bate ots ‘ SWARU San we eee : es Ware acaver ial a a > we Se Xe. co HARVARD UNIVERSITY eS Library of the Museum of Comparative Zoology fia a ok *, ‘ Ls i ** is ods) aaj % page ry aor ts ‘3 eS ri oe vim ar : 4 => Miotw 85 : ‘ Cielo) h | Fl ie r ’ i Pray : j | if ti ’ oe aa : : : ; / ; it ; ee aly at Mee) y Me 7 y, ee ma er Bt) Se DEVEILArS ey hae br Th Uy j aie ye) iW A) Tk Roe ens ea ae 1 a a J ] “ hs hd ae j mi) a ae iy y ? | ae mI 4 f ; } v i Dhan a Ne 7 yt | . ih iy: Lf Mii wae \ “a : -' i mn i ' 1) pe aap at f r a A ai ‘ rt t! hy tae’ @ aw AS BREVIORA MUSEUM OF COMPARATIVE ZOOLOGY HARVARD UNIVERSITY NUMBERS 231-264 1965-1967 CAMBRIDGE, MASS., U.S.A. 1967 Edited By NELDA E. WRIGHT No. No. No. code 9232. PRE? 234. 5 72! S237: 238: CONFEENES BREVIORA MUSEUM OF COMPARATIVE ZOOLOGY ‘ 5 NUMBERS 231-264 1965 A new anole (Sauria, Iguanidae) from Puerto Rico. By Ernest E. Williams, Juan A. Rivero and Richard Thomas. 18 pp. October 29. Hispaniolan giant anoles (Sauria, [guanidae): new data and a new subspecies. By Ernest E. Williams. 7 pp. October 29. South American Anolis (Sauria, Iguanidae): two new species of the punctatus group. By Ernest E. Williams. 15 pp. October 29. Studies on neotropical Pompilidae (Hymenoptera). I. The genus Agenioideus Ashmead in South America. By Howard E. Evans. 7 pp. October 29. A new salamander of the genus Chiropterotriton (Cau- data: Plethodontidae) from Mexico. By George B. Rabb. 8 pp. November 19. . Variation in the number of marginal tooth positions in three species of iguanid lizards. By Clayton E. Ray. 15 pp. December 10. A new species of the ant genus Dacetinops from Sarawak. By Robert W. Taylor. 4 pp. December 15. 1966 An evaluation of Jamaican Dromicus (Serpentes, Colu- bridae) with the description of a new species. By Donald W. Buden. 10 pp. February 25. No. No. 7S). 240. 241. . 242. . 243. . 244, . 245. . 246. . 247. . 248. . 249. 250: South American anoles: Anolis biporcatus and Anolis fraseri (Sauria, Iguanidae) compared. By Ernest E. Williams. 14 pp. February 25. Gymnothorax galetae, a new moray eel from the Atlantic Coast of Panama. By Ira Rubinoff. 4 pp. February ZS, Avocettinops yanoi, a new nemichthyid eel from the southern Indian Ocean. By Giles W. Mead and Ira Rubinoff. 6 pp. February 25. The supposed “sponge spicules” of Merrill, 1895, from the Lower Cretaceous (Albian) of Texas. By William A. S. Sarjeant. 15 pp. February 25. Quaternary fish fossils from west of Lake Rudolf, Kenya. By Keith Stewart Thomson. 10 pp. April 29. A new species of Ashmunella from West Texas (Mol- lusca: Pulmonata). By W. J. Clench and W. B. Mil- ler. 6 pp. April 29. Notes and descriptions of new Urocoptidae from Cuba and Hispaniola (Mollusca: Pulmonata). By William J. Clench. 14 pp. April 29. Pseudanthessius procurrens n.sp., a cyclopoid copepod associated with a cidarid echinoid in Madagascar. By Arthur G. Humes. 14 pp. May 3. The Chanares (Argentina) Triassic reptile fauna. I. In- troduction. By Alfred Sherwood Romer. 14 pp. May 3h, A Triassic ammonite from the Hindubagh region, Baluch- istan, West Pakistan. By Bernhard Kummel. 5 pp. Sully 29. Additional notes on the amphisbaenids of greater Puerto Rico. By Richard Thomas. 23 pp. July 29. The brain of the emu Dromaeus novaehollandiae. II. Anatomy of the principal nerve cell ganglia and tracts. By Stanley Cobb. 27 pp. November 4. No. No. No. yw 2aily. B22: - 253: . 254. 255): n250: H257. E258: 725) 3). 260. 5261: Chronological survey of the tetrapod-bearing Triassic of Argentina. By J. F. Bonaparte. 13 pp. November 4. The Chanares (Argentina) Triassic reptile fauna. II. Sketch of the geology of the Rio Chanares-Rio Gualo region. By Alfred Sherwood Romer and James A. Jen- sen. 20 pp. November 4. A new Hispaniolan gecko. By Richard Thomas. 5 pp. November 4. | Preliminary descriptions of new Honey-Eaters (Aves, Meliphagidae). By Finn Salomonsen. 12 pp. Novem- ber 4. A revision of the fossil selenodont artiodactyis from the Middle Miocene Thomas Farm, Gilchrist County, Florida. By Vincent Joseph Maglio. 27 pp. Decem- ber 6. 1967 Anolis chocorum, a new punctatus-like anole from Dari- én, Panama (Sauria, Iguanidae). By Ernest E. Wil- liams and William E. Duellman. 12 pp. February 3. A review of the Clark Fork vertebrate fauna. By Roger C. Wood. 30 pp. February 3. Biology of the parthenogenetic fungus beetle Cis fuscipes Mellié (Coleoptera: Ciidae). By John F. Lawrence. 14 pp. February 3. Expanding the palpi of male spiders. By William A. Shear. 27 pp. February 3. Monograph of the genus Spiroceramus (Mollusca: Pul- monata: Urocoptidae). By William J. Clench. 10 pp. February 3. The monticola group of the lizard genus Anolis in His- paniola. By Richard Thomas and Albert Schwartz. 27 pp. March 31. No. 262. A phylogenetic survey of moljuscan shell matrix proteins. By Michael T. Ghiselin, Egon T. Degens, Derek W. Spencer, and Robert H. Parker. 35 pp. March 31. No. 263. The hydroid of Vannuccia forbesii (Anthomedusae, Tubulariidae). By Anita Brinckmann-Voss. 10 pp. March 31. No. 264. The Chanares (Argentina) Triassic reptile fauna. III. Two new gomphodonts, Massetognathus pascuali and M. teruggii. By Alfred Sherwood Romer. 25 pp. April 6. INDEX OF AUTHORS BREVIORA MUSEUM OF COMPARATIVE ZOOLOGY ‘ NUMBERS 231-264 1965-1967 BONAPARTE Ji es holt enets teats atic ok Sale Gok Lo and eee 251 IB RINCKMANNGV.OSS VANIDAL vis0s sn 5 5 es wad lee eee mee ee 263 BEDE NR IDONATED 2 Watters oh kere as a2 Vosecie chs Cec aan ahaa See ey a 238 CRENCH aWIELIAM Ms 42255 amtod es 2) 6 aa tet ncnee DAE 2455260 OBR aS CAINE Yer eer War ee gis Ohare eee Oe MANE en 250 IDEGENS WE GONPAE SS kate cee lo einen he fp Mere eee 262 DWE IWIENMANFe WEE TANEIES. fates cps. cuss bape aye, © ord Beene ee ee 256 EVANS IOWARD Beers oan ree sarge ene ee oe eee 234 KG TASE TAEN EINE CEVA Mia Decl. Be cars co a At Sayre ar nee ee 262 PERWINDE Sse NREIORG Gi oe Boer tc sek ca ine shavers Were! Stee eer 246 EN SEN: AIMEES Ate aps fache~ ot cv eceis eve Se eae ee ee DIS OMMET OBE RNETARD (0s 5 hc ahr, oh stand coarse 248 IO ASNIRE GES NOHING Hea. ce urpy ce sees) ty See hen ecw oy eee yeh 258 AGIOS WWINGENT JOSEPH: <4. 56) ors coches eos th oie eee ee DISS) AE AD) SCRIBES Wier 202 ye uid chtd Sale cer ME Ae Oe 241 IVERICTE BIR TN CPt enact nse okie) 2 su eke Pay Marist, CR cr ty «cae a Baht Bea pws 244 RARKE RA IROBE RSD: gbi.d iG cake ake: ae ele, Gere tater ie vega ets 262 EO NBIBMIGTE OR GIE sites tt a yoo Peat a Aad Se er a ee oes 235 EAs IGAY TONS SO eke 3 SU SON te bn ay honk atoke cet eee 236 FOIE R On WAN PAS trata creche: salts ake Race ots Coe eemoeee oes ee ae 23 ROMER. ALERED: SHERWOOD jis. = oss lake Aco ke DAT. 252, 204 FUBINORE OGURA ES e co sik: & Ghea so ee ree 240, 241 SATOMONSENS@ISINING dope cee ct = epeead eae ees Sie ee 254 SARE AINSI 5 WYLIE LE TAINWAG, “So yio..c. 3% o, o spans eae ast eh regenera 242 SCHWARTZ. OAUEB ERG sks fais chic Ke Sachse «, CROCE eae eee Ren Se 261 SHEAR: BWiDETDANGeAGtE ys. PRY. 088 TRC PEL ce Shee eee meee 259 SPENGERY DEREK Ws Stn eee oe Magis Senter 262 SWAVTORS ROBE RI Wiseds 4\.0 oie <.kvoehs Soho Sie rere 230) INEONMASIRIGHARD fscsc 30-6 ocr acorrs-e ae < eens oes 231, 2495253: 261 IRHOMSON: UKEITH: STEWART 565) dak as «= uss oe ee 243 NWWHISEeTANS ERNEST I. .osca sex aul 2351, 2232! 2334239256 WOOD ROGER NE@S 5 bo isc dS cuddles os Sie ontos Me Ce DS BREVIORA Museum of Comparative Zoology CAMBRIDGE, Mass. OcTroBEn 29, 1965 NUMBER 231 A NEW ANOLE (SAURIA, IGUANIDAE) FROM PUERTO RICO By ErRNeEst #. WILLIAMS, JUAN A. RIvERo and RicHARD THOMAS PART I. DESCRIPTION By Ernest EK. WIuuiAMs Museum of Comparative Zoology and JUAN A. RIVERO University of Puerto Rico, Mayaguez On June 16, 1963, Juan Rivero was collecting at Cerro La Punta at an approximate elevation of 1200 meters when his wife and son, who had separated from the rest of the party, shouted that they had what appeared to be a new species of lizard. The animal was first seen in the axilla of one of the outer leaves of the bromeliad Vriesia (Thecodactyllum) sintenisu, which al- though usually epiphytic, was growing abundantly, together with Guzmana berteroniana, on the forest floor alongside the road. Although the cloud forest is fairly heavy in this area, trees and tree ferns had been eut for about 10 or 15 meters along the margin of the road, leaving an open although some- what shaded strip on each side. It was in the marginal area between the forest and the cleared area that the new anole was collected. When capture of the specimen was attempted, it Jumped out of the bromeliad and climbed a small bush nearby. The move- ment was described as slow, but the observers are not certain if it walked, jumped or crawled. Search for other specimens in and out of bromeliads was fruitless. i) BREVIORA NomZal On July 9, Mr. Francis Rolle collected a second specimen at the Maricao Reserve Forest. The animal was at the entrance of a hole on a partially rotten buttress of a tree at about 1.8 meters from the ground, and 600 to 750 mm from a bromeliad. Mr. Rolle used a fly swatter to stun the animal (which later died enroute) and its behavior could not be observed. Except for its lighter color, this specimen does not differ materially from the type. The first specimen was kept in the laboratory for several days, but although it was eating well (Drosophila) and appeared to be in good condition it was eventually preserved in view of the danger that it might die and decay over a week end when it was not under observation. The animal expanded the dewlap on one occasion, and when given the opportunity to move on the floor, it always did so by jumping, or, if forced to move about until tired, by moving the hind limbs simultaneously, as if it were swimming. A 16 mm black and white film forms part of the type material filed at Museum of Comparative Zoology. During its first few days in captivity, the new anole slept by lying flat on one of the side panels of the aquarium which served to hold it. All four legs were on the glass, but the tail, on the floor of the aquarium, apparently served to prop the animal upwards. Later, it slept on a twig, in typical Anolis fashion. it was never seen to enter a small bromehad that was provided inside the small tank, but its tail sometimes encircled twigs or small branches with its tip, although the animal was never seen to hang from its tail. No further specimens were obtained until early 1965 when Richard Thomas and later he and Albert Schwartz collected the new species in numbers at four localities in the mountains of southern Puerto Rico. Thanks to this material more is known both of the habitat and distribution of the new species. The material of the new species has been divided among a number of institutions: the Museum of Comparative Zoology (MCZ), the American Museum of Natural History (AMNH), the Carnegie Museum (CM), the Museum of Zoology University of Michigan (UMMZ), and the United States National Museum (USNM). A number of the Thomas specimens have been re- tained in the Albert Schwartz collection (ASE'S) or the Richard Thomas collection (RT). In allusion to its long concealment from scientific record we call the new species by the Latin adjective which means “‘hid- den’’: 1965 NEW ANOLE FROM PUERTO RICO Se <<, — & : Dewlap of MCZ 83666 (9 ) f type, MCZ 80303. Bottom 1ew oO Dorsal v Fic. 1. Anolis occultus new species. Top showing inset (‘‘slotted’’) margin. (ee) 4 BREVIORA No. 231 ANOLIS OCCULTUS new species Holotype. MCZ 80503, adult female, on Rd. 143, midway be- tween Cerro La Punta (1338 m) and Cerro Maravilla (1183 m), Puerto Rico, J. Rivero ecoll., 16 June, 1963. Paratypes. MCZ 83735, male, Maricao Reserve Forest, Fran- cis Rolle coll., 9 July, 1963; ASFS-V5489-91, V5494, AMNH 94560-61, 13.7 km N Sabana Grande, 2800 feet (850 meters), A. Schwartz, Richard Thomas eoll., 25 February, 1965; MCZ 83661-63, same locality, Richard Thomas eoll., 28 February, 1965; ASFS-V6196-97, 10.6 km SSE Villa Pérez, 3400 feet (1040 meters), Richard Thomas coll., 27 March, 1965; MCZ 83664-67, same locality, Richard Thomas coll., 28 March 1965; MCZ 83656, 18.6 km NE Guayama, 2000 feet (610 meters), Rich- ard Thomas ecoll., 31 January, 1965; MCZ 83657-59, ASFS- V4891-92, V4901, CM 40691-92, UMMZ 126021-22, USNM 157107-08, 20.9 km NNE Guayama, 2300 feet (700 meters), Richard Thomas ecoll., 2 February 1965; MCZ 83660, same lo- eality, Richard Thomas coll., 5 February, 1965; ASFS-V5017-19, same locality, Richard Thomas ecoll., 3 February, 1965. ASFS- V6662-65, RT 1832, 13.7 km S Palmer, Richard Thomas coll. 29 July 1965; ASFS-V6670-71, same locality, Richard Thomas coll. 30 July 1965. Diagnosis. An Anolis distinguished from all others by the wholly granular supraciliary margin, with no enlarged or elon- gate scales, and by the extreme reduction of the canthus, of which only the two posterior scales can be said to be differenti- ated from surrounding scales, as well as by the dorsal scales of the tail which are very small and smooth. Description (paratype variation in parentheses). Head: Nar- row, elongate. Head scales small, smooth, ca. 13 (9-13) scales across snout between the hardly developed second canthals. A very shallow frontal depression. Nostril oval, nasal scale sep- arated from rostral by one small round scale. Supraorbital semicircles weakly developed, separated by 2 (2-4) scales, supraocular area with enlarged scales medially, grading into granules laterally. Entire supracilary margin granular: neither any elongate supraciliaries nor any series of enlarged squarish supraciliary scales. Canthal ridge barely ap- parent, its posterior inception indicated by two slightly enlarged squarish seales, hardly larger than the adjacent dorsal snout scales. First two canthals continued by slightly differentiated rectangular scales to a point below the naris. Loreal rows 4 (2-6, 1965 NEW ANOLE FROM PUERTO RICO 5 usually 4), all scales subequal. Supratemporal scales subgranu- lar, flattened, grading upward into irregularly enlarged scales surrounding interparietal. Interparietal round, small, larger or smaller than ear, separated from the supraorbital semicircles by 3 (2-6, usually 4) scales. Ear small, subround, placed far ventrally, directly behind the commissure of the mouth. Suboculars in contact with supralabials, anteriorly grading into loreals, posteriorly grading into supratemporals. Ten to eleven supralabials to the center of the eye, the posterior supra- labials very low and small. S Sw Ses Fig. 2. Anolis occultus new species. Dorsal, lateral and ventral views of head of type, MCZ 80303. 6 BREVIORA No. 231 Mentals small, about as deep as wide, in contact with 4 (4-6, usually 4) seales between the infralabials. Several rows of throat seales medial to infralabials and posterior to mental somewhat enlarged, grading into granular scales on center of throat; all throat scales smooth. Trunk: Middorsal scales smooth, flat, not larger than flank scales. Ventrals larger than dorsals, smooth, round, juxtaposed, in transverse rows. Gular fan: Large, present in both sexes and well developed even in juveniles, lateral margins inset (‘‘slotted’’) in general skin of neck, all scales granular, smaller than throat scales, much smaller than ventrals, edge scales especially small, lateral seales small but well developed in well separated rows (2) or these scales weakly developed, or almost absent ( ¢ ). Limbs and digits: Limbs short, flattened dorsoventrally, tibial length not exceeding distance snout-to-middle-of-eye. About 16 (14-20, usually 16) lamellae under phalanges ii and 111 of fourth toe. Seales of limbs smooth, much smaller than ventrals, supra- digital scales smooth. Tail: Round, no dorsal crest. No enlarged postanals in males. Seales on dorsal surface very small, smooth, imbricate. Seales behind vent smooth, ca. 4-6 ventral rows on distal half of tail, enlarged and keeled. Verticils indistinct, but apparently 11 dorsal granules above, 5 keeled scales below, per verticil. Color. In the living type the head was gray, with sparse, dark mottles; the eyelids of a lighter, yellowish gray color; general color of dorsum gray, with dark vermiculations and a lght bluish cast on the neck region; flanks above the shoulders yel- lowish green, with black longitudinal spots or vermiculations ; rest of flanks yellowish gray turning greenish towards the two extremities; an ochraceous stain on the dorsum, on the midline, Fic. 3. Scales around nostril compared in Anolis occultus (right) MCZ 83661 and Anolis evermanni (left) MCZ 61223. 1965 NEW ANOLE FROM PUERTO RICO 7 between the anterior limbs; two small, round and well defined yellow spots on each side of the tail base, just below the sacral bones. Below, white, except for some speckling on the throat and tail. The animal could change colors easily, becoming darker or lighter with ease. The preserved type is of a darker gray color, with lighter reti- culation and spotting; there is a light colored frontal band in front of the eyes and a couple of light spots at the base of the tail. Venter and one-third of tail whitish with dark spots and specks, throat with dark areas along the margin of the lip and speckles and spots in the center. Size (In mm). Holotype: snout-vent, 34; tail 36. Largest specimen: snout-vent, 42. DISCUSSION Anolis occultus is far more distinct from other Puerto Rican anoles than a mere tabulation of conventional scale counts and characters (Tables 1 and 2) would indicate. It is a rather Ss = 10 < sjeroduisy SapoITOTUIOS pus [eyoueds0zur Ud9M4Oq SO[BOS Sap DITOTUIES u99M4aq Se[Bos s[eyyuBo puodoes u9d9M4aq So[BIS stosoqoud No. 233 BREVIORA 14 Sulpvis ‘posiepus [Bstopplul 13 fanbijqo = 0 fasdaAsuvay = } ‘posodeyxn{ = f fozworiquitqns (1) ‘so[BoS YURY OFUL oPBOLIQUIL = 1 fyyoows = $_ apma ST (1)4s qysoio Ou ‘mod a[qnop (13) ¥-£ GEG snyoauyoubru apIa opin LT 61 fos ys + + (qso10 IB[NBO.LI1) Id I G G VG €-G syruassip svosoqoud apa aD 92-SG 4s qsa10 OU ‘Mod a[qnop 0 snurysoppfiyd MOLIBU opmM uoisuvdxe 904 é OS FG 904 YFP IBI[OULL] {ys ys | S[BIQUOA 4so10 OU 4so.10 OU ‘MOL OTGnop ‘MOI o[qnop Sold [VB] posie[ue SMOL [BSLOppru 0 0) r-& s[eIqe[qns us0M4aq [ByUoW GAIN joRyUOD UT saTBos Iv[ns uvIpeu G S]BIGBLBAyUL YQLM yoRzU00 ul sTeIqey]qns $1aav)} snyopund sajoue dnois snpojaund JO SloyOBIVYD 9[BIG ponuyu0y — T 2eqBy 1965 ANOLIS PUNCTATUS GROUP 15 Table 2 Character range in South American anoles alpha anoles beta anoles number of toe lamellae 14-30 (mode 18-21) 10-27 (mode 14-18) scales across snout 4-25 7-20 scales between semicircles 0-5 (7 *) 0-4 dorsal scale rows enlarged 0-2 0-12 * Though several alphas range up to five scales between the supraorbital semicircles, the maximum reported here occurs in two exceptional specimens of A. princeps Boulenger (= ? frenatus Cope). ter | BREVIORA Museum of Comparative Zoology Campripez, Mass. OcToBER 29, 1965 NuMBER 234 STUDIES ON NEOTROPICAL POMPILIDAE (HYMENOPTERA) I. THE GENUS AGENIOIDEUS ASHMEAD IN SOUTH AMERICA By Howarp E. Evans The rich pompilid fauna of the neotropics has unfortunately been subjected to a good deal of bad taxonomy. The reasons are the usual ones: workers have been content to erect new species and genera without having seen the types of described species, which are widely scattered throughout Europe and North and South America; each worker has tended to use his own system of classi- fication; and many parts of the neotropics remain very inade- quately collected. The two papers of Nathan Banks on the South American Pompilidae [Bull. Mus. Comp. Zool., 96: 311-525 (1946), and 99: 371-486 (1947)] are useful, but suffer from all three of these deficiencies as well as from Banks’ failure to cite references and his failure to cover quite a number of described species. It will be many years before the taxonomy of the neotropical Pompilidae can be brought to a reasonably high level. In the mean- time, I hope to publish a series of short papers covering such segments of the fauna as I am able to work out to my satisfaction. In addition to the specimens which Banks studied, I have seen much additional material from southeastern Brazil collected by Fritz Plaumann, several collections made in Chile and Peru by Luis Pefia, and the material from a recent trip to Peru, Chile, and Argentina by C. C. Porter of Harvard University. I shall also make use of other material as available, including specimens from my own collecting in Mexico, Central America, and the West Indies. A review of the subfamily Pompilinae in Mexico and Central America is being published elsewhere (Mem. Amer. Ent. Soc., in press). 2 BREVIORA No. 234 The genus Agenioideus is primarily characteristic of the warmer parts of the Holarctic region. There is considerable structural diversity within the genus, but the group nevertheless holds together well on the basis of wing venation, the weak development of the pulvillar pad and comb, and other features which I outlined in 1950 (Trans. Amer. Ent. Soc., 75: 189). One of the four North American species is confined to Mexico, two others range from southern United States into Mexico, and another (humilis Cresson) ranges all the way from southern Canada to Panama. Banks did not regard Agenioideus as occurring in South America. However, three of the four species which he included in Serico- pompilus do, in fact, belong in Ageniozdeus. The fourth species, exilis Banks, has only two submarginal cells and other features characteristic of the genus Euplaniceps, to which genus this species is here reassigned. To the best of my knowledge, Sericopompilus does not occur in South America. The three species of A genizoideus (which I regard as only two) are closely related and somewhat divergent from the North American Ageniozdeus, although still falling readily within that genus. Although the females lack a tarsal comb, as in the subgenus Gymnochares, the over-all resem- blance is perhaps greatest to Agenioideus sensu stricto. A new sub- genus, Hnbanksza, is here proposed for this group. This group may represent the descendants of a stock of Agenioideus which succeeded in entering South America sometime during the Tertiary and which underwent a small radiation there. A review of this small group seems justified at this time not only to clarify the correct generic position of these species, but a!so to properly characterize the male sex. Of the males assigned by Banks to South American Sericopompilus, only one (the allotype of accoleus Banks) properly belongs with this group. Also, Banks made several errors of fact which should be corrected; for example, he stated that the clypeus of the type of accoleus is “fully three and one-half times as broad as long”’ when in fact it measures 2.4 X as broad as long. He also did not have sufficient material to appreci- ate the variation in some of the characters used, for example in the dentition of the claws. ENBANKSIA new subgenus Type species. — Sericopompilus accoleus Banks, 1947. Subgeneric characters. — Small wasps (4-9 mm), the females (and males of one species) with banded wings, both sexes with the tibial spurs entirely white and with a white spot near the base of the hind tibiae; males with the apical abdominal tergite white; 1965 AGENIOIDEUS IN SOUTH AMERICA 3 integument smooth and polished, except the front sometimes micropunctuate; body virtually without erect setae but extensively clothed with silvery pubescence, which is especially conspicuous on the posterior slope of the propodeum. Clypeus wider than lower face, truncate, witha slightly raised apical rim; antennae relatively short for the genus, third segment in the female not equal to the upper interocular distance, in the male not much if any longer than second segment; ocelli in a broad, flat triangle. Thoracic dorsum forming a rather smooth are except the scutellum and metanotal dise prominent, somewhat compressed ; pronotum sloping smoothly in front, its posterior margin broadly angulate or subangulate; postnotum somewhat polished, constricted on the midline; propo- deal slope smooth and even; middle and hind tibiae strongly spinose, but the front tarsus without a comb, the apical tarsal seg- ments not spined beneath. Claws dentate, the tooth sometimes close to the outer ray, such that the claws appear almost bifid; fore tarsal claws of male alike, both dentate. Wing venation similar to that of other Agenioideus (see Evans, 1950, Trans. Amer. Ent. Soe., 75: 190, fig. 58). Abdominal segments showing no strong tendency to telescope; male subgenital plate moderately com- pressed, simple (Figs. 3, 4); male genitalia without basal hooklets, but with some stout setae arising from a lobe at the base of the digitus (Figs. 1, 2). Remarks. — This subgenus is named for N. Banks, who described all its presently known components. KEY TO SPECIES AND SUBSPECIES Females 1. Antennae short, third segment 2.0-3.5 > as long as thick, equal to .35-.45 > the upper interocular distance; wings with only one strong band, sometimes with a weak second band over the basal vein; legs be- yond the basal part of the femora rufous......... 1. minutus (Banks) Antennae longer, third segment 4.5-5.0 xX as long as thick, equal to .65-.80 the upper interocular distance; wings strongly twice-ban led; lecsiiuscoussatlleastaimularcespahiene sere mie ee eter ene er 2 2. Pronotum and mesoscutum ferruginous, except the former with some p.le yellow markings; front relatively narrow, middle interocular distance HOM LLansiacialadistancen ees: 2a. accoleus accoleus (Banks) Pronotum and mesoscutum fuscous, except the former with pale yellow markings on the collar and often along the posterior margin; front broader, middle interocular distance .59-.61 X transfacial distance... . OO Oe 2 Ae Oe POD EL Dee eT ote Cente ae ee 2b. accoleus lucanus (Banks) 4 BREVIORA No. 234 Males 1. Wings clear hyaline, with a whitish bloom; legs rufous beyond basal parts of femora; third antennal segment wider than long; subgenital plate narrowly truncate apically (Fig. 3).............. 1. minutus (Banks) Wings banded; legs mostly fuscous; third antennal segment longer than wide; subgenital plate subacute apically (Fig. 4)................... 2 Pronotum, mesoscutum, scutellum, metanotum, and upper part of meso- pleura ferruginous; middle interocular distance .59 X transfacial dis- GANCC). sce yan a RLS ee aoe re 2a. accoleus accoleus (Banks) Thorax black; middle interocular distance .62-.66 > transfacial dis- CANCELs Se”. Ee RARER Ae MN eee) PE Pen 2b. accoleus lucanus (Banks) i) 1. AGENIOIDEUS (ENBANKSIA) MINUTUS (Banks) new combination Sericopompilus minutus Banks, 1947, Bull. Mus. Comp. Zool. Harvard, 99: 435. [Type: 2, BRAZIL: Tres Lagoas, Matto Grosso, 6-10 Dec. (Cornell Univ. Exped.) (Cornell Univ.)] Female. — Length 5.0-6.5 mm; fore wing 4.6-6.0 mm. Black, except as follows: pronotum with pale yellow markings on the collar and along the posterior margin; clypeus and mandibles rufo-testaceous, the latter darker apically; antennae rufo-testa- ceous except second and apical few segments usually somewhat infuscated; legs beyond the trochanters (or at least beyond the Fig. 1. Male genitalia of Agenioideus (Enbanksia) minutus (Banks), ventral aspect. Fig. 2. Same of A.(E.) accoleus lucanus (Banks), left side omitted. Fig. 3. Tip of subgenital plate of male A.(H#.) minutus (Banks). Fig. 4. Subgenital plate of male A.(Z.) accoleus lucanus (Banks). 1965 AGENIOIDEUS IN SOUTH AMERICA 5 middle of the femora) bright rufo-castaneous; fore wings with a strong brown band from the marginal cell through the outer discoidal cell, some specimens weakly tinged with brown around the basal vein. Clypeus 2.9 as wide as high. Front broad, middle interocular distance .65-.69 X transfacial distance; upper interocular distance .88-.94 X lower interocular distance; POL : OOL = 4:3. Third antennal segment three only slightly if at all longer than four, measuring 2.0-3.5 X as long as thick, .35-.45 X upper interocular distance. Propodeum somewhat more strongly convex and abdomen stouter than in the species which follows. All claws weakly dentate. Male. — Length 5-6 mm; fore wing 4-5 mm. Black; pronotum marked with pale yellow on the collar and in a broad band along the posterior margin; coxae, trochanters, and basal parts of femora black, legs otherwise ferruginous except tarsi dusky; apical half of mandibles rufo-testaceous; antennae dark brown except scape pale beneath; wings clear hyaline, with a whitish bloom, veins and stigma brown. Clypeus 3 X as wide as high, truncate. Head broad, vertex forming a strong are above the eye tops; front broad, middle interocular distance .67—.69 X transfacial distance; upper and lower interocular distances subequal, but middle inter- ocular distance 1.2-1.8 & upper interocular distance; POL: OOL = 7:4. Third antennal segment very small, slightly wider than long, not longer than segment two or more than half the length of four. Postnotum slightly shorter than metanotum; propodeum in profile forming a somewhat higher are than in the following species. Sub- genital plate narrowly truncate apically (Fig. 3); genitalia as shown in Figure 1. Distribution. — Southern Brazil, Paraguay, and eastern Peru. Specimens examined. —8 2 2,30°'c%. BRAZIL: 7 292,14, Nova Teutonia, Santa Catarina, Nov.—Feb. (F. Plaumann) [Mus. Comp. Zool., Cornell Univ., Coll. G. R. Ferguson]; 1 9, Tres Lagoas, Matto Grosso [type, Cornell Univ.]. PARAGUAY: 1 o', Caacupe, Oct. 25, 1955 (F. Schade) [Coll. G. R. Ferguson] PERU: 1 o&, Avispas, Madre de Dios, 400 meters, Sept. 10- 30° 1962 (L. Pefia) [Mus. Comp. Zool.]. 2a. AGENIOIDEUS (ENBANKSIA) ACCOLEUS ACCOLEUs (Banks) new combination Sericopompilus accoleus Banks, 1947, Bull. Mus. Comp. Zool. Harvard, 99: 433. [Type: 9, BRAZIL: Maracaji, Matto Grosso, Apr._May 1937 (G. Fairchild) (Mus. Comp. Zool.)| 6 BREVIORA No. 234 Female. — Length 8.5 mm; fore wing 8.7 mm. Dark brownish- fuscous except as follows: mandibles and clypeus pale ferruginous, the latter with yellow blotching on each side; basal two antennal segments pale ferruginous, the remainder brownish; pronotum and mesoscutum ferruginous, except the collar marked with pale yellow and the posterior pronotal margin indistinctly marked with yellowish; front tibiae and tarsi testaceous; wings hyaline, fore wing with a strong brown band across the basal vein and a broader band across the wing at the marginal cell. Clypeus 2.4 as wide as high. Front relatively narrow, middle interocular distance .56 X transfacial distance; upper interocular distance .80 x lower interocular distance; POL : OOL = 8:5. Third antennal segment 5 & as long as thick, .8 as long as the upper interocular distance. Slope of propodeum low and even, median line somewhat impressed. Claws strongly dentate, the inner ray acute, in the front tarsi the two rays rather close together, subparallel. Male. — Length 7 mm; fore wing 6.5 mm. Dark brownish- fuscous except pronotal collar marked with pale yellow, remainder of pronotum ferruginous; mesoscutum, scutellum, metanotum, and upper two-thirds of the mesopleura also ferruginous; antennae brown except scape whitish below, flagellum testaceous below; legs brown except for the usual spot on the hind tibiae and the pale spurs; fore wings with a dark band at the marginal cell as in the female, but with only a weak, narrow infuscation at the basal vein. Clypeus 2.5 X as wideas high. Front rather narrow, middle inter- ocular distance .59 X transfacial distance; upper interocular dis- tance very slightly exceeding lower interocular; POL : OOL= 7:5; vertex weakly humped at the oceilar triangle. Third antennal seg- ment rather short although longer than second, measuring about 1.2 < as long as thick. Postnotum nearly as long as metanotum; slope of propodeum very low and even. Terminaia as described under accoleus lucanus. Distribution — Brazil (states of Matto Grosso and Sao Paulo). Specimens examined. — Only the type and allotype, the latter from Campinas, Sao Paulo, March 1924 (F. X. Williams) [Mus. Comp. Zool.]. 2b. AGENIOIDEUS (ENBANKSIA) ACCOLEUS LUCANUS (Banks) new status, new combination Sericopompilus lucanus Banks, 1947, Bull. Mus. Comp. Zool., Harvard, 99: 434. [Type: 9, BRAZIL: Nova Teutonia, Santa Catarina, 25 Jan. 1939 (F. Plaumann) (Mus. Comp. Zool.)| 1965 AGENIOIDEUS IN SOUTH AMERICA i Female. — Length 6.5-8.5 mm; fore wing 6.5-9.0 mm. Black, except as follows: mandibles and clypeus pale ferruginous, the latter sometimes with obscure yellowish blotching, sometimes infuscated on the upper half; basal 2.0-2.5 antennal segments testaceous; pronotum marked with pale yellow on the collar and usually with a more or less complete yellowish band along the posterior margin; legs fuscous, front tibiae and tarsi somewhat paler than the remainder; wings banded as in the nominate sub- species. Middle interocular distance .59-.61 x transfacial dis- tance; upper interocular distance .82—.86 > lower interocular. Third antennal segment 4.5-5.0 X as long as thick, equal to .65— .75 X upper interocular distance. Claws of front tarsus variable, the inner ray close to the outer ray or somewhat removed from it. Other features as described for a. accoleus. Male. — Length 6.0-6.5 mm; fore wing 5.0-5.5 mm. Dark brownish-fuscous except mouthparts and clypeus rufo-testaceous, basal two antennal segments and under side of segment three (and sometimes four) testaceous, front legs beyond the trochanters and also middle tibiae sometimes rufo-testaceous; wings banded as in female but the band over the basal vein weak. Middle interocular distance .62-.66 X transfacial distance; upper and lower inter- ocular distances subequal. Third antennal segment 1.3-1.6 x as long as thick, sometimes barely shorter than fourth segment. Sub- genital plate subacute apically (Fig. 4). Genitalia differing from those of minutus in having both the aedoeagus and the digiti more broadly expanded apically (Fig. 2). Distribution — Southern Brazil (Santa Catarina). Specimens examined. —13 2 2,4 7%, all from Nova Teutonia, Santa Catarina, Oct.—March (F. Plaumann) [Mus. Comp. Zool., Coll. G. R. Ferguson]. (Received 14 June 1965) BREVIORA Museum of Comparative Zoology CAMBRIDGE, Mass. NovEMBER 19, 1965 NUMBER 235 A NEW SALAMANDER OF THE GENUS CHIROPTEROTRITON (CAUDATA: PLETHODONTIDAE) FROM MEXICO. By Grorce B. Rass Chicago Zoological Park, Brookfield, Illinois As Martin (1958:12) has noted, the karst caves of the Sierra Madre Oriental of Mexico are important amphibian habitats. Most of the salamander species found in such caves are widespread forms, but at least one described species from San Luis Potosf, Chiro- pterotriton mosaueri, is probably restricted to caves. Another species apparently in this category is described below. For it, I propose the name: CHIROPTEROTRITON MAGNIPES Sp. Nov. Figs. 1, 2 Holotype. — Museum of Comparative Zoology 30607, a young adult male collected in the Cueva de Potrerillos, about 2 km WSW of Ahuacatlan, which is approximately 8 km SW of Xilitla, San Luis Potosi, Mexico, on 18 November 1955 by Alejandro Villa- lobos F. Paratypes. — MCZ 44033, CNHM 142457, and four uncata- logued specimens from the type locality in the collection of the Instituto Politecnico Nacional, Mexico, collected on 20 January 1952 by F. Bonet; UMMZ 125423 (GH 9878), collected at the type locality in July 1964 by Kraig Adler; CNHM 142458, UMMZ 125179, and three uncatalogued males in the Instituto Politecnico collection from Cueva del Madrofio, near Laguna Colorada, Municipio de Landa de Matamoros, Querétaro, Mexico, collected on 16 January 1952 by F. Bonet. Diagnosis — A species distinguished from all others in the genus by its very fully webbed feet, large size (adults 40-60 mm snout- vent length) and greater number of teeth (average of 79 maxillary- premaxillary in adult males). 2 BREVIORA Nor 235 Description of the holotype. — General habitus long and de- pressed. Tail relatively short (slightly less than snout-vent length), rounded in cross section. Basal constriction of tail not indicated. Limbs relatively long; adpressed limbs overlap by two or more costal spaces. Feet broad; toes and fingers joined by thick webbing. On foot, webbing indented between digits 2-8, 3-4, and 4-5 to level of distal end of penultimate phalanx of digits 3 and 4 (Fig. 1). Terminal pads, subtending digits 2-5 on foot and 2-4 on hand, present but not prominently developed, palmar surfaces of hand and foot otherwise smooth. Contours of digits not prominent dorsally, except for pigmentless areas marking some joints. Head flat and wide, eyes large and bulging. Prominent subocular groove extending in an arc from anterior corner of eye to position beneath posterior corner of eye. Postocular and other cephalic and nuchal grooves weakly developed. Viewed from above, snout truncate. Labial protuberances small. Total number of maxillary-premaxil- lary teeth 80; 64 countable mandibular teeth (dentary broken on left side near symphysis). Vomerine teeth in slightly arched single rows numbering 13 on each side. Teeth on posterior vomerine shelf beneath parasphenoid 93, arranged in two obovate patches slightly separated in midline. Skin rather thin. Mental gland not prominent, although marked by less pigment than rest of chin. Cloacal papillae well developed. Two symmetrical pockets with slit-like openings in dorsal posterior cloacal wall. Testis unilobate, pigmentless. Color of body in aleohol uniform dark lavender- brown above, somewhat lighter on ventral surfaces and feet. Palmar surfaces pigmentless. Post-iliac gland weakly indicated by light area. Anterior ends of ceratohyals blade-like, relatively narrow. ‘Transverse processes of first caudal vertebra extend anteriorly to midpoint of preceding postsacral vertebra (see Fig. 2). Distal tarsal 5 large, articulating with fibulare and centrale 2-3; distal tarsal 4 accordingly not in contact with fibulare. Terminal phalanges with T-shaped distal ends. Measurements in mm, fol- lowing techniques in Rabb (1958): Total length, 78; snout-vent length, 40.2; head width, 8.0; eye width, 3.2; nostril width, 0.4; snout-eye length, 1.8; snout-angle of jaw length, 7.1; arm length, 13.8; leg length 14.5; hind foot width, 6.5. Variation. — The holotype is the smallest of the specimens stud- ied. The others are variously shrunken by preservative, but the measurements are generally comparable (Table 1). Sexual dimor- phism is apparent in the longer snout of the larger males, and perhaps in their relatively wider heads. Females evidently reach a larger size than the males, since three of the four available are 1965 NEW CHIROPTEROTRITON 3) longer than the oldest and largest male. The tail is proportion- ately longer in males. The average number of vomerine and maxillary-premaxillary teeth is greater in the females, possibly reflecting the larger size and probably greater age of the females. There is no marked difference detectable in the numbers of teeth in the three age classes represented among the males, although the high vomerine count of the largest male suggests a slight increase with age. The two larger males from Cueva de Potrerillos have some melanophores in the covering of the testes, but there is no such pigmentation in those from Cueva del Madrofio. The number of glands in the mental cluster ranged from about 240 to 390 in the three males examined. There is slight variation in the amount of webbing of the feet, but the average arrangement is that described for the type. Habitat. — The type locality was well described by Bonet (1953). The cave is at an altitude of about 1300 m, and apparently in what was originally a cloud forest zone. According to the field notes of Dr. Villalobos, the topography of the cave in 1955 was considerably changed from that seen by Bonet in 1952. Bonet (op. cit.) also provided data on the Cueva del Madrofio, which is at an altitude of 1810 m in mixed pine-oak woods con- taining madrofio (Arbutus) and Carya. The stalagmites in the cave were actively growing in 1952. Judging from Bonet’s account, Chiropterotriton magnipes was common in both the Cueva de Potrerillos and the Cueva del Madrono along with various cavernicolous arthropods. However, C. magnipes is not the only cave-inhabiting salamander in the region. A specimen collected by Dr. Villalobos at Cueva de la Hoya, which is only 3 km from the Cueva de Potrerillos, is refer- able to C. arboreus (MCZ 30605). Relationships and comparisons. — The genus Chiropterotriton is composed of a group of Central American boletoglossal pletho- dontids that has a distinctive, relatively delicate, general habitus but few diagnostic characteristics. Taylor (1944) defined the group on the basis of the partial webbing of the feet. Tanner (1952) described the throat anatomy in various species, and Rabb (1956) described the skull and other anatomical features of the apparently most primitive species of the group. David Wake (in litt.) has informed me that the tarsal arrangement mentioned for C. magnipes distinguishes many members of the genus from the species groups in Pseudoeurycea. Although the foot shape is like that of certain species of Bolitoglossa, C. magnipes clearly belongs to the genus 4 BREVIORA No. 235 Chiropterotriton in respect to general habitus, tarsal pattern, sub- lingual fold, form of the ceratohyal, and such features of its skeletal anatomy as can be determined from X-rays and one cleared and stained specimen (CNHM 142458). Wake and Brame (1963) indicated that fully-webbed conditions of the feet developed several times within the genus Bolitoglossa. The occurrence of a similar condition in a species of Chiropterotriton lends indirect support to their idea. The feet of Chiropterotriton magnipes are unique in the genus in fullness of webbing and pad-like structure. The closest approach to the condition in C. magnipes is in an unnamed form from El Chihue, Tamaulipas (illustrated in Rabb, 1958, pl. II). In proportions, C. magnipes differs from other Chiropterotriton in having a somewhat wider head (average, 19.1 per cent of sn out- vent length in males). In relation to head length, C. magnipes has a large eye (average percentage, 35.5, o'o7; 37.9, 2 2), although the proportion is matched in C. multidentatus from Rancho del Cielo, Tamaulipas and El Chico, Hidalgo, and in the short-snouted species C. bromeliacia. In adults of most forms, this proportion is 29 to 33 per cent. The tooth count averages in C. magnipes exceed the maxima for all other forms. However, the size of individual teeth in C. magnipes is small. The largest premaxillary teeth in males project about 0.25 mm from the gum, and the ordinary maxillary and dentary teeth are about half this size. The foot structure and long limbs of Chiropterotriton magnipes indicate scansorial habits. The lack of testis pigmentation and the large eye suggest that the species is truly troglodytic and not a casual cave inhabitant. In these respects the species is slightly more specialized than the cave-dwelling C. mosauert and the Tamaulipan Chiropterotriton with a large foot. Presumably its closest relationships are with these and other large, arboreal or scansorial northern Mexican species of the genus (C. multidentatus, C. arboreus). Acknowledgements. — I am grateful to Dr. E. E. Williams for the opportunity to deseribe this salamander. Dr. Alejandro Villa- lobos F. of the Instituto de Biologia, Universidad Auténoma de Mexico, has kindly supplied data on the type locality. Dr. F. Bonet of the Instituto Politecnico Nacional, Escuela Nacional de Ciencias Biologicas (IPN), loaned his series of C. magnipes for study and graciously allowed deposit of specimens in major U. 8. collections. George Zug sent information on species in the University of Michigan Museum of Zoology (UMMZ). X-rays were made with 1965 NEW CHIROPTEROTRITON 5 the beryllium-window machine at the University of Michigan by Kraig Adler. The prints are the work of Raymond Simpson. Abbreviations besides those indicated above are: MCZ, Museum of Comparative Zoology, and CNHM, Chicago Natural History Museum. LITERATURE CITED Bonet, F. 1953. Espeleologia Mexicana. Cuevas de la Sierra Madre Oriental en la region de Xilitla. Bol. Univ. Nac. Aut. Mexico Inst. Geol., 57: vi+96 pp. Makrtin, P. 8S. 1958. A biogeography of reptiles and amphibians in the Gomez Farias Region, Tamaulipas, Mexico. Misc. Publ. Mus. Zool. Univ. Michigan, 101: 1-102, 7 pls. Rasp, G. B. 1956. A new plethodontid salamander from Nuevo Leon, Mexico. Fieldiana: Zoology, 39: 11-20. 1958. On certain Mexican salamanders of the plethodontid genus Chiropterotriton. Occ. Papers Mus. Zool. Univ. Michigan, No. 587: 1-37, 3 pls. TANNER, W. W. 1952. A comparative study of the throat musculature of the Pletho- dontidae of Mexico and Central America. Univ. Kansas Sci. Bull., 34: 583-677. Taytor, E. H. 1944. The genera of plethodont salamanders in Mexico. Pt. I. Univ. Kansas Sci. Bull., 30: 189-232. WakgE, D. B. and A. H. Brame, Jr. 1963. The status of the plethodontid salamander genera Bolitoglossa and Magnadigita. Copeia, 1963: 382-387. No. 235 BREVIORA OTT 16 9€ $9°6E oY NdI £01 92 LOSE 98ST IF 61 NdI $8 OOT OOT LL LL €8 GE &% Iv 6VGF O00F O€ OF GMOS) COE LOSE OS'ST O8'6T OSI 9106 OF TS 0G 61 9°99 9°E¢ G9 91g 0°0¢ Zige CL ¢L [LP 6L1SZI S8SPF2rI NdI ZNWO WHNO o0'¢s LSI Ze'61 + 9€ ULP ne) NdI £6 FI F8 61 rete OG SPIE I8SE SPE PETE OG6I L921 OIG ans 81g 009 fo) 5 NdI NdI ‘By Usyorq 10 YUNOS 9jo[durOoUT SozRoTpUT [OQUIAS SN[g "SOOT SIVSE} JO JOQUINU 9}BOTPUT STPOQUIAS 9[BUT 10}JB STRIOUINN ‘pondoezyBouN 919M SudUeds OoTUdeP[Og OFNYNSUT | I €8 69 L9 Cet) 82 GL GP 8S 9G 6ECE LG°8% IS Lé Sess L108 GG VE VISl 9621 60°61 OS 21 8291 GL 0G G9 0°09 9°8¢ 0°6S 06S IL teks o) EL fe) LSPZPEL ©SSOrF NdI WHNO ZOW $9 O8 99°21 G OF To L090€ TOW y}00} Arejuep JO 1aquUINNY q}00} Areytxeuroad -AIV][IXBUL JO JaquIn 4390} PULIOUIOA JO JOAQUIN pia poy /aXo 07 Jnoug Mvl 03 ynous/aXo 07 ynoug y43us] JUOA-jnNOUS/ IPM poy (2%) quod 04 qnous/Mel 07 ynoug (uu) YQSUNT [IBY (WU) YySuUeT YUNA-ynoug oyoIpRIY Jap BAND SO]T][LlaIVOg ap BANND sadwubpu U0j217019}d0.11Y,) JO Solos odA} oY} UO BYSp DATPBITUBNY) [ PIaV I, 1965 NEW CHIROPTEROTRITON 7 Fig. 1. Chiropterotriton magnipes sp. nov., holotype MCZ 30607. X 2.1. No. 235 BREVIORA BI[IXBUL BUSY ‘er AG ‘oyeuUay 4[NpB uB “(8/86 HO) &ZF [ X ‘owiqoyea Apoq oaimosqo yns oy} ut spodoryyty “Woyxorq st C Yond AG L ZININD ‘sadvubpw wojrsjosajdosy) jo ydeisorpsey ZS BREVIORA Museum of Comparative Zoology CAMBRIDGE, Mass. DECEMBER 10, 1965 NUMBER 236 VARIATION IN THE NUMBER OF MARGINAL TOOTH POSITIONS IN THREE SPECIES OF IGUANID LIZARDS By CuayTon E. Ray U.S. National Museum Although tooth counts (more accurately, the number of tooth positions) are customarily given in the description of fossil lizards, their possible taxonomic value is generally unassessable for want of quantitative data on adequate samples of modern lizards. Further, the possible developmental and adaptive implications of tooth number have been inadequately explored. Edmund (1960, p. 66) notes an apparent lack of correlation between wave length in tooth replacement and number of tooth positions in iguanids, but the condition in juveniles of species with long wave lengths, such as Ctenosaura pectinata with 11 teeth per wave, would be especially interesting. Hotton (1955, p. 97) in a study of adaptive relations of dentition to diet in iguanids, including C. s¢milis, states regard- ing tooth number only that there are ‘‘slight tendencies toward fewer teeth in the maxillary row . . . in smaller individuals.”’ The availability of adequate, though not impeccable, samples of cranial material (assembled for other purposes) representing Ctenosaura similis, C. pectinata, and Anolis carolinensis suggested the desirability of characterizing these samples statistically, in order to compare ontogenetic and taxonomic variability in number of tooth positions, and strength of correlation between number of tooth positions and length of tooth row among two closely related taxa and one distantly related taxon. bo BREVIORA No. 236 MATERIAL Ctenosaura similis: Well-preserved, dissociated elements repre- senting no less than 223 individuals, collected from deposits in the ruins of Mayapan (20°37'48"N, 89°27'42”E), Yucatan, during expeditions (1950-1956) of the Department of Archaeology, Carnegie Institution of Washington. Among the more than 6000 bones identified from these deposits, over a third pertain to C’. similis, the only lizard represented (Pollock and Ray, 1957). The material utilized in the present study undoubtedly dates almost entirely from the period of major occupation of the city, the Maya Resurgence, estimated (and in part confirmed through radiocarbon dating) to have extended from about 1200 through 1450 A.D. Thus the sample was accumulated during an interval of perhaps 250 years or more. The specimens probably were secured within close proximity of the city. The sample is undoubt- edly biased in favor of large individuals. Early juveniles are absent altogether. Although there is certainly high representation of left and right sides of single individuals among the dissociated maxillae and dentaries, the correspondence is far from complete and can be demonstrated in no given instance. The material was made avail- able for study and was deposited in the Museum of Comparative Zoology at Harvard College through the courtesy of Dr. H. E. D. Pollock. Ctenosaura pectinata: Cleaned skulls of 17 modern specimens collected over many years at widespread localities through the range of this nominal species. Specimens are housed in the col- lections of the Museum of Comparative Zoology, American Mus- eum of Natural History, Chicago Natural History Museum, and University of Michigan Museum of Zoology, and were made avail- able by the curators of those collections. Anolis carolinensis: Macerated skulls of 58 individuals collected during 1955 in the vicinity of Gainesville, Alachua County, Florida. This sample, like those of Ctenosaura, does not include very young individuals, and thus is not ideal for the study of growth phe- nomena. In other respects it is for statistical purposes the most desirably constituted sample of the three. The specimens were made available by Dr. Walter Auffenberg and are deposited in the Museum of Comparative Zoology. METHODS The following symbols represent the statistics calculated in this study: N = the number of individuals in the sample. 1965 TOOTH VARIATION IN IGUANID LIZARDS 3 x = the mean of x, here the straight-line length of tooth row, maxillary or dentary. y = the mean of y, here the number of teeth or tooth posi- tions per maxilla, dentary, or premaxilla (fused pre- maxillae). sx = the standard deviation of x. sy = the standard deviation of y. V. = the coefficient of variation of x. V, = the coefficient of variation of y. OR, = the observed range of x. OR, = the observed range of y. r = the coefficient of correlation, here between x and y. a = the growth ratio. co, = the standard error of a. b = the initial growth index. Sa = the standard deviation of the diagonal distances of points from the reduced major axis; a measure of absolute dis- persion of points about that axis. Da = the coefficient of relative dispersion about the reduced major axis. z = statistic computed in testing for significant difference in slope of growth lines. Most of these are the familiar statistics of univariate and bivari- ate analysis dealt with in standard statistical handbooks. However, Imbrie’s paper (1956) is especially valuable regarding the basis for utilization of the reduced major axis rather than regression lines, as well as for a lucid interpretation of all of the statistics used herein. Premaxillae: The fused premaxillae were treated as a single median element. Only the number of tooth positions was deter- mined for this element, as adequate means of measuring the (short) length of the tooth row, especially in Anolis, were not at hand when the samples were assembled. Thus only the standard univariate statistics are presented (Table 1, Figure 1). Mawillae and dentaries: All of the statistics listed above were computed for both left and right maxillae and dentaries of the two samples of Ctenosaura, and for the right side only of Anolis (Tables 2,3, and 4). Asa check on symmetry, the frequency distribution for length of tooth row and number of tooth positions was in every case represented by histograms, one example of which is presented here (Fig. 2). Scatter diagrams of the two variates were prepared in each case, from which it was determined that the dis- tribution of points approximates a straight line closely enough to 4 BREVIORA No. 236 permit representation of the growth line by the equation: y=ax+b. The scatter diagrams for all elements of the right side are presented in Figures 3 and 4. RESULTS Premaxillae: As might be anticipated, the two species of Clenosaura are very similar, in fact essentially identical, in statis- tical characteristics of tooth number, and differ considerably from Anolis carolinensis (Table 1). Actually, these representatives of the two genera differ much more strikingly than the statistics alone reveal. For practical purposes it can be said that the variate will never assume an odd value in Anolis!, whereas it does so in 70 per cent or more of the Ctenosaura sampled (Fig. 1). The tooth count in an Anolis premaxilla more likely than not will be 10, and in Clenosaura 7. The fact that premaxillary tooth counts in Anolis can assume only even values might suggest that the coefficient of variation (V,) for Anolzs is artificially high. However, if only one side of the fused element is considered, with tooth counts of 3-6, V, is not altered. The V for tooth counts here and in the maxillae and dentaries is of little interest except that it reflects rather low variation in tooth number over the broad size range represented in the samples. Maxillae and dentaries: The frequency distribution for number of tooth positions is in general approximately symmetrical (Fig. 2B), although it is skewed considerably to the right in the right dentary of Anolis. The length of tooth row in no case approaches a normal curve, and a polymodal frequency distribution is sug- gested for Ctenosaura similis (Fig. 2A) and Anolis, particularly the former; N is not adequate to suggest a pattern in C. pectinata. It is tempting to suppose that broadly overlapping frequency dis- tributions of different sexual or age classes or both are responsible for the polymodality, analogous to that demonstrated by Klauber (1937, fig. 4) for Crotalus v. viridis, but this could not be tested on the basis of samples at hand. The number of tooth positions per se is not very useful in dis- tinguishing the three taxa tested. It is of no value in separating the two species of Ctenosaura, and its power of resolution is very low even between taxa as remote as Ctenosaura and Anolis. Tooth ' Reference to “Anolis” and to “Ctenosaura”’ is to be understood as a shert- hand device applying only to the samples tested here, and in no case to either genus as a whole. 1965 TOOTH VARIATION IN IGUANID LIZARDS 5 count alone in an isolated fossil specimen would be of taxonomic interest only if it lay within the portion of the Ctenosaura dis- tribution well above the OR of Anolis. The sample of Anolis is separable from those of Ctenosaura on the basis of absolute length of toothrow, which reflects the valid biological distinction that Ctenosaura is a much larger lizard, although the true population parameters unquestionably overlap. The two samples of Ctenosaura are not clearly distinguishable on the basis of any statistic pre- sented in Tables 3 and 4. The relatively high values of Vx, Vy, sa, and D4, in Ctenosaura pectinata reflect principally its greater ORx but may result in part also from geographic variation, not a factor in the other two samples. The most interesting features of the relation between length of tooth row and number of tooth positions in the samples studied are the relative strength of correlation between the two variates and the slope of their growth line?. The two samples of Clenosaura are essentially identical in these features. Both samples exhibit a strong correlation between the variates, both visually (Figs. 3A, B, and 4A, B) and statistically (minimum value of r = .83). The probability is greater than .05 that the minor deviations in slope of the growth lines are attributable to chance, for the values of z are in every case less than 1.96 (Table 4). Anolis, on the other hand, differs radically from Ctenosaura both in strength of correlation and in slope of growth line. Cor- relation is not visually detectable (except in a mildly suggestive linear clustering of points within the wide scatter) in Figures 3C and 4C, and is only weakly indicated by the correlation coefficient. However, the probability of obtaining a calculated r as great as .24 if the sample was drawn from a population in which p = 0 is less than .10, or, if, as is reasonable on developmental grounds, it is assumed that r is necessarily positive, less than .05 (Simpson, Roe, and Lewontin, 1960, appendix table V, p. 426). The slope of the growth line is very steep in Anolis, very gentle in Ctenosaura. The difference in slope is statistically highly significant (Table 4), but biologically means only that the teeth of Anolis are con- siderably smaller than those of Ctenosaura. Thus, more teeth per unit increase in length of tooth row can be accommodated in Anolis than in Ctenosaura. 2 Strictly speaking, the line is not purely a ‘‘growth”’ line as the scatter of points on which it is based undoubtedly represents in part non-ontogenetic variation. Designation of the line as a growth line rests on the assumption that regression in individuals of different sizes is the same as in individuals of different ages. for) BREVIORA No. 236 Although so few samples provide the basis for no more than suggested interpretations, it may be supposed that a difference such as that observed in correlation between number of tooth positions and length of tooth row has adaptive significance. The most fundamental adaptive distinction relating to dentition between Anolis and Ctenosaura is in diet; Anolis is a predatory insect eater, and Ctenosaura an extreme vegetarian. In Ctenosaura the maxillary and dentary tooth rows constitute continuous dental palisades in which the broadly expanded, strongly cusped crowns (more pronounced posteriorly) succeed one another in close order, and tend to overlap, an arrangement apparently advan- tageous in leaf-chopping, whereas in Anolis the tooth rows con- stitute open series in which the broadly based, apically narrow teeth with weak cusps are separated by variable gaps, possibly adaptive in the apprehension and puncturing of insect prey (Fig. 5). If in a strict vegetarian there is selective advantage in maintaining a continuous tooth row, it would be reasonable to expect a con- sistent addition of tooth positions with increasing size, 1.e., a strong positive correlation between the two phenomena. The slope of the growth line would depend in part on the size of addi- tional teeth and on the degree of increase in size of successive replacement teeth. In a form such as Anolis carolinensis with no apparent selection pressure toward a closed tooth row, and perhaps with positive selection for an open tooth row, one might expect only a loose correlation between number of tooth positions and length of tooth row. Obviously these suggestions require testing on broader ontogenetic series of the taxa studied, and on similar series of other taxa. I wish to thank Drs. Thomas Frazzetta, Nicholas Hotton, George Simpson, and Ernest Williams for reading the manuscript. LITERATURE CITED Epmunp, A. G. 1960. Tooth replacement phenomena in the lower vertebrates. Roy. Ontario Mus., Life Sci. Div., Contrib. No. 52: 1-190, 58 figs. Horton, Nicnouas, III 1955. A-survey of adaptive relationships of dentition to diet in the North American Iguanidae. Am. Midland Nat., 53: 88-114. IMBRIE, JOHN 1956. Biometrical methods in the study of invertebrate fossils. Bull. Am. Mus. Nat. Hist., 108: 211-252, 10 figs. KLAvuBER, L. M. 1937. ) @ fo) TOOTH COUNT ine 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 LENGTH OF TOOTH ROW (MM_) ‘: 20 22 24 26 2 30 32 34 36 LENGTH OF TOOTH ROW (MM.) 22 ° TOOTH COUNT 10 6 7 8 9 LENGTH OF TOOTH ROW (MM.) Fig. 3. Scatter diagrams showing tooth counts (y) and length of tooth row (x) in the right maxillae of (A) Ctenosaura similis, (B) C. pectinata, and (C) Anolis carolinensis. The broken line represents the reduced major axis; the open circle, the joint mean. 1965 TOOTH VARIATION IN IGUANID LIZARDS 13 187 ZO 220024: 2b 2801952349 36 a LENGTH OF TOOTH ROW (MM) ; : 32 oe ie TOOTH COUNT \ 18 22524265528 20 30. 32 34 36 LENGTH OF TOOTH ROW (MM.) | 24 nN Oo TOOTH COUNT ie iy) 20 10 6 é 8 9 LENGTH OF TOOTH ROW (MM.) Fig. 4. Scatter diagrams showing tooth counts (y) and length of tooth row (x) in the right dentaries of (A) Ctenosaura similis, (B) C. pectinata, and (C) Anolis carolinensis. The broken line represents the reduced major axis; the open circle, the joint mean. 14 BREVIORA No. 236 i TOOTH VARIATION IN IGUANID LIZARDS 1965 ** G][-¢ ‘pouva ovog “qoodse [engsury UL MOL YYOO} Arequop FYB fo UOIOd LOLL04ysod IZVEIG TOW ‘sisuaurposp9 “VW (A) ‘goods yenZury ut Mod YOO} ALeyUOp FYB Jo uoTYZLOd coLtoysod ‘ZOROE ZOIW ‘seems *-) (aq) ‘qoodse [ensurly UL MOL Y}00O} Arequop ys peywed ‘xzZVere ZOIN ‘sisuaurpoupo *y (q_) “oodse ensury ut Mod YOO} Arequep yYyst eEyaed ‘ZO89E ZOIN ‘syiumis *) (Q) ‘qoodse yRiqry UL MOL YOO} Areypixeur yyste peEyaed ‘qzgere ZOIN ‘sisuaupowo “py (q) “qoodse [Riqvy UL MOL YQOO, Arvypixeur yas peaked ‘ZORS9E ZOIN ‘sues “OQ (YW) “syoup pu dinvsouaz) Jo suUOUOGd *G “BTy BREVIORA Museum of Comparative Zoology CAMBRIDGE, Mass. D&cEMBER 15, 1965 NUMBER 237 A NEW SPECIES OF THE ANT GENUS DACETINOPS FROM SARAWAK! By Rosert W. Taytor Biological Laboratories, Harvard University The singular myrmicine genus Dacetinops Brown and Wilson was discovered only ten years ago. The original series included twelve workers and three dealate females collected by Edward O. Wilson in rain forest at the lower Busu River, near Lae, N. E. New Guinea. These specimens were described as D. cibdella by W. L. Brown and E. O. Wilson (1957), and a larva collected with them was characterized by G. C. and J. Wheeler (1957). A second species, described below, has recently come to hand among material collected in Sarawak by Professor A. E. and Mrs. Eleanor Emerson. This significant and unexpected record robs Dacetinops of its apparent status as a New Guinean endemic. In the light of this discovery it would not be surprising if the genus was found to be more widespread in the Indo-Australian area. DACETINOPS CONCINNA new species Type locality. SARAWAK: Kapit District, Third Division, Nanga Tekalit Camp, 1° 38’ N, 113° 35’ E, January 29, 1963, A. E. and E. Emerson. The holotype and paratypes were taken together in a dead log on the floor of virgin upland rain forest at an elevation of about 180 meters. Nanga Tekalit was a remote field camp established by the Chicago Natural History Museum, situated about 70 miles above Kapit on the Batang Balleh—Sungei Mengiong River system in western Sarawak. 1 Research supported by U. 8. National Science Foundation Grant No. GB 1634. 2 BREVIORA No. 237 Type deposition. Holotype, two paratypes, and fragments of a third, all workers, deposited in the Museum of Comparative Zoology (type No. 31171). Description. The 3 intact type specimens have the following dimensions (holotype cited first). The parameters of measurement and the abbreviations are those of Brown (1953); measurements are given in millimeters: TL 4.8, 4.5, 5.0; HL 1.15, 1.12, 1.22; HW 1.02, 1.01, 1.09; CI 88, 90, 89; ML 0.35, 0.34, 0.38; MI 32, 30, 31; WL 1.47, 1.42, 1.57; petiole L 0.49, 0.45, 0.52; index of cephalic depression 59, 60, 59. Form of head and mandibles as shown in Figure 2. Mandibles longer and more slender than in D. cibdella; outer borders distinctly concave; inner borders with about 15 very small, obtuse, back- wardly directed teeth; apical tooth largest, more acute than remainder, which increase slightly in size towards the base. Antennal funiculus slightly incrassate, lacking a segmentally differentiated club. Palpal formula: maxillary 2, labial 2, strue- tural details of palopmeres as in czbdella (Brown and Wilson, 1957, fig. 2). Antennal scrobes as in czbdella; longitudinal dividing carinae feeble; upper scrobe margins more broadly lamellate, almost completely obscuring posterior parts of scrobes in frontal view. Eyes proportionately larger and more convex, maximum diameter in holotype about 0.17 mm. Occipital lobes well de- veloped. Mesosomal profile (Figure 1) indented at metanotal area. Pro- notal dorsum not margined anteriorly, submarginate at the sides. Propodeal structure and dentition as in czbdella, bases of teeth separated by slightly less than twice their length. Propodeal spiracles directed posterolaterally, placed well back on sides of propodeum, not as far back as in czbdella. Petiolar node longer than in cibdella, its dorsal profile more broadly and evenly arched. Postpetiole less transverse, subcircular in dorsal view. Gastric structure generally as in cibdella; first segment relatively narrow, elliptical in dorsal view, about 0.8 X as broad as long. Yellowish white spongiform material developed on _ petiole, postpetiole and gaster, its distribution as in cibdella, the festoons less massive; tergal postpetiolar blocks separated from sternal mass on each side. General sculpture consisting of very coarse longitudinal costula- tion, with a tendency towards reticulation due to the costular crests being slightly irregular in outline, and to the development of weak transverse ribs in the inter-costal grooves. Reticulation more distinct on posterior part of frons, occipital lobes, and especially 1965 ANT GENUS DACETINOPS 3 Dacetinops concinna sp. n., holotype worker. Fig. 1, mesosoma, petiole and postpetiole in lateral view. Fig. 2, head in frontal view; sculpturation omitted, and only hairs nearest the periphery shown. on front of pronotum and postgenal areas of cranium; these parts also with a finely shagreened microsculpture superimposed on the costulae, which are generally smooth and shining elsewhere. About 10 costulae cross the cranium at its broadest point. Costae of mesosomal dorsum less cleanly cut than in c7zbdella; 4 distinet median ones along pronotal dorsum, with weak traces of another at the edge on each side. Sides of mesosoma coarsely and some- what irregularly rugose. Propodeal declivity smooth and shining, with several weak transverse carinae. Petiole and postpetiole moderately coarsely rugo-reticulate. First gastric segment moderately shining, with fine reticulate microsculpture; its basal half with about 10 sharp 4 BREVIORA Now23 straight longitudinal costae, the median four or five strongest, those at extreme sides almost vestigial and more closely spaced. Mandibles smooth and shining. Antennae very finely sha- greened. Antennal scrobes and coxae coarsely shagreened. Re- maining leg segments finely shagreened, and with fine irregular longitudinal striae. Long whitish body hairs extensively developed, proportionately longer than in czbdella (0.12—-0.65 mm) but similarly distributed; many on head (particularly its margins), mesosoma and nodes occupying positions homologous to those of czbdella. Shorter erect to suberect hairs on legs and gaster, finer subadpressed ones form- ing a pilosity on underside of head, anterior faces of fore-coxae and tip of gaster. Color almost exactly as in czbdella, deep reddish brown, mandi- bles, antennae, and legs dark yellowish brown. Discussion. D. concinna is readily distinguished from D. cibdella by numerous characters, most of which are apparent on comparison of the published descriptions and figures of the two species. None- theless these ants constitute a compact and clearly delimited genus, and the generic diagnosis of Brown and Wilson requires only slight modification at this pomt. The statements concerning the antennal club structure, the mesosomal profile, and the mandibular shape and sculpturation need qualification, but these are minor matters. REFERENCES Brown, W. L., JR. 1953. Revisionary studies in the ant tribe Dacetini. Amer. Midland Natur., 50 (1): 1-137 (pp. 7—15). Brown, W. L., Jr., and E. O. WILson 1957. Dacetinops, a new ant genus from New Guinea. Breviora, No. ds let WHEELER, G. C., and J. WHEELER 1957. The larva of the ant genus Dacetinops Brown and Wilson. Brevi- ora, No. 78: 1-4. (Received 15 September, 1965) BREVIORA Museum of Comparative Zoology CAMBRIDGE, Mass. FEBRUARY 25, 1966 NUMBER 238 AN EVALUATION OF JAMAICAN DROMICUS (SERPENTES, COLUBRIDAE) WITH THE DESCRIPTION OF A NEW SPECIES By Donatp W. BupDEN Department of Biological Sciences University of Idaho, Moscow, Idaho Grant (1938) evaluated the populations of Dromicus on the island of Jamaica and distinguished two species, D. callilaemus (Gosse, 1851), and D. funereus (Cope, 1862). His review was based upon the examination of only seventeen individuals, sev- eral of which lacked adequate locality data. The original de- scriptions incorporated data from only two specimens of D. callilaemus and three specimens of D. funereus. Recently, much more material has been made available for study. A long series of Dromicus collected by Schwartz and others during the summer of 1961 together with material bor- rowed from several museums (a total of 227 specimens) has been utilized in the present study. The examination of a large num- ber of individuals has made possible an analysis of the popula- tions as well as an accumulation of data heretofore not present in the literature. A re-evaluation of the classification is necessary. The population regarded by Grant (1938) as D. funereus actually consists of at least two forms. Present evidence indicates that these forms are best treated as distinct species. All measurements are in millimeters. The methods used in obtaining scale counts have been taken from Dowling (1951), and the terminology in regard to hemipenial descriptions has been derived mainly from Dowling and Savage (1960). Several paratypes of the new species have been deposited in the Ameri- can Museum of Natural History (AMNH), and the University of Illinois Museum of Natural History (UIMNH), as well as in other collections. 2 BREVIORA No. 238 DROMICUS CALLILAEMUS (Gosse, 1851) Natrix callilaema Gosse, 1851, Naturalist’s Sojourn in Jamaica, p. 384. Type locality: Bluefields, Jamaica. Since only two of the three (fide Boulenger, 1894:143) syn- types of D. callilaemus were examined for the present study, it seems expedient to designate one of these specimens as a lecto- type, thus eliminating the possibility of confusion should the third specimen prove to be representative of a form other than D. callilaemus. Description of the lectotype: British Museum (Natural His- tory) 1946.1.5.90, Bluefields, Westmoreland Parish, Jamaica; juvenile, total length 157; tail length 57; ventrals 136; paired subeaudals 112; supralabials 7/7; infralabials 9/9; prefrontals 2; loreals 1/1; preoculars 1/1; postoculars 2/2; temporals 1 + 2/1 + 2; dorsals smooth, dorsal row formula 19-19-17. Inter- nasal as lone as broad, internasal suture considerably shorter than prefrontal suture ; frontal longer than broad, much broader than supraoculars, about as long as its distance to the tip of the snout. A dark cap on the frontal scale, accentuated by darker pigmentation on the anterior portions of the supraoculars and the posteromedial borders of the prefrontals; cap narrowing posteriorly along the parietal sutures, diffusing into darker coloration on the neck; dorsum pale except for darker pigmen- tation in the area of the head and neck, pattern consisting of two rows of dark paramedian punctations on a lght ground color and a dark lateral stripe incorporating scale rows 3, 4, and 5; venter pale except for dark mottling in the area of the chin and throat; dark pigmentation more concentrated along the sutures of the infralabials, supralabials and the anterior chin shields. Because of its small size the specimen could not be sexed; the subcaudal count, however, correlates with those of males from other localities. Distribution: Jamaica, apparently island-wide. Redefinition of species: A relatively small, slender, brick red snake with a distinct head pattern; ventrals 133-141 in males, 130-140 in females; anal divided; subcaudals paired, 106-122 in males, 92-109 in females; ventrals + subcaudals 242-262 in males and 231-242 in females; largest male 445, largest female 465; ratio of tail length to snout-vent length 0.57-0.75 in males and 0.50-0.73 in females; dorsal pattern usually consisting of three indistinct stripes, the lateralmost found on scale rows Oo 1966 JAMAICAN DROMICUS 3, 4 and 5, occasionally restricted to 3 and 4 or 4 and 5; pigmen- tation of stripes darker than ground color; venter pale, un- marked except for dark stippling or mottling in area of chin and throat, anterior scales often with light centers (Fig. 2, left) ; labials pale but mottled, flecked or stippled with darker pigmen- tation; a narrow dark bar extending from naris to the anterior border of the eye, bar continuing as a broad postocular stripe (Fig. 1, top); a dark cap on the frontal scale variously di- lated to include portions of the supraoculars, prefrontals and Fig. 1. Top: Dromicus callilaemus, lateral view of head, MCZ 69070, from Mona, St. Andrew Parish, Jamaica. Bottom: Dromicus funereus, lat- eral view of head, ASFS 15922, from Christiana, Manchester Parish, Jamaica. nasals; cap usually narrowing posteriorly along the parietal sutures and continuing as a middorsal stripe, at times diffusing into the darker ground color of the neck; prefrontals, nasals, and supraoculars often with a dark spot; rostral pale, usually with a dark inverted U-shaped bar; pattern more pronounced in juven- iles; young specimens also with many flecks or punctations on a lhghter dorsum; supralabials 7/7; infralabials 9/9; pre- frontals 2; loreals usually present 1/1; preoculars 1/1; postocu- lars 2/2; temporals 1 + 2/1 + 2; dorsals smooth, dorsal row formula 19-19-17. 4 BREVIORA No. 238 Hemipenis (based upon the incompletely everted left organ of MCZ 69078) as follows: organ slightly bilobed, extending to the level of the ninth or tenth subcaudal; sulcus spermaticus bifureating at the level of the third subcaudal; large spines near basal and midportions of the organ, remainder of organ covered with small spines. The faded condition of the specimens examined precludes a detailed color description. The following comments have been appended from Barbour (1910): ‘‘All the Kingston specimens are red when adult with scattered marks on the sides of the head and throat. The young have a blackish stripe along the body and are spotted with dark brown. The throat and head are heavily marked with brown. The ground color, however, is the same as in the full-grown examples, viz., brick red.’’ Variation: Two specimens (USNM 108531 and MCZ 73761) have 8/9 and 5/8 infralabials, respectively ; MCZ 7374 lacks both loreal scales, and MCZ 69073 maintains only the right loreal, considerably reduced in size; USNM 102650 has a dorsal row formula 19-17-17; temporal scale counts are generally 1 + 2/1 + 2, but these scales may be variously fused or divided. Minor variation in color and pattern intensity is frequent. The head and dorsal pattern may be reduced, nearly absent, or somewhat obscured by darker pigmentation. This condition Fig. 2. Left: Dromicus callilaemus, ventral view of head, MCZ 52255, from Galloway Gap, St. Andrew Parish, Jamaica. Right: Dromicus funereus, ventral view of head, ASFS 15922, from Christiana, Manchester Parish, Jamaica. 1966 JAMAICAN DROMICUS 5 occurs in several specimens from Montego Bay, Kingston, and Port Antonio. Specimens examined: Jamaica, Westmoreland Parish, Blue- fields, two (BMNH 1946.1.5.90 [lectotype], 1946.1.5.91 [para- lectotype|); St. James Parish, Montego Bay, two (MCZ 52254, USNM 42881), Greenwood, one (SMIJ); St. Elizabeth Parish, Malvern, one (SMIJ); Manchester Parish, Mandeville, one (UMMZ 85948); St. Andrew Parish, Kingston and immediate vicinity, 47 (CM 33100, 33112; MCZ 7374, 18887, 44207-15, 44903-04, 52255, 53147, 66893, 69072-81, 69181, 73761; USNM 36662-63, 108079-80, 102650, 38947-48; UMMZ 85949, 85951; SMIJ [8 specimens] ) ; St. Thomas Parish, Arntully, two (USNM 79637-38) ; Portland Parish, Boston Bay, two (USNM 108530- 31), Windsor, one (USNM 117675). DROMICUS FUNEREUS (Cope, 1862) Alsophis funereus Cope, 1862, Proc. Acad. Nat. Sei. Philadelphia, p. 77. Type locality: ‘‘ Jamaica.’’ In view of the fact that the only specimen of D. funereus cited by number, by Cope, has been lost, as indicated by Stejneger in Grant (1938:85), the designation of a lectotype from the two extant syntypes is in order. Description of the lectotype: USNM 12372, adult male; total length 378; tail length 140; ventrals 134; anal divided; paired subeaudals 90; supralabials 7/7; infralabials 9/9; prefrontals 2; loreal 1/1; preoculars 1/1; postoculars 2/2; temporals 1 + 2/1 + 2; dorsals smooth, dorsal row formula 19-19-17; internasals as long as broad; internasal suture slightly shorter than prefrontal suture; frontal slightly longer than broad, about twice as broad as supraoculars, about as long as its distance from the tip of the snout; loreal small, somewhat square; pigmentation dark, speci- men appears somewhat faded; no discernible pattern. Distribution: Western Jamaica, eastward as far as Bog Walk and Port Maria. Definition: A relatively small, stout, brownish black snake, generally lacking a conspicuous head or dorsal pattern; ventrals 123-139 in males and 128-140 in females; anal divided; sub- caudals paired, 78-91 in males, 62-78 in females; ventrals + subeaudals 203-224 in males and 192-218 in females; largest male 479, largest female 462; ratio of tail length to snout-vent length 0.438-0.59 in males, and 0.38-0.47 in females; tail length/snout- vent length ratio somewhat lower in juveniles than in adults; —_ Ye BREVIORA No. 238 dorsum generally dark, unpatterned ; venter usually dark, similar in intensity to the dorsum, paler anteriorly in area of chin and throat with a suffusion of darker pigmentation (Figs. 2 and 4) ; juveniles often exhibiting a dorsal pattern of longitudinal stripes; head with a dark cap generally darker than the sur- rounding ground color, and venter darker than ground color of dorsum; supralabials 7/7; infralabials 9/9; prefrontals 2; loreals 1/1; preoculars 1/1; postoculars 2/2; temporals 1 + 2/1 + 2; dorsals smooth, dorsal row formula 19-19-17. An account of the hemipenial configuration based upon the examination of the right extruded organ of ASFS 15172 is as follows: a single relatively slender, clavate organ extending to the level of the eighth subeaudal; suleus spermaticus bifurcating on the apical surface with each branch terminating on opposite sides of a fleshy rowel (Fig. 3, left); organ virtually nude from the base to the level of the seventh subcaudal, thence be- coming spinose to the apex, the spines generally small though somewhat larger on the non-suleate surface. Variation: Anomalies in numbers of certain head scales were occasionally encountered; scales involved were: supralabials, Fig. 3. Left: Dromicus funereus, hemipenis, view of suleate surface, ASES 15172, from 1 mi. S Reading, St. James, Jamaica. Right: Dromicus polylepis, hemipenis, view of suleate surface, ASFS 13536, from Port Antonio, Portland Parish, Jamaica. 1966 JAMAICAN DROMICUS i infralabials, prefrontals, preoculars and postoculars. Color varia- tions appear negligible. ASFS 15991 and 15994 were noted as being partially erythrystic in life. Juveniles differ considerably from adults in possessing a relatively well marked dorsal pattern on a light ground color. Specimens examined: Jamaica, St. James Parish, Montego Bay, two (USNM 108337-38), 5 mi. W Montego Bay, seven (MCZ 44897-900; UMMZ 85941-43), 1 mi. S Reading, 59 (ASFS 147438, 14932, 15060-63, 15066-67, 15146-48, 15159-77, 15264-92) ; St. Elizabeth Parish, Balaclava, two (USNM 73276-77) ; Trelawny Parish, Quick Step, one (USNM 117676), Windsor, two (USNM 83719-20) ; Manchester Parish, Christiana, 39 (ASFS 15827-29, 15916-25, 15982-16007), Mandeville, 12 (MCZ 13294-96, 44895-96, + two untagged specimens) ; USNM 31096, 108240-41; UMMZ 85946-47) ; Williamsfield, one (USNM 108533) ; Clarendon Par- ish, Grantham, eight (ASFS 16091-98); St. Ann. Parish, at mouth of Roaring River, one (MCZ 28092); St. Mary Parish, 4 mi. S Port Maria, two (MCZ 44901-02) ; St. Catherine Parish, Bog Walk, two (UMMZ 85944-45) ; Jamaica (no other locality designation) one (USNM 5780, paratype). DROMICUS POLYLEPIS new species Holotype: MCZ 81020, an adult male, from Port Antonio, Portland Parish, Jamaica, collected 19 June 1961 by native Jamaican collector. Original number ASFS 13122. Paratypes: Portland Parish, Jamaica, all from Port Antonio, as follows: AMNH 94171-75, native collector, 22 June 1961; AMNH 94176, USNM 152589, native collector, 23 June 1961; USNM 152590-92, native collector, 26 June 1961; UIMNH 56911-14, native collector, 27 June 1961; ASFS 13533-36, native collector, 28 June 1961; ASFS 13606-09, native collector, 30 June 1961; MCZ 81021, native collector, 29 June 1961; MCZ 7363-64 plus two untagged specimens (field tags 39 and 66), A. E. Wright, no date. Associated specimens: Portland Parish, Port Antonio, one (field tagged specimen, number 65, duplicate of series MCZ 7363-64) ; St. Andrew Parish, Constant Spring, one (SMIJ), Cross Roads, one (SMIJ). Distribution: Eastern Jamaica: known only from Portland Parish, Port Antonio, and St. Andrew Parish, Constant Spring and Cross Roads. 8 BREVIORA No. 238 Diagnosis: A relatively small (largest male 455 mm, largest female 412 mm), slender (juveniles) to stout (adults), brownish black snake, usually lacking any conspicuous pattern; ventrals + subeaudals 230-245 in males, and 222-237 in females; hemi- penis single, somewhat bulbous, and heavily spinose. Description of holotype: An adult male, total length 387, tail length 150; ventrals 137; anal divided; paired subecaudals 106 ; head scalation typical of a generalized colubrid; internasals as long as broad, internasal suture slightly shorter than prefrontal suture; frontal longer than broad, broader than supraoculars, as long as its distance to the tip of the snout; loreal nearly square ; supralabials 7/7; infralabials 9/9; prefrontals 2; loreals 1/1; preoculars 1/1; postoculars 2/2; temporals 1 + 2/1 + 2; dorsals smooth, dorsal row formula 19-17-17; dorsum dark brown to black with a pair of ight brown lines of blotches, lower two or three scale rows paler; venter black, suffused with brown anteriorly. Variation: Ventrals 136-141 in males, and 137-146 in females ; paired subeaudals 92-106 in males, 76-93 in females; ventrals + subeaudals 230-245 in males, 222-237 in females; ratio of tail length to snout-vent length 0.44-0.63 in males, and 0.42-0.54 in females; supralabials 7/7; infralabials 9/9; prefrontals 2; tem- porals 1 + 2/1 + 2; a dorsal row formula of 19-19-17, occasionally 19-17-17; shght deviation from the normal number of head scales usually involving supralabials, infralabials, and postoculars. An account of the hemipenial configuration based upon the ex- amination of the right everted organ of ASFS 13536 (Fig. 3, right) is as follows: organ single, relatively stout and somewhat bulbous; apex extending to the level of the eighth or ninth sub- eaudals; sulcus spermaticus bifurecating at a distance from the base approximately equivalent to three-fifths the total length of the organ, each branch terminating in a fleshy apical papilla: a series of relatively large spines extending along either side of the unbranched portion of the sulcus, diverging laterally just proximal to the bifurcation of the sulcus and becoming confluent on the non-suleate surface, forming four or five distinct rows at the level of the sixth and seventh subcaudals; spines some- what reduced in size and number near the base of the organ; remainder of organ covered with small undifferentiated spines ; a depression lateral to each apical papilla forming a fleshy ridge connecting the non-suleate and suleate surfaces. Dorsum generally dark, unpatterned; venter usually dark, similar in intensity to the dorsum, paler anteriorly in area of 1966 JAMAICAN DROMICUS 9 chin and throat with a suffusion of darker pigmentation; varia- tions in color intensity negligible, young specimens tending to be somewhat lighter. Comparisons: Although quite similar to fuwnereus in colora- tion, polylepis may be distinguished on the basis of the total number of underbody scales and the configuration of the hemi- penis. The ventral plus subcaudal counts of polylepis (230-245 in males, 222-237 in females) are more than those of funereus (203- 224 in males, 192-218 in females) in specimens of the same sex, and the hemipenis is stout, bulbous, and heavily spined as op- posed to the relatively slender, clavate, lightly spined and rowelled organ of funereus. In addition, males of polylepis have a higher subcaudal count (92-106) than males of funereus (78- 91). There is no indication of overlap in distribution between the populations of these two species. Occurring sympatrically with callilaemus in eastern Jamaica, polylepis may be distin- guished from the latter by its dark coloration and lack of con- spicuous head and dorsal patterns. In addition to differences in the structure of the hemipenis, specimens of polylepis generally exhibit a lower subeaudal count (males 92-106, females 76-93) than specimens of callilaemus (males 106-122, females 92-109). The single-structured hemipenis of polylepis distinguishes it from the remaining forms of West Indian Dromicus, all of which are characterized by a bilobed organ. Remarks: D. polylepis is the third species of Dromicus de- scribed from Jamaica. Except for this instance, not more than one species of Dromicus occurs on any one island in the West Indies. In general morphology, coloration, and pattern, polylepis is strikingly similar to funereus, and the two may be regarded as sibling species. In addition to D. polylepis, a lizard (Anolis grahami aquarum Underwood and Williams, 1959), and a hum- mingbird (Trochilus polytmus scitulus Brewster and Bangs; ef. Bond, 1956) have undergone differentiation in extreme eastern Jamaica, although in neither of these cases has the differentia- tion reached the status of species. ACKNOWLEDGEMENTS I wish to express my thanks to Dr. Albert Schwartz for use of specimens in the Albert Schwartz Field Series (ASFS), for obtaining the loan of the remainder of the specimens examined, and especially for his assistance in the laboratory; to Richard Thomas for his assistance in the laboratory as well as for the 10 BREVIORA No. 238 illustrations of the hemipenes of funereus and polylepis; and to Ronald F. Klinikowski for the illustrations of the lateral and ventral head views of callilaemus and funereus. For the loan of comparative material I wish to extend thanks to the following: James C. Bohlke and Edmond V. Malnate, Academy of Natural Sciences of Philadelphia; Miss Alice G. C. Grandison, British Museum (Natural History) (BMNH); Neil D. Richmond, Carnegie Museum (CM); Ernest E. Williams, Museum of Comparative Zoology (MCZ) ; C. Bernard Lewis, Sci- ence Museum, Institute of Jamaica (SMIJ); Charles F. Walker and George R. Zug, Museum of Zoology, University of Michigan (UMMZ), and Doris M. Cochran and James A. Peters, United States National Museum (USNM). LITERATURE CITED BARBOUR, THOMAS 1910. Notes on the herpetology of Jamaica. Bull. Mus. Comp. Zool., 52(15) :273-301. BOND, JAMES 1956. Checklist of birds of the West Indies. Acad. Nat. Sci. Philadel- phia, ed. 4, ix + 214 pp. BOULENGER, G. E. 1894. Catalogue of the snakes in the British Museum (Natural His- tory). London, vol. 2, xiii + 382 pp. DOWLING, HERNDON G. 1951. A proposed standard system of counting ventrals in snakes. British Jour. Herpetology, 1(5) :97-99. ,and JAY M. SAVAGE 1960. i Veneers A Wiitetrerd Ss CSL i ( L/ t Fig. 1. Anolis fraseri, MCZ 70227. Dorsal view of head. 4 BREVIORA No. 239 TABLE | South American biporcatus Head scales, at least anteriorly, keeled. 2-3 scales between circumnasal and rostral. Supraorbital semicircles sepa- rated by 0-3 scales. Supraocular seales keeled. 2-3 more or less elongate supra- ciliary scales followed by a dou- ble series of scales which become smaller posteriorly and merge into granules like those medial to the series. 6-9 loreal rows. Interparietal small (but some- times larger than the ear which is usually very small). Subocular scales usually com- pletely and always at least partially separated from the supralabials by a row of inter- vening scales. 7-12 supralabials to center of eye. Dewlap scales strongly keeled, almost as large as ventrals, rather uniformly distributed on the small fan. Ventral scales strongly keeled, mucronate. 22-26 lamellae under phalanges ii and 11 of 4th toe. fraserv ITead scales never keeled, usu- ally flat, pavementose, but some- times swollen or wrinkled. 1-2 scales between ecircumnasal and rostral. Supraorbital senucircles sepa- rated by 2-4 scales. Supraocular scales smooth. Usually no elongate supracili- ary scales; mstead, the whole supraciliary margin of round or squarish scales distinetly larger than the granules medial to them. 4-9 loreal rows. Interparietal larger (larger than the ear which is moderate in size). Subocular scales broadly im con- tact with supralabials. 6-9 supralabials to center of eye. Dewlap scales smooth, small, almost granular, grouped im rows which are separated by naked skin. Ventral scales smooth or weak- ly keeled. 20-24 lamellae under phalanges i1 and iii of 4th toe. 1966 ANOLIS BIPORCATUS AND ANOLIS FRASERI 5 in one species completely overlaps the known range of the other, I have omitted this character. The most striking differences are emphasized by italicizing the pertinent characters for fraseri. Both are large, short-legged, stout-bodied, relatively short- headed forms. The similarities are indeed amazing since the two species are, according to Etheridge, osteologically in totally different sections of the genus. The similarities would be even greater if Mexican and other northern biporcatus were com- pared with fraserz since the northern animals tend to have 3-4 scales between the supraorbital semicircles (cf. Stuart, 1955). Overlapping variation in many features, as shown in Table 1, is characteristic of many sympatric species in Anolis and is one of the several reasons why this genus is described as ‘‘difficult.’’ As in all such eases, the recognition of the valid species depends upon recognition of the constant (or almost constant) associa- tion of characters, however trivial. Smooth or keeled scales, sub- oculars in contact with or separated from supralabials, inter- parietal large or small are none of them characters universally useful in Anolis. The distinctive supraciliaries of fraseri are more useful because a substantial morphological gap between this condition and that shown by biporcatus will always distin- euish these two species. (The frasert supraciliaries occur, how- ever, also and in a more exaggerated degree in latifrons; the condition is not unique to fraser.) a a 0) i ZE PE ES ; Oueeezes LEELA] BRO CEG Sho eeaess Ee le AMO EE eee ~ ¢ ic ~ Yr OY = Cl 7 aya C BOCES Fig. 2. Anolis biporcatus parvauritus, new subspecies, MCZ 78935. Dorsal view of head of type. 6 BREVIORA No. 239 INTRASPECIES VARIATION: A. BIPORCATUS While I have not endeavored to examine all specimens of A. biporcatus from all parts of its range, I have sampled the northern populations (Panama, Costa Rica, Guatemala, Mexico), and I have had the especial advantage of examining the series of 122 specimens taken by Slevin near Boquete, Panama (Slevin, 1942) 1 Having seen this series I am impressed by the similarities of the various populations. There is indeed considerable variation within A. biporcatus, as I understand it. Thus, the number of scales between the semicircles, the number of seales between interparietal and semicircles, the number of loreal rows, the amount of keeling of the head scales, the contact or separation of suboculars and supralabials, and the length of the hind limbs all appear to be quite variable from individual to individual. Yet the species has a strong habitus character and is quite recog- nizable in spite of very striking variation. A very few characters appear to sort out geographically. The supraocular scales and those surrounding the interparietal are extraordinarily small in Ecuadorian and southern Colombian specimens. In these same populations the ear opening is circular and very small in contrast to the rather oval opening of northern specimens. Since there seem also to be some color differences of both body and dewlap in these same populations — which are precisely those sympatric with frasers—it seems permissible to regard the Keuadorian and southern Colombian animals as representing a distinct subspecies. Some discussion of available names is there- fore required. Anolis solifer Ruthven, type locality La Concepcién, Santa Marta Mountains of northern Colombia, is the only name thus far proposed for a South American population of this species. That solifer does belong in biporcatus and may provisionally be synonymized with the nominate race seems clear from the com- parison of the type with Panamanian and more northern material. The type of solifer can be matched by Panamanian material (except for the exceptionally short hind limbs) and Panamanian material is not separable by any constant characters from more northern specimens. The character of the very short hind limbs may define a population in northeast Colombia and Venezuela. One of the two available Venezuelan specimens has limbs as 1] am indebted to Dr. Allan Leviton of the California Academy of Sciences for this opportunity. 1966 ANOLIS BIPORCATUS AND ANOLIS FRASERI if short as those of the solifer type but the other does not. In any event, the name solifer is clearly not available for the distinctive Eeuadorian and south Colombian populations. In the critical characters solifer resembles northern biporcatus. The other names referable to the species biporcatus are equally unavailable. None of them refer to South American material. The nominate race biporcatus Wiegmann 1834 has the type locality ‘‘Mexico,’’ now restricted to Piedra Parada, Chiapas. A. copet Bocourt 1873 has the type locality Santa Rosa de Pansas, Guatemala. Anolis brevirostris Peters 1873, next in date, has the type locality Chiriqui Prov., Panama. Anolis brevipes Boettger 1893 came from Cairo Plantation, La Junta near Limén, Costa Rica. Every one of these seems at present best grouped under the name biporcatus. The southern race is thus without a name. In reference to the small ear it may be ealled: ANOLIS BIPORCATUS PARVAURITUS new subspecies Holotype: MCZ 78935, an adult ¢, ‘‘banana plantation, woods and penal colony camp, northern Gorgona Island, Cauea, Colombia, 5-45 meters altitude,’’ collected by F. Medem and lL. Salazar G., 1 to 23 February 1961. Paratypes: Colombia. Cauca: MCZ 78933-34, 78936-41, same data as type. Narino: MCZ 79142, Rio Mataje. Ecuador. Es- meraldas: USNM 157105, Cachavi; BM 1901.6.27.1, Carondolet ; < awe sO) e Ss Ss cn we “ Se i o ne, SS = IPS Ss PY AW ESI Se SSS TR SST YO SSS Se Sana Fig. 3. Anolis biporcatus biporcatus, MCZ 56248, from Los Diamantes, Costa Rica. Dorsal view of head. 8 BREVIORA No. 239 AMNH 6967, Rio Capayas; BM 1907.7.29.10-11, ZMB 18213, VM 12755, Rio Sapayo; BM 1901.8.3.1, Salidero; BM 1901.6.27.1, San Javier; Orces 3898, San Lorenzo. Imbabura: BM 98.4.28.18, Paramba. Pichincha: Orces 4596, Santo Domingo de _ los Colorados. ‘‘ West Ecuador’’: BM 1946.8.13.21 (fraseri syntype). ‘*Heuador’’: USNM 20610, ANSP 7910. Diagnosis. A subspecies of Anolis biporcatus distinguished by the small size of the scales surrounding the interparietal which are smaller than or only equal to middorsal scales (instead of distinctly larger than middorsal scales), by the small round ear rather ventrally placed (instead of a large oval ear extending well dorsally), and by a black-edged dewlap. Comment. None of the conventional numerical counts in Anolis (e.g. scales across snout, scales between semicircles, scales between interparietal and semicircles, ete.) reveal the distinet- ness of parvauritus, though it is visually apparent immediately on inspection of the specimens (Figs. 2 and 3). The number of scales between interparietal and semicircles might be expected to eorrespond neatly to the very striking difference in the size of seales surrounding the interparietal in the nominate race and in parvauritus. In faet, however, there is wide overlap, and no discrimination could be based on the numerical character. How- ever, comparison of the scales around the interparietal and the scales on the midline of midbody of the same animal gives instant conviction of the difference between the two races. The character of the small size of supraocular scales is equally apparent on inspection. Here there is in parvauritus an absence of clear definition of any supraocular disk. This, however, 1s true also in many specimens of the nominate race. Here, again, comparison with middorsal scales (or direct confrontation with representatives of the northern race) convinces one of the exist- ence of a difference. It is, however, a difference which is some- what less striking than in the case of the scales surrounding the interparietal. The size of the ear is again a character easily appreciated visually (Fig. 4), but it does not reveal itself by comparison with the size of the interparietal, which itself varies and may be very small. Comparison with the temporal area is more useful but the size and shape of the ear are sometimes distorted by preservation. Certain specimens seem intermediate: MCZ 79656-57 from Villa Arteaga, Antioquia, Colombia, have the ear round but rather large. In other regards (scales around interparictal, supraocular scales) these are northern in type. MCZ 79842 from Sincelejo, 1966 ANOLIS BIPORCATUS AND ANOLIS FRASERI 9 They EL ee | >= Fig. 4. Ear size in A. biporcatus biporcatus (top) and re I) POPPE Z OW Yjsuel [t }0J UT Ut QZO PALA ELL AAAS od£yoloy “oun dowy)a . 0 Up I ONS 6 BREVIORA No. 241 LITERATURE CITED BERTIN, LEON 1947. Notules ichthyologiques. 1.— Compléments sur le genre Avocet- tinops. Bull. Soc. Zool. France, 72:54-55. BOuLKE, JAMES E., and FRANK S. CLIFF 1956. A discussion of the deep-sea eel genus Avocettinops, with notes on a newly discovered specimen. Copeia, 1956:95-99. NORMAN, J. R. 1939. Fishes. Sci. Rep. ‘‘ John Murray’’ Expedition, 7(1) :116 pp. RouLE, Louis, AND LEON BERTIN 1924. Notice préliminaire sur la collection des nemichthydés recueillie par l’expédition du ‘‘Dana,’’ (1921-1922), suivie de considéra- tions sur la classification de cette section des poissons apodes. Bull. Mus. Nat. Hist. Natur., Paris, 30:61-67. 1929. Les poissons apodes appartenant au sous-ordre des Nemichthydi- formes. Rep. ‘‘Dana’’ Exp., 1920-1922, No. 4, 113 pp., 19 pls. (Received 9 November 1965.) r ‘ BREVIORA Museum of Comparative Zoology CAMBRIDGE, MAss. FEBRUARY 25, 1966 NuMBER 242 THE SUPPOSED ‘‘SPONGE SPICULES” OF MERRILL, 1895, FROM THE LOWER CRETACEOUS (ALBIAN) OF TEXAS By WituiaM A. S§. SARJEANT Geology Department University of Nottingham, Nottingham, England INTRODUCTION In 1895, a paper was published by J. A. Merrill giving the results of a study of chert nodules from Texas. In this he de- seribed and figured a number of microscopic structures which were interpreted as sponge spicules and referred, hesitantly, to six known genera; eight new species were proposed, all of which were doubtfully attributed to the living genus Geodia. This work eseaped attention until 1945, when Glaessner noted that the supposed spicules of Geodia were in fact hystricho- spherids (p. 20). As such, they constitute the first Lower Cre- taceous hystrichosphere assemblage and hence the first record of Mesozoic dinoflagellate cysts from the United States. Subse- quently, the present author proposed the provisional reallocation of Merrill’s species, on the basis of study of his figures and de- scriptions, to various dinoflagellate cyst genera (Sarjeant, 1964, pli). Through the courtesy of Dr. H. B. Whittington, the author has been permitted to examine Merrill’s type material, from the collections of the Museum of Comparative Zoology, Harvard. The results of this study are presented here. STRATIGRAPHY AND MATERIAL The specimens studied by Merrill were nodules of chert col- lected by Mr. Edward E. Cauthorne from ‘‘a quarry near Austin, Texas.’’ They were obtained from the Caprina Chalk (now known as the Edwards Limestone), a member within the D BREVIORA No. 242 Fredericksburg Formation, Comanchean Stage. This formation is considered to be equivalent to the middle and the lower part of the Upper Albian of Europe (Murray, 1961, p. 349). The material in the Harvard collections consisted of three shdes and seventeen partial nodules or slices of chert (MCZ 185A). Two of the three slides measure 2” by 1” and bear sections of flint cut to a thickness of between two and three times that of a normal petrological section (0.03 mm); these were unlabelled and are here referred to as slides I and II. The third shde consisted of a square of glass measuring 3” by 2” by 1,” thick, bearing a wedge-shaped chert slice (varying in thick- ness from approximately eight times to approximately twice the thickness of a normal petrological section) which was surrounded by Canada balsam in an irregular ring. The shape of the balsam ring indicated that the chert slice had originally been fully twice its present size, but that one half had become detached and lost at some period. The thickness of the glass mount was too great to permit satisfactory examination; the surviving portion of chert was therefore remounted for study (shde IIT). Merrill states that ‘‘several slides’? were made from each nodule ; it is therefore possible, in absence of any clear labelling, that the three surviving slides do not contain the holotype mate- rial of his species. As will be detailed later, a number of speci- mens showed a general correspondence to one or other of Mer- rill’s figures, but the figures are unaccompanied by photographs and are not drawn with sufficient accuracy to permit any certain linking of specimen to figure. There remains a residue of figures which do not correspond even in a general fashion to any speci- men located; it must be presumed in these eases, either that the holotypes were contained in other slides now lost, or that they were contained in the missing half of flint section ITI. With Dr. Whittington’s courteous permission, sections were made from the remaining chert specimens. Nine further mounts were made and examined (slides IV-XII); they were cut to a thickness of between three times and twice that of a normal petrological section. It was hoped that the lithological character would prove sufficiently variable to enable correlation of the new with the old slides and thus to link the latter to their source nodules; unfortunately, this did not prove possible, the sections showing insufficient distinctive features. Merrill also mentions (p. 10) having made a comparative study of English cherts; three sections were made by Merrill, but these proved virtually barren of microorganisms and thus 1966 CRETACEOUS CHERT FOSSILS 3 cannot correspond to the surviving slides. Similarly, since Ene- lish cherts are not mentioned on the specimen label, it must be assumed that the surviving chert specimens are entirely from Texas. GENERAL DISCUSSION OF MERRILL’S RESULTS The earliest studies of microfossils in chert flakes were made by C. G. Ehrenberg, in 1836; he assumed wrongly that the hystri- chospheres were silicified and placed them in the freshwater desmid genus Xanthidium. The name ‘‘xanthidia’’ was used also in descriptions of similar fossils by a group of English microscopists, but G. A. Mantell conclusively demonstrated their organic rather than siliceous composition and thus precluded placing them in Xanthidium (1845). Merrill was clearly unaware of these earlier studies and re- peated Ehrenberg’s error of assuming his microfossils to be silicified. He states (p. 8): ‘‘...we must suppose either that the spicules have been replaced by amorphous silica, or that they are still in the hyaline or colloidal state as found, and coloured yellow by some organic agent perhaps.’’ He noted the presence in the cherts of foraminifera (rotaliids and textularids), either as ‘‘ehosts’’ entirely replaced by silica, or with the organic shell linings persisting; the latter he described as ‘*. . . replaced by amorphous silica, the outlines remaining in a dark substance which has the appearance of an organic residue.’’ This is one of several indications that Merrill was unaware of, or discounted, the possibility of structures formed of organic substances incor- porated into the cherts and persisting almost unchanged. He effectively assumed the whole microfossil content of the chert sections to be originally siliceous or secondarily silicified. This attitude clearly orientated his thinking during his study, for, as will be shown in the ensuing section, the microfossils he described as sponge spicules include spores, pollen, hystrichospheres, wood and plant fragments, and even carbonate crystals. Merrill noted the relatively low concentration of microfossils in the chert (p. 6):‘*...the number of organic remains is few, and the massive silica greatly in excess.’’? This comment is wholly endorsed by the present author; the number of microfossils other than Foraminifera encountered in the nine new slides (IV-XII) averaged less than two. Foraminifera were comparatively abun- dant, averaging more than a dozen per slide. 4. BREVIORA No. 242 DISCUSSION OF THE FORMS FIGURED BY MERRILL In the presentation of his results, Merrill described the forms sequentially according to their arrangement in his plate (here reproduced as Text-fig. A). This procedure is again followed here: the numbers, and names given by Merrill (in quotes), are first stated, and a reinterpretation of the specimens follows. 1. and 2. ‘‘Monactinelid: Axinella ? sp.’’ The figured spec- imens were not relocated; however, they appear to consist of sections through shell fragments which have been entirely re- placed by silica. Merrill noted (p. 7) the finding of supposed shell fragments replaced by silica, but described them as a ‘bright transparent yellow’’ and thus was probably misinter- preting organic tissue fragments. 3. and 4. ‘‘Monactinellid: Reniera ? sp.’’ Also probably oblique sections through shell fragments replaced by silica. The figured specimens were not positively relocated, but a number of similar fragments were noted. 5. ‘‘Monactinellid; acuate spicule.’’ Neither description nor figure is sufficiently informative to permit either certain recog- nition of the figured specimen or precise determination of its true nature. However, it is most probably a fragment of organic tissue (?plant) ; such fragments are quite common in the cherts. 6. ‘‘Monactinellid.’’ The figure and description support the presumption that this is a wood fragment ; such fragments occur infrequently in the cherts. 7. ‘‘Monactinellid.’’ Described by Merrill (p. 13) as ‘‘the most abundant of the sponge spicules found in the flint’’; micro- organisms of this type, varying considerably in the detail of size and shape, were encountered in all twelve cherts examined; the figure is not adequate to permit certain recognition of the holo- type. The presence of septae crossing the cone at right angles to its long axis makes it probable that these are the shell linings of unilocular foraminifera (PI. 1, fig. 1). 8. ‘‘Monactinellid.’’ Also most probably a wood fragment; not relocated. 9. ‘*Monactinellid: Esperites ? sp.’’ Although stated by Mer- rill to be ‘‘a very common form’’ (p. 13), this figure was not correlated with certainty with any microstructure seen in the slides. It may possibly represent a replaced shell fragment or a fragment of organic tissue. 1966 CRETACEOUS CHERT FOSSILS 5 10. ‘*‘Monactinellid: Renicra ? sp.’’ Whilst neither figure nor description permit certainty, it is most probable that this repre- sents a fragment of organic tissue (?plant). Plant fragments of similar appearance, some arcuate, were encountered with fair frequency. 11. ‘‘Tetractinellid: Geodia ? austini n. sp.’’ This species was tentatively transferred to the genus Systematophora in an earlier paper (Sarjeant, 1964, p. 175). A specimen closely resembling the holotype is present in slide I (specimen IA: Pl. 1, fig. 10). From examination of both specimen and figure, there can be no question that this species is a junior synonym of Hystricho- sphaera ramosa (Khrenberg, 1838) O. Wetzel 1933, emend. Davey and Williams 1965a. The species name austini must therefore be rejected, under Article 63 of the ‘‘ International Code of Botan- ical Nomenclature.’’ 12. ‘‘Tetractinellid: Geodia ? cretacea n. sp.’’ This species was transferred to the genus Baltisphaeridium in an earlier paper (Sarjeant, 1964). Several representatives of this species have been encountered in slide III. The holotype cannot be dif- ferentiated with complete certainty; the specimen most similar to the figure is specimen III J (PI. 1, fig. 6). Hystrichospheres of this type are frequent in the Upper Cretaceous. Ehrenberg (1838) was the first to record such forms, applying the name Xanthidium hirsutum, which is that of a living desmid. Lejeune- Carpentier (1941) redescribed Ehrenberg’s fossil material, under the name Hystrichosphaeridium hirsutum, but Deflandre (1946) correctly pointed out that the trivial name hirsutum could not be legitimately transferred to another genus and applhed to the fossil forms. Deflandre had earlier (1937) proposed a new species H. striolatum for similar Upper Cretaceous forms having numer- ous slender, simple or branching spines, and a striate shell sur- face; there can be little doubt that Merrill’s species cretaceum is a senior synonym of striolatum, although confirmation of the striate nature of the shell surface did not prove possible. Under strict application of the rules of priority (Article 63 of the ‘‘ International Code of Botanical Nomenclature’’) the name cretaceum is senior and should be retained; the name striolatum is Junior and should be rejected. However, the species striolatum was soundly based, fully described and adequately figured; the species cretaceum was, in contrast, profoundly misinterpreted and inadequately figured and described. Moreover, complete cer- tainty in the recognition of the holotype is not possible; and full lop) BREVIORA No. 242 Text-figure A A reproduction of Merrill’s plate (1895). The specimens are here labelled according to the interpretation given in the present paper. Fuller discussions of the assignations are given in the text. 1-4, Shell fragments sectioned at varying angles. A fragment of organic material. Wood fragments. Shell lining of a unilocular foraminifer (?). Character not clear. Fragment of organic material. Hystrichosphaera ramosa (Ehrenberg, 1838). Exochosphaeridium cretaceum (Merrill, 1895) = E. striolatum (Deflandre, 1937). Hystrichodinium pulchrum Deflandre (?). Oligosphaeridium complex (White, 1842). Hystrichosphaera ramosa (Bhrenberg, 1838). Spore. Character not clear. Baltisphaecridium texanum (Merrill, 1895). Chlamydophorella sp. (2). Hystrichosphaera ramosa (Ehrenberg, 1838). Hystrichosphaeridium tubiferum (H. H. White, 1842). Detached spines of Hystrichosphaera ramosa (Ehrenberg, 1838). Character not clear. ““Brown bodies’’ of Bryozoa (?). Carbonate crystals. Dinoflagellate cyst, probably Microdinium sp. (?). Character not clear. Probable sponge spicules. Fragment of Hystrichosphaera ramosa (Bhrenberg, 1838 ) (2). Fragment of organic material. CRETACEOUS CHERT FOSSILS 1966 8 BREVIORA No. 242 and accurate redescription of the holotype, even if certainly re- located, would not be possible because of the difficulty in studying chert-enclosed specimens at high magnification. The name strio- latwm is in widespread use; the name cretaceum has not been reused since its original publication, other than in taxonomic lists. Were a generic name concerned, an adequate case for conserva- tion could be made under Article 14; unfortunately, this Article does not apply to species names. Nonetheless, it is here suggested that striolatum should continue to be used, in preference to the hame cretaceum, im view of its more adequate description. The correct generic assignation is currently to Hwochosphaeridium Davey, Downie, Sarjeant and Williams, 1966. 13. **Tetractinellid: Geodia ? spini-curvata n. sp.’’ This species was transferred to the genus Baltisphaeridium in an earler paper (Sarjeant, 1964, p. 175). No specimen comparable to the holotype or attributable, on grounds of general morphology, to this species, was encountered. The figure shows a form with relatively few, simple spines and havine a precingular archae- opyle; it is probably referable to the genus Hystrichodinium De- flandre 1935 emend. Sarjeant 1965, perhaps to H. pulchrum De- flandre 1935, a species very frequent in Enelsh cherts. How- ever, since the holotype is lost and since neither description nor figure permit certain statement of its characteristics, the species name sprni-curvatum cannot continue in use. 14. ‘‘Tetractinellid: Geodia ? irregularis n. sp.’’ This species was transferred to the genus Hystrichosphaeridium in an earlier paper (Sarjeant, 1964, p. 175). No specimen comparable to the holotype or attributable, on grounds of general morphology, to this species, was encountered. However, despite Merrill’s state- ment (p. 15) that ‘‘Nothing similar to this has been found figured,’’ there can be little doubt that this species is a junior synonym of Oligosphacridium complex (H. H. White, 1842), Davey and Williams 1966a, a species frequent in the Cenomanian and known to range down well into the Lower Cretaceous. 15. ‘‘Tetractinellid: Geodia ? tripunctata n. sp.’’? Rejection of this species on the grounds that it is a junior synonym of Hystrichosphaera ramosa (Ehrenberg, 1838) O. Wetzel 1933, emend. Davey and Williams 1966a, has already been proposed (Sarjeant, 1964, p. 175). A number of representatives of the species H. ramosa were encountered, including perhaps Merrill’s holotype (specimen IIIC: Pl. 1, fig. 4); the earlier judgement is fully supported. 1966 CRETACEOUS CHERT FOSSILS i) 16. ‘*Tetractinellid: Hymeraphia ? sp.’’ The figured specimen was not found; however, Merrill’s figure shows a trilete mark, in- dicating that his specimen was in all probability a spore. Spores and pollen occur infrequently in the chert sections; two some- what similar forms are specimens IIIF (which resembles the figured specimens, but lacks a comparable clear trilete mark) and WelcAn GRIMS fies 5): 17. **Tetractinellid: Chondrilla ? sp.’’ The figure and deserip- tion of this form are so indefinite as to render impossible deter- mination of its true character. One specimen encountered (IIC: Pl. 1, figs. 2, 3), a dinoflagellate cyst (questionably referable to the genus Gonyaulacysta and seen in oblique antapical view) may perhaps be Merrill’s specimen, but this is incapable of confirm- ation. 18. ‘‘Tetractinellid: Geodia? texrana n. sp.’ This species was transferred to the genus Baltisphaeridium in an earlier paper (Sarjeant, 1964, p. 175). Neither the holotype nor any com- parable specimen was located; the morphology, as illustrated, is insufficiently characteristic to permit fuller comment on its prob- able affinities. Since the holotype is lost, a fuller analysis of the characteristics of the species is not possible; the name teranwm cannot, therefore, continue in use. 19. ‘‘Tetractinellid: Hymeraphia ? sp.’’ The figure is not especially informative, but shape, spine cover, and a faint indi- cation of a cingulum combine to suggest a dinoflagellate cyst. The combination of characters strongly suggests a species of the genus Chlamydophorella; unfortunately, this probable assignation cannot be confirmed, since the figured specimen was not relocated. 20. ‘‘Tetractinellid: Geodia? spinipansata n. sp.’’? This species was transierred to the genus Hystrichosphacridium in an earlier paper (Sarjeant, 1964, p. 275). It is probable that a speci- men encountered in slide IIT (specimen III B: PI. 1, fig. 7) is the holotype. There can be no question that this is yet another speci- men of Hystrichosphaera ramosa (Ehrenberg, 1838) O. Wetzel 1933, emend. Davey and Williams 1966a; this species shows some range of variation in morphology (ef. Davey and Willams, 1966a), and also varies in appearance according to orientation. The spines are not tubular, so that assignation to Hystricho- sphaeridium is incorrect. The species spinipansatum must be re- jected, as a junior synonym of ramosa. 10 BREVIORA No. 242 21. ‘‘Tetractinellid: Geodia? hilli n. sp.’’ This species was transferred to the genus Hystrichosphaeridium (Sarjeant, 1964). No specimen comparable to the holotype or attributable, on grounds of general morphology, to this species, was encountered. However, despite Merrill’s repetition of his statement, earlier apphed to No. 14, ‘* Nothing similar has been found figured,’’ the specimen he figures is without question attributable to the species Hystrichosphacridium tubiferum (BKhrenberg, 1838) O. Wetzel 1933, emend. Davey and Williams 1966a, a species common throughout the Cenomanian and likely to be present in the Al- bian. The species hilli must, therefore, be rejected as a Junior synonym of tubiferum. 22. **Tetractinellid: Geodia tripunctata ? n. sp. Fragments resulting from solution.’’ Unquestionably these are detached spines of Hystrichosphaera ramosa. A number of such detached spines were seen, none corresponding precisely to the grouping shown in the figure, which might in any case be idealised. 23. ‘‘Lithistid ? Flesh spicule.’’ This is described as ‘‘ yellow- ish’’ and must be presumed to be a plant or wood fragment. The figure, by it generalised nature, precludes certain recognition ; similar fragments were by no means uncommon. 24, 25, 26. ‘‘Tetractinellids; Hymeraphia ? sp.’’ Over-simpli- fied figures and descriptions again render certain recognition difficult. Pollen grains of comparable morphology were noted, and also (in slide X) a species of dinoflagellate cyst questionably attributable to the genus Prolixosphaeridium Davey, Downie, Sarjeant and Williams 1966, with a similarly elongate shape and cover of simple spines (PI. 1, fig. 8). like structures were encountered in several cherts (e.g. VIT B; Pl. 1, fig. 11). Similar structures occur widely in palynological material from the Upper Mesozoic; Merrill’s description (p. 17) is quite accurate. Their interpretation is doubtful; Otto Wetzel (1961, pl. 1, figs. 8-10) has figured similar forms as ‘‘brown bodies’ of Bryozoa; it is also possible that they represent remains of colonial algae, possibly Chlorophyta. 27. ‘*Tetractinellid: Geodia ? sp.’’? Groups of circular, dise- 28. ‘*Tetractinellid: dermal spicules ?’> These are abundant in most of the cherts examined and are undoubtedly rhomb-shaped carbonate (? dolomite) crystals. Carbonate crystals are fre- quently encountered in cherts and are sometimes considered to 1966 CRETACEOUS CHERT FOSSILS 1] indicate that the chert was formed by replacement of a pre-exist- ing limestone. Merrill’s misinterpretation of them is surprising GELS fie), 29. ‘*Tetractinellid: Geodia sp.’’? The figure may correspond to a specimen in slide III, which resembles it in general features (specimen IITH). Although full study is difficult because of its situation deep in the chert, this is certainly a dinoflagellate eyst, and most probably a species of the genus Microdinium, species of which are known to occur in the Cenomanian. 30. ‘*Tetractinellid: globo-stellate of dermal layer.’’ This structure was not relocated and is of dubious reference, possibly pollen, 31. ‘‘Geodia ?”’ The figure and description are inadequate to permit recognition of this form. 32. ‘‘Geodia ? Pyramidal or zone spicule.’’ Not relocated ; most probably correctly interpreted as sponge spicules. 33, 34. ‘‘Hexactinellid. Stauractinellia ? sp.’’> Not relocated ; most probably correctly interpreted as sponge spicules. 35. ‘‘Frame work of a Hexactinellid?’’ This appears to be a fragment of a Hystrichosphaera furcata, although the figure is extremely generalised and precise interpretation is difficult. A sinilar fragment was encountered in slide III (specimen IIIE). 36. ‘‘Monactinellid: Reniera ? sp.’’ The sausage-shaped ob- ject in Merrill’s figure is most probably to be interpreted as a fragment of organic tissue (? plant); many similar fragments were seen. CONCLUSIONS A re-examination of Merrill’s text and figures, in conjunction with study of the three surviving chert sections (one broken) and of new sections made from surviving fragments of cherts, makes it clear that, of the 36 microscopic structures he figures and de- scribes, only three (none confirmed) can be regarded as represent- ing sponge spicules. Nine of his figures certainly, and four others possibly, represent dinoflagellate eysts (mainly — hystricho- spheres) ; these include his eight proposed new species, of which five are junior synonyms of previously described species. The three remaining species retain technical validity, but the holo- types of two are lost, the descriptions and figures being insuffi- cient to enable their accurate characterisation, and the holotype of the third species does not permit full study. 12 BREVIORA No. 242 The remaining 20 microscopic structures figured and described by Merrill are of very diverse character. Four are considered to represent shell fragments sectioned in varying directions; one is thought to be a foraminiferal shell lining; one is a spore; three possibly pollen; one represents an association of dise-like strue- tures, possibly ** brown bodies’’ of Bryozoa; one comprises carbon- ate crystals; two represent wood fragments; four represent frag- ments of organic (? plant) tissue of varying shape; and three are of indeterminate character. The currently valid species of dinoflagellate cysts recognised in Merrill’s figures and material are Hystrichosphaera furcata, Hystrichosphaeridium tubiferum, Exochosphaecridium striolatum, Oligosphaeridium complex, Hystrichodinium pulchrum (doubt- ful), and undetermined species of the genera Gonyaulacysta, Pro- lirosphaeridium, Hystrichodinium, Chlamydophorella and Micro- dinium. Although the number of individuals encountered is small, and a full picture of the dinoflagellate cyst assemblage cannot be said to have been obtained, the species represented are all ones which would be likely to occur in Middle to Upper Albian prep- arations. REFERENCES Davey, R. J., C. Downigz, W. A. S. SARJHANT and G. L. WILLIAMS 1966. Fossil dinoflagellate cysts attributable to Baltispraeridium. In Studies of Mesozoic and Cainozoie dinoflagellate cysts. Bull. Brit. Mus. (Nat. Hist.) Geol., Suppl. 3 (in press). Davey, R. J. and G. L. WILLIAMS 1966a. The genera Hystrichosphaera and Achomosphaera. In Studies of Mesozoic and Cainozoic dinoflagellate cysts. Bull. Brit. Mus. (Nat. Hist.) Geol., Suppl. 3 (in press). 1966b. The genus Hystrichosphaeridium and its allies. Jn Studies of Mesozoic and Cainozoic dinoflagellate cysts. Bull. Brit. Mus. (Nat. Hist.) Geol., Suppl. 3 (in press). DEFLANDRE, G. 1935. Considérations biologiques sur les micro-organismes d’origine planectonique conservés dans les silex de la craie. Bull. Biol. Fr. Belg., vol. 69, pp. 213-44, pls. 5-9. 1937. Microfossiles des silex crétacés II. Ann. Paléont., vol. 26, pp. 51-103, pls. 8-18. 1946. Remarques sur la systématique des hystrichosphaeridés. C. R. Somm. Soc. Géol. Fr., No. 7, pp. 100-111. 1966 CRETACEOUS CHERT FOSSILS 13 EHRENBERG, C. G. 1838. Uber das Massenverhaltnis der jetzt labenden Kieselinfusorien und iiber ein neues Infusorien-Conglomerat als Polierschiefer von Jastraba in Ungarn. Abh. K. Akad. Wiss. Berlin (1836), vol. 1, pp. 109-35, pl. 1. GLAESSNER, M. F. 1945. Principles of micropalaeontology. Melbourne: University Press, 296 pp., 14 pls. LEJEUNE-CARPENTIER, M. 1941. L’étude microscopique des silex (9 iéme note). Sur Hystricho- sphaeridium hirsutum (Ehr.) et quelques formes voisines. Ann. Soc. Géol. Belg., vol. 63, no. 3, pp. 71-92. MANTELL, G. A. 1845. Notes of a microscopical examination of the Chalk and Flint of southeast England. Ann. Mag. Nat. Hist., vol. 16, no. 103, pp. 73-88. MERRILL, J. A. 1895. Fossil sponges of the flint nodules in the Lower Cretaceous of Texas. Bull. Mus. Comp. Zool., Harvard. (Geology Ser. IIT), vol. 28, no. 1, pp. 1-26, 1 pl. Murray, G. E. 1961. Geology of the Atlantic and Gulf Coastal Province of North America. New York: Harper, 692 pp. SARJEANT, W. A. S. 1964. Taxonomic notes on hystrichospheres and acritarchs. J. Paleont., vol. 38, pp. 173-7. 1966. Dinoflagellate cysts with a Gonyaulax type tabulation. In Studies of Mesozoic and Cainozoie dinoflagellate cysts. Bull. Brit. Mus. (Nat. Hist.) Geol., Suppl. 3. (in press). WETZEL, O. 1933. Die in organischer Substanz erhaltenen Mikrofossilien des Bal- tischen Kreide-Feuersteins. Palaeontographica, vol. 77, pp. 141- 88; vol. 78, pp. 1-110, pls. 1-7. 1961. New miecrofossils from Baltic Cretaceous flintstones. Micro- palaeontology, vol. 7, no. 3, pp. 337-50, pls. 1-3. Wuits, H. H. 1842. On fossil Xanthidia. Microsc. J., vol. 11, pp. 35-40, pl. 4. (Received 25 October 1965) 14 BREVIORA No. 242 Plate 1 Fig. 1. Presumed shell lining of a unilocular Foraminifer (Fig. 7 **“Monactinellid’’ of Merrill). & 500. Figs. 2, 3. Gonyaulacysta sp. indet., seen in oblique apical view (Fig. 2) and, by transparency, in oblique antapical view (Fig. 3) —the antapex is at left. Figure 2 closely resembles ‘‘Tetractinellid; Chondrilla 2? sp.’’ of Merrill. & 500. Fig. 4. Hystrichosphaera ramosa (Ehrenberg). The presumed holotype of Merrill’s invalid species Geodia ? tripunctata. > 500. Fig. 5. Hystrichosphaera ramosa (Ehrenberg). Possibly the holotype of Merrill’s invalid species Geodia ? spini-pansata. >< 500. Fig. 6. Hxochosphaecridium cretaceum (Merrill). The specimen may be the holotype. & 500. Fig. 7. Hystrichosphaera ramosa (Ehrenberg). A specimen in terminal view, illustrating the high degree of variation in appearance given by orientation. & 500. Fig. 8. Prolixosphaeridium sp. indet. (Possibly corresponding to Merrill’s Fig. 26b ‘‘Tetractinellid; Hymeraphia ? sp.’’). ©300. Fig. 9. Spore, gen. et sp. indet. (Similar to Merrill’s Fig. 16 ‘‘ Tetracti- nellid; Hymeraphia ? sp.’’?; but lacking a clear trilete mark.) & e300. Fig. 10. Hystrichosphaera ramosa (Ehrenberg). Possibly the holotype of Merrill’s invalid species Geodia ? austini. *& 500, Fig. 11. Group of disc-like organisms, possibly ‘‘brown bodies’’ of Bryozoa (equivalent to Merrill’s Fig. 27 ‘‘Tetractinellid; Geodia ? sp.’’). x 500. Fig. 12. A carbonate (? dolomite) crystal. Similar crystals are described by Merrill as ‘‘ Tetractinellid: dermal spicules?’’ > 500. 1966 CRETACEOUS CHERT FOSSILS 15 BREVIORA Museum of Comparative Zoology CAMBRIDGE, Mass. AprRIL 29, 1966 NuMBER 243 QUATERNARY FISH FOSSILS FROM WEST OF LAKE RUDOLF, KENYA By KeityH Srewart THOMSON Department of Zoology, University College London.1 INTRODUCTION The fossils to be described in this paper were collected during the summer of 1963 by a party from the Museum of Comparative Zoology, consisting of Professor Bryan Patterson, Messrs. A. D. Lewis, C. T. Williams, and the author. It is a pleasure to ac- knowledge the assistance given to us by Dr. L. 8. B. Leakey, and by Mr. John Walsh of the Mines and Geological Department, Kenya, who called our attention to the deposits discussed in this paper. The late Pleistocene and Quaternary deposits west of Lake Rudolf, North Kenya (Turkana District), were visited by us during a survey of the Miocene formations in the region. Ac- counts of earlier visits to the region are given by Arambourg (e.g. 1943), and Fuchs (1939). A general view of the area con- cerned is shown in the accompanying map. A range of hills (in- cluding Losidok) west of the lakeshore is Tertiary in age, consist- ing of grits and volcanics. Along the base of this range there are exposed large and extensive beds of Pleistocene age — largely unfossiliferous lake beds, including conglomerates of small clay pebbles with, in places, small rocks derived from the voleanies. Away from the bases of the hills —and here we explored only eastwards — are flat plains covered with gravel and windblown sands; close to the lake large sand dunes occur. These plains extend from the hills down to the lakeshore. Nearer the lake, as Fuchs has noted (1939), there is evidence of at least two former beaches, some 220 and 90 feet above the current level of the lake. Along these beaches fossils may be found on the surface, although their derivation is difficult to determine, and ‘‘reefs’’ of the _1 Present address: Dept. of Biology and Peabody Museum of Natural History, Yale University, New Haven, Conn. DZ BREVIORA No. 2438 freshwater oyster, Aetheria elliptica, also occur. We were able to follow the beaches for several miles and found fossils even down to the lakeshore. In several places drainage from the plains has cut into the deposits and we found a number of large gullies or washes along the sides of which the bedding of the deposits — apparently with a shg¢ht eastward dip — was visible, and fossils could be collected in situ. The principal place from which fossils were collected was an extensive exposure at Kangatotha (see map), where the Turkwell River has made a large cut through some hundred feet of the lake deposits. Patterson and Williams measured the exposed section as follows: Upper unit, bed 10 Surface white sands fSo9 4 feet black clay OG 3 16 feet white coarse sands with gravel lenses i SAS feet grey clay SO 6 10 feet micaceous sand CG 15s 4 feet black clay Oe 4: 1% to 5 feet coarse sand and gravel SG 5 feet black clay 34 foot white sand 2% feet black clay are 9 feet interbedded black elays and grey sands CC eal 12 feet black clays Lower unit, bed 5 7 feet fine grey sand oO 4, 346 feet fine grey sand highly indurated with iron Cr 3) 5 feet fine white sands, black micaceous flakes SG 4 feet medium yellow sands, black micaceous flakes co 1) «(12% feet coarse grey sands with large clay pebbles in upper unit (base con- cealed ) ‘‘Upper unit’’ and ‘‘lower unit”’ are informal, nou-committal terms; formal stratigraphic naming should follow, not precede, a general survey of the Lake Rudolf Quaternary. The two units are readily distinguishable, the upper being predominantly dark in colour, the lower light. Shehtly indurated fillings of plant roots are abundant in the lower unit. Partly due to the swollen state of the river and partly due to lack of time, a similar cliff on the south bank of the watercourse, here about half a mile wide, was not visited. The tip of northern cliffs corresponds to the 220 foot level of Fuchs. 1966 QUATERNARY FISHES FROM KENYA 3 Molluse, fish, reptilian and mammalian remains are common in the upper unit of the exposures, molluse and fish in the lower. A human skull, a partial mandible, and various skeletal fragments were found isolated in bed 8 of the upper unit, and poorly pre- served parts of a human skeleton were found in the overlying black clay of bed 9. Artifacts and pottery fragments are common on the surface and both were encountered in situ in bed 7 of the upper unit. o Muruangering Nuees tt - . Muruadou 3000 v Karsa Spring Ferguson : Gul Gentral L ksland ie a6 ° Glendays Ss E Camp Chavakono "2500 ° Goritei Kholotum o Kangatotha e Miles 20 15 I i fo) 5 Sketch map of Lake Rudolf and the Lodwar region (Turkana District, Kenya) showing position of the Kangatotha locality. 4 BREVIORA No. 243 FOSSIL FISHES All the fish remains collected consisted of isolated bones, never associated, and very often fragmentary. Few showed great signs of weathering or of having been ‘‘rolled.’’ The bones were highly mineralized, often showing considerable iron content, despite their relatively recent age (see below). From lower to upper strata, the fish were distributed as fol- lows: Lower unit, base of bed 5, seven feet below junction with upper unit Order Ostariophysi Suborder Siluroidea Family Mochokidae Synodontis sp. Seven pectoral spine fragments Family Clariidae Clarias sp. One articular (posterior portion), one quadrate (articular portion ) Suborder Cyprinoidea Family Cyprinidae cf. Barbus Two dorsal pterygiophore fragments Order Percomorphi Suborder Percoidea Family Centropomidae | Lates ef. Lates niloticus Cuvier and Valenciennes | One incomplete parasphenoid (length 150 mm), — three vertebrae (diam. 25, 33, 46 mm), four fin spines (88, 52, 55, 62 mm) Family Cichlidae Tilapia sp. Two fragments of pterygiophores, six fin spines (2 nearly complete, 55 and 47 mm), two branched fin rays, and one vertebra 1966 QUATERNARY FISHES FROM KENYA 5 Lower unit, bed 5 (general) Order Ostariophysi Suborder Cyprinoidea Family Cyprinidae ef. Barbus Tentatively referred to this genus are sections of two fin spines. Order Percomorphi Suborder Perecoidea Family Centropomidae Lates ef. Lates niloticus Cuvier and Valenciennes Four dorsal fin spines (28, 42, 45, 51 mm) Family Cichlidae Tilapia sp. One pterygiophore — first interhaemal (70 mm), two dorsal fin spines (incomplete), four ver- tebrae: two abdominal (diam. 10, 11 mm), one caudal (diam. 15 mm), one terminal (diam. 8 mm, length 26 mm) Upper unit, bed 5 Order Ostariophysi Suborder Siluroidea Family Bagridae Bagrus cf. Bagrus bayad (Geoffroy ) Posterodorsal portion of a single occiput (estimated height 28-30 mm) Family Mochokidae Synodontis cf. Synodontis schall Cuvier Humeral process of left cleithrum (49 mm long), one posttemporal bone, and fragments of four pectoral spines Family Claridae Clarias ef. Clarias lazera Cuvier and Valenciennes Posterior portion of dermethmoid bone (breadth 25 mm), median portion of supraoccipital (max. breadth 35 mm) and eleven vertebrae (diam. 8, 12S, ta Ib bs 6. Te oli am) BREVIORA No. 2438 Suborder Cyprinoidea Family Cyprinidae Barbus cf. Barbus bynni Cuvier and Valenciennes Three dorsal pterygiophores (incomplete), one pharyngeal (length approx. 37 mm), one ver- tebra (diam. 6 mm) Order Percomorphi Suborder Percoidea Family Centropomidae Lates ef. Lates niloticus Cuvier and Valenciennes The most abundant fish in this unit; twenty-three vertebrae (diam. 12, 13, 16, 19, 20, 20, 24, 25, 26, 26, 26, 27, 27, 30, 34, 35, 35, 37, 42, 42, 50, 51, 53 mm), the articular regions of four quad- rates, one incomplete preoperculum, four in- complete fin spines, three fragments of hyo- mandibulars, one dermethmoid, one partial parasphenoid, two left premaxillae (proximal ends), one maxilla (distal end), and two iso- lated neural arches (probably from anterior vertebrae ) Family Cichlidae Tilapia sp. One dorsal pterygiophore, two spines, nine verte- brae (diam. 8, 7, 6, 6, 6, 6, 5, 5, 5 mm) Upper unit, bed 8 Order Ostariophysi Suborder Siluroidea Family Mochokidae Synodontis ef. Synodontis schall Cuvier. The humeral process of one cleithrum (length 55 mm), one dorsal fin spine, three pectoral fin spine fragments ef. Synodontis Fragments of three spines 1966 QUATERNARY FISHES FROM KENYA il. Family Clariidae Clarias ef. Clarias lazera Cuvier and Valenciennes Three dermethmoids (2 complete, lengths 34, 45 mm; one incomplete, anterior width 50 mm), one posttemporal, two sections of supraoccipital (one anterior, one posterior), one pectoral fin spine fragment, two vertebrae (diam. 13, 19 mm) Suborder Siluroidea indet. One vertebra (diam. 23 mm), two dorsal fin spines (80, 48 mm) and several spine fragments Suborder Cyprinoidea Family Cyprinidae Barbus ef. Barbus bynni Cuvier and Valenciennes Four pharyngeal bones (approx. length 30, 30, 40, 45 mm), two (?dorsal) pterygiophore frag- ments ef. Barbus Two vertebrae (diam. 14, 15 mm) Order Percomorphi Suborder Percoidea Family Centropomidae Lates ef. Lates niloticus Cuvier and Valenciennes Six articular bones (posterior sections), five incom- plete quadrates (articular sections), four in- complete preoperculars, the proximal ends of five premaxillae, the anterior part of one right palatine, one dermethmoid, five dorsal fin spines, twenty-two vertebrae (1 second verte- bra — diam. 25 mm; 3 third vertebrae — diam. 20, 23, 27 mm; 18 abdominal vertebrae — diam. 17, 20; 20, 20, 22, 24, 24, 28, 28, 28, 29, 29, 32, 33, 34, 40, 48, 49 mm), plus a number of shat- tered and unidentifiable fragments of large cranial elements 8 BREVIORA No. 243 Upper unit, bed uncertain These fishes were collected at one of the smaller exposures near the lake shore, about two miles north of the mouth of the Turkwell. Order Ostariophysi Suborder Siluroidea Family Bagridae Bagrus cf. Bagrus bayad (Geoffroy ) One occiput, incomplete (max. depth 69 mm) ?Family Clariidae Dermethmoid portion of skull, incomplete Order Percomorphi Suborder Percoidea Family Centropomidae Lates cf. Lates niloticus Cuvier and Valenciennes Posterior portion of a skull (depth of occipital re- gion 80 mm), one premaxilla (length 127 mm), three vertebrae (diam. 29, 30, 88 mm). Note the extreme size of these specimens. Collected on lake shore, horizon not known, but probably upper unit Order Percomorphi Family Centropomidae Lates ef. Lates niloticus Cuvier and Valenciennes Eight vertebrae (diam. 14, 18, 20, 24, 25, 26, 27, 30 mm) and three dorsal fin spines (40, 44, 46 mm ) Family Cichlidae Tilapia sp. One incomplete fin spine DISCUSSION Of the living species to which the fossils are most closely com- parable — Bagrus bayad, Synodontis schall, Clarias lazera, Lates niloticus, Barbus bynni, and Tilapia nilotica — only Bagrus bayad, according to Worthington and Riccardo (1936), does not 1966 QUATERNARY FISHES FROM KENYA 9 frequent sheltered or shallow waters. Synodontis schall is dis- tributed universally in Lake Rudolph. With the possible excep- tion of the bigger Tilapia and Lates, the others are confined to the sheltered and shallower areas of the lake, such as, for ex- ample, Ferguson’s Gulf. A tentative conclusion from examina- tion of the fossils would therefore be that bed 5 of the upper unit was laid down in more open and less shallow waters than bed 8 of this unit or the upper bed of the lower unit. An interesting feature of the fish fauna is the very great size of the largest vertebra of Lates from the fifth locality (lake- shore, horizon uncertain) ; this is far larger than is known from present-day Lates. Daget (1959) and White (1926) have de- seribed similar giant Quaternary Lates, and it is interesting to discover that such forms were still extant in such relatively very recent times, as shown in these Rudolf materials. The nature of the fish fauna does not afford any direct evi- dence as to the age of the Kangatotha beds, since the Nilotic na- ture of the Rudolf fish fauna has been stable at least since the Lower Pleistocene disconnection of the lake basin from the Nile drainage (see Fuchs, 1939, for general chronology of the area). The mammalian fauna, so far as studied, is of Recent aspect. The Kangatotha beds belong to the very latest stage in the his- tory of the Rudolf basin. Their level and position would suggest that they were laid down while the lake was at its largest extent after the Gamblian Pluvial, presumably during the Makahan or Nakuran ‘‘wet phases’’ (see Cooke, 1957, 1963). A radio-carbon date of 4,800 + 100 years was obtained from a sample of Etheria elliptica from the ‘‘beaches’’ on the lake shore, noted above. This attests to very recent age of the deposits and implies an ex- tremely rapid fossilization. It is of great interest to note that the siant forms of Lates, described by Daget, White, and in this paper, survived to a very recent date. Finally, with respect to the time scale, a note may be added here about one of the arti- facts collected on the surface of bed 7 of the upper unit. This is a uniserially barbed bone harpoon (now in the Peabody Mu- seum, Harvard University). De Heinzelin (1957) lists the dis- tribution of similar harpoons and gives a scheme of the evolution of harpoon design with an estimate of the equivalent culture. (To de Heinzelin’s list should be added a specimen from the late Upper Kenya Capsian of Gamble’s Cave described by Oakley, 1961). At first view, the Kangatotha harpoon seems to resemble the penultimate stages of design as shown by specimens from the “S.F.M.’’ and ‘‘G.Y.’’ levels at Ishango (de Heinzelin, 1957), 10 BREVIORA No. 243 whereas the ‘‘N. tuff.’’ specimens from Ishango, the specimen found by Arambourg (1943) at Nanoropus near the end of Lake Rudolf, and the Gamble’s Cave specimen are of a more advanced style. ACKNOWLEDGMENTS The expedition was financed by NSF Grant No. GP 1188. In addition to the acknowledgment made in the introduction, it is a pleasure to thank Dr. P. H. Greenwood for his enthusiastic as- sistance in the study of the material and Professor Bryan Patter- son for valuable comments on the manuscript and for providing the sketch map (Fig. 1). LITERATURE CITED ARAMBOURG, C. 1943. Mission Scientifique de 1’?Omo 1932-1933. Tome 1. Géologie- Anthropologie. Fase. II. Mus. Nat. Hist. Natur. (Paris). Pp. 1-74. Cook, H. B. S. 1957. Observations relating to Quaternary environments in east and southern Africa. Trans. Geol. Soe. 8S. Africa, 60 (Annex.), 73 pp. 1963. Pleistocene mammal faunas of Africa, with particular reference to southern Africa. Pp. 65-116 in F. C. Howell and F. Bourliére, eds., African Ecology and Human Evolution. Aldine Co., Chi- cago. DAGET, J. 1959. Restes de poissons du Quaternaire saharien. Bull. Inst. Frane. Afr. Noire, 23: 182-191. DE HEINZELIN, J. 1957. Les feuilles d’Ishango. Exploration du Pare Albert, 1957: 46- 61. IMU CHSSaViy LE. 1939. The geological history of the Lake Rudolf basin, Kenya Colony. Phil. Trans. Roy. Soc. London, (B) 229: 219-274. OAKLEY, K. P. 1961. Bone harpoon from Gamble’s Cave, Kenya. Antiq. J., 41: 86-87. Waits, EH. I. 1926. The geology and paleontology of the Kaiso bone-beds. Part II. Palaeontology. Fossil Pisces. Oecas. Paper Geol. Surv. Uganda, 2: 45-51. WORTHINGTON, HE. B. AND C. K. RICARDO 1936. Scientific results of the Cambridge expedition to the East Afri- ean lakes, 1930-1, no. 15. The fish of Lake Rudolph and Lake Baringo. J. Linn. Soe. London, Zool., 39: 353-389. (Received July 26, 1965.) BREVIORA Museum of Comparative Zoology CAMBRIDGE, Mass. AprIL 29, 1966 NUMBER 244 A NEW SPECIES OF ASHMUNELLA FROM WEST TEXAS (MOLLUSCA: PULMONATA) By W. J: CLENCH AND W. B. MER! The first specimens of this new species were collected by our colleague, Dr. J. C. Bequaert, in the Davis Mountains, West Texas, in May and September 1961. All of the specimens were dead shells. During June 1965 live specimens of this species were collected by Dr. J. C. Bequaert and W. B. Miller from which the anatomical drawings were made by the Junior author. ASHMUNELLA BEQUAERTI new species Plate 1; text figures 1, 2 Holotype: MCZ 260274, taken from a rockslide about 14 mile up a tributary canyon (locally known as Goat Cave Canyon) to Little Aguja Canyon near Buffalo Trail Boy Scout Camp at the base of the northeastern slope of Black Mountain in the Davis Mountains, Jeff Davis County, Texas. Elevation 4,900 ft. (J. C. Bequaert and W. B. Miller, 5 June 1965). Igneous rockslides line both sides of the canyon, below verti- cal cliffs; Ashmunella bequaerti was found on both sides of the canyon, at elevations from 4800 ft. to 4900 ft., along with Humboldtiana palmert Clench and Rehder, in association with Quercus terana, YQ. hypoleucoides, Acer grandidentata and Rhus trilobata. 1 University of Arizona, Dept. of Zoology. 2 BREVIORA No. 244 Paratypes: MCZ 260275 from the same loeality. Greater Lesser Height Diameter Diameter mm mm mm aay) 1 ISIE Holotype 3.8 13.0 Male Paratype 3.7 12.4 ANG) “e Bie) 12.2 ele = 3.0 UNEZh 105 i 3.4 11.3 10.1 a 3.2 10.7 9:5 28 3.0 10.5 9.3 i Description: Shell lenticular, shghtly convex above, moder- ately convex below, acutely carinate, widely umbilicate, thin, pale brown. The surface, above and below and into the umbilicus, is sculptured with fine growth wrinkle-striae from which project numerous cuticular scales giving a pilose appearance to fresh clean shells; the scales also extend well into the umbilicus. In live and recently dead animals, the seales usually hold dirt and debris, giving a dusty, dirty appearance to the surface. In older specimens where the scales have worn off, there remain raised hyphen-shaped papillae, parallel to the growth striae, giving a granular appearance to the surface. There are six whorls, the first half whorl of the embryonice shell glossy with only a few incipient radial striae, the remainder with a silken appearance, the fine growth wrinkle-striae and the papillae and scales. The first two and a half whorls are convex; subsequent whorls flat- tened. Last whorl descends shghtly to the aperture and _ is deeply guttered close behind the outer and basal margins of the lip. The under surface of the body whorl is marked additionally by very fine, microscopic, spiral striae. The aperture is small and very oblique; peristome white to light brown, reflected except near the upper insertion of the lip, the terminations connected by a slightly raised callus. Parietal wall with two teeth, the larger one situated basally, sinuous, diverging posteriorly toward the smaller upper one, thicker anteriorly ; the smaller one close to the upper insertion of the lip, raised posteriorly. Outer mar- gin thickened, with a flat-topped rectangular tooth set trans- versely across the aperture, spanning and overlapping the gap between the parietal teeth. On the basal margin, two longitudi- nally compressed teeth are connected by a raised ridge along the peristome. Interdental intervals are about equal. The umbilicus, 1966 NEW SPECIES OF ASHMUNELLA 3 measured from lower suture of body whorl just behind peristome to opposite side of body whorl in the pit, is contained about four times in the greater diameter of the shell. The mantle over the lung is clear except for rare, very small groups of light grey pigment spots. Jaw with ten ribs. Genitalia. Penis with lower sac short and wide, upper sac longer and narrower. In the upper sae there are three longitudi- nal ridges, in addition to a small nodule which is attached to the wall of the penis at its Junction with the epiphallus. In whole mounts, this nodule gives the appearance of a short verge when seen by transparency ; dissection of the penis, however, reveals it as a growth on the side of the penial wall. It is possible that this nodule, together with the tip of the muscular walls of the epiphal- lus, acts as a papilla when the penis is completely everted. The lower sac has two thickened processes which come together near the lower end and form a constriction within the penial cavity. There is a short penial retractor on the epiphallus which is in- serted in the floor of the lung cavity ; loose strands of connective tissue connect the penis with base of epiphallus. Talon of the multilobar type. Length of penis 5.0 mm, epiphallus 24.0 mm, flagellum 1.5 mm, penial retractor 1.3 mm, spermatheca and duct 31.0 mm, vagina 4.0 mm, free oviduct 2.5 mm, atrium 1.0 mm. Remarks: Ashmunella bequaerti is the most easterly known species of Ashmunella in the U.S. It is most closely related to the mearnsi group, in shell characteristics as well as geographi- cally. In paratypes, the spire varies from moderately raised conical to completely flat; the diameter varies from 10.5 mm to 13.0 mm. Its nearest relative appears to be A. hebardi Pilsbry and Vanatta, with which it agrees in the lenticular shape of the shell, the sharply carinate periphery, and the general shape of the teeth, including the two parietal teeth. In A. bequaerti, however, the shell is generally more flattened, the edge of the parietal callus is appressed and very weak, not raised into a low, free ridge as in hebardi, and the upper insertion of the peristome is only weakly descending, nearly straight, not strongly descend- ing as in hebardi. Anatomically, the internal nodule at the upper end of the upper penial sac has not been seen or reported in any other Ashmunella. It remains to be seen trom additional dissections, however, whether this is a consistent characteristic. The new species is named after Dr. Joseph C. Bequaert, life- long malacologist and entomologist, who first discovered this 4. BREVIORA No. 244 5mm. Fig. 1. Ashmunella bequaerti Clench and Miller. Dissection of reproduc- tive system with illustrations made from stained whole mount. a, Entire system. b, Enlargement of the distal end of the epiphallus at its junction with the vas deferens where the two are bound together. This may be called the flagellum portion of the epiphallus. There is no free flagellum. 1, Genital orifice; 2, Atrium; 3, Vagina; 4, Oviduct; 5, Prostate; 6, Uterus; 7, Albumen gland; 8, Talon; 9, Hermaphroditic duct; 10, Sperma- thecal duct; 11, Lower sac of penis; 12, Upper sae of penis; 13, Penial retractor; 14, Epiphallus; 15, Vas deferens; 16, Connective tissue; 17, Flagellum bound with vas deferens; 18, Penis papilla. 1966 NEW SPECIES OF ASHMUNELLA 5 snail on 2 May 1961, obtaining quantities of dead shells but no live animals. On 5 June 1965, he and the junior author returned to the locality and were successful in obtaining 4 live adults and 3 live immatures. Dead shells were very numerous. One of the live adults was designated the holotype. Two others were dis- sected to corroborate anatomical findings. The fourth live adult and the three live immatures are being kept alive in a terrarium at the University of Arizona, in the hope of obtaining additional studies on the adult anatomy for comparison. 14 5mm. Fie. 2. Ashmunella bequaerti Clench and Miller. Lower portion of repro- ductive system enlarged. (Numbering as in Figure 1.) REFERENCES Piussry, H. A. 1940. Land Mollusca of North America. Academy of Natural Sciences, Philadelphia, Monograph No. 3, vol. 1, pt. 2, pp. 912-977. 1948. The genera Humboldtiana, Sonorella, Oreohelix and Ashmunella. Proc. Acad. Nat. Sci., Philadelphia, vol. 100, pp. 199-203. (Received 24 January, 1965. ) 6 BREVIORA No. 244 2 Fies. 1-2. Ashmunella bequaerti Clench and Miller, from Goat Cave Canyon, Black Mt., Davis Mts., Texas. Fig. 1. Paratype, MCZ 260275 (4 X). Fig. 2. Holotype, MCZ 260274 (5.4 X). BREVIORA Museum of Comparative Zoology CAMBRIDGE, Mass. APRIL 29, 1966 NUMBER 245 NOTES AND DESCRIPTIONS OF NEW UROCOPTIDAE FROM CUBA AND HISPANIOLA (MOLLUSCA: PULMONATA) By WILLIAM J. CLENCH The following notes and descriptions of new species are based upon material in the Museum of Comparative Zoology which has been collected over the past several years. All species are mem- bers of the family Urocoptidae, one of the most dominant fami- les in the West Indies and in a lesser way in Central America and in the southwestern United States. This family occurs also in northern South America, but species are few and present only a very minor element in the molluscan fauna there. UROCOPTIDAE ARCHEGOCOPTIS Pilsbry Archegocoptis Pilsbry, 1903, Manual of Conchology (2) 15:301 (type species, Cylindrella crenata Weinland and Martens, original designation ). So far as known, this genus is confined to the southwestern or Tiburon Peninsula of Haiti. Specific localities for the few spe- cies are all from the Département du Sud. According to Pilsbry, Archegocoptis is nearest in relationship to Eucalodium Crosse and Fischer, a genus limited in distribu- tion to southern Mexico and Guatemala. The known species are as follows: ARCHEGOCOPTIS BARBOURI Clench Archegocoptis barbouri Clench, 1935, Proc. Boston Soc. Nat. Hist. 41:5, pl. 1, figs. B, F (Tardieu, Mt. LaHotte, Haiti, + 3000 feet). [Holotype, MCZ 108593. ] BREVIORA No. 245 ARCHEGOCOPTIS CRENATA (Weinland and Martens) Cylindrella crenata Weinland and Martens, 1859, Malakozoologische Blit- ter 6:54 (Jérémie, Haiti); Pfeiffer, 1869, Novitates Conchologicae 3:440, jl, A, ane Boi, acs (alemhn)). ARCHEGOCOPTIS DARLINGTONI Clench Archegocoptis darlingtoni Clench, 1935, Proc. Boston Soc. Nat. Hist. 41:6, pl. 1, figs. C, G (Tardieu, Mt. LaHotte, Haiti, = 3000 feet). [Holotype, MCZ 108592. } ARCHEGOCOPTIS DECAPITATA (R6dinge) oD Cerion decapitatum Roding, 1798, Museum Boltenianum, p. 90 (no loeality given [Haiti] ). [See below. ] ARCHEGOCOPTIS EXIMIA (Pfeiffer) Cylindrella eximia Pfeiffer, 1857, Malakozoologische Blatter 4:232 (lo- cality unknown) ; Pfeiffer, 1869, Novitates Conchologicae 3:439, pl. 97, figs. 30-32 (Haiti). ARCHEGOCOPTIS TIBURONICA Cleneh Archegocoptis tiburonica Clench, 1935, Proc. Boston Soc. Nat. Hist. 41:6, pl. 1, fig. I. (Tardieu, Mt. LaHotte, Haiti, between 3-4000 feet). [Holo- type, MCZ 108591. | ARCHEGOCOPTIS HAITIENSIS new species Plate 1, figures 1, 2 Holotype. MCZ 252056, from the top of Morne Rochelois, 25 km WSW of Miragoane, Dépt. du Sud, Haiti, about 3000 feet, collected by W. J. Eyerdam, July 1927. Paratypes. A series of 5 paratypes from the same loeality, MCZ 252057. Height Width mm mm 41 10 Holotype (see Remarks ) 27 10 Paratype 26.9 10 os 26.5 10 ne Description. Shell medium in size, reaching 42+ mm in length (decollated), minutely umbilicate and coarsely sculptured. 1966 NEW SPECIES OF UROCOPTIDAE 3 Color a more or less uniform chocolate-brown and shining on the coarse, axial riblets and dull in between. Whorls 15+ and flat sided. Suture well defined. Spire extended, the upper portion slightly coneave. Aperture holostomatous, subcircular, slightly flaring and reflected. Axis simple. Sculpture consisting of a series of somewhat irregular axial costae which are diagonal and fine, irregular axial threads in between the costae. There is a well defined basal ridge. Remarks. In relationship, this species is nearest to Archego- coptis tiburonica Clench, differing, however, by being much lar- ger, lacking the whorl shoulder and in having a much larger basal ridge. The sculpture is similar in both species, though coarser in A. tiburonica. Many species in the family Urocoptidae lose many of the upper whorls when approaching the adult stage or during their adult life. They produce a plug in one of the mid or earlier whorls and later the whorls above the plug are lost. The fracture is a mechanical one and not automatic, as quite frequently a few adult specimens in a single population may still retain their early whorls. It is of considerable interest to note that A. havtien- sis apparently does this twice, as three of the five paratypes have plugs seven whorls above the aperture; the remaining two paratypes are young specimens and have lost the early three or four whorls. Only the holotype has retained the mid-whorls. ARCHEGOCOPTIS DECAPITATA (Roding) Cerion decapitatum Roding, 1798, Museum Boltenianum, p. 90 (refers to Chemnitz, 1786, Conchylien-Cabinet (1)9: pl. 136, figs. 1256-1257 [fig. 1255, in error] ). Cyclostoma fasciata Lamarck, 1816, Eneyelopédie Méthodique, Atlas, p. 12, pl. 461, fig. 7. Helix truncata Dillwyn, 1817, Descriptive Catalogue of Recent Shells, London 2:948 (Santo Domingo). Helix fasciata Lamarck: Férussac, 1822, Tableaux Systématiques, p. 61, no. 503 (The Antilles). Cylindrella fasciata ‘Chemnitz’ Pfeiffer, 1862, Malakozoologische Blatter 9:199 (Corail [Haiti] ). Cylindrella fasciata ‘Chemnitz’ Crosse, 1891, Jour. de Conch. 39:134 (Co- rail, near Jérémie, SW Haiti). Urocoptis truncata (Dillwyn): Pilsbry, 1903, Manual of Conchology (2)15:154, pl. 39, figs. 27-28 (Corail, near Jérémie, Haiti). Cyclostoma fasciata Lamarck: Mermod, 1952, Revue Suisse de Zoologie 59:46, fig. 111. 4 BREVIORA No. 245 Remarks. This species has had a long and varied history. The only specific locality previously known was that of Weinland who had collected a single worn specimen in a small water course at Corail, Haiti (Crosse, 1891). From this, Pilsbry had assumed that it had been carried down this stream from the mountains above. In 1927, W. J. Eyerdam collected several dead specimens under stones at the base of a cliff in a forested area on Grande Cayemite, a small island about 644 miles NE of Corail. It is possible that the specimen found by Weinland may have drifted from Grande Cayemite. The specimens collected by Eyerdam are certainly in the genus Archegocoptis and appear to be A. decapitata Roding, differing only in that the color band of dull red is on a whorl lower than those figured by both Chemnitz and Lamarck. All specimens so far as known were collected dead. Urocoptis (GONGYLOSTOMA) WHITTUMI new species Plate 2, figure 1 Holotype. MCZ 59580, from Guabairo, near Soledad, Cien- fuegos, Las Villas, Cuba, collected by Clench, Crozier and Navez, October 1928. Paratypes. MCZ 59284, 86289, 105148, all from the above lo- eality, and collected by various students and visitors to the Harvard Tropical Gardens. Height Width mm mn 11.5 2.7 Paratype 16 2.9 eS 14.8 3 zt 16 3 = 17.3 3.2 Holotype Ts 3 Paratype 18.5 3 20.5 3.9 at Description. Shell somewhat fusiform, usually decollated, nar- row, widest point a little below the middle in the decollated speci- mens. Color very pale reddish brown, darker on the earher whorls. Whorls 18 to 22 (not always decollated at the same whorl). Whorls rather flat, evenly tapering towards the spire tip. Last whorl free for one-quarter to one-third of its length. Spire acute, first two whorls a little larger than the third and fourth. Aperture holostomatous, cireular, white and reflected. 1966 NEW SPECIES OF UROCOPTIDAE 5 Columellar axis stout, encircled by a single, wide lamella which bears numerous deep, spine-like crenulations. Sculptured by nu- merous strong, almost straight riblets, about 17-18 on the second whorl. Remarks. This species belongs to the group of U. canteroiana, but differs in having fewer and stronger riblets, different shape and proportions, and by its internal axis, which is stouter; it is also encircled by a larger lamella, the margin of which possesses a greater number of tooth-like processes. This shell is very variable in size, but very constant in its rela- tive proportions. The riblets are approximately the same in number on the different sizes. Named for Mr. Walter Whittum, now of Springfield, Massa- chusetts, whose kindness and courtesy made many trips possible in the rich collecting ground about Guabairo, at the time he was in charge of a colonia for the Soledad Sugar Company in 1928. Urocopris (GONGYLOSTOMA) EXQUISITA new species Plate 2, figure 2 Holotype. MCZ 59286, from one-half mile E of Guabairo, Soledad, Cienfuegos, Cuba, W. J. Clench collector, December 1927. Paratypes. MCZ 59285, 59579, from the same locality. Entire specimens Decollated specimens Height Height mm mm 14.4 Holotype eT 14.4 1S) 12.9 eg Description. Shell fusiform, slender, generally entire; when decollated, limited to only a very few of the early whorls. Widest point about midway in decollated shells. Color light horn, not uniform, but generally somewhat marbled with patches of lighter and darker areas, occasionally albinistic ; surface somewhat glis- tening, often translucent in certain areas. Whorls convex, evenly tapering towards both the spire tip and the aperture. In complete shells whorls number 21-23; in decollated specimens 15-18 whorls. Spire acute, first two whorls slightly larger than the third and fourth. Last whorl free, developed with a long, slightly curved neck. Aperture holostomatous, circular, white and well reflected. 6 BREVIORA No. 245 Axis of columella thin and sinuous with a single spiral thread. Suture somewhat impressed. Sculpture consisting of very fine oblique axial riblets except on the free whorl which has many relatively coarse riblets, irregularly spaced and nearly encircling the whorl except on the face or aperture side. Remarks. This species belongs in the subgenus Gongylostoma Albers, near Urocoptis barbourt Torre and Clench. It differs from U. barbourt, however, in being smaller, and with a larger number of whorls in the decollated specimens, and is relatively more slender and more fusiform. The color is very similar, though the shell is a little more translucent in exrquisita. The free aperture whorl is also much longer than in U. barbourt. The habitat of this species is quite peculiar for the genus in this region, as specimens were found in more or less heavy woods, on the rocky walls of cave entrances, or in other large rocky fissures. The other urocoptids occupied more open areas, espe- cially U. livida barbourt and U. livida atkinsi Torre and Clench. Urocoptis (GONGYLOSTOMA) CANTEROIANA (Arango) Plate 2, figure 4 Cylindrella canteroiana ‘Gundlach’ Arango, 1875 [1876], Anales Real Academia Ciencias Medieas, Fisicas y Naturales, Habana 12:284 (La Vigia, Trinidad [Cuba]); Arango, 1878, Fauna Malacologica Cubana, Habana, p. 117 (environs of Vigia, Trinidad). Urocoptis canteroiana ‘Gundlach’ Arango: Pilsbry, 1903, Manual of Conchology (2)15:254. Height Width mm mm 1225 2.0 Lectotype Lectotype. Here selected, MCZ 189252, from La Vigia, Trini- dad, Cuba, collected by Gundlach and sent to MCZ by Rafael Arango. Remarks. This species has not been figured previously. It is comparatively rare as only a few specimens have been collected. We have specimens from La Vigia, Finea Cantero, and Finea La Pastora, all in the vicinity of Trinidad, Las Villas, Cuba. Urocoptis (GONGYLOSTOMA) DIAGONALIS new species Plate 2, figure 3 Holotype. MCZ 59292, from Mina Carlota, Sierra de San Juan, 8 miles S of Cumanayagua, Las Villas, Cuba, W. J. Clench and Calvin Goodrich collectors, November 30, 1927. a ( 1966 NEW SPECIES OF UROCOPTIDAE Paratype. A single paratype from the same locality is in the Museum of Zoology, University of Michigan. Height Width mm mm 16 3.3 Holotype 15 3 Paratype Description. Shell reaching 16 mm in height (decollated), fusiform, imperforate and sculptured. Color a dull gray. Whorls 16 (remaining), very shehtly convex, the last free for about one- fifth of a whorl. Suture indented. Spire decollated, a loss of 12-14 early whorls. Aperture holostomatous, white and flaring. Seulpture consisting of numerous fine, axial riblets which are shghtly diagonal and are flatly sigmoid in shape. On the free portion of the last whorl these riblets become rings and are closer together. Remarks. This species is related to U. canteroiana (Arango), but differs in several of its characters. U. diagonalis is larger, has finer sculpture of diagonal riblets which are straight in can- terovana and flattened sigmoid in shape in diagonaiis. Urocoptis (UROCOPTOLA) CAYEMITENSIS new species Plate 1, figure 4 Holotype. MCZ 254666, from NE Grande Cayemite, a small island about 614 miles NE of Corail, Département du Sud, Haiti, collected by W. J. Eyerdam, July 1927. Paratypes. MCZ 254667, population 1; and MCZ 254668, population 2. From the same locality as the holotype. Height Width mm mm aha 6.5 Holotype, population no. 1+ 18.5 i Paratype om 17.3 5.8 a 7 15.8 6 ea es 20.5 (ical os population no, 21 20.4 7 a a 20.1 7 i os Description. Shell oblong, glossy, having the greatest diameter above the middle, imperforate, sculptured and rather thin in structure. Color a pale brownish pink. Whorls very shghtly 1 Both populations from the same general area. BREVIORA No. 245 (© e} convex and tapering above the mid-whorls to the truncated sum- mit. Last whorl with a distinct basal ridge. Spire extended. Aperture nearly circular, the parietal area adnate to the whorl above, lip flaring. Suture moderately impressed. Sculpture con- sisting of numerous, slightly diagonal, axial costae, which do not produce crenulations at the suture. Axis with a single twist. Remarks. This species is a member of the U. sericea (Pfeiffer ) complex which is widely distributed along coastal Haiti within the Gulf of Gonave. Urocoptis cayemitensis differs from U. seri- cea by being much smaller, imperforate, and in having a uniform sculpture throughout. The spire is truncated between the sev- enth and eighth whorls from the aperture. Urocoptis (UROcOpTOLA) EKMANI new species Plate 1, figure 5 Holotype. MCZ 260871, from La Source, NW Gonave Island, Haiti, collected by W. J. Eyerdam, August 1927. Paratypes. MCZ 260872, from the locality of the holotype. Height Width mm mm 25 113) Holotype 26 11 Paratype 22.5 10.2 ie Description. Shell large, elliptical in shape, remaining whorls reaching 26 mm in height, imperforate, rather solid and sculp- tured. Whorls 8 (remaining) and moderately convex. Flesh colored with a few, small, irregular, dark brownish red flecks on all of the whorls. Spire extended and tapering rapidly from the mid-area and produced at an angle of about 55°. Aperture sub- circular, holostomatous, one specimen out of eight being adnate above. There is a slight and closed rimation at the umbilical area. Suture slightly indented. Sculpture consists of diagonal and straight, fine axial riblets on the upper whorls and diagonal and arcuate axial riblets on the lower whorls. Protoconch un- known. Remarks. This species is related to Urocoptis bencomoi Clench, but is smaller, proportionally narrower, and has the hp holosto- matous. Named for Dr. Erik L. Ekman, the Swedish botanist who was with W. J. Eyerdam during his trip to Gonave Island. 1966 NEW SPECIES OF UROCOPTIDAE 9 Urocoptis (UROCOPTOLA) CARIBBAEA new species Plate 1, figure 3 Holotype. MCZ 260873, from 1 mile E of Pointe a Raquette, south central Gonave Island, Haiti, collected by W. J. Eyerdam, July 1927. Paratypes. MCZ 260874, from the same locality. Height Width mm mm 24.8 12 Holotype 22 11.6 Paratype 22.5 lal: “s 22.5 11.4 se Description. Shell large, reaching about 25 mm in height, im- perforate, rather solid and sculptured. Whorls 7 (remaining) and moderately convex. Flesh colored with a few small, irregu- lar, dark brownish red flecks on all of the whorls. Spire extended and tapering to the apex and the base from the mid-area and produced at an angle of about 60°. Aperture subcireular and adnate on the parietal area. Suture slightly indented. Sculpture consists of diagonal and straight, fine axial riblets and somewhat arcuate on the last whorl. Riblets on the two remaining early whorls more widely spaced. Protoconch unknown. Remarks. This species is closely related to U. ekmamni, differing in being shorter, with different proportions, having all specimens adnate on the parietal area and in having the axial riblets more widely spaced on the remaining early whorls. Urocoptis (IpIoSTEMMA) PERPLICATA (Beck) Helix (Cochlodina) perplicata Férussac, 1821 [1822], Tableaux Sys- tématiques Animaux Mollusques, p. 61 or 65, no. 506 (The Antilles) [nomen nudum). Brachypodella perplicata Férussac: Beck, 1837, Index Molluscorum, p. 89 [refers to Férussac’s pl. 163, fig. 9]. Clausilia perplicata Férussac: Deshayes [in] Lamarck, 1838, Animaux sans Vertébres (2)8:216. Cylindrella perplicata Férussac: Philippi, 1847, Abbildungen Conchylien 2:217, pl. I (Achatina), fig. 9 (The Antilles). Cylindrella perplicata Férussac, 1851, Histoire Naturelle Générale et Particuliére 2:229, pl. 163, fig. 9 (The Antilles); non Cylindrella perplicata Férussac. Pfeiffer 1840 [in] Weigmann, Archiv fiir Naturgeschicte 1:41 (Fundador, Matanzas, Cuba). 10 BREVIORA No. 245 Cylindrella fastigiata ‘Gundlach’ Pfeiffer, 1860, Malakozoologische Blat- ter 7:20 (Baracoa, Mata, and Yunque, Cuba); Pfeiffer, 1865, Novitates Conchologicae 2:263, pl. 65, figs. 23-25 [Syntypes, MCZ 26639, 86621]. Urocoptis fastigiata Pfeiffer: Pilsbry, 1903, Manual of Conchology (2)15:171, pl. 45, figs. 36-41 (Baracoa, El Yunque, Mata, Oriente, Cuba). Brachypodella perplicata Férussac: Pilsbry, 1903, Manual of Conchology (2) 16:83, pl. 7, figs. 23-24 (Antilles). Remarks. Considerable confusion has existed concerning the proper name for this species. Pilsbry retained both perplicata and fastigiata in two different genera but noted under perplicata that ‘‘the ribs follow one another from whorl to whorl, as in Urocoptis fastigiata,’’ but he gave no indication that the two names applied to the same species. Since Pilsbry’s Manual of Conchology was published in 1903 far more material in this family is available for study, not only from Cuba but from the entire West Indies as well. The credit for the name of this species must go to Beck rather than to Férussac, as Férussae did not publish this name until 1851. The plates were issued much earlier but there were no names attached to them. Through some source, other than publi- cation, these names were available, and Beck was the first to cite a plate and figure for this species. Specimens examined. CUBA: OrtenTE, Farallones de Ba- rigua; Yunque de Baracoa; Silla de Baez, W of Baracoa; Sa- bana, Cabo Maisi; Zona de la Caleta, Baracoa; Arriba Mandinga, Baracoa; Arriba Guandao, Mandinga; Mata, Punta Maisi; Boca de Taco, Nibujon, Baracoa. BRACHYPODELLA (BREVIPEDELLA) IMITATRIX Pilsbry Brachypodella (Brevipedella) imitatrix Pilsbry, 1903, Manual of Conchol- ogy (2)16:47, pl. 8, figs. 54-55 (Port-au-Prince, Sans-Souci, St. Mare, and La Ferriére, Haiti). Remarks. This species has a much wider distribution than most species in the Urocoptidae, extending as it does from Haiti to the Samana peninsula in eastern Santo Domingo. A closely related species, B. angulifera (Gundlach), occurs in eastern Cuba. Specimens examined. Hartt: Bizonton, Lasbaines; Port-au- Prince and Diquini. SAnto Dominco: Mt. Isabel de Torres, Puerto Plata, at 1200 to 1600 feet; Sanchez; Penon de Basiles and Penon de Maria Luisa, both Santa Barbara de Samana, and Punta Lirio, 2 miles E of Santa Barbara de Samana. 1966 NEW SPECIES OF UROCOPTIDAE 1] BRACHYPODELLA (GYRAXIS) SAMANA new species Plate 2, figure 5 Holotype. MCZ 57214, from Penon de Maria Luisa, Santa Barbara de Samana, Républica Dominicana, collected by W. J. Clench, H. D. Russell and R. A. McLean, August 1937. Paratypes. MCZ 57218, from the same locality; and MCZ 57213 from Penon de Basiles in the same general area. Height Width mim mnt 14.5 2.0 Holotype 12.5 2.4 Paratype 14.5 2.9 i 12.8 2.4 = Description. Shell slender, imperforate, generally entire, widest at the mid-section of the shell and finely sculptured. Color marbled in yellowish and light brown with some areas white, the brownish areas translucent, the white and yellowish areas somewhat opaque. Whorls 13 to 14 in entire specimens and moderately convex with the last whorl free. Suture well defined. Spire acute, aperture holostomatous, subcircular with the lip flaring slightly. Axis gyrate. Sculpture consisting of numerous, fine, axial, slightly diagonal and slightly curved costae; the first two whorls finely and axially costate. On the early whorls certain of these costae in small groups impinge shghtly on the suture. Remarks. This species is closely related to Brachypodella (Gyravris) sericata Pilsbry. It differs by being larger, having slightly coarser thread-like costae and having the marbled colora- tion, sericata being a dull white. Brachypodella sericata Pilsbry was collected originally by William M. Gabb during 1869-71 while on a geological survey of Santo Domingo. Unfortunately, the recent mollusks collected by him seldom had specific locality data. We collected this species in some numbers at San Lorenzo Bay, Bahia de Samana. This locality can be considered the type locality as Gabb surveyed this area about the Bahia de Samana. (Received December 1, 1965.) 12 BREVIORA No. 245 Plate 1 Fias. 1-2. Archegocoptis haitiensis n. sp. Morne Rochelois, 25 km WSW of Miragoane, Dépt. du Sud, Haiti, at 1000 meters. Fig. 1, Paratype, MCZ 252057; fig. 2, Holotype, MCZ 252056 (both 3X). Fie. 3. Urocoptis (Urocoptola) caribbaea n. sp. 1 mile E of Pte. a Ra- quette, S central Gonave Island, Haiti. Holotype, MCZ 260873 (3X). Fie. 4. Urocoptis (Urocoptola) cayemitensis n. sp. NE Grande Cayemite, Dépt. du Sud, Haiti. Holotype, MCZ 254666 (4X). Fig. 5. Urocoptis (Urocoptola) ekmani n. sp. La Souree, NW Gonave Island, Haiti. Holotype, MCZ 260871 (3X). 13 NEW SPECIES OF UROCOPTIDAE 1966 PLATE 1 14 BREVIORA No. 245 Plate 2 Pie. 1. Urocoptis (Gongylostoma) whittumi n. sp. Guabairo, Soledad, Las Villas, Cuba. Holotype, MCZ 59580 (5.5 X). Fig. 2. Urocoptis (Gongylostoma) exquisita, n. sp. % mile E of Guabairo, Soledad, Las Villas, Cuba. Holotype, MCZ 59286 (5.5 X). Mina Carlota, Sierra Fig. 3. Urocoptis (Gongylostoma) diagonalis n. sp. Holotype, MCZ de San Juan, 8 miles S of Cumanayagua, Las Villas, Cuba. 59292 (5.5 X). Fia. 4. Urocoptis (Gongylostoma) canteroiana (Arango). (La Vigia, Trinidad [Las Villas, Cuba]). Lectotype, MCZ 189252 (5.5 X). Fic. 5. Brachypodella (Gyraxis) samana n. sp. Pefon de Maria Luisa, Santa Barbara de Samana, Reptblica Dominicana. Holotype, MCZ 57214 (SHY DONe =a | roy e ) —- uw = BREVIORA Museum of Comparative Zoology CAMBRIDGE, Mass. May 3, 1966 NuMBER 246 PSEUDANTHESSIUS PROCURRENS N.SP., A CYCLOPOID COPEPOD ASSOCIATED WITH A CIDARID ECHINOID IN MADAGASCAR By ArtHur G. HUMES Boston University, Boston, Massachusetts and Associate in Marine Invertebrates, Museum of Comparative Zoology INTRODUCTION During an extensive search in 1963-64 for copepods associated with marine invertebrates at Nosy Bé, in northwestern Mada- gascar, 109 adults and 9 copepodids of the new hchomolgid cope- pod described below were recovered from the sediment obtained after washing 30 large pencil urchins, Phyllacanthus imperialis (Lamarck), in weakly alcoholized sea water. (The host echinoid is widespread in the Indo-Pacific region, where it occurs, for ex- ample, in Australia, the Marshall Islands, the Philippine Islands, Ceylon, the Red Sea, and Zanzibar). This new form brings the total number of species known in the genus Pseudanthessius to 24 (including the 22 species listed by Stock, Humes, and Gooding, 1963, and a new species from a polychaete annelid in Madagascar whose description by Humes and Ho is in press). ACKNOWLEDGMENTS The copepods were collected by the author while participating in the 1963-64 activities of the U.S. Program in Biology of the International Indian Ocean Expedition. The study of the specimens has been aided by a grant (GB- 1809) from the National Science Foundation of the United States. I wish to thank Dr. H. Barraclough Fell, Professor of In- vertebrate Zoology at the Museum of Comparative Zoology, for the identification of the echinoid host, and to acknowledge the assistance to the field work given by the staff of the Centre d’Océanographie et des Péches at Nosy Bé, 2 BREVIORA No. 246 DESCRIPTION Family LICHOMOLGIDAE Kossmann, 1877 Genus PSEUDANTHESSIUS Claus, 1889 PSEUDANTHESSIUS PROCURRENS! n.sp. Figures 1-29 Type material. —16 females, 9 males, and 1 copepodid from washings of 3 pencil urchins, Phyllacanthus imperialis (la- marek), in 1 meter depth among dead coral (Acropora) at Pte. Ambarionaomby, Nosy Komba, near Nosy Bé, Madagascar. Col- lected November 28, 1963. Holotype female, allotype, and 19 paratypes (13 females and 6 males) deposited in the U. S. Na- tional Museum, Washington, and the remaining paratypic adults (dissected) together with the copepodid in the collection of the author. Other specimens (all from Phyllacanthus imperialis collected in 1963 at the type locality). — 2 females from 1 host, July 3; 10 females and 7 males from 3 hosts, July 18; 14 females and 5 males from 6 hosts, August 23; 13 females, 6 males, and 4 cope- podids from 7 hosts, October 30; and 16 females, 11 males, and 4 copepodids from 10 hosts, December 14. This last collection is deposited in the Museum of Comparative Zoology. Female. — The body (Figs. 1 and 2) has a broadened pro- some. The length (excluding the setae on the caudal rami) is 0.95 mm (0.90-1.01 mm) and the greatest width is 0.44 mm (0.42-0.46 mm), based on 10 specimens. The ratio of the length to the width of the prosome is 1.28:1. The segment bearing leg 1 is almost completely fused with the head, the only indication of separation being a short weak crease on each side. The lateral areas of the metasomal segments are rounded. The segment of leg 5 (Fig. 3) is expanded laterally, being 50 p. in length and 133 » in width. Between the segment of leg 5 and the genital segment there is a ventral intersegmental sclerite (see Fig. 2). The genital segment (Fig. 3) is 127 » long. Anteriorly its lateral margins form 2 rounded, strongly sclerotized lobes (the width of the segment at this level being 115 ».). The width at the level of the dorsolateral areas of attachment of the egg saes is 107 uw. Behind each attachment area the segment is slightly 1 The specific name procurrens, from Latin procurrere, meaning to bulge out or project, alludes to the outer expansion on the coxa of leg 1 and to the 2 rounded lobes on the anterior part of the genital segment in the female. 1966 NEW COPEPOD FROM AN ECHINOID 3 constricted with nearly parallel margins (the width in this re- gion being 86 ».). Each egg sae attachment area (Fig. 4) bears anteriorly a slender, slightly haired seta (21 » long) and just posterior to it a short naked seta (6.5 ». long) composed of an ex- panded sclerotized basal portion and a slender hyaline distal part. Medial to the latter seta there are 2 small spinelike proc- esses. The three postgenital segments are 44 x 81, 39 x 75, and 65 x 72 v. from anterior to posterior. The anal segment bears on each side on its distal margin a dorsal and ventral row of spinules. The caudal ramus (Fig. 5) is elongated, with a terminal ventral expansion whose margin bears a row of spinules. The length along the inner side of the ramus to the end of the ex- pansion is 114 wu, along the outer side 104 pw, and the width at the level of the outer seta is 24 vy. The ratio of length to width is about 4.5:1. The outer seta, inserted 70 ». from the base of the ramus, is naked and 56 v. in length. The pedicellate dorsal seta is 33 wand slightly haired. The outermost terminal seta (100 yu) and the innermost terminal seta (72 y.) are haired. The 2 long median terminal setae are 177 and 250 y in length respectively and haired. A minute lateral setule is borne on the outer basal margin of the ramus. The dorsal and ventral surfaces of the ramus bear a few refractile points. The dorsal surface of the prosome and the dorsal and ventral surfaces of the urosome bear minute setules and refractile points. In addition, the outer ventral areas of the head carry a sub- marginal row of refractile points (Fig. 6). The ratio of the length of the prosome to that of the urosome is 1.53:1. The ege sacs (Fig. 1) are moderately elongated, often rather pointed posteriorly, and contain numerous eggs. In one female the egg sacs measured 385 x 220 wu, with each egg about 57-60 v. in diameter. The rostral area (Fig. 7) is moderately well developed. Be- tween this area and the front of the labral region there is a slight protrusion on the ventral surface of the head. The first antenna (Fig. 8) is 7-segmented, with the lengths of the segments (measured along their posterior non-setiferous mar- gins) 24 (39 » along the anterior margin), 103, 21, 39, 33, 20, and 17 uw respectively. The formula for the armature is 4, 13, 6, 3, 4 + 1 aesthete, 2 + 1 aesthete, and 7 + 1 aesthete. All the setae are naked. The second antenna (Fig. 9) is 4-segmented, with the last seg- ment elongated (69 » along the shorter ventral margin, 93 pv. along 4 BREVIORA No. 246 the dorsal margin, and 19 ». in width). Each of the first two seg- ments bears a small ventral seta, the third segment bears 4 slender setae (one of them very small), and the last segment bears 2 unequal slender recurved claws (47 and 25 y. respectively along their axes) and 5 setae, one of them very long (99 ».). The extremity of the last segment is swollen, so that the 2 claws insert at one side rather than directly on the tip of the segment. All the setae are naked. The labrum (Fig. 10) consists of 2 diverging, pointed lobes with their medial edges straight and finely dentate, both arising from a large, conspicuous, sclerotized area which projects (Fig. 2) from the ventral surface of the head. On the posterior wall of the labrum, in front of the mouth area, there is a pair of small sclerotized lobes. The surface of the labrum lacks fine ornamenta- tion. The mandible (Fig. 11) has on the concave side of the blade an oblique row of spinules followed distally by a spinelike process lying parallel to the blade and evidently not articulated with it. The convex side of the blade bears a fringe of graduated spinuli- form structures without definite articulations. The paragnath (Fig. 12), lying medial to the base of the first maxilla (as in Fig. 16), is a rounded lobe bearing a small sclerotized outer process, a small postero-inner knob, and a posterior group of hairs. The first maxilla (Fig. 13) is a single segment bearing 4 naked elements, comprising terminally 2 obtuse subequal spines (10 and 8 » long) and a shorter pointed spine (5 y.) and sub- terminally a naked hyaline seta (11 »). The second maxilla (Fig. 14) is 2-segmented. The first segment is unarmed but has a small sclerotized protuberance on its expanded margin. The sec- ond segment is produced to form a long bilaterally spinulose lash; the dorsal surface bears a proximal seta 29 » in length (bearing lateral spinules along one edge) and a distal row of 3-5 spinules. The maxilliped (Fig. 15) is 3-segmented. There are 2 naked setae on the second segment. The terminal segment bears proximally a spine with unilateral spinules, a hyaline naked setule, and near the base of the latter a minute setule; the seg- ment is produced to form an attenuated spinelike structure with a row of long spinules along one side and a row of minute spinules near the opposite margin. The postoral area (Fig. 16) shows between the paragnaths a shield-shaped area which projects ventrally to form a low median process. Posterior to the selerotization which almost joins the 1966 NEW COPEPOD FROM AN ECHINOID 5 bases of the maxillipeds, the ventral surface of the cephalosome protrudes slightly (best seen in a lateral view, as in Fig. 2). Legs 1-4 (Figs. 17, 18, 19, and 20) have trimerous rami except for the endopod of leg 4 which is unimerous. The armature of the legs is as follows (the Roman numerals indicating spines, the Arabic numerals setae) : Esl a protopod) | .0=1-- 1-0 exp | 0. It; Sih endie Ot 0-1. 6 Tes Fo, Sprotopods O-l- A-0) exp > 10.) ik. iio end 0-1; 0-2; JTIL3 Po protopod. 0-1; 1-0. exp) 4-0- eb Thr end, 03-9) (0-2, Ieit-2 EA protopod: J0-1-, 91-0) .exp £0. > iets Tis end II The inner seta on the coxa of legs 1-3 is long and plumose, but in leg 4 this seta is minute (7 » long) and naked. In the first 3 legs the inner margin of the basis bears a short row of hairs, but in leg 4 these hairs are lacking. The outer coxal margin of leg 1 is expanded to form a prominent lobe (Fig. 17) ; in legs 2 and 3 this expansion is much less prominent and in leg 4 it is ap- parently absent. The tips of the outer spines on the exopods are slightly recurved posteriorly and the more proximal ones have minute terminal flagella. The 3 spines on the last segment of the endopod of leg 2 are 14, 10, and 13 wu in length from proximal to distal, with the middle one having a short terminal flagellum. In leg 4 the exopod is longer than in any of the preceding legs. The endopod measures 53 x 16 wu. and has nearly parallel margins without a constriction or notch. It bears a row of hairs on its outer proximal third and an anterior row of small spinules near the insertions of the 2 divergent terminal spines (the outer 18 yu. long with a minute flagellum, the inner 37 » long with a more strongly spinulose flange along the outer side than along the inner side). Leg 5 (Fig. 21) consists of a strong spine (33 y. long) and an adjacent seta (21 p), together with a dorsal seta (22 y), orna- mented as in the figure. External to the spine and seta there is a row of minute blunt spinules at the apex of the segment. AI- though, as in other species of Pseuwdanthessius, there is no free segment of leg 5, it is likely that the spine and its adjacent seta correspond to the terminal armature in other lichomolgids. Leg 6 is probably represented by the 2 setae near the attach- ment of each egg sae (see Fig. 4). 6 BREVIORA No. 246 The color in life in transmitted hght is translucid, the eye red, the intestine black, the ovary gray, the ege sacs opaque gray. (Although in specimens preserved in 70 per cent ethyl alcohol the color is an opaque grayish brown, the color changes quickly to a bright red when the copepods are placed in lactic acid. ) Male.— The body (Figs. 22 and 23) has a much narrower prosome than in the female, but otherwise resembles that sex in general form. The length (not including the setae on the caudal rami) is 0.76 mm (0.73-0.78 mm) and the greatest width is 0.27 mm (0.25-0.28 mm), based on 10 specimens (including the allo- type, 8 paratypes, and 1 specimen from Pte. Ambarionaomby on October 30). The ratio of length to width of the prosome is eon aene The segment of leg 5 is similar to that of the female, and meas- ures 31 x 94». The genital segment (Fig. 24) is nearly as long as wide, 104 x 110 y, and in dorsal view has a subspherical out- line. In lateral view (Fig. 25) the anteroventral part of the segment projects noticeably. There is no intersegmental sclerite between the segment of leg 5 and the genital segment. The 4 postgenital segments are 39 x 52, 32 x 51, 30 x 50, and 43 x 53 pv. from anterior to posterior. The caudal ramus resembles that of the female, but is a little shorter, the inner length being 78 pv, the outer length 73 yu, the width 23 1, and the ratio of length to width 3.26:1. The surfaces of the prosome and urosome bear minute setules as in the female. The ratio of the length of the prosome to that of the urosome is 1.31 :1. The rostral area, first antenna, second antenna, labrum, man- dible, paragnath, and first maxilla are like those in the female. The second maxilla also resembles that of the female, but lacks the small sclerotized protuberance on the expanded margin of the basal segment. The maxilliped (Fig. 26) is much elongated, slender, and 4-segmented (assuming that the fourth segment is represented by the proximal part of the claw). Its entire length including the claw is about 300 wu. The first segment is unarmed. The second bears 2 unequal inner setae and 2 rows of hairs, one along the inner margin distal to the setae and another starting on the proximal inner margin and passing obliquely to the distal posterior surface of the segment. The very short third segment is unarmed. The terminal recurved claw, 135 y. in length along its axis, bears a conspicuous terminal lamella. The slightly crenated fringe along its concave margin is interrupted about midway. Near the base of the claw on its posterior surface there is a seta 1966 NEW COPEPOD FROM AN ECHINOID 1 42 y. long with minute lateral spinules, and on its anterior sur- face there are 2 small naked setules, one 10 yp, the other 4 y long. The postoral area is like that of the female. Legs 1-4 resemble those of the female except that the last seg- ment of the endopod of leg 1 is more elongated (Fig. 27), and the terminal segment of the endopod of leg 2 is more elongated, and its 3 spines are longer (24, 22, and 25 » from proximal to distal), as seen in Figure 28. Leg 5 is similar to that of the female. Leg 6 (Fig. 29) consists of a posteroventral flap on the genital segment. Beyond the rim of the segment the leg projects con- spicuously (see Fig. 24) as a large, pointed, ventral sclerotized process dorsal to which there is a shorter rounded process bearing 2 naked setae 22 and 24 v. in length. The spermatophore, seen only inside the body of the male (Fig. 24), is oval, about 72 x 45 yp, with a short neck. The color in life in transmitted light resembles that of the female. RELATIONSHIP TO OTHER SPECIES IN THE GENUS Twelve species of Pseudanthessius may be readily distin- euished from P. procurrens in that they lack the prominent outer expansion on the coxa of leg 1 and do not bear the two rounded sclerotized lobes on the anterior part of the genital segment in the female. These species are: P. aestheticus Stock, Humes, and Gooding, 1963, P. concinnus Thompson and A. Scott, 1903, P. deficiens Stock, Humes, and Gooding, 1963, P. dubius G. O. Sars, 1918, P. graciloides Sewell, 1949, P. luculentus Humes and Cressey, 1961, P. mucronatus Gurney, 1927, P. nemertophi- lus Gallien, 1935, P. notabilis Humes and Cressey, 1961, P. pec- tinifer Stock, Humes, and Gooding, 1963, P. tortuosus Stock, Humes, and Gooding, 1963, and a new species (Humes and Ho, in press) from a polychaete annelid in Madagascar. Hight other species may be separated from P. procurrens in that they lack the two lobes on the anterior part of the genital segment in the female and have a different armature on the last segment of the second antenna. (Unfortunately, in the original descriptions of these species, definite information on the con- dition of the outer coxal margin of leg 1 was not given.) These species are: P. gracilis Claus, 1889, P. latus Illg, 1950, P. ob- scurus A. Scott, 1909, P. sawvagei Canu, 1892, P. spinifer Lind- berg 1945, P. tenuis Nicholls, 1944, P. thorelli (Brady, 1880), and P. weberi A. Scott, 1909. 8 BREVIORA No. 246 Three species remain to be compared with P. procurrens: P. liber sensu Sewell, 1949, P. liber (Brady, 1880), and P. assimilis G. O. Sars, 1917. The species referred to as P. luber (Brady and Robertson) by Sewell (1949) is probably a new and unnamed species of the genus (see Humes and Cressey, 1961, pp. 80-81). It differs from P. procurrens chiefly in the presence of two long elements (aesthetes ?) on the basal segment of the first antenna, in the inwardly curving terminal spine on the last segment of the exopod of legs 2-4, and in the notch on the outer margin of the endopod of leg 4. No information was given for this form regarding the nature of the genital segment in the female, and the condition of the outer coxal margin of leg 1 is not clearly shown in Sewell’s text figure 32 D. P. liber (Brady, 1880)" lacks the two lobes on the anterior part of the genital segment of the female, has a caudal ramus about twice as lone as wide, and has a different armature on the last segment of the second antenna. Brady did not mention an outer coxal expansion on leg 1, but Sars (1917) both mentioned and figured such an expansion in specimens taken in Norway. The expansion illustrated on Sars’ plate XCIV is, however, less pronounced than in P. procurrens. The armature for the last segment of the endopod of leg 3 in the female is, according to Sars, I,11,3, instead of I,I1,2 as in the Madagascan species. P. assimilis G. O. Sars, 1917, is said by Sars to be closely allied to P. liber (Brady, 1880). Like the latter species it lacks the two lobes on the genital segment of the female and has an armature on the last segment of the second antenna unlike that of P. pro- currens. Sars stated that the legs are ‘‘almost exactly as in P. liber,’’ implying that there is a similar outer coxal expansion on leg 1. Judging from the available information, P. procurrens appears to be near both P. liber (Brady, 1880) and P. assimilis G. O. Sars, 1917, resembling them in having an outer coxal expansion on leg 1, but differing in having two lobes on the anterior part of the genital segment in the female and in having a different armature on the last segment of the second antenna. SUMMARY The new species Pseudanthessius procurrens is associated with the cidarid echinoid Phyllacanthus imperialis (iamarck) in the 2 Not Brady and Robertson, 1876. See Stock, Humes, and Gooding, 19638, p. 10, footnote. 1966 NEW COPEPOD FROM AN ECHINOID 9 region of Nosy Bé, northwestern Madagascar. Within the genus the copepod may be recognized by the conspicuous expansion on the outer coxal area of leg 1 and by the two rounded lateral lobes on the anterior part of the genital segment in the female. The species nearest to the new form appear to be P. liber (Brady, 1880) and P. assimilis G. O. Sars, 1917. Although the host echinoderm is widely distributed in the Indo-Pacific area, the new copepod associated with it is known at present only from the type locality in Madagascar. REFERENCES CITED Brapy, G.S. 1880. A monograph of the free and semi-parasitic Copepoda of the British Islands. Ray Society, London, 3:1-83. BrabDy, G. S. AND D. ROBERTSON 1876. Report on dredging off the coast of Durham and North York- shire in 1874. Rept. British Ass. Adv. Sci. (Bristol), 45: 185- 199, CaNnu, E. 1892. Les Copépodes du Boulonnais, morphologie, embryologie, taxono- mie. Trav. Lab. Zool. mar. Wimereux-Ambleteuse (Pas-de- Calais), 6:1-354. GLAS; C- 1889. Uber neue oder wenig bekannte halbparasitische Copepoden, insbesondere der Lichomolgiden- und Ascomyzontiden-Gruppe. Arb. Zool. Inst. Univ. Wien, 8(3) :327-370. GALLIEN, L. 1935. Pseudanthessius nemertophilus nov. sp., copépode commensal de Lineus longissimus Sowerby. Bull. Soc. Zool. France, 60:451-459. GURNEY, R. 1927. Zoological results of the Cambridge expedition to the Suez Canal, 1924, XXXIII. Report on the Crustacea: — Copepoda (littoral and semi-parasitic). Trans. Zool. Soc. London, 22(4) :451-477. Humes, A. G. AND R. F. CRESSEY 1961. Deux nouvelles espéces de Pseudanthessius (Copepoda, Cyclo- poida) parasites des oursins @ Madagascar. Mém. Inst. Sci. Madagasear, sér. F, 1959, 3:67-82. Humes, A. G. anp J.-S. Ho In press. New cyclopoid copepods associated with polychaete annelids in Madagascar. Bull. Mus. Comp. Zool. Inne, P. 1950. A new copepod, Pseudanthessius latus (Cyclopoida: Lichomolgi- dae) commensal with a marine flatworm. Jour. Washington Acad. Sci., 40(4) :129-133. 10 BREVIORA No. 246 LINDBERG, K. 1945. Un nouveau copépode poecilostome de 1’Inde de la famille des Lichomolgidae; Pseudanthessius spinifer, n. sp. Bull. Soe. Zool. France, 70:81-84. NICHOLLS, A. G. 1944. Littoral copepods from South Australia (IL). Calanoida, Cyclo- poida, Notodelphyoida, Monstrilloida and Caligoida. Ree. So. Austr. Mus., 8(1) :1-62. Sars, G. O. 1917. An account of the Crustacea of Norway with short descriptions and figures of all the species. Vol. 6, Copepoda, Cyclopoida, pts. 11 and 12, Clausidiidae, Lichomolgidae (part), pp. 141-172. Bergen Museum, Bergen. 1918. An account of the Crustacea of Norway with short descriptions and figures of all the species. Vol. 6, Copepoda, Cyclopoida, pts. 13 and 14, Lichomolgidae (concluded), Oncaeidae, Coryecaeidae, Ergasilidae, Clausiidae, Eunicicolidae, Supplement, pp. 173-225. Bergen Museum, Bergen. Scort, A. 1909. The Copepoda of the Siboga Expedition. Part I. Free-swim- ming, littoral and semi-parasitic Copepoda. Siboga Exped., 29a)31-323. SEWELL, R. B. S. 1949. The littoral and semi-parasitic Cyclopoida, the Monstrilloida and Notodelphyoida. John Murray Exped. 1933-34, Sci. Repts., 9(2) :17-199. Stook, J. H., A. G. HUMES, AND R. U. GooDING 1963. Copepoda associated with West Indian invertebrates IV. The genera Octopicola, Pseudanthessius and Meomicola (Cyclopoida, Lichomolgidae). Studies Fauna Curacao and other Caribbean Is., 18(77) :1-74. THOMPSON, I. C. AND A. ScorTr 1903. Report on the Copepoda collected by Professor Herdman at Cey- lon in 1902. Rept. Govt. Ceylon Pearl Oyster Fish. Gulf of Manaar, pt. I, Suppl. Rept. No. 7:227-307. (Received December 10, 1965.) EXPLANATION OF THE FIGURES All the figures have been drawn with the aid of a camera lucida. The let- ter after the explanation of each figure refers to the scale at which the figure was drawn. Abbreviations: a; = first antenna, md = mandible, p = paragnath, mx; = first maxilla, mxg = second maxilla, mxpd = maxilliped, pi = leg 1. 1966 NEW COPEPOD FROM AN ECHINOID 11 scale D. 0.05mm scale A, 0.5mm scale E.0.2mm scale B. 0.2 mm scale C, 0.05 mm = FIGuRES 1-7. Pseudanthessius procurrens u.sp., female. 1, dorsal (A); 2, lateral (A); 3, urosome, dorsal (B); 4, area of attachment of egg sac, dor- sal (C); 5, caudal ramus, ventral (D); 6, edge of cephalosome, ventral (E); 7, rostral area, ventral (F). 12 BREVIORA No. 246 G. 0.03 mm scale scale H.0.1mm + Figures 8-16. Pseudanthessius procurrens n.sp., female (continued). 8, first antenna, anterodorsal (F); 9, second antenna, inner (F); 10, labrum, ventral (D); 11, mandible, inner (D); 12, paragnath, ventral (D); 13, first maxilla, outer (G); 14, second maxilla, dorsal (D); 15, maxilliped, inner (D); 16, postoral area, ventral (H). 1966 NEW COPEPOD FROM AN ECHINOID i133 Ce i) : } AS ATE\ AN. /// \ y Figures 17-21. Pseudanthessius procurrens n.sp., female (continued). 17, leg 1, anterior (F); 18, leg 2, anterior (F); 19, leg 3, posterior (F); 20, leg 4, anterior (F); 21, leg 5, dorsal (D). 14 BREVIORA No. 246 FIGURES 22-29, Pseudanthessius procurrens n.sp., male. 22, dorsal (A); 23, lateral (A); 24, urosome, dorsal (B); 25, genital segment, lateral (H) ; 26, maxilliped, anterior (F); 27, endopod of leg 1, anterior (F); 28, endo- pod of leg 2, anterior (F) ; 29, leg 6, ventral (D). BREVIORA Museum of Comparative Zoology CAMBRIDGE, Mass. May 3, 1966 NUMBER 247 THE CHANARES (ARGENTINA) TRIASSIC REPTILE FAUNA I. INTRODUCTION By ALFRED SHERWOOD ROMER THE 1964-65 LA PLATA-HARVARD EXPEDITION This is the first of a proposed series of papers giving the scien- tific results of a fossil-collecting expedition to western Argentina made during the summer of 1964-65 under the joint auspices of the Universidad de La Plata and Harvard University. The proj- ect grew out of the realization, during an earlier expedition of this sort, that further work in this area might result in im- portant additions to the early tetrapod fauna of South America. In 1958 a group of Harvard paleontologists under the direction of Prof. Bryan Patterson and myself, in cooperation with the Museo Argentina de Ciencias Naturales ‘‘Bernardino Riva- davia’’ de Buenos Aires, spent some months in exploring for vertebrate fossils in western Argentina. In part, Tertiary mam- mals were collected, and in part, fishes and brachyopid labyrin- thodonts from the Triassic Cacheuta series of Mendoza Province (ef. Rusconi, 1951, ete., and Romer, 1960). Most especially, how- ever, we explored and developed the Triassic beds of the Ischi- gualasto region in northeastern San Juan Province (cf. Romer, 1962) ; these proved exceedingly productive. Work on them has been continued by the Instituto Lillo of Tucuman, at first under Dr. Osvaldo Reig, currently under Sr. José F. Bonaparte. The Harvard members of the 1958 party included, in addition to Prof. Patterson and myself, Arnold D. Lewis, chief preparator of the Museum of Comparative Zoology, James A. Jensen, then of the Harvard staff but now in charge of geological and paleon- tological collections at Brigham Young University (Provo, Utah), and Mrs. Ruth H. Romer. Prof. Patterson’s interests have now shifted to another field. The others of the group, however, 2 BREVIORA No. 247 became deeply interested in the problems of the early history of tetrapod life in South America, and felt that, if the opportunity should be presented, a second trip of exploration in western Argentina might quite possibly result in a major addition to this story. Such an opportunity arose in 1964 through a grant from the National Science Foundation, from which could be met most of the costs of a field trip of the sort proposed. Accordingly, an agreement to this end was drawn up between the Harvard Mu- seum of Comparative Zoology and the Museo de Universidad Nacional de La Plata. Under Argentinian law 9080 of 1911, all fossils are declared the property of the national government, un- der the Ministry of Education. A commission of three directors of leading museums is empowered by the Ministry to regulate the collection and disposition of fossil materials. Our agreement with the Museo de La Plata was approved by this commission. As provided by law, type specimens and other representative materials, or casts thereof, are to be deposited in the Museo de La Plata; other specimens remain the property of the Harvard Museum. Further, we are to be allowed to have for two years exclusive right to the study and description of specimens from any new fossil deposits discovered during the course of the work. Our objective was to find fossil localities for reptiles and am- phibians in horizons in the Permian or in Triassic beds earher than that of Ischigualasto. Although the position of the Ischi- gualasto fauna is somewhat debatable, it is essentially middle Triassic in character, predominantly one of rhynchosaurs and cynodonts of gomphodont type; there are known a few advanced thecodonts, but only the bare beginnings of the dinosaur fauna which dominates the scene in the late Triassic. Faunas com- parable to that of Ischigualasto are known in Brasil (Santa Maria) and in Africa (Manda beds) ; but while antecedent Per- mian and Lower Triassic faunas are abundant in Africa, almost nothing was known of earlier tetrapods in South America. On that continent no tetrapod has ever been reported from the Permian of Argentina, and elsewhere nothing is known except for the aberrant aquatic Mesosaurus of southern Brasil (and South Africa) and a single labyrinthodont from the Brasilan Permian (Price, 1948). The earlier Triassic was, before our re- cent exploration, almost equally unknown as regards reptilian faunas. It is possible that, despite general beliefs as to its late Triassic position, the Cacheuta series may be of early Triassic age (cf. Reig, 1961), but the evidence is far from conclusive, and 1966 CHANARES REPTILES 3 little reptilian material is known from this series. Recently, two forms, a dicynodont and a gomphodont, were discovered in south- ern Mendoza Province and are now being studied by Sr. Bona- parte; these are probably early Triassic in age. This short list is a complete roster of all tetrapod vertebrates known from any part of South America in deposits older than the relatively late Triassic of the Ischigualasto Formation and the essentially comparable Santa Maria beds of Brasil. A great deal of stratigraphic work has been done in western Argentina by the geologists of the Instituto Nacional de Geologia y Mineria, the Yacimientos Petroliferos Fiscales and the Comisién Nacional de Energia Atomica, and some thousands of square miles of con- tinental deposits have been mapped as presumably Permian or Triassic in age. But apart from a few finds of fishes, not a single vertebrate fossil was discovered in any of these seemingly barren beds. It was thus not impossible that our trip might have been a complete failure, and that the entire four months planned for field work might have ended without our finding a single speci- men, Geologists, however, intent on stratigraphic work, are no- toriously liable to overlook fossil materials for which the eye of the paleontologist is trained. Our obvious plan of campaign was to visit, one after another, ‘‘sample’’ areas in which Permian or Triassic rocks had been mapped, and to hope that, ultimately, one or more of these widespread regions might yield a worthwhile fauna. We found the geologists of the various Argentinian sery- ices mentioned exceedingly generous and helpful in giving us their advice and free access to unpublished maps, manuscripts, and field notes to aid in our laying out a plan of campaign and in discussing matters of stratigraphy and faunal succession. We wish to thank these friends for their help. They include, amongst others: Director General Félix Gonzalez Bonorino, Dr. Roberto V. Tezon, Vocal del Directorio del Instituto, Dr. Juan C. Turner, Dr. Eduardo Holmberg, Dr. Francisco Fidalgo, Dr. Guillermo Furque, Dr. Emilio Gonzalez Diaz, and Sr. Carlos Turcogreco, of the staff of the Instituto Nacional de Geologia y Mineria; Dr. Eduardo Rolleri, Dr. Miguel Flores, Dr. Alberto Mingramm, Dr. Pedro Criado Roque, Dr. Héctor F. de la Mota, of the Yacimien- tos Petroliferos Fiscales; Profs. Sergio Archangelsky, Angel V. Borrello, and Alfredo Cuerda, of the Universidad de La Plata; Prof. Arturo Amos of the Universidad de Buenos Aires; Dr. Pedro N. Stipanicie, director of geological work for the Comisién 4 BREVIORA No. 247 Nacional de Energia Atomica; and Sr. José F. Bonaparte of the Instituto Miguel Lillo de Tucuman. The major part of the funds which made the expedition pos- sible was furnished by grant GB-2454 of the National Science Foundation. The cost of the motor vehicles and a fraction of the travel expense were supplied from private sources. We wish to thank Brigham Young University for allowing Mr. Jensen to ac- company us, and the Museum of Comparative Zoology for mak- ing it possible for Mr. Lewis to take part. Publication of this paper was aided by NSF grant GB-4615. Our collaboration with the Museo de La Plata proved exceed- ingly advantageous. Professor Rosendo Pascual aided us in every way before, during, and after our months in the field, and for a large fraction of the time was an active member of the field party. We are very grateful to Dr. and Senora Pascual for their hospitality in La Plata before and at the close of the field season. Dr. Mario Teruggi, Dean of Natural Sciences and Director of the La Plata Museum, aided greatly in solving our difficulties with the provincial government. Very welcome was the hospitality given us by the University at Samay Huasi, ‘‘the house of re- pose.’’ This had been the country estate of Dr. Joaquin B. Gon- zalez, famous Argentinian author, statesman, and educator, and is located near the city of Chilecito in La Rioja Province. Willed by him to the University at his death, it is now used as a resort for artists, writers, and faculty members. An irrigated desert oasis of some 40 acres or so, with splendid groves of trees, vine- yards, and gardens, it contains a spacious and well-equipped house, with a permanent staff. Most of our work proved to he in La Rioja Province; Chilecito was the best available town to visit periodically for supphes, and we greatly enjoyed on these occa- sions the amenities afforded by Samay Huasi through Sr. and Sra. Amoretti and their staff, and Prof. Bongiorno of the Univer- sity who has charge, in addition to other duties, of a museum at Samay Huasi. Our equipment for the trip was carefully selected by Mr. Lewis, on the basis of his experience on the previous expedition, to the end of attempting to make our work under the somewhat trying conditions of summer in the arid regions of western Ar- gentina as efficient and as comfortable as possible. For rolling stock we took with us a 4-wheel-drive Jeep and a 1% ton stake- bodied truck, likewise with 4-wheel traction, and equipped with a 75-gallon water tank and spare gasoline tanks. Most useful, as it proved, was a ‘‘tote-gote’’ —a motoreyclette of a type recently | | | i) | 1966 CHANARES REPTILES developed in the western United States for traversing rough country and steep grades, and able to carry a modest load of ma- terials in addition to the rider. Without this, our work would have been severely handicapped, for part of our best collecting areas proved to be one or two miles from the closest point we could reach with a jeep; the laborious business of attempting to carry in on our backs many hundreds of pounds of plaster and hundreds of gallons of water, and return with some thousands of pounds of plaster blocks, would have greatly reduced our results. Our party arrived in Argentina the latter part of October, and set out for the field after a short stay in La Plata and Buenos Aires to complete our equipment and obtain all possible data and advice. We then began our planned series of visits to typical areas of sediments mapped as presumed Permian or Triassic in western Argentina. As the discussion above would suggest, our expectations of success at any given locality were none too san- guine, and for the first two months of the trip our gloomiest foreboding appeared to be realized ; during this time we failed to find the slightest scrap of tetrapod bone in any new area. The Santa Clara Formation occupies a considerable area in the precordillera between Mendoza and San Juan. It has sometimes been thought to be Triassic in age, but is believed by Dr. H. J. Harrington (1941, 1956) to be Permian. A few fishes (Bordas, 1944) and a single footprint (Rusconi, 1951a, 1956, and Peabody, 1955) had been described from the Santa Clara, but no tetrapod fossils. We first camped near the puesto of Santa Clara, near the north end of the area covered by the formation, at the east margin of the precordillera in southern San Juan Province, and explored this area. Here the Santa Clara shales and sandstones, some thousands of feet thick, form a series of mountains sepa- rated by deep valleys. We found further remains of fishes in the area of original discovery, and a further fish locality in shales north of the Quebrada de la Montana a short distance east of the boundary between the Santa Clara beds and the Villavicencio group. In addition, a series of trackways of rather poor quality was found and cast; these were situated a short distance up a senda leading south from the Quebrada de la Montana to Puesto Alojamiento. But as far as our sampling of these vast exposures went, they appeared to be entirely barren of tetrapod osseous remains. A second camp was made in the southern area of the Santa Clara Formation, in the canyon of the Rio de las Peas, between 6 BREVIORA No. 247 Cerro Pena and Cerro Palado de las Higueras. Footprints had been previously reported from this area. We learned later that Sr. Bonaparte had made a visit to this area previous to ours. Like him, we found the region to be exceedingly rich in reptilian trackways, in a fashion comparable to areas of the famous Con- necticut Valley Triassic. The trackways, however, are in the main obviously those of dinosaurs or advanced thecodonts, and the beds are presumably middle or late Triassic in age, despite the presence of ‘‘ Chirothertum’’-like footprints. As in the north- ern part of the Santa Clara Formation, we failed to find any bones; not even fish specimens were discoverable. Bodenbender (1911, ete.) had described at length as ‘‘ Paganzo Il’’ a very thick series of red sandstones, presumably Permian in age, from La Rioja Province. These are present in the region of the Sierra de Sanogasta, its southern continuation, the Sierra de Vilgo, and the country south of the mountains, between the village of Paganzo and the railway junction of Patquia to the east. We spent many weary days walking these sandstones from the region of Patquia northward towards Chilecito, and in the region of Los Colorados railway station. We found not the slight- est trace of animal life in these abundant exposures of ‘‘ Paganzo IL’’; not a bone, not even a fish scale or a footprint. Following this discouraging period, we took time out for a brief visit to Tucuman, where we were hospitably received at the Instituto Lillo by Director Willink, and profited by discussion with Sr. Bonaparte. Permo-Triassic beds had been studied in the Sierra de las Qui- jadas in northwestern San Luis Province by Sr. José Roman Guinazu of San Luis, and footprints from this region were de- seribed by Lull (1942). The group spent some time there. The best exposures lay in a deep amphitheatre, difficult of access, west of Puesto Hualtaran. From the nearest approach by car there was necessary a long, tortuous walk, followed by a horse- trail descent of several hundred meters into the base of the amphitheatre ; the return at the end of a day involved the weari- some ascent of the canyon walls. Here, as in other areas pre- viously explored, not a bone was found, although a number of footprints of middle to late Triassic type were seen in the deepest and lowest beds of the area. Permian sediments had been reported in the Sierra de los Llanos of southern La Rioja (Bracaccini, 1946). Here we found massive red conglomerates and grey plant-bearing shales. Nei- ther type of sediment appeared promising as a possible source of —————— 1966 CHANARES REPTILES 7 tetrapod remains, and only a very short time was spent in this area. Having thus prospected for the better part of two months in some five areas with completely negative results, we had become somewhat disheartened. As some slight alleviation, we decided to visit for a short time our old ‘‘stamping ground’’ of Ischigua- lasto, for the purpose of collecting fresh material of rhyncho- saurs, which by agreement with Sr. Bonaparte, I am studying. We had, in our discouragement, half come to believe that perhaps we had lost the ability to see fossil bones even if present. Ischi- gualasto disabused us of that apprehension; a good selection of rhynchosaur specimens was obtained. The Christmas season had now arrived, and we stopped work to spend a few days at Samay Huasi. But despite the beauty and amenities of that place, the holidays were far from joyful for us. Half of our field time had been consumed, and we had thus far met only with utter failure. But a complete reversal of fortune now occurred. Among the places remaining for possible investigation, one which appeared particularly attractive was an area in western La Rioja Provinee, west and south of Chilecito, west of the Sierra de Sanogasta, and about three hours distant from Chilecito by car. Here, at the east margin of the barren Campo de Talampaya, along the rivers Ta- lampaya, Chanhares, and Gualo, geological reconnaissance by Bo- denbender (1911, ete.) and, in recent years, by de la Mota and Fidalgo (unpublished) had shown the presence of a varied series of sediments of Permian and Triassic age. No fossils had been seen, but in view of the considerable range in time occupied, and the probability of a variety of sediments in the succession of formations which appeared to be present, one might not unrea- sonably hope that vertebrates might somewhere be found. This hope was abundantly justified. We first pitched camp in the southern part of the area, near the elevation known as the Mogote del Gualo and near the only spring of any importance in this arid region. On our first day of prospecting, fragmentary ver- tebrate remains were discovered ; the next day several good speci- mens of dicynodonts came to light; from that day on until we stopped work some seven weeks later, not a day passed without the discovery of worthwhile reptilian material. From our first camp we presently moved some miles to the northwest to camp near the point where the Rio Gualo emerges from the hills into the flat Campo de Talampaya. Finding that still richer fossil de- posits were present farther to the north, in the valley of the Rio 8 BREVIORA No. 247 Chanares, we camped for the remainder of our stay on the banks of that stream. In the following article in this series there will be given a resumé of the geology of the area. The region is only a short distance from the Ischigualasto Valley on the opposite, western, side of the Campo de Talampaya, and several of the later Trias- sic formations in the Chanares-Gualo region could be readily identified with the Los Colorados, Ischigualasto, and Los Rastros formations with which we had become familiar on our earlier trip. No fossils were found by us in the limited exposures of the first two; in the carbon-bearing Los Rastros a small amount of fish material was seen, and more could probably have been col- lected had we not found other, more profitable work. In rather lower beds, perhaps of Permian age, which we are terming the Tarjados Formation, fragmentary reptile remains were found in two areas, and not improbably intensive work would result in further finds. Nearly all our finds, however, were derived from a series of volcanic tuffs underlying the Los Rastros beds, which we are terming the Chanares Formation. These in- clude some 75 meters of distinctive blue-white beds, easily trace- able across the region despite the presence of numerous faults. Almost all the fossil remains, we presently discovered, occurred in the lowest 10 meters or so of these beds, and in most areas a careful following of the base of the Chanares resulted in the dis- covery of fossil specimens at frequent intervals. Some of the specimens were found simply buried in the tuffs. In a large pro- portion of cases, however, they were enclosed in grey-brown colored nodules of coarser tuffs. A fraction of such nodules proved to be barren; in some areas, however, almost every nodule would contain a skull or skeleton, and in some cases, two or three individuals would be represented in a single nodule. Never had any of us encountered a fossil-collecting area of equal richness. To add to our pleasure, it was obvious in the field that prepara- tion of the specimens (unlike much of the material collected at Ischigualasto on our earher expedition) could be readily ae- complished. Apart from a large dicynodont, almost all the specimens were forms of small size; the skulls averaged but a few inches in length. More than 150 good skulls were collected and, in addi- tion, large amounts of postcranial materials. At the time this is written, little of the material has been prepared, but apart from 1966 CHANARES REPTILES 9 a relatively small fraction of ‘‘problematica,’’ the general sys- tematic position of most of the specimens was evident when col- lected. There appears to be but a single type of dicynodont, ap- parently of the general Kannemeyeria type. A considerable frac- tion of the finds were those of small thecodonts, some, at least, appearing to be forms on the pattern of Euparkeria of the Cynog- nathus zone of South Africa (recently redescribed by Ewer, 1965). The greater part of the collection, however, consists of small cynodonts; part of them are carnivorous, but the greater number are ‘‘gomphodonts,’’ in which the cheek teeth form a grinding battery. oe THE TRIASSIC FAUNAL SEQUENCE The position of the Chanares fauna in the Triassic sequence will, I trust, become apparent as its components are described in future numbers of this series. Meantime, a brief discussion of our current knowledge of the sequence of Triassic continental verte- brate faunas seems merited. The Triassic is commonly divided into Lower, Middle, and Upper portions stratigraphically — or in chronologic terms, early, middle, and late — in agreement with the three-fold division of the European Triassic (as indicated by the period name) into Bunter, Muschelkalk, and Keuper. On the basis of marine faunas, a series of stages — Scythian, Anisian, Ladinian, Carnian, Norian, and (as a transition to the Jurassic) Rhaetic — have been erected. It is frequently attempted to fit the vertebrate faunas into this sequence of stages. For the most part, however, this cannot be done with any degree of accuracy, because of the rarity of actual continuity of beds containing con- tinental vertebrate faunas with marine strata which can be as- signed a definite place in the stage sequence; most of the stage names are derived from Alpine beds far removed from continen- tal areas. The only major exceptions have to do with the ‘‘Kotriassic’’ Scythian zone. As both Watson (1942) and Lehman and others (1959) have shown, the Cynognathus and Lystrosau- rus zones of South Africa are definitely to be regarded as ‘‘Kotriassic’’ and Seythian in age! mainly on the basis of laby- rinthodonts; the Moenkopi of the early North American Triassic and the Bunter of Europe are equally well assignable to the 1 Huene (1956) has at times tabulated the Cynognathus zone as middle Triassic. but has cited no reason for this. 10 BREVIORA No. 247 Seythian on the basis of their content of labyrinthodonts — laby- rinthodonts of types not too far removed from those of the ‘‘Kotriassic’’ of Greenland and Madagascar. At present it is perhaps best to avoid commitment of specific faunas to specific marine stages in attempting to define a continental vertebrate sequence. In the early years of this century our knowledge of a Triassic vertebrate sequence was very limited. For the early Triassic, we knew little of continental forms (apart from amphibians) ex- cept in the Upper Beaufort (Lystrosaurus and Cynognathus zones) of South Africa. Here, therapsids, most notably a variety of ecynodonts, constituted the bulk of the fauna; archosaurs, in the form of primitive or aberrant thecodonts, were present in modest numbers; other reptiles were small in numbers and va- riety. At the other end of the Triassic sequence, a late reptilian assemblage of very different nature, consisting almost entirely of dinosaurs and advanced thecodonts, was known from a variety of redbeds formations, such as the Keuper of Europe, the Newark, Chinle, ete. of North America, and the redbeds of the Upper Stormberg of South Africa. Between, continental faunas of pos- sible middle Triassic age were practically unknown, for the Euro- pean Muschelkalk is marine, other middle Triassic beds of the northern continents are also marine, and the Molteno of South Africa had yielded no fossils except plants. With the predominance of therapsids in the early Triassic, and of thecodonts and dinosaurs toward the end of the period, it was generally assumed that middle Triassic reptile faunas, if and when discovered, would exhibit a simple transition between these two extremes, with therapsids on the wane and with thecodonts advancing and showing the beginning of the dinosaur radiation. We are now in process, however, of gaining from Africa and South America a seemingly representative series of middle Trias- sic faunas — from the Santa Maria beds of Brasil, the Manda beds of Kast Africa and, most recently, the Ischigualasto beds of Argentina. To some degree these faunas show expected inter- mediate characters, for they contain advanced thecodonts and the beginnings of the dinosaurs. But the faunas show striking posi- tive characters. Therapsids are rather restricted in variety, but show a surprising abundance of ‘‘gomphodont’’ cynodonts. Equally surprising and striking in these faunas is the overwhelm- ing abundance of rhynchosaurs, almost unknown in earlier or later beds. 1966 CHANARES REPTILES af While it is premature, I think, to tie them down to specific marine stages, or to characterize them as ‘‘early,’’ ‘‘middle’’ or ‘‘late’’ in any restrictive way, I believe it is clear that there are three successive stages in the faunal history of continental Trias- sic vertebrates : (A) An early assemblage, of which the faunas of the Lystro- saurus and Cynognathus zones of South Africa, and the Wuh- siang beds of Shansi (China) are typical. There are abundant therapsids, principally cynodonts— mainly carnivorous, but with a few ‘‘gomphodonts’’ with molar-like cheek teeth. Theco- donts include generalized primitive forms such as Euparkeria and aberrant archaic types such as Erythrosuchus and Chasmato- saurus. Primitive rhynchosaurs are present but rare. (B) The type of assemblage found in the Maleri, Santa Maria, and Ischigualasto beds. Carnivorous cynodonts persist, but are ereatly overshadowed by abundant gomphodonts. Typical rhyn- chosaurs constitute a large fraction of the fauna. Thecodonts of advanced type are present, and a few early dinosaurs. (C) The typical assemblages of the late Triassic, such as the Upper Keuper, South African redbeds, Lufeng series of Yunnan, Newark, Chinle, and Dockum of North America, and Los Colo- rados of Argentina. Therapsids are close to extinction. There are various advanced thecodonts, such as aetosaurids and (in the northern continents) phytosaurs, and numerous and varied dino- saurs, but rhynchosaurs are generally extinct. One would, of course, expect assemblages transitional between these stages. It is probable, for example, that the Chanares fauna will prove to be intermediate between (A) and (B), and the same may be true of the Ntawere fauna of Zambia (Brink, 1963 ; Kitch- ing, 1963). Of the faunas which we have listed under (B), that of Ischigualasto is perhaps relatively late, since several forms have already crossed the thecodont-saurischian boundary. The typical (C) faunas are characterized in general by an abundance and variety of dinosaurs, the presence, in the North- ern Hemisphere, of metoposaurs and phytosaurs, and the absence of rhynchosaurs and practical absence of therapsids. Certain beds usually placed in this category appear, however, either to be of earlier age than that of the typical (C) faunas, or else indicate regional late survival of certain of the forms characteristic of the (B) faunas. For example, the Maleri beds of India (Jain, Robin- son, and Chowdhury, 1964) are generally assumed to be of rela- tively late Triassic age because of the presence of metoposaurs and phytosaurs. However, there is, in the restricted fauna 12 BREVIORA No. 247 known, only one possible dinosaur, and rhynchosaurs persist. We may have here a (B)-(C) intermediate. Baird and Take (1959) have recently discovered a reptilian fauna in the Wolfville sandstone of Nova Scotia, generally re- gvarded as a basal member of the Upper Triassic Newark group. But quite in contrast with typical Newark faunas, Dr. Baird in- forms me that not only rhynchosaur remains, but a gomphodont, as well, are present. Quite surely we are here dealing with a fauna transitional between (B) and (C). Rhynchosaurs are present in the Triassic fauna of the region of Elgin, Scotland, currently being restudied by Walker. Be- cause of the close resemblance between Stagonolepis of Elgin and Actosaurus of the German Keuper, Walker (1961) believes the Elgin fauna to be late Triassic in age. There is, however, no rea- son to belheve that the aetosaurid type of thecodont was particu- larly short-lived; it was already present in the Ischigualasto fauna (Casamiquela, 1960). Possibly the Elgin fauna is tran- sitional between (B) and (C); but the absence of dinosaur re- mains, except for Ornithosuchus (Walker, 1964) —a form so primitive that it has generally been regarded as a thecodont — strongly raises the question as to whether the Elgin fauna may not really belong in the (B) category and be of middle Triassic age. Rhynchosaurs are known from several English localities which are classified as ‘‘ Keuper.’’ But it will be noted that there is no English equivalent of the Muschelkalk of the continent ; the early Triassic ‘‘Bunter’’ equivalents are immediately followed by redbeds of ‘‘Keuper’’ type, and it is quite possible that part or all of the English rhynchosaurs are of middle Triassic age. In sum, it appears to be a reasonable working hypothesis that the Triassic included a sequence of three successive continental vertebrate faunas characterized by a dominance of (A) therap- sids, (B) gomphodonts and rhynchosaurs, and (C) dinosaurs. We cannot, of course, expect completely clean-cut distinctions between the three; but a clear pattern seems to be emerging. LITERATURE CITED BAIRD, D., AND W. F.. TAKE 1959. Triassic reptiles from Nova Scotia. Bull. Geol. Soc. Amer., 70: 1565-1566. BODENBENDER, G. 1911. Constitucion geolégica de la parte meridional de La Rioja y regiones limitrofes (Republica Argentina). Bol. Acad. Nae. Ciene. Cérdoba, 19: 1-220. 1966 CHANARES REPTILES 13 Borpas, A. F. 1944. Peces tridsicos de la Quebrada de Santa Clara (Mendoza y San Juan). Physis, 19: 453-460. BRACACCINI, O. 1946. Los estratos de Paganzo y sus niveles plantiferos en la Sierra de Los Llanos (Provincia de La Rioja). Rev. Soe. Geol. Argent., 1: 19-61. BRINK, A. S. 1963. Two eynodonts from the Ntawere Formation in the Luangwa Valley of Northern Rhodesia. Palaeont. Afr., 8: 77-96. CASAMIQUELA, R. M. 1960. Noticia preliminar sobre dos nuevos estagonolepoideos argen- tinos. Ameghiniana, 2: 3-9. Ewer, R. F. 1965. The anatomy of the thecodont reptile Huparkeria capensis Broom. Phil. Trans. Roy. Soc. London, B 248: 379-435. HARRINGTON, H. J. 1941. Investigaciones geolégicas en las Sierras de Villavicencio y Mal Pais, Provincia de Mendoza. Bol. Direccién Minas y Geologia, Rep. Argent., 49: 1-31. 1956. Argentina. In: Handbook of South American Geology. Mem. Geol. Soc. Amer., 65: 131-186. HUENE, F. 1956. Paldiontologie und Phylogenie der niederen Tetrapoden. Jena, 716 pp. JAIN, S. L., P. L. ROBINSON, AND T. K. R. CHOWDHURY 1964. A new vertebrate fauna from the Triassic of the Deccan, India. Quart. Jour. Geol. Soe. London, 120: 115-124. KITCHING, J. W. 1963. The fossil localities and mammal-like reptiles of the Upper Luangwa Valley, Northern Rhodesia. South Afr. Jour. Sei., 59: 259-264. LEHMAN, J.-P., C. CHATHAU, M. LAURAIN, AND M. NaucHE 1959. Paléontologie de Madagascar. XXVIII. Les poissons de la Sakamena moyenne. Ann. Paléont., 45: 177-219. Tot, RS: 1942. Triassic footprints from Argentina. Amer. Jour. Sci., 240: 421- 425. PRABODY, F. E. 1955. Occurrence of Chirotherium in South America. Bull. Geol. Soc. Amer., 66: 239-240. PRICE; Ty: I. 1948. Um anfibio labirintodonte da formaecao Pedra de Fogo, Estado do Maranhao. Ministerio da Agricultura. Bol. Serv. Geol. Min. Brasil, 124: 1-32. 14 BREVIORA No. 247 ReiG, O. A. 1961. Acerea de la posicién sistematica de la familia Rauisuchidae y del género Saurosuchus (Reptilia, Thecodontia). Publ. Mus. Cienc. Nat. Mar del Plata, 1: 73-114. Romer, A. S. 1960. Vertebrate-bearing continental Triassic strata in Mendoza re- gion, Argentina. Bull. Geol. Soc. Amer., 71: 1279-1294. 1962. The fossiliferous Triassic deposits of Ischigualasto, Argentina. Breviora, Mus. Comp. Zool., No. 156, 1-7. RuScOoNnI, C. 195la. Laberintodontes tridsicos y pérmicos de Mendoza. Rev. Mus. Hist. Nat. Mendoza, 5: 33-158. 1951b. Rastros de patas de reptiles pérmicos de Mendoza. Rev. Soe. Hist. Geog. Cuyo, 3: 1-14. 1956. Mares y organismos extinguidos de Mendoza. Rev. Mus. Hist. Nat. Mendoza, 9: 2-88. WALKER, A. D. 1961. Triassic reptiles from the Elgin area; Stagonolepis, Dasygna- thus and their allies. Phil. Trans. Roy. Soc. London, B 244: 103- 204, 1964. Triassic reptiles from the Elgin area: Ornithosuchus and the origin of carnosaurs. Phil. Trans. Roy. Soc. London, B 248: 53- 134. Watson, D. M.S. 1942. On Permian and Triassic tetrapods. Geol. Mag., 79: 81-116. (Received January 6, 1966.) BREVIORA Museum of Comparative Zoology CAMBRIDGE, Mass. 29 Juty, 1966 NUMBER 248 A TRIASSIC AMMONITE FROM THE HINDUBAGH REGION, BALUCHISTAN, WEST PAKISTAN By BERNHARD KUMMEL Ammonites are extremely rare in the thick development of Triassic strata between Quetta and Hindubagh in West Pakistan. Only five specimens, mostly very poorly preserved, have been re- ported to date. Because of their rarity the documentation of one additional specimen of rather good preservation is warranted. The first record of Triassic ammonites from this region of Baluchistan was by Mojsisovies (1896, p. 611; 1899, p. 44) who described a single specimen, collected by C. L. Griesbach from a loose block, 7 miles south of Hindubagh. This specimen was described as Didymites afghanicus Mojsisovies (1896; 1899, pl. 20, fig. 9) and assigned a Norian age. Vredenburg (1904), on the basis of surveys carried out in the summer of 1901, con- tributed some regional, stratigraphic, and paleontological data on these Triassic formations. This author reported the presence of Monotis cf. salinaria in fair abundance at a few horizons and localities and illustrated one specimen. In addition, he illustrated an ammonite-Halorites sp. The specimen was not found in place but from local conditions Vredenburg concluded that it had not been transported very far. These specimens plus others collected by Vredenburg were submitted to C. Diener who published his results in 1906. In this small collection Diener (1906) was able to recognize the following forms: Monotis salinaria Schlotheim Halorites sp. ind. atf. swbcatenato Mojsisovies Celtites sp. ind. (group of ? acuteplicati) Paratibetites sp. ind. ex aff. tornquisti Mojsisovies Dittmarites or Distichites ? sp. ind. Rhacophyllites vredenburgi Diener 2 BREVIORA No. 248 The generally poor state of preservation of the specimens avail- able to Diener is well reflected in the uncertainty in the identi- fications. Of these specimens only Monotis salinaria and Halorites sp. ind. aff. subcatenato were obtained from the spot where they probably weathered out. All the rest of the specimens were picked up amongst transported boulders from the beds of rivers within the Triassic outcrop area. On the basis of this scanty data, both Vredenburg and Diener concluded that the strata en- closing these fossils are of Norian age. Additional geological observations on the Triassic formations of the Quetta-Hindubagh area have been made by Gee (in Heron, 1939, p. 26-27), by Williams (1959), and by the Colombo Plan Project (1961), but in these reports there is little additional paleontological data. I have had the opportunity to spend a few days in the region between Quetta and Hindubagh examining these Triassic forma- tions but was not successful in finding any fossils. Mr. S. A. Bilgrami entrusted to the writer a specimen found in a stream bed near a mine of the Pakistan Chrome Ltd., two miles east of Gwal Railway Station. This specimen is fairly well preserved and can be assigned with confidence to Arietoceltites arietitoides Diener, a species first described from the Tropites limestone ex- posed at Kalapani, Byans, in the Himalayas. This new specimen thus provides an additional link in the correlation of these Trias- sic formations of Baluchistan with the much better documented Himalayan sequence. The Tropites limestone of Byans_ has yielded 168 species of invertebrates of which 155 are ammonites. The fauna as a whole shows very strene affinities to the Upper Triassic faunas of the Mediterranean region. All of the fossil species were collected from a three foot bed of limestone; of special interest is the mixture in this bed of typical Carnian and Norian ammonite species. Sinee Arietoceltites arietitoides was a unique species previously known only in Byans, and its presence in a mixed fauna at that, it does not help in refining the age determination of the Baluchistan Triassic formations. Diener’s (1912, p. 150) final words on the age of these Triassic rocks were that they were Norian in age. Oo 1966 TRIASSIC AMMONITE SYSTEMATIC DESCRIPTION Family TROPICELTITIDAE Spath, 1951 Genus ARIETOCELTITES Diener, 1916 ARIETOCELTITES ARIETITOIDES (Diener) Figures la, b Tropiceltites arietitoides Diener, 1906, p. 156, pl. 3, fig. 12; Diener, 1912, p. 125; Diener, 1916, p. 101; Spath, 1951, p. 94; Kummel, in Arkell, 1957, p. L171, fig. 201, 4. This species is represented by a single specimen of 56.5 mm in diameter, 13.1 mm for the height of the adoral whorl, 13.8 mm for the width, and 35 mm for the diameter of the umbilicus. In degree of involution and general proportion, this specimen is very much like the type specimen. There are shght differences in ornamentation: in the Baluchistan specimen the ribs are rusiradiate in the adoral 1.5 volutions, whereas in the Byans specimen the ribs are more radial. In both specimens the ribs turn strongly forward on approaching the marginal furrows aligning the median keel. I do not believe that these differences in ornamentation are of any taxonomic significance. The Byans material available to Diener (1906) consisted of the type speci- men plus a few fragments. With this one additional specimen there is insufficient data available to evaluate the range of vari- ability in these features. However, on the basis of experience with populations of comparable forms, one should expect a range of variability that would include that expressed by these two specimens. The suture was only partially exposed on Diener’s type specimen and is not preserved in the specimen recorded here. Occurrence. — From stream bed near mine of Pakistan Chrome Mines Ltd., two miles east of railway station at Gwal, presumably from Alozai group, between Quetta and Hindubagh, West Pak- istan. Repository. — Geological Survey of Pakistan, Quetta, West Pakistan. Acknowledgements. —I wish to thank Mr. 8. A. Bilgrami for the loan of the specimen and for his hospitality during a visit to Hindubagh. My visit to Pakistan was made possible by the aid of National Science Foundation Grant No. G19066. 4 BREVIORA No. 248 REFERENCES ARKELL, W. J. ef al. 1957. ‘Treatise on invertebrate paleontology, Part L, Molusea 4, Ceph- alopoda, Ammonoidea. Geol. Soc. Am. and Uniy. Kansas Press, 490 pp. COLOMBO PLAN PROJECT 1961. Reconnaissance geology of part of West Pakistan. Toronto, 590 pp. DIENER, CARL 1906. Himalayan fossils. Fauna of the Tropites Limestone of Byans. India Geol. Surv., Palaeont. Indica, (15) 1: 1-201, pls. 1-17. 1906. Notes on an Upper Triassic fauna from the Pishin District, Baluchistan, collected by E. Vredenburg in the year 1901. Ree. Geol. Sury. India, 34: 12-21, pls. 3, 4. 1912. The Trias of the Himalayas. Mem. Geol. Surv. India, 36: 1-159. 1916. Kinige Bemerkungen zur Nomenklatur der Trias-cephalopoden. Centralbl. Mineral. Geol. Palaiont., No. 5: 97-105, Herron, A. M. 1939. General report of the geological survey of India for the year 1938. Ree. Geol. Surv. India, 74: 1-132. Mossisovics, E. v. 1896. Beitrige zur Kenntniss der obertriadischen Cephalopoden- Faunen des Himalayas. Denkschr. Akad. Wiss. Wien, 63: 575- 701, pls. 1-22. 1899. Upper Triassic Cephalopoda faunae of the Himalaya. India Geol. Surv., Palaeont. Indica, (15) 3: 1-157, pls. 1-22. SparH, L. F. 1951. Catalogue of the fossil Cephalopoda in the British Museum (Natural History). Part 5, The Ammonoidea of the Trias. London, 228 pp. VREDENBURG, E. 1904. On the oceurrence of a species of Halorites in the Trias of Baluchistan. Ree. Geol. Sury. India, 31: 162-166, pl. 17. WILLIAMS, M. DEAN 1959. Stratigraphy of the Lower Indus Basin, West Pakistan. Fifth World Petroleum Congress, Sec. 1, Paper 19: 1-15. 1966 TRIASSIC AMMONITE 5 Figure 1. Side and ventral view of Arieloceltites arietitoides (Diener) from Triassic formation near Hindubagh, West Pakistan. * 1. BREVIORA Museum of Comparative Zoology CAMBRIDGE, Mass. 29 JuLy, 1966 NuMBER 249 ADDITIONAL NOTES ON THE AMPHISBAENIDS OF GREATER PUERTO RICO By RicHarp THomas! The overall faunal similarity of Puerto Rico and the Virgin Islands has led to the use of the term ‘‘Greater Puerto Rico’’ for their combined areas (Schmidt, 1928, used the term for a sup- posed continuous land mass of the past). The amphibaenids of this region have been dealt with thoroughly by Gans and Alex- ander (1962), in their comprehensive work on the Antillean members of the family. More recently, Gans (1964) has de- scribed a new species, Amphisbaena schmidti, from Puerto Rico. Thorough as it has been, the work of these authors has been hampered by a dependence upon specimens with insufficient or doubtful locality data. Herpetological and ornithological collections made under the sponsorship of Dr. Albert Schwartz in the Virgin Islands (sum- mer, 1964) and Puerto Rico (January-Mareh and July 1965, and previous visits) have added new data to the knowledge of the distribution, taxonomy and ecology of the Greater Puerto Rican amphisbaenids with many new specimens and one new species. PHYSIOGRAPHIC FEATURES OF PUERTO RICO Before proceeding with the discussion of the species, a brief outline of the major physiographic features of the island of Puerto Rico is pertinent. Puerto Rico may be broken into three general physiographic regions, each of which is, in a broad sense, ecologically uniform. 1. The northern limestone region extends west from roughly the city of San Juan to the western extremity of the island at Aguadilla. This region at its extreme width is one-third that of the island. Most of it is oceupied by an extremely eroded to- pography of close set, conical mogotes (to above 1500 feet), often called the Haystack (or Pepino) Hills; but some areas of it are 110,000 SW 84th St., Miami, Florida 33143. No. 249 BREVIORA ‘SOJLONTY Op BLBD S9}PBoIpUL MOAIL OY, “SOUT, [BJUOZTAIOY OU Aq AprMUYyos ‘yp jo oSuvt oy} pue ‘sour, [vuoseIp ouy Aq Diaw “Pp JO IsuvI oY} + SOUT] [vuoseIp os1v09 AQ Po}BVOIPUL ST Sp1oded yuasaid pue snotaoid uo poseq tuaynq ‘py Jo OSuBI UMOUY OY, *(zonsvAvy 1oJ oatpdumsoad ATUO) Spsodar otapedurss quasordad spoqurdS J[VY OM} YYIM SopoITQ ‘*(SopodITo MOT[OY) wWprWwYyos *P puB ‘(SasSO1o YIM Sopa) Daw *p ‘(SJOp YIM Sept) rwayng *P ‘(SepPalo pYOS) vIap9 "Pw AOF SpLOdat APTROT MoU SuLMOYS OoTY OJON FO dey ‘LT ‘sty KAS ~ 7 1966 AMPHISBAENIDS OF PUERTO RICO 3 rolling or flat, especially along the coastal margin. The ter- mination of this region to the south is usually abrupt and fre- quently evident as an escarpment. This limestone region varies from xeric to mesic; parts are moist enough for the growing of coffee. 2. The central highlands, primarily igneous in lithology, com- prise the main mass of the island and are composed of a very complexly subdivided system of mountains in which few major units or trends are evident. Heights range to over 4000 feet in places, but generally the elevations are moderate (1000-2000 feet). Except for the eastern and western coastal margins it is predominantly mesic and at one time was doubtless almost com- pletely forested. 3. The southern coastal plain is directly continuous with the central highlands but is in contrast ecologically in its aridity. Outstanding faunistically in this region is the area to the south of the Valle de Lajas in the extreme southwest; it is comprised in general of a discontinuous range of low coastal hills, which at one time probably constituted a chain of offshore keys or a penin- sula joining the main island in the vicinity of Guayanilla. This southern coastal region, particularly the portion just mentioned, is probably the strongest area of endemism in Puerto Rico, where such forms as Phyllodactylus wirshingi, Sphaerodactylus roosevelti, Ameiva wetmorei, Anolis poncensis, Anolis cooki (= A. cristatellus cooki Grant), and the very distinctive Typhlops granti have their primary distributions. SPECIES DISCUSSION AMPHISBAENA BAKERI Stejneger Specimens examined: new localities: PUERTO RICO: 4.2 km S Mora, 500’ (150 m), ASFS V5335; 7 km W Maricao, ca. 300 meters, ASFS V6424. Previously recorded localities: ‘‘Las Marias,’’ MCZ 66512 (= UPR 12); ‘‘Maricao,’’ MCZ 66513 (= UPR 22); °‘Mayagiiez,’’ UMMZ 55824. Amphisbaena bakeri is known only from a few specimens (12 including the new records cited herein) from a cluster of locali- ties in the north-central portion of extreme western Puerto Rico (Fig. 1). The previous localities for this species were all merely town names. From some of the localities were also recorded specimens of caeca, but due to the imprecise nature of the locali- ties, it could not be determined whether the two forms were truly sympatric. A. caeca was unknown from the northwestern 4 BREVIORA No. 249 corner of the island, and the only localities from which it was known ‘‘with’’ bakeri were those which were possibly peripheral to the range of bakeri (ef. Gans and Alexander, 1962, fig. 4). A. bakeri differs from the wide-ranging A. caeca principally in being larger in size, in having a higher number of body annuli, longer internasal suture, absence of a third row of chin segments, and lack of caudal autotomy (Gans and Alexander, 1962). Only the second (although possibly the third) has been recognized as an absolute difference from caeca. The specific dis- tinctness of bakeri has therefore not been established with cer- tainty. One of the recent specimens (ASF'S V5335) represents only a minor range extension but constitutes the first record from the northern limestone region; ASFS V6424 is from within the pre- viously known range. There is no evident explanation for the small, very restricted, apparent range of bakert. The fact that a number of specimens of caeca (and no bakerv) were collected in the vicinity of Utuado indicates that bakert probably does not occur much farther east than it has been previously reported. The coloration in life of these two specimens of A. bakeri was noted as being ‘‘pale pinkish gray’’; ASFS V6424 was noted as being ‘‘ pink below’’ (ventrally). The right hemipenis of ASFS V6424 is everted and appears to be very similar to the hemipenis of A. caeca (Fig. 2). \ A B Fig 2. Hemipenes of A, Amphisbaena schmidti (ASFS V5871), and B, Amphisbaena caeca (ASFS X4111); view cephalad with organs normal to body axis. ~ 1966 AMPHISBAENIDS OF PUERTO RICO 5 ASFS V5335 from the northern limestone region was col- lected in a small grassy pasture, which formed the bottom of a doline in a mesic portion of the Haystack Hills. It was found beneath a rock where it was seen disappearing into a vertical burrow in moist clayey soil from which it had to be dug. ASES V6424 was taken in a typical coffee producing area of the cen- tral highlands. It was found in a rotten log along the edge of a clearing overgrown with short, weedy vegetation. The nature of the ecological preferences of bakeri cannot, of course, be judged from just two encounters. The locality for the first specimen (4.2 km S Mora) is clearly within the range of caeca, the nearest locality for that species being less than three kilo- meters to the south. A. schmidti also occurs in this part of the limestone hills; the nearest locality for that species was 0.2 kilometers to the north of the locality for the first specimen of bakeri. Thus, in this area, the three known Puerto Rican species occur sympatrically, whatever their ecological preferences (see below for caeca and schmidt). AMPHISBAENA CAECA Cuvier Specimens examined: new localities: PUERTO RICO: 13.3 km E Utuado, ASFS V4443-48; 17.7 km NE Utuado (= approx. 8 km straight line distance), 1100’ (835 m), ASFS V4482, V4499, V4500-01, V4623, V4449, V5860; 8 mi. (13.5 km) NW Utuado, 1500’ (407 m), ASFS V4548; 9 mi. (15 km) E Lares, ASFS V4656; 7.0 km S Mora, 800’ (244 m), ASFS V5152, V5329, V0332; 5 mi. (8 km) NE Lares, 1200’ (366 m), ASFS V5267; Ramey Air Force Base, Rifle Range Beach, ASFS V5430-31 ; 2.3 km E Juana Diaz, 200’ (61 m), ASFS V5755; 3 km NE San German, ASFS V6407; 4.6 km W, thence 4.6 km NW Juana Diaz, ASFS V6620-28, V6631-35; 3.6 km W, thence 9.7 km N Juana Diaz, ASFS V6658-59; 3.6 km W, thence 8 km N Juana Diaz, ASFS V6660-61; 10.2 km E Dos Bocas, ASFS V5861-62; 18.8 km SW Arecibo, 800’ (244 m), ASFS V5908- 09; 112 km NW Utuado, 1300’ (400 m), ASFS V5922; 1 mi. (1.6 km) NW Dorado, ASFS V5939; 4.1 km NE Villa Pérez, 2200’ (670 m), ASFS V5973; 2 km NE Barranquitas, 2100’ (640 m), ASFS V6027; 1.8 km S Adjuntas, ASFS V6207; 12.3 km SE Patillas, ASFS V4798; 10 km SE Patillas, ASFS V4802; 12.9 km SW Fajardo, 800’ (244 m), ASFS V5081. VIEQUES: Cayo de Afuera, ASFS V6168. Previously recorded localities: PUERTO RICO: Isla Verde, ASFS X937-43, X4104- 6 BREVIORA No. 249 25, X7381-98; 2.2 mi. SW Sabana, ASFS X7483-34; Mayagiiez, UPR 84e, 87, 88, 90, 92, 140. Large numbers of A. caeca were examined by Gans and Alex- ander, but they had large series from the eastern part of the island only and relatively few specimens from the west. The species was not then known from the northwestern corner, an area bounded roughly by Mayagtiez, Lares and Arecibo. The present collection also has an eastern bias due to the large series from Isla Verde, but a more thorough coverage has been made in the west, including the formerly blank northwest portion. With the new material I have not undertaken a detailed analysis of trends such as that done by Gans and Alexander (fig. 29). However, I have compared trends in number of body annuli of selected eastern (Isla Verde and east), central (Juana Diaz region), west central (Utuado-Adjuntas region), and west- ern (Ramey Air Force Base to south of Mora) ‘‘samples’’ (these do not include all specimens examined) with those shown by Gans and Alexander for this character (in which the strong- est trends were evident). An agreement in direction of trends was found. Both pre- and postcloacals average fewer in eastern specimens; this is also in agreement with the observations of Gans and Alexander. Additionally, internasal suture length (expressed as a fraction of interprefrontal suture length) aver- ages longer in the eastern specimens. The largest specimen (ASFS V6207) is a male that measures 291 mm total length (tail 21 mm), which appears to be the largest caeca recorded (the largest seen by Gans and Alexander was 269 mm total length). A body annulus count of 214 (ASFS V4120) is lower than the low value of 217 recorded by Gans and Alexander. The Vieques specimen is, in a way, the first record for that island. The specimen was not taken on the main island but on an offshore key, Cayo de Afuera. However, the species may be assumed to occur on Vieques proper. It has long been a matter of speculation as to which species of Amphisbaena would be found on Vieques and Culebra, caeca or fenestrata. Some have tacitly assumed the former. Of course, the finding of one spe- cies does not preclude the possibility that the other might also occur on the island. The Cayo de Afuera specimen has a high number of ventral midbody segments; it is also near the upper extreme of caeca in body annuli. 1966 AMPHISBAENIDS OF PUERTO RICO 7 A. caeca seems to prefer a mesic, usually shady, habitat (see below for comparisons with other species) ; it occurs from mon- tane cafetales at elevations of at least 2200 feet (ASFS V5973, apparently the highest elevation record for caeca) to the coast. It still appears to be absent from the extremely arid southern regions; the record of Fowler (1918) for Guanica probably does not pertain to this form (vide infra). One specimen (ASFS V4798) was collected approximately 40 feet from the ocean’s edge and shed skins were found in the vicinity. Three specimens were collected in rotting wood, two in logs that were not even in firm contact with the ground. The most xeric habitat in which I have found caeca was on Cayo de Afuera; even there it was in a wooded and somewhat shady, although not moist, situation. AMPHISBAENA XERA new species Holotype: MCZ 81019, an adult male from 7 km E of Guanica, Puerto Rico, elevation 600’ (183 m), one of a series collected on 6 Mareh 1965 by Albert Schwartz and Richard Thomas. Paratypes: PUERTO RICO: UIMNH 56910, ASFS V5646, USNM 152588, 7.3 km E Guanieca, 2 March 1965, R. Thomas; ASFS V5659, 7.1 km E Guanica, 3 March 1965, R. Thomas; ASFS V5662, 7.3 km E Guanica, 3 March 1965, R. Thomas; ASFS V5722-23, AMNH 94170, same data as type (foregoing localities are at the same elevation as that of the type) ; UMMZ 73844, Sabana Grande; ASFS V5800, 6.2 km E Juana Diaz, 10 March 1965, R. Thomas; ASFS V6408, 5 mi. (8 km) ESE San German, south slope Cerro Algarrobo, 9 July 1965, R. Thomas; ASFS V6148, 5 mi. (8 km) ESE San German, south slope Cerro Algarrobo, 10 July 1965, R. Thomas; ASFS V6423, 1.5 km WNW San German, 11 July 1965, R. Thomas; ASFS V6427, 3 km NE San German, 12 July 1965, R. Thomas; ASFS V6587, 2.3 km E Juana Diaz, 26 July 1965, R. Thomas; ASFS V6616-18, 4.6 km W, thence 1.1 km NW Juana Diaz, 27 July 1965, R. Thomas; ASFS V6619, 4.6 km W, thence 4.6 km NW Juana Diaz, 27 July 1965, R. Thomas; ASFS V6636-40, 28 July 1965, R. Thomas; ASFS V6646-57, 4.6 km W, 1.9 km NW Juana Diaz, 28 July 1965, R. Thomas; CM 40577-78, RT 1333-34, 4.6 km W, thence 1.9 km NW Juana Diaz, 29 July 1965, R. Thomas ; UMMZ 73846, 3 mi. (4.8 km) E Juana Diaz, 27 Novem- ber 1931, Chapman Grant; MCZ 36301, 10 mi. (16 km) E Juana Diaz, 27 November 1931, C. Grant; UPR 84a-b, Mayagtiez, 10 8 BREVIORA No. 249 December 1957, J. A. Ramos; UPR 86, Mayagiiez, 8 September 1949, Nestor Nazario; UPR 91, no data. Diagnosis: A form of Amphisbaena lacking fusion of head seales, of closest affinities to the species bakeri and caeca, differ- ing from the former in having a low number of body annuli (225-234), the presence of a third row of chin segments (most specimens), and much smaller size. From caeca it differs in pos- sessing a relatively longer internasal suture (32 to 53 per cent of the length of the interprefrontal suture), lack of a postmalar row of chin segments, fewer average midbody segments, caudal autotomy absent or ill-defined, and smaller size. Fig. 3. Dorsal, lateral and ventral views of the head of the type of A. wera (MCZ 81019). Line represents 1 mm to scale. Description of type (Fig. 3): Meristic and mensural data for type and paratypes are found in Table 1. The type is a mature adult male with the hemipenes partially everted. The first two body annuli correspond to three dorsal half-annuli; the median 1966 AMPHISBAENIDS OF PUERTO RICO 9 segments of the second dorsal half-annulus are enlarged to form a pair of parietals in broad contact at the midline; the median segments of the third dorsal half-annulus are not enlarged to form a second pair of parietals. In the chin region two of the three segments of the first postgenial row are fused with two in the second row so that the second row is represented by only one segment (Fig. 3) ; the mental is partly fused with the genial. The postmalar! row contains four scales and is bounded laterally by the malars which abut on the first body annulus. The dorsal groove is weakly indicated by an alignment and some medial rounding of the posterior paravertebral segments. Variation: In size, only five specimens exceed the type; 42 (90 per cent) are smaller. The head scalation described for the type is characteristic; a second pair of parietals is indicated in only four specimens. The internasal sutures range from 32 to 53 per cent of the length of the interprefrontal sutures. In the chin region the most variability is encountered in the second row of postgenials. Its scales are often variously fused with the first and third rows, or it may be represented only by one or two small and irregular scales crowded between the first and second rows. Eighteen specimens have only two postgenial rows and thereby have the chin configuration of bakert. It is as if the second postgenial row were being lost. The lack of a postmalar row is characteristic; one specimen (ASFS V6637) appears to have this row but the malar is abnormally divided. Autotomy con- strictions are evident in a few specimens at the level of the fifth to seventh caudal annulus; slightly heavier pigmentation of a caudal annulus may mark the autotomy level in others. Other specimens either definitely do not have an autotomy constric- tion or the presence of one is indeterminate. Cloacal pores are four, with the exception of one specimen which lacks two on one side. Hemipenes are very similar to those of bakeri and caeca (hie, 2). Coloration: The dorsal coloration in life of this form is typi- cally a pale tan, very reminiscent of the two living specimens of bakerit I have seen. The coloration lightens gradually towards the ventral surface and pigmentation is almost absent on the four ventralmost rows. Dropping out of pigmented ventral seg- ments takes place in some specimens but is not prominent be- cause of the fading out of all pigmentation ventrally. Dorsally 11 have objected elsewhere (Thomas, 1965) to the use of the term postmalar ; however, this was due to my misunderstanding of the terminology as it applied to Amphisbaena innocens. 10 BREVIORA No. 249 and ventrally the coloration is darker (browner) anteriorly. In life the venter is pink due to the translucence resulting from lack of pigmentation. The temporal region is deep reddish purple, probably another product of superficial vascularity and pigmentation. The snout (including prefrontals) is ight (buffy) as is the posterior half of the edge of the lower jaw. Comparisons and discussion: The most pertinent comparisons of wera are with caeca and bakeri (Fig. 4). Seven specimens, having no postmalar row of chin segments and regarded by Gans and Alexander as A. caeca, are here regarded as A. xera. Aside from one specimen I have not seen (SU 7775, examined by Gans and Alexander) which may also be referable to zera, and a possible exception noted above, the absence of a postmalar 1.5 Oh A Lo © @® ® © ®O®O®OO ® ©OO © OOF OOo © © ©990°999 $ © O5 OOOO OO & OOOO 29 O99OD ro } TIVSVNYSLNI 2.0 2.5 3.0 p 4.0 MM MM 3.5 INTERPREFRONTAL Fig. 4. Seatter diagram of internasal versus interprefrontal suture lengths for A. caeca (stippled squares), A. bakeri (solid triangles), and A, xera (hollow circles). Numbers in the wera symbols indicate the number of specimens possessing that value. Values were not plotted for all speci- mens of caeca examined, only enough to show variation including extremes. The symbols at the extreme right for vera and caeca represent the values for the largest specimen of each of these two forms. 1966 AMPHISBAENIDS OF PUERTO RICO 11 row distinguishes zera from caeca in all cases. Internasal suture length varies from 32 to 53 per cent of the interprefrontal suture in xera; in caeca the longest interprefrontal sutures (29 per cent of the interprefrontal length) are found in four specimens from eastern Puerto Rico, where the internasal suture length averages higher than in material from the rest of the island (and interest- ingly an area where sympatry with the two other long internasal suture forms is absent). A. xera is characterized by a low num- ber of segments around midbody; one specimen (from Maya- gtiez) has 32 midbody segments, otherwise the range is 26 to 30; a count of 30 or below is found in less than 20 of over 300 speci- mens of caeca examined by Gans and Alexander and by me. In the low average number of cloacal scales xera is more like eastern caeca than like those from other parts of the island. In coloration xera 1s paler than most specimens of caeca which are typically brown or gray-brown rather than pale (or even reddish) tan (although this color does occur in caeca). Specimens of xera are usually identifiable on first sight by their paler color and more slender proportions (caeca is characteristically stouter, doubtless a correlative of the greater number of midbody seg- ments). The tails are tapered towards the tip and are not so cylindrical and rounded terminally as is typical of caeca; this is a distinguishing character in living specimens. The pale snout of zera is also not found in caeca (or is not so extensive). A. caeca and A. xera are allopatric for the most part; they have been taken together at four localities (Mayagiiez, 3 km NE San German, 4.6 km W, thence 4.6 km NW Juana Diaz, 2.3 km E Juana Diaz), all peripheral to the range of xera and appar- ently to that of caeca. At these localities there is no evidence of intergradation. Relatively short internasal sutures are found in the four Mayagiiez xera and in the one from Sabana Grande (30 to 33 per cent of the length of the interprefrontal suture), but these do not overlap with internasal suture lengths for caeca from any part of the island. One Mayagiiez specimen (UPR 84a) has 32 segments at midbody (16/16), the only xera with counts higher than 30. A. rera is strikingly like bakeri in several respects (internasal suture, lack of caudal autotomy, lack of a postmalar row, pale coloration), although in others it is significantly different (fewer body annuli, size, gradual fading of color on venter). In size, 9 of the 12 known specimens of bakeri are larger than 42 (90 per cent) of the specimens of xera. The configuration of the chin segments may well be thought of as intermediate between the 2 BREVIORA No. 249 conditions in bakert and caeca but tending more towards the former. The 18 xera with but two postgenial rows are even more suggestive of a trend towards bakerv. Schmidt (1928) commented on three specimens from Maya- giiez that appeared to be intermediate between bakeri and caeca. Two of these (noted as having an internasal length of about one- half the interprefrontals) almost certainly pertained to A. xera (these specimens do not appear to be extant). The other speci- men (CNHM 12473) was identified by Gans and Alexander as A. bakeri. Two specimens of bakert are known from Mayagiiez ; one (CNHM 12473) has a body annulus count of 240, which is near the extreme low for that form; another has a count of 255 which is at the upper extreme. A specimen (UPR 7, Gans and Alexander), which has no locality data other than Puerto Rico, is probably from the Mayaguez region; it has a body annulus count of 239. It should also be recalled that three of the Maya- giiez (and the one from Sabana Grande) specimens are at the upper extreme in size for xera. More specimens with precise data from the Mayaguez vicinity may show that bakerv and xera are Subspecifically related. I am reluctant to so allocate these forms now because bakeri is so little known. A. schmidti is distinguished from zxera by its lower number of body annuli (192-207 versus 225-234), higher number of cau- dal annuli (18-22 versus 10-14), the presence of caudal autot- omy, larger size, different proportions, and overall darker color- ation. The two are additionally distinguished by differences in relative size of some of the head scales and by different chin segment configurations. Range: Known from the central to western parts of the south- ern xeric coastal plain of Puerto Rico (Fig. 1) and as far north as Mayagiiez, San German, Sabana Grande, and to the northwest of Juana Diaz. In all probability it occurs on Caja de Muertos,? whence there are shed skins yielding midbody segment counts of 14/16 (1) or 14/14 (2) and eaudal annuli of 12 (2), 13 (1) or 14 (1). The specimen recorded by Fowler (1918) from Guanica may well have been of this new form. Habitat: The topotypical paratypes and type (7.0-7.3 km E Guanica) were collected in the xeric woods which form the more luxuriant covering of the limestone hills of this region. 1Heatwole, Torres and Heatwole (Stahlia, 1965, no. 4, p. 1) recorded taking a specimen of Amphisbaena on Caja de Muertos. I have not examined this speci- men, but almost certainly it pertains to A. wera, as it was noted to lack the post- malar row of chin segments and also agreed with vera in other counts given. te 1966 AMPHISBAENIDS OF PUERTO RICO 13 The soil varied from sandy with moderate organic (plant) con- tent to loamy; there was a uniform cover of leaf litter. At the type locality, A. xera occurs syntopically with Typhlops richardi and Typhlops grantt. The specimen from 6.2 km E Juana Diaz was collected in a region of xeric foothills; it was found at night under a rock which was not at all imbedded in the soil. The area was exposed and had very sparse vegetation. In the heat of the day it was doubtless a most inhospitable place, and the chance of finding an Amphisbaena near the surface would have been negligible. At the following three localities I was able to personally col- lect and observe caeca and xera together. 3 km NE San German. This loeality is in the foothills to the north of the San German valley; it is transitional between the arid region to the south and the mesic highlands, but closer to the latter. The single specimen of caeca was collected on a boulder-strewn and somewhat wooded hillside surrounded by pasture; the specimen of xera was collected near a small stream at the base of the hillside. No ecological differences were dis- cernible. 4.6 km W, thence 4.6 km NW Juana Diaz. This locality is in the valley of the Rio Inabén; it too is transitional between the xeric south and the mesie interior. On the first visit, nine speci- mens of caeca and one of sera were found in a patch of sandy, mesic, river-border woods. The surrounding regions (where nat- ural growth was evident) were covered by semi-arid serubby woods; the spiny palm (Acrocomia), a typically semi-xeric form, was seen on some of the nearby hillsides. On the second visit an attempt was made to collect in a more open, scrubby, overgrown field nearby. Five specimens of xera and five of caeca were ob- tained. Not quite three kilometers to the south, another area of shady (but distinctly more xeric) riverine woods was visited. Eighteen specimens of xera (two of which have not been exam- ined for this paper) were collected and no specimens of caeca. Less than a kilometer more to the south in still arid country, three more specimens of xera but no caeca were taken. 2.3 km E of Juana Diaz. This locality is a mesic, wooded ra- vine surrounded by typically xeric to semi-xeric regions. One specimen each of both species was found in the ravine. 14 BREVIORA No. 249 AMPHISBAENA SCHMIDTI Gans Specimens examined: new localities: PUERTO RICO: 5.5 mi. (9 km) N Lares, 1100’ (346 m), ASFS V5197-99, V5258-60; 5 mi. (8 km) NE Lares, 1200’ (866 m), ASFS V5268; 4 km § Mora, 500’ (153 m), ASFS V5334; Ramey Air Force Base, Rifle Range Beach, ASFS V5432-37; 3 mi. (4.8 km) W Las Llanadas, 600’ (183 m), ASFS V5871; 18.8 km SW Arecibo, 800’ (244 m), ASFS V5910; 11.2 km NW Utuado, 13800’ (400 m), ASFS 75923-24; 1 mi. (1.6 km) NW Dorado, ASFS V5940; 3.5 mi. (5.4 km) W Playa de Vega Baja, ASFS V5959; 2.5 km SW Florida, ASFS V5999; 5.6 km NE Morovis, 400’ (122 m), ASFS V6028-29. Amphisbaena schmidti was described (Gans, 1964) on the basis of six specimens from four more or less coastal localities, three in the northwest section of the island and one (‘‘Salinas’’) presumably on the south coast. Professor Manuel J. Vélez, col- lector of the type and one paratype of schmidti, has informed me (in litt.) that the type locality of A. schmidti should cor- rectly be ‘‘Isabela, orilla (cuneta) Carretera Cano.’’ The lo- cality is in the district of Isabela about six kilometers southeast of the town of Isabela by the side of Carretera Cano (= road number 113). The distribution of schmidti was given as ‘‘coastal Puerto Rico.’’ The new localities cited here are restricted to the north- ern limestone section (up to elevations of 1300 feet), and the species is in all likelihood endemic to this region (Fig. 1). No specimens were found elsewhere on the island. Of the numbers of amphisbaenids collected in the southern part of Puerto Rico (A. caeca and A. xera) and of the shed skins that were often found there, none pertained to A. schmidti. The apparent absence of schmidti from the southern mar- ginal region plus its absence from the coastal regions east of San Juan (in spite of the large series of caeca from there) indi- cate that its distribution is not coastal Puerto Rico and, at the same time, throw considerable doubt on the Salinas record. Con- ceivably it refers to another Salinas (a not uncommon coastal name in Latin America), possibly Punta Salinas on the north coast to the west of San Juan, which is probably in the range of schmidti. The new specimens of schmidti agreed in coloration (in life) very closely with the coloration recorded by Gans; they were usually noted as being purplish brown dorsally with darker 1966 AMPHISBAENIDS OF PUERTO RICO 15 brown heads and tails. The venters are heavily pigmented but sometimes pinkish. In general proportions and appearance this species is very distinct; the blunted and almost triangular head is characteristic and is accentuated by the deeply constricted neck. Indeed, this species, in proportions, color and scutellation, is possibly the most ‘‘different’’ of the species of the Greater Puerto Rican region. The enlarged parietals and enlarged scales at the end of the third row of chin segments are very character- istic. Completely everted hemipenes (Fig. 2) are present in two specimens. The organs are relatively slender and deeply bilobed ; the suleus spermaticus forks in the crotch, and the apices are flattened, a condition that has been noted for innocens and not for caeca which has rounded apices (Thomas, 1965). Autotomy level varies from the sixth to eighth caudal annulus. Different ecological preferences were observed for schmidti and caeca in the jointly occupied parts of their ranges. A. schmidti was found in more open and exposed areas, the edges of pastures, the cleared sides of gradually sloping dolines — areas where the tree cover is lacking, or high and sparse. A. caeca, although wider in its ecological range, prefers shady areas such as coffee groves or wooded ravine bottoms. The following table indicates briefly the type of situation in which either snecies was collected at particular localities within their joint range; the number of specimens at each locality is indicated in paren- theses. schmidti caeca acacia serub (6) coastal almond (Terminalia) woods (2) moderate to well-exposed, scrubby ravine sides (6) moderately, exposed hillside (Gl 55. as. 2 od 5 Ee ee ee ee ee coffee grove (3) 2 a an ee ee coffee grove (1) exposed hillside pasture (2) mesic woods adjacent to pasture (1) high, open ravine woods (1) exposed ravine side (1), shady ravine woods (1) moderately exposed edge of interior of coffee grove (1) coffee grove (1) 16 BREVIORA No. 249 schmidti caeca CANE GANISEIR mie 9 | Nase omadmbo ceed bac doline pasture (1) moist, open pasture (2) Ay A OY acta Ee cleareduslopeyot dolines@l) 9) 9) Paces ae eee moderately exposed roadside shady coastal woods (1) ditch (1) Qoan lnkeiieycarm Gy) eh dba E SS Doo eo AMPHISBAENA FENESTRATA Cope Specimens examined: new localities: VIRGIN ISLANDS. St. John: Lameshur Bay, VINP 1478; Annaberg Road near Frederickdal, VINP 1479; Caneel Hill Trail, VINP 1485; Coral Bay, VINP 1490; Windberg ruins, ASFS V7504-05; Frederick- dal ruins, ASFS V7564, V8066. Tortola: Baugher’s Bay, ASFS V7921; Jackass Gut, ASFS V7940, V7984-85, RT 977. Virgin Gorda: SW slope of Gorda Peak, ea. 500’ (150 m), ASFS V3805- 06. Great Camanoe: between Lee Bay and Cam Bay (shed skin). Previously recorded localities: St. John: VINP 1496-97. In the report on A. fenestrata by Gans and Alexander, ‘‘ Doro- thea’’ and ‘‘ West End of Great St. James”’ are listed in “‘local- ity records’’ under St. John. Dorothea is on St. Thomas and Great St. James is a small satellite island of St. Thomas. When the data from the new specimens are placed with the old data from the various islands, the trends noted by Gans and Alexander are almost completely obliterated. The following ranges summarize the new data (data from Gans and Alexander are in parentheses) : body caudal ventral Island annuli annuli segments postcloacals Staelhomasiy sane (242-247) .... (2-14)... (14-16) 2. . (8-20) St. John 241-251 (247-249) 12-13 (12-14) 14-17 (14-15) 8-11 (9-10) Tortola 243-248 (236-242) 13-14 (13-14) 14-16 (14-16) 8-11 (9-12) Virgin Gorda 240-245 ...... ZAMS) 9. bee MGA ee ee) The shed skin from Great Camanoe is too fragmentary to yield counts. The Virgin Gorda and Great Camanoe records are the first for those islands. The two specimens from Virgin Gorda have higher dorsal midbody counts than do those of the other islands; the total midbody counts are higher than all but one of the specimens from the other islands. 1966 AMPHISBAENIDS OF PUERTO RICO n7. Gans and Alexander have said (1962:95) that the dorsal groove in West Indian Amphisbaena ‘‘may be suggested by an alignment alone.’’ However, in well-injected specimens of A. fenestrata the dorsal groove is structurally evident on approxi- mately the posterior two-thirds to three-fourths of the body. The paramedian segments are rounded medially and the result- ing space is occupied by ‘‘fold granules.’? The appearance of this groove at its fullest development is much the same as that of the lateral grooves. It becomes increasingly indistinct ante- riorly and ends at the approximate level of the thirty-sixth to sixty-seventh postoral annulus. No other specimens of West Indian species seen by me (caeca, schmidti, bakert, cubana, imno- cens, gonavensis, manni) have so prominent a dorsal groove, although one specimen of baker: has a moderately well indicated dorsal groove. Amphisbaena fenestrata inhabits islands that are preponder- antly xeric; however, it prefers (or at least is most easily col- lected in) the most mesic areas that can be found. When we visited Virgin Gorda the weather was unseasonably dry. Shed skins of these animals were found regularly under rocks in the more shady parts of the low, xeric woods that cover much of the island. After almost a week of persistent rock turning, two specimens were obtained on a heavily wooded hillside after a heavy morning shower. Typhlops richardi, which was more abundantly collected in all parts of the island, was also found in moist piles of coconut husks near the coast, where Amphis- baena fenestrata was not. This preference for the more mesic areas was noticed also on St. John and Tortola. The presence of a soil of high organic content seems to be an important fac- tor. A brief visit to the island of Anegada yielded no specimens or signs of amphisbaenids, although six specimens of Typhlops were collected. This species is now known from the islands of St. Thomas, Great St. James, St. John, Tortola, Great Camanoe, and Virgin Gorda. Its true range is probably all of the Virgin Islands large enough to bear a suitable habitat, St. Croix excepted. Anegada is geologically distinct and well set off from the rest of the Vir- gins, but its fauna is not radically different; fenestrata may be found there. ACKNOWLEDGEMENTS The collecting that preceded this study and the study itself were supported by Dr. Albert Schwartz, to whom I am most in- 18 BREVIORA No. 249 debted. Sr. Felix Inigo of the Department of Agriculture and Commerce of the Commonwealth of Puerto Rico has been prompt and courteous in his attention to requests for collecting permits ; Mr. Jesse E. Williams was of great assistance to our collecting on Ramey Air Force Base, Puerto Rieo. Dr. Carl Gans and Mr. A. Allan Alexander have helpfully read and eriticized the manu- seript. I wish to express appreciation to Mr. Donald W. Buden, Mr. Gerald D. Gagnon, Mr. David C. Leber, and Mr. Ronald F. Klinikowski for their most able assistance in the field. Mueh appreciation is due to the following people who loaned material for study: Mr. Hugh B. Muller, Virgin Islands Na- tional Park (VINP); Dr. Charles F. Walker and Mr. George R. Zug, University of Michigan Museum of Zoology (UMMZ) ; Dr. Ernest E. Williams, Museum of Comparative Zoology at Harvard (MCZ) ; Dr. Juan A. Rivero and Mr. Horacio Mayorga, University of Puerto Rico at Mayagiiez (UPR). The following institutions have received paratypes of Amphisbaena xrera: the U.S. National Museum (USNM), the University of Illinois Mu- seum of Natural History (UIMNH), and the Carnegie Museum (CM). RT indicates the Richard Thomas private collection. LITERATURE CITED GANS, CARL 1964. Amphisbaena schmidti, a third species of the genus from Puerto Rico (Amphisbaenia, Reptilia). Breviora, Mus. Comp. Zool., No. 198, pp. 1-11. GANS, CARL AND A. ALLAN ALEXANDER 1962. Studies on amphisbaenids (Amphisbaenia, Reptilia). On the amphisbaenids of the Antilles. Bull. Mus. Comp. Zool., Vol. 128, no. 1, pp. 65-158. Fow Ler, Henry W. 1918. Some amphibians and reptiles from Porto Rico and the Virgin Islands. Pap. Dept. Marine Biol. Carnegie Inst., vol. 12, pp. 1-15. ScHMIDT, KARL PATTERSON 1928. Amphibians and land reptiles of Porto Rico, with a list of those reported from the Virgin Islands. Sci. Surv. Porto Rico and Virgin Islands, vol. 10, part 1, pp. 1-160, 4 pls. THOMAS, RICHARD 1965. Two new subspecies of Amphisbaena (Amphisbaenia, Reptilia) from the Barahona Peninsula of Hispaniola. Breviora, Mus. Comp. Zool., No. 215, pp. 1-14. (Received 2 November, 1965) iT | 1966 AMPHISBAENIDS OF PUERTO RICO 19 KEY TO GREATER PUERTO RICAN SPECIES OF AMPHISBAENA le bo Nasals separated by posterior extension of rostral which contacts pre- frontals; body annuli 236-249 ............... ee es ChLestnaua Nasals in contact with one another ........... Be er. Wee Number of body annuli low (192-207); high number of caudal annuli (CL S522) eee ree Menta ks oy ) Wi: arma Ee hye OO Ue. ee ae schmidti Number of body annuli higher (more than 214); caudal annuli lower GLROTSICSS)) Lp eee a a ae OO ee ee ee ee ae 3 Internasal suture short (less than one-third interprefrontal suture) ; bodyaranm ule AS eee aes ac S ‘LNAA “310 N Zi SITWU.LNAD ‘Ida 'N GIODILYOD “YAdaud ‘LIVIN LSYAdAH WY 1 ;. 3. cuboids: MCZ, 77.76, 4; naviculars: MCZ, 7777. 22 proximal phalanges, MCZ 7778, 19 second phalanges; MCZ 7779, 5 ungual phalanges. Horizon and locality: Middle Miocene, Hawthorne Formation, Thomas Farm, 8 miles north of the town of Bell, Gilchrist County, Florida. Diagnosis: Equal in size to Synthetoceras rileyi rileyi* Frick, but P; is 13 per cent smaller, and median cusp of Py somewhat narrower, so that internal depression is flatter and deeper. Description: The mandible is long and slender, remaining rather shallow beneath the molars. The angle of the jaw is a smooth, continuous curve, as in Protoceras. Anterior to P., the jaw begins to flare outward so that P, is inclined at about 35 degrees from the vertical. The alveoli for the incisors indicate that they were small, closely spaced, and quite procumbent. That for the canine is larger and also procumbent and close behind I,, suggesting that it was a functional member of the incisor series, as in Protoceras. P, follows after a short diastema equal to M, in length. In the + Frick’s collection data is not precise, but the type specimen appears to have been collected in the Catahoula sandstone in Walker County, Texas. Although the age of this formation is still not settled, Fisk (1940) considers it of Miocene age, and Gidley (see Renick, 1936) of Middle Miocene age. 18 BREVIORA INOs255 type and MCZ 3656 and 7637, P, is a large, single-rooted, re- curved, caniniform tooth laterally compressed and with blade- like anterior and posterior edges. In MCZ 3708, 7635, and 3655 the alveolus for the first premolar indicates a very tiny vestigial tooth which could not have been functional to any great extent in the living animal. In size and dental morphology these two groups are identical. Since dimorphs of this kind are known among artio- dactyls, and with no evidence to the contrary, it is preferable to consider these as sexual morphs rather than two distinct taxa, the latter group being the females. A long diastema separates P, and P.. The latter is a two-rooted, simple, robust tooth equal to Ps in length (Fig. 2). Ps is propor- tionately smaller than the corresponding tooth in S. rileyi rileyi, but similar to it in structure. The anterior edge is somewhat re- curved onto the internal face forming a moderate stylid. The high median cusp is connected with the anterior and posterior cusps by narrow crests. A posterior intermediate crest extends backward nearly to the edge of the tooth, but fails to fuse with the posterior crest, resulting in a partially closed lake which is open postero- medially. P, is similar to P; with a stronger parastylid and a lake which is completely closed near the base of the crown. It differs from S. rileyi rileyi in having a narrower median cusp. The molars are indistinguishable from those of S. r. rileyi. My, and M, have narrow crescents, strong parastylids, and weak metastylids. The median ridges are well developed, especially on Ms. Mg is similar but with the posterior edge of the metaconid slightly overlapping the entoconid. The hypoconulid is C-shaped forming two functional grinding surfaces, and is quite variable in Size. g SSS SS =, —S MCZ 3654 cna MCZ 3660 EIS MCZ 3656 Figure 2. Synthetoceras (Prosynthetoceras) rileyi australis, composite crown view of right lower cheek teeth, X1. In the upper dentition, the referred palate most closely resem- bles that of Synthetoceras francisi Frick. M:; is longer than broad with the posterior pair of crescents narrower than the anterior pair. The metastyle is weak and the mesostyle prominent, but less so 1966 | SELENODONT ARTIODACTYLS FROM THOMAS FARM 19 than in S. francisi. The parastyle is as strong as the mesostyle. The internal crescents fuse early and extend as a narrow process between the external crescents, which they do not fuse with until late in wear. A median pillar, check-shaped in cross section, is present between the internal crescents, with the longer arm along the protocone. A prominent elongate pillar rises from the cingu- lum along the anterior side of the protocone. M? is smaller than M® and shorter anteroposteriorly than in Synthetoceras tricornatus; the styles are as prominent as in that species and the median ridges are low and indistinct. M! is the smallest of the molar series and, although worn, appears to be similar to M? in character. P* is triangular, as in S. francisi, with a very strong anteriorly curved metastyle. Unlike conditions in the corresponding tooth of that species, however, the parastyle is nearly absent. P® re- sembles that of S. francisi, but has a stronger metastyle and a strong and complete internal cingulum. A small wear facet on the anterior edge of P® indicates that P* was present, as in S. francisi. On the basis of relative size, a number of foot bones are re- ferred to this species. The phalanges are short and heavy as in Protoceras. Discussion: As can be seen from Table 3 the range of variability of S. australis is sufficient to include the type of S. rileyi. In the character of the molars there are no differences which could not be attributed to intraspecific variation. The premolars, except for the differences already noted, are nearly identical in the two forms. The subspecific rank of S. rileyi australis is proposed in view of the premolar differences, and the distance between Walker County, Texas, and Gilchrist County, Florida (nearly 800 miles), and be- cause of the possibility of a slight difference in horizon between these localities. Since both forms are so poorly known at present it is advisable to retain a subspecific distinction, at least until the extent of variation in each is better known. Frick (1937) was apparently uncertain as to the subgeneric reference of S. rileyi, questionably placing it in the subgenus Syn- thetoceras. The subgeneric distinction between S. (Synthetoceras ) tricornatus and S. (Prosynthetoceras) francisi is based on the smaller size of the latter, its shorter crowned teeth, the retention of P?, and the prominence of the upper canine alveolus. S. rileyi is referred to the subgenus Prosynthetoceras because ot its small size and lower crowned teeth as compared with S. tri- cornatus, and because of the retention of P? in the referred palate, MCZ 4065. 20 BREVIORA TABLE 3 No. 255 Comparative measurements of Synthetoceras (Prosynthetoceras ) rileyi australis in mm Width of palate between M! P® to M® length P® length P? width P*# length P* width M! length M! width M? length M2 width M® length M® width Depth of jaw under Po Depth under P4 Depth under M; P2 to Mz length P; to Py diastema Ps length* Ps width* P; length Ps width Py length P, width M; length My, width Mo length Mo» width M3 length Ms width MCZ 3654 MCZ 3708 PIMeg2 DESI 23.6 a 29M DY 85.4 76.6 42.0 46.0 — 8.5 — 4.0 10.5 10.5 6.5 6.3 13.0 12.2 10.0 OD 18.0 16.1 11.0 10.5 YES) 25.0 11.4 IN dleal Synthetoceras (P.) rileyi australis MCZ 4065 Synthetoceras (P.) rileyi australis 64.5 64.8 9.0 6.1 9.0 3:6 13.6 14.8 V7/e2 15.9 17.0 16.8 MCZ 7635 Pp bcs) 23.0 28.4 76.7 10.1 Soi! 14.4 9.4 15.4 21.6 21.6 10.5 Synthetoceras (P.) rileyi rileyi AMNH 34181, type 1933 20.5 77.4 8.2 Siell 9.7 4.3 10.4 6.4 12.4 9.4 16.7 22.4 22.4 10.1 “ Known only from MCZ 3656; length, 8.2 mm, width, 4.0 mm. 1966 | SELENODONT ARTIODACTYLS FROM THOMAS FARM P| MERYCOIDODONTIDAE MERYCHYINAE cf. MERYCHYUS sp. An oreodont is represented in the Thomas Farm fauna by a single left upper molar, MCZ 7765, and an incomplete tarsus, MCZ 7766. In structure and size the molar (Fig. 3) most closely resembles Merychyus minimus, but is also quite similar to M. elegans of a later age. The external faces of the outer crescents are transversely concave and longitudinally convex, so that their apices overhang the internal crescents in typical oreodont fashion. The posterior crescents are narrower than the anterior pair and are relatively longer anteroposteriorly, as in M® of M. elegans. A strong cingulum is present on the anterior side of the protocone and a weaker one on the posterior side of the hypocone. Figure 3. Oreodont upper left molar; cf. Merychyus. External view (left), and crown view (right), X1%. There can be little doubt of the oreodont affinities of the tarsal elements, which probably belong to a single individual. The cal- caneum is relatively short, strong, and laterally compressed with a strongly truncated proximal end which bears a deep groove for the tendon of the gastrocnemius. The sustentaculum is very small and only slightly projecting beyond the medial surface of the tuber. The astragalar facet on the sustentaculum is relatively long and narrow. The astragalus is wide, low, and quite asymmetrical. The proxi- mal trochlear is inclined at an angle to the distal one, and is placed more lateral in position as is typical of oreodonts. The sustentacu- lar facet is narrow, occupying little more than half of the plantar surface. 22, BREVIORA INO: 7255 These fragments are tentatively referred to the genus Merychyus because of the similarities to M. minimus in both size and morphol- ogy. However, a more certain reference must await more com- plete material. The discovery of oreodont remains in Florida is a significant, although not a completely unexpected find. The apparent rarity of these animals in this region may not reflect their actual abun- dance. Schlaikjer (1935, p. 173) suggested that a gregarious habit combined with localized distribution and/or species habitat re- quirements of oreodonts could account for their great abundance in some areas, and relative rarity in others of the same age. As suggested by Thorpe (1937, p. 9) and Patterson (1949, p. 270), the absence of sediments deposited under appropriate conditions or their subsequent removal by erosion could be responsible for a meager oreodont fauna in certain areas. The idea, as proposed by Romer (1947, p. 10), that Florida was ecologically unfavor- able for these animals is less attractive since the environment of the Florida Miocene was inferred partially by the absence of oreo- donts. The rarity of this group from Florida beds is probably due to a combination of the first two factors: localized distribution and scarcity of sediments. CERVIDAE BLASTOMERY™*. (PARABLASTOMERYX ) FLORIDANUS White, 1940 Hypodigr.:: MCZ 3626, type, partial left mandible with P4, M,, the ante: or external crescent of M., and the alevoli for Ps and P;; MCZ 3627, paratype, upper left molar; MCZ 3706, right mandible with dP, to M2; MCZ 3707, left mandible with dP. to Ms; MCZ 3912, left mandible with dP, to Ms; MCZ 4220, left mandible with P,, dP; to M3; MCZ 4221, right mandible with P; to M3; MCZ 4222, right mandibie with P., M, to M3; MCZ 7795, 23 astragali; MCZ 7796, 24 calcanea; MCZ 7797, 17 cubo- naviculars, MCZ 7798, 103 proximal phalanges; MCZ 7799, 53 second phalanges; MCZ 7800, 24 ungual phalanges; MCZ 7801, 6 metapodiais; MCZ 7802, left tibia, MCZ 7803, right radius. Horizon and locality: Middie Miocene, Hawthorne Formation, Thomas Farm, Gilchrist County, Florida. Diagnosis: White (1940, p. 34) diagnosed this species as fol- lows: “A large Parablastomeryx, paraconid of Py with very small stylid, metaconid large and placed about the mid-length of the tooth, entoconid projecting at nearly right angles to the axis of the tooth, crest of hypoconulid extending along posterior border of the tooth nearly to the inner border.” 1966 SELENODONT ARTIODACTYLS FROM THOMAS FARM 23 Discussion: Since White’s diagnosis, many new specimens have been collected. These demonstrate a remarkable degree of vari- ation in the premolars and reveal the character of the lower mo- lars and milk dentition, which were not available to him. The molars (Fig. 4) tend to confirm White’s observation that B. flori- danus is similar to B. primus. MCZ 7763 MCZ 4349 Figure 4. Blastomeryx floridanus. a, Composite crown view of left lower cheek teeth, X12. b, MCZ 3706, right dP4, X3. The hypoconulid of M; has a prominent anterointernal cusp, and a strong palaeomeryx fold is present on the protoconid. The internal stylid of P, which White described as being “very small” on the type varies from strong on MCZ 7763 to absent on MCZ 4221. In Ps of most of the specimens the entoconid is joined to the hypoconid by a semicircular crest around the posterior margin of the tooth, forming an enciosed lake. However, in MCZ 4349 the lake remains open posteriorly and is divided from the ento- conid by a strong median ridge, which extends posteriorly to the end of the tooth and, after turning mediaily, ends in a minute cap- sule. These and other variants must be considered to be within the range of variation of this species since intermediate types link every extreme. The lower milk teeth, dP, to dP, are known from two speci- mens, MCZ 3706 and 3707. DP, (Fig. 4b) is five cusped with an elongated pillar between the anterior inner cusps and a conical pillar between the posterior inner cusps. DP. is very similar to Ps in structure and length, but is much narrower. There is little difference between dP» and Po». 24 BREVIORA INo#255 Because of the palaeomeryx folds, hypoconulid cusp, and rela- tively large premolars, this species is referred to the subgenus Parablastomeryx. Referred postcranial material includes numerous tarsal bones, phalanges, a tibia, a radius, and several cannon bones. The tarsus is typically cervid. The tuber of the calcaneum is light and later- ally compressed; the cuboid and navicular are solidly fused. The proximal phalanx is lightly built and considerably more slender dorsoventrally than in Recent cervids. The metatarsals are fused along their entire length and there is a deep longitudinal groove on the dorsal side between metatarsal III and metatarsal IV. The distal facets are completely keeled, and together are considerably wider than the shaft. The metacar- pals are fused and somewhat shorter than the metatarsals. The dor- sal groove is shallow and indistinct. The tibia is long and slender, but less curved laterally than in Recent forms. TABLE 4 Comparative measurements of Blastomeryx (P.) floridanus in mm MCZ MCZ MCZ MCZ 7763 4221 4349 3626 Jaw and lower dentition: Jaw length — plis2 — — Ps to M3 — 60.8 66.9 — Ps length 5.6 — — — Ps width 2.6 -- — — Ps length 7.3 8.0 8.2 —- P3 width 4.3 4.3 4.6 — P, length 8.9 8.4 8.7 8.2 Py width 5.4 5.0 5.0 4.9 M; length — 10.1 10.3 9.0 M, width — ES 7.0 6.3 Mp length — 11.8 silo z — Me width — 8.2 TS) —- M3 length — 16.2 eS — Ms width — 8.2 Vel =H “estimated from incomplete specimen. 1966 SELENODONT ARTIODACTYLS FROM THOMAS FARM 25 MACHAEROMERYX GILCHRISTENSIS White, 1941 Hypodigm: MCZ 3651, type, partial right mandible with P, to Mz, and alveoli for P2, P; and M3; MCZ 3652, right mandibular fragment with M,, and a left upper molar; MCZ 3709, partial right mandible with P; to Mz, and part of M;; MCZ 3710, partial right palate with M! and M2; MCZ 7804, 37 astragali; MCZ 7805, 1 calcaneum; MCZ 7806, 3 cubo-naviculars; MCZ 7807, 24 proximal phalanges; MCZ 7808, 5 second phalanges; MCZ 7809, 2 ungual phalanges; MCZ 7810, 1 metacarpal. Horizon and locality: Middle Miocene, Hawthorne Formation, Thomas Farm, Gilchrist County, Florida. Diagnosis: Premolar series relatively longer than in M. tragulus, My shorter, upper molars less Square with outer crescents consid- erably wider than inner ones, median pillars of upper molars only vaguely suggested in contrast to M. tragulus. Discussion: No new information can be added to the diagnosis since this species remains so poorly represented. Aside from the differences in proportion, the teeth are very similar to M. tragulus. The few tarsal remains and a referred metcarpal are essentially like those of Blastomeryx, but much smaller. CONCLUSION The fossil selenodont artiodactyls from Thomas Farm, Gilchrist County, Florida, indicate faunal communication with other parts of the North American continent during the Lower Miocene. Synthetoceras r. australis is at most subspecifically distinct from the western S. r. rileyi, and Blastomeryx floridanus and Machaero- meryXx gilchristensis are close to western species. The specialized camelids, Floridatragulus and Nothokemas, are so far unknown in the West, but they or closely related forms have been recorded from the Oakville of the Texas coastal plain (Wilson, 1962). This contradicts White’s contention that north-central Florida was an island until the end of Oak Grove time. Continuity, even if intermittent, between Florida and the mainland certainly seems to have occurred during the earlier Miocene. On geological grounds, Cook (1945) suggests that a narrow peninsula persisted through much of this period, and Bader (1956), on the basis of his comparisons of the Thomas Farm equids with those of Texas, States that isolation was, at best, very ephemeral. The opportunity for faunal movement between these areas, then, must have been considerably greater than suggested by White (1942). 26 BREVIORA Nossa») REFERENCES BADER, R. S. 1956. A quantitative study of the Equidae of the Thomas Farm Miocene. Bull. Mus. Comp. Zool., 115: 49-78. Cook; CW. 1945. Geology of Florida. Bull. Florida Geol. Surv., No. 29: 1-339. DouGHERTY, J. F. 1940. Skull and skeletal remains of the camel Paratylopus cameloides from the John Day, Oregon. Carnegie Inst. Washington Publ. No. 514: 49-58. Fisk, HN: 1940. Geology of Avoyelles and Rapides parishes. Bull. Louisiana Geol. Surv., 18: 1-240. FRIck, C. 1937. Horned ruminants of North America. Bull. Amer. Mus. Nat. Hist., 69: 1-669. McKenna, M. C. 1966. Synopsis of Whitneyan and Arikareean camelid phylogeny. Amer. Mus. Novit. No. 2253: 1-11. PATTERSON, B. 1949. Rates of evolution in taeniodonts. Pp. 243-278. In G. L. Jepsen, G. G. Simpson, and E. Mayr eds., Genetics, Paleon- tology, and Evolution, Princeton Univ. Press. Purl, H. S. and R. O. VERNON 1964. Summary of the geology of Florida and a guidebook to the classic exposures. Florida Geol. Surv., Special Publ. No. 5: 1-312. RAv, CE: 1957. A list, bibliography, and index of the fossil vertebrates of Florida. Florida Geol. Surv., Special Publ. No. 3: 1-175. RENICK, B. C. 1936. The Jackson group and the Catahoula and Oakville formations in a part of the Texas Gulf coastal plain. Univ. Texas Bull. No. 3619: 1-104. RoMER, A. S. 1947. The fossil mammals of the Thomas Farm, Gilchrist County, Florida. Quart. Jour. Florida Acad. Sci., 10: 1-10. SCHLAIKJER, E. M. 1935. Contributions to the stratigraphy and paleontology of the Goshen Hole area, Wyoming, Part 4. Bull. Mus. Comp. Zool., 74: 97-189. SIMPSON, G. G. 1932. Miocene land mammals from Florida. Bull. Florida Geol. Surv., No. 10: 11-14. 1966 SELENODONT ARTIODACTYLS FROM THOMAS FARM af THORPE, M. R. 1937. The Merycoidodontidae, an extinct group of ruminant mam- mals. Mem. Peabody Mus. Nat. Hist., 3: 1-428. WHITE, T. E. 1940. New Miocene vertebrates from Florida. Proc. New England Zool. Club, 18: 31-38. 1941. Additions to the Miocene fauna of Florida. Proc. New Eng- land Zool. Club, 18: 91-98. 1942. The Lower Miocene mammal fauna of Florida. Bull. Mus. Comp. Zool., 92: 1-49. 1947. Additions to the Miocene fauna of northern Florida. Bull. Mus. Comp. Zool., 99: 497-515. WILSON, J. A. 1962. Field trip no. 10, Tertiary formations between Austin and Houston, with special emphasis on the Miocene and Pliocene. In Geology of the Gulf Coast and Central Texas and a guide- book of excursions. Houston Geol. Soc., 1962: 342-354. (Received 1 November, 1966.) Errata Page 6, Fig. la, for X4/5 read X2/5. Page 10, Table 1, for P! length and width read P?. Page 11, Table 1. M; length, MCZ 4086, for 13 read 25; MCZ 4291, for 13 read 24. Page 11, Nothokemas, line 8 (MCZ 4322) for Pz to Mz read P? to M?. Page 15, Table 2, M, width, AMNH 8197, for 14 read 9.5. Page 17, lines 9, 10. MCZ 3655, for P, to Mz and partial alveoli for I, to canine, read M,-Mz and alveoli for P,-P4,. Page 17, line 11, for MCZ 3666 read 3660. Page 20, Table 3. M2 width, MCZ 7635, for 21.6 read 10.3; AMNH 34181, for 22.4 read 10.4. Page 22, Blastomeryx, Hypodigm. Add MCZ 7763, partial left jaw with P,-P,, and MCZ 4349, left jaw with P;-Ms. qh Tobe OWvst sae = ua = BREVIORA Museum of Comparative Zoology CAMBRIDGE, MAss. 3 FEBRUARY, 1967 NUMBER 256 ANOLIS CHOCORUM, A NEW PUNCTATUS-LIKE ANOLE FROM DARIEN, PANAMA (SAURIA, IGUANIDAE) By Ernest E. Williams and William E. Duellman'! INTRODUCTION In the course of a survey of the herpetofauna of Panama, the junior author, accompanied by Charles W. Myers, Tomas Quin- tero, and Linda Trueb, travelled in a dugout canoe from Santa Maria de El Real, Darién, to the upper part of the Rio Tuira Basin in July, 1965. Most of the month of July was spent working out of a base camp at the confluence of the Rio Tuira and the smaller tributary, the Rio Mono. Collections also were made at the mouth of Quebrada La Plata lower on the Rio Tuira, and on Cerro Quia on the Panamanian-Colombian border. Six specimens of a previously undescribed species of anole are included in the collection; these plus one specimen obtained by G. B. Fairchild at Tacarcuna, Darién, and three specimens obtained more recently from the Serrania de Pirre, form the basis for the present descrip- tion. A long previously collected specimen from Colombia in the American Museum of Natural History (AMNH) proves to be- long to the same species. The junior author is grateful to his field companions, whose combined efforts made the trip a great success. The survey of the herpetofauna of Panama is being conducted in cooperation with the Gorgas Memorial Laboratory in Panama and is supported by a grant (NIH GM-12020) from the National Institutes of Health. This paper is one of a series of studies of the genus Anolis pre- pared by the senior author with the support of National Science Foundation Grant GB-2444. The material of the new species has been divided between the Museum of Natural History, Kansas University (KU) and 1 Museum of Natural History, University of Kansas, Lawrence, Kansas. 2: BREVIORA No. 256 the Museum of Comparative Zoology (MCZ). We name the new anole for the Choco Indians of the region: ANOLIS CHOCORUM new species Type. KU 96934, adult ¢, Rio Tuira at Quebrada La Plata, 100 m, Darién, Panama, collected by a native for Charles W. Myers, 26 July 1965. Paratypes. Panama: Darién. KU 76027, young ¢é, Tacarcuna Village on Rio Tacarcuna, 550 m, G. B. Fairchild coll., 9 July 1963; KU 96931, Rio Tuira at Rio Mono, 130 m, William E. Duellman coll., 12 July 1965; MCZ 82546, same locality, William E. Duellman coll., 13 July 1965; MCZ 82547, same locality, Charles W. Myers coll., 13 July 1965; KU 96932, Cerro Quia, 740 m, Charles W. Myers coll., 9 July 1965; KU 96933, same locality, Charles W. Myers coll., 26 July 1965; KU 98520, north end, Serrania de Pirre, 320 m, C. W. Myers coll., 15 January 1966; KU 98521, south ridge, Cerro Cituro, Serrania de Pirre, 1100 m, C. W. Myers coll., 23 January 1966; KU 98522, same locality, C. W. Myers coll., 25 January 1966. Colombia: Choco. AMNH 18235, Atrato Drainage, Quesada River, R. D. O. John- son coll., 6 November 1920. Diagnosis. An Anolis related to A. punctatus Daudin and A. transversalis A. Duméril but differing from the first in coloration, and in having a large dewlap in the female, and the snout not swollen in the male; it differs from the second in lacking sexual dichromatism in dorsal pattern, and in a higher number of loreal rows, and in coloration. It differs from both species in a lower number of lamellae under phalanges 2 and 3 of fourth toe. See also Tables 1 and 2. Description. (Paratype variation in parentheses.) Head. Head scales moderate to small, smooth. Fourteen (11-14) scales across snout between second canthals. Weak frontal ridges outlining a shallow frontal depression. Scales in frontal depression equal to or distinctly smaller than surrounding scales. Seven (7-9) scales border rostral posteriorly. Nasal scale separated from rostral by one to two scales. Six (5-7) scales between supranasals. Supraorbital semicircles separated by three (1-3) scales, sepa- rated partly or wholly from the weakly differentiated supraocular disks by one (1-2) row of scales. One (1-3) elongate supraciliary continued posteriorly by granules. Canthus distinct, nine (8-10) canthal scales, the anteriormost very small. Loreal rows eight (7-9), the uppermost abruptly larger. 1967 NEW ANOLIS FROM PANAMA 3 SS SOQ5c See RES S = IS Fig. 1. Anolis chocorum, new species. Head scales of type, KU 96934. Actual snout-vent length 79 mm. Temporal and supratemporal scales finely granular, several lines of weakly enlarged granules at the angle between temporal and supratemporal areas. Interparietal larger (or equal to, or smaller) than ear, separated from the supraorbital semicircles by four (2-4) somewhat enlarged scales. Laterally, slightly smaller scales grade into the supratemporal granules. Behind the interparietal the still smaller scales grade very quickly into granules as minute as those of the supratemporal area. Three to four (2-4) suboculars in contact with the supralabials anteriorly, in contact with (or separated by 2-3 scales from) the canthus, posteriorly decreasing abruptly in size. Nine (8-9) supra- labials to the center of the eye. Limbs and digits: Hind foot scales multicarinate. Scales on anterior face of upper and lower arm, thigh and lower leg uni- Carinate, those at knee multicarinate. The larger scales of the hind limb as large or larger than the ventrals. About 20 (17-20) lamellae under phalanges 2 and 3 of fourth toe. Tail: Tail slightly compressed. No dorsal crest, lateral and dorsal scales subequal. Two ventral rows larger, keeled. Verticils not evident. Postanal scales enlarged in male. Scales immediately behind vent smooth. 4 BREVIORA No. 256 aa tae Oy UY) == SS ao. ‘ x \) N' SS S. 7, Se, S Se SO fas K nN) ea “ta \) \S Rae ae vBC Fig. 2. Anolis chocorum, new species. Side view of head of type, KU 96934. Actual snout-vent length 79 mm. Size (snout/vent length): 4 (Type) 79 mm; longest @ 73 mm. Color. Color in life is summarized in Table 3. The blue gray belly spotted with white is quite distinctive and differentiates this species immediately from any previously known. As indicated in Table 3, the sexes differ most prominently in the color of the dewlap; it is in fact uncertain whether any other sex-correlated color differences exist. Both sexes vary somewhat in color; the occurrence in both of uniform and patterned dorsa is especially conspicuous. It is possible, however, that the pat- tern when present is feebler in females than in males; it is true in the specimens as preserved. Whether it was true in life is not recorded. Whether also some peculiar color variants (the light middle of the belly in some males and the gray tongue of one female) imply a range of variation in one sex not found in the other is unknown. Ecology. The Darién localities from which A. chocorum is known are in a broad expanse of relatively undisturbed tropical 1967 NEW ANOLIS FROM PANAMA 5 evergreen forest that is moderately well stratified and characterized by a lower story of slender palms in many places. The area has a definite, but not severe, dry season. Anolis chocorum is arboreal. All individuals observed by day were in trees. One was obtained from the branches of a large tree immediately after it was felled. Another individual dropped from a tree during a storm at night. Two individuals were sleep- ing on bushes at night, one in a large fern on the steep bank of a mountain stream. Several species of Anolis are associated with A. chocorum in eastern Darién; these include two large arboreal species (A. biporcatus and A. frenatus), the moderate-sized arbo- real A. capito, the smaller bush-dwelling A. chloris, limifrons, tropidogaster, and vittigerus, and the semi-aquatic A. poecilopus. On July 10, 1965, Charles W. Myers found a striated egg partly concealed under loose bark on top of a rotting log on Cerro Quia. The egg was 16 mm x 12 mm; it hatched on July 26. The hatch- ling (KU 96933) had a snout-vent length of 30 mm and a total length of 92 mm. In coloration the hatchling resembled adults of A. chocorum by being green above with lateral diagonal rows of blackish brown spots. The tail was green with indistinct grayish brown crossbands. The venter was white with a faint bluish cast to the belly and a pale dull yellow cast to the median part of the dewlap. Relationships. The series to which A. chocorum belongs is clearly an endemic South American one (see discussion of the alpha and beta series of Etheridge, 1960, in Williams, 1965). Its discovery in series in Darién is probably only an accident of activity and attention: similarly thorough and careful collecting has not been done in the adjacent areas of Colombia. However, a single specimen (AMNH 18235) records its presence there. This is a distribution resembling that of a number of other anoles. A. chloris, for example, associated in Darién with A. chocorum, is likewise a South American autochton, well known in Chocoan, Colombia (there known under the names A. palmeri and A. gor- gonae), and in trans-Andean Ecuador. The distribution of 4. chocorum may well be more like that of A. chloris than we now know and extend into western Ecuador. However, the range of A. chocorum, as at present known, is only eastern Darién Province in Panama, where it occurs at elevations of from 100-1100 m, in the upper Rio Tuira Basin and on the Pacific slopes of Cerro Quia and the Serrania del Darién, and in the Colombian Choco. Of the endemic South American group, Anolis chocorum is very clearly most similar to Anolis punctatus Daudin of Brazil, the 6 BREVIORA No. 256 Guianas, Peru and Ecuador. The most immediately striking dif- ferences are in the coloration of the belly (blue gray with white spots) and in the lack of any snout swelling in the male. The presence of a much lower lamellar count and a large dewlap in the female are soon discovered as further differences. A. punctatus is a species of the Amazonian-Guianan forests and the closest approach to the Darién-Choc6 form thus far known is N. end Serraniae de Pirré 5S < =~ Q P ry e Cerro Cituro Tacarcuna Village Quebrada fla UTE 20 30 40 Kilometers Fig. 3. Map of the distribution of Anolis chocorum, new species. 1967 NEW ANOLIS FROM PANAMA 7 represented by a specimen from the Rio Apaporis in eastern Colombia. Over most of its enormous range A. punctatus has smooth ventrals like A. chocorum. Some of the Peruvian, all the Ecuadorian, and the Colombian (Apaporis) populations have keeled ventrals. For these the subspecific name boulengeri is available. Overlapping a substantial part of the western range of A. punc- tatus, all of boulengeri, and part of Peruvian typical punctatus is A. transversalis. Structurally, this differs from A. chocorum in lower loreal and toe lamellar counts and in the peculiar mental- postmental scale pattern. The very striking throat pattern of the female is very different from the condition of either sex in A. cho- corum and the bold crossbanding of the dorsal surface of the female transversalis is also very different. The dorsal pattern of the male transversalis (originally described as a distinct species, buck- leyi), however, does somewhat resemble the irregular spotted condition mentioned above as seen in some A. chocorum. A. transversalis, so far as known, gets geographically no closer to A. chocorum than does A. punctatus. There is the very sub- stantial gap in Andean Colombia. The similarity in many scale characters of the three species is seen in Table 1 and is reinforced by the general similarity in habitus and size. A. chocorum, as shown in Tables | and 2, in many ways connects A. punctatus with A. transversalis. The oblique rows of dark blotches on the dorsum of some specimens (well shown in the type, Fig. 1) resemble the dorsal pattern of the giant anoles, A. frenatus, A. princeps, A. squamu- latus, A. latifrons. All of these have considerably higher scale counts in all respects than A. chocorum, and differ also in their much larger size (over 100 mm snout-vent length rather than 70-80 mm). They are, however, quite clearly closely related, and in mental shape, for example, are more like A. chocorum than are either A. punctatus or A. transversalis. These apparent affinities of A. chocorum in several directions seem to effectively subvert the recent attempt of the senior author to define a punctatus group from which A. transversalis was expressly excluded as a “fringe species” and from which the giant anoles were considered still more remote. This newly discovered anole clearly and firmly connects the so-called “fringe species” (and the giant anoles as well) with the punctatus group. Thus we seem to be faced with a single series of species which, how- ever, exhibit striking and peculiar trends in several adaptive lines. The presence of still surviving intermediates helps to connect up and to indicate the affinities of the more peculiar species. 8 BREVIORA TABLE 1 Scale Counts Anolis chocorum and Relatives No. 256 punctatus (including boulengeri) chocorum transversalis Scales across snout between second canthals 8-14 10-14 4-8 Scales between semicircles 0-2 1-3 0-1 Scales between interparietal and semicircles 0-4 2-4 0-3 Loreals 4-7 6-8 3-6 Labials to center of eye 6-10 8-10 6-9 Lamellae, 4th toe 23-30 17-20 22-27 TABLE 2 Qualitative Characters Anolis chocorum and Relatives punctatus (including boulengert) chocorum transversalis Snout Swollen in 2 Not swollen in Not swollen in either sex either sex Ventrals Smooth (punc- Smooth Smooth tatus) or keeled (boulengeri) Mental At least as deep Wider than deep Deeper than wide as wide Sublabials | = + (large) Dewlap (4 ) Moderate (lateral Large (lateral Large (lateral scales large, elon- scales small, scales small, gate, in close- weak, in multiple weak, in multiple packed rows ) rows widely sep- rows widely sep- arated by naked arated by naked skin) skin) Dewlap (2 ) Absent Large as in male Large as in male but differently but differently colored colored Dorsal pattern Not sexually Not sexually Strong sexual dichromatic dichromatic dichromatism 1967 NEW ANOLIS FROM PANAMA 9 Color in life of A. chocorum DEWLAP Edge anteriorly white, posteriorly orange Sides pale orange with several faint bluish white or green lines Base pale green DORSUM Green, uniform or with oblique rows of dark green blotches on flanks VENTER Green of dorsum continued on edge of belly. Center of belly light or blue gray. Both blue and green areas heavily spotted with white EYELIDS Edges yellow IRIS Pale brown (coppery) with- out dark markings TONGUE Yellow orange, sometimes with tip dark gray LINING OF THROAT Black Edge white Sides green with white lines or pale green Base pale yellow or gray As in As in 6, but center of belly not light in any specimen As in 4 As in 4, or reddish brown with gray periphery As in ¢, or pale flesh gray or pale greenish yellow or pale yellow As in REFERENCE CITED WILLIAMS, ERNEST E. 1965. South American Anolis (Sauria, Iguanidae): Two new species of the punctatus group. Breviora, No. 233:1-15. (Received 5 April, 1966.) No: 256 BREVIORA 10 ysuUa| JUSA-jnOUS [eNIOW ‘F696 AM ‘eddy Jo MOIA [vIIIeT tn eae eee : ee Sa PE ea Ef I ‘saroads Mou “LUNAOIOYI STOUR NEW ANOLIS FROM PANAMA Bi 1967 “UIW 6/ YSUd] JUSA-jnNOUS [eNIOy “PE696 AM ‘odAy spew eR. eet 1 a mam jo usoyed y[emus, CSS ort tetera a ‘sotoods MoU ‘wNsO20YI slouP ‘¢ Sl No. 256 BREVIORA ‘WIL Bf YISUa| JUDA-JNOUS [eENIDY “1€696 AM ‘edAzesred ojeway Jo usayjed esjuaA ‘soiseds Mou ‘wNsoI0Y9 soup ‘9 “SIF SSS a — eee BREVIORA Museum of Comparative Zoology CAMBRIDGE, MaASss. 3 FEBRUARY, 1967 NUMBER 257 A REVIEW OF THE CLARK FORK VERTEBRATE FAUNA By Roger C. Wood INTRODUCTION The Clark Fork fauna, hitherto regarded as the standard for one of the provincial ages of the North American Continental Ter- tiary (Wood ef al., 1941, p-. 9 and plate I), is known from only one area in the Bighorn Basin of northwestern Wyoming. In con- trast to other Paleocene faunas representing provincial ages, no Clarkforkian quarry or pocket has yet been discovered. Con- sequently, this fauna has never been as well characterized as others in the standard sequence, and the validity of this provincial age has not been very well substantiated. This paper reviews previ- ous work on the fauna and discusses some associated problems. Conclusions are based on field observations, a re-evaluation of the described collections, inspection of the original field notes, and unpublished annual reports of the Department of Vertebrate Paleontology of the American Museum of Natural History. Professor Glenn L. Jepsen suggested this problem as a senior thesis topic while I was an undergraduate at Princeton. His generosity in permitting me access to Princeton’s valuable collec- tion of Paleocene fossils is here gratefully recognized. Dr. Mal- colm C. McKenna of the American Museum of Natural History most kindly made available for study the specimens upon which the original Clarkforkian faunal description was based; I have also profited from discussions with him. I am much obliged to Professors Bryan Patterson, George Gaylord Simpson, Albert Wood, and Dr. Leigh Van Valen for criticisms and suggestions concerning this manuscript. In addition I am grateful to Dr. Giles MacIntyre, who found the map used by Walter Granger in the Bighorn Basin during the field seasons of 1910 through 1912. A portion of this map is reproduced in Figure 2. The hospitality of the Churchill family of Powell, Wyoming, added considerably to the enjoyment of 2 BREVIORA No: 257 doing field work during the summers of 1961 and 1962. I would also like to thank Mrs. Frances Wood and Miss Margo Hayes for typing the numerous revisions of this paper. Financial assistance provided from the John Boyd Fund of Princeton University enabled me to spend the latter part of the summer of 1961 and the entire field season of 1962 in the Big- horn Basin, working in the type areas of the Clark Fork fauna. Support from the National Science Foundation training grant for evolutionary biology at Harvard University was helpful during the final stages of my research. Abbreviations used in this paper are as follows: AMNH, American Museum of Natural History; PU, Princeton Univer- sity; FMNH, Field Museum of Natural History; USNM, United States National Museum. HISTORY OF STUDY During the summers of 1910 through 1913, and again in 1916, field parties sponsored by the American Museum of Natural His- tory conducted extensive investigations of the Tertiary sediments of the Bighorn Basin. Some of the results were published by Sinclair and Granger in 1912; of interest to the present study was the recognition of a new vertebrate fauna occurring near the top of the Fort Union." Specimens comprising the fauna were obtained from three different localities whose approximate posi- tions, determined from the following descriptions, are labelled A, B, and C on the accompanying map (Fig. 1). These areas were described (Sinclair and Granger, 1912, pp. 59-60) as being “. . . on the southwest slopes of McCulloch Peak . . . about a mile due east of the point where the Fort Union-Wasatch contact line crosses the Shoshone River, 245 feet stratigraphically below the contact with the red-banded beds [A]; on the north side of the Shoshone River in the bluffs opposite Ralston station [B]; to the northwest of Ralston on Big and Little Sand Coulee [C].” Only the McCulloch Peak locality (A) was regarded as being unquestionably Paleocene; some uncertainty was expressed con- cerning the exact stratigraphic position of the other two areas. This doubt led the authors to conclude “. . . we feel that further examination of the stratigraphy is desirable. Should the beds in question prove to be older than the Knight [early Eocene], and it be deemed advisable to give them a formation name, they may be referred to as the Ralston beds or Ralston formation.” 1 The Paleocene rocks in this area have been referred to as the Polecat Bench Formation; see Jepsen, 1940. CLARK FORK VERTEBRATE FAUNA 1967 sejiw GI | | | | +SUIOJUNOIW = YyjOOjuDaG «ayy Ul PUD UIDJUNOW 4ysDaH 4D pasodxa 9jJD Sy904 d10Z028|/Dg pud uDIyqwodeid = youeg j09a/0dq = d POOMIJIM = M Kuousajon® = © A, A3y OIHLNT -P®4I0||09 O10M S/I1S60} 4104 y4DID Whesnw uDdoWewWY kpoo A” ww zo1u | #0 Ajisofow ey}, YyoIyM WoO) sDe1D ed © puo © 0) ‘ pl6l Ul saBuni9 fq peqiosep saijy020; ya04 Y4DIDQ« |DUOIZIPPD 4O SuUO!yIsod eyoW!xoaddo pub @ + $01}]{D90] 4104 yad])D = jDUIBI1u0— & soBuDIO PUD 41D]}9UIS 40 SUO!4!80d eyDWIxosddd (©) puD ® @ $ [ Duojuow -€ @1nb\y ul pabsojue $8: seul) paysop Kq pepunoiins uoibey -uispg y404 y4D/D @Y} S8O1DD 49DJU0D 8U9905-aUGaI03/D4g 4O U01}090) Be,;OW!NXOAddD eul; pe}jog (S961) 2943!g PUD (J H61) 0/6403 puod Bd10e\4 SMOJpUyY Woss UeHDL 031931109 N338 3AVH SN3WIDSdS WHOS WYVID HOIHM WONS VWaY"V | 3YNdIS 49e19100g | a |] Nos 1 4 BREVIORA No. 257 A more precise definition of this fauna was subsequently pub- lished by Granger in 1914. Further collecting had increased the number of specimens to fragments “representing nearly 200 in- dividuals” (Granger, 1914, p. 204). Characterized by the absence of perissodactyls, artiodactyls, rodents and primates (the last two not then known from Paleocene deposits), the fauna was com- posed predominantly of representatives of two genera, Phena- codus and Ectocion, which constituted over three-fourths of the collection. The rest of the fauna included some Paleocene genera, whose range was known to extend into the overlying Eocene Gray Bull beds, and genera of Gray Bull age. Assignment of the fauna to an appropriate epoch was supposedly aided by the abun- dant presence of the reptilian genus Champsosaurus, which was at the time considered to be a distinctly Cretaceous and Paleocene genus. This evidence, coupled with the absence of the four orders of mammals previously mentioned, inclined Granger to regard these beds as being of late Paleocene rather than early Eocene age. Because the name “Ralston” was preoccupied by a group of Pennsylvanian rocks, Granger substituted in its place the term “Clark Fork” beds. These sparsely fossiliferous strata were esti- mated to attain a maximum thickness of 500 feet. Besides the McCulloch Peak locality (A), exposures were prominent “In the bluff [Polecat Bench] in the northern part of the Bighorn Basin” where “characteristic fossils were found as far east as a point north of Powell [i.e. in beds extending northeast from B].” In addition, two new localities were described: “In the Clark Fork basin the fossils were obtained from both sides of the wagon road where it drops down to the Big Sand Coulee from the Bighorn Basin divide, also from . . . the opposite side of Clark Fork River between the mouths of Line and Little Rocky Creeks” (bid., p- 204). These last areas, not mentioned in Sinclair and Granger’s paper, are marked on the map as localities D and E. Descriptions of the various new types ascribed to the Clark Fork fauna appeared in a series of papers by Matthew (1915a, b, c; 1918) and Granger (1915), and in one by Simpson (1929). It is interesting to note Simpson’s remark (p. 2) that some of the specimens being described might be from the lowest Eocene Sand Coulee horizon rather than from the Clark Fork. However, no reason was given for this statement. Not until 1930 was a complete faunal list published (Jepsen, 1930b, pp. 492-493), which included several species collected by Princeton University field parties subsequent to the American Museum’s initial discoveries. Three new forms were described, 1967 CLARK FORK VERTEBRATE FAUNA 5 the most significant for this study being Dipsalodon matthewi, the only genus restricted to the Clark Fork. In addition, note was made of the fact that Champsosaurus could no longer be used to demarcate the upper limit of Paleocene beds, thus invalidating Granger’s premise that this genus could be used for distinguishing between Paleocene and Eocene sediments. Revision of the Clark Fork fauna was undertaken by Simpson (1937). Several forms from the original collection were for the first time described. In addition, one species, Ambloctonus pris- cus Matthew (= Palaeonictis occidentalis Osborn; Denison, 1938, p. 175), was removed from the fauna on the grounds that field rec- ords associated with the specimen clearly indicated that it belonged with the Sand Coulee fauna. A further deletion from the fauna was made by Jepsen, who removed Parectypodus, the only multi- tuberculate that had been regarded as part of the fauna, because its original inclusion had resulted from “an error in locality identi- fication” (Jepsen, 1940, p. 324). No additions or deletions to the fauna have subsequently been published. INTERPRETATION OF MAPS Since Granger used only verbal descriptions, rather than sec- tion, township, and range data for indicating the locations of specimens and type areas, great value must be attached to the recent finding of his original field map, part of which is reproduced as Figure 2. Its discovery permits much greater confidence in the determination of some of the areas where the original Clark Fork collections were made. No date appears on the map, which was taken from Fisher’s 1906 report. Perusal of the field notes, how- ever, indicates that Granger was in this part of the Bighorn Basin only in 1911 and 1912. Therefore, it seems reasonable to infer that the hachures sketched on his map, undoubtedly intended to indicate badlands, correspond to areas prospected in those years. These badlands, consequently, must represent the type localities for the Clark Fork and Sand Couiee beds described by Sin- clair and Granger (1912) and Granger (1914). Some of these hachures coincide with the placement of letters B, C, D, and E on Figure 1, thus confirming the positions of most of Granger’s type lo- calities. Unfortunately, uncertainty must still remain as to the exact location of the area on the southwest slopes of McCulloch Peak (in- dicated approximately in Figure 1 by the letter A) because no hachures were drawn on this part of Granger’s map. Confusion may be averted by noting the following problems concerning the same name for different places on various maps. No. 257 BREVIORA \ S Buipuo < sojiw ¢ 2 | ° $00} QOS |[PAseju! unoyuos | : $30V1d WHSASS NI G3Y¥3LL13134 N35@ SVH ONILINM QNVYH S,YSONVYD ALINVID 30 3NWS SHL HOs *S3ILI1W901 337N0D ANVWS 3adAL SYSONVYD JO 3NO ANSS3Yd3Y SH33YD VYVH,O Ld ONY LNIVd 40 SHLNOW 3H1 N33SML3@ S3YNHOVH (-LX3L 33S *Q3GN10NI LON W3NV NV3d HOOTINOOW) - 26! Y3SGOL90 - YSEGW31d3S GNV 116: LSNONY ONIENA G319317109 3Y3M STISSOJ WHOS HYW1D 3Y3HM S3ILITVOO1 ONIMOHS dVW O79I4 NISV@ NYOHDIE SYSONVYD 4O LYVd JO NOILONGONdSY +: 2 3yYNDdIZ N9C 1 H 44019 Se) spaq pepuog pa Mojag N 2G) J CLARK FORK VERTEBRATE FAUNA 1967 Sa == (F961) 9940Id v-v-v (Lb61) 1D 49 SMespUuY x-x-x-x—x-x-x (861) MO1S eeseccooocesee (OE 6]) uesdep youeg 4099/0d = d POOMIIIM = Ai0usajonH = © ~----------- (9061) 4eus!4 Fox oust Sau!) joDju0D 40} hex ajDos say G v £ 2 ! (0) NISWS YYOsS WHV10 3HL SSOYNDY LOVLNOO 3N3003-3N30031Vd 3H1 40 NOILV9O) >€ 3YNSIS 8 BREVIORA No. 257 On Fisher’s map (1906; see Fig. 2), Bennett Creek is shown as a tributary of Little Rocky Creek. All subsequent maps, however, depict Little Rocky Creek as being a tributary of Bennett Creek (see Fig. 1). Furthermore, Fisher’s map indicates Little Sand Coulee as being a branch of Big Sand Coulee, but in the Annual Report of the Department of Vertebrate Paleontology for 1912, Granger remarks that “. . . What is known locally as Little Sand Coulee [is] not the Little Sand Coulee of Fisher’s map.” The actual location of the coulee is written on Granger’s field map (see Fig. 2), and later maps all follow Granger’s positioning of the coulee. A final problem is the question of where the Paleocene-Eocene contact in the Clark Fork Basin is situated. The first geological map of the area was published by Fisher in 1906. Presumably the contact there drawn (refer to Figure 3 for position of this and subsequent contact lines) represented only a rough estimate, as the line across the basin is unnaturally straight, and no discussion of the means by which the contact was detected are included in Fisher’s report. Sinclair and Granger (1912) apparently did not take Fisher’s contact very seriously, since, with the exception of the McCulloch Peak locality, the beds they describe as having yielded their presumably Paleocene Clarkforkian material are located several miles to the southwest of Fisher’s line, well into his Eocene, as can be seen by referring to Granger’s field map (Fig. 2). Their 1912 paper, however, includes a sketch map which shows a contact coinciding with that of Fisher’s map. In 1930 Jep- sen published a map whose Paleocene-Eocene contact lay several miles southwest of and parallel to that recognized by Fisher. The line was drawn more precisely on the basis of paleontological evi- dence, using the conventional criterion that the lowest Eocene beds are marked by the first appearance of Hyracotherium. Agree- ing relatively closely with Jepsen’s contact is one drawn by Stow (1938) on the basis of the presence or absence of certain types of heavy minerals that he thought permitted differentiation be- tween Paleocene and Eocene sediments. Although apparently un- aware of Jepsen’s and Stow’s works (as their papers are not indi- cated as being sources of data), Andrews, Pierce, and Eargle (1947) drew a contact closely paralleling these previous ones, but generally slightly to the northeast. For that part of the basin shown on the map, Pierce (1965) indicates minor deviations from his previously drawn contact except in the area directly east of the mouth of Little Sand Coulee, where his boundary diverges appreciably from the earlier one. Thus, the boundary between 1967 CLARK FORK VERTEBRATE FAUNA 9 Paleocene and Eocene beds has been reasonably clearly estab- lished within rather narrow limits. The area where the greatest divergence exists between the various lines, however, is to the east of the mouth of Little Sand Coulee, a fact which is of some importance concerning the provenance of about half the specimens in the Princeton collection (see p. 21). THE AMERICAN MUSEUM COLLECTION Data concerning the American Museum’s Clark Fork collection are summarized in Table I. One fact readily apparent from this table is that there are only 89 specimens in the original collection, not the “nearly 200” mentioned by Granger. This observation may indicate that many specimens, in addition to the one discussed by Simpson (1937; see above, p. 5), have been at one time or another removed from the fauna without this action having been noted in any publication. On the other hand, it may merely indi- cate that Granger had overestimated the size of the collection. At any rate, the basis available for defining and characterizing the type collection is less than half as large as published data indicate. An additional feature worthy of note is the nature of the speci- mens from the McCulloch Peak area (A, Fig. 1); these are described in Table I, footnote 10. The identifiable fragments in- clude no elements restricted to the Paleocene. This fact lends itself to some interesting speculation, as this locality is the only one of the three originally described by Sinclair and Granger as representing unquestionable Paleocene sediments. In this connec- tion some of Granger’s comments in the 1911 Annual Report of the Department of Vertebrate Paleontology are most informative: “These beds [at Ralston] proved to be of the same age as those [from the base of McCulloch Peak] described above and although fossils were scarce some SO individuals were obtained. The small phenacodonts constituted about 12 the entire number and there still was no trace of Perissodactyla, of Artiodactyla, nor Rodentia. The presence of Bathyopsis, Limnocyon, and a Metacheiromid were surprising, considering that the beds lie below the Wasatch... . The dip of the [Ralston] beds where exposed would carry them below the Wasatch at the base of McCulloch Peak, allowing no fault or change in dip.” In other words, Granger apparently con- sidered that the fauna contained in the beds north of Ralston included various forms that he would have expected to find in typical early Eocene sediments. Thus the Clark Fork beds were defined on the basis of negative evidence, i.e., what had not been RE EE EEE i I peynuepr Ajjeoytoads jou = I I smupniods "td wn I I suapyv] Qn I ¢ Lad UDA 2 Auoyisq = I C v rl SISUIUOIS]VA SNUDIUAOGSO UO1LI0JIA (6 suuunuapad snovssiq I -Maylpu uopojvsdig 9 Z 9 snajoid snuajosd syonuxpig I I ‘ds uopoyd&ksog I snpyiu snosiaydp (deur uo z e# Ayeooq) | (dew uo 7# AryR907) 2 UONBULIOJ uoj\s[ey jo Ya0ID PILH.O is Io ,‘porsed,, | ‘nN sap sary} eq JO ynow (dew uo L# AyesoT) (‘Aqeoo] yova Wo.jy sud i ‘spoq IO ‘N fuojsTey JO ‘JSbO sor ea[noD -loads Jo s9quinu 9y} a}vOIpUI JayjO | yIO- yaRjO JO AN yniq 2014} 10 sea pues 31g Jo peoy sweu yovea ayisoddo syesowny ) SHILITVOOT SHIOddS ‘(€r6l pure LE6]) uosdung pue (8€6]) UOsIUad Wo] Osje pue ‘sustIdads oY) YIM pojeIOOsse sjoqr| P[ey Wo Udye} UONeUTIOJU] NOID3ATIOD AYO WYVID WNASNW NVOISWV SHL | dav 10 | CLARK FORK VERTEBRATE FAUNA 1967 ‘snuajoad ‘q wioj [ng AviH oy} 0} a[qeviajo1 Bulog suaumoods {[e ‘payepleAur 9q 0} Wes p[nom saldadsqns [ng AviH pue «IJ YAPIO,, 9} usemjoq uoNouNsIp sty} ‘spoq [Ing Avi wo oe suowoads ,yIO{ YIRIO,, 94} JO WWOS jsKO] We g0UIg ‘so1dedsqns jouNsIPp B Jo JUOWYsI[qeiso oy} posmnbor yoy oy YoIyM ‘aX JO SJUOWOINseoW Woy ‘sOUDpP “IAD SO}ID BY “IDAOMOHZ ,,"UOISUdWIp AJOAD Ul [[Ng AIH 94) WoIJ as0Yy} Jo oduLI OY) UTYIIM ore sudUTIONds yIO4 YARD UL, wu (st ‘d ‘Z€61) paasosqo uosdung “(gy ‘LZ ‘dd) punoy o19M suaurtoeds yYSI0 I9yYI0 OY} spoq Yor UI SUTWII}AP 0} a[qQIssod jou st yy *(p} ‘d aas) spag fing Avery oy) uryUM Ajeoo] & ,f90;IN0D pues SIg Jo peoy ‘spoq YO YAe]O,, woz ose ‘addy oy Burpnyout ‘suaumoeds x1g ‘aay ATUO Jo GdUd}SIX9 OY) SUOTJUOW (QT ‘d ‘7¢6T) uosdung ysnoyjye ‘sardedsqns sty} 0) Sursuojeq sev paTjoqey a1e UOT}D9[09 YO YAe[O oy} ul suswoeds u99}1N0J , tl vV o IVLOL AN _-— = NY Net oN sunbiyuv uopoovid kay] osingavad sisdokyiogoig smignp sidppvisald 10y009 sidvpvisajd smipautdaqul "yd snaavuiad siaapuiid "yo snaapunad snpoovuay dg smojound sajaksoavjog snpnaiwd uopouvavvd sndkypjd puavkxe suapinbav puavkxQ uinoslid WN MaYysJoosua snuvisiads snjkuojdvy 12 BREVIORA INO 257 Footnotes for Table I (continued) Denison (1938, p. 175) states: “Amer. Mus. No. 16068 from the Clark Fork horizon, described by Matthew .. . as ‘?Oxyaena sp. innom.’, is referred here doubtfully to Dipsalodon. It includes much worn upper teeth of approximately the same size as D. matthewi.” Van Valen (1966, p. 82) has referred this specimen to D. matthewi. 3 AMNH No. 18668 has been tentatively referred to this genus, but specific identification is difficult because of the scrappy nature of the specimen. 4Simpson (1937) recognizes three, or possibly four, species of Esthonyx from the Clark Fork beds: E. bisulcatus, E. grangeri, E. latidens, and per- haps E. spatularius. AMNH No. 16065 is labelled as “E. cf. spatularius (type),” but is figured in Simpson (1937, p. 6, fig. 1) as E. ?bisulcatus. How- ever, this specimen could not be the type for either species since it was found in 1912, while the two species in question were described by Cope in 1874 and 1880, the type for E. spatularius being AMNH No. 4809 and that for E. bisulcatus being USNM No. 1103 (Gazin, 1953, pp. 17, 21). None of the specimens in the collection are presently labelled as representing E. bisulcatus. >Jepsen (1930b, p. 493) listed this species as being included in the Ameri- can Museum’s collection although Simpson (1937) does not include it in his Clark Fork faunal list. However no mention or reason for its deletion is given. Gazin (1962, p. 67) notes that “the type of M. priscum from the Clark Fork beds could not be located in the American Museum collections.” Personal investigation has confirmed this report. ® Denison (1938, p. 167) referred Matthew’s type of Dipsaladictis platypus (AMNH No. 15857) to this species, stating that “it is definitely an oxyaenine, which differs from other primitive species of Oxyaena only in its smaller size.” One additional specimen (AMNH No. 18667) is labelled as belonging to this species. ‘This specimen (AMNH No. 15850) was listed as “Nyctitheriidae genus and sp. indet.” by Simpson (1937, p. 4), but has been described as Palae- oryctes punctatus by Van Valen (1966, p. 56). SMany Phenacodus specimens are represented as disarticulated bones or miscellaneous tooth and jaw fragments rather than as associated sets of teeth or bones. Individual fossils were labelled variously as Phenacodus, P. primaevus, P. intermedius, and P. hemiconus, and thus do not concur with the two subspecies included in Simpson’s 1937 faunal list. ®Only two specimens in the collection are labelled as Probathyopsis prae- cursor. Four other specimens in the collection are labelled Bathyopsis or Bathyopsidae. The latter fossils were clearly intended by Simpson to be included in the new genus since one of them comes from opposite the mouth of Littlke Rocky Creek, one of the localities mentioned as an area where specimens of this species had been found, and it is in fact the only speci- men of any kind from this area. For another of the four localities listed 1967 CLARK FORK VERTEBRATE FAUNA 13 as having yielded P. praecursor remains, three miles east of the mouth of Pat O’Hara Creek (locality 2, Fig. 3), no Clark Fork specimens could be located. Simpson suggested (1929, p. 2) that “Some of this material may be from the Sand Coulee horizon.” Four specimens, including the type, are from locality 1 (Fig. 3), confirming the idea that at least some of the specimens are part of an early Eocene fauna rather than a late Paleo- cene Clark Fork one. In addition to the specimens listed in the chart, the following material is included in the collection: three boxes of “miscellaneous bones and teeth” from the west slope of McCulloch Peak, containing fragments identifiable only as Champsosaurus, Phenacodus, Ectocion, and a pantodont; four boxes of miscellaneous fragments labelled only as being from the “Clark Fork Basin,’ and another from “north of Ralston.” None of this fragmentary material would appear to affect the conclusions based on analysis of the better specimens in the collection, and hence has not been discussed. found in them, rather than on fossils actually contained in them which could be characterized as having a distinctive Paleocene aspect. Since the type fauna was principally found in strata whose position in the section was by no means certain, and since the fauna from the only site regarded by Granger as being definitely of Paleocene age had no obvious Paleocene affinities, it is not at all certain whether or not a Clark Fork fauna can in fact be distinguished. Examination of the labels accompanying the specimens in the American Museum’s Clark Fork collection reveals that over 85 per cent of the fossils, acquired in the years 1910-1916, were found at three sites. These localities were: (1) the head of Big Sand Coulee (Area D, Fig. 1); (2) east, or three miles east, of the mouth of Pat O’Hara Creek (Area C, Fig. 1); (3) north, or three miles north of Ralston; or bluff, or base of bluff, northeast of Ralston (Area B, Fig. 1). For convenience of reference, American Museum specimens from these areas will henceforth be designated merely as being from locality number 1, 2, or 3 (see Fig. 1). Locality number 1 is in the region characterized by Granger as the type area of the lowest Eocene Sand Coulee beds (1914, p- 205). “Near the head of Big Sand Coulee in the Clark Fork basin there is a series of about 200 feet of red-banded shales lying 14 BREVIORA No. 257 between the gray shales! of the Clark Fork and the Systemodon- bearing gray shales above which are referred to the Gray Bull beds. The outcrop of these red-banded beds extended, from the high bluff on the south side of the Coulee, in a northwesterly direc- tion across the basin for several miles.” Furthermore, in this same paper, Granger describes (p. 203) the Gray Bull beds and men- tions various areas where they outcrop. “In the Clark Fork basin they are exposed in the southwestern part along the heads of Big and Little Sand Coulees, where they are of a uniform gray color instead of the usual gray, red, and yellowish banding.” Reference to all the geologic maps of this part of the basin shows that the Paleocene-Eocene contact line crosses the southern branch of Big Sand Coulee approximately two miles below its head, and the western branch nearly five miles from its head. The eastern prong of the coulee, of which the one on the map is only one of many, is probably not that referred to on the labels, since its head originates at the base of Polecat Bench and such a fact would have made a convenient reference for the labels. As Granger noted the misidentification of Little Sand Coulee on his field map (refer to p. 8 and Fig. 2), he may have been talking about either the southern or western branches of Big Sand Coulee. However, whichever of these two is chosen, the locality would be well into Eocene Gray Bull or Sand Coulee strata. In view of the fact that Granger’s field map indicates an extensive area of prospected exposures at the head of the southern branch of Big Sand Coulee, and none are marked near the head of its western fork, it is most probable that this area is the one in question. There is no doubt at present that these are Gray Bull beds. In addition, many specimens in the American Museum collec- tions labelled as “Wasatch,” “Sand Coulee,” or “Gray Bull” are noted also as having come from “the head of Big Sand Coulee.” For example, the label accompanying the type of Apheliscus nitidus (AMNH No. 15849, listed in Table I) reads “Clark Fork forma- tion, Clark Fork, Wyoming, 1911, head of Big Sand Coulee.” That accompanying another specimen of the species, AMNH No. 16925, reads “Wasatch formation, Sand Coulee Beds, Clark Fork, ‘Jepsen has pointed out that it is impossible to use color banding for distinguishing Clark Fork from overlying beds, since “. . . red strata occur well down in the Clark Fork beds” (1940, p. 237; see also 1930b, p. 493). 1967 CLARK FORK VERTEBRATE FAUNA £5 Wyoming, 1913, head of Big Sand Coulee.” All other specimens of A. nitidus are similarly labelled as being from Gray Bull beds. It would therefore appear that the two cited specimens, although labelled as being from two different horizons, are most probably from the same one, the Gray Bull. Consequently, based on the evidence revealed by the specimen labels, it is probable that all supposed Clark Fork fossils labelled as having been found at the “head of Big Sand Coulee” should more properly be referred to the lower Gray Bull or Sand Coulee beds. Since half of Sinclair and Granger’s original Clark Fork fauna (see Table I) was obtained from the vicinity of the head of the Coulee, it is likely that these specimens are part of one or more early Eocene faunas rather than of a late Paleocene Clark Fork fauna. “East [or three miles east] of the mouth of Pat O’Hara Creek” (locality 2) is an area lying in an extensive series of Willwood exposures. This locality, as is true for locality 1 also, was never mentioned by Granger as one of the five type areas (A through E, Fig. 1) for the Clark Fork beds. Granger’s field map (Fig. 2) shows two sets of badlands sketched along a north-south axis in this region, the southern one being labelled “red-banded beds” and the northern one “below red-banded beds.” These may be interpreted in the light of Granger’s remarks in the 1913 Annual Report of the Department of Vertebrate Paleontology: “Section across Clark Fork Basin from head of Little Sand Coulee northeast to mouth of Big Sand Coulee: Wasatch — gray shales — 500+ feet | Sand Coulee] — red-banded shales — 200 feet Ralston — gray shales — 500+ feet.” Clearly, the red-banded beds must represent some of Granger’s Sand Coulee exposures. Consequently, the beds farther north can only be regarded as Clark Fork beds, and it seems reasonable to assume that locality 2 corresponds to this area, which has always been mapped as being covered by Eocene beds (Fig. 3). It seems curious that part of a fauna purported to be of latest Paleocene age should be derived from sediments whose presumed early Eocene age has never been questioned. With respect to this para- dox, an examination of the specimens from locality 2 is most instructive. Reference to Table II reveals that all of the Clark Fork fossils from locality 2 are also known from the Sand Coulee or Gray Bull, and often from both of these horizons. In fact, the one specimen of Phenacodus hemiconus from this locality was apparently regarded by Granger as being of Gray Bull provenance, (gd ‘ectel ‘moyney) [Ing Aeip ‘aag]NOD pues ‘yO yare[O (pse “d ‘(ST6I ‘“lesueIH) [Ing AvIH ‘QgINOD pues ‘yIO4 YAeIO (oe “d ‘S161 ‘ledueInH) [Ing AvIH (or ‘d ‘L¢61 ‘uosdumg) sg[NOD pues ‘yoy yre[D (6 ‘d ‘L¢61 ‘uosdung) s9[NOD pues ‘yoy yAe[D ‘osuel S}I jnoqe pres aq uv Suryjou sny} ‘peyn -uapt Aypeoyioads JON (ese vd ‘SI6I ‘Jasueip) [[ng Aeipn “2a]NOD pues ‘yx1O4 YIeIO (zed “SI6L ‘losueID) [Ing Avein ‘aa[NOD pues *yIO4 Yeo No:257) BREVIORA NMON® SI SHIOddS AHL HOIHM WOdd NISVd NAO MAVIO AHL NI SNOZIYOH 9LO9T ‘SLO9T peiaquinuun fT "6S09T ‘PSO9T 9SO9T 99091 L6S6E peloquinuun | 19091 “9P09T LyO9T (HNWYV) SNHWIOddS dads snnbyuv uopoonjd kay I snaavuiid snpoasvuayd snuognuay snpospuayd suapyy] xKuoyssy 1adsupss xKUOYIST u01d0JIq SISUIUOJS]DA UOIIOJID SNUDIULOGSOg UO1I0JIQ (s[Issoy 94} YIM payeloosse sjaqey oy} Woly udye} o1e suOT}ROyNUAPy ) SdIOddS < 7¢ ALITVOOT WOW NOLLOATIOO WHOA WAVIO WNASAW NVOIMAWV AHL NI SNAWIOddS I AlaVL 1967 CLARK FORK VERTEBRATE FAUNA Ly as he states (1915, p. 339): “Eleven specimens from the Gray Bull beds and one from the Lost Cabin are referable to this small variety.” Thus there is nothing distinctive about the specimens from this locality to suggest their association with the Clark Fork fauna; on the contrary, based on the presence of Phenacodus hemiconus, there is a slight suggestion that these specimens may actually be of Gray Bull derivation. At any rate, since this area was not described as one of Granger’s type localities, and since its faunule does not differ significantly from those of other Sand Coulee or lower Gray Bull sites, it seems highly probable that “east” or “three miles east of Pat O’Hara Creek” is in Sand Coulee or lower Gray Bull, and not Clark Fork, strata. Specimens from locality 3 are too vaguely labelled to be of any value in this study. Faunas from the southeastern side of Polecat Bench have been found ranging from the Silver Coulee through the lower Gray Bull. Without knowledge of section, township, and range, it is virtually impossible to infer to what fauna these Ameri- can Museum fossils belong. It is very likely that elements from several faunas are actually represented by these fossils. There- fore, on the basis of the foregoing inferences, none of the Ameri- can Museum materials from localities 1, 2, or 3 can certainly be ascribed to what have been regarded as Clarkforkian strata. Rather, those from localities 1 and 2 can be considered as per- taining in all probability to one or more early Eocene faunas, while specimens from locality 3 cannot be definitely assigned either to a late Paleocene or to an early Eocene fauna. No specimens in the collection are labelled as having come from either side of the wagon road dropping into Sand Coulee Basin (area D on map). However, any such specimens would definitely be of early Eocene age. Listed in Table I is another group of specimens that may be considered as a unit for the purposes of this review. In contrast to the previous three groups discussed, which were character- ized by similarities in locality data, this fourth group is distin- guished by a total lack of useful locality information. Labelled variously as coming from “Clark Fork beds” or “formation,” or even “Clark Fork Period,” these fossils clearly can be of no use in defining the composition of the fauna for the following reasons. First, as has already been shown, if these fossils came from locali- ties 1, 2, or 3, they probably should not be included in the Clark Fork fauna. Second, if these specimens were not found at locali- ties 1, 2, or 3, then they could have been picked up anywhere by 18 BREVIORA No. 257 chance in a basin where four faunal horizons (Silver Coulee, Clark Fork, Sand Coulee, and Gray Bull) have been described, as the labels associated with these specimens give no guidance as to the specific area in which they were collected. The five remaining specimens from the type collection, not excluded from being Clarkforkian by virtue of the considerations already discussed, are listed in Table I under the column headed “other.” Of the fossils in this category, only two specimens, one of Coryphodon sp., and one of Ectocion osbornianus ralstonensis, were found at one of Granger’s type areas, “between Little Rocky and Line Creeks,” marked as locality E on the map. The hachures in section 3, T 57 N, R 102 W of Granger’s field map probably represent the exposures in which these specimens were collected. Andrews, Pierce, and Eargle (1947) showed both Paleocene and Eocene sediments in this region. However, Pierce (1965) indi- cates that only Eocene and Quaternary deposits are found within this section. Very probably, therefore, the two specimens under consideration should be associated with the Sand Coulee or Gray Bull faunas. Also in this last category is one specimen of Phenacodus pri- maevus that was found “6 miles north and one mile east of Powell, Wyoming, at base of bluff,” unquestionably an area of Paleocene deposits. This specimen may be regarded as coming from the Poiecat Bench Formation. Accompanying a single specimen of Probathyopsis praecursor (AMNH 16063) are the following locality data: “southeast side of Clark’s Fork, opposite mouth of Little Rocky Creek, Ralston [Clark Fork] formation.” The most recent maps show Little Rocky Creek as being a tributary of Bennett Creek, and the sedi- ments opposite its mouth, on the west side of the Clarks (or Clark) Fork River are mapped as Quaternary in age. However, Fisher’s map shows Bennett Creek as being a tributary of Little Rocky Creek (see p. 8), and the rocks opposite the latter creek’s mouth on the east bank of the Clarks Fork River were mapped as being of Eocene age. Therefore, it would seem that the locality designated by the label refers to the area opposite the mouth of Bennett Creek as its course is presently mapped. Hachures in this vicinity on Granger’s field map (Fig. 2) would appear to confirm this deduction. Excluding the Quaternary deposits immediately to the east of the Clarks Fork channel, Jepsen (1930b), Stow (1938), Andrews, Pierce, and Eargle (1947), and Pierce (1965) have all regarded the sediments of this area as being of Paleocene age. In this review, therefore, this one specimen of Probathyopsis will be regarded as being from upper Paleocene beds. 1967 CLARK FORK VERTEBRATE FAUNA 19 In the case of the only fossil not yet discussed, the locality data are relatively specific. The type of Oxyaena aequidens is from “Clark Fork beds, Clark Fork Basin, 1912, about seven miles east of Pat O’Hara Creek.” Rocks in this part of the Clark Fork Basin are extremely flat-lying, and it is rather difficult to establish the contact between the conformable Paleocene and Eocene strata, as can be seen from the fact that the contact lines drawn by different mappers differ markedly one from another (see Fig. 3). The Oxyaena specimen was probably found somewhere in this contact region. However, considering both the fact that the distance esti- mate made in 1912 could have been at best only an approxima- tion, and that there is no agreement as to exactly where the contact lies in this region anyway, it can only be a matter of speculation as to whether this specimen belongs to a late Paleocene or an early Eocene fauna. Thus, the specimen is of no help in defining the Clark Fork fauna. Of the original Clark Fork collection of 89 specimens, therefore, only two, one of Phenacodus primaevus and one of Probathyopsis praecursor, may be certainly regarded as being from the Polecat Bench Formation. The remainder of the specimens must be ex- cluded from this category for one of two reasons. Either the fos- sils are of a different age from what they were originally con- sidered to be, or the locality data associated with the specimens are not specific enough to be of any stratigraphic utility. As it would be unwarranted to try to define a fauna on the basis of only two species, each represented by only one specimen, it is necessary to analyze other evidence pertaining to this problem. THE PRINCETON UNIVERSITY COLLECTION Comparison of Tables I and III reveals several marked dif- ferences between the Princeton and the American Museum Clark Fork collections. First, there are only about one-fifth as many specimens in the Princeton collection as in that of the American Museum. Second, there are somewhat better documented locality data for the Princeton material. Still another contrasting factor is that the preponderance of the American Museum specimens was collected during the two field seasons of 1910 and 1911, while Princeton parties have been adding to the collection continuously for over 30 years. Only genera added to Clark Fork fauna by Jepsen in 1930 (with the exception of Parectypodus, see above, p. 5) will be considered in the present review, as these elements are the ones that have formed part of the basis on which the fauna has usually been defined. a a aaa aaa aa MT 6 |: (9LT-pLI “aa ‘ep6l ‘ZZ-61 ‘dd ‘L¢6]) snussd oy) Jo uONeN{RAd-a1 suOsdUTIg YIM 9018 0} posayfe udeq sey soy posn AdojouTW -19} OY, ‘SISUAUOJS]DA U0190JIY SB VUOSIJOI Uo}OdUIIg oY} UI suoUMIdeds oy} poynuepl (¢6p “d ‘qocEI) Uesdoy , ‘TOG Ul UOTDa][OO AjIsIOATUG, UOJDDUTIg 94} UI ~ a s]oqry oy) isureSe poyooyo osm pur ‘(Q96]) SUOUUIS pu ‘(qQEg6]) Uesdof Woy UdyR} dIe SUONROYyNUAEP! sYIDAdg , (e) Z 8I ‘SNHWIOddS SO AHENON IVLOL (sidvpvisajd “€6ZEL ‘ON SB Bore [eI9UNd oUKS) YINOUL s}I IVIU daNOD puLE e]}}IT Jo 3sva SLEEI ‘ds sisdoxyjnqotg dgfnoD pues oI} JO yYJNoW jo jsva SOEEI dgJNOD pues aI} JO yNoW jo jsvo LO€ET M IOLU‘N LS L ‘ze Ss (ed4t) c6zEl 1ay009 sidppvisald M 66 UN 9S L ‘8S v66rl ag[nog pues s[}IT JO yJNou saoqe oyrw T ‘xoidde SIEEl snaapunsd snpoopuayd < Fs da[NOD purg aI JO you Jo sea pZEEl ‘ds xXuoyssq > (49[NH $,191804) M 96 UN PS L ‘TIS TL8¢1 es d9[NOD purg aT JO yINow Jo ysea = ZI EET ma a9[NoD pues a1] JO YINoW Jo ysva OLEEl -SISUIUOJS]DVA uonipedxe gz6] ‘¢ dues MS aw 9IZEL SNUDIULOGSO U0190JIq a9g[NOD pues aI Jo yJNow Jo ysvo 60EEI da[nNoD purs aI] JO yNoW Jo jsva S6CEI snyunuavad snovssig s9[NOD pues HVT JO YNou Jo 3svo [BNE tS)! M 101 U‘N 9S L ‘pt S (edA1) ZSTET ImayspUL Uopojpsdiq BURJUOJ “OD UOGIeD ‘YooID IvIg JO AAS SOIT ¢ OOrET sniajoid uopoydkiog (YaIND s,19}804) M 96 UN S$ L ‘ZS eSsrl dajnoD pues aI JO yYynow 1eou PSTEl smignp sajsajodavg NAHWIOddS HLIM GALVIOOSSV ‘ON Nd SHIOddS Ss VLVd ALITVOOT Q 1GHdSVd@ NHHad AAVH NOILOATIOO WAOA WAV TO NOLAONTYd AHL JO SLYOdda GHAHSITANd HOIHM NOdN SNAWIOddS Wl JIgVvL 1967 CLARK FORK VERTEBRATE FAUNA Das Over one-half of the specimens in the Princeton collection were found in an area designated on the labels as “east of |or near] the mouth of Little Sand Coulee” (see Table III). How restricted an area this description represents is a moot point since the topog- raphy of this region offers few outstanding features to serve as convenient reference points and, as noted in the previous section, various mappers have differed in their opinions of exactly where the contact between the Paleocene and Eocene should be drawn in this region. Consequently, these fossils may represent strictly Paleocene forms, strictly Eocene forms, or some of both ages. In this case, not only a lack of accurate locality data, but also a lack of agreement as to where the contact is, hinder any useful evalua- tion of this part of the collection with respect to defining the Clark Fork fauna. The remainder of the specimens in the Princeton collection represent isolated finds scattered across a broad region. Nearly fifty miles separate the farthest spaced of the specimens attributed to the Clark Fork. Representatives of two species, Ectocion osbornianus ralsto- nensis and Phenacodus primaevus, comprise roughly one-third of the Princeton Clark Fork collection. Both of these species are found in overlying faunal horizons. in addition, the presence of Phenacodus sp. and Ectocion sp. has been reported from the Silver Coulee faunal horizon immediately below the Clark Fork beds (Jepsen, 1930b, p. 491). Reference of most of the American Museum specimens, previously regarded as being of Clarkforkian age, to one or more of the early Eocene faunas was shown in the previous section to be virtually certain. Some aspects of the phylogeny of these species, originally reconstructed from studies of American Museum materials, have thus been deduced on the basis of erroneous stratigraphic assumptions. Restudy is required before they can be used in the definition of any fauna. Conse- quently, specimens of Phenacodus and Ectocion are at present of no value in helping to define a characteristic Clark Fork fauna. The specimen originally designated as Coryphodon (species undetermined) by Jepsen (1930b, p. 493) has been described by Simons (1960, p. 13) as Coryphodon proterus. Only the type can be assuredly ascribed to this species. Definite inclusion of this specimen in the Clark Fork fauna is not possible. It was found at the foot of the Beartooth Mountains where unconformities between Tertiary and Paleozoic rocks prevent accurate determi- nation of stratigraphic position. No other fossils were found asso- ciated with Coryphodon proterus, so that no faunal clues are available to help make stratigraphic inferences. In fact, the only 22 BREVIORA No, 3257 reason for referring this specimen to a Paleocene horizon is that it was found in rocks mapped, presumably on non-paleontological grounds, as being of Paleocene age. Whether C. proterus is of Paleocene or Eocene age cannot be determined on the basis of available evidence. Simuarly, the one specimen of Esthonyx sp. known from the Princeton collection must be excluded from further con- sideration as an element of the Clark Fork fauna, since it is im- possible to determine whether the specimen came from Paleocene or Eocene beds (see Table III, text p. 21). However, since all the American Museum Clarkforkian representatives of this genus have been shown to be most probably of Lower Eocene derivation, it seems reasonable to assume, contrary evidence being lacking, that the Princeton Esthonyx is likely to be of the same age. A single specimen recovered from the sediments “east of the mouth of Little Sand Coulee” was identified by Jepsen (1930b, p. 493) as Probathyopsis sp. Since its stratigraphic position re- mains in question (see above p. 21), it is unsuitable in helping to define any fauna. For the same reason, the two specimens of Dissacus praenuntius listed in Table III can not be used to dis- tinguish between different faunal horizons. The specimens of Carpolestes dubius attributed to the Clark Fork horizon have been recovered from two widely spaced locali- ties. Although the age of the rocks at one of these localities is problematical, the other site is undoubtedly of Paleocene age. However, Carpolestes dubius is known from horizons both above and below the Clark Fork level (Jepsen, 1930b, p. 491; Van Houten, 1945, p. 450), the type being from the Silver Coulee faunal horizon. Thus this species is so wide-ranging through time that its presence is of no use as an indication of Clarkforkian age. There is no reason to question that the type specimen of Dip- salodon matthewi was found in late Paleocene sediments. Unfor- tunately, the only other specimen was collected in the problematical region “east of the mouth of Little Sand Coulee.” However, Pat- terson (personal communication) informs me that a specimen of ?Dipsalodon sp. is present in the Plateau Valley fauna of Tif- fanian age. Patterson’s specimen (FMNH No. P26095) is a frag- ment of the right maxilla with P*, the broken stump of M', and the root of M’. It can be compared with AMNH No. 16068, con- sisting of P* and M!, which Denison (1938, p. 175) has doubt- fully referred to D. matthewi. (This specimen is listed as an oxyaenid in Table I.) But since Jepsen’s two specimens of D. 1967 CLARK FORK VERTEBRATE FAUNA 23 matthewi are both represented by lower teeth only, no direct com- parison can be made between them and Patterson’s specimen. (PU No. 13152 (type) has P,4, M:2; PU No. 13311 is an M,.) Patterson’s manuscript notes state that his “specimen is smaller than the type of D. matthewi and AMNH No. 16068. It probably represents a new species.” In spite of not being able to make direct comparisons between Patterson’s and Jepsen’s specimens, however, it is possible to say that they are roughly of the same size and morphology, as might be expected in forms related at the generic level. Consequently, Dipsalodon can no longer be regarded as the only genus restricted to the Clark Fork fauna. Precise locality data are associated with only one of the sup- posedly Clarkforkian representatives of Plesiadapis cookei (Table III). But this specimen, the type, was found in the region east of the mouth of Little Sand Coulee where disagreements exist con- cerning the location of the contact between Paleocene and Eocene sediments. The specimen’s uncertain stratigraphic derivation pre- cludes its being clearly assigned either a Paleocene or an Eocene age. The two other specimens of P. cookei were also found in the same general area, and the same uncertainty applies to them. Thus, on the basis of published evidence it is not possible to include P. cookei definitely in the Clark Fork fauna. As already stated, over half of the Princeton collection must be eliminated from consideration as elements of the Clark Fork fauna solely on the basis of insufficiently precise locality data. Because of wide-ranging stratigraphic distribution and a lack of reliable knowledge regarding phyletic relationships, other specimens are similarly not suitable for defining the fauna. Only two described specimens in the Princeton collection, the type of Dipsalodon matthewi and one of Carpolestes dubius, may still be considered as having certainly come from the latest Paleocene strata in this area. Together with one specimen of Phenacodus primaevus and one of Probathyopsis praecursor in the American Museum collec- tion, these fossils represent the only certain components of what has been described as the Clark Fork fauna. As previously men- tioned, however, Phenacodus is at present of no use for fine age discriminations between rocks in the mid-Paleocene to early Eocene time span. The single specimen of Probathyopsis prae- cursor can hardly be diagnostic of a latest Paleocene faunal assem- blage when all the other described specimens of this species are probably of early Eocene age. Carpolestes dubius is already known to range from Silver Coulee to Gray Bull. Since nothing 24 BREVIORA No. 257 is known about the rates of deposition of the Polecat Bench sedi- ments, it is possible that the “Clark Fork” specimens of Phena- codus primaevus, Dipsalodon matthewi, Carpolestes dubius and Probathyopsis praecursor may be either essentially contemporane- ous with or of a younger age than the fossils in the Silver Coulee fauna. But until a larger number of fossils from the latest Paleo- cene sediments are found, it is not possible in practice to differen- tiate a Clark Fork fauna from those of under- and overlying strata. At any rate, to base one of the provincial ages in the standard section of the North American Continental Tertiary on a “fauna” consisting of only four specimens representing four species seems unwarranted. As a direct consequence there also do not seem to be sufficient grounds for continuing to regard the Clark Fork as a discrete uppermost Paleocene faunal zone or as a member of the Polecat Bench Formation. 1 SOME RELATED PROBLEMS If it is accepted that a definitive Clark Fork fauna is not a reality, then a number of questions arise. For example, most of the American Museum material of Probathyopsis praecursor, in- cluding the type, clearly is from lower Gray Bull beds. In addi- tion, there is no certainty as to whether the one specimen of Probathyopsis sp. was found in Paleocene or Eocene deposits. These facts complicate the evolutionary sequence proposed by Jepsen (1930a, p. 129) of (1) P. sp., (2) P. praecursor, and (3) P. successor. The evidence strongly suggests that P. praecursor and P. successor actually existed contemporaneously rather than that the former gave rise to the latter. Furthermore, it is entirely possible that P. sp. might also have coexisted with the other two forms. Wheeler (1961, p. 21) states that P. newbilli from the Plateau Valley beds “is quite distinct from both P. praecursor and P. successor and is not closer to one than to the other.” Such a statement suggests that P. newbilli could be ancestral to the two later, divergent forms. However, this scheme would not clarify the relationship of P. sp. to the other forms with which it may have been contemporaneous. Jepsen (1930a, p. 129) felt that P. ‘The possibility of course exists that a post-Tiffanian Paleocene provin- cial age may be recognized in the future. In that event the age would require a new name. “Clarkforkian” could be resurrected only if a real fauna of such age were to be found in the Clark Fork area. In view of the intensive search for one that has been made there over the past thirty-five years I regard this possibility as most unlikely. 1967 CLARK FORK VERTEBRATE FAUNA 2S sp. is more similar to P. praecursor than to P. successor, but until there is better stratigraphic control it will be difficult to resolve the problem of Probathyopsis taxonomy. The classification of Ectocion has received attention from Granger, Simpson, and McKenna. From the material available to him, Granger (1915, pp. 348-354) described four species: E. parvus, E. ralstonensis, E. osbornianus, and E. superstes. Simpson (1937, pp. 19-22; 1943, pp. 174-176), however, felt that Granger’s four species should be regarded as successive subspecies grouped under the specific name E. osbornianus. More recently, McKenna (1960, pp. 102-103) discussed Ectocion classification in his paper on the Four Mile fauna. His remark (p. 103) that “the locality data of practically all existing collections of Willwood mammals are inadequate for detailed stratigraphic (and hence evolutionary ) analysis” is equally applicable to specimens attributed to the Clark Fork fauna. But neither Simpson’s nor McKenna’s proposed clas- sification is clearly acceptable because of the probability that most of the American Museum’s Clarkforkian specimens are of Eocene rather than Paleocene age. E. osbornianus ralstonensis, for exam- ple, presumably existed contemporaneously with E. 0. complens, formerly thought to be the evolutionary successor of this species. Restudy of early representatives of the genus Phenacodus will also be required. Two species, P. primaevus and P. intermedius, whose ranges were listed as extending down into Clark Fork beds, were described by Granger (1915, pp. 337, 342). Subsequently Jepsen (1930b, p. 491) reported the presence of P. sp. in his Silver Coulee fauna. Phenacodus primaevus grangeri is also known from the Plateau Valley fauna (Patterson, personal com- munication). Simpson (1937, pp. 17-19) published a statistical analysis of the lower Gray Bull and Clark Fork phenacodonts in the American Museum collection. He thought that possibly the Clark Fork and some of the Gray Bull specimens represented a single species. Further, he felt that perhaps the two groups might be differentiated on a subspecific level, concluding (p> 19) sthat the Clark Fork specimens are a nearly or quite homogeneous sample of one subspecies and that this subspecies may also occur in the Gray Bull . . . or may be distinct. These alternatives cannot at present be adequately checked.” In view of the probability that most of the supposed Clarkforkian phenacodonts were actually found in Gray Bull beds, it would not be surprising if the two samples represented only one species. Conceivably the type and only known specimen of Plesiadapis dubius (AMNH No. 16073) could have come from either late Paleocene or early Eocene strata, having been found at locality 26 BREVIORA No. 257 number 3. In appearance it most closely resembles P. fodinatus, originally described (Jepsen, 1930b, p. 515) from the late Paleo- cene Silver Coulee horizon. The character separating these two species was noted by Jepsen as follows: “The anterior crest of P, is developed into a minute cuspule, and a small ridge appears on the anterolingual part of the protoconid, quite in contrast to the distinct metaconid on P, of P. dubius.” Unfortunately, P4 of the type of P. fodinatus (PU No. 13278) has been broken off and is now missing, so that it is no longer possible to make the compari- son except from Jepsen’s illustrations. The abundant quarry ma- terial of P. fodinatus, most of which has been obtained since 1930, reveals, however, that the presence or absence of a distinct meta- conid is a variable feature and not a specific character. That only one specimen of P. dubius has ever been found seems curious, in view of the rather extensive amount of field work in the upper Paleocene and lower Eocene beds of this region. Since the range of P. fodinatus has been, subsequent to 1930, extended up into the lower Gray Bull by continued collecting, it appears entirely possible that P. fodinatus is a synonym of P. dubius. The Sand Coulee beds were defined by Granger (1914, p. 205) as follows: “This horizon does not contain Systemodon but it does contain the Perissodactyl genus Eohippus in abundance as well as Artiodactyls, Rodents, and Primates, and marks the first ap- pearance of these four orders. It also marks the last appearance of the primitive order Multituberculata which is represented in this horizon by a genus of Plagiaulacids.” Many of the criteria used to characterize this faunal zone have in the past fifty years been shown not to be definitive. Rodents and primates were later found in the Paleocene. Granger’s “plagiaulacids,” now classified as ptilodonts, are known from levels higher than the Sand Coulee. ! When Homogalax (Systemodon) was found in the Sand Coulee beds subsequent to Granger’s definition of them, Jepsen (1930a, p. 119) suggested suppression of this term and included the beds ' However, multituberculates are very rare in Gray Bull and later deposits. Van Houten (1945, p. 448) reports one species, Ectypodus simpsoni, extend- ing into middle Gray Bull beds. Robinson, Black, and Dawson (1964, p. 810) have described a multituberculate from upper Eocene deposits near Badwater, Wyoming. However, their specimens, representing “an undescribed species related to Ectypodus hazeni, . . . comprise a small per- centage of the fauna, certainly less than 1 percent . . .” These two cases are the only known occurrences of multituberculates above the Sand Coulee strata. 1967 CLARK FORK VERTEBRATE FAUNA 27 in the Gray Bull. Simpson (1937, p. 1), however, noted that “the Sand Coulee fauna was not wholly defined . . . on the absence of Homogalax but also by the generally slightly less advanced character of its mammals.” Now that most of the American Muse- um’s Clarkforkian material can be probably relegated to lower Eo- cene beds, it may be possible to redefine a Sand Coulee fauna. For example, this might be done on the basis of the presence of Plesiadapis cookei and a species of Probathyopsis, the last abun- dant appearance of the multituberculates, and a more primitive aspect in those forms common to both the Sand Coulee and Gray Bull. Reappraisal of the Sand Coulee fauna is desirable in order to determine conclusively whether or not such a horizon can be distinguished. SUMMARY Some of the published data concerning the American Museum’s Clark Fork collection do not agree with the locality information on specimen labels. For example, there are no fossils from either side of the road descending into the Sand Coulee Basin from the Bighorn Basin divide (area D), although it was described as one of the Clarkforkian type localities. The majority of the American Museum collection comes from three localities (1, 2, and 3, Fig. 3). Review shows that it is doubtful that most of these specimens from localities 1 and 2 should be referred to a late Paleocene Clark Fork horizon, but that they are rather of early Eocene age. Other specimens (those from locality 3 or labelled as being from Clark Fork “beds,” “period,” or “formation”) are associated with such vague stratigraphic information that it is not possible to determine whether they are from upper Paleocene or lower Eocene beds. Because of insufficient locality data it is also impossible to say whether or not approximately one-half of the described Princeton collection belongs to a late Paleocene or an early Eocene fauna. A wide ranging stratigraphic distribution or a lack of adequate knowiedge regarding phylogenetic relationships precludes using most of the remaining fossils to characterize the fauna. Thus, of 107 specimens (representing at least 23 and perhaps 26 species) in both the American Museum and Princeton collec- tions that have been described as elements of the Clark Fork fauna, a total of only four specimens representing four species (Carpolestes dubius, Phenacodus primaevus, Probathyopsis prae- cursor, and Dipsalodon matthewi) can be surely ascribed to a possible latest Paleocene faunal horizon in the Polecat Bench Formation. However, only one of these species (Dipsalodon 28 BREVIORA INO257 matthewi) may be unique to what has been described as the Clark Fork fauna. Carpolestes dubius is known from both the Silver Coulee and the Gray Bull. Specimens of Phenacodus primaevus are known from the Gray Buil, and Phenacodus sp. is present in the Silver Coulee. Until the phylogeny of this genus is reviewed, it is not possible to use species of Phenacodus for horizon markers. Probathyopsis praecursor is much more commonly found in early Eocene than late Paleocene beds. Thus this species can hardly be considered as diagnostic of the Clark Fork fauna. If ?Dipsal- odon sp. from the Tiffanian Plateau Valley deposits of Colorado is correctly referable to this genus, then it invalidates the supposed restriction of this genus to the Clark Fork. Therefore, the Clark Fork “fauna” consists of four specimens referable to four mammalian genera, each from a different locality. None of the genera (and not more than one of the species) are restricted to a lithically distinguishable Clark Fork horizon. Such evidence scarcely warrants recognition of the Clark Fork as a pro- vincial age, faunal zone, or member of the Polecat Bench Formation. LITERATURE CITED ANDREWS, D. A., W. G. PIERCE, AND D. H. EARGLE. 1947. Geologic map of the Bighorn Basin, Wyoming and Montana, showing terrace deposits and physiographic features. Oil and Gas Investigations Preliminary Map 71. Core, ED: 1874. Report on the vertebrate fossils discovered in New Mexico, with descriptions of new species. Extract from appendix FF of the Annual Report of the Chief of Engineers, 1874, Wash- ington, Gov. Print. Office, pp. 4-18. 1880. The northern Wasatch fauna. Amer. Natur., vol. 14, pp. 908-909. DENISON, R. H. 1938. The broad-skulled Pseudocreodi. Ann. N. Y. Acad. Sci., vol. 37, pp. 163-256. FISHER, C. A. 1906. Geology and water resources of the Bighorn Basin, Wyoming. U. S. Geol. Surv. Prof. Pap. No. 53, IlI-VI, pp. 1-72, 1 map. Gazin, C. L. 1953. The Tillodontia: an early Tertiary order of mammals. Smith- son. Misc. Coll., vol. 121, no. 10, pp. 1-110. 1962. A further study of the lower Eocene mammalian faunas of southwestern Wyoming. Smithson. Misc. Coll., vol. 144, no. 1, pp. 1-98. 1967 CLARK FORK VERTEBRATE FAUNA 29 GRANGER, W. 1914. On the names of lower Eocene faunal horizons of Wyoming and New Mexico. Bull. Amer. Mus. Nat. Hist., vol. 33, pp. 201-207. 21914. Annual Report of the Department of Vertebrate Paleontology, American Museum of Natural History, for the years 1911, 1912, and 1913. MS on file in the department. 1915. A revision of the lower Eocene Wasatch and Wind River faunas. Part III. Order Condylarthra, families Phenacodonti- dae and Meniscotheriidae. Bull. Amer. Mus. Nat. Hist., vol. 34, pp. 329-361. JEPSEN, G. L. 1930a. New vertebrate fossils from the lower Eocene of the Big- horn Basin, Wyoming. Proc. Amer. Phil. Soc., vol. 69, pp. 117-131. 1930b. Stratigraphy and paleontology of the Paleocene of northeastern Park County, Wyoming. Proc. Amer. Phil. Soc., vol. 69, pp. 463-528. 1940. Paleocene faunas of the Polecat Bench formation, Park County, Wyoming, Part I. Proc. Amer. Phil. Soc., vol. 83, pp. 217-341. MATTHEW, W. D. 1915a. A revision of the lower Eocene Wasatch and Wind River faunas. Part I. Order Ferae (Carnivora), suborder Creo- donta. Bull. Amer. Mus. Nat. Hist., vol. 34, pp. 4-103. 1915b. A revision of the lower Eocene Wasatch and Wind River faunas. Part I]. Order Condylarthra, family Hyopsodontidae. Bull. Amer. Mus. Nat. Hist., vol. 34, pp. 311-328. 1915c. A revision of the lower Eocene Wasatch and Wind River faunas. Part IV. Entelonychia, Primates, Insectivora (part). Bull. Amer. Mus. Nat. Hist., vol. 34, pp. 429-483. 1918. A revision of the lower Eocene Wasatch and Wind River faunas, Part V. Insectivora (continued), Glires, Edentata. Bull. Amer. Mus. Nat. Hist., vol. 38, pp. 565-657. MATTHEW, W. D. and W. GRANGER. 1915. A revision of the lower Eocene Wasatch and Wind River faunas. Introduction. Bull. Amer. Mus. Nat. Hist., vol. 34, pp. 1-4. McKEnNna&«, M. C. 1960. Fossil Mammalia from the early Wasatchian Four Mile fauna, Eocene of northwest Colorado. Univ. Calif. Publ. Geol. Sci., vol. 37, pp. 1-130. PIERCE, W. G. 1965. Geologic Map of the Clark Quadrangle, Park County, Wy- oming. U.S. Geol. Surv. Quadrangle Map GQ-477. ROBINSON, P., C. C. BLACK, and M. R. DAWSON. 1964. Late Eocene multituberculates and other mammals from Wy- oming. Science, vol. 145, pp. 809-811. 30 BREVIORA Nos257 Stmons, E. L. 1960. The Paleocene Pantodonta. Trans. Amer. Phil. Soc., vol. 50, pp. 1-98. SIMPSON, G. G. 1929. A new Paleocene uintathere and molar evolution in the Amblypoda. Amer. Mus. Novit., No. 387, pp. 1-9. 1937. Notes on the Clark Fork, Upper Paleocene fauna. Amer. Mus. Novit., No. 954, pp. 1-24. 1943. In Bogert, et al., Criteria for vertebrate subspecies, species, and genera. Criteria for genera, species, and subspecies in zoology and paleontology. New York Acad. Sci., vol. 44, pp. 145-178. 1945. The principles of classification and a classification of mammals. Bull. Amer. Mus. Nat. Hist., vol. 85, pp. 1-350. SINCLAIR, W. J., and W. GRANGER. 1912. Notes on the Tertiary deposits of the Bighorn Basin. Bull. Amer. Mus. Nat. Hist., vol. 31, pp. 57-67. Stow, M. H. 1938. Dating Cretaceous-Eocene tectonic movements in Bighorn Basin by heavy minerals. Bull. Geol. Soc. Amer., vol. 49, pp- 731-761. VAN HOUTEN, F. B. 1945. Review of latest Paleocene and early Eocene mammalian faunas. Jour. Paleont., vol. 19, pp. 421-461. VAN VALEN, L. 1966. Deltatheridia, a new order of mammals. Bull. Amer. Mus. Nat. Hist., vol. 132, pp. 1-126, 8 plates. WHEELER, W. H. 1961. Revision of the uintatheres. Bull. Peabody Mus. Nat. Hist. Yale Univ., No. 14, iii-vi, pp. 1-93, 14 plates. Woon, H. E., 2ND et al. 1941. Nomenclature and correlation of the North American con- tinental Tertiary. Bull. Geol. Soc. Amer., vol. 52, pp. 1-48, | plate. BREVIORA Museum of Comparative Zoology CAMBRIDGE, MaASss. 3 FEBRUARY, 1967 NUMBER 258 BIOLOGY OF THE PARTHENOGENETIC FUNGUS BEETLE CIS FUSCIPES MELLIE (COLEOPTERA: CIIDAE) By John F. Lawrence Thelytoky, or female-producing parthenogenesis, has been re- ported for a large number of animal groups, and a considerable amount has been written on its occurrence in various groups of insects, such as the Psychidae, Aphididae, Tettigoniidae, Phasmi- dae, and Curculionidae (White, 1954; Suomalainen, 1962). Aside from the well known cases in Ptinus clavipes and several European weevils, there have been few references to thelytoky in the Coleop- tera; yet this type of reproduction is probably fairly common throughout the order, judging from the number of groups in which unusual sex ratios have been noted. The step from the pre- liminary observation to experimental evidence is often a big one, however, because of the difficulties encountered in rearing many insects. While endeavoring to straighten out a taxonomic problem concerning some small fungus-feeding beeties in the family Ciidae, I discovered an apparent case of thelytoky in the species Cis fuscipes Mellié, and was able to confirm this by a series of rear- ing experiments discussed below. To my knowledge, this is the first record of its occurrence within the large superfamily Cucu- joidea. In connection with a study on the systematics and biology of western North American Ciidae, an attempt was made to distin- guish between two apparent sibling species, Cis fuscipes Mellié and Cis impressus Casey. The two species are widespread in North America, have broadly overlapping distributions, and feed on the same species of fungi. The only characters given in the literature to separate the two are the form of the male pronotum and ab- domen; in C. impressus the pronotum of the male is impressed anteriorly and the abdomen bears a setigerous pore on the first visible sternite, while both characters are apparently absent in the “male” of C. fuscipes. Females of the two species are apparently indistinguishable (Casey, 1898; Dury, 1917). nN BREVIORA No. 258 In an attempt to solve this problem, I examined about 1500 museum specimens from various parts of North America, collected large series of adults and larvae from both dimorphic and mono- morphic populations in California, and dissected series from Cali- fornia and Minnesota, in order to examine genital characters. No characters could be found to separate females or larvae of the two species, nor could any external sexual characters be found in C. fuscipes. When dissections were made of relatively large samples, it was found that the monomorphic series consisted entirely of females. This led to the hypothesis that thelytoky was occurring in C. fuscipes and that the name impressus had been applied to male specimens of the same species. The following rearing experiments were then conducted. REARING EXPERIMENTS Larvae of Cis fuscipes and fresh pieces of Polyporus versicolor were collected in Marin Co. and Alameda Co., California. Each larva was isolated in a petri dish with a piece of fungus, and water was added from time to time. Each resulting female was kept in isolation and observed periodically for the presence of eggs and larvae. Several rearings were attempted, but only one will be dis- cussed here. One larva was isolated on December 9, 1962, and by January 26, 1963, it had pupated and eclosed, producing an adult female. By February 7, eggs had been laid and early instar larvae were seen boring in the fungus. By April 1, 15 females were seen in the medium; 7 of these were preserved, while the other 8 were isolated in separate dishes. Eggs were seen in most of these dishes by May 4, and on May 31, all the adult females were removed and preserved, leaving only the F. larvae in the dishes. By July 19, adult F. females were present in all 8 dishes. Five of the clones were preserved, and the numbers of individuals contained in each were as follows: 51, 50, 43, 42, and 47. The remaining clones produced an Fs generation by October 24, and one of these was retained to produce F,’s by the end of the year. Because of intermittent field work, the exact generation time could not be recorded, but the F, generation was produced in about 60 days, which appears to be a reasonable figure when compared with observations made on several other ciid species. Two attempts were made to cross females from these partheno- genetic clones with males collected in the same area; all of these were unsuccessful in that only female progeny resulted. Mating probably did not take place, but many more trials will be necessary 1967 PARTHENOGENETIC FUNGUS BEETLE 3 before one can conclude that parthenogenetically-produced females will not cross with males from bisexual populations. It is obvious that some form of thelytoky is occurring in this species, but the actual cytological mechanisms involved have not been studied. Chromosome counts made by Mr. Norihiro Veshima showed that the diploid number is 14 both in the male and in the parthenogenetic female. Thus polyploidy does not occur in the California populations examined. GEOGRAPHICAL AND ECOLOGICAL RANGE Cis fuscipes is the most widespread and common species of Ciidae in North America, ranging from northern British Columbia to southern California on the Pacific Coast, east across southern Canada to Nova Scotia, and south throughout the eastern and mid- western United States (east of the 100th meridian) to Browns- ville, Texas, and Dunedin, Florida (Fig. 1). It is apparently absent from the more arid parts of the continent, such as the Great Plains, the Great Basin, and the southwestern deserts. A single record from Provo, Utah, indicates its possible occurrence in the poorly collected Rocky Mountain region. Specimens have also been seen from Mexico (no specific locality), Cuba, Hawaii, and from the island of Madeira in the eastern Atlantic. These isolated records will be discussed below. Throughout its range, the species occurs in association with sev- eral species of bracket fungi (Basidiomycetes: Polyporaceae), where both adults and larvae feed within the fruiting bodies. It appears to be restricted to members of the Polyporus versicolor group (Paviour-Smith, 1960), which have small, relatively thin sporophores with whitish, punky context and a trimitic hyphal sys- tem and which usuaily cause white rot in dead or dying angio- sperms. It is especially common in Polyporus versicolor, which may be considered its “headquarters” (Elton, 1949; Paviour-Smith, 1960), but also feeds on related fungi. Out of the 92 host records which I have accumulated for this species, 72 are from P. versi- color, 7 from P. hirsutus, 6 from Lenzites betulina, 2 from P. con- chifer, and a single record each from P. squamosus, Trametes suaveolens, Fomes fraxineus, Ganoderma brownii, and P. gilvus. In the western part of its range, Cis fuscipes is usually found associated with Cis versicolor Casey, Cis vitulus Mannerheim., Sulcacis curtulus (Casey), Ennearthron californicum Casey, and Octotemnus laevis Casey. In California, extensive coliecting has revealed that C. fuscipes (as well as C. vitulus and O. laevis) 4 BREVIORA No. 258 requires a somewhat more humid environment than C. versicolor, S. curtulus, or E. californicum, and it may be replaced by these species in drier environments. In the eastern part of its range, the species occurs with a number of other ciids, including Cis pis- toria Casey, Cis falli Blatchley, Sulcacis lengi Dury, Strigocis opaci- collis Dury, and Octotemnus laevis Casey, and other fungivorous beetles, such as Neomida bicornis (Fabricius) and Euparius mar- moreus (Oliviér). It is relatively common in most areas and appears to be successfully competing with other species inhabiting the same fungi. Like many species of polypores, P. versicolor and its relatives are widely distributed and are found on many different tree species, although they are normally restricted to angiosperms. One might expect a fairly continuous distribution of these fungus species within any deciduous forest, but the actual occurrence of fruiting bodies will depend upon the forest composition, number of dead trees, and physical factors affecting the establishment of the fungus and the production of sporophores. Even in a continuous forest, one may find P. versicolor replaced in certain areas by P. gilvus or P. pargamenus, each of which support different ciid species. This breaking up of the habitat will be more evident, of course, where the hardwoods themselves are scattered or rare. The absence of Cis fuscipes from the more arid part of western North America may be due to the inability of the species to tolerate drier environments, as was suggested by observations on California pop- ulations, rather than to the lack of hardwood forests, or to physical factors affecting the host fungi. The relative rarity of the beetle in the more humid parts of the Sierran and Cordilleran ranges, how- ever, may be related to the scattering of the habitat, since large stands of conifers separate riparion situations in which alders and other angiosperms provide a suitable medium for the host fungi. DISTRIBUTION CF BISEXUAL POPULATIONS The abundance of male specimens varies considerably through- out the range of this species. Since intensive population sampling was not attempted, it is not possible to present a detailed analysis of the sex ratios, but a general idea of the distribution of bisexual populations may be derived from the data recorded for over 2500 specimens (from 405 localities) collected by myself or bor- rowed from various museums. For the entire range, 12 per cent of the specimens examined were males, and these represented 21.7 per cent of the localities. In Figure 1, the localities from which PARTHENOGENETIC FUNGUS BEETLE 1967 ‘SOLOS OY} UI B[VU OUO svoT 3v JO DOUaSOId OY) O}LdIPUT S1v]s Spop1OdI1 Ud9q DALY STB -o} A[Uo YoIyM WoT sarRdo] JUdsaidar sapIIIO Youtg “sPEW Sadiosn{f s19 Jo uoNNgLYSIP POM MON oy} Jo dey aun! 6 BREVIORA No. 258 males were taken are represented by stars, while those from which only females were recorded are represented by black dots. Two things that are immediately apparent are: 1) the preponderance of male records in the northwestern part of the continent, as com- pared to their rarity on the east coast and complete absence in the Midwest, and 2) the scattered occurrence of males through- out the range as far south as Florida and Texas. From California, Oregon, Washington, British Columbia, and Alberta, 918 speci- mens (36.2%) from 77 localities (19.0%) were examined, and of these 221 specimens (24.1%) from 51 localities (66.2% ) were males. From the remainder of Canada and the United States, 1618 specimens (63.8%) from 328 localities (81.0%) were exam- ined, and of these 84 specimens (5.2%) from 37 localities (12.7% ) were males. Of the total male specimens examined, 72.4 per cent were collected in the Pacific states and western provinces. Although the sample sizes vary greatly and were often small, (decreasing the reliability of negative evidence), I think there is little doubt that bisexual populations are more abundant in the Northwest and become increasingly rare to the east and south. Parthenogenetic populations occur throughout the range, and in northern coastal California both types appear to be common in any one area. Males have been taken at several localities along the east coast and in the Gulf Coast states, but I have seen no records of their occurrence in the Midwest. SYSTEMATIC RELATIONSHIPS AND PROBABLE ORIGIN Cis fuscipes is closely related to a group of Old World species which were placed in the subgenus Hadraule Thomson by Reitter (1902), Chujo (1939), Miyatake (1954), and others. The name Hadraule, however, does not apply to the group, since its type species, Cis elongatulus Gyllenhal, has been excluded (Lohse, 1964; Lawrence, 1965). The species seem to form a natural unit and are characterized as follows: stout, oblong to subparailel body form; pronotum with fairly broad lateral margins and produced and rounded apical angles, which are preceded by a lateral in- flexion of the disc; dua! and subseriate elytrai punctation; vesti- ture of stout, erect bristles; prosternum slightly tumid but not cari- nate; apex of the protibia strongly dentate; and the male with two weakly developed clypeal piates, anteriorly impressed pro- notum, setigerous pore on the third abdominal sternite, and rela- tively simpie aedeagus. The species group extends from eastern 1967 PARTHENOGENETIC FUNGUS BEETLE | Siberia through southeastern Asia and India to Africa and Mada- gascar, being absent in northern Africa and Europe, but the species which most closely resemble C. fuscipes occur mainly in Siberia and Japan. These include Cis seriatopilosus Motschulsky (Siberia: Amur, Sakhalin; Japan); Cis seriatulus Kiesenwetter (Japan); Cis heiroglyphicus Reitter (Siberia: Khabarovsk; Japan); Cis tai- wanus Chujo (Taiwan, Loo Choo Islands, Japan?) ; Cis subrobus- tus Miyatake (Japan); and Cis japonicus Nobuchi (Japan). Those species which have been studied biologically occur on the same fungi as C. fuscipes. Cis seriatopilosus and C. taiwanus have both been recorded from Polyporus versicolor, while C. sub- robustus has been collected on P. hirsutus (Chujo, 1939; Fukuda, 1940; Miyatake, 1954). Although most of these beetles appear to be specifically distinct from the North American species, a single Siberian specimen identified by Reitter as C. seriatopilosus is very similar to individuals of C. fuscipes. Perhaps further collecting in eastern Siberia will reveal the presence of fuscipes there. The presence of several close relatives in Asia and the abun- dance of males in the northwestern part of the range suggest that Cis fuscipes or its ancestor dispersed into North America across the Bering Strait from a point of origin in the eastern Palaearctic. Since there is no evidence of unusual sex ratios among the Japanese and Siberian species, it is probable that the parthenogenetic mode of reproduction originated in northwestern North America and the species subsequently spread east and south across most of the continent. In the absence of a fossil record, there is no direct evidence relating to the time of dispersal of C. fuscipes into North America. Some indirect support may be derived from Linsley’s analysis of the cerambycid beetle fauna (1958), in which he considers all of the northern or Holarctic elements of the fauna to be post- Pleistocene in origin. Several of the examples of thelytoky in animais have been demonstrated to be glacial phenomena; this is discussed by Suomalainen (1962) for insects, and by Darewski and Kulikova (1961) for lacertid lizards. The recent origin of parthenogenesis has been argued by some authors on theoretical grounds, because of the inherent long-term disadvantages of this type of genetic system. SPREAD AND OVERSEAS DISPERSAL The ability of a parthenogenetic species to colonize and spread rapidiy throughout a new area is well known and has been dis- cussed in various texts, such as White (1954), and Mayr (1963). 8 BREVIORA No: 25's A system which suspends sexual reproduction not only makes pos- sible the perpetuation of a successful genotype as soon as it is formed, but doubles the fecundity by eliminating the “reproductive wastage” characteristic of a population in which roughly half of the individuals are males. Where the suitable habitat for a species is broken up, a parthenogenetic form will be able to spread more rapidly, since only a single female need reach a favorable micro- environment. In an area which is initially unsuitable for a bisexual species, the development of a parthenogenetic system may speed up the process of adaptation to the new environment by imme- diately fixing and reproducing a favorable genotype. Stebbins (1950) and others have stressed the importance of apomixis in the rapid colonization of new habitats by plant species. A comparison of the ranges of North American Ciidae shows that only Cis fuscipes occurs throughout the northern part of the continent and yet is also an important element in the fauna of the southeastern and midwestern United States. Several other species have northern distributions and similar affinities to Palaearctic species. Some of these, such as Cis horridulus Casey and Doli- chocis indistinctus Hatch extend southward only in montane regions and may occur in the southern Appalachians or in the mountains of Arizona. Sulcacis curtulus (Casey) extends into southern California but is rare in the northeastern and northern midwestern states. Eridaulus levettei (Casey) is fairly wide- spread in the eastern part of the continent, but does not occur in the West, while Octotemnus laevis Casey is common along both coasts and in parts of the Midwest. Neither of these species is as common or widespread as C. fuscipes, and both are absent from the Southern Coastal Plain and Gulf Coast. Two other wide- ranging species, Ennearthron californicum Casey and Ennearthron thoracicorne (Ziegler), occur in the western and eastern states respectively. Both of these, however, are of southern origin, hav- ing their closest relatives in the Neotropical Region, and neither extends far into Canada. The comparison of C. fuscipes with other northern or Holarctic members of the North American fauna may seem to imply that all of these have similar relationships to Palaearctic species and thus represent Old World invasions of roughly the same age. This does not appear to be the case. All of the other northern species have Old World counterparts which extend throughout the Palaearctic from Europe to Siberia and Japan, whereas only C. fuscipes belongs to a group restricted to the eastern Palaearctic and Oriental regions and having some relationships with palaeotropical forms. It may be that C. fuscipes 1967 PARTHENOGENETIC FUNGUS BEETLE 9 represents the most recent element in our fauna, but there is no direct evidence for this. I think there is little doubt that the evolution of thelytoky in C. fuscipes is responsible for the apparently rapid spread of the species and for its present wide distribution in North America. In addition to the obvious advantages of the parthenogenetic sys- tem to dispersal and colonization, the increased fecundity has probably contributed to its success in competing with the large and diverse fauna of southern origin occupying the southern peri- phery of the range. The only large area in North America which is suitable for C. fuscipes, but in which the forest cover is con- siderably broken up, due partly to agriculture, is the Midwest. It is interesting that this is the only region in which males are totally absent. The presence of males in scattered localities along the east coast and at the southern extremities of the range raises several questions which can be answered only by more detailed popula- tion studies. Is the parthenogenetic form the result of a single evolutionary event, an obligate thelytokous form being comptetely independent of the bisexual form? Is a residual bisexual popula- tion continually giving rise to parthenogenetic clones? Is partheno- genesis facultative in this species, so that occasional females can give rise to normal males? It is hoped that future studies will provide the answers. Mellié’s original series of Cis fuscipes included several speci- mens from the island of Madeira. It would seem improbable that the Madeiran series represents the same species, but an examina- tion of Mellié’s and Wollaston’s specimens revealed that they are conspecific with the North American fuscipes, and that they are all females. According to Wollaston (1854, 1865), the species has become well established in cultivated areas at low elevations around Funchal and is likely to be an accidental introduction. Although it is possible that a specimen of fuscipes could have rafted from the Caribbean to Madeira, it is more reasonable to assume that the species was introduced by man. Recently, a series of female specimens of fuscipes were collected at Olinda on the island of Maui, Hawaii, and presented to the Commonweaith Institute of Entomology. Aithough Perkins, Swezey, and others have collected extensively in Hawaii, this is the first record of this species there, and probably represents a recent human introduction. In addition to this, two female speci- mens in the Reitter coilection were apparently collected in Cuba. Whether by natural means or by human transport, Cis fuscipes 10 BREVIORA No. 258 has dispersed from North America to these various islands, and has become established on at least one of them. The advantages of parthenogenesis in the establishment of a species having been introduced into a new area by long-distance dispersal have been discussed by several authors, including Longhurst (1955) for Crustacea, and Baker (1955) for plants. TAXONOMIC STATUS Because of its variability and wide distribution, Cis fuscipes has been given a number of names by different authors. Mellié (1848) described the species from a series of 5 specimens from Boston in the Chevrolat collection and 4 more collected by Wollas- ton on Madeira. In the same paper, three other names were applied to the species: Cis atripennis, also from Boston, Cis chevrolatii from ““Nouvelle-Orleans,” and Cis dubius from the latter locality. Leconte, Horn, and others applied the name fuscipes to the common North American species, and it has been used consistently by later authors. Casey (1898) described two more species, Cis carolinae (North Carolina) and Cis pallens (Montana), which were distinguished from fuscipes on the basis of color and relative lengths of antennal segments. He also described a third species, Cis impressus (Pacific Coast), differing from fuscipes only in the form of the male pronotum, which was said to be “broadly impressed at apex.” Dury (1917) noted that Cis fuscipes occurred throughout North America, that C. impres- sus occurred on both coasts, and that the females of the two species could not be distinguished from one another. He also considered chevrolatii and carolinae to be synonyms of fuscipes. During the summer of 1963, I examined the Casey types at the United States National Museum, Washington, D. C., and in the spring of 1966 I had the opportunity to study the types of Mellié in the Salié and Pic collections at the Muséum National d’Histoire Naturelle, Paris, and in the Wollaston collection at the British Museum (Natural History), London. All of the above names, with the exception of Cis impressus Casey, are based on female specimens of the variable Cis fuscipes. The name fuscipes is here selected as the senior synonym because of its continual usage in the North American literature. Being based on a bisexual population, Casey’s name impressus presents certain difficulties. Since it has not yet been established whether this species is faculta- tively parthenogenetic or rather composed of a bisexual species and one or several obligate parthenogenetic clones, it could be 1967 PARTHENOGENETIC FUNGUS BEETLE 11 argued that the name impressus should be applied to populations in which males occur. The true nature of the biological situation will be made clear only after intensive population analyses, rear- ing experiments, and cytological studies. In any case, I preter to consider impressus a synonym of fuscipes on purely practical grounds, since no characters have been found to distinguish the larvae or females of the two forms. I therefore propose the fol- lowing synonymy: CIs FUSCIPES Mellié Cis fuscipes Mellié, 1848: 271, pl. 2, fig. 23; Wollaston, 1854: 281; Wollaston, 1865: 234; Cas2y, 1898: 78; Blatchley, 1910: 898; Dalla Torre, 1911: 11; Dury, 1917: 11; Leng, 1920: 246; Weiss, 1920: 138 (host); Weiss and West, 1920: 8 (host); Weiss and West, 1921b: 169 (host); B6ving and Craighead, 1931: 270-71, pl. 92, figs. K-R (larva); Arnett, 1962: 827, fig. 1.98; Hatch, 1962: 231; Lawrence, 1965: 279. Type locality: “Bos- ton.” Lectotype, female, Pic collection, Mus. Natl. Hist. Nat., Paris. Cis atripennis Mellié, 1848: 258, pl. 2, fig. 15; Casey, 1898: 77; Dalla Torre, 1911: 6; Leng, 1920: 246. Type locality: “Boston.” Holotype, female, Pic collection, Mus. Natl. Hist. Nat., Paris. NEW SYNONYMY. Cis chevrolatii Mellié, 1848: 249; Blatchley, 1910: 898; Dalla Torre, 1911: 8; Dury, 1917: 11; Leng, 1920: 246. Type locality: “Nouvelle-Orleans.” Lectotype, female, Pic collection, Mus. Natl. Hist. Nat., Paris. Cis dubius Mellié, 1848: 273; Dalla Torre, 1911: 10; Leng, 1920: 247. Type locality: “Nouvelle-Orleans.” Lectotype, female, Salle collection, Mus. Natl. Hist. Nat., Paris. NEW SYNONYMY. Cis carolinae Casey, 1898: 78; Dalla Torre, 1911: 3.) Dury, 1917 11 keng: 1920: 246. Type locality: “North Carolina (Asheville).” Holotype, female, Casey collection, U.S. Nat. Mus., Washington, D.C. Cis impressa Casey, 1898: 79; Pallawome slo = 12. Dany oly. 1; Blatch- ley, 1918: 54; Leng, 1920: 246; Weiss and West, 1921a: 61 (host); Blatchley, 1923: 19; Hatch, 1962: 231. Type locality: “Vancouver Island.” Holotype, male, Casey collection, U. S. Nat. Mus., Washington, D. C. NEW SYNONYMY. Cis pallens Casey, 1898: 78; Dalla Torre, 1911: 15; Leng, 1920: 246. Type locality: “Montana (Missoula).” Holotype, female, Casey collection, U. S. Nat. Mus., Washington, D. C. NEW SYNONYMY. ACKNOWLEDGMENTS This investigation was supported in part by a grant from the National Science Foundation (GB-4743), grants-in-aid from the Woodrow Wilson Foundation, Society of the Sigma Xj, and Uni- versity of California Committee on Research, and fellowships from the National Science Foundation and Woodrow Wilson 12 BREVIORA No. 258 Foundation. Sincere thanks are extended to E. G. Limnsley, R. L. Usinger, C. D. MacNeill, and H. V. Daly for their advice dur- ing the early part of the study, to I. Taveres, R. L. Gilbertson, and J. A. Stevenson for the identification of host fungi, to N. Ueshima for the preparation and interpretation of cytological material, and to E. Mayr and P. J. Darlington, Jr. for their criti- cal review of the manuscript. I am indebted to the curators and staff of the following institutions for making their collections available to me: Academy of Naturai Sciences, Philadelphia, Pa.; American Museum of Natural History, New York, N. Y.; Brigham Young University, Provo, Utah; British Museum (Natural His- tory), London, England; California Academy of Sciences, San Francisco, Calif.; California Insect Survey, University of Cali- fornia, Berkeley, Calif.; Canadian National Collection, Entomology Research Institute, Ottawa, Canada; Carnegie Museum, Pitts- burgh, Pa.; Chicago Natural History Museum, Chicago, IIl.; Cin- cinnati Museum of Natural History, Cincinnati, Ohio; Cornell University, Ithaca, N. Y.; Hlinois Natural History Survey, Urbana, Ill.; Museum of Comparative Zoology, Cambridge, Mass.; Muséum National d’Histoire Naturelle, Paris, France; Oregon State Uni- versity, Corvallis, Oregon; United States National Museum, Wash- ington, D. C.; University of Alberta, Edmonton, Alta.; University of California, Davis, Calif.; University of Washington, Seattle, Wash.; University of Kansas, Lawrence, Kansas; Utah State University, Logan, Utah. Thanks are also extended to the num- erous individuals who have collected material for me or made their own collections available for study. LITERATURE CITED ARNETT, R. H. 1962. The beetles of the United States (A manual for identification). Part V. Fascicle 98. Cisidae. Catholic Univ. Amer. Press, Washington, D. C. Pp. 827-830, pl. 1.98. BAKER, H. G. 1955. Self-compatibility and establishment after “long-distance” dis- persal. Evolution, 9(3): 347-348. BLATCHLEY, W. S. 1910. An illustrated descriptive catalogue of the Coleoptera or beetles (exclusive of the Rhynchophora) known to occur in Indiana. Bull. Indiana Dept. Geol. Nat. Res., 1, 1385 pp., 590 figs. 1918. On some new or noteworthy Coleoptera from the west coast of Florida. — IV. Canad. Ent., 50: 52-59. 1923. Notes on the Coleoptera of southern Florida with descriptions of new species. Canad. Ent., 55: 13-20, 30-36. 1967 PARTHENOGENETIC FUNGUS BEETLE 13 BOvING, A. G. and F. C. CRAIGHEAD 1931. An illustrated synopsis of the principal larval forms of the order Coleoptera. [part] Entom. Amer., N.S., 11(4): 258-351, pls. 86-125. (GASEY- alles 1898. Studies in the Ptinidae, Cioidae, and Sphindidae of America. New York Ent. Soc., 6(2): 61-93. CHUJO, M. 1939. Onthe Japanese Ciidae (Coleoptera). Mushi, 12(1): 1-10. DALLA Torre, K. W. VON 1911. Cioidae. Pars 30. Jn W. Junk and S. Schenkling, edit., Coleop- terorum Catalogus. Junk, Berlin. 32 pp. DarEwWSKI, I. S. and W. N. KULIKOWA 1961. Natiirliche Parthenogenese in der polymorphen Gruppe der kaukasischen Felseidechse (Lacerta saxicola Eversmann). Zool. Jahrbuch., 89(1): 119-176, 35 figs. Dury, C. 1917. Synopsis of the coleopterous family Cisidae (Cioidae) of America north of Mexico. J. Cincinnati Soc. Nat. Hist., 22(2): 1-27. ELTON, C. 1949. Population interspersion: an essay on animal community pat- terms Je Ecole sii: ail-23: Fukupa, A. 1940. A study of the Japanese Ciidae, Coleoptera (II). [In Japanese} Trans. Nat. Hist. Soc. Formosa, 30: 48-53, 5 figs. Hatcu, M. H. 1962. The beetles of the Pacific Northwest. Part III: Pselaphidae and Diversicornia I. Univ. Washington Press, Seattle. ix + 503 pp., 66 pls. LAWRENCE, J. F. 1965. Comments on some recent changes in the classification of the Ciidae (Coleoptera). Bull. Mus. Comp. Zool., 133(5): 273- 293): LENG, C. W. 1920. Catalogue of the Coleoptera of America north of Mexico. Sher- man, Mount Vernon, N. Y. x + 470 pp. LINSLEY, E. G. 1958. Geographical origins and phylogenetic affinities of the ceram- bycid beetle fauna of western North America. In C. L. Hubbs, edit., Zoogeography. Amer. Assoc. Adv. Sci., Publ. 51. Pp. 299-320. Louse, G. A. 1964. Die in Mitteleuropa vertretenen Gattungen der Cisidae. (1. Beitrag zur Kenntniss der mitteleuropaischen Cisidae.) Ent. Blatter, 60(2): 116-122. 14 BREVIORA No. 258 LONGHURST, A. R. 1955. Evolution in the Notostraca. Evolution, 9: 84-86. Mayr, E. 1963. Animal species and evolution. Harvard Univ. Press, Cambridge, Mass. xiv + 797 pp. MELLIE, J. 1848. Monographie de l’ancien genre Cis des auteurs. Ann. Soc. Ent. France, Ser. 2, 6: 205-274,313-396, pls. 9-12. MIYATAKE, M. 1954. Studies on the Japanese Ciidae, I (Coleoptera). Sci. Rep. Matsuyama Agr. Coll., 14: 40-67, 11 pls. PAVIOUR-SMITH, KITTY 1960. The fruiting-bodies of macrofungi as habitats for beetles of the family Ciidae (Coleoptera). Oikos, 11(1): 43-71, 1 pl. REITTER, E. 1902. Analytische Uebersicht der palaearctischen Gattungen und Arten der Coleopteren-Familien: Byrrhidae (Anobiidae) und Cioidae. Verh. Naturforsch. Ver. Briinn, 40: 1-61. STEBBINS, G. L. 1950. Variation and evolution in plants. Columbia Univ. Press, New York. xix + 643 pp. SUOMALAINEN, E. 1962. Significance of parthenogenesis in the evolution of insects. Ann. Rev. Ent., 7: 349-366. Weiss, H. B 1920. Coleoptera associated wiih Polyporus versicolor L. in New Jersey. Psyche, 27: 137-139. WEIss, H. B. and E. WEsT 1920. Fungous insects and their hosts. Proc. Biol. Soc. Washington, 33: 1-20, 1 pl. 1921a. Additional fungous insects and their hosts. Proc. Biol. Soc. Washington, 34: 59-62. 1921b. Additional notes on fungous insects. Proc. Biol. Soc. Washing- ton, 34: 167-172, 1 pl. WHite, M. J. D. 1954. Animal cytology and evolution. 2d ed. Cambridge Univ. Press. 454 pp. WOLLASTON, T. V. 1854. Insecta Maderensia; being an account of the insects of the islands of the Madeiran group. Van Voorst, London. xliii + 634 pp., 13 pls. 1865. Coleoptera Atlantidum, being an enumeration of the coleopter- ous insects of the Madeiras, Salvages, and Canaries. Van Voorst, London. xlvii + 526 pp., appendix, 140 pp. (Received 5 July, 1966). BREVIORA Museum of Comparative Zoology CAMBRIDGE, MaAss. 3 Fepruary, 1967 NUMBER 259 EXPANDING THE PALPI OF MALE SPIDERS' By William A. Shear? INTRODUCTION Genitalia and secondary sexual structures are important in the determination of spiders to the species level. This is especially true in the male, where the palpus has become modified as an intromittent organ. In many genera, the females are virtually in- separable into species, and small differences in shape, size, rela- tive position, and complexity of the parts of the highly modified and specialized male palpus are used to distinguish between closely related species of a genus. In families where the male palpus is highly specialized, the palpal organ has become so folded and convoluted that some parts are hidden in the normal retracted palpus. While these parts are not always of taxonomic importance, they nevertheless need to be examined in detail, especially in characterizing new species. Many publications concerned with spider taxonomy include figures of male palpi that have been treated in order to unfold or to expand the organ. In other publications, figures are of un- treated and unexpanded palpi and, as a result, confusion and un- certainty arise in comparing species and in making identifications, because frequently there are striking differences between ex- panded and unexpanded palpi of the same species. In an ex- panded palpus, sclerites, which are hidden in an unexpanded pal- pus, are revealed and the serial relationship of the sclerites be- comes more apparent. My purpose is not to decide what parts of the palpus are of value in taxonomy, but merely to determine what additional parts can be seen when the palpi of representative species are expanded. ‘Part of a thesis submitted to the faculty of the University of New Mexico for the degree of Master of Science > Department of Biology, Concord College, Athens, West Virginia tN BREVIORA No. 259 I wish to express my indebtedness to Dr. C. Clayton Hoff for his guidance throughout the period of this study, and especially to Dr. Herbert W. Levi of the Museum of Comparative Zoology at Harvard University, Cambridge, Massachusetts, for many helpful suggestions. I appreciate the use of space and facilities made available by Dr. James S. Findley, Dr. William C. Martin, and Dr. William G. Degenhardt, curators of the Museum of South- western Biology, Albuquerque, New Mexico. Some of the speci- mens used for this study were donated by Dr. Andrew A. Weaver of the College of Wooster, Wooster, Ohio, and by Mr. Joseph A. Beatty of Southern Illinois University, Carbondale, Illinois. METHODS In The Spider Book, Comstock (1940, p. 111) described a process for expanding the complex palpal organs of male spiders. This method was probably the one used by Comstock in prepar- ing specimens for a 1910 paper on spider palpi, because the text of the 1910 publication is the same as the section on palpi in The Spider Book. The method he described involves immersing the palpus in boiling 10 per cent “caustic” for 10-15 minutes. After treatment, the palpus is rinsed in water and preserved in glycerine. Engelhardt (1910) challenged this method on the grounds that the boiling solution was somewhat destructive. He was working on the more delicate female genitalia at that time. In 1925, Petrunkevitch, in a similar study, concluded that Engelhardt’s ob- jections were not substantiated by results, and further stated that palpi prepared by boiling corresponded closely to those he observed in actual use by the male spider. Gering (1953) prepared a large series of palpi using boiling potassium hydroxide solution, but for studies of the natural mode of expansion placed preserved palpi in concentrated potassium hydroxide solution for a few minutes before a rapid transfer to distilled water, where the palpi were inflated by osmotic pressure. Levi (1961) used hot sodium hydroxide solution, followed by distilled water. Osmotic pressure expanded the organ. In 1965, he suggested the use of boiling 85 per cent lactic acid solution, on the grounds that it is far less destructive than sodium hydroxide solution. He also mentioned that satisfactory results can be ob- tained by using fine needles and forceps to expand the parts of the palpal tarsus. 1967 PALPI OF MALE SPIDERS 3 The method used in this study was similar to that used by Gering (1953) and Levi (1961). Palpi removed from specimens preserved in 70 per cent ethanol solution were placed in 10 per cent KOH solution for 6-8 hours at room temperature and were completely expanded in distilled water. The expanded palpi were then returned to alcohol. Twenty-nine species, distributed among 22 genera and 10 fam- ilies of higher spiders, were selected as test samples on the basis of the number of specimens available, completeness of taxonomic definition, and number of available references. Also, each of these species has palpi similar in degree of complexity to the palpi of a number of other species and genera in their respective families. Before expanding the palpus of one of the species, a number was assigned and placed in the vial containing the specimens. In most cases, the right palpus of an individual was removed and treated, leaving the left as a control, but in some small species, the whole male was treated to avoid damage to the palpus in re- moval. After expansion, the palpus (or the whole specimen) was placed in alcohol in a separate vial and labelled with the same number as was previously assigned. The vial containing the pal- pus was then bound to the original vial of specimens with a rub- ber band. These specimens in alcohol were used in making the freehand drawings and in writing the descriptions. TERMINOLOGY The course taken by the evolution of secondary sexual struc- tures in spiders is open to speculation. Comstock (1940), Bar- rows (1925), Savory (1928), Gertsch (1949), Alexander and Ewer (1957), and Levi (1961) have expressed ideas on the sub- ject. Until we can be reasonably certain of the phylogeny of the sclerites of the male palpus, names applied to these sclerites do not necessarily indicate homologies. Comstock (1940) effectively reviewed previous research on spider palpi and devised a system of terminology widely accepted by most authors. Comstock’s terminology will be followed in this study. The following description refers to the expanded palpus of Araneus sericatus (Clerck) (Figs. 20, 21). The segments of the palpus proximal to the tarsus are only rarely modified. In most spiders, the coxa of the palpus bears an endite that functions as a mouth part, crushing and straining food material. The trochanter can bear a small projection of unknown 4 BREVIORA No. 259 function. The femur of the palpus of a male spider is sometimes furnished with sclerotic picks, which rub against stridulating areas on the sides of the chelicerae, producing squeaking or barking noises. Gering (1953) demonstrated in the genus A gelenopsis that apophyses on the tibia and patella can serve as locking de- vices in copulation. It is the palpal tarsus, especially the tarsal claw, that has been strongly modified (Barrows, 1925). The body of the tarsus has become modified in males of the more specialized families to form a cup-shaped cymbium (cy), the cavity of which is the alveolus. The cymbium is basal in position and the palpal organ itself folds into the alveolus when at rest (Figs. 18, 19). The most proximal portion of the organ is called the basal division. Actually attached to the wall of the alveolus is a hollow, bladder-like structure termed the basal hematodocha (bh). The cavity of the basal hematodocha is continuous with that of the cymbium, allowing the hematodocha to be inflated by increasing blood pressure. At the distal end of the basal hematodocha is a smooth or annulated ring of sclerotic material, the subtegulum (st). The middle division consists of the tegulum (7), an inappropri- ately named ring of heavily sclerotized tissue connected to the subtegulum by the middle hematodocha (mh). Articulated with the tegulum is a usually small sclerite, the median apophysis (ma). In some male spiders, this appears as a knob or process on the tegulum. The apical division is the most complex of the three divisions. The embolus (e) is the functional part of the organ, bearing at its tip the opening of the ejaculatory duct. The conductor (c) is a usually membranous outgrowth from the base of the embolus, serving to protect the embolus. The radix (1) and stipes (s) are small, and are often absent in less complex palpi. Both radix and stipes arise near the base of the embolus, although they may ap- pear more distal. The palpal organs in some families do have parts of the organ actually distal to the embolus. These parts in- clude a distal hematodocha (dh) and the lateral subterminal apophysis (/sa). The terminal apophysis (fa) is located at the very tip of the organ. OBSERVATIONS Order ARANEAE Family THERIDITDAE The theridiids, or comb-footed spiders, have recently been the subject of an extensive revision by Levi from 1952 to 1964. In 1967 PALPI OF MALE SPIDERS 5 1961, Levi published his ideas on the evolution of the palpal organ in this large and complex family, and concluded that the original type of palpus was probably simple, that most members of the Theridiidae have a complex palpal organ, and that secondary simplification can be seen in some otherwise highly specialized theridiids. For my purposes, I selected four species that represent various degrees of complexity within the family. Steatoda borealis (Hentz) Pigss 12S Levi (1957a) gives excellent figures of the palpus of Steatoda borealis in his revision of the genus Steatoda. Both expanded and unexpanded palpi are shown, and some individual sclerites are figured. In a mesal view of the unexpanded right palpus (Fig. 1), the embolus (e) can be seen, but the basal portion is hidden. The conductor (c), and the radix (7), both of which support the em- bolus in this species, are fairly obvious, but the shape of the radix is obscured by other sclerites lying over its ectal part. In a mesal view of an expanded right palpus (Fig. 2), the rotate base of the embolus (e) is exposed, and its true relationship to the conductor and radix (r) is revealed. In an ectal view of the same palpus (Fig. 3), expansion has caused a general ectad rotation of all the sclerites, bringing the median apophysis (ma) from behind over- lying sclerites. Enoplognatha tecta (Keyserling ) Figs. 4, 5 When Levi (1957b) revised the genus Enoplognatha, he gave two figures of an unexpanded palpus and one of an expanded palpus of this species. In an ectal view of an unexpanded left palpus (Fig. 4), the median apophysis and subtegulum are completely covered by overlying sclerites, although the shape of the embolus (e) and its relationship to the conductor (c) and radix (r) are obvious. In an ectal view of an expanded palpus (Fig. 5), the median apophy- sis (ma) has been rotated ectad and ventrad and the subtegulum (st) is visible. In both Figures 4 and 5, the hooklike projection of the anterior portion of the cymbium (cy) is seen. This is the paracymbial hook, which in some families is developed as a sep- arate sclerite. 6 BREVIORA No. 259 Theridion differens Emerton Figs. 6, 7 Levi (1957b) reviewed the genus Theridion, a very large and geographically widespread assemblage of, in many cases, remark- ably similar species. Levi did not figure an expanded palpus of Theridion differens, although he figured expanded palpi of sev- eral other species of the genus. When an unexpanded right palpus in ventral view (Fig. 6) is compared with an expanded right palpus (Fig. 7) seen from the same aspect, it is clear that little more is visible after expansion. Theridion frondeum Hentz Figs. 8, 9 This species is figured by Levi in his revision of Theridion (1957b). As in Theridion differens, an expanded palpus (Fig. 9) shows very little that is not readily visible in the unexpanded pal- pus’ (Fig. 3). Family LINYPHITDAE Recent literature on this family has been restricted to a few occasional papers describing new species and giving new distribu- tional records (Chamberlin and Ivie, 1943; Gertsch and Davis, 1946; Gertsch, 1951; and others). Blauvelt (1936) studied the palpus of Linyphia. Because the palpi of members of this family are somewhat uni- formly complex, two of the more common species were selected for this study. Drapetisca alteranda Chamberlin Figs, 10514 In a mesal view of an unexpanded right palpus (Fig. 10) of Drapetisca alteranda, the paracymbium (pc) is seen as a free sclerite, loosely articulated with the cymbium (cy). The lateral subterminal apophysis (/sa) is highly developed and forms a shield covering the entire palpal organ. In a mesal view of an expanded right palpus (Fig. 11), the general ectad rotation caused by expansion is readily noticed. Parts, including the sub- tegulum (st) and tegulum (ft), which were completely hidden before treatment, are now easily seen. The bulky base of the embolus (e) is now visible, and the shape of the lateral sub- terminal apophysis (/sa) is more satisfactorily observed. 1967 PALPI OF MALE SPIDERS 7 Linyphia marginata C. L. Koch Pips. 23 In this species, studied by Blauvelt (1936), the lateral sub- terminal apophysis (/sa), as seen in an ectal view of the unex- panded right palpus (Fig. 12), is not so highly developed as it is in the complex shield of Drapetisca. The tegulum (7) and sub- tegulum (st) are at least partially visible, but the conductor (c) and embolus (e) are obscured. In an ectal view of an expanded right palpus (Fig. 13), an ectad rotation brings into clearer view the position of the median apophysis (ma), which seems much more distal in the unexpanded palpus, and the exact shape and size of the unusually heavy embolus (e). The conductor (c) is readily observed, with its supporting sclerotic bar along one side. The lateral subterminai apophysis (/sa) is reflexed, appearing as if it were proximal, rather than distal, to the embolus. Family MICRYPHANTIDAE This family has not been studied as a unit since the publication of a long series of papers by Crosby, Crosby and Bishop, and Bishop and Crosby from 1905 to 1938 (listed by Kaston, 1948). Because of the seemingly simplified structure of the male palpus, there is some controversy about the proper taxonomic position of the Micryphantidae. The family is such a large one and shows so many degrees of palpal complexity that it would be impossible to treat the Micryphantidae satisfactorily in a paper of this nature. A recent study by Merrett (1963) of the palpal anatomy of this family treats the Micryphantidae as the subfamily Erigoninae of the Linyphiidae, but Merreit’s work was confined to British species. The species I studied were selected on the basis of their inter- mediate type of palpal organ. Soulgas corticarius (Emerton) Figs. 14, 15 This species was treated by Crosby and Bishop (1936), who gave a figure of an unexpanded palp, along with the characteristic cephalic modification found in this group. An ectal view of an unexpanded right palpus (Fig. 14) shows the obviously simple structure, including a free paracymbium (pc), subtegulum (st), tegulum (¢), and embolus (e). The shape of the tibia (f7) is unusual, having a modification that in at least some degree occurs in all male micryphantids. In mesal view, an 8 BREVIORA No. 259 expanded right palpus (Fig. 15 shows two features that cannot be seen in an unexpanded palpus. The great development of the basal hematodocha (bh) is obvious. A small sclerite (ma) pos- sibly represents a rudimentary median apophysis. Ceratinopsidis formosa (Banks) Figs l6;pal7 Bishop and Crosby (1930) published a figure of the unex- panded palpus of Ceratinopsidis formosa. The unexpanded pal- pus (Fig. 16) has features similar to those of Soulgas corticarius. In an expanded palpus (Fig. 17) observed in dorsoectal view, the greatly developed basal hematodocha (bh) is seen, but no sclerite corresponding to a median apophysis is apparent. Family ARANEIDAE Archer (1951, p. 3) defined this family as follows: “Paracymbium vertical or divergent from the axis of the cym- bium, but its basal face not resting on the apical face of the tibia, instead usually being separated from it by a distinct gap. Attach- ment of the genital bulb to the cymbium ranging from universal- median and then by all degrees of migration to frankly basal (as in the Theridiidae). Tegulum greatly overbalancing the sub- tegulum, the latter ranging from ring-like to a vestigial knob. Position of the cymbium and genital bulb normal.” If the Araneidae are considered in the strict sense, excluding the Tetragnathidae and the Theridiosomatidae, then all members of the family have complex palpi, but some do show various de- grees of reduction in certain sclerites. I selected eight fairly rep- resentative species. Araneus sericatus Clerck Figs: 18,19 20521 The complicated palpus of this species was described in the section on TERMINOLOGY. In a ventral view of the unex- panded right palpus (Fig. 18), little can be seen except the sub- tegulum (st), tegulum (ft), terminal apophysis (fa), and median apophysis (ma). In mesal view (Fig. 19), some additional struc- tures are visible, but their relationships to definitely identifiable structures are obscure. A number of interesting structures are revealed in an ectal view of an expanded right palpus (Fig. 20). The anterior portion of the subtegulum is completely unsclero- tized, so that the basal and middle hematodochae (bh, mh) are 1967 PALPI OF MALE SPIDERS 9 fused anteriorly. The tegulum (¢) is a heavy rounded structure articulating closely with the large median apophysis (ma). The parts of the embolic subdivision are visible distal to the tegulum. The spatulate stipes (s) and small conductor (c) articulate with the base of the embolus (e), but the radix (r) has been dis- placed proximad. In a mesal view of the same palpus (Fig. 21 Dr the hook-like lateral subterminal apophysis (/sa) is fused with the large plate of the terminal apophysis (ta), and both are sep- arated from the embolic subdivision by the distal hematodocha (dh). Araneus nordmanni (Thorell) The palpus of Araneus nordmanni is very similar in general plan to that of Araneus sericatus, differing only in the shape and position of some of the sclerites. Expansion in this species has the same effect as expansion has on the palpus of Araneus seri- catus. Argiope trifasciata (Forskal ) Figs. 22, 23 The main difference between the palpus of this species and those of species of the genus Araneus lies in the reduction of some of the sclerites of the embolic subdivision. In a mesal view of an unexpanded right palpus (Fig. 22), the subtegulum is not seen, but in its place a portion of the basal hematodocha (bh) appears as the most basal part. The tegulum (t), embolus (e), and well-developed conductor (c) are all read- ily visible. The median apophysis (ma) is large and has a ser- rated ventral edge. In a dorsomesal view of an expanded right palpus (Fig. 23), the subtegulum (st) has been moved into view. The ectad rotation of the median apophysis (ma) has revealed a small hook on the dorsal surface. The embolus (e) and con- ductor (c) are fused for about two-thirds of their length. Other sclerites of the embolic subdivision seem to be absent, although the swollen base of the embolus possibly represents a reduced radix. Araniella displicata (Hentz) Figs. 24, 25 Chamberlin and Ivie (1942) removed this species from Ara- neus and made it the type species of their new genus Araniella. Unfortunately, they did not present figures of the type species. 10 BREVIORA No. 259 I studied the palpus in detail. In an ectal view of an unexpanded right palpus (Fig. 24), the subtegulum (st) appears cupped, as if it has taken on the function of the conductor in this species. The very large tegulum (ft) is easily visible. The basal part of the embolus (e) seems to be fused to the tegulum. The basal part of the median apophysis (ma) is hidden by the cymbium (cy). In an ectal view of an expanded right palpus (Fig. 25), the re- duction of the subtegulum (st) characteristic of this family, as described by Archer (1951), is clearly seen. The basal part of the median apophysis (ma) is visible and there is a small “hema- todocha” around the base of the embolus (e) and median apophy- sis. This area of lightly sclerotized tissue may represent a vesti- gial distal hematodocha, but due to the absence of any sclerites morphologically distal to the embolic subdivision, this ‘“‘hema- todocha” is best considered an articulating membrane between the tegulum (ft) and the median apophysis. Singa pratensis Emerton Figs; 26,27 The members of the genus Singa are structurally close to the theridiids (Kaston, 1948), but the structure of the palpus is closer to Araneus than are some species formerly considered to be in Araneus. According to Kaston (1948), who figures the pal- pus and epigynum, the species of Singa resemble one another so closely in form and appearance that they can be separated only by examining the external genitalia. By comparing a mesal view of an unexpanded right palpus (Fig. 26) with the palp of Araneus sericatus (Fig. 19), one sees that the relative sizes of the structures correspond closely. The subtegulum (st) is reduced. The tegulum (¢), median apophysis (ma), and terminal apophysis (7a) are all visible. The expanded right palpus in ectal view (Fig. 27) bears close resemblance to the expanded palpus of Araneus sericatus (Fig. 20). The sclerites of the embolic subdivision are all visible in essentially the same posi- tions as in Araneus sericatus, but the radix of Araneus sericatus is in a slightly different position (Fig. 21). Neoscona arabesca (Walckenaer ) The genus Neoscona is closely allied to Araneus. The males bear compiex palpal organs built around the general Araneus plan. Expansion delineates the relationships of the various scler- ites in much the same way as in Araneus sericatus, Araneus nord- manni, and Singa pratensis. 1967 PALPI OF MALE SPIDERS 11 Mangora gibberosa (Hentz) Figs. 28, 29 Mangora gibberosa has a complex palpus that, due to the dis- placement of the sclerites, seems even more complex than it really is. In an ectal view of an unexpanded right palpus (Fig. 28), the subtegulum (st) is visible and is large for a member of the family Araneidae. The tegulum (ft) and terminal apophysis (ta) are separated by a large radix (r), the sheet-like, membranous conductor (c), and the enlarged base of the embolus (e). In a ventroectal view of an expanded right palpus (Fig. 29), the hema- todochae seem to appear very much reduced, and most of the sclerites occupy positions different from those of the previously described araneids. In addition, the conductor is much enlarged. Cyclosa conica (Pallas) Kaston (1948) gave a figure of the palpus of the male of Cyclosa conica. In the degree of palpal complexity and reduction of hematodochae, this species lies somewhere between Argiope trifasciata and Mangora gibberosa. As in Mangora, the sclerites are much displaced, making them difficult to identify. Expansion causes an ectad rotation and further displacement, but, after treatment, individual sclerites can be seen as entities and at least tentatively identified. The median apophysis is bifurcate, as in many araneids, and the distal lobe is sigmoid. As in Mangora, the conductor is large and membranous and interferes with identifica- tion of the sclerites in the unexpanded palpus. Family AGELENIDAE The genitalia of some agelenids have been studied intensively with regard to their form and function (Gering, 1953). The palpus is moderately complicated, with a reduced subtegulum and a well-developed conductor that serves, at least in Agelenopsis, as a locking device during copulation (Gering, 1953). Wadotes calcaratus (Keyserling ) Figs, 30593), 32 Muma (1947) revised the genus Wadotes and figured only individual sclerites rather than complete palpal organs. He con- sidered the terminal apophysis to be of unusual value in taxo- nomic decisions. 12 BREVIORA No. 259 In a ventral view of an unexpanded right palpus (Fig. 30), the subtegulum (st), tegulum (¢), and embolus (e) are easily seen, as is the terminal apophysis (ta). In an ectal view of an ex- panded right palpus (Fig. 31), the distal portion of the embolus (e) is extensive and threadlike. Muma (1947) refers to the em- bolus as lying along a distinct conductor, but it seems to me that the only structure one can suggest as a conductor is solidly fused to the terminal apophysis. In a mesal view of an expanded right palpus (Fig. 32), the feature most noticeable is the distinctly annulated subtegulum (sf). Coras lamellosus (Keyserling ) Muma (1946) revised the North American members of the genus Coras and figured the palpi of a number of species, includ- ing Coras lamellosus. The palpus is very much like that of Wadotes calcaratus in the general arrangement of sclerites. After expansion, a general ectad rotation brings the long, thin embolus into full view. The conductor can be seen as separate from the terminal apophysis, and there is a twisted median apophysis that before expansion was concealed by the conductor and terminal apophysis. A gelenopsis utahana (Chamberlin and Ivie) Figs. 33, 34 This is one of the species used by Gering (1953) in a monu- mental study of the structure and function of the external geni- talia of the genus Agelenopsis, and he gives several very useful figures of the expanded palpus. In an ectal view of the unexpanded right palpus (Fig. 33), the embolus (e) is seen to be long, heavy, and spirally coiled. The tegulum (ft) is somewhat indented to accommodate the embolus. In a mesal view of an expanded right palpus (Fig. 34), the fused basal and middle hematodochae (bh-mh) are visible. The sub- tegulum (st) has been pushed into view, and the conductor (c) has rotated mesad. Gering (1953) refers to the basal part of the embolus as the radix, and this may be morphologically correct. Tegenaria domestica (Clerck ) Figs: 355/36 Tegenaria domestica is a common house spider often referred to as Tegenaria derhami (Scopoli), and its habits have been studied extensively. The legs and palpal segments of this species are elongate. 1967 PALPI OF MALE SPIDERS 13 In a ventral view of the unexpanded right palpus (Fig. 35), the cymbium (cy) bears a long distal finger. The tegulum (¢), median apophysis (ma), and embolus (e) are fused into one sclerite capping the hematodochae. In an ectoventral view of an ex- panded right palpus (Fig. 36), the fused basal (bf) and middle hematodochae (mh) are inflated. The “cap” has been rotated ectad, and the hook-like conductor (c) has been extended on a short articulating membrane. No subtegulum was evident in this palpus, but an area on the basal hematodocha that is more heavily sclerotized than the rest of the hematodocha may correspond to the subtegulum. Cicurina robusta Simon Figs. 37; 38 Exline (1936) and Chamberlin and Ivie (1940) examined rep- resentatives of the genus Cicurina, and both gave excellent figures, concentrating on the tibia, which in males of this genus is uniquely developed. Frequently they neglected to provide figures of the palpal organ itself. In an ectal view of an unexpanded left palpus (Fig. 37), the hooked conductor (c), the radix (r), and the embolus (e) are all visible, at least in part. The unusual tibial apophysis (ti) ex- tends up the ectal side of the cymbium (cy). In an ectal view of an expanded right palpus (Fig. 38), the flattened subtegulum (st) has become visible, and the cupped tegulum (ft) accommodates the radix (r) and the base of the embolus (e). The long and spirally wound embolus tapers to a hairlike tip. The hooked con- ductor (c) is extended. Family LYCOSIDAE This is a large and homogeneous family of which Lycosa is the principal genus. The palpi of most lycosids are similar to one another in general plan. Lycosa ammophila Wallace Wallace (1942) studied the Jenta group of the genus Lycosa and described a number of new species, including Lycosa ammo- phila, from southeastern United States. Wallace gave figures of the palpus in ventral view, but emphasized the median apophysis, which he also figured separately in outline. In the unexpanded palpus, most of the features are readily visible. Expansion does, however, clarify relationship among the sclerites. 14 BREVIORA No. 259 Family CLUBIONIDAE The palpi of this family are simple, except for those of certain genera belonging to the subfamily Liocraninae and none of these are considered here. The palpal organs of most liocranines are similar to those of the Agelenidae. Clubiona bryantae Gertsch Figs. 39, 40 Gertsch (1941) proposed Clubiona bryantae as a new name for Clubiona agrestis Emerton, but he gave no figures. The new name was accepted by Edwards (1958), who gave several figures of the male palpus. In an ectal view of an unexpanded right palpus (Fig. 39), the subtegulum (st), tegulum (¢), and embolus (e) are subequal. The expanded right palpus (Fig. 40) shows the ectad rotation of the embolus (e¢), exposing its fingerlike tip. Aysha gracilis (Hentz) Figs. 41, 42 Aysha and a number of related genera (Anyphaena, Anyphae- nella, Gayenna, and others) are often placed in a separate family, the Anyphaenidae, on which no recent work has been done. In a ventral view of an unexpanded right palpus (Fig. 41), the tegulum (ft) is large and extends below the alveolus or cavity of the cymbium (cy). The embolus (e) has a large, rotate, flattened base, and then tapers to a fine point, cupped by a projection of the cymbium. In a mesal view of an expanded right palpus (Fig. 42), all the parts appear to have been pushed ectad, and the subtegulum (sf) is now visible. The track of the internal sperm duct through the tegulum is prominent. Family SALTICIDAE The Saiticidae is considered by some authorities to contain the most advanced of spiders (Kaston, 1948), but the palpi are highly simplified. Metacyrba undata (De Geer) Figs. 43, 44 In a ventral view of an unexpanded right palpus of Metacyrba undata (Fig. 43), little is visible except the tegulum (7), which is enlarged and greatly extended proximally into a hollow on the 1967 PALPI OF MALE SPIDERS 15 ventral surface of the palpal tibia (ti), and the fused embolus- conductor (e, c). In the ectodorsal view of an expanded paip (Fig. 44), the basal hematodocha (bh) is large and displaces all the sclerites ectad. The subtegulum (st) is visible as a small, sclerotic ring. Tutelina elegans (Hentz) Figs. 45, 46 Kaston (1952) gave figures showing the general appearance of this species, but he did not illustrate the genitalia. In a ventral view of an unexpanded right palpus (Fig. 45), the general ap- pearance is essentially the same as in Metacyrba undata, but no recognizable conductor is present on or near the embolus (e). Expansion of the palpus (Fig. 46) brings about an ectad rotation and, in an ectodorsal view, the subtegulum (st) is exposed. Salticus scenicus (Clerck ) This species is distributed throughout the United States and Europe and has been widely studied. The palpus is very simple and follows the general plan of the two previously described spe- cies of the family Salticidae. In the expanded palpus, the basal hematodocha is greatly expanded, and the subtegulum is brought into view. Family ULOBORIDAE This family was recently revised by Muma and Gertsch (1964), and a number of new species were described. Figures were given of unexpanded palpi and individual segments, but an error was made in the figure titles. The palpi of these spiders are remarkable for the extreme development of certain sclerites. The habits of the Uloboridae are unique, but among some species the modes of web-building converge with those of the Araneidae. This behavioral converg- ence is not reflected in the morphology of the palpal organs of the males. Hyptiotes cavatus Hentz Figs. 47, 48, 49 Muma and Gertsch (1964) illustrate the entire palpus of Hyp- tiotes cavatus and also give a figure of what they call the “tip of the embolus.” Close examination of actual specimens and of the textual description given by Muma and Gertsch leaves little doubt 16 BREVIORA No. 259 that this figure actually represents the tip of the median apophysis. In an ectal view of an unexpanded right palpus (Fig. 47), the most striking feature is the large, elongated median apophysis (ma) with its membranous tip. The tegulum (t) and subtegulum (st) are almost completely covered by the enlarged membranous conductor (c). The embolus (e) originates mesally and coils around the other sclerites of the palpal organ. In a mesal view of the same palpus (Fig. 48), the origin of the embolus (e), as well as its attenuated distal part entering the conductor (c), is easily seen. The conductor is accompanied by the long, hooked radix (r). An ectoventral view of an expanded right palpus (Fig. 49) shows that expansion has resulted in little change, but the whole of the distal part of the embolus (e) can now be seen and its relationship to the median apophysis (ma) is clear. Family DICTYNIDAE Chamberlin and Gertsch (1958) revised this family, of which Dictyna is the largest genus. While the palpal organs of all North American species of Dictyna are similar, Chamberlin and Gertsch separated species on the basis of the structure of the tip of the embolus. Dictyna sublata (Hentz) Figs: 50) 51 In a mesal view of an unexpanded right palpus (Fig. 50) of Dictyna sublata, the embolus (e) appears heavy and coiled and distally enters the large shieldlike conductor (c). The path of the sperm tube can be traced through the tegulum (ft) and the base of the embolus. A ventral view of an expanded right palpus (Fig. 51) shows that an almost complete ectad rotation of the base of the embolus has occurred. The subtegulum (st), as well as the basal and middle hematodochae (bh, mh), is visible. DISCUSSION For purposes of discussion, the males of species used in this study can be separated into three groups: those with very complex palpi, those with moderately compiex palpi, and those with simple palpi. Drapetisca alteranda, Linyphia marginata, Araneus sericatus, Araneus nordmanni, Neoscona arabesca, Cyclosa conica, Singa pratensis, Mangora gibberosa, and Hyptiotes cavatus fall in the 1967 PALPI OF MALE SPIDERS 17 group having very complex palpi. In this group almost all pos- sible sclerites are present in a more or less highly developed form. These species are especially characterized by the number of scle- rites distal to the embolus. Males of species of Drapetisca and Linyphia have palpi in which the lateral subterminal apophysis is developed into a large shield covering the whole organ. In species of Araneus and Singa, the male palpus has many structures con- cealed by other sclerites and by the manner in which the palpus is folded. In males of Mangora, Cyclosa, and Hyptiotes displace- ment of the sclerites has caused confusion when unexpanded palpi are used in the description of species. Expanding the palpus can be of great value in this group. In the group with moderately complex male palpi are: Enoplo- gnatha tecta, Theridion differens, Theridion frondeum, Steatoda borealis, Argiope trifasciata, Araniella displicata, Wadotes calcara- tus, Agelenopsis utahana, Coras lamellosus, Cicurina robusta, Lycosa ammophila, and Dictyna sublata. This group is character- ized by the approximately equal development of most sclerites. The apical division is poorly represented, and some of the sclerites of the embolic subdivision are reduced. Usually the basal hema- todocha is well developed. When the palpus is moderately com- plex, one has to use his own judgment whether or not expansion will be worthwhile. The process of expansion has the value of making relationships clear and exposing portions of sclerites that may be hidden, because in most species in this group the full com- plement of sclerites (except the subteguium) is usualiy visible be- fore expansion. The third group, species with simple male palpi, includes: Soulgas corticarius, Ceratinopsidis formosa, Tegenaria domestica, Clubiona bryantae, Aysha gracilis, Metacyrba undata, Salticus scenicus, and Tutelina elegans. These spiders have palpi charac- terized by the loss, fusion, or reduction of sclerites. Expansion seems to be of little utility in this group. The process of expansion seems to have its greatest value in studying species with complex palpi. The complex palpus may indicate specialization and occurs in three very highly specialized families (Theridiidae, Linyphiidae, Araneidae). However, spiders of a fourth very specialized family, the Salticidae, have a simple type of palpus. If the hematodochae of a complex palpus are reduced, expan- sion does not show structures that were not visible in the unex- panded palpus. In such cases, a delicate dissection of the palpus 18 BREVIORA No. 259 may be indicated. The need for expansion may also be partly ob- viated by the presence of other taxonomically useful sexual modi- fications, such as the ornate third leg of some salticid males or the tibial apophyses and cephalic modifications of the micryphantids. SUMMARY 1. Several methods have been suggested for expanding the palpi of male spiders for taxonomic purposes. The method used in this study involved immersing the palpus for 6-8 hours in a 10 per cent KOH solution, followed by a quick transfer to distilled water. Expanded palpi were stored in 70 per cent ethanol solution. 2. The expanded palpus of each of 29 species of spiders was compared with the unexpanded palpus of the same species to determine if any parts not observable in the unexpanded pal- pus could be seen in the expanded palpus. 3. The process of expansion was found to be most useful in species with complex palpi and least useful in species with simple palpi. 4. Under certain conditions, such as the reduction of hemato- dochae, many features of complex palpi were still obscure. 5. The experimenter must use his own judgment in deciding when the process of expansion is worthwhile, but certainly more detailed examination of palpi after expansion can be of benefit in the present confused state of spider taxonomy. LITERATURE CITED ALEXANDER, A. J., and D. W. Ewer. 1957. On the origin of mating behavior in spiders. Amer. Natur. Ui tees Mikes il 7/e ARCHER, A. F. 1951. Studies in the orbweaving spiders (Argiopidae). 1. Amer. Mus. Novit. No. 1487: 1-52. Barrows, W. M. 1925. Modification and development of the arachnid palpal claw, with especial reference to spiders. Ann. Entomol. Soc. Amer. 18: 483-516. BisHoPp, S. C., and C. R. Crossy. 1930. Studies in American spiders: genera Ceratinopsis, Ceratinop- sidis, and Tutaibo. J. New York Entomol. Soc. 38: 15-34. BLAUVELT, H. H. 1936. The comparative morphology of the secondary sexual organs of Linyphia. Festschr. Embrik Strand 2: 81-171. 1967 PALPI OF MALE SPIDERS 19 CHAMBERLIN, R. V., and W. J. GERTSCH 1958. The spider family Dictynidae in America north of Mexico. Bull. Amer. Mus. Natur. Hist. 116: 1-152. CHAMBERLIN, R. V., and W. IvIE 1940. Agelenid spiders of the genus Cicurina. Bull. Univ. Utah, Biol. Ser. 5(9): 1-108. 1942. A hundred new species of American spiders. Bull. Univ. Utah, Biol; Ser. -7@1) 2 1-187. 1943. New genera and species of North American linyphiid spiders. Bull. Univ. Utah, Biol. Ser. 7(6): 1-39. ComsTOock, J. H. 1940. The spider book, revised and edited by W. J. Gertsch. Com- stock Publishing Company, New York. 721 pp. Crospy. C. R., and S. C. BISHOP 1936. Studies in American spiders: miscellaneous genera of Erigo- neae. Festschr. Embrik Strand 2: 52-64. Epwarps, R. J. 1958. The spider subfamily Clubioninae of the United States, Can- ada and Alaska (Araneae: Clubionidae). Bull. Mus. Comp. Zool. 118: 365-436. ENGELHARDT, VY. 1910. Beitrage zur Kenntnis der weiblichen Copulationsorgane eini- ger Spinnen. Z. Wiss. Zool. 96: 32-117. EXLINE, H. 1936. Nearctic spiders of the genus Cicurina Menge. Amer. Mus. Novit. No. 850: 1-26. GERING, R. L. 1953. Structure and function of the genitalia in some American agelenid spiders. Smithson. Misc. Collec. 121(4): 1-84. GertTscH, W. J. 1941. New American spiders of the family Clubionidae. II. Amer. Mus. Novit. No. 1148: 1-18. 1949. American spiders. D. Van Nostrand Publishing Company, Inc., New York. 285 pp. 1951. New American linyphiid spiders. Amer. Mus. Novit. No. 1514: 1-11. GERTSCH, W. J., and I. Davis 1946. Report on a collection of spiders from Mexico. V. Amer. Mus. Novit. No. 1313: 1-11. KasTOoN, B. J. 1948. Spiders of Connecticut. Connecticut State Geol. Natur. Hist. Surv. Bull. 70: 1-874. 20 BREVIORA No. 259 KasTON, B. J. 1952. How to know the spiders. Wm. C. Brown Company, Dubuque, Iowa. 220 pp. Levi, H. W. 1957a. The spider genera Crustulina and Steatoda in North America, Central America, and the West Indies (Araneae, Theridiidae). Bull. Mus. Comp. Zool. 117: 367-424. 1957b. The spider genera Enoplognatha, Theridion, and Paidisca in America north of Mexico (Araneae, Theridiidae). Bull. Amer. Mus. Natur. Hist. 112: 1-123. 1961. Evolutionary trends in the development of palpal sclerites in the spider family Theridiidae. J. Morphol. 108: 1-9. 1965. Techniques for the study of spider genitalia. Psyche 72(2): 152-158. MERRETT, P. 1963. The palpus of male spiders of the family Linyphiidae. Proc. Zool. Soc. London 140(3): 347-467. Muma, M. 1946. North American Agelenidae of the genus Coras Simon. Amer. Mus. Novit. No. 1329: 1-20. 1947. North American Agelenidae of the genus Wadotes Chamberlin. Amer. Mus. Novit. No. 1334: 1-12. Muma, M., and W. J. GERTSCH. 1964. The spider family Uloboridae in North America north of Mexico. Amer. Mus. Novit. No. 2196: 1-43. PETRUNKEVITCH, A. 1925. External reproductive organs of the common grass spider, Agelena naevia Walckenaer. J. Morphol. 40: 559-573. Savory, T. H. 1928. The biology of spiders. Sidgwick and Jackson, London. 376 pp. WALLACE, H. K. 1942. A study of the lenta group of the genus Lycosa with descrip- tions of new species (Araneae, Lycosidae). Amer. Mus. Novit. No. 1185: 1-21. 1967 PALPI OF MALE SPIDERS ABBREVIATIONS USED IN FIGURES basal hematodocha conductor cymbium distal hematodocha embolus lateral subterminal apophysis median apophysis middle hematodocha paracymbium radix stipes subtegulum tegulum terminal apophysis tibia i i) BREVIORA No. 259 Fics. 1-3. Steatoda borealis. 1, Mesal view of an unexpanded right pal- pus. 2, Mesal view of an expanded right palpus. 3, Ectal view of an expanded right palpus. Fics. 4, 5. Enoplognatha tecta. 4, Ectal view of an unexpanded left palpus. 5, Ectal view of an expanded left palpus. Fics. 6, 7. Theridion differens. 6, Ventral view of an unexpanded right palpus. 7, Ventral view of an expanded right palpus. Fics. 8, 9. Theridion frondeum. 8, Ventral view of an unexpanded right palpus. 9, Ventral view of an expanded right palpus. 1967 PALPI OF MALE SPIDERS 23 Fics. 10, 11. Drapetisca alteranda. 10, Mesal view of an unexpanded right palpus. 11, Mesal view of an expanded right palpus. Fics. 12, 13. Linyphia marginata. 12, Ectal view of an unexpanded right palpus. 13, Ectal view of an expanded right palpus. Fics. 14, 15. Soulgas corticarius. 14, Ectal view of an unexpanded right palpus. 15, Mesal view of an ex- panded right palpus. Fics. 16, 17. Ceratinopsidis formosa. 16, Ectal view of an unexpanded right palpus. 17, Dorsoectal view of an expanded right palpus. 24 BREVIORA No. 259 Fics. 18-21. Araneus sericatus. 18, Ventral view of an unexpanded right palpus. 19, Mesal view of an unexpanded right palpus. 20, Ectal view of an expanded right palpus. 21, Mesal view of an expanded right palpus. Fics. 22, 23. Argiope trifasciata. 22, Mesal view of an unexpanded right palpus. 23, Dorsomesal view of an expanded right palpus. Fics. 24, 25. Araniella displicata. 24, Ectal view of an unexpanded right palpus. 25, Ectal view of an expanded right palpus. 1967 PALPI OF MALE SPIDERS DS SB) Fics. 26, 27. Singa pratensis. 26, Mesal view of an unexpanded right palpus. 27, Ectal view of an expanded right palpus. Fics. 28, 29. Mangora gibberosa. 28, Ectal view of an unexpanded right palpus. 29, Ventroectal view of an expanded right palpus. Fics. 30-32. Wadotes calcaratus. 30, Ventral view of an unexpanded right palpus. 31, Ectal view of an expanded right palpus. 32, Mesal view of an expanded right palpus. Fics. 33, 34. Agelenopsis utahana. 33, Ectal view of an unexpanded right palpus. 34, Mesal view of an expanded right palpus. 26 BREVIORA No. 259 Fics. 35, 36. Tegenaria domestica. 35, Ventral view of an unexpanded right palpus. 36, Ectoventral view of an expanded right palpus. Fics. 37. 38. Cicurina robusta. 37, Ectal view of an unexpanded right palpus. 38, Ectal view of an expanded right palpus. Fics. 39, 40. Clubiona bryantae. 39, Ectal view of an unexpanded right palpus. 40, Ectal view of an ex- panded right palpus. Fics. 41, 42. Aysha gracilis. 41, Ventral view of an expanded right palpus. 42, Mesal view of an unexpanded right palpus. Fic. 43. Metacyrba undata, ventral view of an unexpanded right palpus. 1967 PALPI OF MALE SPIDERS 27 Fic. 44. Metacyrba undata, ectodorsal view of an expanded right palpus. Fics. 45, 46. Tutelina elegans. 45, Ventral view of an unexpanded right palpus. 46, Ectodorsal view of an expanded right palpus. Fics. 47-49. Hyptiotes cavatus. 47, Ectal view of an unexpanded right palpus. 48, Mesal view of an unexpanded right palpus. 49, Ectoventral view of an expanded right palpus. Fics. 50, 51. Dictyna sublata. 50, Mesal view of an unex- panded right palpus. 51, Ventral view of an expanded right palpus. a ‘ ve ¥ : arti f ; ade te) by, Any 4 J BREVIORA Museum of Comparative Zoology CAMBRIDGE, Mass. 3 Fesruary, 1967 NUMBER 260 MONOGRAPH OF THE GENUS SPIROCERAMUS (MOLLUSCA: PULMONATA: UROCOPTIDAE) By William J. Clench The last complete monograph of the family Urocoptidae was that of H. A. Pilsbry, 1902-1904, in the second series of the Man- ual of Conchology. At that time only a single species was cred- ited to the genus Spiroceramus. This species was S. amplus (Pfeiffer) from the region of Bayamo in the south of Oriente Province in eastern Cuba. Since then, C. G. Aguayo has de- scribed a species from the northern portion of Oriente Province, and I have described two species from the Bahamas, one from Little Abaco in the northern part, and one from Acklin’s Island in the southern part of the Bahamian Archipelago. Two new species are described in this report from the Sierra de Cubitas, an isolated mountain system in north-central Camagiiey Province in central Cuba. With comparatively few exceptions the West Indian Urocop- tidae are strictly calciphiles, living on limestone rocks where they feed on the encrusting lichens. A very few are arboreal. The Bahamian species were living on or under stones; the Cuban spe- cies on the limestone ledges. Under optimum conditions they can exist in incredible numbers, usually preferring shade, but they also exist on the limestone with no shade at all. They are active at night and during periods of rain. UROCOPTIDAE Genus SPIROCERAMUS Pilsbry and Vanatta Spiroceramus Pilsbry and Vanatta, 1898, Proc. Acad. Nat. Sci. Philadel- phia, 50: 281 (type species, Macroceramus amplus Pfeiffer, original desig- nation). Shells thin, cylindrical, and composed of many narrow whorls. Spire acute, the apex entire, not deciduous. Colored a pale brown- ish with patches of white. Sculptured with fine oblique striae. 2 BREVIORA No. 260 Whorls 9 to 15, convex, early whorls straight or slightly angular. Aperture somewhat oblique, subcircular, and the lip narrowly re- flected. Suture indented, smooth or crenulate. Axis twisted or encircled with a thin spiral lamella. Pilsbry (1904, p. 172), on the basis of the then known single species S. amplus, stated that it had the shape of Holospira, the axis of Arangia, and the aperture and protoconch of Micro- ceramus. This statement can still stand with only the slight modi- fication that all species do not have a spiral lamella on the axis and certain species possess a crenulated suture. Subgenus SPIROCERAMUS Pilsbry and Vanatta Species with or without the spiral lamella on the axis and hay- ing the oblique sculpture of fine striae similar on all whorls. The suture with or without crenulations. SPIROCERAMUS (SPIROCERAMUS) AMPLUS (Pfeiffer) Plate 1, figures 6, 7 Macroceramus amplus Gundlach, 1856, [in] F. Poey, Memorias Sobre la Historia Natural de la Isla de Cuba, Habana 2: 8, no. 378 (Cuba). [Nomen nudum. | Macroceramus amplus ‘Gundlach’ Pfeiffer, 1858, Malakozoologische Blatter 5: 44 (Guisa [Bayamo, Oriente], Cuba); Pfeiffer, 1859, Mono- graphia Heliceorum Viventium 4: 689; Pfeiffer, 1868, Novitates Con- chologicae 3: 383, pl. 89, figs. 12-14; Arango, 1878, Contribucion a la Fauna Malacologica Cubana, Habana, p. 84 (Guisa and San Andres, Bayamo, Cuba). [Syntypes, MCZ 39318.] Microceramus (Spiroceramus) amplus Pfeiffer. Pilsbry, 1904, Man. of Conch: (2) 16: 172," pl: 26, figs. 24°" 29: Measurements Height Width mm mm 10.5 3:9 Syntype. Guisa, Bayamo, Cuba it 4. Syntype. Guisa, Bayamo, Cuba 11.1 3:3 Syntype. Sao Arriba, Holguin, Cuba Description: Shell reaching 11.1 mm in height, minutely um- bilicate, cylindrical, attenuate, the upper third rapidly tapering to the early whorls and sculptured. Whoris 14 to 15, convex, those on the upper one-third slightly angled. Color a light grayish brown and mottled with a dull white. Spire extended, the lower two-thirds nearly parallel sided, the conic one-third forming an 1967 THE GENUS SPIROCERAMUS 5 angle of about 35°. Aperture subcircular, with the outer lip slightly refiected. Inner or parietal lip reflexed over the minute umbilicus. Axis twisted and forming a broad lamella. Columella short and slightly angled. Suture deeply impressed. Sculpture con- sisting of numerous and fine, oblique striae. Protoconch with 112 whorls and sculptured with fine, oblique striae. Remarks: This is rather a remarkable species for any member of the Urocoptidae as it occurs in two distinct and well separated areas. Holguin is about 70 kilometers NE of Bayamo with much relatively flat country between them. Specimens examined: CuBA. Oriente Prov.: Guisa, Bayamo; Cerro Moncada and Cerro San Juan, Sao Arriba, both Holguin. SPIROCERAMUS (SPIROCERAMUS) BARBOURI Aguayo Plate 1, figure 5 Spiroceramus amplus barbouri Aguayo, 1935, Mem. Soc. Cubana Hist. Nat., 9: 126, pl. 9, figs. 3-4 (Paradones junto a la Curva de la Campana, Gibara [Oriente], Cuba). [Holotype, MCZ 237877. Additional paratypes from the same locality, MCZ 110570.] Measurements Height Width mm mm 55) 2: Holotype 4.5 1.8 Paratype 4.75 2, Paratype DES D Paratype Description: Shell reaching 5.8 mm in height, imperforate, cylindrical, attenuate, the upper third rapidly tapering to the early whorls and sculptured. Colored a light brown with most of the oblique costae and crenulations white. Whorls 9 to 11 and con- vex, the early 5 whorls straight-sided and vertical. Spire extended, the lower two-thirds nearly parallel sided, the conic one-third forming an angle of about 40°. Aperture subcircular, the outer lip simple and but slightly reflected. Axis twisted and forming a narrow lamella. Columella short and slightly angled. Suture impressed. Sculpture consisting of numerous and fine, oblique striae, many of which terminate above with crenulations at the suture. Protoconch with 1%2 whorls and sculptured with fine and straight striae. Remarks: Originally described as a subspecies of S. amplus, this present entity is a distinct species. S. barbouri is only one-half 4 BREVIORA No. 260 the height of amplus and has a crenulated suture which amplus does not have, and the axial lamella is much narrower. So far as now known, this species is limited to the type locality. Specimens examined: Holotype and paratypes. SPIROCERAMUS (SPIROCERAMUS ) PILSBRYI new species Plate 1, figure 3 Holotype: MCZ 188845, from Cerro de Tuabagiiey, Sierra de Cubitas, Camagiiey Province, Cuba. Paratypes: MCZ 256082, from the type locality, and MCZ 188847 from Cueva del Circulo, Sierra de Cubitas, Camagiiey Province, Cuba. Measurements Height Width mm mm Sail 4.8 Holotype 13 4.1 Paratype MCZ 188847 Description: Shell reaching 15 mm in height, minutely umbili- cate, tapering from the ninth whorl to the protoconch and sculp- tured. Whorls 13 to 14 and slightly convex. Color a light horn with numerous, irregular patches of white. Spire extended and forming an angle of about 36°. Aperture subcircular, the outer lip slightly reflexed and the inner or parietal lip reflected over the minute umbilicus. Axis twisted and forming a broad lamella. Columella angled and somewhat broadened. Suture indented and coarsely crenulated. Sculpture consisting of exceedingly fine, ob- lique striae. Protoconch with two very finely lirate whorls. Remarks: §. pilsbryi and S. vanattai, both from the Sierra de Cubitas are very different from one another. S. pilsbryi differs from vanattai by being much larger, having the broad axial lamella, coarse sutural crenuiations and very much finer sculpture. The conic taper toward the protoconch starts from near the center (ninth whorl) in pilsbryi, while in vanattai the entire shell from the body whorl tapers toward the protoconch. This must represent quite a rare species, as only a very few specimens are known. Specimens examined: Holotype and paratype. 1967 THE GENUS SPIROCERAMUS 5 SPIROCERAMUS (SPIROCERAMUS) VANATTAI new species Plate 1, figure 4 Holotype: MCZ 256083, from Los Cangilones, Sierra de Cubitas, Camagiiey Province, Cuba. Paratypes: From the same locality (MCZ 131387; MCZ 80765), and from El Cercado de Cubitas, near Cueva del Circuto (MCZ 131388; MCZ 188846); Finca Santa Gertrudis, Minas (MCZ 80766); Corrales de Cangilones (MCZ 131390); Paso de los Burros (MCZ 131386); Paso Paredones (MCZ 131391); El Tuabagiey, near the Cueva del Indio (MCZ 188844), all Sierra de Cubitas, Camagiiey Province, Cuba. Measurements ! Height Width mm mm 9.4 4.6 Holotype 10 4.5 Paratype 8.6 4.5 Paratype 9 4.2 Paratype Description: Shell reaching 10 mm in height, minutely umbili- cate, conic, tapering from the body whorl to the protoconch and sculptured. Whorls 10 to 11 and moderately convex. Color light horn with irregular patches of white which cover mainly the oblique striae. Spire extended and forming an angle of 30°. Aperture subcircular, the outer lip slightly reflexed. Inner or parietal lip reflected over the minute umbilicus. Axis twisted but lacking a lamella. Columella short and slightly angled. Suture moderately indented and finely crenulate. Sculpture consisting of numerous and prominent oblique striae. Protoconch with 2 whorls and smooth. Remarks: See Remarks under S. pilsbryi. Named for E. G. Vanatta who had been associated with H. A. Pilsbry at the Acad- emy of Natural Sciences, Philadelphia. Specimens examined: Holotype and paratypes. Subgenus INSULACERAMUS new subgenus Species with the axis twisted but not encircled with a spiral lamella. Early non-protoconch whorls rather coarsely lirate, later 1 All specimens measured are from the type locality. 6 BREVIORA No. 260 whorls finely lirate. Suture noncrenulate. So far this subgenus is limited to the Bahama Islands. Speci- mens are exceedingly rare, as only three are known to represent the two species. Type species: Microceramus (Spiroceramus) greenwayi Clench. SPIROCERAMUS (INSULACERAMUS) ROBERTSONI (Clench) Plate 1, figure 2 Microceramus (Spiroceramus) robertsoni Clench, 1963, Bull. Mus. Comp. Zool. 128: 406, pl. 3, fig. 1 (Delectable Bay Settlement, Acklins Island, Bahama Islands). [Holotype, MCZ 225313.] Measurements Height Width mm mm 6.7 120 Holotype Description: Shell reaching 6.7 mm in height, thin, imperforate, dull and sculptured. Color probably gray, whorls 11 and convex. First 5 to 6 whorls forming the conic portion of the spire. Aper- ture subcircular. Lip simple. Columella nearly straight. Axis simple and not twisted. Suture indented. Sculpture consisting of numerous, fine, oblique axial riblets. These riblets are somewhat stronger on the first 5 to 6 whorls. Remarks: This species is closely related to S. greenwayi Clench from Fox Town, Little Abaco Island, Bahamas. It differs by being a little more coarsely sculptured and having the protoconch a little smaller. SPIROCERAMUS (INSULACERAMUS) GREENWAYI (Clench) Plate 1, figure 1 Microceramus (Spiroceramus) greenwayi Clench, 1938, Mem. Soc. Cu- bana Hist. Nat., 12: 329, pl. 25, fig. 1 (Fox Town, Little Abaco Island, Bahama Islands). [Holotype, MCZ 116688. ] Measurements Height Width mm mm 6.5 1.9 Holotype Description: Shell thin, small and imperforate. Color a dull isabelline (true color, however, cannot be given as the single 1967 THE GENUS SPIROCERAMUS 7 specimen was found dead). Whorls 11, early 6 whorls which form the cone strongly convex; later whorls much less so. Aper- ture subcircular, lip simple and slightly flaring. Columella con- cavely arched, not distinct but forming the inner margin of the aperture. Axis simple with only a very slight twist. Sculpture: nuclear whorl faintly costate; next five strongly costate with oblique somewhat sinuous riblets, generally whitish; remaining whorls finely costate. Suture deeply indented but not crenulate. Aperture cast at an angle of 40° from a base line. Remarks: See under S. robertsoni (Clench). Specimens examined: Holotype. REFERENCES ARANGO, RAFAEL 1878. Contribucion a la Fauna Malacologica Cubana, Habana, pp. 1-315. PILsBRY, H. A. 1904. Manual of Conchology. (2) 16: 1-329. (Received 2 August, 1966.) BREVIORA No. 260 Plate 1 Fig. 1. Spiroceramus (Insulaceramus) greenwayi (Clench), Fox Town, Little Abaco Island, Bahama Islands (18.4 X). Holotype, MCZ 116688. Fig. 2. Spiroceramus (Insulaceramus) robertsoni (Clench), Delectable Bay, Acklins Island, Bahama Islands (10 X). Holotype, MCZ 225313. Fig. 3. Spiroceramus (Spiroceramus) pilsbryi n. sp., Cerro de Tuabagiiey, Sierra de Cubitas, Camagiiey, Cuba (5.3 X). Holotype, MCZ 188845. Fig. 4. Spiroceramus (Spiroceramus) vanattai n. sp., Los Cangilones, Sierra de Cubitas, Camagiiey, Cuba (5.3 X). Holotype, MCZ 256083. Fig. 5. Spiroceramus (Spiroceramus) barbouri Aguayo, Curva de la Com- pana, Gibara, Oriente, Cuba (5.3 X). Holotype, MCZ 237877. Figs. 6-7. Spiroceramus (Spiroceramus) amplus (Pfeiffer), Guisa, Bayamo, Oriente, Cuba (5.3 X). Syntypes, MCZ 39318. Figure 7 shows the axis and the spiral lamella. 1967 THE GENUS SPIROCERAMUS 9 10 BREVIORA No. 260 Plate 2 Map showing distribution of the species. 1. Spiroceramus (Spiroceramus) amplus (Pfeiffer), Guisa, Bayamo, and Sao Arriba, Holguin, both Oriente, Cuba. 2. Spiroceramus (Spiroceramus) barbouri Aguayo, near Curva de la Campana, Gibara, Oriente, Cuba. 3. Spiroceramus (Spiroceramus) pilsbryi n. sp., Cerro de Tuabagiiey, Sierra de Cubitas, Camagitiey, Cuba. 4. Spiroceramus (Spiroceramus) vanattai n. sp., Los Cangilones, Sierra de Cubitas, Camagiiey, Cuba. 5. Spiroceramus (Insulaceramus) robertsoni (Clench), Delectable Bay, Acklins Island, Bahama Islands. 6. Spiroceramus (Insulaceramus) greenwayi (Clench), Fox Town, Little Abaco Island, Bahama Islands. me BREVIORA Museum of Comparative Zoology CAMBRIDGE, MASss. 31 Marcu, 1967 NUMBER 261 THE MONTICOLA GROUP OF THE LIZARD GENUS ANOLIS IN HISPANIOLA By Richard Thomas and Albert Schwartz! In 1962, Williams summarized the then available data on three allied species of Hispaniolan anoles: A. monticola Shreve, A. chris- tophei Williams and A. etheridgei Williams (= A. darlingtoni Coch- ran). Of the three species, monticola was named on the basis of a single male from “the northern and eastern foothills, Massif de la Hotte, 1000-4000 feet, Haiti.” Additional specimens of this species were reported by Williams from the general region about and to the north of Camp Perrin and from Ile Grande Cayemite off the north coast of the Tiburon Peninsula of southwestern Haiti. The third species (described as A. darlingtoni by Cochran, but later changed to A. etheridgei by Williams [1962], for nomencla- tural reasons) was named from seven specimens collected by P. J. Darlington at four localities in the Cordillera Central of the Republica Dominicana. Finally, A. christophei from the Citadelle Laferri¢re in northern Haiti was based upon two females, both of which had been rather long in preservative. Williams (1962) reported on the coloration and pattern in life of A. monticola, and described the males of A. christophei. The latter species was still known only from the vicinity of the type locality. As for A. ether- idgei, the coloration and pattern of this species were still unknown, and its range remained as delimited by the original holotype and paratypes. The three species in the monticola group share a community of scale characteristics; those that distinguish them from other Hispaniolan anoles are the ventrals in transverse rows plus the 1 10,000 SW 84th St., Miami, Florida 33143. 2 BREVIORA No. 261 subocular scales separated from the supralabials by a row of inter- vening scales (Williams, 1962:6). There is also agreement in other and more minor details of scalation, but the limits given by Williams (1962) for these data have been somewhat expanded by our examination of far more material than was available to him at the time of that paper. Perhaps the most significant observation on the monticola group is that of Williams (1962:7), who stated that, “Whether there are any more members of this small sub-group of Hispanio- lan anoles will have to be determined by more thorough search of the island . . . If related species are equally local in distribu- tion, they may well have been missed.” In the spring of 1966, Thomas secured a series of still another member of the monticola group in north-central Haiti, rather close to the type locality of A. christophei. Since we now have large series of the members of this group and much additional information on variation, distribu- tion, and natural history, it is appropriate to summarize all of our data in the present paper and to describe the new northern Haitian species. We are very grateful to Miss Patricia A. Heinlein, Donald W. Buden, Ronald F. Klinikowski and David C. Leber for assistance in the field in the Republica Dominicana. It also gives us very great pleasure to acknowledge Mr. Leber’s work on our behalf in the careful execution of the color portraits reproduced herein; with three exceptions (the races of A. monticola and the new species) these were done in the field from living lizards. We most readily acknowledge the assistance of Ernest E. Williams in the present endeavor, not only for suggesting that we name the new species and for the loan of recently accumulated material in the Museum of Comparative Zoology of a group in which he is per- sonally interested but also for his guiding our way through the at times impenetrable mass of Hispaniolan anoles. We have fol- lowed his techniques in taking of data and in description, so that the present information may be more readily compared with his own descriptions of several Hispaniolan anoles. Finally, Dr. Williams has graciously allowed that portions of a projected paper by himself and Schwartz be withdrawn and utilized herein, so that information on all the members of the monticola group might be more compactly available. Our data are based primarily on specimens in the Albert Schwartz Field Series (ASFS), but we have also borrowed ma- terial from the Museum of Comparative Zoology (MCZ), the Field Museum of Natural History (FMNH), and the American Museum 1967 ANOLIS MONTICOLA GROUP 3 of Natural History (AMNH); for the courtesy extended us in these matters, we wish to thank Ernest E. Williams, Robert W. Inger, Hymen Marx, Charles M. Bogert, and Richard G. Zweifel. Dr. Zweifel’s assistance in supplying copies of pertinent field notes is also much appreciated. William G. Hassler kindly allowed us to examine his photographs and field sketches of some of the anoles concerned. This study has been supported in part by National Science Foundation grant GB 2444 to Dr. Ernest Williams. ANOLIS CHRISTOPHE! Williams The apparently most primitive member of the christophet- etheridgei-monticola trio of Hispaniolan anoles, A. christophei, was described from two female specimens from the Citadelle Laferriere, Dépt. du Nord, Haiti (Williams, 1960). Later, this description was amplified by a series of 20 additional specimens from the type locality (Williams, 1962). The species has thus been known only from topotypical specimens and its distribution has otherwise been unknown. In March 1963, four of these lizards were taken at Paraje La Palma, La Vega Province, Republica Dominicana, by C. E. Ray and R. Allen; the species was thus known to inhabit the eastern Cordillera Central as well as the Bonnet-a-l’Eveque in Figure 1. Dorsal view of head of Anolis christophei, ASFS X9193. Snout-vent length 45 mm. 4 BREVIORA No. 261 the Chaine de Marmelade in the north. The elevation of the type locality is 2840 feet (865 meters); the elevation of the La Palma specimens is unknown. The coloration and pattern of christophei have not been described from living specimens; additional material collected by ourselves and associates in 1963, 1964 and 1966, and in 1964 by J. D. Lazell, Jr. has contributed information not only on these characters but also on the ecology and habits of the species. A. christophei is now known to occur in four regions, and doubtless is distributed in the intervening areas, with one prob- able exception. It is a fairly common lizard in the Cordillera Central in the Republica Dominicana. Specimens have been col- lected from the northeast of Jarabacoa, south to west of Con- stanza, but not in the Valle de Constanza itself. Not only does the Species occur in the uplands, but it is also to be found on the steep eastern escarpment of the Cordillera between El Rio and Jayaco. Elevations of occurrence in this general area vary be- tween about 2000 feet (610 meters) and 4250 feet (1296 meters) (Fig. 14). A second major area of occurrence is in San Cristobal Province, near El Cacao. This locality lies on the southern slopes of the Cordillera Central; the lizards were taken at elevations of 1200 to 1400 feet (366 to 427 meters). The third area of occur- rence is in the Cordillera Septentrional; at only one locality in this range, A. christophei was collected at an elevation of 2200 feet (671 meters). The final region whence A. christophei has been taken is the Chaine de Marmelade in Haiti, at elevations of 2840 feet (865 meters) to 3500 feet (1068 meters). The altitudinal distribution of the species is known to encompass elevations of from 1200 to 4250 feet (366 to 1296 meters) and its geographi- cal distribution includes the Massif du Nord (presumably) in northern Haiti, the Cordillera Central in central and southern Republica Dominicana, and the Cordillera Septentrional in the northern portion of the latter country. The first three areas are presumably confluent, whereas the lizards in the Septentrional are apparently completely cut off from their southern and western rel- atives by the arid portion of the Valle de Cibao. It is barely pos- sible that A. christophei occurs to the east in the Republica Dom- inicana and thus circumvents the inhospitable part of the Valle de Cibao, thereby bridging the apparent gap between the Cordillera and Septentrional populations. Certainly the valley of the Rio Yuna is sufficiently mesic, and even now reasonably well forested, to support A. christophei. However, we consider it unlikely that these wet lowlands are presently inhabited by such a confirmedly montane lizard. 1967 ANOLIS MONTICOLA GROUP 5 A. christophei, although observed with some frequency during the day, is much more readily collected at night. Most specimens were taken sleeping on ferns (especially tree-ferns), herbs, and shrubs along the margins of streams in gallery forest in the pine- clad uplands of the Cordillera Central. The greenish brown dorsal coloration renders them relatively inconspicuous under such cir- cumstances, but the long dark almost filamentous tail often re- veals their presence. The hindlimbs are flexed in the sleeping posture, and the lizards sleep more often across the leaflets and/or leaves rather than aligning themselves along the stems of branch- lets; often they give the appearance of having been suddenly overtaken by the lowering temperatures of nightfall and of having stopped abruptly in their tracks. They give the impression of being draped across the greenery rather than having deliberately chosen a resting place for the night. This posture is in direct SZ LE —~ Df Giee yA ZG QZ a d CATY Fae ae We : = — = > x EY IAS KS OSS GSS ZEEE OCLs ae = = SS = S eee Figure 2. Lateral view of head of Anolis christophei, ASFS X9193. Snout-vent length 45 mm. contrast to that of such larger anoles as A. cybotes, A. distichus, and 4. ricordi which are encountered with them, but is quite simi- lar to that of A. etheridgei which occupies similar situations at night. The habitat of this anole is moist, shady, montane woods, par- ticularly stream-associated forests and mesic ravines. During the day A. christophei was observed on tree trunks (four to six feet above the ground) along wooded edges of the Rio Jimenoa below Paso Bajito, and on tree trunks and fallen but sloping logs near Paso Bajito itself. One was found on a fence post in riparian 6 BREVIORA No. 261 woods near El Rio. The Haitian specimens were collected in dense second-growth woods, where they were found on small trees, two to five feet above the ground. At night along a very narrow rivulet adjacent to a platanal east of El Rio, A. christophei occupied the usual stream-side plants, and was not encountered on larger shrubs and trees nor on the banana plants themselves. The specimens from near El Cacao in the southern Cordillera Central were collected at night as they slept on bushes and ferns along the roadside in a region of coffee and cacao plantations. In the Cordiliera Septentrional, specimens were also taken at night from stream-side shrubs and vines (the only place where they were encountered on the latter type of plants); on the vines, the lizards slept on the leaves and not on the woody stems. In no instance was there any attempt at concealment. Perhaps the most noteworthy characteristic of A. christophei is the extremely large dewlap for so small and relatively slender a lizard (Pl. 1, upper left). The dewlap is pale, grayish purple (PI. 47E2, Maerz and Paul, 1950, for Haitian specimens) or violet in life, with widely separated rows of pale yellow to whitish, or grayish, or bluish scales. The dewlap skin usually has a strangely metallic lustre which is distinctive. The dorsum is greenish brown to yellow-brown, with the heads distinctly greener than the back. There are four pairs of darker brown, bronzy or distinctly green- ish brown paramedian blotches which often are fused across the midline of the back to give four butterfly-shaped figures between the shoulders and the sacrum. The spaces between blotches are not clear but are variously marbled or marked with darker green or brown, so that the whole effect is extremely cryptic. There is a yellow-green to yellow-buff shoulder stripe which continues down the sides as a vague, paler, lateral stripe. There is a yellow to greenish yellow subocular semicircle and a black to very dark brown postocular smudge. The chin and throat are dark gray; the ventral color is yellow to greenish yellow, and the tail is ter- minally black. In the dark phase, the dorsal ground color is dark gray with a purplish tinge, and the markings are very dark gray (nearly black) with a reddish or bronzy tinge. There is neither sexual dichromatism nor ontogenetic change in coloration or pat- tern; a single tiny juvenile (snout-vent length 21 mm) is colored and patterned like adults. The iris is blue. Although A. christophei is a small anole, it is not obviously slim and attenuate; the long tail adds to the impression of attenuation. A. christophei is not strongly sexually dimorphic in size, al- though males reach a slightly larger size (49 mm snout-vent 1967 ANOLIS MONTICOLA GROUP Th length; ASFS X8835) than females (45 mm snout-vent length; ASFS V1957). There are 6 to 10 (mode 8) scales across the snout at the level of the second canthal, and 4 to 8 (mode 6) rows of loreals. The supraorbital semicircles are separated by one or two scales (mode 1), and there are 3 to 7 (mode 4) scales be- tween the interparietal and the semicircles. Fourth toe lamellae on phalanges II and III vary from 18 to 25 (mode 22, with 16 speci- mens, but 15 specimens have 23 lamellar scales). Figure 3. Mental region of head of Anolis christophei, ASFS X9193. Snout-vent length 45 mm. Specimens examined: HAITI, Dépt. du Nord, Citadelle Lafer- riere, 20 (MCZ 66900-19); Dépt. de l’Artibonite, 8 to 9 km W Marmelade, 3500 feet (1068 m), 13 (ASFS V9900-12); REPUB- LICA DOMINICANA, La Vega Province, Municipio Constanza, Paraje La Palma (not mapped), 15 (MCZ 75142-43, 79349-51, 79353-62); 6 km W Constanza, 4250 feet (1296 m), 2 (ASFS X8834-35); 11 km NE Jarabacoa, ca. 2000 feet (610 m), 6 (ASFS V4198-201, V4326-27); 10 km NE Jarabacoa, 1 (ASFS V4216); 4 km E Paso Bajito, Rio Jimenoa, 2700 feet (724 m), 10 (ASFS X8850-57, V1866-67); 4 km SW El Rio, 4000 feet (1220 m), 5 (ASFS X8552-55, X8737); 4 km SW EI Rio, 3500 feet (1068 m), 3 (ASFS X8886-88); 6 km E El Rio, 3600 feet (1098 mi); 7 CASES -X9205-11); 7.1 mi: (1-4 kmoee, El Rio; 3500 feet (1068 mm) CASES X8113))223- km, E. El Rio, 3050 feet (930 m), 3 (ASFS X9193-95); Espaillat Province, 2 km N Puesto Grande, 2200 feet (671 m), 6 (ASFS V1956-61); San Cristébal Province, 15.5 km SE El Cacao, 1400 feet (427 m), 2 (ASFS V2498-99); 2.1 km SE El Cacao, 1200 feet (366 m), 6 (ASFS V2492-96, V2502). s BREVIORA No. 261 ANOLIS ETHERIDGE! Williams Anolis etheridgei Williams (= Anolis darlingtoni Cochran, 1939) was described from four localities in the Cordillera Central in the Republica Dominicana: Loma Vieja, south of Constanza; Loma Rucilla; Valle Nuevo; and Constanza. The included eleva- tions for these four localities are from 3000 feet (915 meters) to 8000 feet (2440 meters). The type series comprised a total of seven specimens. From this it might be assumed that A. etheridgei is uncommon; this is far from the case since it is one of the more abundant anoles of these interior uplands, although its total dis- tribution is still somewhat circumscribed and surely incompletely known. No previous information on coloration and pattern in life has been reported. New specimens (three of which were taken by J. D. Lazell, Jr., in 1964, and the balance by ourselves and associates in 1963) indicate that the known distribution of A. etheridgei is confined to the interior highlands of the Cordillera Central in the Republica Dominicana, from Loma Rucilla on the west to east of El Rio on the east, from Paso Bajito on the north to southeast of Constanza on the south (Fig. 14). Although one of the paratypes is re- corded from Constanza (and this town is within the limits of the range of etheridgei as noted above), no additional specimens have = 5 ie”. yA Me Ee ws — SIE oT k = SARE SSS SSE mecaccor Figure 4. Dorsal view of head of Anolis etheridgei, ASFS X9146. Snout- vent length 42 mm. 1967 ANOLIS MONTICOLA GROUP 9 been taken from the Valle de Constanza, despite careful search for it there. Recently collected specimens indicate an altitudinal range of 3050 feet (930 m) to 6100 feet (1890 m); these elevations are embraced by the data for the type series, although in almost every case of the latter, the elevation for each locality is an inclusive range. A. etheridgei is similar to A. christophei in habitat preferences, and although it was encountered frequently sleeping on tree ferns, shrubs and herbs in gallery forest at night, it is not confirmedly re- stricted to this sort of situation. Specimens were collected in a pro- tected and secluded depression (about 7 meters in diameter) in rainforest north of Constanza; the depression was protected by a dense thicket along its margins, and the lizards were sleeping on bushes in the center of the depression. During the day occasional specimens were encountered in forest on bushes or small trees ad- jacent to paths, but many more were observed and collected with facility while sleeping at night. The discussion of the sleeping habits of A. christophei applies equally well to A. etheridgei. Sleeping lizards are found draped over the leaflets of ferns and tree-ferns, the leaves of herbs and shrubs, and are not aligned along the branches or twigs. Their long tails reveal their pres- ence with ease. None was encountered sleeping on vines, and usually the lizards were seen within two feet of the ground sur- face. The legs are flexed in the sleeping posture. <> >= SSS = Figure 5. Lateral view of head of Anolis etheridgei, ASFS X9146. Snout-vent length 42 mm. 10 BREVIORA No. 261 A. etheridgei, in contrast to A. christophei, is sexually dichro- matic. Males are transversely crossbanded with four darker cross- bands between the neck and sacrum. These bands may be fairly prominent or much obscured due to interband pigmentation. The neck is brown to bronzy above, with bright pea-green on the sides of the head and neck, blending into a duller green on the sides. The back is some shade of tan to brown, always dull in hue. The sides of the head and neck may be pale powder-blue rather than green. The ventral color is tannish opalescent to opalescent. The general coloration of males is a combination of tans, browns, and greens of varying hues. The dewlap is small, all white or white with a grayish basal area (Pl. 1, upper right). Females have a dorsal longitudinal bronzy zone, bordered dorsolaterally by rich dark brown longitudinal bands which extend onto the postorbital area. The dorsal bronzy band expands on the head and forms a U-shaped bronzy nuchal figure which abuts against the upper eye- lids. Females at times have a series of four or five middorsal diamonds superimposed on the dorsal band; the halves of each diamond may be staggered to give a more complex dorsal pat- tern. The sides of the neck are white, and often females lack green or any greenish tints at all. The venters are like those of males except that occasional specimens have the venter light brown with a pinkish tinge. Both sexes have a cream subocular spot (which is more prominent in males) and the iris is blue. The chin and throat are very pale green in males and cream with brown striae in females. Figure 6. Mental region of head of Anolis etheridgei, ASFS X9146. Snout-vent length 42 mm. 1967 ANOLIS MONTICOLA GROUP i Of the seventy-eight lizards available to us, the two largest males and the largest female (all with snout-vent lengths of 43 mm) come from a series of four lizards from 9.1 mi. (14.6 km) N of Constanza. Despite the long series from elsewhere in the Cordillera, strangely the three largest lizards are from the same small series; the difference in bulk of animals in this small lot of lizards, compared to those from elsewhere, was rather striking in life as well. There are from 8 to 15 (mode 10) scales across the snout, and there are 5 to 10 (mode 6) rows of loreals. The scales between the supraorbital semicircles vary from O (semicircles in contact) to 4 (mode 2), and there are from 3 to 6 (mode 4) scales be- tween the interparietal scale and the semicircles. Fourth toe lamel- lae on phalanges II and III vary from 15 to 21 (mode 18). Specimens examined: REPUBLICA DOMINICANA. La Vega Province, Municipio Constanza, Paraje La Palma (not mapped), 3 (MCZ 79345-47); 7.2 mi. (11.4 km) S Constanza, 5000 feet (1525 m), 1 (ASFS X8241); Loma Vieja, 6000 feet (1830 m), south of Constanza, 1 (FMNH 73378); 12.6 mi. (20.2 km) SE Constanza, 6100 feet (1891 m), 1 (ASFS X9146); 16 km N Constanza, 6000 feet (1830 m), 1 (ASFS X8950); 9.1 mi. (14.6 km) N Constanza, 3500 feet (1068 m), 4 (ASFS X8791- 94); 4 km SW El Rio, 4000 feet (1220 m), 30 (ASFS X8522- 51); 4 km SW EI Rio, 3500 feet (1068 m), 3 (ASFS X8889-91 ); 6 km E El Rio, 3600 feet (1098 m), 1 (ASFS X9212); 23 km EEE Rio; 3050 feet (930) m), 14 (ASES X9179-92): 22) km NW Bonao, 3900 feet (1189 m), 11 (ASFS V4282-92); 23 km NW Bonao, 4100 feet (1251 m), 1 (ASFS V4272); 19.6 km NE Bonao, 3300 feet (1007 m), 6 (ASFS V4296-301); 11 km E Paso Bayjito, 4500 feet (1372 m), 1 (ASFS X8849). ANOLIS MONTICOLA Shreve In 1962 Williams reviewed the then available material of A. monticola and recorded its colors in life, based on notes taken by W. G. Hassler in 1935, and by A. S. Rand and J. D. Lazell in 1960. He then indicated that, on the basis of very recent material not included in that study, there appeared to be differences be- tween populations of the northern and southern slopes of the Massif de la Hotte. Our own material, plus additional speci- mens (MCZ 74866-70), shows that there are indeed two forms of monticola: one possessing the four prominent dark ocellar patches as shown by Williams (1962: fig. 2), and one lacking the pair of nape patches. 12 BREVIORA No. 261 Williams (1962) cited five specimens collected by W. G. Hass- ler (AMNH 49818, 49845, 50108-09, MCZ 65139, formerly AMNH 50110) with the locality given as “25 miles north of Aux Cayes on Jeremie Road” and MCZ 56140 from “mountains on Jeremie road about 8 miles from Camp Perrin.” Reference to the map of Haiti published by the Service de Géodésie et Cartog- raphie 1:100,000 shows that 25 miles north of Les Cayes lies well onto the northern slope, as does the locality 8 miles from Camp Perrin. We have consulted Hassler’s original field notes and find that the true situation is otherwise: AMNH 49845 is in actuality listed as being from “high Mts. on Jeremie road 32 miles from Aux Cayes 2000-3000’ approx.”; AMNH 49818, 50108-10 are said to be from “about 4 miles from Camp Perrin.” ft is thus evident that these localities have suffered in the transcrip- tion to the catalogue. The specimens from 4 miles north of Camp Perrin are from the south slope, if the distance is accurate; the Tombeau Cheval specimens cited by Williams, 1962, are apparently from south of the high point on the Jérémie road but near it. The specimens of these two series agree in having the four ocellar patches. Recent material from farther west, obtained by Thomas, lacks the nape ocellar patches and agrees in this respect with MCZ 74866-70 from Trou Bois, 115 miles south of Beaumont on the Les Cayes-Jérémie road, and with AMNH 49845 from 32 miles from Les Cayes on the Jérémie Road (which is probably in the vicinity of the locality for the Trou Bois series). The type speci- men of monticola, although with the color pattern obscured by preservation and age, lacks the pair of nape patches (the neck patches and dorsal dark bands can still be seen), and thereby agrees with the material just mentioned. The locality for this specimen is not precisely known, but “northern and eastern foot- hills of the Massif de la Hotte” is in reality the northern and eastern foothills of Pic Macaya (P. J. Darlington, pers. comm. to E. E. Williams), the highest peak of the range, which lies to the west of the Tombeau Cheval region (Fig. 10). We are convinced that two distinct populations of Anolis monticola have been sampled: one, of the northwestern and western extreme of the Massif de la Hotte, and another farther to the east (and possibly to the south). We regard the Grande Cayemite record for monticola (MCZ 58026, a female, collected by W. J. Eyerdam) as dubious. Thomas has visited this island and the habitat is for the most part very arid, not at all similar to habitats occupied elsewhere by monticola. 1967 ANOLIS MONTICOLA GROUP 13 This is not in itself incontrovertible evidence against the occurrence of the species on that island, and it may be noted that Diploglossus sepsoides Gray, normally an inhabitant of distinctly moist situa- tions, does occur there. However, the single Eyerdam specimen of Sphaerodactylus copei Steindachner was noted as being indis- tinguishable from S. copei picturatus Garman from the nearby mainland (Schwartz and Thomas, 1964:326), yet the recently obtained series of S. copei from Grande Cayemite is undeniably a distinctive new subspecies. The Eyerdam collection is also re- sponsible for the aberrant locality of the Citadelle Laferricre for specimens of Anolis hendersoni Cochran (Williams, 1963), a trenchantly south island species. We will not be apodictic but regard the Grande Cayemite specimen with suspicion. Figure 7. Dorsal view of head of Anolis monticola quadrisartus, type, MCZ 62998. Snout-vent length 43 mm. Anolis monticola shows strong sexual dimorphism in size; males reach a maximum of 55 mm and females 39 mm. There are 7-11 (mode 9) scales across the snout at the level of the second canthal and 6-9 (mode 7) loreal rows. The supraorbital semicircles are separated by 2-4 (mode 3) scales, and there are 3-6 (mode 4) scales between the interparietal and the semicircles. Fourth toe lamellae in phalanges IT and II vary from 17-21 (mode 18 or 19). 14 BREVIORA No. 261 Plate 1. Lateral views of heads of males of four species of Hispaniolan Anolis, as follows: upper left, Anolis christophei, ASFS X8113, 7.1 mi. E El Rio, La Vega Prov., Republica Dominicana, painted from living specimen; upper right, Anolis etheridgei, ASFS X8522, 4 km SW El Rio, 4000 feet, La Vega Prov., Reptblica Dominicana, painted from living spec:'men; center left, Anolis monticola monticola, ASFS V9624, ca. 5 km SSE Marché Léon, Dépt. du Sud, Haiti, painted from Thomas’ field notes and color transparencies of living specimens; center right, Anolis monticola quadrisartus, MCZ 63004, paratype, Tombeau Cheval, Dépt. du Sud, Haiti, painted from color noted by collectors, as quoted by Williams (1962:3-4); lower, Anolis rimarum, AMNH 96469, paratype, 8 to 9 km W Marmelade, Dept. de l’Artibonite, Haiti, 3500 feet, painted from Thomas’ field notes on living specimens. 1967 ANOLIS MONTICOLA GROUP 15 ANOLIS MONTICOLA MONTICOLA Shreve Diagnosis: A subspecies of Anolis monticola characterized by the absence of paired, large, black, light-centered ocelli on the nape, and by a yellow to reddish orange dewlap; females have a relatively straight-sided (in contrast to wavy or scalloped) mid- dorsal zone. Sens : I () ' Ka nan fi 0, CH ith Ve f \) te i ’ } i My () ; t SSeS so. 4s (ee = TRS SS S= 2S Se SS OHS SS SSIES — <= Ss, Soe, SS OSES SSS EB: 1962. The amino acid composition of the organic matrix of some recent and fossil shells of some west coast species of Myzrilus. Doctoral thesis, California Institute of Technology. INABA, A. 1959. Cytological studies in molluscs. II. A chromosome survey in the stylommatophoric Pulmonata. J. Sci. Hiroshima Univ. (B1)18: 71-93. KNIGHT, J. B., R. L. BATTEN and E. L. YOCHELSON 1954. Status of invertebrate paleontology, 1953. V. Mollusca: Gas- tropoda. Bull. Mus. Comp. Zool., 112: 173-179. 22 BREVIORA No. 262 Marcus, E., and E. Marcus 1956. On the tectibranch gastropod Cylindrobulla. Acad. Brasil. Ci. An., 28: 119-128. MARGOLIASH, G., and E. L. SMITH 1964. Structural and functional aspects of cytochrome c in relation to evolution. Internat. Congr. Biochem., 6th, New York: 206-207. MCALEsTER, A. L. 1964. Preliminary suggestions for a classification of nuculoid bi- valves. J. Paleont. 38: 397-400. NEWELL, N. D. 1965. Classification of the Bivalvia. Amer. Mus. Novit., No. 2206: 1-25. NISHIKAWA, S. 1962. A comparative study of the chromosomes in marine gastropods, with some remarks on cytotaxonomy and phylogeny. J. Shi- monoseki Coll. Fish., 11: 149-186. ODHNER N. HJ. 1912. Morphologische und phylogenetische Untersuchungen Uber die Nephridien der Lamellibranchien. Z. Wiss. Zool., 100: 287-391. PELSENEER, P. 1911. Les Lamellibranches de l’expédition du Siboga, partie ana- tomique. Siboga Expéditie, Monogr. 53a: 125 pp. PERRIER, R., and H. FISCHER 1911. Recherches anatomiques et histologiques sur la cavité palléale et ses dépendences chez les Bulléens. Ann. Sci. Nat., Zool., (9)14: 1-190. PURCHON, R. D. 1957. The stomach in the Filibranchia and Pseudolamellibranchia. Proc. Zool. Soc. London, 129: 27-60. 1958. Phylogeny in the Lamellibranchia. Proc. Centenary Bicen. Congr. Biol. Singapore, 1958: 69-82. 1959. Phylogenetic classification of the Lamellibranchia, with special reference to the Protobranchia. Proc. Malacol. Soc. London, 33: 224-230. REMANE, A. 1956. Die Grundlagen des natiirlichen Systems, der vergleichenden Anatomie und der Phylogenetik. 2 ed., Geest and Portig, Leip- zig: 364 pp. RIGBY, J. E. 1965. Succinea putris: a terrestrial opisthobranch mollusc. Proc. Zool. Soc. London, 144: 445-486. ROBERTSON, R. 1963. The hyperstrophic larval shells of the Architectonicidae. Ann. Rep. Amer. Malacol. Union, 1963: 11-12. ——— and A. S. MERRILL 1963. Abnormal dextral-hyperstrophy of post-larval Heliacus (Gas- tropoda: Architectonidae). Veliger, 6: 76-79. 1967 MOLLUSCAN SHELL PROTEINS 23 RUDALL, K. M. 1963. The chitin/protein complexes of insect cuticles. Advances Insect Physiol., 1: 257-313. SOKAL, R. R. and P. H. A. SNEATH 1963. Principles of numerical taxonomy. Freeman, San Francisco: xvi + 359 pp. SPENCER, W. P. 1949. Gene homologies and the mutants of Drosophila hydei. In: G. L. Jepsen, G. G. Simpson and E. Mayr (eds.), Genetics, paleontology and evolution. University Press, Princeton, Ch. 3: xiv + 474 pp. STASEK, C. R. 1963. Synopsis and discussion of the association of ctenidia and labial palps in the bivalved Mollusca. Veliger, 6: 91-97. TayLor, D. W. and N. F. SOHL 1962. An outline of gastropod classification. Malacologia, 1: 7-32. THIELE, J. 1929- Handbuch der systematischen Weichtierkunde. Gustav Fischer, 1935; Jena. 2 vols. WILBUR, K. M. 1964. Shell formation and regeneration. Jn: K. M. Wilbur and C. M. Yonge (eds.), Physiology of Mollusca. Academic press, New York, Ch. 8: xili + 473 pp. (Received 13 September, 1966.) BREVIORA 1967 MOLLUSCAN SHELL PROTEINS Fe) PLATE | GASTROPODA Figure 1. Achatinella lorata Pfeiffer, 1848, Oahu, Hawaiian Islands, Pease collection, MCZ (Mollusk Department) — aperture view. X 1. Figure 2. Acmaea pustulata (Helbling, 1799), Puerto Sosya, Dominican Republic, MCZ, coll. 1937 — top view. X 2/3. Figure 3. Akera soluta (Gmelin, 1791), Zanzibar, MCZ No. 4444 — aperture view. X 1/3. Figure 4. Architectonica nobilis (Roding, 1798), Middle Atlantic Coast, 2-10 m, MCZ No. 90867 — aperture view. X 2. Figure 5. Astraea caelata (Gmelin, 1798), Pelican Shoals, Florida, MCZ, coll. 1939 — aperture view. X 2/3. Figure 6. Bulla striata (Bruguiere, 1792), Puerto Vieja, Dominican Re- public, MCZ, coll. R. H. Parker, April 1965 — aperture view. X 2/3. Figure 7. Cavolina tridentata (Forskal, 1776), pelagic, West Indies, Alexander Agassiz coll. 1879, MCZ (uncat.) — aperture view. X 2. Figure 8. Cerion regium (Benson, 1849), northeast end of Castle Island, Crooked Island Group, Bahamas, B.W.I.—coll. Robertson and Scott, 1958, MCZ (uncat.) — aperture view. X 2/3. Figure 9. Crepidula fornicata (Linné, 1767), Hadley Harbor, Elizabeth Islands, Massachusetts, Sta. P-2211-65, Systematics-Ecology Program, Ma- rine Biological Laboratory — interior view; 9A, exterior view, X 1 1/3. Figure 10. Crepidula plana (Say, 1822), Hadley Harbor, Elizabeth Is- lands, Massachusetts, Sta. P-275-65, Systematics-Ecology Program, Marine Biological Laboratory — interior view; 10B, exterior view. X DY Bs Figure 11. Colus trophius (Dall, 1919), off San Francisco, California, 1874-1929 m, coll. R. H. Parker, MCZ (uncat.) — aperture view. X 1 1/3. Figure 12. Cypraea zebra (Linné, 1758), Miami, Florida, MCZ (uncat.) — aperture view. X 1/3. Figure 13. Dolabella callosa (Lamarck, 1801) = scapula (Martyn), Calapan, Mindoro, Philippine Islands, MCZ No. 96107 — side view. X 2/3. Figure 14. Epitonium angulatum (Say, 1831), near mouth of New Brazos River, Freeport, Texas, MCZ No. 230893 — aperture view. X 2. Figure 15. Fissurella barbadensis (Gmelin, 1791), Romey Point, Bonn- guen Air Force Base, Puerto Rico, MCZ No. 1956— top view (exterior). X23: 26 BREVIORA No. 262 PLATE | (Continued) Figure 16. Haliotis cracherodi (Leach, 1814), San Diego, California, MCZ No. 58659 — exterior view. X 1/3. Figure 17. Helisoma trivalvis (Say, 1817), La Porte, Indiana, MCZ No. 63234 — side view; 17B, aperture view. X 1 1/3. Figure 18. Janthina janthina (Linné, 1767), Cape Florida Key, Biscayne, Florida, MCZ No. 155173 — aperture view. X 5/6. Figure 19, Littorina littorea (Linné, 1758), Hadley Harbor, Elizabeth- Islands, Massachusetts, Sta. P-0610-65, Systematics-Ecology Program, Ma- rine Laboratory — aperture view. X 1. Figure 20. Lunatia triseriata (Say, 1826), Hadley Harbor, Elizabeth Islands, Massachusetts, Sta. P-247-65, Systematics-Ecology Program, Marine Biological Laboratory — aperture view. X 2. Figure 21. Murex brevifrons (Lamarck, 1822), Mayaguez, Puerto Rico, coll. M. R. Carriker, Systematics-Ecology Program, Marine Biological Laboratory — aperture view. X 1/3. Figure 22. Nassarius trivittatus (Say, 1822), Hadley Harbor, Elizabeth Islands, Massachusetts, Sta. P-0605-65, Systematics-Ecology Program, Ma- rine Biological Laboratory — aperture view. X 2. Figure 23. Nerita plexa (Dillwyn, 1817), Mauritius, MCZ No. 139024 — aperture view. X 1/3. Figure 24. Oxynoe viridis (Pease, 1864), from Pease coll., MCZ (uncat.) aperture view. X 2 1/3. Figure 25. Polinices duplicatus (Say, 1822), Provincetown, Massachu- setts, MCZ (uncat.) aperture view. X 1 1/3. Figure 26. Siphonaria alternata (Say, 1826), Bermuda, coll. Bryant, 1903, MCZ No. 24213 — top view (exterior). X 1 1/3. Figure 27. Succinea ovalis (Say, 1817), bank of Grand River, W. Bridge Street Ferry, Ottawa County, Michigan, MCZ No. 166697 — aperture view. XG. Figure 28. Turritella terebra (Lamarck, 1822), Manila Bay, near Cavite, Luzon, Philippine Islands, MCZ No. 138070—aperture view. X 1/3. Figure 29. Umbraculum indicum (Lamarck, 1819), Hawaii, C. B. Adams coll., MCZ No. 1173 —top view (exterior); 29A, side view. X 1/3. Figure 30. Urosalpinx cinerea (Say, 1822), Hadley Harbor, Elizabeth Islands, Massachusetts, Sta. P-0612-65, Systematics-Ecology Program, Marine Biological Laboratory — aperture view. X 4 2/3. 1967 MOLLUSCAN SHELL PROTEINS 749 PLATE | (Continued) Figure 31. Viviparus georgianus (I. Lea, 1834), Lake Woodruff near Fick Island, Volusa County, Florida, MCZ No. 186800 — aperture view. Ke2 3. AMPHINEURA Figure 32. Chaetopleura apiculata (Say, 1830), Hadley Harbor, Eliza- beth Islands, Sta. P-7-63, Systematics-Ecology Program, Marine Biological Laboratory — top view (exterior). X 2. No. 262 BREVIORA 28 1967 MOLLUSCAN SHELL PROTEINS 29 PLATE Il CEPHALOPODA Figure 33. Argonauta hians (Dillwyn, 1817), southern Tropical Atlan- tic, MCZ (uncat.) — side view of egg case. X 1/3. Figure 34. Nautilus pompilius (Linné, 1758), Southwest Pacific Ocean, MCZ (uncat.) — side view. X 1/3. Figure 35. Spirula spirula (Linné, 1758), Saint Kitts Island, B.W.L., MCZ (uncat.) — side view. X 1 1/3. PELECYPODA Figure 36. Aequipecten irradians (Lamarck, 1819), Hadley Harbor, Eliz- abeth Islands, Massachusetts, Sta. P-65-64, Systematics-Ecology Program, Marine Biological Laboratory — exterior view, right valve. X 1. Figure 37. Anadara transversa (Say, 1822), Hadley Harbor, Elizabeth Islands, Massachusetts, Sta. P-77-64, Systematics-Ecology Program, Marine Biological Laboratory — exterior view, left valve. X 9 1/3. Figure 38. Arctica islandica (Linné, 1758), Georges Bank, Massachu- setts) MCZ No. 14325 — interior view, right valve. X iW/ 3. Figure 39. Corbicula consobrina (Cailliand, 1827), Nile River, Egypt, MCZ No. 14676 — interior view, right valve. X 1 1/3. Figure 40. Crassostrea virginica (Gmelin, 1791), Holding Tank, Supply Department, Marine Biological Laboratory, Woods Hole, Massachusetts — interior view, right valve. X 1/3. Figure 41. Laevicardium mortoni (Conrad, 1831), Hadley Harbor, Eliz- abeth Islands, Massachusetts, Sta. P-141-64, Systematics-Ecology Program, Marine Biological Laboratory — exterior view, left valve. X 8 1/3. Figure 42. Limopsis compressus (Dall, 1908), off Salina Cruz, Mexico, 1020-1240 m, coll. R. H. Parker, MCZ (uncat.) — exterior view, left valve. e273: Figure 43. Lyonsia hyalina (Conrad, 1831), Hadley Harbor, Elizabeth Islands, Massachusetts, Sta. P-200-64, Systematics-Ecology Program, Ma- rine Biological Laboratory — exterior view, left valve. X 5. Figure 44. Macoma tenta (Say, 1834), Hadley Harbor, Elizabeth Is- lands, Massachusetts, Sta. P-91-64, Systematics-Ecology Program, Marine Biological Laboratory — exterior view, left valve. X 2 1/3. Figure 45. Malletia, species “M,” Atlantic Abyssal Plain, 4970 m, coll. H. L. Sanders, in coll. at Woods Hole Oceanographic Institution — exterior view, left valve. X 2 1/3. 30 BREVIORA No. 262 PLATE Il (Continued) Figure 46. Mercenaria mercenaria (Linné, 1758), Holding Tank, Supply Department, Marine Biological Laboratory, Woods Hole, Massachusetts — exterior view, left valve. X 2 1/3. Figure 47. Mulinia lateralis (Say, 1822), Hadley Harbor, Elizabeth Is- lands, Massachusetts, Sta. P-0902-65, Systematics-Ecology Program, Marine Biological Laboratory — exterior view, left valve. X 2 1/3. Figure 48. Mytilus edulis (Linné, 1758), Hadley Harbor, Elizabeth Is- lands, Massachusetts, Sta, P-148-64, Systematics-Ecology Program, Marine Biological Laboratory — exterior view, left valve. X 2 2/3. Figure 49. Neotrigonia margaritacea (Lamarck, 1803), New South Wales, Australia, from coll. of A. L. McAlester, Yale University — exterior view, right valve. X 2/3. Figure 50. Nucula proxima (Say, 1822), Hadley Harbor, Elizabeth Is- lands, Massachusetts, Sta. P-92-64, Systematics-Ecology Program, Marine Biological Laboratory — exterior view, left valve. X 2 1/3. Figure 51. Nucula proxima variety truncula (Dall, 1898), Buzzards Bay, Massachusetts, Sta. “R,” coll. H. L. Sanders, Woods Hole Oceanographic Institution — exterior view, left valve. X 4 2/3. Figure 52. Periploma leanum (Conrad, 1831), Martha’s Vineyard, Chap- paquidick Island, Massachusetts, MCZ No. 192934 —§interior view, left valve. X 1/6. Figure 53. Petricola pholadiformis (Lamarck, 1818), Black Beach, West Falmouth, SEP-949, George M. Gray Museum, Systematics-Ecology Pro- gram, Marine Biological Laboratory — exterior view, left valve. X 2 1/3. Figure 54. Pitar cordata (Schwengle, 1951), off Port Aransas, Texas, 80 m, MCZ No. 194372 — interior view, right valve. X 2/3. Figure 55. Pitar morrhuana (Linsley, 1845), Hadley Harbor, Elizabeth Islands, Massachusetts, Sta. P-189-64, Systematics-Ecology Program, Ma- rine Biological Laboratory — exterior view, left valve. X 2 1/3. Figure 56. Saxidomus nuttalli (Conrad, 1837), Gulf of Georgia, British Columbia, Canada, MCZ No. 5235 — interior view, right valve. X 1/3. Figure 57. Solemya velum (Say, 1822), Hadley Harbor, Elizabeth Is- lands, Massachusetts, Sta. P-225-65, Systematics-Ecology Program, Marine Biological Laboratory — exterior view, left valve. X 2 1/3. Figure 58. Tagelus divisus (Spengler, 1794), Hadley Harbor, Elizabeth Islands, Massachusetts, Sta. P-21-63, Systematics-Ecology Program, Marine Biological Laboratory — exterior view, left valve. X 2 1/3. Figure 59. Yoldia limatula (Say, 1831), Hadley Harbor, Elizabeth Is- lands, Massachusetts, Sta. P-267-65, Systematics-Ecology Program, Marine Biological Laboratory — exterior view, left valve. X 2 1/3. 1967 TABLE | DISTANCE BETWEEN GENERA WHEN CONSIDERED AS POINTS IN FACTOR SCORE SPACE JONAXO SNTYWSSVN VILWNN1 VINd1d3u9 XNId1wSOuN WNIdOLLT1 wantivHy | Sl lolol Alle MOLLUSCAN SHELL PROTEINS (SNOISN3WIO 8) vaOdO4LSY9 ARCHITECTONICA ACHATINELLA ASTRAEA LITTORINA UROSALPINX CREPIDULA SIPHONARIA CAVOLINA FISSURELLA NASSARIUS SUCCINEA POLINICES APLYSIA UMB RACULUM JANTHINA DOLABELLA EPITONIUM NERITA MELANELLA HELISOMA MUREX HALIOTIS VIVIPARUS VINOSIYLOIN SN139VL SNWOGIXVS o yadLsOssvad Cal VI1LI8v BERREEEE BB“BBEE LAEVICARDIUM AEQUIPECTEN PETRICOLA MERCENARIA LIMOPSIS MALLETIA CRASSOSTREA SAXIDOMUS CORBICULA (SNOISN3WIG 6) WO0dAD3913d PERIPLOMA NEOTRIGONIA 2 {3 fells] __vwwanem z(ele : E 2 iz (2| N3193d INDI < BE WOIGHWIIA3WT Pe e[ sua a VISNOAT SISdOWI1 v1091ulad VINITOW . VavaVNy TURRITELLA AKERA 31 No. 262 BREVIORA *sueaw OLJawyylue aULWeSOXeYy/ULa}Odd Ob° Scr O9L 6l2 ble el oe @ Ge 08 0s v2 Le 68 82 8 92 ve S S LL if b°0 8 2g Le 9L BL 6L 62 2 SS ¢ == €°0 os L°0 ~°0 b°0 9 o¢ Ob 62 Zt 8b Of 0S 92 G2 82 L 02 v €€ Zl 89 6b lb 86 8Y SS 92 S2 69 8z 1€ Lv 9 él Sl Sl v2 l 9 Sl 9 02 S v Ol S9 vad bl Le 02 Se 9L $9 € -- 2 LL el b°0 6 20 SLL 89 0s Lye O2L vr LE? BSL Ell €£L 9LE 682 eSp 002 Bee LEL c0l €6 6S vl G2 Lt 8 LSL v6 ell 6€ BE Ge 6S 60 vv L6 69 6S 86 el OLL OLL (As) LS €9 €2 SL 9L 22 tl Ov L8 20l 9 2ll 69 b9l €8 bl juawnbaquy | yuawnbazuy wndeuzsoluad (uad) ParplLopej |] parseopeouny YFLM LLeys LLeusS} LLeUs LLEUS LLeys}] = LLe4s *sqeuy plunz4od PULL LGOON | SNL LIAW] ELNINN | SLZOLLEH I snLLznen OBL [07 "PpalsLo_eo g SpatyLo_eoun satoads ¢ - sueaWw ILuzawoan y (auoq W10L OOOL Y3d SOILVY GIDV ONIWV ZT AAV; 2 2 22 G9 6b v BE le ne ral l2 L2 09 G2 1g eb 02 G2 Ol ol Ob 6€ 82 le 90L 99 o2k e0L LOL LL el aa 86 pol eb Lg 9LL O€l 2133N2) LLey4s LLey4s eidas | epnards 86°0 LE L€ as) 0€ él St bl 88 LEL £6 86 89 09 82l LLE4S eunald -07yaeYD 96°St 0S aque eunald -dJaeUD SauLWweSOXa} uLayOdd auluLbay SULPLASLH auLsAy auLsh]-HO auLuele|Auaud DULSOUAL auLonay duLoNaLoOs] BULUOLYZOW SULLEA [$Ley] aurysé9 auLue Ly BuLoAl[9 BULLOUd Ploy otweyn[o auLuas duLuoadyy ploy otquedsy aul LO4d-HO 1967 MOLLUSCAN SHELL PROTEINS 3)3) TABLE 3 FACTOR SCORES. GASTROPODA. (Shell Tissues) ] 2 3 4 5 6 7 8 Phylogenetic Positive Loading OHPRO Relationships | Varimax Factors Negative Loading 2] SIPHONARIA HELISOMA PLANORBIS ACHATINELLA SUCCINEA CAVOLINA UMBRACULUM ARCHITECTONICA AKERA OXYNOE APLYSIA BULLA DOLABELLA — mNNnN W OOO mM OW Cony ony Ups" Cony ins) (=) own rw OT ON W oOo fF Oo Mw W PM Somos 2 Sw & —- ON NN — OC — 58) 2 .6 4 2 ] al 6 4 4 -6 1 0 0 4 3 EPITONIUM JANTHINA TURRITELLA MUREX UROSALPINX NASSARIUS COLUS MELANELLA POLINICES LUNATIA LITTORINA CYPRAEA CREPIDULA PLANA CREPIDULA FORNICATA FISSURELLA ACMAEA VIVIPARUS NERITA ASTRAEA HALIOTIS SS ——— —) Dwnn o1w nn FN —] FP S|] Wm MD OWWnN DWH HM CO ("NH WHO MY W WO OO ON WKN OWN DY FON OO MH W © OM —]| WHaanant WN? NY —| NY W OO M~aMnDO DAO 1H oO fF fF ON 34 BREVIORA No. 262 TABLE 4 FACTOR SCORES. PELECYPODA. (Shell Tissues) ] 2 3 4 5 6 7 8 9 Phylogenetic Positive ISOLEU A : Loadin LEU Relationships 9 VAL Varimax HIS Factors LAEVICARDIUM MACOMA MULINIA TAGELUS CORBICULA ARCTICA AEQUIPECTEN CRASSOSTREA MYTILUS ANADARA LIMOPSIS NEOTRIGONIA PERIPLOMA LYONSIA NUCULA 1967 OH-Proline Aspartic Acid Threonine Serine Glutamic Acid Proline Glycine Alanine Cystine [half] Valine Methionine Isoleucine Leucine Tyrosine Phenylalanine OH-Lysine Lysine Histidine Arginine Protein Hexosamine Nerita Marine 66 MOLLUSCAN SHELL PROTEINS AMINO ACID RATIOS PER 1000 TOTAL Viviparus | Archi- Fresh Water 1060 tectonica Marine 22 TABLE 5 Siphonaria Marine 23 Helisoma|Planorbis Fresh Water 1] Fresh Water 114 Succinea Land 77 35 Achatinella Land 82 BREVIORA Museum of Comparative Zoology CAMBRIDGE, Mass. 31 MARCH, 1967 NUMBER 263 THE HYDROID OF VANNUCCIA FORBESII (ANTHOMEDUSAE, TUBULARIIDAE) By Anita Brinckmann-Voss Stazione Zoologica, Napoli, Italy, and Institute of Animal Science, University of Manitoba, Winnipeg, Manitoba, Canada ABSTRACT: This paper deals with a hydroid of the order Anthomedusae (Athecatae), family Tubulariidae, formerly described as Hybocodon forbesii Mayer 1894. The structure of the hydroid proves that the species does not belong to the genus Hybocodon; a new genus, Vannuccia, is therefore intro- duced. The hydroid, the budding medusae, the asexual reproduction and the young medusae are described in this paper. INTRODUCTION During the preparation of a monograph of the Anthomedusae and athecate hydroids from the Gulf of Naples a tubularian hy- droid was found on mud in the Bay of Naples. This hydroid liberated medusae, in the laboratory, which appeared to belong to the species Hybocodon forbesii Mayer 1894. But since the hydroid does not have the structure of other Hybocodon hydroids and does not resemble any other hydroid of the Tubulariidae, a new genus had to be established for the former Hybocodon forbesii. The new genus is named Vannuccia, in rec- ognition of Dr. Martha Vannucci, the head of the plankton section of the Oceanographic Institute of Sao Paulo, who has done ex- tensive research on the plankton around the Brazilian coast. The find reported here is the first and only case in which Van- nuccia forbesii (Mayer) has been recorded from the Mediter- ranean. Only one specimen of the hydroid stage could be found and no free medusae were collected from the sea. However, the species proved easy to keep in the laboratory, and numerous medusae as well as hydroids were liberated from the one specimen. The research at Naples was supported by a grant from the U.S. 2 BREVIORA No. 263 Office of Naval Research, No. 2100(00). Completion of the re- search was made possible by a grant from the National Science Foundation, No. GB2358 under the sponsorship of the Museum of Comparative Zoology, Harvard University, Cambridge. VANNUCCIA 0. gen. Type species. Vannuccia forbesii (Mayer) 1894: 236, pl. I, fig. 1; the gender of the new generic name is feminine. Generic characters. Tubulariidae with solitary hydroids with one whorl of moniliform oral tentacles and an aboral whorl of filiform tentacles. With endodermal stem canals and basal tuft of rooting filaments. Medusa buds borne on blastostyles just above the aboral tentacle whorl. Free medusae with one tentacle and no exumbrellar cnidocyst rows. VANNUCCIA FORBESII (Mayer) Hybocodon forbesii Mayer, 1894: 236, pl. I, fig. 1. Specific characters. Hydroid with 12 to 14 oral tentacles and 16 to 20 aboral tentacles. Asexual reproduction by transverse fission of the stem. Free medusae with slight asymmetrical bell, one fully developed tentacle with a large terminal cnidocyst swelling. Distribution. Tortugas, Bahama Islands, and Florida (Mayer, 1894; 1910); Trivandrum and Madras, India (Nair, 1951); south- ern Japan (Uchida, 1927); Bay of Naples, Mediterranean. Description of the species. The hydroids of V. forbesii are very similar to those of Corymorpha nutans. The hydranth of V. forbesi consists of a stem or hydrocaulus and the hydranth body itself. The base of the stem or the “foot” (Fig. 6) shows a similar ar- rangement of rooting filaments and endodermal canals to that seen in Corymorpha nutans. The whole stem is surrounded by a flexible perisarc which extends slightly below the hydranth body. There the perisare is usually marked by a slight constriction of the stem (Fig. 1). The length of hydranth and stem measures 2-3 cm for speci- mens with medusae buds, but there is great variability in the length of the stem. The oral tentacle whorl consists of 12 to 14 tentacles each of them carrying 4 to 6 cnidocyst swellings. The aboral tentacle whor] consists of 16 to 20 tentacles. Each aboral tentacle is very long and has a small swelling at its tip which is more or less pronounced depending on whether the tentacles are contracted or expanded. 1967 HYDROID OF VANNUCCIA FORBESII 3 PHAN AR GE 1 WAH AN wan DURES: Wieletecsct/, bs ere ava 1mm ee! Figure 1. Vannuccia forbesii, adult hydroid with medusae buds. Drawn by Ilona Richter. The medusa buds develop in clusters on short blastostyles. As the buds are naked, the one tentacle of each bud can be seen in the more advanced stages of the medusa formation. The liberated medusae are 1.8-2.0 mm high. The exumbrella is apple shaped and somewhat asymmetrical. Young medusae show all the features of the specific characters of the adult medusae with the exception of the gonads, which develop later. This is typical of the Cory- morpha-Tubularia group. 4 BREVIORA No. 263 ©:2 5mm Fig. 2. Vannuccia forbesii, male medusa two days old. Drawn by Ilona Richter. Medusae of V. forbesii carry one fully developed tentacle and three tentacle bulbs. The tentacle bulb which is opposite the fully developed tentacle is larger than the other two (Fig. 2). This bulb starts to grow out on the second day after liberation to form a small rudimentary tentacle. Mayer, who gives a descrip- tion of a fairly young and an adult medusa of V. forbesii (Mayer, 1894; 1910), shows that this tentacle may grow longer in fully mature specimens (Mayer, 1910, pl. IIT). The one fully developed tentacle has a large swelling at its tip, which is provided with numerous cnidocysts. The manubrium is club shaped without a peduncle and extends to the velum. Usually the manubrium does not stay in the center of the subumbrellar cavity, as is generally expected of anthomedusan species, but is located slightly nearer the tentacle bulb opposite the fully developed tentacle. The gonads start to develop soon after the liberation of the medusa. The stomach is completely encircled by the gonads. As the medusae are difficult to feed, I did not manage to get specimens with fully mature gonads. The velum is thin and narrow. Sunod jo sasvys eyuswdojaaap da.1y) ‘“Msaqsof PloonuUuDA *¢ dINBIA HYDROID OF VANNUCCIA FORBESII 6 BREVIORA No. 263 Reproduction and development. All of my medusae of V. for- besii were males. Therefore, the sexual reproduction could not be observed. However, there is a very interesting type of asexual reproduction of the hydroid: the base of the foot of the hydroid thickens slightly and is then constricted off from the remainder of the stem (Fig. 3), and is freed from the perisarcal sheath. It develops four oral and six aboral tentacle buds after it is freed (Fig. 4). Sometimes, however, the tentacle buds are already de- veloped inside the perisarc of the old hydranth. A similar type of | Figure 4. Comparison between the asexual budding of Euphysa aurata and Vannuccia forbesii. 1, Euphysa aurata; Wl, Vannuccia forbesii. a, Hydranth stem; b, young buds which are constricted off from the foot of the hydranth; c, developing young hydranths; d, perisarc. For the clarity of the figure the rooting filaments are omitted in II. 1967 HYDROID OF VANNUCCIA FORBESII 7 budding is known from Euphysa aurata (Forbes) and Hypolytus peregrinus Murbach. But in Euphysa aurata the oral tentacles of the bud always develop distal to the aboral ones, in relation to the “mother” stem (Fig. 3). The liberated hydranths do not settle immediately but move around slowly until they affix themselves after some hours or even one or two days. This asexual reproduction may become so prolific that a whole tuft of hydroids is formed out of one speci- men in about a week. Ecological observations. When found in May 1962. the young hydroid of V. forbesii had only six oral and ten aboral tentacles. During its development the number of tentacles increased. The young one was found in mud at 30 m depth. In the same region I found Euphysa aurata, but no other hydroids or medusae of V. forbesti. The hydroid was kept at 18°C. The asexual reproduction started at once, after the hydroid had been brought to the lab- oratory. Medusae production took place in autumn and winter. The asexual reproduction did not cease while the development of medusa buds was taking place. SYSTEMATIC DISCUSSION The medusae which were liberated from the above described hydroid match Mayer’s description of Hybocodon forbesii (Mayer, 1894), with one exception: There is a velum in our specimens. But then, while Mayer says in his original description “without velum” in 1894, his redescription, based on more abundant ma- terial, in his “Medusae of the World” (Mayer, 1910:42) men- tions: “The velum is narrow.” Mayer placed his medusa species in the genus Hybocodon on account of the slightly asymmetrical bell and the one tentacle. Kramp is the first one to question the placement of Hybocodon forbesii in the genus Hybocodon, suggesting however that it should provisionally remain in that genus (Kramp, 1959, 1961). The medusa stage is somewhat reminiscent of Euphysa and Corymor- pha as well. Regarding the hydroid stage, difficulties arise about placing the species in the proper systematic position. The morphology of the hydroid demands that the species (up to now known from its medusa stage only) not remain in the genus Hybocodon any more, because it is certainly not a Hybocodon hydroid. The estab- lishment of the new genus Vannuccia for the former Hybocodon forbesii is suggested therefore. 8 BREVIORA No. 263 The combination of moniliform oral and filiform aboral tenta- cles in V. forbesii hydroids is unique among the Tubulariidae. Yet, upon checking the different solitary tubulariid hydroids, either from literature or preserved material, I find that the hydroid of V. for- besii bears some relationship to Hypolytus obvoluta (Kramp) 1933; this was first described as Corymorpha obvoluta, and later placed in the genus Hypolytus by Rees, being cogeneric with H. peregrinus Murbach 1899, on account of the moniliform tentacles in both sets (Rees, 1957). However, if one compares the figures of H. peregrinus Murbach and H. obvoluta (Kramp ), it seems that the aboral tentacles of H. obvoluta are much less moniliform than the aboral tentacles of H. peregrinus. This points to a certain | I Figure 5. Oral tentacles of young Vannuccia forbesii (1), and young Corymorpha nutans (11). order insofar as there are: (1) moniliform tentacle whorls in both sets in H. peregrinus Murbach, (2) “reduced” aboral monili- form tentacles in H. obvoluta (Kramp), (3) filiform aboral and moniliform oral tentacles in Vannuccia forbesii, and (4) filiform tentacles in both whorls, slightly moniliform in the oral set when young, in Corymorpha nutans. In this comparison, Vannuccia is placed between Hypolytus and Corymorpha. There are also other characters indicating the relationship of Vannuccia forbesii with H. obvoluta, on the one hand, and with Corymorpha nutans, on the other: Hypolytus obvoluta shows faint endodermal channels; they are more apparent in V. forbesii and strongly developed in C. nutans. Similarly, the anchoring filaments are feebly developed 1967 HYDROID OF VANNUCCIA FORBESII 9 in Hypolytus obvoluta, become more apparent in Vannuccia for- besii, and are strongly developed in Corymorpha nutans. In summary, the genus Vannuccia is related to the genera Hypolytus and Corymorpha, representing an intermediate stage be- tween them. Medusa and hydroid of Vannuccia forbesii are de- cidedly different from those of either genus, so that the new genus, Vannuccia may be justified. [2854 Figure 6. Vannuccia forbesii, base of hydranth stem. The rooting fila- ments form a dense tuft, but in order to make the figure clearer only a few of them are drawn in this figure. A, Gelatinous perisarc; B, ectoderm; C, endodermal ridge; D, developing rooting filament; E, rooting filament. The genus Vannuccia is closer to Corymorpha than to Tubularia. If the family Tubulariidae is divided into two families, the Cory- morphidae and Tubulariidae (Rees, 1957; Brinckmann-Voss, MS in preparation), the genus Vannuccia would have to be placed in the Corymorphidae. 10 BREVIORA No. 263 ACKNOWLEDGMENTS I wish to thank Mrs. Sofia Giaquinto-Boag for her most efficient help in maintaining the polyps, and the director of the Naples Zoological Station, Dr. P. Dohrn, for his generous support. REFERENCES CITED ALLMAN, G. J. 1872. A monograph of the gymnoblastic or tubularian hydroids. Ray Society, London: 1-450, 84 figs., 32 pls. KRAMBP, P. L. 1933. Coelenterata, Ctenophora and Chaetognatha. The Scoresby Sound Comm. 2nd East Greenland Exped. in 1932. Meddel. om Gronland., 104(11): 1-20. 1949. Origin of the hydroid family Corymorphidae. Vidensk. Medd. fra Dansk Natur. Foren., 111: 183-215. 1959. The Hydromedusae of the Atlantic Ocean and adjacent waters. Copenhagen. Dana Report No. 46: 1-283. 1961. Synopsis of the medusae of the world. J. Mar. Biol. Assoc. United Kingdom, 40: 1-469. Mayer, A. G. 1894. An account of some medusae obtained in the Bahamas. Bull. Mus. Comp. Zool., Harvard College, 25(11): 235-242, 3 pls. 1910. Medusae of the world. Part I, Hydromedusae. Carnegie Inst. Washington, Publ. 109(1): 1-230, 29 pls. MursacH, L. 1899. Hydroids from Wood’s Holl, Mass. Hypolytus peregrinus, a new unattached marine hydroid. Quart. J. Microscop. Sci., (N.S.) 42: 341-360. Nair, K. K. 1951. Medusae of the Trivandrum Coast, Part I, Systematics. Bull. Centr. Res. Inst. Univ. Travancore (Trivandrum), 2 C (Nat. Sci.) No. 1: 47-75. REES, W. J. 1957. Evolutionary trends in the classification of capitate hydroids and medusae. Bull. Brit. Mus. (Nat. Hist.), Zool. 4(9): 453-534. Ucnipa, TH. 1927. Studies on Japanese Hydromedusae. I, Anthomedusae. J. Imp. Univ. Tokyo, Sect. 4, Zool., 1: 145-241, 47 figs., 2 pls. (Received 26 September, 1966.) BREVIORA Museum of Comparative Zoology CAMBRIDGE, Mass. 6 ApriL, 1967 NUMBER 264 THE CHANARES (ARGENTINA) TRIASSIC REPTILE FAUNA Ill. TWO NEW GOMPHODONTS, MASSETOGNATHUS PASCUALI AND M. TERUGGII By Alfred Sherwood Romer In previous numbers of this series an account was given of the 1964-1965 expedition of the Museo de la Plata and the Museum of Comparative Zoology to the Permo-Triassic of western Ar- gentina, and the Permo-Triassic geology of the Chafares-Gualo region was briefly described (Romer, 1966; Romer and Jensen, 1966). This and succeeding papers will be devoted to descrip- tions of the new genera and species contained in the Chanares fauna. Owing to a sequence of political incidents which need not be detailed here, the greater part of the collection has only recently (December 1966) reached the Museum of Comparative Zoology after the lapse of nearly two years. The one box of specimens which reached Cambridge earlier and is now partially prepared, contains, however, a considerable variety of reptilian remains, par- ticularly skulls and jaws. Represented, in addition to indeterminate fragmentary materials, are a small pseudosuchian, a moderately large dicynodont with powerful tusks, two carnivorous cynodonts, and at least four (possibly more) gomphodont cynodonts. For most of these forms, the materials at hand are none too complete, even as regards crania and dentitions, and it hence seems wise to defer description until the arrival of the main body of the collec- tion, in which further specimens are almost surely present. Two of the gomphodonts, however, are represented by a number of well- preserved skulls and jaws, and will be described here as Masse- tognathus pascuali and M. teruggii. As will be seen from the descriptions and illustrations, the two are comparable in many ways, and are obviously closely allied. The skulls assigned to M. teruggii are on the average about 43 per cent larger than 2) BREVIORA No. 264 those of M. pascuali, and the differences between the two are in great measure differences which might reasonably be associated with the contrast in size. I was at first inclined to believe that we were possibly dealing with growth stages of a single form. How- ever, the skulls sort clearly into two size groups, rather than form- ing the graded series expected had they represented growth stages of a single species. The difference in size between the two is con- siderably greater than that expected in sex differences, and it is hence reasonable to believe that we are dealing with two species of a single genus. MASSETOGNATHUS gen. nov. Type species: M. pascuali sp. nov. Diagnosis. Traversodontid gomphodonts of modest size, known skull lengths (snout to condyles) ranging from 82 to 138 mm. The skull is short and broad, the width about seven-tenths the length to condyles. Muzzle expanded laterally so that the maxilla is broadly exposed ventrally lateral to the cheek tooth row. As in gomphodonts generally, the zygomatic arch is broad but with a constricted posterior base; the arches widely expanded for their full length in a line parallel to the long axis of the skull; the epipterygoid-pterygoid bar extending broadly back to the quadrate region; the angular process of the dentary well developed and thickened at its ventral border; posterior to this, the angular-sur- angular well exposed laterally. No suborbital process of the jugal; external exposure of the squamosal on the zygomatic arch re- stricted to upper and posterior borders; no posterior projection of the angular process of the dentary. Four upper, three lower in- cisors; canines modestly developed; 12-15 upper, 11-13 lower cheek teeth; anterior teeth smaller and simpler in structure than main series, but no distinct “premolar” series. Teeth of typical traversodont pattern; upper “molars” with two lateral cusps, the more posterior one the more prominent; a high posterior transverse ridge with cusps close to and at the posteromedial corner. Teeth subquadrate, but extending farther laterally toward the posterior corner; anterior and posterior margins essentially transverse and parallel, but slightly convex in outline anteriorly toward the ex- ternal margin and similarly concave posteriorly, so that each tooth “shoulders” to a slight degree into the area of its next anterior neighbor. The generic name (masculine) refers to the obviously excellent chewing powers of the jaws. 1967 TWO NEW GOMPHODONTS 3 MASSETOGNATHUS PASCUALI sp. nov. Holotype. No. 65-XI-14-1, Museo de la Plata, skull and jaws, collected by the 1964-1965 expedition of the Museo de la Plata and the Museum of Comparative Zoology. The specific name is in honor of Dr. Rosendo Pascual, Pro- fessor of Paleontology in the Universidad de la Plata, who accom- panied the expedition during much of its stay in western Argentina, and was exceedingly helpful to us before, during, and after our field work. Horizon and locality. From an exposure in the Triassic Cha- fares Formation about two miles north of the point where the Chanares River debouches into the Campo de Talampaya, in west- ern La Rioja Province, Argentina. Diagnosis. A small form, with a mean length, snout to con- dyles, of about 87 mm in available specimens. Posterior portion of frontals and anterior portion of parietals form a flat triangular dor- sal area, the posterior end of the triangle extending back to a parietal foramen of modest size; the postorbitals extend back on either side to a point about opposite the parietal foramen. The angular process of the dentary descends to form approximately a right angle with the ventral margin of the bone. Twelve upper and 11 lower cheek teeth in available specimens. Description. Massetognathus pascuali appears to be an exceed- ingly common member of the Chanares fauna; even in the re- stricted material currently available from our collection there are some seven fairly good skulls, as well as less complete skull and jaw specimens and considerable postcranial material. This is a relatively small animal, the mean length of available specimens, from premaxilla to occipital condyle (allowing for minor anterior or posterior deficiencies), being 87.3 mm, with a range from 82 to 95 mm. As noted above, the first thought that these specimens might be growth stages leading to M. teruggii appears to be negated by the fact that the series here assigned to M. pasuali seems to clus- ter closely about the size mean, with an average deviation of but 3.6 mm, or about 4 per cent. The general nature of the skull is obvious from Figures 1-3. Its structure is for the most part that familiar in advanced cynodonts generally and gomphodonts in particular, and in many regards does not call for detailed description. In its general proportions it is relatively broad and short. The seeming shortness is exaggerated by the most distinctive feature of the genus — the “swollen” muzzle. As in all typical gomphodonts, the two tooth rows of the cheek 4 BREVIORA No. 264 battery are close together on the ventral surface of the skull. In most gomphodonts there is little lateral extension of the muzzle beyond the line of the tooth row, and (as may be seen, for ex- ample, in Diademodon or Exaeretodon) the snout region is slen- der. Here, however, the maxillae extend far out dorsally (with a downward slope) to a point about opposite the lower margin of the orbits, then, curving downward and inward, present a broad ven- tral surface lateral to the tooth rows. Figure 1. Lateral view of the skull of Massetognathus pascuali. Natural size. The skull is low, and the orbits face somewhat more dorsally than laterally and slightly anteriorly. The prefrontal and post- orbital meet broadly above the orbits. The nasals and frontals form an essentially flat dorsal platform atop the skull. On either side, the prefrontals and postorbitals are somewhat raised above the flat surface of the frontals and anterior part of the parietals. Posterior to the orbits, the ridges formed by the postorbitals con- verge to leave a triangular wedge of flat dorsal surface formed by the back portion of the frontals and the anterior portion of the parietals. This narrows posteriorly, with the rather small parietal foramen at the apex of the triangle. Posterior to this, the con- joined parietals form a sharply developed parietal crest. The post- orbitals on either side extend backward, sheathing the parietals laterally, to about the level of the parietal foramen. The lateral surfaces of the braincase appear to be constricted in typical cynodont fashion; I shall postpone description of this region and the interior of the orbit until further material is available for “dissection” and sectioning. The occipital surface of the skull is of normal cynodont construction. Laterally, as in gomphodonts gen- erally, but in contrast to other cynodonts, the squamosal descends 1967 TWO NEW GOMPHODONTS 5 Figure 2. Dorsal and ventral views of the skulls of Massetognathus pas- cuali. Natural size. 6 BREVIORA No. 264 ventrally so that the connection between the portion of the bone forming the lateral wing of the occipital region and that associated with the zygoma is only by a narrow neck ventrally; in posterior view, the dorsal margins of the two portions form a V, widely open above. The pterygoid-epipterygoid bar reaches broadly back to the quadrate region above a pterygo-paroccipital fenestra. The squamosal forms the boundaries, medial and lateral, of a distinct Figure 3. Posterior view of the skull of Massetognathus pascuali. Natu- ral size. external auditory meatus. The postorbital bar is slender; the zygo- matic arch, however, is deep dorsoventrally. In contrast to cynog- nathids, in which the arch slants diagonally outward and backward from the orbital region to attain its greatest length posteriorly, the arch, as in gomphodonts generally, attains its full breadth at the posterior margin of the orbit, and runs back from this point parallel to the main axis of the skull, allowing a high development of the temporal musculature. The external exposure of the squamosal on the arch is restricted to a narrow area at the top of the arch, and, curving downward posteriorly, along the anterior margin of the external meatus. Ventrally, the two rows of cheek teeth are close together an- teriorly, and diverge gently posteriorly. The secondary palate ex- tends back most of the length of the tooth row and (although the sutures are none too certain) the palatines appear to form the most posterior portion of the secondary palate. Back of the secondary palate, little of sutures can be made out in the material available. On either side, the lateral margins of the pterygoids extend back- ward as raised ridges which terminate in sharply-pointed flanges di- rected downward and backward. From the median bar formed posteriorly by the conjoined pterygoids and basicranial axis, a pair of ridges run diagonally forward and laterally toward the posterior ends of the tooth rows, and a median ridge runs directly forward. 1967 TWO NEW GOMPHODONTS 7 The posterior ventral structures are of the type commonly seen in cynodonts, with an excavated basioccipital-sphenoid area bounded by a pair of ridges; on either side, there is a well-developed vagal foramen, and farther forward, the fenestra ovalis. A V-shaped channel, bounded medially by the braincase wall and laterally by the pterygoid-epipterygoid bar, represents the primitive cranio- quadrate passage; anteriorly it is roofed below by a median ex- tension of the pterygoid. In most specimens the quadrate (plus quadratojugal, loosely attached to the skull) has dropped out, leaving a pair of notches to represent its area of articulation. In the jaw the dentary is highly developed, with a high and broad coronoid process, a posterior process which reaches back nearly to the articular area, and a pronounced angular process of the dentary which is thickened and rounded ventrally. Posteriorly, in contrast to non-gomphodonts, the angular and surangular are well exposed on the outer surface. The elements of the jaw apart from the dentary are incompletely preserved in the available material, and I shall reserve description of the inner surface of the mandible for a future occasion. There are four upper and three lower incisors of modest size, somewhat chisel-shaped and with longitudinal striations. The canines are relatively little developed. The cheek teeth are sep- arated from the canines by a short diastema; both above and be- low, the members of each cheek series (Figs. 4, 6) are crowded close to one another, and the two rows, rather close together in front, curve gently outward posteriorly. As is now well known in other gomphodonts (Crompton, 1955, etc.), the composition of a gomphodont cheek battery may change during “adult” life by the eruption of additional teeth posteriorly and resorption of anterior ones, so that no precise tooth count is possible. In our material there is no positive evidence of anterior resorption, but in most specimens there is evidence of posterior addition in the presence, at the back end of the battery, of either a newly erupted but unworn tooth, one well formed by not yet erupted, or a partially developed tooth still in its alveolus. In the specimens currently available to us there are generally 12 maxillary teeth, and 11 in the only two mandibles now in our possession. Although, in contrast to dia- demodonts, there is no sharp contrast between small “premolars” and a “molar” series, the more anterior cheek teeth are smaller and with a simpler pattern. The wear on the teeth is obviously great, so that little detail of the crown pattern can be seen except in teeth at the back end of the series which are as yet unerupted or very freshly erupted. 8 BREVIORA No. 264 Figure 4. Left (above) and right (below) upper cheek tooth series of a specimen of Massetognathus pascuali. X 3. Such a tooth from the maxillary series is shown in Figure 5. In surface view the crown is subquadrate, nearly twice as broad as long anteroposteriorly. The inner margin is curved; anterior and posterior margins are nearly straight and parallel to one another, except that towards the outer border the anterior margin is some- what convex, the posterior border opposite somewhat concave. Each tooth thus “shoulders” somewhat into the proper area of its next anterior neighbor — a feature which is much more pronounced in the more advanced traversodonts of the later Ischigualasto horizon. On the outer border, the base of the cone forming the postero- external cusp extends considerably outward and backward, so that the external boundary is a diagonal line. The outer portion of the tooth is a high anteroposterior ridge, on which a major cusp de- velops posteriorly; a minor cusp, only slightly separated from it, 1967 TWO NEW GOMPHODONTS 9 is developed along the downward anterior slope of the ridge. Me- dial to the two cusps, the external ridge descends abruptly into a basin which occupies most of the crown area. This basin, apart from its straight lateral border, is an essentially oval concave area, without internal markings. It is surrounded on posterior, medial, and anterior borders by a curved ridge, considerably higher along the posterior margin than medially and anteriorly. An elongate cusp is present about two-thirds the way in along the pos- terior ridge, a second cusp near the posteromedial corner, and a lower cusp part way along the anterior ridge. In a freshly erupted tooth a series of tiny bead-like denticulations may be seen along Figure 5. A right upper “molar” of Massetognathus pascuali. A, Crown view. 8B, Posterior and somewhat ventral. C, Medial. D, Anterior and somewhat dorsal. E, Lateral. X 4. most of the extent of the ridges surrounding the basin posteriorly, mediaily, and anteriorly, and a few may even be present along the external ridge. With wear, the denticulations rapidly disap- pear; with somewhat further wear, the anterior ridge is ground down, leaving the posterior ridge projecting, but with distinction between its two cusps lost, and the internal ridge still prominent. With still further wear, the posterior ridge, too, becomes worn away, and the basins of successive teeth become a continuous sur- face, flat anteroposteriorly but concave as seen in end view, rising medially to a persistent medial ridge, and rising still higher laterally to the outer longitudinal ridge. The general position of the lower cheek series corresponds to that above. (Because of imperfections in available specimens of M. pascuali, I have shown instead, in Figure 6, that of M. teruggii.) As in the upper jaw, the most anterior cheek teeth are smaller and simpler in pattern than the main series, although there is no sharp division into “premolars” and “molars.” Just as in the case of 10 BREVIORA No. 264 Figure 6. Right cheek tooth series of a specimen of Massetognathus ter- ugegil. X 2. the upper “molars,” wear tends to obliterate rapidly most of the details of the crown pattern. A freshly erupted tooth is shown in Figure 7. The lower “molars” are of the diagnostic traversodont pattern. This pattern, as regards position in the jaw, gives them a build contrasting sharply with those of the upper jaw. Curi- ously, however, if one imagines a lower tooth rotated 90°, so that the anterior border faces, instead, to the outside, the pattern be- comes highly comparable to that of the upper tooth. The lower cheek teeth are subquadrate in shape. The anterior portion forms a high transverse ridge, partially subdivided into two cusps, of which the more external is the larger and higher. Just as the ex- ternal ridge of the upper tooth descends steeply on its inner sur- face to the basin of the tooth, so here the posterior face of the anterior transverse ridge descends abruptly into the subquadrate, smoothly concave tooth basin. As in the case of the upper tooth, the basin is surrounded by a continuous ridge on the other three Figure 7. A right lower “molar” of Massetognathus teruggii. A, Crown view. B, Lateral. C, Anterior. D, Medial. E, Posterior. X 4. 1967 TWO NEW GOMPHODONTS 11 sides; the external ridge is much the highest, the internal the low- est. As in the margins of the upper basin, so here a nearly continu- ous series of bead-like cuspules is present in the unworn tooth, but in contrast to the upper teeth, no conspicuous cusps are present, even in the unworn tooth. With wear, the denticulations soon dis- appear, and except for the strong anterior cross-ridge, the tooth is reduced to a somewhat concave basin with a somewhat raised rim externally. MASSETOGNATHUS TERUGGII Sp. nov. Holotype. No. 65-XI-14-2, Museo de la Plata, skull and jaws, collected by the 1964-1965 expedition of the Museo de la Plata and the Museum of Comparative Zoology. The species is named for Professor Mario Teruggi, who was most helpful in the carrying out of our field work and in our post- expeditionary difficulties. Horizon and locality. From the same horizon and locality as the type of M. pascuali. Diagnosis. A larger species than M. pascuali, with a mean skull length of about 125 mm in available specimens. The flat- tened dorsal area formed by the frontals and anterior portions of the parietals is more restricted than in M. pascuali, the roof con- stricting to a sagittal crest in front of the parietal foramen, which is reduced to a tiny slit. The postorbitals extend backward on either side of the crest to a point well back of the foramen. The angular process is not as sharply developed as in M. pascuali, the posterior and ventral borders of the process meeting at an obtuse angle. Thirteen to 15 maxillary, and 11 to 13 mandibular cheek teeth in available specimens. Description. M. teruggii seems, next to M. pascuali, to be the most common of Chanares reptiles; our present restricted materials contain some five skulls of this form. As noted above, this species is considerably larger (more than 40 per cent) than M. pascuali. This is, of course, a considerably greater difference than would be expected if we were dealing with sex differences. The skulls at- tributed to M. teruggii, like those of M. pascuali, seem to form a compact series, ranging from 115 to 138 mm in length, with an average deviation from the mean of about 9 mm, or about 7 per cent only. The skull of this species (Figs. 8-10) is very similar in most regards to that of its associate, M. pascuali. Since we have de- scribed, above, the skull of this smaller form in some detail, we 12 BREVIORA No. 264 Figure 8. Laterial view of the skull of Massetognathus teruggii. X %4. can here avoid unnecessary and monotonous repetition of descrip- tion of their numerous similar features, and call attention only to the few points of difference. These are in part associated with size. As is, I think, well known (and should be obvious), the chewing apparatus of a large form must be disproportionately large as com- pared with that of a relative of smaller size. Here the differences in this feature have to do with a somewhat greater development of a sagittal crest. This is accomplished by a greater forward develop- ment of the crest, so that the flattened area of the top of the skull is “pinched in” to a greater degree than in M. pascuali, and a median ridge is, in contrast, formed anterior to the area of the parietal foramen, which is reduced almost to the vanishing point. The posterior flanges of the postorbitals which clamp in upon ither side of the parietals here reach farther back than is the case in M. pascuali, to extend to a point well back of the parietal foramen. As may be seen from Figure 8, the angle of the dentary is less sharp than in M. pascuali, the posterior border of the dentary sloping downward and forward at an angle, rather than descending vertically to the apex. As noted, the tooth count of the cheek bat- tery is somewhat higher than in M. pascuali. I find no essential dif- ference in the pattern of the “molars” in unworn teeth of the two species. DISCUSSION “Gomphodont” cynodonts, with a masticatory series of cheek teeth, were early discovered in the Cynognathus Zone of the early Triassic of South Africa. Diademodon (of which Gomphognathus and several other generic names are synonyms) is the best known, and has been well described by Watson (1911, 1913), Broili and 1967 TWO NEW GOMPHODONTS 113) Figure 9. Dorsal and ventral views of the skull of Massetognathus ter- uggil. X %. 14 BREVIORA No. 264 Figure 10. Posterior view of the skull of Massetognathus teruggii. X %. Schréder (1935), and most recently by Brink (11955; etc*): Trirachodon, Protacmon, Trirachodontoides, Gomphodontoides, and Inusitatodon are less well known forms from the Cynognathus Zone; a recent discovery is Cragievarus (Brink, 1965). Until recent decades, no remains of gomphodonts were known in other areas or in later Triassic horizons, and it was assumed that the gomphodonts were a short-lived and unimportant branch of the cynodont stock. This, however, was due to our lack of adequate knowledge of Middle Triassic faunas. Explorations from the nineteen-thirties onward, of beds of this age in South Amer- ica and Africa, are now revealing the fact that the gomphodonts long remained a flourishing group which formed a major com- ponent of the Middle Triassic faunas — at least those of the south- ern continents. From the Santa Maria beds of southern Brazil, von Huene (1928, 1942) described Gomphodontosuchus and Traversodon. From collections made in the Manda beds of East Africa by Par- rington and for von Huene have been described Theropsodon (Huene, 1950), and Aleodon, Cricodon, and Scalenodon (Cromp- ton, 1955). In South Africa, the Middle Triassic Molteno beds, long considered barren, have begun to yield fossils to Crompton, including a gomphodont Scalenodontoides (Crompton and Ellen- berger, 1957). Brink (1963) has recently described a Diademodon specimen from Northern Rhodesia at a level which he considers equivalent to the Cynognathus Zone, and a new gomphodont. Luangwa, from a horizon in the Ntawere Formation which is prob- ably a Molteno equivalent. Large gomphodonts, not yet described, are said to persist into the lower part of the Upper Triassic Red- beds of Basutoland (Crompton, 1964). A recent British expedi- tion resulted in the discovery of further gomphodont material in Northern Rhodesia and Tanganyika. In Argentina a visit by Fren- guelli to the Middle Triassic deposits of Ischigualasto produced a 1967 TWO NEW GOMPHODONTS 15 few scraps of gomphodont skull to which Cabrera (1943) gave the name Exaeretodon and Theropsis. Exploration of this region by a Harvard-Buenos Aires Museum expedition in 1958 revealed the presence of a rich deposit of Middle Triassic vertebrates. As a result of further extensive collections made by the Instituto Lillo of Tucuman, Bonaparte (1962, 1963c) has been able to give a comprehensive description of Exaeretodon (with which Theropsis proves to be synonymous ), and has described two further genera of gomphodonts, Proexaeretodon and Ischignathus (Bonaparte, 1963d, 1963a). Sr. Bonaparte informs me that he has under de- scription a small diademodont gomphodont from the early Triassic Puesto Viejo Formation of southern Mendoza Province. According to Dr. Donald Baird, a gomphodont, as yet unde- scribed, is present in the lower beds of the Newark series of Nova Scotia — the only gomphodont as yet reported from the Northern Hemisphere. Most recent of gomphodont-bearing faunas to be discovered is that of the Chanares Formation. Forms of this nature are here exceedingly abundant, and, as seen in the field, appear to make up a very large proportion (? one-half) of the total number of specimens recovered. In the small fraction of our collection now available for study there appear to be at least two further distinct genera present, in addition to the forms described above. The gomphodonts are obviously an herbivorous side branch of the advanced cynodont stock. The circumstances which led to this radical departure from the typically carnivorous mode of life of other cynodonts and, indeed, of all other theriodonts generally, are far from clear. In the late Permian a host of dicynodonts, large and small, were the typical herbivores of the times; as anyone who has collected in the Endothiodon or Cistecephalus zones of South Africa knows, 90 per cent or so of all specimens found in these beds are members of the dicynodont group. By the time we reach the Cynognathus Zone, the dicynodonts are much reduced in numbers and variety, and survive then and later in the Triassic only in the form of such large types as Kannemeyeria, Ischigual- astia, and Placerias. In such typical Middle Triassic faunas as those of Ischigualasto and Santa Maria, the common herbivores are, instead, of two types in about equal numbers — rhynchosaurs and gomphodonts. In the Chanares fauna we are dealing with an earlier Meso-Triassic stage. Rhynchosaurs must have been in existence (a few ancestral types are found in the Cynognathus Zone fauna), but presumably were little developed. (We have found no trace of a rhynchosaur in the sample of the collection currently 16 BREVIORA No. 264 available to us, and saw no sure rhynchosaur material while col- lecting.) Gomphodonts are extremely abundant and apparently varied and, in the way of percentages, occupy much the same posi- tion in the Chanares that the dicynodonts did in the Upper Per- mian. Seemingly, it was only later in the Middle Triassic that rhynchosaurs became fully evolved and came to rival the gompho- donts in numbers. Judging by the very considerable number of new forms and the amount of new data unearthed during the last few years, we are currently undergoing, so to speak, an explosive period in our acquisition of knowledge concerning gomphodonts. For this rea- son, and because our knowledge of many of the forms concerned is very incomplete, it is inadvisable at this time to attempt any comprehensive study of gomphodont classification and phylogeny. Pertinent facts, however, may be discussed. In 1954 Haughton and Brink included all African forms then known in the Diademodontidae. In our 1956 classification of therapsids, Watson and I included all known gomphodonts except Gomphodontosuchus in the single family Diademodontidae, and I followed the same pattern in my “Osteology of the Reptiles,” published the same year. Lehman, in 1961, in the Piveteau “Traité” also included all gomphodonts (including Gomphodonto- suchus) in the Diademodontidae, although effecting a subdivision into subfamilies.1 Von Huene, however, had erected a second fam- ily, Traversodontidae, for the Brazilian genera Traversodon and Gomphodontosuchus, and Crompton and Ellenberger advocated a division of the gomphodonts into two major groups, Diademodon- tidae and Traversodontidae. Study of South American gomphodonts brought me in 1961, in contrast to my earlier opinion, to the conclusion that diademodonts and traversodonts (to which Exaeretodon obviously belonged) were distinct groups, but also to feel that the Trirachodon-Cricodon tooth type indicated the presence of a third group. Bonaparte (1963b) advocated two families, but (contra Crompton) would include Gomphodontosuchus, as well as the newly discovered Ischigualasto genera, in the Traversodontidae. Brink (1963) would unite these families into a superfamily Gomphodontoidea (melior Diademodontoidea?) of advanced cynodonts, in contrast to the carnivores — superfamily Cynognathoidea. 1 His term “Eudiademodontinae” is, of course, inadmissable, not being based on the name of a genus. 1967 TWO NEW GOMPHODONTS Ly. Actually, it would seem, on the basis of present knowledge, that three, rather than two, groups of gomphodonts are present in the Lower and Middle Triassic. DIADEMODONTIDAE Here I would place a series of Cynognathus Zone forms typified by Diademodon, and a series of named genera identical or closely related, including: Cyclogomphodon, ?Cynochampsa, Diastemo- don, Gomphodontoides, Gomphognathus, Microhelodon, Octagom- phus, and Protacmon. The entire anatomy is, as mentioned above, adequately known only in Diademodon. The skull is relatively long, as in typical cynognathids, with a rather long and slender facial region, and with the greatest width toward the posterior end. The squamosal is deeply incised medial to the external auditory meatus, giving a V-shaped dorsal outline to the bone in posterior view; this is quite in contrast to the condi- tion in primitive cynodonts and even cynognathids, in which the two portions of the squamosal are broadly connected at the pos- terior margin of the temporal fenestra. In cynognathids the skull shape, as seen from above, is that of an isosceles triangle, the breadth gradually increasing backward to its maximum at the posterior end of the zygoma, whereas in diademodonts there is a major increase in width at the orbits, so that the zygomatic arches lie nearly parallel to the main axis of the skull for their entire lengths. The arches are deep, and are mainly formed by the jugals; the squamosals have but a narrow external exposure pos- teriorly and dorsally. A peculiar feature is a prominent ventral flange below the orbit formed by the jugal. In contrast to some cynognathids there persists in Diademodon (but not in Protacmon) a primitive condition in that the pterygoid-epipterygoid bar extends back to the quadrate. Certain of the features here listed for Diademodon appear to be present in gomphodonts generally; how- ever, the proportions of the “muzzle” are variable, the ventral jugal flange may be absent in some groups, and a broader external exposure of the squamosal is reported in some cases. Four incisors and a stout canine are present above and below; beginning after a short diastema, the diademodont dentition in- cludes a long row of cheek teeth on either side, the two tooth rows diverging with an outward curvature posteriorly. Thirteen or more cheek teeth may be present; here and in at least certain gom- phodonts, teeth may be added posteriorly to the cheek battery during “maturity,” and in Diademodon small anterior teeth may 18 BREVIORA No. 264 be lost and their alveoli closed. The most anterior cheek teeth are peg-like, the most posterior retain a primitive, laterally compressed shape like that of carnivorous cynodonts. The main portion of the dental battery consists of gomphodont “grinders.” Here, as in other gomphodonts, there is a strong tendency for the de- velopment of rows of small bead-like denticulations along the ridges and cusps of the “molars”; however, these rapidly disappear with wear and are seldom seen, except in teeth in process of erup- tion. The evolution of the typical “molars” appears to have taken place by an inward growth, from the original compressed carnivore tooth type, of a “heel” overlapping the lower teeth. The original longitudinal tooth line is represented by two stout cusps along the outer margin; variable cusps may be developed medially. There may be some development of small cusps along anterior and pos- terior margins. A low ridge develops crossing the center of the “heel” from side to side; this, when unworn, shows a row of tiny denticles, but never a developed cusp. The lower “molars” are circular to subquadrate in shape, with a ring of cusps and cuspules surrounding a central basin. There is some suggestion of compari- son with the upper “molars” in that the marginal cusps tend to be more highly developed on the medial and lateral margins, and there is some development of a transverse ridge crossing the basin. The diademodont tooth type is, as far as known, mainly con- fined to forms from the Cynognathus Zone of South Africa. How- ever, Crompton (1955: 656-659, fig. 14 A-D) has described frag- mentary jaws from the Manda beds with mandibular “molars” of diademodont pattern. Further, the imperfect skull and jaws from these same beds which form the type of Aleodon brachyrhamphus (Crompton, 1955) are suggestive, as Crompton notes, of diade- modont relationships, but the teeth are so worn that almost noth- ing can be made of their pattern. ‘TRIRACHODONTIDAE Trirachodon and Trirachodontoides of the Cynognathus Zone and Cricodon of the Manda beds are forms with a dentition of very distinctive type. There are ten to eleven cheek teeth in Cri- codon, but only six reported in Trirachodon. The cheek tooth rows are nearly straight, but with a slight outward curvature pos- teriorly in Cricodon. In this genus the most anterior cheek teeth are small subcircular “premolars” and, at least in the lower jaw, the most posterior teeth, as in Diademodon, “revert” to the narrow cynognathid type. In the Cynognathus Zone forms the lower 1967 TWO NEW GOMPHODONTS 19 “molars” are unknown. The upper cheek teeth of Trirachodon and both uppers and lowers of Cricodon are transversely broadened, but narrow anteroposteriorly. There is a single internal cusp and a single external one, and connecting the two a transverse ridge which with wear develops a prominent longitudinal cusp. It seems clear that this tooth type has developed independently of that of the diademodonts, by medial lateral development from an ancestral type with but a single major cusp; the diademodont type, on the other hand, expanded laterally from a tooth with greater anteroposterior length. Both Crompton and Ellenberger (1957) and Bonaparte (1963d) would include these forms in the Traversodontidae. But while the upper “molars” of typical traver- sodonts might perhaps be derivable from the type seen in the pres- ent group, I think it is impossible to conceive of the lower “molars” of the traversodonts having been derived from the radically differ- ent type seen in Cricodon and presumably present in Trirachodon. A minor but perhaps significant feature suggesting at least a distant trirachodont-diademodont relationship is the presence in Trirachodon of a suborbital process of the jugal comparable to that of Diademodon and Protacmon. Probably of no significance is the fact that the fragment of skull from the Upper Triassic Redbeds named Tritheledon by Broom (1912) shows a row of transversely widened multicuspidate teeth, somewhat suggestive of a trirachodont. TRAVERSODONTIDAE At present this family (as here restricted ) definitely includes two African forms — Scalenodon of the Manda beds and Scalenodon- toides of the Molteno — and a series of South American forms — Traversodon from Santa Maria, Exaeretodon, Proexaeretodon, and Ischignathus from Ischigualasto, and from the Chanares the two species of Massetognathus described above, as well as other forms to be described later. Many of these are incompletely known — Scalenodontoides from a lower jaw alone. Scalenodon from numer- ous but fragmentary remains, Traversodon from isolated elements from which von Huene has restored a skull; Proexaeretodon and Ischignathus are represented by incompletely preserved skulls. As yet Exaeretodon and Massetognathus are the only traversodontids in which adequate skull material is known. Considerable variation in skull structure appears to have been present in the group. All appear to have had, as seemingly in gomphodonts generally, zygo- matic arches placed far laterally in a plane parallel to the long 20 BREVIORA No. 264 axis of the skull. In general the muzzle is narrow; only in Masse- tognathus, as far as known, do we find a “swollen” maxillary region. In all forms where data are available, the angular process of the dentary is more pronounced than in other cynodonts (includ- ing Diademodon) and the angular-surangular more exposed lat- erally. The number of cheek teeth is variable — 8 maxillary teeth, for example, are figured in Scalenodon and Traversodon, 9 in Exaeretodon, 11 in Proexaeretodon, 7 in Ischignathus, 11 to 15, as noted, in Massetognathus. The “molars,” and most especially those of the mandible, with a pair of prominent crests across the anterior border (behind which is a broad “heel’’), are the most clearly diagnostic feature of the group. The mandibular teeth are not at all comparable with those of Cricodon (or those expected in Trirachodon), and offer a major reason for separating those genera from the Traversodontidae. As far as they are known in members of the South American fauna, the upper “molars” are nearly equally distinctive. There are here two external cusps, the posterior the major one; a prominent transverse ridge along the posterior margin, with a typical develop- ment of two cusps toward and at the internal edge; anterior to the posterior crest a concave basin, bounded in front and laterally by a ridge much lower than that behind. The cheek teeth are set closely in series, and even in Massetognathus each tooth “shoulders” slightly into the one ahead somewhat medial to the lateral border. This “shouldering” is much more pronounced in Proexaeretodon and Exaeretodon, so that in the latter genus the tooth appears to consist of two portions, external and internal, set at a considerable angle to one another. Of the two African forms, Scalenodon and Scalenodontoides, definitely assignable to the traversodonts, both exhibit lower “molar” patterns characteristic of the family. The upper dentition is unknown in the latter genus, but it is present and well described by Crompton in the former. The maxillary “molars” are basically comparable to those of South American forms, but differ in that there is but a single external cusp, without development of the accessory anterior one, and, further, in that the prominent trans- verse ridge, bearing two cusps toward and at the inner margin, lies not at the posterior border of the tooth but some distance anterior to it. These differences suggest that Scalenodon is rather more primitive than the described South American forms. Possibly we may see here a stage in the evolution of the traversodont upper “molar” from one resembling that of Trirachodon, but as remarked 1967 TWO NEW GOMPHODONTS 21 above, such an ancestry seems out of the question as regards the lower cheek teeth. However, there is evidence in the Manda beds of a form with more advanced upper “molars” — this is represented by a maxilla which Crompton (1955: 659-660, fig. 14 E) compared tentatively with the Brazilian Gomphodontosuchus, but which is obviously distinct (there are alveoli for at least ten cheek teeth, whereas the short-jawed Gomphodontosuchus has but six). The two “molars” preserved are somewhat worn, but show definitely the presence of two external cusps, a central basin, a low, crenulated, anterior transverse ridge, and a high posterior one. Further, there is a slight posterior marginal concavity on each tooth, indicating a be- ginning of the “shouldering” which became more prominent in some of the Argentinian genera. It appears probable that we are dealing with a form closely related to, and not impossibly belonging to the genus Massetognathus. Significant in this connection is the fact that in this specimen there is, as Crompton notes, a pro- nounced outward “bulge” of the maxilla lateral to the tooth row, as in Massetognathus but not reported in any other described gomphodont. FORMS OF DOUBTFUL ASSIGNMENT Apart from genera which seem clearly to belong to one or an- other of the three groups distinguished above, there are a number of forms which cannot be assigned, mainly because of lack of knowledge of molar pattern. Inusitatodon (Brink and Kitching, 1953) is a small Cynognathus Zone form represented by an incomplete skull. The jaws are tightly locked and the crown pattern of the cheek dental battery — which includes two “premolars” and apparently eight “molars” —is unknown. Brink and Kitching compare the skull with that of Diademodon in various regards, but, without giving reasons for this, state that Trirachodon is the closest relative. In the newly described Cragievarus (Brink, 1965) from the Cynognathus Zone, the cheek tooth pattern is, unfortunately, not described. The “molars” are subquadrate in outline, suggesting reference to the Diademodontidae, and Brink notes various points of agreement with Diademodon. However, the cheek tooth row is short (? nine teeth above, seven below), the skull lacks the sharp expansion in width at the orbits usual in gomphodonts, the zygomatic arch is weak, the jugal lacks the ventral process seen in Diademodon and Protacmon, and the angular process of the dentary is less developed than in typical gomphodonts. Possibly 22 BREVIORA No. 264 this form is a primitive but somewhat aberrant diademodontid. Luangwa (Brink, 1963: 89-93), from the Ntawere Formation of Northern Rhodesia, is obviously, from its expanded cheek teeth, a gomphodont, but the crown pattern is unknown. The shortness of its tooth row (only five “molars” and a miniscule “premolar’’) is unmatched, except in Trirachodon and Gomphodontosuchus, but the shape of its subquadrate “molars” is quite different from those of the former (narrow anteroposteriorly) and from the triangular “molars” of the latter genus. The general skull pattern and the “incipient” ventral jugal process suggest the Diademodontidae; pos- sibly we are dealing with a juvenile. Brink states that its closest ally is Scalenodon, but gives no reasons. Aleodon from the Manda (Crompton, 1955) is, again, surely a gomphodont, but, again, we have almost no data on tooth cusps. The skull is little known. The rather “fat” shape of the “molars” suggests the Diademodontidae. Theropsodon of the Manda beds was described by von Huene (1950) on the basis of a nearly complete but poorly preserved skull. It is definitely a gomphodont, according to von Huene, but the tooth pattern cannot be determined. Possibly it may be identi- cal with either Cricodon or Scalenodon, from these same beds, or may represent a true diademodont, of which individual teeth are known to be present there. Gomphodontosuchus from the Brasilian beds is apparently a somewhat aberrant form, with a short face, short tooth row, and massive jaws. The crown pattern of the “molars” is imperfectly preserved. Watson and I, in 1956, assigned this form to a separate family; Crompton and Ellenberger (1957) assigned it to the diade- modonts; Bonaparte advocates a position in the Traversodontidae. The last is perhaps the best suggestion, but better material is needed. Bonaparte (1963b) has suggested that Belesodon of the Santa Maria beds might prove to be a traversodont. However, the Chanares fauna includes a form, as yet undescribed, which is obvi- ously related and probably antecedent to Belesodon, in which teeth of cynognathid type are preserved. Watson, Kiihne, Crompton, and Bonaparte have all successively suggested that the tritylodonts were of gomphodont derivation. This may well be the case. In no other therapsid group do we find any appreciable trend for the development of multicuspidate molars. In forms such as [schignathus we find a shortened, com- pact, “molar” series and a considerable postcanine diastema — features to be sought in a tritylodont ancestor. But although 1967 TWO NEW GOMPHODONTS 23 Crompton has hopefully pointed out that it would be possible to derive tritylodont molars from those of traversodonts, there is still a very considerable structural gap to be bridged. It is to be hoped that further exploration of Middle Triassic deposits will yield forms bridging this gap. A further possibility is that the gomphodonts may be in some way related to the ancestry of the monotremes. Their ancestry is a complete blank (despite a suggestion at one time of docodont rela- tionships) and it is generally felt that they probably evolved from therapsids of some sort quite separately from other mammals. The curiously multicuspidate “molars” of the forms here described are suggestive in general (although not in detail) of the likewise multi- cuspidate molars of Ornithorhynchus. It may well be that mono- tremes are descended from gomphodonts of some sort. LITERATURE CITED ATTRIDGE, J.. H. W. BALL, A. J. CHARIG, and C. B. Cox 1964. The British Museum (Natural History) — University of London joint paleontological expedition to Northern Rhodesia and Tan- ganyika, 1963. Nature, 201: 445-449. BONAPARTE, J. F. 1962. Descripcion del craneo y mandibula de Exaeretodon frenguellii, Cabrera y su comparacion con Diademodontidae, Tritylodonti- dae y los cinodontes sudamericanos. Publ. Mus. Munic. Cienc. Nat. Tradic. Mar del Plata, 1: 135-202. 1963a. Descripcion de Ischignathus sudamericanus n. gen. n. sp., nuevo cinodonte gonfodonte del Triasico Medio superior de San Juan, Argentina. Acta Geol. Lilloana, 4: 111-118. 1963b. La familia Traversodontidae. Acta Geol. Lilloana, 4: 163-194. 1963c. Descripcion del esqueleto postcraneano de Exaeretodon. Acta Geol. Lilloana, 4: 5-52. 1963d. Un nuevo cinodonte gonfodonte del Triasico Medio superior de San Juan, Proexaeretodon vincei n. gen., n. sp. Acta Geol. Lil- loana, 4: 129-133. BRINK, A. S. 1955. A study on the skeleton of Diademodon. Palaeont. Afr., 3: 3-39. 1963. Two cynodonts from the Ntawere Formation in the Luangwa Valley of Northern Rhodesia. Palaeont. Afr., 8: 77-96. 1965. A new gomphodont from the Cynognathus zone of South Africa. Palaeont, Afr., 9: 97-105. BRINK, A. S. and J. W. KITCHING 1953. On some new Cynognathus zone specimens. Palaeont. Afr., 1: 29-48. 24 BREVIORA No. 264 BROILI, F. and J. SCHROEDER 1934. Beobachtungen an Wirbeltieren der Karrooformation. II. Uber den Cynodontier Tribolodon frerensis Seeley. Sitzungsber. Bayer. Akad. Wiss. Munchen, 1934: 163-190. 1935. Beobachtungen an Wirbeltieren der Karrooformation. [X. Uber den Schadel von Gomphognathus Seeley. Sitzungsber. Bayer. Akad. Wiss. Munchen, 1935: 115-182. Broom, R. 1912. On a new type of cynodont from the Stormberg. Ann. S. Afr. Mus., 7: 334-336, pl. XXII. CABRERA, A. 1943. El primer hallazgo de terapsidos en la Argentina. Notas Mus. La Plata, Paleont., 8: 317-331. CROMPTON, A. W. 1955. On some Triassic cynodonts from Tanganyika. Proc. Zool. Soc. London, 125: 617-669. 1964. A preliminary description of a new mammal from the Upper Triassic of South Africa. Proc. Zool. Soc. London, 142: 441- 452. CromPTon, A. W. and F. ELLENBERGER 1957. On a new cynodont from the Molteno beds and the origin of the tritylodonts. Ann. S. Afr. Mus., 44: 1-14. HAUGHTON, S. H. and A. S. BRINK 1954. A bibliographical list of Reptilia from the Karroo beds of Africa. Palaeont. Afr., 2: 1-187. HUENE, F. VON 1928. Ein Cynodontier aus der Trias Brasiliens. Centralbl. Min. Geol. Paladont., 1928, Abt. B: 251-270. 1942. Die fossilen Reptilien der sidamerikanischen Gondwanalandes. Munich, 332 pp. 1950. Die Theriodontier des ostafrikanischen Ruhuhu-Gebietes in der Tiibinger Sammlung. Neues Jahrb. Geol. Palaeont., Abh., 92: 47-136. 1956. Palaontologie und Phylogenie der niederen Tetrapoden. Jena, 716 pp. LEHMAN, J. P. 1961. Cynodontia. Jn: Triaté de Paleontologie, ed. J. Piveteau, VI, vol. I: 140-191. RomeER, A. S. 1956. Osteology of the reptiles. Chicago, 772 pp. 1961. Synapsid evolution and dentition. Jn: International colloquium on the evolution of mammals. Kon. Vlaamse Acad. Wetensch. Lett. Sch. Kunster Belgié, Brussels, 1961, Part 1: 9-56. 1966. The Chafares (Argentina) Triassic reptile fauna. I. Introduc- tion. Breviora, No. 247: 1-14. 1967 TWO NEW GOMPHODONTS 25 Romer, A. S. and J. A. JENSEN 1966. The Chanares (Argentina) Triassic reptile fauna. II. Sketch of the geology of the Rio Chanares-Rio Gualo region. Breviora, No. 252: 1-20. WATSON, D. M. S. 1911. The skull of Diademodon with notes on those of some other cynodonts. Ann. Mag. Nat. Hist., (8)8: 293-330. 1913. Further notes on the skull, brain, and organs of special sense of Diademodon. Ann. Mag. Nat. Hist., (8)12: 217-228. 1920. On the Cynodontia. Ann. Mag. Nat. Hist., (9)6: 506-524. WATSON, D. M. S. and A. S. ROMER 1956. A classification of therapsid reptiles. Bull. Mus. Comp. Zool., 114: 37-89. (Received 26 September, 1966.) RAEC |W Bee al oN I — }) oe 1 ee 1\¢y, « oe ; 7 Ope . hi] eg mee ME OE DT ATE wy TI 302 746 Pe tae Se So. ee ey a ate pn Ye en ie me Pe Ae a kA at ne In oh eiead aaeR pet Te eke ers be