==^= ru ^^^ r^ Q— 3- § a -- - " □ NEW YORK ACADEMY OF SCIENCES SCIENTIFIC SURVEY OF Porto Rico and the Virgin Islands VOLUME X NEW YORK : Published by the Academy 1930 CONTENTS OF VOLUME X Page Title-page. Contents ^ Dates of Publication of Parts " List of Illustrations iv Amphibians and Land Reptiles of Porto Rico, with a List of Those Reported from the Virgin Islands. By Karl Patterson Schmidt 1 The Fishes of Porto Rico and the Virgin Islands — Branchiostomidae to Sciae- nidae. By J. T. Nichols 161 The Fishes of Porto Rico and the Virgin Islands — Pomacentridae to Ogcoce- phaUdae. By. J. T. Nichols 297 The Ascidians of Porto Rico and the Virgin Islands. By Willard G. Van Name 401 Index 5 ' 3 Dates of Publication of Parts Part 1, November 22, 1928. ^ Part 2, September 10, 1929. ^"^ *7 jL mL. Part 3, March 15, 1930 Part 4, August 1, 1930 (iii) iv SCIENTIFIC SrEVEY OF POETO EICO LI^T OF ILLUSTRATIONS Figures Page Head of toad (left) contrasted with LeptodacUjlus (right) 30 Foot of Leptodadylus (left) contrasted with foot of Eleutherodadylus (right). Compare slender and expanded tips of digits 30 Head of Bufo lemur 32 Habitus of juvenile Bufo lemur (A), with side view of head (B). A. M. N. H. No. 1015L Natural size 33 Bufo marinus. Maj-agiiez 35 Leptodadylus alhilabris, to show variation in color pattern and in form of snout. A. M. N. H. No. 10125 (A and B); A. M. N. H. No. 10143 (C and D). Natural size 39 Leptoiadylus dominicensis. A. M. N. H. No. 20952 for comparison with L. alhilabris. Three-fourths natural size 39 Lateral view of tadpole of Leptodadylus alhilabris 42 Mouth parts of tadpole of Leptodadylus alhilabris 43 Inside of mouth of Eleutherodadylus portoricensis (left) and of E. richmondi (right), showing different arrangement of the vomerine teeth 45 Three common color variants of Eleutherodadylus portoricensis. A, A. M. N. H. No. 10139; B, No. 10243; and C, No. 10249. Natural size 46 Embryo of EleidherodadT/lus portoricensis. A. M. N. H. No. 10302. Six times natural size 48 Peter's figures of the embryo of Eleutherodadylus cortoricensis 49 Eleutherodadylus gryllus. A. M. N. H. No. 10226. Twice natural size 51 Eleutherodadylus locustus. A. M. N. H. No. 10240. Twice natural size 53 Eleutherodadylus cramptoni. A. M. N. H. No. 10305. Twice natural size. ... 55 Eleutherodactylus antillensis. A. M. N. H. No. 10019. Twice natural size 57 Eleutherodactylus brittoni. A. M. N. H. No. 10318. Twice natural size 60 Eleutherodadylus wightinanae. A. M. N. H. No. 10220. Twice natural size . . 61 Eleutherodadylus richmondi. A. M. N. H. No. 10237. Twice natural size. ... 63 Eleutherodadylus monensis. A. M. N. H. No. 24463 65 Eleutherodadylus unicolor. Size of head (left), top of head (center), and inside of mouth (right) of type 66 Eleutherodactylus unicolor. U. S. N. M. No. 26963, type. Twice natural size. 67 Digits of Hemidactylus mahouia (left) and Sphaerodadylus macrolepis (right) contrasted 68 Head of Hemidactylus mabouia 70 Head and shoulders of Sphaerodactylus macrolepis. A. M. N. H. No. 13037 (A) and No. 13697 (B), showing two common types of pattern. Two and a half times natural size 72 Heads of Porto Rican Anolis. Anolis cuvieri (left of top row), Anolis cristatellus (center of top row), Anolis gundlachi (right of top row); Anolis evermanni (left of middle row); Anolis stratulus (center of middle row), Anolis krugi (right of middle row); Anolis pulchellus (left of bottom row), Anolis poncensis (right of bottom row) 76 Caudal crest of Anolis cuvieri (left), of A. gundlachi (center), and A. cristatellus (right) 78 CONTENTS OF VOLUME X v Page Dorsal scales of Porto Rican Anoles related to Anolis pulchellus. Left to right: A. krugi, A. pulchellus, and A. poncensis 94 Head of Cyclura stejnegeri (type) 103 Celestus pleii. A. M. N. H. No. 13133. Natural size 106 Ameiva wetmorei. A. M. N. H. No. 13820. A. Head from above. B. Head from side. C. Head from below. D. Arm from in front. E. Posterior face of leg. F. Foot from above. G. Preanal scales. Three times natural size 109 Ameiva wetmorei. A. M. N. H. No. 13821. Natural size 110 Ameiva alboguttata. A. M. N. H. No. 14003. Mona Island. Natural size. . . 116 Head of AmpMsbaena caeca from above and from side 118 Head of Amphisbaena bakeri from above and from side 120 Head of Mabuya sloanii from above. A. M. N. H. No. 14007 (A) and A. M. N. H. No. 14006 (B). To show variation in pattern. Twice natural size. 122 Head of Mabuya sloanii from side. A. M. N. H. No. 14007. Twice natural size 122 Pattern of tail of Typhlops Plalycephalus (left) contrasted with that of T. rostellatus (right). A. M. N. H. Nos. 13336 and 13179. Natural size.. . 126 Heads of Porto Rican Typhlops. Left to right (upper), T. platyecphalus; T. rostellatus; (lower), T. monensis. (First two species from Stejneger; last from Schmidt.) 127 Head of Epicrates inornatus from above (left), after Stejneger. Head of Epicrates monensis from above (right), after Schmidt 131 Dromicus stahli, head from above and from below 135 Color-pattern of Dromicus stahli 136 Head of Dromicus exiguus from above and from the side 137 Alsophis antillensis, head from above and from side 140 Color-pattern of Alsophis antillensis 140 Head of Alsophis portoricensis from above, showing scale-pits in dorsal scales. . 143 Color-pattern of Alsophis portoricensis. A. M. N. H. No. 8435. Twice natural 144 size ^*^ Color-pattern of Alsophis variegatus. A. M. N. H. No. 13774 145 Carapace and plastron of Pseudemys stejnegeri. One-half natural size 148 Head of Pseudemys stejnegeri from below and from side, to show color-pattern. . 150 Branchiostoma caribaeum 1*0 Asymmetron lucai/anum. Johns Hopkins Univ. Studies. Biol. Lab. V 181 Ginglymostoma cirratum. From Zoologica, IX 181 Galeocerdo tigrinus. From Zoologica, IX 18? Carcharhinus falciformis. From Zoologica, X 183 Carcharhinus limbatus, From Zoologica, X 183 Sph-yrna zygaena. From Zoologica, IX 184 Pristis pectinatus. Breder's Field Book of Marine Fishes (Putnam) 185 Dasyatis americana. Breder's Field Book of Marine Fishes (Putnam) 186 Dasyatis say. Breder's Field Book of Marine Fishes (Putnam) 187 Aetobatus narinari. From Zoologica, X 188 Anguilla rostrata. From Zoologica, IX 189 Leptocephalus conger. From Zoologica, X 190 VI SCIENTIFIC SURVEY OF PORTO RICO Page Mayerina mayeri ^^^ Muraenesox savanna Aphthalmichthys caribbeus ^^1 Myrophis longleii ^^^ Chilorhinus suensonii -^^^ Schagebranchus ophioneus ^^3 Myrichthys oculatus. From Zoologica, X 193 Myrichthijs acuminatus. From Zoologica, X 194 Myrichthys keckii ^94 Ophichthus gomesii 195 Gymnothorax moringa. From Zoologica, X 196 Gijmnothorax funebris. From Zoologica, X 196 Gijmnothorax albimentis 19 ' Gymnothorax jordam Echidna catenqta Tarpon atlanticus. From Zoologica, IX 198 Flops saurus. From Zoologica, IX 199 Albula vulpes. From Zoologica, IX 200 Jenkinsia lamprotaenia. From Zoologica, X 201 Sardinella anchovia. From Zoologica, IX 201 Harengula sardina. From Zoologica, X 202 Harengula macrophthalma. From Zoologica, X 203 Opisthonema oglinum. From Zoologica, IX 203 Anchovia perfasciata. From Zoologica, IX 204 Anchovia brownii. From Zoologica, IX 205 Anchovia choerostoma. From Zoologica, X 205 Anchovia lyolepis. From Zoologica, X 206 Cetengraulis edentulus. From Zoologica, X 2C6 Trachinocephalus myoos. From Zoologica, IX 207 Synodus intermedius. From Zoologica, X 207 Synodus foetens. From Zoologica, IX 208 ■ 20Q Carassius auratus 91 r\ Fundulus fonticola *■ 210 Poedlia vtvipara 211 Tylosurus notatus 919 Tylosurus ardeola " Tylosurus raphidoma. From Zoologica, X 212 Tylosurus acus. From Zoologica, IX 213 Hyporhamphus unifasciatus. From Zoologica, X 214 Hemiramphus brasiliensis. From Zoologica, IX 214 Parexocoetus brachypterus. From Zoologica, IX 215 Cypselurus bahiensis. From Zoologica, X 215 Aulostomus maculatus. From Zoologica, X 216 Fistularia tabacaria. From Zoologica, IX 216 Syngnathus mackayi. From Zoologica, X 217 Syngnathus floridae Syngnathus elucens. From Zoologica, X 218 Hippichthys cayorum COXTEXTS OF VOLUME X vii Page Hippichthys ensenadae 219 Doryrhamphus sierra 219 Hippocampus punctulatus. From Zoologica, X 220 Atherina stipes. From Zoologica, X 220 Atherina araea. From Zoologica, X 221 Mugil brasiliensis 221 Mugil curema. From Zoologica, IX 222 Mugil trichodon 222 Agonostomus moniicola. From Zoologica, X 223 Sphyraena barraaida. From Zoologica, IX 224 Sphyraena guachancho. From Zoologica, X 224 Sphyraena picudilla. From Zoologica, X 225 Polynemus virginicus. From Zoologica, X 225 Myripristis jacobus. From Zoologica, X 226 Holocentrus ascensionis. From Zoologica, X 226 Holocentrus vexillarius. From Zoologica, X 227 Upeneus maculatus. From Zoologica, X 227 Upeneus parvus 228 Upeneus martinicus. From Zoologica, X 228 Auxis thazard. From Zoologica, IX 229 Scomberomorus maculatus. From Zoologica, IX 229 Scomberomorus regalis. From Zoologica, IX 230 Scomberomorous cavalla. From Zoologica, IX 230 Trichiurus lepturus. From Zoologica, IX _. 231 Oligoplites saurus. From Zoologica, IX 232 Seriola falcata 232 Decapterus punctatus. From Zoologica, IX 233 Trachurops crumenophthalmus. From Zoologica, IX 233 Caranx ruber. From Zoologica, X 234 Caranx bartholomaei. From Zoologica, IX 235 Caranx hippos. From Zoologica, IX 235 Caranx crysos. From Zoologica, IX 236 Caranx latus. From Zoologica, X 237 Vo7ner setapinnis. From Zoologica, IX 237 Vomer setapinnis cubensis. From Zoologica, X 238 Selene vomer. From Zoologica, IX 239 Chloroscombrus chrysurus. From Zoologica, IX 239 Trachniotus glaucus. From Zoologica, X 240 Trachinotus falcatus. From Zoologica, IX 241 Trachinotus carolinus. From Zoologica, IX 241 Nomeus gronovii. From Zoologica, IX 242 Peprilus paru. From Zoologica, IX 243 Apogon selUcauda 244 A pogon conklini 244 Apogonichthys alutus 245 Apogonichthys stellatus. From Zoologica, X 245 Centropomus undecimalis. From Zoologica, X 246 Centropomus parallelus 246 viii i : Be-^ides the material collected by the Survey of Porto Rico and the Virgin Islands, all of which is deposited in The American Museum of Natural History, I have had the privilege of examining, thanks to the courtesy of Dr. W. C. F. McClure, the Porto Rican collection preserved at Princeton University. ., Dr. Stuart T. Danforth, of tlie University of Porto Rico, at Maya- giiez, kindly sent me both his personal collections and those of the Uni- versitv for examination in connection with this report. . ,, .. 8CHM1UT, AMPHIliIAX8 OF PORTO. RICO 5 Plan of Work ''The Herpetology of Por;to Rico"' bv Dr. Leonhard Stejneger (1904) is a work of exceptional merit. It remains a model for the exact and complete description of an insular famia, and sets a high standard for systematic zoology in general. It is a pleasure to record here the use- fulness of this volume. A copy accompanied me to Porto Eico in 1919 and proved most serviceable as a field manual, making possible the identification of most of the species and thus facilitating all phases of field study. It was my first plan to prepare merely a supplement to Dr. .Stejneger's report, embodying only the additions to our knowledge of the Porto Rican herpetological fauna since 1904. After a review of the necessary additions, in conference Math Dr. PI. C. Crampton, it was decided, how- ever, to enlarge the scope of the work and present a renewed "complete account" both for the sake of increased usefulness to future students and to bring it into better accord with the similarly complete reports of other contributors to the Survey. The existence of Stejneger's report has greatly simplified the preparation of the present one. In the case of the numerous species whose definition has required no change, I have followed Stejneger's . descriptions closely or quoted them ver- batim, and I have availed myself of a large number of his text figures, especially for the illustration of key characters. The figures drawn for the present paper are designed to present the habitus of a number of species, and thus supplement Stejneger's otherwise complete illustra- tion of the fauna. These figures are the work of Mrs. E. L. Beuten- mliller, whose drawings have embellished so many herpetological papers. The half-tone figure of Eleuiherodaciylus unicolor was supplied through the courtesy of Dr. Stejneger. I have adopted a conservative position on one phase of nomenclature. Excellent arguments might be advanced for treating several of the Porto Rican forms as subspecies rather than as full species. Such a nomen- clature would reflect more information as to the actual relations of the forms concerned than binomial treatment. The species of Tijphlops allied to jamaicensis, the fresh water turtle, and the Mona Island Ameira and Cydura are cases in point. It is very difficult, liowever. to draw a line between insular subspecies and insular species, and our knowledge of many forms is manifestly imperfect. Any attempt at a trinomial arrangement of Eleutherodactylns is obviously impossible. I have accordingly left the matter for future consideration, preferably in connection with a new list of the West Indian fauna as a whole. Q SCIENTIFIC SURVEY OF PORTO RICO So much work still remains to be done on the herpetological fauna of Porto Eico by some resident naturalist, especially with reference to the discrimination of the small tree frogs and their life histories, that the present account of the fauna is hardly more likely to be "final" than was that of Dr. Stejneger more than twenty years previously. Acknowledgments It is a pleasure to acknowledge the cordial furtherance of the present work by the meml)ers of the Porto Rico Committee of the New York Academy of Sciences, by my various sometime colleagues of The Ameri- can Museum of Natural History who took part in the Survey of Porto Eico and the Virgin Islands, and by nearly everyone with whom we came in contact in the course of the herpetological field-work. We were especially indebted, when in Porto Eico, to Mr. and Mrs. B. A. Wall, of Bayamon ; to Mr. E. M. Bruner, Forester of Porto Eico ; to Mr. Marc Lejeune, of Mayagiiez, to whom I owe the visit to Mona Island ; and to Colonel George A. Shanton, Chief of the Insular Police. In the course of the preparation of the report I have had the most cordial aid from various herpetologists. I ha\e applied for information, for specimens or for advice to Dr. Leonhard Stejneger and Miss Doris Cochran, of the United States National Museum ; to Dr. Thomas Bar- bour and Mr. Arthur Loveridge, of the Museum of Comparative Zool- ogy; to Dr. G. K. Noble and Mr. Clifi:ord H. Pope, of The American Museum of Natural History ; to Dr. Emmett Reid Dunn, of Smith Col- lege; to Dr. Stuart T. Danforth, of the University of Porto Eico; and to Mr. H. W. Parker, of the British Museum (Natural History). The friendly criticism and interest of Mr. Herbert F. Schwarz, now editor of the reports of the Survey of Porto Eico, have improved the pres- ent paper at innumerable points, both in minor details and in more im- portant matters. My thanks (and still more those of the reader) are due to him for great patience with a difficult manuscript. Porto Eican Heepetology since 1904 Stejneger presents an excellent historical review of the growth of our knowledge of the amphibians and reptiles of Porto Eico (1904, pp. 553- 559). The small but interesting collection secured by W. W. Brown, Jr., on Mona Island in February, 1893, has since come to light and was reported on by myself (192G). Subsequent to the collections made for the United States National Museum in 1899-1901, no mention of Porto Eican herpetology appeared SCHMIDT, AMPHIBIANS OF PORTO RICO 7 until 1913, when Stejneger described the unusually interesting and ex- tremely distinct Ameiva wetmorei from Eio Loco, near Guanica. The type was collected by Dr. Alexander Wetmore in the course of his in- vestigations of the Porto Rican bird fauna. The collections made by Mr. Charles F. Silvester, while on the staff of the expedition of the Carnegie Institution to Porto Rico in 1915, were reported upon by Fowler in 1918. Fowler figures Ameiva wetmorei and discusses variation in other species. The discovery of bones referable to an extinct species of Cyclura in a cave near Ciales by Dr. Glover M. Allen and James Lee Peters, in 1917, filled an important gap in the distribution of this typically Greater Antillean genus. The species was described by Barbour (1919), the type being the extremities of a left humerus, with numerous additional limb-bones, jaws and vertebrae. Similar material was collected for the Scientific Survey of Porto Rico and the Virgin Islands by H. E. An- thony in 1916. The herpetological collecting of the various workers who have taken part in the Scientific Survey of Porto Rico and the Virgin Islands has been described above. Dr. E. Greywood Smyth, Entomologist for the Porto Rican Agricul- tural Experiment Station at Rio Piedras, has paid some attention to the amphibians and reptiles and in 1920 published a brief account of the food habits of the Anoles. The food habits of Porto Rican lizards were subsequently analyzed in some detail by George N. Wolcott, in a paper published in 192-1 in the Journal of the Department of Agriculture of Porto Rico. A small collection made in the course of ornithological investigations in 1924-1925 was reported upon by Stuart T. Danforth (1925 and 1926). This material was subsequently purchased by the Field Mu- seum of Natural History. Mr. Danforth has also collected on Desecheo Island, adding Ameiva exsul to the list from that island in 1926. Lists of the Amphibiaxs axd Land Reptiles of Poeto Rico and THE Adjacent Islands I. PORTO RICO 1. Bufo lemur 5. Eleutherodaciylus gryllus 2. Bufo marinus* 6. Eleidh erodactylus locustus 3. Leptodactylus alhlJabri^ 7. Eleutherodactylus cramptoni 4. Eleutherodaciylus portoricensis 8. Eleutherodaciylus antillensis * Introduced. 8 SCIENTIFIC SURVEY OF PORTO RICO 9. Eleutlierodaciylus hrittoni 10. Eleutherodacf i/his wlglitmanae 11. Eleutherodactylux richmondi 12. Eleutherodacf yJ us unicolor 13. Hemidactylus mahouia 14. Sphaerodactylus macrolepis 15. Anolis cuvieri 16. Anolis cristateUus 17. Anolis gundlachi 18. Anolis evermanni 19. Anolis stratulus 20. Anolis krugi 21. Anolis pulchellv^ 22. Anolis poncensis 23. -fCydura portoricensis 24. Celestus pleii 25. Ameiva exsul 2'6. J_?neii/'rt. wetmorei 27. Amphisbaena caeca 28. AmpJiisbaena haJceri 29. Mabuya sloanii 30. Typhlops platycephalus 31. Typhlops rosfellatus 32. Epicrates inornafus 33. Dromicus stahli 34. Also phis portoricensis 35. Alsophis antillensis 36. Pseudemys stejnegeri II. MOXA ISLAND The fauna of Mona Island, which adds six species to the above list, is as follows : 1. Eleutherodact ylus monensis 2. Sphaerodactylus macrolepis 3. Anolis cristateUus 4. Cyclura stejnegeri 5. Ameiva alboguttata 6. Mabuya sloanii 7. Typhlops monensis 8. Epicrates monensis 9. Alsophis variegatus III. DESECHEO ISLAND Desecheo Island is rarely visited. Herpetological specimens were se- cured by Bowdish in 1901. by Lntz in 1914 and by Danforth in 1926. The species known are : 1. Anolis cristateUus 3. Alsophis portoricensis 2. Ameiva exsul IV. VIEQUES ISLAND Ten species, all of them identical with Porto Rican forms, are known from the island of Vieques. These are : 1. Leptodactylus albilahris 2. Eleuth erodactylus antillensis 3. Sphaerodactylus macrolepis 4. Analis cristateUus 5. Anolis stratuilus 6. Anolis pulcheUus 7. Anolis cuvieri 8. Ameiva exsul 9. Mabuya sloanii 10. Alsophis antillensis HVnMIDT. AMPHIBIANS OF PORTO RICO 9 V. CULEBRA ISLAND The Culebra fauna lacks Sphaerodactyhis, wliieli has doubtless merely been overlooked. It adds a Virgin Island form, Dromicus exiguus, to the fauna under consideration. Its species are : Leptodactylus albilabris Ameiva exsul Eleutherodacfylus antUlensis Mabuya sloanii Anolis cristateUus Dromicus exiguus Anolis stratulus Also phis anUllensis Anolis pulchelhis VI. CAJA DK ML'ERTOS ISLAND Anthony and Goodwin secured four lizards and a snake from this island during their field-work in 1926. These represent three species: Anolis cristateUus Alsophis portoricensis Ameiva wetmorei Habitat Associations and Faunal Subdivisions Porto Eico includes a wide range of habitat conditions, from the ex- tremely wet mountain rain forest of the Luquillo, where mountain palms and hardwoods are hung with lianas and draped with moss that never dries out, to the opposite extreme of aridity on the southwest corner of the island (near Guanica and Ensenada), where a cactus flora predominates. Some of the types of habitat, wath distinct associations of reptiles and amphibians, appear to be the following: I. Northern Coastal Plain (Collections secured from Santurce, Rio Piedras, Bayamon and Mayagiiez). II. Coffee Belt, 900-2000 ft. (Collections from Aibonito and Maricao). III. Deforested Hilltops, above 2000 ft. (Collections secured at Aibo- nito and Maricao). IV. High Rain Forest, 1200-3485 ft. (Collections secured from El Yunque, Luquillo Forest Reserve). V. Pepino Limestones (Collection from Catano). VI. Arid Limestones, southwestern Porto Rico (Collections from Coamo Springs, Ensenada and Salinas). This list is quite inadequate from an ecological standpoint and in it only II, III and IV approach the definition of Biotopes, with recogniz- able Biocoenoses. Turning first to the distribution of the fauna in Porto Rico itself, a 10 SCIENTIFIC SURVEY OF PORTO RICO number of corrections are necessary in the account of the vertical distribution given by Stejneger. These will be presented in detail be- low, in the systematic discussion of the species. In general, recent observations show that altitude in itself has played a relatively small part in determining the distribution of the fauna. Thus Anolis piil- chellus, which Stejneger believed to be confined to the coastal plain, be- low 500 feet, is present at all altitudes, at least up to 2000 feet, in open fields; and Anolis hrugi, for the most part confined to the coffee belt, is found as far down as Coamo Springs (500 ft. alt.) where the conditions of moisture and shade are suitable. The species which are abundant at the lower altitudes (i. e., on the coastal plain) and extend in varying de- grees into the higher are the following: 1. Bufo lemur* 12. Ameiva exsul'* 2. Leptodactylus alhilahris 13. Ameiva wetmorei* 3. Eleutherodactylus portoricensis 14. Amphisbaena caeca 4. Eleutherodactylus antillensis 15. Mabuya sloanii* 5. Hemidactylus mabouia''' 16. Typhlops platycephalu-r-' 6. Spliaerodactylus macrolepls 17. Typhlops rostellatus 7. Anolis Guvieri 18. Epicrates inornatus 8. Anolis cristatellus 19. Dromicus stahli 9. Anolis stratulus 20. Alsophis portoricensis 10. Anolis pulchellus 21. Alsophis antillensis* 11. Anolis poncensis* 22. Pseudemys stejnegeri* Of these only nine (marked with an asterisk) are, so far as known, confined to the coastal plain, or to altitudes below 500 feet. Bufo marinus may now be added to this list. The species, on the other hand, which do not occur on the coastal plain or at least only as stragglers, are the following : 1. Eleutherodactylus gryllus 7. Eleutherodactylus unicolor 2. Eleutherodactylus locustus 8. Anolis gundlaclii 3. Eleutherodactylus cramptoni 9. Anolis evermanni 5. Eleutherodactijlus hrittoni 10. Anolis l-rugi 5. Eleutherodactylus richmondi 11. Celestus pleii 6. Eleutherodactylus iciglitmanae 12. Amphisbaena bakeri Of these Eleutherodactylus cramptoni, E. unicolor and E. richmondi and E. locustus are confined, so far as known, to the peak of El Yunque ; the others are probably most abundant in the coffee belt. Since nearly two-thirds of the coastal-plain species overlap the coffee belt in dis- SCHMIDT, AMPHIBIANS OF PORTO RICO H tribution, it seems obvious that the distribution in altitude offers little basis for a faunal division. The changes due to cultivation, it may be assumed, have played an important part in the present distribution. The clearing of lowland forests, for example, has undoubtedly driven species to the coffee belt and to the residual forests, while the clearing of the hills has probably afforded access to the higher altitudes in the case of species originally confined to the more open spaces in the coastal plain. The coastal-plain fauna, however, is not a homogeneous one. AnoUs poncensis and Ameiva ivetmorei and possibly Alsophis antillensis are confined to the arid or semiarid southwestern part of the island, and Ehutherodactylus antillensis, Anolis cuvieri, TypJilops platycephalus and Typlilops rostellatus have not been recorded from that part of the island. Anolis poncensis and Ameiva wetmorei are two of the most peculiar and striking species in the entire fauna, the latter being more closely related to species in Hispaniola and St. Croix than to other Porto Eican forms. I propose, then, to divide Porto Eico faunally into a humid district, comprising the greater part of the island, characterized by the presence of Eleutherodaciylus antillensis, Anolis cuvieri and Typlilops rostellatus (besides the species of Eleutherodactylm confined to El Yunque) ; and an arid district, including the southwestern corner, characterized by the presence of Anolis poncensis and Ameiva wetmorei. Various cacti form the most characteristic element in the flora of the arid district (Plate I), while the humid district was probably originally a forested area (Plates II and III), bordered by open spaces along the coast. The contrast in habitat conditions between the arid area to the south- west and the dripping cloud forest of the Luquillo is extreme. The cloud forest affords ideal conditions for the tree frogs, and these are extraordinarily abundant in the moister belt above 1200 feet altitude. The amphibian chorus in the rain forest on El Yunque is the most extraordinary I have heard. As one stands at the Forester's Cabin, at about 1300 ft. altitude, a roar of sound comes from tlie wooded ravine adjoining, and from the slopes above, making a veritable Babel of frog notes. One by one the individual voices can be dissociated from the general confusion. Those of Leptodactylus alhilahris and Eleutherodac- tylus portoricensis, become separated first, since these are already familiar from the first night in Porto Eico. E. portoricensis here appears to have added several variations to its lowland notes, but in general its voice proves readily distinguishable. N'ext to these, the most insistent ele- ment in the chorus is a rapid click-clicking not unlike that of a tele- graphic instrument, with a very insect-like quality. This proves to be 12 SCIENTIFIC SURVEY OF PORTO RICO the note of the tiny Eleutherodactylus gnjllus, and it was undoubtedly this note which Stejneger ascribed to tbe young of E. [lortoricen.sis. This clicking note comes also from the lower branches of the trees, probably up to a height of twenty feet. A fourth note, carefully run down, proves to be that of a large, green, long-horned grasshopper, and to the sur- prise of the collector another succession of sounds, even more character- istically grasshopper-like, beginning with a shrill prolonged note and ending with a series of clicks, proves to issue from the distended throat of still another Eleutherodactylm. Directing the attention, now, as much as possible away from the known elements of the chorus, one may distinguish a strikinglv different element. A sad little series of wdiis- ties descending in the scale and becoming successively fainter proves to belong to a very distinct species of small Eleutherodactylus (E. wight- inanae), which sits on the ground or on the lower leaves of plants, and is certainly a most difficult species to discern, even when it is singing a foot away from the collector's ear. Another tiny species has a slow clicking note, — the sixth to be distinguished. There is still an undifferentiated chorus awaiting investigation, and three species of tree frogs {E. ricli- mondi, E. unicolor and E. cramptoni) are known from El Yunque, whose notes I did not trace. In the arid southwestern section there is no such wealth of amphib- ians a]ul, while this is obviously due to the lack of moisture and hence is primarily an ecological difference, the differentiation of very distinct species confined to this area bears witness to so long a history of similar relations between topography and moisture that here habitat conditions have dominated the faunal history. The fact that this section of Porto Rico appears to be intimately related to the island of St. Croix, figures in my argument below on the relations of the faunae. Origin and Relations of the Porto Rican Hertetological Fauna I. THE west INDIAN FAUNA The origin of the West Indian fauna, specifically of the Greater An- tillean fauna, has been a controversial topic among zooge.ographers for a generation. Arguing from herpetological evidence, Stejneger (1904) and Barbour (1910, 1914, 1916) have maintained that the fauna is de- rived from the mainland by migration over land connections, and An- thony (1918) supports the same view from the standpoint of mammal- ogy. Matthew (1915, 1919) has been the chief exponent of the alter- native theory that the Antilles have received their fauna through fortui- tous dispersal without such connection. HVHMIDT, AMPHIBIANS OF PORTO RICO 13 Anthony (1925), in summarizing the evidence from the mammalian fauna in an earlier volume of the present series, adopts a modified form of the "'land-connection'' hypothesis, and Matthew himself (1919), has agreed that the Greater Antillean islands may at some time have been united. The West Indian amphibians and reptiles appear to me to afford evidence supporting Anthony's conclusions, at least in a general way. Comparison between the distribution of amphibians and re])tiles and the distribution of mammals is made difficult by the much greater age of amphibian and reptile stocks. The arrival of the bulk of the West Indian reptile fauna may be contemporary with that of the earliest of the mammals, the insectivores, whose mammalian contemporaries are ex- tinct. Reptilian distribution frequently affords clues to pre-mammalian faunal history. Thus Madagascar and New Zealand may be allowed to be oceanic islands so far as their mammalian faunae are concerned, while their Pre-Tertiary contacts with continental faunae are reflected in their amphibians and reptiles. From a general review of the distribution of the reptiles I am con- vinced that they support the general theses of Matthew regarding the trend of dispersal from Holarctic centers and the want of evidence for Antarctic connections. I am equally convinced that reptilian dis- tribution fails to support some of his secondary theses, especially with regard to the oceanic nature of the faunae of Madagascar and the West Indies. It is embarrassing to be so thoroughly an eclectic zoogeographer, and one finds oneself exposed to the fire of both schools. My own general conclusions with regard to the West Indian fauna, based primarily on the herpetological evidence, are : 1. That the Greater Antilles received their fauna from Central Amer- ica at a time so early that the continental fauna has subsequently under- gone great changes, probably in Eocene or even in Pre-Tertiary time. 2. That the Greater Antillean fauna gives us a somewhat ol^scure representation of this earlier Central American fauna, most of which, in accordance with Matthew's general hypothesis, has moved on to South America. 3. That there has been a union of the larger islands during part of their existence, which has produced the uniformities in their faunae. 4. That the Lesser Antillean fauna is derived from South America, that it is a genuinely fortuitous one and that no land-bridge has existed through this chain in Tertiary time. By way of general review of the Greater Antillean herpetological^ fauna, I have drawn up a list of the genera in tabular form. '1 .■;'^ L» i, • ^ ^^ i5i ?l Y; 14 SCIENTIFIC SURVEY OP PORTO RICO List of Genera Amphibians 1. Bufo*i 2. Hijla*\ 3. Le-ptodactylus* ] 4. Eleutherodactylus *t 5. Sminthillus f Reptiles 1. Gonatodes *t 2. Sphaerodactylus *t 3. Hemidactylus f 4. Aristelliger * 5. Tarentola 6. Thecadadylus *t 7. Anolis *t 8. Norops *t 9. Deiroptyx 10. Chaniaeleolis 11. Chamaelinorops 12. Xiphocercus 13. Iguana 14. Cydura*] 15. Leiocephalus f 16. Hispaniolus 17. Celestus* 18. Sauresia 19. Wetmorena 20. Cricosaura 21. A-meiva*\ 22. Amphisbaena] 23. Cadea 24. Mabuya*^ 25. Typhlops\ 26. Tropidophis*^ 27. Epicratesf 28. Trelanorhinus* 29. Arrhyton 30. Alsophis\ 31. Drojnicus ?t 32. Uromacer 33. Hypsirynchus 34. laltris 35. Pseudemys* 36. Crocodylus*] Total species Total genera Endemic genera Non - endemic genera not found in other islands . . . Number of species native on the Greater Antillean Islands Cuba 5 1 16 1 1 5 1 1 1 25 1 1 1 1 5 1 1 1 1 2 1 4 1 2 3 3 2 1 2 91 29 5 Jamaica 4 8 1 6 1 1 6 1 1 2 1 1 1 1 1 2 1 1 40 18 1 His- paniola 1 4 1 9 1 5 2 1 13 2 3 8 1 3 1 1 8 3 1 2 2 3 3 5 5 1 1 1 1 92 29 7 Porto Rico 1 1 10 1 1 8 1 1 3 2 1 3 2 3 2 1 41 16 Virgin Islands 1 1 2 1 1 1 6 1 1 2 1 1 1 2 1 23 15 * Central American. t South American. SCHMIDT, AMPHIBIANS OF PORTO RICO 15 The number of species in this table is somewhat unsatisfactory for comparison on account of the inclusion of vicarious forms from out- lying islands — the Cayman Islands with Cuba; Tortuga, Gonaives, ISTavassa and Beata with Hispaniola, and Mona with Porto Rico. The Amphibians and Reptiles of the Greater Antilles represent 41 genera. Two of these, Igiuma and Thecaductylus enter the region only in the Virgin Islands, and are present in the Lesser Antilles, They are consequently an alien element in the fauna, the more so as they are not specifically differentiated; it is extremely likely that Iguana was intro- duced by the Indians in the course of their wanderings, while the gecko is probably fortuitous through non-human agencies, A third genus, Tarentola, is represented only in Cuba and is otherwise African, spe- cifically Mediterranean, in distribution. This still more alien form is well differentiated from its congeners and represents one of the most curious of genuinely discontinuous distributions, I suppose it to be an ancient "flotsam-jetsam" arrival. Of the remaining 37 genera, 14 are endemic; 11 are generally dis- tributed on the four larger islands, and 20 are represented on three or more of the islands. It is a curious fact that the endemic genera, with the exception of Cyclura, are confined to single islands, and thus do not contribute to the hypothesis of a former union. The 20 more widely distributed genera, however, all have vicariating forms from island to island, and a number of sections of genera, such as the giant Anoles, come near to being widely distributed endemic genera, like Cyclura. The endemic forms are chiefly minor end-stages or divergent branches which have arisen by local evolution, such as Chamaeleolis, Deiroptyx, Chamaelinorops, Xiphocercus, Hispaniolu^, Sauresia, Wetmorena and Arrhyton. A few, however, are plainly relict forms, notably the Xan- tusid Cricosaura, the Iguanid Cyclura, the Hispaniolan snakes Uromacer and laltris, the Brachycephalid frog Sminthillus. Five genera, Lepto- dactylus, Sminthillus, Norops, Leiocephalus and Tretanorhinus, are neither endemic nor widely distributed, and this is a very heterogenous list, with no appreciable parallelism in distribution. Eighteen genera occur both in the Greater Antilles and Central America, but 14 of these are likewise represented in South America, and these, with the 7 genera common to South America and the larger West Indian islands but absent in Central America, make the faunal relation with South America appreciably more intimate than with Central America. This very fact seems to me to accord best with the theory of the Central American origin of the fauna, on the supposi- -^Q SCIENTIFIC SURVEY OF PORTO RICO tion that the South American fauna is mainly of northern origin, as pointed out by Matthew in his general scheme of dispersal. The degree of differentiation ])etween the continental and West Indian representatives varies greatly, and at first glance appears to indicate varying ages of origin. Some of this variation, however, may be due to otheT factors than time of separation. Such an archaic-looking relict as Cricosaura, widely distinct from its continental allies, may perhaps represent about the same amount of evolution as has occurred in Anolis and its derived genera, the difference being the contrast between a de- clining group and an expanding one. The crocodiles, on the other hand, seem to belong to quite different invasions, C. rJiomhifer and moreletu being assignable to an earlier arrival, their ranges now entirely circum- scribed by that of the modern wide-ranging, semi-marine Crocodylus acuius, whose wide distribution evidently has little bearing on the prob- lem of land connections. In a more detailed discussion of the genera I shall try to show that the faunal picture presented accords with a derivation from Central America at an early date, on the hypothesis of a southward trend in the migrations of the world as a whole, and that it is direct faunal relations witli Central America, such as that of the Xantusiidae, which require explanation rather than the discontinuity in range of AmfliisUena or Leiocephalus. Of the genera of Amphibians, Bufo, Hyla, Eleutherodadylus and Leptodactylus have a wide Neotropical distribution. The anomalous nature of the distribution of Leptodactylus will be discussed below. Sminthillus has a single Cuban species, and two others, Peruvian and Brazilian, have since been described. The discovery of additional species in this genus (originally described as monotypic) contributes to the likelihood that it is a natural group. Among the reptiles, geckos are notable for erratic distributions, though when critically examined their rantjes are often found to be closely parallel to those of other groups. The Antillean geckos, however, are really heterogeneous in distribution. Tarentola and Thecadadylus have already been mentioned. Gonatodes is widely distributed in Central and South' America, apparently ranging into the Antilles from the west. Spkaerodactylus has a wide neotropical distribution, but its wealth of Antillean species distinguishes it as an autochthonous genus, and its development is very like the other characteristically West Indian forms, such as Amek)a, Dromicus or Eleutherodactylu-^. E emidactylus , with H. mahouia on all the islands, appears to be a house-gecko, and human SCHMIDT, AMrniBlAyS OF PORTO RICO 17 agency may well have played a part in its distribution. It is somewhat remarkable that the genus Hemidactylus is unknown in Central Amer- ica. I am not at all convinced that the African geckos commonly re- ferred to mahoiiia are con-specific with the Antillean form. The East African and Madagascan species does not seem to me to be identical even with tlie West African one ! Hemidactylas hrookii, on the other hand, in Hispaniola, would appear to l)e an African form introduced by the slave-trade. A rIstelUger is confined to Central America, Jamaica and Hispaniola. It is included in my list as Cuban because it reaches the Cayman Islands, whose faunal affinity is primarily Cuban. On the coast of Yucatan this s|)eeies is characteristic of the fringe of cays, and its occurrence in the West Indies offers no anomaly. The Igiianid genera include the monotypic and endemic Deiroptyx, ChamaeleoUs, Chamnelinorops and Xiphocercus, Iguana, already men- tioned, and AnoJis, Norops, Cyclura, Leiocephaliis and Hispmiiolus. Norops seems to be a more jirimitive form than Anolis, with three con- tinental species, and is a declining group in contrast with the expanding Anolis. The Cuban species is thus plainly a relict. Leiocephalus, with a numl)er of species in Cuba and Hispaniola, is otherwise best developed in western South America, and is absent in Central America. I regard this also as a relict distribution, but of a group that is holding its own. Cyclura. is even more interesting. The curious '"combs'' on its toes, though rather a trivial character, quite definitely ally its species more closely to the Galapagan Conolophus and Amhlyrhynckus (and the other Pacific genus as well, the Fijian Brachijlophus) than to tlie Central American Ctenosawa. Ctenosaura extends southward as a wedge sepa- rating these allied forms, and I have endeavored elsewhere* to show that the Ctenosaura have spread southward from the great Southwestern shield in Xorth America. Leiocephalus lends itself to this interpreta- tion if it be visualized as retreating before more advanced Iguanid genera, such as Sceloporus. Anolis, in the full flower of expansion, ob- scures distributional argument by its wealth of forms and closely-knit ranges. The only species of Anolis that is supposed to be common to Central America and Cuba is A sagrei, an inhabitant, like Aristelliger, of the off-shore cavs in the Bav of Honduras. The endemic genera re- quire no comment except that Xiphocercus is represented in Colombia by a related or parallel form. The Anguidae are represented in all four islands by Celestus. Two additional genera, Sauresia and ^yetlllore^}a, moiiotypie "end-stage" * 1922. Bull. Amer. Mns. X;it. Hist., XLVI. p. cn. j3 SCIENTIFIC SURVEY OF PORTO RICO forms, are confined to Hispaniola. Celestus also occurs in Central America and its close ally, Diploglossus, is found in both Central and South America. Celestus and Diploglossus are plainly primitive genera, and the modern representatives of the family, the plated lizards (Gerr- lionotus), have the same spatial relations with them as exist between Ctenosaura and Gyclura, or Sceloporus and Leiocephalus. The Xantusiidae are a declining group composed of the North Ameri- can genus Xantmia, the Central American Lepidophyma and the Cuban Cricosaura. This distribution is not in accord with the above-cited southward migrations, but this is a recurrent anomaly which requires a modification of the Matthewsian hypothesis of the dispersal of primi- tive forms. It must be recognized that evolution in the direction of habitat restriction may strictly parallel an evolution in which the primi- tive forms become peripheral by retreat in space. This is an obvious phenomenon among the Xaniusids, which inhabit areas adjacent to what I have regarded as the probable center of dispersal of American lizards, but are plainly relicts among more modern and progressive forms. The species of Xantusia are curiously restricted as to habitat — X, vigilis by its association with the Yucca, .Y. henshaivi by its rock-dwelling huhii — while both are doubtless nocturnal, as is Lepidophyma. The mainland Xantusids have retreated owing to habitat restriction, while the Cuban genus represents the other alternative, that of actual retreat, and appears as a true relict, though also rigidly confined to a single habitat. The case is directly comparable to that of the Central African lemurs, which escape their modern competitors by their nocturnal habits, while the Madagascan lemurs have survived through actual migration and the timely separation of their retreat. The Teiidae are represented only by Ameiva, though the West Indian species are divisible into two rather distinct sections. Ameivas are widely distributed on the South and Central American mainland, but the continental species are fewer than the West Indian. I suspect that the genus Cnemidophorus bears the same relation to Ameiva as Sceloporus does to Leiocephalus, namely, that it is a more modern group of species, with Ameiva more or less in retreat. The Amphisbaenidae are well represented in the Antilles, with both Cadea and Amphishaena in Cuba and Amphishaena extending out to the Virgin Islands. Except for Bipes, which is present in west ^Mexico, the family is wanting In Central America, and the Antilleau forms are thus relicts of a former type of dispersal. The evidence for the southward migration of the Amphisbaenians seems to me ample, even SCHMIDT, AMPHIBIANS OF PORTO RICO 19 without the direct evidence of the Oligocene fossil forms, Rhineura, confined to Florida, is quite obviously one of the many curious forms accumulated in the southeastern United States as a result of divergent migration from holarctic dispersal centers. The nearest relative of Cadea seems to be Venezuelan, while Ainphishaena itself is well repre- sented almost throughout South America. The only Scincoid genus is j\lahiii/a, generally distributed in the tropics of the world, but nowhere sj)eeiating in the Americas as its does in the Old World. Its range in both hemispheres is nearly exclusive of that of the more northern and obviously more recent Eumeces. Among snakes the Typhlopidae afford no especial evidence of faunal relation. The Antillean Typhlops hinihricalis was long supposed to be a widespread species occurring also in South America. Cochran (1924) and I (1920) have brought the distribution of the West Indian forms into harmony with that of other groups. The most notably primitive genus, Anoinalepis, is Central and South American (or at least Pana- manian and Peruvian), and not Antillean. The Boidae are represented by two genera. Epicrates has a species on each of the larger islands and has split into three species in Hispaniola, with a separate species on Mona Island and another distinct form in the Bahamas (confined to Turk's Islands). Tropidophis fails to reach Porto Eico, and its principal radiation occurs in Cuba. Epicrates is M'anting in northern Central America, but it reappears in South Amer- ica. Tropidophis is said to have both South and Central American allies, but they are little known. The relationship of the Colubrine genera are vague, but their nearest allies seem to be South American, with the exception of Tretnorhinus, which is found in Cuba and Central America. Pseudemys, the single genus of fresh-water turtles, is quite as easily derivable from the Central American representatives as from the Florid- ian, and the existence of insular differentiation, which I am able to show for the Porto Eican specimens, makes it unnecessary to regard Pseudemys as a strictly recent arrival. The absence of other fresh-water turtles is highly remarkable, in view of the ancient character and great diversity of the American turtle fauna. It is no less anomalous to find in Cuba a fossil Testudo related to the Galapagan species, though its presence adds to the faunal relations between the Antilles and the Galapagos. Crocodylus, finall}^, adds a distinctively Central American form to the West Indian fauna. The broad-snouted Cuban Crocodylus rhonihifer is directly allied to C. moreletii of the adjacent parts of Mexico and the Yucatan peninsula. The wide-ranging, undifferentiated C. acutus 20 SCIENTIFIC SURVEY OF PORTO RICO floods over the ranges of these earlier forms. Crocodiles do not range beyond the Orinoco basin in South America, and evidently are more recent arrivals than the caimans or the alligators. In support of my proposition (2) above, I have contrasted the distri- bution of such a group as Sceloporus, an essentially modern genus, with that of a more ancient Iguanid genus, Leiocephalus. Sceloporus is essentially Sonoran, with a wealth of ISTorth American species, and a broad overflow into Central America. Leiocephalus is West Indian and South American. Allowing for discrepancies and irregularities such as tliose I have discussed for the Xantusiida?, the list of such pairs of genera is impressive : Ancient, West Indian Modern, Sonoran Leiocephnhis Sceloporus Cijclum Ctenosaura Celestus Gerrhonotus A meiva Cnemidophorus Mahuya Eumeces Cnemidophorus, among the genera listed as Sonoran, ranges widely into South America. Otherwise its development is so closely similar to that of the other Sonoran genera that I am disposed to search for an explanation of this anomaly rather than remove it from the Modern, Sonoran list. I am fully convinced that the fauna of the Greater Antilles reached these islands from Central America, and that the majority of the endemic forms represent a nearly contemporary faunal invasion. That an actual landbridge existed over wliich the migration took place, is my somewhat more hesitant l^elief . The existence of mammals and amphibians, even as a depauperate fauna, is evidence in favor of continental connection. The amphibian and reptile fauna exhibits a relatively greater diversity than does the maunnalian. The sixteen families represented are: Amphibians Bufoniilae Hylidae Brachycephalidae Eeptiles Gekkonidae Amphisliaenidae Boigidae Iguanidae Scincidae Emydidae Anguidae Typhlopidae Crocodilidae Xantusiidae Boidae Teiidae Colubridae SCHMIDT, AMPHIBIANS OF PORTO RICO 21 The Central x^merican launa has thirty-two families, iuit several of these are isolated groups which could scarcely be expected in the West Indies, — the Helodermatidae and Xenosauridae, for example. Others are obviously more recent arrivals and for this reason their presence is not to be expected; instances in point are the Ranidae, Plethodontidae and Cro- talidae. The disproportion between the continental and Antillean fauna? in number of families is accordingly much reduced, perhaps about 26:16. It is a striking and important fact that the South American fauna is actually poorer in families of amphibians and reptiles than the Central American 1)y four or five. If the Chelydridse, Crocodilida^ and IMetlio- dontida;>. which are essentially Central American and only enter South America at the northwest, are also excluded, the genuinely South American families number only twenty-five. If the date of the supposed continental connection of tlie W-est -Indies be placed at the close of the Mesozoic. the relative wealth of am])hibians and reptiles and the poverty in mammals are completely ex})lained. T'n- fortunately, a connection so early in geological histdvy does not account for the more recent members of the mammalian fauna, for which a Mio- cene date of arrival is indicated. The two families of insectivores agree with the reptiles as to^'early date of entry, while the remaining mammals appear to represent at least two later immigrations. One is tempted to suppose a very early continental connection for ampliibians and i*e|)tiles, insectivores, etc., and to recognize Mattliew's argument that the remaining mammals are accidental. Geological conclusions based on zoogeographic evidence so fragmentary and contradictory are evidently of little real value. One set of conclusions, however, from a general consideration of the fauna, seems well founded. This is my proposition (3), that the larger islands were connected at an early stage in the development of their fauna : that they have subsequently been separated, more probably by block-faulting than by any great change of level, that Porto Pico and the Virgin Islands were the last to be cut off, and that tlie Virgin Islands were connected with Porto Rico as recently as the Pleistocene.' ' The evidence of long isolation of the three western islands is plainly to be seen in the independent radiation which has taken place in the •elements of their fauna\ Bufo has evolved 5 species in Culja, while Jamaica with no native toad, and Hispaniola with'a single Bufo, have each produced 4 species of Hyla, independently, in tlie opinion of Dunn, who has lately examined the Jamaican species in detail. Dunn's conclu- sions are somewhat hesitatingly accepted by Xoble (1027). W}>ether or 22 SCIENTIFIC SURVEY OF PORTO RICO not the radiation in Hyla has been independent, the fact that both Jamaica and Cuba have only a single species of Ameiva while Hispaniola has 8, and 4 on the main island alone, certainly illustrates an inde- pendent evolution. Long separation is equally evident in the wealth of Cuban Anolis Avith no less than 3 related genera represented, Norops, Chamceholis and Deiroptyx, equalling the number of species on the other three islands together. The Jamaican fauna, poor in some genera, has no less than 6 well-established species of Sphcerodactylus. surely a sharp contrast with the 2 in all Central America ! Among the snake genera, Tropidophis has developed 4 species in Cuba, while Epicrates has 3 in Hispaniola, and an extra species on Mona. The Hispaniolan Dromicus, Alsophis and Uromacer fall in line with the other genera. The total impression of the herpetological fauna is plainly one of a fundamental unity, obscured only by the long evolu- tion during subsequent isolation. The fauna of the Lesser Antilles has been effectively contrasted by Anthony with that of the larger islands. His argument from the mam- mals that the animal population of the Lesser Antilles is fortuitous from South America and of relatively recent origin agrees exactly with my impression based upon the herpetological fauna. The fauna of Trinidad itself is far from rich in comparison with that of the mainland. Its reptiles and amphibians amount to about 80 species — almost all of them specifically identical with those of northern South America. This South Amjerican fauna disappears rather in proportion to distance from the mainland than in relation to size of landmass, for Tobago has 24 species with little endemism, Grenada 17 with about 4 endemic forms, St. Vincent 10 with 6 endemic. In the next four islands the fauna ranges only from 10 to 14 species, with 5 to 10 endemic forms. En- demic forms in the whole chain are very slightly differentiated from their very obvious relatives. The species may be grouped as mainland forms, with a haphazard distribution on the islands, endemic species of main- land genera, slightly differentiated, and endemic species of genera which range throughout the chain with vicariating forms from island to island. There is little or no "radiation," which is so marked a characteristic of the Greater Antilles. Examples of the haphazard distribution are afforded by Leptodactylus and Iguana, probably transported by the Indians as food animals, and by the snakes in general, though the faunae in question may be imperfectly known, Leptotyphlops hilineatus, for example, occurring on Barbados and St. Lucia, Cloelia clelia on Grenada, St. Lucia and Dominica, although not recorded from the intervening SCHMIDT, AMPHIBIANS OF PORTO RIVO ^3 islands, St. Vincent and Martinique. The fer-de-lance skips Tobago, Grenada and St. Vincent, to appear only on St. Lucia and Martinique. The northern group of twelve small islands, from Anguilla to Mont- serrat, has a fauna impoverished in genera, but rich in endemic species. Leptodacttjhts pentadaciylus and Iguana iguana appear to represent Pre- Columbian introduction by natives. Eleutherodactylus martinicensis is recorded from five islands. Its status requires re-investigation. Ty- phlops is known from St. Kitts and Antigua, the species doubtless unde- scribed. LepfotypJilops albifrons is reported only from Antigua. Its wide range and apparently haphazard arrangement seems to indicate a facility for fortuitous dispersal. Thecadactylus rapicaudus, widespread in the Lesser Antilles, is recorded from five of the northern islands. These irregular distributions contrast strongly with 3 endemic species of Sphaerodactylus on three islands, 7 Anolis on nine islands, and 8 Amei- va on ten islands. This portion of the fauna, which perhaps should in- clude Also phis,, appears to represent an older nucleus, and I interpret its relations as representative of the uniformity of a typically oceanic fauna plus endemism induced by a considerable lapse of time. The most obvious differences theoretically to be expected between continental and oceanic insular faunae are (1) heterogeneity and (2) impoverishment in the oceanic islands. The presence of relict forms seems to me to be strong evidence of a land connection remote in time. Impoverishment may obviously occur in a continental fauna by extinc- tion ; and "fortuitous dispersal" may act as a screen allowing only cer- tain forms to pass, so that extreme uniformity of fauna, instead of heterogeneity, may be a result of truly oceanic dispersal, as is to be seen in Polynesia, where island after island is inhabited by the same five species of lizards. Such uniformity is complicated by the age of the islands. It is thus curiously difficult to frame criteria whereby an in- sular fauna derived from land connection may be distinguished from one produced by "fortuitous dispersal." II. THE PORTO RICAN AND VIRGIN ISLAND FAUNA Turning to the more detailed consideration of the Porto Pican fauna, it is interesting to note that important advances have been made in our knowledge of the amphibians and reptiles of this area since Stejneger's discussion of their origin and relations in 1904. The herpetological dis- coveries bearing directly on this problem have been (1) the description of Ameiva wetmorei, (2) the finding of Bufo and Cyclura on the outer Virgin Islands, (3) the discovery of fossil remains of Cyclura on St. 24 SCIENTIFIC SURVEY OF PORTO RICO Thomas aud Porto Eico, (4) additions to the fauna of Hispauiola, especially the discovery of Leptodactylus dominicensis, (5) additions to the fauna of Porto Pico, especialh' Alsophis antillensis and Eleutliero- dactylus antillensis. which ally it more intimately to the fauna of the Virgin Islands, and (6) the elucidation of the relations of the Greater Antillean Typlilops. All of this new information has tended to empha- size the essentially Greater Antillean character of the Porto Pican fauna. Stejneger divides the Porto Pican herpetological fauna into South American and Central American elements, including in the former the genera Ameira. Amphishaena, Typlilops, Alsophis and Dromicus. These genera are all represented in Hispaniola, and tlieir immediate presence in Porto Pico is amply accounted for by a union with Hispaniola. I have endeavored to show above that tliis apparent relation of the West Indian fauna with the South American is best explained by a land connection with Central America, when the time relations and larger outlines of faunal migration are considered. Yaughan (1919, Bull. U. S. Xation. Mus., Xo. 103, pp. 547-612) has given an excellent resume of the geological history of the West Indian area as far as known. In advocating the existence of former land con- nections with South and Central America, his paper cuts incisively into the more speculative maze of zoogeographic controversy. The only flaw is the fact that he appears to base his conclusions in ])art on zoogeo- graphical data (p. 610), whereas I should like to accei)t them as a basis for zoogeography. For the present purpose however, — an examination of the immediate origin of the Porto Pican reptile and amphibian fauna, — the outline of tlie geologic history advanced by Vaughan is highly satisfactory, and it is possible to crystalize conclusions on the relations between Porto Pico and the Virgin Islands with each other and with Hispaniola into a more definite statement than has hitherto been possible. According to Vaughan's physiographic history of the area in question, the Greater Antilles were Joined to one another in late Miocene time, the resulting landmass including Porto Pico and the Virgin Islands as its easternmost extension. Tlie scanty zoogeographic ties between the Vir- gin Islands and the Lesser Antilles exclude the presence of a contempo- raneous land bridge to South America, or at least the continuation of any such bridge for a time commensurable with that of the union of the Greater Antilles. During the period of this uplift, the genera of rep- tiles and amphibians which may properly be regarded as "Greater Antillean" (through presence on three or more of the larger islands) acquired their distribution. SCHMIDT. A.UI'HnH.WS OF PORTO RICO 25 In order to connect the A'irgin Islands with Porto IJico, no great degree of emergence is necessary, as they are separated ])y water not ex- ceeding twenty fathoms in depth. St. Croix is included in the same way, with no greater amount of emergence, because its present separation by greater depths of water is believed to l)e due to faulting. The water between Porto Eico and Santo Domingo is much deeper (reaching 318 fathoms), but still shallow in comparison with the depths to the north and south. Even for this connection, however, the amount of emergence necessary is no greater, for there is important evidence of faulting, as I have shown elsewhere (1926). The earthquakes of October, 1918, which caused great damage to the cities of Mayagiiez and Aguadilla. on the west coast of Porto Rico, were probably caused by adjustments in this faulted area. The sharp truncation of the eastern end of Santo Domingo doubtless represents another fault line. Point Jiguero and Desecheo Island appear to represent the older period of mountain-mak- ing of the general Hispaniolan-Porto Eican axis, (i. e., an Eocene or Lower Oligocene connection) while Mona Island, almost exactly half way between the southwest corner of Porto Eico and Saona Island, and topographically almost exactly similar to Saona, may be a remnant of the Upper Miocene (or later?) land bridge itself. The rapid under- cutting of the north and east sides of Mona now in progress indicates a considerable recent reduction of its area. The date of the faulting which separated Porto Eico from Santo Domingo is placed in the Pliocene by Yaughan (p. 611)), and the separation of St. Croix from Porto Eico probably took place during the same period, but perhaps at a later date. It is likely that Porto Eico and the remaining Virgin Islands were separated by a submergence in the Pliocene, but they were reunited in the Pleistocene, perhaps by the withdrawal of the water for the conti- nental ice sheets, to form a "Greater Porto Eico" to which the common fauna of the islands now separated corresponds. The present configura- tion of Porto Eico and the Virgin Islands is (geologically) extremely recent. The very evident peneplanation of the greater part of the mountain area at a height between 1500 and 2000 feet appears to date at least from the early Tertiary, and implies long-continued existence as a land area. The interpretation of the existing faunal relations of Porto Eico, in the light of the geological hypothesis, becomes relatively simple. The fauna of the Virgin Islands stands in the same relation to the Porto Eican as does that of Porto Eico to the Hispaniolan. The degree of dif- ference in each case corresponds to tlie relative length of time since theif 36 tSCIENTIFIC SURVEY OF PORTO RICO respective separations, aud the degree of impoverishment to their rela- tive areas. The fauna of the Virgin Islands consists of 22 species, of which 2, Iguana iguana rliinolopha and Thecadactylus rapicaudus, belong to genera foreign to Porto Eico. The iguana was probably introduced by man; the gecko probably is a fortuitous arrival. Of the remaining 20 species, 11 are identical with Porto Eican forms: 1. Leptodactylus alhilabris 7. Anolis stratulus 2. Eleutherodactylus antillensis 8. Anolis pulcheUus 3. Sphaerodactylus macrolepis 9. Ameiva exsul 4. Hemidactylus mabouia 10, Mabuya sloanii 0. Anolis cuvieri 11. Alsophis antillensis 6. Anolis cristatellus The remaining 9 species are directly related to Porto Eican species and are for the most part simply vicariating forms : Virgin Islands Porto Rico 1. Bnfo iua-pis Bufo lemur 2. Eleutherodactylus lentus Eleutherodactylus richmondi 3. Anolis acutus Anolis poncensis ( ?) 4. Cyclura pinguis \Cyclura portoricensis (?) 5. Ameiva polops Ameiva wetmorei 6. Amphishaena fenestrata Amphishaena caeca 7. Typhlops richardii • Typhlops platycephalus 8. Dromicus exiguus Droniicus stahli 9. Alsophis sancti-crucis Alsophis portoricensis Examined more in detail the chief questions which require discussion are: (1) the impoverishment of the Virgin Island fauna, in which many species of Porto Eican Eleutherodactylus and Anolis are unrepresented, while Celestus, Epicrates and Pseudemys are entirely wanting; (2) the apparently haphazard distribution of Bufo, Anolis cuvieri and Cyclura; (3) the position of St. Croix in relation to the other islands and Porto Eico, and (4) the origin of the species common to several of the Virgin Islands but absent in Porto Eico. The absence of forms may be explained as original or secondary. The discovery of the remains of recently extinct Cyclura in both Porto Eico and St. Thomas, coupled with the presence of living Cyclura on Mona and Anegada, obviously indicates that in this genus a process of extinc- tion is taking place. The same factor probably accounts for the isolated SCHMIDT, AMPHIBIANS OF PORTO RICO 27 occurrence of a Bufo on Virgin Gorda and of AnoUs cuvieri on Tortola. The absence of Celestus and of Anolis gundlachi and Anolis hrugi, may indicate, on the other hand, that these forms never reached the Virgin Islands. There is no reason to believe that, if the whole land area were elevated 150 or 200 feet and so reunited, the species which now avoid the coastal plain in Porto Eico would be able to reach the Virgin Islands. The fact that of the entire Virgin Island fauna only Eleutherodactylus lentils is related to "coffee belt" species in Porto Eico indicates that this factor has probably operated as an important one in the past. The mere fluctuation in size of these islands has an important influ- ence on the rain fall and humidity (and evaporation), i. e., the most im- portant climatic factors affecting the fauna. The complete submer- gence of an islet would not be necessary to exterminate the greater part of its fauna, and it is a differential extermination of this nature which I believe to be the chief cause of the impoverishment of the Virgin Island fauna, and possibly of the West Indian fauna in general. St. Croix presents something of a problem. The Amphishaena fenes- trata from that island should be compared again with specimens from St. Thomas, and with A. caeca. Anolis acutus, Ameiva polops and Alsophis sancti-crucis are decidedly less closely allied to Porto Eican species than are the species from St. Thomas and even the outermost of the northern Virgin Islands. On the other hand, Ameiva polops indicates a relation- ship with the arid district of Porto Eico. If the "Greater Porto Eico" at any time included St. Croix, that area must have belonged to the ex- tended arid district, which influenced distribution in the "Greater Porto Eico" exactly as it does in the present. The separation of St. Croix in Pliocene time by faulting (as suggested by Vaughan) doubtless excluded it from union with Porto Eico in the Pleistocene, while a Pleistocene (Glacial period) connection of the other Virgin Islands with Porto Eico seems highly probable. Three species — Eleutherodactylus lentus, Amphishaena fenestrata and Dromicus exiguus — are common to two or more of the Virgin Islands and are absent from Porto Eico. Their development may be explained as due to a differentiation of the fauna of the lower-lying eastern end of the "Greater Porto Eico," or to differentiation during the hypotheti- cal Pliocene separation. The Porto Eican herpetological fauna differs from that of Hispaniola chiefly in the absence of the following genera.* * Oedipus is excluded from the Hispaniolan fauna pending verification of its occur- rence. Dunn regards the Haitian origin of Peter's Oedipus infuseatus as mythical. 28 SCIEM'IFIC .SURVh:y OF /'ONTO RICO 1. Hyla 8. Sauresia 2. Gonatodes 9. Tropidophis 3. Aristelliger 10. Uromacer 4. Leiocephalus 11. Hypsirhynchus 0. Hispaniolus 12. laltris 0. Chamaelinorops 13. Crocodylus Seven of these, Chamaelinorops, Hispaniolus, Wetmorena, Sauresia, Uromacer, Hypsirhynchus and laltris are confined to Hispaniola, while the other six are found in Cuba and Jamaica and for the most part in Central America. An extensive impoverishment of the fauna of Porto Rico is obviously its most conspicuous characteristic. Eecent extinction may well be admitted as a considerable factor in this impoverishment, in view of the discovery of the remains of an extinct Porto Eican Cyclura as well as by analogy with the extinction of the mammalian fauna. This may be due to two factors, the restriction of habitat formations due to increased cultivation, and the changes in climate due to past emergence aud submergence. On the other hand, the much greater altitudes of the mountains of Hispaniola, and the great diversity of habitat conditions of that island, of which perhaps the most remarkable is the stratification of the vegetation on the mountains, makes it highly probable that a number of forms have developed in situ, and had not acquired a suf- ficient range before the separation of Porto Rico to reach it, even if the habitat conditions of the intervening area were not unfavorable. If the late Miocene uplift was not extensive, and if Mona Island is a remnant of the gctual land connection via Saona and southwest Porto Rico, the habitat conditions of the land-bridge must have been such as to prevent the spread of many forms. It is more difficult to explain the differences in the development of such genera as Sphaerodactylus, Celestas, Arneiva and Epicrates, which have several species on Hispaniola and only one on Porto Rico. It is possible that Hispaniola was broken up into several islands dur- ing the Miocene, as is indicated by the Miocene deposits which compose the plain between the Central Sierra and the Monte Cristi Range, and Ijy the "through valley" of the saline lakes to the southwest. Only two species are common to Porto Rico and Hispaniola, one of which, Heniidactylus mahouia, is a house-gecko and plainly fortuitous, while the other, Mahuya sloanii, requires critical study. I have seen no Hispaniolan specimens. The numl)er of species which are closely related on the two islands is large : SCHMIDT, AMl'HWTAKS OF PORTO RICO 29 Porto Kico Hispaniola 1. Biifo lemur Biifo gutturosus 2. Leptodadylus alhUahris Leptodactylus dominicensis 3. Eleutherodactylus portoricensis Eleutherodactylus auriculatoides 4. Eleutherodactylus richmondi Eleutherodactylus iveinlandi 5. Sphaerodactylus 7nacrolepis Sphaerodactylus dijficilis 0, Anolis cuvieri Anolis ricordu 7. Anolis cristatellus Anolis cybotes 8. Anolis pidchellus Anolis semilineatus 9. fCyclura portoricensis Cyclura cornuta 10. Celestu-s pleii Celestus sp. 11. Ameiva exsul Ameiva vittipunctata 12. Ameiva icetmorei Ameiva Uneolata 13. Amphishaena caeca Amphisbaena ireinlandi 14. Typhlops platycephalus Typhlops sp. 15. Epicrates inornatus Epicrates striatus 16. Dromicus stahli Dromicus parvifrons 17. Alsophis portoricensis Alsophi-s melanichnus 18. Pseudemys stejnegeri Pseudemys palustris The Mona Island species, especially Cyclura stejnegeri and Epicrates monensis, add important links to this relation. The remaining- Porto Riean species, mostly Eleutherodactyhix and Anolis, which are clearly more closely related to other forms in the Greater Antilles than to South American or Lesser Antillean species, may be regarded as the individual development of the Greater Antillean fauna on Porto Eico, whose mountains occupied a relatively isolated position during any land connections that may have existed, certainly since the early Tertiary. The general conclusion is that the herpetological fauna of the "Greater Porto liico"' is simply an impoverished Greater Antillean fauna. Its resemblances to the fauna of Hispaniola are due to land connection, the date of which is placed by geologists in the Upper Miocene. The ditfer- ences between the Porto Rican and PTispaniolan faunas are due : ( 1 ) to a process of extinction still continuing: (2) to the isolated position of Porto Rico at the eastern end of the land mass, the habitat conditions of the supposed land-bridge being unsuited to the spread of many forms ; (3) to the differentiation of specifically Porto Rican forms, (a) through- out the Tertiary, the mountains of Porto Rico being a center of differ- entiation for autochthonous forms, as I suppose those of Hispaniola to have been, and (b) during post-Pliocene time, since the separation of Porto Rico from Hispaniola. 30 SCIENTIFIC SURVEY OF PORTO RICO SYSTEMATIC ACCOUNT OF THE SPECIES Class AMPHIBIA Order SALIENTIA Family Bufonidae^ Key to the Genera of Porto Ricax Frogs and Toads A. No teeth ; a large parotoid gland on each side of the neck ; skin very rough and warty ^^f^ Fig. 1— Head of toad (left) contrasted with Leptodactiilus (right). (From Stejneger.) KiG. 2. — Foot of Leptodactyhis (left) con- trasted with foot of EleutherodactyliiK (right). Compare slender and expanded lips of digits. (From Stejneger.) A A. Maxillary and vomerine teeth present; no parotoid gland; skin relalively smooth. iThe Bufonidae in the broad sense of Noble's revision of the families of Salientia (Noble, 1922. p. 22) includes the family Leptodactylidae, which, defined in contrast with the Bufonidae in the restricted sense of older authors, will be found in Stejneger. 1904, p. 5G9. SCHMIDT, AMPHIBIANS OF PORTO RICO 31 B. Tips of fingers not dilated, tapering Leptodactylus BB. Tips of fingers dilated into adhesive disks Eleutherodactylus Bufo Ijaureuti Key to the Porto Rican Species or True Toads A. Head with prominent bony crests; parotoid gland small and rounded— Bufo lemur AA. Head without long bony crests ; parotoid glands enormous, subtriangular— Bufo marinus * Bufo lemur (Cope) Sapo Concho Text Figs. 3-4. Peltaphnjnc lemur Cope, 1868, Proe. Acad. Nat. Sci. Phila., p. 311. Bufo lemur Stejneger. 1004, Rept. U. S. Nat. Mus.. 1902, p. 570, Figs. 1-5.— Barbour, 1914. Mem. Mus. Comp. Zool., Vol. XLIV, p. 242: 1917, Proc. Biol. Soc. Wash., Vol. XXX, p. 102.— Schmidt, 1920, Ann. N. Y. Acad. Sci., Vol. XXVIII. p. 108. Bufo (Peltaphryne) gutturosus Peters, 1876, Mon. Ber. Akad. Wiss. Berlin, p. 709. Bufo gutturosus Gundlach, 1881, Anales Soc. Espafi. Hist. Nat., Vol. X, p. 314.— Stahl, 1882, Fauma Puerto-Rico, p. 71, p. 161.— Garman, 1887, Bull. Essex Inst., Vol. XIX, p. 16 (part). The native name, Sapo concho, referring to the ridged head of the adults of this species, is an appropriate one. Unfortunately, due to the rarity of the form, it is practically unkno'WTi to the majority of Porto Ricans and the native boys apply the name indifferently to large speci- mens of the frog-like Leptodactylus or even of the tree frog. Type locality. — Porto Eico. Distribution. — Previous to 1919, this species was known from the north side of Porto Eico, the only exact localities given being Arecibo, Bayamon and Vega Baja. Its occurrence at Coamo Springs, nearly at the opposite side of the island, proves that it is widely distributed. Its rarity is perhaps due to the mongoose. The close relations of Bufo turpis Barbour, from Virgin Gorda (British Virgin Islands), with Bufo lemur constitute important evidence for the close faunal affinity of the Virgin group with Porto Eico. The relations with the Hispaniolan Bufo gutturosus are evidently close, and the three species gutturosus, lenur and turpis are much more closely allied among themselves than any one of them is to Bufo peltacephalus of Cuba, although peltacephalus must be considered their Cuban representative. Specimens collected. — 5 : Coamo Springs. 32 SCIENTIFIC SURVEY OF PORTO RICO Diagnosis. — A true toad with rough skin, rounded parotoid glands on the shoulders, and high bony ridges on the liead. Original descnption. — "This is a toad of singular aspect, owing to the extraordinary development of the bony crests of the cranium and the large size of its eyes. "The muzzle is short and very much flattened, projecting much beyond the mouth. The upper lip forms indeed a strongly projecting bony rim all round the mouth. Loreal region very concave, canthus concave and verv close together. The superciliary crests are extraordinarily elevated, having an arched outline, and descending steeply to the loreal region. It is angulate posteriorly, joining the almost equally developed supra- tympanic ridge. The crown of the head is thus a deep basin, widened above the tympana, and obstructed hy a cross-elevation in front. Strong ridffes behind and before the orbit ; nostrils vertical, a short bony longi- tudinal ridge below them. According to the characters of the genus Fig. 3.-^IIead of Bufo hnnir. (From Stejueger.) there is no derm on the head. Tympanum vertically oval. Parotoids broad oval, directed obliquely downwards, covered like the remainder of the upper surfaces of the body and liml)S, with numerous closely 1 su1)round tubercles, with rugose surfaces. Feet rather short, wit .nail tubercles, and only one remarkably weak metatarsal tubercle, the inner. A strong corneus ridge on the inner margin of the tarsus. The heel reaches the middle of the parotoid. The toes are about half weighed, and have a strong dermal margin. Two strong carpal tubercles. Under surfaces studded with small tubercles, with acute points. Tongue ob- ovate, largely free. "The color above is a blackish-brown, the top of the head yellow shaded; two longitudinal brown spots on the frontal region. A light vertebral line disappears on the back and reappears on the coccyx, and another light line passes round the inside of the parotoids and diverges on the scapular region. Limbs yellowish cross-banded, below dirty white, below the vent blackish. "This curious animal was found by George Latimer, the correspondent of the Smithsonian Institution in Porto Pico, W. I." The bony crests of the head of this toad distinguish it from the intro- SCHMIDT, AMPHIBIANS OF PORTO RICO 33 duced marine toad at a glance, and indeed from all of the remaining tail- less amphibians of Porto Eico. Coloration in life* — '"Iris pale brassy, sprinkled with black. General color above dull clay-colored Avith a strong olive wash; blackish brown markings and an ill-defined hourglass-shaped mark between shoulders; also a larger blackish spot on each side of the coccyx, which is marked by a pale streak; indications of blackish cross bands on legs; underside dirty white, becoming tlesh-colored behind and strongly reddish flesh color on underside of femur and nearest portion of belly; tips of toes Fig. 4. — Habitus of juvenile Btifo lemur (A), with side view of head (B). A. M. N. H. No. 10151. Natural size. dark brown ; tips of warts on back black, those between shoulders partic- ularly large. "Another specimen was colored as follows : Upper side olive, strongly suffused with 'gallstone yellow,' which is particularly noticeable over the insertion of the fore limbs : very few traces of dusky markings, but the pustules are.< black, especially anteriorly; an intensely orchraceous- rufous spot on the middle of the back ; on the underside the yellow suf- fusion invades the white ground-color on the portion nearest to the flanks. "The third large specimen was quite similar to the last, though with- out any rufous spot on the back, which seems to be an anomaly. Whole upper surface darker olive, and flanks, including space at base of fore * Quoted from Sejoeger. 1004. p. .">72. 34 SCIENTIFIC SURVEY OF PORTO RICO limb and below the ear, more intensely and more well-defined yellow; underside dirty yellowish white. "A young specimen was colored as follows : General color al)o\ e drab, more Isabella-colored on head ; dark markings blackish, those on shoul- ders pale-edged externally; flanks with a purplish suffusion and indica- tions of a broad longitudinal band, well-defined and pale-edged above, but gradually fading below into the pale Isabella color of the belly; under- side with a network of coarse dark-gray mottlings and marblings." RenwA-s. — The five half-grown specimens (collected by myself) are so nearly uniform and were found in so circumscribed an area that they probably are members of a single brood. They agree in coloration with the juvenile specimen described by Stejneger. All show the dark mark of hour-glass shape on the shoulders. The dimensions of one of these specimens may be compared with those of an adult recorded by Stej- neger. A. M. N. H. U. S. N. M. Parts measured No. 10151 No. 27148 Tip of snout to vent 3''' ^^- ^^ ™™- Tip of snout to posterior edge of tympanum 12 12 Greatest width of head 13 32 Foreleg from axilla 21 51 Hind leg from vent 37 99 These five specimens were found under limestone boulders on the artificial terrace in the moist gorge immediately behind the bath houses of the Coamo Sprijigs Hotel. When exposed, they relied on immobility and their resemblance to the soil for protection, and they were, in fact, extremely difficult to see. ' The first one was discovered accidentally while I was catching a Leptodactylus under the stone beneath which it was secreted. A large boulder sheltered three toads and one Leptodactylus. The stomach contents of these specimens included ant remains, beetle wings, an insect larva and segments of a small millipede. Nothing is knovm of the breeding habits of this species. Bufo nmrinus (Linne) Marine Toad Fig. 5 Rana marina Linne, 1758, Syst, Nat., I, p. 211. Bufo marinus Schneider. 1799, Hi.st. Amphib., Fasc. I. p. 219.— Danforth, 1925, Copeia, No. 147, p. 77. Type locality. — America. SCHMIDT, AMPHIBIANS OF PORTO RICO 35 Distribution. — Central America, northern South America. Xative in Trinidad, introduced in several of the Lesser Antillean islands and in Porto Eico and Jamaica. Fig. 5. — Dufo 77tarinus ?. Mayagiiez. (Danforth Collection.) During my stay in Porto Eico in 1919 I heard reports of the introduc- tion of this species, but I did not see any specimens. Danforth's record 36 .WIENTIFIC SURVEY OF PORTO RICO from Mayagiiez seems to be the only reference in herpetological litera- ture to the occurrence of the species on Porto Rico. With certain reservations as to the need for revisionary studies, Bufo marimis has an extremely wide range, extending from southern Mexico to southern Brazil and the northern Argentine. Dmgnosis. — A large toad with unusually large parotoid glands; bony ridges of head well developed but low, not higher than the eyelid ; no horizontal labial ridge. Origmal description. — "A frog with swollen shoulder glands and warty hinder parts ; hands with four digits, completely divided ; feet with five digits, partly divided." r Descripti-on of a Porto Rican specini&n. — (S. T. Danforth collection, $ , Mayagiiez.) Top of head bony, with well developed crests, the height of the supraorbital crest about 2 mm. above the interorbital space, not higher than the eyelid ; orbital crests continued toward the snout by well-defined ridges on the canthi; an anteorbital and postorbital ridge from the supraorbital ridge about half way to the labial border; no true labial ridge; a short supratympanic branch connects the supraorbital ridge with the parotoid gland ; no parietal branches from the orbital crests ; snout prominent, lores well-defined by the pre-orbital and canthal ridges ; eyes prominent ; tympanum large, subcircular ; parotoid large, distinctly outlined, its lower edge in line with the labial border ; a postrictal gland ; first finger longer than second, second as long as the fourth; toes about one-third webbed ; subarticular tubercles single : metatarsal tubercles small, rounded ; a sharp ridge on the tarsus continuous with the outer web of the inner toe ; no tibial gland ; back and sides densely covered with warts of varying size; the warts, the parotoid glands, and even the skin between the warts are densely covered with small black spines ; ven- tral surfaces nearly smooth ; upper sides of first and second fingers cov- ered with nuptial asperities forming a fine rugose pad. Brownish gray above; venter gray with a faint tinge of yellow, with numerous round darker spots ; throat dark gray. A female specimen from the same locality has larger and fewer warts, which are entirely smooth, as are the deeply pitted parotoid glands ; cranial crests somewhat lower ; ventral surfaces finely granular, the granulation finely spinose ; parotoid gland more sharply outlined ; fore- limb much more slender than in the male. SVHMIDT, AMPHIBIANS OF PORTO RICO 37 Measurements of Male and Female $ 9 Snout to vent 103 mm. 9o mm. Snout to posterior edge of tympanum 31 29 Greatest width of liead 40 37 Foreleg from axilla ■ 67 55 Hindleg from vent 134 121 Tibia 42 38 These two specimens are the only ones I have seen from Porto Rico. They are widely different from the Biifo marinus of the southern half of the South American continent. Lutz, in describing- Bufo paracnemis, suggests the composite nature of the wide-spread Bufo marinus of aiithors. There can be no doul)t. however, that the toad here described is the one introduced in Porto Pico, and marinus probably applies best to the ma- rine toad of northern South America. The marine toad is stated, by Wolcott (1924, p. 35), to have been introduced about 1920 at Mayagliez, by Dr. D. W. May. By 1924 it had spread over a radius of about four miles and is apparently securely estab- lished. A second introduction was made in 1924 at Rio Piedras, at the suffcrestion of Sr. Menendez Ramos, former director of the Insular Ex- periment Station. Leptodactylus Fitzinger Rana ; Sapo Text Figs. G-9 Leptodaet J lus albilabris (Giinthcr) Cy St iff not h US ulhihihris Giinther, 1859, Ann. Mag. Nat. Hist.. (3), Vol. IV, p. 217.— Reinhardt and Luetken, 1803, Vid. Me2 mm. Tip of snout to posterior edge of tympanum 18 20 Greatest breadtli of head 19 23 Foreleg from axilla 29 35 Hind leg from vent 69 80 -& The most frequent coloration is a grayish brown of varying shade, sometimes reddish, sometimes nearly black. This may be uniform or Quoted from Stejneger, 1904, p. 585. 48 SCIENTIFIC SURVEY OF PORTO RICO mottled with darker. In the lighter specimens there is nearly always a dark interorbital mark, and in a few the snout is white in front of this either with a broad transverse white band or completely light to the tip of the snout. There is usually also a dark sub-canthal mark, interrupted by the eye, and continued over the ear for a short distance. In a few cases the dorsum is spotted irregularly with vivid white spots. In 18 out of 194 specimens a light line, beginning at the snout and passing over the edge of the eyelid to the ear, continues as a broader light dorso- lateral band to the thigh. In 19 specimens there is a sharp median white stripe (compare Fowler, 1918, Fig. 2). In 5 there is a broad median light band, about four times as broad as the more common narrow line. The hind legs are occasionally distinctly barred, more usually indistinctly barred or uniform. The concealed surfaces of the thighs are often Fig. 12. — Embryo of Eleutherodaciylus portori- censis. A. M. N. H. No. 10302. Six times natural sizp. bright pink or red. The venter is usually light and unspotted, occasion- ally it is spotted with groups of dark-brown punctations. In no specimen were the concealed surfaces of the thighs reticulated with the fine or coarse dark network of E. antillensis. The color variation has also been noted by Stejneger, but I am sure that the juvenile variant described by him from El Yunque must refer to the dwarf species, E. gnjllus. The majority of the specimens collected were taken at night, either singing or merely sitting about on vegetation, sometimes several feet from the ground. Others were found concealed under logs and stones, especially in coffee plantations. The most usual hiding place of this species in the daytime is beneath the sheaths of the outer leaves of the banana. Specimens in this situation are almost without exception uniformly colored and nearly black. About every third banana plant examined was inhabited by one or more of these tree frogs. Persistent search about the banana plants failed to discover the eggs of this species, and it was not until the writer visited the Luquillo .SCHMIDT. AMPHIBIANS OF PORTO RICO 49 •Forest that a single egg-mass was discovered in a basal leaf of an air plant, just at the surface of the water in the lower part of the leaf. A large E. aurioitlatus in the same plant, but not on the same leaf as the eggs, escaped. There were thirty-six eggs, with well-advanced embryos, adhering in an oval mass from which individual eggs could easily be detached. The eggs measure 6-8 mm. in greater diameter, being some- what elongated in the axis of the embryo. The young of this species are extraordinarily abundant, and it is difficult to understand why the eggs are so infrequently observed. It is possible that at the time of the writer's visit (August-October) the height of the breeding season was past. The only date of breeding previ- ously recorded is that noted by Gundlach, May 24 (Peters, 1877, Monats- ber. Akad. Wiss. Berlin, 1876, p. 709). Professor Johnson found a mass of eggs, with embryos at about the stage of those observed by the writer, ill the same bunch of moss in which the giant female specimen, men- FiG. 13. — Peter's figures of the embryo of Elctttlivrodartijlus iiortoricenxis. tioned above, was collected, July 8, 1919. Gundlach (loc. cit) also ob- served a female sitting on the egg-mass received by him, while Bello y Espinosa (Martens, 1871, Zool. Garten, Vol. XII, p. 351) records that in the case noted by him the parent frog remained in the neighborhood of the eggs "as if to guard them." From these several observations it appears not unlikely that the female does remain in the neighborhood of the eggs until they are hatched, Ijut further observations on this point are desirable. Euthven (1915, Occ. Papers Mus. Zool. Univ. Michigan, No. 11), observing the breeding habits of E. cruentus Peters, in Colombia, found no evidence of such a habit. The chorus of this species, or the isolated notes of single males, are among the most familiar and insistent sounds in Porto Eico. The call is a clear, whistled co-qui, co-qui, varied by co-qui-qui-qui-qui-qui. Two males often call alternately from neighboring stations. A favorite situation for the singing male is the base of the leaf of a lilliaceous plant or the center of a whorl of leaves on shrubs, etc., which appears to magnifv the sound and thus increase its ventriloquial character. 50 SCIENTIFIC SURVEY OF PORTO RICO For its relations with other West Indian Eleutherodactylus, and the history of the species, I may quote my remarks when proposing a name for this species : "It is paradoxical to write of the most abundant species of tree-frog of Porto Eico as new to science, for this form is one of the best known in the genus, represented in many museums b}^ large series of specimens. It is really famous, for its eggs and embryos were the basis for the article by Peters, describing the direct development, with suppression of the tadpole stage, which is a general character of the genus Eleutherodac- tylus. Peter's figures have found their way into great numbers of text- books, usually under the original designation, Hylodes martinicensis. "The name now proposed for this well-known and well-characterized species is, nevertheless, the first to be based on Porto Eican specimens. The confusion of this form with other Antillean species has been due to the weight of authority that has identified it first with the Lesser Antillean Eleutherodact ylus martinicensis (Tschudi) — Peters, Gundlach, Garman, and Boettger — and, later, with the Cuban Eleutherodactylus auriculatu-s (Cope) — Boulenger, Stejneger, and Barbour. In dealing with this species in 1920, I accepted the identification witli auriculatiis without question. "Stejneger, in 1902, accepted Boulenger^s record of auriculatus from Santo Domingo, and its occurrence on that island would of course make its presence in Porto Eico much more probable. The Nobles secured an allied species in the Dominican Eepublic, described by Dr. Noble as Eleutherodactylus auriculatoides (1923, Amer. Mus. Novitates, No. 61, p. 3) and it is probable that this species represents auriculatus in Santo Domingo. "The renewed and more intensive study of the Greater Antillean amphibian faunae was, to some extent, initiated by my field work in Porto Eico in 1919, which added six species of Eleutherodactylus to the supposedly well-known herpetolngical fauna of that island. This was followed by the work of Dr. and Mrs. G. K. Noble in the Dominican Eepublic in 1922, which added five new Eleutherodactylus and a new Hyla to the Hispaniolan fauna. The recent additions to the Cuban tree- frogs (eight species) and to the Jamaican fauna (six Eleutherodactylus and a Hyla) by the field work of Dr. Emmett E. Dunn in 1924 and 1925 were, consequently, scarcely surprising, though it may be emphasized that all of these islands were supposed to be well explored herpetologically. The new crop of novel species was due to the application of a simple technique of collecting by voice at night, using an electric flashlight. SCHMIDT. AMl'HIBI \\S OF I'Oh'TO h'ICO 51 ''A l)etter knowledge of the old species has inevitably accompanied the recognition of the new forms, and it is now evident that there are no native species of this genus generally distributed in the Greater Antilles. The Cuban Eleutlierodactylus cmricidaiui^ is now well known through Dr. Dunn's field work. He writes me that tliis species does not breed in bromeliads, and that its cry resembles the syllables "chi-leen." The repeated "coqui" of the Porto Rican species, wdiich gives it its native name, is one of the most characteristic sounds of the nocturnal chorus in Porto Rico. "All of this contributes little by little to the certainty that the common Porto Rican tree-frog is specifically distinct from any other West Indian form.'" Eleutherodaetyius gryllus Schmidt Text Fig. 14 EleuthcnnUntiiluK gruUns .Selimidt. 1!>2(X Ann. N. Y. Acad. Sei., Vol. XXVIII, p. 172, Fig. 3. There is no native name for this species. Tjipe locality. — El Yunque, near the Forester's Cabin, about 1300 feet altitude. Pig. 14. — EleutfieroflactiiUts gryllus. A. M X. H. No. 10226. Twice natural size. Distribution. — Eleiitherodactylus gryllus is confined to Porto Rico, where it is known from the Luquillo Forest and from Maricao. It is probably confined to the more humid higher parts of the island. Specimens collected. — 16: Maricao and EI Yunque. Diagnosis. — Distinguished from Eleutherodaetyius portoricensis by its shorter snout, less granulate venter, and its minute size. Description of type. — "Habitus of Eleutherodaetyius portoricensis, but with a distinctly shorter snout, its length equal to the diameter of the eye (in E. portoricensis the diameter of the eye equals its distance from 52 SCIENTIFIC SURVEY OF PORTO RICO the nostril), and to the interorbital space; canthus rostralis rounded; nostril one-third the distance from tip of snout to eye; tympanum scarcely distinct, one-fourth the diameter of the eye, its distance from the eye equal to its diameter ; toes without vestige of web ; digital disks well- developed ; first toe as long as the second ; an inner and outer metatarsal fold; vomerine teeth in two oblique patches behind and within the choanae: tongue large, slightly nicked behind; skin smooth above, but apparently much more glandular than in E. portoricensis ; venter strongly granulate; a large subgular vocal sac, "Middle of the back, beginning with an interorbital line, dark gray, enclosing a light spot on the occiput ; sides and snout lighter, the darker color everywhere consisting of minute black punctations, especially evi- dent on the limbs and throat; venter light." Measurements Tip of snout to vent 16 mm. Tip of snout to posterior border to tympanum 5.5 (Greatest breadth of head 6 Foreleg from axilla H Hind lee from vent 24 Tibia ^-^ KemaH-s.— The type is a male, taken singing at night, with the usual pale night coloration. Specimens taken in the daytime (concealed under moss) are very dark in color and exhibit considerable variation in pattern, two having a light median dorsal line. In a specimen taken in an air plant (No. 10291) the dorsal dark area is cinnamon brown and the sides bright pale green, the legs with dark bars ; this coloration has been described by Stejneger (1904, Eept. U. S. Nat. Mus., 1902, p. 586) as a variant coloration of juvenile E. auriculatus ( =^ portoricensis) . The darker specimens have narrow light crossbands on the limbs. The granulation of the venter in the female specimens is faint, though still evident. This species was very numerous at Maricao and on El Yunque, sing- ing frequently from trees, at least ten feet from the ground. On El Yunque specimens were collected in air plants, near the peak, and under moss on the rocks of the peak itself. The song is a rapid succession of shrill clicks, very insect-like, the chorus sounding not unlike the rapid clicking of a telegraphic instru- ment. Were it not for the minute size of the singers, and the extremely dis- tinct note, this species might well be considered the young of E. poiiori- SVHMJDT. AMPHIBIANS OF I'OKTO RICO 53 censis; I am unable to agree with Stejneger's supposition that its note is made by juvenile males of the latter species. The oronads, at any rate, appear to be those of an adult in the specimens examined, differing in form and pigmentation from those of young E. poiioricenxls of similar size. Nothing has been added to the knowledge of this species since its de- scription in 1920. ?]leutherodaft.vlus locustus Schiuidt Text Fig. 15 ElcuthtroiUicttilus hxii.sliis .Schmidt, 1020, Ann. N. Y. Acid. 8ci.. Vol XXVIII, p. 174, Fig. 4. No native name is known for this species. Type locality. — El Yunque, near the Foi-ester's cabin, ahoiit I.'IOO feet altitude, Luquillo Forest Reserve, Porto Eico. Bistnbution. — Known only from the type locality. Specimens collected. — 1 : El Yunque. Fig. 15. — Eleutheroihutiilus loeustus. A. M. N. H. No. 10240. Twice natural size. Diagnosi.^. — Size small; snout obtuse; nostril much nearer to the end of the snout than to the eye; tympanum small, indistinct, one-fourth the diameter of the eye, se])arated from the eye by a little more than its diameter; vomerine teeth in two oblique series, behind and within the choanae ; toes free ; digital disks well-developed ; tibiotarsal articulation reaching the posterior border of the eye ; heels overlapping when the legs are placed at right angles to the body; skin rugose above, with scattered round tubercles, especially on the eyelid ; venter smooth ; inner face of thighs finely rugose. Original description. — "Head slightly longer than broad, slightly nar- 54 SCIENTIFIC SURVEY OF PORTO RICO rowei- than the body; snout moderately obtuse, its length anterior to the eye exceeding the interorbital space; nostrils one-fourth the distance l)e- tween eye and tip of snout from the latter; tj'mpanum scarcely distinct, one-fourth the diameter of the eye, se})arated from the eye by a little more than its diameter ; canthus rostralis rounded ; elbow and knee pressed along the side, overlap; heels overlap when the legs are placed vertically to the axis of the body ; tibiotarsal articulation reaching the posterior border of the eye ; disks of fingers and toes well-developed ; toes without vestige of web; inner and outer metatarsal tubercles present; no tarsal fold ; first toe as long as the second ; vomerine teeth in two linear obli()ue patches, converging posteriorly, well-separated on the median line, behind and within the choanal by about the diameter of the choana; tongue large, slightly nicked behind ; skin rugose above, with rounded tubercles ; a well-marked mid-dorsal ridge from snout to vent ; eyelid strongly rugose; venter smooth (faintly rugose under the lens) thighs slightly rugose ; male with a large subgular vocal sac. "Dorsum gray mottled with grayish brown ; a well-defined inter- orbital dark band : sides of canthus with a dark mark, interrupted by the eye, extending over the tympanum; legs not barred, with dusky mark- ings; venter uniform, light." Measurements Tip of sikhU to veut 19 uim. Tip of snout to posterior edge of tympanum 7 Greatest breadth of head 6.5 Foreleg from axilla 12 Hind leg from vent 29 Tibia 9 Remarks. — This species was discovered by accident, singing on a leaf some three feet from the ground. Its song is the most distinctive of any noted in Porto Eico, beginning with a shrill continuous note almost at the limit of audibility, which is followed by a succession of clicks. So closely does this note resemble a familiar type of note produced by long- horned grasshoppers that the writer neglected to search for the author of the sound, and watched the present specimen repeat the song several times before being convinced that it really proceeded from an EleufJiero- clacfyhi.9. Eleutherodactylus locustus is closely related to E. portoricensis, from which it is distinguished by its small size and smooth venter. Even more closely related to the still smaller new species E. gri/llus, it is nevertheless readily distinguished by its smooth venter and moi-e rugose dorsum, as well as by its remarkable voice. f^CHMIDT, AMPHIBIANS OF PORTO RICO 55 This species has not been rediscovered since its description in 19 SO. I might entertain a doubt as to its validity, in view of its being based on a single specimen, had not Dr. and Mrs. Xoble since observed an Eleuth erodactylus with a similarly peculiar note in Hispaniola. •Eleutherodac'tjius cramptoni* Schmidt Text Fig. 16 Eleuth rrodactylus crampton i XXVIII, p. 176, Fig. 5. Schmidt. 1920, Ann. N. Y. Acad. Sci.. Vol. No native name is available for this species. Type locality. — Peak of El Yunque, 3485 feet altitude, Luquillo Forest Eeserve, Porto Rico. Distribution. — Known only from the type locality. Specimens collected.— o : El Yunque. Fig. 16. — Eleutherodactylus cramptoni. A. M. N. H. No. 10305. Twice natural size. Diaynosis. — Size small; habitus stout; hind legs short; snout very obtuse, canthus rostralis rounded; dorsum very rugose with rounded tubercles ; vomerine teeth in two oblique linear series, extending laterally as far as the inner border to the choans; digital disks large; uniform dark brown above, light brown beneath. Original description. — "Habitus stout, compact; snout short, obtuse, canthus rostralis rounded; nostril one-third the distance from tip of snout to eye ; heel reaching the anterior border of the orbit ; heels meet but do not overlap when the legs are placed at right angles to the body ; both anterior and posterior limbs notably stout, nearly twice as thick as * Named for Professor Henry E. Crampton, whose active interest and investigation have greatly furthered the zoological work of the Scientific Survey of Porto Rico and the Virgin Islands. 56 SCIENTIFIC SURVEY OF PORTO RICO those of E. portoricensis of the same bod}' length ; vomerine teeth in two linear, oblique series, extending laterally as far as the choanae; tympanimi small, distinct; dorsum covered with rounded tubercles, extending onto the eyelids and snout; venter finely granular; digital disks large, first toe as long as the second ; no subgular vocal sac. "Color uniform brown above, lighter brown below, slightly variegated with lighter punctation." Measurements Tip of snout to vent 1'^ ^^■ Tip of snout to posterior edge of tympanum 6.5 Greatest breadth of head ' Foreleg from axilla 1- Hiud leg from vent -'^ Tibia ^ Remarks. — The two paratypes are similar in every respect to the type, with the single exception that one of them is slightly more mottled with light, and has the hind legs indistinctly barred. All three specimens were taken under moss in the crevices of the rocks on the peak of El Yunque. The species is a well-differentiated one, characterized by the stoutness of its limbs, the obtuseness of the snout and the extreme rugosity of the dorsum. Like the other novelties in my collection of 1919, this species awaits the verification of further field-study. Eleutherodact J his antillensis (Reinbanlt & Luetken) Coqui Text Fig. 17 Hi/Iodcs antillensis Reinbardt and Luetken, 1863, Vidensk. Med. naturh. For. Kjobenhavn, 1862, p. 209. Eleutherodact plus antillensis Stejneger, 1904, Rept. U. S. Nat. Mus., 1902, p. 591, Figs. 20-24.— Barbour, 1914, Mem. Mus. Comp. Zool., Vol. XLIV, p. 247; 1917, Proe. Biol. Soc. Wash., Vol. XXX. p. 102.— Schmidt, 1920, Ann. X. Y. Acad. Sci.. Vol. XXVIII, p. 178, Fig. 6. Hylodes martimccnsis Peters, 1876, Monatsber. Akad. Wiss. Berlin. 1876, V\ .1, Fig. 6. This species is not distinguished by the Porto Ricans from the E. ■por- toricensis, and when observed at all is called a "coqui." Type localiti/.- — St. Thomas, Virgin Islands. Di^ribuiion. — This species was long known only from the Virgin Islands, St. Thomas and Tortola, and from A'ieques Island. It is SCHMIDT, AMPIIlBlANti OF PORTO RICO ot doubtfully reported from St. John and St. Croix. I collected it on Porto Rico and Culebra in 1919. Specimens collected. — 30 : Aibonito, Bayamon^ Maricao, Santurce and Culebra Island. Diagnosis. — Limbs shorter than in Eleutherodactylus portoricensis, the heels just meeting when the legs are bent at right angles to the body. Posterior surfaces of the thighs with a dark reticulation. Venter more coarsely granulated and digital disks smaller than in portoricensis. Original descrpAion. — "An Hylodes with a verrucose venter, palatine Fig. 17. — Eleutherodactylus antil- Iciisis. A. M. N. H. No. 10019. Twice natural size. teeth moderately separated, each series wedge-shaped, the two forming an angle open forward; [palatine teeth] widel} separated from the border of the lip ; digital disks rather large." Remarks. — Stejneger* gives a detailed description of this species, based on a Vieques Island specimen : "Tongue rather broad, heart-shaped, slightly nicked behind; vomerine teeth in two club-shaped oblique series, some distance behind but not laterally beyond the choanse, converging backward and well separated ; nastril much nearer the tip of snout than the eyes, their distance from the eye less than the diameter of the latter; upper eyelids narrower than the interorbital space; tympanum a little less than one-half the diam- eter of the eye, its distance from the eye less than one-half its diam- * stejneger, 1904, p. .592. y '■."--' •"5;''Sv V'V^- k:Ji L3 \.-r u-^'\ 58 SCIENTIFIC SURVEY OF PORTO RICO eter; fingers with rather small disks, first equalling second; disks of toes not smaller than those of the fingers; tip of first toe reaching disk of second; two metatarsal tubercles, the outer being rather small and obscure; series of plantar tubercles corresponding to metatarsals; no tarsal fold; the bent lim])s being pressed along the side, knee and elbow, fail to meet; hind limb being extended along the side, heel reaches the eye; hind limbs being placed vertically to the axis of the body, the heels barely meet; skin above with scattered granules and a very narrow raised median line from tip of snout to vent; throat and chest smooth, belly and posterior aspect of femur strongly granular; a strong fold across the breast between the axillae." The original description is sufficiently diagnostic as the only other species on St. Thomas is the smooth-skinned E. lentus, which has the vomerine teeth in long transverse series. Color. — In coloration this species is less variable than E. portoricensis but the median white dorsal line may be present or absent. It is de- veloped in twelve of the present specimens. The usual color is grayish brown, with faint dusky markings, and a sharply defined black canthal line which extends over the ear and a short distance beyond it, outlined above in most cases by a very narrow white line on the canthus extending over the eyelid. The concealed surfaces of the legs are reticulated with black, which affords a fairly good character for distinguishing this species in the field from E. portoricensis. One specimen, No. 10001, a male, is violet-red above, has a very heavy black canthal and supra-auricular mark, and the concealed surfaces of the legs black with sharply defined Avhite spots. The measurements of the largest specimens of each sex follow: Parts measured No. 10117 d 10082 ? Tip Of snout to vent 24 mm. 33 mm. Tip of snout to posterior edge of tympanum 10 13 Greatest breadth of head 11 14 Foreleg from axilla 16 19 Hind leg from vent 38 48 Eabiis. — This species prefers slightly lower herbage and- slightly wet- ter situations than E. portoricensis, which is often associated with it. It ranges to an altitude of nearly 2000 feet at iVibonito. Nothing is known of the breeding habits of this species. The song of E. antillensis may be readily distinguished from that of E. portoricensis by its more metallic quality, and the frequent series of uniform notes ki-ki-ki-ki-ki . The males conceal themselves more SCHMIDT, AMPHIBIAXS OF PORTO RICO 59 carefully wheu calling, making it exasperatingly difficult to locate the singer. They often sing from a position in the axils of the leaves of lilliaceous plants. In Sauturce along the railroad and trolley embank- ments north of the Hotel Eureka, I found this species more abundant than E. portoncetisis. The single specimen from Culebra agrees closely with the Porto Eican series. The discovery of this form in Porto Rico proper greatly reduces the differentiation of the Virgin Island fauna. EJeutlierodactylus brittoni* Schmidt Text Fig. 18 EleuthcrodmtuluH brittoni Schmidt, 1920, Ann. N. Y. Acad. Sci. Vol. XXVIII, p. 179, Fig. 7. No native name is available for this species. Type locality. — El Yunque, near the Forester's Cabin, about 1300 feet altitude, Luquillo Forest Reserve, Porto Rico. Distribution. — Known only from El Yunque and Maricao. Specimens collected. — 4 : El Yunque and Maricao. Diagnosis. — Derived from Eleutherodactylus antillensis, from which it is distinguished by its small size, its sharp canthus rostralis, which is continued as a dorso-lateral angle some distance behind the ear, and its more posteriorly placed nostril. Original description.— -^^H'dhiiw?, slender, head narrower than the body, legs rather short, snout sharp-pointed; nostril two-fifths the distance from the end of the snout to the orbit ; canthus rostralis sharp : inter- orbital space broader than the eyelid; heel reaching the anterior border of the or])it ; heels meeting but not overlapping when the legs are at right angles to the l)ody; top of snout fiat, as is the anterior half of the back behind the eyes, the side of the body being vertical anteriorly ; vomerine teeth in two small rounded patches, behind and within the choana^ : tympanum indistinct, separated from the eye by less than its diameter; dorsum smooth, venter coarsely granulate; digital disks small, as long as wide : a well-defined tarsal fold ; a well-developed sitbgular vocal sac. "Dorsum light grayish brown, venter lighter. Two black spots between the eyes, one on the middle of the back, and three posteriorly on the back, above the groin ; legs with a single faint darker bar on the femur ; concealed surfaces of the femur not reticulated; a black subcanthal streak, continued below the dorso-lateral angle behind the eye." * Named for Dr. Nathaniel !>. Britton. Chairman of the Committee for the Scientific Survey of Porto Rico and the Virgin Islands, New York Academy of Sciences. 60 SCIENTIFIC SURVEY OF PORTO RICO Measurements Tip of snout to vent 16 nun. Tip of snout to posterior edge of tympanum Greatest breadth of head G Foreleg from axilla Hind leg from vent 23 Tibia S Remarks. — The three paratypes are closely similar in size and struc- tural characters to the type. Two have the black subcanthal and shoulder mark outlined with white above. One lacks the dorsal black spots. The single specimen from Maricao was taken singing in herbage along tlie roadside, together with E. portoricensis and E. antillensis. Two Fig. 18. — Eleuthcrodactylvs brittoni. A. M. N. II. No. 10318. Twice natural size. were taken singing on El Yunque, likewise in low herbage, and the last was found by accident in collecting E. wightmanae. The song of this species is a succession of clicks, less shrill and less rapid than in E. gryllus. This species stands in the same relation to E. antillensis a.- E. gri,li'iis does to E. portoricensis. Eleutherodactylus wightmanae* Schmidt Text Fig. 19 Eleutherodactylus wightmanae Schmidt, 1920, Ann. N. Y. Acad. Sei., Vol. XXVIII, p. 181, Fig. 8. No native name exists for this species. Type locality. — El Yunque, near the Forester's Cabin, about 1300 feet altitude, Luquillo Forest Reserve, Porto Rico. * Named for tlie author's wife, Margaret Wightman Schmidt, whose loyal assistance contributed largely to the success of the work in Porto Rico in 1919. SCHMUrr, AMPHIBIANS OF PORTO RICO 61 Distrihution. — Known from El Yunque and Maricao, at opposite ends of the island. Specimens collected. — 13 : El Yunque and Maricao. Diagnosis. — Size small; snout pointed; nostril much nearer to the tip of the snout than to the eye; tympanum small, distinct, separated from the eye by about its own diameter; vomerine teeth in two straight series, in the same line, extending as far laterally as the choan«, and about the diameter of a choana behind them ; toes free, digital disks well- developed; tibio-tarsal articulation reaching the anterior border of the eye; heels overlapping when the legs are placed at right angles to the body; skin rugose above, with elongate folds and ridges; venter rugose; thighs granular. Original description. — "Head as long as broad, narrower than the body; snout pointed, its length anterior to the eyes once and a half the Fig. 19. — Eleutherodactylus wi(jht- manae. A. M. N. H. No. lO^-JO. Twice natural size. interorbital width ; nostrils one-third the distance between eye and tip of snout from the latter ; tympanum distinct, small, about one-third the diameter of the eye, separated from the eye by a little more than its own diameter; canthus rostralis sharp; elbow and knee pressed along sides overlap : heels overlap when the legs are placed at right angles to the body; tibio-tarsal articulation reaching the anterior border of the eye; disks of fingers and toes well-developed; digits slender, free; first toe distinctly shorter than the second ; no tarsal folds ; vomerine teeth in two straight series, separated in the median line, extending laterally as far as the outer border of the choana?, and about the diameter of a choana behind them; tongue large, slightly nicked behind; skin rugose above, with longitudinal lines or folds, the most distinct of which orig- inate behind tlie orbits and extend backward about two-thirds the length 62 SCnrXTIFfC .PURVEY OF PORTO RICO of the back ; a less distinct mid-dorsal ridge from snout to vent ; venter and onter face of the thighs moderately rugose : a subgular ^'(»cal sac. "Brown above, with a black subcanthal line, extending over the ear half way along the sides ; a black spot on each side of the back over the groin ; venter uniformly light ; a single dark crossband on the radius ; one on the femur, tibia, and tarsus (in line when the legs are folded), and a dark spot on the metatarsus; anterior and posterior faces of the thigh dusky. j> Measurements Tip of snout to vent 20 mm. Tip of snout to posterior edge of tympanum 7.5 Greatest breadth of head 7.5 Foreleg from axilla H Hind leg from vent 30 Tibia 1<^ Remarks. — In structural characters the twelve paratypes agree closely with the type. Two specimens are light gray, instead of brown, with only indications of the black spots; in most specimens the postocular dark streak is broken up into a series of spots; one specimen is light brownish gray on each side, the area between shar]ily darker : the bars on the legs are distinct in all specimens. The plaintive, diminuendo note of this small species is one of the most characteristic sounds in the amphibian chorus of the Luquillo Forest. Its song consists of a series of six or eight whistled notes, each slightly lower in pitch and a little fainter than the previous one. The creature sings habitually on the ground or in the lowest leaves of plants. To lo- cate its position from its song it is particularly difficult, partly because it is usually well concealed, partly on account of the peculiar ventriloquy of its voice. Nothing has been added to our knowledge of this species since its dis- covery. Eleutherodaetylus rirhinoiidi Stejneger Text Figs. 10 and 20 El€uthero(Jacti/lus richmondi Stejneger, 1904, Rept. U. S. Nation. Mus., 1902, p. 593, Figs. 25-29. — Barbour, 1914. Mem, Mus. Comp. Zool., Vol. XLIV, p. 247.— Schmidt, 1920, Ann. N. Y. Acad. Sci., A'ol. XXVIII, p. 183, Fig. 9. No native name exists for this species. Type locality. — Catalina Plantation, about 890 feet altitude, eastern slope of El Yunque, Porto Eico, SCHMIDT, AMPHIBIANS OF PORTO RICO 63 Distribution. — Known only from Porto Eico and apparentl}' confined to El Yunque. Specimens collected. — 11 : El Yunque, from 1300 feet to the peak. Diagnosis. — ^"Toes free without a vestige of web; belh^ smooth; tym- panum distinct, less than one-half the diameter of the eye; vomerine teeth in two long angular transverse series, extending beyond the external border of the inner nares and some distance behind them; digital disks small ; nostril much nearer tip of snout than eye ; hind limbs not cross-barred." Fig. 20.^Eleuthero(lactylus richmondi. A. M. N. H. Xo. 10237. Twice natural size. Original description. — -"Tongue narrow, somewhat emarginate behind; vomerine teeth in two angular series behind the choanal, their distance from the choanffi greater than the diameter of the latter; inner arm of each vomerine series longer, outer extending laterally beyond the choange ; nostril much nearer the tip of the snout than the eye, the distance from the eye slightly less than the diameter of the latter; upper eyelids somewhat narrower than interorbital space, tympanum slightly less than one-half the diameter of the eye, its distance from the latter slightly less than its diameter; disks of fingers rather small, first finger shorter than second; disks of toes small, first toe short, only reaching ,^,4 SCIENTIFIC SURVEY OF PORTO RICO subarticular tubercle of second; subarticular tubercles well devel- oped; two well developed metatarsal tubercles; no plantar tubercles; no tarsal fold; the bent limbs being pressed along the sides, knee and elbow overlap; hind limb being extended along the side, heel reaches center of eye; hind limbs being placed vertically to the axis of the body, the heels overlap; skin above and on flanks granular, underside smooth; posterior aspect of femur areolate." Remarks. — The coloration in life has been described by Stejneger as follows: — "Back dusky chestnut, lighter on sacrum; from each nostril along canthus rostralis, edge of eyebrow and sides of back a narrow dirty bluish-white stripe somewhat wider on sides of back than on canthus rostralis; sides of face and flanks below this stripe blackish, legs black- ish; fore legs marbled with pale drab, hind legs with dull pale chest- nut ; under side dull greenish gray, with an ill-defined yellow spot in each groin, and marbled with dusky brown on throat and under side of hind legs. Iris blackish, brassy above pupil." Like the larger series examined by Stejneger, the specimens collected on El Yunque by myself are extremely uniform in structural characters and in coloration. The only variation noted is the occasional lightening of the chestnut color of the dorsal area between the light dorso-lateral lines. The proportions are quite different in this species from the other Porto Eican species of the genus : — A.M.N.H. U.S.N.M. Parts moasured No. 19233 No. 26884 Tip of snout to vent 32 mm. 38 mm. Tip of snout to postei-ior edge of tympanum 13 15 Greatest breadth of head 12 15 Foreleg from axilla 21 24 Hind leg from vent 51 62 Two extremely small specimens, measuring 9 and 11 mm., respectively, probably are recently transformed. They are colored like the adults. All of the specimens known were found under stones or palm leaves on the trail or on damp ground, associated with E. portoricensis. The males of this species were not discovered singing and its voice is un- known. The slender digits and small adhesive disks probably indicate that it is more terrestrial in its habits, and the eggs may prove to be laid on the ground, like those of E. luteolus of Jamaica. This species is allied by the form of its vomerine teeth to E. lenius of the Virgin Islands, E. monensis of Mona Island, E. weinlandi of His- paniola and perhaps to E. jamaicensis of Jamaica. This group of species thus composes an interesting series of vicariating forms. SCHMIDT, AMPHIBIANS OF PORTO RICO 65 Eleutherodactylus inonensis ( Meerwarth ) Text Fig. 21 Hylodes monensis Meerwarth, 1901, Mitt. Naturli. Mus. Hamburgli, Vol. XVIII, p. 39, PI. 1 (Fig. 11), PI. 2 (Figs. 4-5). Eleuthei-odactylus monensis Stejneger, 1904, Rept. U. S. Nation. Mus., 1902, p. 595, Figs. 30-34.— Barbour, 1914, Mem. Mus. Comp. Zool., Vol. XLIV, p. 247.— Schmidt, 1926, Puhl. Field Mus. Nat. Hist., Zool., Vol. XII, p. 154. No native name exists for this species other than "coqui." Type locality. — Mona Island, West Indies. Distribution. — Confined to Mona Island. Specimens collected. — 41 : Mona Island. Diagnosis. — Vomerine teeth in long arched transverse series behind the choanae; belly smooth; soles of feet tubercular; hind foot nearly as long Fig. 21. — Eleutherodactijlus mo- nensis. A. M. N. H. Xo. 24463, as fore leg; color of back and sides pale, with brownish markings. Original description. — "This species is closely allied to the previous one [Hylodes lentus Cope] and differs from it in the following features : in H. lentus the femur, measured from the ischio-pubic crest to the edge o| the knee is shorter than the tibia and at most as long as the distance from axilla to groin, while in H. monensis it is as long as the tibia and longer than the distance from axilla to groin; the tongue is wedge- shaped; the vomerine teeth are also in two rows behind the choana?, but G6 SCIENTIFIC SURVEY OF PORTO RICO the angle in each row is more obtuse, and the outer portion (beyond the ankle) is about half as long as the inner, instead of subequal, as in H. lentils. "The color of the belly is uniform whitish yellow, that of the upper side a grayish flesh-color with small brown spots sparsely distributed on the back, sides, and limbs, and a more or less well-defined star-shaped figure formed from larger brown spots between the shoulders. A brown line extends from the nostril to the eye." Remarks. — This species is well characterized in the original descrip- tion and figures. The series of specimens secured by Anthony in 1936 agrees with the Hamburg originals and the single specimen in the National Museum described in detail by Stejneger. Eleutherodactylus unicolor Stejneger Text Figs. 22 and 2.3 Eleutherodacti/lus unicolor Stejueger, 1904, Rept. U. S. Nation. Mus., 11)02, p. 597, Figs. 35-39. No native name. Tijpe locality.— Cam]) on El Yunque at 29 T8 feet altitude, Luquillo Forest Eeserve, Porto Rico. Distribution. — Known only from the type locality. Duignosis. — "Toes free without a vestige of Aveb ; belly granular; tympanum distinct, one-third the diameter of eye; vomerine teeth in .iii(i. (From Stejneger. ) Dr. E. Graywood Smyth of the Agricultural Experiment Station at Rio Piedras to frequent the vicinity of the arc-lights in Rio Piedras at night to prey upon the insects. Beyond this meagre information nothing is known of its distribution on the island. In the West Indies as a whole it is widespread— from Trinidad and Parbadoes to Cuba and Jamaica. It is curious that it has not been recorded from the northern grouj) of smaller islands in the Lesser Antilles. I have little doubt of its identity as a species through- out this range. A West African species, Heinidactylm hrookii, occurs in Hispaniola (Port-au-Prince), where it was doubtless introduced by the slave-ships. The ]\lediterranean Hemidacfi/lus iiircirus. it is said, has become establislied at Key West. Sphaerodaetyhis Wagler Spha erodactj lus niaerolepis Giinther Text Figs. 24 and 26 Salaniandni. SalaiiiJindrita ; Salaraanqua, Salamanquita ; Lucia (?); Ranita ; I^agartija cabeza de muerte SCHMIDT, AMPHIBIANS OF PORTO RICO 71 SpJiacrodactt/lus inacrolcpis Giinther, 1<859, Ann. Mag. Nat. Hist.. (3), Vol. IV, p. 21.3, PI. 4, Fig. B. — Barbour, 1914, Mem. Miis. Comp. Zool., Vol. XLIV, p. 270; 1915, Proc. Biol. Soc. Wasli., Vol. XXVII, p. 72: 1917, Vol. XXX, p. 08.— Schmidt. 192(K Ann. N. Y. Acad. Sci., Vol. XXVIII, p. 1^.— Bar- bour. 1!)21, Mem. Mus. Comp. Zool., Vol. XLVII, p. 253, PI. 6, Figs. 2-3, I'l. v.). Figs. 5-8. — Wolcott, 1924, Journ. Dept. Agric. Pto. Rico, Vol. VII, p. 13. Spluierodactiilus nutcrohpix moncnsis Meerwarth, 1901, Mitt. Naturh. Mus. Hamimrg. Vol. XVIII. p. 20. SpharrrMlartiiliis monensh Stejueger, 1904, Kept. U. S. Nat. Mus., 1902, p. 007.— Barbour, 3!»14. Mem. Mus. Comp. Zool.. Vol. XLIV, p. 270. Sphdcrodartylus grandinquamis Stejueger, 1904, Kept. U. S. Nat. Mus., 1902, p. 002. Figs. 40-52.- Barbour, 1914, Mem. Mus. Comp. Zool., Vol. XLIV, p. 270.— Fowler, 1918, Publ. Carnegie Inst. Wash., No. 252, p. 7. The species of this genus are usually known as Salamandritas (or Salamanquitas) to the Spanish-speaking West Indians. The wriggling mode of progression of these tiny geckos is quite salamander-like and the name is to that extent appropriate. Stejneger found that the name "Lucia"' or "Santa Lucia'' was applied to this species by the children in Luquillo. This, however, must have been an unfortunate localism, as these names are used everywhere in Porto Eico for the Mahuya, a totally different creature. The equally local use of the name "Eanita" at Aibo- nito illustrates the tendency to apply a name which exists in the language to any unidentified species. Type locality. — St. Thomas, Virgin Islands. Distrihidion. — ^Yidespread on Porto Eico, M'here it is recorded from Aibonito, Bayamon, Cantaiio, Coamo Springs, Ensenada, Luquillo, Maricao, Ponce and El Yunque. It is evidently not confined to the coastal plain as supposed by Stejneger. On the outlying islands it is known from Mona and Vieques and occurs on most of the Virgin Islands. Specimens collected. — 45, from Aibonito, Bayamon, Cataho, Coamo Springs, Ensenada, Maricao, El Yunque and Mona Island. Diagnosis. — A small geckoid lizard with unexpanded digits which are provided with an enlarged flat circular scale beneath the tip; dorsal scales keeled, imbricate, somewhat variable in size ; no vertebral series of small scales. Original description. — "Body surrounded by about forty longitudinal Series of scales of rather large size; no vertebral streak of smaller ones, those of the back keeled, of the belly smooth. Trunk and tail uniform blackish ])rown, in younger individuals some scales with blackish tips; head greyish brown, marbled with black ; jaws and throat striolated with blackish. 72 SCIENTIFIC SURVEY OF PORTO RICO "The snout is of moderate extent, and slightly pointed; all the upper surface of the head and the sides are covered with scales of moderate size; there is an exceedingly small horn-like spine above the middle of the orbit. The rostral shield is low, and bent backwards on the upper surface of the snout; the sides of the jaw are margined with three elon- gate labials; the nostril is situated above the posterior extremity of the rostral shield and first labial, and exceedingly small. The anterior lower labial is single; a series of three other shields covers the lateral margin of the lower jaw. The scales of the throat are small, those of the breast and of the extremities keeled. The ear-opening is very small, one-third only of the width of the eye. The fingers and toes have an en- FiG. 26. — Head and shoulders of Sphaerodactylus macrolepis. A. M. N. H. No. 13037 (A) and No. 13697 (B). showing two common types of pattern. Two and a half times natural size. tire and unarmed disk. The tail is covered with smooth scales, rather smaller than those of the trunk; there is a series of larger ones, plate- like, along the lower medial line. No femoral or anal pores. "I add to the statement of the coloration given above that the belly is uniform dirty white, and the tail minutely dotted with blackish. Two specimens were in the collection." Remarks. — Barbour (1917, p. 98), after examining a considerable series of Spaerodactylus macrolepis from the Virgin Islands, expressed a measure of doubt as to the distinctness of 8. grandisquamis. Both Bar- bour and I have subsequently reached the conclusion that it is untenable. Stejneger separated S. grandisquamis and S. monensis from *S'. macrolepis solely on the size of the scales, which he gives as 34-38 about the body in S. grandisquamis, 46-48 in S. monensis. In the series from Porto Eico collected by myself the variation is as follows : SCHMIDT, AMPHIBIANS OF PORTO RICO 73 Scales about the body 32 3G 40 44 48 52 5G Number of specimens 2 8 3 4 5 6 2 In five specimens from Mona Island the numbei- of scales varies from 44 to 52 ; as S. macrolepis is intermediate between ^S*. grandisquamis and S. monensis, it is evident that the variation in the present series includes all three supposed forms. There is probably a somewhat different range of variation on the several islands but the extremes are certainly in- cluded in that of the Porto Rican series. Reproduced tails have a much widened series of median ventral scales. Measurement of A. M. N. H. No. 1.3324 Total length 06 mm. Tail 34 Tip of snout to posterior edge of ear 8 Breadth of head 5.5 Foreleg from axilla 7.5 Hind leg from groin 10 The smallest specimen measures 12 mm. from snout to vent. This species is highly variable in coloration. The absence of the elab- orate head pattern appears to be due to the general darkening of the coloration, as all of the lighter specimens have it in some form or other. Only two specimens are without a trace af the black shoulder band with its two white spots. In the remaining specimens its development is very irregular, the white spots persisting even when the black band is indis- tinguishable. The darker dorsal spots are usually in rows, and in some specimens form longitudinal lines. The throat is miiform or heavily dotted with dark spots. Juvenile specimens are usually dark in color, with the white spots of the scapular band distinct. One specimen (N"o. 13811) is light brown in ground color with five light gray longitudinal stripes; one, median from between the eyes to the base of the tail; two, dorso-lateral from the upper posterior corner of the eye to the base of the tail, and two, lateral from the eye through the ear to the groin, the latter only narrowly separated from the light venter. In specimens from Mona Island, the scapular black band is larger, outlined with light color, the brilliant white spots transverse, forming a continuous line in one specimen. Habits. — The usual habitat of this species is the ground cover of dead leaves in coffee plantations and forest. Elsewhere it is found under stones and logs. Along the base of the limestone cliffs on the hills back of Catano, Sphaerodactyliis were numerous, scurrying for the crevices at the base of the cliff when alarmed. One specimen was found under a leaf 74 SCIENTIFIC SURVEY OF PORTO RICO sheath, and others on the trunks, of banana plants. The specimens from Bayamon were found in picking up rubbish on freshly cleared land. These reptiles frequently venture forth in broad daylight, but are prob- ably essentially crepuscular or nocturnal. Four of the specimens collected on Mona Island were taken among the limestone boulders on the west side of the island, and of these three were secured at night wdth the hand lamp. Two were taken beneath pieces of coral on the flat terrace on the south side. Their extreme agility makes them difficult to catch, and the tail is often broken or the skin torn. When caught, they frequently turn a uniform light gray, becoming brown again in the collecting bag. An egg, probably of this species, was found under a log at Aibonito, August 21, 1919. It is white, discolored by stains, with a hard and smooth shell, 6 x 4.5 mm. Iguanid.ae Anolis Daudin The lizards of the genus AnoliSj characterized by expanded digits that liave a raised terminal claw-like portion, and by the presence of a throat- fan, usually brightly colored, which may be distended vertically, are the most ubiquitous of West Indian lizards. The multitude of species in the West Indies is paralleled in Central America and in northwestern South America. Wherever lizards of the genus Anolis occur, a few species are usually extremely abundant — thus Anolis crisiatellus on Porto Eico, A. cyhotes on Hispaniola, .4. sagrei on Cuba and at Belize. The power of color change is highly developed in the Anoles, and they are frequently miscalled ''chamaeleons" or "cameleones." As Stejneger remarks, "Anolis"' might well be adopted as a vernacular name. In spite of their color-change, the species are almost invariably distinguishable by some feature of their coloration, and I have drawn up a key to the males of the species on this basis to supplement tlie synopsis based on structural characters devised by Stejneger, which must be referred to when the color characters fail. Key to the Species of Anolis Recorded from Porto Rico A. Dorsal scales entirely separated from each other by several circles of granules ; size large ; male with tail crest A. cuvieri AA. Dorsal scales juxtaposed or imbricated ; size moderate or small. B. Dorsal scales (all, or with the exception of two rows on the median line) granular or tubercular, differing liut little, if at all, from laterals, but very much from the much larger ventrals, which are smooth or feebly keeled. SCHMIDT, AMPHIBIAN^; OF PORTO RICO 75 C. Two, or more, shields or scales between the superciliaries and the supraocular semicircle boi-dering the supraocular granules anteriorly. Tail of male crested. I). Supraocular semicircles separated by at least two scale rows; occipital shield separated frr)m supraocular semicircles by at least five scale rows (Fig. — ) A. gundlacM I>r>. Supraocular semicircles in contact or with at most a single series of scales between; occipital shield separated from » supraocular semicircles by at most four scale rows A. cristatellus CC. One shield between the superciliaries and the supraocular semi- circle bordering the supraocular granules anteriorly. D. Width of head as great, or greater than distance from tip oi snout to center of eye ; anterior femoral scales keeled, gradually diminishing ; color greenish A. evermanni DD. Width of head less than distance from tip of snout to center of eye; anterior femoral scales smooth, abruptly larger than the others ; color brownish or grayish 1. stratulus BB. Dorsal scales large, flat, keeled, imbricate, very much like the ventrals, which are very strongly keeled, the keels forming continuous ridges. C. Lateral scales granular. D. Width of head much more than half the distance from tip of snout to ear-opening ; four to six median dorsal scale rows more or less abruptly larger than the others ; skin of dewlap in male. orange 1. krufri DD. Width of head about one-half the distance from tip of snout to ear-opening ; dorsal scales gradually increasing in size from the laterals toward the median rows ; skin or dewlap in male, crimson A pidchcUus CC. Lateral scales imbricated, keeled 1. poncensis Color Key to Porto Rican Angles A. No longitudinal stripes. B. Color brown or gray, never bright green. C. No short transverse saddle-shaped vertebral spots. D. Iris dark brown ; skin of dewlap greenish yellow, its edge brown- ish orange A. rristd tell us DD. Iris metallic blue; skin of dewlap orange olive, with distant yellow scales A. guudlaehi CC. Short black saddle-shaped spots on the mid-line of the back ; dew- lap bright orange 4. stratulus BB. C(dor bright green (if not green, no transverse bands or spots). C. Size large, front of head flat, bony .1 cuvieri CC. Size moderate, front of head concave A evermanni AA. Light longitudinal stripes present. B. Throat-fan white A. poncensis BB. Throat-fan crimson A. pulchellus BBB. Throat-fan orange A. Irugi SCIEXTIFIC SURVEY OF PORTO RICO Fig. 27. — Heads of Porto Rican Anolis. Anolis cuvieri (left of top row). AiioHs crista- tellus (center of top row), AnoMs gumllachi (right of top row) ; Anolis evermanni (left of middle row), Anolis stratulus (center of middle row), Anolis krugi (right of middle row) ; Anolis pulchellus (left of bottom row), Anolis poncensis (right of bot- tom row). (From Stejneger.) SCHMIDT, AMPHIBIANS OF PORTO RICO 77 Anolis cuvieri Menem Lagarto, Chipojo Text Figs. 27 and 28 Anolis cuvieri Merrem, 1820, Syst. Amphib., p. 45. — ISoulenger, 1885, Cat. Lizards, Brit. Mus., Vol. II, p. 23.— Garman, 1887, Bull. Essex Inst., Vol. XIX, p. 27.— Stejneger, 1904, Kept. U. S. Nation. Mus. 1902. p. 627, Figs. 81-84, 87.— Barbour. 1914, Mem. Mus. Comp. Zool., Vol. XLIV, p. 273.— Fowler, 1918, Papers Dept. Marine Biol., Carnegie Inst., A'ol. XII, p. 7. — Schmidt, 1920, Ann. N. Y. Acad. Sci., Vol. XXVIII, p. 185.— Smyth, 1920, Rev. Agric. Pto. Rico, Vol. IV, p. 19.— Wolcott, 1924, Journ. Dept. Agric. Pto. Rico, Vol. VII, p. 14. Anolius velifer Cuvier, 1829, Regne Anim., 2nd Ed., Vol. II, p. 29, PI. 5, Fig. 1.— Guerin, 1830, Icon. Regne Anim., Rept., PI. 12, Fig. 1. Anolis velifer Dumeril and Bibron, 1837, Erpet. Gen., Vol. IV, p. 164. — Dumeril, 1851, Cat. Method. Rept. Mus. Paris, Vol. I, p. 59.— Reinhardt and Luetken, 1863, Vid. Meddel. Naturh. Foren., Copenhagen, 1862, p. 260.— Cope, 1868, Proc. Acad. Nat. Sci. Phila., p. 312.— Peters, 1876, Monatsber. Akad. Wiss. Berlin, p. 705. — Gundlach, 1881, Anales Soc. Espan. Hist. Nat., Vol. X, p. 308.— Stahl, 1882, Fauna Puerto-Rico, p. 69, p. 159. Xiphosurus velifer Cope, 1861, Proc. Acad. Nat. Sci. Phila., p. 208. Type locality. — Jamaica (erroneously). Distribution. — Anolis cuvieri has been taken at Aibonito, Catalina Plantation (El Yunqiie), Ciales, Hmnacao, Luquillo, Mayagiiez and Utuado. It is probably not found in the arid southwestern corner of the island, but ranges quite generally over the remaining part of Porto Eico. It is recorded from Vieques and Tortola of the Virgin Islands. Its absence from the other Virgin Islands is perhaps due to difEerence in the habitat conditions or perhaps to extinction. It is nearly allied to Anolis ricordii of Hispaniola. Specimens collected. — 11. Aibonito. Original description. — "The rayed fin extending from the base to the middle of the tail, with twelve to fifteen rays. Throat fan extending to the breast." Remarks. — Stejneger's discussion of the historical aspect of the taxonomy of this species, and of its relations with the Hispaniolan Anolis ricordii is a model of completeness and clarity. His description of a male specimen follows : "Top of head flat, with only shallow depressions on prefrontal and occipital region, the scales being rather small and roughly keeled and tuberculated, even those on top of the snout, but especially those of the supraorbital semicircle and frontal ridges; about nine enlarged supra- oculars, flat, keeled, and in contact with the semicirculars ; supraorbital 78 SCIENTIFIC SURVEY OF PORTO RICO semicircles separated by about three scale rows from each other and ivhm the occipital, which is barely noticeable ; scales surrounding the occipital depression on the sides and behind rather large, fiat, polygonous, each with a strong keel; six loreal rows, the scales composing the lower row next to the supralabials largest; one row of large keeled suboculars; 7-8 supralabials to under the center of the eye; temporals flat, with a low tubercle, all the scales of the sides of the head being more or less rugose or wrinkled; ear-opening rather small, upright, oval; back and Fig. 28. — Caudal crest of Anolis cuvierl (left), of A. gundlachi (center), and A. cristatellus (right). sides covered with uniform scales tuberculated or keeled, separated from each other by one or more rings of minute granules ; on the median line of the neck and back a series of about fifty triangular spines forming a saw-tooth ridge scarcely connected with the caudal crest; ventral scales about same size as dorsals, though more closely set, but not keeled or distinctly tuberculated except on the flanks; scales on chin and throat more elongate, distinctly keeled or tuberculated; scales on upper side of fore limbs larger than dorsals, juxtaposed or imbricate, keeled, be- coming larger and multicarinate toward the hand; scales on upper side of hind limb similar, though less sharply keeled ; scales on rmder side of femur slightly larger than ventrals, indistinctly tuberculate; digital expansion well developed, about thirty-three lamellae under second and third phalanges of the fourth toe; tail strongly compressed, basal half with a high fin-like crest supported by about fourteen bony *'rays," the HiCHMlDT, AMPHIBIANS OF PORTO RICO 79 elongations of the neural spines of the caudal vertebrge; scales covering sides of tail flat, keeled, those on the fin between the "rays" elongate, three to four rows between rays, about fourteen longitudinal rows on side of tail at the level of the fiftli ray from the base; gular appendage very large, with distant rows of small tuberculate scales on the naked skin, the edge being rounded, thickened, and scaly; large postanal plates."' Female specimens lack the caudal "fin" and have a smaller dewlap with the scales set more closely. Stejneger's description of the colora- tion in life is as follows : "Iris hazel, with a bright brassy ring bordering the pupil; general color above greenish gray; back clouded with brownish and sides with blackish dots, the dusky of the back and the black spots on the sides ar- ranged in four perceptible, though indistinct, cross-bands; eyelids black- ish, with a citron-yellowish spot above and behind the eye and a smaller one in front ; under the eye a long semilunar white spot barely invading the posterior supralabials; several whitish spots on temples and sides of neck; underside white with dark-gray mottlings and spots; dewlap delicately IsTaples-yellow, scales on the edge white; legs indistinctly erossbarred with dusky bands more or less spotted with blackish. Tongue pale cadmium orange, whole interior of mouth of same color, but duller." Stejneger records only a single specimen as being emerald gi-een, while ten of the eleven collected by me were a uniform green. The eleventh was gray mottled with brown and black as described above. It lay closely pressed to a small branch of which it seemed an integral part until the continued pointing of my small boy assistant enabled me to distinguish it. As a protective coloration this pattern is an extraor- dinary success. Color-change is evidently very complete in this species. There is little variation in the Porto Rico Survey series. In A. M. N. H. No. 13234 the tail crest is unusually high, fully as high as in A. ricordii of Hispaniola, but the scale characters which distinguish cuvieri from ricordii are perfectly constant. The measurements of the largest and smallest specimens are as follows : Parts measured t^\f-.?- ^■^'■^■^■ No. 13236 No. 13138 Total length 418 mm. 304 mm. ^oI. N. II. riuts uieasmed No. 132:21 No. 12908 Total length 43 mm. 18(i mm. B(xl.y IG 65 Tail 27 121 Length of head G 20 Breadth of head 4 14 Arm 8 32 Leg 14 56 Eecently hatched specimens have a crossbanded coloration. In No. 13221. hatched in captivity, the dorsal ground color is gray, crossed by a sharp transverse chocolate-colored band, with a wavy margin over the eyes ; a horseshoe-shaped band from the upper posterior corner of the eye over the occiput behind the occipital ; a diagonal streak of the same color from the lower posterior corner of the eye to the shoulder ; eyelid with seven radial bands (including those above mentioned) ; five dark cross- bands on the back to the base of the tail, widest at the sides; tail with about eight crossbands. This pattern is evidently the foundation of the crossbanded phase of the adult. Females nearly always have a broad light mid-dorsal band. Stejneger describes the coloration in life as follows : "Iris dark brown ; edge of eyelids light yellowish ; general color above bronzy greenish gray; head and several faint longitudinal irregular spots on the sides of the back more brownish; on each side of the median dorsal line between the insertion of the hind legs a better de- fined and larger spot of irregular outline, pale brownish edged with brownish black and a light line outside the dark margin; on the middle line of the tail a series of dusky spots located at the base of' the largest spines ; throat whitish ; rest of underside suffused with greenish yellow, most intensely in the preanal region; dewlap greenish yellow verging into brownish orange toward the edge." Habits. — It is evident that the differences in distribution between this species and A. gundlaclii are not due to altitude preference, as supposed by Stejneger, but to habitat conditions, of which light seems to be one of the determining factors, A. cristatellus being the species of open fields and roadsides, A. gundlaclii of the thickly planted coffee plantations and of the forests. This species is the one most frequently seen on fence SCHMIDT, AMPHIBIANS OF PORTO RICO 85 posts, where it rests head down on tlie shady side, usually a single speci- men to a post. We have it from the deforested hills above Maricao, taken at an altitude of 2500 feet. . .- The examination of 100 stomachs yields the following information a.s to food habits: empty, 23; imidentified insect remains, 15; beetle re-- mains, 20 (larvae and adults; a species of weevil, Diaprepcs, very abun- dant) ; Orthoptera, 1(3 (cockroaches, grasshoppers, a single cricket and a mantis); ants, 10; caterpillars, 9; bugs, 5 (mostly Heteroptera ; one large cicada); flies, 3; spiders, 3; vegetable matter, 9 (mostly bright- colored seeds) ; vertebrates, 2 (AnoKs sp.). Wolcott, 1924, p. 27, giv^srJi more detailed account of the food of this lizard. ■•3?:-. The eggs are two or three in number, about 10 x 6 mm., uniforflify oval, the surface white and striate. They are frequently found und^r th(i' edo-es of logs or stones, or in debris about the base of banana plants. ''' •: Anolis guiidlachi Peters '>'•■' Text Figs. 27 and 28 Anolis gundlachi Teters, 1876, Mouatsber. Akad. Wiss. Berlin, p. 705.— Gund- lach, 1881, Anales Soc. Espafi. Hist. Nat., A^ol., X, p. .308.— Stahl, 1882. Fauua Puerto-Rico, p. 69, p. 159.— Stejiieger, 1904, Kept. U. S. Nation. Mus., 1902, p. 633, Figs. 89-91.— Barbour, 1914, Mem. Mus. Comp. Zool., Vol. XLIV, p. 273.— Fowler, 1918, Papers Dept. Marine Biol., Carnegie Inst., Vol. XII, p. 9.— Schmidt, 1920, Ann. N. Y. Acad. Sci., Vol. XXVIII, p. 188.— Wolcott, 1924, Journ. Dept. Agric. Pto. Rico, Vol. VII, p. 33. Anolis gundlacUi Bonlenger, 1885, Cat. Lizards, Brit. Mus., Vol. II, p. 25.— Garman, 1887, Bull, Essex Inst., Vol. XIX, p. 27. Type lomlity. — Utuado, Porto Eico. Dts^ri&ufw/i.— Recorded from Adjuntas, Aibonito, 18 km. south of Ciales, between Lares and Eio Blanco, Maricao, Santa Catalina, Utuado and El Yunque. The parallel with A. Tcrugi is emphasized by the fact that both species are strictly confined to Porto Eico, Avhile both pulchellus and cristatellus range widely through, ihe Virgin Islands. Specimens collected..— AS : Adjuntas, Aibonito, Maricao and El Yunque. Duignosis.—BoYsal scales granular, like laterals; ventrals larger, im- bricate and keeled; supraocular semicircles separated by two rows of scales, widely separated from the occipital; tail of male with a high fin-like crest. Original description. — ^^'Ventral scales convex or weakly keeled, lateral and dorsal scales small, granular, those of the two median dorsal and 8(j SCIENTIFIC SURVEY OF PORTO RICO nuchal rows enlarged and forming a low double keel ; a high fin on the basal half of the tail, supported by the dorsal vertebral processes, con- tinuous with a fold of skin beginning on the neck. The supraorbital semicircles separated by two or three scale rows. Supraocular area com- posed of fifteen to seventeen keeled scales. The supraocular semicircles are continued to the middle of the snout as sharply diverging supra- rostral keels. Scales of the upper surface of the snout polygonal, mostly keeled. Nostrils lateral. Distance of the ear-opening from the eye little less than the length of the snout. Occipital large, somewhat smaller than the ear-opening, separated from the semicircles by four or five rows of small flat scales. Length of head to ear-opening as long as the tibia. Throat-fan moderately large, covered with widely separated, rather large keeled scales. Lower and posterior faces of upper arm and thigh finely granular. Tail with larger scales arranged in rings, and the tail-fin covered with similar scales. "Grayish green, with small black spots forming three or four irregu- lar cross-bands on the body; a few rounded dark-edged, white spots on the sides of the body and the base of the tail. Snout and tail-fin black- ish-green, each scale with a yellowish green spot. Scales of the throat- fan lemon-yellow, its skin blackish; belly and under surface of head greenish yellow." Remarks. — Stejneger remarks on the discrepancies between Peters' description and his own, but these are due in part to variation and in part to a different method of description. The ventrals are normally sharply keeled. The occipital may be separated from the supraocular semicircles by as many as nine small scales. The keeled scales of the supraorbital disk are normally somewhat fewer than in Peters' specimens, ten in Stejneger's description. The coloration in life is described by Stejneger as follows : "General color dark olive above, with five wide lateral nearly black cross bands, which barely meet on the median line, while on the .sides they are very close together, being only separated by an oblique series of small yellowish spots ; a wide postocular blackish-brown band passes above the ear and joins its fellow of the other side on the back of the neck ; top of head densely marbled with indistinct spots of brown edged with dusky ; edge of eyelids, semicircular line formed by the keels of the suboculars, as well as alternating spots on the supralabial sutures lemon- ^'ellow; underside dull olive-yellow, chin bright lemon-yellow, the entire under surface densely marbled with blackish; underside of limbs similar, but paler ; liml)S-iibove cross-barred olive and blackish, like back ; SCHMIDT, AMI'HIBIANS OF PORTO RICO 87 tail similarly crossbarred, but slightly browner in the basal half or a little beyond the compressed elevated portion, followed by a median uni- form blackish portion and a terminal part which is uniform pale brown- ish olive; feet nearly uniform dusky; dewlap very large, wath thickened edge, the color of the skin being a dull orange-olive, the distant scales straw yellow; iris blackish brown; tongue plumbeous." This species is very distinct from A. cristateUus, but obviously directly related to that species. Its range and habitat are much more restricted, and the amount of variation is accordingly smaller. In the Survey of Porto Rico Series, the height of the tail crest (at its highest point) reaches a maximum of three times the diameter of the tail at the same point. The measurements of an adult male and female and of the type are as follows : A. M. N. H. A. M. X. H. Parts measured No. 131-_'6 d No. 13029 ? Type cT Total length 161 mm. 121 mm. 165 mm. Body 13.5 16.5 Breadth of head 11.5 8.5 — Arm 30 21 30 Leg 51 36 50 Habits. — Beyond the interesting facts concerning its habitat prefer- ence, nothing is known of the habits of this species. It is most abundant in the coffee-belt and in the higher forested areas, but reaches the coastal plain at Utuado (Stejneger, 1904, p. 537) and Arecibo (Fowler, 1918, p. 9). It replaces Anolis crlstatellus m the more shaded situations, and these two species thus form a dovetailing pair exactly like Anolis pulchel- lus ami Anolis hnigi. Stejneger records an egg-measurement as 11 by 5 mm. Anolis stratulus Cope Lagarti.ia manchada Text Fig. 27 Anolis striatulus (misprint) Cope, 1861, Proc. Acad. Nat. Sci. Phila., p. 209.— Garman, 1887, Bull. Essex Inst., Vol. XIX, p. 29. Anolis stratulus Reinhardt and Luetken, 1863, Vid. Meddel. Naturh. Foren., Copenhagen, 1862, p. 255.— Bocourt, 1873, Miss. Sci. Mex., Zool. Kept., Livr. 2, PI. 14, Fig. 11. — Peters, 1876, Monatsber. Akad. Wiss. Berlin, 1876, p. 706.— Gundlach, 1881.— Anales Soc. Espafi. Hist. Nat.. Vol. X, p. 310.— Stahl, 1882. Fauna Puerto-Rico, p. 69, p. 159.— Boulenger, 1885, Cat. Lizards Brit. Mus., Vol. II, p. 27.— Meerwarth, 1901, Mitt. Naturh. 88 SCIENTIFIC SURVEY OF PORTO RICO Mus. Hamburg, Vol. XVIII, p. 22.— Stejneger, 1904, Kept. U. S. Nation. Mus., 1902, p. G51, Figs. 105-107.— Barbour, 1914, Mem. Mus. Comp. Zool., Vol. XLIV, p. 274; 1917, Proc. Biol. Soc. Wash., Vol. XXX, p. 99.— Fowler, 1918, Papers Dept. Marine Biol., Carnegie Inst., Vol. XII, p. 11.^ Schmidt, 1920, Ann. N. Y. Acad. Sci., Vol. XXVIII. p. 188.— Smyth, 1920, Rev. Agric. Pto. Rico, Vol. IV, p. 18.— Danforth, 1925, Copeia, No. 147, p. 77.— Wolcott, 1924, Journ. Dept. Agric. Pto. Rico, Vol. VII, p. 24. Anolis dorsomaculatus Reinhardt and Luetken, 1863, Vid. Meddel. Naturh. Foren., Copenhagen, 1862, p. 255. Type locality. — St. Thomas, Virgin Island. Distribution.- — On Porto Rico proper this species is recorded from Adjuntas, Aibonito, Arroyo, Caguas, Catalina Plantation, Cayey, Coamo Springs, Condado, Enseuada, Humacao, Maricao/ Manati, Ponce, Rio Piedras, San Juan, Utuado and El Yunque. It is absent on Mona Island, but otherwise its range closely parallels that of Anolis cristatellus. It is recorded from Culebra and Vieques, St. Thomas, Tortola and Jost Van Dyke. Specimens collected. — 58 : Aibonito, Coamo Springs, Ensenada, Mari- cao. El Yunque, Vieques and Culebra islands. Diagnosis. — Dorsal scales juxtaposed, granular, like the laterals; one shield bordering the supraocular granules anteriorly between the super- ciliaries and the supraocular semicircle; width of head less than dis- tance from tip of snout to center of eye ; anterior femoral scales smooth, abruptly larger than the others; color brownish <' • Original description.- — ''^Size small; form elongate." ' -Head rather elongate, depres^eid, inuch as in .4. alligator Dura.''Bibr. 'Tail once and two-thirds the length of the body, moderately compressed, weakly verti- cillate, irregularly serrate. No dorsal dermal fold; an imperfect fold upon the nape, where two or three rows of scales appear to be a little larger than those upon the dorsal and lateral regions of the body. Anterior femoral and anti-brachial scales large, smooth, similar to those of the belly. Superior humeral, antibrachial, femoral and tibial similar to those of the back. Occipital shield separated from the superciliaries by small scales; the latter usually in contact medially, four or five in number on each side. .Palpebral disc rather round in outline, composed of nine smooth scales. - Facial rugae weak, soon obsolete, covered by three scales anterior to the last superciliary. The space between these as far as the end of the muzzle, covered with small smooth scales. Rostral plate bordered by five scales, the median one fitting into an emargination be- tween two mucronatious.. , Nostrils lateral. Canthus rostralis slightly SCHMIDT, AMPHIBIANS OF PORTO RICO 89 concave, very obtuse anteriorly. Superior labials eight. Loreal rows five. Anterior half of inferior labials in contact with an infedor series of plates, which are longer than broad, the anterior smaller than the first inferior labial. Goitre rather large. Two or three small plates be- hind the vent; scales at the base of the tail smooth. Extended posterior extremity not reaching beyond the posterior border of orbit. Total length 4 in. 7 lin.; tail 2 in. 11 lin. ; head from shoulder 8 lin. "From alcoholic specimens it appears that the color is greenish gray above, with very numerous darker marblings. The head and chin are darker. The medial dorsal line is crossed by four deep brown spots bor- dered with white. The anterior of these, on the interscapular region, is narrow and more transverse. There is a fifth spot at the base of the tail. The latter is clouded with brown superiorly, and the extremities are cross-barred with the same. Thighs dark, varied posteriorly. Goitre red-orange, abdomen greenish, femora and vent golden." Remarks. — Stejneger found that about half of this specimens had the supraocular semicircles in contact, and half separated by a single scale- row. In our series of twenty-six specimens the majority have the semi- circles in contact. One (A. M. N. H. No. 13282) has only a single row of scales between the occipital and the semicircles. In recently hatched specimens the dorsal markings are invariably very distinct. Stejneger describes the color of a male specimen in life as follows : "Iris dark brown; general color above light yellowish gray, much lighter below ; the saddle-shaped spots on back very pronounced blackish brown bordered by whitish ; on sides an irregular series of burnt-umber brown spots, also with white margins; throat and adjacent portions of underside of neck of a delicate pale bluish green; skin of the dewlap deep orange, the distant scales canary yellow, those on anterior edge more whitish." Habits. — Anolis stratulus, while frequently found in the same sitaa- tions as Anolis cristatellus, is noticeably more frequent on trees. .In Vieques and Culebra it was found only in trees, and at Coamo Springs and Ensenada it was much more common on the Ceiba or Almacigo trees than on the fence posts which are the chief resort of A. cristatellus. An examination of twenty-five stomachs indicates that ants form a much larger proportion of the food than in A. cristatellus. The contents are classified as folloM^s : empty, 3 ; unidentifiable insect remains. 4 ; ant remains, 12 ; beetle remains, 5 ; spiders, 2 ; cockroach, 1 ; earwig, 1 ; flies, 1; lizard skin (doubtless its own), 1. Wolcott (1924, p. 24) gives a more detailed analvsis of the stomach contents of this species. 90 SCIENTIFIC SURVEY OF PORTO RICO Anolis evennanni Stejneger Lagartija vei-de. Green Anolis Text Fig. 27 Anolis evennanni Stejneger, 1904, Rept. U. S. Nat. Mus., 1902, p. 647, Figs. 102-104.— Barbour, 1914, Mem. Mus. Comp, Zool., Vol. XLIV, p. 284.— Schmidt, 1920. Ann. N. Y. Acad. Sci., Vol. XXVIII, p. 188.— Smyth, 1920, Rev. Agric. Pto. Ric. Vol. IV. p. 18.— Wolcott, 1924, Journ. Dept. Agric. Pto. Rico, Vol. VII, p. 34. Type locality. — Catalina Plantation, east slope of El Yunque, Porto Rico, 890 feet altitude. Distribution. — Confined to northeastern Porto Eico, and usually to the more humid higher altitudes. Occasional on the northern coastal plain. Recorded from Adjuntas, Aibonito, Jajome Alto, Maricao, Utuado and on El Yunque from 890 feet to 2978 feet. Specimens collected. — 37 : Adjuntas, Aibonito, Maricao and El Yunque. Diagnosis. — Dorsal scales granular or tuberculated, juxtaposed, slightly larger than the laterals, all much smaller than the ventrals; superciliary ridge not defined on the posterior half of the eyelid ; a single shield between the superciliaries and the supraorbital semicircle border- ing the supraocular granules anteriorly; width of head as great or greater than distance from tip of snout to center of eye ; anterior femoral scales keeled, gradually diminishing; usual color in life vivid green. Original description.'^' — "Top of head with two diverging frontal ridges, disappearing before reaching the nostrils, and inclosing a frontal hollow ; head scales keeled or wrinkled ; rostral low, much narrower than the mentals ; seven narrow scales in a row between the nostrils ; one shield of each supraocular semicircle in contact, the others separated by one scale; occipital somewhat smaller than ear-opening, separated from the supraocular semicircles by three or four rows of scales ; supraocular disk consisting of ten or twelve polygonal keeled shields, separated from the semicircle by one row of granules ; one large shield in front of the supra- ocular granules between the superciliaries and the supraocular semi- circle; canthus rostralis sharp, consisting of five elongated shields in- creasing gradually in size posteriorly, superciliary ridge consisting of one narrow elongated shield and one similar l)ut very small one, but not followed by a differentiated series of small scales, the granules of the supraocular disk continuing uninterruptedly into the granules surround- * U. S. N. M. No. 26855 ; Catalina plantation, about 890 feet altitude ; February 21, 1900. aVHXJIDT, AMJ'HJBIANS OF PORTO RICO 91 ing the eye ; loreal rows five or six ; subocular semicircle keeled, broadly in contact with the supralabials ; supralabials nine, the suture between seventh and eighth being under the center of the eye; temporal gran- ules about the size of dorsals, a well-marked double series of small scales forming the supratemporal line; dorsals coarse, keeled granules, with- out any median enlarged series, laterals smaller but similar; ventral scales rather small, slightly imbricate, rounded behind, flat, those on the throat granular; fore legs above with small keeled scales, about three series on the anterior face of the lower arm being greatly enlarged, more than twice as large as the ventrals ; anterior scales of femur enlarged, keeled, gradually diminishing posteriorly and below; scales covering hands and feet above multicarinate ; digital expansion wide, about twenty-eight lamellje under phalanges ii and iii of fourth toe; tail moderate, slightly compressed, with fairly well-marked verticils, every eighth or ninth vertical row being somewhat enlarged and sur- mounted by a strongly serrated edge of enlarged triangular spines, the fourth or fifth corresponding to the enlarged vertical scale row being larger than the others; dewlap naked, with distant series of scales, edge not thickened; postanal scales slightly developed. "The dermal folds on upper neck and back are very low, especially the latter, but there is a distinct depression between them on the shoulder region/' Remarks. — This species is so adequately characterized in Stejneger's description that little remains to be added from an examination of the 37 specimens in the collection made by the Survey of Porto Eico. As in A. stratidus, the scale between the supraciliaries and supra- orbital semicircles, anterior to the supraorbital granules, is remarkably constant. It is double on one side in only 1 specimen out of 31 ex- amined. The semicircles may be broadly in contact (3 specimens), nar- rowly in contact (9), or separated by a single row of scales (19). The scales between the semicircles and the occipital vary from two to four, as in Stejneger's series. The coloration in life has been described by Stejneger as follows : "Iris, dark brown ; eyelids, abruptly flesh-colored ; general color, bright emerald green without markings ; abdomen, underside of hind legs, and thick basal portion of tail below, pale glaucous green; ter- minal third of tail, black, tip, pale; dewlap, gamboge yellow; scales, pale yellow, no thickened edge. "When handled the animal changed from green to Avax-yellow with numerous dusky spots and marblings on body and crossbars on tail, 92 SCIENTIFIC SURVEY OF PORTO RICO as well as longitudinal dusky stripes on throat; when reassuming its normal color the dusky markings disappeared before it turned green." The measurements of a representative of each sex are as follows : A. M. N. U. A. M. N. H. Parts measured ^^ 132i8 ^ No. 13388 ? Total length 175 mm. 114 mm. Body 70 45 Tail 105 60 Length of head 21 1.3 Breadth of head 13 7.5 Arm 32 20 Leg 51 33 Habits. — Beyond the fact that it is more or less restricted to the more humid higher parts of Porto Rico by habitat preference, almost nothing is known of the habits of this species. Specimens of this species were extremely abundant on the slopes of El Yunque, and specimens were taken from within the Forester's Cabin. Elsewhere, this species was one of the less common forms, found more often on larger trees, and at a considerable height from the ground. The results of the examination of the contents of twenty stomachs are as follows: empty, 3; beetle remains, 11; wasps, 2; ants, 1; caterpillars, 1; spiders, 1; skin of Anolis, 2 (doubtless their own) ; juvenile Anolis evermanni, 1. Anolis piilrhelliis Dumeril and Bibron Lagartija Rayon Text Figs. 27 and 29 Anolis pulchellus Dumeril and Bibron, 1837, Erpet. Gen., Vol. IV, p. 97. — Dumeril, 1851, Cat. Method. Kept. Mus. Paris, Vol. I, p. 56.— Reinhardt and Luetken, 1863, Vid. Meddel. Naturh. Foren., Copenhagen, 1862, p. 257.— Bocourt, 1874, Miss. Sci. Mex., Zool. Rept., Livr. 3, PI. 16, Fig. 28.— Peters, 1876, Monatsber. Akad. Wiss. Berlin, 1876, p. 706.— Gundlach, 1881, Anales Soc. Espaii. Hist. Nat., Vol. X, p. 310.— Stahl, 1882, Fauna Puerto-Rico, p. 69, p. 159. — Boulenger, 1885, Cat. Lizards Brit. Mus., Vol. II, p. 67.— Carman, 1887, Bull. Essex Inst., Vol. XIX, p. 48.— Meer- warth, 1901, Mitt. Naturh. Mus. Hamburg, Vol. XVIII. p. 25.— Stejneger, 1904, Rept. U. S. Nation. Mus., 1902, p. 660, Figs. 112-116.— Barbour, 1914, Mem. Mus. Comp. Zool., Vol. XLIV, p. 295; 1917, Proc. Biol. Soc, Wash., Vol. XXX, p. 99.— Fowler, 1918, Papers Dept. Marine Biol., Car- negie Inst., Vol. XII, p. 12.— Schmidt, 1920, Ann. N. Y. Acad. Sci., Vol. XXVIII, p. 189.— Smyth, 1920, Rev. Agric. Pto. Rico, Vol. IV, p. 17.— Danforth, 1925, Copeia, No. 147, p. 78.— Wolcott, 1924, Journ. Dept. Agric. Pto. Rico, Vol. VH, p. 15. BCHMIDT, AMPHIBIANS OF PORTO RICO 93 Type locality.— MsLvtiniqwe (erroneously) . Distribution. — If the localities of Stejneger's list be combined with those of the Porto Eico Survey, and those given by Wolcott, this species is known from Aibonito, Aiiasco, Boqueron, Caguas, Camuy, Catano, Cataxo, Coamo Springs, Ensenada, Hucares, Juncos, Mameyes, Manati, Maricao, Mayagliez, Ponce, San Antonio, San German, San Juan, San- turce, Toa Baja and Utuado. Anolis p'ldchellus is recorded from nearly all of the Virgin Islands, including Anegada and St. Croix. Except for its absence from Mona Island, it has therefore the same distribution as A7iolis cnstatellus. Specimens collected.— 'i^ '. Aibonito, Cataiio, Coamo Springs, Ensen- ada, Maricao, Mayagiiez, Santurce, Culebra and Vieques Islands. Diagnosis. — Dorsal scales large, flat, keeled and imbricate, like the ventrals; the lateral scales granular; width of head about half the dis- tance from snout to ear-opening; dorsal scales gradually increasing in size from the laterals ; skin of throat-fan crimson. Original description. — "The length of the head is double its width behind, which is little more than the height of the occiput; the head, viewed from above, has the form of an isosceles triangle; the upper surface of the head slopes forward, but is not a plane ; a swelling between the nostrils borders a hollow which extends to the tip of the snout; this hollow is bordered on either side by a rounded ridge representing the continuation of the supraorbital ridge, outside of which is a longitudinal groove; the nostrils, suboval, and directed obliquely backward, are situ- ated at each side of the tip of the muzzle ; each nostril is bordered above by an oblong arched scale in contact with the rostral, with a similar scale below, and three or four small granules behind it. The canthus rostralis is well-marked, continuous with the superciliary border ; all the upper head scales are slightly keeled ; six or eight equal, irregularly poly- gonal scales on the tip of the snout, disposed in pairs ; behind these are four to six without any symmetrical arrangement; of these the middle one is perhaps the largest; a double row of several-sided small plates paves the frontal hollow ; a single scale separates the supraorbital ridges ; the polygonal scales which form these two ridges are twice as large as the other head shields; each supraocular area exhibits a disk of eight or nine small polygonal scales, the rest of it being covered with small granular scales; the triangular loreal region is slightly depressed; its scales are quadrilateral, disposed in three longitudinal rows; the rostral is wider than high, its lower border straight, its upper border arched; the two mental scales are equilateral, three sided : one counts six oblong 94 SCIENTIFIC SURVEY OF PORTO RICO scales on each lip; the posterior part of the head is covered with small swollen and keeled scales in the midst of which is the enlarged circular occipital scale; the temples are granular, as are the eyelids, which have a double row of small tubercles on their edges ; the tympanic membrane Ventrals. Yentrals. Ventrals. Laterals. Laterals. Laterals. Middle of back. Middle of back. Middle of back. I -aterals. Ventrals. Laterals. Ventrals. Laterals. Ventrals. Fig. 29.— Dorsal scales of Porto Rican Anoles related to Anolis pulchellus. Left to right: A. krugi, A. pulcheUus, and A. poncensis. tiVHMWT, AMI'HIBIANS OF PORTO RICO 95 is set a little within the auricular hollow, which is sub-oval and entirely without denticulation ; the skin forms a large throat-fan, which extends as far as the middle of the breast ; the body is compressed, the back some- what roof-shaped ; laid along the body the fore-limbs reach almost to the groin, the hind-limbs to the back of the eye ; the tail is compressed enough to have a sharp upper edge ; it is a third longer than head and body, with a low denticulated crest its entire length; the upper and lateral sides of the tail are covered with small granular scales ; the scales of the upper part of the back are imbricate and appear circular, though actually polygonal; they are in about twenty-four longitudinal rows; they are not positively keeled, but one might say they have their sides depressed and center raised ; the sides of the body are covered with small, smooth, imbricate suboval or subcircular scales, smaller than those of back and l)elly ; the under side of the head is covered with rounded, thickened scales, subimbricate, and smooth or feebly keeled; scales of the throat fan rather large, keeled, and rhomboidal ; the ventral scales are imbricate, slightly keeled, and lozenge-shaped, with rounded angles; the limbs are covered with sub-hexagonal keeled scales ; the thighs are granular : rhomboidal scales, strongly keeled, and arranged in longi- tudinal rows extend the whole length of the tail." Remarks. — Stejneger comments on the vagueness of this original description, but there can be little doubt that it is correctly assigned to this most slenderly-built of the Porto Eican Anoles. The number of loreal scales in a vertical row is usually four (five or six in A. krugi) ; in 85 specimens, 69 have four loreal rows, 15 have five, 1 has six. The scales separating the occipital from the supraorbital semicircles number one in 1 specimen, two in 29 specimens, three in 51 specimens, four in 4 specimens. The semicircles are in contact in 17 specimens, separated by one scale row in 66 specimens, by two scale rows in 2 specimens. I find no important variation fn my series of 17 specimens from Culebra Island and 8 from Vieques. The measurements of a specimen of each sex are as follows : A. M. N. H. A. M. N. H. No. 13988 d No. 13186 § Length 182 mm. 136 mm. Body ... 47 36 Length of heart 15.5 11.5 Breadth of head 8 6 Foreleg 1» 14 Hind leg 33 27 Color. — Stejneger describes the variation in coloration of this species. 96 SCIENTIFIC SURVEY OF PORTO RICO An adult male in life was colored as follows: "Iris dark brown; upper surface dull clay-colored, more dusky along the median line; head darker; more brownish; from eye to half way down the side of neck a broad black line, and another on the edge of the lower lip; a third blackish line, but considerably fainter, on lower edge of mandible, being more distinct between ear and shoulder; flanks and underside Naples- yellow, a stripe on upper labials over ear to shoulder more primrose-yel- low; on flanks a series of oblique, elongated spots of brightest gamboge- yellow narrowly margined with black; skin of dewlap bright crimson anteriorly verging into dark rose-pink, posteriorly into orange, the dis- tant scales arranged in rows and colored gamboge-yellow." Habits.- — Little is known of the habits of this species. I have seen juvenile specimens perched at the extreme ends of dry branches, where they were perhaps waiting for small insects to alight. Stejneger supposed that Anolis pulchellus was confined to the coastal- plain area, rarely going above 500 feet in altitude. In the course of the present survey, it was found to be everywhere abundant, up to an altitude of at least 2,000 feet, but strictly confined to open fields. It is usually associated with Anolis cristatellus, but probably does not as a rule range so high. Wolcott gives a detailed analysis of the food of this species. In two series ants formed, respectively, 11.2 per cent and 20 per cent of the total. Only Anolis stratulus eats a larger proportion of ants. Anolis krugi Peters Text Fig. 29 Anolis krugi Peters, 1876, Monatsber. Akad. Wiss. Berlin, 1876, p. 707. — Gundlaeh, 1881, Anales Soc. Espau. Hist. Nat., Vol. X, p. 310.— Stahl, 1882, Fauna Puerto-Rico, p. 69, p. 159.— Boulenger, 188.5, Cat. Lizards Brit. Mus., Vol. II, p. .37.— Stejneger, 1904, Kept. U. S. Nation. Mus., 1902, p. 655, Figs. 108-111.— Barbour, 1914, Mem. Mus. Comp. Zool.. Vol. XLIV, p. 284.— Schmidt, 1920, Ann. N. Y. Acad. Sci., Vol. XXVIII, p. 190.— Danforth, 1925, Copeia, No. 147, p. 78.— Wolcott, 1924, Journ. Dept. Agric. Pto. Rico, Vol. VII, p. 22. Anolis hrugii [misprint] Smyth, 1920, Rev. Agric. Pto. Rico, Vol. IV, p. 18. Type locality. — Porto Kico. Distribution.— This species is confined to Porto Rico. It has been recorded from Ad juntas, Aibonito, Catalina Plantation, Cayey, Ciales, Coamo Springs, Lares, Mameyes, Maricao, Utuado and El Yunque. I have examined the specimens from Ensenada referred to Jcrugi by Fowler and find them to be pulchellus. SCHMIDT, AMPHIBIANS OF PORTO RICO 97 Speci»i€ns collected. — 62 : Adjuntas, Aiboiiito, Coamo Springs, Maricao and El Yunque. Diagnosis. — Dorsal scales imbricate, keeled, like ventrals; lateral scales granular; width of head much more than half the distance from tip of snout to ear-opening; loreal scales in five to seven rows; skin of throat-fan in male orange. Original description. — "Ventral scales strongly keeled, lateral and dorsal scales granular, with the exception of four median distinctly keeled rows which appear on a longitudinal dermal fold. A row of large, hexagonal keeled scales on the upper side of the tail. ''The supraorbital semicircles almost completely separated by a row of intermediate scales. The supraorbital ridges extend on the snout as two keels placed twice as far apart as their distance from the canthus rostralis. Supraocular area composed of four to six larger keeled scales and a few smaller ones. The distinct occipital is larger than the trans- verse ear-opening and separated from the semicircles by two or three rows of scales. The loreal area has five rows of scales at its middle. Length of head to ear-opening a little longer than the tibia. "Olive green; punctate and vermiculate with black on the back and on the sides below a lateral yellow longitudinal line ; white dots on the neck ; ventral surface greenish yellow." Remarks. — Peter's description requires modification principally with reference to the rows of keeled scales on the middle of the back, which are six rather than four, with two or three rows of small keeled scales adjoining them, so that the transition to the granular laterals is not quite so sharp as would seem to be indicated. Among sixty specimens the number of loreal scales in a vertical row is as follows : four in 1 specimen, five in 34, six in 23, seven in 2. The number of scales between the occipital and the supraorbital semicircles varies from one to six, one in 1 specimen, two in 18, three in 25, four in 13, five in 2, six in 1. The supraorbital semicircles are in contact in 2 specimens, separated by a single scale row in 34, by two scale rows in 19, by three in 5. ' Color.— The coloration of an adult male in life is described by Stej- neger as follows : "General color bright yellowish olive-green, sides of back and flanks with minute black spots, larger on back, but none along the median area occupied by the enlarged scales ; from under eye through ear to groin a broad and very distinct line of canary yellow ; brightest, nearly lemon vellow, on middle of flanks ; a black spot immediately behind eye, but 98 SCIENTIFIC SURVEY OF PORTO RICO no postocular band ; underside paler, more buffy ; immediately belo^v the lateral yellow band the color is more olive, with minute black specks; hind legs posteriorly suffused with ferruginous; tail crossbarred with dusky ; dewlap yellowish, gradually deepening to orange toward the edge ; eye dark brown, nearly black, with a faint silvery edge to the iris; eye- lids edged with whitish." This species is often difficult to distinguish from A. jmlchellus without direct comparison ; the color of the dewlap, which in life is orange in- stead of crimson, is distinctive. In alcoholic specimens of A. Tcrugi the narrower band of enlarged dorsal scales is the most satisfactory char- acter for separating it from A. pidchellus. Other characters are at best comparative, useful only for a series of specimens. The measurements of a representative specimen of each sex of .4. l-rugi and of the type are as follows : A. M. N. H. A. M. N. H. Parts measured No. 13360 d No. 13207 ? Type Length 203 mm. 134 mm. 170 mm. Body 51 36 46 Length of head 16 11 18 Breadth of head 10.5 6 — Foreleg 21 15 20 Hind leg 39 29 38 Habits. — Little is known of the habits of this species. Stejneger had the impression that Anolis krugi is characteristic of the intermediate altitudes, from 500 to 1500 feet. The specimens in the present series from Coamo Springs are from an altitude of less than 300 feet, while specimens from Aibonito reach an altitude of at least 2000 feet. The specimens from Coamo Springs supply the clue as to the determining factor in the distribution of the species, for at that locality A. hrugi was abundant among the ferns and vines of the moist dark gorge back of the bath houses, but was seen nowhere else. At Aibonito ^.nd Maricao Anolis pulchellus was noted on the bare hilltops or in open fields, while a few steps within the borders of the coffee plantations only .4. Tcrugi was to be found. Moisture and shade, therefore, are the habitat requirements of Anolis Tcrugi. Anolis cristatellus and Anolis gundlacTii have an ex- actly parallel distribution. The males of this species are inveterate fighters, and it is not un- common to find specimens with injured dewlaps and often with injured mouths, probably due to their habit of locking jaws. When the lizards are fighting, the dermal folds of the neck and back are raised to the highest degree. A pair so engaged on a tree at Maricao was quite ob- SCHMIDT, AMPHIBIANS OF PORTO RICO 99 livious of the approaching collector. They manoeuvered especially to gain the superior position and finally fell to the ground with jaws locked. Anolis poncensis Stejneger Lagartija Text Fig. 29 AnoUs poncensis Stejneger, 1904, Rept. U. S. Nation. Mus., 1902, p. 6.55, Figs. 117-121.— Barbour, 1914, Mem. Mus. Comp. Zool., A'ol. XLIV, p. 284.— Fowler, 1918, Papers Dept. Marine Biol., Carnegie lust.. Vol. XII, p. 12. AnoJis pensensis Schmidt. 1920, Ann. N. Y. Acad. Sci., Vol. XXVIII, p. 191 (misprint). Type locality. — Hills three miles east of Ponce. Distribution. Confined to Porto Eico, and on Porto Rico confined to the southwestern arid section. It is known from Coamo Springs, En- senada, Guanica and Ponce. Specimens collected. — 38: Coamo Springs and Ensenada. Diagnosis. — Dorsal scales imhricate, large, flat, keeled, much like the ventrals, which are very strongly keeled, the keels forming continuous ridges; laterals smaller than dorsals and ventrals, but imbricate and keeled, much larger than the laterals of Icrngi or pulchellus; dewlap small, white, covered with large keeled scales. Original descripiion. — "Top of head with two slightly curved and low frontal ridges; frontal hollow very shallow; head scales more or less wrinkled or keeled; four scales in a row between the nostrils; supra- ocular semicircles broadly in contact; occipital slightly larger than the ear-opening, separated from the supraocular semicircles by one row of flat scales; supraocular disk consisting of five or six polygonal keeled shields conspicuously larger than the surrounding scales, which are also keeled, and of which one row separates the disk from the semicircle ; two scales in front of the supraocular scales between the superciliaries and the supraocular semicircle; canthus rostralis consisting of four elongate narrow shi.elds, the first one very small, the third longest; the super- ciliary ridge consisting of an anterior long and narrow shield and a series of scales contrasting in size with the granules surroimding the eye, but not with the small scales of the supraorbital region; loreal rows three or four; subocular semicircle keeled, broadly in contact with the supralabials, not bending upward behind the orbit; supralabials seven, the suture between fifth and sixth being under the center of the eye; central temporals large granules; a well-developed double row of scales forming a supratemporal line ; dorsal scales rhomboidal, imbricate, 100 SCIENTIFIC .SURVEY OF PORTO RICO sharply keeled, the keels forming continuous parallel ridges, large, much larger than the laterals and nearly as large as the ventrals ; laterals simi- lar to the dorsals, but much smaller and less sharply keeled ; ventrals like the dorsals, only more pointed and slightly larger; scales on throat and chin similar, keeled, only considerably smaller, though larger than the laterals: arms and legs covered with similar imbricated, keeled scales nearly as large as the ventrals; hands and feet above with pluricarinate scales; digits long and slender, expansion moderate; 18 lamellge under phalanges ii and iii of fourth toe; tail very long, more than twice head and body, moderately compressed, covered with large keeled scales forming continuous ridges, with scarcely any indications of verticilla- tion, the upper edge being but faintly serrated ; dewlap entirely covered with imbricated, pointed, and keeled scales nearly as large as the ventrals, edge not thickened ; postanal plates very small." RemarJcs. — Anolis poncensis is a highly unique species, not only in its lepidosis, but in the extremely small size of its throat-fan, which is scarcely one-third as large as that of A. pulchellus or A. krugi when fully extended. There is little variation in the series of 38 specimens collected by myself. The loreal rows in a vertical line are three in 18 specimens, four in 20. The scales between the occipital and the supraorbital semi- circles are none in 2 specimens, one in 21 and two in 14. . The supra- orbital semicircles meet in 32 specimens and are separated by a single scale in 6. Thus the type comes much closer to being a fully normal specimen than was supposed by Stejneger from the variation in the 6 specimens before him. The females invariably have a broad mid-dorsal light band. Color of living specimens — "Ground color above drab verging on tawny-olive on the tail and strongly washed with cinnamon on the sides, middle portion of back about five scales wide, uniform without spots, but on the sides of back and on flanks there are three longitudinal series of dusky spots on each side, about seven spots in each series from axilla to groin; these spots are not permanent, but appear and disappear at intervals ; a pale supratemporal line, washed with pale rufous, from pos- terior edge of supraocular disk ; below this an elongate blackish spot in- volving the eye and part of loreal triangle strongly tinged with tawny on the latter and on temples; edge of eyelids deep rufous; below the dark spot a pure white line on the lower row of scales of loreal triangle, suboculars and lower temporals to above the ear; several oblique whitish lines, which proceeding from the throat join on side of neck under the * stejneger. 1904. p. 608. SCHMIDT, AMPHIBIANS OF PORTO RICO 101 ear, and a short line behind the shoukler form a lateral whitish stripe which disappears at the anterior thircl of the distance between shoulder and groin; a dusky line below the white one, involving the upper and lower labials and continued to a little beyond the lower edge of the ear ; a faint dusky stripe across upper arm and on side behind the axilla bor- dering the pale lateral neck stripe below; underside whitish, washed faintly with tawny, the throat with several longitudinal series of narrow, disconnected, dusky stripes; a faint dusky stripe along the median line of the belly; tail underneath whitish, strongly washed with tawny-olive, the pale color anteriorly extending upward on the sides of the tail so as to form a series of numerous pale crossbands which do not reach the median line above ; the posterior half of tail regularly barred with wide dark and pale rings: limbs above like the back, the hind limbs with indistinct dusky markings washed with rufous ; a small dewlap perfectly covered with large white scales, so as to entirely hide the skin imderneath, even when highly distended, the color of which, however, appears to be whitish; iris blackish brown." The measurements of a specimen of each sex are as folloAvs : A. M. X. H. A. M. N. H. rarts measured No. 13784 d" No. 13845 ? Lt'iigth 160 mm. 112 mm. Bodv ...V.V 44 40 Tail 116 72 Length of head 13.5 11 Breadth of head 8.0 6.5 Foreleg l'*^ 1"* Hind leg 31 27 Hahits. — This species was found associated with Anolis cristatellus and with an occasional A. pulchellus at both Coamo Springs and Ensenada. Broadly speaking, it replaces A. pulchellus in the southwest- ern part of the Porto Rico, inhabiting fences and grazing land much as .4. jiuchellus does in the remaining portion of the island. A few speci- mens were noted on the arid cactus-covered hilltops about Ensenada. Near Coamo Springs this species occurred in colonies, sometimes a mile or more apart. Nothing more is known of its habits. Cyclura Harlan The "Rock Iguanas'" of the West Indies are one of the most con- spicuous elements in their endemic fauna. They appear to me to be a very archaic group of lizards, veritable relics from the "Age of Reptiles." The genus is extinct on Porto Rico proper, but the bones of a fossil 102 SCIENTIFIC SURVEY OF PORTO RICO species are abundant in cave deposits. The fossil form {Cyclura portori- censis Barbour) will not be described here. Another extinct form {Cyclura mattea Miller) is known from St. Thomas, and the living C. pinguis from Anegada shows that the existence of these fossil species might have been predicted. Cyclura stejnegeri Barbour and Noble Rock Iguana Text Fig. 30 Metopoceros coniutus Meerwarth, 1901, Mitt. Naturh. Mus. Hamburg, Vol. XVIII, p 2G (part). Cijclum coniuta Stejneger. l!Kt4. Kept. U. S. Nation. Mus., 1902, p. 670, Figs 122-126.— Barbour, 1914, Mem. Mus. Comp. Zool., Vol. XLIV, p. 299 (part). Cyclura stejnegeri Barbour and Noble, 1916, Bull. Mus. Comp. Zool. Vol. LX. p. 163. PI. 12.— Schmidt, 1920, Ann. N. Y. Acad. Sci., Vol. XXVIII, p. 191 ; 1926, Publ. Field Mus. Nat. Hist., Zool., Vol. XII, p. 155. Type locality. — Mona Island. Distribution. — Confined to Mona Island. Specimens collected. — 2 : Mona Island. Diagnosis. — A large lizard with a dorsal crest of spine-like scales; an enlarged conical scale or low horn on the frontal region; other head scales variously enlarged; second and third toes with a curious comb- like scale. Original description. — "Very similar to C. cornuta from which it may be distinguished by the following characters : nasals in contact with the rostral ; two, and in part three rows of scales between the nasals. Pre- frontals separated from the enlarged median frontal scale by two rows of scales. A single large elongate canthal scale preceded by three small precanthals. Dorsal crest much reduced between the shoulders, abso- lutely interrupted on the rump, fifty-one scales in the crest from shoulder to rump. Limiting row of each verticil not much wider than the other rows of the verticils. Color somewhat faded, uniform dark olive-green." BemarJcs. — The original description is comparative, and for the present paper may be supplemented by Stejneger's earlier description, which refers to the same specimen later made the type by Barbour and Noble : — "Eostral wide, as wide as mental, broadly in contact wdth nasals; nasal large, ovoid, perforated by a large nostril of the same shape; on each side of the top of the snout, immediately behind and adjoining the nasal, a series of three large shields, strongly convex, the posterior pair particularly so, and almost keeled ; the series are separated by numerous SCHMIDT, AMPHIBIANS OF PORTO RICO 103 small scales anteriorly about three in a row, posteriorly four ; the anterior pairs subequal, the posterior one nearly as long as the two others together, those of each series broadly in contact without any intervening scales; separated from these prefrontal series by two rows of scales there is a large rounded, median, frontal shield, its center on a line with the an- terior edge of the orbit, convex and wrinkled radially from the center; supraocular semicircles evident, though the component keeled scales hardly exceed the similar scales which form the supraorbital disk ; semi- circles separated by about four rows of smaller keeled scales; occipital located well forward between the semicircles, from which it is separated fi'iG. 30. — Head of Cyclura stejnegeri (type). (From Stejneger.) by three rows of scales, on a line between the posterior borders of the orbits, smaller than the nasals ; one large keeled canthal scale nearest the orbit, the anterior ones biit slightly developed ; a well-developed series of strongly keeled suboculars continued backward as a supratympanic series to above the ear; ten supralabials, the suture between the last two under the center of tlie eye; a series of small scales separating the suboculars and the supralabials; above the angle of 'the mouth and in front of the lower edge of the ear a large tubercular shield and above it about the middle of the front edge of the ear another shield, convex and almost as large; tympanum elliptical, erect, large; eleven lower labials to the center of the eye; a series of enlarged malar scales, the posterior ones 104 SCIENTIFIC SURVEY OF PORTO RICO strongly keeled and separated from the lower labials bj' several rows of small scales; dorsal and ventral scales small, about eleven contained in the vertical diameter of the tympanum, rhomboidal, obliquely keeled, the keels pointing toward the median line; from the occiput along the median line of the neck and back a series of enlarged strongly keeled scales forming a low serrated crest, which is much reduced between the shoulders, absolutely interrupted on the rump, and consequently not continuous with the caudal crest ; length of the crest scales on the middle of the back three to the vertical diameter of the tympanum, 51 in the dorsal crest from shoulder to rump; throat covered with scales similar to the ventrals but smaller; sides and underside of neck with numerous folds, a large median one almost large enough to be called a dewlap, joining posteriorly a strong transverse fold; upper surface of limbs with slightly imbricated, keeled, posteriorly pointed scales, somewhat larger than the dorsals, on the lower arm about seven, on the tibia about four to the vertical diameter of the tympanum; a single series of about eighteen femoral pores; inner side of second toe with one 'comb,' of third toe with two 'combs' (see fig. 125) ; tail compressed, covered with obliquely keeled scales in vertical rows forming faintly indicated verticils, about four rows of the larger scales to a verticil ; tail surmounted by a crest of enlarged, pointed triangular scales forming a strongly serrated edge." The single very old male specimen collected by myself has the irregu- lar development of the large tubercular scales of the head characteristic of old individuals of this group. The nasal is separated from the rostal on one side by a space filled with very small scales, on the other by a large tubercular shield. The smaller head shields are widely separated by very small intervening scales. There are 3 prefrontals on one side, 4 on the other; 3 large tubercular shields in front of the tympanum on one side, 1 on the other. There is a large fleshy fold on the neck, sur- mounted by a low crest, which passes into the higher dorsal crest. There is an additional fleshy lobe on one side behind the ear and others grow out of the irregular subgular folds. The cheeks are enormously swollen below the angle of the jaws. The 21 low spines on the nuchal fold are followed by 51 higher dorsal spines, the highest 20 mm., and after an interspace on the rump there are 50 spines on the base of the tail gradually becoming smaller to a point where the height of the spine equals its length. The highest spines on the tail measure 21 mm. The frontal tubercle measures 14 mm. in width and 9 mm. in height. There are 3 rows of femoral pores, 18 in the anterior row on each side. A SCHMIDT, AMPHIBIANS OF PORTO RICO 105 third ''comb" is plainly distinguishable on the third toe. The scales of the reproduced tip of the tail are not arranged in verticils. The measurements of this old male are as follows : Parts measured A. M. N. H. No. 13775 Length 910 mm. (tail repi-oduced ) Body 470 Length of head 127 Breadth of head 79 Foreleg 173 Hind leg 260 In spite of the separation of the nasal shield from the rostral in this old individual, I have retained the name stejnegeri, as it may well be that, although the young of the three related species, cornuta, nigerrima and stejnegeri, are clearly distinguishable, in the adults the characters are obscured. In other respects this individual accords well with the previously described specimens from Mona. Additional material of cornuta, however, is required to establish satisfactorily the status of the forms on Mona and ISTavassa. Habits. — The Eock Iguana is confined to the table land on Mona Island, descending occasionally to the tillable area to feed on the young corn and cotton. The extreme rockiness of the habitat of this species is scarcely conceivable unless one has visited the region or the similar areas in Porto Eico, Santo Domingo and Cuba. The limestone is weathered into cup-like hollows bounded by knife edges and spear points ; the handfuls of soil in the hollows are largely occupied by cactus, and a single day's tramp demolishes a pair of shoes. The natives of the island who engage in hunting the wild goats, pigs and cattle, wear several pieces of pigskin strapped to the soles of their feet, and even so, their ability to thread their way through the cactus is little short of marvelous. The specimen taken was the only one seen by our party of seven men in an eleven hours' tramp, and was caught and wounded by the dogs. The native hunters report that the rock iguanas take refuge in the cracks and holes in the rocks, and that they are somewhat more abundant than would be inferred from our experience. Celestiis Gray Celestus pleii Dumeril and Bibron Culebra de cuatro patas Text Fig. 31 Diploglossus pleii Dumeril and Bibron, 1839, Erpetol. Gen., Vol. V, p. 605. — Dumeril, 1851, Cat. Method. Kept. Mus. Paris. Vol. I, p. 154. — Boulenger, 1885, Cat. Lizards Brit. Mus., Vol. IL p. 294 (part). - '- -. '., ..- '■V Li LI 106 SCIENTIFIC SURVEY OF PORTO RICO i^r^ OS to CO o o Celestus pleii Cope, 1868, Proc. Acad. Nat. Sci. Phila., p. 124.— Stejneger. 1904. Rept. U. S. Nation. Mus., 1902. p. 622, Figs. 74-79.— Barbour, 1914, Mem. Mus. Comp. Zool., Vol. XLIV, p. 304; 1919, Proc. New England Zool. Club, Vol. VII, p. 13.— Schmidt, 1920, Ann. N. Y. Acad. Sci., Vol. XXVIII, p. 192. DiplOfflos.su s (Celestas) pleil Bocourt, 1879, Miss. Sci. Mex., Zool., Livr. 6, p. 381, PI. 22, Fig. 4. Celestas deffener Cope, 1868, Proc. Acad. Nat. Sci. Phila., 1868, p. 124. Diplofflossiis plei Peters, 1876, Monatsber. Akad. Wiss. Berlin, 1876, p. 708— Gundlach, 1881, Anales Soc. Espan. Hist. Nat., Vol. X, p. 811.— Stahl, 1882, Fauna Puerto-Rico, p. 69. Type locality. — Martinique (erroneously). Distribution. — Confined to Porto Rico, where it has been recorded only from Ad- juntas, Aibonito, Catalina Plantation and Lares. The seven specimens secured at Aibonito all came from a single hilltop in a coffee plantation. Specimens collected. — 7 : Aibonito. Diagnosis. — An elongate lizard with short weak limbs, smooth skinklike scales, and with four median shields on the head, a very large frontal, broad prefrontal, narrow in- terparietal and small occipital. Original description. — "Nasal plates very small, entirely lateral; two pairs of. supra- nasals in contact; an octagonal inter-nasal, transversely widened; no f ronto-nasals ; six supra-oculars on each side; frontal large, subquadrangular, oblong; two very small f ronto-parietals, not in contact ; an inter- parietal in the form of an isosceles triangle; two oblong parietals; a triangular occipital with rounded posterior border; a very small freno-nasl (^postnasal) ; two loreals, the first higher tlian long; two freno-orbitals SCHMIDT, AMPHIBIAXS OP PORTO RICO 107 (^ preorbitals) ; a triangular suborbital; ear rather large, sub-circular, exposed, with entire border; body anguiform; limbs short, stout; tail C3^clo-tetragonal ; lower eyelid scaly; scales with six to eight striae; back with wavy brown markings on a fawn-colored ground, with a brown band on each side. ''This species is most closely allied to Diploglossus sagrae, but differs in having a much larger ear opening, by the more strongly marked and fewer striae of the dorsal scales, which do not exceed ten, while there are fifteen in that species." Remarhs. — There is no possibility of confusing this species with any other Porto Eican lizard. Stejneger gives a more exact description and effectively clears up the identity of the species, which was described from the Plee collection as coming from ]\Iartinique. He describes the color as follows : "Color above (living and in alcohol) walnut l)ruwii. with numerous more or less interrupted and anastomosing dusky cross bands which do not reach the lateral longitudinal band. The latter is of a dark brownish gray with a sharply defined crenelated upper edge, gradually fading into the pale color of the underside which is clay colored washed with orange ; lower lips and throat spotted with dark brownish gray." The series of seven specimens in the Porto Rico Survey Collection is so uniform in scale characters as to suggest that the individuals are members of a single family. The proportion of the length of the fore- limb to that of the body varies from .12 to .15. The scales ahout the body are 34 in one, 35 in one, 36 in four, 38 in one. The measurements of the largest specimen are as follows : A. M. N. H. Parts measured • ^O- ISl'^^ Length 210 mm. Body ^^ Length of head ^'^ Breadth of head ^^ Foreleg Hind leg 1*^ All of the specimens found by me were under logs in a coffee planta- tion. When they were uncovered, their attempt to escape consisted of burrowing in the loose leaf mold. When grasped, they squirm violently, and because of their very muscular and smooth body it is difficult to hold them. The two female specimens contain respectively one and three well- advanced embryos. The egg measures 18 by 11 mm. The completely 108 SCIENTIFIC SURVEY OF PORTO RICO formed embryo rests on a very large yolk mass. The head and legs of the embryo are proportionately larger than in the adult, while the tail is shorter, Teidae Ameiva Meyer The Ameivas are typical representatives of the family Teidae, readily distinguishable from all other Porto Rican forms in having the rectangu- lar plates of the belly arranged in both longitudinal and transverse rows, while the dorsal scales are minute and granular. Although this genus is well represented on the mainland (by Ameiva ameiva, A. hifrontata, A. undulata, A. f estiva and their allies), the West Indies have the maxi- mum number of species and also exhibit the greatest amount of differ- entiation. Hispaniola alone, including Navassa and Beata islands, has eight species. The Ameivas are apparently especially adapted to insular conditions, for on a number of the smaller islets they occur alone or almost alone (Sombrero, Eedonda). Synopsis of the Porto Rican Ameivas A. Ventral plates in eight rows; caudal scales oblique, smooth; fronto- parietals united a. wetmorel AA. Ventral plates in ten or twelve rows; caudal scales straight, keeled; fronto-parietals distinct. B. Generally white-spotted only on the posterior half of the back. .A. ecesul BB. Back with white spots to the neck A. albnguttata Ameiva wetmorei Stejneger Text Figs. 32 and 33 Ameiva wetmorel Stejneger, 1913, Proc. Biol. Soc. Wash., Vol. XXVI, p. 69.— Barbour, 1914, Bull. Mus. Comp. Zool., Vol. XLIV, p. 311.— Fowler. 1918. Papers Dept. Marine Biol., Carnegie Inst., Vol. XII, p. 12. Fig. 6, PI. 1.— Schmidt, 1920, Ann. N. Y. Acad. Sci., Vol. XXVIII, p. 193.— Danforth, 1925, Copeia, No. 147, p. 78. Type locality. — Above Rio Loco, Guanica, Porto Rico. Distribution. — Ameiva wetmorei appears to be confined to the region near Ensenada and to Caja de Muertos Island. It probably ranges westward toward Cabo Rojo and eastward toward Ponce on the lime- stone hills. It was secured on Caja de Muertos by H. E. Anthony in 1926. Ameiva lineolata, its relative in Hispaniola, appears to be simi- larly confined to the more arid parts of that island, and arid or semi- arid conditions prevail also on Great Inagua and St. Croix, each of SCHMIDT, AMPHIBIANS OF PORTO RICO 109 which is inhabited by a related species. These four species form a highly interesting group of Ameiva, characterized by the oblique scales of the tail, a distinctive habitus and a lineolate type of coloration. Specimens collected. — 30: Ensenada and Caja de Muertos Island. Fig. 32. — Ameiia tcetmorei. A. M. N. H. No. 13820. A. Head from above. B. Head from side. C. Head from below. D. Arm from in front. E. Posterior face of leg. F. Foot from above. G. Preanal scales. Tbree times natural size. Diagnosis. — An Ameiva with eight rows of ventral ])lates; caudal scales oblique, smooth ; fronto-parietals united ; seven distinct longitudi- nal light lines, the median beginning on the snout. Original description.* — "Nostril between the two nasals; anterior * Type from Oiianica, above Rio Loco. U. S. N. M. No. 49731. 110 SCIENTIFIC SURVEY OF PORTO RICO Mi Fig. 33.— A. M. Natural - Ameiva wetmorei. N. H. No. 13821. size. nasals broadly in contact behind rostral ; fronto-nasal broader than long, in contact with nasals, loreal, and prefrontals; pre- frontals pentagonal, broadly in contact; frontal pentagonal, in contact with first and second supraoculars, not touching the third; a single hexagonal fronto-parietal broadly in contact with the third and very narrowly with second supraocular; three occipitals, the outer two very large, squarish, the median one long and narrow, almost rectangular; five super- ciliaries; three supraoculars, the first in con- tact with the first superciliary, the others separated from the sunperciliaries by a single row of fine granules; loreal undivided; seven supralabials, first in contact with posterior nasal only, second with posterior nasal and loreal, third largest, fifth and sixth in con- tact with a long subocular ; temples with small flat irregular scales; mental followed by a large, unpaired postmental; six large in- fralabials, third largest; four pairs of chin- shields, first pair in contact, second pair half separated by granules of chin; between in- fralabials and chin-shields posteriorly a single line of flat scales, the two posterior ones large, the anterior small, not reaching first pair of chin-shields; chin and throat covered with small scales or granules diminishing in size posteriorly ; mesoptychium with a 'median patch of enlarged scales, the larger ones about four times the size of the chin granules ; back, sides, and upper surface of limbs covered with granules which are slightly enlarged into small hexagonal scales on the median line of the back ; underside of body with eight longi- tudinal and thirty-five transverse rows of rec- tangular plates; the outer row less than one half the size of the next one; one large preanal plate, preceded by one much smaller, and this by two still smaller placed trans- SCHMIDT, AMPHIBIANS OF PORTO RICO m versely ; on the lower arm two rows of large antebracials, separated from the much smaller single row of brachials by small scales, on the lower edge of the upper arm a single series of enlarged plates; under- side of thigh covered with two series of large scales or plates and three smaller ones; thirteen or fourteen femoral pores; under- side of tibia covered entirely across by three plates, of which the upper is the largest and larger than the other two together; upper side of wrist with three series of enlarged plates; fifth (outer) toe extending far beyond the first (inner), almost to the claw of the second; tail cov- ered with smooth scales in rings, the scales being oblique with parallel sides except the median row, which is wedge-shaped; about tv/enty-two scales in the fifteenth ring from the base. Coloration (in alcohol) above dark brownish olive with seven distinct greenish white longitudinal lines, the median one somewhat wider than the others and starting from the tip of the snout while the others originate in front of the eye, and continuing some distance on the tail, except the outer row, which terminates in the groin ; upper sides of limbs also dark olive brown with very distinct round greenish-white spots ; underside greenish-white dark- ening on tail." Remarks. — Little need be added to Stejneger's description. Fowler describes and figures the second specimen known. His color plate rep- resents the alcoholic coloration rather than that of the living reptile. In the series of 37 specimens in the Porto Rican Survey collection the prefrontals are broadly in contact in 21 specimens, meet at a point in 1 and are separated by a suture between the frontal and frontonasal in 3. The number of supraciliaries varies from five to seven, normally six. The interparietal is horizontally divided in 1 specimen. There are usually two or three transversely enlarged postoccipitals. On the whole, there is a remarkably small degree of variation. The measurements of a male and female specimen : A. M. N. U. A. M. N. H. Parts measured No. 13821 d No. 13828 $ Length 169 mm. (tail reproduced at tip) 147 mm. Body 52 45 Length of head 12.5 11 Breadth of head 8.5 6.5 Foreleg 16 14 Hind leg 30 26 Habits. — Almost nothing is known of the habits of this species. It was found only on or near the tops of the limestone hills back of Ensenada, associated with a few Ameiva exsul. 113 SCIENTIFIC SURVEY OF PORTO RICO The extreme fragility of the tail (25 out of 27 specimens having broken or regenerated tails) seems to be correlated with its brilliant color and the capacity for violent contortion of the separate member. It has frequently been suggested that the fragile tails of lizards might serve as a protective device, the broken and active tail occupying the pursuer while the owner escapes. In this case the assumption is made more than usually plausible by the combination of color, motion and noise — the tail making a noticeable rattling in the dead leaves of the ground cover. Mdiile the lizard moves quite silently. Ameiva exsul Cope Iguana Ameiva plei var. exsul Cope, 1862, Proc. Acad. Nat. Sci. Phila., p. 66. Ameiva extil Stejneger, 1904, Rept. U. S. Nat. Mus., 1902, p. 612, Figs. 59-66.— Barbour, 1914, Mem. Mus. Comp. Zool., Vol. XLIV, p. 310; 1917, Proc. Biol. Soc. Wash., Vol. XXX, p. 99.— Fowler, 1918, Papers Dept. Marine Biol., Carnegie Inst., Vol. XII, p. 12.— Danforth, 1925, Copeia, No. 147, p. 78. Ameiva exsul Barbour and Noble, 1915, Bull. Mus. Comp. Zool., Vol. LIX, p. 439.— Schmidt, 1920, Ann. N. Y. Acad. Sci., Vol. XXVIII, p. 192.— Woleott, 1924, Journ. Dept. Agric. Porto Rico, Vol. VII, p. 10. Ameiva plei Cope, 1862, Proc. Acad. Nat. Sci. Phila., p. 65 (not of Dumeril and Bibron). — Peters, 1876, Monatsber. Akad. Wiss. Berlin, 1876, p. 708. — Gundlach, 1881, Anales Soc. Espan. Hist. Nat., Vol. X, p. 311.— Stahl, 1882, Fauna Puerto-Rico, p. 69, p. 158. Ameiva riisei Reinhardt and Luetken, 1863, Vid. Meddel. Foren., Copenhagen. 1862, p. 2.32.— Bocourt, 1874, Miss. Sci. Mex., Zool. Rept., livr. 4, PI. 20 B. Fig. 3. Ameiva riisi Boulenger, 1885, Cat. Lizards Brit. Mus., Vol. II, p. 354 — Gar- man, 1887, Bull. Essex Inst., Vol. XIX, p. 11.— Meerwarth, 1901, Mitt. Naturh. Mus. Hamburg, Vol. XVIII, p. 30, PI. 2, Figs. 7-9. Ameiva vittipunctata Baettger, 1893, Kat. Rept. Mus. Senck., Vol. I, p. 74 (not of Cope). The transfer of the common name "iguana'' to this species illustrates the usual fate of vernacular names when they become incorporated into tlie language — they are transferred to any convenient animal without reference to zoological relationship. Type lomlity. — Water Island, near St. Thomas, Virgin Islands. Distribution. — Recorded from Arecibo, Arroyo, Caguas, Cataiio, Can- grejos Pt., Caya Santiago, Coamo Springs, Ensenada, off Humacao, Luquillo, Palo Seco Point, Ponce, Rio Piedras, San Juan, Santurce and TJtuado, Neither Stejneger nor I secured this species in Vieques. My record iiVHMIDT, AMPHIBIAN,"^ OF PORTO RICO HP, from Culebra appears to be the first from that island. In the A^irgin Islands it is recorded from St. Thomas, St. John, Tortola, Virgin Gorda, Anegada and St. Croix. Specimens collected. — 57 : Coamo Springs, Ensenada, Palo Seco Point, Santurce and Culebra Island. Diagnosis. — An Ameiva of moderate size, with ventral plates in ten longitudinal rows; generally white-spotted only on the posterior half of the back; femoral pores averaging 15.5 on each side; number of scales in fifteenth tail segment averaging 45; median gulars formmg a group of slightly enlarged scales; plates on upper arm much wider than long. Original description. — "This form differs in possessing a narrow bright yellow band on each side, extending from the superciliary ridge to a point on the anterior part of the tail. The anterior extremity extended backward exceeds the extremity of the appressed femur. Total length 7 in. 6 lin.; exclusive of tail, 3 in. 1 lin. (Probably young)." Remarks.— CoTpe's inadequate description accompanies remarks on the Virgin Island and Porto Rican Ameivas, which he refers to A. plei Dumeril and Bibron. A number of points in the description of A. pleii seem to exclude the possibility that it is based on Porto Rican specimens, as were so many of the other species collected by Plee, and Cope's varietal name has priority over the name riisei proposed by Reinhardt and Luetken. Stejneger gives a very detailed description of the species. His descrip- tion of the coloration in life is as follows : "Ground color above of old skin (the specimen was shedding) 'pea green,' of new skin more olive green, the difference being slight, how- ever; underside pale 'pearl gray' with a decided wash of 'turquoise blue' on the groin and tail; dark markings blackish; eyelids edged with whitish; iris very dark brown. "A somewhat larger specimen had the ground color above tawny olive becoming olive gray on the tail; head not colored differently from back; lower back with a broken network of black meshes; flanks with a series of vertical black spots on a slightly browner ground alternating with a double or triple series of pale dots, which continue indistinctly on hind legs and sides of tail ; tip of snout and of lower jaw pink flesh color; sides of head pale drab; underside whitish with a bluish cast, which is strongest on the sides and under hind legs and tail. "A young specimen had head and neck uniform tawny olive ; ground color of back similar, but becoming duller toward the tail, which is drab above ; ground color of flanks similar, though more russet near the light lateral line, especially anteriorly ; sides of head and neck nearly uniform 114 SCIENTIFIC SURVEY OF PORTO RICO pale ciimanioii ; a narrow pale cream-buif line from superciliaries slightly broadening on the back and fading out at about the posterior third of the back, margined with blackish on both sides; on back and flanks a series of narrow blackish crossbars becoming obsolete on the lower back, the interspaces filled with roundish Isabella-colored spots; lower back similarly spotted, as are also the upper side of the legs; tail above with these spots more faintly indicated; underside whitish with a turquoise- blue suffusion on both sides of abdomen and under the tail; underside of thighs pale gray dappled with white round spots like those on the back:" This species attains a large size, apparently much exceeding lialf a meter, but the larger individuals are exceptionally wary and I was unable to secure them. The largest seen were on Culebra Island. In nearly all the specimens examined an additional row of ventral plates on each side is enlarged to a varying degree, in some cases to such an extent that there are distinctly twelve longitudinal rows of ventrals. Habits.— Ameiva exsul distinctly prefers a sandy soil, and is every- where more abundant on sand. The reptiles frequently make shallow burrows under stones or other loose objects. Near Santurce they were especially abundant along a disused tramway, burrowing under the ties. At Ensenada they occasionally made burrows under the concrete side- walks. This species is almost entirely confined to the coastal plain, but it follows the river bottoms into the interior of the island. It is found on the tops of the limestone hills of southwestern Porto Rico, but not on those of the northern side of the island. It was very abundant on Culebra, where the mongoose has not been introduced, while on Vieques the inhabitants informed me that the "aldea" had exterminated the 'iguana" as well as the "lucia" and the snakes. The common report in Porto Rico that the "iguana" eats the shoots of young corn appears to be supported to a degree by an examination of stomach contents. Of 20 stomachs examined, one was empty; 11 con- tained vegetable matter, chiefly large numbers of red-coated seeds ; 5 had unidentifiable insect remains; 2 had crickets; 3 had small crabs; 3 had eggs of a lizard; 1 had the tail of a large Anolis cristateUus; 6 had parasitic worms. Wolcott, in a much more detailed examination of stomach contents, finds only 6.7% of vegetable matter, consisting of mushrooms. Wolcott found two batches of eggs, numbering 4 and 7 respectively, buried four or five inches deep in a pile of humus in the garden of the Experiment Station at Rio Piedras. The eggs were faint pink to bright SCHMIDT, AMPHIBIANS OF PORTO RICO 115 pink in color, measuring 20 to 22 mm. in length and 13 to 15.5 mm. in siiort diameter. Anieiva albogiittata Boulenger • Text Figs. 34 and 35 Ainciia dlhui/uftutu Houleuger, 1890, Jahresber. Naturw. Ver. Maj^deburt;. 1S94-1S06, p. 112.— Meerwarth, 3901, Mitt. Naturli. Mus. Hamburg, Vol. XVIII, p. 32, PI. 2, Figs. 6-8.— Stejneger, 1904, Kept. U. S. Nation. Mus., 1JK)2, p. 618, Figs. 67-72.— Barbour, 1914, Mem. Mus. Comp. Zool., Vol. XLIV. p. 311.— Barbour and Noble, 1915, Bull. Mus. Comp. Zool., Vol. I.IX, p. 440.— Schmidt. 1920, Ann. N. Y. Acad. Sci.. Vol. XXVIII, p. 193; 1926. Publ. Field Mus. Nat. Hist., Zool., Vol. XII, p. 156. 2^i//)e loculUy. — Mona Island. Distribution. — Confined to Mona Island. Specimens collected. — 46 : Mona Island. Diagnosis. — Closely allied to the Porto Eican Anieiva exsul, but back with white spots to the neck; femoral pores averaging 13.2 on each side; number of scales in fifteenth tail segment 34 ; median gular group of scales less differentiated, often confluent with the adjoining scales; plates on upper arm slightly wider than long. Original description. — "N"ostril between two shields; five or six oc- ci})ital shields in a transverse row, bordered behind by small, irregular scales; three or four supraoculars, the fourth, when present, very small; seven or eight supraciliaries ; loreal imdivided; six supralabials ; five infralabials ; an unpaired and four or five paired chin-shields ; a broad middle zone of slightly enlarged gulars ; mesoptychial scales larger ; body scales finely granular, smooth ; ventrals in ten longitudinal rows, the outer very small, and in thirty-two or thirty-three transverse rows ; three large anals ; brachial plates in a single row, completely separated irom the antebracliials ; four or five rows of femoral scales ; a row of very large tibial shields with two smaller rows on the inner side ; twelve to fifteen femoral pores on each side; scales of the tail straight, the u])per ones keeled. Dorsal side light grayish-brown, with a dark brown lateral i)an(l extending from the shoulder to the base of the tail, outlined above by a light line and spotted with black; the back and sides with the excep- tion of the head and neck are thickly spotted with white ; underparts white, throat of male red." Remarl-s. — Ameiva alhoguttata is extremely close to Amclra e.rsul, but may be distinguished by the more spotted dorsum. The femoral pores in forty specimens average 13.2, in forty A. exsul the average is 15.3, almost exactly as in the smaller series examined by Stejneger. The Mona 116 SCIENTIFIC SURVEY OF PORTO RICO Fig. 34. — Ameiva alhogut- tata. A. M. N. H. No. 14003. Mona Island. Natu- ral size. Fio. 35. — Ameiva alhoguttata. A. M. N. H. No. 14003, Mona Island. Natu- ral size. SCHMIDT, AMPHIBIANS OF PORTO RICO 117 Island form does not exhibit the tendency to enlargement of an additional row of ventral plates, one specimen having only eight longitudinal rows of ventrals (No. 13739). Habits. — Like the Porto Eican species, the Mona Island Ameiva is found chiefly on the sandy land of the terraces on the west and south sides of the island. An occasional specimen, however, may be seen on the rocky table land. The results of the examination of 20 stomachs are as follows : empty, 4; vegetable matter (chiefly red coated seeds), 8; unidentifiable insect remains, 3; beetles, 3; crickets, 2; land snails, 2; Anolis cristatellus (juv.), 1. This species is preyed upon by Alsophis variegatus. Amphisbaenidae Ainphisbaena Linnaeus The limbless lizards of this genus are immediately distinguishable by their soft skin, which is marked off into rectangular scale-like areas. The two species in Porto Eico may be distinguished as follows : A. Body rings 220-2.30; suture between nasal shields very short, one-fifth or less of the prefrontal suture ; one temporal A. caeca. AA. Body rings about 250; nasal suture long, more than one-third of the prefrontal suture ; no temporal A. hakeri. Ainphisbaena caeca Cuvier Culebra ciega; vibora Text Fig. 36 AmvhiHhaena caeca Cuvier, 1829, Regne Anim., 2nd Ed., Vol. II. p. 73. — Dumeril and Bibron, 1839. Erpet. Gen., Vol. V, p. 492.— Dumeril, 1851, Cat. Method. Kept. Mus. Paris, Vol. I, p. 148.— Peters, 1876, Monatsber. Akad. Wiss. Berlin, p. 708.— Gundlach, 1881, Anales Soc. Espan. Hist. Nat., Vol. X, p. 312.— Stahl, 1882, Fauna Puerto-Rico, p. 70, p. 160.— Strauch, 1883, M61. Biol. Acad. Sci. St. Petersburg, Vol. XI, p. 405.— Boulenger, 1890, Proc. Zool. Soc. London, p. 79.— Stejneger, 1904, Rept. U. S. Nation. Mus., 1902, p. 676, Figs. 129-132.— Barbour, 1914, Mem. Mus. Comp. Zool., Vol. XLIV, p. 319.— Fowler, 1918, Papers Dept. Marine Biol.. Carnegie Inst., Vol. XII, p. 14.— Schmidt, 1920, Ann. N. Y. Acad. Sci.. Vol. XXVIII, p. 194.— Camp, Bull. Amer. Mus. Nat. Hist., Vol. XLVIII, p. 317, Fig. F. Amphishaena hakeri Danforth, 1925, Copeia, No. 147, p. 78 (not of Stejneger). Type locality. — Martinique (erroneously). Distrihution.— Confined to Porto Eico, where it has been found at 118 SCIENTIFIC SURVEY OF PORTO RICO Aibonito, Bayamon, Catalina Plantation, Lares, Luquillo, Mayagiiez, Eio Piedras and Utuado. Specimens collected. — 18 : Aibonito, Bayamon and Eio Piedras. Diagnosis. — An Amphishaena with 220 to 230 body rings; one temporal scute; suture between the nasals very short, 16-19 rings on tail. Original description. — "There is one in Martinique, entirely blind {Amphisbaena caeca, Cuv.), May not this be A. verniiciilaris Spix, XXV, 2 ? He says eyes conspicuous, while I do not find any at all. He uses the same expression for his A. o.ryura." Remarks. — The original description is evidently quite useless, and this species might more justly be cited as described by Dumeril and Bibron, who redescribed the typical material. However, I follow them and Stejneger in ascribing the species to Cuvier. Stejneger's description is much the most useful and may be quoted in full : "Eostral small, triangular, the portion visible from above short, about Fig. 36. — Head of Amphishaena caeca from above and from side. (After Stejneger.) equaling the suture between the nasals ; prefrontals very long, the suture between them longer than the one between the frontals and five times as long as the nasal suture; ocular moderate, quadrangular, smaller than the postocular and the third supralabial; in the angle shields a well- developed temporal, between and behind the latter two only slightly smaller than the ocular; eye plainly visible through ocular; a pair of occipitals, broader than long, in contact behind the frontals ; three supra- labials, the second as long as the other two together ; three lower labials, the second longer than the other two together; mental followed by a large median postmental, twice as long as broad ; behind the second lower labial a large malar shield; just behind the postmental and between the nialars three scales in a transverse row (postgeneials) ; 226 rings on the body and 17 on the tail; the segments of each ring longer than broad on the back, broader than long on the imder side, 16 above and 18 below the lateral line; anal shields, 6; preanal pores, 4. Color, flesh color, with a squarish brown spot, darkest on the back, occupying the middle of each segment, these spots being absent on many of the ventral SCHMIDT, AMPHIBIANS OF PORTO RICO 119 segments of the posterior half of the body ; top of head uniform brownish, except rostral and. nasals, which are colorless." Stejneger's discussion of the history and relations of this species is a model of exact taxonomy. Tlie variation in the series of eighteen specimens in the Porto Kico Survey collection falls well within the limits established by Stejneger, who examined nineteen. One specimen has a small supraocular plate on each side. It is consequently extremely surprising to find that the specimens listed l:)y Danforth from Mayagliez as A. lakeri are intermediate between lal-eri and caeca. One of these specimens (now F. M. N. H. N"o. 12473) has the nasal suture about one-fourth that of the prefrontals, temporal present, and 238 body-rings. The two other specimens have 234 body rings, nasal suture about one-half that of the prefrontals, and one of them lacks the temporal on one side. Tl!e three specimens are thus some- what closer to A. caeca than to hakeri, but it is very evident that examina- tion of adequate series from Mayagiiez and Lares and the intervening area may alter our conception of the Porto Eican AmpMsbaena radi- cally. It is curious that no Ampliisbaena is known from southwestern Porto Rico. The largest specimen measures 233 mm., tail 18 mm. Habits.— M\. of the specimens were found burrowing in the ground, most of them uncovered by cultivation. One was located about three inches beneath an ant's nest under a log, in the course of digging up the eggs of Leim ado phis. When the creature is killed in formalin, the head is bent abruptly to one side, indicating apparently a special development of the muscles of the neck, which doubtless is of advantage to the Am- phishaena in burrowing. Three eggs were obtained, — one beneath a termite nest, the other two under the log where the above-mentioned adiilt was dug up. The largest egg measured 42 mm. by 11 mm, Ampliisbaena bakeri Stejneger Text Fig. 37 Ampliisbaena hakeri Stejneger, 1904, Ann. Kept. U. S. Nation. Mus.. 1902, p. 681, Figs. 13.3-137. Type locality. — Lares, Porto Rico. Distrihui ion. —Confined to Porto Rico, where it is known only from Lares. 120 SCIENTIFIC SURVEY OF PORTO RICO Diagnosis. — An Ampliishaena with about 250 body rings; nasal suture long^ more than one-third the prefrontal suture; no temporal. Original description. — "Eostral small, triangular, the portion visible from above short, nearly one-third the suture between the nasals; pre- frontals long, the suture between them slightly longer than the one be- tween the frontals and but slightly more than twice the nasal suture; ocular moderate, quadrangular, the anterior angle very long and pointed ; eye not visible; a pair of occipitals, longer than broad (the one on the left side abnormally divided), broadly in contact behind the frontals; three supralabials, the second longer than the other two together; three lower labials, the second longer than the other two together; mental followed by a large median postmental, much longer than broad; behind Fig. 37. — Head of Ampliishaena iaJieri from above and from side. (After Stejneger.) the second lower labial a large triangular malar shield ; behind the post- mental and between the malars 3 scales in a transverse row (post- geneials) ; 249 rings on the body and 16 on the tail; the segments square, slightly longer than broad on the back, the 6 median rows on the ab- domen broader than long, especially the middle pair; 16 above and 16 below the lateral line; anal shields or segments, 6; preanal pores, 4. Color light flesh, with a brownish spot in the center of each segment, rather indistinct, especially on the lower surface. Dimensions Tip of snout to vent ■ 260 mm. Tail 18 Diameter of bodj' 9 "Apart from a slight oscillation in the relative length of the sutures on the head the variability is insignitieant. As in 4. cceca, the occipitals appear most subject to variation, but they seem to be longer than broad, as a rule, even in the clearly abnormal specimen (No. 25537) as shown in Fig. 134. The number of rings varies only between 249 and 251 in the three specimens at hand." SCHMIDT, AMPHIBIANS OF PORTO RICO 121 Remarks. — I have discussed above imder cceca the unexpected vari- ations of A. cceca in the direction of hakeri. I am fully convinced of the specific validity of hakeri. Habits. — Nothing is known of the habits of this species. SCINCIDAE Mabuya Fitzinger Mabuya sloanii (Daudin) Text Figs. 38 and 39 Lucia ; Santa Lucia Scincus sloanii Daudin. 1903, Hist. Nat. Kept., Vol. IV, p. 287, PI. 55, Fig. 2. Eumeces sloanii Dumeril and Bibron, 1839, Erpet Gen., Vol. V, p. 639.— Dumeril, 1851, Cat. Method. Rept. Mus. Paris, Vol. I, p. 156. Mahmja sloanii Bocourt, 1879, Miss. Sci. Mex., Rept., p. 401, PL 22 B, Fig. 3.— Stejneger, 1904, Ann. Rept. U. S. Nation. Mus., 1902, p. 608, Figs. 56-58.— Barbour, 1914, Mem. Mus. Comp. Zool., Vol. XLIV, p. 320; 1916, Proc. Biol. Soc. Wash., Vol. XXIX, p. 219.— Fowler, 1918, Papers Dept. Marine Biol., Carnegie Inst., Vol. XII, p. 7.— Schmidt, 1920, Ann. N. Y. Acad. • Sci., Vol. XXVIII, p. 194.— Wolcott, 1924, Journ. Dept. Agric. Pto. Rico., Vol. VII, p. 13.— Schmidt, 1926, Publ. Field Mus. Nat. Hist., Zool., Vol. XII, p. 156. Mabuia sloanii Boulenger, 1887,, Cat. Lizards Brit. Mus., Vol. Ill, p. 193 (part) ; 1896, Jahresber. Naturw. Ver. Magdeburg, 1894-1896, p. 113.— Meerwarth, 1901, Mitt. Naturh. Mus. Hamburg, Vol. Ill, p. 135. Euprepes semitaeniatus Wiegmann, 1937, Arch. Naturg., Vol. Ill, p. 135. Tiliqua richardi Gray, 1838, Ann. Nat. Hist., Vol. II, p. 292. Mahouya sloanei Gray, 1845, Cat. Lizards. Brit. Mus., p. 94. Mabouia aenea Giinther, 1859, Am. Mag. Nat. Hist. (3), Vol. IV, p. 212 (not of Gray). Mabuia cuprescens Cope, 1862, Proc. Acad. Nat. Sci. Phila., p. 186. Gonfjylus (Eumeces) agilis Reinhardt and Luetken, 1863, Vid. Middel. Naturh. Foren., Copenhagen, 1862, p. 229 (not M. agilis). Mabuya fulgida Cope, 1868, Proc. Acad. Nat. Sci. Phila., p. 311 (not M. fulgida Cope, 1862). Euprepes (Mabuia) spilonotus Peters, 1876, Monatsber. Akad. Wiss. Berlin, 1876, p. 708 (not of Wiegmann, 1837).— Stahl, 1882, Fauna Puerto-Rico, p. 159. Euprepes spilonotus Gundlach, 1881, Anales Soc. Espaii. Hist. Nat., Vol. X, p. 311. Mabuia nitida Garman, 1887, Bull. Essex Inst., Vol. XIX, p. 51. Type locality. — St. Thomas, Virgin Islands. Distribution.— The Porto Rico Survey secured this species from Baya- mon and Ensenada, on Porto 'Rico, and from Mona and Culebra islands. 123 SCIENTIFIC SURVEY OF PORTO RICO No other definite localities have been recorded for Porto Eico. On Mona Island the species occurs on the low terrace as well as on the rocky plateau. In the Virgin Islands this form is recorded from St. Thomas, St. Croix, St. John and Jost van Dyke. I am not at all convinced that the Hispaniolan skink is identical with this species. To avoid possible con- fusion I hereby restrict Garman's Mabui-a nitida to the Porto Eican form. Specimens collected. — 7: Bayamon, Ensenada, Mona Island and Cule- bra Island. Diagnosis. — A typical lizard in body-form, with weU-developed limbs, large plates on the top of the head, no occipital shield, ventral scales like the dorsals and laterals, all very smooth and shiny. Fig. 38. — Heud of Mahuya sloanii from above. A. M. N. H. No. 14007 (A) and A. M. N. H. No. 14006 (B). To show variatiou in pattern. Twice natural size. A. B. Fig. 39. — Head of Mabuya sloanii from side. A. M. N. H. No. 14007. Twice natural size. Original description. — "Sloan's skink is a slender and slini-waisted form, resembling the five-lined skink of North America but differing from it in certain notable characters. The elongate head is covered with plates ; the body is a little narrowed, and is covered, like the limbs *and the anterior third of the tail, with small imbricate rounded scales ; the rest of the tail is covered with rings or veritable verticils of scales. I have noticed under each thigh a row of small pores, but am unable to count them as some of the scales have been lost. The feet each have five slender clawed digits. "This skink is brown above and whitish beneath and is easily recognized by means of the four black longitudinal lines which begin on the end of SCHMIDT, AMPHIBIANS OF PORTO RICO 123 the snout, namely a broader one passing above each arm and prolonged to the thigh, and two others,- a little narrower, extending to the sides of the back." Remarls. — Daudin's mention of "petits grains poreux" beneath the thighs may be explainable as referring to the tubular canals of the scales, which are frequently conspicuous, for no skink known has femoral pores. In other respects his description is vague, but the identity of the species is satisfactorily fixed by the fact that the origin of the type, collected by Eichard on St. Thomas, is made known by Dumeril and Bibron. Stejneger has satisfactorily cleared up the synonymy, and I am able to add Wiegmann's Euprepes semitaematus. With seven specimens before me— three from Culebra, three from Porto Eico, and one from Mona — I am unable to find differences corres- ponding to the separate localities, other than the difference in color de- scribed below. In all specimens there are two pairs of chin-shields in contact behind the unpaired postmental. The supranasals form a suture in four specimens. The prefrontals are narrowly or widely separated by a suture between the frontal and the fronto-nasal. The supraoculars are three on one side in one specimen. Another specimen has three large occipitals on one side. The scales about the body are 33 in the specimens from Culebra and Mona, and in one from Porto Eico, 30 in the remain- ing two. The coloration is highly interesting. The three specimens from Porto Eico agree with the description of Stejneger (1904, p. 611) in hav- ing a narrow black border above the dorso-lateral light line. In the specimens from Culebra, this is increased anteriorly to include the whole of the head, neck and shoulders, leaving, however, a sharply de- fined median light line from the frontal to the shoulders, where it merges into the dorsal color. This pattern is approximated also in the specimen from Mona Island. It is evident that the type of Euprepes semitaeniatus AViegmann, described by Stejneger (1901, p. 610), corresponds accurately with the Culebra specimens. It is therefore possible that several insular forms may be distinguishable when adequate series become available. In view of the close approach of the Mona specimen to those from Culebra, I prefer to retain the use of sloanii for the entire series for the present. The measurements of the only specimen with a complete tail are as follows : 124 SCIENTIFIC SURVEY OF PORTO RICO A. M. X. H. Parts measured No. 14007 Length ISO mm. Body 67 Length of head 15 Breadth of head 10 Foreleg 17 Hind leg 25 The largest specimen (from Culebra) measures 90 mm. from snout to vent. Habits. — This species is presumably viviparous like its American congeners. Nothing definite is known about its breeding habits. The "Lucia" is everywhere reported to enter houses, where its presence is supposed to bring good luck. It is most abundant in the more arid localities, and its persistence on Mona and Culebra may be in part due to the favorable habitat conditions. It was not reported at Aibonito, and probably does not enter the coffee plantations. The specimens seen in life were foimd among rocks (3), on the base of a cocoanut palm (1), in a knot hole in a fence post (1), and in the cracks of a rotten log (1). The specimen taken on Mona was in a vertical rock-fissure on the tableland. SERPENTES Suborder Synopsis of the Genera of Porto Rican Snakes A. Eyes covered by scales ; ventral scales small like dorsals ; tail extremely short (Typhlopidfe) Typhlops AA. Eyes well-developed ; ventral scales transversely enlarged ; tail elongate. B. Subcaudals undivided (Boidae) .Epicrates BB. Subcaudals in pairs ( Colubridse ) . C. Dorsal scales with one pore at the tip or none Dromicus CC. Dorsal scales with a pair of prominent pores at the tip. .Alsophis Typhlopidae Typhlops Oppel Ket to the Porto Rican Species of Typhlops A. Rostral very narrow, one-fifth to one-sixth the width of the head ; a white spot beneath the head and another beneath the tail T. rostellatus. AA. Rostral wider, one-third to one-fourth the width of the head. B. Scales on mid-line of back 365-430; venter light, with an angular notch or ring of the white color just in front of the tail T. platvcephalus. BB. Scales on mid-line of back 313-321; brown above, nearly white beneath, without a white caudal ring or notch T. monensis. SCHMIDT, AMPHIBIANS OF PORTO RICO 135 Tjphlops platycephalus Dumeril and Bibron Culebra ciega Text Figs. 40 and 41 Typhlops platycephalus Dumeril and Bibron, 1844, Erpet. Gen., Vol. VI, p. 293. Typhlops luinbricalis (nee Linne) Peters, 1876, Monatsber. Akad. Wiss. Berlin, 1876, p. 708.— Gundlach, 1881, Anales Soc. Espau. Hist. Nat., Vol. X, p. 312.— Stahl, 1882, Fauna Puerto-Rico, p. 70, p. 160.— Stejneger, 1904, Kept. U. S. Nation. Mus., 1902, p. 684, Figs. 141-144 (part).— Barbour, 1914, Mem. Mus. Comp. Zool., Vol. XLIV, p. 322 (part). Typhlops 1-ichardii (nee Dumeril and Bibron), Dumeril, 1851, Cat. Method. Rept, Mus. Paris, p. 205, (part).— Schmidt, 1920, Ann. N. Y. Acad. Sci., Vol. XXVIII, p. 195.— Danforth, 1925, Copeia, No. 147, p. 78. Typhlops jamaiccnsis Cochran, 1924, Journ. Wash. Acad. Sci., 14, p. 177 (part). Type locality. —Martmique (erroneously). Distributwn. — Confined to Porto Rico, where it has been collected at Aguadilla, Bayamon, and Mayagiiez. . Specimens collected. — 19 : Bayamon, Diagnosis. — A Typhlops with body scales in 22 rows anteriorly and 20 at mid-body; 365 to 420 scales from head to tip of tail; tail with a white ring or notch. Original description. — ^"This Typhlops, as its name indicates, has the head much more depressed than any of its congen,ers; the sides of the snout, in vertical profile, slope together at a subacute angle, rounded at the apex; the diameter of the body at the middle is contained almost forty-five times in the total length; the length of the tail is about one thirty-ninth of the total ; the upper part of the rostral is narrower than in the preceding species; the angle made by the posterior sides of the fronto-nasals is more acute than in T. lumhricalis or in T. richardii; the preoculars are proportionately larger and the oculars narrower than in these latter species; the upper labials are also less developed and none of them extends upward on the side of the head; the tail is a moderately slender cone, but its terminal spine is distinctly longer and more compressed; the rostral is a longitudinal band whose lower portion is shorter than the upper; the upper portion is narrow and very obtuse behind; the lower part is wider in front than behind, angulate on each side, and with a small projection enclosed by the first pair of labials ; the anterior frontal, (the true frontal), the interparietal, and the post- interparietal are similar, hexagonal, widened, and about one-half the size of the lower portion of the rostral; the supraoculars are a little larger tlian the frontal, hexagonal, and somewhat oblique to the trans- 126 SCIENTIFIC SURVEY OF PORTO RICO verse line ; the parietals have the same shape, but they are strictly trans- verse, and in large part behind the oculars; a pair of post-parietals par- allels them; the nasals are scalene triangles rounded on their posterioi angles. The fronto-nasals, which do not meet behind the rostral, each has the shape of a <, with rather wide branches which narrow rapidly posteriorly; the preoculars are a little shorter than the fronto-nasals, subequilateral, two of their sides fitting into the chevron-shaped fronto- nasal; the oculars are a little higher and much narrower than the pre- oculars ; their peaks are acute and their bases rather enlarged and rounded behind; the four upper labials are oblong, the first very small, quadrilateral, higher posteriorly; the second is quadrangular, strongly rounded on its lower border, as are the third and fourth, which are triangular and increase gradually in size ; the eyes are perfectly distinct, lateral, just beneath the surface; the body scales are in twenty longi- „„,,_ Fig. 40. ^Pattern of tail of TiiiMopt %''>-vM'^^. t." ^ ^:^ PlatiMcphaJus (left) contrasted with that of T. rostcUatus (right). A. M. N. H. Nos. 1333G and 13179. A- "^^ B. ^? Natural size. * tudmal and three hundred and fifty transverse rows, with twelve trans- verse rows on the tail," Bemarl's. — Cochran (1924, p. 176) has given reasons for attaching this description to the Porto Rican species. This extremely satisfactory allocation of platycephahis has against it only the slightly low scale count given by Dumeril and Bibron, and the narrowness of the rostral. Exami- nation of types in the Museum d'Histoire Naturelle in Paris may upset this arrangement, but such a result will not affect the interesting results of Cochran's analysis of the West Indian Typhlops, which shows that the Virgin Island richardii, the Porto Eican platycephalus and the Jamaican jamaicensis constitute a series of related forms, united by her as jamaicensis. In view of their insularity and the gap due to' the lack of an Hispaniolan species of this group, I prefer to retain the three forms as distinct. The appropriateness of a subspecific nomenclature for such a series is somewhat in dispute, but where a series of more than two forms inhabit a chain of islands and exhibits overlapping but not identical characters, I see no reason against the extension of the sub- specific category to include them. The scale rows are 22-20-20 in seven specimens, 22-20-18 in seven, the reduction to 20 rows occurring a little anterior to the middle of the body. The scales from head to tip of tail on the mid-dorsal line vary SCHMIDT, AMPHIBIANS OF PORTO RICO 127 from 365 to 415, and by iucluding Cochran's counts of the U. S. National Museum specimens, this range is increased to 365-420. The coloration of this species is characteristic. The color above is brown, each scale darker on its posterior two-thirds; the underside is whitish, the dividing line between the dorsal and ventral color being extremely irregular ; the ventral color forms a notch or a complete ring about the level of the vent. The total length in fourteen specimens varies from 216 to 310 mm., average 266 mm. The diameter of the body -is contained in the total length from 34 to 44 times. Habits. — The specimens of the present series were discovered in the course of cultivation on the farm of Mr. B. A. AVall. The single speci- men secured by me personally was burrowing in the loose earth around Fig. 41. — Heads of Porto Rican Typhlops. Left to right (upper), T. platycephalus, T. rostellatus ; (lower). T. wonensis. (First two species from Stejneger ; last from Schmidt.) an old stump, in which both Typhlops and Lewiadophis eggs were found. The living specimen coiled tightly about my hand and was able to inflict a considerable prick with the sharp tail-spine. The tail is definitely manoeuvered for this purpose, and this habit has probably given rise to the many superstitions about the existence of snakes with a tail sting. Three eggs of this species, containing well-developed embryos, were removed from the soil about the same stump. The egg is elongated, like a slightly bent cylinder with rounded ends, and has a perfectly smooth, white surface. The embryo measures 98 mm. in length and 3 mm. in diameter. The smallest hatched specimens measure 114 mm. Three of the smallest specimens in the collection are in every way like the adults except that they are pale grayish white. This appears on examination to be caused by the opacity of the skin, which is nearly 128 SCIENTIFIC SURVEY OF PORTO RICO ready to be shed, probably for the first time. An adult Cuban specimen in the collection has the same appearance, and the underlying skin proves to be normally colored. A certain number of the cases of sup- posed albinism in T. lumhricalis may be due to this appearance. Typhlops rostellatus Stejneger Text Figs 40 and 41 Typhlops rostellatus Stejneger, 1904, Kept. U. S. Nation. Mus., 1902, p. 686, Figs. 145-147.— Barbour, 1914, Mem. Mus. Comp., Zool., Vol. XLIV, p. 322.— Schmidt, Ann. N. Y. Acad. Sci., Vol. XXVIII, p. 197. Type locality. — Lares, Porto Eico. Distribution. — Confined to Porto Eico, where it has been collected at Aibonito, Bayamon and Lares. Specimens collected. — 11 : Aibonito, Bayamon. Diagnosis. — Snout rounded ; nostrils lateral ; preocular in contact with the third labial only ; nasal completely divided ; two post-oculars ; rostral very uarroM-, one-fifth to one-sixth the width of the head; scale rows 18-20 ; a sharply defined white spot beneath the tail. Original description. — "Head blunt, not depressed, snout projecting, rounded laterally; nostrils lateral; rostral narrow, about one-sixth the width of the head (1 :6.4), not extending as far back as a line between the anterior edge of the eyes; nostril on a suture completely dividing the nasal, the lower anterior part in contact with first and second, the upper posterior nasal in contact with second and third; preocular wider than ocular, its anterior angle much produced and rather acute, in con- tact with third supralabial only ; ocular with the anterior border strongly convex, in contact with third and fourth supralabials ; supralabials four, the posterior two large and reaching high up on the side; prefrontal, frontal, and interparietal scale-like, subequal; supraoculars and parietals enlarged, especially the latter; eye distinctly visible; 18 scale rows round the body; about 333 scales on the middle line of the body underneath from chin to vent, and 13 under tail ; tail ending in a spine. Color uni- form dark brown, slightly paler underneath; through the dark ground color a distinct blackish network can be traced, the meshes of which an- teriorly coincide with the outline of the scales, but becoming more and more discordant posteriorly; rostral and anterior nasal brown above, margined with whitish, underneath whitish; a very abrupt whitish spot occupying the anal region and the under side of the tail." Remarks. — This species is readily distinguished from T. platycephalus by its nearly uniform coloration above and below, and the sharply defined SCHMIDT, AMPHIBIANS OF PORTO RICO 129 white subcaiidal spot. There is little variation in the present series. The scales about the body number 30 in ten specimens, 18 in one. Stejneger had three specimens with 18 scales and onh' one with 20. The measurements of the largest specimen are as follows : A. M. N. H. Parts measured No. 13345 Total length 205 mm. Tail 5 Greatest diameter 4.5 Habits. — Xothing is known of the habits of this species, beyond the fact that it is subterranean like all Typlilops. The specimens taken by myself at Ail^onito were found under fallen logs in a coffee planta- tion. Typlilops moiiensis Schmidt Text Fig. 41 Tijphlops lumhilcalis Meerwarth, 1901, Mitt. Naturh. Mus. Hamburg, Vol. XVIII, p. 5. Tijphlops monensis Schmidt, 1926, Publ. Field Mus. Nat. Hist., Zool., Vol. XII, p. 157, Fig. 1. Type locality. — Mona Island, West Indies. Distrihutio?}. —Ivnown only from Mona Island. Diagnosis — Allied to Typlilops lumhricalis Linne as defined by Cochran (1924, p. 17-1) by the number of scales from head to tail; distinguished by the more pointed snout, depressed head, and the successive increase in size of the three median scales behind the rostral. Original description. — -"Head depressed, snout strongly projecting, pointed when viewed from above ; diameter contained in the total length about 40 times, varying from 3.8 mm. anteriorly, to 4.8 mm. near the tail. Eostral broader than the first median scale behind it, not extend- ing as far back as a line drawn between the anterior borders of the eyes; nostril slightly below the rostral edge, on a suture which extends from the middle of the upper edge of the second upper labial to the rostral at the lateral angle ; preocular a little wider than the ocular, in contact with the third labial; eye very distinct; ocular large, with a nearly straight anterior edge, in contact with the third and fourth upper labials; three median scales behind the rostral (prefrontal, frontal, and interparietal) successively larger, the last nearly as large as the "parietal'' (or posterior supraocular) which separates it from the ocular; four upper labials, the last largest; nasals narrowly in contact behind the rostral. "Scale rows 20 anteriorly, 20 at mid-body, and 18 posteriorly; 321 scales from rostral to tail-spine on the vertebral line. 130 SCIENTIFIC SURVEY OF PORTO RICO "Color nearly uniform white, terminal part of each scale faintly dusky. "Total length 182 mm., tail 3 mm.'' Remarks. — The single paratype, Hamburg Museum No. 2039, agrees with the type in number of scale rows and in the details of head-scales.^ except that the nasals are narrowly separated by a rostral-prefrontal suture. The color is brown above, nearly white beneath, the brown pigment confined to the distal two-thirds of each scale. There is no trace of a white caudal ring or notch. Scales from rostral to tail-spine 313. It is interesting to find the Mona Island Typlilops related to the Cuban and Hispaniolan forms rather than to the Porto Eican. These two specimens were loaned for study through the courtesy of the ISTatur- historisches Museum of Hamburg. They are the only ones known. BOIDAE Epicrates Wagler The snakes of this genus are principally Greater Antillean, with rela- tives in South America rather than in Central America. There is a single species on each island, with the exception of Hispaniola, which has no less than three species. An anomaly of its distribution lies in the fact that the Bahaman specimens are conspecific with the Hispaniolan E. striatus. As in other boas, the snakes of this genus exhibit the claws at the sides of the vent which are the only external indication of the vestigial hind limbs. Synopsis of the Porto Rican and Mona Island Epicrates A. Indistinct dark dorsal markings 70-80; supraoculars about one-third as broad as the frontal E. inornatus. AA. Distinct dorso-lateral spots 51-57 ; supraoculars about one-half as broad as the frontal E. monensis. Epicrates inornatus (Reinhardt) Culebron Text Fig. 42 Boa inornata Reinhardt, 1843, Danske Vid. Selsk. Afhandl., Vol. X, p. 253, PI. 1, Figs. 21-23. CMlahothrus inornatus Jan., 1864, Icon. Ophid., Livr. 6, PI. 5, Fig. B, 1865; idem, text, livr. 2, p. 65.— Cope, 1868, Proc. Acad. Nat. Sci. Phila., p. 312.— Peters, 1876, Monatsber, Akad. Wiss. Berlin, p. 708.— Stahl, 1882, Fauna Puerto Rico, p. 70, p. 126, p. 160.— -Garman, 1883, N. Amer. Rept. I, Ophid., p. 132 ; 1887, Proc. Amer. Philos. Soc, Vol. XXIV, p. 279. Chilohothrus inornatus Gundlach, 1881, Anales Soc. Espan. Hist. Nat., Vol. X, p. 312. SCHMIDT, AMPHIBIANS OF PORTO RICO 131 Epicrates inornatus Stejneger, 1901, Proc. U. S. Nation. Mus., Vol. XXI II, p. 470 ; 1904, Kept. U. S. Nation. Mus., 1902, p. 688, Figs. 148-150.— Fowler, 1918, Papers Dept. Marine Biol., Carnegie Inst., Vol. XII, p. 14. Piesigaster boettgeri Seone, 1881, Abh. Senck. Ges., Vol. XII, p. 218, PL 1. Type locality. — Porto Eico. Distribution. — Confined to Porto Eico, where it is recorded from Bayamon, Caguas, Humacao and El Yunqne. Diagnosis. — More than nine shields on the top of the head; supra- oculars about one-third as broad as the frontal; ventrals 261-271; sub- FiG. 42.— Head of Epi- crates inornatus from above (left), after Stejneger. Head of Epicrates monensis from above (right), after Schmidt. caudals 67-75 ; scale rows at mid-body 38-42 ; about 75 dark spots in the dorsa,l row from head to vent. Original description. — "A boa with the head covered with irregular plates; eyes and nostrils lateral; labial scutes flat; dark in color (obso- lete fusca) with irregular diffuse markings posteriorly; ventral plates 264-271, subcaudals 67-69; dorsal scales at mid-body 39-41." RemarTcs. — Even the above brief description is entirely adequate, as there is only one boa in Porto Eico. In the twelve specimens known, the ventrals range from 261 to 271, the subcaudals from 67 to 7 dorsal scale rows from 36 to 43. -^5, the 132 SCIENTIFIC SURVEY OF PORTO RICO I am indebted to Mr. B. A. Wall for the photograph repro- duced as Plate IV, which constitutes a record of this species from El Yuuque. Color. — Stejneger describes the coloration in life as follows : "Nearly uniform 'bistre' with ventrals and subcaudals darker, narrowly pale- edged behind; above numerous indistinct crossbars (70-80 from neck to ^ent) of dusky color witli one or two scales nearly black, thus emphasizing the spots, of which all the component rows (dorsal, dorso- lateral, lateral and ventrolateral) are recognizable; the crossbars in- crease in width posteriorly; a blackish postocular band indistinctly con- nected witli a niedio-lateral faint longitudinal line on the neck; supra- labials fading into pale brownish gray at the commissure; slight traces of rufous on rostral and other shields of face; iris silvery gray clouded with dusky. "A somewhat smaller individual (1,500 mm. total length) brought home by Mr. Bowdish is very similar in coloration, only the underside is more slate color, and the pattern much more distinct, the crossbars show- ing paler centers with blackish margins; the spots of the lateral series show a tendency to form a lateral blackish line on the anterior third of the body. "Another specimen showed hardly any traces of bars or spots; gen- eral color al)ove, chestnut, darkest on the median region and tail, gradu- ally becoming lighter toward the ventrals; the latter brownish-slate color with pale edges; throat and chin mottled dull rufous and brownish slate; scattered obscure dusky spots on flanks." Habits. — Nothing is known of the habits of this species, Disfribution. — Confined to Porto Eico, where it is recorded from Bava- mon, Caguas, Humacao and El Yunque. Epicrates nionensis Zenneck Epic-rates moncnsis Zenueck, 1898, Zeitsclir. Wiss. Zool., Vol. LXIY, p. 64, PI. 3, Figs. 58-62.— Stejneger, 1904, Kept. U. S. Nation. Mus., 1902. p. 692, Figs. 153-157.— Schmidt, 1926, Publ. Field Mus. Nat. Hist., Zool., Vol. XII, p. 158, Figs. 2-3. Epicrafcs fordii var. monensis Meerwarth, 1901, Mitt. Naturh. Mus. Hamb., Vol. XVIII, p. S. Type locality.- — Mona Island. Distribution. — Entirely confined to Mona Island, where, with Cydura, it is a conspicuous link with the Hispaniolan fauna. Diagnosis. — More than nine head shields on the top of the head; supraoculars about one-half as broad as the frontal; ventrals 259-267; SCHMIDT, AMPHIBIANS OF PORTO RICO I33 subcaudals 79-88 ; scale rows at mid-body 33-43 : about 55 dark dorsal spots in the vertebral row on tlie body. Original description. — "The dorsal pattern on the body consists of the two uppermost rows of spots, the individual spots almost all united transversely. These dorsolateral spots are more irregular in sliape tlian in E. fordii. Their number ranges from 51 to 57. There is a single row of rather large spots on the sides, which frequently unite with the upper ones to form cross-bands. A faint postocular stri])e seems to be a continuation of the lateral row of spots. The fact tliat the lateral spots extend far downward toward the belly may indicate that they include a component part of the lower lateral row, but such a row is nowliere indicated. The dorsal rows of spots continue on the tail, while the lateral rows break off at the anus, (^n the head the lateral stripe may be absent or only very weakly indicated. The first of the dorsolateral spots lie on the posterior part of the head, and there arc very faint indications of a pair of stripes on the top of the head. Tlic dorsal ground color is light yellowish brown in juvenile specimens, the markings dark brownish black. In older specimens the ground color is much darker, so that the pattern is less sharply defined. "The distinctive peculiarities with reference to E. fordii are: a. Xuraber of spots in the dorsolateral row 51-57, as compared with 69-78. b. A single row of lateral spots instead of two. c. The frequent union of' the lateral and dorsolateral spots into crossbands. d. Head pattern very faint, except posteriorly. "The ventrals, in four specimens range from 259 to 263, the sub- caudals in two from 79 to 82. The dorsal scale rows range from 38 to 42." Remarks. — I have recently redescribed this species from a s])ecimen in Field Musemn of Natural History, collected by W. W. Brown in LS92. It is consequently more than thirty years since a specimen has been collected. Habits. — Xothing is known of the habits of this species. The tail of an Anolis cristatellus was found in the stomach of the specimen described by me. COLUBRIDAE Droniieus Kihion A word is necessary regarding the use of the generic names l)roini(iis and Alsophis. Stejneger shows that the type of Dromicus is D. cursor of 134 SCIENTIFIC SURVEY OF PORTO RICO Cuba, but he employs Leimadopliis for this snake and its allies because he regards Dromicus as preoccupied by Dromica Dejean, 1826. The International Commission has subsequently ruled that such a name as Dromica does not preoccupy Dromicus. It is therefore necessary to employ Dromicus for the snakes referred to Leimadopliis by Stejneger in 1904. Dunn, meanwhile (1922, p. 219), has shown that Dromicus, Alsopliis and Rhadinea intergrade in such a confusing way that he proposes to unite them all under the oldest name^ Dromicus. This results in a very large and very unwieldly group of snakes and, as the West Indian forms fall into two perfectly natural groups, I have in the present paper re- tained Alsophis as a distinct genus for the snakes with sharply defined scale-pits. Unfortunately I have complicated the synonymy by using Dromicus in Dunn's inclusive sense in describing Alsophis variegatus from Mona Island. Dromicus Stahli (Stejneger) Cnlebra Text Figs. 43 aud U Dromicus parrifrons (nee Cope). Peters, 1876, Monatsber. Akad. Wiss. Ber- lin, p. 708. — Gundlacli, 1881, Auales Soc. Espafi. Hist. Nat., Vol. X, p. 312.— Stahl, 1882, Fauna Puerto Rico, p. 70, p. 160. LeimadopMs stahli Stejneger, 1904, Rept. U. S. Nation. Mus., 1902, p. 695, Figs. 161-165.— Schmidt, 1920, Ann. N. Y. Acad. Sei., Vol. XXVIII, p. 198. Type locality. — Bayamon, Porto Eico. Distribution. — Confined to Porto Eico, where it is known from Ad- juntas, Aibonito, Bayamon, Caguas, Catalina Plantation, Ensenada, Humacao, Mayagiiez and El Yunque. Specimens collected. — 21 : Aibonito, Bayamon and Ensenada. Diagnosis. — Tail less than one-fourth of the total length ; eight supra- labials, three entering orbit; scales in nineteen rows; ventrals 116-166; subcaudals 83-97. Original description. — "Rostral much broader than high, scarcely visible from above; internasal suture shorter than prefrontal suture; frontal longer than its distance from end of snout, shorter than parietals, widely separated from preocular; supraocular narrower than frontal; nasal divided, longer than its distance from eye; loreal small, as high as broad, pentagonal ; one large preocular ; two postoculars ; one large an- terior temporal followed by two smaller ones; eight supralabials, second in contact with posterior nasal, loreal, and preocular; third, fourth, and SCHMIDT, AMPHIBIANS OF PORTO RICO 135 fifth supralabials in contact with eye ; eight lower labials, four in contact with anterior chin-shield, two in contact with posterior; anterior chin- shields much shorter than posterior ones; scales smooth, without pores, in 19 rows; ventrals, 157; anal divided; 89 pairs of subcaudals. "Color pattern : On a brownish ground a narrow dusky lateral line covering the adjacent edges of the fourth and fifth scale-rows; above this line a pale longitudinal band covering the remaining part of fifth, the whole of the sixth, and other * half of seventh rows ; a median dorsal darker band of six scale-rows is thus set off, a series of elongated dusky spots on the seventh row, three scales apart, forming the limit as a line of dashes ; head above with numerous dusky spots, and a longitudinal line on the middle of the frontal and the parietal suture which in com- bination with a spot on the posterior half of each supraocular, form a fleur-de-lis-shaped figure, the median line continuing some distance down Fig. 43. — Dromicus stahli, head from above and from below. (After Stejneger.) the back; a dusky, black-edged band on the side of the head from rostal through nostril and eye over temporals and connected with the continuous dark lateral line on fourth and fifth scale-rows; labials whitish with a dusky spot on the middle of each, and a dusky oblique band from the eye to the commissure crossing the suture between fourth and fifth supralabials ; underside whitish, dusted over with minute dusky specks, which show a tendency to congregate near the ends of the ventrals so as to form a line of ill-defined spots on each side of the abdomen." Re marls. — The Survey of Porto Eico collection contains twenty-four specimens of this species. The range in number of ventral plates is slightly greater in this series than in Stejneger's series of 146-166 in twenty-three specimens. The subcaudals range from 83-94. The sexes are scarcely distinguishable by these characters. The tail-length varies from .29 to .34 of the total (.29-. 31 in $ , .32-.34 in $ specimens). The scales about the l)ody are * Probably "lower half." 136 SCIENTIFIC SURVEY OF PORTO RICO imifonuly 19-19-17, The lower labials are nine, (eight in the original description). Freslily hatched specimens show the color pattern most distinctly, especially the median black markings on the head. The largest specimen, a female, measures 580 mm., tail 178 mm, Habifs. — This species proves to be still fairly abundant in Porto Rico, but its secretive habits have preserved it from scientific collectors as well as from the mongoose. It was noted at Aibonito beneath a log in a pasture, at the base of an extremely rotten stump in a coffee plantation, and one was observed that had been killed on the road by an auto- mobile. At Bayamon it was located by turning over stumps of trees which had been grubbed out, but which were quite firmly imbedded in the ground. Only one specimen contained identifiable remains in its stomach — a tail and egg of Anolis pulchellus. The tail was so coiled about tlie egg Fig. 44. — -Color-pattern of Dromicus stahli. (From Stejneger. ) that it appears probable that the egg was forced out after or during the swallowing of the lizard. It is likely that the lizard had attempted egg-laying under the edge of the stump which concealed the snake. Eggs of this species were found in three places : under a log in a pasture, and under an old termite nest in a coffee plantation, at Aibonito, and in the loose soil under a stump at Bayamon, One lot contained 7 eggs, another 13, and the third 40. The adult female found with the largest number of eggs contained 6 well-developed eggs. The eggs in this place were in three lots — IS old and discolored, in two clusters; 6 loose, somewhat different in ap- pearance; and two clusters of 6 and 10 eggs very fresh and white. Examination of the eggs showed that they contained emljryos in at least three stages, the fresher eggs having scarcely begun development, the oldest containing embryos nearly ready to hatch. The eggs found under the termite nest were also in two clusters, one of 7 eggs with advanced embryos, the other of 6, with no apparent development. The SCHMIDT, AMPHIBIANS OF PORTO RICO 137 older eggs are slightly larger, ranging from 21 to 25 mm. in length and from 12 to 15 mm. in diameter. The surface is finely striate, very white in the fresher specimens. It appears that the adult females of this species take up a location from which they do not wander far, and in which they lay successive batches of eggs, from 6 to 18 (?) in numl^er. The largest "nest" contained the remains of still older eggs which were either infertile or from which the young had hatched. The eggs are laid in clusters of 6 to 10, the individual eggs adhering firmly to the mass. The rate of reproduction is evidently fairly rapid. Droinicus exiguus Cope Text Fig. 45 Dromicun cjriguns Cope. 1862, Proc. Acad. Nat. Sci. Phila.. p. 79. — Reinhardt and Lnetken, 1863, Vid. Meddel. Naturh. Foren. Copenhagen, 1862, p. 216.— Garman, 1887, Proc. Amer. Philos. Soc, Vol. XXIV, p. 282.— Bou- lenger, 1894, Cat. Snakes Brit. Mus., Vol. II, p. 126.— Meerwarth, 1901, Mitt. Naturh. Mus. Hamburg, Vol. XVIII, p. 14. Leimadophis exiguus Stejneger, 1904, Kept. U. S. Nation. Mus., 1902, p. 698, Figs. 167-169.— Barbour, 1914, Mem. Mus. Comp. Zool., Vol. XTJV, p. 339. Type locality. — St. John and St. Thomas, Virgin Islands, Distribution. — St. Thomas, St, John and Culebra islands. Diagnosis. — Closely allied to Dromicus stahli, from which it differs in having fewer ventrals, 134-146 instead of 151-159. Fig. 45. — Head of Droinicus exi- guus from above and from the side. (After Stejneger.) Original description.— "f>ize small; body stout; head little distinct, flat above, muzzle prominent, Eostral plate broad, presenting no su- perior surface. Prefrontals well-developed. Vertical elongate, lateral borders straight, the posterior long, forming an acute angle. Occipitals 138 SCIENTIFIC SURVEY OF PORTO RICO well developed, the median or common suture shorter than the vertical plate, obtuse posteriorly, bounded by one large and five small temporals on each side. Postoculars two ; preocular one, rather broad ; loreal small. Postnasal longer than prenasal. Eight superior labials, third, fourth, and fifth entering orbit. Nine inferior labials, fourth and fifth largest. Scales in nineteen longitudinal rows. Total length of largest of five specimens 17 in. 1 lin. ; tail 5 in. 4 lin. "Above light brown, sometimes yellowish, densely punctuated with darker. ^The median dorsal region is of a deeper shade; distant dark brown spots sometimes form two parallel series, one of each side of it. A dark brown band along the fourth row of scales nearly to the end of the tail; it is sharply defined only superiorly; it is continuous with a head band which passes through the eye. Beneath yellowish, punctu- ated with brown, especially toward the extremities of the gasterosteges." Remarks. — This is the only species of snake on our list which has not passed through my hands. Stejneger's description of a Culebra specimen and his remarks on variation in this species are as follows : "Eostral scarcely visible from above; internasal suture shorter than prefrontal suture; frontal long, longer than the parietal suture, but shorter than the parietals; loreal (abnormally) joined to prefrontals; one preocular; two postoculars; one long anterior temporal, and two smaller posterior ones; eight supralabials, third, fourth, and fifth enter- ing eye (on left side nine, fourth, fifth, and sixth entering eye) ; posterior chin-shields longer than anterior ones; 19 rows of smooth scales without pores; 144 ventrals; anal divided; subcaudals, 82 pairs. Color as de- scribed under Leimadophis stahli, p. 695, but paler. Dimensions Tip of snout to tip of tail 310 mm. Vent to tip of tail 100 "The above specimen is abnormal in having no loreal. Ordinarily the loreal is very small, sometimes even rudimentary, and Eeinhardt and Luetken mention a specimen having none on the left side, preocular and postnasal being in contact. Garman mentions a specimen in the mu- seum at Cambridge, Massachusetts, having the prefrontals fused on the median line. The normal number of supralabials is 8, but the Culebra specimen described has 9 on one side. Meerwarth describes a similar specimen from St. Thomas. Ventrals (in 19 specimens recorded) vary between 134 and 146, subcaudals between 79 and 86 pairs. "The only essential difference between L. exiguus and L. stahli seems to be the lower number of ventrals in the former. Altogether 29 speci- SCHMIDT, AMPHIBIANS OF PORTO RICO 139 mens of both species have been examined and recorded, and in these the difference is marked and constant." Habits. — Nothing is known of the habits of this species. Alsophis Fitziuger Alsophis antillensis (Schleged) Text Figs. 46 and 47 Psammophis antillensis Sclilegel, 1837, Pbys. Serp., Vol. II, p. 214. Dromicus antillensis Dumeril and Bibron, 1854, Erpet. Geu., Vol. VII, Part 1, p. 659.— Giinther, 1858, Cat. Colubrine Snakes Brit. Mus., p. 129.— Cope, 1860, Proc. Acad. Nat. Sci. Pbila., p. 560.— Jan.. 1867. Icon. Ophid., livr. 25, PI. 1 ,Fig. 1.— Boulenger, 1894, Cat. Snakes Brit. Mus., Vol. II, p. 123.— Meerwartb, 1901, Mittb. Naturb. Mus. Hamburg, Vol. XVIII, p. 12, PI. 1, Fig. 13. Alsophis antillensis Cope, 1862, Proc. Acad. Nat. Sci. Pbila.. p. 76.— Rein- bardt and Luetken, 1863, Vid Meddel. Naturb. Foren., 1862. p. 218.— Gar- man, 1887, Proc. Amer. Pbilos. Soc, Vol. XXIV, p. 282.— Stejneger, 1904, Kept. U. S. Nation. Mus., 1902, p. 704, Figs. 171-174.— Barbour, 1914, Mem. Mus. Comp. Zool., Vol XLIV, p. 336; 1917, Proc. Biol. Soc. Wasb., Vol. XXX, p. 101.— Fowler, 1918, Papers Dept. Marine Biol., Carnegie Inst. Wasb., Vol. XII, p. 14.— Scbmidt, 1920, Am. N. Y. Acad. Sci., Vol. XXVIII, p. 199. Alsophis anegadxe Barbour, 1917, Proc. Biol. Soc. Wasb., Vol. XXX, p. 102 (type locality, Anegada, outer Virgin Ids.). Type locality. — St. Thomas, Guadeloupe, Martinique and Cuba. Re- stricted to St. Thomas by Giinther. Distribution. — A^irgin Islands and Porto Eico. It is known from St. Thomas, St. John, Virgin Gorda and Anegada, from Vieques and Cule- bra, and from Coamo Springs on Porto Eico. Specimens collected. — 2 : Coamo Springs. _ Diagnosis. — Dorsal scales usually in nineteen, occasionally in seven- tween, rows; a pair of pits present at the tip of each scale; a row of black spots on the lower half of the fifth scale-row; ventral plates 170-189, caudals 116-144. Original description.— ''This species has the habitus of Psammophis inoniligev, to which it is also somewhat allied by its color pattern, but its head is much wider at the base and sharply conical; the narrow snout ends in a blunt point; the head-shields are almost like those of Psam- mophis moniliger, with the exception of the frontal, which is usually less elongate. ''The Antillean Psammophis has all its, teeth equal in length; it reaches a length of three feet; three specimens of diverse ages measure 140 SCIENTIFIC SURVEY OF PORTO RICO respectively (in mm.): 760 + 290, 490 + 180, and 180 + 80; the ventrals var}^ from 178 + 100 to 204 + 144; there are seventeen to nine- ten rows of smooth, lanceolate scales. "The lower parts are yelloAv ; upper parts yellowish brown ; back with three narrow black bands, the median composed of two fine serrated f-iG. 46. — Alsoptiis antillcnsis, head from above and from side. (After Stejneger.) lines, the two on the sides of a great number of small dots; the lateral lines pass through the eye to the sides of the snout; this pattern is most distinct in young specimens; in older specimens the lines are obscure or converted into a network formed by the black borders of the scales." Fig. 47. — Color-pattern of Alsophis anfillensis. (From Stejneger.) Be marks. — The original description, based on specimens from St. Thomas, Guadeloupe, Martinique and Cuba, is a "composite species." The name has come to be restricted to the Virgin Island form by the con- sensus of opinion among herpetologists. Lidth de Jeude records Dromi- cus antiUensis from Curacao, and this is repeated by Euthven (1923, p. 9). This can scarcely be the Virgin Island species. SCHMIDT, AMPHIBIANS OF PORTO RICO 141 The identification of the two specimens collected by me at Coamo Springs as this species removes the element of geographical distinctness from the allied A. portoricensls. The male specimen has only seven- teen scale rows, and so might be identified with A. portoricensis, were it not that the coloration of both specimens is nearly typical of A. antil- lensis, while the female has nineteen scale rows at mid-body. In view of the higher number of ventral plates and the distinct coloration, I prefer to retain portovicensis and antiUensis as distinct species. The two specimens agree closely in coloration with the color variety described by Barbour fron Anegada and, as I do not wish to admit of a discontinuous distribution of .4. anegadce, it seems best to include both Porto Eican and Anegadian specimens with A. antiUensis. The measurements and scale characters of the two Porto Eican speci- mens are as follows : AM. N. H. Parts measured No. 13305 d 13306 ? Length 707 mm. 820 mm. Tail 24.5 " 270 " Tail length 35 -33 Ventral plates 184 185 Subcaudals 134 132 Dorsal scales 17-17-15 17-19-15 Much the best description extant is that of Stejneger, based on a Culebra specimen, which I quote in full : "Eostral much broader than high, barely visible from above ; internasal suture scarcely shorter than the prefrontal suture ; frontal broader than supraocular, about equaling its distance from the tip of the snout and the parietal suture; nostril between two nasals; loreal moderate, trapezoid, the posterior border being strongly convex ; one preocular separated from frontal ; two postoculars, the lower one very narrow ; temporals 1 + 3 ; 8 supralabials, third, fourth, and fifth entering eye, the fifth and follow- ing ones abruptly much higher than the anterior ones; 5 lower labials in contact with anterior chin-shield, which is much shorter than the posterior; 19 rows of smooth scales with two conspicuous apical pores; 183 ventrals; anal double; 118 pairs of subcaudals. Color (in alcohol) above brownish drab, the individual scales irregularly tipped and edged with dusky; underneath whitish with dark drab mottlings on chin and throat and a series of similarly colored dots on the lateral canthus of each ventral shield, forming a dotted line on each side of the abdomen, each ventral, moreover, posteriorly more or less irregularly edged with brownish drab; a few brownish irregular spots on the labials and upper 142 SCIENTIFIC SURVEY OF PORTO RICO head shields, with a double series of elongate brownish spots on the upper neck ; from anterior nasal through eye a dark-brownish streak continuing on the sides of neck and body as a broken line of elongate spots; these spots which on the sides of the body occupy the lower half of every second or third scale in the fifth scale row, the upper half being whitish or decidedly paler than the ground color." Habits. — Little is known of the habits of this species. One of my specimens was shot in an arroyo behind the Coamo Springs Hotel, the other was found at the base of the clifE behind the bath houses, beneath a stone. This specimen bit fiercely when captured. The stomach of one specimen contained the tail of an Anolis. Alsophis portoricensis ( Reinhardt and Luetken) Culebra Text Figs. 48 and 49 Alsophis portoricensis Reinhardt and Luetken, 1863, Vid. Meddel. Naturh. Foren. Copenhagen, 1862, p. 221. — Peters, 1876, Monatsber. Akad. Wiss. Berlin, p. 708.— Gundlach, 1881, Anales Soc. Espan. Hist. Nat., Vol. X, p. 313.— Stejneger, 1904, Rept. U. S. Nation. Mus., 1902, p. 700, Fig. 170. (part).— Barbour, 1914, Mem. Mus. Comp. Zool., Vol. XLIV, p. 335.— Fowler, 1918, Papers Dept. Marine Biol. Carnegie Inst., Vol. XII, p. 14.— Schmidt, 1920, Am. N. Y. Acad. Sci., Vol. XXVIII, p. 199 (part).— Danforth, 1925, Copeia, No. 147, p. 79. N"o common name other than "culebra." Type locality. — Porto Eico. Distribution. — Confined to Porto Eico, Desecheo Island and Caja de Muertos. On Porto Eico it is recorded from Adjuntas, Bayamon, Hu- macao, Mayagiiez and Utuado. Specimens collected. — 3 : Adjuntas and Caja de Muertos. Diagnosis. — ]N"ine large shields on top of head; dorsal scales in seven- teen rows, usually 17-15-14; dorsal scales with a pair of distinct pits at the tip of each; dorsal scales dark brown outlined with black, ventrals each with black posterior border. Oi-iginal description. — "Color (in alcoholic specimen), rufous brown above, the individual scales black bordered; yellow beneath, the abdomi- nal and subcaudal plates black bordered ; scales in seventeen rows, ventral plates 175, subcaudals 122." Remarks. — The N". Y. Academy's three specimens were secured by H. E. Anthony. Through the courtesy of Dr. Stuart T. Danforth, I have been enabled to examine the specimen of this species mentioned by him, and he also loaned for study a specimen secured by him on Desecheo SCHMIDT, AMPHIBIANS OF PORTO RICO 143 Island in 1926. Mr. M. Graham Netting of the Carnegie Museum kindly called my attention to the existence of two Porto Eican snakes of this species in the collections of the Carnegie Museum. I am further in- debted to Miss Doris Cochran for information on the specimens in the U. S. National Museum, and to Mr. Arthur Loveridge for notes on the two specimens in the Museum of Comparative Zoology. In all I have notes on thirteen specimens from Porto Eico, on one from Caja de Muertos and on two from Desecheo Island. My belief in the distinctness of the Mona Island species, which I dis- tinguished in 1926 primarily on what appeared to be a constant differ- PiG. 48. — Head of AlsopMs portoricensis from above, showing scale-pits in dorsal scales. (From Stej- neger.) ence in coloration, was rudely shaken by the discovery of pale colorations in Porto Eican specimens, especially Carnegie Museum No. 1333, from Ad juntas, and the Caja de Muertos specimen. An unexpected distinc- tion turns up, however, in the fact that the normal scale reduction in Porto Eican specimens is 17-15-14 or 17-15-13. When I first observed this scale count in the two Ad juntas specimens, I supposed it to be an anomaly. Ten of the thirteen specimens have the formula 17-15-14, two have 17-15-13, and one has 17-15. The Caja de Muertos specimen re- duces to 14 scale-rows, and in the Desecheo Island specimens one has the formula 17-15-14, the other 17-15. Eight Mona Island specimens have 15 scale rows anterior to the anus. .x^S ^ t i s^ R A k Yj 144 SCIENTIFIC SURVEY OF PORTO RICO It is not unlikely that a pale form of true portoricensis formerly oc- curred in the arid district of southwestern Porto Eico, and the Muertos Island specimen is evidence to this effect. The Desecheo Island specimens agree quite closely in coloration with the Mona Island, species, and their relations are evenly divided by their scale counts. Additional specimens are evidently required to settle the principal affinities of the Desecheo population of this species. .'Xl/MTiWl Fig. 49. — Color-pattern of Alsophis porto- ricensis. A. M. N. H. No. 8435. Twice natural size. The range of ventrals and subcaudals in male specimens is 169-179 and 125-134. In females these counts are 175-183 and 115-128. The typical coloration was described by Eeinhardt and Luetken, and our Ad Juntas specimens and Stejneger's Humacao specimens agree excel- lently. The dorsal scales are dark brown, each scale outlined with black, and all except the anterior ventrals are heavily margined with black at their free edges. Habits. — Nothing is known of the habits of this species. Alsophis variegatus (Schmidt) Text Fig. 50 Dromicus sanctae-crucis var. portoricensis Boulenger, 1896, Jahresb. Naturw. Ver. Magdeburg, 1894-1896, p. 313; Meerwartli, 1901, Mitt. Naturh. Mas. Hamburg, Vol. XVIII, p. 11. Dromicus sanctae-crucis Bouleuger, 1896, Cat.. Snakes Bi-it. Mus., Vol. Ill p. 634 (not of Giinther). SCHMIDT, AMPHIBIANS OF PORTO RICO 145 Alsophis portoricensis Stejneger, 1904, Rept. U. S. Nation. Mus., 1902, p. 700, Fig. 170, (part).— Schmidt, 1920, Ann. N. Y. Acad. Sci., Vol. XXVIII, p. 199 (part). Dromicus variegatus Sclimidt, 1926, Publ. Field Mus. Nat. Hist., Zool., Vol. XII, p. 160, Figs. 4-5. Type locality. — Mona Island. Distribution. — Confined to Mona Island. Specimens collected. — 3 : Mona Island. Diagnosis. — Allied to Dromicus portoricensis in scale characters, and distinguished chiefly by its coloration, in which the regular reticulation of black of the dorsal scales and the black borders of the ventrals are absent. Dorsal scale formula 17-15. Original description. — "Habitus unspecialized ; venter weakly angu- late; head large and well distinguished from the neck, somewhat de- FiG. 50. — Color-pattern of Alsophis varie- gatus. A. M. N. H. No. 13774. pressed ; body rather slender. Eostral wider than high, just visible from above ; internasal suture two-thirds that of the prefrontals ; frontal longer than its distance from the end of the snout, as long as the parietal suture ; parietals large; nasal divided; loreal small, 5-sided; a single large pre- ocular, extending to the upper side of the head, not in contact with the frontal; two postoculars, the lower much the smaller; temporals }^, with a well marked groove between them and the labials; upper labials eight, the third, fourth and fifth entering the eye; lower labials ten; chin shields slender, the posterior pair much longer than the anterior. ''Dorsal scales 17-17-15; ventrals 173; tail incomplete. 146 SCIENTIFIC SURVEY OF PORTO RICO "Top of head with a few brown spots on a lighter ground color ; a black lateral line from the nostril through the eye, extending some dis- tance on the neck; a short nuchal black line from the parietal suture to the constriction of the neck; upper and lower labials and chin light, punctulate with brown dots; venter immaculate anteriorly, with slight brown markings on the angle and on the posterior margins of the ventrals toward the tail; subcaudals Avith narrow brown markings parallel to but not at the rear border ; posterior margins of the dorsal scales with irregu- lar black markings, tending to form zig-zag cross-bands." Bemarls. — The two specimens collected on Mona Island and men- tioned in my paper in the Annals of the New York Academy of Sciences, A. M. N. H. Nos. 13773 and 13774, may be named as para- types of this species. They agree with the type described above, with the wavy dorsal cross-bands somewhat better developed. They show no approach to the coloration of the specimens of portoricensis examined by me, and Stejneger (loc. cit).) contrasts his Porto Eican specimens in the same way with one from Desecheo Island. The ventrals, caudals, and measurements of the two male paratypes are as follows : ventrals 177, 179; caudals 125, 113; total length 661 mm., 780 mm.; length of tail 213 mm., 2-48 mm. Another specimen, a female, collected by An- thony in 1926, agrees with the coloration described above. It has 179 ventrals. Meerwarth's notes on the coloration of the twenty-five Mona Island specimens examined by him confirm the constancy of this character. The extremes and averages of the ventral and caudal counts on record are as follows : No. of specimens Ventrals 41 Caudals 30 There is certainly an average difference in both caudals and ventrals in the two sexes, but the extremes appear nearly to coincide. Mr. H. W. Parker writes me that the four Mona Island specimens in the British Museum have the scale formula 17-15, and this agrees with that of the four specimens examined by me and affords the most con- stant distinction from Alsophis portoricensis. Habits. — The two specimens secured by myself on Mona Island were found hidden beneath rubbish on the low terrace of cultivated land on the south side of the island. The stomach of one specimen contained the remains of two Ameiva tails, and that of the other one tail of the same lizard. Extremes Average 170-181 176 112-126 120 SCHMIDT, AMPHIBIANS OF PORTO EICO 147 Order TESTUDINATA t Emydidae Pseudemys Gray Each of the islands of the Greater Antilles is inhabited by a species of fresh-water turtle belonging to the genus Pseudemys. This genus has a large development in Eastern North America and in C'entral America. Pseudemys stejiiegeri, sp. nov. Text Figs. 51 and 52 Emys rugosa Stahl, 1882, Fauna Puerto-Rico, p. 68.— Garman, 1887, Proc. Amer. Pliilos. Soc, Vol. XXIY, p. 286. * Clemmys dccussata Peters, 1876, Monatsber. Akad. Wiss. Berlin, p. 705.— Gundlach, 1881, Anales See. Espan. Hist. Nat., Vol. X, p. 307. Pseudemys palustns Stejneger. 1904. Kept. U. S. Nation. Mus., 1902, p. 710, Figs. 179-186. Type locality. — San Juan, Porto Eico. Distribution. — Recorded only from C'aguas, San Juan, Desengaiio (Cartagena Lagoon) and Guanica Lake. Diagnosis. — A Pseudemys closely allied to the Pseudemys palustris of Jamaica and Hispaniola, from which it is distinguished by smaller size and by having the axillary and fifth marginal shields usually not in contact. Type.—U. S. X. M. Xo. 25642, San Juan, Porto Rico. Adult female collected by the U. S. Fish Commission ''Fish Hawk" Expedition. Description of type. — ' ''Shell moderately convex, the height being more than one-half the greatest width; length of carapace less than two and a half times the height of the shell and about one and one-third times its greatest width; carapace faintly keeled and with longitudinal wrinkles crossed by radiating ridges, which are especially strong on the anterior costals ; nuchal narrow ; first vertebral shield urceolate, anterior and posterior sutures of same length; lateral sutures of second, third, and fourth vertebrals much longer than the anterior and posterior sutures; vertebrals much narrower than costals; posterior margin of carapace slightly serrate, each of four posterior marginals on each side being faintly emarginate; carapace broader behind than in front, the posterior marginals flaring out considerably ; plastron less than two- thirds and more than one-half the greatest width of the carapace; the posterior lobe a trifle wider than the anterior, its length much less tlum 1 Quoted from Stejneger, 1904, p. 711. 148 SCIENTIFIC SURVEY OF PORTO RICO the width of the bridge ; abdominal suture longest, equaling those of the pectorals and femorals together; humeral suture shortest; gulars pro- jecting, cut off square anteriorly; plastron slightly emarginate behind; axillars and inguinals large, latter largest; head moderate; snout short, pointed, feebly projecting; upper jaw with a very slight median notch, no cusps; jaws feel)ly denticulated; alveolar surface broad, with a deep notch behind on the median line; symphysis of mandible as broad as one-half the longest diameter of the orbit; digits connected with broad webs. Color (in alcohol) of carapace above nearly uniform tawny olive ; plastron yellowish, with obscure dusky symmetrical sinuous mark- ings all over; top of head without markings; yellowish lines narrowly Fig. 51. — Carapace and plastron of Pseudemys stejnegeri. One-half natural size. (From Stejneger.) edged with blackish on sides and under surface of head and neck, one from the nostrils crossing the upper jaw obliquely and ending abruptly at the posterior angle of the mandible, another from above the nostrils, crossing the eye of the lower posterior edge of the orbit, and thence obliquely down and backward to the corner of the mouth, continuing backward under the tympanum down the side of the neck; two fainter lines, one between the two just described and one above the transocular line, crossing the tympanum; a line on the symphysis of the mandible bifurcating on the chin and a third median line originating on the chin a short distance behind the fork, the three continuing parallel SCHMIDT, AMPHIBIANS OF PORTO RICO 149 down the under side of the neck ; two similar bnt wider lines on the upper side of the fore legs and two on the under side of the hind legs." Dimensions mm. Length of carapace 232 Width of carapace anteriorly 150 Width of carapace posteriorly 170 Height of shell 95 Width of anterior plastral lobe 90 Width of posterior plastral lobe 931 Width of bridge 88 Width of head 31 ♦ It must be added that the relation between the axillary and fifth marginal shields in this specimen is the normal one, i. e., that they are widely separated by a suture of the fourth marginal with the pectoral, as is illustrated in Stejneger's figure of another specimen, reproduced here (Fig. 51). Remarks. — Stejneger makes the comment with reference to this turtle that "There are indications at hand that there may be some constant differences between those inhabiting the different islands, but the ma- terial at my disposal is not sufficient to warrant an attempt to separate them." I have seen nineteen Porto Eican specimens and thirteen His- paniolan, but I have nevertheless hesitated at separating the Porto Eican specimens as a distinct species. I feel that there are now certain indica- tions at hand that there are two forms of this turtle in Hispaniola, which lack of material prevents me from distinguishing; and this un- certainty as to the Hispaniolan forms does not clarify the relations of the Porto Eican species. The character chosen as distinctive, the con- tact of the axillary shield with the fifth marginal or its exclusion from the fifth marginal by a contact of the fourth marginal with the pectoral shield is a trivial one. The specimens examined vary in this respect as follows : AxiUary reaching Not i-eaching: Localities 5th marginal .5th marginal Porto Rico 3 15 Hispaniola 11 2 Cuba 11 1 Jamaica 2 In the series of paratypes— U. S. N. M. No. 256-13, 25644 and 25653, A. M. K H. No. 15186, and F. M. N. H. No. 12476-12489 inclusive (the latter ex Danforth collection) — the length of the carapace of the type is not exceeded. The extremes of the Danforth -series are 150 SCIENTJFIC SURVEY OF PORTO RICO 100-179 mm., the average 124: mm. Three Hispaniolan specimens measure 333, 334 and 341 mm., and 1 have seen much hirger specimens at Monte Cristi. Additional information as to the adult size of Porto Eican specimens is, however, much to be desired. Fowler has discussed the color dimorphism in this species in some de- tail, and Danforth comments on it as follows : "There is a popular idea that there are two species, a green and a black one, but I have seen inter- grades between the two.'" This feature of the Porto Eican species is un- known in the Hispaniolan Pseudemys. Fig. 52. — Head of Pseudemys stejneyeri from below and from side, to show color-pattern. (After Stejneger.) Habits. — Nothing is known of the habits of this species except for the observations of Danforth (1935), which I quote: "By April they were laying eggs. For that purpose they come out on land at night, and the natives choose that time to hunt them with the aid of lights. They are sold in the markets for food. These turtles are only rarely seen sun- ning themselves." a hand-list of the amphibians and eeptiles of the vieCtIn islands In view of the fact that the Virgin Islands are frequently visited by naturalists en route for other localities, I have drawn up a table of the known distribution of the species, and added artificial keys and notes on some of the questions of interest which remain for investigation, in the hope that they may be found useful. SCHMIDT, AMPHIBIANS OF PORTO RICO 151 Distributional List of the Amphibians and Reptiles of the Virgin Islands Bufo turpis Leptodactylits albilabrifi . . . . Eleutherodactylus antillensis Eleutherodactylus lentus... Hemidactylus mabouia .... Thecadactylus rapicaudus . . Sphaerodactylus viacrolepis . Anolis cuuieri Anolis cristatellus Anolis stratnlu.s Anolis pidchelhis Anolis acutus Iguana iguana Cyclura pinguis Ameiva exsul Ameiva polops Amphisbaena fenestrata . . . . Mabuya sloanii Typhlops richardii Dromicus exiguus Alsophis antillensis Alsophis sandi-crucis , Total No. Species 22 o Eh o PL, X X X X X X X X X 12 a. 3 X X X X X X X X X 10 X X X X X X X 10 03 o X X X X X X X X X X X X X X 16 c o 1-5 m X X X X O H X X X X X X X Q c > CO O •-5 X .X X 03 O O M ;_ X X X o3 -o 03 be OJ C < X X X X X o o 02 X X X X X X X X X X X 13 XOTES ON HeRPETOLOGY OF THE YlKGIN ISLANDS 1. Amphibians. 1. Bujo iiirpis Barbour. Beadily recognized as the only toad in the islands. Known only from a single specimen collected on Virgin Gorda by James Lee Peters in 1915. Additional specimens for further com- parison with the Porto Rican toad are much to be desired. 2. Leptodadylus alhilabris (Giinther). General aspect frog-like. Barbour has called attention to apparent differences in specimens from St. Croix, and suggests that the species is adapting itself to burrowing habits there. 3. Eleutherodactylus antillensis (Eeiuhardt & Luetken). Coloration variable, usually uniform grayish brown, the concealed surface of the thighs reticulated with black. 4. Eleutherodactylus lentus (Cope). The uniformly mottled colora- 153 SCIENTIFIC SURVEY OF PORTO RICO tion and the light dorsolateral lines readily distinguish this species. Its note is imdescribed, and its breeding habits are quite unknown, II. Eeptiles. 1. Hemidactylus mahouiu (Moreau de Jounes). It is difficult to understand why this introduced form has not become more common. It may be looked for at night on the walls of buildings near electric lights. 2. Thecaductylus rapicaudus (Houttuyn). It is not known whether this species has become established in St. Thomas and St. Croix. 3. Sphaerodactylus macrolepis Giinther. The range of variation in size of dorsal scales should be determined for Virgin Island specimens, for comparison with the data given above for the Porto Rican series. 4. Anolis cuvieri Merrem. The giant Anolis is recorded only from Tortola. It may be extinct even there, as there is no recent record. It might be looked for on St. John. 5. Anolis cristatelhis Dumeril & Bibron. The common Anolis of fence posts and open brush. 6. Anolis stratulus Cope. Often associated with A. cristatellus but a little more arboreal in its habits in Porto Rico. It should be looked for on St. Croix. 7. Anolis pulchellus Dumeril & Bibron. Also associated with A. cris- tatellus, this species is readily recognized by its more slender body and longer tail. Another species {Anolis richardii Dumeril & Bibron) with a slender body, keeled ventral scales and the occipital scale in contact with the scales bordering the orbits, was described from Tortola. Special search by Mr. Peters, who explored the outer Virgin IsJands for the Museum of Comparative Zoology, failed to re-discover this species. Anolis l-rugi, another closely allied species, might be looked for in the more shaded and moist localities on St. John. 8. Anolis acutus Hallowell, Related to Anolis pulchellus but confined to St. Croix, this species has not recently been recorded. It should be compared with Anolis poncensis of the arid district in Porto Rico for possible relationship. 9. Iguana iguana (Linne). This species is much used for food in many localities, which probably accounts for its introduction in St. Thomas. It does not appear to have become well established, but Barbour records a specimen of iguana from Water Island, near St. Thomas, in 1917. SCHMIDT, AMPHIBIANS OF PORTO RICO 153 10. Cyclura pinguis Barbour. Known from a single specimen secured on Anegada by Mr. Peters. It should be further compared with the extinct Cyclura mattea Miller, from St. Thomas. 11. Ameiva exsul (Cope). Apparently exterminated in St. Thomas by the mongoose, the ground lizard is still found on the adjacent Water Island. 12. Ameiva polops Cope. Known only from the type from St. Croix. It should be looked for on the tops of the limestone hills, in the same habitat as that of A. luetmbrei of Porto Rico. 13. Ampliishaena fenestrata Cope. This species may be looked for wherever there is tillable soil. Specimens from St. Croix should be com- pared with those from St. Thomas for possible differences. 14. Mabuya sloanii (Daudin), A rare species. Virgin Island speci- mens have a somewhat different coloration from those of Porto Eico. 15. TypMops richardii Dumeril and Bibron. This burrowing blind snake can usually be secured through people who are cultivating or plowing. A series from both St. Thomas and St. Croix would be of interest for comparison with the Porto Eican specimens described above. 16. Dromicus exiguus Cope. This species like L. stalili of Porto Eico, may prove more abundant than is believed to be the case. It prob- ably is found in similar situations. 17. Alsophis antillensis (Schlegel). Formerly abundant on St. Thomas, now apparently rare. 18. Alsophis sancti-crucis (Cope). The present status of this species on St. Croix is unknown. It was not found by Noble or Euthven, who visited the island in 1914. ARTIFICIAL KEYS TO THE SPECIES I. Frogs and Toads fSkin rough; head with bony ridges Bufo turpis ^' ^ Skin smooth ; head without bony ridges ... 2 fTips of digits not at all dilated ; thigh with J dark crossbars LepodacUjhis alhilalris "■ 1 Tips of digits slightly or considerably di- lated ; thighs mottled, not barred 3 Tips of digits slightly dilated; belly smooth ; back mottled Eleutherodactylus lent us . Tips of digits well dilated ; belly granular ; back usually uniform or with narrow light line in the middle Eleutherodactylus antillen- sis 154 SCIENTIFIC SURVEY OF PORTO JiJ CO o J II. Lizards and Snakes CLimbs well developed 2 ■ ) Limbs absent 14 No eyelids ; skin soft, often broken in catching; digits more or less dilated.... .3 Eyelids present ; skin firm ; digits dilated or not 5 'Digits dilated only at tip, with circular plate beneath ; skin covered with over- lapping scales ; size small, less than three inches SpJiaerodactylus macrolcitis Digits broadly dilated, with transverse lamellae beneath ; skin of back covered with granular scales not overlapping ; adults larger, exceeding three inches.... 4 'Slender terminal joint bearing claw be- yond the expanded portion of digits Hemidactylus niahouia No slender terminal join on digits : claw concealed in slit Itetween expanded sides of digits Theeadactylus rapicaudns i i Digits dilated, with slender terminal joint beyond dilation ( A;io//.s-) 6 Digits not dilated 10 Scales on back (closely examined) consist of larger scales entirely surrounded by smaller granules Anolis cuvieri Scales on back not as above 7 ["Ventral scales keeled ; dorsal scales more J or less enlarged in vertebral region 8 Ventral scales smooth 9 'Enlarged occipital scale (largest median scale on head behind orbits) separated from enlarged scales bordering orbits by one or more scales Inolis pulchellus Occipital scale in contact with scales bor- dering orbits Anolis acutus 'Back with four or five well-defined trans- verse spots ; throat fan of male uniform orange Anolis stratultis 9. J Coloration extremely variable ; female usually with light longitudinal mid-dorsal I stripe; throat fan of male greenish i yellow, brownish orange at edge Anolis cristatellus 8. i 10. SCHMIDT, AMPHIBIANS OF PORTO RICO J 55 Uuder side of body covered with large plates in regular longitudinal and trans- verse series H I Under side of body covered with overlap- I ping scales like those of sides and back. .12 f Eight rows of ventral plates Imeiva polops ^^' )Ten or twelve rows of ventral plates Ameiva exsul f No dorsal fold or crest Mabuya sloanii ^~' ) Well-defined dorsal crest of spines 13 CDorsal crest continuous Iguana iguana ^^- ) Dorsal crest interrupted on rump Cyclura pinguis f Eye concealed beneath skin 15 ^^- ) Eye distinct 16 rBody covered with overlapping scales Typhlops rkhardii 15. J Skin divided into small rectangular seg- I meuts, arranged in regular rings Amphisbacna fcnestrata 'No pits at tips of scales ; dorsal scales in 19 rows; ventral plates 134-146; sub- 16. \ caudals 79-86 Dromicus exiguus A pair of distinct pits or pores near tip of each scale 1'7 ["Dorsal scales in 17 rows ; ventral plates I 191-195 ; subcaudals 145-147 Alsoph is sancti-crucis -|7 J ■ 1 Dorsal scales in 19 rows ; ventral plates I 170-189 ; subcaudals 116-144 AUophis anUllensis BIBLIOGRAPHY ^ Barbour, Thomas 1910. Notes on the herpetology of Jamaica. Bull. Mus. Comp. Zool., Vol. LII, pp. 271-301, Pis. 1-2. 1914. A contribution to the zoogeography of the West Indies, with es- pecial reference to amphibians and reptiles. Mem. Mus. Comp. Zool., Vol. XLIV, pp. 200-359, PI. 1, Text Fig. 1. 1915. Recent notes regarding West Indian reptiles and amphibians. Proc. Biol. Soc. Wash., Vol. XXVII, pp. 71-78. 1916. Additional notes on West Indian reptiles and amphibians. Idem., Vol. XXIX, pp. 215-220. 1917. Notes on the herpetology of the Virgin Islands. Idem, Vol. XXX, pp. 97-104. 1919. Herpetological notes. Notes on Celestus. Proc. New England Zool. Club., Vol. VII, pp. 11-13. 1 See H. E. Anthony, 1926, Mammals of Porto Rico, Living and Extinct, Vol. IX, Part 2, of the present series, for references on zoogeography of the West Indies in general. 156 SCIENTIFIC SURVEY, OF PORTO RICO 1919a. A new rock iguana from Porto Rico. Proc. Biol. Soc. Wash., Vol. XXXII, pp. 145-148, PI. I. 1920. Sphacrodactylus. Mem. Mus. Comp. Zool., Vol. XLVII, pp. 215-278, Pis. 1-26. 1923. West Indian investigations of 1922. Occ. Papers. Mus. Zool. Univ. Mich. No. 132, pp. 1-7. 1923a. Reptiles in the East and West Indies — and some digression. Amer. Nat., Vol. LVII, pp. 125-128. Baeboub, Thomas, and Noble, G. Kingsley. 1915. A revision of the lizards of the genus Ameiva. Bull. Mus. Comp. Zool., Vol. LIX, pp. 417-479. 1916. A revision of the lizards of the genus Cyclura. Idem, Vol. I.X, pp. 139-164, Pis. 1-15. Barboub, Thomas, and Ramsden, Charles T. 1919. The herpetology of Cuba. Mem. Mus. Comp. Zool., Vol. XLVII, pp. 69-213, Pis. 1-15. Bello y Espinosa, [Domingo] 1871. Zoologische Notizen aus Puerto Rico. . . . Nach dem Span- ischen frei bearbeited von Herrn E. von Martens in Berlin. Zool. Garten. Vol. XII, pp. 348-351. BOETTGEK, OSKAR 1892. Katalog der Batrachier-Sammlung im Museum der Seucken- bergischen Naturforschenden Gesellschaft in Frankfurt-am-Maiu. pp. X + 73. 1893. Katalog de Reptilieu-Sammlung im Museum der Senckenbergischen Naturforschenden Gesellschaft in Frankfurt-am-Main. Pt. I, pp. X + 140. 1898. Idem, Pt. 2, pp. X + 160. Boulenger, George Alfred 1882. Catalogue of the Batrachia, Salientia s. Ecaudata in the collection of the British Museum (Natural History), 2ud Ed. 1885-7. Catalogue of the lizards in the British Museum (Natural History), Vols. I-III. 1890. First report on additions to the lizard collection in the British Museum (Natural History). Proc. Zool. Soc. London, pp. 77-86, Pis. 8-11. 1893-4. Catalogue of the snakes in the British Museum (Natural History), Vols. I-IT. 1896. Ueber einige Reptilien von der Insel Mona (West Indien). Jah- resber. Naturw. Ver. Magdeburg, 1894-1896, pp. 112-114. Cochran, Doris 1924. Typhlops lum'bricaUs and related forms. Journ. Wash. Acad. Sci., Vol. XIV, pp. 174-177. SCHMIDT, AMPHIBIANS OF PORTO RICO 157 Cope, Edward Dbinkee 1S62. Synopsis of the species of Holcosus and Ameiva, with diagnosis of new West Indian and South American Colubridae. Proc. Acad. Nat. Sci.. Phila., pp. 60-82. 1862a. Contributions to neotropical saurology. Idem. pp. 176-188. 1868. Sixth contribution to the herpetology of tropical America. Idem. pp. 305-313. 1885. On the species of Iguaninae. Proc. Amer. Philos. See., Vol. XXIII, pp. 261-271. CuviER, Georges 1829. La regne animal, distribue d'apres son organisation. Nouvelle Edi- tion, Vol. II. Danforth, Stuart T. 1925. Porto Rican herpetological notes. Copeia, No. 147, pp. 76-79. 1926. Birds of the Cartagena Lagoon, Porto Rico. Journ. Dept. Agr. Porto Rico, Vol. X, pp. 1-136, 45 figs. DUMERIL, A. M. C, AND BiBRON, G. 18.34-1854. Erpetologie general ou histoire naturelle des reptiles. Vols. I-IX, and atlas. DUMERIL, AUGUSTE, BOCOURT, F., AND MOCQUARD, F. 1870-1909. Etudes sur les reptiles. Miss. Sci. Mex. Amer. Centr., Pt. 3, Section 1, text and atlas. Fowler, Henry W. 1918. Some amphibians and reptiles from Porto Rico and the Virgin Islands. Papers Dept. Marine Biol., Carnegie Inst. Wash., Vol. XII, pp. 1-15, I'l. 1, Text Figs. 1-6. Garman, S. W. 1887. West Indian Batrachia in the Museum of Comparative Zoology. Bull. Essex Inst., Vol. XIX, pp. 13-16. 1887a. On West Indian reptiles in the Museum of Comparative Zoology at Cambridge, Mass. Proc. Amer Philos. Soc, Vol. XXIV, pp. 278-286. Gray, John Edward 1838. Catalogue of the slender tongued saurians with descriptions of many new genera and species. Am. Nat. Hist., Vol. II, pp. 287-293. 1.845. Catalogue of lizards in the British Museum. GUNDLACH, J. 1881. Apuntos para la fauna puerto-riquena, III. Anfibios. An. Soc. Espan. Hist. Nat., Vol. X, pp. .305-317. 158 SCIENTIFIC SURVEY OF PORTO RICO GtJNTHER, Albert 1859. On the reptiles from St. Croix, West Indies, collected by Messrs. A. and E. Newtou. Ann. Mag. Nat. Hist. (3), Vol. IV, pp. 209-217. 1880. Observations on some rare reptiles and a batrachian now or lately living in the Society's menagerie. Trans. Zool. Soc. London, Vol. XI, pp. 215-222, Pis. 42-46. Jan, Giorgio 1860-1881. Iconographie generale des ophidiens. Meerwarth, Hermann 1901. Die westindisclien Reptilien und Batrachier des Naturhistorischeu Museums in Hamburg. Mitt. Naturh. Mus. Hamburg, Vol. XVIII, pp. 1 + 41, Pis. 1-2. I Merrem, Blasius 1820. Tentamen systematis amphibiorum. Miller, Gerritt 1918. Mammals and reptiles collected l)y Theodoor de Booy in the Virgin Islands. Proc. U. S. Nation. Mus., Vol. LIV, pp. 507-511, PI. 81. MOREAU DE JONNES 1818. Monographic du mabouia des murailles ou gecko mabouia des Antilles. Bull. Soc. Philom. I'aris, pp. 138-139. NiEDEN, Fritz 1923. Anura I. Das Tierreich, Ifg. 46, pp. XXXI I -(-584, 380 text figs. Peters, W. 1876. Ueber eine von Hrn. Viceconsul L. Krug und Dr. J. Gundlach auf der Insel Puertorico gemachte Sammlung von Saeugethiereu und Amphibien, so wie ueber die Entwickelung eines Batrachiers, Hylodes martinicensis Dum. Bibr., ohne Metamorphose. Mouats- ber. Akad. Wiss. Berlin, pp. 703-714, PI. 1. Reinhardt, I. Th. 1843. Beskrivelse af nogle nye Slangearter. Dauske Vidensk. Selsk. Afhandl., Vol. X. pp. 233-279, 3 pis. Reinhardt, J., and Luetken, C. F. 1863. Bidrag til det vestindiske Geriges og navuligeu til de dansk-vestin- diske Oers Herpetologie. Vidensk. Meddel Naturh. Foren., Copen- hagen, 1862, pp. 153-291. ROSENFELD, A. H. 1925. The food of Porto Rican lizards. J. Econ. Ent., Vol. XVIII, pp. 422-423. SCHMIDT, AMPHIBIAXS OF PORTO RICO 159 SCHLEGEL, HeINRICH 1837. Physionomie des Serpents, Vol. II. Schmidt, Karl Patterson 1920. Contributions to the herpetology of Porto Rico. Anm. N. Y. Acad. Sci., Vol. XXVIII, pp. 167-200, Text Figs. 1-9. 1926. The herpetology of Mona Island, West Indies. Publ. Field Mus. Nat. Hist., Zool., Vol. XII, pp. 149-16.3. Kgs. 1-5. 1927. A new EleutheroAactylus from Porto Rico. Araer. Mus. Novitates, No. 279, pp. 1-3. Seoane, Victor Lopez « 1881. Neue Boiden-Gattung und-Art vou den Philippinen. Abh. Senck. Naturf. Ges., Vol. XII, pp. 217-224, PI. 1. 1890. Nouvelle espece de batracien anoure des iles Philippines. Mem. Soc. Zool, France, Vol. III., pp. 206-210, PI. 6. Sein, Francisco, Jr. 1927. El sapo. Revista de Agricultura, Vol. XIX, pp. 2.38-2.39, 1 fig. Smyth, E. Greywood 1920. Nuestro amigo el AnoUs. Rev. Agric. Puerto Rico, Vol. IV, pp. 11-21. Stahl, a. 1882. Fauna de Puerto-Rico. Clasificacion sistematica de los animales que corresponden a esta fauna y catalogo del gabinete zooligico de Dr. A. Stahl en Bayamon. Stejneger, Leonhard 1901. On the herpetology of Porto Rico. Tageblatt V, Intern. Zool. Congr., No. 8, Aug. 26, 1901, p. 28. 1904. The herpetology of Porto Rico. Rept. U. S. Nation. Mus., 1902, pp. 549-724, PI. 1, Text Figs. 1-196. 1905. Blunders in the scientific records. Science (2), Vol. XXI. p. 472. 1911. A new lizard from Porto Rico. Proc. Biol. Soc. Wash., Vol. XXVI, pp. 69-72. Strauch, a. 1883. Bemerkungen iiber die Eidechsenfamilie der Amphisbaeniden. Mel. Biol. Acad. Sci. St. Petersb., Vol. XI, p. 405-407. 1887. Bemerkungen iiber die Geckouiden-Sammlung im Zoologischen Mu- seum der Kaiserlichen Akademie der Wissenschaften in St. Peters- burg. Mem. Acad. Sci. St. Petersb. (7), Vol. XXXV (No. 2), pp. 1-72, 1 pi. WoLcoTT, George N. 1924. The food of Porto Rican lizards. Journ. Dept. Agric. Porto Rico, Vol. VII, No. 4, pp. 5-37. 160 SCIENTIFIC SURVEY OF PORTO RICO WiEGMANN, A. F. A. 1837. Herpetologische Notizen. Arch. Naturg., Vol. Ill, pp. 123-136. Zenneck, I. 1898. Die Zeichuung der Boiden. Zeitsclir. Wiss. Zool., Vol. LXIV, pp. 1-384, Pis. 1-8. THE FISHES OF PORTO RICO AND THE VIRGIN ISLANDS Branchiostomidae to Sciaenidae By J. T. Nichols CONTEXTS rage Introduction 169 Field work in Porto Rico 160 Acknowledgments l^O Previous knowledge of Porto Rican fishes 171 Plan of work 171 Explanation of technical terms 172 Faunal discussion 175 The faunae represented in Porto Rican waters 175 Limits and relationships of the West Indian shore fauna 175 The tropical pelagic fauna 176 Tabular orientation of faunae repi'esented 177 The Gulf Sti-eam current system 177 Analysis of ranges of members of the West Indian shore fauna 178 Ontogenic adaptation to different faunae 178 Origins of the West Indian shore fauna 178 The slight Bermudian sub-fauna ISO Systematic account of the species 180 Branchiostomidae ISO Branchiosto-ma Costa 180 Branchiostoma caribaeum Sundevall 180 Asymmetron Andrews 181 Asymmetron lucayanum Andrews 181 Ginglymostomidae ISl Ginijlymostouia Miiller and Ilenle 181 Ginglymostoma cirratum (Gmeliu) 181 Galeidae 1S2 Galeocerdo Miiller and Henle 182 Galeocenlo tif/riniis Miiller and llenle 182 Carcharhinus Blainville 183 Carcharhimis falciformis (Bibron) 183 Cdfcharhinus limbatiis (Miiller and Henle) 1S3 Sphyrnidae ^^ Sphyrtia Rafinesque 184 Sphyrna zygactta (Linnaeus) 184 Pristidae 1S5 Pristis Latham 185 Pristis pectinatits Latham 185 Dasyatidae 186 Dasyatis Rafinesque 186 1(53 SCIENTIFIC SURVEY OF PORTO RICO Page Dasi/atis americana Hildebrand and Schroeder 186 Dasiiafis say (Le Sueur) 187 Myliobatidae 188 Actohatus Blaiuville 188 Aetobatus narinari (Euphrasen) 188 Anguillidae 189 AnyuiUa Shaw 189 Anguilla rostrata (Le Sueur) 189 Leptocephalidae 189 Leptoccphalus Scopoli 189 Lcptocephalus conger (Linnaeus) 189 Mayerina Silvester 190 Mayerina mayeri Silvester 190 INIuraenesoeidae 191 MnracDCSox McClelland 191 Muracncsoje savanna (Cuvier) 191 Moringuidae 191 ApJithalmichthys Kaup 191 AphthalniicMhys caribbeus Gill and Smith 191 Myridae 192 Myrophis Liitkeu 192 3Iyrophis longleii Silvester 192 Chilorhmus Liitken 192 Chilorhinus siiensonii Liitken 192 Ophichthyidae 19.3 Sph agebranchus Bloch 193 Sphagebranchus ophioneus Evermann and Marsh 193 Myrichthys Girard 193 MyricMhys oculatus (Kaup) 193 Myiichthys acunmiatiis (Grouow) 194 Myrichthys keclcii Silvester , 194 Ophichthus Thunberg and Ahl 195 Ophichtus gomesii (Castelnan) 195 Muraenidae 195 Gymnothorax Bloch 195 Gymiwthorax moringa (Cuvier) 195 Gymnothorax funcbris Ranzani 196 Gymnothorax albimcntis (Evermann and Marsh) 197 Gymnothorax jo^rdani (Evermann and Marsh) 197 Echidna Forster 198 Ech idna catenata ( Bloch ) 198 Elopidae 198 Tarpon Jordan and Evermann 198 Tarpon atlanticiis (Cuvier and Valenciennes) 198 Flops Linnaeus 199 Flops sa urus Linnaeus 199 NICHOLS, PORTO RICO AND THE VIRGIN ISLANDS 163 Page Albulidae 200 Albula Bloch and Schneider 200 Alhula inilpes (Linnaeus) 200 Clupeidae 201 Jenkinsia Jordan and Everniann 201 Jenkinsia lamprotacnia (Gosse) 201 Sanlinella Ciivier and Valenciennes 201 Sardinella anchovia Cuvier and Valenciennes 201 Harengula Cuvier and Valenciennes 202 Harengula sardina Poey 202 Harengula macrophthalma (Ranzani) 202 Opisthonema Gill 203 Opisthonema oglinum (Le Sueur) 203 Ilisha Gray 204 Illsha hleekeriayia (Poey) 204 Engraulididae 204 Anchovia Jordan and Evermann 204 Anchoma perfasciata ( Poey ) 204 Anchovia cubana ( Poey ) 204 Anchovia hrownii (Gmelin) 205 Anchovia choerostoma (Goode) 205 Anchovia lyolepis (Evermann and Marsh) 206 Anchovia product a (Poey) 206 Cetengraulis Giinther 206 Cetcngranlis cdentiilus (Cuvier) 206 Synodontidae 207 Trachinocephalus Gill 207 Trachinocephalus mi/ops (Forster) 207 Synodiis Bloch and Schneider 207 Synodus intermedins (Agassiz) 207 Synodus foetens (Linnaeus) 208 Aulopidae 208 Chlorophthalmus Bonaparte 208 ChlorophthaliHus rhalybcius (Goode) 208 Cyprinidae 209 Carassius Nilsson 21)9 Carassiiis auratus ( Linnaeus) 209 Poeciliidae 209 Fundulus Lacepede 2(19 Fundulus fonticola Cuvier and Valenciennes 209 Poecilia Bloch and Schneider 210 Poecilia vivipara Bloch and Schneider 210 Esocidae 211 Tylosurus Cocco 211 Tylosurus notatiis (Poey) 211 Tylosurus timucu (Walbaum) 211 Tylosurus ardeola (Cuvier and Valenciennes) 211 1(34 SCIEXTIFIC SURVEY OF PORTO RICO Page Tjilosurus euryops Bean and Dresel 212 TiiloNiirux raphidoma (Ranzani ) 212 Tiflomiriis (km,s (Laeei^ede) 213 Hemirauiijliidae 213 Hyporhamphus Gill 213 Hyporhumphus %imf(iHc'uitu)< (Ranzani) 213 Hcmlmmphus Cuvier 214 Hcmiramphus hrasilicnsia (Linnaeus) 214 Exoeoetidae 215 Parexocoetus Bleeker 215 Porexococtus hrachyptcnis (Solander) 215 Cyp-sclurns Swainsou 215 Cyi>f^6 Holocentridae Myripristis Cuvier — " Myripristis jacohu.s Cuvier and Valenciennes 226 Holoccntrus Seopoli — " Holocentriis ascensionis (Osbeck) --^ Holocentrus vexillarius Poey -- ' 007 Mullidae "" Upeneus Cuvier — '_ Upeneus maculatus (Blocli) — ' Jjpeneus parrus I'oey — '^ Upeneus martinicns Cuvier and Valenciennes 228 Scombndae " • /-. • 229 Auais Cuvier Auxis tMzard (Lacepede) "--^ Scomheromorus Lacepede — '' Scotuhcromonis maculatus ( Mitcliill ) —9 ScomheromoniH reynHs Blocli —^^ Scomheromorus ca valla (Cuvier) 230 001 Trichiuridae "' 001 Trichiurus Linnaeus -" Trk-hiurus Jepturus Linnaeus 2?.l 000 Carangidae ~^~ 000 Oligoplites Gill — '- Oliooplitcs snunts (Blocli and Schneider) 232 Serioki Cuvier -'^- Seriola fakata Cuvier and Valenciennes 232 Decapterus Bleeker -'"' Decapterus puiictatiis ( Agassiz) -J-J Trachurops Gill -"*" Trachurops crumenophthalmus (Bloch) 233 Caranx Lacepede — ^"* Caranx ruber ( Blocli ) 234 Caranx hartholomaei Cuvier and Valenciennes 234 Caranx Mpi)os ( Linnaeus ) — ^'^ Caranx crysos ( Mitcliill ) — ^" Caranx kttiis Agassiz -'^^ Tomer Cuvier and Valenciennes --^^ Vomer setaphutis setapinnis (Mitchill) '-'' Vomer setapinnis cuhensis Nichols 238 Sekne Lacepede - Sekne router ( Linnaeus ) — ^• Chlorosconihriis Girard -^•' Chkjroscomhrus chrysurus (Linnaeus) 239 Trachmotus Lacepede -^^ Ige SCIENTIFIC SURVEY OF PORTO RICO Page Trachinotus glaucus (Bloch) 240 Trochhwtus falcaUis (Linnaeus) 241 Trachinotus caroUnus { Linnaeus) 241 >s'omeidae 242 Nomeus Cuvier 242 Nomeus gronovii ( Gmelin ) 242 Stromateidae 243 Pcprilus Cuvier 243 Feprilns paru ( Linnaeus ) 243 Clieilodiptei-idae 243 Apogon Lacepede 243 Apoyon seUicauda Evermann and Marsh 243 Apogon conklini ( Silvester) 244 Apogonlchthys Bleeker 245 Apogonichthys alutus (Jordan and Gilbert) 245 Apogonichthys stellatus Cope 245 Centropomidae -^^ Centropomus Lacepede 24b Centropomus undecimalis ( Bloch) 246 Centropomus parallelus Poey • 246 Centropomus pectinatus Poey 247 Serranidae -^' Petrometopon Gill ^^ ' Petronictopon eruentatus cruentatiis ( Lacepede) 247 Petrometopon cruentatus coronatus (Cuvier and Valenciennes).. 248 Cephalopltolis Bloch and Schneider 248 CephalophoUs fulvus ruber (Bloch and Schneider) 248 CephalopJiolis fulvus punctatus ( Linnaeus) 249 Epincphelus Bloch 249 Epinephelus adscensionls (Osbeck ) 249 Epineplmlus striatus (Bloch) 250 Epincphalus guttatus { Linnaeus) 2.51 Epincphelus morio (Cuvier and Valenciennes) 251 Alphcstes Bloch and Schneider 252 Alphcstes chloropterus ( Cuvier and Valenciennes) 2.52 Mucteroperca Gill -j- MyctC7'operca bonaci (Poey) -•^- Mycteroperca bowersi Evermann and Marsh 2.53 Hypoplectrus Gill — ^^ Hypoplcctrus unicolor chloruriis (Cuvier and Valenciennes) 254 Hypoplectrus unicolor guttavarius (Poey) 2.o4 Diplectrum Holbrook -5o Diplectrum radiale ( Quoy and Gaimard) 2oo Prionodes Jenyns -^^ PriOHodcs baldwini Evermann and Marsh 255 Bules Cuvier 2o6 Dules dispilurus (Giinther) — j" Paranthias Guichenot -^' NICHOLS, PORTO RICO AND THE VIRGIN ISLANDS 167 Page Paranthias furcifer (Cuvier and Valenciennes) 257 Rypticus Cuvier 257 Rj/pticiis saponaceus (Bloch and Schneider) 257 Rypticus coriaceus ( Cope) 258 Rypticus Mstrispinus (Mitchill) 258 Lobotidae 259 Lohotes Cuvier 259 Loliotes surinamensis (Blocli) 259 Priacantliidae 259 Priacantlius Cuvier 259 Priacantlius arenntus Cuvier and Valenciennes 259 Priacantlius criientatus (Lacepede) 260 Lutiauidae 260 Lutianus Bloch 260 Lutianus cyanopterus (Cuvier and Valenciennes) 260 L utianus griscus { Linnaeus ) 261 Lutianus jocu (Bloch and Schneider) 262 Lutianus apodus ( Walbaum) 263 Lutianus aya (Bloch) 263 Lutianus vivanus (Cuvier and Valenciennes) 264 Lutianus analis (Cuvier and Valenciennes) 265 Lutianus megalophthalnius (Evermann and Marsh) 265 Lutianus synagris (Linnaeus) 266 Lutianus mahogoni (Cuvier and Valenciennes) 267 Ocyurus Gill 267 Ocyurus chrysurus ( Bloch ) 267 RhmihopUtes Gill 268 RhombopUtes aurorubens (Cuvier and Valenciennes) 268 Apsilus Cuvier and Valenciennes 268 Apsilus ilcntatus Guichenot 268 FAelis Cuvier and A' alenciennes 269 Etelis oculatus (Cuvier and Valenciennes ) 269 Haemulidae -"^ Haemuion Cuvier 269 Haeniulon album Cuvier and Valenciennes 269 Hacmulon niacrostomus Giinther 270 Haemulon bonaricnse Cuvier and Valenciennes 271 Haemulon parra (Desmarest) 271 Haemulon carbonarium Poey 272 Haemulon melanurum (Linnaeus) 273 Haemulon sciurus ( Shaw ) - ' "^ Haemulon plumieri (Lacepede) 2(4 Haemulon flaiolineatum (Desmarest) 274 Bathy stoma Scudder - ' 'j Bathystoma rimator (Jordan and Swain) 275 Bathy stoma auroUneatum (Cuvier and Valenciennes) 275 Bathystoma striatum (Linnaeus) 2<6 Anisotremus Gill - ' " 1(58 SIOIEXTIFIC f 284 Xystaema Jordan and Evermann 284 Xystaema cinereum ( Walbaum) 284 Xystaema havana Nichols 285 Diapterus Ranzani 286 Diapterus rhombeus (Cuvier and Valenciennes ) 286 Diapterus oUsthostomus (Goode and Bean ) 286 Diapterus plumieri (Cuvier and Valenciennes) 287 Diapterus plumieri (Cuvier and Valenciennes) 287 Kyphosidae 288 Kyphosus Laeepede '. 288 Kyphosus incisor (Cuvier and Valenciennes) 288 Kyphosus sectatrix (Linnaeus) 288 Sciaenidae 289 Cynoscion Gill 289 Cynoscion jamaicensis ( Vaillant and Bocourt ) 289 Larimus Cuvier and Valenciennes 290 Larimus breviceps Cuvier and Valenciennes 290 Ofloufosrion Gill 290 Oiloiitoscioii dented- ( Cuvier and Valenciennes) 290 Corvtila Jordan and Eigenmann 291 Corvula sanctae-luciae Jordan 291 Corvula batahana ( Poey) 291 Bairdiella Gill 291 MCHOLS, PORTO RICO AND THE VIRGIN If< 207 Distribution. — West Indian fauna, rather common about the Greater Antilles, but apparently not common in Porto Rican waters. Diagnosis.— Head 3.3 to 3.3; depth 3.3 to 3.4; eye 3.5 to 4. Dorsal 14 to 15; anal 23; scales 42. Attains a length of 6 inches. Synodontidae Trachinoeephalus Gill TrachinoceplialiLS inyops (Forster) Ground spearing; lagarto Salmo my ops Forster, 1801, MS., Bloch and Schneider, Syst. Ichtb., p. 421. Trachinoeephalus myops Evermann and Marsh, 1902, p. 91. Fig. ^'1.— Trachinoeephalus myops ^^ ^^^-_il_ &> _ ^ _ ^., •--;-- • From Zoologica, IX ^^^s^. ) ^^^^ ^'^^ ' ^ ' Type locality. — St. Helena. Distribution. — West Indian fauna from South Carolina to Brazil, and southward, occasionally north to Massachusetts. Eare in Porto Eican waters. Dia^gnosis.— Head 3.4; depth 5; eye 6.4. Dorsal 12; anal 14; scales 58. Attains a length of about 9 inches. Synodus Bloch and Schneider Synodus interniedius (Agassiz) Sand diver ; lizard-fish Sauriis intermedins Agassiz, 1829, in Spix, Piscium Brazil., p. 81, PI. XLIV. Synodus intermedius Evermanu and Marsh, 1902, p. 92. Fig. 43. — Synodus intermedius ^^ ^JMy ^^ From Zoologica, X Type locality. — Brazil. Distribution. — Southern Florida to Brazil, generally rather common. Fairly common in Porto Eican waters. Diagnosis. — Head 3.8 ; depth, 7 ; eye 7. Dorsal 11 ; anal 11 ; scales 49. This species attains a length of from 10 to 12 inches. Remarhs. — Lizard-fishes have little or no food value. 208 SCIENTIFIC SURVEY OF PORTO RICO Synodus foetens (Linnaeus) Galliwasp ; lizard-fish ; lagarto Salmo foetens Linnaeus, 1766, Syst. Nat, ed. 12, p. 513. Synodus foetens Evermaun and Marsh, 1902, p. 92, Fig. 16. Fig. 44. — Synodus foetens From Zoologica, IX Type locality. — South Carolina. Distribution. — West Indian fauna, Cape Cod to Brazil, very common from South Carolina southward. Not uncommon in Porto Eican waters. Diagnosis. — Head 4; depth 8; eye 8. Dorsal 10; anal 12; scales 60. This species grows to be a foot in total length. HaMts. — The lizard-fish frequents sandy shores. It rests on the bot- tom, prepared to attack with lightning speed and devour any luckless smaller species that swims within range. Its mottled gray or brown color gives it a low visibility on the bottom, of aggressive rather than protec- tive usefulness. Young an inch or two long are translucent, with pairs of oval pigment spots along the midline, and may have habits somewhat like those of the related lantern fishes. AULOPIDAE Chlorophthabiius Bonaparte Chlorophthalmus chalybeius (Goode) Green-eye Hyphalonednis chalybeius Goode, 1881, Proc. U. S. Nat. Mus. for 1880, p. 484. Chlorophthalmus chalybeius Evermann and Marsh, 1902, p. 93. Type locality. — Gulf Stream, off Block Island. Distribution. — Off the Atlantic coast of America from Ehode Island to Porto Eico, at depths of from 85 to 220 fathoms. A single specimen (21/2 inches long), Mayagiiez Harbor, P. E., at a depth of 220 fathoms. Diagnosis. — Head 3; depth 6; eye 2.5. Dorsal 9; anal 6; scales 51. Remarl-s. — The Porto Eican specimen of this interesting fish is so identified by Evermann and Marsh, though showing slight differences from the type taken off Block Island. NICHOLS, PORTO RICO AND THE VIRGIN ISLANDS 209 Cypkinidae Carassius Nilsson Carassius auratus (Linnaeus) Goldfish Cyprinus auratus Liumvus, 1V58, Syst. Nat., ed. 10, Vol. I, p. 322. Carassius auratus Nichols, 1915, Bull. Amer. Mus. Nat. His., Vol. XXXIV, p. 141. Fig. 45. — Carassius auratus Type locality. — China and Japan. Distribution. — A native of eastern Asia, introduced cosmopolitan m temperate fresh waters. In 1914 the goldfish was said to he abundant at Isabella, P. E., in the northwest corner of the island. It had recently been placed in a small pond in the hills above Guayama, where it seemed to be doing well. Specimens collected. — 1 : Governor's House near Guayama. Diagnosis.— HeAd 3.6; depth 2.4; eye 3.4 (specimen of about 4 inches standard length). Dorsal II, 16 to 20; anal II, 5; scales 26 to 28. Dorsal in the middle of the back; last simple dorsal and anal rays spinous, serrate behind. No teeth in the mouth. The usual striking golden color is replaced by inconspicuous olivaceous shades in the very young and some grown feral individuals. POECILIIDAE Fundulus Lacepede Fuiiduliis fontit'ola Cuvier and Valenciennes Porto Rican killifish Fundlus fonticola Cuvier and Valenciennes, 1846, Hist. Nat. Poiss.. Vol. XVIII, p. 198. Fundulus fonticola Evermanu and Marsh, 1902, p. 96. 210 SCIENTIFIC SURVEY OF PORTO RICO Fig. 46. — Fiindulus fonticola Type locality. — Porto Rico. Distnhution. — Known only from mountain springs in Porto Pico; rare. Diagnosis. — Head broad, little depressed. Tail more slender and body deeper than in Fundulus heteroclitus. Body plump, with a long caudal peduncle. Dorsal 11; anal 12; scales 37. Poecilia Bloch and Schneider Poecilia vivipara Bloch and Schneider Viviparous tooth-carp; top-minnow Poecilia vivipara Bloch and Schneider, 1801. Syst. Ichth., p. 452, PI. 86. Fig. 2. Poecilia vivipara Everninnn and Marsh. 1902. p. 97. Fig. 47. — Poecilia vivipara Type locality. — Surinam. Distribution.— ^vesh waters in Brazil, Guiana, Martinique and Porto Pico. In Porto Pico found near Ponce and Fajardo, and at Arroyo and Hucares, — abundant. Specimens collected.— 20: Mouth of the Loiza Piver. 18: Guanica. Diagnosis.— Ue&d 3.6 ; depth 3.5 ; eye 3.5. Dorsal 7 ; anal 8 ; scales 25. Length 3 or 4 inches. Habits. This species belongs to that division of the tooth-carps or killifishes in which the eggs hatch within the body cavity of the mother fish, and young are brought forth "alive." A^iviparous tooth-carps are very plentiful in the middle Americas in respect to individuals, species and o-enera. Their small size and the numerical abundance which they attain in restricted sluggish fresh waters renders them an effective factor in mosquito control. This is the only species of top-minnow that occurs in Porto Pico unless others have been recently introduced. NICHOLS, PORTO RICO AND THE VIRGIN ISLANDS o^i ESOCIDAE Tylosurus Cocco Tjiosurus notatus (Poey) Needlefish ; long- jaws ; agujon BeJone notata Poey, 1860, Memorias, Vol. II, p. 293. Tylosurus notatus Silvester, 1916. Yearb-Carn. Inst. Wash, for 1915, Vol. XIV, p. 215. Porto Rico. Fig. 48. — Tylosurus notatus Type locality. — Havana. Distribution. — West Indies, north to Pensacola, Fla. Diagnosis.— Re^di 2; depth 5 (in head). Dorsal 13; anal 13 (count- ing developed rays only) ; scales 150, 85 before dorsal. Body robust, not compressed. Peduncle compressed, deeper than broad, without trace of a keel. Length about 20 inches. Tylosurus timucu (Walhaiim) Needlefish ; agujon Esox timucu, Walbaum, 1792, Artadi Piseium, Vol. Ill, p. 295; after Marc- grave. Tylosurus timucu Evermann and Marsh, 1902, p. 99. Type locality. — Brazil. Distribution. — Florida Keys to Brazil. Not uncommon in Porto Pican waters. Diagnosis. — Head 2.7 to 2.8; depth T (in head) ; eye 2.3 to 2.8 in post- orbital part of head. Dorsal 15; anal 17; scales 225, about 150 before the dorsal. Caudal peduncle compressed, deeper than broad, without trace of a keel. Length about 18 inches. Tylosurus arcleola (Cuvier and Valenciennes) Needlefish ; agiajon Belone ardeola Cuvier and Valenciennes, 1846, Hist. Nat. Poiss., Vol. XVIII, p. 425. Tylosurus ardeola Evermann and Marsh, 1902, p. 99. 212 SCIEXTIFW SURVEY OF PORTO RICO mrnmr. 3*;^^ ■^, -.- ■- ■ — _^ ^^g^ Fig. 49. — Tylosurus ardeola Type locality. — Martinique. Distribution. — West Indian fauna, not common. Three or four rec- ords in Porto Eican waters. Diagnosis. — Head 3.7 to 3.8; depth 8 (in head) ; eye 7. Dorsal 13; anal 17; scales moderate, about 150 before the dorsal. Caudal peduncle depressed, broader than deep, with a keel. Length a foot or more. Tjiosurus euryops Bean and Dresel Needlefish ; long-jaw ; agujon TylosurKs euryops Bean and Dresel, 1884, Proe. U. S. Nat. Mus. for 1884, p. 168. Tylosurus euryops Nichols, 1915, Bull. Amer. Mus. Nat. Hist., Vol. XXXIV, p. 142. Porto Rico. Type locality. — Jamaica . Distribution. — Cuba and Jamaica, not common. A few small ones re- corded from San Juan Ba}^, P. R., in summer, possibly the young of T. timucu. Specimens collected. — 1 : San Juan. Diagnosis. — Head 3 ; depth in head 6 ; eye in postorbital part of head 2.2. Dorsal 15; anal 17; scales 200. Eegion above base of pectoral without a black spot; 140 to 150 scales before the dorsal fin; peduncle compressed, deeper than broad, without trace of a keel. Tylosunis raphidoina (Ranzani) Neefllefish ; hound-fish : agujon • Beloiic raphidoina Kauzani. 1842, Nov. Comm. Ac. Nat. Sci. Inst. Bonon., Vol. V, p. 359, PI. XXXVIII, Fig. 1. Tylosurus raphidoma Evernianu and Marsh, 1902, p. 99, Fig. 17. S[t]r(tii!iylura raphidoma Beebe and Tee Van, 1928, Zoologiea, Vol. X, p. 63. ,^ .T — ^^aeaszaa^,^^ T^iG. 50. — Tijlosuriis raphidoma ' ^"^ — '^^ R'- :l:^.5s s^''~^^, From Zoologiea, X Type locality. — Brazil. Distribution. — Generally abundant in the West Indies, from the Florida Keys to Brazil. The young drift northward abundantly to the Capes of the Carolinas, and occasionally to Xew Jersey. Tylosurus NICHOLS, PORTO RICO AND THE VIRGIN ISLANDS 213 rapliidoma is the most numerous species of the genus in Porto Eican waters, where it is generally distributed. Diagnosis. — Head 3.3 ; depth 3.4 in head ; eye 7 in snout, 2.66 in postorbital part of head. Dorsal I, 20 to 24; scales about 350. Caudal peduncle slightly depressed, with a slight black keel; caudal fin un- equally lunate. Attains a length of from 3 to 5 feet. Habits. — Much preyed upon by barracudas. The young found among broken drifting eel grass, to which they bear a concealing resemblance (E. W. Gudger). Tylosurus acus (Lacepede) Houndfish ; agiijon Sphyraena acus Lacepede, 180.3, Hist. Nat. Poiss., Vol. V, p. 6, PI. I, Fig. 3. Strongylura acus Fowler, 1928, Proc. Ac. Sci. Phila., p. 463. Porto Rico. Fig. 51. — Tylosurus acus From Zoologica, IX Type locality. — Martinique. Distribution. — West Indies, occasionally northward to Buzzards Bay. A Mediterranean species may be identical. Recorded from Porto Eico by Fowler. Diagnosis. — Head 2.G ; depth 18.5 ; eye about 2.5 in postorbital part of head. Dorsal 23; anal 21 to 22; scales 380 to 400. Keel on the peduncle strong, black; caudal deeply and unequally emarginate; no definite silvery lateral band. Eeaches a length of 4I/2 feet. Habits. — Houndfishes swim near the surface, hovering on the out- skirts of schools of smaller surface fishes, which seek to escape their periodic piratical raids by scattering in all directions, often leaping into the air. This and other big needle-fishes are themselves proficient in leaping. Their long bodies shoot out of the water and through the air as might a thrown spear or javelin, and there are said to be instances of their thus by chance striking and wounding fishermen in small boats. Hemiramphidae Hyporliaiiiphus Gill Hyporhamphus unifasciatiis (Ranzani) Half-beak ; escribano Hemirhamphus unifasciatiis Ranzani, 1842, Nov. Comui. Ac. Sci. Bonon., Vol. V, p. .326. Hyporhamphus unifasciatiis Evermann and Marsb, 1902, p. 101, Fig. 18. 214 SCIENTIFIC SURVEY OF PORTO RICO ••' ^'^^'■■.^^>-. '!'^^~rrX ^^ From Zoologica, X Type locality. — Brazil. Distribution. — West Indian fauna, Key West to Eio de Janeiro, com- mon. Common in Porto Rican Waters. Diagnosis.— ]1qa(\. 4.6; depth 6.8; eye 4. Dorsal 14 or 15; anal 14 to 16 ; scales 52. Caudal fin equally lunate. Attains a length of about a foot. RemarJcs. — Extensively used for food. Habits. — More numerous in shallow water near shore than is the fol- lowing species, less given to leaping. Heniirainphus Cuvier Heniiraniphus brasiliensis (Linnaeus) Half -beak ; ballyhoo ; balao Esox hrastUensis Linnaeus, 1758, Syst. Nat., ed. 10, p. 314 ; after Browne. Hemirhamphiis hrasiUeiisis Evermann and Marsh, 1902, p. 102, Fig. 19. -r^!=r r^ TT wT^*^"^^ Fig. 53. — Hemiramphtis brasiliensis ^-^V, ^- ^ ^ u-L;;;^-^-^^^^^ From Zoologica, IX Type locality. — Jamaica. Distribution. — West Indian fauna, common off shore. Generally dis- tributed in Porto Rican waters. St. Croix. Diagnosis. — Head 4.2; depth 6.2; eye 3.8. Dorsal 13 or 14; anal 12 or 13 ; scales 53. Caudal fin unequally forked, the lower lobe the longer. Attains a length of 15 inches. Habits. — The Spanish name, "balao,'' refers to the habit of the fish to skitter over the surface of the water propelled by the strong lower lobe of the caudal fin. Some such fish was ancestral to the flying-fishes, and it is easy to understand how, given this habit, the pectoral fins, at first used only to raise and steady the half-beak's head, should have in- creased in size so as to lift the flying-fish into the air. The ballyhoo feeds chiefly on algae, its peculiar spear-like lower jaw being presumably an adaptation for collecting small floating particles of food at the surface. This character of an elongate lower jaw is shared by certain young needlefish, ancestral to the half-beaks, and by certain young flying-fishes, to which the half-beaks are ancestral. Ac- cording to our interpretation it is a character which serves a method of NICHOLS, PORTO RICO AND THE VIRGIN ISLANDS 215 feeding more favorable for small than for larger fishes, and hence has had greater permanence in the young than in the adults of this single evolutionary series. EXOCOETIDAE Parexoceetus Bleeker Parexoooetus brachypterus (Solander) Short-winged flying-flsh ; volador Exocoetus hraclnjiiicnis Solander, 1846, in Richardson Ichthyol. Chin., Proc. Brit. Assoc, 1846, p. 265. Parexocoetus mesogaster Evermann and Marsh, 1902, p. 103. Parcxocoetus hrachmttrniH Meek and Ilildebrand, 1923, Fishes of Panama, Pt. 1. Fig. 54. — Parexocoetus irachypterus From Zoologica, IX Type locality. — Otaheite. Distribution. — Common in tropical seas, East Indies, West Indies, Hawaiian Islands. Apparently the commonest flying-fish about Porto Eico. Diagnosis. — Head 4.4; depth 5; eye 3. Dorsal 12; anal 13; scales about 38. Attains a length of T inches. Cypseluriis S\Aainson Cypselurus bahiensis (Ranzani) Bahia flying-fish ; volador Exocoetus 'bahiensis Ranzani, 1842, Nov. Comm. Ac. Sci. Inst. Bonon., Vol. V, p. 362, PI. XXXVIII. Cypsiliiriis hahiensis Evermann and Marsh, 1902, p. 104. Fig. 55. — Cypselurus hahiensis From Zoologica. X Type locality. — Bahia. Distribution. — Off tropical American coasts in both Atlantic and Pacific, plentiful. Xorth to Cuba, and one Porto Eican record. 21(3 SCIENTIFIC SURVEY OF PORTO RICO Diagnosis. — Head 4; depth 5; eye 3.1. Dorsal 13; anal 9; scales about 50. Attains a length of 8 to 12 inches. AULOSTOMIDAE Aulostonius Lacepede Aulostomus maculatus Valenciennes Ti-umpet-flsh ; trompetero Aulostoma maculatum Valenciennes, abt. 1845, in Cuvier's Illst. Poiss., PI. XCII, Fig. 2. Aulostomus maculatus Everniann and Marsh, 1902. p. 105, Fig. 20. - ^ . - . .i.fA^^A^ ^ — ^^^ PiQ 56 — Aulostomus maculatus F'rom Zoologica, X Type locality. — Uncertain. Distribution.— CsiTTihheaLn Sea, north to southern Florida. Appar- ently not common about Porto Eico. Eecorded from St. Croix. Diagnosis. — Head 3 ; depth 11 ; eye 3 to 2.5 in postorbital part of head. Dorsal X-23 ; anal 25 ; body covered with fine ctenoid scales. Attains a length of 2 feet. FlSTULARIIDAE Fistularia Linnaeus Fistularia tabacaria Linnaeus Cornet-fish ; trompetero Fistularia tahacarin Linnaeus, 1758. Sys. Nat., ed. 10, p. .312. Fistularia tabacaria Evermann and Marsh, 1902, p. 106. Fig. 57. — Fistularia tabacaria From Zoologica, IX Type locality. — "America." Distribution. — West Indies and neighboring seas, generally common, occasionally northward to Florida and Carolina, casually to Massachu- setts. Not common in Porto Eican waters. Diagnosis. — Head 2.8. Dorsal 14; anal 13; scale-less, but with bony plates, mostly covered by skin. Attains a maximum length of 6 feet, usuallv much smaller. NICHOLS, PORTO RICO AND THE VIRGIN ISLANDS 317 Syngnathidae Syngnathus Linnaeus Syngnathus maokayi (Swain and Meek) Mackay's pipe-fisti Siphostoma mackayi Swain and Meek, 1884, Proc. U. S. Nat. Mus. for 1884, p. 239. Siphostoma mackayi Evermann and Marsli, 1902, p. 107. Fig. 58. — Syngnathus mackayi From Zoologica, X Type locality. — Key West. Distribution. — Gulf of Mexico and West Indies, known from off Pen- sacola, Key West, Haiti, Cozumel, Yukatan, etc. One of 7 inches re- corded from Mayagiiez, P, E. Diagnosis. — Head 5.66 to 6.25 in total length. Dorsal 39 to 32; rings 18 + 33 to 36. Syngnathus floridae (.Jordan and (lilbert) Florida pipe-fisti Siphostoma floridae Jordan and Gilbert, 1882, Proe. U. S. Nat. Mus. for 1882, p. 263. Siphostoma floridae Evermann and Marsb, 1902, p. 107. Fig. 59. — Syngnathus floridae Type locality. — Pensacola, Fla. Distribution. — North Carolina to Texas. Three specimens seined in Ensenada del Boqueron, P. E. Diagnosis. — Head 6 to 6.5 in total length. Dorsal 27; rings 17 to 18 + 31 to 32. Habits. — This is a rather common species on sandy continental shores. Syngnathus elucens Poey Pipe-fish Syngnathus elucens Poey, 1867, Synopsis, p. 443. Havana. Siphostoma elucens Evermann and Marsh, 1902, p. 108. 318 SCIENTIFIC SURVEY OF PORTO RICO Fig. 60. — Syngnaihus elucens From Zoologica, X Type locality. — Havana. Distribution. — Known from Havana, Haiti, and Porto Eico, whence one 6 inches long is recorded. Diagnosis. — Head 7 in total length; depth 3.6 in head. Dorsal 23, on 1 + 4 rings; rings 17 + 32. Sjngiiathus jonesi Giinther Jone's pipe-flsh SyngnatJius jonesi Giinther, 1874, Ann. Mag. Nat. Hist., (4), Vol. XIV, p. 8. Siphostoma jonesi Evermann and Marsli, 1902, p. 108. Type locality. — Bermuda. Distribution. — Bermuda and Porto Eico. Diagonsis. — Head and snout short, the latter somewhat bent upward, shorter than postorbital part of head. Dorsal 18, on 1 + 5 rings; rings 17 + 32. Habits.— The pipefishes are small attenuate fishes, encased in rings of bony armature. They hide among weed and often drift considerable distances in ocean currents. Xumerous species are plentiful on con- tinental shores adjacent to the West Indies, and others are known only from the islands. The few species and individuals recorded from Porto Eican waters may have drifted there from distant centers of greater abundance. The male pipe-fish carries the eggs and small young in a pouch placed under his tail. Two small males of this species, their pouches with eggs, were taken in 11: fathoms of water between Vieques and Culebra. Hippichthys Bleelier Hippichthys cayoriun (Evermann and Kendall) Short-nosed pipe-fish Corythroichthys cayorum Evermann and Kendall, 1898, Bull. U. S. Fish. Comm. for 1897, p. 128, PI. VII, Fig. 7. Corytliroiclitliys cayorum Evermann and Marsh, 1902, p. 108, Fig. 21. Fig. 61. — Hippichthys caiioritm Type locality. — Key West. Distribution. — West Indian fauna, from Key West to Porto Eico, rare. Diagnosis. — Head 8.6; depth 12.6; eye in head, 4.33. Dorsal 21, on iy2 + 31/2 rings; anal 3; rings 17 + 26 ==43. Length from 4 to 5 inches. NICHOLS, PORTO RICO AND THE VIRGIN ISLANDS ;> 1 !) Hippichthys ensenadae (Silvester) Short-nosed pipe-fish Corythroiclithys ensenadae Silvester, 1916, Yearbook Carn. Inst. Wash, for 1915, Vol. XIV, p. 215. Fig. 62. — Hippichthys ensenadae ,-f"^' ' mh i n i iniiniM hniuinHdiii Type locality. — Ballenas Point, P. E. Distribution. — Known only from Porto Eico. Diagnosis. — Head 9 ; eye 5. Dorsal 19, on 1 -|- 4 rings ; anal 2 ; rings 18 + 33. Body with 23 yellow and 22 brown color rings, which break up on ventral surface of belly. Length about 4 inches. Remarlcs. — Known only from the type. Habits. — The fish was secured from a bunch of coral. Doryrhamphus Kaup Doryrhamphus sierra Nichols Rough pipe-fish Doryrhamphus sie>'ra Nichols, 1915, Bull. Amer. Mus. Nat. Hist., Vol. XXXIV, p. 142, Fig. 1. . g3;gqrr -|-)-H-l-W FTTFyHl Fig. 63. — Doryrhamphus sierra Type locality. — Mouth of Loiza Eiver, Porto Eico. Distribution. — Known only from the type locality, where several speci- mens were taken. Specimens collected. — 12 : Mouth of the Loiza Eiver. Diagnosis. — Head in length to base of caudal 5.9 ; depth in head 5.6 ; eye in head, 8.8. Dorsal about 43, on 2^/2 + 5I/2 rings; rings 20 -|- 25. Snout 1.7' in head; caudal large, a little longer than head without snout (specimen 79 mm. standard length). Eidges on body high, sharp, ser- rate. Eemarhs. — In pipe-fishes of this genus the brood-pouch of the male is located on the abdomen, not the tail. They are apt to frequent the mouths of rivers but are rare in West Indian waters. There is so much difference between young and adults that the nunil:)er of good species is uncertain. 320 SCIENTIFIC SURVEY OF PORTO RICO Hippocampus Rafinesque Hippocampus punctulatus Guichenot Sea-horse; cabalito del mar Hippocampus punctulatus Guichenot, 1860, in Sagra's Cuba, Poiss., p. 174, PL V, Fig. 2. Hippocampus punctulatus Evermann and Marsh, 1902. p. 109. Fig. 64. — Hippocaitiptis punctulatus From Zoologica. X Type locality. — Cuba. Distribiiiion . — Tropical parts of the Atlantic, common in southern Florida, the West Indies, Brazil and Western Africa. Uncommon in Porto Eican waters. Specimens collected. — 1 : San Juan. Diagnosis. — Snout about 2.6 in head; eye 2 in snout. Dorsal 18, cov- ering 2^ -|- 1 rings; rings 12 -|- 30. Length of this species from 3 to 5 inches. Habits. — The grotesque- little sea horse, with head and shoulders like a knight in the game of chess, is related to the pipe-fishes, and shares with them their peculiar reproductive habits. It swims in an upright position and has a finless prehensile tail. Atherinidae Atherina Linnaeus Atherina stipes Miiller and Troschel Broad-headed silverside ; cabezote Atherina stipes Miiller and Troschel, 1848, in Schomburgk, Hist. Barb., p. 671. Atherina stipes Evermann and Marsh, 1902, Fig. 22. Hepsetia stipes Beebe and Tee Van, 1928, Zoologica, Vol. X, p. 88. Atherina laticeps Evermann and Marsh, 1902, p. 111. Fig. 65. — Atherina stipes From Zoologica, X NICHOLS, PORTO RICO AND THE VIRGIN ISLANDS 221 Type locality. — Barbados. Distribution. — West Indian fauna, north to Western Florida, abund- ant in Porto Rican waters as elsewhere. Specimens collected. — 2 : San Juan. Diagnosis.— Head 3.5 to 3.8; depth 4.6 to 5.2; eye 2.2 to 2.4. Dorsal IV or V-I, 8 to 10 ; anal I, 10 to 12 ; scales 36 to 38. Length about 3 inches. Remarl-s. — A most important food of larger fishes. Atherina araea Jordan and Gilbert Silverside Atherina araea Jordan and Gilbert, 1884, Proc. U. S. Nat. Mus. for 1884, p. 27. Atherina araea Evermann and Marsh, 1902, p. Ill, Fig. 23. Atherina stipes Evermann and Marsb, p. 110, text, not the figure. Fig. 66. — Atherina araea ^S) ^ ^^^"\\Vt ^^^ From Zoologica, X ^~-=ic ^r) >sss? ^^^^ Type locality. — Key West, Fla. Distribution.— West Indian fauna, north to Florida. Uncommon in Porto Rican waters. Diagnosis.— Head 4. to 4.6; depth 5.3 to 6.6; eye 2.4 to 2.8. Dorsal V or VI-I, 9 to 10; anal I, 10 to 13; scales 41 to 45. Length 2 to 3 inches. Remarhs. — In most localities less plentiful than the preceding, but reported to be the most abundant species in Haiti. Probably only ra- cially distinct from .1. harringtonensis of Bermuda. MUGILIDAE Mugil Linnaeus Mugil brasiliensis Agassiz Southern mullet; liza; le brancho Mugil brasiliensis Agassiz, 1829, in Spix, Pise. Brasil., p. 2.34, PI. LXXII. Muffil hrasiUensis Evermann and Marsh, 1902, p. 112. Fig. 67. — Muyil brasiliensis 232 SCIENTIFIC SURVEY OF PORTO RICO Type locality. — Atlantic Ocean off Brazil. Distribution. — Cuba to Patagonia, common throughout the West In- dies and along the coast of Brazil. Abundant in Porto Rican waters. St. Croix. Diagnosis. — Head 4; depth 5; eye 5.75. Dorsal IV-I, 8; anal III, 8; scales 35. Vertical fins almost scaleless. Length from 16 to 18 inches. Eemarls. — An important and esteemed food fish abundant in the markets of Porto Pico. Mugil curema Ciivier and Valenciennes White mullet ; liza ; Josea Muoil curema Cuvier and Valenciennes, 1836, Hist. Nat. Poiss., Vol. XI, p. 87. Brazil, Martinique and Cuba. Mugil curema Evermann and Marsh, 1902, p. 113. £\ ~-.-,.__^^ Fig. 68. — Mugil curema '■-''' i ■ ,"-" ': -' -■ ■ ' 3— -^^^^iv From Zoologica, IX ■ « Type localities. — Brazil, Martinique and Cuba. Distribution. — West Indian fauna, north to Cape Cod; and tropical American Pacific waters, south to Chile; particularly common in the tropics. Plentiful aliout Porto Rico, where it enters the Rio Grande River to Caguas. Specimens collected. — 3 : San Juan. Diagnosis. — Head 4 to 4.4; depth 3.9 to 4.5; eye 4, Dorsal IV-I, 8 ; anal III, 9 ; scales 38. Soft dorsal and anal fins well scaled. Length a foot or less. Remarks. — An important food fish. Mugil trichodon Poey Fan-tail mullet ; liza Mugil trichodon Poey, 1875, Ann. Lye. Nat. Hist. N. Y., Vol. XI, p. 66, PI. VIII, Figs. 4 to 8. Mugil trichodon Evermann and Marsh, 1902, p. 113. '(Si ^^ '^^^f Fif5- QQ-— Mugil tricJiodon NICHOLS, PORTO RICO AND THE VIRGIN ISLANDS 223 Type locality. — Cuba. Distrihution.— Florida. Keys to Brazil. Uncommon in Porto Eican waters. Diagnosis.— Head 4.2; depth 3.6; eye 4. Dorsal lY-I, 8; anal III, 8; scales 33. Soft dorsal and anal fins well scaled. Attains a length of from 6 to 10 inches. Habits.— This is a rather small species and a very active jumper. It seems to be more numerous on continental shores than in the West Indies. Agonostomus Bennett Agonostomus montioola (Bancroft) Fresh-water mullet; dajao Mugil monticola Bancroft, 1836, In Griffith's edition of Cuvier's Animal King- dom, Fishes, p. 367, PI. XXXVI. Agonostomus monticola Evermann and Marsh, 1902, p. 114, Fig. 25. Fig. 70. — Agonostomus vionticola From Zoologica, X Type locality. — Uncertain. Distribution.— Fresh waters of the West Indies and eastern :\rexico. Abundant in the fresh-water streams of Porto Rico. Diagnosis. — Head 3.5; depth 3.8; eye 6. Dorsal IV-I, 8; anal III, 9; scales 42. Attains a length of 11 inches; usually smaller. Remarks. — ^Is used for food. Habits. — A proficient jumper. Sphyraenidae Sphyraena Bloch and Schneider Sphyraena barracuda (Walbaum) Great barracuda ; picuda Esox barracuda Walbaum, 1792, Artedi Piscium, Vol. Ill, p. 94 ; after Catesby. Spliyraena haiTacuda Evermann and Marsh, 1902, p. 11."., Fig. 20. 224 SCIENTIFIC SURVEY OP PORTO RJCO Fig. 71. — Sphyraena hnrracuda From Zoologica, IX Type locality. — Bahamas. Distribution. — AYest Indian fauna, north to Charleston, S. C, and Bermuda, south to Brazil. Common and generally distributed in Porto Eican waters. Specimen collected. — 1 : San Antonio Bridge, San Juan. Diag 710 sis. —Head 3.3; depth i (2 in head) ; eye 6.6. Dorsal V-I, 10; anal I, 8 ; scales 83. Body compressed ; maxillary reaching past front of orbit; teeth large. Irregular inky black spots on the flanks are a good distinguishing mark for this species of barracuda. Attains a very large size, commonly 3 feet or more long, sometimes 5 feet and perhaps even 10. Remarl-s. — This fish is valued for food in Porto Rico, where the prejudice on the ground that it is unwholesome does not attach to it as in Cuba. There are well authenticated instances of poisoning as a result of eating large individuals. Init no such case is known from Porto Rican waters. Habits. — A fierce voracious fish, paralleling in the sea the habits and appearance of the fresh-water pikes. There are a few well authenticated instances of its attacking man in the water. Sphyraena guachancho Cuvier and Valenciennes Small barracuda ; Guachanohe Sphyraena (juachancho Cuvier and Valenciennes, 1829, Hist. Nat. Poiss., Vol. Ill, p. .342. Sphyraena yuachaneho Evermann and Marsh, 1902, p. 116. ^ — ■^TZ^ ^^«t ^^^^ Fi(5. 72. — iS/jhiiroeiKi fiiiiiiliancho ^^^^ J))'^ ~^I^1^^^~^§^ From Zoologica, X Type locality. — Havana. Distribution. — West Indian fauna, north to Florida; generally com- mon, but rare in Porto Rican waters. Diagnosis. — Head 3 ; depth 7 ; eye 5.4. Dorsal V-I, 9 ; anal I, 8 ; scales 115. Body little compressed or sub-terete: pectoral reaching front of spinous dorsal ;' maxillary reaching front of orbit. Length 2 feet. Bemarl's. — Valued for food where abundant. MCHOL^S, I'ORTO RICO AXD THE VJNGIX It "^v^^^^ From Zoologica. A -^ ■ Ti/pe locality. — Havana. Distribution. — West Indian fauna from Cuba to Babia. TiichKb^d in tbe Porto Eican list on the authority of Dr. Stabl. Diagnosis. — Head 3.1 to 3.2 ; depth 2.2 to 2.3 in bead ; eye about 5. Dorsal V-I, 9; anal I. 9: scales 110. Body approaching sub-terete; pectoral not reaching front of first dorsal; maxillary not reaching front of orbit. Length 18 inches. Remarl's. — Valued for food and, where abundant, of commercial im- portance. POLYNEMIDAE Pclyneimis TJnnaens Polynenius virginieus Linnaeus Thread-fin ; barhndo I'oUiHvmus virginicitii Linnaeus. 175S, Syst. Xat.. ed. 10, \\. :',17. Folydactulus virginieus Evermann and iNIarsh. 10n2, i>. 117. I'lC. 74. — Poliincmiix ririjiiiicKx From Zoologica. X Type locality. — ^" America." Distribution. — West Indian fauna, north to the Florida Keys but not to Virginia. Rather common and generally distributed in Porto Rican waters. Specimens seen. — San Juan (peddled cooked). Diagnosis.-— Head 3.3; depth 3.2 to 3.3; eye 5. Dorsal \'1II-I, 12 anal III, 13; scales 60. x\ttains a length of about a foot. Ee marl's. — A useful food fish. 226 SCIENTIFIC SURVEY OF PORTO RICO HOLOCENTRIDAE Myripristis Cuvier Myripristis jacobus Cuvier aud Valenciennes Big-eyed squirrel-fish; Frere Jaques; candil Myripristis jacobus Cuvier and Valenciennes, 1829, Hist. Nat. Poiss., Vol. Ill, p. 162. Myripristis jacobus Nichols, 1915, Bull. Amer. Mus. Nat. Hist., Vol. XXXI V, p. 143. Porto Rico. Fig. 7.5. — Myripristis jacobus From Zoologica, X Type locality. — Martinique. Distribution. — West Indies to Brazil, generally common. Not com- mon about Porto Rico. Sp&cimens collected. — 2 : San Juan. Diagnosis. — Depth 3; head 4; eye large. Dorsal X-I, 14 or 15; anal IV, 13; scales 36 to 38. Preopercle without conspicuous spine at its angle ; color red. Attains a length of about a foot. Habits. — Seems to spend the daytime in hiding or in deeper water. Holocentrus Scopoli Holooeiitrus asoensionis (Osbeck) Common squirrel-tish ; cartinau Perca ascensionis Osbeck, 1771, Iter Chineusis, p. 388. Holocentrus ascensionis Evermann and Marsh. 1002. p. 118, PI. III. Fig. 76. — Holocentrus ascensionis From Zoologica, X Type locality. — Ascension Island, Distribution — Eocks and reefs of the tropical Atlantic, north to Florida, and numerous at Bermuda, generally common in the West Indies. Common about Porto Eico, and recorded from St. Croix. yiCHOLS, PORTO RICO AM) THE VIRGIN ISLANDS 327 Specimens seen. — San Juau Market. ^ Diagnosis. — Head 3 to 3.1 ; depth 3 to 3.3 ; eye 2.8 to 3.3. Dorsal XI, 15 ; anal IV, 10 ; scales 48. Attains a length of 1 to 2 feet. Remarks. — About Porto Eico this fish is reputed to be unwholesome, but it is occasionally to be seen in the market. Holocentrus vexillarius Poey Black-barred squirrel-fish Holocentrum vcxillariitm Poey, 1860, Memorias, Vol. II, p. 158. Holocentrus vexiUarius Everuiann aud Marsh, 1902, p. 119. Fig. 77. — Holocentrus vexillarius From Zoologica, X Type locality. — Cuba. Distribution. — West Indies, not common, recorded from Porto Eico. Diagnosis.— Redid. 2.8; depth 2.6; eye 2.7. Dorsal XI, 13; anal IV, 9; scales 40, Maxillary extending to below first third of eye; dorsal with black markings. Size small. MULLIDAE Upeneus Cuvier Upeneus niaculatus (Bloch) Red goat-fish ; salmonete Mullm mactilatus Bloch, 1793, Ausl. Fische, and Ichth., PI. 348. Upeneus maculatus Evermann and Marsh, 1902, p. 120, PI. IV. Fig. 78. — Upeneus waruintus From Zoologica, X Type locality. — Brazil. Distribution. — West Indian fauna from the Florida Keys to Brazil. Abundant about Porto Eico. St. Croix. Specimens collected. — 2 : Ponce ^larket. 228 XCIENTIFIV SURVEY OF PORTO RICO JU'uKjiiosis. — Head 3.2; depth o.7 : eye 4 to 5. Dorsal YIII-I, 8; anal 11, 7 ; scales 30. Teeth in both jaws uniserial; side with black blotches. Lenutli !> to 10 inches. Eeinarl-s. — Extensively used and esteemed for food. I'peneus parvus Poey ►Small gfoat-fish VpcncHH iiamis I'oey, 1851, Meniorias, Vol. I, i». 226. VpencKx iKirvii.i Kvenaann and Marsh, 1002, p. 121. l-"iG. TH. — UtJCiicuii itarvus Type locality. — (^il)a. Distrihufion. — Known only from the type, which was obtained by Poey in Cuba, and from a specimen recorded from Porto Eico by Stahl. Diagnosis. — Dorsal \'I1-1. 8; anal II, 6; scales JrO. Teeth of both jaws biserial, at least in front; dorsals and caudal with dark cross-bands. Upeneus niartiiiious Cuvier and Valenciennes Yellow soat-fish: salmonete amarilla IJpencii.s jinirti)ticiifi Cuvier and Valeneienues, 3820, Hist. Nat. Poiss., Vol. Ill, p. 483. Vpenvu.s iiKirtiiiicii.s Everniann and .Marsh. 1902. p. 121. PL V. Fig. so. — L'pencus martinicus From Zoologlca, X Typ& locality. — Martinique. Distribution. — West Indies, north to Florida. Less abundant in Porto Rico than is the red goattish. Spedmens collected. — 1: C'ondado Pocks. San Juan. Diugno.m. — Head 3.3 ; depth 4 ; eye 3.4 to 3.5. Dorsal YIII-I, 8 ; anal IT, G; scales 37. Teeth of both jaws biserial, at least in front; dorsals and caiulal ])lain yellow. Attains tlie lenath of a foot. NICHOLAS, PORTO RICO AXD THE VIRGIX LSLAXD.^ 229 SCOMBRIDAE Auxis Cuvier Auxis thazard ( Lacepede) Frigate mackerel ; albacora Scomber thazard Lacepede, 1802, Hist. Nat. Poiss., Vol. IT, p. 9. Auxis thazard Evermann and Marsli, 1902, p. 122, Fig. 27. Fig. 81. — Auxin thazard From Zoologica, IX T^jpe locality.— BeUveen G° and 7° S. lat. off coast of Xew Guinea. Distribution.- — All warm seas, occasionally northward to Cape Cod. Probably not rare about Porto Eico. Diagnosis.— Bead 3.8; depth 4.4; eye 6. Dorsal X-12-VIII : anal 13-VII. Body mostly scaleless posteriorly; anteriorly covered with small scales, those of the pectoral region enlarged, forming a corselet. Attains a length of about 15 inches. Habits. — Frequently travels in large schools, and is very erratic as to presence or absence at a given locality. Scomberomonis Laoepede Sooinberomorus maoulatus (Mitchill) Spanisli mackerel ; carita Scoiiihcr iiiaciilafiis Mitcliill, 181.", Trans. I.lt. and I'liilos. i>. PI. VIII. Fig. 88. — Dvcaiitcnis luinctatus From Zoologica, IX Type locality. — Brazil. Distribidion. — West Indian fauna, Cape Cod to Brazil, only occa- sional northward. Xot uncommon ahout Porto Eico. Diaffnosis.—TLe&d 4; depth 5.6; eye 3.3. Dorsal VIII-I, 30-1; anal II-I, 2T-I; scutes 39. Common up to 6 inches long; rarely attains the maximum of 12 inches. Trachiirops Gill Trachurops ('rumenophtlialinus (Bloch) Goggle-eyed scad ; chicliarro Scohiber criiiiiciiDitlithdlinuN lUoeli, 179:!, Ausl. Fiscbe, and Iclitli., PI. CCCXLIIl. Trachurops crunioiophtliahnus Evermann and Marsh. 1902. p. 129, Fig. .30. Sclar crumciiophtliiilnius Meek and Hildebrand. 1925, Fishes of Panama, Pt. 2. Fig. Sy. — Trachurops criimenoiilithnliniix From Zoologica, IX Type locality. — Acara in Guinea. Distribution. — Cosmopolitan in tropical seas; found on both coasts of America ; nortli to Cape Cod in the Atlantic. Common in Porto Rican waters. 234 SCIEXTJFIC SURVEY OF PORTO RICO Diagnosis.— ILe&d 3.2 ; depth 3.8 ; eye 3.2. Dorsal VIII-I, 25 or 26 ; anal II-I, 22 ; scutes 35. Shoulder girdle with a deep cross furrow, and a fleshy projection above the furrow. Rarely attains a length of 2 feet, usually 10 inches or less in length. Caranx Lacepede Caranx ruber (Bloch) Skip-jack; cibi mancho Scomber ruber Bloch, 1793, Ausl. Fische, and Ichth., PI. CCCXLII. Caranx ruber Evermann and Marsh, 1902, p. 130. Fig. 90. — Caranx ruier From Zoologica, X Type locality. — St. Croix. Distribution. — West Indies, casually north to I^forth Carolina. Not plentiful about Porto Eico. St. Croix. Specimens collected. — i : San Turce, San Juan. Diagnosis.— ne&d 3.5; depth 3.5; eye 5.4. Dorsal VIII-I, 26 to 27; anal II-I, 23 to 24 ; scutes 25 to 30. Soft dorsal and anal little elevated in front. Gill-rakers on lower limb of first arch 31 to 33. Specimens down to 4 inches in length, have depth 3 times, more or less, in length, with only slight tendency to become more slender with increasing size. Attains a length of about 15 inches. Remarks. — Riiljer is a misnomer for this fish, due to a wrongly colored figure ; it is more or less blue, never red. Habits. — An active, swift-swimming species, more plentiful about small islands and reefs than mainland shores. Caranx bartholoniaei Cuvier and Valenciennes Yellow-jack ; cibi amarillo Caranx bartholoniaei Cuvier and Valenciennes, 1833, Hist. Nat. Poiss., Vol. IX, p. 100. Caranx bartholomaei, Evermann and Marsh, 1902, p. 131, Fig. 33 (mis- titled). NICHOLS, PORTO RICO AND THE VIRGIX ISLANDS 235 Fig. 91. — Caranx bartJiotomaei From Zoologica, IX Type locality. — St. Bartholomew. Distribution. — West Indies, north to Florida and the Carolinas, rarely to Massachusetts. Common about Cuba, but not very common in Porto Rican waters. Diagnosis. —Head 3.3; depth 2.8; eye 4.8. Dorsal VIII-I, 26 to 27; anal II-I, 22 to 23 ; scutes about 30. Soft dorsal and anal little elevated in front. Specimens under 6 inches in length (to base of caudal) have the depth 2.5 or less in this length ; from six inches to a foot long, depth falls off very rapidly and at the length of a foot there is no depth differ- ence between this species and the preceding. Gill rakers on lower limb of first arch, 17 to 19. Attains a length of about 15 inches. Remarl'S. — The yellow-jack is closely related to the skip-jack, and occurs more or less in association with it in the West Indies. Its young, which are notably deeper-bodied, drift northward in the Gulf Stream, and have a prettily mottled concealing coloration among the gulf weed. Relative depth is a good criterion to separate small individuals of yellow- jack and skip-jack but, when a length of about a foot has been reached, there is no depth difference. The yellow-jack has decidedly coarser gill- rikers. Caraiix hippos (Linnaeus) Common jack ; crevalle Scomber hippos Linnaeus, 1766, Syst. Nat., ed. 12, p. 494. Caranx hippos Evermann and Marsh, 1902, p. 131, Fig. .31. Fig. 92. — Caranx hippos From Zoologica, IX Type locality. — Charleston, S. C. Distribution. — Tropical Atlantic and eastern Pacific, generally abun- dant, found on both coasts of America, north to Cape Cope and the Gulf 236 SClEyTlFIC SURVEY OF PORTO RICO of California. Not common in Porto Rican Avaters. Recorded from St. Croix. Specimens collected. — 8 : San Juan. Diagnosis.— Head 3; depth 2.5; eye 3.8. Dorsal VIII-I, 20; anal II-I, 17; scutes about 30. Soft dorsal and anal more or less falcate in front. Canines well developed. Breast naked with a small rhombic scaled area before ventrals. Attains a length of about 21/2 feet and a weight of 20 pounds. RemarJcs. — A good game fish but does not rank high as a food fish. Habits. — The commonest species of its kind on continental shores, fre- quently penetrating estuaries to fresh water. An active predaceous fish, hunting singly or in small, loosely organized companies. Caraiix erysos (Mitchill) Hardtail-jack ; runner ; cojinuda Scomber criisos Mitchill, 1815, Trans. Lit. and Phil. Soc. N. Y., Vol. I, p. 424. Caranx cryms Everniann and Marsh, 1902, p. 132, Fig. .32, PI. IX. Fig. 93. — Caranx rrusos From Zoologica, IX Type locality. — New York. Distribution.— \Yest Indian fauna, from Cape Cod to Brazil, generally abundant southward. Not very numerous at Porto Rico. A similar fish on the west coast of Middle America is very closely related if not the same. Specimens collected.— o : Santurce and Fort San Geronemo, San Juan. Diagnosis.— Read 3.5; depth 3.2; eye 5.6. Dorsal VIII-I, 24; anal II-I, 20; scutes about 45. Soft dorsal and anal more or less falcate in front, breast completely scaled. Attains a length of about 2 feet and a weight of 4 to 6 pounds. Remarl-s. — A well-known food fish. Caranx latus Agassiz Horse-eyed jack; jurel Caranx latus Agassiz, 1829, Pise. Bras., p. 105. Caranx hitus Evermann and Marsh, 1902, p. 132, not Fig. .33 (mistitled) . NICHOLS, PORTO RICO AMJ THE VIRGIN ISLANDS 337 Fig. 04. — Carana: latus From Zoologica, X Type locality. — Brazil. Distribution. — All tropical seas, abuudant in the West Indies. By far the commonest species of Caranx in Porto Rico. Specimens collected. — 1 : San Juan. Diagnosis. — Head 3.3 to 3.4; depth 2.7 (more than 3 at a standard length of 3 feet) ; eve 4. Dorsal VIII-I. 22 : anal II-I, 18 ; scutes about 35. Soft dorsal and anal more or less falcate in front : breast completely scaled. Occasionally reaches a length of 3 feet or more. In such cases it is less deep-bodied than in the more usual, shorter representatives of the species. Bemarks. — Various species of Caranx named in different parts of the world are referable to C. latus, unless they be taken for the types of local races of it, which may prove differentiable. Habits.— The preferred habitat of this species seems to be the shores of islands, and in the West Indies it more or less replaces the common jack, which it resembles in appearance. Vomer Cuvier and Valenciennes Vomer setapinnis setapinnis (Mite hill) Moonfish ; corcobado Zeus setapinnis Mitehill. ISlo. Trans. Lit. and I'liilos. Soc. N. Y. for 1815, p. 384. Vomer setipimiis Cope, 1871, Trans. Amer. Phil. Soc, Vol. XIV, p. 472. St. Croix. Vomer setapinnis setapinnis Nichols. 1918, Bull. Amer. Mus. Nat. Hist.. Vol. XXXVIII, p. 669-676. Torto Rico. Fig. 'Jii.^Voiner setapinnis From Zoologica. IX Type locality. — Xew York. Distribution.— Cape Cod (casually Maine) to the West Indies. Re- corded from Porto Rico and St. Croix. 238 SCIENTIFIC SURVEY OF PORTO RICO Diagnosis.— Uead 3.2 to 3.3; depth 2 in adult, 1.2 to 1.8 in young. Dorsal VIII-I, 21 or 22; anal II-I, 19 or 20; scutes 20 Remarks. — There is doubt whether one or both forms of Vomer seta- pinnis occur in Porto Rican waters. In recording V. setipinnis from St. Croix, Cope, 1871, presumably did not differentiate between the races. Nichols, 1918, refers small specimens from Porto Rico to V. s. setapinnis, but there is a chance of error due to the small size of the speci- mens examined. Evermann and Marsh were correct in their reference of Porto Rican material to V. gabonensis Jordan and Evermann, if the measurements given are based on some of their larger specimens from that region. Both forms apparently occur in Cuba (Nichols, 1912, Bull, Amer. Mus. Nat. Hist., Vol. XXXI, p. 186), and probably also in Porto Rico. Vomer setapinnis cubensis Nichols Deep moonfish ; corcobado Vomer setapinnis cuhensis Nichols, 1918, Bull. Amer. Mus. Nat. Hist., Vol. XXXVIII, p. 672. Vomer gahonensis Jordan and Evermann, 1896, Bull. U. S. Nat. Mus., Vol. XLVII, Pt. 1, p. 934, not of Guicheuot. Vomer gahonensis Evermann and Marsh, 1902, p. 133. Fig. 96. — Vomer setapinnis cubensis From Zoologica, X Type locality. — Cuba. Distribution.— West Indies, probably (this form) to Brazil. Common about Porto Rico. Diagnosis.— TLead 2.7 to 2.8; depth 1.5 to 1.6 (specimens 4 to 6 inches standard length) ; eye 3.8. Dorsal YIII-I, 22 ; anal I, 18; scales minute. In Porto Rico the moonfish attains the weight of a pound or more. Remarl-s.— The moonfish is one of the compressed, silvery Caranx-like fishes, deep-bodied, particularly when young, at which stage it drifts widely in ocean currents. It is much used for food. NIVHOLH, PORTO RICO AXD THE VIRGIN ISLANDS 239 Selene Lacepede Selene vomer (Linnaeus) Angel moonfish ; silver angelfish ; jorobado Zeus vomer Linnaeus, 1758, Syst. Nat., ed. 10, p. 266. Selene vomer Evermann and Marsb, 1902, p. 135, Fig. 34 and 35. Fig. 97. — Selene vomer From Zoologlca. IX Type locality. — '''America." Distribution. — Both coasts of tropical America, from Cape Cod to Brazil, and from Lower California to Peru. Included in the Porto Eican list on the authority of Poey and Stahl. Diagnosis. — Head 3; depth 1.5 (the young much deeper) ; eye 2.5 in preorbital (4 in head, obliquely) ; scales minute, lateral line wholly un- armed. Dorsal VII-I, 23 ; anal II-I, 18. Soft dorsal and anal lobes produced, more or less filamentous. Young have elongate ventrals and some of the dorsal spines filamen- tous, both fins becoming reduced with age; but the lol:)es of the dorsal and the anal are relatively longest in the adult. Attains a total length of about a foot and a weight of about 2 pounds. Remarks. — A delicious panfish. Chloroscombrus Girard Cliloroseombrus chrysurus (Linnaeus) bumper ; casabe Scomher chrj/siirus Linneaus. 1766. Syst. Xat., ed. 12, p. 494. Chloroscombrus chrijsnrus Evermann and Marsh, 1002, p. 136, Fig. 36. Fig. 98. — Chloroscombrus ehrysurus From Zoologica, IX 240 SCIENTIFIC SURVEY OF PORTO RICO Type locality. — Charleston, S. C. Distribution. — West Indian fauna, Cape Cod to Brazil, common on the south Atlantic coast of the United States and about Cul^a, appar- ently not common in Porto Eican waters. Diagnosis.— Hesid 3.7 to 3.8; depth 2.3 to 2.4; eye about 3. Dorsal VIII-I, 26 ; anal II-I, 26 ; lateral line unarmed Curve of the abdomen greater than that of the back. Attains a length of about 10 inches. Remarks. — ISTot valued as food, the flesh being thin and dry, the bones large. Trachinotus Lacepede Trachinotus glaucus (Bloch) Gaff-topsail pompano ; palometa Chaetodon glaucus Bloch, 1787, Ausl. Fische, and Ichth., PI. 210 ; based on a drawing by Plumier. Trachinotus glaucus Evermann and Marsh, 1902, p. 137, Fig. 37. Fig. 99.^ — Trachniotus glaucus From Zoologica, X Type locality. — Martinique. Distrihution.—West Indian fauna, from Virginia to the Caribbean. Common from the Carolinas to Florida ; common in Porto Eican waters. St. Croix. Specimens seen. — San Juan. Diagnosis.— Head 4; depth 2 to 2.2; eye 3.6. Dorsal I (procumbent) VI-I, 19; anal II-I, 17 to 18; scales fine. Body much compressed; sides with narrow black cross-bars ; lobes of vertical fins elongate, extend- ing past middle of caudal fin in adult. Attains a length of a foot or more. RemarTcs. — In Porto Eico the gaff-topsail pompano seems to rank with the species of Caranx in food value. It is not much prized for food else- where. NICHOLS, PORTO RICO AND THE VIRGIN ISLANDS 241 Tra«?hinotus falcatus (Linnaeus) Round pompano ; palometa Lal)rus falcatus Linnaeus, 1758, Syst. Nat., ed. 10, p. 284. Trachinotus falcatus Everniann and Marsli, 1902, p. 138, Fig. 38. Chaetodon rhomhoides Bloch, 1787, Ich., PI. 209. Martinique. Trachijnotus rhomioides Cope, 1871, Trans. Amer. Pliil. Soc, Vol. XIV, p. 472. St. Croix. Fig. 100. — Trachinotus falrntus From Zoologica, IX Type locality. — "America." Distribution. — West Indian fauna, Cape Cod to Brazil, couiinon south- ward; apparently only the young are carried by the Gulf Stream as far north as Woods Hole, Massachusetts. Common in Porto Eican w