann. 1923. Nautilus, 36: 130. Strcjpliifiis subvexiis (Conrad). Frierson. 1927. Check List North American Naiades, p. 23. Strophitus spillnuinii (Lea). Vander Schalie, 1940. Lloydia. 3: 197. Anodontoides elliotti (Lea), partini. Clench and Turner. 1956. Bull. Florida State Mus.. 1: 182. Description. Shell medium in size, reaching about 100 mm in length. Outline subrhomboidal, becoming subelliptical or trape- zoidal with age. Valves inequilateral, subcompressed to rather inflated, thin. Anterior end regularly rounded, posterior end obliquely subtruncate above, rounded or subtruncate below. Ven- tral margin almost straight, often becoming slightly arcuate with maturity. Dorsal margin short, forming a broad angle with the obliquely descending posterior margin. Hinge ligament short but prominent. Posterior ridge generally quite broad and very faintly doubie. Posterior slope flat or slightly concave. Umbos moder- ately swollen, slightly raised above the hinge line, located from about the center to the anterior third of the valves, their sculpture consisting of a few very strong ridges running parallel to the growth lines, somewhat interrupted at the posterior ridge. Surface of the shell smooth, generally shiny, waxy yellow, greenish yellow, brownish, often rayless, though sometimes with green rays espe- cially on the posterior slope. Occasionally the surface has numerous broad green rays which give the whole shell a greenish appearance. Left valve with a stumpy, vestigial pseudocardinal tooth directly below the umbo, sometimes with traces of first and third teeth, before and behind it. No lateral teeth. Right valve with one rather well developed pseudocardinal, triangular and compressed, or tubercular, knob-like and stumpy. Umbonal cavities moderately deep with dorsal muscle scars. Length Height Widtl mm mm mm 59 36 25 62 36 24 77 52 37 92 56 39 93 51 40 116 54 38 4 BREVIORA No. 270 Anterior adductor muscle scars well impressed, posterior ones less so. Pallial line faint. Nacre bluish, dull salmon or purplish, often spotted with yellow. Measurements Topotype of S. suhrcxiis (Conrad) MCZ 146655. Holotype of M. c:>niui.sau^aen.sis Lea Holotype of M. tombigheensis Lea Holotype of M. gesnerii Lea Holotype of M. spillnumii Lea Holotype of M. alabamensis Lea Remarks. Strophitus siihvexiis (Conrad) of the Gulf drainage, shows some of the same diversity of shape as does Strophitus undu- kitus (Say) of the Atlantic drainage and interior basin. It is read- ily distinguished from itndiilaius which has entirely rudimentary pseudocardinals, represented, if at all, by slight swellings. In S. siibve.xus there is at least one pseudocardinal in each valve, and while small, they are usually well developed, triangular and com- pressed, or tubercular, knob-like and stumpy. It has also been confused with Anodontoides radiatus (Conrad), but that shell is always regularly elliptical with rays over the entire surface, while subvexus is subelliptical, inclined to be biangulate behind, seldom with rays over the entire surface, and with pseudocardinals that are less rudimentary. Vander Schalie (1940) correctly reports this species from two localities on the Chipola River as Strophitus spillmanii. Clench and Turner ( 1956) assumed that Vander Schalie's specimens were Anodontoides elliottii (Lea) \= radiatus (Conrad)], but all of their records of radiatus from the Chipola River are S. subvexus. Ortmann (1923) reviewed some of the names applied to the forms of this species, but did not include those from the Tombigbee River drainage. S. subvexus (Conrad), like its synonyms con- nasaugaensis Lea and tombigbeensis Lea, was founded on rather trapezoidal inflated individuals, while other names were given to the more common subelliptical specimens. Conrad's figured type which was about 2 inches (51 mm) in length, was specifically stated to be in the Academy of Natural Sciences of Philadelphia, but it has been lost. He did not give a precise locality, but in all 1967 GULF DRAINAGE UNIONIDS 5 probability the type came from the Black Warrior River in the vicinity of Tuscaloosa, Tuscaloosa Co., Alabama, since Conrad is known to have been there in June of 1833 (Wheeler, 1935: 38). Range. Gulf drainage, from the Pascagoula River system of Mississippi to the Apalachicola River system of Georgia and Florida. SPECIMENS EXAMINED PASCAGOULA RIVER SYSTEM Mississippi: Edahoma Creek, 6 mi. NW Soso, Jones Co. ALABAMA COOSA RIVER SYSTEM Tonibigbee River Drainage. — Mississippi: Tombigbee River, Columbus, Lowndes Co. Alabama: Coalfire Creek, Coalfire. Pick- ens Co. Bodka Creek, 5 mi. NW Gainsville, Sumter Co. Okatuppa Creek, 4 mi. SE Toxey, Choctaw Co. Black Warrior River. Tusca- loosa Co. Cahawba River Drainage. — Alabama: Black Creek, St. Clair Co. Shoal Creek, Montevallo, Shelby Co. Coosa River Drainage. — Georgia: Teloga Creek; Chattooga River, Summersville; both Chattooga Co. Alabama: Mills Creek, Cherokee Co. Tennessee: Conasauga River, Conasauga, Polk Co. Georgia: Conasauga River, Gregorys Mill, 10 mi. N Eaton, Mur- ray Co. Conasauga River, Gilmer Co. (USNM). Conasauga River. 1 .4 mi. N Resaca, Gordon Co. Black Creek. St. Clair Co. Choc- ciocca Creek, 3 mi. S Lincoln; Talladega Creek; both Talladega Co. Tallapoosa River Drainage. — Alabama: Uphaupee Creek, Macon Co. Alabama River Drainage.- — Alabama: Cub Creek, Pine Hill, Wilcox Co. APALACHICOLA RIVER SYSTEM Chipola River Drainage. — Florida: Big Creek, 8 mi. W Malone, Jackson Co. Cowarts Creek, nr. Florida state line, Hou- ston Co. (MZUM). Reedy Creek, 6 mi. W Malone; Chipola River. 1 mi. N Marianna; both Jackson Co. Chipola River. Scotts Ferry, Ca'hounCo. (MZUM). Chattahoochee River Drainage. — Georgia: \ Chattahoochee River] below Columbus. Muscogee Co. (USNM). Flint River Drainage. — Georgia: Kinchafoonee Creek, 4 mi. N Bronwood, Terrell Co. Ichawaynochaway Creek. 3 mi. N Morgan. Calhoun Co. 6 BREVIORA No. 270 Anodontoides radiatus (Conrad) Plate 2, figures 1-4 Alusinidonta ladiata Conrad 1834, Amer. Jour. Sci., 25: 341, pi. 1, fig. 10 (small streams in South Alabama; measured holotype ANSP 41147 labeled, Greene Co. ) Margahtana elliottii Lea 1858, Proc. Acad. Nat. Sci. Phila., 10: 138 (Chat- tahoochee River (below Uchee Bar] near Columbus [Muscogee Co.] Georgia). Lea, 1859, Jour. Acad. Nat. Sci. Phila.. (2)4: 226, pi. 31, fig. 108. Hgured holotype USNM 86257. Lea, 1859. Obs. Unio, 7: 44. Margaiitana clUpticci Lea 1859, Proc. Acad. Nat. Sci. Phila., 11: 113 (Tombigbee River, Columbus [Lowndes Co.] Mississippi). Lea, 1862. Jour. Acad. Nat. Sci. Phila., (2)5: 106, pi. 18, fig. 254; figured holo- type USNM 86258. Lea. 1862, Obs. Unio, 8: 110.^ Anodonta showcdterii Lea 1860, Proc. Acad. Nat. Sci. Phila.. 12: 307 (Coosa River Wetumpka [Elmore Co.] Alabama). Lea, 1862, Jour. Acad. Nat. Sci. Phila., (2)5: 215, pi. 33, fig. 284: figured holo- type USNM 86487. Lea, 1863, Obs. Unio, 9: 37.^ Strophitiis radiatus (Conrad). Frierson, 1927, Check List North American Naiades, p. 23. Anodontoides elliotti (Lea), pariini. Clench and Turner, 1956, Bull. Flor- ida State Mus., 1: 182, pi. 5. fig. 3. Description. Shell generally small to medium in size, reaching 72 mm in length. Outline elliptical. Valves inequilateral, inflated, thin. Anterior end regularly rounded, posterior end rather pointed. Ventral margin slightly curved. Dorsal margin slightly curved scarcely forming an angle with the obliquely descending posterior margin. Hinge ligament covering most of the dorsal mar- gin. Posterior ridge generally broad and rounded, occasionally subangulate. Posterior slope flat. Umbos moderately swollen, slightly raised above the hinge line, located toward the anterior third of the shell, their sculpture consisting of a few moderate ridges that curve up sharply behind. Surface of the shell smooth, brownish or brownish green, generally with bright or dark green rays of different widths over the entire surface of the shell. Left valve with a rudimentary, laminate, somewhat bifurcated, pseudocardinal tooth. Hinge plate very narrow. No lateral teeth. Right valve with one long, narrow, low, pseudocardinal tooth, anterior to the umbo. Umbonal cavities rather shallow with dorsal muscle scars. An- terior adductor muscle scars clearly outlined, posterior ones ob- scure. Pallial line faint. Nacre bluish white, often with yellow spots toward the umbonal cavities, iridescent posteriorly. !967 GULF DRAINAGE UNIONIDS Length Height Widtli mm mm mm 50 29 19 64 36 22.5 71 35.5 25.5 Holotype of M. clliptica Lea Holotype of A. radkita Conrad Mosquito Creek, Chattahoochee, Gadsden Co., Florida. 72 43 31 Holotype of /I. i/n^nY/Z/i^r/V Lea Remarks. Anodontoides radiatus (Conrad) of the Gulf drain- age, shows the same consistency of shape as does Anodontoides jerussacianus (Lea) of the interior basin. It is readily distinguished from jerussacianus which is toothless, by the presence of a vestigial tooth in each valve. A. radiatus is, in general, more consistently elliptical, and has brighter, more clearly defined rays. In the Gulf drainage. A. radiatus has been confused with Strophitus siibvexus (Conrad). Clench and Turner's (1956:183) records of elliotti \= radiatus] from the Chipola are all suhvexus. A. radiatus differs from S. suhvexus by being a smaller species with a consistently elliptical shell, having bright green rays over the entire surface of the shell, whereas siibvexus is subelliptical. inclined to be biangulate behind, seldom with rays over the entire surface. The hinge teeth of radiatus are laminate and delicate, the left valve with a single bifurcated tooth, the highest portion often being posterior to the umbo, the right valve with one long narrow tooth, parallel to the anterior dorsal margin. The hinge teeth of siibvexus are less delicate, the left valve with a stumpy, vestigial, pseudocardinal directly below the umbo, sometimes with the trace of a first and third tooth before and behind it, the right valve with a triangular, or knob-like tooth which is not parallel to the anterior dorsal margin. Frierson ( 1927:23) gives the correct synonymy of this species. Conrad's holotype of Alasniidonta radiata is labeled as coming from Greene Co.. Alabama. In all probability the type is from a tributary of the Black Warrior River near Erie | a town, no longer extant, about 10 mi. W Greensboro, Hale Co.. Alabama] since Conrad is known to have been there around the first of June 1833 (Wheeler. 1935:38). Range. Gulf drainage, from the Alabama-Coosa River system of Alabama to the Apalachicola River system of Georgia and Florida. Absent from the intervening Choctawhatchee River sys- tem and the Chipola River of the Apalachicola River system. 8 BREVIORA No. 270 SPECIMENS EXAMINED ALABAMA COOSA RIVER SYSTEM Tonibigbee River Drainage. — Mississippi: Tombigbee River, Columbus, Lowndes Co. (USNM). Akibanui: [Tributary of Black Warrior River] Greene Co. (ANSP). | Tombigbee River] Mar- geno Co. Coosa River Drainage. — Alabama: Coosa River, Wetumpka, Elmore Co. (USNM). Alabama River Drainage. — Alabama: Pine Barren Creek, Camden, Wilcox Co. ESCAMBIA RIVER SYSTEM Conecuh River Drainage. — Alabama: Oakywoods Creek, But- ler Co. Dry Creek, Brantly, Crenshaw Co. APALACHICOLA RIVER SYSTEM Chattahoochee River Drainage. — Georgia: Chattahoochee River, West Point, Troup Co. (USNM). Chattahoochee River, Uchee Bar, below Columbus, Muscogee Co. (USNM). Big Uchee Creek, Young's Bridge, 6 mi. NE Scale; Big Uchee Creek, below Fort Mitchell; both Russell Co. Apalachicola River Drainage. — Florida: Mosquito Creek, 1 mi. S. Chattahoochee, Gadsden Co. Alasmidonta wrightiana (Walker) Plate 2, figure 5 Stropliitiis wriglitiaiiiis Walker 1901, Nautilus, 15: 65, pi. 3 (Tributaries of the Flint River, Baker Co., Georgia; holotype MZUM 74938. Type locality corrected to Ochlockonee River, Florida, on the authority of Walker, by Simpson, 1914, Cat. Naiades, 1: 357). AUisnudoiita triangiilala (Lea), partiin. Clench and Turner, 1956, Bull. Florida State Mus., 1: 180. Remarks. In the Gulf drainage, Alasmidonta wrightiana (Wal- ker) has been confused with A. triangulata (Lea) but it differs from that species by being elliptical, rather than triangular, by having a posterior ridge that is rounded rather than sharp, and especially by having a very characteristic sculpture on the posterior slope which consists of a number of ridges that extend from the posterior ridge to the dorsal and posterior margins. A. triangulata is not sculptured on the posterior slope. Long known only from the holotype, the Museum of Compara- tive Zoology is grateful to the Florida State Museum for a more recently collected specimen. 1967 GULF DRAINAGE UNIONIDS 9 Range. Gulf drainage, restricted to the Ochlockonee River, Florida. SPECIMENS EXAMINED OCHLOCKONEE RIVER SYSTEM Ochlockonee River Drainage. — Florida: Ochlockonee River, bridge, 8 mi. W Tallahassee, Leon Co. Obovarfa rotulata (B. H. Wright) Plate 3, figure 1 Unio rotulatiis B. H. Wright 1899. Nautilus, 13: 22 (Escambia River, Escambia Co., Florida: holotype USNM 159969, figured by Simpson, 1900. Proc. Acad. Nat. Sci. Phila., p. 78, pi. 4. fig. 2). OhoYurici rotiihita (Wright). Simpson, 1914, Cat. Naiades, 1: 297. Remarks. Obovaria rotulata (Wright) of the Escambia River, differs from Obovaria imicolor (Lea) of the Alabama-Coosa River system by being almost circular in outline, the latter being elliptical or ovate. Overlooked by Clench and Turner, rotulata is the easternmost representative of Obovaria in the Gulf drainage. Formerly known only from the holotype, the Museum of Zoology, University of Michigan, has recently acquired specimens of this species collected by William H. Heard. Range. Gulf drainage, restricted to the Escambia River, Florida. SPECIMENS EXAMINED ESCAMBIA RIVER SYSTEM Escambia River Drainage. — Florida: Escambia River, Escam- bia Co. (USNM). Lampsills BiNOMiNATUS Simpsou Plate 2. figure 6; Plate 3, figure 2 Unio lineatiis Lea 1840, Proc. Amer. Philos. Soc, 1: 287 (Chattahoochee River, Columbus [Muscogee Co.] Georgia). Lea, 1842, Trans. Amer. Philos. Soc, 8: 206, pi. 12, fig. 20; figured holotype USNM 84884. Lea, 1842, Obs. Unio, 3: 44, non Unio lineata "Valenciennes' Bory de St. Vincent 1827. Lcunpsilis hinoininatus Simpson 1900, Proc. United States Natl. Mus., 22: 528, new name for Unio lineatiis Lea 1840. Lanipsilis e.xcavatiis (Lea), partiiu. Clench and Turner, 1956, Bull. Florida State Mus., 1: 198. 10 BREVIORA No. 270 Description. Shell rather small in size, not reaching over 58 mm in length. Outline of male elliptical; of female obovate. Valves inequilateral, somewhat inflated, and rather thin. Anterior end regularly rounded, posterior end distinctly pointed in the male, somewhat truncated in the female. Ventral margin straight or slightly curved in males, often extending below the base line in females. Dorsal margin short and straight, forming a broad angle with the obliquely descending posterior margin. Hinge ligament short but prominent. Posterior ridge faintly double, rather rounded and poorly defined. Posterior slope slightly concave, usually with a few faint ridges and wrinkles. Umbos moderately swollen, slightly raised above the hinge line, located toward the anterior third of the shell, their sculpture not observed. Surface of the shell smooth and shiny, waxy yellow, greenish yellow, sometimes brownish, with very fine, slightly interrupted, dark green rays over the entire sur- face of the shell. Left valve with two serrated pseudocardinals, one in front of the other, the anterior one somewhat triangular, the hinder one much lower. Hinge line short but with a considerable interdentum, before two short straight lateral teeth. Right valve with two pseudocardinals, separated by a deep pit, the anterior tooth ves- tigial, the hinder one rather chunky and triangular; one lateral tooth. Umbonal cavities rather deep with dorsal muscle scars. An- terior adductor muscle scars well impressed, posterior ones less so. Pallial line faint. Nacre bluish white, or salmon colored Measurements Length Height Width mm mm mm Holotype of U. lineatus Lea, female Chattahoochee River, Columbus, Musco- gee Co., Georgia, female Ibid., male Line Creek, Vi mi. W Digbey, Spaulding Co., Georgia, male Remarks. Lanipsilis binominatns Simpson of the upper Flint River drainage, has been confused with Latnpsilis excavatus (Lea) which ranges from the Amite River in Louisiana to the Escambia River of Alabama and Florida. L. binominatns appears to be a much smaller species. When specimens of the same size of both 33 23 16 37 28 19 38 25 18 58 41 32 1967 GULF DRAINAGE UNIONIDS 11 species are compared, the males of hinominatus are more ellipti- cal, and specimens of both sexes have narrower, sharper, darker green rays which are consistently linear and do not broaden and become diffuse as do the rays of L. e.xcovatus. Clench and Turner did not find L. excavatus east of the Escambia River, and since they did not recognize L. bmominatus, they as- sumed its type-locality to have been in error. Subsequent collect- ing has substantiated the original locality. Range. Gulf drainage, limited to the Upper Apalachicola River system. SPECIMENS EXAMINED APALACHICOLA RIVER SYSTEM Chattahoochee River Drainage. — Georgia: Chattahoochee River, Westpoint, Troup Co. (H. D. Athearn, formerly Boston Soc. Nat. Hist. 3526). Chattahoochee River, Columbus, Muscogee Co. Flint River Drainage. — Georgia: Line Creek, Vi mi. W Dig- bey, Spaulding Co. (H. D. Athearn). Lampsilis jonesi Vander Schalie Plate 3, figure 5 Lampsilis jonesi Vander Schalie 1934. Nautilus, 47: 125, pi. 15, figs, la, lb, 2, 3a, 3b (Pea River, Priston's Mill, Dale Co., Alabama; holotype, Alabama Mus. Nat. Hist.) Lampsilis aiistralis Simpson, partini. Clench and Turner, 1956, Bull. Florida State Mus., 1: 182. Lampsilis jcmesi Vander Schalie. Athearn, 1964. Nautilus, 77: 138. De.scription. Shell generally small to medium in size, reaching 48 mm in length. Outline subovate or elliptical, moderately elon- gate. Valves inequilateral, not much inflated, and rather solid. Anterior end regularly rounded, posterior end generally distinctly biangulate. Ventral margin slightly curved. Dorsal margin slightly curved, forming a sharp angle with the obliquely descending, doubly scalloped, posterior margin. Female shells show general swelling of the disc below the posterior ridge, and the ventral mar- gin is more angular than that of the male shell. Hinge ligament short. Posterior ridge broad, but distinctly double. Posterior slope slightly concave. Umbos moderately swollen, slightly raised above the hinge line, located toward the anterior quarter of the shell, their sculpture not seen. Surface of the shell smooth, shining. 12 BREVIORA No. 270 olivaceous green or olivaceous yellow, usually with irregularly distributed green rays. Left valve with two solid, compressed, jagged pseudocardinals. a slight interdentum, and two thin slightly granular lateral teeth. Right valve with two pseudocardinals, the more posterior one rudi- mentary, the anterior one well developed, stumpy, and jagged; one lateral tooth. Umbonal cavities shallow, with traces of dorsal muscle scars. Anterior muscle scars well impressed, posterior ones less so. Pal- lial line distinct anteriorly. Nacre bluish-white, thickened anter- iorly; thinner and slightly iridescent posteriorly. Measurements Length Hci^lit Width mm mm mm 46 22.5 16 48.5 23 18 Lectotype of L. jonesi Vander Schalie Allotype Remarks. Lampsilis jonesi Vander Schalie of the Choctawhat- chee River system has been confused with Lampsilis australis Simpson, a species which occurs in the Choctawhatchee and Es- cambia drainage systems. The posterior ridge of L. jonesi is double, often with a distinct tertiary ridge above. The extremities of the ridges form a scalloped edge posteriorly, a unique character. The posterior ridge of australis is poorly defined and ends in a blunt point. The periostracum of jonesi is somewhat coarse, while that of australis is often smooth and glossy. L. jonesi is somewhat cylin- drically shaped, australis is more elliptical and less inflated. Females of jonesi exhibit a swelling of the disc below the posterior ridge, rendering the ventral margin slightly angular, while those of australis only show a greater rounding of the ventral margin. Vander Schalie compared jonesi to Lampsilis subangulata (Lea) which he said occurred with it in the Choctawhatchee River system, but these were actually specimens of australis since subangulata is restricted to the Apalachicola and Ochlockonee River systems. The two records of Clench and Turner (1956:198) from the Choctawhatchee River system listed as subangulata are australis. Lampsilis subangulata and australis are allopatric. Lampsilis subangulata is generally more attentuate and pointed posteriorly, more inflated, with a sharper posterior ridge and a more excavated posterior slope. The surface is more highly polished and the green rays are sharper and brighter. Vander Schalie did not specifically designate a holotype for L. jonesi though he makes it clear that it is in the Alabama Museum 1967 GULF DRAINAGE UNIONIDS 13 of Natural History. It is assumed that he intended the male specimen shown on his plate 15, figures la, lb, to be the holotype. It is here selected as the lectotype. Range. Gulf drainage, restricted to the Choctawhatchee River system of Alabama and Florida. SPECIMENS EXAMINED CHOCTAWHATCHEE RIVER SYSTEM Pea River Drainage. — Alabama: Pea River, Elamville, Bar- bour Co. Pea River, Priston's Mill. Dale Co. Pea River, Flemings Mill, Coffee Co. Choctawhatchee River Drainage. — Alabama: East Fork Choc- tawhatchee River, Midland City, Dale Co. East Fork Choctawhat- chee River, 8 mi. W Abbeville, Henry Co. West Fork Chocta- whatchee River, 7 mi. SE Ozark; Choctawhatchee River. 1 mi. N Newton; both Dale Co. Lampsilis HADDLETONi Athcam Plate 3, figure 3 Lampsilis haddletoni Athearn 1964, Nautilus. 77: 135. pi. 9, figs. g. h. (Choctawhatchee River, West Fork. 7 mi. SW Ozark, Dale Co.. Ala- bama; holotype. National Museum of Canada 20095 ). Remarks. Lampsilis haddletoni Athearn is known only from the two specimens collected at the type locality. It "appears some- what similar to L. ochracea (Say) |of the Atlantic drainage]. It is smaller than that species, less inflated, the shell thicker, and the cardinal teeth are much larger. The species differs from Villosa choctawensis Athearn by its orbicular outline, flesh colored nacre and in particular in the coloration of the epidermis [periostracum]. The I periostracum I of L. haddletoni is darker anteriorly on the disc while it is lighter near the ventral margin and on the posterior slope. Rays are prominent only on the posterior slope. The [periostracum] of adult V. choctawensis is dark throughout except in the vicinity of the umbones where it appears lighter and exhibits fine rays." (Athearn) Villosa choctawensis Athearn Plate 3, figure 4 Villosa choctawensis Athearn 1964, Nautilus, 77: 137, pi. 9, figs, c, d. e, f. (Choctawhatchee River, 2 mi. SW Caryville; about 1 mi. downstream from U. S. Highway 90. Holmes Co., Florida; holotype National Museum of Canada 20096). Pleiirobema strodeanuin ( B. H. Wright), parfini. Clench and Turner, 1956, Bull. Florida State Mus.. 1: 161. 14 BREVIORA No. 270 Remarks. "Villosa choctawensis has probably been m'staken for Pleurobcma .siroclccuium B. H. Wright by collectors in the past. The males of V . choctawensis are strikingly similar to that species. However, they lack the consistently well defined posterior ridge of P. strodeanuni. The male and female shells of P. strodeanum are essentially alike as is characteristic of shells of the genus Plcuro- bema. Sexual dimorphism is well defined in specimens of V. choctawensis. "The nacre of the posterior area of V. choctawensis lacks most of the bluish appearance found in that area of P. strodeanum. V . choctawensis is in some respects similar to V. villosa B. H. Wright but is much shorter than that species." (Athearn) The records beiow are supplemental to those given by Athearn. Range. Gu'f drainage, restricted to tlie Choctawhatchee River system. SPECIMENS EXAMINED CHOCTAWHATCHEE RIVER SYSTEM Pea River Drainage. — Alabama: Pea River, Vi mi. SW Geneva, Geneva Co. Choctawhatchee River Drainage. — Florida: Choctawhatchee River, 8 mi. W Miller's Cross Roads, Holmes Co. LITERATURE CITED Athearn. Herbert D. 1964. Three new unionids from Alabama and Florida and a note on Lampsilis jonesi. Nautilus, 77: 134-139. pi. 9. Clench, William J. and Ruth D. Turner 1956. Freshwater mollusks of Alabama. Georgia, and Florida from the Escambia to the Suwannee River. Bull. Florida State Mus., 1: 97-239, pis. 1-9. Frierson, Lorraine S. 1927. A classified and annotated check list of the North American Naiades. Bavlor Univ. Press, Waco, Texas, 111 pp. (errata et corrigenda. 1 p. ). Johnson, Richard I. 1965. A hitherto overlooked AnoiloiiUi (Mollusca; Unionidae) from the Gulf drainage of Florida. Breviora. Mus. Comp. Zool., No. 213: 1-7. pis. 1-2. Ortmann, Arnold E. 1923. The anatomy and taxonomy of certain Unionidae and Anjdon- tinae from the Gulf drainage. Nautilus, 36: 130-132. 1967 GULF DRAINAGE UNIONIDS 15 Vander Schalie, Henry 1940. The naiad fauna of the Chipola River, in northwestern Florida. Lloydia. 3: 191-208, pis. 1-3. Wheeler, Harry E. 1935. Timothy Abbott Conrad with particular reference to his work in Alabama one hundred years ago. Bull. Amer. Paleont.. 23(77): 1-157, 27 pis. 16 BREVIORA No. 270 PLATE 1 Fig. I, Strophitiis stihve.xii.s (Conrad). Black Warrior River, Tuscaloosa Co., Alabama. Topotype MCZ 146655. Length 59, height 36, width 25 mm (nat. size). Fig. 2, Srrophitus siihrcxns (Conrad). |Magbys Creek of] Tombigbee River, near Columbus | Lowndes Co.] Mississippi. Holotype USNM 86278 of Mari^aritaiui spillnuinii Lea. Length 93, height 51. width 40 mm (slightly reduced ). Fig. 3, Stropliiiiis suhvcxus (Conrad). Cowarts Creek, nr. Florida state line, Houston Co., Florida. MZUM 138467. Length 86, height 47. width 37 mm (slightly reduced). 1967 GULF DRAINAGE UNIONIDS 17 18 BREVIORA No. 270 PLATE 2 Fig. 1, Anodontoides rcidiutiis (Conrad). [Tributary of Black Warrior River, Greene Co., Alabama]. Holotype ANSP 41147. Length 64, height 36, width 22.5 mm (slightly reduced). Fig. 2, Anodontoides radiotus (Conrad). Chattahoochee River [below Uchee Bar] near Columbus [Muscogee Co.] Georgia. Holotype USNM 86257 of Marf^ctrilaiKi clliottii Lea. Length 45, height 26, width 18 mm (slightly reduced). Fig. 3, Anodontoides radiatiis (Conrad). Tombigbee River, Columbus [Lowndes Co.] Mississippi. Holotype USNM 86258 of Margarilmui ellip- tica Lea. Length 50, height 29, width 19 mm (slightly reduced). Fig. 4, Anodontoides radiutiis (Conrad). Coosa River, Wetumpka ] El- more Co.] Alabama. Holotype USNM 86487 of Anodonta showalterii Lea. Length 72, height 43, width 31 mm (slightly reduced). Fig. 5, Alasmidonta wrightiaiui (Walker). Ochlockonee River, Florida. Holotype MZUM 74938. Length 54, height 38. width 31.5 mm (nat. size). Fig. 6, Lampsilis hinoni'natns Simpson. Chattahoochee River, Columbus (Muscogee Co.] Georgia. Holotype USNM 84884 of Unio lineatns Lea. Length 33. height 33, width 16 mm. female (slightly reduced). 967 GULF DRAINAGE UNIONIDS 19 20 BREVIORA No. 270 PLATE 3 Fig. 1, Ohovaria rotitlata (B. H. Wright). Escambia River, Escambia Co., Florida. Holotype USNM 159969. Length 48, height 45, width 32 mm (nat. size). Fig. 2, Lampsilis binominatus Simpson. Line Creek, Vi mi. W. Digbey, Spaulding Co., Georgia. Collection of H. D. Athearn. Length 58, height 41, width 32 mm, male (slightly reduced). Fig. 3, Lampsilis haddletoni Athearn. Choctawhatchee River, West Fork, 7 mi. SW Ozark, Dale Co., Alabama. Holotype National Museum of Can- ada 20095. Length 30, height 23, width 12.5 mm (slightly enlarged). Fig. 4. Villosa choctawensis Athearn. Choctawhatchee River, 2 mi. SW Caryville; about I mi. downstream from U. S. Highway 90, Holmes Co., Florida. Holotype National Museum of Canada 20096. Length 37, height 26, width 16.5 mm, male (slightly enlarged). Fig. 5, Lampsilis jonesi Vander Schalie. Pea River, near Elamville, Bar- bour Co., Alabama. While not exactly from the type locality, this specimen was received from Vander Schalie labeled as a paratype, MCZ 98802. Length 48, height 24, width 16 mm, female (slightly enlarged). 967 GULF DRAINAGE UNIONIDS 21 BREVIORA MmseMinti of Comparative Zoology Cambridge, Mass. 17 November, 1967 Number 271 CHELONIA DEPRESSA GARMAN RE-INVESTIGATED Ernest E. Williams, Alice G. C. Grandison,' and Archie F. Carr, h.^ The pantropical sea turtle genus Chelonia is characterized by the tendency to mass at restricted breeding places — islands or bits of shore — where copulation as well as nesting takes place. Desul- tory nesting by females going separately ashore occurs in inter- vening regions, but by far the greater part of the reproduction of the genus takes place in aggregations. Results of extensive recent tagging studies have revealed that members of a nesting as- semblage may converge from several different year-around feeding grounds, some of which may lie as far as a thousand miles away. This site tenacity in breeding is so strong that one wonders how new nesting colonies are ever established. The genus may there- fore be strongly fragmented genetically, and nascent species, or even quite genuine species, may currently be confused under the collective name Chelonia mydas. The surprising thing is that this isolation has produced so little well-marked morphologic di- vergence. One example of the degree of differentiation that occurs is shown by Carr and Hirth (1962). but material representing the numerous isolated breeding populations of the genus is still too meagre to support an extensive revision of the group. One local population, however, is morphologically so distinct that it may be tentatively regarded as a species, even though only about 40 speci- mens are at present known to have been preserved in museums. This sharply distinguished taxon is Chelonia depressa Carman, described in 1880. The type of C. depressa is a mounted adult male in the Museum of Comparative Zoology (MCZ 4473), with the locality "North Australia." It is a flat-shelled turtle with short flippers, and with ^ British Museum (Natural History), London S.W. 7, England. - Department of Zoology, University of Florida, Gainesville, Florida. 2 BREVIORA No. 271 only three postoculars on either side. The squamation of both tore and hind limbs is notably weak. In superficial appearance this could be equally well a freakish specimen of Chclonia mydas or some sort of Lepidochelys. The type is still one of the few adults known. However, our own examination of the type and of hatchlings and young turtles from the vicinity of the type locality and data generously provided for specimens in Australian museums (see Table 1 ) leave little doubt that there exists a distinct north and northeast Australian popula- tion of Chelonia, clearly differentiated from other known forms of the genus, for which the name dcprcssa is available. Boulenger in 1889 placed C. dcprcssa in the synonymy of C. nixdas, although he had one juvenile specimen of the Carman species at hand, in the British Museum collections. A year later, believing that Boulenger was in error, Baur proposed elevating dcprcssa to generic rank on the basis of trivial features of the lower jaw seen in the type. Baur's opinion was generally ignored, and C. inxdas and C. dcprcssa have almost universally continued to be regarded as conspecific. Probably the principal reason for the rejection of Carman's species by the majority of herpetologists has been the absence from most institutions of any material of the form. Even the careful re-appraisal of the characters of dcprcssa by D. B. Fry in 1913. on the basis of seven specimens from north- ern Australia and adjacent New Cuinea waters, did not carry the day. The issue may have been somewhat clouded by McCulloch's (1908) attempt to erect a new genus and species {Natator tessel- latus) for a specimen which Fry was able to show was clearly identical with dcprcssa. However, it was also unfortunate that Fry, after providing a superlative case for the distinctness of dcprcssa on the basis of external characters, went on to place emphasis on certain skull characters (he had only one subadult skull of dc- prcssa) which are subject to considerable individual variation. Skepticism was only reinforced when Barbour (1914) revealed that Carman's own original series was composite and that the young specimen (MCZ 1413) was a typical mydas according to Fry's newly provided characters. The down-grading of Fry's study — much more than the neglect of Carman's inadequate description — was truly unfortunate. More than 50 years later we can add no new really useful charac- ters to those offered by Fry. Most of Fry's characters are indeed subject to some individual variation and others are difficult to state objectively, but in aggregate thev — at least those based on externals — build a clear diagnosis of dcprcssa and suggest that 1967 CHELONIA DEPRESSA CARMAN 3 it may be the most trenchantly ditTerentiated local population of any of the five sea turtle genera anywhere in the world. We have checked 1 5 characters in the material examined. We list these characters below, with the condition characteristic of depressa in parenthesis in each case. 1. Shape of the carapace (oval in depressa). 2. Contact of first vertebral and first marginal (rarely present in depressa). 3. Postanal plastral scutes (frequent in depressa). 4. Brachial plates (often absent in depressa). 5. Shields between first inframarginals and humeral (one only in depressa). 6. Relative length of pastral plates (femoral longest in de- pressa ) . 7. Size of forelimb (short in depressa). 8. Scales of forelimb (wrinkled skin overlying phalanges in depressa ) . 9. Scales of hind limb (wrinkled skin overlying phalanges in depressa ) . 10. Number of postoculars (3 in depressa). 11. Prefrontal length relative to supraocular (equal or less in depressa ) . 12. Contact of prefrontal and maxillary sheath (no or very lim- ited contact in depressa). 13. Number of postparietals ( 1 or 3 in depressa. not symmet- rically divided). 14. Upper eyelid scales (scales uniformly small in depressa). 15. Number of scales posterior to postoculars (subtcmporals) (numerous in depressa). Each of these characters is of some use for the recognition of depressa but most are not key characters. It is clear from the statement of some of them that there is individual variability (e.g. marginal vertebral contact, postanals). Others, such as the shape of the carapace or the size of the forelimbs, are relatively diflficult to make objective since allometry will modify the condition at various sizes. Still others (e.g. upper eyelid scales) are obvious when comparative material is available but in the absence of readily repeatable counts are sometimes equivocal. Two characters in combination, however, do appear absolutely and unequivocally to define depressa as against other Green Turtles anywhere. These are the postocular scale count of 3. plus the areas of wrinkled skin distally on fore and hind limbs. In BREVIORA No. 271 reality, the last character may be sufficient, since we know of no overlap or approach in this feature at all. There is, unhappily, rare overlap in the postocular count, al- though this also is usually diagnostic. All known depressa have 3 postoculars. In all other populations sampled by us and in material available in the British Museum, the Museum of Com- parative Zoology, the United States National Museum, and the American Museum of Natural History a count of 3 postoculars on each side, without evidence of aberrant fusion with other scales, is very rare indeed, almost inconsequential (but see below). The extreme infrequency of 3 postoculars in non-depressa popu- lations is shown with great clarity in Table 2, in which counts from 3,000 specimens from one East Pacific and two Atlantic lo- calities are given; at least one other aberrant count in the table is a more frequent variation than the 3-3 count. Only one British Museum specimen, one from the Great Barrier Reef, is equivocal in these two crucial characters. It has postocular counts of 3-4 and counts across the forelimb of 6-7. It seems, however, on most characters to be nndas, and the three other members of the small series collected at the same time are unquestionable mydas, though one other has a 3-3 postocular count. (One Museum of Compara- tive Zoology specimen | MCZ 947 1 | from Murray Id., Torres Strait, likewise has a 3-3 postocular count.) Figure 1. Forelimbs of Cheloiiia mydas (A) and Clwlonid (U'[)rc.yui (B) compared to show suggested places for counting scales across limb. (Modi- fied from Fry, 1913.)"^ 1967 CHELONIA DEPRESSA GARMAN 5 The disturbing element here is not the occasional overlap in characters but that the overlap occurs in the relatively small sample we have from the known range of depressa. The dearth of information on the distribution of breeding grounds of Clwlonici in Australia makes it difficult to judge whether, and to what de- gree, depressa and mydas may nest sympatrically. There is an important nesting colony of mydas-Wkt turtles on Capricorn Reef at the southern end of the Great Barrier Reef — especially on Heron Island and Northwest Island, at 23)°2)0'S (Moorhouse, 1933). The islands of Torres Strait are. or used to be. heavily used for nesting by one or more forms of Chelonki, and Stokes (1846) reported abundant nesting both in the Gulf of Carpentaria and at Baron Island on the western Australian coast, at 20°45'S, 11530' W. In none of these cases is it known whether the turtle involved is mydas or depressa} The presence of the areas of wrinkled skin on the flippers of depressa is an even better method of diagnosing depressa than the three postoculars. The condition can clearly be shown in a compar- ative figure (such as Fry, 1913. fig. 46 a and b. and our photo- graphs). However, a simple count will solve the problems of those who may feel, in the absence of actual comparative ma- terial, that they have an ambiguous or intermediate condition in front of them. A count of scales across the dorsal surface of the forelimb. two scales distal to the claw, will in C/wlonia nixdas 1 Harold Cogger (Australian Museum) writes in a letter of March 9, 1967, "I'm afraid that I can give you little information re breeding sites of this species or of C. ?nyclas along the north coast of Australia. C. depressa certainly breeds in many areas, and at the moment David Lindner, who collected our adult, currently has several depressa nests under observation at Port Essington. Despite frequent reports of mydas breeding in the area, we have no authentic records. As you know we have no specimens of mydas sympatric with our dozen or so specimens of depressa. I have writ- ten to David Lindner asking him to obtain all possible information on the occurrence of mxdas in the area." For the northeast coast of Australia — the other portion of the known range of C. depressa — J. T. Woods, Director of the Queensland Museum, reports sympatry between C mydas and C. depressa over a stretch between Townsville and Heron Island (the latter approximately opposite Keppel Bay) but no overlapping records further north. Data that Mr. Woods has generously provided indicate that the eastern Queensland specimens of C. mydas adjacent to C. depressa localities are quite typical. However, since breeding range and residence range are for these turtles usually quite dis- tinct, further information is much to be desired. 6 BREVIORA No. 271 almost always be 4 or 5, while in Chelonia clepre.s.sa the same count (including always the small weak scales between the larger ones) will be 7 to 10. Most of the characters of depressa involve what appears to be a general weakening or reduction of the scalation. Besides the fewer postoculars, the lateral temporals and brachials are reduced in number, and the scales of the eyelid and forelimb are reduced in size. Series of specimens of depressa of several size groups will be required for final evaluation of certain of the other dif- ferentiating characters; for example, the broader and shorter fore flipper, the greater width of the head and the special texture of the areolar areas in the laminae of the hatchling. A survey of the breeding and feeding range of depressa will reveal aspects of its ecology that will surely help clarify the status and relationships of the population. The short foreleg and broad skull, for example, if fully verified, may prove to be features of a carnivorous forager, like Caret ta and Lepidochelys which poke about rocky places preying on crabs and mollusks, and hence do not make regular journeys between underwater pastures of turtle grass or algae and sleeping places and nesting beaches as typical populations of Chelonia do. That depressa may be a carnivorous Chelonia is also suggested by Fry's citing (1913:165) of the opinion of Hugh Christie that its meat is disagreeable in taste. However, it should be noted that Christie insisted that, "C. depressa is purely a vegetarian as far as my observations go. ..." As Fry stated (1913, footnote, p. 165), the disagreeable taste may well have kept depressa out of turtle markets and thus out of the sight of zoologists. Green turtles with inferior eating qualities turn up in other places too, and these seem consistently to have a tendency to feed on invertebrates instead of plants. While the dietary divergence is, in other populations, not correlated with strong morphological differentiation, this may simply mean that depressa, with its putatively peculiar feeding regimen, has been isolated for a longer period of time than they. *Studies of the feeding habitats and habits of depressa as well as of its sexual cycle and nesting ethology are needed; this work when done will greatly augment our understanding of a long neglected species. ACKNOWLEDGMENTS For the opportunity to examine material or for critical infor- mation on specimens not borrowed, we are indebted to L. D. 1967 CHELONIA DEPRESSA CARMAN 7 Brongersma (Leiden Museum), R. Inger (Field Museum. Chi- cago), H. Cogger (Australian Museum, Sydney), J. L. Woods (Queensland Museum). W. D. L. Ride. G. M. Storr (Western Australian Museum). M. Tyler (South Australian Museum). Data on Atlantic and Caribbean Clielonici were obtained with the support of the National Science Foundation, the Office of Naval Research, and the Caribbean Conservation Corporation. TABLE 1 The known specimens of Chelonia depressa^ '■'I. Type, MCZ 4473, "North Australia" — adult *2. BM 87-5.16.3, "N. W. Australia" — juvenile "3. FMNH 97086, Green Island near Cairns, Great Barrier Reef, Queensland, Australia — juvenile *4. MCZ 54713 = J 1111, Queensland Mus., no data — juvenile — not seen by Fry but the fourth Queensland Museum speci- men cited by him •=5. MCZ 54714 - J 3065, Queensland Mus., Cape Bowling Green, mid-East Queensland — juvenile *6. Type, Natator tessellatiis — Port Darwin, Australian Mus. — Northern Territory. Australia = Fry specimen No. 6 — juve- nile *7. Australian Mus. — Port Darwin, Northern Territory, Australia = Fry specimen No. 1 — juvenile "8. Australian Mus. — Torres Strait = Fry specimen No. 4 — juvenile 9. .Australian Mus. — Murray Island = Fry specimen No. 7, Torres Strait (skull removed, figured by Fry) — half grown 10-11-12, Queensland Mus., J 184, J 185 (2), no data — not "New Guinea?" as stated by Fry = Fry specimens Nos. 2, 3, 5 — destroyed since Frv"s time because of poor condition — juve- niles 13-14. Queensland Mus., J 1109, J 1110, Keppel Bay. mid-East Queensland — seen by Ogilby — "cannot be located in the present collections" — juveniles 15-16-17. Queensland Mus.. J 3066-68, Cape Bowling Green. mid-East Queensland — -juveniles 18-19-20. Queensland Mus., J 8551 (cast only). J 8575, J 8577, Thursday Island, Torres Strait — adult '^'Specimens examined by E. E. Williams 1 MCZ = Museum of Comparative Zoology; BM = British Museum (Natural History); FMNH = Field Museum of Natural History. 8 BREVIORA No. 271 TABLE 1 (Cont.) *21-30. Australian Mus., R.81 15-16, 8909-13, 9012, Sir Edward Pellew Group, Northern Territory — juveniles *31. Australian Mus.. R.I 1756, Bathurst Island, Northern Territory — juveniles -32. Australian Mus., R. 25691, Cape Don, Northern Territory — juvenile 33. Western Australian Museum, R 773, Cape Don, Northern Territory — adult *34-37. Western Australian Museum. R 682-85, "N W Coast" — juveniles *38-39. Western Australian Museum, 10423-24, "from the Northwest" — ^ juveniles 40. Australian Mus., R. 26347, Port Essington near Cape Don, Northern Territory — adult 1967 CHELONIA DEPRESSA GARMAN CN CO < c -a .2 ^ C y: < ^ ■(— > *^ si 3 DO lU 3 J= O &0 >^ E ? o U Oh o c o .3 cs 3 cj cd !;! £ ■« 3 ^^ OX) ■-*=; CQ C O o •-^ CJ >-, O (u y =- "" S 3^ ?j oj -a 03 u::5 ^ t w-1 I I I I I 5J I r4 :s (^) ON o •<^ oo (N VO o >o (N ojsnSnjJox «s (N puEisj uoisuassy (3961 'IRAVPIBO lUOJJ BlBp) Biujojii^D i^fBa JO j[no 10 BREVIORA No. 271 PAPERS CITED Barbour, T. 1914. On some Auslralasian reptiles. Pioc. Bidl. Soc. Washington. 27: 201-206. Baur. G. 1890. The genera of the Cheloniidae. Amer. Nat., 24: 486-487. BOULENGER, G. 1889. Catalogue of the chelonians, rhynchocephalians and crocodiles in the British Museum (Natural History). London, x + 3 1 1 pp. Caldwell. D. K. 1962. Sea turtles in Baja Californian waters (with special reference to those of the Gulf of California) and the description of a new subspecies of northeastern Pacific Green Turtle. Los Angeles County Mus. Contrib. Sci.. No. 61 : 1-31. Carr, Archie and Harold Hirth 1962. The ecology and migrations of sea turtles. ?. Comparative features of isolated Green Turtle colonies. Amer. Mus. Novit., No. 2091: 1-42. Fry. D. B. 1913. On the status of Clwlonlci clcprcssci Garman. Rec. Austral. Mus.. 10: 159-185. Garman, S. 1880. On certain species of Chelonioidae. Bull. Mus. Comp. Zool.. 6: 123-126. McCulloch, a. R. 1908. A new genus and species of turtle from North Australia. Rec. Austral. Mus.. 7: 126-128. MOORHOUSE, F. W. 1933. Notes on the Green Turtle (Chclonia Diydas). Great Barrier Reef Committee Reports, Brisbane, vol. I, pt. 1. 23 pp. Stokes, J. L. 1846. Discoveries in Australia; with an account of the coasts and rivers explored and surveyed during the voyage of H.M.S. Beagle. London, 2 vols. (Received 6 January 1967.) 967 CHELONIA DEPRESSA GARMAN 11 Figure 2. Distribution of Chelonia deprcssa in Australian waters. 12 BREVIORA No. 271 Plate 1. Clielonia dcpres.sa, the British Museum hatchhng. Top: dorsal view of head to show asymmetrical division of postparietal, and prefrontals equal in length to supraoculars. Bottom: dorsal view of body to show rela- tive shortness of forelimbs and their weak scalation. 1967 CHELONIA DEPRESSA CARMAN 13 14 BREVIORA No. 271 Plate 2. Clielonia depressa, the British Museum hatchling. Top: lateral view of head to show three postoculars and small number of subtemporal scales. Bottom: ventral view of body to show single brachial scale and weakness of scalation on forelimbs. !967 CHELONIA DEPRESSA CARMAN 15 BREVIORA Museiiijm of Comparative Zoology Cambridge, Mass. 17 November, 1967 Number 272 THE ECOLOGICAL DISTRIBUTION OF THE ANOLINE LIZARDS AROUND KINGSTON, JAMAICA A. Stanley Rand ' In each of the Greater Antilles and the islands of the Great Bahama Bank it is usual to find several species of Anolis very common in a small area. Though taken in the same habitat or even from the same tree, the species differ in where they live within the habitat. Differences in microhabitat have been de- scribed for the four species on Bimini by Oliver (1948); on Cuba for five species in Camaguey and Oriente by Ruibal (1961), and for five species near Havana by Collette (1961); for seven of the ten species on Puerto Rico by myself (Rand, 1964); and for the three commonest species on Hispaniola (Rand, 1962). No comparable study has been published for Jamaica, though the sys- tematics of the anoline lizards have been described by Underwood and Williams (1959), and some notes on their ecology given by Grant (1940) and Underwood (1951). This paper attempts to fill this gap in our knowledge by presenting data on the dif- ferences in microhabitat among five Jamaican anoles. I restrict myself to the anoles near the city of Kingston because micro- habitats do differ geographically in Jamaican anoles — particularly in A nolis lineatopus. THE JAMAICAN ANOLES Seven species of the genus Anolis are known from Jamaica: Anolis valencienni, gannani, opalinus, grahami, lineatopus, recon- ditus, and sagrei. The last of these, sagrei, occurs widely on Cuba and the Bahamas and may have been introduced by man into Jamaica. It is restricted to the west end of the island. Anolis reconditus is known from only a few specimens from the Blue 1 Smithsonian Tropical Research Institute, Box 2072 , Balboa, Canal Zone 2 BREVIORA No. 272 Mountains and is closely related to lineatopus. The remaining five species are widespread on the island. According to Etheridge (1960), all five are quite closely related, and four of them, gar- mani, opalinus, grahami, and lineatopus, are more closely related to one another than they are to species occurring elsewhere. Among these four, opalinus and grahami are particularly close (Underwood and Williams, 1959). Except for sagrei, only grahami occurs outside of Jamaica. It has been introduced into Bermuda, and a subspecies or related species occurs on Grand Cayman. I have followed the nomenclature of Underwood and Williams except in considering valencienni a member of the genus Anolis rather than of Xiphocercus; in this I follow Etheridge cited in Williams (1962). METHODS The observations, summarized, on the ecological distribution of the five species of Anolis occurring in the vicinity of Kingston were made during a ten month stay. Kingston, on the south coast of Jamaica, in the rain shadow of the Blue Mountains, has a climate which is tropical and semi- arid. Probably completely covered by forest or scrub originally (Asprey and Robbins, 1953), there are now no completely un- disturbed areas in the vicinity of the city. Present habitats range from moderately tall forest through dense, thorny second growth to open grassy pastures, and include a variety of edificarian situa- tions, including well-watered gardens and parks. The distributions of the Anolis are not uniform through these habitats. Each species occupies a different microhabitat though these overlap widely and in some places all five species are pres- ent. I have seen four diflferent species in one tree at the same time and individuals of two species within a few inches of one another. This overlap between species is striking and tends to obscure differences in their ecological distribution. The microhabitats of species around Kingston differ in at least two ways: in the sorts of perches which they use, and in the sorts of habitats where they are common. To show these differences, censuses were run in four diflferent areas selected to cover the most frequent habitats around King- ston. Two of these areas were much modified by man, the other two less so. 1967 ECOLOGY OF JAMAICAN ANOLES 3 The first area, "Barbican, open," was in a new residential suburb north of Kingston. This had recently been cleared of thorn scrub second growth and built up to new houses with moderate to extensive grounds. There were some small trees but few large ones, many fences, hedgerows, grassy lawns, and buildings. It was the most open area censused. The second area, "Mona, park," was part of the grounds of the University of the West Indies at Mona, east of Kingston. This was park-like with quite well-kept lawns and many larger trees, most of these scattered and isolated from one another but some growing in small groves. There were not many smaller trees and relatively few bushes. The third area, "Barbican, bush," was dense thorny scrub sec- ond growth joining the first area. There were scattered larger trees but most were only 15-30 feet tall with much undergrowth and many tangled vines. The ground was covered with dry leaf litter. The fourth area, "Mona, bush," was taller forest at the base of Long Mountain near the second area. This was also second growth, though older than that at Barbican, and may never have been clean felled. There were many tall trees, 40-60 feet, form- ing a nearly complete canopy, and the undergrowth, though quite dense, was more open than in the Barbican second growth and in one place had been cleared completely leaving only the larger trees. There were many vines and a litter of dead leaves on the ground. Each area was carefully searched repeatedly, and each lizard recorded where it was first seen. No area was censused more than once a day and a lizard was only counted once per census even if it changed its position radically. Even with binoculars, many of the lizards present in an area were not seen, and those seen are not a random sample of those present. In all species the males are larger and more con- spicuous than the females. Certain species are more easily seen than others. The large green garmani is more conspicuous than any other; the slow-moving grey valencienni is the least con- spicuous. A lizard on a bare tree trunk at eye level is not apt to be overlooked, but one on top of a branch high overhead fre- quently escapes notice. The further up a lizard is, and the further out among the foliage, the less likely it is to be seen. Because of this bias, the census results are not fully valid representations of the distribution of the lizards, but since this bias acts equally for each species, the differences between them should be real. This 4 BREVIORA No. 272 censusing technique and its weaknesses have been discussed more fully in my paper (Rand, 1964) on the Anolis of Puerto Rico. Of the various possible characteristics of the lizards' perches, two were selected for analysis: the height above the ground at which each lizard was seen and the diameter of its perch. The data on height of perch and perch diameter did not seem to vary significantly from habitat to habitat and these have been combined in Table 1. The statistical significance of the differ- ences in perch height is given in Table 2. The abundance of the various species in the difl'erent habitats is given in Table 3. In the following accounts of the species, the census data are supplemented by a summary of more casual observations on these species, both within and outside of the census areas. OBSERVATIONS Anolis lineatopus lineatopus This form is moderate in size (males reaching 70 mm snout- vent length) and a strongly patterned brown in color. This is probably the most abundant species in the Kingston area and certainly the most commonly seen. Anolis lineatopus occurs in the widest variety of habitats of all of the Kingston anoles. As Table 3 shows, it was common in all four census areas, but least common in the Mona bush. It was also common in old shady gardens, in fencerows through grazed pastures, and was the only species seen in the edge of the man- grove swamps on Palisadoes. It was quite commonly seen on porches and verandas and sometimes entered houses. Like the other Anolis, lineatopus spends most of its waking time sitting on a perch. Unlike the others, it is commonly seen on a perch where it could not climb higher than a couple of feet. Individuals were seen on trees, on fence posts, on stone walls, on rocks both large and small, on ruined houses and on the walls of occupied dwellings, as well as in hedges and bushes. Of all the species it was most frequently seen on the ground (6 during cen- sus), though individuals spent only short periods there. Most Anolis lineatopus recorded during censuses were close to the ground and on moderate to large diameter perches (Table 1). Seventy-six per cent were below 6 feet and about 83 per cent on perches larger than V2 inch. Large individuals (adult males) usually perched somewhat higher (only 26 7^ below 3 feet) than did smaller individuals (adult females and sub-adults, 60% below 3 feet). This difference is significant at the 1 per cent level 1967 ECOLOGY OF JAMAICAN ANOLES 5 (Table 3). The maximum height at which a large individual was seen during census was 9 feet. Individuals were rarely seen higher than this at any time. Larger lineatopus were less fre- quently seen on small diameter perches (13% on perches less than Vi inch) than were small ones (207^ on perches less than V2 inch). Very young lineatopus seemed to avoid larger trees and were usually seen close to the ground in bushes, hedges, brush piles, or sometimes perching on large dead leaves on the ground. Anolis grahami grahami This species is of moderate size (males reaching 70 to 75 nam snout- vent length), predominantly light green in color but can change to dark brown. There is also a common finely spotted phase. Where it occurs, grahami may rival or even surpass lineatopus in abundance. However, as Table 3 shows, it is common only in the more open areas both at Mona and in Barbican, and quite rare in the bushy or forested areas. It occurs in grazed pastures. It was also common in the shady, well-watered gardens around Kingston as well as in the newer more open ones, and it was this species which most commonly entered houses, climbing on the walls and hiding behind pictures, though usually venturing outside the house during the day. This species is most frequent on trees and walls and, when seen on fence posts and stones and in brush heaps, there is usu- ally the opportunity to climb to a fair height. Individuals, par- ticularly smaller ones, were not infrequently seen in hedges. Individuals were seldom seen on the ground and none during census taking. Most Anolis grahami (Table 1) recorded during census were well above the ground and on moderate to large diameter perches. Sixty-six per cent were above 5 feet and 92 per cent on perches of a diameter greater than Vi inch. Large individuals (adult males) and smaller ones (adult females and sub-adults) show litde difi'erence (not significant at the 5% level. Table 2) in the height at which they were seen, though there is a curious and unexplained small number of smaller lizards seen between 6 and 10 feet. There is some tendency for smaller in- dividuals to be seen on small diameter perches more frequently than are larger individuals. The maximum height at which large individuals were seen during census was 25 feet, for smaller individuals 30 feet. Such heights were not infrequently seen during casual observations. 6 BREVIORA No. 272 Very young grahami were seldom seen and usually were up well above the ground. Anolis opalinus This is a small species (adult males reaching between 45 and 50 mm snout-vent length), grey or brown in color, usually reticu- lated, with a prominent light lateral stripe. Though quite common in the habitats which it occupies, it is much less conspicuous than either grahami or lineatopus, in part because of its small size and drab coloration and in part because it is less common around houses. Anolis opalinus is primarily a lizard of bushy or forested areas. Grant reports that it is called "coffee lizard" for this reason. As Table 2 shows, it was completely absent from the residential area in Barbican and rarely seen in the park-like area in Mona. These census results are substantiated by other observations, and this species was never seen around the houses in Mona and Bar- bican and only rarely on the University of the West Indies campus at Mona, and then only in one group of tall trees, with dense foliage, grown close together and close to the forested area. It was quite common in both bushy study areas, less in the dense second growth at Barbican than in the more open and taller forest at Mona. This species was never seen in open pasture situations around Kingston and curiously enough it was rare or absent in the old, well shaded gardens in the residential areas we examined. Around Kingston, this species was almost always seen on trees or bushes, though sometimes on hanging vines. It was seldom seen on rocks, logs, or stumps even in the areas where it was quite common. It was also seldom seen on the ground and never so during censusing. Most opalinus recorded during census were well above the ground and on moderate to large diameter perches (Table 1). Only 11 per cent were recorded below 3 feet and about 63 per cent were recorded above 5 feet. About 94 per cent were on perches of a diameter greater than Vi inch. Large individuals (adult males) and smaller ones (adult females and sub-adults) show somewhat similar distributions in height, though the dif- ference is significant at the 1 per cent level (Table 2). More of the smaller individuals were seen on small diameter perches (11% on perches less than Vi inch) and were less frequently seen more than 6 feet above the ground. No very young individuals were seen close to the ground. 1967 ECOLOGY OF JAMAICAN ANOLES 7 Anolis garmani This species is the largest of the Jamaican Anolis (males reach 120 mm in snout- vent length). It is usually bright green in color but can change to dark brown. In the Kingston area this species is nowhere near as abundant as the three preceding species. An hour or two of searching was usually required to find even a single individual, and during most census periods none were seen at all. As Table 2 suggests, this species was most common in the bushy or forested areas, and it was seen during census in neither the open residential area at Barbican nor the park area at Mona. However, during casual observation, occasional individuals were seen in both these latter areas. Individuals of this species were never seen on isolated trees in open pastures, even on large trees. They were, however, seen occasionally in the older, shady, well-established gardens in Mona. Though occasionally seen in bushes or vine tangles, and once on a telephone pole, most individuals were seen on large trees. Though they do occur in gardens occasionally, none were seen to enter houses or on stone walls or ruins. None were seen on the ground. Data from individuals collected to determine temperatures have been added to those collected during censuses, in Table 1. The data are still meagre but suggest that garmani occurs primarily high in the trees, as 61 per cent of the records are over 6 feet and 44 per cent over 10 feet. All individuals seen were on moderate to large perches. These data accord with general observations that this species is one which lives well above the ground. No very young individuals were seen. Anolis valencienni This species is second in size only to Anolis garmani (adult males reach about 80 mm snout- vent length). It is a mottled grey in color. Though seldom seen and certainly less common than any spe- cies except garmani, this species occurs in a wide variety of habitats. It is very cryptically colored and usually slow moving so that it is seldom seen and is certainly more common than the census figures indicate. As Table 2 shows, it was taken in Mona both in the park area and in the forest. None were seen during censusing at Barbican but during casual observations they were noted occasionally in both the open residential area and in the dense second growth. Almost everywhere we watched Anolis, 8 BREVIORA No. 272 this species eventually turned up — except in the mangroves and in the open pastures. It certainly occurred in the old established gardens in the residential areas. This species was almost always seen on trees, though one was seen on a fence post, and occasionally in taller bushes. This species is most unlike the others in its movements and posture for it is a slow-moving animal which spends much of its time pressed close to its perch. The few census records (to which have been added data col- lected during temperature studies) show that this species occurs over a wide range of heights and with a less clear height prefer- ence than the other species (Table 1). The census data suggest that it uses mostly moderate to large diameter perches. This is the only point where my casual impressions contradict the census results. My impression, based on casual sightings and on watch- ing individuals, is that this species spends much of its time out among the smaller branches and twigs far from the trunk, and that in this it differs from the other species. The few very young individuals seen were on trees and bushes several feet above the ground. DISCUSSION A comparison of the three common species of Anolis in the Kingston area shows that each occupies a different microhabitat, though overlap occurs between them. These differences are of two types, differences in the usual perch height and differences in the type of habitat occupied. Each species of Anolis spends most of its time on certain kinds of perches. The characteristics which these perches have in common describe the structural niche of the species. Important among these characteristics is perch height. The types of habitats frequented by a species seem related to the amount of sun avail- able and in turn to temperature. This aspect of the distribution can be called a climatic niche. These terms have been discussed at greater length with respect to the Anolis of Puerto Rico (Rand, 1964). In perch height, (Table 1), Uneatopus differs from both grahami and opalimis in perching closer to the ground, while grahami and opalinus are almost identical in their perch height distribution. Table 2 shows the differences between male Uneatopus and male grahami, female grahami, and male opalinus significant at the 0.1 per cent level. The differences between female Uneatopus 1967 ECOLOGY OF JAMAICAN ANOLES 9 and both sexes of opalinus and grahami are significant at the 0.1 per cent level. The differences between grahami and opalinus are nowhere significant at the 1 per cent level, though that between female grahami and female opalinus is significant at the 5 per cent level. Though there are differences among these species in an- other aspect of structural niche, i.e. perch diameter, these are small and probably of much less importance. The difference in perch height is probably more important than these figures indicate. Much of the food which Uneatopus catch is on the ground, while the food of the other two species tends instead to be insects which are either on the lizard's perch or on nearby branches or leaves. Two of these three species show intraspecific differences in perch height. In both Uneatopus and opalinus small individuals perch closer to the ground than do large ones. The difference in Uneatopus is striking, in opalinus less so. Where these two species occur together the intraspecific differences assure that the greatest overlap between them involves the large individuals of the larger species {Uneatopus) and the small individuals of the smaller species (opalinus). Those individuals which are closer in size, large opalinus and smaller Uneatopus, overlap least. In grahami there is a slight tendency for the smaller individuals to range higher than do the large ones, so that where grahami and Uneatopus occur together the overlap is greater between the large individuals (adult males) of each species and less between the smaller individuals (females and sub-adults). A study of the social behavior of Uneatopus (Rand, 1967) has shown that adult males have much larger territories (defended home ranges) than do females and sub-adults. A. grahami behave similarly. It was also found that female Uneatopus and grahami defend their terri- tories against other lizards of their own size regardless of species while the males seem to be more tolerant of other species. Where Uneatopus and grahami occur together, the overlap is greatest between the individuals with the largest home ranges and least between those with the smaller home ranges. The spatial separation between the smaller grahami and Uneatopus may be re- inforced by their interspecific territorial defense. The other differences between these three species are in the habitats which they occupy. Anolis Uneatopus occurs throughout the habitats studied around Kingston and so overlaps com- pletely the other two species. However, opalinus and grahami oc- cupy quite different habitats and show relatively little overlap. In general, opalinus occurs in bushy or forested areas while 10 BREVIORA No. 272 grahami occurs in more open areas and in gardens. Differences of this sort between other species of Anolis on other islands have been correlated with differences in amount of shade and with the preferred temperatures of the lizards, by Ruibal (1961) for several Cuban species, and by Rand (1964) for the Puerto Rican species. Certainly the forested areas where opalinus is most common are more shaded than the open areas where grahami occurs. Some very scant data on temperature indicate that these areas are also cooler. Rand (1964) found that in Puerto Rico the species which lived in the shade at low elevations ranged much higher into the mountains than did the species which lived in the sun at low elevations. In Jamaica, opalinus reaches much higher elevations than does grahami (Underwood and Williams, 1959). Data on the body temperatures of these three species were col- lected at Kingston. Lizards were noosed and their cloacal tem- peratures taken with a Schulthies Museum Special Thermometer, within a few seconds after capture. The temperature of the air was then taken in the immediate vicinity. Temperatures were taken only during sunny periods when the lizards had the opportunity to thermoregulate. These data are shown in Figure 1. This figure shows that grahami has an appreciably higher body temperature than does opalinus. This suggests that the difference between them in habitat distribution may be associated with their dif- ferent temperature preferences, opalinus with a lower body tem- perature preference living in more shaded forest areas and grahami with a higher preferred body temperature living in more open, sunnier areas. The occurrence of grahami and not opalinus in some of the very shady gardens in residential areas does not fit this picture but may be due to the small extent of each garden. The body temperature of lineatopus, whose habitat range over- laps those of both the other tv.'o species, might be expected to be intermediate. It is clearly lower than that of grahami. Associated with this is the fact that it becomes active earlier in the morning than does grahami and even more noticeably moves out of the sun earlier in the day. The opalinus temperatures appear lower than those of lineatopus, though too few to be really conclusive. To summarize, the three common species of Anolis in the Kingston area differ from one another in their ecological distri- bution. One species, Anolis lineatopus, differs from the other two in one aspect of its structural niche, living much closer to the ground than do the other two. These two have very similar struc- tural niches but one of them, opalinus, lives in the forested areas 1967 ECOLOGY OF JAMAICAN ANGLES 11 and the other, grahami, in the more open areas. A. opalinus has a lower preferred body temperature than does grahami, which suggests that the difference between them is one of a climatic niche. The relationships of the two rarer species of Anolis, garmani and valencienni, between themselves and to the three common species, are less well documented than are the relationships among the three common species. The largest species, garmani, lives high in trees and thus sel- dom comes in contact with lineatopus. The difference in perch height between male lineatopus and garmani is significant at the 1 per cent level; between smaller lineatopus and garmani at the 0.1 per cent level (Table 2). Like opalinus, garmani is most com- mon in the forested or bushy areas and avoids the more open areas, though it occurs in shaded gardens with grahami. There is a suggestion that garmani lives even higher in the trees than do opalinus and grahami, and also that it is much more closely restricted to large trees than they are. It is probable that in Anolis the size of the prey caught is strongly correlated with the size of the lizard (Schoener, 1967). It, therefore, may be important that the two species which overlap most widely (garmani and opalinus) are the most different in size, garmani being the largest of the Jamaican species and opalinus the smallest. The census data on the final species, A. valencienni, suggest that this species overlaps both the climatic and structural niches of all of the other species, though overlapping lineatopus least. However, the census data are few and, as mentioned before, I have a strong impression that this species spends much more time out among the smaller branches than do any of the others. It is rather different from the other species in its slow motion and in its appearance, and Underwood (1951) reports: "I have kept alive a number of these lizards, and it seems to me that they prefer somewhat larger insects than do other anoles of comparable size." The situation described for Kingston seems to be repeated all along the dry south coast of Jamaica from about the Morant River in the east nearly to Mandeville in the west and a short way up into the foothills of the mountains. A similar situation may exist on the north coast from about Port Antonio west to near Lucea. However, over the rest of the island the situation is com- plicated by the replacement of ^. /. lineatopus by other forms of A. lineatopus with rather different habitat preferences. Unfortu- nately, there is not enough information to describe these situations adequately. 12 BREVIORA No. 272 The Jamaican anoles can be compared with those of Puerto Rico (Rand, 1964), an island of comparable size and with a simi- lar number of species, but ones that are only very distantly related to those on Jamaica (Etheridge, 1960), the Jamaican animals belonging in the beta section of the genus while the Puerto Rican species are members of the alpha section. With respect to the structural niche, two Puerto Rican species (A. cristatellus and gimdlachi) are very similar to A. lineatopus; two species {A. evermcmni and stratulus) are very similar to A. opalinus and grahanii; and one species {A. cuvieri) is apparently like A. gannani. There is on Puerto Rico no species comparable to A. valeucienni in structural niche, while on Jamaica there is no species comparable to the Puerto Rican grass and twig living spe- cies A . krugi, pulcheUiis and poncensis. As in Jamaica, the most closely related Puerto Rican species occupy similar structural niches but differ in microclimatic niches. In Puerto Rico, two pairs of species and a triplet do this, while in Jamaica only one (possibly two) such pair exists. The eccritic temperatures of the Puerto Rican species are similar to those of their Jamaican counterparts but not identical. Thus the anoline faunas of these two islands, though only dis- tantly related, show similarities in their microhabitats. Certain of the structural and climatic niches occupied on both Jamaica and Puerto Rico are very similar. Even more striking is the similarity in the kinds of differences in microhabitats between anoles. On each island very closely related species are paratopic or nearly so and differ in climatic niche and eccritic temperature, while syntopic species are less closely related and usually occupy different structural niches. There has been considerable parallelism in at least certain of the ecological aspects of the independent adaptive radiations of the Anolis on Puerto Rico and on Jamaica. ACKNOWLEDGMENTS I am indebted to Professor David Steven and the staff of the Zoology Department of the University of the West Indies for allowing me to use their facilities and for helping me in many other ways during my stay in Jamaica. I must also thank Dr. E. E. Williams for his advice and criticism throughout this proj- ect, and my wife Patricia for her help both in the field and in preparing the manuscript. This study was supported in part by National Science Foundation Grant No. G- 16066. 1967 ECOLOGY OF JAMAICAN ANOLES 13 SUMMARY During a ten month field study in the vicinity of Kingston, Jamaica, four different habitats were censused for comparative data on perches used (structural niches) and habitats occupied (climatic niches) by five species of Anolis. In the Kingston area there are three very common species: lineatopus, opalinus, and grahami. Anolis lineatopus lives near the ground on moderate to large diameter perches, female and young perching lower and on smaller perches than adult males. It ranges from the most densely forested through the most open habitats. The eccritic temperature range seems intermediate between those of the following species. The closely related species A. opalinus and grahami both use moderate to large perches and perch higher than does lineatopus. A. opalinus lives in forest and grahami in the more open habitats. They replace one another with relatively little overlap. A. grahami has a higher eccritic temperature range than does opalinus. Of the two less common species, A. garmani lives high in trees in the forest. This is the largest of the Jamaican species and it may be significant that the common species that it overlaps most widely is the smallest, opalinus. Anolis valencienni ranges through all habitats and occurs on a variety of perches, probably most frequently on the branches of the crown. This species differs from the others in shape and in its slow movements, and there is some indication that it also takes larger prey. Though the Jamaican anoles are only distantly related to those of Puerto Rico and represent the results of independent adaptive radiations, there has been considerable parallelism in the parti- tioning of the habitat with respect to both structural and climatic niches on the two islands. 14 BREVIORA No. 272 TABLE 1 Structural niche distribution in Kingston anoles based on data collected during the censuses described on page 2. The number of individuals seen is expressed as a percentage of the total num- ber of individuals of that species (and size group in the first three species). Anolis lineatopus Perch Large individuals diameter > 3" 1/2 -3" 10ft 5.27o 6-10 ft 7.8 20.8 11.7 % 3- 5 ft fe < 3ft 1.3 15.6 11.7 13.0 1.3 6.5 3.9 N=77 leaf Total 7.8 29.9 36.4 1.3 26.0 Smaller individuals >3" 1/2-3" 3" 1/2 >10ft 19.4% 6-10 ft 25.8 3- 5 ft 14.5 < 3 ft 6.5 ■3" <'/2" 9.7 1.6 9.7 1.6 8.1 3.2 leaf Total 40.7 37.1 22.6 9.7 >3" 27.5% 2.5 20.0 5.0 1/2-3" 10.0 12.5 5.0 2.5 < 1/2 "leaf 10.0 2.5 2.5 Total 37.5 25.0 27.5 10.0 Anolis opalinus Large individuals N=46 >3" 1/2-3" 10ft 13.07o 23.9 36.9 6-10 ft 28.3 10.9 39.2 3- 5 ft 6.5 6.5 13.0 < 3 ft 10.9 10.9 Smaller individuals N=64 >3" 1/2-3" 3" 1/2-3" 10ft 33.3% 11.1 44.4 6-10 ft 11.1 5.6 16.7 3- 5 ft 11.1 11.1 22.2 < 3 ft 11.1 5.6 Anolis valencienni All individuals N=23 >3" 1/2-3" a — ^27 / •• N ~ O 2S Anolis grahami 23 /i 1 1 1 1 1 1 1 i 1 1 I 23 25 27 29 31 33 Air T»mp»rature 35 "C 35»C 33 3 fe 31 ft I ^ %27 25 23 J I I 1 I L Anolis opalinus J I I \ \ I 23 25 27 29 31 33 35'C Air Temperature Figure I. Anole body temperature plotted against ambient air tem- peratures. BREVIORA Miaseiiiiii of Coinniparative Zoology Cambridge, Mass. 17 November, 1967 Number 273 STUDIES ON NEOTROPICAL POMPILIDAE (HYMENOPTERA) ADDITIONAL NOTES ON EPiPOMPILUS KOHL Howard E. Evans Epipompilus is one of the more exciting genera of Pompilidae, not only because these are among the most brilliantly colored of spider wasps, but also because of the many unusual and apparently primitive structural features of the genus. A few years ago the group was known from about six specimens, all females. In 1961 (Psyche. 68: 25-37), I described the male and reviewed the American species, basing my review on 18 specimens representing nine species. I have now accumulated more than 100 specimens and feel in a position to make some important additions to knowl- edge of the genus. Three species are described as new, all from South America, and the males of three species are described for the first time (nigribasis, aztecus, and pulcherrimus) . The last- named species is the only one known from the United States, and I am indebted to R. W. Dawson of Washington State University for sending me an excellent series collected at the Archbold Bio- logical Station in Florida. For much new South American material I am indebted to Fritz Plaumann, of Nova Teutonia, Brazil, and to Charles C. Porter of Harvard University. Thirteen American species of Epipompilus are now known. Seven of these are known from both sexes, five from females only, and one from males only. All species are included in the following key, but only those species are discussed in the text which are described as new or for which new information is available. I have placed the species in two groups on the basis of an apparent concordance of the shape of the male subgenital plate with the nature of the convergence of the eyes of the female. I am hopeful that this may assist in making future sex associations. BREVIORA No. 273 KEY TO SPECIES Females Eyes converging at the top, the upper interocular line distinctly less than the lower interocular line; front fairly wide, middle interocular line at least about 0.57 x width of head (aztecus group) 2 Eyes parallel or diverging above, the upper interocular line equal to or exceeding the lower; front barely wider than the two eyes taken together, middle interocular line at most about 0.55 x width of head {delicatus group ) 6 Wings wholly fuliginous; abdomen wholly rufous; middle and hind tibiae with numerous short spines above pulcherrimus (Evans) Wings hyaline, bifasciate; abdomen black with whitish spots; middle and hind tibiae not or exceedingly weakly spinose 3 Abdomen with whitish spots only on tergites two and five, none on sternites aztecus (Cresson) Abdomen with whitish spots on tergites two through five or six, also on some of the sternites 4 Greater part of thorax and propodeum rufous; apical tergite black bijasciatus (Ashmead) Greater part of thorax and propodeum black; apical tergite with whit- ish markings 5 Pronotum entirely rufous; hind femora black; propodeum, in profile, nearly flat in front, obliquely declivous behind inca n. sp. Pronotum mostly black, with a whitish band along the posterior mar- gin and with whitish and rufous markings anteriorly; hind femora ferruginous; propodeum, in profile, forming a smooth arc jocosus n. sp. Body and legs wholly black except for a pair of small whitish spots on second tergite; pronotum short, subangulate behind williamsi (Banks) Body and legs variously marked with rufous and whitish; pronotum arcuate or very weakly subangulate behind 7 Thorax wholly rufous (or with limited black markings on sides of pronotum or on venter) (Central and South American species) 8 Thorax with extensive black markings (South American species) .... 10 Front angle of ocellar triangle less than a right angle, postocellar and ocello-ocular lines subequal; second abdominal tergite yellowish brown insolitus Evans Front angle of ocellar triangle greater than a right angle, postocellar line much exceeding ocello-ocular line; second tergite with a whitish band or spots 9 1967 NOTES ON EPIPOMPILUS 3 9. Postnotum a very narrow band; face and clypeus testaceous, antennae wholly testaceous; length of fore wing under 4 mm .nigribasis (Banks) Postnotum medially about half as long as metanotum; face and clypeus black, antennae infuscated on apical third; length of fore wing 5-6 mm delicatus Turner 10. Tergites one and two mostly black, the latter with white markings; pronotum with extensive pale markings; postocellar line only slightly exceeding ocello-ocular line tuciimaniis n. sp. Basal 1.5-2.5 tergites rufous, second sometimes with white markings; pronotum wholly black; postocellar line much exceeding ocello-ocular line 11 1 1 . Propodeum black except for limited whitish markings; antennae black; radial vein angulate at second intercubital vein haiipti (Arle) Propodeum ferruginous; antennae ferruginous except darker apically; radial vein arcuate excelsus (Bradley) Males 1. Subgenital plate with a broad, truncate base from which arises a slender, parallel-sided spine-like process (as in Figs. 1, 3); known species either with the thorax extensively ferruginous or the femora and tibiae almost wholly ferruginous {azteciis group) 2 Subgenital plate tapering gradually to a point from a relatively narrow base (as in Fig. 5); known species without ferruginous markings on the thorax and with the hind legs, at least, mostly fuscous {delicatus group) 5 2. Length of fore wing 4.5 mm or less; transverse median vein of hind wing erect, nearly perpendicular to anal and median veins; body black except clypeus and pro- and mesonota contrastingly ferruginous pidcherrimns (Evans) Length of fore wing 5.5 mm or more; transverse median vein of hind wing oblique; body color not as above 3 3. Thorax black, with limited whitish markings on pronotum and propo- deum; hind femora ferruginous; aedoeagus very long and slender (Fig. 2) jocosus n. sp. Thorax in considerable part rufous; hind femora black; aedoeagus not as above 4 4. Wings strongly twice-banded; middle legs mostly fuscous; basal third of antennae light ferruginous aztecus (Cresson) Wings weakly once-banded; middle femora and beyond light reddish brown; antennae brownish basally innubus Evans 5. Eyes strongly hairy; length of fore wing not over 3 mm; aedoeagus exceedingly short (Fig. 6) nigribasis (Banks) Eyes with very short, barely noticeable hairs; length of fore wing over 3 mm; aedoeagus longer than above 6 4 BREVIORA No. 273 6. Front coxae with at least the basal 0.4 black; mesopleura rather dull, the punctures coarse, separated for the most part by less than their own diameters; parameres slender, somewhat tapering (Brazil) .... excelsiis ( Bradley) Front coxae with at most the basal 0.3 black; mesopleura shining, finely punctate; parameres broader and more coarsely setose 7 7. Third submarginal cell barely wider than high; parameres blunt apic- ally, volsellae moderately slender (Central America) delicatus Turner Third submarginal cell much wider than high; parameres acute apic- ally, volsellae unusually slender (Argentina) tucumaniis n. sp. AZTECUS SPECIES-GROUP Epipompilus pulcherrimus (Evans) This species has been known from two females, one from southern Florida and the other from Andros Island in the Bahamas. R. W. Dawson collected one female and five males in a Malaise trap at the Archbold Biological Station, Lake Placid, Florida, in the winter of 1965-1966 (29 Oct.-Jan. 6). The female is small (fore wing 4.5 mm) and differs from the two females previously known in having the pronotum wholly black except the collar, the remainder of the thorax and propodeum wholly black. Description of male. — Length 4.5-5.0 mm; fore wing 3.5-4.5 mm. Head black except clypeus and mandibles in large part testaceous; antennae black; palpi brown; thorax black except entire pronotum, mesoscutum, and scutellum bright ferruginous (these areas with some black blotching in one specimen); legs black except all spurs whitish, front femora and tibiae suffused with testaceous, hind tibiae with a basal white spot; wings clear hyaline, with dark veins and stigma. Body clothed with short, whitish hairs, these hairs rather long on posterior slope of propodeum. Clypeus truncate, twice as wide as high. Head 1.15 X as wide as high; front broad for the genus, middle interocular line 0.62- 0.65 X width of head; eyes rather weakly convergent below, lower interocular line 0.95 X upper interocular line; ocelli in a broad triangle, postocellar line about 1.5 X ocello-ocular line. Front strongly alutaceous, with shallow but rather large, setigerous punctures. Antennae compact, segments five and beyond slightly produced below, so that the flagellum is crenulate in profile; first four antennal segments in a ratio of about 16:6:11:10, segment three about 1.5 X as long as thick. Setae of eyes exceedingly small. Pronotum short, its posterior margin arcuate; mesonotum closely 1967 NOTES ON EPIPOMPILUS Fig. 1. Subgenital plate of male Epipompilus jocosiis n. sp., paratype. Fig. 2. Genitalia of same specimen, ventral aspect. Fig. 3. Subgenital plate of male E. piilcherrimus (Evans). Fig. 4. Genitalia of same specimen, ventral aspect. Fig. 5. Subgenital plate of male E. nigribasis (Banks). Fig. 6. Genitalia of same specimen, ventral aspect. 6 BREVIORA No. 273 punctate; postnotum nearly as long as metanotum, polished, de- pressed and slightly contracted medially. Propodeum shining, the setigerous punctures small; slope of profile low and even. Middle and hind tibiae with numerous short spines above; claws with a weak, erect tooth, except front tarsal claws more strongly toothed, the tooth subparallel to the outer ray, the outer claws of the front tarsus strongly curved. Fore wing with basal vein arising well beyond transverse median vein; radial vein arcuate, with only a faint angulation at the second intercubital vein; hind wing with transverse median vein erect, nearly perpendicular to anal and median veins, anal lobe large for the genus, half the length of the submedian cell. Abdomen rather small, depressed; subgenital plate in the form of a slender, setigerous process arising from a broad plate (Fig. 3); genitalia as shown in Figure 4. Epipompilus aztecus (Cresson) This species has been known only from southern Mexico. I have recently studied a male from Guatemala which represents the first known male of this species, as well as two females from Panama. The two females were collected on Barro Colorado Island, Canal Zone, one in April 1926 by C. T. Greene and one in May 1939 by J. Zetek [both U. S. National Museum]. They are relatively small for this species (fore wing 7.5-8.0 mm) and somewhat darkly colored: the middle femora are fuscous, and the propodeum and metapleura are wholly fuscous in one specimen, fuscous except at the anterior margin in the other. The male is from Santa Adelaida, Guatemala, collected at 1000 meters eleva- tion in March and April, 1931, by J. Bequaert [coll. of Mus. Comp. Zool.]. All three of these specimens were borrowed many years ago by the late R. R. Dreisbach and have only recently become available for study. Dreisbach made a slide of the male terminalia, but it was poorly prepared, and in my opinion the subgenital plate on the slide (and perhaps the genitalia) do not properly belong with this specimen. I have therefore omitted men- tion of the terminalia in the short description that follows. Description of male. — Length 7.0 mm; fore wing 5.7 mm. Head black, except lower inner orbits with a whitish streak, antennal sockets, face below sockets, and clypeus wholly light ferruginous; mouthparts largely light ferruginous, basal segments of maxillary palpi slightly infuscated; antennae ferruginous, apical two-thirds somewhat infuscated on upper surface; pronotum, mesoscutum, scutellum, upper half and posterior margin of meso- pleura, and a small blotch on the metapleura, light ferruginous, 1967 NOTES ON EPIPOMPILUS 7 remainder of thorax and propodeum fuscous (posterior corners of propodeal rim white; disc of scutellum and lateral posterior mar- gins of scutum slightly infuscated); front legs wholly light fer- ruginous, middle and hind legs fuscous except spurs white, tibiae with a white spot, and posterior coxae with a large white apical spot; abdomen fuscous, shining; fore wings strongly twice-banded much as in female. Body clothed with short, whitish hairs, these hairs rather coarse and conspicuous on the front and on the posterior part of the propodeum. Clypeus large, measuring 1.85 X as wide as high. Head 1.12 X as wide as high, the vertex between the eyes forming a rather strong arc; front broad, middle interocular line 0.60 X width of head; inner orbits strongly convergent below, lower interocular 0.78 X middle interocular line, 0.93 X upper; ocelli in a broad triangle, postocellar line twice the ocello-ocular line. Front dull, with subcontiguous punctures on an alutaceous background. An- tennae moderately long for the genus, weakly crenulate in profile; first four segments in a ratio of about 17:6:10:10, segment three 1.6 X as long as thick. Eyes not hairy. Pronotum short, with an abrupt anterior declivity, its posterior margin arcuate; mesonotum dull, closely punctate; postnotum considerably shorter than metano- tum; propodeum with a strong median impression, in profile the slope very low. Hind tibiae with a few weak spines which barely extend above the pubescence; tarsal claws as described for pul- cherrimus. Fore wing with the basal vein arising very slightly beyond the transverse median vein; radial vein slightly angulate at second transverse cubital vein; maximum width of third sub- marginal cell 1.6 X its maximum height; hind wing with transverse median vein oblique; length of anal lobe about 0.4 that of sub- median cell. Epipompilus bifasciatus (Ashmead) This species has been known only from the type female from Bahia, Brazil. I have recently studied a second specimen from a locality not far distant: Corema, Paraiba, Brazil, June 1957 (no collector given) [coll. Dept. Zool., Sec. Agricultura, Sao Paulo]. It is slightly smaller than the type (fore wing 7.4 mm), but there are no important differences in color or structure. Epipompilus jocosus new species Holotype. — $ , URUGUAY: Florida, Feb. 1952 (no collector stated) [MCZ, No. 31321]. 8 BREVIORA No. 273 Description of type female. — Length 9.5 mm; fore wing 7.5 mm. Head black, except mandibles, clypeus, and lower half of front light ferruginous, the ferruginous extending up the inner orbits to their point of greatest emargination; lower outer orbits, malar space, and labium also ferruginous, the palpi partially infuscated; pronotum black, except collar whitish and posterior margin with a whitish band (including the posterior lobes), the anterior margin, laterad of the collar, ferruginous; remainder of thorax and propodeum black, except for paired whitish spots just above the middle and hind coxae; abdomen black, with paired ivory-white spots on tergites two through six and sternites two through five, all of these spots partially or completely enclosing a black spot within them; antennae and front legs wholly ferru- ginous except somewhat dusky apically; middle and hind coxae black with whitish apices; middle and hind femora and middle tibiae ferruginous; hind tibiae fuscous, middle and hind tarsi dusky ferruginous; middle and hind tibiae with whitish spots near the base, tibial spurs whitish. Wings hyaline, the fore wing with two prominent brownish bands. Body uniformly clothed with pale pubescence. Clypeus shallowly emarginate, measuring about 2.5 X as wide as high. Head 1.23 X as wide as high; eyes covered with short hairs; front moderately broad, middle interocular line 0.60 X width of head; eyes convergent above, lower interocular line 1.1 X upper; ocelli in a broad, flat triangle, postocellar line 2.2 X ocello-ocular line. First four antennal segments in a ratio of about 20:8:18:21, segment three 0.47 X upper interocular line. Front shining, closely punctate. Pronotum fairly long, broadly subangulate behind; mesoscutum closely punctate; postnotum wholly concealed dorsally; propodeum, in profile, smoothly con- vex. Front femora weakly incrassate; middle and hind tibiae with some very short spines intermingled with the pubescence; claws with the tooth long, erect. Fore wings with the radial vein quite strongly angled at the second intercubital vein, hind wing with the transverse median vein oblique, the submedian cell very sparsely and weakly setose, the anal lobe about 0.3 as long as the sub- median cell. Allotype. — $ , BRAZIL: Nova Teutonia, Santa Catarina, 23 Feb. 1961 (F. Plaumann) [Mus. Comp. Zool.]. Description of allotype male. — Length 8.5 mm; fore wing 7.5 mm. Body black, with a bluish cast, marked with ivory-white as follows: apical three-fourths of mandibles, apical half of cly- peus, spots on inner orbits opposite bases of antennae, pronotal 1967 NOTES ON EPIPOMPILUS 9 collar, band along posterior margin of pronotum, interrupted medially, sides of posterior rim of propodeum, apical margin of last abdominal tergite; antennae ferruginous, first two segments black above, flagellum moderately infuscated above, especially apically; coxae black, with apical white markings; trochanters black; femora ferruginous, tibiae also of this color except middle and hind tibiae infuscated at extreme base and toward apex and with a whitish spot on the outer side near the base; spurs whitish; tarsi fuscous. Wings lightly tinged with brownish, with a faint darker cloud in the marginal and second and third submarginal cells, also at the extreme wingtip. Body pubescence light brown to silvery, especially long and semierect on posterior slope of propodeum. Clypeus weakly emarginate, 2.1 X as wide as long. Head 1.16 X as wide as high, the vertex arched very weakly above the eye tops; front narrow, middle interocular line 0.58 X width of head; inner orbits weakly convergent below, lower interocular line 0.95 X upper; ocelli in a broad triangle, post-ocellar line 1.7 X ocello- ocular line. Eyes with only extremely short, inconspicuous setae. Front alutaceous and uniformly punctate. Antennae rather long, the flagellum crenulate in profile; first four segments in a ratio of about 17:7:14:13, segment three 1.8 X as long as thick. Pronotum of moderate length, broadly subangulate behind; meso- scutum closely punctate; postnotum shining, depressed medially, about two-thirds as long as metanotum; slope of propodeum low and even, the median line weakly impressed. Middle and hind tibiae with a large number of short spines above; tooth of claws strong, sloping outward somewhat, outer claws of front tarsus strongly curved, bifid. Features of wings as in female, except anal lobe of hind wing larger, about 0.4 as long as submedian cell. Subgenital plate and genitalia as shown in Figures 1 and 2; the very long setae at the apex of the parameres protrude from the apex of the abdomen even when the genitalia are in resting position. Paratypes. — BRAZIL: 3 S S, same data as allotype except dated 10-15 Feb. 1964, Jan. 1965 (F. Plaumann) [Mus. Comp. ZooL, U. S. Nat. Mus.]. Variation. — The paratypes resemble the allotype closely in color except that one of them has the whitish band on the prono- tum obsolescent. The fore wing varies from 6.5 to 7.7 mm in length; the middle interocular line varies from 0.56 to 0.59 X the head width, the lower interocular line from 0.85 to 0.95 X the upper interocular line. 10 BREVIORA No. 273 Epipompilus inca new species Holotype. — 9 , PERU: Machu Picchu, 1900 meters, 4-19 Sept. 1964 (C. C. Porter) [MCZ, No. 31322]. Description of type female. — Length 13.5 mm; fore wing 11 mm. Head black except lower inner orbits with a whitish streak, apical third of clypeus testaceous, mandibles testaceous except darker apically, palpi fuscous; pronotum wholly ferruginous; re- mainder of thorax and propodeum black except for paired small whitish spots on lower posterior angles of mesopleura and on upper metapleura, and much larger spots at the posterior angles of the propodeum; abdomen black, with large, paired whitish spots on tergites two through six and on sternites two through four, with much smaller spots on the last two sternites; antennae ferruginous, somewhat dusky apically; front legs ferruginous, the tarsi dusky; middle and hind coxae black, with white apices; middle femora ferruginous, the middle tibiae ferruginous except for a black streak on the outer surface; hind femora and tibiae black, the middle and hind tarsi also nearly black; hind tibiae with white spots near the base; tibial spurs dusky ferruginous. Wings hyaline, fore wing with two prominent brownish bands. Entire body clothed with short, pale pubescence. Clypeus shallowly emarginate, measuring about 2.5 X as wide as high. Head 1.2 X as wide as high; eyes densely short-haired; front moderately broad, middle interocular line 0.60 X width of head; eyes convergent above, lower interocular distance about 1.1 X upper; ocelli in a broad triangle, postocellar line 1.6 X ocello-ocular line. First four antennal segments in a ratio of about 29:10:23:27, segment three 0.51 X upper interocular dis- tance. Front shining, very closely punctate. Pronotum rather long, its posterior margin subangulate; mesoscutum closely punctate; postnotum complete, only about a third as long as metanotum. Propodeum, in profile, sloping very weakly on the anterior two- thirds, then abruptly declivous behind; surface of propodeum covered with fine, closely parallel transverse striae. Front legs weakly incrassate; middle and hind tibiae not at all spinose; claws with the tooth strong, erect, the outer ray curved so as to be nearly parallel to it. Basal and transverse median veins of fore wing interstitial; radial vein strongly angulate at second intercubital vein; transverse median vein of hind wing strongly oblique, the submedian cell almost devoid of setulae; anal lobe small, about 0.3 as long as submedian cell. 1967 NOTES ON EPIPOMPILUS 11 Remarks. — This large species appears closely allied to bifascia- tus and to jocosus, but there are prominent color differences as well as differences in the postnotum, shape of the propodeum, spines of the hind tibiae, and so forth. Although innubus belongs to this group and is known from the male only, from a locality only a few hundred miles from the type locality of inca, it seems much too small and differently colored to represent the male of this species. DELICATUS SPECIES-GROUP EPIPOMPILUS DELICATUS Turner The male of this species was described by Evans, 1966, Mem. Amer. Ent. Soc, 20: 33. The genitalia are most similar to those of piilcheirimus, although the subgenital plate is very different and much like that of excelsus Bradley. As a result of other new sex associations in the genus, I feel more confident that this male is correctly associated with the female delicatus than I did when describing it. I have seen one additional female of this species, collected at Turrialba, Costa Rica, by Kenneth Cooper [collection of U. S. National Museum]. This specimen is of the same size as the type and is very similar structurally, but there are some minor color differences: the greater part of the second tergite is rufous, upon which two whitish spots are imposed; tergite five also has large, paired whitish markings against a rufous background (these may be present but concealed in the type); the legs are mostly rufous but are irregularly tinged with brownish. EPIPOMPILUS NiGRiBASis (Banks) This species was transferred to Epipompilus by Evans, 1966, Mem. Amer. Ent. Soc, 20: 31. It has been known only from the type female, from Panama, but I have recently discovered two males which unquestionably belong here. One is from Colombia: Bonda, August, Ace. no. 1999 [Acad. Nat. Sci. Philadelphia], the other from Rio de Janeiro, Brazil, 7 March 1966 (H. and M. Townes) [coll. H. K. Townes]. The small size and close coin- cidence of the wing venation with that of the female lead me to feel confident of this association. The eyes of the female are unusually strongly hairy, and these are the only known males having strongly hairy eyes. Although the two males are from widely separated localities, the genitalia and all other features are closely similar. 12 BREVIORA No. 273 Description of male. — Length 2.5-3.5 mm; fore wing 2.2-3.0 mm. Head black; thorax and abdomen dark castaneous to fuscous, without white or rufous markings; antennae Hght brown, the fiagel- lum darker on the upper surface; coxae and hind femora and tibiae dark brown, legs otherwise light to medium brown, the spurs brownish; wings subhyaline, the fore wings with a brownish cloud at the marginal and submarginal cells. Body extensively clothed with short, light brown setae; eyes densely covered with short setae. Clypeus truncate, slightly wider than lower front. Head about 1.10 X as wide as high, the vertex forming an even arc above the eye tops; eyes strongly convergent below, lower interocular distance only 0.8 X upper interocular; middle interocular line 0.57-0.61 X width of head; ocelli in an obtuse triangle, post-ocellar line 1.2-1.6 X ocello-ocular line. Front alutaceous and with shal- low punctures. Antennae rather long, the flagellum crenulate in profile; first four segments in a ratio of about 6:2:5:6, segment three 1.6 X as long as thick; flagellar pubescence coarse. Prono- tum very short, arcuate behind; pro- and mesonota somewhat shining, with shallow, irregular punctures; postnotum very short, concealed medially by the overhanging metanotum; slope of pro- podeum low and even. Hind tibiae with some unusually long spines for the genus; claws with the tooth short, erect, except outer claws of front tarsus strongly curved, bifid. Basal vein of fore wing arising well beyond transverse median vein; marginal cell large, the radial vein angulate at second intercubital vein; hind wing with transverse median vein weakly oblique, almost perpen- dicular to median vein; anal lobe very small, hardly more than a minute scale with a fringe of setae (as it is in the female). Sub- genital plate shaped much as in excelsiis, but with fewer setae (Fig. 5). Genitalia with the aedoeagus exceedingly small; para- peniai lobes abruptly expanded on apical half; parameres short, setose (Fig. 6). Epipompilus tucumanus new species Holotype. — 9 , ARGENTINA: Horco Molle, Tucuman, 25 March-30 April, 1966 (C. C. Porter) [MCZ, No. 31320]. Description of type female. — Length 8.5 mm; fore wing 7.0 mm. Head black, except antennal lobes with ivory-white spots, apical half of mandibles testaceous; pronotum black, its posterior margin with a whitish band which does not include the posterior lobes but extends broadly across the shoulders to the anterior margin, the bands irregularly bordered with rufous; mesoscutum 1967 NOTES ON EPIPOMPILUS 13 black except ferruginous on the midline posteriorly; remainder of thorax and propodeum entirely ferruginous; abdomen black except basal third of first tergite as well as basal 1.3 sternites ferru- ginous, and with the following ivory-white: large, paired spots on tergite two, large, connected spots on tergite five, and a median spot toward the base of tergite six; antennae fuscous except each of the first three segments with an ivory-white spot above; front coxae black, with a whitish spot apically, middle and hind coxae ferruginous, with black and white markings apically, middle and hind femora ferruginous except black apically, legs otherwise fuscous (including spurs) except middle and hind tibiae with whitish markings on outer surface. Fore wings strongly twice- banded, faintly luteous between bands, but apices hyaline; hind wings hyaline, the tips clouded. Body wholly clothed with short, pale pubescence. Clypeus arcuately emarginate, measuring 2.3 X as wide as its median length. Head 1.2 X as wide as high; eyes wholly covered with short hairs; front narrow, middle interocular line 0.56 X width of head; eyes convergent below, lower interocular line 0.9 X upper; front angle of ocellar triangle slightly exceeding a right angle, postocellar line 1.2 X ocello-ocular line. First four antennal segments in a ratio of about 19:7:18:24, the third segment 0.53 X upper interocular line. Front somewhat shining, the punctures rather shallow. Pronotum arcuate behind; mesoscutum minutely, shallowly punctate; postnotum a very narrow transverse band; slope of propodeum low and even, the median line not impressed. Front femora not incrassate; hind tibiae with numerous short, dark spines; claws with the tooth strong, erect. Fore wing with the basal and transverse median veins interstitial; radial vein not at all angled at second intercubital vein; hind wing with the trans- verse median vein oblique, the anal lobe very small, about 0.3 the length of the submedian cell. Allotype. — 6 , ARGENTINA: Horco Molle, Tucuman, 15-21 May 1966 (L. Stange) [Coll. H. K. Townes]. Description of allotype male. — Length 5.0 mm; fore wing 4.7 nmi. Head black, except marked with whitish around antennal insertions, on extreme lower inner orbits, and on basal half of clypeus; thorax and propodeum black, with some bluish reflec- tions, except posterior margin of pronotum white (not reaching lateral lobes); abdomen black, with bluish reflections; antennae dark brown, blackish on upper surface; palpi light brown to testa- ceous; front coxae black on basal third, middle section light fer- ruginous, apex white, middle coxae black at base but mostly light 14 BREVIORA No. 273 ferruginous, hind coxae black with white tips; all trochanters fuscous; front legs beyond trochanters light ferruginous, also mid- dle femora; middle legs beyond femora lightly infuscated, also middle spurs; hind femora, tibiae, tarsi, and spurs black, except tibiae with a basal white spot; wings hyaline, fore wings weakly infuscated on outer third. Body pubescence pale, silvery. Clypeus arcuate emarginate, measuring twice as wide as high. Head 1.1 X as wide as high, the vertex elevated in an even arc above the eye tops; inner orbits converging below, lower interoc- ular line only 0.7 X upper; middle interocular line 0.60 X width of head; ocelli in an obtuse triangle, postocellar line only 1.1 X ocello-ocular line. Eyes not hairy. Front alutaceous, punctate. Antennae moderately elongate, weakly crenulate in profile; first four segments in a ratio of about 13:5:9:8, segment three nearly twice as long as thick. Pronotum short, its posterior margin broadly arcuate; mesoscutum strongly punctate; median portion of postnotum nearly as long as metanotum, smooth and polished; slope of propodeum very low. Claws dentate except outer claws of front tarsi strongly curved, bifid; hind tibiae with a number of spines extending above the coarse pubescence. Wing venation not differing appreciably from that of female. Subgenital plate essen- tially the same as in excelsus (Bradley) (see Evans, 1961, Psyche, p. 34, fig. 3). Genitalia differing from those of excelsus in only a few details: the volsellae are considerably more slender and elongate, and the parameres are slightly broader, tapering to a subacute point on the apical third, and slightly more strongly setose. Par city pes. — 4 9 9, same data as type [Mus. Comp. Zool., U. S. Nat. Mus., and Inst. Miguel Lillo, Tucuman]. 1 9 , Jujuy, Argentina, 15 Jan. 1966 (H. and M. Townes) [Coll. H. K. Townes]. Additional specimen (not designated a paratype). — 1 9 , San Esteban, near Puerto Cabello, Venezuela, 14 Jan. 1940 (P. J. Anduze) [Coll. H. K. Townes] . Variation. — The paratypes vary in length from 7.0 to 8.5 mm, fore wing from 6.0 to 7.0 mm. The third antennal segment varies from 0.45 to 0.55 X the upper interocular line, the lower interoc- ular fine from 0.85 to 0.95 X the upper. There are no differences in color or structure worthy of note. The Venezuela specimen is similar to the type series in many respects, and it seems to me very probable that it belongs with this species. In size and all standard measurements it falls within the ranges expressed above, and the wing venation is similar, although the wings have more of a yellow tinge toward the base. 1967 NOTES ON EPIPOMPILUS 15 There are several color differences: the mandibles are mostly testaceous; the antennae are medium brown except the first three segments yellowish below, white above; the pronotum has a very broad ivory-white band, though occupying the same position as in the types; the front femora are brownish, the middle and hind femora fuscous, the coxae rather extensively mottled with black; the first abdominal segment is wholly black, the spots on tergite two confluent medially. EPIPOMPILUS EXCELSUS (Bradley) This species is characteristic of southeastern Brazil, and appears to be relatively common for the genus. Fritz Plaumann, of Nova Teutonia, Santa Catarina, has now sent me 13 females and 33 males from that locality, collected August through April. I have also seen one female and 16 males from Rio de Janeiro, collected in March, 1966, by Henry and Marjorie Townes, as weU as 10 males from Alto da Serra, Morretes, Parana, collected by the Townes in February 1966 [Coll. H. K. Townes]. I now feel convinced that the male I described in 1961 does in fact go with this species. The series from Rio de Janeiro differs somewhat from specimens from Parana and Santa Catarina, but the male genitalia are identical and other structural differences minor. The female has a pair of white spots superimposed on the rufous base of tergite two, and the ocellar triangle is a little less broad, the postocellar line being only about 1.5 X the ocello-ocular hne. The males from this locality are rather small (fore wing 3.3-4.4 mm) and also tend to have a less broad ocellar triangle; they also have somewhat more white on the legs, the front coxae having the apical 0.3-0.5 white. The genitalia of the specimens examined proved to resemble very closely the figure in my 1961 paper (p. 34, fig. 4). (Received 8 March 1967) BREVIORA Mmseiiimi of ComparatiYe Zoology Cambridge, Mass. 17 November, 1967 Number 274 THE GENUS HOPS (ARANEAE, CAPONIIDAE) IN PANAMA AND THE WEST INDIES Arthur M. Chickering The Caponiidae include a comparatively small group of genera now known to be distributed over much of Africa, South America, Central America, southwestern part of the United States, and the West Indies. For many years I have had a continuous interest in this family and during this period I have accumulated a rather large collection of certain genera in the family. I have had a special inter- est in the genus Nops and this seems to be a convenient time to record my understanding of this genus as it is now believed to occur in Panama and the West Indies. Again, I wish to express my appreciation and gratitude for the aid and encouragement in the pursuit of my studies extended by the staff of the Museum of Comparative Zoology for many years. My thanks are also extended to Dr. W. J. Gertsch, American Museum of Natural History, for the loan of numerous Caponiidae from Cen- tral America and the West Indies. Grants GB-1801 and GB-5013 from the National Science Foundation have made it possible for me to continue my studies in the Museum of Comparative Zoology and to spend a total of nearly eleven months collecting in Panama and the West Indies. With the exception of the type of Nops gertsch i sp. nov., all types named in this paper together with all other specimens belonging to the genus Nops in my personal collection will be deposited in the Museum of Comparative Zoology. Genus Nops MacLeay, 1839 Type species is Nops guanabacoae by monotypy. The genus Nops was established on the basis of a female from Cuba. Since that time a dozen additional species have been recog- nized from South and Central America and the West Indies. The 2 BREVIORA No. 274 most important features of the genus observed during my study of the group may be stated as follows: Carapace low; widest opposite or just behind the second coxae; considerably narrowed at anterior end; median thoracic groove or pit obscure. Eyes: two, on a slightly raised tubercle; separated by less than their diameter. Chelicerae of moderate size; with no teeth along fang groove but with a rela- tively large, soft lobe on the promargin (Fig. 11); fang evenly curved; with conspicuous striations on lateral surfaces. Maxillae: robust; lobate at base; strongly convergent; obtusely terminated dis- tally and not narrowed as in Caponina; somewhat concave near middle. Lip: usually firmly united to sternum; about as broad as long. Sternum: with a well marked margin; raised somewhat from margin and then flat throughout; nearly oval in outline; widest opposite interval between second and third coxae. Legs: 4123 in order of length; coxae 1243 in order of length; all tarsi with three claws each but third c'aw often hidden; first and second tarsi with a proximal, ventral, membranous lobe; all tarsi bisegmentate; first and second metatarsi with a median, ventral, membranous fold nearly or quite throughout the segment; first and second coxae lobate at base. Palp: femur in both sexes with a small, prolateral cusp at its base in opposition to the cheliceral striations; with a more or less conspicuous, tibial brush on the prolateral side near distal end in both sexes; in both sexes the tarsus bears a dorsal, distal, oval pad of fine hairs; in females the tarsus lacks a terminal claw; in males the tarsus is simple but usually more or less distinc- tive. Abdomen: elongate; ovoid; with four spiracles clearly shown; six spinnerets with anterior and median pairs nearly in a transverse row; females usually with a moderately well developed scutum from genital groove to pedicel but this is usually poorly developed in males; females have no definite epigynum. I have been obliged to transfer Caponina blanda Bryant to the genus Nops on the basis of what I regard as very clear and definite evidence. After a period of uncertainty I have also come to the conclusion that Nops coccineus Bryant belongs to Nops blandus. Caponina darUngtoni Brvant, known only from a female taken in the Dominican Republic, has been rather carefully examined with the conclusion that it probably deserves to be placed in a new genus. Its status can be determined only when males are available for study. The total list of species belonging to the genus Nops, as presented in this paper, may be given as follows: Nops blandus (Bryant); Nops coccineus Simon; Nops craneae sp. nov.; Nops flutillus sp. nov.; Nops gertsclii sp. nov.; Nops glaucus Hasselt; 1967 NOPS IN PANAMA AND THE WEST INDIES 3 Nops guanabacoae MacLeay; Nops largus sp. nov.; Nops simla sp. nov.; Nops toballus sp. nov.; Nops ursumus sp. nov. With the exception of Nops gertschi sp. nov., all new species listed above are known from both sexes. Key to male Nops from Panama and the West Indies la. Species with embolus nearly as long as or longer than the cymbium (craneae, guanabacoae, largus, simla) 2 lb. Species with embolus distinctly shorter than the cymbium (blandus, coccineus, flutillus, gertschi, toballus, ursumus) 5 2a. Species with distinct dorsal abdominal spots on a light back- ground {craneae, simla) 3 2b. Species without distinct dorsal abdominal spots (guanabacoae, largus) 4 3a. Abdomen with a central, longitudinal, irregular, grayish stripe and and a series of four irregular, grayish spots on each side . . craneae, p. 6 3b. Abdomen with a series of paired purplish spots along the dorsal surface; without a central stripe simla, p. 13 4a. Embolus somewhat sinuous largus, p. 1 1 4b. Embolus not sinuous; somewhat swollen near middle guanabacoae, p. 1 1 5a. Species with at least somewhat definite dorsal abdominal spots (coccineus, gertschi, ursumus) 6 5b. Species with no definite dorsal abdominal spots (blandus, flutillus, toballus) 8 6a. Abdomen with a central, dorsal, indented, brownish stripe consisting of four interconnected, irregular angular spots; each lateral side with a brownish stripe coccineus, p. 5 6b. Abdomen lacking such a color pattern as given above (gertschi, ursumus) 7 7a. Abdomen with a series of five pairs of dark grayish, dorsal spots ursumus, p. 16 7b. Abdomen grayish in general, with two light greenish bars in an- terior two-thirds; without any such distinct series of spots as given above gertschi, p. 9 8a. Embolus short; about half as long as tarsal bulb blandus, p. 4 8b. Embolus longer; nearly or quite as long as tarsal bulb (flutillus, toballus) 9 9a. Abdomen purplish dorsal ly flutillus, p. 7 9b. Abdomen gray dorsally toballus, p. 14 Except for the females with a distinct abdominal color pattern this sex is conspicuously lacking in characters with can be used in constructing a key for their separation into species. For this reason no satisfactory key for their separation has yet been devised. 4 BREVIORA No. 274 Nops BLANDUS (Bryant) Figures 1-2 Caponina blonda Bryant, 1942: 328, pi. 2, fig. 17. The male holotype from St. Croix, U. S. Virgin Islands, is in the Museum of Comparative Zoology. Nops coccineus, — Bryant, 1948: 344. (Not Nops coccineus Simon.) Caponina blanda Bryant must be transferred to Nops because it very clearly exhibits the characters of the latter genus. Until recently I have considered the specimens called Nops coccineus by Miss Bryant, from Haiti, as a new species and intended to describe it as new. Recently, however, following careful comparison with N. blandus, I have been forced to place them together. There are minor differences but, in my judgment, no greater than are com- monly found among individuals of the same species (Figs. 1-2). Figure 1. Nops blandus (Bryant). Left male palp; prolateral view. Fig. 2. Nops coccineus Bryant (= Nops blandus (Bryant)). Left male palp; prolateral view. I had hoped to collect specimens of Nops blandus during my two recent visits to St. Croix but I failed to find any members of the genus. I now have two males from St. John together with several immature specimens collected in July, 1966; I also have a female together with two immature specimens collected in February, 1964. These are all tentatively placed in Nops blandus pending more data on the Nops population of these islands. During my recent visit to Virgin Gorda, British Virgin Islands. I collected two mature males, one female, and two immature specimens. These were all taken on August 18, 1966. and are also tentatively placed in this species. 1967 NOPS IN PANAMA AND THE WEST INDIES A female Nops and several immature specimens collected on St. Thomas in February, 1964, and in July, 1966, are not yet definitely assigned to species. Nops coccineus Simon Figure 3 Nops coccineus Simon, 1891, fig. 18, pi. 42. Male and female syntypes from St. Vincent, B. W. I., are in the British Museum (Natural History). Simon, 1893: 324; Petrunkevitch, 1911: 133; Lutz, 1915: 82; Berland, 1932: 339; Roewer, 1942: 316; Bonnet, 1958: 3114. While working in the British Museum in the summer of 1958, I had an opportunity to examine both sexes of Nops coccineus Simon. The following notes were written during my examination of these specimens. Female: length from anterior border of clypeus to posterior end of anal tubercle 8.78 mm; carapace 2.99 mm long; 2.26 mm wide opposite interval between second and third coxae where it is widest; regularly rounded posteriorly but much nar- rowed in front; about 1.3 mm tall. Two eyes circular and separated by a little more than the radius of one. Height of porrect clypeus 3.25 times the diameter of an eye. Chelicerae with a fleshly lobe just promarginal to fang; no teeth observed; markedly striate along Figure 3. Nops coccineus Simon. Left male palp; prolateral view. Fig. 4. Nops craneae sp. nov. Left male palp; prolateral view. outer surfaces; opposing palpal, femoral cusps moderately devel- oped. Tarsi 1 and 2 with a small third claw; tarsi 3 and 4 with a somewhat more prominent third claw; all tarsi bisegmentate; tarsi 1 and 2 with the usual ventral, proximal lobe. Metatarsi 1 and 2 6 BREVIORA No. 274 with the usual ventral, median, longitudinal, membranous fold. Sternum only extended to bases of fourth coxae. Epigynal area lightly cornified. Color in alcohol: cephalothorax very reddish; legs somewhat lighter; abdomen yellowish dorsally with a central, in- dented, brownish stripe consisting of four connected, irregularly angular spots. Male: essential features of the male palp shown in Figure 3; total length from anterior border of porrect clypeus to posterior end of posterior spinnerets 5.6 mm; color in general like that of female but clearer; each lateral side of abdomen with a brownish stripe, very irregular; venter yellowish. NOPS CRANEAE Sp. nOV. Figure 4 Holotype. The male is from Simla, Arima Valley, Trinidad, W. I., in the close vicinity of the Wm. Beebe Tropical Research Station, April 17, 1964. This species is named after Miss Jocelyn Crane, director of the station. Description. Total length from anterior border of clypeus to posterior end of extended spinnerets 7.8 mm; length from anterior border of clypeus to posterior end of abdomen 7.21 mm. Carapace 3.25 mm long; 2.34 mm wide opposite interval between second and third legs where it is widest; only about 0.55 mm tall; only a slight indication of a median, thoracic fovea; surface very finely granulate. Eyes: two as usual on a slightly raised tubercle; sepa- rated from one another by slightly less than the radius of one; height of clypeus equal to nearly 2.75 times the diameter of an eye. Chelicerae, maxillae, lip and sternum essentially as given in state- ment of characters of the genus. Legs: tibial index of first leg 13, of fourth leg 9; no spines observed on legs; typical of the genus in regard to tarsal bisegmentation, tarsal claws, tarsal basal, ventral lobes and metatarsal ventral, longitudinal folds. Palp: essential features shown in Figure 4; cusp at base of femur on prolateral surface; oval, dorsal, tarsal pad of short, fine hairs as usual; all segments simple except the tarsus. Abdomen: typical of the genus. Color in alcohol: carapace and sternum a dark reddish; consider- able black pigment in ocular area; legs and mouth parts reddish brown in general but with variations; metatarsi 1 and 2 and the proximal divisions of tarsi 1 and 2 very light yellowish. Abdomen: light greenish with dark gray spots; an irregular, median, dark gray stripe extends throughout the dorsum; on each dorsolateral side there is a series of 4 irregularly elongated, dark gray spots the last of which spreads and unites with the median stripe; the venter has 1967 NOPS IN PANAMA AND THE WEST INDIES 7 two small and very irregular, dark gray spots and a pair of larger gray spots in a ventrolateral position; the greenish coloration is somewhat darker in the region of the tracheal spiracles. Female paratype. The described female paratype measures 10.34 mm from anterior border of clypeus to posterior end of abdomen; from anterior border of clypeus to posterior end of spinnerets 11.25 mm. Carapace 4.03 mm long; 3.06 mm wide op- posite second coxae where it is widest; otherwise essentially as in male holotype. Eyes: two, separated by about Vs of their diameter; height of clypeus equal to about 3.5 times the diameter of an eye. Chelicerae, maxillae, lip and sternum essentially as in male. Legs: tibial index of first leg 14, of fourth leg 10; tarsal claws, tarsal, basal, ventral lobes, metatarsal ventral, membranous folds and other features as stated for the genus all typical. Abdomen: es- sentially as in male; scutum only moderately developed anterior to genital groove. Color in alcohol: nearly identical to that of male; abdominal color pattern clear and definite; venter with three very irregular, dark spots on each ventrolateral side; the first of these is lateral to the scutum, the second is a little behind the middle and somewhat nearer the middle line, while the third is anterior to the spinnerets and still nearer the middle line. Records. The described female paratype is also from Trinidad, W. I., with no date of collection (N. A. Weber). Two females are in the collection with the described female paratype; two immature specimens from Balandra Bay, Trinidad, April, 1922 (Reynolds collection). I took eight specimens in April, 1964, in the vicinity of Simla, Arima Valley, Trinidad, W. I. One immature specimen from Piarco (A. M. Nadler), January, 1955, and a female from Gas- paree, November, 1944 (T. S. Jones) are tentatively placed here. NOPS FLUTILLUS Sp. UOV. Figures 5-6 Holotype. The male holotype and described female paratype are from the Dutch West Indies, Curagao, Sint-Nicolaas; St. Marthaa, December 29, 1962 (B. de Jong and H. W. Levi). The specimens referred to this species were tentatively placed under the name Nops gloiicus Hasselt — a species based upon a female from the West Indian island of Bonaire. After a period of some uncertainty the decision has been made to regard the species as new. The name of the species is an arbitrary combination of letters. 8 BREVIORA No. 274 Description. Total length of male from anterior border of cly- peus to posterior end of abdomen 5.79 mm. Carapace 2.5 mm long; 1.87 mm wide opposite second coxae where it is widest; about 0.55 mm tall; otherwise essentially typical of males of the genus. Eyes: two as usual; separated by about %4 of their diameter. Chelicerae, maxillae and lip essentially typical of males of the genus. Sternum: longer than wide in ratio of about 6:5; somewhat more oval in outline than usual; fourth coxae separated by about %o of their width; otherwise quite typical of the genus. Legs: tibial index of first leg 14, of fourth leg 10; tarsi typical with respect to bisegmentation, claws, and ventral, proximal, membranous lobes; metatarsi also typical with respect to the ventral, medial, membra- nous fold; no spines but with many hairs and bristles; trichobothria observed on tarsi, metatarsi, tibiae and, possibly, also on palpal segments. Palp: essential features shown in Figures 5-6. Abdomen apparently typical of the genus in all essential features. Color in alcohol: all parts except abdomen essentially as described for N. largus sp. nov.; abdomen purplish over most of the dorsum, be- coming darker and broader posteriorly; lateral sides yellowish, be- coming purplish posteriorly; venter with a fairly well defined scutum anterior to genital groove and with light green color around a broad margin; remainder of venter yellowish with purplish dots, becoming more definitely purplish toward posterior end. Figures 5-6. Nops fliitilliis sp. nov. Fig. 5. Left male palp; prolateral view. Fig. 6. Tip of embolus, more enlarged; retrolateral view. Figs. 7-8. Nops gertschi sp. nov. Fig. 7. Right male palp; prolateral view, Fig. 8. Tarsal bulb, more enlarged; retrolateral view. 1967 NOPS IN PANAMA AND THE WEST INDIES 9 Female paratype. The described female paratype has a total length of 8.45 mm from anterior border of porrect clypeus to pos- terior end of abdomen. Carapace 4.03 mm long; 3.04 mm wide op- posite posterior border of second coxae where it is widest; 0.96 mm tall; otherwise essentially typical of the genus. Eyes: two as usual on a low tubercle; separated by nearly % of the diameter of an eye; height of porrect clypeus equal to about 4.5 times the diameter of an eye. Chelicerae, maxillae, lip and sternum all essentially typi- cal of females of the genus. Legs: tibial index of first leg 18, of fourth leg 13; tarsi typical with respect to claws, bisegmentation, and basal, ventral lobes; metatarsi typical with respect to ventral, medial, longitudinal folds; claw tufts moderately developed; with no palpal tarsal claws; with the usual, distal, prolateral, tibial, palpal brush. Abdomen: essentially as in male; a well defined scutum oc- cupies the whole area of tracheal spiracles continued to the pedicel; what has frequently been termed the epigynum is a slightly re- curved slit between anterior spiracles. Color in alcohol: essentially as in male holotype with minor variations; with considerable light green color in parts of the ventral scutum. Records. One male paratype, two female paratypes and two im- mature specimens taken with the holotype. One male paratype and eleven immature specimens from several localities on the island of Curagao collcted by Dr. and Mrs. H. W. Levi and Mr. B. de Jong in December, 1962. NOPS GERTSCHI sp. nOV. Figures 7-8 Holotype. The male holotype is from Loma Cibao, La Vega, Dominican Republic. August 9, 1956, 600-1600 m elevation (Dr. Allan F. Archer, E. B. M.). The holotype will be deposited in the American Museum of Natural History, New York, N. Y. Until recently the specimen described here as a new species was regarded as another specimen of Nops blandus (Bryant) but study has convinced me that it must be regarded as new. The species is named after Dr. W. J. Gertsch. Description. Total length from anterior border of clypeus to posterior end of abdomen 5.98 mm; from tip of somewhat por- rect chelicerae to posterior end of abdomen 6.18 mm. Carapace 2.86 mm long; 2.08 mm wide opposite second coxae where it is widest; narrowed to about 0.78 mm at anterior end. Eyes: two as usual on a low tubercle; separated by their diameter; surrounded 10 BREVIORA No. 274 by black pigment, lighter between the eyes; height of porrect clypeus equal to nearly 3.75 times the diameter of an eye. Chelicerae, maxillae, lip and sternum all essentially typical of males of the genus; fourth coxae separated by their width. Legs: tibial index of first leg 14, of fourth leg 9; tarsi typical as stated in de- scription of the genus; metatarsi 1 and 2 also typical with respect to the ventral, longitudinal, membranous fold; typical with respect to other observed features. Palp: essential features shown in Fig- ures 7-8; palpal, tibial, prolateral brush apparently lacking on left palp. Abdomen: essentially typical of males of the genus. Color in alcohol: carapace and sternum an orange yellow; with consider- able black pigment in ocular area; legs nearly the same as carapace dorsally but lighter ventrally; lip and chelicerae like sternum; maxillae lighter; palp like legs with variations. Abdomen: dorsum grayish in general but with two irregular, light greenish, dorsolateral bars in anterior two-thirds; lateral sides dark grayish with ventral extensions of the lighter dorsal, irregular bars; venter yellowish with a faint greenish tint anterior to genital groove; posterior to genital groove light greenish with irregular dark spots for nearly three-quarters of the distance to the spinnerets; the area im- mediately in front of the spinnerets is dark like most of the dorsum. Nops GLAUCUS Hasselt Nops glaiiciis Hasselt, 1887: 77, figs. 1-4, 10. Female holotype from Bonaire, Dutch West Indies, in the Natural History Museum, Leiden. Simon, 1892: 449; Petrunkevitch, 1911: 134; Roewer, 1942: 316; Bonnet, 1958: 3114. Through the courtesy of Dr. van der Hammen of the Natural History Museum in Leiden I have been able to examine the holo- type of this species from which I have derived the following: total length 9.17 mm from anterior border of clypeus to posterior end of abdomen; carapace 3.9 mm long and 2.92 mm wide opposite interval between second and third legs. Eyes: two as usual on a low tubercle; separated by nearly three-fourths of the diameter of an eye; height of clypeus nearly equal to four times the diameter of an eye. Legs: typical of the genus in all respects as given in the description of the genus. The usual palpal, tibial brush is present. The color appears to have faded somewhat during long preserva- tion. Legs and palps light yellowish; carapace, sternum, and other mouth parts all reddish brown; abdomen light grayish with a pur- plish tint dorsally and yellowish ventrally. I consider the original drawings very good with one exception; in figure 1 the second femur 1967 NOPS IN PANAMA AND THE WEST INDIES 11 is over-inflated; in the holotype under observation the second femur is only normally robust. Simon reported this species from Venezuela in 1892. Apparently it has not been definitely recorded since that time. NoPS GUANABACOAE MacLcay Figures 9-10 Nops guanabacoae MacLeay, 1839: 1-14, pis. 1-2. The female holotype from Cuba is probably in the British Museum (Natural History), London. Walckenaer, 1842: 442; Hasselt, 1887: 71; Simon, 1893: 328; Petrunkevitch, 1911: 133; Bryant. 1940: 271; Roewer, 1942: 316; Bonnet, 1958: 3114. The male was unknown until 1940 when Miss Bryant described it and had both sexes for study. As a result of my examination of these specimens I am obliged to disagree with several of the pub- lished statements concerning the male. These may be briefly stated as follows: eyes separated from one another by nearly % of the diameter of one of them; chelicerae with the usual fleshy lobe on the promargin of the fang groove; lip firmly united to sternum and only a little longer than wide; legs 4123 in order of length; all tarsi with three claws; tarsi 1 and 2 with the typical ventral, proxi- mal lobe; metatarsi 1 and 2 with the ventral, median, membranous fold; the usual palpal, tibial brush is present in both sexes. Es- sential features of the male and female palps are shown in Figures 9-10. There are now two mature males together with several fe- males and immature specimens from several localities in Cuba, and the species is, apparently, known only from that island. Nops largus sp. nov. Figures 11-15 Holotype. The male is from the Panama Canal Zone Forest Pre- serve, January 21, 1958. The name of the species is a Latin ad- jective referring to its abundance in its natural habitat. Description. Total length 6.7 mm from anterior border of cly- peus to posterior end of abdomen; from anterior border of slightly extended chelicerae to posterior end of spinnerets 6.89 mm. Cara- pace 3.15 mm long; 2.34 mm wide opposite second coxeae where it is widest; about 0.7 mm tall. Eyes: two as usual on a slightly raised tubercle; separated by slightly less than their radius; height of porrect clypeus nearly equal to 2.66 times the diameter of an eye. Chelicerae, maxiUae and lip essentially 12 BREVIORA No. 274 typical of the genus; cheliceral fleshy lobe shown in Figure 11. Sternum: surface rather coarsely granulate; coxae 1 and 2 basally lobed as usual; terminated just posterior to bases of fourth coxae which are separated by about their width (Fig. 12). Legs: tibial index of first leg 14. of fourth leg 11; spines lacking; tarsal claws, ventral, proximal lobes on tarsi 1 and 2 and ventral, medial. Figures 9-10. Nops giianabacoae Macleay. Fig. 9. Left male palp; pro- lateral view. Fig. 10. Left female palp; prolateral view. Figs. 11-15. Nops largus sp. nov. Fig. 11. Left chelicera of male. Fig. 12. Sternum of male. Fig. 13. Left male palp; prolateral view. Fig. 14. Tip of embolus; en- larged. Fig. 15. Left female palp; prolateral view. 1967 NOPS IN PANAMA AND THE WEST INDIES 13 longitudinal folds on metatarsi 1 and 2 all typical of the genus. Palp: essential features shown in Figures 13-14. Abdomen: quite typical of the genus; ventral scutum rather poorly developed; with a copious supply of short hair. Color in alcohol: carapace and sternum a bright orange red; legs generally somewhat lighter dorsally and still lighter ventrally; ail coxae and mouth parts yel- lowish with variations. Abdomen: greenish gray with many nar- row, light colored lines running in a variety of directions; lateral sides grayish with the gray areas extending to ventral surface and covering the posterior fourth of the venter with the remainder nearly white with greenish spots around spiracles. Considerable variation in coloration has been noted among paratypes with a tendency for a purplish color to replace the greenish gray. Female paratype. Total length of described female paratype 8.58 mm from anterior border of clypeus to posterior end of abdo- men; from anterior border of slightly porrect chelicerae to pos- terior end of spinnerets 8.78 mm. Carapace 3.84 mm long; 2.86 mm wide opposite second coxae where it is widest; 0.58 mm tall; posterior end transversely rugulose; with a series of light colored, fine, intricately branched lines covering entire surface; otherwise as in male. Eyes: two as usual; separated by about % of their di- ameter; height of porrect clypeus equal to about 3.3 times the di- ameter of an eye. Chelicerae, maxillae, lip and sternum essentially as in male and typical of females of the genus. Legs: tibial index of first leg 15, of fourth leg 9; essentially as in male with respect to the ventral lobe on tarsi 1 and 2, the median, ventral fold on metatarsi 1 and 2, and tarsal claws. The palp lacks a tarsal claw but has the usual tibial brush and the femoral, prolateral, basal cusp. Abdomen: essentially as in male. Color in alcohol: es- sentially as in male with minor variations. Records. The described female paratype is from Barro Colorado Island, Panama Canal Zone, January, 1958. About forty mature males together with many females and immature specimens as- signed to this species have been taken in several localities in the Canal Zone and in Panama proper in El Valle, El Volcan, and Arraijan. I have found it especially abundant on Barro Colorado Island, Canal Zone. NoPS SIMLA sp. nov. Figures 16-17 Holotype. The male is from Simla, Arima Valley, Trinidad, W. I., in the close vicinity of the Wm. Beebe Tropical Research Sta- tion, March 31, 1964. The name of the species is a noun used in apposition after the locality where the holotype was collected. 14 BREVIORA No. 274 Description. Total length 4.94 mm. Carapace 1.95 mm long; 1.45 mm wide opposite second coxae where it is widest; 0.44 mm tall; median thoracic fovea barely indicated. Eyes: two as usual; separated by about %i of their diameter; with a moderate amount of black pigment around eyes; height of porrect clypeus nearly equal to 2% times the diameter of an eye. Chelicerae, maxillae and lip with only minor variations from the typical. Sternum: very granulate; obtusely terminated between bases of fourth coxae which are separated by about ^;-, of their width. Legs: tibial index of first leg 14, of fourth leg 10; tarsi typical of the genus with re- spect to bisegmentation and claws; tarsi 1 and 2 typical with re- spect to ventral, basal lobe; metatarsi 1 and 2 essentially typical with respect to ventral, medial, longitudinal fold except that this seems to be somewhat reduced in prominence. Palp: essential features shown in Figure 16; embolus very long. Abdomen es- sentially typical of the genus. Color in alcohol: carapace, lip and sternum a medium reddish brown; palp and other mouth parts yellowish with variations; all coxae light yellowish; other segments of legs reddish brown with variations; legs 1 and 2 somewhat the darkest. Abdomen: background color light yellowish; dorsum with a series of five pairs of light, irregular, purplish spots (Fig. 17); at posterior end there is a pair of elongated, lateral, purplish spots extending forward from the spinnerets; venter generally yellowish. Female paratype. Total length 5.92 mm. Carapace 2.15 mm long; 1.69 mm wide opposite posterior border of second coxae where it is widest; about 0.44 mm tall. Eyes essentially as in male. Chelicerae, maxillae, lip and sternum essentially as in male. Legs: tibial index of first leg 16, of fourth leg 10; other features of legs essentially as in male and typical of the genus. Abdominal features and color essentially as in male. Records. The female paratype was taken in the same locality as the holotype on April 25, 1964, together with an immature speci- men. The following have been on loan from Dr. W. J. Gertsch: two males from Arima, Trinidad, May, 1953 (N. L. H. Kraus) and an immature specimen from Piarco, Trinidad, January, 1955 (A. M. Nadler). Nops TOBALLUS sp. nov. Figures 18-19 Holotype. The male holotype is from St. Catherine Parish, near May Pen, Jamaica, W. I., November 22, 1957. The name of the species is an arbitrary combination of letters. 1967 NOPS IN PANAMA AND THE WEST INDIES 15 Figures 16-17. Nops simla sp. nov. Fig. 16. Left male palp; prolateral view. Fig. 17. Abdomen: dorsal view. Figs. 18-19. Nops tobaUus sp. nov. Fig. 18. Left male palp; prolateral view. Fig. 19. Left female palp; pro- lateral view. Figs. 20-21. Nops iirsitmus sp. nov. Fig. 20. Male carapace; dorsal view. Fig. 21. Left male palp; prolateral view. Description. Total length 3.94 mm. Carapace 1.87 mm long; 1.47 mm wide opposite second coxae where it is widest; 0.51 mm tall; surface very finely granulate; with no definite thoracic fovea. Eyes: two as usual, separated by about their radius; height of por- rect clypeus equal to a little less than three times the diameter of an eye. Chelicerae, maxillae, lip and sternum essentially typical of 16 BREVIORA No. 274 the genus with minor differences; fourth coxae separated by shghtly more than their width. Legs: tibial index of first leg 15, of fourth leg 10; with very slender spines, hardly more than robust bristles; tarsal claws, tarsal ventral lobes and metatarsal, ventral folds on legs 1 and 2 typical of the genus. Palp: essential features shown in Figure 18. Abdomen: ovoid; with many short hairs; surface very granular; otherwise quite typical of the genus. Color in alco- hol: carapace a bright orange with eyes on a black spot divided in front; legs nearly the same color above but lighter below; sternum somewhat lighter than carapace; abdomen with dorsum and lateral sides gray, venter nearly white anterior to genital groove, greenish from genital groove to near spinnerets where it becomes grayish. Female paratype. Total length 6.18 mm, exclusive of somewhat extended posterior spinnerets; 6.5 mm long to tip of posterior spin- nerets. Carapace: 2.73 mm long; 2.02 mm wide opposite second coxae where it is widest; about 0.65 mm tall; otherwise essentially as in male. Eyes, chelicerae, maxillae, lip and sternum essentially as in male with minor differences. Legs: tibial index of first leg 16. of fourth leg 10; essentially typical of the genus in respect to other features. Abdomen also essentially typical of the genus. Records. The described female paratype is from Clarendon Parish, 3 miles north of May Pen, Jamaica, W. L, November 10, 1963. Additional specimens are in the collection from Jamaica as follows: a female from Blue Mts., Main Range, August, 1934 (P. J. Darlington, Jr.); a male from St. Thomas Parish, Morant Bay, May, 1956 (C. C. Hoff); several immature specimens from St. Andrew, St. Catherine, and St. Thomas parishes taken in 1956, 1957 and 1963. Nops URSUMUS sp. nov. Figures 20-21 Nops maculata, — Banks, 1929: 53-96, 4 pis. (Not A'^. maciilata Simon). Holotype. The male holotype was taken in a Berlese funnel on Barro Colorado Island, Panama Canal Zone, between April and June, 1950, by Dr. James Zetek. The name of the species is an arbitrary combination of letters. Description. Total length from anterior border of clypeus to posterior end of abdomen 4.94 mm; from anterior border of clypeus to posterior end of spinnerets 5.14 mm. Carapace 2.07 mm long; 1.52 mm wide opposite posterior border of second coxae where it is widest; 0.59 mm tall; with no discernible thoracic 1967 NOPS IN PANAMA AND THE WEST INDIES 17 fovea; surface finely granulate; with a sparse covering of hair (Fig. 20). Eyes: two as usual; separated by about %3 of their diameter; height of porrect ciypeus about equal to 2.5 times the diameter of an eye. Chelicerae, maxillae, lip and sternum apparently quite typical of the genus as observed in this study. Legs: tibial index of first leg 13, of fourth leg 8; legs 1-3 only slightly different in length; typical of the genus with respect to tarsal claws, ventral, proximal lobes on tarsi 1 and 2 and ventral, median fold on metatarsi 1 and 2. Palp: essential features shown in Figure 21. Abdomen: all features essentially typical of the genus. Color in alcohol: carapace and sternum a rich, dark orange with a moderate amount of black pigment in ocular region. Mouth parts and legs reddish brown with variations. Abdomen: background color of dorsum a light greenish with five pairs of dark grayish spots; the first pair of spots are broad and extend posteriorly for a consider- able distance and are separated by a narrow, greenish stripe in the middle; the remaining four pairs of spots are diagonal bars reach- ing forward and then downward for a very short distance ventro- laterally; the last two pairs are somewhat united and the fourth pair is considerably extended posteriorly; shortly anterior to the anal tubercle there is a single, small, irregular spot somewhat united to the most posterior pair of bars; the venter is whitish with a green- ish tint; a short distance anterior to the spinnerets there is a short, median line of irregular gray dots and lateral to these is an irregu- lar dotted area on each side. Female paratype. Total length from anterior border of ciypeus to posterior end of abdomen 9.42 mm; length from anterior border of ciypeus to posterior end of spinnerets 9.94 mm. Carapace 3.45 mm long; 2.67 mm wide; 0.85 mm tall; otherwise essentially typi- cal of females of the genus. Eyes: two as usual; separated by about their radius; height of ciypeus nearly equal to 3.5 times the dia- meter of an eye. Chelicerae, maxillae, lip and sternum essentially typical of females of the genus; sternal suture line unusually clear but probably does not indicate freedom of movement for the lip. Legs: tibial index of first leg 12, of fourth leg 9; tarsal claws, ventral, proximal lobe on tarsi 1 and 2 and ventral, median fold on metatarsi 1 and 2 all typical of the genus. Abdomen: essentially typical of females of the genus; region from genital groove to pedicel rather poorly supplied with a scutum. Color in alcohol: essentially as in male with minor variations; here the two posterior, dorsal, abdominal bars are united to one irregularly shaped spot; the venter has scattered dark spots, the largest of which occur as a 18 BREVIORA No. 274 pair of very irregular spots shortly anterior to the spinnerets. Records. The described female paratype was taken in the same locality as the holotype, January, 1958. One mature male para- type was taken with the holotype; another male was taken by Dr. Zetek in the same locality and by the same method, Oct. -Dec, 1941. A very few females which may be mature and nearly three dozen others in various stages of immaturity are in the collection from Barro Colorado Id., C. Z. Forest Preserve, Summit Gardens, and Pedro Miguel, all in the Canal Zone. The immature specimen reported by Banks (1929) as N. macidata Simon quite clearly be- longs in N. ursumus sp. nov. The following have been on loan from Dr. VV. J. Gertsch: a male and a female from Barro Colorado Id., Dec. 1937-Jan. 1938 (F. J. Ryan); three immature specimens from the same locality taken July, 1930, and July-August, 1938; an immature specimen taken in Santa Rosa, Colon Prov., Panama, Sept., 1945 (C. D. Michener). BIBLIOGRAPHY Banks, Nathan 1929. Spiders from Panama. Bull. Mus. Comp. Zool., 69: 53-96, 4 pis. Berland, Lucien 1932. Encyclopedic Entomologique. 16. Les Arachnides. Paris. Bonnet, Pierre 1958. Bibliographia Araneorum. Toulouse. Vol. 2(4). Bryant, Elizabeth B. 1940. Cuban spiders in the Museum of Comparative Zoology. Bull. Mus. Comp. Zool.. 86(7); 249-532. 22 pis. 1942. Notes on the spiders of the Virgin Islands. Ibid., 89(7): 317- 363, pis. 1-3. 1948. The spiders of Hispaniola. Ibid.. 100(4): 331-447, 12 pis. Cambridge, F. O. P.- 1897- Arachnida-Araneida. Vol. II. //;: Biologia Centrali-Americana. 1905 Dulau & Co., London. Hasselt, a. W. M. van 1887. Etudes sur le genre Nops. Tijds. Entomol., 30: 77-86, figs. 1-4, 10. LuTZ, Frank E. 1915. List of Greater Antillean spiders with notes on their distribu- tion. Ann. New York Acad. Sci., 26: 71-148. Macleay, W. S. 1839. On some new forms of Arachnida. Ann. Natur. Hist. 2: 1-14, pis. 1-2. Petrunkevitch, Alexander 1911. A synonymic index-catalogue of spiders of North, Central, South America, etc. Bull. Amer. Mus. Natur. Hist., 29: 1-809. 1967 NOPS IN PANAMA AND THE WEST INDIES 19 ROEWER, C. Fr, 1942. Katalog der Araneae. Vol.1: 1-1040. Bremen. Simon, Eugene 1891. On the spiders of the Island of St. Vincent. Pt. 1. Proc. Zool. Soc. London, Nov. 17, 1891: 549-575. 1892- Histoire naturelle des Araignees. Deuxieme edition. Vol. 1. 1895. Librairie Encyclopedique de Roret, Paris. 1892. Voyage au Venezuela. Ann. Soc. Entomol. France, 1892: 449. Walckenaer, C. a. Baron de 1842. Historie naturelle des insectes: Apteres. Vol.2. Paris. (Received 20 March 1967.) BREVIORA Mmiseiiiiim of Connparative Zoology Cambridge, Mass. 17 November, 1967 Number 275 A NEW SCiNCID LIZARD FROM THE NORTHERN SOLOMON ISLANDS Allen E. Greer and Fred Parker' During a term of duty by Parker (1962-1963) and a subse- quent ( 1966) collecting trip by him to Bougainville and neighbor- ing islands in the extreme northern Solomon Islands, a previously undescribed species of skink was discovered on Bougainville and Shortland Islands. A single specimen of the new species from Choiseul is also known but has previously been reported in the hterature as Sphenomorphus solomonis (Burt and Burt, 1932:544). On the basis of current, but not necessarily phylogenetically correct, generic concepts, the new species is assigned to the genus Sphenomorphus and may be known as Sphenomorphus tanneri- new species Holotype: Museum of Comparative Zoology 76551; collected by Fred Parker at Kunua, Bougainville, on 8 June 1963. Pamtxpes (423 specimens) :"nORTHERN BOUGAINVILLE: KUNUA (100 feet above sea level): MCZ 76483, 76484, 22.vi, ll.vii.l962; MCZ 76486-76487, 25-30.xii.1962; MCZ 76488- 76550. 76552-76623. 77308-77361, 78091, 78299, 84140, + 2 untagged specimens, 12.v-29.vii.l963; MELILUP {ca. 3000 feet a.s.l.)^ MCZ 89593. 92295, 92326-92339, 92366-92368, 12- 20.V.1966; MUTAHI (2700 ± 500 feet a.s.l.): MCZ 87570- 87577, 88799, 89102-89126, 89594, 91430-91437. 92203-92228, 92269-92294, 92296-92302, 92304-92325, 92340-92365, 92372- 92377, 9-20.V.1966; RAMAZON RIVER (1600-2400 feet a.s.l.): ^ Department of District Administration, Kimdiawa, Territory of New Guinea. "The new species is named for Charles Tanner, Honorary Curator of Reptiles at the National Museum, Melbourne, who introduced one of us (Parker) to reptiles several years ago and has since helped in many ways. 2 BREVIORA No. 275 MCZ 92229-92236, 19.V.1966; TOPANAS (500 feet a.s.l.): MCZ 88443-88453, 8.V.1966. SOUTHERN BOUGAINVILLE: MATSIOGU (2100 feet a.s.l.): MCZ 92379-92382, 92522, 2.iv. 24.iii.1966; PAMAUITA (1200 feet a.s.l.): MCZ 92521, 92523, 23. ii, 2.iv.l966; TURI- BOiRU (260 feet a.s.l.); MCZ 87578-87580, 87727-87732, 92496-92501, 92520, 92990-92992, 9-21. iii. 1966; MCZ 92369- 92371, 26-27.iv.I966. SHORTLAND: KOLIAI (10 feet a.s.l.): MCZ 89736, 1 l.iv.l966. CHOISEUL: American Museum of Natural History 44004. Dioij^nosis: S. kinneri differs from all other species with 28-32 midbody scales in its species group (Table 4) in the following combination of characters: small size (snout-vent length 52 mm or less); prefrontals forming a median suture; adpressed limbs widely separated; no symmetrical pairs of nuchal scales; m!ddorsal scales subequal rather than with the scales of the two vertebral rows transversely enlarged; dorsum rich dark brown to brownish black with small paler spots on body which often coalesce on the anterior dorsolateral line to form a distinct light line. S. tanneri is most similar in squamation to S. nigriventre de Rooij (1915: 214-215) from southern New Guinea, and S. antoniorum Smith (1927: 216-217) from Timor. It differs from S. nigriventre in being much smaller in size (snout-vent length 90 mm for nigriventre) and lacking the dorsal transverse series of light, dark- edged spots of this species. It differs from S. antoniorum in having the prefrontals meeting medially (prefrontals separated or just touching one another in antoniorinn), and in lacking the light brown dorsum with a heavy clustering of darker brownish spots along the dorsolateral line. Description (Fig. 1): A cryptic skink ranging in snout-vent length from 23-52 mm; tail slightly longer than snout-vent length; head obtusely conical; limbs pentadactyl, failing to meet when adpressed to body. Rostral slightly wider than deep, projecting well onto dorsal surface of snout; external naris placed well forward and ventral in a single large nasal; supranasals lacking; frontonasal wider than long, forming a short suture with rostral but separated from frontal by prefrontals, i.e., prefrontals paired and meeting medially (ex- cept in one individual); single anterior and posterior loreals; fron- tal about as long as frontoparietals and interparietal measured along midline, obtusely rounded posteriorly and in contact with 1967 NEW SOLOMON ISLANDS SKINK B Figure 1. Splienoinoiplius taiincii (MCZ 92304, paratype): (A) Dorsal view of the head and nape, showing two successive "nuchal" scales on the right side of the nape; (B) nasal area on the left side of the head, showing the external naris situated anteroventrallv in the large nasal scale. 2 anteriormost supraoculars; 4 supraoculars; lower eyelid scaly, scales of eyelid separated from supralabials by a complete row of subocular scales; frontoparietals and interparietal distinct, approx- imately subequal in size; parietals meet behind interparietal; no distinct, symmetrical series of nuchais; 6-7 supralabials, 4th or 5th supralabial below center of eye (Table 1 ). Ear opening oval, tympanum slightly sunk below surface; auric- ular lobes lacking; midbody scales smooth, in 28-32 (usually 30- 32, see Table 1 ) longitudinal rows; midbody scales of dorsum subequal in size; a pair of enlarged preanals; 3 median rows of subcaudal scales subequal in size; 13-20 (usually 15-19, see Table 1 ) obtusely keeled lamellae beneath 4th (longest) toe; dorsal surface of 4th toe covered distally by a few single scales, medially by 3 longitudinal rows of scales and proximally by 4 or 5 rows (Group IV of Brongersma, 1942). Color (Fig. 3): In life the color pattern is relatively constant. The dorsum is a rich dark brown to brownish black with small paler spots on the body and tail. Larger light spots on the anterior 4 BREVIORA No. 275 dorsolateral line often coalesce to form a noticeably distinct line. The flanks are marked as the dorsum but usually lighter. The chin and throat are heavily marked with black to brown; the an- terior venter is whitish to yellowish; the posterior venter and ventral surface of the tail arc occasionally translucent yellow, but more often, translucent dull reddish. Some specimens have dark spots under the tail and others show a bar of fine spots on the posterior part of the throat. Preserved specimens lose the yellowish and reddish colors. Some individuals from the southern Bougainville (Turiboiru. Pamauita and Matsiogu) populations lack the light dorsal spotting and tend to have the dark pigment of the dorsum and sides con- centrated through the center of the scales. This gives the appear- ance of dark longitudinal stripes on an only slightly lighter back- ground. Distribution (Fig. 2): S. tanncri has been collected at both the north and south ends of Bougainville (422 specimens), as well as on Shortland (1 specimen) and Choiseul (1 specimen). Altitud- inally, the species is known to range from the coastal lowlands up to about 3000 feet. Variation: The variation of several meristic characters is sum- marized in Table 1 . The degree of contact between the prefrontals ranges from a broad median suture to a bare "point contact" at their inner angles. In only a single specimen of the whole series, however, do the prefrontals just fail to meet at their inner angles. In the ma- jority of specimens, the line of contact is at least one-third to one- half the length of the prefrontal. Large, symmetrical nuchal scales are definitely lacking, although the sporadic occurrence of one or more (serial) large, transverse scales in the 4-5 scale rows posterior to the parietals ( Fig. 1 ) sug- gests that perhaps the genetic basis for the formation of nuchal scales is still being variably expressed. Such a high degree of variability might well obtain in a previously well-ordered structure that was being secondarily lost. Either the 4th or 5th supralabial may be below the center of the eye. When it is the 4th supralabial, either the 2nd or 3rd supra- labials have fused to form a scale as long as the first supralabial or a small wedge-shaped scale may project between the 2nd and 3rd supralabial, thus separating them except along the edge of the lip where they remain in contact. In most individuals, however, the 5th supralabial is below the center of the eye. 1967 NEW SOLOMON ISLANDS SKINK The 4th supralabial occurs below the center of the eye with greater frequency in the populations from southern Bougainville (Turiboiru. Pamauita and Matsiogu = 37% ) than in the'popula- tions from northern Bougainvihe (Kunua, Mutahi, Topanas, 155° E 156° E 30 miles #Topanas .Ramazon River ■6°S 7°S 6°S SHORTLAND I FAURO 7°S- I55°E MONO 1 (^ 156° E Figure 2. Bougainville and surrounding islands, showing the localities from which Splwiioniorpluis tcinncri is presently known. There is also a single specimen from an unknown iocalitv on Choiseul. 6 BREVIORA No. 275 Melilup and Ramazon River = \VA%). The frequency of this trait is probably subject to clinal variation from one end of Bou- gainville to the other, but the rate of change per unit distance will not be known until populations from the central two-thirds of Bougainville are sampled. This particular difference in squamation between northern and southern populations on Bougainville is made even more interest- ing by noting its correlation with the difference in color patterns between the two populations (see above) — the striped pattern occurring in some individuals in the south but never in the north. These differences may be taxonomically important, but their sig- nificance can not be properly evaluated until populations from intervening localities are studied. Type specimen: The holotype has 30 scales around midbody, 16-16 lamellae beneath the 4th toe, and the 5th supralabial is below the center of the eye. Shortland Island specimen: The single specimen from Shortland Island (Fig. 2), the only other island besides Bougainville and Choiseul from which the species is known, does not differ substan- tially from the Bougainville specimens. There are 30 scales around midbody and 17-17 lamellae beneath the 4th toe. Comparisons between the Shortland population and the northern and southern populations on Bougainville would be interesting, but can hardly be carried very far at present with only one speci- men available to represent the Shortland population. It is interest- ing to note, however, that while the Shortland specimen lacks the striped pattern characteristic of some individuals from southern Bougainville, it does have the 4th supralabial below the center of the eye on both sides of the head. This condition, as discussed above, tends to occur more frequently in lizards from southern Bougainville (37^r) than those from northern Bougainville {WVi'c). Choiseul Island specimen: The single specimen from an un- known locality on Choiseul (AMNH 44004) is poorly preserved but appears to differ in no significant way from Bougainville and Shortland specimens. Habitat: On Bougainville the species is found in its greatest numbers in the swampy lowlands near the coast where small creeks spread out and intermix and where the swampy species of the Pandanus plant grows. The land is subject to occasional inun- dation during heavy rains, but there are always dry knolls onto which the reptiles and amphibians can retreat. The usual micro- habitat in the swampy lowlands is under and in rotten logs and 1967 NEW SOLOMON ISLANDS SKINK 7 thicker layers of decaying vegetable matter, while on higher ground and in the mountain ranges the species is usually found only living under and in rotten logs. The Shortland Island specimen was found under a pile of decay- ing fronds in a coconut plantation. Habits: S. tcmneri is a burrower and shows a preference for very damp country. Other than this, little is known about the species' habits. No individual was ever seen out in the open of its own accord. It escapes by burrowing when cover is removed and it is disturbed. Reproduction: S. tcmneri is oviparous and remarkably consistent in laying only 2 eggs at a time. Of the 49 females found to be gravid with oviducal eggs, 47 had snout-vent lengths ranging from 41-50 mm (Table 1). Given the large total sample size (424 specimens) from which these females were taken, the low end of the size distribution may represent the size at which females become capable of reproducing. Of the 49 gravid females, 47 contained 2 eggs; 46 of these held 1 egg in each oviduct and a single specimen contained 2 eggs in the right oviduct alone. Of the 2 remaining females, one con- tained 3 eggs, 2 in the right oviduct and 1 in the left, and a second female contained only a single well-developed egg in the right ovi- duct. In all cases in which there were eggs in both oviducts (47), those eggs in the right oviduct were invariably situated more an- teriorly; this is apparently an accommodation for the stomach, which lies on the left side. The 49 gravid females were collected in March (2), May (20), June (5 ), July (21 ), and December ( I ). Eggs collected in the field on 17 May 1966 measured 14.8 x 8.0 (with an embryo), 13.0 x 7.5, 13.4 x 7.0 mm (unopened). Morphological comparisons n-ith Bougainville relatives: The only other sympatric members of tanneri's species group (Table 4) on Bougainville are Sphenomorphus solomonis and S. cranei. Both species can be easily distinguished on the basis of the char- acters listed in Table 2. Tanneri and solomonis are similar in their dorsal color patterns which consist of white flecking on a dark ground color. In tanneri, however, the ground color is a rich dark brown to brownish black while in solomonis the color of the dorsum ranges from brownish black to gray. The white spots along the dorsolateral line never coalesce to form a light line in solomonis as they do in tanneri. The ventral color patterns of tanneri and solomonis are quite distinctive. In tanneri the venter of the body is immaculate while 8 BREVIORA No. 275 the chin and throat are heavily pigmented. Solomonis, on the other hand, usually has the venter of the head and body uniformly spotted throughout, each ventral scale of the body containing a central dark spot. The dorsal ground color in cranci varies from very dark dull brown to blackish brown and is patterned with thin, whitish to yellow transverse bars from the nape of the neck onto the tail. In life the belly and underside of the tail in tanneri and cranei are often suffused with a translucent reddish color. In soloinonis the ventral surfaces lack any reddish coloration. Ecological comparisons with Bougainville relatives: The species most similar in ecology to tanneri on Bougainville is Sphenomor- phus solomonis, one of tanneri's closest relatives on the island. Both species, along with two species of Tribolonotus (blanchardi and an as yet unnamed species), are the most extreme burrowers in the Bougainville lizard fauna. Solomonis is much more common than tanneri and has a slightly greater altitudinal range — from the coast to at least 3000 feet (and perhaps up to 4000 feet) in the mountains. Solomonis occurs in a wider variety of habitats than tanneri, being most commonly found in the top few inches of soil and humus and under logs in open gardens, secondary scrub, tall pri- mary forest, dry coastal country and swampy areas. In the moun- tains, solomonis is much reduced in numbers as the altitude in- creases and is found mainly on the ridges and in secondary scrub. In the lowlands very large numbers of solomonis can be taken when logs and vegetation are being cleared off the ground for native gardens and when the topsoil is first worked. As fewer cranei have been collected than solomonis and tanneri, field observations for this species are less complete than for its two closest relatives on Bougainville. Cranei, however, also seems to be a burrower, although much less so than solomonis and tanneri. Unlike solomonis and tanneri, cranei has not been collected in the lowlands, but is known from the mountains and steep-sided valleys above 500 feet where there is tall, cool primary forest. Cranei is much more highly dependent on water and moisture than tanneri and solomonis and is most usually found under stones, wood, or debris in small creek beds with running water or soaked beds. In this habitat it is found in company with Tribolonotus blanchardi and Sphenomorphns (Parotosannis) concinnatns, but cranei is by far the least common of the three species. 1967 NEW SOLOMON ISLANDS SKINK 9 Tanneri and solonionis are both oviparous, but the mode of reproduction in cranei is unknown. As noted above, tanucri is re- markably consistent in producing only two eggs at a time. Soloni- onis, on the other hand, produces clutches with 3-6 eggs. Discussion: Current research (by Greer) on the delimitation and relationships of skink taxa indicates that there are several species groups now included in the genus Sphenomorphus. The two larg- est species groups of the genus are discussed below. One of these, which may be called the variegatus species group (Table 3), also includes skinks of the genera Otosaurus, Paroto- sauriis, and Insulasauriis and is characterized by the following suite of external characters: well developed digits and limbs, which overlap in most species when adpressed to the body; frontal gen- erally in contact with 3 or more of the anteriormost supraoculars; generally 5 or more supraoculars; supranasal scale (large in Ofo- sourus, small in Parotosauriis) or double anterior loreal present in most species, but some species with no supranasals and only a single anterior loreaP; nuchals lacking; high number of scales around midbody (usually 32 or more); dorsal scales subequal in size. The other large species group within Sphenomorphus, the one to which tanneri belongs and which may be called the fasciatus species group (Table 4), is characterized by the following external char- acters: digits and limbs usually less well developed, the limbs gen- erally not overlapping, or just slightly overlapping when adpressed to the body; frontal in contact with the anterior 2 supraoculars; 4 (much less frequently 5) supraoculars; a single anterior loreal; no supranasals; usually a series of two or more symmetrical nuchal scales; fewer scales around midbody (generally 36 or fewer); scales of the 2 vertebral rows usually broader than the other dorsal scales, i.e., they are transversely enlarged. Whole skulls of several species of both species groups have also been examined. Both species groups display several important similarities in their skull morphology, but a discussion of these ^ As the small supranasal of some Paiotosaunts and the double anterior loreal of some Sphenomorphus may be variable in shape and form, or even present on one side of the head and absent on the other in a single individual {Parotosauriis concinnaliis. personal observation and Burt and Burt, 1932:542), it does not seem justifiable to exclude from the variegatus species group those species which agree with the diagnosis except for lacking a supranasal or a double anterior loreal. 10 BREVIORA No. 275 similarities is best deferred to a later date. The three important differences in skull morphology between the variei^afus and solo- monis groups are as follows: variegatus group 1.) Postorbital bone lacking or very small. 2. ) Supratemporal fenestra usu- ally lacking or very small. 3.) No anteriorly projecting ec- topterygoid process to the palatine which would ex- clude the palatal ramus of the pterygoid from a posi- tion on the edge of the in- fraorbital vacuity. solomonis group 1.) Postorbital bone usually long and thin. 2.) Supratemporal fenestra generally well developed. 3.) Two subgroups within the fascia t us group: solomonis subgroup (Greer, 1967) with ectopterygoid process to palatine which excludes all or most of pterygoid from a position on infra- orbital vacuity; fasciatus subgroup without anteriorly projecting ectopterygoid process to palatine.^ The two species groups may also be distinguished ecologically and geographically. The skinks of the variegatus group appear to be primarily surface dwellers, while those of the fasciatus group are, by and large, secretive burrowers as is tanneri. Geographically, the variegatus group ranges from southeast Asia and the Greater Sunda Islands northeast to the Philippines and east through Celebes and the Lesser Sunda Islands to New Guinea and the Solomons. The group is not found, however, in Australia. The center of abundance for the group appears to be the western part of the Indo-Australian archipelago and. perhaps to a lesser extent. New Guinea. The fasciatus group has its center of abundance on New Guinea, but extends northwest to the Philippines (fasciatus), west into the Lesser Sunda Islands (emigrans and antoniorum) , south into north- ern and eastern Australia (e.g., crassicauda, elegantuluin, punctu- latum and mjobergi), and east into the Solomon Islands (cranei, solomonis and tanneri). 1 S. tanneri lacks the ectopterygoid process to the palatine and is therefore a member of the fasciatus subgroup of the fasciatus species group of SpiH'in'>inorphus. 1967 NEW SOLOMON ISLANDS SKINK 11 The two species groups considered here appear to be mono- phyletic. although convergent tendencies in both groups (e.g. the loss of nuchal scales, of transversely enlarged vertebral scales, and of the postorbital bone in the jasciaius group and the presence of only 4 supraoculars in the variegatiis group) make formal taxo- nomic treatment difficult. For this reason the ranking and further delimitation and subdivision of the species groups will be consid- ered at a later date. ACKNOWLEDGEMENTS Our thanks go first and foremost to Miss A. G. C. Grandison of the British Museum (Natural History) who made extensive and detailed comparisons between S. tanneri and several types in her charge. Dr. Ernest E. Williams of the Museum of Comparative Zoology has read the manuscript in several drafts and has offered many helpful criticisms. Mrs. Patricia H. Kerfoot made the line drawing for Figure 1 and Mr. Rick Stafford took the photograph for Figure 3. Travel expenses incurred by Greer in the course of his reseaich for this paper have been defrayed by support from the Evolutionary Biology Fund which is administered for the National Science Foun- dation by the Evolutionary Biology Committee of the Biological Laboratories. Harvard University. LITERATURE CITED Brongersma, L. D. 1942. On the arrangement of the scales on the dorsal surface of the digits in Lygosonui and allied genera. Zool. Meded. Leiden, 24 (1-2): 153-158. Burt, C. E. and M. D. Burt 1932. Herpetological results of the Whitney South Sea Expedition. VL Bull. Amer. Mus. Nat. Hist.. 63 (5):461-597. Greer, A. E. 1967. A new generic arrangement for some Australian scincid lizards. Breviora, No. 267:1-19. RooiJ. N. DE 1915. The reptiles of the Indo-Australian archipelago. L Lacertilia, Chelonia, Emydosauria. Leiden, E.J. Brill Ltd., 384 pp. Smith, M. A. 1927. Contributions to the herpetology of the Indo-Australian region. Proc. Zool. Soc. London, 1927:199-225. (Received 23 March 1967.) 12 BREVIORA No. 275 TABLE 1 Frequency distribution for several nieristic characters in Spheno- morphus tarmeri. N = number of specimens examined. Midbody Scales (N — 156) X 28 29 30 31 32 f(x) 5 5 115 11 20 Supralabial below Center of Eye (N = 415) X 5-5 4-5 4-4 f(x) 337 43 35 Number of Lamellae beneath 4th Toe (N = 228) X 13 14 15 16 17 18 19 20 f(x) 2 3 20 62 90 32 17 2 Snout-Vent Length of Gravid Females (N = 47) x 41 42 43 44 45 46 47 48 49 50 f(x) I 3 5 8 7 6 7 5 3 2 TABLE 2 Morphological differences between S. tanneri and its two closest relatives on Bougainville. tcinnevi solomonis cranei Snout-vent length 23-52 mm 27-72 mm 29-1 \ mm Scales around 28-32 24-30 32-41 mid-body (usually 30) (usually 2( 6-30) (usually 34, 36) Prefrontals meet Yes, except in 1 No Yes, except in 2 of 424 specimens of 23 specimens Supralabial below 4th (137f) or 4th (< 1/2%) or 5th {W/c ) or center of eye 5th{879r) 5th (> 991/2%) 6th (897r ) Nuchal scales Absent Present, 2-4 pairs Present, 3-5 pairs Subdigital lamellae 13-20, 13-17 20-29, most (4th toe) usually 15-19 usually 22-28 967 NEW SOLOMON ISLANDS SKINK 13 Figure 3. Holotype of Sphcnomorpluis tanncri (MCZ 76551). showing the light spotting on the dark brown dorsum. 14 BREVIORA No. 275 CQ < 1 -a t/5 kH '4—' VO r- o C C3 C3 L- rt (U ji: a. (U o c/5 > _o dJ > c •5og C3 -^ C T3 o Ti ^ (y3 ■(-> -o O OJ >■ ^ (M ~^l 52 C t/3 3 C '^ s: c CO a. ~ .i; "t: .— £ ^ c o 5^ OX) 5 t^ o 1/3 ^3 dJ O O 53 C := 0, ^ t;; rt •-= -o ^ (U * S 2: O S2 '^ S " K! c;r r- O oj Ji 'U '=^ V5 — i: > 3 > > O W OJ 03 Z Z CI On joqiny — + + + + + ri o o ir, ON C c: 2 O ~ c S ~ t^ 00 + + + ^ TT 4 '^ 4J o 0) Xi «j x> OJ c aj u 4J o 4J U 03 U ca± O- i- m^ 00"^' 05 a a O :; ac 1967 NEW SOLOMON ISLANDS SKINK 15 + (^1 00 ) 3C I/-, sC 30 + + + 1 + + + +1 + + + + + + + + + + + ++ + ++ + + + + + + ON oo ri ri 30 On rj I/-, — — ri ri vC sC 00 r'l ri ri r I — ri n — >/-. o r~- n n — I 00 00 I I (^1 — n n -1- ^ r«-. ^ -^ ^ ■? oc oc »a 03 rt u (U c c o ^ O •o X) == 5 Oj ^ u w ^.2 4. U -1^ CO C5 -1^ 03 C/3 O c « O = _c '5. a c c z z c o o a z c<3 C •-. — c ^ -^ = ^ L- o o b c a :^ O t: — 03 c/5^ 00 P. Z 03 4J 0/J U 00 ?;3- 0/J c OJ O 00 222^2: m22 =i =x = M o £1 sC ON 00 On t^ " 1j ■- E 4> c II C/5 — ON 1) c ti 00 ^ S 00 O Ov 0- — K> c/5 — oa — 1^ 00 c OJ 00 c OJ 00 c OJ ON .^ ^> ^i> 5 — t^— O— "^i— <^. o 2 03 =; ON >"ri 2 o y ooo OaoSjjs Ojjv « 03— 03— h— 03— 03 ■*5 1^ s: •Si Si -^ * S ^ to C ■5 c: o 16 BREVIORA No. 275 r<-i ^^ ^ m ^-K ri ^^ ^ , . r*". . . r'l r<-, (N ' — ' ro ^^ -nC r^. r^. r'l t ^^ r'l \D ^ — ^ -J r^ ■ — ■ "-' ' — " ■ — ' ■ — ■ '/~i ■ — ■ >/-. ir, ^ O ^x: r-- r- v~, iri r- 4 ^ + + + u + + + + 1 + + +1 Di O H^ OS 1 + + UJ + + + + + + _1_ 1 + Z < >/-. r~- ON fN "/-i O — -J CQ ri ri ri ri r 1 1^ :5V. o E Q c o ^2 CON OJ ^ ON oc o ■X, 'cc 1- o CLl 0H y5 ^ "Jj ^c:. 1967 NEW SOLOMON ISLANDS SKINK 17 +1 I +1 I + + + + + + + + + o rj n ^ r^, \o VO r 1 -r rj 00 00 ri 00 6 1 O 53 £ no ■« < ox: x: C/J « > rt 1) Cu z 5^ v5 Z J3 P?, U o«0O ^•- C O^ -iH— O— > CQ— PoacQ — 18 BREVIORA No. 275 UJ I < 1/5 -t-; ■•s i O ^ o (U a 3 en rt p ^ C o on ^4— > <1J s u r* o D- (U 1/2 c ^ ■>-' T3 ~ c3 ::: o 2^ S ^-^ ^ S o c^ ^ T3 C/5 rt g 3 p O C C o (U c l- "^ '-^ y • — ■ uo f3 C (U >- •- 3 rt '^ O 73 •5 r^ -C O^ '« I- o t/^ en on x: .Si C^ "^ *1 -i _ ^- /~i ^ SJIKd , I >^i [RqonM + +1 1 + + 1 (IBjuojj SuiqDnoi) — — sjB|nDORjdns 3 1^ SlK)UOJJ3JJ J33UJ squiij passajdpv 301 qjt' uo anipLutq + + + + -t O "vC — rj r) '^ iBjiSipqns 3 ?l ::: i^ ri — Apoqpau ^ _ punoji; rr> ^ joqjnv 3^ ♦-' — -a c - c ri — r^ c Of- + n + I + .— Ci S31EDS r<-, CS Ct! CS « rt rt c c 1) C ■a c — C 3 2 2 ^ 3 "TZ sSUBy u 0 -0 0 a 0 a 0 'J-. ^ 0 C ^ ^ ^ s 5 — u 0 "* ^» a> C/5— 03— C/)— T3— 03— ffl — ^ — "^ t: t— ^ — ^ — > ^- — • 00 O ON .? ON .? 00 OQ — 1967 NEW SOLOMON ISLANDS SKINK 19 + l|^-+l+ + ^. +++ +++++ + vc ri ir, a\ — o o — ri r^. ■Xi o or-~ ri3C — r) <: ""i 1/-, r~~ t~- 0^ ^ vC o^ 1^1/", -x, ^ >/-, 'y, \C <^' ON I I I I I I I I I I +I++I + I ++++ +++++ + rj n f"' n n ri n n ri n ri n ri n n n ri ^ ^ '!:)■'* vC'^ ^ ir-, 'Tf ^ -f "^ ^-^ ^1/-. '^i^r + + IM I+ + +I+I I II++I + + I++I II I 1+1 + r- c ON OO c; >o ^c o (^ — n — D ri — n ri vc r«-, •/^i -t vC vC oc 'J-, n r4 -t sC 'j: 00 sC rj 00 OC n ri M ;-, o c OS c^, r^- r^ ri^, r^. r<-i r-l c 00 CO — , — c O' (3 3 OC OC vC vC OC 00 00 00 OC r^, rl ri r<-. r^, r^, r<-j r^> r<", ri r J ri r) ri ri «3 n n rt a ,— _C v: rt Ki a ^ '•* — z < Z ^ Z z c/) r~- 00 _4J OS _c' OS ..I, oJ c ._i^ 00 •— ' >> OS OJ an OS, c^ OC' OS ■— ^ c D. 'oj ^~~ 'o C/J 1) '5 If C3 ^" c *~~ 'o flj V5 J- , o « o ' _aj r 1) o y: a ^ o o f— c Q. O ON 00 o OJ OS > 00 rt oo a. o n OS OJ OS >, a; O OS 00 00 O o OS 00 5 Q O CO ^ :/^ 1^ > "^ m — 1^ ■a — 6 m > 03 ai c > tj — 20 BREVIORA No. 275 + c~. ^. c-. + + + + + C-. ++ +^- + 1/-, >y-i v,D oo r-- o\ — )- r— T ""• "'i ""' + +I+ + + + + + + + + ++ + + + iri O '>C ■^ r<-, tT O —I o oo "/". "* r<-, r^, r^, r^, r^i ^. -* Tt- r^, ^ tT rl — n rj n n n ' ■ ri ri rj Ci ^ r^l rJ n ri JT]" */"j ^ ^ •* -t "=1- "* ^ 1- C-- '^ tT ^ Tf t -* '^ + I + I I I I I I I +1 I I I I +1 I + I c^. I I I I I + + I I I I III II I ,-.1111 t oc oc n oc "'. ~ ri oc oc t, < >/-, >/-. (^i r-~- — ri ^ vO "^ o O n rj ri (N cs ri (N (N n n ri n ri n ri n n (n ■ crj2._-~ — S'^ bo — "" C^'CCCf, c ■- ^ ^ c g_vc 2°i--£t3— r-— OC— '*~f^ —'^ Mit/jE— - i-o"^'— f~~ rt'aj 1^— Di-o — hJ^^QS— CQ— 03— ^— CO— >>Q_Oii_l<— OQ— Ucu I - t •= ^ -y, c o < c Q. ^ c ~ s =^ .s O >; i •;:; _^ -c: >* ^ ■*•• c-.r -^ ^ \r: —t^ ^ ^ '^ ^ "Sj ^ ■^ • ^ :::; ^ -O N? •5 ^ ^•~-5-SS-"^ ~5=:5 5T^ ~$S BREVIORA Miasemim of Coimparsitive Zoology Cambridge, Mass. 24 November, 1967 Number 276 THE GENERIC RELATIONSHIPS OF THE AFRICAN SCINCID GENUS EUMECIA Allen E. Greer In 1870 Bocage described a new skink (anchietae) from Africa that was peculiar in having an elongate body form, small ap- pendages with a reduced number of digits (2 fingers and 3 toes), a pair of supranasals meeting behind the rostral and a spectacle in the movable lower eyelid. Bocage placed the new species in a distinct genus which he named Eumecia, apparently to emphasize a similarity which he believed to exist with Eumeces. The three subsequent revisionary studies on scincid genera rightly recognized the closer afiinity of Eumecia with what have come to be called the lygosomine skinks rather than with Eumeces or its subfamily.^ Boulenger (1887) placed anchietae in the Riopa section of the genus Lygosoma, and in 1897 described a species (johnstoni) from Nyasaland that was distinguishable from anchietae primarily on the basis of a further reduction in the number of digits (1 finger and 2 toes in johnstoni; 2 fingers and 3 toes in anchietae) . Like anchietae, johnstoni was placed in the section Riopa of the genus Lygosoma. Smith (1937) revived Bocage's name Eumecia for a subgenus of the genus Riopa and included in it only anchietae and johnstoni. Mittleman (1952) agreed with Smith's conception of the taxon Eumecia but gave it full generic rank in his classification. Loveridge (1953 and 1957) placed both anchietae and johnstoni in the genus Riopa without recognizing subgenera, thus adopting ^ The subfamily Lygosominae is characterized by a single frontal bone, and palatine bones which meet along the midline of the palate to form a secondary palate. Eumeces is considered to be a scincine, a subfamily characterized, in part, by a divided frontal bone, and palatines which do not meet along the ventral midline of the palate (Greer, MS). 2 BREVIORA No. 276 a basically Boulengerian conception of the species' taxonomic po- sition. The same author (1953) pointed out the very real pos- sibility that jolinstoni was simply a race of ancluetae and further noted that some anchietae had three fingers instead of two. That anchietae and jolinstoni should be accorded separate supra- specific rank as recognized by Bocage (1870), Smith (1937). and Mittleman (1952) is suggested by the fact that they are the only African lygosomine skinks with supranasals to have fewer than 5 fingers and 5 toes. Indeed it was on this criterion that Eumecia has been regarded as distinct from Riopa and its supposed relatives. In this paper, the two closely related species anchietae and johnstoni^ are considered to constitute a distinct genus Eumecia, for reasons that will become apparent in the following discussion, and the genus is shown to be most closely related to Mabuya and not Riopa. For the purposes of this discussion Riopa is understood in the sense of Smith's (1937) subgenus Riopa and Mittleman's (1952) genera Riopa, Squamicilia and Mochliis collectively. It is not clear to me why Eumecia was always thought by Boulenger and later authors to be more closely related to Riopa and its sup- posed relatives than to Mabuya. On the basis of externals there is no one character that will serve to distinguish all Riopa from all Mabuya. There is however one external character that will dis- tinguish some Mabuya from all Riopa, namely the relative posi- tion of the prefrontal scales, and in this regard Eumecia is like Mabuya rather than Riopa. The prefrontals are never in contact in Riopa, but they do meet medially in Eumecia and in about one-third of the species of Mabuya (surveyed from Boulenger, 1887). Comparisons of the skulls of 30 species of Mabuya and 13 spe- cies of Riopa indicate that there are important differences in skull morphology between Mabuya and Riopa, and on the basis of skull characters Eumecia is clearly more closely related to Mabuya than to Riopa (Table 1). The single greatest difference between Mabuya and Riopa is in the relationships of the bones of the palate (Fig. 1). In both ^ I have seen neither a skull nor an alcoholic specimen of jolinstoni, which is still known only from the type. I am assuming throughout this paper that the two species are so similar (conspecific?) that, unless stated other- wise, observations made on the two skulls and alcoholic specimens of anchietae are also valid for jolinstoni, and therefore for the whole genus Eumecia. 1967 GENERIC RELATIONSHIPS OF EUMECIA Figure 1. Ventral view of the secondary palate of: A) Eumecia anchie- tae (MCZ 41562), Kaimosi, Kakamega, Kenya: B) Mahuya polytropis (MCZ 8103), Krilii Cameroon; C) Riopa punctata (MCZ 3238). 70 miles SW of Amballa, India: D) Leptosiaphos hlochmanni (untagged MCZ specimen). Upper Mulinga, Idjwi, Id., Congo. Abbreviations: E, ectopterygoid; P, palatine; PT, pterygoid. A and B drawn to one scale and C and D drawn to another scale. 4 BREVIORA No. 276 Mabuya and Riopa the palatine bones meet along the ventral midline of the palate to form a secondary palate separating the air (above) and food (below) passages, a structure that is char- acteristic of the subfamily Lygosominae (Greer, MS). In Riopa the development of a secondary palate is more extensive than in Mabuya in that the palatal rami of the pterygoids approach but do not touch along the midline of the palate. The pterygoids are separated from each other by a pair of medial, posteriorly project- ing processes from the palatines which are closely applied to the inner edges of the palatal rami and which themselves touch along the midline to close the gap between the pterygoids. In Mabuya, on the other hand, and in Eumecia, the palatal rami of the pterygoids are widely separated and divergent, and there are no posteriorly projecting processes from the palatines (Fig. 1). In Eumecia there is a broad suture between the prefrontal and nasal bones which thus separates the frontal from the maxilla. In 12 of the 23 species of Mabuya examined for this feature, the pre- frontal articulates with the nasal, but in Riopa only 2 of the 12 species checked show a similar relationship of the prefrontal and nasal. Similarly. Eumecia shows a short longitudinal series of pterygoid teeth, a feature shared with 17 of the 30 species of Mabuya ex- amined for this character. However, Riopa bowringi was the only one of the 1 3 species of Riopa examined which was found to have pterygoid teeth. Another somewhat statistical difference that aligns Eumecia with Mabuya is the number of teeth on the maxilla. The two specimens of Eumecia anchietae examined had 24-26 teeth on the maxilla. Only one species of Mabuya (occidenfalis) of the 29 examined had fewer than 20 maxillary teeth; all others had 20 or more. Of the 13 species of Riopa examined for this character, however, only 4 species had 20 or more teeth on the maxilla. The one significant similarity between Riopa and Eumecia in the gross morphology of the skull is in the common absence of the postorbital bone. Eumecia anchietae and the 13 species oi Riopa examined lacked this bone while in all 29 species of Mabuya studied, a small, but discrete, postorbital bone was present. On the weight of the total evidence presented here, however, it would seem as if the postorbital of Eumecia had been lost independently of its loss in Riopa. Although it seems established, on the basis of the evidence dis- cussed above, that the relationships of Eumecia lie closer to Ma- buya than to Riopa, it is still reasonable to ask if there might not 1967 GENERIC RELATIONSHIPS OF EUMECIA 5 be a group even more closely related to Eumecia than is Mabuya. In short, this does not seem to be the case. In Africa the only lygosomine skink which has the palatal rami of the pterygoids widely divergent and which lacks medial palatine processes pro- jecting posteriorly as in Mabuya and Eumecia is Ablepharus bou- toni, a relatively recent immigrant from the Australian Region. This species, however, lacks pterygoid teeth and supranasals which make it an unlikely representative of a stock ancestral to Eumecia. In all African lygosomines other than Mabuxa, Eumecia and Ablepharus, the palatal rami of the pterygoids are more closely apposed along the ventral midline of the palate, and medial pala- tine processes project posteriorly between the pterygoids. Indeed, in some African lygosomine groups with this basically Riopa-VikQ palate (African Ablepharus, with the exception of A. boutoni; African Leiolopisma; Panapsis; Leptosiaphos). a further special- ization has been the deep posterior emargination of the palatal rami of the pterygoids. The only other groups of skinks with both supranasals and a palatal pattern generally similar to that of Mabuya and Eumecia are the genera Emoia and Eugongylus. Although these two genera are closely related^ and are very close relatives of Mabuya, it seems less likely that Eumecia has arisen from an Emoia-Eugongylus stock than from a Mabuya ancestry. Eumecia has pterygoid teeth and a broad surface suture be- tween the prefrontal and nasal bones while none of the 20 species of Emoia or 2 species of Eugongylus examined had pterygoid teeth or a broad prefrontal-nasal suture. Many Mabuya, on the other hand, have pterygoid teeth { 17 of 30 species examined) and a broad suture between the nasal and prefrontal (12 of 23 species examined). Eumecia also has only 9 teeth on the premaxillae whereas all the Emoia and Eugongylus studied have 11 or more (12) pre- maxillary teeth. However, only a few Mabuya (6 of 30 species ^ The species of Eugongylus have also been treated as relatives of Riopa in the three most recent classifications dealing with lygosomine skinks. BoLilenger (1887) referred the species of Eugongylus to the section Riopa of the genus Lygosonia. Smith (1937) recognized the taxon as a subgenus of Riopa, and Mittleman (1952) gave the taxon full generic rank. Again, the reasons for supposing that the relationships of Eugongylus lay in the direction of Riopa are as unclear to me as those for aligning Eumecia with Riopa. On the basis of skull morphology, partly discussed here, Eugongylus is strikingly similar to Emoia. This similarity and its taxonomic sig- nificance will be discussed elsewhere. 6 BREVIORA No. 276 available) have 10 or more (11-12) teeth on the premaxillae, all others have 9 (or less frequently, 8). Furthermore, Eumecia is live-bearing in its mode of reproduc- tion, as are about half the species of Mabuya (10 of 21 species for which information is available). In contrast, the 17 species of Emoia and 2 species of Eugongylus for which the mode of repro- duction is known are all egg laying (Greer, personal observation). Emoia and Eugongylus do lack the postorbital bone, as does Eumecia, and although Mabuya invariably has a small postorbital, it seems most likely, on the weight of other evidence presented above, that Eumecia has lost the bone independently of its loss in Emoia and Eugongylus. The present distribution of Mabuya and Eumecia certainly sup- ports the derivation of Eumecia from a Mabuya ancestry. Mabuya has obviously been in Africa a long time — long enough to have evolved numerous and diverse species. Part of this diversity is manifest in the evolution of Eumecia — a "Mabuya" with reduced appendages. In that it is the only "Mabuya" to have lost the post- orbital bone, and to have reduced the number of fingers and toes from the primitive number of 5-5, the taxon clearly deserves generic rank. ACKNOWLEDGEMENTS My thanks go first and foremost to Dr. Ernest E. Williams of the Museum of Comparative Zoology (MCZ) who has been most generous in his policies regarding the preparation of skulls from alcoholic specimens in the herpetological collections of the Museum. The following people have also contributed greatly to the present study by making available to me skulls of species not represented, or only poorly represented in the Museum of Comparative Zoology collections: Dr. Charles M. Bogert and Dr. Richard G. Zweifel, American Museum of Natural History (AMNH); Dr. Alan E. Leviton and Dr. Steve C. Anderson, California Academy of Sciences (CAS); Dr. Robert F. Inger and Mr. Hymen Marx, Field Museum of Natural History (FMNH); Dr. Charles F. Walker, University of Michigan Museum of Zoology (UMMZ). This study was supported in part by a grant from the Evolu- tionary Biology Fund which is administered for the National Sci- ence Foundation by the Evolutionary Biology Committee of the Biological Laboratories, Harvard University. 1967 GENERIC RELATIONSHIPS OF EUMECIA 7 MATERIALS Complete skulls of the following species have been examined. RIOPA: ajer (MCZ 41517, 41519, 71881); bowringi (1 from the series MCZ 26501, 26512); femaudi (MCZ 49696); laeviceps (MCZ 71889); lineata (AMNH 46379); mabiiiifonnis (MCZ 40267); pembana (MCZ 46106); popae (MCZ 44706); punctata (MCZ 3238); sunder alii (MCZ 41537, 41543); tanae (MCZ 40256); vinciguerrae (MCZ 17892); albopunctata (MCZ 8360). EMOIA: adspersa (AMNH 29227); atrocostata (MCZ 15074, 15080, 26476, 26419); boettgeh (MCZ 22014); callisticta (MCZ 67203, 612>0^);cyanogaster (CAS 100684, MCZ 15121, 15135, 72278, 72287); cyanura (MCZ 14582, 14584, 14586, 75954, 75956); flavigularis (MCZ 65869); kordoana (MCZ 48603); kuekenthali (FMNH 134594); loveridgei (MCZ 49321); macu- lata (MCZ 49501, 49505 lot); mivarti (MCZ 73807, 75984); nigra (MCZ 15153, 15157,67770, 72510, 72514, 72515. 72517, 72523, 75522); paUidiceps (MCZ 79856); physicae (AMNH 95772), ruficauda (MCZ 26482, 2 specimens, 26492); sanjordi (AMNH 40169); submetallica (AMNH 59015); sorex (MCZ 1105); samoensis (MCZ 16931). EUGONGYLUS: albojasciolatus (MCZ 4097, 72703); rufes- cens (MCZ 49341). EUMECIA: anchietae (MCZ 41557, 41562). MABVYA: aurata (MCZ 56550); bayoni (MCZ 39731); ben- soni (MCZ 22583); binotata (MCZ 22421); blandingi (MCZ 55171); brachypoda (MCZ 71410); brevicollis (MCZ 41306); capensis (MCZ 21433); comorensis (MCZ 24151, 2 specimens, 24155); dorsovittata (MCZ untagged specimen); elegans (MCZ 67954) ; ^'/^g/t'/ (MCZ untagged specimen) ; fasciata (MCZ 37835+ 2 untagged specimens); gravenhorsti (MCZ 11609); hildebrandtl (MCZ 70254, 70248); lacertijormis (MCZ untagged specimen); longicaudata (MCZ 25191); m«/>i/vfl (CAS 71456^ UMMZ 1047, MCZ 32040, 36617, 38935, 54201. 81182, 81184); macrorhyn- cha (MCZ 49551, 49552); maculaha (MCZ 3926); maculilabris (MCZ 24820. 24821); megalura (MCZ 47611); multifasciata (CAS 60692 + 2 untagged specimens, UMMZ S 1830, 1831, MCZ 25198, 25199, 37843); occidentalis (MCZ 43180); per- rcteti (MCZ 19711); planijrons (MCZ 85545); /^o/y^ro/^/^ (MCZ 8103, 54559); qidnquetaeniata (MCZ 52424, 2 specimens, 55179, 67838, 67840); sulcata (MCZ 21645); vw/fl (MCZ 18658, 18668, 50823, 50824, 85543). 8 BREVIORA No. 276 LITERATURE CITED BOCAGE, J.V.B. DU 1870. Description d'un "Saurien" nouveau de TAfrique occidentale. J. Sci. Lisboa, 3:66-68. BOULENGER, G. A. 1887. Catalogue of the lizards in the British Museum (Natural History). London, Vol. Ill, xxi + 575 pp. 1897. List of the reptiles and batrachians collected in northern Nyasa- land by Mr. Alex Whyte, F.Z.S., and presented to the British Museum by Sir Harry H. Johnston, K.C.B.; with descriptions of new species. Proc. Zool. Soc. London 1897: 800-803. Greer, A. E. ms A subfamilial classification of scincid lizards. LOVERIDGE, A. 1953. Zoological results of a fifth expedition to East Africa. 111. Reptiles from Nyasaland and Tete. Bull. Mus. Comp. Zool., 110(3): 143-322. 1957. Check list of the reptiles and amphibians of East Africa (Uganda; Kenya; Tanganyika; Zanzibar). Bull. Mus. Comp. Zool., 117(2): 153-362 + xxxvi. MiTTLEMAN, M. B. 1952. A generic synopsis of the lizards of the subfamily Lygosominae. Smithsonian Misc. Coll.. 117(17): 1-35. Smith, M. A. 1937. A review of the genus Lygosoma (Scincidae: Reptilia) and its allies. Rec. Indian Mus., 39(3): 213-234. (Received 14 April 1967.) 1967 GENERIC RELATIONSHIPS OF EUMECIA TABLE 1 Comparison of certain external scale (1) and skull (2-6) characters in the genera Mabuya, Eumecia and Riopa (scnsu Smith's (1937) sub- genus Riopa and Mittleman's (1952) genera Riopa, SquamiciUa and Mochhis collectively). Mabuya 1. Prefrontal scales meet medially in about Vs of the species. 2. Palatal rami of ptery- goids separated along midline of palate; no posteriorly projecting medial processes from the palatines ( Fig. 1). 3. 12 of 23 species examined with a surface suture be- tween prefrontal and nasal bones to separate frontal and maxilla. 4. 17 of 30 species examined have pterygoid teeth. 5. 28 of 29 species examined have 20 or more teeth on the maxilla. 6. Small to minute postor- bital bone present. Eumecia Prefrontals meet medially. As in Mabuya (Fig. 1). Surface suture be- tween prefrontal and nasal bones to sepa- rate frontal and maxilla. Pterygoid teeth present. 24-26 teeth on maxilla. Postorbital bone lack- ing. Riopa Prefrontals always sepa- rated. Palatal rami of pterygoids separated along midline of palate by posteriorly projecting medial pro- cesses from the palatines (Fig. 1). Only 2 of 13 species ex- amined with a surface suture between prefrontal and nasal bones to sepa- rate frontal and maxilla. Only 1 {R. bowringi) species of 13 examined has pterygoid teeth. Only 4 of 13 species ex- amined have 20 or more teeth on maxilla. Postorbital bone lacking. BREVIORA MiLHseuiitM of Comparative Zoology Cambridge, Mass. 24 November, 1967 Number 277 LESTIDIUM BIGELOWh A NEW SPECIES OF PARALEPIDID FISH WITH PHOTOPHORES Michael J. F. Graae^ ABSTRACT Two specimens of a new luminous paralepidid fish were taken in the Indian Ocean during Cruise VI of the R/V Anton Bruiin. Their main distinguishing feature is the presence of three separate and distinct, circular photophores along the ventral midline that are unlike the light organs found in other paralepidids. The two specimens are described here as a new species, and the structure of one photophore is examined in detail. INTRODUCTION Two evolutionary lineages, the myctophoids and the alepisau- roids, are generally recognized within the pelagic members of the Myctophiformes (Iniomi). The myctophoid families Myctophidae and Neoscopelidae have discrete photophores lying in rows on each side of the ventral midline, and it has been shown recently (Haneda, 1958; Rofen, 1965) that a few members of the alepi- sauroid family Paralepididae have elongated luminous glands im- mediately on, or adjacent to, the ventral midline. One genus also has a small finger-like projection, before the eye, that may be luminous, but this organ does not appear to have the structure found in the myctophoids. This paper is concerned with a unique new species of paralepidid which, though quite similar to other paralepidids in most characters, is distinguished from them by the discrete photophores on the ventral midline. 1 1702 San Andres St., Santa Barbara. California. 2 BREVIORA No. 277 I am indebted to the National Science Foundation for its or- ganization of the American Program in Biology, International Indian Ocean Expedition, and to Dr. John H. Ryther of the Woods Hole Oceanographic Institution, the Scientific Director of the program. I also gratefully acknowledge the support of NSF Grant (jF-147 to Harvard University, which aided this study. I thank Dr. Giles W. Mead, Chief Scientist during Cruise VI of the Bruun, for his counsel during the course of this study and for his review of the manuscript. I am especially indebted to William O'Day, who so kindly and patiently prepared, sectioned, and photographed the photophore described herein. I also wish to extend my appreciation to those who have helped me with advice and encouragement, especially Dr. Basil Nafpaktitis and Mrs. Myvanwy M. Dick. This work is part of a Harvard senior honors dissertation presented in April of 1966. Lestidium bigelowi, new species Holotype: MCZ 44881, 37.2 mm in standard length; collected during the International Indian Ocean Expedition, R/V Anton Bruun, Cruise VI, trawl 339A; APB label 7227; hydrographic station 339; collected in 10' Isaacs-Kidd Midwater Trawl, depth 0-61 5m; 30-V-64. 0140-0645 hrs; 4"01'-4° 14'S, 65°00'-65°02'E. Paratype: MC/ 44880, 28.4mm in standard length; R/V Anton Bruun, Cruise VI; trawl 340B; APB label 7241; hydro- graphic station 340; 10' IKMT, 0-746m; 31-V-64, 1945-0155 hrs; 5°55'-6'^08'S, 64°48'-64"^58'E. Diagnosis: This species is the only paralepidid known to have light organs that are small and discrete, as opposed to long and tubular. There are three of these; all are directed ventrally. One is at the isthmus, one is between the bases of the pectoral fins, and one is just posterior to the bases of the ventral fins. In most characters, Lestidium bigelowi is almost intermediate between the genera Lestidium and Lestidiops (see Table 1). Description: The following description is based on both types. Body long and slender, very compressed. Keel present for short distance ventrally between anus and anal fin. Anus between tips of appressed ventral fins. Skin smooth; scales absent, with lateral line ossicles insufficiently ossified to be observed. Lateral line ex- tending along side of body from edge of operculum lo a vertical with first of procurrent caudal rays, and slightly above middle of body. 1967 NEW PARALEPIDID FISH E E DO •a c CO to E E (U c 5 c -2 c OJ > ■T3 C ■ — * « -™ 00 « oo ■o ^ c75 'I- N _; U S /-, ri n d vO sD r- >/". ri >/-i — I-- "/-l r- 'T so so 00 ^ V~i vO vO 00 r<-i OS *— r~^ >o "O r^ ^ r4 ' — 1 Tt — o ri SO On OS r-- >/-i ^ — ; 00 CO -rr 00 ri d OS vb v-i so rj — ^ ir, >o t-~ v-i OS ri ri Tf d ri — so so I I 1 I o OS ^ r4 r 1 r 1 1 OS oc 1 ■ OS ri oo o so so OS m o o OS ^ •/-- >/"i r^ so c _c x: =3 •a c O c — c c c > o /-) C q oo 0\ rr, ^ — rj vO r^ iri r-- oo r- r<-i ^ rJ i/-i 'O r<-, sC ly^, r4 vO r4 (■<-, 00 o r^ >/-i "/"i r^i — o rj I oo OS c OS — OS r) OO ri v-i r-~ >o t^ 'i- O rj sO OS oo tT t- ri Ti- 1 ri 1 OS "/-. 1 "O 1 rj ^ OS O tT v"j rt v-, so >/-, ri r«-i — ' — (-<". »/"* >/-l r^, — 00 . — , 00 OS ri OS ■^1 sC o^ 00 Tl- Vt ^-^ 00 1 '-r q I »/"J "" 00 V-1 rj o r- t^ _^ — OS r.i 00 rr, "^ OS o ^ o r»-, r«-, Os so 1 OO IT, 00 Ov f. 00 00 ^ 13 2 "O 3 aj — - c;^<^;^ c o CO >^Q- c/5 Q Q Q < wj 2 c — u C c c ■a o c OJ C3 X3 — • X ■- '^ .■. j; E ^ CQ U UJ D d. Q c c C O C3 .3^ t. « ■o o c < < Q < c o T3 CQ O •a OJ c C/5 ■" c ■a CO o c^- 01 c ~ CJi Ol c (rt • ■- 3 c c o c cu cr o o 0) n c in «) - 0) a> <1> > o l-l OX) O) e E -O "Z o ^ W. ^ o -^ XJ o o c ■o o Cj ♦ D a M A * [X >> 1- XI s L> tf) o c ^ **" 3 o o in u T3 ' w en c O .2 (/) - o = = = E n n o — OJj f- o tn n LU O Q ++ ® • + o BREVIORA MiaseiuaM of Coiraipsirative Zoology Cambridge, Mass. 5 April, 1968 Number 281 ADDITIONAL NOTES ON BATOID FISHES FROM THE WESTERN ATLANTIC Henry B. Bigelow and William C. Schroeder Introduction Order Batoidei . Suborder Torpedinoidea . Family Torpedinidae . Torpedo nobiliana . Diplobatis pictus . Suborder Rajoidea Family Rajidae . Raja bullisi . Raja cervigoni . Raja clarkii . Raja floridana . Raja fuliginea . Raja lentiginosa . Raja olseni . Raja teevani . Breviraja plutonia . Breviraja sinusmexicanus Breviraja spinosa . Cruriraja riigosa . Family Pseudorajidae . Pseudoraja atlantica . Family Anacanthobatidae A nacan thobatis american us Anacanthobatis longirostris Springeria jolirostris . References cited .... Page 2 2 2 2 2 3 4 4 4 5 8 9 11 13 14 14 15 15 16 17 18 18 18 18 19 20 22 'Contribution No. 1875 from the Woods Hole Oceanographic Institution. BREVIORA No. 281 INTRODUCTION During the last fifteen years we have received for study a large number of interesting batoids, together with some cyclostomes, sharks, and chimaerids, trawled by U.S. Fish and Wildlife vessels, chiefly the "Oregon" and the "Silver Bay" but also the "Combat," "Pelican," and "George M. Bowers." Most of these were taken in the Gulf of Mexico, the Caribbean, and along the coast of South America as far south as 01°49'N off the mouth of the Amazon, with a few from along our eastern coast to the southward of Cape Hatteras. Bottom trawling depths ranged from 40 to 750 fathoms (65-1370 m), chiefly between 100-500 fathoms (183-915 m). Many new species are included in our various reports from 1951 to 1965, listed in the references cited. The present account includes eighteen species of batoids, none of them new, but they yield additional information as to range, depths of capture, size, etc. They were taken at 95 "Oregon" and 10 "Silver Bay" stations which fall within the following depth ranges: 10 at 40-100; 11 at 101-200; 60 at 201-300; 13 at 301-400; 8 at 401-500; and 3 stations at 501-600 fathoms, ranging from northern Florida to Venezuela, including the Gulf of Mexico. Also included is a summary of previously published capture records for each species. We thank the various investigators who collected and preserved these specimens, and Harvey R. Bullis, Jr., of the U. S. Fish and Wildlife Service for placing them at our disposal. Order BATOIDEI Suborder TORPEDINOIDEA Family TORPEDINIDAE Torpedo NOBiLiANA Bonaparte 1835 Our recent accounts (1962, 1965) of this species include the capture of 12 specimens, ranging in length from 211 to 700 mm, trawled in various localities ranging from North Carolina to the Caribbean coast of Panama (9°00'N, 81°23'W), including the Gulf of Mexico, in depths of 10 to 290 fathoms ( 18 to 530 m). It is known in the western Atlantic from as far north as the offing of southern Nova Scotia and in the eastern Atlantic from northern Scotland to the Mediterranean, Azores, Madeira, and tropical West Africa. 1968 ADDITIONAL BATOID FISHES 3 The present collection includes 12 specimens, 205 to 393 mm long, 3 from the east coast of Florida, 1 from the east coast of Hispaniola, and 8 from off the coast of Venezuela, in 40 to 240 fathoms (73 to 440 m), "Oregon" stations 4393, 4394, 4401, 4402, 4418, 5037, 5101, 5109 and "Silver Bay" stations 5188, 5530. An examination of a number of recently obtained specimens, necessitates a modification of two descriptive characters given in Fishes of the Western North Atlantic, Part 2 (Bigelow and Schroeder, 1953, p. 94), stated as "disc appreciably shorter (less than 80 per cent) than its breadth" and "height of caudal fin not greater than distance from its own upper origin to origin of first dorsal." Four males 215-900 mm long and seven females of 211-613 mm have a disc length ranging from 77 to 90 per cent of the disc width, only two individuals falling below 80 per cent. The height of the caudal is greater, 1.10 to 1.38 times, than the distance from the upper origin of the caudal to the origin of the first dorsal on 10 specimens and less, 0.95 times, on only one. In color, this ray has been described as dark chocolate to purplish brown above, either uniform or with a few obscure darker spots, and white below, but with the edges of the disc and pelvics of the same hue as the upper surface, the tail with irregular dark markings. Our specimens likewise are so colored above but the markings below on several are grayish. It is reputed to reach a weight of about 200 pounds. DiPLOBATis PiCTUS Palmer 1950 The type of this species was taken off Georgetown, British Guiana, a female 137 mm in total length. Our recent account (1962) includes 24 specimens, 77-164 mm long, from the mouth of the Amazon 02°29'N to British Guiana 09°3rN in 9-50 fathoms (16-92 m). The present collection includes 3 males of 85-159 mm and 3 females of 140-196 mm from 12°19'-12°37'N, 70°34'-71°10'W, off the mouth of the Gulf of Venezuela, in 40-65 fathoms (73-119 m), "Oregon" stations 4393, 4394, 4401, 4402. In our 1962 account we mentioned certain characters that supposedly differentiated D. pictus from D. guamachensis Martin 1957. Thus we stated that in 21 specimens of pictus, 11 to 164 mm long, the width of disc ranges from 38.8 to 50.0 per cent, the length 4 BREVIORA No. 281 of disc from 38.8 to 48.0 per cent, and the breadth of pelvics from 28.7 to 36.6 per cent of total length; also that the lateral folds originated between the rear end of the first dorsal base and the origin of the second dorsal. In a personal communication received from Dr. Martin in 1959, based on 4 specimens of guamachensis 56.5-70.9 mm long (including the type), from the Gulf of Cariaco, Venezuela, the above proportions are given as 53.0 to 58.1, 50.3 to 54.5, and (on two specimens) 23.8-26.0, respectively, the lateral folds originating "immediately behind the origin of first dorsal." The color and markings on pictus vary considerably and our earlier collection does not include any with the design of guama- chensis with its irregular dark bars and spots. However, in our present collection of 6 specimens, two males 85 and 196 mm long, respectively, are marked much like Martin's illustration, though somewhat more elaborately. This variation in the markings on pictus recalls that of Narcine brasiliensis. Furthermore, they fall in line with the proportional dimensions of all our other pictus and also confirm that the lateral folds originate between the dorsal fins and not immediately behind the origin of the first dorsal. We therefore consider D. guamachensis a synonym of D. pictus. The known range of this species now extends from the offing of the Gulf of Venezuela to that of the Amazon River and in depth from 9 to 65 fathoms (16-119 m). Suborder RAJOIDEA Family RAJIDAE Raja bullisi Bigelow and Schroeder 1962 Specimens previously recorded (1962, 1965), 166 to 478 mm long, range from Dry Tortugas, Florida, 24°36'N, to the coast of Surinam, 07°15'N, in 110-300 fathoms (201-549 m). Three specimens are included in the present collection: a 205 mm female from off Santa Marta, Colombia, 11°08'N, 74°23.8'W, in 100 fathoms ( 183 m), "Oregon" station 4856; a 280 mm female 11°31'N, 60°5rW, in 240-250 fathoms (440-457 m), station 5029; and a 370 mm male 11°36.5'N, 62°46.5'W, in 200-240 fathoms (366-440 m), station 5037, from the offing of the Gulf of Paria, Venezuela. Bullisi closely resembles Raja teevani but its persistent nuchal thorn, somewhat broader tip of snout, and somewhat greater space between the dorsals (0.23 to 0.45 times the first dorsal base) 1968 ADDITIONAL BATOID FISHES 5 distinguish it from the latter which lacks a nuchal thorn, has a very sharp tip of snout, and dorsals that are confluent or with an inter- space seldom exceeding 0.15 times the first dorsal base. Raja cervigoni Bigelow and Schroeder 1964 Previous records of this species include: an immature male of 357 mm, the holotype, from 10 miles (16 km) northeast of Carupano, in 20-30 fathoms (37-55 m), Museo Historia Natural La Salle, Venezuela, No. 873; a male of 206 mm, and a female of 229 mm from Punta Araya, Estado Sucre, in about 20 fathoms (37 m), off the eastern part of Venezuela; and a male of 235 mm from off the Guianas, 07°25'N, 54°35'W, in 75-80 fathoms (137-146 m), "Oregon" station 2289. The present collection includes 2 males, 144 and 506 mm long, respectively, the latter mature, and a female of 418 mm, from off the Peninsula de Guajira, Venezuela, 12°29'N, 71°54'W, in 95 fathoms (174 m), "Oregon" station 5685. Description of mature male. Proportional dimensions in per cent of total length. Disc. Extreme breadth 69.5; length 50.2. Snout length. In front of orbits 10.7; in front of mouth 13.5. Orbits. Horizontal diameter 4.4; distance between 4.2. Spiracles. Length 2.8; distance between 6.1. Mouth. Breadth 8.1. Exposed nostrils. Distance between inner ends 7.9. Gill openings. Length 1st 1.7; 3rd 1.7; 5th 1.5; distance between inner ends, 1st 13.8; 5th 7.3. First dorsal fin. Height 2.8; length of base 4.9. Second dorsal fin. Height 2.8 length of base 4.9. Pelvics. Anterior margin 12.1. Distance. From tip of snout to center of cloaca 46.5; from center of cloaca to 1st dorsal 36.5; to tip of tail 53.5; from rear end of 2nd dorsal to tip of tail 3.4. Interspace. 1st and 2nd dorsals 3.8. Disc 1.38 times as broad as long; maximum angle in front of spiracles 99°; anterior rays of pectorals extending 50 per cent of distance from level of orbits toward tip of snout, the latter pro- jecting; rostral process firm, extending to tip of snout; anterior margins of disc slightly convex in front of spiracles, thence concave and straight toward outer corners which are sharply rounded, the width of disc across anterior edge of orbits 25 per cent of total length of specimen; posterior margins and corners and inner 6 BREVIORA No. 281 margins all rounded. Axis of greatest breadth 71 per cent of distance back from tip of snout to axils of pectorals. Tail rather slender, the lateral folds low down, narrow, originating a little in advance of tips of pelvics, reaching tip of tail; length of tail from center of cloaca to origin of first dorsal 0.78 times as great, and to its tip 1.15 times as great as distance from center of cloaca to tip of snout. A row of 9 thorns along anterior and inner margins of one orbit, 10 along the other, and 3 minute thorns opposite inner margin of each spiracle with several small ones over tip of rostrum. Three prominent thorns along midline in the nuchal region, followed by a space and the next thorn, a tiny one, opposite the axils of pectorals, with 6 more to opposite tip of anterior pelvic lobe, then 22 thorns alternating large and small in an unbroken series to the first dorsal fin; 2 thorns in front of second dorsal. A row of 21 thorns low down each side of the midrow on tail, more or less alternating in size, beginning a little in advance of tips of posterior pelvic lobe and reaching to opposite origin of first dorsal. Malar thorns prominent, sharp, backward pointing, each patch consisting of 3 to 4 rows 40 mm long and 16 mm wide. Alar thorns well developed, each patch about 60 mm long and 13 mm at widest, the rows irregular, as many as 5 rows anteriorly reducing to a single thorn posteriorly. A band of fine prickles along edge of disc from opposite posterior malars to outer angle. Upper surface otherwise smooth. Lower surface with a patch of prickles on end of snout merging with a narrow band of prickles along edge of disc extending to about opposite mouth; otherwise smooth below. The claspers are massive, extending 75 mm beyond the tips of pelvics. Color. Upper surface plain medium brown. A prominent ocellus on each side of disc, situated a little posterior to the greatest axis of disc, its center from the midline of disc a distance about equal to that from tip of snout to center of orbit; distance between centers of ocelli 133 mm, and between centers of ocelli and centers of orbits 127 mm. The ocelli have a pale center surrounded by a roundish area of solid dark brown the diameter of which is about 9 to 1 1 mm, thence a pale area surrounded by a narrow, somewhat broken dark brown circle about 21 mm in diameter and finally with an outside narrow pale circle. Lower surface of disc and pelvics margined with a band of pale grayish, the tail with brownish blotches. 1968 ADDITIONAL BATOID FISHES 7 The 144 mm male has a disc 1.45 times as broad as long; maximum angle in front of spiracles 109°; contour of disc about the same as the 506 mm male, except that width across anterior edge of orbits is 34.7 per cent of total length of specimen; axis of greatest breadth 72.6 per cent; length of tail to origin of first dorsal 0.72 times as great and to its tip 1.08 times as great as distance from center of cloaca to tip of snout. Two thorns along anterior margin and 1 at inner posterior margin of each orbit; 1 nuchal thorn; a row of 16 large, sharp, thorns in midline, evenly spaced, from opposite axils of pectorals to first dorsal and 2 between dor- sals; a row low down each side of midrow, from opposite tip of pelvics to interdorsal space, 11 on one side 12 on the other, in size about the same as those in midrow; otherwise smooth above. Smooth below except for a single row of 6 to 7 very small spines along edge of disc between tip of snout and nostrils. Claspers not reaching tip of pelvics. The distance of the center of each ocellus from the midline of disc is equal to that from tip of snout to an- terior margin of orbit; distance between centers of ocelli 40 mm and between centers of ocelli and centers of orbits 35.5 mm. The ocelli have a dark brown center and an outside solid dark brown circle 6 mm in diameter, with a pale area between and another surrounding the colored areas. The lower surface is marked with a broad outer band of gray from opposite mouth to axils of pec- torals, its inner margin irregular, and with a pair of irregular grayish blotches at the origin of the anterior pelvic lobes and on the tips of the posterior lobes, the tail with two vague grayish areas pos- teriorly. The 418 mm female agrees closely with the small male in pro- portional dimensions. The ocular and nuchal thorns are of the same number as on the 506 mm male but the midline thorns are more numerous, there being 6 very small ones from opposite the ocelli to the axils of pectorals followed by 30, alternating large and small, to the first dorsal, with 2 small thorns and 1 large between the dorsals. The side row contains about 35 thorns extending from near axils of pectorals to opposite the second dorsal, more or less alternating in size, but the posterior 5 or 6 very small. The prickles along the edge of disc anteriorly, above and below, are similar to those of the large male. The distance between the centers of the ocelli is 1 17 mm and between centers of ocelli and centers of orbits 98 mm. Color above and below similar to the large male. The tooth count on the 7 specimens thus far examined ranges from 3 2 '■^ 4 0* 8 BREVIORA No. 281 Raja clarkii Bigelow and Schroeder 1958 Our previous records of R. clarkii (1958, 1962, 1905) include a total of 4 males 228-665 mm long, and 4 females of 176-747 mm. Four were trawled in the northern part of the Gulf of Mexico, 3 off the coast of Nicaragua and 1 off Panama, in depths ranging from 200 to 300 fathoms (366-549 m). The present collection includes 3 males of 330-710 mm from the north central part of the Gulf of Mexico, "Oregon" stations 4701, 4703, 4716; 6 males of 265-525 mm and 2 females of 256-370 mm from off Colombia, stations 4841, 4882, 5689, 5690, 5722; and a male of 351 mm and 2 females of 250-275 mm off Venezuela, stations 4438, 4456. These stations range in depth from 220 to 500 fathoms (400-915 m). This species is characterized among western Atlantic rajids by the presence of a band of formidable and very sharp thorns extending along the margin of the lower surface from the tip of the snout almost to the outer corners of the disc. An amplification of pre- vious descriptions follows. The median row of thorns, which is continuous from the nuchal region to the first dorsal fin, number from 30 to 43 of which 4 or 5 are in the nuchal-scapular area followed by about 7 to opposite axils of the pectorals. The total number of thorns does not nec- essarily increase with the size or age of the skate, for a male and a female, 256 and 324 mm long, respectively, have 43 thorns while a male and a female of 710 and 747 mm have only 37-38 in the median row. There is a triangular patch of thorns in the nuchal- scapular area which may have but 1 thorn on each side of the mid- row or as many as 4 or 5. In arrangement, the orbital thorns on most of our specimens, including the largest, consist of 2 anterior, 1 posterior, and from 1 to 3 aside the spiracles. The color of the upper surface of all the specimens we have seen is brown with darker punctulations scattered over the disc, pelvics, and more or less on the tail. Some have conspicuous white spots while others lack them. Thus 5 specimens, 250-747 mm long, have 3 pairs of spots, 8 of 265-580 mm have 1 pair, and 8 of 176-525 mm lack them. One specimen of 710 mm has 1 pair anteriorly and 1 spot on one side posteriorly. It is noteworthy that the spots are located in the same parts of the disc. Thus the anterior pair is opposite the spiracles, the median pair, slightly closer together, opposite the scapular region, and the posterior pair a little in ad- vance of axils of pectorals and about in line with the anterior pair. Where but one pair is present it is located anteriorly. In shape the 1968 ADDITIONAL BATOID FISHES 9 spots may be barlike, oval, or spherical, and in size, if roundish, from less than to about equal to the area of the orbit, or, if barlike, about as long or longer than the diameter of the orbit. The lower surface is white with a broad grayish band, its inner margin ir- regular (Bigelow and Schroeder, 1958, fig. 9; 1965, fig. 1) ex- tending on most specimens from a little in advance of the outer angles of disc to the axils of the pectorals and along the rear margin of the pelvics, but on several this band originates near the tip of the snout. Also, a few small gray spots or blotches may be present here and there on the disc, including one each side of the cloaca. The largest male, 710 mm in length, is approaching maturity as the claspers are well developed, extending 56 mm beyond the tips of the pelvics, and a few of the alar thorns are exposed. The present known range of this species is from the northern part of the Gulf of Mexico to the offing of Colombia (12°30'N, 72^08'W) in 200-500 fathoms (366-915 m). Raja floridana Bigelow and Schroeder 1962 The original account of this species includes 27 specimens 158- 448 mm long, and 1, the holotype of 772 mm, trawled between Cape Lookout, North Carolina, and Dry Tortugas, Florida, 34° 21' to24°17'N, in 170-225 fathoms (31 1-412 m). '' The present collection includes 32 specimens, 145-585 mm long, males and females about equally divided, taken at 12 "Oregon" stations (5097, 5098, 5100, 5101. 5102, 5106, 5107, 5113, 5119, 5234, 5295, 5482) and "Silver Bay" station 5454. These fall within a small area between the offings of Cape Kennedy and St. Augustine, Florida, 28°00'-29°58'N, 79'51'-80°10'W, in 160- 213 fathoms (293-390 m). In general appearance floridana resembles R. laevis Mitchill, 1815, but it differs from laevis in a proportionately longer snout, wider disc, usually a shorter tail, and smaller dorsal fins. Thus, on 6 males and 3 females of floridana 298-772 mm in total length, the distance from the tip of snout to the mouth is 19.5-24.9 per cent, the width of the disc 73.2-81.2 per cent, the length of the tail from center of cloaca 43.8-47.1 per cent, and the distance from origin of first dorsal fin to the rear end of base of second dorsal 8.4-12.1 per cent of total length. These proportions on 2 males and 1 fe- male of laevis, 506-542 mm long, are 17.0-18.0 per cent, 68.0-71.0 per cent, 49.7-51.7 per cent and 14.5-14.8 per cent, respectively. On 7 floridana 158-240 mm long the distance from snout to mouth is 18.0-21.1, length of tail 46.2-50.5 and there are 8 to 10 tail 10 BREVIORA No. 281 thorns, whereas on 5 laevis of 177-195 mm these proportions are 14.1-17.8 and 54.2-58.7 and the tail thorns 13 to 18. We are indebted to George C. Miller of the U. S. Fish and Wild- life Service for pertinent data on a collection of 17 specimens, 155- 252 mm long, and 1 of 465 mm, all taken between the offing of South Carolina and the Straits of Florida in 180-220 fathoms (329- 402 m), except one from 29°04'N, 88°31'W, "Oregon" station 3724, the latter extending the range to the northern part of the Gulf of Mexico. Four of these specimens have tail lengths of 50.9- 53.8 per cent of total length, a little greater than our 50.5. The tail thorns number 8 to 12 except on the one of 465 mm which has 15. Three have 2 anterior orbital thorns and 1 posterior, the others 1 anterior and 1 posterior. The armature of the upper surface, at least on sizes up to 585 mm, is limited to the orbital thorns, a single midrow of rather in- conspicuous thorns from somewhat in advance of the axils of pectorals to the first dorsal fin, and to 1 or 2 or no thorns between the dorsals. But there is not necessarily a progressive increase in the number of thorns with the size of the skate. Up to a length of about 400 mm most of our specimens have 1 anterior and 1 pos- terior orbital thorn, but a few, including the smallest, have 2 anterior and 1 posterior. One of 585 mm has 10 or 11, and in addition 2 aside each spiracle, while another of 488 mm has only 5 or 6 thorns along the margin of each orbit with 1 aside each spiracle. The thorns in the midrow usually number from 8 to 14 on specimens up to a length of about 400 mm, but two of 432-435 mm have only 3 (there is evidence that some thorns were shed and others still embedded), while one of 585 mm has 23, the anterior 13 being smaller than the posterior 10. The lower surface is smooth. The immature 772 mm holotype has 12 small thorns along the anterior and inner margins of each orbit with 3 very small ones in advance of these while the midrow contains 25 small thorns of assorted sizes, with a few rather large ones. The lower surface is densely prickly in advance of axis through the mouth, except for a nearly smooth area immediately in front of mouth. A band of prickles extends along the margin of the disc from the snout to about opposite the 4th or 5th gill openings, the remaining lower surface, including the tail, being virtually smooth. The maximum angle of the disc in front of spiracles ranges from 81° to 94°, except 74° on the holotype. The teeth range in number 1968 ADDITIONAL BATOID FISHES 11 from 27 to 38 in each jaw, and have a low triangular cusp, even on the smallest specimens of both sexes. In color, the smallest specimens are light brown above, with vague dark brown irregular spots about half as large to as large as the orbit, and the tail has 5 dark bars about equally spaced. The lower surface is whitish with a narrow dusky margin on the outer angles of disc. Our larger specimens lack the spots, and below are usually plain grayish white, some of them partly brownish. All, small to large, have dark-ringed mucous pores on the under surface, most numerous anterior to the axis through the mouth and immediately below the lower jaw. It is of interest that all of the many specimens so far collected have been taken within the rather limited depth range of 160-225 fathoms (293-412 m), although many trawl hauls, both shoaler and deeper, have been made within the known latitudinal range of this species. Raja fuliginea Bigelow and Schroeder 1954 UntU recently this species had been known only from the holo- type, a juvenile male 306 mm long, USNM No. 163367, taken at "Oregon" station 534, in the northwestern part of the Gulf of Mexico, 27°32'N, 93°02'W, in 400-450 fathoms (732-823 m). A second specimen, a female of 330 mm, was trawled at "Oregon" station 4147, about 45 miles southwest of Dry Tortugas, Florida, 24°12'N, 83°32'W, in 500 fathoms (915 m) (Bigelow and Schroeder, 1965). We now have 14 more specimens, as follows: 9 from the western and northern parts of the Gulf of Mexico, "Oregon" stations 4701, 4711, 4713, 4802; 2 off Isla del Rosario, Colombia, station 4883; 1 from about 40 miles northwest of Caracas, Venezuela, station 4449, and 2 off Curagao, station 4416, males and females evenly divided, total lengths 224-445 mm. The range in latitude is from 27°42' to 10°15.5'N, in longitude from 96°00' to 67°38'W. All were taken in 500-600 fathoms (915-1098 m). In our original account (1954), fuliginea was compared with R. bathyphila Holt and Byrne 1908 which it most closely resembles in general arrangement of thorns and prickles, in proportional dimensions and in the very dark coloration of the lower surface of disc and tail. But fuliginea differs from bathyphila in a more obtuse anterior contour and especially in the fact that the entire lower surface of the tail, apart from a very narrow median stripe, is densely prickled from base to tip, whereas it is naked in bathyphila. 12 BREVIORA No. 281 Also, the prickles on the upper surface of disc and tail are coarser than on bathyphila and there are no naked areas. Proportional dimensions in per cent of total length of all 14 specimens: Disc. Extreme breadth 43.7-51.8; breadth across anterior margin of orbits 24.6-28.6; length 37.5-49.0. Snout length. In front of orbits 9.1-10.3; in front of mouth 10.0-13.2. Orbits. Horizontal diameter 3.4-4.6; distance between 2.7-3.6. Spiracles. Length 2.1-2.9; distance between 5.7-7.3. Mouth. Breadth 5.3-7.8. Exposed nostrils. Distance between inner ends 6.2-7.3. Gill openings. Length, 1st 1.4-1.8; 3rd 1.4-1.8; 5th 1.1-1.5; distance between inner ends, 1st 10.7-14.8; 5th 5.8-8.5. First dorsal fin. Height 2.5-3.6; length of base 4.6-6. 1 . Second dorsal fin. Height 2.5-3.5; length of base 4.5-6.3. Pelvics. Anterior margin 10.1-13.4; origin to tip 13.4-17.5. Distance. From tip of snout to center of cloaca 35.2-44.0; from center of cloaca to 1st dorsal 42.4-48.8; to tip of tail 56.0-64.8; from rear end of 2nd dorsal to tip of tail 2.7-4.5. Interspace: 1st and 2nd dorsals 0.0. In armature, the upper surface of the disc, tail, and posterior lobe of pelvics, is densely covered with small sharp prickles. The dorsals, caudal fin, and the skin over eyes also are prickly. The orbital thorns, in various combinations, range from 2 at the inner margin of each orbit to 5 or 6 at the anterior and inner margins. There may be, in addition, a thorn aside one or both spiracles, or this thorn may be lacking at both. All these thorns are usually rather prominent but are very small on several specimens. The thorns on the nuchal-scapular region are in most cases arranged in a triangular patch, with 2 to 4 in the midrow and 1 or 2 on each side of these, the exceptions being 3 on each side, or none, only 1 thorn in the midrow and, on a male of 445 mm, all the nuchal- scapular thorns are lacking. The thorns in the midrow, including the nuchal region, number from 15 to 29, the fewest (15 and 19) being on the larger specimens, a male of 445 mm and a female of 412 mm in total length, and the most (28 and 29) on a male of 224 mm and a female of 293 mm. They may continue from the scapular region with no interruption, or after a short space, or resume as far rearward as opposite the axils of the pelvics, and extend to the dorsal fin or fall short of the dorsal a distance as much as that between the spiracles. In addition to the midrow there is 1968 ADDITIONAL BATOID FISHES 13 a siderow of thorns, usually irregular, beginning opposite the axils of the pectorals or as far rearward as about the tips of the pelvics and ending, when complete or nearly so, about opposite the origin of the first dorsal. But on some specimens the row runs out well in advance of the dorsal, and on one, a male of 428 mm, the row is virtually nonexistent, there being only 2 or 3 thorns on each side of the midrow. The thorns in the side row may be of the same size as those in the midrow, or somewhat larger or smaller. The lower surface of disc and pelvics is smooth but the tail is densely set with prickles or thornlets, similar to those of the upper surface, but on some specimens these are lacking or sparse along a very narrow median stripe. The color of the disc and pelvics is plain grayish above, ranging from light to rather dark. The tail may be of the same hue or some- what paler. Below, the disc and pelvics are dark gray (3 speci- mens), dark brown (7 specimens), or blackish brown (4 speci- mens) some with pale areas around the entrance to cloaca and/or on end of snout, around the jaws, and at tips of pelvics. The tail is of the same shade, or slightly darker than it is above. Raja lentiginosa Bigelow and Schroeder 195 1 Many specimens of this skate, 80 to 435 mm in total length, have been trawled by the "Oregon" in the Gulf of Mexico, where it is widespread, southward to the offing of Nicaragua (11°27'N, 83°11'W) within depths of 29-305 fathoms (53-558 m). The present collection includes 6 more of 226-418 mm, from the north- ern Gulf taken in 100-240 fathoms (183-439 m), "Oregon" stations 4614, 4616, 4696, 4705. There is a band of thorns along the median zone of back and tail, in 3 rows anteriorly, thence merging to 5 rows, and again in 3 rows as the tail narrows. These thorns have been described as originating in the nuchal region but we have since seen many speci- mens on which the band begins posterior to the nuchal region as much as the distance from tip of snout to rear margin of orbits. In some instances the median row originates slightly in advance of the side rows, but usually it begins posterior to them. The upper surface everywhere is densely freckled with light to dark brownish and whitish spots, including the tail, pelvics and claspers, distinguishing lentiginosa from R. gannani (Whitley, 1939) which it closely resembles, the latter having fewer spots which are grouped mostly in a distinct rosette pattern. 14 BREVIORA No. 281 The lower surface is white, variously marked with grayish which may be in the form of a group of blotches, mostly fused, on each pectoral, or a broad band with an irregular inner margin extending from opposite the first pair of gill openings to as far as the inner margin of the pectorals. Some have an elongate blotch on the claspers and a blotch on the anterior and/or posterior pelvic lobes. Tail white. In the original account of lentiginosa the second dorsal and the caudal fin are shown as confluent, but we have found a few speci- mens with a short space between these fins. Raja olseni Bigelow and Schroeder 1951 Previous accounts (1951b, 1953, 1962) of this species include 15 specimens, 151-568 mm long, trawled in 53-130 fathoms (97- 238 m) at 9 stations by the "Oregon" and at 1 by the "Silver Bay." All were taken in the northern and northwestern part of the Gulf of Mexico. The present collection includes two males of 161-196 mm, and a female of 335 mm, from "Oregon" station 4695, 26°16'N, 90°13'W, in 50-53 fathoms (92-97 m), in the same region as above. We include this record to emphasize how restricted is the present known range of this species in spite of the many bottom trawl hauls made by the "Oregon" and other vessels in other parts of the Gulf and along the outer coast between North Carolina and the offing of the Amazon. Raja teevani Bigelow and Schroeder 1951 Accounts of this species (1951b; 1965) include 6 males and 11 females, 175-635 mm long, of which 4 were trawled off Pensacola, Florida (3 "Oregon" stations), 1 at the northwest edge of Great Bahama Bank, and 1 in Santaren Channel ("Silver Bay" stations), and 1 1 off the coasts of Honduras and Nicaragua, the latter between 16°43' and 12°25'N (9 "Oregon" stations), in depths of 240 to 400 fathoms (439-732 m). We now have 5 more, all males, 306- 840 mm long, of which 4 were taken in the northwestern Gulf of Mexico ("Oregon" stations 4703, 4709, 4729, 4800) in 300-400 fathoms (549-732 m) and 1 of 500 mm from off Riohacha, Colombia, at ir50'N, 73°05'W (station 4911), in 175-190 fathoms (320-348 m). Teevani closely resembles R. floridana from which it differs in having fewer orbital thorns when about half grown and larger. Thus, on sizes up to about 600 mm there are, with few exceptions, 1968 ADDITIONAL BATOID FISHES 15 1 or 2 anterior thorns and 1 posterior, and on the largest, of 840 mm, only 5 thorns along the margin of one orbit and 3 at the other. Whereas young floridana up to about 300 mm have 1 anterior and 1 posterior thorn; at 450 mm there may be as many as 8 or 9 at each orbit and on one of 772 mm, the largest seen, there are 12 along the anterior and inner margins and 3 more in advance of these. Minute prickles are present on the upper surface of the disc on teevcmi of all sizes, on some anteriorly only, but on others scattered all over the disc. They are lacking on our floridana, except for the large one of 772 mm on which the end of the snout is densely covered with coarse prickles and small thorns blending into minute prickles to about opposite the spiracles, the rest of disc smooth. On the lower surface of teevani prickles are present anterior to the mouth and in a band along the edge of disc from near the end of snout to about opposite the first pair of gill openings, on small in- dividuals as well as large, but floridana is perfectly smooth below, up to a length of at least 450 mm, though prickles are present anteriorly on the 772 mm specimen, Teevani and floridana v/ere in no case taken at the same station. The known range of the latter is from the offing of Cape Lookout, North Carolina to Dry Tortugas, Florida, in 160-225 fathoms (293-412 m). Breviraja PLUTONiA (Carman) 1881 Many specimens of this species have been recorded (Bigelow and Schroeder, 1953, 1962) from the offing of Cape Lookout, North Carolina, and southward to an area centering about 30 miles southwest of Dry Tortugas, Florida, in depths of 160-400 fathoms (293-732 m), in size from 62 to 270 mm. The present collection includes 100 specimens 76-257 mm long, trawled between St. Augustine and Vero Beach, Florida, in 188-215 fathoms (344- 393 m), "Oregon" stations 5075, 5077, 5089, 5097, 5115, 5231, 5233, 5483, 5484, and 7 specimens west to southwest of Dry Tortugas in 190-310 fathoms (348-567 m), stations 4335, 4356, 4558. One of the latter, 275 mm, is the largest we have seen. Breviraja sinusmexicanus Bigelow and Schroeder 1950 All the previous captures of this skate have been recorded (1950, 1953, 1962, 1965) from the northern part of the Gulf of Mexico, except for one station in the Gulf of Campeche. There are 26 16 BREVIORA No. 281 specimens in the present collection, also from this region and from the western part of the Gulf off the Mexican coast. These were trawled at "Oregon" stations 3681, 4580, 4581, 4583, 4606, 4614, 4616, 4697, 4703, 4709, 4729, 4730, 4776, 4800, 4808, 4809, 4816. The range in length, including specimens taken on all cruises, is from 78 to 360 mm, and in depth of capture from 170 to 500 fathoms (311 to 915m), except for one station at 100 fathoms (183 m). Breviraja spinosa Bigelow and Schroeder 1950 Previous accounts of this species include several specimens taken by the "Albatross" during the years 1885-1886 off Delaware Bay and North Carolina, stations 2624 and 2730, in 258 and 727 fathoms (472-1330 m), respectively. These were in very poor condition, their identity as spinosa was doubtful, and they cannot now be found for re-examination. Published reports (1950, 1953, 1962) record a large series of spinosa ranging from the offings of Charleston, South Carolina (32°58'N), to the Guianas (07°05'N), in length 77-424 mm, trawled in 150-400 fathoms (274-732 m). Eighteen more have been taken at 7 "Oregon" stations (4428, 4854, 4912, 5039, 5689, 5690, 5722), from off Colombia, from south of Curagao, and southwest of Grenada, 115-528 mm long, from depths of 257-400 fathoms (470-732 m). Following is a description of a mature male of 528 mm, much larger than any male previously seen (332 mm). Proportional dimensions in per cent of total length. Disc. Extreme breadth 58.5; length 46.2. Snout length. In front of orbits 8.7; in front of mouth 10.4. Orbits. Horizontal diameter 4.6; distance between 3.5. Spiracles. Length 4. 1 ; distance between 6. 1 . Mouth. Breadth 7.4. Exposed nostrils. Distance between inner ends 7.0. Gill openings. Length 1st 1.3; 3rd 1.3; 5th 1.1; distance between inner ends 1st 15.5; 5th 7.6. First dorsal fin. Height 2.6; length of base 4.2. Second dorsal fin. Height 2.8; length of base 4.6. Pelvics. Anterior margin 13.6. Distance. From tip of snout to center of cloaca 42.7; from center of cloaca to 1st dorsal 44.7; to tip of tail 57.3; from rear end of 2nd dorsal base to tip of tail 3.8. Interspace 1st and 2nd dorsals 0.0. Disc obtusely rounded in front, thence concave toward outer corners which are broadly rounded; posterior margin rounded; 1968 ADDITIONAL BATOID FISHES 17 maximum angle in front of spiracles 135°; width across anterior edge of orbits 28.5 per cent of total length of specimen. Axis of greatest breadth 75 per cent of distance rearward from tip of snout to axils of pectorals. Thorns in the nuchal-scapular region in a triangular patch, con- tinuing in a band of 5 rows along back, without interruption, and on to tail, where they are somewhat larger and in 5-6 rows an- teriorly, thence narrowing to 2-3 rows, ending at first dorsal; anterior and inner margins of orbits with a row of 7 or 8 thorns, also a pair opposite each spiracle; a patch of prominent and sharp malar thorns opposite the spiracles and orbits, the length of patch about 70 mm and greatest width 20 mm; alars in as many as 4 rows, the longest row 70 mm; disc and tail covered with fine close- set prickles, coarsened in advance of the short triangular rostral process; prickles present on skin over eyes and on dorsal and caudal fins, also a few on posterior lobes of pelvics. Lower surface of disc and tail smooth. The claspers extend from the axil of pelvics a distance a little greater than from tip of snout to rear margin of orbits. Jaws strongly arched centrally. Teeth ^ , uppers in center of jaw crowded, with slender to narrowly triangular cusps, those to- ward corners of jaw arranged in parallel rows, not in quincunx; lowers similar, but those in median sector larger than correspond- ing upper teeth. Color above plain medium brown, dorsals and caudal fins blackish; below, black except pale around jaws, on nasal flap, edges of gill openings, in areas near end of snout, and at tip of anterior pelvic lobes. Cruriraja rugosa Bigelow and Schroeder 1958 This skate appears to be relatively common within its known geographic and depth ranges. We have previously recorded (1958, 1962, 1965) 103 specimens from numerous localities extending from the northern part of the Gulf of Mexico to the northern coast of Panama, the latter at 9^03'N, 81°18'VV. These ranged in length from 90 to 485 mm and were trawled in depths of 200-500 fathoms (366-915 m). The present collection includes 19 rugosa of 100-404 mm, from 250-550 fathoms (457-1007 m), taken at "Oregon" stations 4414, 4662, 4663, 4730, 4731, 4816, 4883, 4902, 5039, and "Silver Bay" station 5142. Most are from within the geographic range 18 BREVIORA No. 281 given above, but included are four additional localities: (1) the north coast of Haiti, (2) the coast of Colombia, (3) southwest of Curagao, and (4) southwest of Grenada at 1 1°40'N, 62°33'W. Family PSEUDORAJIDAE PsEUDORAJA ATLANTICA Bigclow and Schrocdcr 1962 Many specimens of this species have been recorded (1962, 1965) from off the Atlantic coast of Nicaragua, 13°31'N, to the offing of the Amazon River, 01°45'N, in depths of 135-350 fathoms (247-640 m), in length 86-481 mm. The present col- lection includes 32 males and 22 females trawled at 25 "Oregon" stations from off the coast of Panama 9°47'N, 79°25'W to the vicinity of Tobago, ll°3rN, 60°5rW, in depths of 200-340 fathoms (366-622 m), in length 90-518 mm. Mature males have a patch of malar thorns along the outer part of the pectorals from end of snout to a little in advance of a line through the orbits. These prominent thorns may have sharp tips or be covered by skin. A few may be present on individuals as small as 350 mm but in sizes upward of about 420 mm, they are well developed. The alar thorns usually start to appear at about 400 mm. On a specimen of 425 mm they are in 2 to 3 rows, on one of 447 mm up to 6 rows, but on the largest male of 491 mm there are only 3 to 4 rows of thorns. The size of the patch of thorns ranges from about 40 to 48 mm in length and 9 to 12 mm in width, usually with but 1 or 2 thorns at the anterior and posterior ends. The claspers on a 230 mm specimen are minute and do not reach the tip of the posterior pelvic lobe; at 340-400 mm they reach a little beyond the pelvic lobe and at maturity, when they have become rigid, they extend for a considerable distance. Family ANACANTHOBATIDAE Anacanthobatis americanus Bigelow and Schroeder 1962 Published accounts (1962, 1965) of this species include 51 specimens taken at 12 "Oregon" stations from off the coasts of British Guiana, Venezuela, Panama, and Honduras, 7° 34' to 16°35'N, in 100-400 fathoms (183-732 m), ranging in length from 95 to 350 mm. The present collection includes 42 specimens taken at 15 "Oregon" stations (4412, 4413, 4415, 4416, 4424, 4841, 4842, 4854, 4855, 4902, 4912, 4925, 5039, 5689, 5690) all from off the coasts of Colombia and Venezuela between 09°02'-12°30'N 1968 ADDITIONAL BATOID FISHES 19 and 76°31'-69°18'W, and "Silver Bay" station 5142 off the north coast of Haiti at 19°52'N, 71°58'W, in depths of 205-500 fathoms (375-915 m), in length 114-362 mm. The size at which males reach maturity varies. Thus, on one of 270 mm the claspers are minute, and while one of 295 mm is mature, with rigid claspers, hooks exposed, pelvics and pectorals separated, another of 320 mm is immature, with flexible claspers and the pelvics and pectorals still united. The alars at maturity are in up to 5 rows, the patch of thorns on the largest male, of 350 mm, being 29 mm long and 10 mm wide. Anacanthobatis longirostris Bigelow and Schroeder 1962 Three specimens of this bizarre skate have been recorded (1962, 1965) : a female of 507 mm, the holotype, from off the Mississippi Delta, 29°09'N, 87°53'W, in 500-575 fathoms (915-1052 m), "Oregon" station 2823; a male of 135 mm from Santaren Channel. 23°59'N, 79°43'W, in 350 fathoms (640 m), "Combat" station 450; and an immature male of 483 mm from the same locality, 23°40'N, 79°13'W, in 290 fathoms (530 m), "Silver Bay" station 2458. Two more have been trawled, both females, of 745 and 630 mm respectively, off the north coast of Haiti at 19°55'N, 72°00'W, in 470-500 fathoms (860-915 m), "Silver Bay" station 5146. Pro- portional dimensions in per cent of total length (exclusive of fila- ment at tip of snout) of both specimens follow; Disc. Extreme breadth 61.2, 56.3; length 74.3, 64.7. Snout length. In front of orbits 33.0, 30.1; in front of mouth 35.5,33.1. Orbits. Horizontal diameter 2.8, 2.7; distance between 3.8, 2.9. Spiracles. Length 1.6, 1.0; distance between 6.1, 5.7. Mouth. Breadth 5.3, 5.1. Exposed nostrils. Distance between inner ends 5.5, 4.9. Gill openings. Length, 1st 1.0, 1.0; 3rd 1.0, 1.0; 5th 0.7, 0.8; distance between inner ends, 1st 1 1.6, 10.8; 5th 8.2, 7.3. Pelvics. Length of anterior limb 15.3, 12.0; distance, origin of anterior limb to tip of posterior lobe 13.3, 13.2. Distance. From tip of snout (from base of filament) to center of cloaca 61.7, 57.1; from center of cloaca to tip of tail 38.3, 42.9. The axis of greatest breadth of the disc is about two-thirds the distance back from tip of snout to axils of pectorals. The upper caudal fin is vestigial on the 745 mm specimen and the lower caudal appears to be missing, while on the one of 630 mm the tip of tail is 20 BREVIORA No. 281 lacking, total length having been estimated on the basis of the disc width. The proportional dimensions of the latter specimen agree closely with the holotype but the one of 745 mm has a longer snout, 33.0 per cent of total length of specimen; vertical length of disc, 74.3 per cent; wider disc, 61.2 per cent; and a greater distance from tip of snout to center of cloaca, 61.7 per cent, these proportions on the holotype being 29.8, 64.7, 56.8 and 57.4 per cent, respectively. Springeria folirostris Bigelow and Schroeder 1951 Many specimens of this species have been recorded (Springer and Bullis, 1956; Bigelow and Schroeder, 1951a, 1965), all of them from the northern part of the Gulf of Mexico, ranging in latitude from 26°46' to 29^30'N, in longitude from 85°09' to 96°20'W, from depths of 164 to 280 fathoms (300-512 m), in length from 125 to 620 mm. In our last account of S. folirostris a description was given of a male, of 576 mm, the first mature specimen to come to our at- tention. This showed that on males the posterior lobe of the pelvics ceases to be adnate to the pectorals when that stage of growth is attained, as occurs with Anaccmthobatis americanus. The present collection includes three females 140-330 mm long and a male of 195 mm, from 27°56'N, 90°36'W, in 220 fathoms (402 m), "Oregon" station 4704, also a nearly mature male of 530 mm from 27°45'N, 93°56'W, in 200 fathoms (366 m), station 4606. Proportional dimensions in per cent of total length of the 530 mm male. Disc. Extreme breadth 52.0; length 54.7. Snout length. In front of orbits 21.0; in front of mouth 22.5. Orbits. Horizontal diameter 3.4; distance between 2.2. Spiracles. Length 1.0; distance between 5.5. Mouth. Breadth 4.7. Exposed nostrils. Distance between inner ends 4.5. Gill openings. Length 1st 0.76; 3rd 0.76; 5th 0.66; distance between inner ends, 1st 9.5; 5th 4.9. Caudal fin. Length of base, upper 4.7; lower 4.2. Pelvics. Length of anterior limb 13.0; origin of anterior limb to tip of posterior lobe 1 1 .0. Distance. From tip of snout (from base of tentacle) to center of cloaca 47.5; from center of cloaca to tip of tail 52.5. These proportions agree closely with those of the holotype. a male 400 mm long. 1968 ADDITIONAL BATOID FISHES 21 The pectorals and pelvics are free, as on a fully mature male. The claspers, which are semi-flexible and with the hooks not yet ex- posed, extend beyond the tips of the pelvics a distance equal to the space between the outer margins of the spiracles. The alar thorns have broken through the skin but are not fully developed, the greatest width and length of the patch, on each side, being 12 mm and 25 mm, respectively, with the thorns arranged in as many as 4 rows. The jaws are rather strongly arched. Teeth f^ , those in center of jaw the smallest and more crowded, with a triangular blunted cusp, those toward corners of jaw with a flat crown. Grayish brown above, whitish below, except for some brownish blotches on the posterior segment of the pelvic limb and a broad band of pale gray along the outer margin of the disc. ADDENDUM Illustrations are included in the following publications by Bigelow and Schroeder, all listed in the references cited. Torpedo nobiliana 1953, figs. 22, 23 Diplobatis pictus 1962, pi. 1 Raja bullisi 1962, fig. 1 Raja cervigoni 1964, fig. 1 Raja clarkii 1958, fig. 8; 1965, fig. 1 Raja fioridana 1962, figs. 2, 3 Raja fuliginea 1954, fig. 4 Raja lentiginosa 1951b, fig. 1; 1953, fig. 48a Raja olseni 1951b, fig. 2; 1953, fig. 54a Raja teevani 1951b, fig. 3; 1962, fig. 3 Breviraja pliitonia 1953, figs. 67, 68 Breviraja simismexicanus .... 1950, pi. 5; 1953, fig. 69 Breviraja spinosa 1950, pi. 6; 1953, fig. 71 Cruriraja rugosa 1958, figs. 10, 1 1; 1962, fig. 10 Pseudoraja atlantica 1962, figs. 11, 12, 13; 1965, fig. 6 Anacanthobatis americamis . . 1962, figs. 14, 15, 16; 1965, fig. 7 Anacanthobatis longirostris . . 1962, figs. 17, 18; 1965, fig. 8 Springeria jolirostris 1951a, fig. 1; 1953, fig. 78a; 1965, fig. 9 22 BREVIORA No. 281 REFERENCES CITED BiGELOw, H. B. and W. C. Schroeder 1950. New and little known cartilaginous fishes from the Atlantic. Bull. Mus. Comp. Zool., vol. 103, no. 7, pp. 385-408, 7 pis. 1951a. A new genus and species of anacanthobatid skate from the Gulf of Mexico. Jour. Washington Acad. Sci., vol. 41, no. 3, pp. 110-113, 1 fig. 1951b. Three new skates and a new chimaerid fish from the Gulf of Mexico. Jour. Washington Acad. Sci., vol. 41, no. 12, pp. 383-392, 4 figs. 1953. Fishes of the western North Atlantic. Mem. Sears Found. Mar. Res., No. 1, part 2, x + 588 pp., 127 figs. 1954. A new family, a new genus, and two new species of batoid fishes from the Gulf of Mexico. Breviora, Mus. Comp. Zool., No. 24, 16 pp., 4 figs. 1958. Four new rajids from the Gulf of Mexico. Bull. Mus. Comp. Zool., vol. 119, no. 2, pp. 201-233, 11 figs. 1962. New and little known batoid fishes from the western Atlantic. Bull. Mus. Comp. Zool., vol. 128, no. 4, pp. 159-244, 23 figs., Ipl. 1964. A new skate. Raja cervigoni, from Venezuela and the Guianas. Breviora, Mus. Comp. Zool., No. 209, 4 pp., 1 pi. 1965. A further account of batoid fishes from the western Atlantic. Bull. Mus. Comp. Zool., vol. 132, no. 5, pp. 443-477, 9 figs., 2 pis. Bonaparte, C. L. 1832- Iconografia della fauna Itahca — , vol. 3, Pesci, 78 pis. (not 1841. numbered) and accompanying text (pages not numbered). [For dates of publication see Salvadori, Boll. R. univ. Mus. Zool. Anat. Comp., Turin, vol. 3, no. 48, 1888.] Garman, S. 1881. Reports on the results of dredging under the supervision of Alexander Agassiz, along the Atlantic Coast of the United States. Report on the selachians. Bull. Mus. Comp. Zool., vol. 8, no. 11, pp. 231-238. Holt, E. W. L. and L. W. Byrne 1908. Second report on the fishes of the Irish Atlantic Slope. Fish- eries, Ireland, Sci. Invest. (1906), No. 5, 63 pp., 5 pis. Martin, S., Felipe 1957. Una nueva especie de Torpedinidae del Golfo de Cariaco, Edo. Sucre, Venezuela. Nov. Cientif., Contr. Ocas. Mus. Hist. Nat. La Salle, Ser. Zool., No. 21, 4 pp., 3 figs. 1968 ADDITIONAL BATOID FISHES 23 MiTCHILL, S. L. 1815. The fishes of New York, described and arranged. Trans. Lit. Philos. Soc. New York, vol. 1, pp. 355-492, pis. 1-6. Palmer, G. 1950. A new species of electric ray of the genus Diplobatis from British Guiana. Ann. Mag. Nat. Hist., ser. 12, vol. 3, pp. 480- 484, 3 figs. Springer, S. and H. R. Bullis, Jr. 1956. Collections by the "Oregon"' in the Gulf of Mexico. U. S. Fish and Wildlife Service. Spec. Sci. Rept.-Fisheries No. 196, 134 pp. Whitley, G. P. 1939. Taxonomic notes on sharks and rays. Australian Zoologist, vol. 9, pp. 227-262, 18 figs. BREVIORA Muiseiiim of Comparative Zoology Cambridge, Mass. 5 April, 1968 Number 282 THE EXTINCT BABOON, PARAPAPIO JONESh IN THE EARLY PLEISTOCENE OF NORTHWESTERN KENYA Bryan Patterson Among the fossils collected during 1966 in the early Pleistocene sediments of the Kanapoi area, southeastern Turkana ( Patterson, 1966), is a specimen of a small baboon. The surface find of a fragment of a right horizontal ramus with M-.;; prompted excava- tion at the spot and sifting of the slope debris. No parts were found in situ but a number of fragments were recovered from the slope and several of these fitted together to form a partial mandible. Good contacts are present from the symphysis back to the base of the ascending ramus on the right side. The bases of the incisors and canines and the anterior root of the right P-, are preserved in the symphysis; complete or nearly complete teeth present are the left P^, Ml and M;., and the right M^.:,. The specimen bears the field number 122-66K, and was found by Mr. Roger C. Wood in the drainage of the Kikimon, a dry wash east of the Kanapoi and. like it, draining into the Kakuryo. All fragments were found within a small area, which, together with the gentle nature of the slope, suggests that the fossil may not have moved far from its burial place. As is shown by the sizes of the incisors and canines relative to the cheek teeth and by the marked overlap of the anterior root of P.-. along the posterolateral face of the canine, the specimen is cer- tainly a male. It agrees so closely with specimens of Panipapio jonesi Broom from South Africa that, if Freedman's (1957, 1960) specific distinctions are accepted, there can be no doubt as to the identification. Frecdman has discussed the structure of this and other species of Panipapio in considerable detail, which makes it unnecessary to do more than comment on a few points. Curiously enough, the new specimen has the most complete symphysis of any male individual BREVIORA No. 282 of P. jonesi yet collected. The anterior surface is not steep, form- ing an angle of approximately 45' with the lower border of the horizontal ramus. The two ridges that converge upward toward each of the median incisors are not very prominent; they enclose a shallow, median depressed area that extends dorsally from the foramen symphyseosum. Between the posterior portion of the ridge, below, and the anterior alveolus of P;., above, is a small, Fig. 1. Parapapio jonesi Broom. Field no. 122-66K. Dorsal view of incomplete mandible of male. X 1- rather rugose depression. Apart from this, the anterior surface is only slightly roughened. The incisal shelf slopes very gently down- ward from the incisor alveoli to a point opposite about the middle of P;j. The symphysis terminates at the level of the anterior end of P4; there is no mental spine. The ascending ramus arises a little 1968 EXTINCT BABOON FROM KENYA 3 behind M.;. The lateral face of the horizontal ramus is very slightly concave beneath M^. and the anterior half of M;.,. but neither here nor in the portion of the bone beneath the left P4 — Mi is there any delimitation of a mandibular fossa as such. The teeth pre- served call for no special comment. Measurements (in mm) Length (mesio-distal) Width Width, anterior Width, posterior Width, hypoconulid Depth of ramus posterior to Ms c P. Mx M. Mc 6.3 6.4 8.2 10.4 12.7 10.0 6.2+ 7.3 8.8+ 9.6 7.1 9.1 8.6 5.0 29.1 All measurements have been taken in accordance with Freed- man's methods. Dimensions of the Kanapoi specimen fall. or. in the case of the widths of P4 and M^. presumably fell, within the observed ranges for males of his sample of P. jonesi (1957. tables 6b, 6d) in all cases but three. These, the anterior and posterior widths of M] and the length of M.-.,. fall within the observed ranges for females. As he stresses, the ranges of the two sexes overlap broadly for P4 — M;., in the genus. His samples for males, further- more, are small, numbering only 5. 4 and 3 in these cases. Parapapio is characteristic of the older part of the South African Pleistocene sequence. P. jonesi is well represented at Sterkfontein Figure 2. Parapapio jonesi Broom. Field no. 122-66K. Right lateral view of incomplete mandible of male. The ramus fragment with Pi — Mi has been reversed from the left side. X 1. 4 BREVIORA No. 282 and Makapan, by 24 and 15 specimens, respectively, and one in- dividual from Taung has been referred to it. The only specimens of Panipapio recorded from Swartkrans and Kromdraai are five individuals, 3 and 2, respectively, of a small form, and these are fragmentary. Freedman placed them with some hesitation in P. jonesi, pointing out that better material might in future require their separation. Two partial female mandibles from Swartkrans have "fairly large and quite deep mandibular fossae," a feature con- spicuously lacking in those from the earlier deposits. The Kanapoi specimen agrees with the earlier South African material and adds one more to the small list of species in common between the early Pleistocene of eastern and southern Africa. The field work was supported by National Science Foundation Grant no. G.A. 425. The drawings are the work of Mr. Arnold D. Clapman. REFERENCES Freedman, L. 1957. The fossil Cercopithecoidea of South Africa. Ann. Transvaal Mus.. 23: 121-262. 1960. Some new cercopithecoid specimens from Makapansgat, South Africa. Palaeont. Afric, 7: 7-45. Patterson, B. 1966. A new locality for early Pleistocene fossils in northwestern Kenya. Nature', 212: 577-581. (Received 19 June 1967.) BREVIORA MmseiLiiimi of Comparative Zoology Cambridge, Mass. 5 April, 1968 Number 283 SCYTHIAN AMMONOIDS FROM TIMOR Bernhard Kummel One of the more important areas of ammonitiferous Scythian (Lower Triassic) strata is the island of Timor. The monograph by Welter (1922) on the Lower Triassic ammonoids of Timor is the standard reference. However, an earlier paper by Wanner (1911) did describe four species of Lower Triassic ammonoids. A number of additional comments and descriptions on these Timor am- monoids were contributed by Spath (1934). In the course of a general study of all Scythian ammonoids I have had the opportunity to examine many collections from Timor including the type speci- mens of the Wanner (1911) and Welter ( 1922) studies. This note was prepared to make some additions and corrections to our knowledge of Scythian ammonoids of Timor. Prosphingitcs austini Hyatt and Smith is recorded for the first time from Timor. This is a common species of Oweuiles Zone age in the circum-Pacific and Arctic regions. Prenkites simdaicus Welter is established as the type of a new genus, Vickohlerites. Previous problems in the interpretation of this species were due to the inaccurate suture illustrated by Welter (1922). The genus Parowcnitcs Spath (1934) was established for an owenitid with a presumed goniatitic suture as illustrated by Welter (1922, pi. 169 (15). figs.^5, 8). Examination of Welter's types clearly shows that the goniatitic suture is the result of excessive grinding and polishing of the specimens. Preparation of other paratypes shows that the suture is ceratitic; hence the genus Par- owenites Spath is a synonym of Owenites Hyatt and Smith. The genus Metadagnoceras Tozer is recognized for the first time from the bed with manganese coated fossils at Nifoekoko. This specimen had previously been noted by Spath ( 1934, p. 269, foot- note) as being "very close to Dagnoceras terbunicuin." 2 BREVIORA No. 283 For the loan of specimens used in this study I am very grateful to Dr. H.K. Erben of Bonn University. Dr. M.K. Howarth of the British Museum (Natural History), and Dr. H.J. MacGillavry of the Geological Institute. Amsterdam University. The laboratory aspects of this study were supported by N.S.F. grant GB-5 109X. SYSTEMA lie PALEONTOLOGY Class CEPHALOPODA Cuvier, 1797 Subclass AMMONOIDEAZittel. 1884 Family PROPTYCHITIDAE Waagen, 1895 Genus Owenites Hyatt and Smith, 1905 Type species, Oweniws koeneni Hyatt and Smith, 1905 Owenites simplex Welter Plate 1. figures 1-9 Owenites simplex Welter. 1922: 153, pi. 169(15), figs. 1-8; Kutassy, 1933: 606. Paiowenites simplex, — Spath. 1934: 187, fig. 5S; Kummel in Arkell et at., 1957: L138, figs. 171-la,b; Kummel and Steele. 1962: 647. Owenites kwangsiensis Chao. 1959: 83, 250, pi. 22. figs. 1-6, text-fig. 26b. Owenites plicatiis Chao, 1959: 85. 251, pi. 22. figs. 19-21, text-fig. 26e. Owenites aff. plicatus Chao. 1959: pi. 22, figs. 24. 25. Owenites costatus Chao. 1959: 83. 249. pi. 22, figs. 10-18. 22, 23. text-fig. 26c. Owenites eostatns var. lenticiiUiris Chao, 1959: 84. 249, pi. 22, figs. 7-9, text-fig. 26d. Spath (1934: 187) established the genus Parowenites primarily on the basis of the goniatitic suture of Owenites simplex as repro- duced by Welter (1922: pi. 169(15). figs. 5, 8). Examination of Welter's type specimens plus a number of paratypes shows that the sutures reproduced by Welter were exposed by grinding and polishing. The specimen which yielded the suture of Welter's figure 8 is apparently lost, as it is not in the collections of the Paleontolo- gical Institute. Bonn University.^ However, the paratype of Wel- ter's plate 169( 15). figures 3-5 is available, and this shows clearly that excessive grinding of the lateral area resulted in the simple, goniatitic suture. One of the unfigured paratypes of Welter (Pi. 1, 1 Abbreviations in this paper: MCZ = Museum of Comparative Zoology; GPlBo =: Geologisch-Palaontologisches Institut. Bonn; BMNH = British Museum (Natural History). 1968 SCYTHIAN AMMONOIDS FROM TIMOR 3 figs. 8, 9 of this report) yielded the suture of Figure 3E. As can readily be seen, the two lateral lobes are ceratitic and the whole aspect of the suture is owenitid. The measurements of 27 specimens of this species are given on Table I. and plotted on the graph of Figure 1. There is consider- able variation in the diameter of the umbilicus which interestingly enough is not obvious on casual inspection of the specimens. What is more interesting is that the diameter of the umbilicus can vary from one side of the conch to the other. For instance, one of the paralectotypes (PI. 1, figs. 3-5) has an umbilical diameter of 3.7 mm on one side and 4.6 mm on the other side of the conch. H 16 15 14 I- 13 12 11 10 9 8 7 6 5 4 3 2 1 0 5 10 15 20 25 30 35 40 DIAMETER Figure 1. Variation in whorl height (H) and whorl width (W) of Oweuites simplex from limestone blocks with Owoiitcs egrediens, Bahati, Timor. w X xk 4 BREVIORA No. 283 The form described by Chao (1959) as Owenites kwangsiensis from the Owenites Zone of Kwangsi, China, I consider to be iden- tical to O. simplex Welter (Figs. 3E, H). From the same horizon and locality which yielded Owenites kwangsiensis. Chao (1959) described O. costatus and O. plicatiis. Of the first of these species Chao had 1 7 specimens which he compared mainly with Owenites pakiingensis Chao; the latter I consider a synonym of Owenites koeneni (Kummel and Erben. 1968). Owenites costatus does have constrictions and ribs but these are stated to be developed on an irregular pattern, thus differentiated from Owenites kwangsien- sis. Examination of the 31 specimens from Timor of Owenites simplex, representing all growth stages, shows that there is great variation in the pattern of ornamentation. Every one of the speci- mens of Owenites costatus illustrated by Chao ( 1959: pi. 22, figs. 10-18, 22, 23) can be directly compared to one of the Timor speci- mens before me. Owenites plicatus Chao was established on two fragmentary specimens. On the basis of the description and illus- trations of these two specimens I can see no reason for separating them from O. simplex. The apparently more simplified suture (fig. 21) was noted by Chao as most probably reflecting imma- turity and not a distinctive new pattern. Owemtids of the type of O. simplex are only known from Timor and Kwangsi. China. Occurrence — Dark red limestone blocks with Owenites egre- diens of Bahati, Timor. Repository — All specimens are in the Paleontological Institute, Bonn University. Family PARANANNITIDAE Spath. 1930 Genus Prosphingites Mojsisovics, 1886 Type species, Prosphingites czekanowskii Mojsisovics, 1886 Prosphingites austini Hyatt and Smith Plate 2, figures 1-16 Prosphingites austini Hyatt and Smith, 1905: 72. pi. 7. figs. 1-4; Freeh. 1908, pi. 63. fig. 5: Krattt and Diener. 1909: 160; Diener. 1915: 233; Smith. 1932: 98, pi. 7, figs. 1-4; Spath, 1934: 195-196; Kummel and Steele, 1962: 683; Kummel 1965: 544. Prosphingites sp. indet. Spath. 1921: 298, 301. Prosphingites spathi Freboid. 1930: 20. pi. 4, figs. 2. 3. 3a; Spath, 1934: 195, pi. 13, figs. 1, 2; Tozer. 1961: 58, pi. 13. figs. 1. 2; Kummel and Steele, 1962: 687; Kummel, 1965: 544. 1968 SCYTHIAN AMMONOIDS FROM TIMOR 5 Prusphiiigites sinensis Chao, 1959: 122. 297, pi. 25. figs. 1-5, pi. 27, figs. 1-17. text-figs. 40a-c; Kummel. 1965: 544. Prosphingites radians Chao. 1959: 123. 298. pi. 28. figs. 12-16. text-fig. 39d; Kummel, 1965: 544. Prosphingites oralis Kiparisova. I960: 137, pi. 33, figs. 2-4; Kiparisova, 1961: 115.pl. 25. figs. 6-9. Prosphingites orientalis Kiparisova. 1961: 117. pi. 26. figs. 1, 2. Prosphingites slossi Kummel and Steele. 1962: 683. pi. 101. figs. 8-17; Kum- mel. 1965: 544. None of the three previous studies on Scythian amnionoids from Timor (Wanner, 1911; Wehcr, 1922; Spath, 1934) contains any mention of the genus Prosphingites. I have available five specimens that can with confidence be assigned to Prosphingites austini Hyatt and Smith. These specimens are from Fatoe Kosat and are part of the Jonker Collection of the Geological Institute, Amsterdam University. Measurements (in mm) on four of these species are as follows: D W H U 34.5 21.7? 13.3 11.8 30.8 16.7 13.6 8.4 27.8 ? 12.8 7.8 27.2? 17.0 11.3 8.2 26.0 18.5 12.3 8.2 The similarity of these Timor specimens with Prosphingites slossi Kummel and Steele (1962: 683. pi. 101, figs. 8-17) and other mid- Scythian prosphingitids is indeed remarkable (PI. 2). Documenta- tion of species of this genus has been slow. The genotype (P. czekanou'skii Mojsisovics, 1886) is from northern Siberia and of late Scythian age. Hyatt and Smith (1905) then described Pros- phingites ciustini from the Meekoceras beds of the Inyo Range, southeastern California. Frebold (1930) and Spath (1934) re- corded the Spitsbergen species — Prosphingites spathi — which they considered late Scythian in age. It was not until the post-war period that descriptions of a number of new species from middle and late Scythian horizons began to appear. These were from Scythian faunas of the Primorye Region (Kiparisova. 1961). southern China (Chao, 1959). New Zealand (Kummel. 1965). and Nevada (Kummel and Steele, 1962). From the mid-Scythian, Owenites Zone, the following species of Prosphingites are known: Prosphingites austini Hyatt and Smith, P. oralis Kiparisova, P. 6 BREVIORA No. 283 orientalis Kiparisova, P. sinensis Chao, P. involutus Chao, P. kwangsianus Chao, P. radians Chao, P. spathi Frebold, and P. slossi Kummel and Steele. Most of these species were introduced over a short period of time (1959-1962) and each author was generally without knowledge of the others' activities. We thus have an accumulation of species in which very little comparative analysis accompanied their original descriptions. When I introduced the species Prosphingites slossi (Kummel and Steele, 1962) I was fully cognizant of its close relationship and possible identity to P. austini Hyatt and Smith (PI. 2, figs. 1-3) ; however, on the argument that Prosphingites austini was known only from a single and not a very well preserved specimen, I con- sidered it best to ignore the species. However, subsequently, on study of all new Scythian species that had been introduced, and through restudy of P. austini, I am now convinced that P. slossi is a synonym of P. austini. Kummel and Steele (1962) presented measurements on 49 specimens of P. slossi from the Meekoceras beds at Crittenden Spring, Nevada. These data show that at least within that population there is considerable variability in conch width and umbilical diameter. This collection also clearly showed that there was considerable variation in the pattern of constrictions. Of the Timor specimens recorded here, only four yielded com- plete measurements. The range of umbilical diameters in these four specimens falls well within the range of umbilical diameters in the Crittenden Spring fauna of Prosphingites slossi. The whorl width tends, however, to be greater in the Timor forms than in the Nevada fauna. This, however, considering the small size of the sample and the identity or close similarity of the other conch features, is not sufficient reason to even suggest specific distinction. The pattern of ornament, and variations in the nature of the um- bilical shoulder and wall are the same. The sutures can be said to be identical (Fig. 2). Most of the other species placed here in the synonymy of Pros- phingites austini are known from very few specimens. Large num- bers of specimens of P. spathi are in the British Museum ( Natural History), but only a few measurements are available and nearly all of this Spitsbergen collection consists of small phragmocones or juvenile specimens. It is uncertain from the text but it appears that the fauna of Prosphingites spathi from Ellesmere Island studied by Tozer (1961) consists of very few specimens, perhaps only three. Both Spath (1934) and Tozer (1961) have emphasized the varia- bility of this species, including its suture. I can find no criteria by 1968 SCYTHIAN AMMONOIDS FROM TIMOR Figure 2. Diagrammiitic representation of the suture of: A, Prosphiiij^itcs aiistini Hyatt and Smitii ( 1905: pi. 7, fig. 4) from Meekoceras beds. Union Wash. Inyo County, California, USNM 75256. at a diameter of 20 mm; B, Prosphingites slossi Kummel and Steele, from Meekoceras beds, Crittenden Spring, Nevada, paratype. at diameter of 18 mm; C. Prosphingites slossi Kummel and Steele, from Meekoceras beds. Crittenden Spring, Nevada, para- type. at a diameter of 30 mm; D, Prosphingites spathi Frebold, from Tozer (1961: pi. 13. fig. Ic) from Meekoceras beds. Blind Fiord Formation, Ellesmere Island, at a diameter of 16 mm; E, Prosphingites spathi Frebold, from Spath (1934: pi. 13, fig. le) from Owenites Zone, Spitsbergen; F, Prosphingites aiistini Hyatt and Smith, specimen in Yonker collection, from Timor, Department of Geology, Amsterdam University, at a diameter of 22 mm; G, Prosphingites oralis Kiparisova (1961: fig. 79) from Owenites Zone, Primorye Region, at a whorl height of 5 mm; H, Prosphingites sinensis Chao (1959: fig. 40a) from Olenites Zone, Kwangsi, China, at a diameter of 20 mm. which one can separate P. spathi from P. austini. In regards to the age of the Spitsbergen forms it has been pointed out by Kummel ( 1961) and Tozer (1961) that the beds containing P. spathi are of mid-Scythian Owenites Zone age rather than late Scythian as be- lieved by Spath. The mid-Scythian, Owenites Zone, strata of the Primorye Region contain, according to L. D. Kiparisova (1961), two species of 8 BREVIORA No. 283 Prosphingites, namely, P. ovalis Kiparisova and P. orientalis Kipari- sova. The first of these species, P. oralis, was said to difi'er from P. spathi in its larger shell, oval coiling, more denticulated lobes, and narrower ventral saddle. On close examination none of these criteria are valid. The Arctic P. spathi so far is known from very few specimens on Ellesmere Island and several but mainly juvenile forms from Spitsbergen. Even so, size is hardly a parameter for species distinction, especially in a case like this. In regards to the suture, an inspection of Figure 2 demonstrates considerable varia- bility in this respect. The Crittenden Spring fauna also shows considerable variability in details of the suture. The same varia- bility has been noted for P. spathi by Spath (1934) and Tozer (1961). Authors continue to place too high a degree of im- portance on minor variations in suture patterns. Time and trouble generally dictate the representation in a description of only one suture pattern. My own experience has amply shown that the suture is as variable as any other conch feature in most if not all Scythian ammonoids. From the same horizon and locality in the Primorye Region, Kiparisova (1961) has described Prosphingites orientalis. From the two illustrations given by Kiparisova (1961 : pi. 26, figs 1,2). P. orientalis differs from P. oralis primarily in the more acute rounding of the umbilical shoulder. The contrast in the nature of the umbilical shoulder of these two species is precisely the same as seen in the fauna of Prospliingites slossi from Crittenden Spring, Nevada (Kummel and Steele, 1962: pi. 104, figs. 8-17). The supposed ovate coiling of P. oralis is suspect, as not all of the illus- trated specimens show this. Even so, variants showing slightly ovate coiling are common in Scythian ammonoids. Again, I can see nothing in the suture, conch shape or pattern of constrictions in these Primorye forms to separate them from P. austini, P. slossi, or P. spathi. Among the very large Scythian fauna of Kwangsi, China, Chao (1959) recognized a number of prosphingitids. Among these, Prosphingites radians must be considered a synonym of P. austini and the other species discussed above. The conch form, pattern of constrictions and suture (Fig. 2) are well within the range of variation of what is here defined as P. austini. There is, in addition, Prosphingites sinensis, of which I am not as confident in placing in the synonymy of P. austini. The description and illustrations of this species leave much to be desired. However, on the basis of avail- able data, I consider it most probable that this species is the same 1968 SCYTHIAN AMMONOIDS FROM TIMOR 9 as P. austini. Even the unusual suture, in terms of the other forms combined with P. austini. may not be definitive (Fig. 2H). Prosphingites kwangsicmus Chao { 1959: 296, pi. 28, figs. 17-22) is reported to be a smooth form lacking constrictions. The species, however, is based on two incomplete specimens of modest preserva- tion. More satisfactory documentation of this species I suspect would show it to be conspecific with all the forms placed here in Prosphingites austini. Occurrence. — The Timor specimens recorded here are from red limestones at Fatoe Kosat, Timor. Repository. — Collection of Lower Triassic ammonoids made by Dr. H. G. Jonker during his Timor expedition in 1915-1916. Specimens are deposited in the Timor collection of the Geological Institute, University of Amsterdam. Genus Vickohlerites n.g. Type species, Prenkites sundaicus Welter, 1922 Conch evolute, with depressed whorl section, broad rounded venter, conch smooth except for growth lines. Suture consists of two denticulated lateral lobes and an auxiliary lobe on the umbili- cal wall. The type species is from an unknown horizon on Timor. Conspecific specimens are present in the Subcolumbites fauna of Chios and an indeterminant species (but not conspecific) is present in the Subcolunibites fauna of Kotal-e-Tera, Afghanistan (Kummel, 1968). Vickohlerites sundaicus (Welter) Plate 1, figures 10. 11 Piciikites sundaicus Welter, 1922: 150, pi. 168 (14), figs. 18-21; Kutassy, 1933: 621; C. Renz, 1945: 301; Renz and Renz. 1947: 60; Renz and Renz, 1948: 29. pi. 12. fig. 1; Chao. 1959: 306. "Prenkites" sundaicus. — Spalh, 1930: 77; Spath. 1934: 188, 209. Welter (1922) described this species on the basis of a single specimen from Noel Niti, Timor, and was quite positive as to the close relationship of his specimen to Prenkites malsorensis Arthaber from the Subcolumbites fauna of Albania. This close relationship is difficult to see. Prenkites is a more involute form, with depressed whorls which contract on the adoral quarter volution. The umbili- cal shoulders are subangular and bear fine nodes. The Timor speci- men has a diameter of 40.7 mm, an adoral whorl of 21.5 mm, a height of the adoral whorl of 13.2 mm and an umbilical diameter 10 BREVIORA No. 283 of 20.3 mm. The conch is evolute, the umbilicus comprising ap- proximately 50 per cent of the conch diameter. The whorls are depressed with the maximum width at the umbilical shoulder. The venter is arched and grades with no perceptible ventral shoulder to the umbilical shoulder which is acutely rounded. The umbilical wall is steep but not vertical. The last half volution of the specimens is body chamber and shows traces of delicate growth lines. The penultimate half volution bears a series of weak forward projecting ridges which are most prominent on the center part of the venter and disappear completely midway between the venter and the umbilical shoulder. It is in Welter's (1922: pi. 14, fig. 21) representation of the suture that one can pinpoint the uncertainty in interpretation of this form. Welter's drawing of the suture covers only the portion from the venter to the umbilical shoulder, but implied that is was a complete suture. A new drawing of the suture is given on Figure 3J. As can be seen, there are two prominent lateral lobes but the umbilical wall bears a good portion of a fairly large auxiliary saddle and a small but very distinct denticulated lateral lobe. The first author to comment on Prenkites simdaicus after Welter was Spath (1930: 77) who remarked *' 'Prenkites' simdaicus Welter, in whorl shape resembles Columbites, but in suture line it is closer to Siib- columbites . . ." In terms of Welter's representation of the suture this statement is correct. In Columbites the second lateral lobe is generally very small consisting of a single prong, whereas in Sub- columbites the second lateral lobe, though much smaller than the first, is more highly developed in terms of its breadth and pattern of denticulation. On the basis of suture alone, Prenkites sundaicus can not be attached to either Columbites or Subcolumbites. There is a general similarity in conch shape of Prenkites sundai- cus with some groups of Subcolumbites. Among the subcolumbitids three distinct groups can be recognized. There is first of all the perrinismit/ii group with a tendency for carination of the venter; secondly, the dusmcmi group with a more marked development of the reticulate ornamentation, and a compressed whorl section, but lacking the tendency toward carination; finally, there is the robus- tus-multijormis group characterized mainly by their depressed whorl section. It is to this last group that Vickohlerites sundaicus has great resemblance in conch form. Within the two subfamilies of the Paranannitidae those genera assigned to the Columbitinae tend to have sutures lacking an auxiliary lobe, whereas within the 1968 SCYTHIAN AMMONOIDS FROM TIMOR 11 Paranannitinae an auxiliary lobe is commonplace, as in Prosphing- ites, Zenoites, Chiotites, etc. It is within this subfamily that Vic- kohlerites sundaicus belongs. Renz and Renz (1948-' 24, pi. 12, fig. 1) have described and illustrated a single specimen from the Siibcoliimbites fauna of Chios as a representative of this species. The general conch form of their specimen is the same as that of the type specimens from Timor. This Chios specimen measures 55.5 mm in diameter, 21.4 mm for the width of the adoral whorl, 19 mm for the height, and 23.5 mm for the width of the umbilicus. The dimensions of the whorl height and umbilical diameter in the two specimens are rea- sonably similar. However, the Timor specimen has a broader whorl than the Chios specimen by approximately 14 per cent. This dif- ference in whorl width is difficult to evaluate, as each locality has yielded only a single specimen. In addition, the Chios specimen, apparently, does not show the suture. The overall similarity of the Chios specimen to that from Timor is such that, in spite of the differences and lack of data mentioned above, the two specimens should be considered as conspecific. An undescribed specimen from a Subcolutnbites fauna at Kotal- e-Tera, Afghanistan, is clearly congeneric with the Timor and Chios specimens discussed above but is not conspecific. There are significant differences in whorl cross-section and in the suture. Occurrence. — The holotype (PI. 1, figs. 10. 11) is from Noel Niti, Timor. Welter (1922: 85, 86) considered this specimen to come from his lowest Triassic horizon. He lists (on p. 150) the forms associated with this specimen as Meekoceras sp. indet. The available evidence does not allow any precise determination of the age, but the biological affinities of this species suggest that it is late Scythian in age. Repository. ■ — The holotype is in the Paleontological Institute of Bonn University; the specimen from Chios is in the Natural History Museum, Basel J 13576. Family MEEKOCERATIDAE Waagen, 1895 Genus Metadagnoceras Tozer, 1965 Type species, Metadagnoceras pulcher Tozer Metadagnoceras freemani n. sp. Plate 1, figures 12, 13. This species is proposed for the specimen from Timor mentioned by Spath ( 1934: 269, footnote) as "an isolated example of a new 12 BREVIORA No. 283 species from Timor . . . which is very close to Dagnoceras terbiini- cum." The specimen measures 47.8 mm in diameter. 14.7 mm for the width of the last whorl, 21.7 mm for the height of the last whorl, and 10.4 mm for the diameter of the umbilicus. The conch is com- pressed with a low, arched venter and rounded ventral and umbili- cal shoulders. The suture (Fig. 3D) has a very large first lateral lobe with denticulations extending all along the vertical side to a narrow irregular first lateral saddle. There is a small denticulated second lateral lobe on the umbilical shoulder and wall. This species does not resemble Dagnoceras terbimicum in conch shape; the differences are centered mainly in the character of the venter. In the suture, the first lateral saddle and the first lateral lobe of the two species are quite similar but the remainder of the suture is quite different. In general conch morphology Metadagno- ceras jreenumi is quite similar to the type species, M. pulcher Tozer. The sutures are similar in basic plan but differ in significant details ( Figs. 3 A, D). The suture is quite similar to an undescribed species from the Tobin Formation, Nevada, but the whorls are of quite different proportions and the conch much more evolute (Fig. 3B). Occurrence. — Nifoekoko, Timor, from bed with manganese coated fossils, including Alhanites, Prohungarites, etc. Repository. — British Museum (Natural History) C33701, holotype. 1968 SCYTHIAN AMMONOIDS FROM TIMOR 13 H Figure 3. Diagrammatic representation of the suture of: A, Metadagno- ceras piilclier Tozer (1965: fig. 9). from British Columbia, at a whorl height of approximately 20 mm; B, undescribed new species from Tobin Formation, Nevada, at a whorl height of 11 mm, MCZ 9638; C, undescribed species of Mctaclagnoceras collected by N. J. Silberling from Star Peak Formation, Nevada, at a whorl height of 22 mm; D, Metadagnoceras frce- imini n. sp., holotype. from Nifoekoko. Timor, at a whorl height of 14 mm, BMNH C33701; E, Olenites simplex Welter, original drawing based on one of Welter's unfigured paratypes (PI. 1. figs. 8, 9 of this report) at a diameter of 13 mm; F, Oneiiites costatus Chao ( 1959: fig. 26c) at a diam- eter of 20 mm; G Owenites costatus var. lentkularis Chao (1959: fig. 26d ) at a diameter of 12 mm; H, Olenites kwungsiensis Chao (1959: fig. 26b) at a diameter of 25 mm; I, Owenites plicatiis Chao (1959: fig. 26e) at a diameter of 30 mm; J, Vickohlerites siindaiciis (Welter), at a diameter of 30 mm, holotype, GPIBo 231. 14 BREVIORA No. 283 TABLE 1 Measureiiiciits of Owenites simplex Welter from Timor N 0. D W H U W/D H/D U/D 1 39.8 10.3 16.4 1 0.0 25.9 41.2 25.1 2 39.6 10.3 15.6 7.2 26.0 39.3 18.2 3 32.8 9.5? 13.0 13.4 28.9? 39.6 40.8 4 28.9 8.9 12.2 7.2 30.8 42.2 24.9 5 25.9 7.8 11.2 6.3? 30.1 43.2 24.3? 6 25.4 8.4 10.0 8.0 33.1 39.4 31.4 7 25.1 •) 10.9 6.3 '7 43.4 25.1 8 24.2 7.3 10.5 5.8 30.2 43.4 23.9 9 22.6 7.4 9.3 5.1 32.7 41.2 22.6 10 21.2 8.0 9.1 5.6 37.7 42.9 26.4 1 1 21.1 6.9 9.7? 4.9? 32.7 45.9? 23.2? i: 19.7 7.1 7.5 6.2 36.0 38.1 31.5 13 18.8 7.0 8.0 4.0 37.2 42.6 21.3 14 17.4 •) 8.4 3.1? •) 48.3 17.8? 15 16.1 6.1 7.0 3.8 37.9 43.5 23.6 16 16.0 5.9 6.8 4.6 36.9 42.5 28.8 17 15.6 6.2? 6.8 3.8 39.7? 43.6 24.4 18 15.5 6.1 7.8 4.0 39.4 50.3 25.8 19 15.1 6.1 6.4 3.7 40.4 42.4 24.5 20 13.3 5.3 6.2 2.7 39.8 46.6 20.3 21 13.3 5.1 5.4 3.3 38.3 40.6 24.8 -) ■» 14.6 5.6? 6.3 3.7 38.4? 43.2 25.3 23 12.4 5.3 5.4 •) 42.7 43.5 •) 24 10.4 4.3 4.8 1 .6? 41.3 46.2 15.4? 25 10.1 4.7 4.7 1.5? 46.5 46.5 14.9? 26 8.0 3.7 3.7 9 46.3 46.3 0 27 7.3 3.8 3.3 1.1? 52.1 45.2 15.1? D = diameter; W = width of adoral whorl; H - height of adoral whorl; U = diameter of umbilicus. No. 2. Lectotype. Welter (1922: pi. 169(15), figs. 1. 2: PI. 1. figs. 1, 2 of this report) GPIBo 234a. No. 8. Paralectotype, Welter (1922: pi. 169(15). figs. 3-5; PI. 1. figs 3-5 of this report) GPIBo 234h. No. 10. Paralectotype. Welter ( 1922: pi. 169( 15), figs. 6, 7; PI. 1, figs. 6, 7 of this report) GPIBo 234c. No. 20. Suture specimen from Bahati, Timor (PI. 1, figs. 8, 9, Te.\t-fig. 3E of this report) GPIBo. 1968 SCYTHIAN AMMONOIDS FROM TIMOR 15 REFERENCES Arkell. W. i. et ul. 1957. Treatise on Invertebrate Paleontology, Ammonoldea. R. C. Moore ed., Pt. L, Mollusca. 4, 490 pp. Chao, Kingkoo 1959. Lower Triassic ammonoids from western Kwangsi, China. Palaeont. Sinica, 145: 1-355. DiENER, Carl 1915. Fossiliiim Catalogus. I. Animalia. Pt. 8, Cephalopoda Triadica. Berlin, 369 pp. Frebold, H. 1930. Die Altersstellung des Fischhorizontes des Grippianiveaiis iind des unteren Saurierhorizontes in Spitzbergen. Skr. Svalb. og Ishavet, 28: 1-36. Frech, F. 1903- Lethaea Geognostica. Das Mesozoicum. I, Trias. Stuttgart, 1908. 623 pp. Hyatt, A. and J. P. Smith 1905. The Triassic cephalopod genera of America. Prof. Pap. U.S. Geol. Surv., 40: 1-394. Kiparisova, L. D. 1960. Novye vidy drennikh rastenii i besopzvonochnykh SSSR, Chast' 2. Novye rannetriasovye nautilus i prosfingit Yuzhnogo Primorya. Vsesoiuznyi nauchno-issIedovateTskii geologicheskii institut (VSEGEI). [New species of ancient plants and in- vertebrates of the U.S.S.R., Pt. 2. New early Triassic nautiloids and prosphingitids from the Primorye Region. All Union Sci. Res. Geol. Insl.. pp. 136-139.] 1961. Paleontologicheskoe obosnovanie stratigrafii triasovykh otloz- henii Primorskogo kraia. Chast'l. Golovonogie moUiuski. Vsesoiuznyi nauchno-issledovatel'skii geologicheskii institut (VSEGEI). trudy. [Paleontological foundation for the strati- graphy of the Triassic deposits of the Primorye Region. Pt. I, Cephalopoda. Trans. All Union Sci. Res. Geol. Inst.. N. Ser., 48: 1-278.) Krafft, a. v. and Carl Diener 1909. Himalayan Fossils. Lower Triassic Cephalopoda from Spiti, Malla Johar, and Byans. Mem. Geol. Surv. India, Palaeont. Indica, Ser. 15,6(1): 1-186. Kummel, B. 1961. The Spitsbergen arctoceratids. Bull. Mus. Comp. Zool., 123: 499-532. 16 BREVIORA No. 283 1965. New Lower Triassic ammonoids from New Zealand. New Zealand J. Geol. Geophys.. 8: 537-547. 1968. Additional Scythian ammonoids from Afghanistan. Bull. Mus. Comp. Zool.. in press. KuMMEL. B. and H. K. Erben 1968. Lower and Middle Triassic cephalopods from Afghanistan, Palaeontographica, in press. KuMMEL, B. and G. Steele 1962. Ammonites from the Meekoceras gracilitatis zone of Crittenden Spring. Elko County. Nevada. J. Paleont.. 36: 638-703. KUTASSY, A. 1933. Fossilium Catalogus. I. Animalia. Pt. 56, Cephalopoda Triadica II. Berlin, pp. 371-832. Mojsisovics, E. V. 1886. Arktische Triasfaunen. Beitriige zur palaeontologischen Char- akteristik der Arktisch-Pacifischen Triasprovinz. Mem. Acad. Imp. Sci. St. Petersburg, Ser. 7, 33: 1-159. Renz. Carl 1945. Beitriige zur Slratigraphie und Paliiontologie des ost-mediter- ranen Jungpaliiozoikum und dessen Einordnung im griechischen Gebirgssystem, Teil I and 11. Eclogae Geol. Helvetiae, 38: 211- 313. Renz, Carl and Otto Renz 1947. Uhersicht ijber eine untertriadische Ammonitenfauna von der Insel Chios (Griechenland ). Verhandl. Naturforsch. Gesell. Basel, 58: 58-79. 1948. Eine untertriadische Ammonitenfauna von der griechischen Insel Chios. Abhandl. Schweiz. Palaont., 66: 1-98. Smith, J. P. 1932. Lower Triassic ammonoids ox North America. Prof. Pap. U.S. Geol. Surv., 167: 1-199. Spath, L. F. 1921. On ammonites from Spitsbergen. Geol. Mag.. 58: 297-305, 347-356. 1930. The Eo-Triassic invertebrate fauna of East Greenland. Medd. omGronland, 83: 1-90. 1934. Catalogue of the fossil Cephalopoda in the British Museum (Natural History). Part IV, The Ammonoidea of the Trias. London, pp. 1-521. Tozer, E. T. 1961. Triassic stratigraphy and faunas, Queen Elizabeth Islands, Arctic Archipelago. Mem. Geo!. Surv. Canada, 316: 1-116. 1968 SCYTHIAN AMMONOIDS FROM TIMOR 17 1965. Latest Lower Triassic ammonoids from Ellesmere Island and northwestern British Columbia. Bull. Geol. Surv. Canada, 123: 1-45. Waagen, W. 1895. Salt Range fossils. Fossils from the Ceratite Formation. Pt. I, Pisces-Ammonoidea. Mem. Geol. Surv. India, Palaeont. Indica, Ser. 13, 2: 1-323. Wanner, J. 1911. Triascephalopoden von Timor und Roti. N. Jahrb. Mineral. Geol. Palaont.. Beil.-Bd.. 32: 177-196. Welter, O. A. 1922. Die Ammoniten der unteren Trias von Timor. Palaont. Timor, 11(9): 83-154. (Received 11 July 1967.) 18 BREVIORA No. 283 EXPLANATION OF PLATE 1 Figs. 1-9. Olenites simplex Welter from Timor. Figs. 1, 2, lectotype (Welter, 1922: pi. 169(15), figs. 1. 2) GPIBo 234a, XI; 3-5 paralectotype (Welter. 1922: pi. 169(15), figs. 3-5) GPlBo 234b, XI. 5; 6. 7, paralecto- type (Welter, 1922: pi. 169(15). figs. 6, 7) GPIBo 234c. XI. 5: 8, 9, paralectotype from which suture of Text-figure 3E was taken, GPIBo, X3. Figs. 10. 11. I'ickohlcrites siindaiciis (Welter) from unknown horizon on Timor. Holotype GPIBo 231. XI. Figs. 12, 13. Metadii^iiocerus frcenuini n. sp. from beds with managanese coated fossils, Nifoekoko. Timor. Holotype. BMNH C337()l. XI. 1968 SCYTHIAN AMMONOIDS FROM TIMOR 19 ^1 10 13 ;!-* - w. *, 11 "t>!. 8 \ > 20 BREVIORA No. 283 EXPLANATION OF PLATE 2 Figs. 1-16. Prosphin^itcs aiistini Hyatt and Smith. Figs. 1-3, from Meekoceras beds Union Wash, Inyo Range. California, holotype USNM 75256, XI. 5: 3, 4. Prosphlngites slossi Kummel and Steele, from Meeko- ceras beds. Crittenden Spring, Nevada, paratype MCZ 5781, XI; 6. 7, Prosphini^ites slossi Kummel and Steele from Meekoceras beds, Crittenden Spring, Nevada, paratype MCZ 5616. Xl; 8, Prospliingites slossi Kummel and Steele from Meekoceras beds, Crittenden Spring, Nevada, paratype MCZ 5615 XI; 9. 10. Prosphingites slossi Kummel and Steele, from Meekoceras beds, Crittenden Spring. Nevada, holotype MCZ 5614. XI; 11-16. Pros- phingites aiistini. Jonker Collection, Timor, Geology Department, University of Amsterdam, the suture of Text-figure 2F was taken from the specimen illustrated here on figures 15, 16. 1968 SCYTHIAN AMMONCIDS FROM TIMOR 21 Z^r^ ^^^ / 4 V- Vi ■ 4, *»»H ,,_-<^«£^H^'''^?^W"f .#; 5^- 8 12 11 5 V 14 t4,'«» /^- i'- ^^: BREVIORA Museiuiini of Comparative Zoology Cambridge, Mass. 5 April. 1968 Number 284 THE RELATIONSHIPS OF ANOLIS OF THE ROQUET SPECIES GROUP (SAURIA:IGUANIDAE) — III. COMPARATIVE STUDY OF DISPLAY BEHAVIOR George C. Gorman ^ INTRODUCTION In separate papers on chromosome cytology and electrophoresis of blood proteins (Gorman and Dessauer. 1966; Gorman and Atkins, 1967), the relationships of the roquet group of Anolis have been discussed in detail. The present paper is an attempt to use comparative ethology as an additional index of relationship and evolutionary history of the group. No attempt is made here to provide an "ethogram"' of any of the forms. Instead, some stereotyped action patterns that vary from island population to island population are compared as taxonomic characters. Taxonomy and Distribution Eight forms are considered in this paper. The species of the roquet group are found on all the islands of the Lesser Antilles from Martinique south to Grenada and west to Bonaire. To the north of Martinique are anoles of the hinuiculatus group; to the south, on Trinidad (a continental island), and to the west, on Curacao and Aruba, are members of the chrysolepis group (classi- fication into species groups follows Etheridge, 1960). These species groups seem to exclude each other, and the only recorded overlaps" (on Trinidad, St. Lucia, and Guayana) appear to be recent introductions (Gorman and Dessauer, 1965, 1966). Figure 1 summarizes the distribution of the group. 1 Present address: Museum of Vertebrate Zoology. University of Cali- fornia, Berkeley, California. BREVIORA No. 284 66° 64° ! 1 62° 60° OCCULATUS ^ Dominica ROQUET ' -A Wl3'''i"iq"e — 14° LUCIAE ^/j sj. Lucia TRINITATIS S. .Vincent GRISEUS EXTRE- ■'Bequia MUS 14 — Barbados — 12° LINEATUS Curacao ^^ BONAIRENSIS ^ Bonaire i AENEUS RICHARDI ,' ^j Grenada 12'- — 1 68 '_ J RICHARDI^ -* Tobacjo CHRYSOLEPIS ,-V, hoiiaireiisis i aeueus J ricluirdi richardi I , ,. . common Hb ( runarcli i^ri.seus Tn the 2n=36 group (primitive) A. r. ric/uirdi and A. r. gn'seus form a natural subunit characterized not only by proteins but by giant size. An hypothesis about the biogeographic movements of the group is being presented in a separate paper (Gorman and Atkins, 1968). The defense of this hypothesis will not be repeated here. However, it is presented as a working model with which the be- havorial results may be compared. A. luciae is considered the most primitive member of the group. Reasons include chromosome number, size (intermediate between the giants and the other five forms), and geographic position (more or less in the center of the group.) From St. Lucia, St. Vincent was twice colonized, leading to two species, trinitatis (small) and gri.sen.s (giant). From St. Vincent, the two reached Grenada and through further differentiation have become aeneiis (small) and richardi (giant). A. ricluirdi has reached Tobago without further difi"erentiation. The small Grenada species colonized Barbados (extremus) and from there Martinique {roquet). Or. alternatively, the small St. Vincent form (trinitatis) colonized Barbados directly, followed by a double radiation from Barbados to Grenada and Martinique. Behavior in Anolis Because of the elaborate displays of these strongly territorial lizards, the Anolis species have been of long-standing ethological interest. Greenberg and Noble (1944) presented the most com- prehensive discussion of a single species, A. carolinensis. This report was based upon four years of observation of approximately 6(X) lizards. Included in their paper are sections on aggression, courtship and mating, egg laying, dominance, subordination and hierarchies, territory, and sex recognition. Other studies, such as Kastle's ( 1963) on Norops f= Anolis) auratus. Carpenter's (1965) on A. townsendi, and Evans' (1938) on A. sagrei, provide valuable descriptions of displays and social behavior in single species. 4 BREVIORA No. 284 The only attempt at a broad comparative survey of Anolis be- havior was by Ruibal (1967), who described aggressive displays in twelve species of Greater Antillean Anolis. The genus is divided into two major groups (alpha and beta), based on an osteological difference (Etheridge, 1960), and sup- ported independently by chromosome differences (Gorman, 1965). Ruibal (1967) was interested in determining if there was any parallel between the alpha-beta dichotomy and behavior. Although Ruibal found species-specific differences, he wrote, "the analysis of the initial response in these anoles demonstrates that the behav- ioral patterns are of limited value in determining relationships within the West Indian species of Anolis." My own study, presented here, indicates the value of behavior as a systematic character within at least one species group. METHODS In 1963-64, I spent 10 months in the Caribbean area, based at the University of the West Indies in Trinidad. Subsequently, several additional brief visits were made to the islands where the roquet group occurs, as well as to neighboring islands. At first, the lizards were simply observed in the field and extensive notes on ecology and behavior were prepared. Later, they were studied in cages at the University of the West Indies. In general, two adult males were placed on opposite sides of an opaque partitioned cage. Several hours to days later, the partition would be removed and social interactions described. To gather quantifiable data the lizards were photographed on 16 mm motion picture film. The bulk of the data presented are from analysis of films made in the field. In some cases the films are of natural interaction, but high intensity terri- torial disputes are not frequent in natural situations; they are easily elicited, however, by tethering an "intruder" to a long pole and placing him within the territory of a resident. Films were analyzed on a Weinberg-Watson single frame analyst. This allows frame-by- frame analysis of the temporal sequence of displays. Amplitudes of movements were not measured precisely but were estimated. No great emphasis has been placed upon amplitude differences in display. Elicitation of Display Iguanids respond to visual cues. The introduction of a mirror into a cage elicited highly stereotyped displays identical to those 1968 DISPLAY BEHAVIOR OF SOME ANOLES 5 seen in the field in natural encounters. Carpenter (1962) has re- ported results of mirror experiments in Sceloporus, and a rather accurate description of an A. roquet extremiis display to a mirror was written for a popular audience by Dr. Maurice Burton (London Illustrated News, 29 August 1964). Neither the opponent's dewlap color nor the dewlap per se seems to be of great importance in eliciting a response from a resident. Greenberg and Noble (1944) found that restraining the dewlap by means of a coUodian coating did not affect the social status of the male. Dominant animals were still able to hold territory and win combats with normal rivals. In addition, normal males dis- played to males unable to extend the dewlap. I have found that members of the roquet group in captivity will display to other species if these are in the same size range. A planned experiment to study releasers by use of models pre- sented to residents in the field was thwarted at the outset. A. trini- tcitis are very abundant and quite willing to display to other anoles in the Botanic Garden in St. Vincent. A life-size wooden model of a male triniiati.s was constructed and. I thought, accurately painted. A slit was made in the throat region for insertion of paper dewlaps of the "right" or "wrong" color. If males responded aggressively to the model, then variations in size and color were planned to elucidate exactly which parameters were important to a resident territorial male. However, responses to the dummy were uniformly ncizative. Experiments involving a decapitated intruder were somewhat more successful. The decapitated head was pinned to a piece of sponge, cut approximately in the shape of a lizard torso. Several resident males attacked and/or displayed actively to this model, and even to the decapitated head alone, when dangled freely. This experiment was terminated when a griseus ran down from the crown of a tree and grabbed and ate the trinitatis head. In a reciprocal trial, a head-sized piece of sponge was pinned to the body of the decapitated lizard. This did not elicit aggressive display. Some males were painted bright red, others bright white. Results in several trials were equivocal. Some residents displayed or attacked, others fled from the intruder. Thus, something in the configuration of the head probably acts as a releaser. Another apparently important component in elicitation of dis- play is movement on the part of the intruder. A tethered male will usually flatten himself on the substratum in a classically submissive 6 BREVIORA No. 284 posture.^ The resident will watch. Slight jiggling of the intruder is sufficient to elicit immediately a high intensity response (biting at- tack or display). Description of Display Carpenter (1962) has standardized his descriptions of iguanid behavior by describing eight characteristics of display — site, posi- tion, posture, movement type, parts moved, units of movement, sequence, and cadence. Ruibal listed and defined the following movements and postures to describe aggressive encounters in Anolis. 1 . Dewlapping — the characteristic erection of the large gular appendage. 2. Bobbins — the vertical motion of the head. 3. Pushups — an extension of the limbs to raise the body off the ground. 4. Tail wags — vertical or horizontal movement of the tail. 5. Tail lashes — a very exaggerated movement of the tail in a broad whipping manner. 6. Tongue — the tongue may be everted and protruded from the mouth and in some species may become purple with en- gorged blood. 7. Nuchal and dorsal crests — the dorsal regional of the neck and the middorsal region of the body are usually raised as thin fins. Application of these descriptions to the roqiict species group would be as follows: A . The eight categories of Carpenter. 1 . Site: All the anoles of the roquet group are strongly arboreal. Most daytime observations are made of animals on tree trunks four to seven feet above the ground. Most displays are given when the lizards are on tree trunks; however, during a battle they will display from the ground. When they display they try to find a slightly elevated site, such as a knot on a tree. 2. Position: Orientation is usually perpendicular to the recipient of the display. If the opponents are parallel, then the heads are facing each other. The position chosen maximizes the opponent's view. 1 Darwin (1872) pointed out that submissive postures were often the antitheses of aggressive ones. Thus, in the anoles in submissive posture the crests are down, dewlap in. body low. 1968 DISPLAY BEHAVIOR OF SOME ANOLES 7 3. Posture: In aggressive encounters, nuchal and dorsal crests are generally raised. As noted by Greenberg and Noble (1944) for A. carolinensis. "The initial hostile reaction is a puffing or gorging of the throat followed almost immediately by lateral flat- tening so that the male appears much larger from side view." This serves as an excellent description for the roquet group and shall be referred to as "engorged head." These three categories are not of further use in this comparative study because they are constant for the species group. 4-6. Movement type, parts moved, and units of movement are encompassed by Ruibal's categories, and are discussed below. 7. Sequence: The order of appearance of units of movement will be exemplified by quotations from notes on actual encounters. These categories again are consistent for the species group. 8. The measured time for the performance of each unit of movement. This is the important category employed in comparing displays within the species group. B. The seven movements of Ruibal. 1. Dewlapping: The dewlap of the roquet group and all its near neighbors is well developed. The cadence of extension differs between the roquet group and the anoles that geographically sur- round it. 2. Bobbing: In the roquet group all eight forms bob the head in high intensity display, and each form shows a temporal se- quence that is unique. I also refer to nodding, in which the head is raised and lowered very slowly. This is a less stereotyped move- ment and seems to occur at lower intensity. 3. Pushups: Again, all forms in the group rise up on all limbs in high intensity display. Some have also been observed to include "rhythmic movements" as a part of the challenge display. This is a series of movements at very precise intervals — it may simply be head bobs, or pushups, or pulsating locomotion. 4. Tail wags: Horizontal movement of the tail is seen under situations of stress. It occurs when two anoles are actually light- ing and is not seen in the clearly dominant animal. It also occurs when an anole is captured by a herpetologist and is held in the hand. 5. Tail lashes: All species of the roquet group may use the tail in a challenge display. The frequency and amplitude of tail lashes is an important character in tracing relationships. 6. Tongue: Several members of the roquet group have been seen to protrude the tongue during display. I do not know how frequent or how widespread this movement is. The tongue is 8 BREVIORA No. 284 engorged with blood and quite bright in color. At the termina- tion of the display, while the dewlap is being retracted, the mouth is opened and closed several times as though a swallowing move- ment were occurring. This has been recorded for A. r. roquet, A. tiinitatis and A. bonairensis and may be present in the other species. 7. Nuchal and dorsal crests: Present and raised in all species during aggressive encounters. THE DISPLAY CATEGORIES OF THE ROQUET GROUP OF ANOLIS There are three obvious categories of display in this species group: activity bobs, courtship, and aggressive display. Only the last was studied in detail. (1 ) The activity bob corresponds to Carpenter's (1962) "As- sertion Display." For example, in changing position a lizard may run a few feet, stop, bob, run a few more feet, stop, bob, etc. This bob is also commonly seen after feeding. I have reason to believe that it is stereotyped and may be species-specific. However, I do not have enoutih data to discuss it further. This bob is most commonly given by females. (2) Courtship was difficult to study under field conditions, as most displays directed by males to females were followed by rapid departure by the female. There are no obvious differences among the species in courtship display, and my observations are con- cordant with those of Greenberg and Noble (1944) on the fol- lowing points: a) There is no sign ot aggression; i.e., engorged head, crest erection, and tail jerks are all absent. b) The approach pattern differs from the aggressive encounter in that there are numerous rapid low amplitude bobs and an irregu- lar, jerking walk. (3) Aggressive display. The most elaborate of the displays are also the most stereotyped, hence the most useful for analysis of evolutionary relationships. Intensity of display obviously varies, as has been pointed out by Kasde (1963) and Ruibal (1967). Ruibal limited his study to what he called the "initial response," which he defined as "the immediate and first response of the test animal to the decoy." Later he states that it is comparable but not identical to the chal- lenge display of Carpenter ( 1962) — "the initial response actually represents the first part of the challenge display." 1968 DISPLAY BEHAVIOR OF SOME ANGLES 9 I agree completely with Ruibal that there is a predictable, stereo- typed portion of lizard display behavior. However, the term "initial" in RuibaFs phraseology is unfortunate, because the lit- erally immediate response of a resident male to an intruder is quite unpredictable: the resident may flee, go into the engorged head posture and wait, attack by biting, or display at varying in- tensities. I prefer to use the term challenge display. All eight members of the roquet group have challenge displays which follow a definite sequence of actions: (1) extend dewlap; (2) rise high on all four hmbs; (3) raise the tail, then lower it; (4) lower the body; (5) give rapid head bob; (6) retract dewlap. The following excerpt from my field notes will exemplify this pattern: A. r. roquet — Martinique — 2 July 1965: A tethered male intruder is placed about 12 inches above a resident male (R) perched on a tree. The intruder starts running away from resident. R slowly moves up several inches, pauses. Intruder jiggled. R backs away. R turns perpendicular to intruder. Intruder moves and R follows, stopping a foot above. Head is in engorged position R moves to two feet above intruder; extends dewlap fully; rises high on all four limbs; 5 tail lashes; lowers body; dewlap half re- tracted. R pauses for several seconds. Extends dewlap fully while body rising high on all four limbs. Tail raised high and then lowered slowly. Bobs head. Retracts dewlap. Display is repeated again. Note that the sequence of movements is the same in the two full displays, but in the first instance head bobbing did not follow the tail movement. The six sequential steps of the challenge display are present or absent in independent fashion, but when they do appear they are evidently never out of order. That is, the dewlap might not be extended but the resident male still lashes his tail and then bobs; or he may omit the bobs, or he may not fully retract the dewlap at the end of the sequence. However, he will never bob, then lash the tail, and then extend the dewlap. Some of the steps are better correlated than others. The tail movement almost always occurs with the dewlap fully extended. And, if there be a rapid bob with dewlap extended, the dewlap is almost always retracted at the end of the sequence unless the dis- playing lizard immediately launches into a new sequence. The high intensity challenge display is interpreted as occurring when there is an intermingling of attack and withdrawal stimuli. The elaborate display is often preceded by backing away from the intruder. Indeed, the displays often follow unsuccessful direct 10 HREVIORA No. 284 attacks. The following encounters may serve as illustrations. 1. A. trinitatis — St. Vincent — 8 July 1966: Intruder is placed three feet below resident who is 10 feet up in a coconut tree. R runs down, bites at intruder. Bites again. They lock jaws. R breaks away and runs back up. Bobs rapidly. Attacks again several times, biting at the head of intruder. R backs away. Extends dewlap half-way. Rapid bob. R attacks again. Grab jaws. R flipped to ground. 2. A. r. roquet — Martinique — 3 July 1965: Tethered male intruder placed four feet below resident who turns immediately, runs down and stops one foot above intruder. Pauses. Then rushes at intruder and bites at head; extends dewlap, bobs. Runs jerkily above intruder; full dewlap extension, head bob, dewlap slowly retracted. This is repeated again. Then R moves toward in- truder and bites at him. K then backs away and moves to one foot above intruder. Dewlap fully extended, head bobs, then dewlap retracted. Repeated several times. Each time there is a slight move- ment of the intruder, there has been a full dewlap extension followed by head bobbing. Resident then runs to one foot above intruder. Faces down and is approximately perpendicular. Rises high on all four limbs, tail comes up slowly, full dewlap extension; body lowered, head bob, dewlap retracted. Then R moves up and away from intruder. In Trinidad. I filmed a long sequence of a male A. aeueus dis- playing to a tethered intruder in which the resident gave the full display, then ran up and away from the intruder. I then moved the intruder up a comparable distance, the resident again quickly dis- played and moved away again. This was repeated several more times. These examples serve to support the hypothesis that the elab- orate challenge display is not simply a ritualization to avoid physi- cal fighting, but rather a display given in a situation with conflicting stimuli of attack and flee. Under natural field conditions these displays have been seen only when the opponents were extremely evenly matched. Usually the engorged head of a dominant resident is enough to stimulate flight in a subordinate or intruding male. Combat is by no means always the initial high intensity response, as the following notes indicate. A. r. roquet — Martinique — 3 July 1965: Intruder placed near male resident who immediately goes into engorged head posture and moves down toward intruder. R extends dewlap fully and rises up on all four limbs; three tail lashes; dewlap slowly retracted. Some encounters are settled without actual biting. I once witnessed a very lengthy dispute between two male A. trinitatis in 1968 DISPLAY BEHAVIOR OF SOME ANGLES H the Botanic Gardens of St. Vincent. Both were on a broad tree. At the start, one was about two feet above the other. After some fifteen minutes with dozens of high intensity displays, they slowly moved towards one another and, when contact was made, they butted heads, snouts pointing downwards. Then one backed off and climbed up and around the tree. Although all these encounters describe males, females also dis- play to females. Although the secondary sexual characters of dewlap and crests are much reduced, and the females are also much smaller than adult males, the postures and motions appear identical to those of males. I never saw a natural interaction be- tween females, but did observe and film female display by use of a tethered intruder. SYSTEMATIC SECTION I. Definition of the roquet species group on the basis of BEHAVIOR Although each of the eight forms has a display which is quite distinct because of variation in cadence, all members of the species group have basically a very similar display. In the use of the dewlap, the roquet group differs from its near neighbors in the chrysolepis and bimaculatus groups. Thus, behavior provides fur- ther evidence that the group is a natural evolutionary unit. In all eight forms, the dewlap is extended and held out for the entire display sequence, and sometimes is not retracted at the end of the sequence. The dewlap is never pumped in and out in any rhythmic manner. This contrasts sharply with the other Lesser Antillean anoles. Figure 2 illustrates the challenge display of A . roquet extremus. In place of the activity bob. members of the binioculatus group seem to have a true assertion display sensu Carpenter. I have watched and filmed A. oculatus on Dominica pumping the dewlap in and out in a slow rhythmic manner — even when no recipient of the display was apparently present. In high intensity challenge, dewlap pumping is an integral part of a complex display which also involves bobbing and tail movements. A. lineatus (chrxsolepis series of beta anoles) of Curacao also pumps the dewlap. This species is quite unrelated to the bimacula- tus group (alpha anoles), and serves to illustrate the problem Ruibal (1967) encountered in a search for a major diagnostic character to split the genus by means of behavior. 12 BREVIORA No. 284 Figure 2. Challenge display of A. roijuet extrciniis. Dewlap fully ex- tended, tail up, and body high off the perch. 1968 DISPLAY BEHAVIOR OF SOME ANOLES 13 ^ o ■"^ -y: o x: ^ ^— o s ■♦-» a. _>< C^ '^ o T3 « C o '5 3 ■J <*- 0) _c o w on ;-i x) -. a Li y: "c. Ic J". s [— '~o ~ c -^ r— C 1) ^ o x: ^ ■♦— • ^ o o ^ . x: SB "Oh H _c y^ i "O c c3 ^ ^g ■5 o ■K C E o C^ r3 D « >. ou J= t. •— c o r3 aj C -a "^^ _o c :C H M c/i y: OJ s r<~i :S p C ^^ (U -.— 1 C -a *^ • ^ c >^ aj o ii^ T3 H yj o 14 BREVIORA No. 284 A. clirysolepis of Trinidad and the South American mainland displays by flashing the dewlap extremely rapidly and in a stereo- typed manner. The entire display appears to consist of dewlapping. The head is held high but is not bobbed. Figure 3 show.v clirysolepis in challenge positions. Because dewlapping is widespread in both the alpha and beta anoles, it is interpreted as a primitive character which has been secondarily lost in the roquet group. Further evidence for this will be presented below. IT. Description of the challenge display of A. luciae A. luciae is considered the most primitive member of the species group. It is thus worth analyzing its display in some detail, to serve as a basis for comparison. Descriptions of interactions 1. Forestiere. St. Lucia — 6 July 1965 Resident male is high in a tree. Tethered intruder placed four feet below him. R rises up on all limbs, extends dewlap, tail goes up once, then down. Bobs head. There are four distinct series of bobs while the dewlap is held out. Dewlap slowly retracted. 2. Vigie Beach. St. Lucia — 6 July 1965 Resident male is in a breadfruit tree. Intruder placed one foot below him. R immediately orients perpendicularly to intruder, head facing head. Resi- dent rises up high. Extends dewlap and bobs head. The head bobs are in four distinct series, with a partial dewlap retraction following each series. Intruder turns, and R turns, maintaining perpendicular orientation. Repeats display, again with four series on a single dewlap. The display of luciae may be characteriezd as distinct from other members of the roquet group by the following: 1 ) Tail — the tail is infrequently used in display sequence. When it is used, it is raised once slowly, then lowered. 2 ) Head bobs — luciae has a head bob sequence that is ex- tremely long in duration and that is clearly divided into several series. That is, a series of bobs is followed by a pause, followed by a series of bobs, followed by a pause, etc. (Figs. 4L and 5L ). 3) Partial dewlap retraction — during the long pauses, the dewlap is very slightly retracted, then re-extended. A. luciae is the only member of this species group to move the dewlap at all 1968 DISPLAY BEHAVIOR OF SOME ANGLES 15 in mid-sequence. This movement might be a vestige of the dewlap pumping common to most Anolis. The following quantitative data were analyzed: 1) Number of peaks preceding a pause. A pause was arbi- trarily defined as 6/16 of a second (N.B. all films were made at 16 frames/second, and time units discussed are in 16th of a second or frame) with no head movement. A peak was defined as the highest point the head reached before the next downward move- ment. Thus sub-peaks were not counted. The distribution in num- ber of peaks per bobbing series proved to be quite variable, with a range from 1-18. However, the mean is 5.2 and 54 per cent of the sequences had 4, 5 or 6 peaks. 2) The time interval of bobbing series was also measured. Again the range is very broad, 2-47 frames. The mean is 18. 3) The pause lengths between bobbing series were also meas- ured and the distribution of these time intervals is shown in Figure 7. Note that the first category is also the largest. This is probably an artifact — i.e. a pause was defined as 6 or more frames. This does not mean that it is necessarily a "real'" pause to the animal. If 10 or more frames were chosen as the arbitrary limit, there would be a gradual rise in frequency to the 30-33 frame class, and then a drop. No relationship could be demonstrated between the number of peaks in a sequence and the length of the following pause. Two other categories also easily lent themselves to quantitative analysis: number of bobbing series in a single display sequence, and total time elapsed in the full bobbing sequence. Because of the repetitious nature of the long displays and the high cost of motion picture film, few full sequences were filmed. Some examples are tabulated below: . bobbing series Total time Full sequence filmed 5 10.5 sec. no 3 8.5 yes 5 19.3 yes 5 12.5 yes 4 12.5 no 3 6.7 yes The total elapsed time for a full bobbing sequence in A. luciae is far greater than for any other member of the species group. 16 BREVIORA No. 284 1968 DISPLAY BEHAVIOR OF SOME ANOLES 17 Figures 4-6. Head bobbing sequences in roquet group AnoUs. Time intervals are one-half second between successive dots on upper and lower borders of each figure. Long pauses are indicated by a break, with the time interval in seconds written in. Figure 4 L. A. Incicie B. A . honiiirensis T. A . irini talis Figure 5 L. A. Iiicicic G. A. ricliardi griseiis RI. A. r. ricliardi Figure 6 E. A. roquet e.xtrenuis RO. A. r. roquet A. A. (iciieiis 18 BREVIORA No. 284 13 11 Vi ^ o 1— >x ^ ae UJ 8 8 bO K o 7 LLm o 6 6 OS LU ea ^s 13 4 3 6-9 1013 14-17 18-21 22-25 26-29 30-33 PAUSE LENGTHS (FRAMES) 34-37 > 37 Figure 7. The frequency distribution of pause lengths which precede subsequent bobbing series in A. Iiiciae. One frame is equivalent to one- sixteenth of a second. ITT. Presumed derivatives of A. luciae A. luciae, for reasons stated earlier, is presumed to be the primi- tive member of the species group. Following the model presented in the zoogeographic study of the roquet group, three forms are believed to be derived directly from luciae: double colonization of St. Vincent from St. Lucia gave rise to trinitatis and griseu.s, and colonization of Bonaire gave rise to bonaireusis. In this section we shall compare the displays of these three forms to those of A . luciae. a) A. bonairensis. The relationships of this species to the remainder of the species group are probably the least clear. Bonaire is far west of the Lesser Antillean chain (see Figure 1), but the anoles on the adja- cent islands just west of Bonaire are certainly unrelated to bonair- ensis. When the species was described (Ruthven, 1923), it was compared with roquet and aeneus. The osteological study of Etheridge placed this form very close to the other members of the 1968 DISPLAY BEHAVIOR OF SOME ANOLES 19 roquet group, and this was confirmed by blood proteins and chromosomes. In our paper on blood proteins (Gorman and Des- sauer, 1966), it wa* pointed out that although boiiaircusis was clearly a member of the roquet group, it was not particularly close to any one form. We speculated that it might be derived from ae- neus, since Bonaire is due west of Grenada'and the general current flow is westerly. However, when we examined chromosomes (Gorman and At- kins, 1967), we found that bonairensis belonged to the more primitive group (2n=36), making it highly improbable that it was derived from aeneus. The geographically intermediate form A. bonairensis blanquillanus has been unavailable for study. It is, of course, possible that Bonaire was colonized independ- ently by a mainland South American Anolis. However, that is not likely. Although the roquet group forms are clearly allied to South American alpha anoles, they share a unique complex of osteological and biochemical characters which most likely evolved in the Lesser Antilles. Thus our hypothetical ancestor would be a ///c/ac-like or trini- tatis-Wkt anole. J. D. Lazell, Jr., who is studying the external mor- phology of Lesser Antillean anoles, relates bonairensis to luciae (personal communication). The display of bonairensis is basically similar to other members of the species group. It may include tail lashes and head bobs. Description of an interaction 4 May 1964 Intruder placed on a tree. Resident male comes running across a fallen stump to stop 2 feet from intruder. He runs at least 15 feet. Head is in engorged posture. 1 jiggled intruder. R immediately extends dewlap fully, rises high on all four limbs, tail is raised once, dropped, body lowered. Bobs head and retracts dewlap. The head bob sequence of bonairensis is unique in the large number of peaks and the great length of time. The bobbing may or may not be broken up by pauses of 6 frames or longer. In 20 filmed sequences, 14 are without pauses. These sequences have a range of 12-23 peaks (mean 17.7) and a time range of 43-95 frames (mean 62.2). Of six sequences with pauses, five have only a single pause and one sequence has two pauses. This last case may be an artifact, for the first pause follows a single slow head rise, which is probably not part of the bobbing sequence. As opposed to luciae, which has pause lengths of as long as 40+ frames, the 20 BREVIORA No. 284 pauses of honairensis are all short, ranging from 6-10 (mean 8.3) frames. Those sequences which have pauses range in total peak number from 13-24 (mean 17.7), which is almost identical to the non- pause displays. The total time intervals range from 47-91 frames (mean 69.0), which is within the range of non-pause displays. Thus the head bobbing sequence of A. bonairensis is best described as a long single sequence with many peaks, occasionally broken into two series. The display of honairensis could be derived from the lucicie display with but one minor alteration: elimination of the pauses which break the bobbing sequence into distinct series. Bonairensis also uses tail lashing in high intensity displays. In this respect the species resembles other members of the roquet group more than luciae. b) A. trini talis. A. trinitatis is the widespread small anole on St. Vincent. It also occurs on Trinidad, where it was probably introduced by man. On Trinidad it hybridizes with aeneus. A detailed analysis of hybridiza- tion is being prepared. The data on trinitatis presented in this paper include St. Vincent and Trinidad specimens, as no discernible difference could be found in their displays. The challenge display of trinitatis consists of very rapid tail lashes followed by rapid bobs. Long pauses are infrequent. A slow tail movement has not been observed. Figure 8 compares the characteristic rapid tail movement of trinitatis with that of ex- tremiis, which is slow. Descriptions of encounters appeared in the introductory part of this paper. The variation in 30 rapid bob sequences of trinitatis is: Rapid bob time interval No. peaks per bobbing sequence range (sixteenths of a sec.) 12-31 2-10 mean 23.0 5.7 56% are 4. 5, or 6 A typical rapid head bob of trinitatis is presented in Figure 4T. The full bobbing sequence in trinitatis resembles a single series of bobs within a sequence in A. luciae. The display of trinitatis differs from that of luciae in the rapid tail lash and in the shorter, non-paused head-bob sequence. The latter, however, could easily be derived from a foreshortening of the luciae display. c) A. ricliardi griseus. Although this is the largest, it is paradoxically the shyest mem- ber of the roquet species group; that is, A. r. griseus is quick to 1968 DISPLAY BEHAVIOR OF SOME ANGLES 21 flee at one's approach. As a result, I have few data. No natural aggressive interactions were seen. Two males did display to tethered intruders, and their several bobbing sequences were recorded on film. These two males were filmed in the summers of 1964 and 1966, respectively, and on opposite ends of St. Vincent. The display action patterns are quite similar. Thus, although lacking details on the total range of variation display, the data are probably valid from the broadly comparative viewpoint. High intensity display consists of dewlap extension, rapid tail lashes, head bobs, and dewlap retraction. The head bobs are in series, as in /I. luciae. Figure 5G illustrates a typical bobbing sequence of griseus. I do not have enough pauses measured to compare in detail, but here too the distribution seems similar to luciae. Of 1 1 measured pauses, 5 are between 6-9 frames, 5 are 20 or more frames, and only one falls between these two intervals (compare with Figure 7). There does seem to be a reduced number of peaks per series; in 18 measured series the mean is 3.3, and 61 per cent of the sequences have 1, 2, or 3 peaks. In luciae the mean is 5.2, and 54 per cent of the series have 4, 5, or 6 peaks. However, the mean time interval per bobbing series ( 16.6 frames) is very close to that of luciae (18.0). This implies that the head bobs of griseus are broader (slower) than those of luciae. Such a trend can be seen by comparing Figures 5L and 5G. The total time of a bobbing sequence, adding all series and pauses, may also be shorter than in luciae. (Compare these data with those on p. 15.) No. bobbinij ; series Total time Ful 1 sequence filmed 2 5.0 sec. yes 2 4.8 yes 5 8.4 yes 4 7.3 yes In summary, the displays of griseus are quite similar to those of luciae. In head bobbing there seems to be a tendency for fewer peaks per series without a significant decrease in the time of the series, whereas the total time of a bobbing sequence is reduced. Also, griseus may use several vigorous tail lashes in the challenge display, whereas luciae does not use the tail frequently, and does not use multiple lashes when it does use the tail. 22 BRHVIORA No. 284 IV. Presumed derivatives of griseus and trinitatis a) A. r. richardi. A. r. richardi is very closely related to griseus. On the basis of blood proteins (Gorman and Dessauer, 1966), the two forms were considered conspecific and distinct from all other members of the species group. Unlike griseus, richardi is not shy, and numerous displays have been observed in Grenada and in Bequia, the northernmost of the Grenadines. No displays were studied in Tobago. Descriptions of encounters 1. St. George's, Grenada - 17 July 1965 A male resident is seven feet up in a mango tree. Intruder placed two feet helow. R immediately into engorged head posture watching intruder. Intruder runs up. R extends dewlap fully, rises up high on all limbs, tail up a bit, then down slowly. Again up and then down slowly, dewlap still out. Intruder removed. This is interpreted as a low intensity response. Male richardi may rise up slowly and the head is sometimes slowly nodded. The fol- lowing is an example of a high intensity response, the type of response that was analyzed. 2. Beausejour. Grenada - 17 July 1965 Intruder placed near resident male. Dewlap extended. No tail motion. Bobs vigorously about 7 times. Dewlap partially retracted. Intruder jiggled. R rises up, lashes tail four times, then a fifth time with lower amplitude. Dewlap is held out. R looking at intruder. Intruder removed. The high intensity display of richardi is characterized by vigorous tail lashing, followed by head bobbing. Unlike griseus, there are not multiple series, but rather a single series, in a bobbing sequence. Figure 5 RI shows an example of richardi head bobs. These bobs are broad, i.e. the rise and fall are relatively slow and the peak is flatter than in griseus. The display appears much more stereotyped than that of either hiciae or griseus, both in number of peaks (range 4-7. mean 5.8) and in total elapsed time in the bobbing sequence (range 39-58, mean 47.3 frames). Although there are only eight filmed displays, numerous other displays observed in the field seemed to conform to these ranges. b) A. aeneus, extremus and roquet. These three forms are very closely related and were considered conspecific by Underwood (1959), although he later changed his mind (1962). They were first studied biochemically by Gorman 1968 DISPLAY BEHAVIOR OF SOME ANGLES 23 and Dessauer (1965), who found that on the basis of electro- phoresis of serum proteins they were closely related, and a linear chain of relationship could be established: aeneiis and extrcmus differ in hemoglobin mobility, extremus and roquet in transferrin mobility, and twneus and roquet differ in both these characters. In a more detailed study of blood proteins encompassing the entire species group (Gorman and Dessauer, 1966), extremus was rec- ognized as a subspecies of roquet, and aeneus was recognized as a Figure 8. Comparison of the slow tail movement of A. roquet extreiniis (upper) with the rapid tail movement of A. trinitatis (lower). Numbers at the tip of the tail indicate frames elapsed from start of display (one frame equal to 1/16 second). A. Tail rise in extrcmus. The tail is slowly arched up. A full two seconds elapse before peak is reached. B. Tail lowering (same display). The tail is dropped more quickly than raised. Multiple dashes at 36 indicate tail was in rapid motion and film was blurred on that frame. Total time interval for sequence is 40 frames {IVi seconds). C. Tail rise in trinitatis. The peak is reached in 3 frames. D. Tail lowering (same display). The tail is down one frame later. Total time interval. 4 frames ( Vi second). The drawings were made by superimposing images from a single frame projector. 24 BREVIORA No. 284 species. The three forms are obviously very close to trinitatis. The chromosome study showed that trinitatis was a likely ancestral form, as it had the primitive chromosome number. In display behavior, there are great similarities among the three in gross movements. That is, the sequences of movements are identical, all have a variable tail movement, slow to moderately fast — but always slower than trinitatis (Figure 8) — and all bob the head. But here there are consistent differences. A. r. extremus has a head bobbing pattern that appears to be the most rigidly stereotyped. There are several rapid peaks and the display is over. There are never any long pauses. Figure 6E illus- trates an example of this head bob. Thirty-two head bobbing sequences were filmed, the number of peaks per sequence ranged from 2-5 (mean 4), and the time interval in frames ranged from 6-14 (mean 10.6). Although in blood proteins c.xtrenuis seemed to be intermediate in this series of three forms, in head bobbing, roquet and aeneus are more similar to each other than to extremus. Both forms have displays characterized by a long pause; thus the total time interval of the bobbing sequence is longer than that of extremus. A. r. roquet tends to have more peaks than does aeneus, while aeneus tends to have longer pauses and thus a longer bobbing sequence. However, there is considerable overlap, as tabulated: Total time of bob sequence Rapid bob No. peaks (sixteenths of a sec. ) no. range mean range mean aeneus (25) 2-4 2.8 27-62 44.6 roquet (28) 2-9 4.3 18-47 33.3 Figures 6A and RO compare rapid bobs in the two forms. The examples depicted accurately show the typical differences in the head bobbing sequences between the two forms. Not only are there fewer peaks and a longer pause in aeneus, but it also may be seen that in the typical display of aeneus a head rise follows the drop after the long pause, whereas in roquet, following the long pause the head is dropped and the bobbing sequence is terminated. A. trinitatis is the presumed ancestor of this group. Chromosome loss being rare, it is highly probable that trinitatis gave rise to only one of these forms. Two alternatives are plausible. 1. Colonization of Barbados, loss of chromosome pair, giving rise to extremus. A. r. extremus then sends out two radiations, one to Martinique where a transferrin difference evolves (roquet), the J968 DISPLAY BEHAVIOR OF SOME ANGLES 25 Other to Grenada where a new hemoglobin evolves {aeneits). The display of extremus is not very different from that of trinitatis. Both have head-bob sequences consisting of several peaks in a short time interval. The sequence of trinitatis is more variable, in both number of peaks and time interval and is occasionally punctuated by pauses. The similarity in display of roquet and aencus would have to be accounted for by convergence. 2. Colonization of Grenada, loss of a chromosome pair, giving rise to aeneus. Evolution of the aeneus type display (increase in tendency for long pause, decrease in number of peaks) is followed by colonization of Barbados {extremus), and from there coloniza- tion of Martinique {roquet). There would be minor change in the head-bob display of roquet, and evolution of a new head-bob display in extremus. There is one other bit of behavioral evidence that can be examined to assess these two alternatives. Members of the roquet group have a movement which I termed "rhythmic motion." This may consist of bobbing, or walking, or climbing in a very jerky manner. The animal will move, pause, move, pause, etc. The pause is long and the movement is very sudden and exaggerated. The actual context of the display is not fully understood by me, but it appears to occur at a somewhat lower intensity than the tail move- ments and the rapid head bob in male-male territorial encounters. This movement was seen most frequently in extremus and aeneus, but was also filmed occasionally in trinitatis and roquet. In A. r. extremus, in eleven different displays, 100 intervals- between-bobs were measured. The range in frames (one frame = 1/16 sec.) was 5-12 (5-9 in 97 per cent of the intervals) and the mean interval between bobs is 7.27. In A. trinitatis in four displays (from same individual) 30 inter- vals range from 7-1 I with a mean of 8.46 frames. In A. r. roquet only one sequence was filmed; there are 21 pauses ranging from 7-9 frames with a mean interval of 7.75. In A. aeneus in ten displays 103 intervals were measured ranging from 9-19 frames with a mean of 1 1.9. If this character has any meaning in terms of relationships, it implies that aeneus is the most divergent member of the series and that trinitatis is closer to extremus. This supports the first alternative above. Table 1 (p. 29) summarizes some of the data for the roquet group. 26 BREVIORA No. 284 DISCUSSION At the level of the species group, the display behavior is useful as an additional taxonomic character. The roquet group can be defined as distinct from its geographic near neighbors by the absence of dewlap pumping within a display sequence. Furthermore, dif- ferences can be delineated which distinguish the eight forms of the species group. Why stereotyped displays in island races of lizards? Why inter- island differences? Do they relate to isolating mechanisms? Among some well-established cases of vocalization in birds, island races may have much more variable, less stereotyped, and less complex calls than the corresponding mainland races (Marler, 1959). The assumption here is that there are fewer sympatric congeners, and that there is therefore reduced selection for specificity. The few studies on island lizards, however, do not show a similar pattern. The display of Anolis townsendi, a species restricted to Cocos Island some hundreds of miles west of the Central American mainland, is quite stereotyped (Carpenter, 1965). Furthermore, Carpenter ( 1966) found, in a study similar to this one, that 1 1 of 12 Galapagos islands have iguanids of the genus Tropidurus with species-specific (or island-specific) display patterns. Each island has only one Tropidurus. These differences were attributed to genetic drift. In the study by Ruibal ( 1967) on comparative anole behavior, island races were discussed, but the species studied were from Greater Antillean islands where each species might have a half dozen sympatric congeners at the same locality. Yet the displays described are no more stereotyped than the ones described here. It could be argued that some of the sympatric roquet group anoles evolved behavioral differences as additional isolating mech- anisms. For example, there was double colonization of St. Vincent from St. Lucia, which led to trinitatis and griseus. Perhaps some of the differences in the displays as seen today reflect selection for species-specific differences. Why are the displays no less stereotyped where there is only one form on an island? Probably the least variable display in the roquet group is that of extremus from Barbados, an island which has and probably has always had only one Anolis. I have attempted to emphasize that certain components in the display of each form characterize that form, yet there is some variation in the temporal sequence of displays. Since each island was colonized by a presumably small number of arrivals from 1968 DISPLAY BEHAVIOR OF SOME ANOLES 27 another island, population differences in display are not at all surprising. There is sampling error to begin with. Then, as the new population adjusts to the new island, there is reconstruction of the gene pool. Displays are almost certainly polygenic, virtually all genes have pleiotropic effects, and thus differences in challenge displays in anoles on isolated islands are the expected result. But selection, too, could be playing a very strong role. When we look at the display of A. chrysolepis and compare it with the roquet group, we find few similarities. A. chrysolepis hardly moves its body, but flashes its cherry-red dewlap quickly; in the roquet group the tail is waved, the head is bobbed, and there are vigorous push-ups. When we look at the ecology of these anoles, this behavior makes perfect sense. A. c/ir\solepis is a small brown forest anole living in deep shade. Body movements would not be conspicuous, the wind might be rustling leaves. But a sudden flash of red color appearing from "nowhere" is quite striking. Anoles in the roquet group are larger and display from more open sites. Thus head and body move- ments are conspicuous sign stimuli. At a subtler level, this line of reasoning could account for display differences within the roquet group. The islands are by no means identical. Bonaire, for example, is sparsely vegetated. St. Lucia has more characteristic lush tropical flora. Could differences in the configuration of the habitat act to select for display differences? In afl probability they could. Perhaps it is advantageous for luciae to repeat himseif by giving multiple head-bobbing series while a single long series suflRces in the more open terrain for A. bonairensis. In cases of sympatry in the Lesser Antilles, display differences may be very important in sexual selection, even though these dis- plays are "territorial," not "courtship," displays. In a study of two sympatric species of grackles, Selander and Ciller (1961) found that males hold mutually exclusive territories and solicit nesting females of either species. It is the females that must select con- specific mates based upon a display difference. Possibly, in AnoUs, females associate only with males giving the "proper" display. Hunsaker (1962) showed that female Sceloporus preferentially associated with mechanical models which gave the species-specific head-bob. Because of the importance of display in the social struc- ture of the population, even on a one-anole island we must expect limited variation, i.e. stereotypy. Despite differences in displays of all the forms considered, it should be emphasized that the basic display patterns are quite 28 BREVIORA No. 284 similar and tiiat this may be used as a valuable taxonomic charac- ter. Quite possibly, in the Greater Antilles where 20 or more species of Anolis may be found on a single island, there might be very strong selection for behavioral differences (including challenge display) as isolating mechanisms in closely related forms. This would render the patterns taxonomically useless, except to show that different species had different displays: it would mask, rather than elucidate patterns of relationship. In general, the behavioral data support my proposed phylogeny of the roquet group based upon chromosomes and proteins. The presumed prototype, A. luciae, has a display which could well be the basis for further elaborations and foreshortenings, leading to the displays of the other forms. It is interesting that luciae and griseus have very similar displays, which differ from ricluirdi. An entirely different approach, marshalling biochemical evidence, has led me to the conclusion that ric/hirdi had to be derived from griseus and not vice-versa (Gorman and Atkins, 1968). The relationship between luciae and bonairensis is less clear. If blanquillanus, the presumed intermediate, had been available for study, this perhaps could have been clarified. Relating trinitatis and luciae closely is not a novel idea. Under- wood (1962) considered them conspecific, although his reasons were not particularly clear. The two forms differ considerably in blood proteins; however, the similarity in karyotype and certain similarities in display behavior tend to support the idea of close relationship. To compare the different island forms I have presented the data in an oversimplified manner. That is, I implied that a display once established is static. Thus, I call luciae ancestral to various other forms and compare the displays of today's luciae to the presumed derived forms. Obviously, the displays of all forms living in isola- tion are subject to change. The minor use of the tail in luciae may well be a secondary loss rather than a primitive character that was later elaborated in all the other members of the species group. Despite the oversimplification, however, comparative behavioral data when combined with other information are valuable in tracing the evolution and relationships of the roquet group. ACKNOWLEDGMENTS I am indebted to Dr. Garth Underwood, formerly of the Uni- versity of the West Indies in Trinidad, for sparking my interest in West Indian Anolis, and for providing space and facilities in the 1968 DISPLAY BEHAVIOR OF SOME ANOLES 29 Zoology Department of the University of the West Indies. I have been aided by the kindnesses of many people living in the West Indies, in particular: Dr. J. L. Price, Dr. J. Kenny, Mr. H. Boos, and Mr. J. Boos of Trinidad; Dr. J. L. Groome of Grenada; Mr. and Mrs. R. Milligan of Barbados; Dr. E. Kirby and Mr. F. Prescod of St. Vincent; Mr. E. Long of St. Lucia; and Dr. A. Krumperman of Bonaire. Field assistance has been provided by Miss C. Raybin. Miss A. Swidler, Mr. A. R. Kiester, and Mr. R. Ross. Illustrations used in this paper were prepared by Mrs. P. Kerfoot. In 1963-1964 I held a travelling fellowship from the University of California. Further field work has been supported by NSF Grant GB-2444 to Dr. E. E. Williams, and grants from the Explorers Club, Sigma Xi, and the Evolutionary Biology Committee of Harvard University. In addition to my travelling fellowship, my graduate studies have been supported by an NSF Predoctoral Fellowship (1964-1966) and a Turtox Fellowship (1966-67). Dr. E. O. Wilson provided valuable criticism of the manuscript. The frequent consultations with Dr. E. E. Williams, and his con- stant interest in this study, have been instrumental in its fruition. TABLE 1 Head bob patterns of roquet group anoles Time Interval in 16thsof Sec. p eaks, /Bobbing Sequence No. filmed Form sequences Range Mean liiciae 1 83 1-18 5.2 honairensis - 14 12-23 17.7 honairensis'-^ 6 13-24 17.7 trinitatis 30 2-10 5.7 griseits i 18 1-8 3.3 ricluirdi 8 4-7 5-8 extrcnuis 32 2-5 4.0 aeiieiis 25 2-4 2.8 roquet 28 2-9 4.3 Range Mean 2-47 18 43-95 62.2 47-91 69.0 12-31 23.0 1-41 16.6 39-58 47.3 6-14 10.6 27-62 44.6 18-47 33.3 1 In liiciae and gri\c'tis. a head bobbing sequence consists of several series of bobs, with pauses between the series. These data are for the individual series. - In bomiirensis there are occasionally paused displays, as in hiciae and griseiis. Here are tabulated non-paused displays. •■■ This tabulates individual series of the paused displays. 30 BREVIORA No. 284 LITERATURE CITED Carpenter, C. C. 1962. Patterns of behavior in two Oklahoma lizards. Amer. Midi. Natur. 67: 132-151. 1965. The display of the Cocos Island anole. Herpetologica 21: 256-260. 1966. Comparative behavior of the Galapagos lava lizards (Tropi- diinis). In: The Galapagos. Proceedings of the Galapagos International Scientific Project, R. I. Bowman, editor (pp. 269- 273). Berkeley and Los Angeles, Univ. of California Press. Darwin, Charles 1872. The Expression of Emotions in Man and Animals. London, Murray, 374 pp. Etheridge, R. 1960. The relationships of the anoles (Reptilia:Sauria:Iguanidae). An interpretation based on skeletal morphology. Ph.D. thesis. University of Michigan. (University Microfilms. Inc., Ann Arbor.) Evans, L. T. 1938. Cuban field studies of territorial behavior of Anulis sagiei J. Comp. Psychology 25: 97-125. Gorman, G. C. 1965. Interspecific karyotypic variation as a systematic character in the genus AnoUs (SauriaTguanidae ). Nature (London) 208: 95-97. Gorman, G. C. and L. Atkins 1967. The relationships of Anulis of the loqiict species group (Sauria: Iguanidae). IL Comparative chromosome cytology. Systematic Zoology 16: 137-143. 1968. The zoogeography of Lesser Antillean Anolis (Sauria:lguani- dae). An interpretation based on chromosomes and lactic dehydrogenases. In press. Gorman, G. C, L. Atkins, and T. Holzinger 1967. New karyotypic data on 15 genera of lizards in the family Iguanidae, with a discussion of taxonomic and cytological im- plications. Cytogenetics 6: 286-299. Gorman, G. C. and H. C. Dessauer 1965. Hemoglobin and transferrin electrophoresis and relationships of island populations of Anolis lizards. Science 150: 1454-1455. 1968 DISPLAY BEHAVIOR OF SOME ANOLES 31 1966. The relationships of AnoUs of the roquet species group (Sauria: Iguanidae). I. Electrophoretic comparison of blood proteins. Comp. Biochem. and Physiol. 19: 845-853. Grhenbhrg. B. and G. K. Noble 1944. Social behavior of the American chameleon {Aiioli.s caroliiicn- sis). Physiol. Zool. 17: 392-439. HUNSAKER, D. 1962. Ethological isolating mechanisms in the Sceloporus toiquatiis group of lizards. Evolution 16: 62-74. Kastle, W. 1963. Zur Ethologie des Grasanolis (Noyops aiiratiis). Z. Tierpsych. 20(1): 16-33. Marler, p. 1959. Developments in the study of animal communication. //;. Darwin's Biological Work, P. R. Bell, editor, (pp. 150-206). Cambridge, Cambridge University Press. RUIBAL, RODOLFO 1967. Evolution and behavior in West Indian anoles. //;.• Lizard Ecology, A Symposium, W. W. Milstead. editor, pp. 1 16-140. Univ. of Missouri Press, Columbia. RUTHVHN, A. G. 1923. The reptiles of the Dutch Leeward Islands. Occ. Pap. Mus. Zool. Univ. Michigan No. 143: 1-10. Selander, R. K. and D. R. Giller 1961. Analysis of sympatry of the great-tailed and boat-tailed grackles. Condor 63: 29-86. Underwood, G. 1959. Anoles of the eastern Caribbean. Part 111. Revisionary notes. Bull. Mus. Comp. Zool., 121: 191-226. 1962. Reptiles of the eastern Caribbean. Caribbean Affairs, N.S. 1: 1-192. (Received 19 July 1967.) E V I Mmsemiii of Comparative Zoology Cambridge, Mass. 5 April, 1968 Number 285 THE GENERA OF PUFFBIRDS (BUCCONIDAE) G. W. Cottrell In 1882, in the last part of A monograph of the jacamars and puff-birds, pages xxxiv-xli, P. L. Sclater divided the family of puffbirds (Bucconidae) into seven genera: Bitcco, Malacoptila, Micromonacha, Nonnula, Hapaloptila, Monacha (Sclater's emenda- tion of Monasa Vieillot 1816), and Chelidoptera. The large genus Bucco, containing twenty of the forty-three species enum- erated for the family, he grouped in two subgenera (following VVagler 1827, Sy sterna avhmi): Bucco, "rostro rotundato ad basin dilatato" (fifteen species), and Nystalus, "rostro compresso et magis elevato" (five species). He saw no grounds for division into subfamilies. By 1914 the trend of the times had swelled the number of genera to thirteen, as set forth by Ridgway in The buds of North and Middle America 6: 371-373, the increment all taking place within the limits of Sclater's genus Bucco. The roster in Ridgway runs Notharchus,^ Argicus, Nystactes, Bucco, Hypnelus, Ecchaunornis, 1 Misprinted "NortharcJius" in the key to the genera, "Notharchiis" else- where. This name, with two of its trivial adjuncts, has had an ill-starred orthographic career. In 1958 Eisenmann, Auk 75: 101, called attention to the misprinting in Peters 1948, Check-list 6: 10, of the subspecific hyperrhynchus (Sclater 1856) as "hypeirynchus." And Ridgway's "Nortlwr- chiis" has just been noted. But there is worse to come. In recent decades (first observed in Griscom 1932, The distribution of bird-life in Guatemala), Notharchus has frequently been printed "Notliarcus," contrary to the original publication, with indicated derivation, of Cabanis and Heine 1863, Museum Heineaniim 4, 1: 149 {Northarchus formed from vwdrjs, sluggish, slothful, and dpxos, leader, chief — though the latter incorrectly accented apxos by Cabanis and Heine). Of less consequence than the generic misspelling is the shift (dating back at least to Wagler 1827, Sy sterna avium) of the ending of the specific name macrorhynchos as first published by Gmelin 1788, Systema naturae 1, 1: 406, from -os to -us, latterly perhaps by attraction to Sclater's hyperrhynchus, originally a specific designation but a subspecies of macrorhynchos since 1939 (Sassi, Temminckia 4: 161). 2 BREVIORA No. 285 Nystalus, Malacoptila, Micronwnacha, Hapaloptila, Monasa, Non- nula, Chelidoptera. The first seven genera (comprising Sclater's Biicco) are grouped as the subfamily Bucconinae, the remaining six as the subfamily Malacoptilinae. Cory 1919, Catalogue of birds of the Americas 2, 2: 390-410, has the same thirteen genera as Ridgway, but without division into subfamilies. Thirty years later the emphasis in classification had shifted from dissimilarity toward similarity: Peters 1948, Check-list 6: 10-24, reduces the seven genera equated by Ridgway with Sclater's Bucco to four, Argicus and Nystactes being merged with Bucco, and Ecchaunornis with Nystalus. The other six genera remain un- changed. No subfamilies or subgenera are recognized. Peters' arrangement has been regularly followed down to the present (e.g., de Schauensee 1966, The species of birds of South America). A new review of the family has led to the conclusion that a full return to Sclater's generic system should be made by merging Notharchus, Nystalus, and Hypnelus, maintained by Peters, with Bucco.'- Since the Check-list does not justify its systematics, one must revert to Ridgway for criteria of differentiation. Ridgway follows Sclater in dividing his subfamily Bucconinae (Sclater's genus Bucco) into two major groups: one with thick bill, "its width at nostrils nearly to quite equal to its depth at same point" (Sclater's subgenus Bucco), the other with compressed bill, "its width at nostrils much less than its depth at same point" (Sclater's subgenus Nystalus). In the thick-billed group he included his genera Notharchus, Argicus, Nystactes, and Bucco, in the com- pressed-billed group the genera Hypnelus, Ecchaunornis, and Nystalus, paralleling Sclater's division except with respect to the species included in Hypnelus, which were placed by Sclater in the thick-billed group. An actual check of specimens shows the amount of compression to be only very slightly greater in Sclater's subgenus Nystalus (most in maculatus) . Lesser criteria of differentiation in Ridgway relate to variations in bill shape and color, tarsus length, length of rictal bristles, length of remiges or rectrices, plumage markings, as set out in his 2 This same merging of genera was proposed and followed by Charles G. Sibley in a seminar syllabus on the Piciformes, Cornell University, ca. 1956, but was not published. Its actual priority is hereby duly recorded. 1968 GENERA OF PUFFBIRDS 3 key."^ Thus, Notharchus is distinguished from Argicus, Nystactes, and Biicco by tarsus shorter than outer anterior toe without claw, upper parts mostly plain blackish, forehead sometimes white, as opposed to tarsus equal to or longer than outer anterior toe with- out claw, upper parts mostly brown or rufous-tawny, forehead never white. We have seen that Peters included Argicus and Nystactes in Bucco. Within his compressed-billed group, Ridgway distinguishes Hypneliis from Ecchaunornis and Nystalus by a strongly bifid maxilla (see below), longer rictal bristles, longest primaries decidedly longer than secondaries, indistinct spotting of upper parts, black pectoral band, plain (unbarred) tail, wholly black (not red. orange, or horn) bill. Peters merged Ecchaunornis with Nystalus. Such morphological criteria, as employed by Ridgway for his seven genera, and tacitly maintained by Peters for the four genera recognized by him, do not seem adequate for separation at the generic level. With respect to ethology, there are no recorded aspects that militate against the proposed merging — though information is in general scanty and fragmentary (also repetitious), and in some species almost totally lacking. The genus Bucco, now re-extended to Sclater's limits (and equivalent to Ridgway's subfamily Bucconinae), is marked by a relatively stout and straight bill, much broadened at the base, with strongly uncinate tip to the maxilla and terminally ascending more or less convex gonys. A striking feature, found to greater or less degree in all but one of the species, is the bifid tip to the maxilla. An examination of specimens shows that this peculiar process, part of the rhamphotheca and not of the underlying bone, is particularly prominent in tamatia and ruficollis (including bicinctus, Phelps and Phelps 1958, Proc. Biol. Soc. Wash. 71: 122), pronounced in macrorhynchos, pectoralis, ordii, tectus, noanamae, striolatus. ^ Ridgway 's key cannot be relied on even within its own frame of refer- ence. There is, for example, no subdivision "bb" to match "b" under "aa," which renders subsequent allocations indeterminate. The genus Hapaloptila appears under the subdivision "aa," characterized as having the tip of the maxilla not uncinate, whereas in fact the tip in Hapaloptila is strongly hooked. According to the key the bill in Monasa, Nonniila, and Chelidoptera is red, orange, or yellow, true in reality of Monasa only. Again, Chelidoptera is keyed as not "tawny" beneath, though actually with cinnamon-chestnut belly. 4 BREVIORA No. 285 maculatus, fairly slight in capensis, slight in radiatus, very slight in chacuru, and not distinguishable at all in macrodactylus. In a study of this bifid process, Boetticher 1936, Festschrift ziim 60. Geburtstage von Professor Dr. Embrik Strand 1 : 74-11, suggested that it might serve as a sort of vise for holding the heads of cap- tured flying insects (especially butterflies) while the wings were sheared off by the mutually adapted tomia. He admitted that there were no specific observational data to support his hypothesis; nor have any been reported since. The treatment of this character in Ridgway is inconsistent and in part inaccurate (notably with regard to extralimital taxa). In the key there is the general statement, applicable to all seven genera, that the tip of the maxilla is sometimes cleft medially, but no discrimination between individual genera (and hence species) until the compressed-billed group, comprising Hypnelus, Ecchaunornis, and Nystcdus, is reached. Here Hypnelus (consist- ing of the currently recognized ruficollis) is correctly signalized as having the tip of the maxilla strongly bifid, and Ecdmunornis (consisting of radiatus and chacuru) equally correctly as having the tip not distinctly if at all bifid. But Nystahis (consisting of striolatus and maculatus) is united in this respect with Ecchau- nornis, whereas in actuality the medial cleft is pronounced in its component species. The subsequent generic description for Nothar- chus (consisting of macrorhynchos, pectoralis, ordii, and tectus) includes a more or less distinctly bifid tip to the maxifla, but there is no later indication as to the situation in Argicus {macrodacty- lus), Nystactes (tamatia), or Bucco (noanamae, capensis), de- scriptions being omitted for extralimital genera. A review of the remaining six bucconine genera as found in Sclater, Ridgway, and Peters reconfirms their distinctness. They all have a less stout, more compressed, less basally broadened bill than the reconstituted Bucco, with the gonys more or less de- curved terminally and, except in Hapaloptila, the maxifla more gradually decurved and non-uncinate. In none of the six does the bifid maxflla occur. Differences in size, in shape and proportions of bill, wings, and tail, and in plumage patterns serve adequately to separate one from another, as described and fllustrated by Sclater, and hence are not detaUed here. Certain ethological fea- tures reinforce the morphological distinctions, as may be seen in the social behavior of Monasa and the mode of flight and flocking of Chelidoptera, this latter being in many respects the most divergent of the taxa included within the family. 1968 GENERA OF PUFFBIRDS 5 Consequent upon the merging proposed above, the genera and species of the Bucconidae stand as follows (order from Peters, species as in Peters except as indicated, subspecies not shown) : Biicco macrorhynchos, pectoralis, ordii, tectus, macrodactylus, tamatia, nocmamae, capensis, radiatus, chacuru, striolatus, maculatiis, ruficollis (includes bicinctus of Peters, following Phelps and Phelps 1958, Proc. Biol Soc. Wash. 71: 122) Malacoptila striata, jusca, semicincta (subspecies of fusca in Peters; distinct species here, following Traylor 1951, Fieldi- ana, Zool. 31: 614), julvogularis, ruja, panamensis, mystacalis Micromonacha lanceolata Nonnula rubecula, sclateri, brunnea, mficapilla (includes jron- talis of Peters, following de Schauensee 1946, Not. Nat. 163: 5), amaurocephala Hapaloptila castanea Monasa atra, nigrifrons, morphoeus, flavirostris Chelidoptera tenebrosa. (Received 26 July 1967.) BREVIORA MiiseiMii of Comparsitive Zoology Cambridge, Mass. 5 April, 1968 Number 286 THE STATUS OF NESOGALAXIAS NEOCALEDONiCUS (WEBER AND DE BEAUFORT) (PISCES, GALAXIIDAE) R. M. McDowall, Museum of Comparative Zoology, Harvard University^ Fisheries Research Division, New Zealand Marine Department- ABSTRACT The validity of the genus Nesogalaxias, proposed by Whitley (1935) for Galaxias neocaledonicus Weber and de Beaufort (1913), is supported by its very reduced pleural ribs, a character which has not been observed in other galaxiids. N. neocaledonicus occurs in mountain lakes in southern New Caledonia and is the most tropical species in the family. It is thought to have dispersed from Australia to New Caledonia when temperatures were lower during the Pliocene and Pleistocene, and the post-Pleistocene warming seems to have confined it to a few mountain lakes. Hither- to recorded only from Lac en Huit, N. neocaledonicus is now also known to occur in Grand Lac. INTRODUCTION In 1913, Weber and de Beaufort described a small fish species in the genus Galaxias from a lake in the mountains of southern New Caledonia. Several workers, e.g. Whitley (1935) and Scott (1936, 1966), have listed this species or commented on its generic position, but none appears to have re-examined it thoroughly. Since its generic status is currently uncertain, a re-examination of the species is necessary. 1 Present address. - On leave. Fisheries Research PubUcation No. 105. 2 BREVIORA No. 286 In 1964, Dr. P. H. J. Castle collected a small sample of this species from Lac en Huit, the type locaHty, and more recently (1966), Mr. W. Skrzynski collected further samples from this lake, and also from Grand Lac, apparently for the first time. These samples have been generously made available to me, en- abling an assessment of the affinities of this species to be made. GENERIC STATUS Although Whitley (1935) proposed a new genus for G. neo- caledonicus, he supplied no generic diagnosis, merely stating that "it is not a true galaxiid and may receive the new generic name Nesogalaxias." Scott (1966) found that "no feature in the origi- nal description would seem to preclude attribution to Galaxias" but apparently not having specimens, he took no further action. Nesogalaxias is correctly included in the family Galaxiidae, its lack of scales, fin positions, lateral line pores on the head, dentition, lack of a mesocoracoid, and jaw structure all conforming to the general galaxiid pattern. N. neocaledonicus was found to come within the range of variation of members of the family in New Zealand, in most of the characters examined. Superficially it difters little from some of the more generalized species of Galaxias. Stokell (1945) revised the generic classification of the Galaxiidae. He suggested that the very long head of the specimen figured by Whitley (1935) indicated deformity and classed the genus as "not recognised." However, Stokell's (1945) conclusion that the very long head in Whitley's (1935) figure is a deformity is incorrect, as the species does have a very long head (see Table 1). The pelvic fins have more posterior insertion than in most galaxiids, but this is a result of the great head length contained in the pre-pelvic dimension. In ratios of both standard length/head length and standard length/pre-pelvic length N. neocaledonicus represents an extreme for the family, but the range of values overlaps the ranges for these ratios in other species, e.g. Galaxias argenteus (Gmelin). Vertebral number (41-43) is very low, but not as low as in G. zebratus (Castelnau), which Jubb (1965) reported as having 39-41 vertebrae. In most osteological characters, N. neocaledonicus is conven- tional for the family Galaxiidae. The caudal skeleton shows the 1968 STATUS OF NESOGALAXIAS irregular variability which is characteristic for the family, and is not distinctive. The pectoral girdle lacks the postcleithrum, but this bone has disappeared in several other unrelated species, e.g. G. divergens Stokell, G. paucispondylm Stokell. However, in the absence of pleural ribs from those vertebrae behind the pelvic fins, N. neocaledonicus is different from all other galaxiids examined. These bones are present on vertebrae between the head and about the level of the pelvic fins, but thereafter disappear quite suddenly from the remainder of the abdominal vertebrae. The ribs present anteriorly, although short, are well developed, appear to be strongly ossified and have well developed heads articulating with the para- pophyses of the vertebrae. There are no epipleural ribs. The habits of A', neocaledonicus are somewhat different from other galaxiids (see p. 5) and this species seems to have entered a somewhat unusual type of niche for the family. These characters suggest that N. neocaledonicus stands apart from the rest of the family Galaxiidae and is best regarded as generically distinct from the more generalized species. The name Nesogalaxias Whitley (1935) has priority and is available. Nesogalaxias Whitley, 1935 Type species: Galaxias neocaledonicus Weber and de Beaufort, 1913, by original designation. Diagnosis: Trunk cylindrical to a little compressed, naked, dorsal fin posterior, above vent. Pelvic fins present, usually seven rayed. Caudal fin forked with 16 principal rays. Teeth uniserial on jaws and entopterygoids, biserial on basihyal. Postcleithrum absent. Pleural ribs lacking from abdominal vertebrae behind the pelvic fins. Figure 1. Nesogalaxias neocaledonicus (Weber and de Beaufort, 1913). Lac en Huit, New Caledonia. Length to caudal fork 52 mm. 4 BREVIORA No. 286 Nesogalaxias neocaledonicus (Weber and de Beaufort, 1913) Galaxias neocaledoniciis 'Weher and de Beaufort, 1913: 173. Nesogalaxias neocaledonicus, Whitley, 1935: 42; Scott, 1936: 85, 1966: 250. Syn types: Zoology Museum, Amsterdam, Holland, no. ZMA 104, 700, 2 specimens (seen). Museum National d'Histoire Natu- relle, Paris, France, nos. 2312-3 (Bertin and Esteve, 1951, not seen). Type locality: Recorded as "Lac en 8, sudliches Neu-Kale- donien" by Weber and de Beaufort. The locality for the paratypes in the Museum National d'Histoire Naturelle is listed as "Lac en Huit" by Bertin and Esteve (1951). Description (Fig. 1, Table 1): A small, rather slender species, trunk deep anteriorly but tapering to a very slender caudal peduncle which is much longer than deep. Greatest body depth somewhat behind pectoral fin bases. Head very long, slender, flattened ventrally, tapering rather abruptly forward dorsally to a slender, long snout. Eye large, projecting above head profile, interorbital concave, very narrow. Jaws sub-equal, long, extending to about anterior third of eye, gape narrow. Profile of lower jaw from ventral aspect deep and U-shaped. Canine teeth weakly developed in jaws; three to four moderately developed entopterygoidal teeth. Gill rakers strongly developed; two weak pyloric caeca. Dorsal and anal fins membranous and short based, rather high. Anal origin a little behind a line through dorsal origin. Pectoral fin inserted rather low, long in interval between pectoral and pelvic fin bases. Pelvic fin long in interval between pelvic base and anal origin, fin tapering to a point medially with middle rays longest. Caudal fin forked, rather short, fin tips rounded; flanges of caudal peduncle rather poorly developed. Size: Weber and de Beaufort listed a maximum size of 76 mm in their sample. Amongst fishes examined by the writer, the largest was only 56 mm long, and most adults and sub-adults were 40-50 mm long. Coloration: Trunk a uniform grayish brown, darkening dorsally and on head; ventrally paler, creamish gray. Pigmentation extends a little on to bases of fins, especially the caudal. Meristic variation: Dorsal fin rays (segmented) 8 (6), 9 (12) 10 (2); anal rays 9 (2), 10 (14), 11 (4); caudal rays (principal) 15 (1), 16 (19); pectoral rays 11 (5), 12 (14), 13 (1); pelvic rays 6 (1), 7 (19); branchiostegals 6 (7), 7 (13); vertebrae 41 1968 STATUS OF NESOGALAXIAS 5 (3), 42 (16), 43 (3); gUl rakers 3-11 (1), 3-12 (0), 3-13 (3), 4-11 (1), 4-12 (6), 4-13 (4), 5-12 (1), 5-13 (3), 6-13 (1). DISTRIBUTION AND HABITAT N. neocaledonicus is known only from two small lakes in the Plaine des Lacs, New Caledonia (Fig. 2) — Lac en Huit and Grand Lac. Skrzynski (personal communication) has kindly sup- plied the following notes on the lakes and the habitats in which he found this species. The Plaine des Lacs is about 100 km from Noumea. There are two main lakes, Lac en Huit and Grand Lac, each at an altitude of about 250 m, and a few square kilometers in area. They are about 1 Vi km apart and separated by a ridge, their outflows merg- ing a few kilometers below the lakes. Both lakes are depressions in an iron pan, their bottoms being solid rock with loose rock and soil in some places. The only aquatic plant growth observed consisted of loose patches of reeds in a few areas and encrusting algae. Neither of these provide much cover for fishes and the only cover in which N. neocaledonicus was ob- served was loose rock on the lake bed and crevices in the bed itself. The fish were found to conceal themselves in this cover during the day, but at night were commonly observed swimming in mid-water. Small unpigmented juveniles about a centimeter long were found to shoal freely in mid-water during the day. RELATIONSHIPS AND ORIGIN N. neocaledonicus is the most tropical species generally recog- nized as a galaxiid. G. indicus Day, from India, has not been re- collected since it was first described in 1888, and is doubtfully a galaxiid. Apart from A', neocaledonicus and the doubtful G. indi- cus, fishes of this family are confined to the southern temperate zone and, like the Northern Hemisphere salmonoids, are primarily cold water fishes. The fact that N. neocaledonicus has found refuge in the mountain lakes of New Caledonia suggests that it too pre- fers or is limited to cold waters. If this is so, it seems probable that it reached New Caledonia at a time of cooling, most likely in the Pliocene-Pleistocene, when lowered temperatures would have facilitated dispersal to New Caledonia. With subsequent post- glacial warming occurring, the populations have become confined to the cooler mountain lakes. BREVIORA No. 286 E in > a '2 o u U 'i f' mmiux Figures. 14-17. Miograinmopes lariindiis sp. nov. Fig. 14. Carapace, dorsal view. Fig. 15. Maxillae and lip. Fig. 16. Sternum, seen from be- low. Fig. 17. Epigynum, seen from below. Fig. 18. Miagrammopes latens Bryant. Epigynum, seen from below. 1968 MIAGRAMMOPES IN PANAMA AND THE WEST INDIES 11 one another by nearly nine times their diameter; separated from LE by about 1.3 times their diameter; with black pigment nearly cover- ing the ocular tubercles. Chelicerae: apparently typical of females of the genus. Maxillae, lip and sternum with features as shown in Figures 15-16. Legs: tibial index of first leg 10, of fourth leg 9; calamistrum as usual in females of the genus; fourth metatarsus and tarsus with the short, stout spines similar to those figured for M. simus Chamberlin and Ivie; first leg nearly twice as long as second leg and nearly three times as long as third leg. Abdomen without a dorsal, medial swelling such as frequently occurs in females of the genus; otherwise typical of females of the genus and without note- worthy features. Epigynum: obscurely distinctive; observable features shown in Figure 17. Color in alcohol: carapace with most of dorsum light yellowish brown but with numerous gray streaks and spots; lateral areas gray with considerable black pigment in ocular regions; sternum light grayish with a yellowish median stripe; mouth parts generally yellowish with a few grayish spots; legs generally brownish. Abdomen with a conglomerate color pat- tern; anterior two-thirds of dorsum yellowish, irregularly reticulated and with two pairs of irregularly shaped darker, oblique spots; pos- terior end of abdomen irregularly brown; lateral sides with several long, narrow, alternating, irregular light and dark lines; venter brownish with a series of stripes leading from genital groove to near the posterior end; there is a narrow, indefinhe, brownish, central stripe, a pair of pale, narrow stripes, a pair of broader brownish stripes, and a pair of light lines most lateral of all. Records. A female collected on May 5, 1964, in the type local- ity appears to be a paratype; an immature male and an immature female also collected in the type locality at about the same time probably belong to this species. MiAGRAMMOPES LATENS Bryant Figure 18 Miagrammopes latens Bryant, 1936: 326, pi. 23, fig. 4. The holotype male is in the Museum of Comparative Zoology; it was taken in Cuba, Sierra del Cobre, Loma del Gato. Bryant, 1940: 330; 1948: 393; Roewer, 1954: 1350; Bonnet, 1957: 2831. The holotype male is from the mountains of Cuba; the palp is still in good condition and well represented by Miss Bryant's plate 23, figure 4. I have not been able to find the specimen regarded by Miss Bryant as the allotype but I have studied a female from the same locahty and, apparently, taken with the allotype. My Figure 12 BREVIORA No. 289 18 is taken from this specimen. I see the epigynum somewhat dif- ferently than represented by Miss Bryant's figure 98 (1940) ; I have noted, however, considerable variation in the appearance of this organ among available specimens assigned to this species. The matter will probably not be clarified until numerous specimens are available for study following extensive collecting in the regions where the species occurs. MlAGRAMMOPES LICINUS Sp. nOV. Figures 19-23 Holotype. The male is from El Volcan, Panama, August 14, 1950. The name of the species is an arbitrary combination of letters. Description. Total length from anterior border of slightly porrect chelicerae to posterior end of spinnerets 3.32 mm. Carapace 1.19 mm long; 0.84 mm wide at level of ocular tubercles; 0.48 mm tall; ocular tubercles quite well delineated by grooves; with a fairly well defined median thoracic fovea; with a pair of depressions shortly behind ME; with a fairly well developed coat of hair which appears whitish in the light used for description (Fig. 19). Eyes: four as usual, in a gently recurved row as seen from above; ME separated from one another by seven times their diameter, from LE by slight- ly more than four times their diameter; ratio of eyes ME : LE = 5 : 7. Chelicerae, maxillae and lip: all apparently typical of males of the genus. Sternum: extremely convex in the middle; fourth coxae contiguous. Legs: tibial index of first leg 12, of fourth leg 13; first tibia with few spines (Fig. 20) and with different numbers on right and left; first metatarsus with spines as shown in Figure 20; first leg slightly more than twice as long as second leg and nearly three times as long as third leg; the calamistrum is lacking but the fourth metatarsus and tarsus bear several robust, ventral spines similar to those which occur on females; tarsal claws as usual. Palp: general features shown in Figures 21-23; the tibia is very distinctive with its dorsal expansion and concavity; copiously supplied with slender spines; tarsus complicated with its parts difficult to dis- tinguish. Abdomen: without special features; apparently typical of males of the genus. Color in alcohol; carapace with a yellowish. median, longitudinal stripe about one-third as wide as this part of the body; this yellowish stripe contains a narrow, grayish stripe which widens near the middle into an irregular diamond-shaped figure; a broad dark gray stripe covers the remainder of the cara- pace on each side; sternum yellowish; abdomen with a fairly broad, 1968 MIAGRAMMOPES IN PANAMA AND THE WEST INDIES 13 grayish, central, dorsal stripe indented along two-thirds of its length and then irregularly broken into a series of five or six somewhat separate spots; on each side of this central stripe there is a yellowish stripe; lateral sides of abdomen irregularly dark grayish; the venter has a grayish central stripe and on each side a narrow, light yellow- ish stripe and a narrow, darker stripe; all five of these ventral stripes are quite inconspicuous; legs in general yellowish; first pair with considerable dark gray coloration on dorsal and prolateral surfaces. Records. The female is unknown and there are no paratypes. Figures 19-23. Miagrammopes licinus sp. nov. Fig. 19. Carapace, dorsal view. Fig. 20. Right first tibia and metatarsus; retrolateral view. Figs. 21-22. Left palp; prolateral and retrolateral views, respectively. Fig. 23. Left palpal tibia; dorsal view. 14 BREVIORA No. 289 MlAGRAMMOPES MOLITUS Sp. nOV. Figures 24-28 Holotype. The male is from Jamaica, W. I., Portland Parish, Hardwar Gap, October 2, 1957. The name of the species is an arbitrary combination of letters. Description. Total length about 3.15 mm (body curled closely). Carapace 1.17 mm long; 0.81 mm wide at level of lateral ocular tubercles; about 0.44 mm tall at level of median eyes where it is tallest; with median thoracic fovea barely indicated; with grooves in ocular region very obscure (Fig. 24). Eyes: four as usual with row gently recurved; ratio of eyes ME : LE ■=6:1; ME separated from Figures 24-28. Miagrammopes molitiis sp. nov. Fig. 24. Carapace, dorsal view. Fig. 25. Left first tibia and patella; dorsal view. Fig. 26. Left first metatarsus; prolateral view. Figs. 27-28. Left male palp: pro- lateral and retrolateral views, respectively. 1968 MIAGRAMMOPES IN PANAMA AND THE WEST INDIES 15 one another by slightly more than five times their diameter, from LE by about 3.3 times their diameter; with only a moderate amount of black pigment around eyes. Chelicerae, maxillae, lip and sternum impossible to observe clearly because of sharply curled position of holotype and its fragility. Legs: tibial index of first leg 11, of fourth leg 10; first leg with a series of spines on tibia and metatarsus as shown in Figures 25-26; these spines are less robust and have a different placement than in other species treated in this study; other legs with few or no spines but with a copious hair covering; no true fringes observed; first legs considerably more than twice as long as second legs, three times as long as third legs and 1.7 times as long as fourth legs. Palp: complicated; parts difficult to observe because of transparency; form essentially as shown in Figures 27-28. Abdo- men: apparently quite typical of males of the genus. Color in al- cohol; carapace with a broad, yellowish, median stripe extending the whole length of this part of the body and occupying the middle third; lateral thirds of the carapace grayish brown; sternum yellow- ish; legs in general yellowish but first legs somewhat grayish brown along dorsal surfaces; abdomen with a median, dorsal, yellowish, longitudinal stripe and a narrower grayish stripe on each side; re- mainder of the abdomen yellowish with variations. Records. The female is unknown; there are no male paratypes; an immature specimen was taken in the same locality and on the same day as the holotype but its identification is uncertain. MlAGRAMMOPES OBLUCUS Sp. nOV. Figures 29-30 Holotype. The female is from Jamaica, Portland Parish, Hard- war Gap, November 22, 1957. The name of the species is an arbitrary combination of letters. Description. Total length from anterior border of somewhat ex- tended chelicerae to posterior end of spinnerets about 5.57 mm (body considerably curled). Carapace 1.58 mm long; 1.12 mm wide at level of lateral ocular tubercles where it is widest; about 0.55 mm tall; median thoracic fovea barely discernible; with a moderately developed coat of hair (Fig. 29). Eyes: four as usual, in a gently recurved row, seen from above; ratio of eyes ME : LE = 8:10; ME separated from one another by nearly 6.25 times their diameter, from LE by about 3.5 times their diameter. Chelicerae, maxillae and lip apparently quite typical of females of the genus. Sternum: essentially as described for M. simus Chamberlin and Ivie except that there is an angular distortion at the swollen and 16 BREVIORA No. 289 ^li'!!# 33 Figures 29-30. Miagrammopes obliicus sp. nov. Fig. 29. Carapace; from above. Fig. 30. Epigynum; seen from below. Figs. 31-33. Miagram- mopes pinopus sp. nov. Fig. 31. Carapace; dorsal view. Fig. 32. Epigynum; seen from below. Fig. 33. Abdomen; lateral view, right side. expanded median part; posterior end not clearly visible but appears to terminate in a narrow, rounded point. Legs: tibial index of first leg 12; of fourth leg 8; first leg slightly more than twice as long as second leg, nearly three times as long as third leg and about 1.4 times as long as fourth leg; calamistrum extends over about five- sixths of the length of the fourth metatarsus; first metatarsus with a definite dorsal fringe; fourth metatarsus and tarsus each with a row of short spines on ventrolateral surface essentially as in M. simus Chamberlin and I vie (Fig. 40). Abdomen: robust, with a slight swelling a little in front of the middle; otherwise quite typical of females of the genus as observed in this study. Epigynum: some- what obscure; essential observed features shown in Figure 30. Color in alcohol: carapace generally grayish brown; with a lighter, median, irregular stripe extending the entire length of this part of the body; the lateral thirds darker and with a narrow, marginal white stripe. Sternum light yellowish brown. Mouth parts yellowish with varia- tions; the lip is largely grayish. Legs: yellowish in general; first and 1968 MIAGRAMMOPES IN PANAMA AND THE WEST INDIES 17 second legs with dorsal and dorsolateral surfaces dark grayish brown; fourth metatarsus dark dorsally and prolaterally. Abdomen: with a central, dorsal, very irregular brownish gray stripe flanked on each side by a lighter, irregular stripe; a pair of black dots is present on the dorsum somewhat in front of the middle, and another pair of black dots occurs just behind the middle; lateral sides irregularly brownish; the venter is a kind of light, reddish brown with a narrow, central stripe flanked on each side by a narrow lighter stripe; two pairs of black dots appear in the posterior half of the venter. Records. The male is unknown. One female was taken on the same day and in the same general locality as the holotype and is tentatively regarded as a paratype, but there are differences which may be due to immaturity. MlAGRAMMOPES PINOPUS Sp. UOV. Figures 31-33 Holotype. The female is from St. John, U. S. Virgin Islands. Centerline Road, 4 miles east of Cruz Bay, March 4, 1964. The name of the species is an arbitrary combination of letters. Description. Total length from anterior border of extended cheli- cerae to posterior end of extended abdomen 6.9 mm. Carapace: 1.76 mm long; 1.07 mm wide at level of lateral ocular tubercles; about 0.55 mm tall; with grooves in ocular region only moderately developed; with median thoracic fovea shallow and inconspicuous (Fig. 31). Eyes: four as usual, in a gently recurved row; ratio of eyes ME : LE = 6.5 : 7; ME separated from one another by about seven times their diameter, from LE by slightly more than half that distance. Chelicerae and maxillae apparently typical of females of the genus. Lip: long, slender, convex; widest just before the middle; distal end a fine point. Sternum: long, slender, except for excessive width between second and third coxae; narrowest between first and second coxae; posterior end very hairy, somewhat tuberculate, rounded and ending just opposite bases of fourth coxae, which are separated by a little more than one-third of their width. Legs: tibial index of first leg 1 1, of fourth leg 7; first leg slightly more than twice as long as second leg, somewhat less than three times as long as third leg, and about 1.4 times as long as fourth leg; all tarsi with claws as usual; calamistrum as usual in females of the genus; fourth metatarsus and tarsus with spines essentially as shown in Figure 40 for M. simiis Chamberlin and Ivie; other spines seem to be lacking on legs. Palp: with well developed claw and with about five long, slender teeth. Abdomen: moderately raised dorsally; with posterior 18 BREVIORA No. 289 end continued into a definite caudal extension above anal tubercle (Fig. 33) and spinnerets; otherwise typical of females of the genus. Epigynum: essential features shown in Figure 32; quite distinctive; resembles that of M. latens Bryant. Color in alcohol: carapace grayish brown with many narrow, darker, irregular lines and spots and with lateral sides somewhat lighter; sternum yellowish brown; legs generally yellowish with first pair gray on dorsal surfaces; mouth parts yellowish. Abdomen: light yellowish in general; very reticulate; with a grayish brown, broken stripe along the median dorsal region and with lateral regions bearing irregular and poorly defined, darker, oblique bands; venter highly reticulate and with a pair of very narrow, dark stripes extending from epigynum to the vicinity of the cribellum. Records. The male is unknown; there are no female paratypes. MlAGRAMMOPES SCOPARIUS SimOU Figures 34-36 Miagrammopes scopariiis Simon, 1891: 555. The types, male and female, are in the British Museum (Natural History) and were taken on St. Vincent, B.W.I. Simon, 1892: 220; Petrunkevitch, 1911: 97; 1930: 232; Lutz, 1915: 79; Roewer, 1954: 1350; Bonnet, 1957: 2831. Lutz (1915) reported this species from Puerto Rico, but this is now regarded as an error in identification and, apparently, the species is known only from the type locality. No figures were pub- lished with the original description, but in 1902 Simon published a figure of the male palp (fig. 164). Through the courtesy of the Department of Zoology, British Museum (Natural History), I have been able to examine eight specimens from St. Vincent. Three of these are males and one of them has been selected for brief de- scription as given below. The five females all appear to be in various stages of immaturity and, for this reason, no detailed de- scription has been prepared. Male. Total length about 3.2 mm (body somewhat curled). Carapace 1.1 mm long; 0.79 mm wide at level of LE; general ap- pearance as shown in Figure 34. Simon gave length of male as 4.5 mm. Eyes: four as usual, in a recurved row as seen from above; ME separated from one another by 4.6 times their diameter, sep- arated from LE by about 2.25 times their diameter; ratio of eyes ME : LE r= 7 : 8. Chelicerae, maxillae, lip and sternum appar- ently all typical of males of the genus. Legs: 1423 in order of length; first leg more than twice as long as second leg, more than three times as long as third leg but a little less than twice as long as 1968 MIAGRAMMOPES IN PANAMA AND THE WEST INDIES 19 Figures 34-36. Miagrammopes scopariits Simon. Fig. 34. Carapace of male from above. Fig. 35. First right tibia; retrolateral view. Fig. 36. Left palpus of male; prolateral view. fourth leg. Tibial index of first leg 11, of fourth leg 9. First tibia with a conspicuous battery of robust spines (Fig. 35). Calamistrum lacking; present in immature male. Palp: essential features shown in Figure 36. Abdomen: typical of the genus; without unusual features. Color in alcohol: carapace light yellowish brown in gen- eral, somewhat lighter in middle of anterior half; whole area ir- regularly reticulated with fine dark lines; legs yellowish in general with variations; first leg grayish dorsally; abdomen yellowish through the center and grayish along lateral sides with venter yellowish. Females. As already indicated, the five females accompanying the males seem to be immature with no developed epigyna. Their general appearance agrees quite well with that of the males. The largest female has a total length of about 4.22 mm (Simon gave length of female as 6 mm). Miagrammopes simus Chamberlin and Ivie Figures 37-44 Miagrammopes simus Chamberlin and Ivie, 1936: 12, pi. 2, figs. 11-12. The female holotype from Barro Colorado Island, Panama Canal Zone, is probably in the University of Utah. Roewer, 1954: 1350; Bonnet, 1957: 2832. Miagrammopes alboguttatus, — Banks, 1929: 89. (Not M. alboguttatus F. P. -Cambridge.) 20 BREVIORA No. 289 I have examined the specimens identified as M. alboguttatus F. P. -Cambridge by Mr. Banks and am convinced that they are M. simus. Because the original description of the holotype female is so brief I have thought it worth while to offer a more detailed descrip- tion of a selected female from the locality where I collected the original specimen and loaned it to the senior author. It is difficult to match males and females in this genus but I believe I have done so in this case. A male has also been selected and described in the following pages. Males and females are by far the most numerous of any species of the genus in my collection. Female. Total length from anterior border of somewhat porrect chelicerae to posterior end of posterior spinnerets 6.63 mm. Cara- pace about 1.76 mm long; 1.41 mm wide at level of lateral ocular tubercles where it is widest; with a conspicuous pair of depressions slightly behind and somewhat medial to ME; with these depressions continued to the lateral margins by grooves in a slightly diagonal direction; with width only slightly narrowed to posterior margin; medial thoracic fovea short and slightly procurved. Eyes: four as usual, in a slightly procurved row, viewed from above; ratio of eyes ME : LE = 10 : 13; ME separated from one another by slightly less than nine times their diameter, from LE by about 1.6 times their diameter. Chelicerae: removal of these organs seems to show that the promargin of the fang groove has a single tooth and several minute cusps each with a stout, long hair; the retromargin of the fang groove appears to have a series of minute cusps each also with a stiff hair. Maxillae and lip: apparently quite typical of females of the genus (Fig. 37). Palpal tarsus with a robust claw having four or five very slender teeth (Fig. 38). Sternum: very angular; swol- len near middle where a conspicuous shelf crosses from interval between second and third coxae; slopes sharply from high region both anteriorly and posteriorly; a second less conspicuous shelf crosses the sternum between third and fourth coxae; posterior end obscure but not extended between fourth coxae which are separated by about one-third of their width (Fig. 39). Legs: tibial index of first and fourth legs 10; all tarsi short; first tarsus slightly more than one-fifth as long as first metatarsus; all legs with fine hair; first tibia with a fringe of fairly long hair; apparently three true claws with several spurious claws on each tarsus; spines on fourth metatarsus and tarsus as shown in Figure 40; trichobothria appear to be present on femora, tibiae and metatarsi; the calamistrum extends over about seven-twelfths of length of fourth metatarsus; first leg nearly 3.5 times as long as third leg. Abdomen: essentially as shown in dorsal 1968 MIAGRAMMOPES IN PANAMA AND THE WEST INDIES 21 Figures 37-44. Miagrcunmopes simiis Chamberlin and Ivie. Fig. 37. Female maxillae and lip. Fig. 38. Female left palpal tarsus; prolateral view (with aid from dissected specimen). Fig. 39. Female sternum; seen from below. Fig. 40. Right fourth metatarsus and tarsus of female; re trolateral view. Fig. 41. Epigynum from below. Fig. 42. Carapace of male from above. Fig. 43. First tibia of male; retrolateral view (a few spines restored from scars). Fig. 44. Left palp of male; retrolateral view. 22 BREVIORA No. 289 view in figure 11 of Chamberlin and Ivie, 1936; cribellum entire; spinnerets six as usual. Epigynum: somewhat obscured by hair; essential features shown in Figure 41 (variations in appearance have been noted among numerous specimens assigned to this species). Color in alcohol: legs generally brown with variations; first pair darkest of all but with tarsus light yellowish brown and metatarsus somewhat darker; carapace light yellowish brown throughout the dorsal region and with lateral sides dark brown; sternum yellowish brown with variations. Abdomen: with a semi- transparent cardiac area and white spotted area posterior to cardiac region; a pair of irregular, white stripes extend along the dorsum nearly the entire length; lateral regions brownish with an irregular white spot a little in front of the middle; venter light yellowish brown with some variation. Considerable variation in color pattern has been noted among the numerous specimens of this species now in the collection. Male. Total length from anterior border of slightly porrect cheli- cerae to posterior end of abdomen about 3.05 mm (body curled and difficult to measure accurately). Carapace 1.06 mm long; 0.99 mm wide at level of lateral ocular tubercles; narrowed to 0.84 mm just behind ocular tubercles; pits and diagonal grooves, so con- spicuous in the female, are here much less prominent; general ap- pearance shown in Figure 42. Eyes: four as in female, in a straight row as seen from above; ME separated from one another by 8.5 times their diameter, from LE by about twice their diameter; ratio of eyes ME : LE = 7 : 10; lateral eyes directed somewhat ven- trally. Chelicerae, maxillae and lip apparently as in female except for great reduction in size; details hidden by curled position and interlaced legs which are too fragile for disentanglement. Sternum: greatly swollen in middle but lacks the conspicuous transverse folds seen so clearly in the female; fourth coxae almost contiguous. Legs: tibial index of first leg 12, of fourth leg 11; calamistrum apparently lacking; first leg more than twice as long as second leg and nearly three times as long as third leg; first tibia with a battery of nearly two dozen robust spines on dorsal and dorsolateral sides essentially as shown in Figure 43; few spines observed elsewhere and the most conspicuous of these are on the first metatarsus. Palp: with femur, patella and tibia short and without special features; tarsus compli- cated and distinctive (Fig. 44). Abdomen: without noteworthy features; typical of males of the genus. Color in alcohol: in general light yellowish with variations; carapace with a somewhat angular, nearly white, central dorsal spot, with remainder of dorsum yellow- ish and with lateral sides grayish and with considerable black pig- ment in ocular region; sternum grayish; abdomen with dorsum light 1968 MIAGRAMMOPES IN PANAMA AND THE WEST INDIES 23 yellowish with numerous white flecks; a broken, grayish stripe ex- tends along each side of abdomen and the venter is irregularly grayish. Legs are yellowish in general but with variations; first legs with femora grayish retrolaterally; first tibiae, metatarsi and tarsi are all yellowish brown; other legs with grayish spots. Records. Specimens assigned to this species have been collected on Barro Colorado Island, Panama Canal Zone, during each of my several periods of field work there beginning in 1934; I have a few records of the species from other localities in the Canal Zone and in Panama proper. The described female is from Barro Colorado Island, Canal Zone, May 18, 1964; the described male is from the same locality, August 2, 1954. MlAGRAMMOPES TONATUS Sp. nOV. Figures 45-49 Holotype. The male is from Jamaica, W. I., St. Andrew Parish, Mt. James at Plaintain River, October 24, 1957. The name of the species is an arbitrary combination of letters. Description. Total length 3.45 mm. Carapace 1.08 mm long; 0.79 mm wide at level of lateral ocular tubercles; not conspicuously grooved as in some species in the genus; depressions just posterior to ME only moderately developed; median thoracic fovea only moderately depressed (Fig. 45). Eyes: four as usual, in a mod- erately recurved row, seen from above; ratio of eyes ME : LE r= 5 : 6; ME separated from one another by about 6.5 times their diameter, from LE by slightly more than four times their diameter; lateral ocular tubercles only moderately developed. Chelicerae, maxillae and lip apparently quite typical of males of the genus. Sternum: with the usual form in this genus; greatly swollen, widened and extended between second and third coxae. Legs: tibial index of first leg 11, of fourth leg 8; calamistrum apparently lacking; first tibiae and metatarsi with few conspicuous, erect spines, others hori- zontal and hidden by hairs (Fig. 46); first leg more than twice as long as second leg, nearly twice as long as fourth leg and about three times as long as third leg; tarsal claws as usual; several tri- chobothria observed on tibiae and other segments. Palp: very short as usual; essential features shown in Figures 47-49. Abdomen ap- parently typical of males of the genus. Color in alcohol: carapace generally a dark gray with lighter spots and darker, irregular lines together with a dull, yellowish, ventral margin; sternum dark gray, lighter at anterior end; first pair of legs fight yellowish brown ventral- ly, darker gray elsewhere except for tarsus which is yellowish brown; 24 BREVIORA No. 289 47 Figures 45-49. Miagrammopes tonatus sp. nov. Fig. 45. Carapace of male from above. Fig. 46. First right tibia; retrolateral view. Figs. 47-48. Male palp; retrolateral and prolateral views, respectively. Fig. 49. Tip of tarsal bulb; nearly dorsal view. legs 2-4 much as first but less conspicuously so; mouth parts yellowish with grayish streaks. Abdomen: with a somewhat in- dented, dark gray, median stripe extending from base for about three-fourths of length of abdomen and then joining a darker area which covers the posterior quarter of the dorsum; lateral to this median region are lighter areas with irregular, yellowish white spots; venter with a central, light, reticular stripe extending from genital groove to posterior end; on each side of this central stripe there is a narrow, irregular, semitransparent stripe and further laterally are reticulated areas; entire color pattern indefinite and difficult to describe adequately. Records. An immature male taken with the holotype appears to belong to this species. The female is unknown. 1968 MIAGRAMMOPES IN PANAMA AND THE WEST INDIES 25 MlAGRAMMOPES UNIPUS Sp. nOV. Figures 50-55 Holotype. The male is from Barro Colorado Island, Panama Canal Zone. June, 1950. The name of the species is an arbitrary combination of letters. Description. Total length from anterior border of chelicerae to posterior end of spinnerets 3.58 mm. Carapace 1.06 mm long; 1.0 mm wide at level of lateral ocular tubercles; about 0.4 mm tall; with general form essentially as in M. simus Chamberlin and Ivie 50 54 55 53 52 Figures 50-55. Miagrammopes unipus sp. nov. Fig. 50. Right first tibia of male; retrolateral view. Figs. 51-52. Left palpal tarsus of male; prolateral and retrolateral views, respectively. Fig. 53. Left palpal femur, patella and tibia; retrolateral view (from male paratype). Fig. 54. Ster- num of described female paratype; seen from below. Fig. 55. Epigynum; seen from below. 26 BREVIORA No. 289 with depressions medial to ME moderately conspicuous but grooves leading to lateral border almost lacking. Eyes: four as usual, in a gently procurved row as seen from above; ME separated from one another by about eight times their diameter, from LE by slightly more than their diameter; ratio of eyes ME : LE = 8 : 10.5. Chelicerae: short and largely hidden by elongated maxillae which are twice as long as lip; all of these parts hidden by fragile sur- rounding structures. Sternum: elongated, narrow, quite convex opposite second coxae; continued laterally between second and third coxae; obtusely terminated between bases of fourth coxae but not continued between the latter, which are separated by about one- fifth of their width. Legs: tibial index of first leg 10, of fourth leg 7; first leg nearly three times as long as third leg; second leg only a little shorter than fourth leg; first tibia with a set of robust spines (Fig. 50); first metatarsus with fewer and less robust spines; tarsal claws as usual in the genus. Palp: distinctive features shown in Figures 51-53; femur of moderate length; patella and tibia extreme- ly short. Abdomen: considerably flattened dorsally; otherwise typical of males of the genus. Color in alcohol: carapace with a narrow, central, somewhat broken grayish stripe; with considerable black pigment in ocular area; with a broad, lateral, grayish stripe on each side from lateral eyes to posterior border; the lower part of this gray stripe becomes lighter, thus adding another stripe to the lateral sides and, finally, the ventral margin is darker gray, thus making a total of seven distinguishable stripes on this part of the body; ster- num yellowish with fairly broad, grayish margins. Abdomen: yellowish dorsally with two pairs of narrow, elongated, grayish spots dorsolateral in position and a series of small grayish spots at pos- terior end above spinnerets; laterally there are several short, gray- ish stripes; the venter has a narrow, median, grayish stripe with a narrow yellowish stripe on each side, and further laterally is a broader, light grayish stripe on each side; shortly in front of the re- duced cribellum there is a narrow, grayish, transverse bar. Legs yellowish in general with several grayish spots and whole segments; fourth legs with most conspicuous grayish parts; fourth patellae dark gray; distal parts of fourth tibiae the same and nearly the entire metatarsi are gray. Female paratype. The described female paratype measures 5.2 mm from anterior border of carapace to posterior end of spinnerets; length from anterior border of somewhat extended chelicerae to posterior end of spinnerets about 5.59 mm. Carapace 1.3 mm long; 1.32 mm wide at level of lateral ocular tubercles; medial to ME there is a fairly conspicuous depression in communication with a 1968 MIAGRAMMOPES IN PANAMA AND THE WEST INDIES 27 narrow groove which extends along the border of the ocular tubercle to the lateral border of the carapace; another narrow groove in front of the ocular tubercle passes to the lateral border of the carapace and is connected to a short, narrow groove extending about half way through the tubercle between ME and LE; otherwise apparently typical of females of the genus. Chelicerae hidden and not clearly visible. Maxillae long, parallel, with rounded lateral borders. Lip: about three-fifths as long as maxillae; central portion swollen; sternal suture may not be a movable joint. Sternum: only moder- ately convex; not conspicuously swollen as in most females of the genus seen in this study; not extended between fourth coxae which are separated by nearly one-fourth of their width (Fig. 54). Legs: tibial index of first leg 10, of fourth leg 9; first leg twice as long as second leg and more than three times as long as third leg; fourth leg about 0.7 as long as first leg; spines appear to be lacking except for the moderately developed set of robust spines on fourth meta- tarsus and tarsus similar to those shown in Figure 40; legs with a fairly well developed coat of hair and first metatarsus has a mod- erately well developed ventral fringe; calamistrum appears typical of females of the genus. Abdomen: with a pronounced dorsal medial swelling; quite typical of females of the genus. Epigynum: obscure and difficult to observe clearly; observed essentials shown in Figure 55; a dorsal injury to the abdomen has permitted eggs to extrude, thus indicating that the female is nearly or quite mature. Color in alcohol: essentially as in male holotype except that the gray parts in that sex are here all considerably extended and blackened. Records. The described female paratype was taken on Barro Colorado Island, Panama Canal Zone, August 22, 1939. Several male paratypes are in the collection from the type locality and taken as follows: August, 1936, 1950; July, 1939, 1950, 1954; February, 1958. One male paratype was taken at Summit, Panama Canal Zone, July, 1950. Three additional females are, with some un- certainty, placed in this species and were taken as follows: July, 1934; August, 1939; May, 1964, all from Barro Colorado Island, Canal Zone. REFERENCES CITED Banks, Nathan 1909. Arachnida of Cuba. Sec. Rept. Centr. Exper. Stat. Cuba. Pp. 150-174, pi. 45. 1929. Spiders from Panama. Bull. Mus. Comp. Zool., 69: 53-96, 4 pis. 28 BREVIORA No. 289 Bonnet, Pierre 1957. Bibliographia Araneorum. Toulouse. Vol. 2 (3). Bryant, Elizabeth 1936. Descriptions of some new species of Cuban spiders. Mem. Soc. Cubana Hist. Nat., 10: 325-332, pi. 23. 1940. Cuban spiders in the Museum of Comparative Zoology. Bull. Mus. Comp. Zool., 86(7) : 249-532. 22 pis. 1948. The spiders of Hispaniola. Bull. Mus. Comp. Zool., 100(4): 331-447, 12 pis. Cambridge, O. P. - 1870. Descriptions and sketches of two new species of Araneidea with characters of a new genus. Jour. Linn. Soc. London, Zool., 10: 401, figs. 1-12. Cambridge, O. P. - and F. P. - Cambridge 1889- Arachnida-Araneida. Vols. I-IL ///.• Biologia Centrali-Ameri- 1905. cana. Dulau & Co., London. Chamberlin, Ralph V., and Wilton Ivie 1936. New spiders from Mexico and Panama. Bull. Univ. Utah, 27 (5) (Biol, series, 3, No. 5): 3-103, 17 pis. Franganillo, Pelegrin Balboa, S. J. 1936. Los Aracnidos de Cuba hasta 1936. Privately printed in Havana, 183 pp. LuTZ, Frank E. 1915. List of Greater Antillean spiders, with notes on their distri- bution. Ann. New York Acad. Sci., 26: 71-148. Petrunkevitch, Alexander 1911. A synonymic index-catalogue of spiders of North, Central, South America, etc. Bull. Amer. Mus. Nat. Hist., 29: 1-809. 1926. Spiders from the Virgin Islands. Trans. Conn. Acad. Arts Sci., 28: 21-78. 1930. The spiders of Porto Rico. Pt. 2. Trans. Conn. Acad. Arts Sci., 30: 159-355, 240 figs. ROEWER, C. Fr. 1954. Katalog der Araneae. Brussels, Vol. 2b: 927-1751. Simon, E. 1891. On the spiders of the Island of St. Vincent. Pt. I. Proc. Zool. Soc. London 1891: 549-575, pi. 42. 1892- Histoire naturelle des araignees. Deuxieme edition. 2 Vols. 1903. Librairie Encyclopedique de Roret, Paris. (Received 28 August 1967.) BREVIORA Meseuinii of Comparative Zoology Cambridge, Mass. 29 May, 1968 Number 290 GEOGRAPHIC VARIATION IN THE HISPANIOLAN FROG ELEUTHERODACTYLUS WETMOREI COCHRAN Albert Schwartz' Eleutherodactylus wetmorei Cochran is a brightly colored lepto- dactylid frog which was described from a series of four specimens collected by Dr. Alexander Wetmore at Fond des Negres, Departe- ment du Sud, Haiti. The holotype and paratypes were taken from a communal nest of the Palm Chat {Dulus dominicus Linnaeus) along with two Hyla dominicensis Tschudi (Cochran, 1932:191). Later, Cochran (1941:76-77) reported 11 additional specimens of E. wetmorei, collected by W. L. Abbott, from Moron, Departe- ment du Sud, Haiti, near the extreme western tip of the Tiburon Peninsula. Both the typical and Abbott specimens were especially poorly preserved, but the frog is a very distinctive one in pattern (and in coloration in life). Two populations of E. wetmorei have the concealed surfaces of the hindlimbs with dark patterns on the brightly colored (orange to yellow) ground color; Cochran pointed out, however, that the Moron frogs seemed to lack this distinctive thigh pattern, and Shreve and Williams (1963:324) commented that two of the Moron series they examined lacked thigh markings, in contrast to the single individual they had from the Petionville area. The two-note voice of E. wetmorei is prominent in nocturnal choruses throughout southern Haiti; the frogs, however, are arbo- real (as the situation for the type series intimates), and collecting series of E. wetmorei is extremely difficult. Through the efforts of Dr. Ernest E. Williams, several excellent lots of E. wetmorei are now available from some Haitian localities, including the south- ern slope of the Massif de la Selle in the vicinity of Thiotte, and from the area between Jeremie and Dame-Marie at the extreme tip of the Tiburon Peninsula. The latter specimens amply confirm the ^Dept. of Biology, Miami-Dade Junior College, Miami, Fla. 33167. 2 BREVIORA No. 290 distinctness of the distal peninsular population of E. wetmorei, and the Thiotte specimens likewise demonstrate characters (which also occur elsewhere along the Dominico-Haitian border) distinctive to the populations in that region. Remarkably, there is now more material of these two peripheral populations than there is of the nominate subspecies, although I have taken specimens of the latter at Camp Perrin on the southern foothills of the Massif de la Hotte. Aside from my own material in the Albert Schwartz Field Series (ASFS), I have examined specimens in the American Museum of Natural History (AMNH), Carnegie Museum (CM), Museum of Comparative Zoology (MCZ), and United States National Museum (USNM); for the loan of these frogs I am grateful to Charles M. Bogert and George W. Foley, Neil D. Richmond and Clarence J. McCoy, Jr., Ernest E. Williams, and Doris M. Cochran. In the field I have had the capable assistance of Ronald F. Klinikowski, David C. Leber, and Richard Thomas. Messrs. Leber and Thomas were successful in securing the first specimen of E. wetmorei from the Republica Dominicana; it was this frog, strikingly different from material I had seen in life at Camp Perrin, which prompted the present study. The excellence of the illustrations is once more due to the work of Mr. Leber. The collections in Haiti made for the Museum of Comparative Zoology were financed by NSF Grant 1 6066 to Dr. Ernest E. Wil- liams. SYSTEMATIC ACCOUNTS Eleutherodactylus wetmorei Cochran, 1932 Eleutherodactyliis wetmorei Cochran, 1932, Proc. Biol. Soc. Washington, 45: 191. Eleutherodactylus auriculatus wetmorei, — Cochran, 1941, Bull. U. S. Natl. Mus., 177:74. Eleutherodactylus varians wetmorei, — Schwartz, 1960, Sci. Publ. Reading Public Mus. and Art Gallery, 11:6. Eleutherodactylus wetmorei, — Schwartz [in press]. Studies Fauna Curasao and other Caribbean Islands. Remarks: Although E. wetmorei was originally named as a dis- tinct species, Cochran later (1941 : 75) considered wetmorei closely related to E. auriculatus Cope (= E. varians Gundlach and Pet- ers). Later still, I (Schwartz, 1960:6) suggested that wetmorei might indeed be correctly regarded as a subspecies of E. varians, since the two forms have several structural and pattern features in common (although chromatically they differ strikingly). The 1968 ELEUTHERODACTYLUS WETMOREI 3 calls of E. varians and E. wetmorei are, however, very different (Schwartz, in press), the call of E. wetmorei resembling that of the Cuban E. eileenae Dunn rather than E. varians. Accordingly, I have once more separated E. wetmorei nomenclatorially from E. varians; the former species is limited to Hispaniola, the latter to Cuba and the Isla de Pinos. ELEUTHERODACTYLUS WETMOREI WETMOREI Cochrau Type locality: Fond des Negres, Dept. du Sud, Haiti; holotype, USNM 72617. Definition: A subspecies of E. wetmorei characterized by anterior and posterior faces of thighs reticulate black to dark brown on an orange ground, and groin orange with similarly colored reticu- lum (Fig. 2A). Remarks: There are only ten specimens which I assign to the nominate subspecies; of these, the two paratypes (USNM 72618- 19) are in extremely poor condition, and three (ASFS X2993-95) are juveniles with snout-vent lengths between 11.5 and 18.2 (all measurements in millimeters). Mensural data for a single male (ASFS X2684) are: snout-vent length 28.7, head length 11.3, head width 12.8, tympanum 2.0, eye 4.0, naris to eye 3.8, femur 12.8, tibia 15.2. fourth toe 11.5; measurements of four females (extremes and means) are: snout-vent length 33.5-35.1 (34.2); head length 12.0-13.5 (12.9); head width 13.3-14.7 (14.1); tym- panum 2.2-2.5 (2.3); eye 3.8-4.8 (4.2); naris to eye 3.7-4.5 (4.2); femur 14.1-15.2 (14.8); tibia 17.1-17.8 (17.4); fourth toe 13.4- 14.3 (13.8). The condition of the topotypes of E. w. wetmorei precludes ab- solute assurance that the material from Camp Perrin is identical in pattern to them. Cochran's figure (1941:75, fig. 24) however, shows quite clearly that the anterior and posterior faces of the thigh are darkly reticulate, and these markings plus the pale interocular bar are still barely visible on the holotype. Although Fond des Negres lies in the extreme northeastern foothills of the Massif de la Hotte and Camp Perrin lies in the southwestern foothills of the same range, the Camp Perrin specimens agree fairly closely in thigh design with the Fond des Negres material. Accordingly I con- sider the Camp Perrin specimens to be E. w. wetmorei, although there is a possibility that, with the acquisition of fresh topotypical material, the Camp Perrin population will be found to differ from that at Fond des Negres. 4 BREVIORA No. 290 Fresh specimens of E. w. wetmorei from Camp Perrin are tan to brownish tan in hfe, with a very faint pair of dorsolateral pale lines (ASFS X2684) present in one individual. The posterior and an- terior faces of the thigh, the groin, and the upper surface of the pes were orange (pi. 2 D 12; all color designations from Maerz and Paul, 1950) in life, and the entire ventral surface was yellow (pi. 9 L 5). Both faces of the thighs, as well as the groin, are overlaid with a dark brown to greenish black reticulum; reticulum remnants more or less outline the orange groin patch and separate the bright color from the more drab dorsal tan. The sides are at times vaguely marked with a brown reticulum which extends, in a diluted fashion, onto the chest as an area of brownish dots or flecks. The throat may be flecked with brown, and there is a pair of yellow-orange glands on the posterior portion of the throat; inguinal glands are absent in the species. The underside of the hindlimbs is usually marked with brown continuations of the anterior and posterior thigh pattern, and the crus also shows some ventral brown reticulations. The ventral crural reticulum is a continuation of the anterior and posterior crural reticulum, occupying the concealed surfaces of the crus. All specimens have a pale (tan to buffy) broad interocular bar, with one exception (MCZ 35199) which now appears to lack this feature. In the juveniles, the entire thigh was yellow in life, but the thigh pattern is identical to that of aduhs. In adults, the iris is golden above and below, whereas in juveniles the iris is metaUic buffy. At Camp Perrin, E. w. wetmorei was abundant in trees about the settlement, and the ringing two-note call made up a prominent portion of each night's choruses. Specimens, on the other hand, were difficult to secure; the three juveniles were taken by Mr. Leber from arboreal bromeliads, whereas the adult male (ASFS V2684) was obtained while calling on a Sansevieha leaf only one foot off the ground. Like the equally arboreal Cuban E. varians, E. wet- morei occasionally can be collected when it calls from low vocaliz- ing sites. As pointed out above, all known localities for E. w. wetmorei are associated with the lower slopes of the Massif de la Hotte. The locality near Miragoane is unlocatable on any current map, so it is possible that this specimen came from a lower elevation than the other E. w. wetmorei; the elevation at Camp Perrin is 1000 feet (305 meters) and that at Fond des Negres is 730 feet (240 meters). Figure 1 shows localities for specimens of E. w. wetmorei. 1968 ELEUTHERODACTYLUS WETMOREI Specimens examined: Haiti, Dept. du Sud, Camp Perrin, 6 (ASFS X2684, X2897, X2962, X2993-95); Fond des Negres, 3 (USNM 72617 — holotype; USNM 72618-19 — paratypes); Bu- tete, nr. Miragoane, 1 (MCZ 35199). Fig. 1. Map of southwestern Haiti, showing the Tiburon Peninsula. Localities for E. wetmorei wetmorei, E. w. williamsi, and E. w. ceraemeriis are shown as solid circles. ELEUTHERODACTYLUS WETMOREI CERAEMERUS new subspecies Holotype: MCZ 36101, an adult female, from Thiotte, Dept. de rOuest, Haiti, one of a series collected April 1962 by G. Whiteman. Paratypes: MCZ 36097-100, 36102-05, USNM 146614, same data as holotype; MCZ 37216-18, same locality as holotype, sum- mer 1962, G. Whiteman; CM 38550-55. same locality as holotype, June 1962, G. Whiteman; MCZ 36107-11, Tete a I'Eau, near Sal- trou, Dept. de I'Ouest, Haiti, April 1962, G. Whiteman; CM 37786, MCZ 34501. Marbial, 21 km NE Jacmel, Dept. de I'Ouest, Haiti, 20-21 April 1961, L. Whiteman; AMNH 44034, ridge of Massif de la Selle, just south of Savane Zombi, on Saltrou road, "4500 feet," Dept. de TOuest, Haiti, 6 April 1935, W. G. Hassler; MCZ 34503, AMNH 44050, Colombia (= Colombier), Dept. de rOuest, Haiti, 6 April 1935, W. G. Hassler; MCZ 34502, AMNH 44036-37, La Mahot. near Colombia {— Colombier), Dept. de rOuest, Haiti. 7 April 1935, W. G. Hassler; MCZ 31733, Boutil- lier Road, Dept. de TOuest. Haiti, 9 August 1959, E. E. Williams BREVIORA No. 290 and A. S. Rand; MCZ 37141, La Boule, Dept. de TOuest, Haiti, 18 April 1960, J. A. Rivero; ASFS V271 1, 7 km SE Los Arroyos, 2200 feet (720 meters), Pedernales Province, Republica Domini- cana, 29 June 1964, R. Thomas. Definition: A subspecies of E. wetmorei characterized by anterior and posterior faces of thigh bright orange with a pattern of black spots rather than a reticulum, and groin orange with scattered black dots and outlined by black (Fig. 2B). Fig. 2 Three subspecies of E. wetmorei, as follows: A, E. w. wetmorei, ASFS X2962, Camp Perrin, Dept. du Sud, Haiti; B, E. w. ceraemeriis, MCZ 36101, holotype, Thiotte, Dept. de I'Ouest, Haiti; C, E. w. williamsi, MCZ 31151 , holotype, Marfranc, Dept. du Sud, Haiti. Description of holotype: An adult female with the following measurements: snout-vent length 35.6, head length 12.7, head width 14.5, longitudinal diameter of tympanum 2.5, longitudinal diameter of eye 4.5, naris to eye 4.0, femur 14.7, tibia 17.5, fourth toe 13.8. Head distinctly broader than long; snout truncate, with nares conspicuous at anterior end of canthus rostralis; diameter of eye slightly longer than distance from naris to anterior corner of eye; diameter of tympanum much less than diameter of eye, dis- tance from tympanum to eye slightly less than diameter of tym- panum. Interorbital distance 5.4, greater than diameter of eye. Digital discs present, large, those on digits 2, 3, and 4 only slightly smaller than area of tympanum. Fingers short, basally with a very small web, 3-4-2-1 in order of decreasing length; subarticular tubercles pale, concolor with palmar surface of hand. Toes moder- ately long, oiily very slightly webbed basally, 4-3-5-2-1 in order of decreasing length; subarticular tubercles prominent, concolor with plantar surface. Heels overlap strongly when femora are held at 1968 ELEUTHERODACTYLUS WETMOREI 7 right angles to body axis. Inguinal glands absent. Dorsum smooth. Throat and chest smooth, belly coarsely granular. Dorsal surface of fore- and hindlimbs smooth. Posterior and ventral faces of thighs covered with moderately sized juxtaposed flattened granules. Tongue large, free and notched behind, its greatest width equal to about two-thirds of that of floor of mouth. Vomerine teeth in two small, almost diagonal patches, enclosed within the median margins of the choanae, separated from them by a distance equal to the diameter of a choana, the two patches separated medially by a dis- tance equal to about half the length of a single vomerine patch. Coloration and pattern of Jiolotype: Dorsum (as preserved) brown and without pattern except for a vaguely discernible paler interocular bar and a very faint middorsal hairline; hindlimbs con- color with dorsum; concealed surfaces of thigh, crus, and pes paler (presumably orange in life) with scattered dark brown dots and outlined with a faintly darker brown line, punctate with dark brown dots along its length; groin spot pale (presumably orange in life) with a few scattered dark dots on the left side (right side immacu- late), the groin spot outlined by a vague darker line and sharply set off from the dorsal color; forelimbs dark brown, concolor with the dorsum but with a sharply distinct pale mark on the wrist and an- other at the forelimb insertion. Venter pale (presumably orange in life), heavily suftused with brown stippling on the sides anteriorly (anterior to the pale groin spot); throat densely stippled with dark brown; underside of forelimbs dark brown on antebrachium, pale on brachium. A pair of moderately prominent glandular areas on the posterior throat anterior to the level of the forelimb insertions. Variation: The series of nine male E. w. ceraemerus has the fol- lowing measurements (extremes and means): snout- vent length 28.5-33.4 (30.4), head length 9.8-11.8 (10.8), head width 10.8- 12.6 (12.2), tympanum 1.7-2.0 (1.8), eye 3.4-4.0 (3.7), naris to eye 2.9-3.9 (3.5), femur 11.6-14.2 (13.2), tibia 14.5-15.9 (15.2), fourth toe 11.2-12.3 (11.7). Twenty females have the following measurements: snout-vent length 26.7-38.3 (34.4), head length 9.6-13.3 (11.9), head width 10.8-15.0 (13.5), tympanum 1.6-2.5 (2.1), eye 3.1-4.7 (4.1), naris to eye 3.2-4.5 (3.8), femur 11.9- 16.8 (14.6), tibia 14.1-18.9 (17.2), fourth toe 10.4-14.3 (13.2). As preserved, the paratypes agree well in color with the holotype, although some specimens are paler (tan rather than brown). In coloration, the concealed hindlimb surfaces and the groin spot are always sharply and conspicuously set off from the remainder of the hindlimb. and the pattern of these areas is regularly one of dots or 8 BREVIORA No. 290 spots, never a reticulum as in E. w. wetmorei. At times the spots or dots are somewhat fused (MCZ 34501), but the condition never approaches that of the nominate subspecies. The groin spot may have some dark flecking within it, or it may be immaculate; in either case it is outlined dorsally by a dark and punctate line as in the holotype. Some specimens show a pale interocular bar (MCZ 36105), but this is not the regular condition in ceraemerus. A very fine median hairline is barely discernible in some specimens (USNM 146614), as it is in the holotype, but most frogs lack this pattern element. Other specimens (MCZ 36102) have some fine and vague stippling along the lower sides above the venter, but this condition likewise occurs only sporadically. Only those frogs which are dorsally as dark as the holotype have the sides of the venter anterior to the groin spot as darkly pigmented as does the holotype. There are only two specimens from north of the Massif de la Selle, those from La Boule and Boutillier Road; the balance of the series is from the southern side of that massif. I do not know if the ranges of these two populations are continuously directly across the massif, but it does not seem likely, since my impression is that E. wetmorei is generally a frog of low to moderate elevations in forested situations. In any event, the two frogs from north of the La Selle agree in thigh pattern details with those from the southern slope except that the degree of fusion between the spots is greater than is customary for southern specimens. It seems quite likely that another subspecies will be found to inhabit the northern slopes of the La Selle and its associated ranges above Port-au-Prince; the thigh pattern of these northern specimens does not approach the reticulum of E. w. wetmorei. Structurally, the paratypes agree with the holotype except that the buccal cavity of the latter is somewhat aberrant. In most specimens of E. w. ceraemerus, the vomerine patches are small, compact, and enclosed well within the inner margins of the choanae, separated from the choanae by a distance equal to that of one vomerine series and separated from each other by a similar distance. In the holo- type, the vomerine patches are closer to one another than in the balance of the series. Comparisons: Comparison of the subspecies wetmorei and ceraemerus is hardly necessary; the two forms can be easily dis- tinguished by the pattern of the concealed surfaces — reticulate in wetmorei, dotted in ceraemerus. Size differences are difficult to assess, since the series of E. w. wetmorei is very small. E. w. ceraemerus seems the larger of the two forms. 1968 ELEUTHERODACTYLUS WETMOREI 9 Although there are no color data on most specimens of E. w. cememerus, Richard Thomas recorded the color in life of the Dominican specimen as follows: dorsum grayish tan (= putty colored), venter bright orange. Forearm insertions, pre- and post- femoral surfaces, and groin bright orange with black dotting. Undersides of legs and dorsal and ventral surfaces of feet bright orange; dark flecking roughly outlining region of contact between orange and tan on both limbs and groin. Remarks: The abundance of specimens of E. w. ceraemerus, in contrast to the paucity of those of the nominate subspecies, is puzzling, although it is possible that the frog is more abundant or more easily secured on the southern slope of the Massif de la Selle than it is farther west. The Dominican specimen was collected while it was calling about ten feet above the ground in a coffee tree; others were heard calling in the same general region, but most were vocalizing high in the trees and thus were inaccessible. The scarcity of specimens from the Morne I'Hopital above Petionville is equally puzzling; the calls of this frog were commonly heard between Petionville and Fermate, but there are only two specimens from this entire region. The species apparently is absent from the Sierra de Baoruco, where the dominant, moderately sized, more or less arbo- real Eleutherodactylus is E. armstrongi Noble and Hassler. This species also occurs south of Port-au-Prince on the Montagnes Noires at Furcy, an area where E. wetmorei is absent. The Dominican specimen is from the extreme edge of the Republica Dominicana in what may well be considered the eastern extreme of the Massif de la Selle. The altitudinal limits of E. w. ceraemerus vary from a low eleva- tion of about 600 feet (200 meters) at Marbial to high elevations of about 4270 feet (1400 meters) near Savane Zombi. These elevations are greater than those of the nominate subspecies to the west. The name ceraemerus is derived from the Greek "keraia," a dot, and "meros," thigh, an illusion to the dotted concealed surfaces of the hindlimbs in this subspecies. Eleutherodactylus wetmorei williamsi new subspecies Holotype: MCZ 37757, an adult female, from Marfranc, Dept. du Sud, Haiti, one of a series collected 26-27 December 1962 by Hill and Vuilleumier. Paratypes: MCZ 37751-56, 37758-63, same data as holotype; MCZ 37587, Perine, near Jeremie, Dept. du Sud, Haiti, December 1962, G. Whiteman; MCZ 37586, Carrefour Sanon, near Jeremie, 10 BREVIORA No. 290 Dept. du Sud, Haiti, December 1961, G. Whiteman; USNM 60627- 35, Moron, Dept. du Sud, Haiti, 20 December 1917, W. L. Abbott. MCZ 28599-600, Moron, Dept. du Sud, Haiti, 24 December 1917, W. L. Abbott. Definition: A subspecies of E. wetmorei characterized by com- plete absence of pattern on anterior and posterior faces of thigh and on groin spot (Fig. 2C). Description of holotype: An adult female with the following measurements: snout-vent length 33.8, head length 11.8, head width 13.0, longitudinal diameter of tympanum 2.0, longitudinal diameter of eye 4.2, naris to eye 4.1, femur 14.2, tibia 16.9, fourth toe 13.1. Head distinctly broader than long; snout truncate, with nares conspicuous at anterior end of canthus rostralis; diameter of eye about equal to distance between naris and anterior corner of eye; diameter of tympanum much less than diameter of eye, dis- tance from tympanum to eye equal to about one-half the diameter of tympanum. Interorbital distance 4.0, slightly less than diameter of eye. Digital discs present, large, those on digits 2, 3, and 4 only slightly smaller than area of tympanum. Fingers short, basally with a very small web, 3-4-2-1 in order of decreasing length; subarticular tubercles pale, concolor with palmar surface of hand. Toes moder- ately long, only very slightly webbed basally, 4-3-5-2-1 in order of decreasing length; subarticular tubercles prominent, concolor with plantar surface. Heels overlap strongly when femora held at right angles to body axis. Inguinal glands absent. Dorsum smooth. Throat and chest smooth, belly coarsely granular. Dorsal sur- face of fore- and hindlimbs smooth. Posterior and ventral faces of thighs covered with moderately sized, juxtaposed, flattened gran- ules. Tongue elongate and narrow, free and notched behind, its greatest width equal to about one-half of that of floor of mouth. Vomerine teeth in two very small patches, well enclosed within the median margins of the choanae, separated from them by a distance equal to the length of one vomerine series, the two patches separated by a distance equal to the length of one vomerine patch. Coloration and pattern of holotype: Dorsum (as preserved) pale tan, heavily stippled with brown on the trunk; head less densely stippled and with remnants of a pale interocular bar, outlined an- teriorly and posteriorly by remnants of brown lines; upper surfaces of hindlimbs concolor with dorsum; concealed surfaces of thigh, crus and pes paler (presumably brightly colored in life) without any dark markings or outlining in darker; groin spot also paler and without dark markings or outlining; forelimbs tan, concolor with dorsum. Venter pale (presumably brightly colored in life), without 1968 ELEUTHERODACTYLUS WETMOREI 11 heavy stippling; throat unstippled but with a pair of darker glan- dular areas anterior to the forelimb insertions. Variation: The single male E. w. williamsi has the following measurements: snout-vent length 32.7, head length 11.5, head width 13.1, tympanum 1.6, eye 4.0, naris to eye 3.9, femur 14.6, tibia 16.4, fourth toe 12.9; the series of 21 females has the following mensural data (extremes and means): snout-vent length 25.2-36.4 (32.7), head length 9.5-12.2 (1 1.5), head width 10.7-14.0 (13.1), tympanum 1.2-2.3 (1.8), eye 3.0-4.7 (3.9), naris to eye 3.1-4.5 (3.5), femur 11.2-15.2 (13.5), tibia 14.0-18.0 (15.7), fourth toe 9.1-14.0 (11.9). The paratypic series includes both long-preserved and freshly collected material; the former is the lot from Moron and the latter material is from the type locality and from near Jeremie. Many recent specimens have the more or less speckled dorsal aspect of the type, but this style of pattern does not occur in the Moron material, a fact which I assume is due to the length of time in preservation. The presence of the interocular bar is variable, but it is demonstrated in both old and new specimens; many frogs show this feature more diagrammatically than does the holotype. A median dorsal hairline occurs in some individuals (MCZ 37759, USNM 60629). There is never any indication of dark pattern on the concealed surfaces nor in the groin, these areas being quite pale; the freshly taken specimens still show a weak pinkish-orange wash on the thigh, so that I assume that these areas were quite brightly colored in life, presumably some shade of yellow or orange like the two other subspecies of E. wetmorei. Structurally, the paratypes agree very closely with the description of the holotype. Comparisons: E. \v. williamsi is easily distinguished from the nominate subspecies and E. w. ceraemerus since it lacks any indica- tion of thigh or groin pattern. As preserved, the specimens of williamsi are regularly paler than those of the other subspecies, and it seems possible that there is a dorsal chromatic difference between williamsi and the other subspecies. E. w. williamsi does not appear to reach so large a size as does ceraemerus but exceeds slightly the maximum size recorded for E. w. wetmorei. The absence of detailed color data in life for E. w. williamsi hampers more detailed comparisons. Remarks: E. w. williamsi occupies the extreme northwestern portion of the Tiburon Peninsula in Haiti. Its range, as now known, is circumscribed, with all localities being associated with the Monts Cartaches to the west of Jeremie. Altitudinally, E. w. williamsi 12 BREVIORA No. 290 occurs somewhat lower than the two more eastern subspecies, from a low elevation of about 130 feet (40 meters) at Moron to 1440 feet (440 meters) at Carrefour Sanon; Ferine is not locatable. The distinct patternless condition of the western population was first recognized by Cochran (1941:76), who pointed out that the series from Moron was light in color; later, Shreve and Williams (1963:324) again commented that the Moron series lacked limb and groin markings, in contrast to their specimen from Boutillier Road {ceraemerus) . The new series in the Harvard collection amply confirms these differences and demonstrates that they are not due to fading or the age of the previously known material from this region. I take great pleasure in naming the far western sub- species of E. wetmorei for Dr. Ernest E. Williams, whose interest in Hispaniolan herpetology, sponsorship of field work on that Antillean island, and generous cooperation with others involved with His- paniolan herpetological problems can be acknowledged only in a token fashion by the use of his name as a patronymic designation for this distinctive subspecies. DISCUSSION E. wetmorei is now known to be widely distributed along the length of the Tiburon Peninsula in Haiti; the altitudinal limits of the species indicate that it is an inhabitant of low to moderate elevations associated with the north and south flanks of the massifs de La Selle and de La Hotte. There is but one record from the Republica Dominicana, along the Dominico-Haitian border in the extreme eastern ranges of the Massif de La Selle. The species remains un- known from the Sierra de Baoruco proper. Doubtless many areas in Haiti whence E. wetmorei remains unknown will ultimately be found to be inhabited by this species; since the frog is predomi- nantly arboreal, it is difficult to collect, and often the species has been heard vocalizing in areas where specimens were impossible to secure. As in several other south island (sensu Williams, 1961) am- phibians and reptiles, E. wetmorei has differentiated into a series of three distinct populations which are easily recognizable. Although there is really too little material to be certain, it is interesting that the specimens from Fond des Negres agree with those from Camp Perrin on the southern slopes of the Massif de La Hotte farther to the west. Such a similarity is remarkably like that shown in Sphaerodactylus copei Steindachner, wherein material from the vicinity of Miragoane (and thus near the Fond des Negres) is like 1968 ELEUTHERODACTYLUS WETMOREl 13 5. c. cataplexis Schwartz and Thomas (at Camp Perrin) rather than like S. c. picturatus Carman (at Jeremie). Apparently the low and forested pass across the Massif de La Hotte in the Aquin-Fond des Negres-Miragoane region has allowed passage of some southern subspecies to the north coast, thereby more or less separating the two north coast populations {S. c. picturatus and S. c. copei sep- arated by S. c. cataplexis, just as E. w. williamsi and E. w. cerae- merus are separated by E. w. wetmorei; see Schwartz and Thomas, 1965, for details of gecko distributions in this region). Such similarities seem hardly due to chance. LITERATURE CITED Cochran, Doris M. 1932. A new frog, Eleiitherodactylus wetmorei, from the republic of Haiti. Proc. Biol. Soc. Washington, 45: 191-193. 1941. The herpetology of Hispaniola. Bull. U. S. Natl. Mus., 177:i-vii, 1-398, 120 figs, 12 pis. Maerz, a., and M. Rea Paul 1950. A dictionary of color. McGraw-Hill Book Co., pp. i-vii, 1-23, 137-208, 56 pis. Schwartz, Albert 1960. Nine new Cuban frogs of the genus Eleiitherodactylus. Reading Public Mus. and Art Gallery, Sci. Publ. 11:1 -50, 6 figs. (IN PRESS). The Antillean Eleiitherodactylus of the auriculatus group. Studies Fauna Curagao and other Caribbean Islands. and Richard Thomas 1965. Subspeciation in Sphaerodactyliis copei. Quart. Jour. Florida Acad. Sci., 27(4) :316-332, 1 pi. Shreve, Benjamin, and Ernest E. Williams 1963. The herpetology of the Port-au-Prince region and Gonave Island, Haiti. Pt. II. The frogs. Bull. Mus. Comp. Zool., 129(5) : 302-342, 5 pis. Williams, Ernest E. 1961. The evolution and relationships of the Anolis semilineatus group. Breviora, Mus. Comp. Zool., No. 136:1-7, map. (Received 9 November, 1967.) BREVIORA MitaseiLam of Comparative Zoology Cambridge, Mass. 29 May. 1968 Number 291 A NEW SPECIES OF TRIBOLONOTUS (LACERTILIA: SCINCIDAE) FROM BOUGAINVILLE AND BUKA, SOLOMON ISLANDS, WITH COMMENTS ON THE BIOLOGY OF THE GENUS Allen E. Greer and Fred Parker' INTRODUCTION The skinks of the genus Tribolonotus are undoubtedly one of the most bizzare taxa of lizards. Indeed, many of their anatomical fea- tures such as abdominal glands (Parker, 1940). volar pores (Roux, 1930 and 1934; Parker, 1940), and single right functional ovi- duct (discussed below) are as yet unreported in other lizards. Among skinks, these features, as well as their anomalous head and body squamation, make them one of the most easily distinguished groups in an otherwise taxonomically difficult family. In this paper we describe an unnamed species of the genus and discuss various aspects of the ecology, reproductive biology, and behavior of several species. DESCRIPTION OF A NEW SPECIES OF TRIBOLONOTUS During a recent period of residence (1962-1963) on Bougain- ville Island and a subsequent return trip (1966) to the northern Solomons. Parker collected a large number of specimens of a hitherto unnamed species of Tribolonotus on the islands of Bou- gainville and Buka. The new species is very similar in many details of squamation to T. ponceleti Kinghorn, 1937, which occurs on Bougainville and Shortland islands, and may therefore be known as ' Department of District Administration. Kundiawa. Territory of New Guinea, 2 BREVIORA No. 291 Tribolonotus PSEUDOPONCELETii new species Holotype: Museum of Comparative Zoology (MCZ) 1291 A, an adult male collected by Fred Parker at Kunua, Bougainville, on 30 December 1962. Paratypes (610 specimens) : BOUGAINVILLE: KUNUA: MCZ 12161-12119, 72781-72785, 72790-72791, 72793-72805, 72807- 72813, 72815-72829, 72831-72863, 72865-72866, 72868-72869. 72871-72873. 72875-72900, 72910-72912. 76148, 76155-76178, 76181-76191, 76193-76201. 76412-76469. 78094-78282. + 4 untagged specimens; American Museum of Natural History (AMNH) 92026-92058, + 10 untagged specimens; Australian Museum (AM) R 26603-26607; British^Museum (BM) 1963.578- 1963.584; Field Museum of Natural History (FMNH) 141982- 141985, 152596; University of Kansas Natural History Museum (UKNHM) 98510-98511; Zoological Museum of Berlin (ZMB) 39155 (2 specimens). BOKU: MCZ 65878-65880, 67723-67726; AMNH 89433. KIETA: MCZ 65875-65877, 67249-67250. MATSIOGU area: MCZ 92423-92426, 92428. 92468. 92492- 92495. MELILUP: MCZ 92429-92433. MUTAHI: MCZ 87615-87623. 88782-88794, 91467-91472, 92394-92422, 92427, 92434-92440. Above POPHEIARAI: MCZ 76179. RAMAZON RIVER (1600-2400 feet): MCZ 92502-92503. BUKA ISLAND: MCZ 67706-67716, 73850-73861; AMNH 89434. KUBA: MCZ 92491. Diagnosis: Pseiidoponceleti differs from blancliardi of the Sol- omons in having two rows of enlarged vertebral scales instead of a single row as in bkmchardi. Sclmiidti of the Solomons also has two rows of enlarged vertebral scales, but whereas these scales extend anteriorly to the single large parietal scale in schmidti, they only extend to a point just anterior to the level of the insertion of the forelegs in pseiidoponceleti. Size is the easiest way to dis- tinguish ponceleti from pseiidoponceleti. The three specimens of ^ The similarity in certain points of squamation between T. ponceleti and the new species, as well as their supposed sympatry (but see Ecology sec- tion of paper below), has caused the new species to be confused with ponceleti both in the field and in museum collections. Fortunately, this confusion has been carried over into the literature only once (Zweifel, 1966). To help clarify the taxonomic confusion, it seems best to take the path of least nomenclatural confusion and call the new species pseiido- ponceleti. 1968 NEW SPECIES OF TRIBOLONOTUS 3 poucelcti known in collections range from 116-125 mm in snout- vent length, whereas the largest of 611 pseudoponceleti examined is only 73 mm in snout-vent length. Gracilis and novaeguineae of New Guinea can be immediately distinguished from pseudoponceleti by means of the greatly en- larged spines on the two vertebral and paravertebral scale rows, and the head casque which, posteriorly, is raised above the level of the nape. Pseudoponceleti differs from annectens of New Britain (known from the holotype and more recently from a second specimen in the Zoologisk Museum, Copenhagen i) in several scale characters which are summarized here from Zweifel (1966; see Fig. 1). "In both species there is a prominent row of enlarged scales paralleling the much larger vertebrals, but where at midbody in pseudopon- celeti there is one of these smaller scales for each vertebral, an- nectens has only one for every two vertebrals. There are two primary temporal scales in pseudoponceleti and three in annectens." The first infralabial in pseudoponceleti is long and thin and extends posteriorly almost to exclude the second infralabial from the edge of the lip, or in a very few cases ( 1 specimen in 46 examined by Zweifel, 1966) the first labial does just exclude the entire second infralabial from the edge of the lip. In annectens the first infra- labial extends posteriorly to exclude the second (in the type) or the second and third (in the Copenhagen specimen) infralabials trom the edge of the lip.- Description (Fig. 2; also see fig. 1 in Zweifel, 1966): T. pseudoponceleti measures between 26 and 73 mm in snout-vent length, with the tail approximately 59 to 65 per cent of the total length. The pentadactyl limbs are well developed and overlap when adpressed to the body. The head is deep and rather triangular in shape when viewed from above. The longitudinal striations on the head, the keeled vertebral and caudal scales, and the enlarged scales on the sides and limbs give the animal a rather rugose appearance. 'The locality data for this second known specimen of annectens are as follows: Yalom, 35 km S.E. of Cape Lambert. Gazelle Peninsula, New Britain. Elevation 1000 meters above sea level. - As these infralabials are excluded from the edge of the lip, they might be called sublabials as Zweifel (1966) notes. We choose to call them in- fralabials here, however, to emphasize the proper homologous relationships. 4 BREVIORA No. 291 The rostral is approximately 2'/2 times as wide as deep. A very large frontonasal covers the dorsal surface of the snout, and pre- frontals are absent. The frontal is slightly shorter than the fronto- nasal and touches this scale throughout its entire anterior length. The frontal is in contact with the 2 anterior supraoculars laterally and the frontoparietals posteriorly. There are 4 supraoculars. The frontoparietals are distinct or fused to varying degrees. A distinct interparietal is lacking. The single large parietal, within which lies the parietal eye, is bounded on both sides by a smaller anterior and posterior parietal scale.' The nostril is contained in a single nasal. A large, roughly rec- tangular loreal extends from the anterior corner of the eye to the nasal and first supralabial. There are 5 supralabials; the first is extremely long and thin and extends posteriorly to the level of the anterior half of the eye, thereby excluding the anterior 2/5 of the large, rectangular second supralabial from the margin of the lip. A large 3rd and smaller 4th and 5th supralabial follow. Both the second and third supralabials border the smaller scales of the en- tirely scaly eyelid. There is a small scale above the 4th supralabial and two superposed scales are above the 5th supralabial. Posterior to these last 3 scales are the two temporal scales, of which the lower is about twice the size of the upper. The mental is followed on either side by 5 or 6 infralabials. The first of these is long and thin and extends posteriorly about as far as does the first supralabial above. The first infralabial excludes approximately Vi-Va of the second infralabial from the margin of the lip. A single large postmental is followed by a pair of very large chin scales which form a suture along the midline and are followed by a second part of chin scales of slightly smaller size. These chin scales are separated by 4 of the smaller gular scales. All the large head scales bear longitudinal ridges or keels. The tympanum is almost level with the surface of the head. Auricular lobes are lacking. ^ Zweifel's (1966) interpretation of the single large median parietal as possibly resulting from the fusion of the interparietal and parietals seems perfectly plausible to us. In all other lygosomine skinks the parietal eye occurs in an interparietal which would argue for the interparietal's being part of the single median parietal of Triholonotiis. Zweifel's observation of the incomplete posterior median suture in the large parietal can cer- tainly be seen in many pscitdoponceleti and again argues for the paired lygosomine parietals having become part of the single shield. 1968 NEW SPECIES OF TRIBOLONOTUS 5 There are 21-26 pairs of enlarged, keeled, vertebral scales be- tween the base of the tail and a point just anterior to the insertion of the forelimbs. Between each of these pairs of enlarged ver- tebrals there is usually a single small median scale. The side of the body between the large vertebral scales and the large ventral scales is covered with many small granular scales and fewer enlarged, almost tubercle-like scales di^sposed in roughly oblique rows. Granular scales predominate in an area adjacent to and as wide as the vertebral rows, except for a row of slightly enlarged, keeled scales that alternate with the vertebrals. The tubercle-like scales extend from the side of the body forward onto the side of the neck and dorsally onto the nape between the enlarged post-parietal scales and the anteriormost pair of enlarged vertebrals. The ventral scales are large and at the level of the abdominal glands are roughly disposed in tranverse rows 8 scales in breadth. Each ventral scale, except for those covering the abdominal glands, bears a small median keel and is drawn out to a point posteriorly. The scales covering the abdominal glands are smooth and lack the keel and posterior point. There are two large, keeled, preanal scales. The scales of the limbs are rather strongly keeled, the keels being drawn up into almost tubercle-like projections on the scales of the upper and outer surfaces of the hind leg. The scales of the tail are keeled and arranged in annuli. The digits are covered above by single scales throughout most of their entire length, although at their base there may be 1 or 2 paired scales (Group I of Brongersma, 1942). There are 20-29 (17 on one apparently normally developed toe; see Table I) subdigital lamellae on the 4th (longest) toe. The lamellae on the basal third of the toe consist of paired scales, one scale being large and cover- ing most of the subdigital surface and the other scale being small and laterally situated. The lamellae on the distal % of the toe are smooth and consist of single scales. Adult males have both palmar and plantar pores. The palmar pores are few in number (2-4), and are distributed in an arc on the anterior edge of the palm. The plantar pores (3-10) are dis- tributed in two rows across the sole: a longer series from the ex- treme base of the 4th toe and a shorter series from the base of the 3rd toe. In larger males there may be one or two pores on the base of the 5th toe. Females always lack palmar pores, and whereas all females have large plantar scales similar to those bearing pores in males, the 6 BREVIORA No. 291 pores are developed in only some females. When present in females the plantar pores are less well developed and less numerous (1-6) than in males. There is a pair of abdominal glands which are covered by 2 pairs of midventral scales lacking keels and the posterior median point. Color (Fig. 2): In preservative, specimens are generally brown above and on the sides and light yellowish brown below. Most individuals are uniformly dark above, but some show various degrees of light tan to whitish mottling on the nape. back, and upper surface of the base of the tail. The lighter mottled pattern tends to be bilaterally symmetrical on the nape and back and is often expressed as 1 or 2 transverse, posteriorly projecting chevrons on the dorsal base of the tail. Variation: There is a slioht difference in the dorsal color of the Bougainville and Buka specimens (Fig. 2). Pseudoponceleti from Bougainville tend to be darker brown dorsally with less light mottling than is found in Buka specimens. In Buka pseudoponce- leti the light dorsal color tends to be more pronounced, with a few individuals being predominately creamy tan on the body with some brown mottling. The general differences in color pattern between the Buka Island and Bougainville populations are correlated with differences in the distribution of certain scale counts between the two populations (Table 1). Using the coefficient of difference (CD.) statistic (Mayr, et ai, 1953) to determine the degree of joint non-overlap between the Bougainville and Buka populations in regard to these differences, the CD. was found to be lower in each case than the minimum CD. of 1.28 and 90 per cent joint non-overlap conven- tion commonly accepted for giving subspecific recognition to two populations. These differences do not. therefore, seem to warrant subspecific recognition for the Bougainville and Buka populations. Distribution (Fig. 1): Tribolonotus pseudoponceleti has been collected at several localities on Bougainville as well as on Buka, just north of Bougainville (Fig. 1). Altitudinally, the species is known from the coastal lowlands up to an elevation of about 4000 feet (in an area 7-8 miles east of Kunua. Bougainville). None of these skinks has ever been taken by Parker or his native collectors on the Buin Plain at the south end of Bougainville, however. In this region the species was only found from about 2000 feet and above in the mountains. 1968 NEW SPECIES OF TRIBOLONOTUS 6°S 7°S 155° E BUKA I 156° E 30 miles 6°S ' o ° SHORTLAND I S ^^Lofung I55°E MONO I cC^ 156° E Figure 1. Map of Bougainville and surrounding islands showing the known collecting localities for Triholonotiis pseudoponceleti (circles) and T. poiiceleti (triangles). 8 BREVIORA No. 291 It is curious that the species was not found on Shortiand in the Bougainville Straits, approximately 6.1 miles south of Bougain- ville. This smaller island is separated from Bougainville by a sea passage not more than 25 fathoms in depth and was probably connected with the mainland during the Pleistocene. Comparisons: As Zweifel (1966) has provided a detailed com- parison between cmnectens and pseudoponceleti (under the name ponceleti) as well as between cmnectens and other species of the genus, the comparisons in this section are limited to cmnectens and ponceleti and are intended to complement Zweifel's recent dis- cussion. In his diagnosis of cmnectens, Zweifel (1966) noted that the "ventral scales number 40 in T. cmnectens; 44 to 54 (mean 48.6) in 45 specimens of T. ponceleti." This particular difference be- tween cmnectens and pseudoponceleti (Zweifel's ponceleti) is, however, less clear cut with the additional data now at hand. There are, indeed, generally more gular and ventral scales' in Bougain- ville (whence came all but one of the specimens of pseudoponceleti (42-54, avg. = 48.4) than in annectens (40 in the type, 46 in the Copenhagen specimen; avg. = 43.0), but the Buka Island popula- tion of pseudoponceleti has a somewhat intermediate number of gular and ventral scales (42-49, avg. = 45.0). Until now ponceleti was known in the literature only from the single individual of the original description. Parker has, however, recently obtained two more specimens for the Museum of Com- parative Zoology, which have allowed us to make direct compari- sons between ponceleti and pseudoponceleti. Other than size (see Diagnosis) there are several subtle differ- ences in squamation between pseudoponceleti and ponceleti. In pseudoponceleti the first supralabial and infralabial, which are long and thin, extend posteriorly to at least the level of the anterior corner of eye, whereas in ponceleti the first supralabial and infra- labial extend posteriorly only to a point about half way between the snout and the anterior corner of the eye. The position of the first supralabial thus permits the second supralabial to form a suture with the nasal in ponceleti, but in pseudoponceleti the first ^ Counted along the ventral midline from the enlarged, paired chin scales to, but not including, the large, paired preanal scales (Zweifel's 1966 standardization). Presumably Zweifel's use of "ventral scales" in the preceding quote of this paragraph is meant to include the gular scales also. 1968 NEW SPECIES OF TRIBOLONOTUS 9 supralabial meets the loreal, separating the second supralabial from the nasal. In pseudoponceleti the enlarged vertebral scales become poorly defined just anterior to the level of the insertion of the fore- limbs, but in ponceleti the enlarged vertebral scales maintain their continuity farther on to the nape. The keels of the vertebral scales and two rows of tubercles just lateral to the vertebral scales are relatively larger and more pronounced in ponceleti than in pseud- oponceleti. The one or two scales situated in the dorsal midline between a preceding and following pair of enlarged vertebrals are also relatively larger in ponceleti than in pseudoponceleti. There are two rows of plantar pores in male pseudoponceleti, one extend- ing from the base of the 4th toe onto the sole and the other extending from the base of the 3rd toe onto the sole. In male ponceleti there is only one row of pores extending from the base of the 4th toe. Breeding: T. pseudoponceleti is oviparous and lays but one ob- long, leathery, shelled egg in a clutch. Furthermore, the left ovi- duct is reduced to a small, vestigial appendage of the cloaca leaving the right as the functioning oviduct. In only two of 101 females with oviducal and/or ovarian eggs collected at Kunua, Bougain- ville, was there a small yolky egg in the reduced left oviduct. These 2 eggs, however, did not appear to be viable. It is noteworthy that the left ovary is still functional. Ten of the 101 females with oviducal and/or ovarian eggs had a single large yolky egg in the left ovary, but in all 1 0 females the left ovary had shifted to the right side, where, presumably, the ovulated eggs would pass into the right oviduct, i.e., the only functional oviduct. In no case was a large ovarian egg found in the left ovary when it was in its normal position on the left side of the body cavity. These data would indicate that as yolking proceeds in a follicle of the left ovary, the ovary is displaced to the right side, perhaps as a result of "crowding" with other internal organs, e.g., the stomach which tends to lie more on the left side of the body. It would appear that ovulation of a second egg may occur just prior to or just after depositon of a preceding egg. This supposi- tion is based on observations of single oviducal eggs the same size as the largest ovarian egg in the right oviduct just anterior to a full term egg in the same oviduct. In other cases there is simply a large ovarian egg and no oviducal egg. Such a situation would per- tain if a full term oviducal egg had been laid just before ovulation of the second egg. The 101 females with large ovarian eggs and/or oviducal eggs ranged in snout-vent length from 44-58 mm (avg. = 51.8 mm). 10 BREVIORA No. 291 The smallest female contained a large, shelled oviducal egg. Given the large total sample size, it would appear that 44 mm is near the snout-vent length at which females first become capable of repro- ducing. Ecology: T. pseudoponceleti is sympatric with T. blanc/iardi on Bougainville, although the two species occur in rather different ecological situations. T. blanchardi is a montane species, and is probably to be found throughout the mountain ranges in the center of Bougainville. Specimens have been collected at 2000-3000 feet in the ranges north of Buin. Other specimens from the ranges east of Kunua were collected in river valleys from about 500-3000 feet. The habitat varies somewhat with altitude. In the lower altitudes where the ridges are still warm and comparatively dry, the species lives only in the steep sided and heavily shaded small creek valleys. At higher altitudes the species occurs farther from the waterways, but is still easier to collect along the creeks. The lizard does not live in the open in any environment, but prefers tall, shaded, primary forest and abundant moisture. In river and creek beds individuals are found under stones, logs and leaf rubbish at, and just above, water level, but only very occasionally may one be found moving on the ground. It often is found in company with Sphenomorphus cranei and S. concinnatus. T. pseudoponceleti has a greater altitudinal range (coast to above 4000 feet) than T. blanchardi, but it is found more frequently at lower altitudes. T. pseudoponceleti is terrestrial and very secretive, living in and under decayed logs in all its types of habitat. It is com- mon in swampy areas, and in some places is common on the plains where it is found in primary forest and in secondary growth. In addition, newly cleared gardens usually contain numbers of these skinks. The species likes moist conditions but not the very wet conditions preferred by blanchardi. In the mountains there is little syntopy between the two species, because while blanchardi is confined to creek beds initially, then the shaded slopes of the steep ridges, pseudoponceleti is to be found only on the sharp ridges where the scrub is more open and the ground drier and warmer. This restriction to the ridges has been confirmed by extensive collecting east of Kunua from 2000-4000 feet. A number of other coastal species which range to about 2000 feet in the mountains are also confined to the drier ridge tops. In the Boku area the natives believe pseudoponceleti prefers the hairy base of a large ground fern as a dwelling place, but this was not borne out by the breaking up of a number of these ferns in a 1968 NEW SPECIES OF TRIBOLONOTUS 11 hunt for the species. The habitat in this area was decaying wood on the ground. Whether T. ponceleti also occurs on Bougainville along with pseudoponceleti and blanchardi is uncertain. The single specimen from which the species was described (Kinghorn, 1937) was said to have come from "Buin, east Bougainville Island, Solomon Is- lands." However, there are no ponceleti among several hundred reptiles that Parker and his native collectors obtained in southern Bougainville near Buin. Upon inquiry, Mr. Harold Cogger of the Australian Museum found the notes that accompanied Father J. B. Poncelet's zoological collection of which the single specimen of ponceleti was a part. The notes consist mainly of a list of the common English and native names for each of the specimens and are prefaced by an unequivo- cal statement regarding the provenance of the collection — "All those {sic) specimens have been collected at Buin (South Bougain- ville), 10 to 15 miles in the interior, during the months of Septem- ber and October 1934, by Father J. B. Poncelet S.M., of the Buin Catholic Mission." Native collectors have obtained the only other two known ponce- leti at Lofung on Shortland Island (Fig. 1), but whether this large Tribolonotus also occurs on Bougainville is still an open question. The presence of ponceleti and the absence of pseudoponceleti on Shortland Island seems well established. It is also fairly clear that pseudoponceleti is absent from the Buin Plain. It is not certain whether ponceleti occurs on the Buin plain, although it is fairly well established that ponceleti does not occur elsewhere on Bougain- ville where large collections have been made (Fig. 1 ). Reasons for the absence of pseudoponceleti from the Buin plain and Shortland Island are obscure as are the reasons for the absence of ponceleti from the rest of Bougainville, if this species occurs on Bougainville at all. There seem to be few ecological differences between the Buin plain, the other lowland areas of Bougainville where pseudo- ponceleti has been collected, and Shortland Island (Parker, per- sonal observation). As no noticeable ecological differences are known to be corre- lated with the ranges of ponceleti and pseudoponceleti, the only likely remaining explanation for their allopatry is possibly compe- titive exclusion — perhaps of a very dynamic nature. But this facile explanation seems implausible, as the great size difference between the two species would certainly suggest a difference in the utilization of the habitat by the two species. 12 BREVIORA No. 291 On the basis of our present information there is Httle more that can be said beyond this description of the problem. The ecologic and geographic relationships of ponceleti and pseucloponcelcti will be interesting field problems for future workers on Bougainville. Little seems to have been published on the ecology or habits of the closely related species T. gracilis and T. novaeguineae, but Mr. Harold Cogger has written to us about his personal experience with T. gracilis. Mr. Cogger says, "I am familiar with this species only on Karkar Island, where I obtained approximately 90 specimens. All specimens collected by me, or in my presence, were found under decaying logs and other vegetation usually on the banks of small non perennial streams, up to an altitude of about 2,000 feet, usually in fairly dense primary or secondary forest." Behavior: T. pseudoponceleti is cryptic in its habits and never moves about in the open. No lizards were ever collected or seen in the open, and none was found moving about at night. In captivity they will come into the open parts of the cage to eat if there is little or no light. When found in its usual habitat under logs pseudopon- celeti makes no attempt to escape but may move away slowly if given time. When an individual is uncovered there seems to be an instinctive reaction to flatten on the ground to avoid being noticed. Usually two adults, a male and a female, are found together, and occasionally, if the log is large enough, another pair or two or three partly grown specimens may be close by. Rarely is a lone individ- ual found — a close search into the ground and log nearby usually reveals a second individual. T. pseudoponceleti is not strictly speaking a burrower but moves through insect burrows and hollows in and under logs. In captivity the species will eat termites. The lizards seem to be keen sighted and can detect movements in a dimly lit room some 10-15 feet away from the cage containing captive skinks. Other species of skinks of different genera seemed to be able to detect similar movements only when within 5-6 feet of the cage. T. blanchardi is very similar to pseudoponceleti in its behavior. It moves slowly, although when being chased it will make short bursts of speed to escape. In captivity it is not given to the quick, nervous movements found in other skinks, e.g., Emoia and Spheno- niorphus species, either when frightened or feeding. When moving, the skink holds the whole body and sometimes the tail off the ground. Virtually nothing is known of the ecology or habits of ponceleti. The two specimens of ponceleti collected for Parker by the natives at Lofuno, Shortland, had dirt in the crevices between the scales 1968 NEW SPECIES OF TRIBOLONOTUS 13 which may indicate that the species is something of a burrower. Many pseudoponceleti also have dirt between their scales when col- lected and this species is known to spend much of its time in close contact with the soil in secretive situations. VOLAR PORES AND ABDOMINAL GLANDS IN THE GENUS IRIBOLONOJUS Among lizards, volar pores and the dermal abdominal glands are unique to the genus Tiibolonotus. The presence of volar pores was first pointed out by Roux (1930) in male T. schmidti when he used them as a character diagnostic of a new genus — Pediporus. Roux (1934) later rectified this mistake when he found similar pores on some of the toes and sole of a male T. novacguineac the type species of the genus Tribolonotus. Parker (1940) extended Roux's observations on volar pores of gracilis and provided additional information on their distribution in schmidti and novaeguineae. Parker also provided the first account of abdominal glands in the genus, having noticed them in male gracilis and novaeguineae. Function: The function of both volar pores and abdominal glands is speculative. Secretion seems to be the primary purpose of both structures, and from the distribution of the glands on the body of the animal — belly and volar surfaces — it would seem as though the secretion would be placed in a new part of the microenviron- ment with every step taken by an animal. What, then, is the value of such a secretion? The only information we have on the ecology of Tribolonotus is that given above for blanchardi, pseudoponceleti, and gracilis. All three species appear to be highly secretive and spend most of their time under objects such as stones, leaf rubbish, and rotten logs — microhabitats that are both dark and relatively moist, and if the secretions are volatile and odiferous, they might be important olfac- tory signals. Furthermore, the fact that both structures are generallv better developed in adults than juveniles and in males rather than females suggests that their function may have some sexual significance. For instance, the secretions provided by the pores and glands may be olfactory markers of territory. Distribution of volar pores in Tribolonotus (Table 2): The two known specimens of annectens are females and both lack volar pores. 14 BREVIORA No. 291 Plantar pores occur in both male and female pseudoponceleti where they are found in a series behind both the 3rd and 4th toe. In some males, but never in females, there may be an additional pore or two on the basal third of the 5th toe. Palmar pores are present in males but not in females. The three known specimens of ponceleti are males and all have a single series of plantar pores directly in line with the 4th toe (Fig. 3). Palmar pores are absent. Both gracilis and novaeguineae females apparently lack volar pores. Males of both species have pores on the basal half of the 3rd and 4th toes and on the sole in a line directly behind the 3rd toe. Male gracilis possess palmar pores (personal communication, Harold Cogger re Karkar Island gracilis), but whether palmar pores are present or absent in male novaeguineae is unclear as yet. Female blanchardi also lack volar pores. Males have both palmar and plantar pores which are distributed mainly on the palm and sole, although occasionally a pore is found on the basal third of one of the digits. In sclunidti, females lack volar pores, but males have both palmar and plantar pores, the latter being distributed in two series: one behind the 4th toe and a second, shorter, series behind the 3rd toe. Distribution of abdominal glands in Tribolonotus (Table 2): In the Copenhagen specimen of the female annectens which we have examined, the glands are covered by four pairs of median ventral scales. Both sexes of pseudoponceleti have a single pair of unlobed ab- dominal glands lying beneath two pairs of midventral scales. The glands are slightly smaller in females than in males. In ponceleti, which is at present known from only 3 males, the abdominal glands are bilobed and covered by one or two pairs of ventral scales (Fig. 3). Female gracilis and novaeguineae lack abdominal glands, but in males of both species the glands are paired and unlobed and extend a distance of three or sometimes four ventral scales along the mid- line. Female blanchardi also lack abdominal glands. Males have four pairs of bilobed abdominal glands which extend the length of four or five ventral scales. The glands are very small and minutely lobed in female schmidti, but the ventral scales covering the glands are not visibly different from the other ventral scales. In males the two pairs of ventral scales covering the glands are characteristically smooth along their posterior edges rather than mucronate as are the surrounding en- larsed ventral scales. 1968 NEW SPECIES OF TRIBOLONOTUS 15 REPRODUCTION IN TRIBOLONOTUS The mode of reproduction is known for all the species of Tribo- lonotus except ponceleti which is known from only 3 male speci- mens. From an examination of gravid cmncctens (also see Zweifel. 1966:2), blanchardi , gracilis and novaeguineae, as well as pseudo- ponceleti, it would appear that these species are oviparous and pro- duce but a single oblong leather shelled egg in a clutch in which little or no embryonic development takes place prior to deposition. Schniidti is atypical of the genus in producing living young but typical in producing but one young at a time. Roux (1930:133) says of a single female he received from Guadalcanal, Solomon Islands, "Elle renferme un embryon bien developpe." The adap- tive significance of live-bearing habits in schniidti are difficult to imagine and will probably remain obscure until more is known of the species' behavior and ecology. It is interesting to note that blanchardi and schniidti are similar to pseudoponceleti in having only one functional oviduct, the right. In blanchardi and schniidti as in pseudoponceleti the left oviduct is reduced to a small vestigial appendage of the cloaca. In blanchardi the left ovary is still functional, and, as in pseudopon- celeti, is usually displaced to the right side prior to ovulation. In schniidti, on the other hand, the left ovary is either absent or greatly reduced in size; in the latter case it is situated well posteriorly in the body cavity and is probably not functional. Unfortunately the state of preservation of the four gravid gracilis and single gravid novaeguineae available for study was such that the condition of the left ovary and oviduct could not be interpreted. These data are of further interest on two counts. First, it may well be that a single egg or young is characteristic of the taxon Tribolonotus. Such specificity in the number of eggs produced is not unusual in some groups, e.g. geckos, but it is unreported in any supraspecific taxa of skinks.^ Second, a single functional oviduct has been previously unknown in reptiles except in some leptotyphlopids and typhlopids (Guibe, 1948; Robb, 1960; Fox and Dessauer, 1962). In these "worm snakes," as in Tribolonotus, it is the left oviduct that is non-func- tional. The left ovary in the leptotyphlopids and typhlopids is slighdy smaller in size than the right ovary, although it is still 'Many Emoia, however, especially the smaller species, have but two eggs in a clutch (Greer, 1968). 16 BREVIORA No. 291 functional. The left ovary seems to have lost all function in Tribo- lonotus sclunidti, but the large ova discovered in the left ovary of a few individuals of both blanchardi and pseudoponceleti would indicate that this organ is in some cases still functional in these species. ACKNOWLEDGMENTS We would like to thank Mr. Harold Cogger of the Australian Museum for his many contributions to the development of this man- uscript. Not only did Mr. Cogger supply us with information on the type of ponceleti and his field experiences with gracilis, he also made many helpful criticisms and corrected several errors in an earlier draft of the paper. Dr. E. E. Williams of the Museum of Comparative Zoology has also read the manuscript in several versions and has smoothed over many "rough spots" with each reading. Part of the research (done by Greer) for this paper was sup- ported by the Evolutionary Biology Fund which is administered for the National Science Foundation by the Evolutionary Biology Committee of the Biological Laboratories, Harvard University. Dr. Richard G. Zweifel of the American Museum of Natural His- tory very kindly read the final draft of the manuscript, and Miss Nancy Uhlar typed the final version of the paper. Publication has been financed by NSF Grant GB 6944 to Dr. E. E. Williams. A part of Parker's collecting expenses during his 1966 field work m the Solomon Islands was covered by a grant from the Science and Industry Endowment Fund of the Commonwealth Scientific and Industrial Research Organization of Australia. LITERATURE CITED Brongersma, L. D. 1942. On the arrangement of the scales on the dorsal surface of the digits in L\gosonui and allied genera. Zoologische Mededeelin- gen. 24 ( 1-2) : 152-158. Fox. W. and H. C. Dessauer 1962. The single right oviduct and other urogenital structures of female Tvplilops and Leptotyphlops. Copeia (1962) No. 3: 590-597. Greer, A. E. 1968. Clutch size in the genus Einoia. Copeia (1968), in press. 1968 NEW SPECIES OF TRIBOLONOTUS 17 GUIBE, J. 1948. Contribution a Fetude de Fappareil genital des typhlopides (Ophidiens). Bull. Soc. Zool. France, 73: 224-228. KiNGHORN, J. R. 1937. A new species of scink from the Solomon Islands. Rec. Australian Mus., 20(1 ) : 1-2. Mayr, E., E. G. Linsley, and R. L. Usinger 1953. Methods and Principles of Systematic Zoology. McGraw Hill Book Co., Inc., New York, 336 pp. Parker. H. W. 1940. Undescribed anatomical structures and new species of rep- tiles and amphibians. Ann. Mag. Nat. Hist., (11) 5: 257-274. ROBB. J. 1960. The internal anatomy of Typhlops Schneider (Reptilia). Australian J. Zool.. 8(2): 181-216. Roux, J. 1930. Note sur un reptile scincide des lies Salomon presentant des pres pediaux. Verhandl. Naturforsch. Ges. Basel, 41: 129-135. 1934. Contribution a la connaissance de la faune erpetologique des lies Salomon. Verhandl. Naturforsch. Ges. Basel, 45: 77-81. ZWEIFEL, R. G. 1966. A new lizard of the genus Triholonotiis (Scincidae) from New Britain. Amer. Mus. Novitates, No. 2264: 1-12. (Received 21 November 1967.) 18 BREVIORA No. 291 TABLE 1 Frequency distribution for several meristic characters in the two island populations of Tribolnotus pseiidoponceleti BOUGAINVILLE Enlarged, paired middorsal scales X 21 22 23 24 25 f(x) 1 3 32 64 13 Gular + ventral scales X 45 46 47 48 49 50 51 52 53 54 f(x) 8 8 17 26 22 10 9 1 3 2 Subdigital lamellae 4th toe X 17 20 21 22 23 24 25 26 27 28 29 N = 211 x = 23.9 f(x) 1 2 18 24 44 41 44 18 15 3 1 cr = 1.83 BUKA ISLAND Enlarged, paired and middorsal scales f(x) f(x) 6 1 N - 114 x = 23.8 a - 0.75 N= 106 x = 48.4 <7= 1.97 f(x) 22 23 24 25 N=:22 x = 23.8 1 6 12 3 Gular + ventral scales S 1 \D PUERTO RICO ^ ijL.^-vSS^ Vx-x. ^>-/-^— ^^ m^^\\) Figure 2. Localities on Puerto Rico sampled for A. c. cristatellus (circles) and A. cooki (squares). The number within each symbol is the number of specimens sampled for that locality. 968 ANOLIS CRISTATELLUS CAYO ICACOS ( 2 1 CAYO LOBOS ( 4 ' JSLETA MARINA '( I ) ( I ) HANS LOLLIK . / J>^^ VIRGIN 'GORDA V ^-""^ BT' ST. THOMAS ( I ) VIEQUES Figure 3. Map of the eastern portion of the Puerto Rico bank; islands where specimens were sampled are named. Number in parentheses is the number of specimens from each locahty. Figure 4. Map of the southern Bahamas and eastern Greater Antilles. The area enclosed in solid black represents the distribution of A. scriptiis. Inagua and Mayaguana are the localities sampled. Dotted areas represent submerged island banks. Arrows show general trend of the currents. BREVIORA No. 293 RESULTS Male meiosis. The four species in the cristatellus group previ- ously studied (gundlachi, pulchellus, poncensis, krugi, see Gorman and Atkins, in press) are all very similar to one another. In diaki- nesis there are six large bivalents, two intermediate-sized bivalents, five small bivalents (13 autosomal bivalents in all), and a sex trivalent. Figure 5a shows this in A. gundlachi. Among the forms reported on here, only A. cooki has 13 autosomal bivalents and a sex trivalent. A. cooki differs from gundlachi and pulchellus only in the detail that the smallest bivalents do not break sharply in size when compared with the intermediates (see Figure 5b). All A . c. cristatellus and both the scriptus subspecies differ from cooki and the other members of the cristatellus series by having one fewer autosomal bivalent. Again there are six macrobivalents; however, there are only six smaller ones, which apparently can be divided into two classes of three intermediate-sized and three small, but certainly with gradation (Figs. 5c, d). It is among the offshore populations of A. cristatellus wileyae that we have made the most unusual finding. Twelve of the animals sampled appeared identical to the Puerto Rican cristatellus and Bahaman scriptus, but two others differ in having a moderate-sized body that might be a univalent. One of these animals is from the small key Cayo Lobos, the second is from Vieques. Three of the four males sampled from each of the two islands had normal meiosis (Fig. 6). Mitosis. The mitotic metaphase of males of cooki is characterized by 29 chromosomes. There are six pairs of metacentric macro- chromosomes and 1 7 smaller elements. Resolution of these ele- ments is difficult, but there appear to be three pairs of small meta- centric chromosomes, four dot-like pairs (presumably acrocentric), and three unpaired chromosomes, the largest of which is clearly metacentric and is presumably the Y (Fig. 7b). Females have 30 chromosomes and lack the unpaired metacentric. The male karyo- type is similar to pulchellus (Fig. 7a), but again, as in meiosis, the break between size classes of chromosomes is greater in pul- chellus. There are six pairs of metacentric macrochromosomes, two intermediate pairs of metacentric chromosomes, five pairs of micro- chromosomes, and, in the male, three unpaired chromosomes (2n=29). A. scriptus and A. c. cristatellus have male diploid numbers of 27; female scriptus have 2nr=28 (female cristatellus were not studied). The male diploid number of 27 is also found m A. c. wileyae (one exception will be discussed below). 1968 ANOLIS CRISTATELLUS Figure 5. Meiosis in cristatellus group anoles. Giemsa stain. A. A. ^iiiidUichi. There are six large hivalents. two of intermediate size, and five distinctly smaller bivalents; a sex trivalent is on the far right. B. A. cooki. There are six large bivalents. seven ranging from interme- diate to small size with no sharp break between them; and a sex trivalent. C. A. c. cristatellus. There are six large bivalents, and only six ranging from intermediate to small size with no sharp break between them: and a sex trivalent. D. A. scripliis Icucophaeiis. Quite similar to cristatellus directly above. In details of karyotype, cristatellus and scriptus are very similar. There are six pairs of metacentric macrochromosomes and six autosomal pairs of small chromosomes, of which the largest three pairs are also metacentric. Males have three unpaired sex chromo- somes (Figs. 7c, d). The one exceptional animal was the A. c. wileyae from Vieques which had the extra body (presumed univalent) in meiosis (no mitotic divisions were seen in the specimen from Cayo Lobos that 8 BREVIORA No. 293 Figure 6. Meiosis in A. cnstatelliis wileyae. Giemsa stain. A and B. Two different males from the island of Vieques. Male A is identical to typical crisicitellus, male B has a dark staining supernumerary body (arrow). C. A male from Cayo Lobos that also shows a dark staining supernu- merary body (arrow). D. A male from St. Thomas, Virgin Islands, which has the typical cristate'lliis complement. had the same meiosis). Only 10 mitotic spreads could be counted from the testis of this animal, of which 4 had 27 chromosomes (the same as other cristatellus) and 6 had 28, or one extra chromo- some. The extra chromosome appears to be a microchromosome. Possibly, the animal is a 27/28 mosaic. However, interpretation is difficult. Because of the small size of the microchromosomes, it is easy to have one masked by a large arm, and it is indeed possible that the real diploid number is 28. Ten counts are far too few to establish the karyotype with certainty. It should be emphasized that the two chromosomally aberrant wileyae did not differ mor- phologically from the other lizards sampled. [968 ANOLIS CRISTATELLUS imi"'"' 9Sli# •!»• • •■ • fl B m^ Jlii ^tS m ii ♦,* • « • * * c ll K H « « <«►«» .^ « • • m^ ^ D Figure 7. Mitosis in males of ciistatelliis group anoles. Giemsa stain. A. Aiiolis pulclieUits (2n=29). There are six pairs of metacentric mac- rochromosomes (top row), two pairs of small metacentric chrcmoscmes (far left of second row), a sharp break between these and the five pairs of microchromosomes, and three unpaired sex chromosomes. B. A. cooki (2n=:29). Macrochromosomes as above (top row). Pairs seven and eight, small metacentrics, are as in piilchelliis, but there is no sharp break between pair eight and pair nine. There are again three un- paired sex chromosomes. C. A. ciistatelliis wileyae (2n=27). Macrochromosomes as in the above species. Microchromosomes most similar to cooki. but one pair fewer. Sex chromosomes similar. D. A. SLiiptiis leiicophacns (2n=27). Resembles c;/.s7{(/(7////s. 10 BREVIORA No. 293 DISCUSSION The chromosomal data support the hypothesis that cooki is a species distinct from cristatellus. This is in accord with observations made by Thomas in the field. There are two alternative explanations which may account for the observed chromosomal complements of these two species and their present-day distribution: \. A. cristatellus has the lowest diploid number in its species group. Reduction in chromosome number is usually the derived condition and. if this be so in this case, cristatellus is a derived form. Chromosomal heterozygotes experience meiotic difficulties, lead- ing to reduced fitness; the resultant structural mutants would be swamped in large populations. Chromosome structural changes, therefore, are probably established in small isolated populations. Thus it is logical to assume that the stock leading to cristatellus evolved in the limited population of a small island off Puerto Rico. Once chromosome loss had been established, cristatellus reinvaded the main island, successfully filling the lowland arboreal niche throughout Puerto Rico. A. cooki has a very restricted range (southwestern Puerto Rico) and is specialized for a very dry rocky area. One hypothesis ac- counting for its origin is that it may have been isolated from its ancestral stock on an insular region corresponding to the present coastal hills of southwestern Puerto Rico. This stock, quite pos- sibly the same one that gave rise to cristatellus (for the two species are remarkably similar in appearance), might have had the prim- itive karyotype for the species group. The diploid number of cooki (29) is the same as that of gund- lachi and the grass anoles of the cristatellus series, but there are some differences in the details of karyotype. The six pairs of macro- chromosomes are comparable in all the forms. In cooki there is then a gradation from pairs 7 to 13. of which the largest three are metacentric (see Fig. 7), while in the grass anoles and gundlachi pairs 7 and 8 are medium-sized metacentrics which break sharply in size with the smaller pairs (9-13). With respect to the microchromosomes, cristatellus resembles cooki in that pairs 7, 8, and 9 are all metacentric and grade in size into the smaller elements. The major difference, of course, is that cristatellus has one fewer pair of microchromosomes. Hence we might postulate that cooki, although morphologically specialized, retains the primitive karyotype. A. cristatellus, then, is similar but has suffered chromosome loss, while A. gundlachi and the three grass anole species are the products of a single 1968 ANOLIS CRISTATELLUS 11 radiation in which a new karyotype was established by reciprocal translocations from a species with the primitive (cooki-Uke) karyo- type. 2. An alternative hypothesis would consider the gundlachi-type of karyotype primitive for the species group. A. cristatellus would again have been derived from a 2n=29 ancestor by chromosome loss, followed by minor rearrangements. The apparently closely related A. cooki might then have been directly derived from cris- tatellus by centric fission, which secondarily raised the male diploid number back to 29. Karyotypic data, in any case, show that giindlachi is closer to the grass anoles than it is to cristatellus. This is not obvious from external morphology but is in accord with serum protein data (Maldonado and Ortiz, 1966) and osteological data (Etheridge, 1960). In karyotype, gundlachi is virtually indistinguishable from the grass anoles. It is a further point of interest that the chromosomal evidence confirms the close relationship of scriptus and cristatellus. No other anoles of the 70 species and subspecies that we have studied have a male diploid number of 27, and even details of karyotype are identical. From a glance at a map, one misht expect that the southern Bahamas would be colonized from Hispaniola rather than Puerto Rico. However, as E. E. Williams pointed out to us. the general current flow is northwesterly, and there are several barely sub- merged banks between Puerto Rico and the southern Bahamas, banks which may have served as stepping stones for colonization. It is therefore not surprising to find a Puerto Rican derivative on these islands (see Fig. 4). The widespread Hispaniolan member of the cristatellus group, A. cy botes, has a diploid number of 36 in both males and females with no heteromorphism and could not possibly be the direct ancestor of scriptus. The finding of apparent supernumerary chromosomes in two island populations of A. c. wileyae is of cytological interest, but lack of data at the present time on the extent of this variation within the populations and on the behavior of the supernumerary precludes discussion. ACKNOWLEDGMENTS We thank Dr. E. E. Williams for his constant interest and advice during the course of this work. Mr. W. P. Hall, III, has carefully read the manuscript and provided valuable criticism. Dr. Albert 12 BREVIORA No. 293 Schwartz supported field work in Puerto Rico done by himself and Thomas jointly and by Thomas alone. Specimens from the small keys ofT Puerto Rico were obtained through the aid of Drs. H. Heatwole and R. Levins, and Mr. F. MacKenzie, in conjunction with their work supported by NSF Grant GB-2906. Specimens were collected by the senior author on field trips sponsored by NSF Grant GB-2444 to Dr. E. E. Williams and the Evolutionary Biology Committee of Harvard University. Additional specimens were sent by Mr. A. Laska. Mr. R. McAndless, Mr. D. Norton, Mr. C. R. Warren, and Miss A. Swidler. Laboratory work was sponsored by Children's Bureau Project No. 906. We thank Mrs. C. Kayavas and Miss L Leone for technical assistance, and Mrs. P. Kerfoot for the illustrations. Additional support was provided by NSF Grant GB-6944 to Dr. E. E. Williams. LITERATURE CITED 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. Etheridge, R. 1960. The relationships of the anoles (Reptilia, Sauria, Iguanidae): an interpretation based on skeletal morphology. Ph.D. thesis. University of Michigan. University Microfilms. Inc., Ann Ar- bor, Michigan. Gorman, G. C. and L. Atkins 1966. Chromosomal heteromorphism in some male lizards of the genus Anolis. Amer. Nat., 100: 579-583. In press. The zoogeography of Lesser Antillean AnoUs. An inter- pretation based upon chromosomes and lactic dehydrogenases. Bull. Mus. Comp. Zool. Gorman, G. C, L. Atkins, and T. Holzinger 1967. New karyotypic data on 15 genera of lizards in the family Iguanidae, with a discussion of taxonomic and cytological implications. Cytogenetics. 6: 286-299. Grant, C. 1931. A new species and two new subspecies of the genus Anolis. J. Dept. Agriculture, Puerto Rico, 15: 219-222. Maldonado, a. a., and E. Ortiz 1966. Electrophoretic patterns of serum proteins of some West Indian Anolis. Copeia, 1966: 179-182. 1968 ANOLIS CRISTATELLUS 13 MooRHEAD, p. S., P. C. NowELL, W. J. Mellman, D. M. Battips, and D. A. HUNGERFORD 1960. Chromosome preparations of leukocytes cultured from human peripheral blood. Exp. Cell Res., 20: 613-616. Rand, A. S. 1962. Anolis scriptiis Garman 1887, an earlier name for Aiiolis leiico- phaeiis Garman 1888. Breviora, Mus. Comp. Zool., No. 153: 1-5. 1964. Ecological distribution in anoline lizards of Puerto Rico. Ecol- ogy, 45: 745-752. Thomas, R. 1966. Additional notes on the amphisbaenids of greater Puerto Rico. Breviora, Mus. Comp. Zool., No. 249: 1-23. (Received 21 November, 1967.) BREVIORA Museiuiinn of Comnparative Zoology Cambridge, Mass. 29 May, 1968 Number 294 DISTRIBUTION AND BIOLOGY OF THE OPISTHOPROCTID FISH WINJERIA TELESCOPA BRAUER 190P Richard L. Haedrich and James E. Craddock^ INTRODUCTION First taken by the German VALDIVIA Expedition in 1898, Winteria telescopa Brauer 1901 remained a unique oddity until Marshall (1960) reported a second specimen, Blache (1963) a third and fourth, and Bertelsen and Munk (1964) a fifth. Recent midwater trawl cruises, in particular those of the Woods Hole Oceanographic Institution in the Atlantic and of the ANTON BRUUN in the Indian and Eastern Pacific Oceans, have added sis- nificantly to this number. Twenty-three specimens are known, of which 18 have not been reported previously. This note records the specimens, compares them taxonomically, and offers some ob- servations on the distribution and natural history of Winteria. ACKNOWLEDGMENTS For access to specimens we extend our thanks to E. Bertel- sen, Carlsberg Foundation's DANA Expedition, Charlottenlund (DANA); Robert J. Lavenberg, Los Angeles County Museum, Los Angeles (LACM); Giles W. Mead, Museum of Comparative Zoology, Harvard (MCZ); Richard H. Rosenblatt and Robert L. Wisner, Scripps Institution of Oceanography, La Jolla (SIO); Frederick H. Berry, Tropical Atlantic Biological Laboratory, Bur- eau of Commercial Fisheries, Miami (TABL); C. Richard Robins, Institute of Marine Science, University of Miami, Miami (UMML); 1 Contribution No. 2031 from the Woods Hole Oceanographic Institution, and Papers from the "Dana" Oceanographical Collections No. 66. - Woods Hole Oceanographic Institution, Woods Hole, Mass., and Museum of Comparative Zoology, Harvard University. BREVIORA No. 294 5 mm I 1 Figure 1. A young Wintcria telescopa. 25.5 mm in standard length, from the Eastern Pacific (MCZ — 33°3rS, 77°29'W). 5 mm Figure 2. Head of an adult male Winteria telescopa, 91.5 mm in standard length, from the Eastern Pacific (USNM — 32"02'S, 73°48'W). Note the retinal diverticulum on the eye. The pinniform structure below the eye is a muscle, probably the adductor mandibulae, seen through the thin cover- ing bone. Drawn by Martha M. Howbert. 1968 NOTES ON WINTERIA 3 and Robert H. Gibbs, Jr., United States National Museum, Wash- ington, D. C. (USNM). Don Dockires, SIO, provided informa- tion on the sex of two Scripps specimens. The manuscript was read by Richard H. Backus and Giles W. Mead. Financial support has been provided in part by a United States Government Grant under the Fulbright-Hays Act to the senior author, and by the National Science Foundation through its support of the U. S. Pro- gram in Biology of the International Indian Ocean Expedition (IIOE) and the Southeastern Pacific Biological Oceanographic Program, and its grants GB-4424 (principal investigator Giles W. Mead) and GB-4431 (principal investigator Richard H. Backus) to the Woods Hole Oceanographic Institution. MATERIAL EXAMINED The following list of material, by ocean, indicates source institu- tion (many specimens are as yet uncatalogued and hence have no catalog numbers), number of specimens, sex (if known), standard length in mm, and (in parentheses) vessel, cruise, station number or numbers, position, date, primary depths fished (no closing gear was used), and gear (IKMT = Isaacs-Kidd midwater trawl): Atlantic — MCZ, 1 spec, $ , 55 mm (CHAIN, Cruise 35, RHB- 962, 5°24'N, 39°55'W, 13 Feb. 1963, 510-860 m, 10' IKMT); MCZ, 1 spec, 9 , 83 mm (CHAIN, Cruise 35, RHB-976, 0°03'N, 27°3rW, 26 Feb. 1963, 565-675 m, 10' IKMT); MCZ. 1 spec, 9, 98.5 mm (ATLANTIS II, Cruise 20, RHB-1206, 11°01'N, 34°18'W, 23 Feb. 1966, 430-490 m, 10' IKMT); TABL 14, 1 spec, 86 mm (GERONIMO. Cruise 2, sta. 82, 3°28'S, 01°14'W, 6 Aug. 1963); UMML 19998, 1 spec, 50 mm (PILLSBURY, sta. 295, 0°25'N, 5°09'E, 23 May 1965, 850 m, 10' IKMT). Indian — DANA, 1 spec, 100 mm (DANA. sta. 3847(3), 12°02'S, 96°43'E, 11 Oct. 1929, 2500 mwo, 300 cm Ring-trawl); MCZ, 1 spec, 9 , 1 15 mm (ANTON BRUUN, Cruise 3, Trawl 10 (AE-14), 2°06'S, 60^02'E, 21 Aug. 1963, 1600 m, 10' IKMT); MCZ, 1 spec, $, 130 mm (ANTON BRUUN, Cruise 3, Trawl 11 (AE-15), 5°03'S, 63°10'E, 23 Aug. 1963, 685 m, 10' IKMT). Pacific — DANA, 1 spec, 82.5 mm (DANA, sta. 3716(3), 19°18.5'S, 120°13'E, 22 May 1929, 2000 mwo, 300 cm Ring- trawl); LACM, 1 spec, $ , 124 mm (ELTANIN, Cruise 14, biol. sta. 1186, 52°10'S, 159°21'W, 4 Aug. 1964, 895-1030 m, 10' IKMT); LACM, 1 spec, 9 , 147 mm (ELTANIN, Cruise 27, biol. sta. 1986, 45°33'S, 147°18'E, 26 Feb. 1967, 2840-3001 m, 5' 4 BREVIORA No. 294 Blake Trawl); MCZ, 1 spec, 25.5 mm (ANTON BRUUN, Cruise 13, collection 40, 33°3rS, 77°29'W, 28 Jan. 1966, 425-820 m, lO'IKMT); MCZ, 1 spec, 27 mm (ANTON BRUUN, Cruise 13, collection 43, 33°25'S, 77°38'W, 28-29 Jan. 1966, 100-380 m, 10' IKMT); MCZ, 1 spec, 69 mm (ANTON BRUUN, Cruise 13, col- lection 58, 33°42'S, 72°17'W, 3 Feb. 1966, 400 m, 10' IKMT); SIO, 1 spec, 33 mm (ANTON BRUUN, Cruise 12, SIO 65-664, 33°31'S, 75°18'W, 17-18 Dec 1965, 530 m, 10' IKMT); USNM, 1 spec, $, 91.5 mm (ANTON BRUUN, Cruise 14, sta. 550-A (RHG-66-9), 32°02'S, 73°48'W, 15 Feb. 1966, 200-500 m, 10' IKMT). Dr. Richard H. Rosenblatt has very kindly sent us counts and X-rays of three specimens in the collections at SIO. Pertinent data are: SIO 61-33, 1 spec, 70.5 mm (Monsoon Exped., sta. III-5, 10°39'S, 98°51'E, 22/23 Nov. 1960, 0-1500 m, 10' IKMT. Counts: D 7, A 7, Pect. 12-12, Pelv. 8-8, vert. 34); SIO 61-37, 2 spec, 121.3 and 144.7 mm (Monsoon Exped., sta. IV-19, 33°19'S, 72°34'E, 19 Dec. 1960, 0-2000 m, 10' IKMT. Counts: D 9 and 7, A 7 and 7, Pect. 12-13 and 12-13, Pelv. 9-9 and 8-9, Lat. line scales 35 and ca. 35, vert. 35 and 35, respectively). DESCRIPTION The following descriptive notes are provided to complement the characterizations of other opisthoproctids by Cohen (1964): Body cylindrical, plump, becoming somewhat compressed at caudal peduncle, not particularly elongate. Belly lacking a sole. A rectal bulb and reflector organ present (Bertelsen and Munk, 1964). Eyes tubular and directed anteriorly. Interorbital very narrow. A dark-colored retinal diverticulum located anterolaterally on each eye. Snout elongate and hyaline; nasal capsules easily discerned. Branchiostegal rays 3, short and blunt. Gill membranes united across the isthmus. Maxillaries very small, thin and scale- like. Premaxillaries appear to be absent. No apparent teeth on the dentary. Swimbladder present (Marshall, 1960). Dorsal adipose fin present. Dorsal, anal, and ventral fins posteriorly placed; rays perhaps prolonged. Pectoral fins on peduncles, horizontally ori- ented high on the sides. Anus between pelvic fins, well before anal fin. Scales very deciduous. Color in life deep blue-black with silvery overtones on the head. Fresh-caught specimens have a very plump body; preserved speci- mens are much shrunken in comparison. 1968 NOTES ON WINTERIA •li ■D o 3aNvy 9861 -nS -1019 c>Vr.-ao — >— CO E , . o o (tl-3V) 01 l^fJi CM.u^Sol-^■-^(^JO^uD^o 1 1 I log^o ir>Icoco<->co Q o'^ £-Nnnaa noinv -mr-r^vovDcj^^ ,,,,_^^ ^ , ^ z ,S "u WIAimWA csiooinr^'— csJO-— PO*o^i^cj> ir^oc\jc% t^ooa3CMO% _J ■S-S adiC^OlOH - « 9021 -ana \£) I CT> r^ <^ O o 0+* com>x}oDir>*3-(\r— i£)vOvouDu-)t i i i i miCDr-*o-)00 h- 02-11 SliNVIlV a> ro >i) r- ^O ^ x: SC-NIVHO ^j E en t} C -D a> o ui9UC ■o ■*-» VNva (. V 1 -§ (D 4J 8? ; U013D3L100 W> i £1- ■wnnaa noinv s- V o 3 ■4-» c- iO 296-8Ha s E S£-NIVHD OJ o. = ^ S62 -BIS lo 3 AMnasnid _o-5 '^-i n I. o 01 1-99-99 OIS a. o 4^ 21- •Nnnaa noinv 2 W - &e I I I I iD^oooO'3-iric\io<— >x>*^ ' ' ' ' \o CO txi m CO h- LD CO CT^ ^ P~-- O- £1 -Nnnaa noinv £i-Nnnaa noinv O) en .— q: c _j 6 BREVIORA No. 294 COMPARISON OF SPECIMENS Morphometric and meristic data for 17 specimens, including Brauer's holotype of Winteria telescopa, are presented in Table 1. The small range of variation in both proportions and counts is strong evidence that there is but one species represented, and thus all our material is referred to Winteria telescopa Brauer 1901. There is perhaps some geographical variation in the number of vertebrae. The range is 33-35 vertebrae in the Atlantic, 34-35 in the Indian, and 35-36 in the Pacific. No pattern of variation is evident in the numbers of fin rays. Little allometry occurs between 25 and 147 mm SL (Table 1). The only regular change is in the length of the head, which becomes relatively shorter with growth. This may be also true in the fins, although the rays are extremely fragile and are usually damaged. As has been observed in the young of other opisthoproctids (Co- hen, 1964), young Winteria (Fig. 1) somewhat resemble Rhyncho- hyalus, the genus linking the two divergent lines of the Opisthoproctidae. Characters found also in Rhynchohyalus are the elongate snout, the tubular eyes, the retinal diverticula ("orbi- tal light organs" of Cohen, 1964, see Bertelsen et al., 1965), the raised horizontal placement of the pectoral fins, and the insertion of the pelvic fins above the anus. With growth, the profile of the head in Winteria changes. The head becomes somewhat deeper in respect to its length, and the angle between the snout and the interorbital region is more pronounced (Fig. 2). This is in con- tradistinction to Brauer (1906: pi. I), who shows a smooth arc from the tip of the snout to the occiput. DISTRIBUTION Four of the previously known specimens of Winteria are from the tropical Atlantic off Africa (Brauer, 1901, holotype; Marshall, 1960, 1 spec; Blache, 1963, 2 spec), and have been referred by their recorders to Winteria telescopa. The fifth specimen was in- cidentally reported from the eastern Indian Ocean by Bertelsen and Munk (1964). The concern of their paper was anatomical, and they made no judgement as to the specfic status of their specimen, referring to it simply as Winteria. As shown above, all specimens so far known can be referred to Winteria telescopa Brauer 1901. The distribution of this species is circumglobal (Fig. 3). The shallowest possible depth-of -capture (ANTON BRUUN 13-43) was certainly greater than 200 meters; 1968 NOTES ON WINTERIA c CL, o (U a, •o u ■*— » o > u o u •5 c -a U Oh a 2 S a .2 T3 C £^ 3 O u< o OX) C/3 u 1) 3 lyi u^ CU a C3 U <*-! O i/> •n Uc C o o (> (U in O cu c ro U ki U, • IM u BREVIORA No. 294 c« C o C/5 ■:?o 3dnivd3di^3i en *-• re ^ V <4-l (l> o o •£ r- -a c o w il JO a fj o *-" k> o ^ r\ . *-• (U > -a g.S ao.S G3 — ; I ^' I ■*^ 1^ 6 3 (U ^ H D- re I/) OJ C = -B ON 1968 NOTES ON WINTERIA 9 the deepest tow (ELTANIN 1986) reached 3000 meters. The majority of captures probably occurred in the range 500-700 meters. Wiuteria then, like other opisthoproctids (Cohen, 1964), is a relatively deep-living fish. Hydrographic casts had been made nearby or were made in con- nection with nine of the trawls which took Winteria. The tem- perature-salinity relationships at these stations, from 200 meters to the maximum possible depth-of-capture, are plotted in Figure 4. With the exception of the two western Indian Ocean stations (IIOE 150, 151), none falls entirely within a discrete water mass. They occur, for the most part, in the transitional regions between water masses, areas where the productivity is probably higher than in the generally poorer waters of the central gyres. On the T-S curves of Figure 4, most probable depths-of-capture are indicated. When at surfaces (density isopleths dependent upon temperature and salinity) are superimposed on the T-S plot, these depths fall within a at range of 27.00 to 27.40. This suggests that the distribution of Winteria may be determined at least in part by density, a not unreasonable proposition. Pickford (1946) found that the distribution of Vampyroteuthis, a bathypelagic cephalopod, was restricted to the at range 27.40 to 27.80. REPRODUCTION The gonads could not be distinguished in fishes smaller than 50 mm SL. Fishes between 50 and 90 mm appeared immature, with the onset of maturity occurring at about 90 mm SL. Both males and females could be recognized, and the gonads appear to be normal. There is no indication of hermaphroditism or vivi- parity. The ovaries of the three large females examined each con- tained two size classes of eggs, a 0.5-1.2 mm diameter group and a 1.8-2.5 mm diameter group (Table 2). These two size classes Table 2. Standard length of female Winteria and numbers and sizes of eggs in one ovary. Std. length 98 mm 115 mm 147 mm N" small eggs 80 101 125 N*^ large eggs 59 102 171 Size range small eggs .8-1.1 mm .9-1.2 mm .5-1.0 mm Size range large eggs 1.8-2.2 mm 2.2-2.5 mm 2.0-2.5 mm 10 BREVIORA No. 294 are present in about equal numbers. The data suggest a direct relationship between size of the fish and the number of eggs. With the exception of the two Scripps specimens (SIO 61-37), all the Winteria were taken as single captures. The larger of the two fish which were taken together is a female containing large orange eggs; the smaller specimen appears to be a ripe or near- ripe male. Winteria is probably a solitary fish, encountering an- other of its kind only occasionally and pairing when ready to spawn. FEEDING The stomachs of the six largest specimens were opened. One was empty; the rest contained a finely divided, light yellowish, pulpy material. No structure could be discerned in this material, and we assume it was not composed of crustacean remains. It most closely resembled the finely shredded tissues of siphonophores, medusae, and salps found by Haedrich (unpublished) in the stomachs of stromateoid fishes. LITERATURE CITED Bertelsen, E. and Ole Munk 1964. Rectal light organs in the argentinoid fishes Opisthopr actus and Winteria. Dana-Report No. 62: 1-21. Bertelsen, E., Birgit Theisen, and Ole Munk 1965. On a postlarval specimen, anal light organ, and tubular eyes of the argentinoid fish Rliyncholiyalus natalensis (Gilchrist and von Bonde). Vidensk. Medd. Dansk Naturh. Foren., 128: 357-371. Blache, J. 1963. Poissons bathypelagiques rares ou peu connus provenant des eaux de I'Atlantique oriental tropical. 1"" Note. Famille des Opisthoproctidae. Cahiers O.R.S.T.O.M., Oceanographie, No. 5: 89-95, 4 figs. Brauer, a. 1901. Ueber einige von der Valdivia-Expedition gesammelten Tiefsee- fische und ihre Augen. Sitzungsber. Gesell. Beford. Ges. Nat- urwiss. Marburg, 1901, No. 8: 115-130. 1906. Die Tiefsee-Fische. I. Systematischer Teil. Wiss. Ergebn. "Val- divia", 15: 1-420, 16 pi. Cohen, Daniel M. 1964. Suborder Argentinoidea in Fishes of the Western North Atlan- tic. Sears Found. Mar. Res., Mem. 1, part 4: 1-70, 20 figs. 1968 NOTES ON WINTERIA H Marshall. N. B. 1960. Swimbladder structure of deep-sea fishes in relation to their systematics and biology. D/icovery Reports, 31: 1-122. PicKFORD, Grace E. 1946. Vampyroteiithis infernalis CHUN. An archaic dibranchiate cephalopod. I. Natural history and distribution. Dana-Report No. 29: 1-40. SvERDRUP, H. U., Martin W. Johnson, and Richard H. Fleming 1942. The oceans. Prentice-Hall, New Jersey, 1087 pp. BREVIORA Mesemm of Contiparative Zoology Cambridge, Mass. 29 May, 1968 Number 295 THE CHANARES (ARGENTINA) TRIASSIC REPTILE FAUNA IV. THE DICYNODONT FAUNA C. Barry Cox^ The dicynodonts described in this paper were collected by the 1964-1965 expedition of the Museo de la Plata and the Museum of Comparative Zoology to the Permo-Triassic of western Argen- tina (Romer, 1966). Nearly all come from the Chanares Forma- tion, the lithology and stratigraphic relationships of which have been described by Romer and Jensen (1966). They state (p. 12) that, "With one exception, all of the numerous vertebrate fossils found in the Chaiiares were from the lowest 10 meters or so of the formation." Eighteen dicynodont specimens were found in this fossiliferous band. One of these belongs to a new genus and species of dicynodont, described here as Chanaha platyceps. Eight speci- mens belong to a new species closely allied to Dinodontosaurus turpior from the Santa Maria Formation of Brazil (Cox, 1965); it is here named Dinodontosaurus brevirostris. Another specimen belongs to a second new species of this genus, here named D. platy- gnathus. Finally, an isolated large scapulocoracoid and clavicle show the presence of another genus of dicynodont. The remaining seven specimens are incomplete or badly damaged and therefore cannot be assigned with certainty to any of the named taxa. In the accompanying figures, oblique shading indicates broken bone surface, horizontal shading indicates the presence of matrix, and broken lines indicate restored outlines of bone. The following abbreviations are used: MCZ — the Museum of Comparative Zo- ology, Harvard University; BYU — the Earth Sciences Museum, Brigham Young University, Provo, Utah; DGM — Divisao de Geologia e Mineralogia, Ministerio das Minas e Energia, Rio de Janeiro. 1 Zoology Dept., King's College, Strand, London W.C.2., England. 2 BREVIORA No. 295 Chanaria gen. nov. Type species: C. platyceps sp. nov. Diagnosis. Dicynodont of moderate size: skull length 33 cm. Teeth absent except for upper tusks. Greatest width of skull is across occipital region. Blunt anterior end to snout. Tusk pro- jects anteroventrally. Lacrimal extends forwards on face, prob- ably meets septomaxilla. No bony bosses on snout or on pineal region. Interorbital bar fairly wide. Frontal extends anteriorly into nasal as rectangular midline projection. Quite large pre- frontal. Preparietal present, forming anterior border of pineal foramen. Postorbital bone extends well behind pineal foramen. Intertemporal bar narrow but concave in cross-section, not ridged. Squamosal does not extend on to intertemporal bar, posterior end of which is formed by interparietal. Short, wide temporal opening. Low, wide occiput; supraoccipital does not extend high up in mid- line. Chanaria platyceps sp. nov. Holotype. No. 65-XI-14-3 Museo de la Plata, La Plata, Ar- gentina, skull. Collected by the 1964-1965 expedition of the Museo de la Plata and the Museum of Comparative Zoology. Horizon and locality. From an exposure in the Triassic Chan- ares Formation about 6 miles east of the point where the Chanares River emerges into the Campo de Talampaya, in western La Rioja Province, Argentina. Description. The type skull is 33 cm long and 30 cm wide. The mid-region of the palate, both quadrate regions, and part of the left side of the skull are missing. Dorsal view (Fig. 1). The premaxilla extends a considerable distance posteriorly in the midline. There is only a short midline nasal suture, for the posterior regions of the nasals are separated by a rectangular anterior projection of the frontals. There is quite a large prefrontal; the posterior end of its suture with the frontal is uncertain, owing to damage. The interorbital bar is of moderate breadth. The postorbital bone extends back to a point halfway between the pineal foramen and the posterior end of the intertemporal bar. The intertemporal bar is quite narrow, but it is slightly concave from side to side, without any midline ridge. The extreme pos- terior end of the bar is formed by the interparietal. 1968 CHANARES DICYNODONT FAUNA The occipital wing of the squamosal extends laterally, not pos- terolaterally. As a result, the posterior edge of the skull is almost straight and runs transversely, and the temporal opening is almost quadrilateral, its width approximately equalling its length. Occipital view (Fig. 2). The occiput is rather low, its midline region being little higher than its more lateral parts. Most of the midline surface above the foramen magnum is formed by the in- terparietal rather than by the supraoccipital. The bones of the central occipital plate are fused together. The quadrate region and outer edge of the lateral wing of the squamosal are missing, but the dotted outline shown is the minimum extent of squamosal necessary to accommodate a quadrate of normal size. Figure 1. Chanaria platyceps, type specimen. Dorsal view of skull, x V3 (for abbreviations, see p. 26). BREVIORA No. 295 Figure 2. Chanaria platyceps, type specimen. Occipital view of skull. X V3 (for abbreviations, see p. 26). Lateral view (Fig. 3). The powerful tusks project antero- ventrally. The squamosal runs forward to meet the maxilla under the front half of the orbit. A groove runs anteroventrally from the anterior corner of the nostril to the edge of the snout. A ridge which forms the lower edge of this groove also extends posteriorly along the maxilla be- hind the nostril. Above this ridge both the maxilla and the lacri- mal curve slightly inwards, so that there is no sharply defined pos- terior border to the nostril. The lacrimal extends a considerable distance anteriorly and, though its anterior end and most of the septomaxilla are missing, it seems very probable that these two bones met, separating the nasal from the maxilla. The palatal surface of the premaxilla bears the pair of anterior ridges and the median posterior ridge that are normal in the dicy- nodonts. The remainder of the ventral region of the skull is too incomplete to show any features of value. Taxonomic position. In an earlier paper (Cox, 1965) I sug- gested that two main families of Triassic dicynodonts could be distinguished on the basis of the shape of the snout, the presence or absence of a median crest along the intertemporal bar, and the proportions of the occiput and of the temporal opening. In Chan- aria the snout is wide and blunt, there is no median crest along 1968 CHANARES DICYNODONT FAUNA Figure 3. Clianaria platyceps, type specimen. Lateral view of skull, x V3 (for abbreviations, see p. 26). the intertemporal bar, the occiput is wide but low, and the tem- poral opening is short. In all these characteristics, Chanaria clearly belongs to the family Stahleckeriidae. Bonaparte (1966a) has since disagreed with my proposed classi- fication of the Triassic dicynodonts into separate families. He be- lieves that these forms are very uniform in structure, and that they are monophyletic in origin. Though I do not agree with either of these beliefs, Bonaparte's views and my own are, in the absence of knowledge of the Permian ancestry of the Triassic forms, sub- jective views of equal validity. The truth will emerge only from further work on the dicynodonts of the late Permian and early Triassic, and further discussion on these alternative classifications may be postponed until relevant material has been found and studied. DlNODONTOSAURUS BREVIROSTRIS Sp. HOV. Holotype. No. 65-XI-14-4 Museo de la Plata, La Plata, Argen- tina, skull. Collected by the 1964-1965 expedition of the Museo de la Plata and the Museum of Comparative Zoology. Horizon and locality. From an exposure in the Triassic Chan- ares Formation about 6 miles east of the point where the Chanares River emerges into the Campo de Talampaya, in western La Rioja Province, Argentina. BREVIORA No. 295 Referred specimens. Specimen Nos. 3452-3457, Museum of Comparative Zoology, Harvard; specimen No. MR- 120, Earth Sciences Museum, Brigham Young University, Provo, Utah. Figure 4. Dinodontosaiirus brevirostris, type specimen. Dorsal view of skull, X V3 (for abbreviations, see p. 26). Description. The skull of the type specimen, and those of speci- mens MCZ Nos. 3453 and 3457 and of specimen BYU No. MR- 120, are all about 30 cm long. Specimen MCZ No. 3454 is 23 cm long and appears to be a juvenile. Other dicynodont remains from the Chanares Formation which cannot be certainly identified but which probably belong to this, the commonest dicynodont in the fauna, show that specimens nearly twice as large as the type were present. 1968 CHANARES DICYNODONT FAUNA Figure 5. Dorsal views of pineal areas of specimens of Dinodontosaunis, reduced to equal size. A, D. turpior, MCZ No. 1628; B, D. turpior, MCZ No. 1687; C, D. brevirostris, MCZ No. 3454; D, D. brevirostris, type speci- men; E, D. turpior, DGM No. 530R; F, D. turpior, DGM No. 213R. 8 BREVIORA No. 295 Dorsal view (Fig. 4). D. brevirostris has the short temporal region and narrow, ridged intertemporal bar which are character- istic of the genus. However, there are some differences between D. brevirostris and the type species, D. lurpior, in the relation- ships of the postorbital to the pineal opening. The postorbital of D. brevirostris approaches quite closely to the side of the pineal opening, and also extends posteriorly well beyond it. Though there is considerable range of variation in these characteristics in D. turpior (Fig. 5), the postorbital in that species is not normally as extensive as that of D. brevirostris. Ventral view. The palate of the type specimen of D. brevirostris is too poorly preserved to show anything of value, but this area is well shown in the smaller specimen, MCZ No. 3454 (Fig. 6). The secondary palate extends posteriorly to a point level with the root of the canine tusks. The vomer and palatines do not form a bony Figure 6. Dinodontosaunis brevirostris, MCZ No. 3454. Ventral view, X 4/9 (for abbreviations, see p. 26). 1968 CHANARES DICYNODONT FAUNA roof to the anterior part of the primary palate, between the anterior rami of the pterygoids. On either side of the median blade of the vomer there is therefore a clear passage dorsally (in the bony skull) to the ventral side of the bones of the skull roof (see also Fig. 9B). Behind the vomer, only a relic of the interpterygoid vacuity re- mains; through it is visible the ventral surface of the processus cultriformis of the parasphenoid. The only specimen of D. turpior in which the median region of the palate is well preserved is specimen DGM No. 530R. In this^ the vomer and palatines form a bony roof over a greater extent of the posterior part of the primary palate, and the median fusion of the pterygoids has extended further forward within the interptery- goid vacuity, so that the processus cultriformis is not visible from Figure 7. Dinodontosaums brevirostris, type specimen. 1/3 (for abbreviations, see p. 26). Lateral view, x iThis specimen is of particular interest because it was described in 1936 by Tupi Caldas under the name Diodontosaurus pedroanum, a name which predates Romer's description of Dinodontosaums (1943). However, Tupi Caldas's description is unfortunately so brief and so poorly illustrated that it cannot be regarded as a satisfactory basis for the identification of a new taxon, and Diodontosaurus pedroanum must therefore be regarded as a nomen nudum. The specimen has recently been more fully figured by Beltrao (1966). 10 BREVIORA No. 295 below. However, it would be unjustifiable to assume that the two species difi'er in these features, as the smaller extent of these bones in specimen MCZ No. 3454 could merely be a result of its relative immaturity. Side view (Fig. 7). The external naris is rather damaged but appears to be shallow, and there are no large foramina in the re- gion of the septomaxilla. Though the edge of the maxilla where it forms the margin of the mouth is somewhat damaged, it is thin and was probably sharp-edged. This edge descends ventroposteri- orly so that it reaches a considerable distance down the anterior border of the tusk. Comparison of D. brevirostris with the specimens of D. turpior in the Museum of Comparative Zoology suggested that the snout of the Argentinian species was very much shorter, and the tusk much more ventrally directed, than in the Brazilian species. How- ever, I was fortunately able to visit Rio de Janeiro and study the excellent material of D. turpior in the Divisao de Geologia e Min- eralogia, Ministerio das Minas e Energia. This material has been prepared under the direction of Mr. L. I. Price, who has also made extensive studies of it. Examination of this material soon showed that the skull of specimen MCZ No. 1670 is very unusual in its long, low shape, and that there is a considerable range of variation in this region in D. turpior (Fig. 8). Nevertheless, there do appear to be two constant difi"erences between the snouts of the two species. Firstly, the external naris of D. brevirostris is much closer to the anterior edge of the premaxilla than is that of D. turpior. Secondly, the tusk of D. brevirostris always projects directly ventrally, so that there is an abrupt angle between the body of the maxilla and the anterior end of the zygomatic arch, whereas there is a smoother transition between these regions in D. turpior. It is worth noting also that the lacrimal extends further forward on the face and meets the septomaxilla in all specimens of D. brevirostris in which the sutures of these regions are visible. This condition is known in D. turpior (Fig. 8F), but in that species these two bones are often separated by the maxilla. There is a prominent boss on the dorsal surface of the palate, anterior to the base of the epipterygoid. This boss lies above a very distinct curving suture which separates it from the pterygoid, and it is therefore formed by the basisphenoid. Restudy of this area in D. turpior shows an identical situation, despite my state- ment that the boss in that species "is clearly part of the pterygoid" (Cox, 1965:482). Camp and Welles (1956:fig. 49) also show this boss as part of the basisphenoid in Daptocephalus leoniceps. 1968 CHANARES DICYNODONT FAUNA 11 Figure 8. Lateral views of snouts of specimens of Dinodontosaiinis, re- duced to equal size. A, D. turpior, MCZ No. 1670; B, D. turpior, MCZ No. 1687; C, D. turpior, MCZ No. 1628; D, D. brevirostris, MCZ No. 3454; E, D. turpior, DGM No. 530R; F, D. turpior, DGM No. 309. 12 BREVIORA No. 295 Vi-;: pmx B Figure 9. Dinodontosaurus brevirostris, MCZ No. 3454. Lateral views, X 4/9, A, of complete skull; B, after removal of suborbital bar, postorbital bar, anterior end of left half of palate, and part of left maxilla. Dotted lines in figure B show ventral outlines of midline regions of premaxilla. vomer and pterygoid (for abbreviations, see p. 26). 1968 CHANARES DICYNODONT FAUNA 13 Further information on the structure of the dorsal roof of the palate and of the median interorbital-internasal septum is provided by specimen MCZ No. 3454 (Fig. 9A, B). This shows clearly that the ectopterygoid is present, confirming my eadier tentative identification of this bone in D. turpior (Cox, 1965). The palatine has an extensive exposure on the dorsal surface of the palate. It forms the lower border of the large foramen which runs forwards into the dental cavity of the maxDIa; it extends dorsally to contact the parasphenoid, and runs back to meet the basisphenoid. Much of the palatine overlies the lateral wing of the vomer, which is visible in front of the palatine and also above its posterior exten- sion, where the vomer forms the lower border of the dorsal end of the canal which runs forwards and upwards from the interptery- goid vacuity. As Klaauw and Roon (1942) have pointed out, this opening is not the simple equivalent of the old interpterygoid vacuity, and they have suggested that it be called the fenestra medio-palatinalis. The canal was probably traversed by a blood vessel. Since no such blood vessel is known in either Sphenodon (O'Donoghue, 1920) or Ctenosaura (Oelrich, 1956), it is im- possible to identify it, though it seems most likely to have been a branch of the palatine artery. Above the fenestra medio-palatinalis lies the medial region of the parasphenoid-basisphenoid complex. The processus cultri- formis of the parasphenoid, as is normal in synapsids, continues anteriorly beyond the basisphenoid. and receives the lower edge of the sphenethmoid. The processus cultriformis is underlain by the vomer. This medial part of the vomer forms the lower part of the interorbital septum and also the posterior part of the internasal septum. It extends ventrally to meet the posterior end of the me- dian palatal ridge of the premaxilla and has a total depth of 5.5 cm. The more dorsal portion of the interorbital septum is formed by the sphenethmoid (Fig. 9B). More posteriorly, much of the rod-like epipterygoid is preserved in the young specimen, but its lower edge is damaged and the posi- tion of its suture with the parietal dorsally is uncertain. There are clear sutures between the prootic and the parietal, and between the prootic and the parasphenoid-basisphenoid complex. All the specimens of D. brevirostris are tusked. Most of the tusks are broken off short, but the whole of the right tusk of speci- men MCZ No. 3453 is preserved and shows clear signs of wear. The outer surface of the distal end of the tusk is worn, so that its tip lies near its posteromedial edge. Traces of other wear facets 14 BREVIORA No. 295 are also visible on the posteromedial surface of this specimen, but are better shown in an isolated tusk, specimen MCZ No. 3452 (Fig. lOA-C). In longitudinal section of a tusk, the dentine appears to consist of a series of V-shaped bands; in a tusk about 15 cm long and 2.5 cm in diameter at its base, these bands are about 2.5 mm thick (Fig. lOE). They are caused by alternating light and dark zones of dentine, which are apparently due to slight variations in the con- centration or thickness of the dentinal tubules. Further bands, which are similar but only a few tenths of a millimetre in diameter, are also visible in thin sections examined under a microscope. A system of cracks, which in general parallel this system of V- shaped bands in the dentine, can also be seen in longitudinal sec- tion (Fig. lOD), and appear as concentric circles in transverse section of the tusk. A similar system has been described in Placerias and Kannemeyeria by Camp and Welles (1956). However, these cracks are not related to the above-mentioned variations in the structure of the dentine, and appear to be post-mortem. The cracks merely reflect the main plane of structural weakness in the dentine, which in turn is related to its mode of deposition. The base of the tusk is open and it probably grew continuously. A series of annular grooves can also be seen around that portion of the tusk which lies within the maxilla; this has also been noted by Camp and Welles (1956). Similar annular grooves and variations in dentinal structure (alternating between columnar and marbled dentine) have been described in the elephant seal Mirounga leonina by Laws ( 1953 ) . He has shown that the pattern of banding in that animal is complex, but that there is a regular annual repe- tition of this pattern. The details of the annual pattern also differ between the sexes, due to their different cycles of activity during the breeding season. It is unfortunately impossible to verify whether these features in Dinodontosaiirus are similarly related to age or sex. This would be expected only if the climate were suffi- ciently seasonal to cause variations in the rate of growth of both sexes, or if variations resulted from such seasonal activities as egg- laying. Postcranial material. Specimens MCZ Nos. 3454, 3455 and 3456 all included postcranial material; that belonging to specimen MCZ No. 3455 is particularly well preserved. However, none of this material shows significant differences from the corresponding bones of D. turpior, or adds to our knowledge of the postcranial skeleton of the genus. 1968 CHANARES DICYNODONT FAUNA 15 B Figure 10. Dinodontosaiiriis brevirostris. A-C, right tusk of specimen MCZ No. 3452 showing wear facets, x %. A, lateral view; B, anterior view; C, medial view. D, E, longitudinal sections of part of tusk of specimen MCZ No. 3456. D, showing system of cracks, x V^; E, showing pattern of bands in dentine, x 1. 16 BREVIORA No. 295 DiNODONTOSAURUS PLATYGNATHUS sp. nOV. Holotype. No. 65-XI-14-5 Museo de la Plata, La Plata, Argen- tina, fragmentary skull and lower jaw. Collected by the 1964-1965 expedition of the Museo de la Plata and the Museum of Com- parative Zoology. Horizon and locality. From an exposure in the Triassic Chah- ares Formation, about P/i miles north-north-west of the point where the Gualo River emerges from the Piano del Gualo, in west- ern La Rioja Province, Argentina. Referred material. Specimen No. 149R, Divisao de Geologia e Mineralogia, Ministerio das Minas e Energia, Rio de Janeiro. Description. The type specimen consists only of the palatal and occipital regions of a large skull, and of an almost complete lower jaw (Fig. 11). The bluntly-ending snout and the presence of downwardly directed canine tusks show that the specimen prob- ably belongs to the genus Dinodontosaurus. The anterior end of the lower jaw, however, is elongated and tapers to a relatively thin Figure II. Dinodontosaurus platygnathus, type specimen. Lateral view of lower jaw, x Vi (for abbreviations, see p. 26). point, unlike that of Dinodontosaurus turpior. It at first seemed likely that this was merely an aberrant type of distortion, but the existence of an almost identical lower jaw (specimen No. DGM 149R) in the Rio de Janeiro collection from the Santa Maria Formation of Brazil makes this explanation less plausible, and sug- gests instead that a different species of Dinodontosaurus may be represented. The fragments of the palate and occiput associated with the type specimen unfortunately do not show any other fea- tures by which the species could be distinguished from D. turpior, but it is felt nevertheless that the characters of the lower jaw merit specific distinction as Dinodontosaurus platygnathus. 1968 CHANARES DICYNODONT FAUNA 17 KANNEMEYERIID REMAINS FROM THE CHANARES FORMATION Specimen MCZ No. 3459 comprises a large left scapulocoracoid and clavicle; it was found about two miles east of the Mogote de' Gualo (see Romer and Jensen, 1966, fig. 2). The scapulocoracoid is badly flattened. As preserved, it has a total length of 63.5 cm, the scapula itself being about 48 cm long (Fig. 12). The scapula is very markedly constricted, being only 8 cm across at its narrowest point but expanding to 25 cm wide at its upper end. Most of the spine down the outer surface of the scapula has been eroded away but, from the width of its base and from the contours of the surrounding bone, the spine was clearly well developed and its dorsal end rose rapidly upwards from the blade. A groove runs down the inner surface of the lower end of the scapula. This groove leads to the coracoid foramen, which lies within the precoracoid bone. The incomplete clavicle measures 33.5 cm along its outer sur- face, which is slightly convex longitudinally. As discussed in an earlier paper (Cox, 1965), the shape of the scapula seems to be one of the diagnostic features of the two main families of Triassic dicynodonts. Short, wide scapulae are found in the Stahleckeriidae, which includes the other Chaiiares dicyno- donts {Chanaria and Dinodontosaurus) . Tall, narrow-waisted scapulae are, on the other hand, characteristic of the Kannemey- eriidae, which includes the genera Kannemeyeria, Parakanne- meyeria, Sinokannemeyeria, Barysoma and Ischigualastia. The scapula of MCZ No. 3459 is very similar to that of Barysoma, which is known from some postcranial material and an occipital plate from the Santa Maria Formation of Brazil (Romer and Price, 1944; Cox, 1965). The scapula of MCZ No. 3459 and that of Barysoma resemble one another closely in shape and in the posi- tion and strength of the spine. The scapula of Barysoma is, how- ever, considerably larger; though incomplete, it is 55.5 cm long. The single scapulocoracoid and clavicle of MCZ No. 3459 do not, of course, provide sufficient evidence to claim that Barysoma was present in the Chaiiares Formation or, alternatively, to erect a new genus or species of dicynodont. The specimen is merely evidence that the kannemeyeriid dicynodonts existed in Argentina during the time that the Chaiiares Formation was deposited. This is not surprising, since kannemeyeriids are known in Argentina both from the earlier Puesto Viejo Formation (Bonaparte, 1966a) and from the later Ischigualasto Formation (Cox, 1965). 18 BREVIORA No. 295 Figure 12. Scapulocoracoid of specimen MCZ No. 3459, x Vs (for ab- breviations, see p. 26). DICYNODONT REMAINS FROM THE TARJADOS FORMATION A few fragmentary dicynodont remains were also collected in the transitional beds underlying the thick white sandstones which form the upper part of the Tarjados Formation. Specimen MCZ No. 3468 was collected just north of the Piano del Gualo, about one mile west of the Mogote del Gualo (see Romer and Jensen, 1966, fig. 2) ; it includes fragments of a skull and lower jaw. The maxilla encloses the proximal portion of a powerful tusk about 3.0 cm in diameter. This tusk shows traces of the alternating bands de- scribed above in Dinodontosaurus brevirostris, but also bears sev- eral longitudinal furrows. The remaining two specimens from the Tarjados Formation may also belong to the Dicynodontia: two fragments of large limb bones (MCZ No. 3469) and fragments of ribs (MCZ No. 3467). 1968 CHANARES DICYNODONT FAUNA 19 THE AGE OF THE CHANARES FAUNA The commonest Chanares dicynodont is Dinodontosaurus hrev- irostris. This species is extremely closely related to Dinodontosaurus turpior, which is the commonest dicynodont of the Santa Maria Formation of Brazil. The two species differ only in that the nostril of D. brevirostris is slightly closer to the front edge of the pre- maxilla. and its postorbital bone extends closer to, and further behind, the pineal foramen. This degree of similarity between their most abundant types of dicynodont is very strong evidence that the two faunas are very close in age. This is supported by the pres- ence in both faunas of the peculiar elongated type of dicynodont jaw, which has here been named Dinodontosaurus platygnathus. The Chaiiares fauna, however, lacks Stahleckeria, a dicynodont which occurs with Dinodontosaurus in the Santa Maria fauna; this suggests that the two faunas are not identical in age. In the slightly greater extent of its postorbital, D. brevirostris is more primitive than D. turpior, and it may well be ancestral to the Brazilian spe- cies; this suggests that the Chanares fauna is slightly earlier than the Santa Maria fauna. Another fauna which appears to be earlier than that of the Santa Maria Formation is that of the Manda Formation of East Africa (Cox, 1965). This includes the dicynodont genus Kannemeyeria, found also in the early Triassic Cynognathus Zone fauna of South Africa, and differs from the Santa Maria fauna in containing a more primitive type of rhynchosaur (Colbert, 1958) and in lacking dinosaurs. Since both the Manda fauna and the Chanares fauna thus appear to be earlier than the Santa Maria fauna, it is necessary next to discuss the relative ages of these two faunas. As just noted, the strong similarity between the dicynodonts of the Chafiares fauna and those of the Santa Maria fauna strongly suggests that these two faunas are very close in age. In view of the considerable differences between the Santa Maria fauna and the Manda fauna, this further implies that the Chafiares fauna is prob- ably later than the Manda fauna. However, Romer (1966) has, on the contrary, suggested that the Chafiares fauna is the earlier of the two. The evidence provided by the other elements of these faunas (cynodonts, pseudosuchians and rhynchosaurs) must there- fore be examined. The traversodontid Massetognathus is the only Chanares cyno- dont so far described (Romer, 1967). Each of the upper cheek teeth of this genus "shoulders" slightly into the one ahead; their crowns each bear two external cusps and a cross-ridge which lies 20 BREVIORA No. 295 at the extreme posterior edge of the tooth so that the crown con- sists of a single basin. The traversodontid of the Manda fauna is Scolenodon, in which the upper cheek teeth do not "shoulder" into one another, the crown of the tooth bears only a single external cusp, and the cross-ridge lies more anteriorly, so that it divides the crown into anterior and posterior basins. The upper cheek teeth of later Triassic traversodontids such as Proexaeretodon and Exaeretodon of the Ischigualasto Formation, on the other hand, are similar to those of Massetognathus, but the "shouldering" of one tooth into the next is more pronounced. As Romer (1967: 20) states, "These differences suggest that Scolenodon is rather more primitive than the described South American forms." Further- more, trirachodontid cynodonts are present in both the Manda Formation and in the earlier Cynognathus Zone fauna of South Africa, but are absent in the Chafiares fauna. The relationships among all these cynodont faunas are thus more easily explained if the Chafiares fauna is later than that of the Manda. Romer (1966) has stated that the Chafiares fauna contains small thecodonts similar to Eiiparkeria of the Cynognathus Zone. Other, more advanced pseudosuchians are also present, however, and our knowledge of the Chafiares pseudosuchians is as yet too incomplete to provide any basis for a dating of the fauna. Rhynchosaurs are absent from the Chafiares fauna, but are found in both the Manda fauna and the Santa Maria fauna. As noted by Romer (1967), this fact is more easily explained if the Chafiares fauna is older than that of the Manda and represents a time at which the rhynchosaurs were not yet developed as a major faunal ele- ment. However, rhynchosaurs are also unknown in two of the three localities in the Santa Maria Formation, and the composition of the fauna of these two localities is exactly the same as that of the Chafiares fauna: dicynodonts (mainly Dinodontosaurus) , cyno- donts and pseudosuchians. Rhynchosaurs are known in the Santa Maria Formation only from a third locality, in which dicynodonts are absent (von Huene, 1935-42; Bortoluzzi and Barberena, 1967). These facts suggest that, though rhynchosaurs and dicyno- donts are found together in the earlier Manda fauna, the ecological preferences of the South American type of rhynchosaur were dis- tinct from those of the dicynodonts of these faunas, so that these groups are rarely preserved together — and it is relevant to note here the extremely specialised dental apparatus of the rhyncho- saurs. If this is true, it is neither surprising nor significant that rhynchosaurs have not been found in the Chafiares fauna. 1968 CHANARES DICYNODONT FAUNA 21 To summarize, none of the evidence from tiie other vertebrate groups provides a convincing reason for rejecting the conclusion (based on the dicynodonts) that the Chanares fauna is only slightly earlier than that of the Santa Maria Formation of Brazil; it further appears to be later than that of the Manda Formation of East Africa. One may next attempt to assess the relative ages of these South American Triassic faunas. The Chafiares Formation of Argentina is somewhat older than the Los Rastros Formation, by which it is conformably overlain. On the other hand, it is also, on the evidence of the species of Dinodontosaurus, somewhat older than the Santa Maria Formation of Brazil, which may therefore be regarded as equivalent in age to the Los Rastros Formation of Argentina. Un- fortunately, vertebrate fossils have not been found in the Los Rastros Formation, so it is impossible to confirm the equivalence by direct faunal comparison. The footprints of Rigalites from the Los Rastros Formation could, however, well have been made by one of the large pseudosuchians of the Santa Maria Formation (Bonaparte, 1966b). There is also no unconformity between the Los Rastros Forma- tion and the Ischigualasto Formation in the Chanares area, but merely a decrease in the amount of sandstone and an increase in the amount of clay and shale. Despite this lack of any break be- tween the two formations, and the similarity between their cynodonts and rhynchosaurs, the archosaurs of the Ischigualasto fauna are considerably more advanced than those of Los Rastros/Santa Maria age. Finally, it is impossible to equate these Gondwanaland verte- brate faunas to the standard German and Alpine divisions of the Northern Hemisphere Triassic with any degree of accuracy (Romer, 1966; Cox, 1967). One can at present only attempt to define the extreme limits between which these faunas must lie. As argued earlier, even the Chafiares fauna is younger than the Manda fauna of East Africa. The latter contains the pseudosuchian reptile Mandasuchus, which is very similar to the genus Ticinosuchus, found in a definitely Anisian level of Monte San Giorgio, Switzer- land (Krebs, 1965; Charig, MS in preparation). All these South American faunas are therefore probably post-Anisian. At the other extreme, even the Ischigualasto fauna is earlier than the Upper Norian, since it contains armoured pseudosuchians but lacks coelurosaurs and large dinosaurs (cf. Chowdhury, 1965). In fact, this fauna is probably somewhat earlier than Norian: Bona- parte (1966b), after reviewing the whole Ischigualasto fauna, has concluded that it is probably of Carnian age. 22 BREVIORA No. 295 These conclusions are similar to those of Stipanicic, which are based upon evidence from fossil floras and ammonites, and which may be briefly considered here. The palaeobotanical evidence consists of the presence in the Los Rastros Formation (among others) of a flora which contains both Gondwanic Dicroidium ele- ments and also a considerable number of Northern Hemisphere species (Stipanicic, 1957). Stipanicic points out (In press) that the northern species are known in the Northern Hemisphere only in deposits of post-Ladinian age. However, the range of these species into the earlier Triassic is still unknown, and this evidence therefore cannot yet be regarded as conclusive. The ammonite evidence consists of the presence of specimens, identified by Barthel (1958) as Ciiccoceras n.sp.aff. cuccense and Beyrichites sp., in beds lying under the igneous Pastos Grandes Group in Chile. According to Stipanicic (1967), this group prac- tically interfingers with the Choiyoilitense Group of Argentina. Stipanicic (In press) believes that the Choiyoilitense Group is part of an igneous complex which is to be found below the various fos- sil vertebrate faunas and below the Dicroidium flora. Since the ammonites mentioned above indicate an Upper Anisian age, the overlying igneous complex and the fossil faunas and floras must all be post-Anisian. However, Kummel (personal communication) feels that the ammonite specimens are too incomplete for a definite taxonomic identification, and this line of argument must therefore be regarded as unproved. SUMAAARY Three new stahleckeriid dicynodonts from the Chaiiares Forma- tion of Argentina are described: Chanaria platyceps gen.et sp.nov., Dinodontosaurus brevirostris sp.nov., and Dinodontosaurus platy- gnathus sp. nov. A few post-cranial remains suggest that a kanne- meyeriid dicynodont was also present. Dinodontosaurus brevirostris is very closely related to Dinodon- tosaurus turpior of the Santa Maria Formation of Brazil, and is probably directly ancestral to it. The Santa Maria Formation there- fore appears to be only slightly younger than the Chanares Forma- tion, and equivalent to the Los Rastros Formation of Argentina. The Chanares fauna is therefore younger than the Manda fauna of East Africa; this conclusion is supported by the cynodonts of these faunas. Though rhynchosaurs are absent from the Chaiiares fauna, this is also true of two of the three localities in the Santa Maria Formation, and their absence is almost certainly merely ecological. 1968 CHANARES DICYNODONT FAUNA 23 The Chanares, Santa Maria and Ischigualasto faunas are, col- lectively, probably of Ladinian to Carnian age. ACKNOWLEDGMENTS I am, firstly, very grateful to Dr. A. S. Romer for generously in- viting me to describe these dicynodonts, which he collected during an expedition which was largely financed by the National Science Foundation, under grant No. GB2454. Most of the material was prepared and illustrated by my research assistant, Mr. P. Hutch- inson; I am greatly indebted to the Natural Environment Research Council for the grant which has allowed him to work with me, and for a special grant which enabled both of us to carry out this work at Harvard. Most of the preparation was done with the aid of NSF grant No. GB4615 to Dr. Romer. My grateful thanks are also due to the Royal Society, whose travel grant made it possible for me to visit Rio de Janeiro. Finally, I should like to thank Mr. L. I. Price of the Divisao de Geologia e Mineralogia, Ministerio das Minas e Energia, Rio de Janeiro, for allowing me to make use of his draw- ings for Figures 5E, F, and 8E, F. REFERENCES CITED Barthel, K. W. 1958. Eine marine Faunula aus der mittleren Trias von Chile. Neues Jahrb. Geol. Palaont. Abh., 106: 352-382. Beltrao, R. 1966. Paleontologia de Santa Maria e Sao Pedro do Sul, Rio Grande do Sul, Brasil. Bol. Inst. Cien. Nat. Univ. Fed. Santa Maria, No. 2: 3-114. Bonaparte, J. F. 1966a. Una nueva "fauna" Triasica de Argentina (Therapsida: Cyno- dontia, Dicynodontia). Consideraciones filogeneticas y paleo- biogeograficas. Rev. Asoc. Pal. Argentina, 4: 243-296. 1966b. Cronologia de algunas formaciones Triasicas Argentinas ba- sadas en restos de tetrapodos. Rev. Asoc. Geol. Argentina, 21: 20-38. BORTOLUZZI, C. A., and M. C. Barberena 1967. The Santa Maria beds in Rio Grande do Sul (Brazil). In: Problems in Brazilian Gondwana Geology, eds. J. J. Bigarella, R. D. Becker, and I. D. Pinto, Curitiba, Brazil, pp. 169-195. 24 BREVIORA No. 295 Camp, C. L., and S. P. Welles 1956. Triassic dicynodont reptiles. Part I. The North American genus Placerias. Mem. Univ. California, 13: 255-304. Chowdhury, T. R. 1965. A new metoposaurid amphibian from the Upper Triassic Maleri Formation of Central India. Phil. Trans. Roy. Soc. London, (B) 250: 1-52. Colbert, E. H. 1958. Relationships of the Triassic Maleri fauna. J. Palaecnt. Soc. India, 3: 67-81. Cox, C. B. 1965. New Triassic dicynodonts from South America, their origins and relationships. Phil. Trans. Roy. Soc. London, (B) 248: 457-516. 1967. Changes in terrestrial vertebrate faunas during the Mesozoic. In: The Fossil Record, ed. W. B. Harland, et al. London (Geo- logical Society), pp. 77-89. HUENE, F. VON 1935-1942. Die fossilen Reptilien des slidamerikanischen Gondwana- landes. Munich, 332 pp. Klaauw, C. J. VAN DER, and J. M. van Roon 1942. Fenestra medio-palatinalis und Interpterygoidalspalt. Arch. Neerland. Zool., 6: 327-340. Krebs, B. 1965. Ticinosuchus ferox nov. gen. nov. sp. Ein neuer Pseudosuchier aus der Trias des Monte San Giorgio. Schweiz. Palaeont. Abhandl., 81: 1-140. Laws, R. M. 1953. A new method of age determination in mammals with special reference to the elephant seal [Miroiinga leonina, Linn.). Falk- land Is. Dep. Surv. Sci. Rep., 2: 1-11. O'Donoghue, C. H. 1920. The blood vascular system of the tuatara, Sphenodon punctotus. Phil. Trans. Roy. Soc. London, (B) 210: 175-252. Oelrich, T. M. 1956. The anatomy of the head of Ctenosaura pectinata (Iguanidae). Misc. Publ. Mus. Zool. Univ. Michigan, 94: 1-122. Romer, A. S. 1943. Recent mounts of fossil reptiles and amphibians in the Museum of Comparative Zoology. Bull. Mus. Comp. Zool., 42: 331-338. 1968 CHAN ARES DICYNODONT FAUNA 25 1966. The Chaiiares (Argentina) Triassic reptile fauna. I. Intro- duction. Breviora, Mus. Comp. Zool., No. 247: 1-14. 1967. The Chaiiares (Argentina) Triassic reptile fauna. III. Two new gomphodonts, Massetognathiis pascuali and M. teruggii. Bre- viora, Mus. Comp. Zool., No. 264: 1-25. RoMER, A. S., and J. A. Jensen 1966. The Chaiiares (Argentina) Triassic reptile fauna. II. Sketch of the geology of the Rio Chafiares — Rio Gualo region. Breviora, Mus. Comp. Zool., No. 252: 1-20. RoMER, A. S., and L. I. Price 1944. Stahleckeria lenzii, a giant Triassic Brazilian dicynodont. Bull. Mus. Comp. Zool., 43: 465-490. Stipanicic, p. N. 1957. El Sistema Triasico en la Argentina. XX Cong. Geol. Int. (Mexico, 1956), Section II: 73-112. 1967. Consideraciones sobre las edades de algunas fases magmaticas del Neopaleozoico y Mesozoico. Rev. Asoc. Geol. Argentina, 22: 101-133. In press. Triasico. Review, I Int. Symp. Gondwana Strat. Paleont. TuPi Caldas, J. 1936. Paleontologia do Rio-Grande-Do-Sul. O fossil de Sao-Pedro. Rev. Inst. Hist. Geogr. Rio Grande do Sul., 1936: 243-249. (Received 5 January 1968.) 26 BREVIORA No. 295 LIST OF ABBREVIATIONS USED IN THE FIGURES a angular art articular bo basioccipital bsp basisphenoid-parasphenoid complex CO coracoid corf coracoid foramen d dentary ect ectopterygoid eo exoccipital ept epipterygoid f frontal fmp fenestra medio-palatinalis ip interparietal J jugal 1 lacrimal mx maxilla n nasal op opisthotic P parietal pal palatine pc processus cultriformis pco precoracoid pmx premaxilla po postorbital PP preparietal prf prefrontal pro prootic psp parasphenoid 1968 CHANARES DICYNODONT FAUNA 27 pt pterygoid q quadrate qj quadratojugal s sphenethmoid sa surangular sc scapula smx septomaxilla so supraoccipital sp splenial sq squamosal V vomer Harvard MCZ Libra 3 2044 066 302 753