y oy Py
o aus ie

RS

the Museum of Comparative Zodlogy
AT HARVARD COLLEGE. Aa

OVE. SIV.) INO: 222

A CONTRIBUTION TO THE ZOOGEOGRAPHY
‘OF THE WEST INDIES, WITH ESPECIAL
Se REFERENCE TO AMPHIBIANS
AND REPTILES.

ENG

THOMAS BARBOUR.

WITH ONE PLATE.

CAMBRIDGE, U.S.A:
Printed for the Museum.
Marcu, 1914.

79 ‘ KN 7 Le My 1, oe
of the Museum of Comparative Zodlogy
| AT HARVARD COLLEGE,

Vou. XLIV. No. 2.

OF THE WEST INDIES, WITH ESPECIAL
REFERENCE TO AMPHIBIANS
AND REPTILES.

BY
THOMAS BARBOUR.

WITH ONE PLATE.

‘CAMBRIDGE, U.S. A.:
Printed for the Museum,
Marcu, 1914.

n iy os af 2
"

TABLE OF CONTENTS.

PaGeE.

INTRODUCTION . i 5 i : : 3 ; ' é A j . 209
NoTE . ; 3 $ & 5 5 : 4 , é 5 : 5 ONS
LIST OF SPECIES INCORRECTLY RECORDED FROM THE WEST INDIES. 217
INTRODUCED SPECIES (Fortuttously or otherwise) : ! : 0 : 2220
ZOOGEOGRAPHY 224
CuBA 224
JAMAICA DT
Hartt AND San Dominco 227
Porto Rico 228
THE VirGiIn ISLANDS 229
Tue Lesser ANTILLES 230
GRENADA 230
CONCLUSIONS 236
ANNOTATED LIST OF THE SPECIES 238
TABLE OF DISTRIBUTION . 347

PLATE

A CONTRIBUTION TO THE ZOOGEOGRAPHY OF THE
WEST INDIES, WITH ESPECIAL REFERENCE
TO AMPHIBIANS AND REPTILES.

INTRODUCTION.

SINCE its earliest years the Museum of Comparative Zodélogy has received
many collections representing the fauna of the West Indian Islands. To men-
tion a few of these, Louis Agassiz and the other scientists on the Hassuer col-
lected at St. Thomas, on their memorable voyage; and later — from 1877 to
1880 — the Biaxe visited very many of the islands. The opportunity to col-
lect upon all of them was eagerly grasped by Mr. Samuel Garman, who was
Assistant Naturalist on the BLakr during part of the time that she was in charge
of Alexander Agassiz. Collections were procured upon other islands than those
at which the BLaxn touched by the kindness of enthusiastic colonists and others.
The material available from Porto Rico consists of some received from Dr. A.
Stahl, long a resident of Bayamon; some collected by Garman; and other speci-
mens got by exchange from the United States national museum, collected by
Stejneger and Richmond. From San Domingo as well as Guadeloupe there are
specimens sent to Louis Agassiz by Duméril from the Paris museum. This
material has a unique value, since it formed in many cases part of the material
studied by Duméril and Bibron for their great Erpétologie général. More re-
cently there has been received material collected by Mr. A. H. Verrill at Santiago
dela Vega. From Haiti the Museum contains the large collection of Weinland,
made at Jeremie and other points in the western district of the island; and from
this collection Cope described many of the species peculiar to the island, although
some were described by Garman at a later date. The most recent Haitian
material comes as a part of the results of Mr. W. M. Mann’s trip to the island
from November, 1912 to February, 1913. From Cuba there is material con-
tributed by Filipe Poey, Samuel H. Scudder, Wirt Robinson, and others. I have
added to the collections the material collected during trips to the Bahamas,
Cuba, and Jamaica. One of the special incentives to writing this paper was
afforded by a large and finely preserved collection received from Grenada.

210 BARBOUR: ZOOGEOGRAPHY.

During a part of the summer of 1910 (August 17 to September 25) Messrs.
G. M. Allen and C. T. Brues worked at Grenada in the interest of the Museum
of Comparative Zoélogy. In view of the fact that in several branches the Mu-
seum already possessed adequate collections from this island, special efforts
were made to collect only the most interesting and little-known elements in
the fauna. Thus Dr. Allen discovered the armadillo, long known by hearsay
only, which has proved to belong to a new race. He found besides several bats,
either new species, or others of special interest. No attempt was made to do any
marine collecting. Only few birds were wanted; and concerning these a short
paper has been published (Barbour, Proc. Biol. soc. Wash., 1911, 24, p. 57-60).
Mr. Brues devoted himself particularly to the insects; and, as a result, some fine
material was procured. Prof. W. M. Wheeler has already written upon the ants
(Bull. M. C. Z., 1911, 54, p. 167-172). Of more importance from a zoégeo-
graphical point of view was the finding, by both Allen and Brues, of specimens
of a new species of Peripatus (Brues, Bull. M. C. Z., 1911, 54, p. 303-318). No
Peripatus had previously been known from Grenada; and its close affinity with
species in Trinidad and Guiana, together with the distinctness due to its isolated
island life, make it important evidence, first, that it was not brought to Grenada
by human agency, or otherwise fortuitously; secondly, that Grenada has never
been completely submerged since its first separation from South America on the
one hand and Antillea on the other. The species is extremely rare, and its dis-
tribution is very closely confined to a small highland area of virgin forest in the
middle of the island. It does not, so far as their careful collecting goes to prove,
ever occur in the low, cultivated lands, or in the Botanic gardens.

Dr. Allen and Mr. Brues made important archaeological collections, so
that the trip was a most successful one.

It has already been mentioned that during the cruises of the Buaxe Mr.
Alexander Agassiz was accompanied by Mr. Samuel Garman. After their
return to Cambridge, Garman published his well-known series of papers dealing
with the reptiles and amphibians of the various islands. Considering the fact
that the time on shore was always limited to the short stay while the ship was
in port, the collections are a monument to Mr. Garman’s prodigious industry.
But more important than the gathering of the material was the advance in
method he made in studying it. He was the first to recognize that the various
species existing in the Antilles did not occur scattered in a perfectly reasonless
way upon various islands; but rather that each island possessed a fauna with
certain well-marked features, and usually as well with well-marked local species

INTRODUCTION. 211

or races related to, but differentiated by isolation from, the forms found on the
neighboring islands. At first, certain of the so-called conservative zoélogists
objected at the making of such a large number of new species. Time, however,
has justified Garman’s work to a large extent.

Since the original writing of this paper Mrs. Barbour and I again visited
Cuba (January—March, 1910) for the purpose of filling as many gaps as possible
in the collection by collecting at localities in the island from which there was
previously no material in the Museum. Cuban material is peculiarly important
in this connection because of the close relationship of many Bahaman and Haitian
species with those on Cuba. There has been question as to the identity or dis-
tinctness of these, and there are also many interesting and important points
regarding the relative abundance and local distribution of species within the
island. I left for Havana January 30th, and went directly to Soledad estate,
near Cienfuegos, where I received the kind hospitality of Mr. Edwin F. Atkins
and his family; and much aid, especially from Mr. R. M. Grey, who is in charge
of the Harvard Botanical Station, and Captain Beal, of CoLtonta GUABAIRO
who was more than kind to Mrs. Barbour and myself. From Soledad we re-
turned to Havana, and were joined by Dr. and Mrs. J. L. Bremer. To Dr.
Bremer I owe many thanks for his aid to my collecting. Together we visited
Herradura, San Diego de los Bafios, and the city and vicinity of Pinar del Rio,
then later Madruga. We returned to Cambridge on March 15th. The notes
on these specimens have been incorporated into the paper which was previously
written. In connection with this Cuban trip, it is a great pleasure to thank
my old friend, Dr. Aristides Mestre, Assistant Professor in the University of
Havana, and Dr. Carlos de la Torre, Professor in the University of Havana,
for much kind advice and other aid, as well as for some very valuable books and
specimens.

A delay in the publication of the paper enables me to include the results of
my stay in Cuba during January—March, 1913. Prof. W. M. Wheeler and Mr.
Louis A. Shaw were with me for the first two weeks and during the entire excursion
J had the very great advantage and pleasure of the company of Dr. dela Torre and
of his assistant Mr. V. J. Rodriguez. Upon this trip we devoted some attention
to birds and mammals with uniform good fortune since many of the rarest and
most interesting species were secured. Our first trip from Havana was to Bolon-
dron where we were hospitably entertained by Mr. Edwin F. Sanborn at the
Incento ARMont«, from here we reached the edge of the Cienaga de Zapata at
Hato Jicarito. From Bolondron we went to Aguada de Pasajeros which we made

212 BARBOUR: ZOOGEOGRAPHY.

a base for several excursions into the neighborhood, spending several days at
the edge of the great Cienaga in a house kindly placed at our use by Sr. Fran-
cisco Morales. Sr. Morales’s cane fields at CoLontaA SAN FRANCISCO are on
the Rio Hanabana and the enormous saw-grass morasses may be conveniently
reached by horse-back. Travel in the swamp is quite a different matter to
reaching the edge, however, although it is remarkable to see through how deep
mud the horses will carry one before they must be left. The uneven hard sub-
stratum upon which the mud, covered with floating vegetation, rests almost
in a liquid state makes riding quite exciting and it is not uncommon for one’s
horse to flounder into a depression almost disappearing in the ooze to be gotten
out with some difficulty.

Crocodilus rhombifer the true Cuban crocodile was abundant about the
Laguna de Punta Gorda which was perhaps our most productive collecting
ground.

By the courtesy of Sr. Freyre and Sr. don Miguel Diaz the railways and
other means of transport on the great estates of Marta Victorta and PErRse-
VERANCIA were at our disposal and our hearty thanks are due them for much
courtesy and aid.

We returned to Havana from Aguada and Messrs. Wheeler and Shaw left
for the North while Mr. Rodriguez and myself made a short excursion to Matan-
zas, Union de Reyes and Alacranes (Alfonso XII). Our object was to secure
blind fishes and Crustacea from the caves of the southern part of Matanzas
province and in this we were successful.

Returning again to Havana Professor de la Torre joined us and we started
at oncefor Bayamo. From here we proceeded to Baire by rail and then by horse-
back to the village of Los Negros where we stayed some time and to Pozo
Prieto. Near the latter locality we spent several days at the cafetal Hl Alto of
Sr. Pedro Diaz which was a fine locality in the heart of the humid forest region
of the Sierra Maestra. Returning to Bayamo our party divided, Sefior Rod-
riguez going to Bueycito and the region of the Pan de Azucar while Professor de
la Torre and myself went to Manzanillo and by various stages south to Cabo
Cruz. Here the successful search was made for Cricosaura and Tarentola.
After this trip we went back to Manzanillo, to San Luis and then to Guan-
tanamo. Here we stayed with Mr. C. T. Ramsden whose study of Cuban birds
is well known. A return to Bayamo and Havana with a few short days there
together with Mrs. Barbour, who came down to meet me ended a most

charming and profitable trip. Our collection was small but the common

INTRODUCTION. 213

species were already well represented and only such species as were especially
desirable were searched for.

The revival of interest in zodgeography has prompted making this paper
more than a simple check list, although such a list of the reptiles and amphibians
of the Antilles has long been needed. The species which are included in this
list are only those which seem to be worthy of recognition; all other recorded
species are considered synonyms or incorrect records for the area. This com-
prises the Bahamas, Greater and Lesser Antilles as far south as Grenada; Swan
Islands and the Cayman group are also included. As stated elsewhere among
the notes on certain species, the practice has been followed of recognizing island
races as far as possible. It is considered that a more incorrect concept is gained
regarding the conditions among these islands by stating that the same species
occurs upon several islands, when there is generally a constant definitive varia-
tion seen in individuals from each island, even though in some cases this varia-
tion appears to be slight. Jordan’s law of evolution by isolation has often been
startlingly effective in producing very distinct species upon nearly adjacent
islands, where these islands have apparently — indeed, one might say certainly —
exactly the same physical and climatic conditions.

No attempt is made to give complete synonymies; and unless some special
comment is made, those species which have been considered synonyms by
Boulenger in his catalogues are so accepted.

This paper must be considered in the nature of a preliminary, since unfortu-
nately press of time and in many cases want of material have prevented the writ-
ing of a herpetology, with descriptions and keys for the identification of species.
This list will show where reptile collecting in the West Indies can be carried on
most profitably, what material from the region is available for study in the
Museum of Comparative Zoélogy; and these records, which have been carefully
culled, serve as a basis for certain considerations regarding the geographical dis-
tribution of Antillean reptiles and amphibians, and regarding the origin of this
fauna.

Norte.— After this paper was about completed, Mr. George Nelson of the
Museum staff made three visits to the Swan Islands thanks to the kindness of
Dr. W. A. Brooks. Together several months were spent there, and collections
made in winter, spring, and summer. For the first time the islands have been
studied by the same person at various times of the year. Mr. Nelson’s material
force the reconsideration of conclusions which maintain that the Swan Islands
had an oceanic origin. Special mention may be made of the very strange fact

214 BARBOUR: ZOOGEOGRAPHY.

that Mr. Nelson’s collections prove definitely that Tiaporus fulginosus Cope, a
teid genus supposedly peculiar to the islands does not occur there. The types
were collected by Townsend in 1887 and there is documentary evidence in his
correspondence with the U. 8. national museum which leaves no doubt as to
the fact that the type specimens were actually taken on Swan Island. If the
species was confined to the larger island, it, being wholly terrestrial in habit,
may have been exterminated by the introduced cats.

In 1911 Mr. Perey R. Lowe published a quite fascinating book, A Naturalist
on Desert Islands, though why ‘“‘desert islands” does not at once occur to one.
Lowe spent three weeks about the Swan Islands on the yacht Zenatpa landing
frequently and making considerable collections. He is firmly convinced that
they have never been in connection with any of the land masses which have in
the past existed in the Caribbean region. The geologic evidence which he brings
forward certainly supports his contention but, so far no borings have been pos-
sible to show the real underlying structure of the islands. The fauna is rich,
equally so upon Little Swan, which has never been inhabited and cultivated, as
upon the greater island, which has.

Remarking upon the indigenous species of hutia Lowe (p. 103) speaks of -—

“A strange race of vegetarian rats (Capromys), tree-dwellers or tree-climbers, and now
almost extinct on other islands in the West Indies, but which found their way across the sea

to Swan Island in the same fashion as the iguana, and there founded a specific race of their
own.”

Again on p. 112-114 we read :—

“This rat is of an extremely mild and almost genial disposition; has a head and body very
much after the style of an enormous guinea-pig; and is covered with rather long and silky
hairs protruding through a thick fur. Its specific name is Capromys thoracatus of True, and
the species thoracatus is restricted to Swan Island. The genus Capromys to which it belongs
is an interesting one; for it comprises arboreal forms which are only found in the larger West
Indian Islands such as Cuba and Jamaica, where it is in imminent danger of becoming extinct.
It is just possible, therefore, that Little Swan Island will, in the future, represent the last
stronghold of this peculiar and old time race of rats, for here they are left absolutely unmo-
lested; and no enemies, human or otherwise, seem likely to disturb them. Scientifically,
these rats are allied to the coypu (Myopotamus) of South America; an animal attaining to the
length of two feet, which lives in burrows near the water, and feeds on aquatic plants. How
these rats came to find their way to Swan Island is a little point in the problem of the distribu-
tion of species which may be worth referring to; for as we have seen, there is every reason to
suppose that Swan Island has never had any connection with the mainland, and is of infinitely
later date, geologically speaking, than the islands comprising the Greater Antilles, being of
quite recent coral origin. Indeed as far as its fauna is concerned, Swan Island might be looked
upon as an oceanic or pseudo-oceanic island.!_ If, as seems most likely, this Swan Island race

1 There are no snakes on Swan Island.

INTRODUCTION. 215

of rats was derived from the race which inhabits Jamaica (C. melanurus) the original progeni-
tors of the Swan Island species must have drifted over a sea space of at least three hundred
and twenty miles; for Swan Island lies exactly that distance in an almost due westerly direc-
tion from the nearest point of Jamaica, along the course of the Gulf Stream. The genus is
not found on the mainland of Honduras, ninety-eight miles to the south, which puts this
comparatively short sea-route out of court.

For the sake of those who may not have given such a subject a thought, we might add
that a journey of this nature across an open sea could only have been rendered possible by
the rats having been carried out to sea by means of a floating island of vegetation, or a mass
of entangled tree-trunks such as one often sees in mangrove swamps.”

This entire account is so inaccurate that one’s confidence in the author’s
interpretations and observations in geology are sadly shaken. This creature
which is in no wise related to the rats except that it is a rodent, is not like C.
melanurus which also, is not found on Jamaica. C. brownz is the Jamaican form,
a wholly distinct species now rare. owing to mongoose ravages. C. melanurus
is a long-tailed species from the high mountains of eastern Cuba. In Cuba two
other species, C’. pilorides and C. prehensilis, are abundant, the former extremely so.
The genus is not, except for the Swan Island species, confined to the Greater
Antilles, but a species, C. ingrahami, is found in the grass of one of the small low-
lying Plana Keys of the southern Bahamas. It is mathematically improbable
that an island of vegetation set afloat from a country like Jamaica where there are
no rivers capable of performing this relatively rare phenomenon would be in-
habited at the same time by a hutia. The island would also have to steer very
carefully to meet Swan Island which is very small, and even granting that there
was a current to carry it along or a suitable wind which would drive it two and
one half miles an hour, without submerging it, one hundred and twenty-one hours
approximately would be necessary for the voyage. Unless more than one hutia
was carried on the island, this voyage would have to be performed by a second
hutia which would have to reach the island before the death of the first. It is of
course, improbable that these events have ever happened. The habits of the
Jamaican Capromys are not such as to render it likely that it would ever have
attempted so precarious a sea trip and there are no physical conditions in Jamaica
which would have forced it to do so. Contrary to Lowe’s assertion there are
snakes upon the Islands, one of which represents an autocthonous race. Of course
the absence of fresh water and other adverse physical features strictly limit the
variety of living forms which the islands can support. Pulmonate gastropods !
occur and a variety of insects — so that the fauna is not surprisingly poor but

1For a notice of the affinities of the extremely interesting species of land shells peculiar to Swan
Island, see Clapp, Nautilus, Jan. 1914, 27, No. 9, p. 97-101.

216 BARBOUR: ZOOGEOGRAPHY.

rather the reverse. It will be surprising indeed if there is not geological evidence,
however difficult it may be to find, which will prove that the Swan Islands are
the remnants of an ancient land mass which once reached to Jamaica and per-
haps, or even probably, also that commissure which once reached from the main-
land to the east end of Cuba, via the Cayman Islands when the east end of
Cuba was an island separate from the rest of what is now Cuba. Oceanic
deeps are certainly not all of ancient origin; and there are many restricted areas
of Cuba, Jamaica, or Florida which if, by some oscillatory or submerging process,
they were left above the sea level and their surrounding areas had vanished
would be as difficult as Swan Island is to prove their earlier connection with
other land, near at hand or far away.

LIST OF SPECIES INCORRECTLY RECORDED FROM THE
WEST INDIES.

The following species which do not properly belong to the Antillean fauna
are recorded by Reinhardt and Liitken in their Bidrag til det vestindiske 6riges
og naynligen til de Dansk-Vestindiske ders herpetology (Vid. Meddel. nat.
foren. Kjébenhavn for 1862, 1863, p. 150-291).
carefully and exhaustively revised and tabulated all the records for this area, and

These authorities have most

it is therefore unnecessary to more than point out those that are evidently based
upon wrongly labeled specimens, and then refer to Reinhardt and Liitken, who

themselves were skeptical regarding very many of their records.

Name used by Reinhardt & Liitken.

Bufo ornatus Spix.

Cystignathus ocellatus (Linné).

Cystignathus typhonius Daudin.

Leiupurus marmoratus Duméril & Bibron

Hylaplesia picta Duméril & Bibron.

Bothrops bilineatus (Wied) var. unicolor Jan.

Crotalus horridus (Schlegel).

Elaps corallinus (Linné).

Elaps fulvius (Linné).

Elaps lemniscatus (Linné).

Geophis punctovittatus Jan.,

Homalocranium melanocephalum (Linné).

Homalocranium semicinctum Duméril &
Bibron.

Elapoides sieboldi Jan.

Streptophorus atratus Hallowell.

Erythrolamprus aesculapii (Linné).

Heterodon platyrhinus (Latreille).

Ischnognathus dekayi (Holbrook).

Gerarda bicolor Gray.

Hipistes fasciatus Gray.

Helicops angulatus (Linné).

Spilotes corais (Cuvier).

Spilotes pullatus (Linné).

Coryphodon constrictor (Linné).

Liophis reginae (Linné).

Lygophis lineatus (Linné).

Cyclophis aestivus (Linné).

Philodryas viridissimus (Linné).

Ahaetulla liocercus (Boie).

Dryiophis acuminatus (Wied).

Name in present use.

Bufo crucifer Wied.

Leptodactylus ocellatus (Linné).

Leptodactylus typhonius (Daudin).

Paludicola marmoratus (Duméril & Bibron).

Dendrobates trivittatus (Spix).

Lachesis bilineatus (Wied).

Crotalus horridus Linné.

Elaps corallinus (Linné).

Elaps fulvius (Linné).

Elaps lemniscatus (Linné).

Atractus trilineatus (Wagler).

Homalocranium melanocephalum (Linné).

Homalocranium semicinctum Duméril &
Bibron.

Geophis chalybaeus (Wagler).

Streptophorus atratus Hallowell.

Erythrolamprus aesculapii (Linné).

Heterodon platyrhinos (Latreille).

Storeria dekayi (Holbrook).

Gerarda prevostiana (Eidoux & Gervais).

Hipistes hydrinus (Cantor).

Helicops angulatus (Linné).

Spilotes corais (Cuvier).

Spilotes pullatus (Linné).

Zamenis constrictor (Linné).

Liophis reginae (Linné).

Aporophis lineatus (Linné).

Cyclophis aestivus (Linné).

Phylodryas olfersii (Lichtenstein).

Leptophis liocercus (Wied).

Oxybelis acuminatus (Wied).

218 BARBOUR: ZOOGEOGRAPHY.

Name used by Reinhardt & Liitken.

Dryiophis fronticinctus (Giinther).
Leptodeira annulata (Linné).
Himantodes cenchoa (Linné).
Ophisaurus ventralis (Linné).
Plestiodon laticeps (Schneider).
Gymnophthalmus quadrilineatus (Linné).
Cnemidophorus murinus (Laurenti).
Cnemidophorus lemniscatus (Daudin).
Cnemidophorus sexlineatus (Linné).
Salvator teguixin (Linné).

Basiliscus americanus Laurenti.
Polychrus marmoratus (Linné).
Tropidolepis undulatus (Latreille).
Hypsibatus agamoides (Gray).
Alligator cynocephalus Duméril & Bibron
Alligator punctulatus Spix.
Podocnemis dumeriliana (Schweiger).
Malacoclemys palustris (Gr.)

Emys scabra (Linné).

Emys concinna (Linné).

Cistudo carolina (Linné).

Name wn present use.

Dryophis fronticinetus Giinther.
Leptodeira annulata (Linné).
Himantodes cenchoa (Linné).
Ophisaurus ventralis (Linné).
Eumeces quinquelineatus (Linné).
Gymnophthalmus quadrilineatus (Linné).
Cnemidophorus murinus (Laurenti).
Cnemidophorus lemniscatus (Daudin).
Cnemidophorus sexlineatus (Linné).
Teius teguixin (Linné).

Basiliscus americanus Laurenti.
Polychrus marmoratus (Linné).

- Sceloporus undulatus (Latreille).

Plica plica (Linné).

Cayman latirostris (Daudin).

Cayman sclerops (Schneider).
Podocnemis dumeriliana (Schweigger).
Malaclemys centrata (Latreille).
Chrysemys scripta (Schoepft).
Chrysemys concinna (Leconte).
Terrapene carolina (Linné).

Subsequent to 1862 but few incorrect records of West Indian species have
been published. The following notes on four species show how all of these
came to be credited to Grenada.

Anolis chrysolepis Dumérit & Brisron.
Dumérit et Brsron, Erpét. gén., 1837, 4, p.94. BouLenasr, Cat. lizards Brit. mus., 1885, 2, p. 89.

This species has been recorded from Grenada on the following grounds:—

Anolis nummifer and A. turmalis, described by O’Shaughnessy (Ann. mag.
nat. hist., 1875, ser. 4, 15, p. 278), do not appear from the description to be the
same. Boulenger has included both in the synonymy of A. chrysolepis. This
question is, however, beside the issue, since no individuals which agree with the
description of any of the species come from Grenada, whence, however, the
types of A. turmalis were said to have been collected. It seems very probable
that a considerable collection from some South American mainland station was
sent to the British museum labeled Grenada. The specimen of Plica plica
(Linné) which Boulenger records from the same locality, is an exactly parallel
case, and this specimen as well as one of Norops auratus probably came to the
British museum at the same time.

This was written before the receipt of a letter dated November 6, 1911,

LIST OF SPECIES INCORRECTLY RECORDED. 219

in which Dr. Boulenger tells me that my surmise was correct; and that a collec-
tion from New Grenada (Colombia) got credited to the island of Grenada. This

affects also the record for the following species.

Plica plica (Linn&).

LINnNE, Syst. nat., ed. 10. 1758, 1, p. 367. Bouencer, Cat. lizards Brit. mus., 1885, 2, p. 180.

From New Grenada, not island of Grenada.

Norops auratus (Daupin).

Daun, Hist. nat. rept., 1802, 4, p. 89. BouLnnerr, Cat. lizards Brit. mus., 1885, 2, p. 95.

A record having the same history as the preceding.

Gonatodes vittatus (Lic#TENnsTErn).

Licutenstretn, Nomencl. reptil., 1856, p.6. Boutrencmr, Cat. lizards Brit. mus., 1885, 1, p. 60.

The same history as the preceding. The record of this species from Domin-
ica, needs verification, and is worthy of special mention in connection with the

foregoing from Grenada.

Tropidodactylus onca (O’SHAUGHNESSY).

O’SHAUGHNESSyY, Ann. mag. nat. hist., 1875, ser. 4, 15, p. 280. BouLEeneer, Cat. lizards Brit. mus.,
1885, 2, p. 97.

The types of this species came from ‘“‘ Venezuela and Dominica.”” In answer
to my query regarding the probable accuracy of the data for these lizards, Dr.
Boulenger writes, ‘‘The Tropid. onea was purchased of a dealer (Mr. Cutter).
The locality Dominica is therefore open to doubt.”’ The species, thus, is obvi-
ously Venezuelan only.

Casual mention may also be made-of two records which Reinhardt and
Liitken evidently overlooked, namely Testudo polyphemus and Cayman sclerops,
which Gray recorded from Cuba (Ann. mag. nat. hist., 1840, 5, p. 115).

Anolis lineatus Dauprn.

Daupin, Hist. nat. rept., 1802, 4, p. 66, pl. 48, fig. 1. Boutmnenr, Cat. lizards Brit. mus., 1885, 2, p. 38. ~

At various times this species has been reported from the Antilles. Duméril
and Bibron (Erpét. gén., 1837, 4, p. 146) recorded it from Martinique. It is
really confined to a region farther south. Meek (Field mus. publ. Zool., ser.
1910, 7, p. 416) records specimens from Aruba and Curacao.

220 BARBOUR: ZOOGEOGRAPHY.

The case of this species is quite different from the others. Still it seems
wisest to consider it here instead of in the list of certainly authenticated species.

Gymnophthalmus luetkenii Bocourt.

Bocourt, Miss. sci. Mex. Reptil., 1881, p. 474. Boutenarr, Cat. lizards Brit. mus., 1885, 2, p. 429.

The type of the species was said to have been collected on St. Lucia by
Bonnecourt, who sent a single specimen to the Paris museum. The specimens
of this genus from St. Lucia reported by Garman and Boulenger all belong to
G. pleti, so that it is probable that the locality of this type is incorrect. Never-
theless, the evident close relationship between the two species suggests that G.
luetkenti will eventually be found on some other of the Lesser Antilles, unless
indeed it exists in St. Lucia alongside of its congener. In this case, it must be
very rare to have escaped capture by so excellent a collector as Garman.

In 1860 Cope described Scolecophis fumiceps (Proc. Acad. nat. sci. Phila.,
1860, p. 371) stating that the type was in the Museum of Comparative Zoélogy
and that it probably came from Cuba. It probably did not come from Cuba,
and I cannot find that the specimen which Cope borrowed was returned. Bou-
lenger considers Cope’s new species a synonym of Homolocranium planiceps, (cf.
Cat. snakes Brit. mus., 1896, 3, p. 226).

Ameiva bifrontata Corr.
Cops, Proc. Acad. nat. sci. Phila., 1862, p. 67.

This species was described from St. Thomas. It is confined to the mainland
of South America.! The type locality was the result of an error in labeling.

INTRODUCED SPECIES. (FORTUITOUSLY OR OTHERWISE).

Bufo marinis (LINNs).
Linné&, Syst. nat., ed. 10, 1758, p. 211. Boutenasr, Cat. Batr. Sal. Brit. mus., 1882, p. 315.
A South American species which has been carried purposely to almost all the

Antilles. We have it from Grenada, Barbados, St. Lucia, St. Kitts, Martinique,
Nevis, Montserrat, and Jamaica, as well as from the Bermudas.

1The status of this species has just been fixed by Ruthven (Occas. papers, Univ. Mich. mus.,
Dec. 27, 1913, no. 2, p. 1-3).

INTRODUCED SPECIES. 221

Hyla rubra Daupin.

Daovpin, Hist. nat. rain., 1803, p. 26, pl.9. Bounpnemr, Cat. Batr. Sal. Brit. mus., 1882, p. 403.

This mainland species has been recorded from St. Lucia by Boulenger
(Proc. Zool. soc. London, 1891, p. 354). It is probable that its introduction
may have been accidental. Hylas are frequently carried about in shipments
of living plants, and this case may easily be parallel to the finding of Hyla infra-
frenata, a Papuan species, in the Botanical gardens at Buitenzorg, Java.

Hyla squirilla Bosc.

Daupin, Hist. nat. rain., 1803, p. 18, pl. 3, fig. 2. Boutenaer, Cat. Batr. Sal. Brit. mus., 1882, p. 398.

The specimens from Stranger’s Cay upon which the Bahaman record for
this species rests have been reéxamined; and while they are not especially well
preserved, there can be no doubt that they belong to this species. They do not
show the markings so characteristic of the closely related H. femoralis, and the
absence of web between the fingers precludes their inclusion in H. carolinensis.
There can be no doubt regarding their artificial introduction, as was suggested
when first the specimens were recorded (Barbour, Bull. M. C. Z., 1904, 46, p. 56).

Constrictor constrictor LINNE.

Linnf&, Syst. nat., ed. 10, 1758, 1, p.215. BouLenemr, Cat. snakes Brit. mus., 1898, 1, p. 117.

This mainland species occasionally reaches the southern Lesser Antilles,
carried over on masses of flotsam and jetsam from the mouths of the large rivers
of the mainland. Wallace (Island life, New York, 1881, p. 71) has said that
‘“A large boa-constrictor was once floated to the island of St. Vincent twisted
round the trunk of a cedar-tree, and was so little injured by its voyage that it
captured some sheep before it was killed. The island is nearly two hundred
miles from Trinidad and the coast of South America, whence it almost certainly
came.”

Such individuals have undoubtedly come to the islands once in a while
since time immemorial, and yet never anywhere in the area has the species be-
come established. Nor has any other species been proved to have been intro-
duced and established by similar means.

For the use of the generic name Constrictor instead of Boa, consult Stejneger,
Proe. U.S. nat. mus., 1902, 24, p. 185.

222 BARBOUR: ZOOGEOGRAPHY.

Hemidactylus mabouia (Morravu DE JONNES).

Moreau DE Jonngs, Bull. Soc. philom. Paris, 1818, p. 138. BouLrncerr, Cat. lizards Brit. mus., 1885.
1, p.122. Sresnecer, Rept. U.S. nat. mus. for 1902, 1904, p. 599, fig. 40-45.

This gekko is quite in a class by itself with regard to its ease of establish-
ment after accidental distribution. It has been found upon many of the West
Indian Islands:— Cuba, Jamaica, Haiti, San Domingo, Porto Rico, St. Thomas,
St. Croix, Just van Dyke, Tortola, Dominica, Sta. Lucia, St. Vincent, Barbados,
Martinique, Petite Martinique, Mustique, Becquia, and Grenada. It is a wide-
spread species carried from place to place in cargoes of wood and fruit. It is
common on the Brazilian coast and elsewhere in South America, along the shores
of East and West Africa, and Madagascar.

I procured a specimen on the wall inside a fisherman’s house at the foot of
Morro Castle, Havana, March 10, 1912. It was lying beside an oil light waiting
for insect prey just as its relatives may so often be seen in the East Indies. I was
told by the people that they never appear about electric lights as the glare is too
strong.

Vague rumors are current upon all the West Indies concerning the occasional
appearance of crocodiles upon their shores. From the following account it will
be seen that this may happen, and perhaps does quite frequently. It is a very
noteworthy fact that there has never been an establishment of any species in the
Antillean region proper, except, of course, upon Cuba, Jamaica, and Haiti,

where they were once very abundant and may still be found.

Crocodilus intermedius GRavrEs.

Graves, Ann. gén. sci. phys., 1819, 2, p.344. BouLenerr, Cat. chelonians, etc., Brit. mus., 1889, p. 280.

What appears to be the only existing record for the capture of a crocodile
in Grenada was secured by Dr. Allen during his visit. He reports that one came
ashore on the east side of Grenada opposite St. George’s, about 6 September,
1910, where it was seen on the beach by a negro. Later it was observed swim-
ming between the beach and a nearby point of land, and a few days after was
found dead on the beach. Its death occurred, unfortunately, while Dr. Allen
was at work in another locality; and before he could secure the specimen, the
teeth had been broken out by natives, and as the body putrefied on the shore,
it was ordered buried by the local authorities. In its mutilated condition, how-
ever, it was photographed by a local photographer, Mr. Smith, who very kindly
gave Dr. Allen a print. This shows at once the elongate, slender snout, almost

INTRODUCED SPECIES. 223

gavial-like, which distinguishes Crocodilus intermedius from any other species
on this continent. Boulenger (loc. cit.) says of this species that the largest known
specimen measures nearly four metres. Mr. Smith, who measured the Grenada
example, reported it as being 14 feet in length. It evidently came from the
Orinoco, whence only this species could have come; and was probably carried
by the strong westerly drift to the island. The most interesting and important
observation to be drawn from this fortuitous occurrence is the fact that so strong
and resistant a creature, aquatic at that, should not have been able to survive
what would appear to have been a rather easy journey. No storm of special
violence was noted, and crocodiles do not need to feed very frequently; yet this
individual lived but a short time. This is one of the best actual observations
bearing on dispersal by ‘‘flotsam and jetsam”’ that has been recorded by compe-
tent witnesses.

Col. Fielden, in the Zoologist (1888, p. 236; 1889, p. 298) records ‘‘the
interesting fact of an alligator being transported alive on the trunk of a tree from
the continent of South America to Barbados in 1886.” There is no suggestion

as to what species may have been represented.
Testudo denticulata Linns.

Garman remarks that this tortoise “feeds readily in captivity, and is kept
about the houses and carried from place to place much as the more common do-
mestic animals.”’ There are examples in the Museum from St. Lucia, St. Vin-
cent, and a very large mounted example said to be from Porto Rico. It has also
been reported from Cuba, upon several occasions. It is, of course, confined to
South America in a wild state.

The three following Cuban records cannot be eliminated from the list with

the same certainty that the foregoing West Indian records can.

Phyllobates bicolor Binron.
Brsron, Sagra’s Hist. Cuba. Rept., 1840, pl. 29, bis. Boutenenr, Cat. Batr. Sal. Brit. mus., 1882, p.
195.
Gundlach, during his long residence in Cuba, never found this species.
No specimen is to be seen in any of the local museums of Havana. Dr. Carlos
de la Torre, a most enthusiastic and indefatigable collector and accurate observer,
told me that he had never in his life-long collecting in Cuba been able to find an
example. Since among the other collections sent to France by Ramon de la

Sagra there were many examples which are now known never to have come from

224 BARBOUR: ZOOGEOGRAPHY.

Cuba, it is almost certain that this record is based on a specimen from some other
locality, probably Mexico. A number of the molluses which d’Orbigny de-
scribed from Cuba in the same work have been shown to have been Mexican and
n t Cuban.

Natrix anoscopa (Corr).

Corr, Proc. Acad. nat. sci. Phila., 1861, p. 299; Proc. U.S. nat. mus., 1892, 14, p. 673. BovuLenceEr,
Cat. snakes Brit. mus., 1893, 1, p. 242.

A Cuban species, the only Antillean representative of its genus. Cope
compares it with N. cyclopion and N. rhombifera. Dr. Stejneger is inclined to
believe that Cope had an African species. It is probably some species which
by mistake got marked with an incorrect locality label. The evidence that
Gundlach ever really found a species of Natrix in Cuba is very far from con-
vincing. There is no Natrix of any species in the Museo Gundlach of the
Instituto de Segunda Ensefianza in Havana.

Kinosternon baurii Garman.
GarMaN, Bull. Essex inst., 1892, 23, p. 141. Srmpenrock, Zool. jahrb. Suppl. 1909, 10, p. 442.

One of the types (M. C. Z., No. 4,050) of this species is said to have come
from Cuba. It was mentioned at an earlier date by Garman; and was said to
have been sent to the Museum, with much other Cuban material, by Filipe Poey
(Proc. Amer. philos. soc., 1887, 24, p. 286). This is the only member of the
family Cinosternidae which has ever been suggested to occur in the West Indies.
On a visit to Washington, April, 1912, I saw in the United States national mu-
seum a specimen of Gerrhonotus! which was also said to have been sent from Cuba
by Filipe Poey. Ofcourse both these specimens may be victims of misplaced
labels.

ZOOGEOGRAPHY.

Instead of discussing the geographical distribution of the genera or species,
it is more convenient to take up the fauna of each island or group of islands
having a homogeneous biota; and then in a general survey, to draw some general

conclusions regarding the main features of origin and dispersal of the fauna.

CUBA.

This island first claims attention because of many peculiar characteristics.
It shows a far more intimate relationship with both North America and Central

CUBA. 225

America than do any of the other islands. A few years ago it would have been
considered unreasonable to assert that Cuba had been long and intimately re-
lated to Central America, and had certainly been for a shorter time connected
with Florida or with the island which has since formed the region where Florida
is now. Only in 1902 did Vaughan (Science, January 24, 1902, 15, p. 148-149)
question the occurrence of remains of the fossil sloth Megaloecnus in Cuba, even
after the Cuban zodlogist, Dr. de la Torre had published his Observaciones geo-
logicas y paleontologicas en la region central de la Isla [de Cuba]. In this paper
the author recorded the finding of specimens at Cardenas, Sagua, and Santo
Domingo, in addition to the type, which came from the now classic locality,
Banos del Ciego Montero. Vaughan added, “I am not able to express an opinion
as to the correctness of these localities or in Torre’s ability to determine fossil
vertebrates. I am inclined to doubt because there has been so much error re-
garding those fossils concerning which we have subsequently been able to procure
definite data.’”’ Asis now known, Dr. dela Torre has not only discovered well
preserved remains of many individuals of Megalocnus; but he has uncovered
besides the remains of a great fauna of rodents, edentates, and other mammals,
as well as of tortoises and crocodiles. I have had the privilege of examining
his collection in Havana; and de la Torre’s final accounts of his discoveries
will be awaited with the greatest interest. Suffice it to say that, stimulated by
Vaughan’s remarks, he has vindicated his previous statements with splendid
success.

These finds, of course, prove Cuba’s intimate relationship with the main-
land; had they never been made, however, the proof offered by the existing
genera of reptiles is equally convincing. In Florida Liodytes alleni has a nearest
ally and possible progenitor in the Cuban species of Tretanorhinus. This genus
occurs upon no other of the West Indies. In Florida, Rhineura, the only repre-
sentative of the Amphisbaenidae in the United States, has possibly been derived
from the ancestors of one of the two Cuban species of Amphisbaena. Although
the genus Rhineura itself is an ancient one being known from the Miocene of
South Dakota (R. hatcheri Baur). The occurrence in Cuba of Phyllobates, which
is found upon no other island is very noteworthy. Tarantola cubana Gundlach
& Peters represents a genus occurring in the Mediterranean region of the Old
World, and affords an interesting parallel to the occurrence of Spelerpes and
Hyla pulchrilineata Cope in Haiti. As yet, no urodele has been discovered in
Cuba; and the allies of the Hyla mentioned are found in Mexico, the south-
eastern United States, and in the Old World, conspicuous among them being
Hyla anderson Baird, Hyla eximia Baird, and Hyla arborea Linné,

oy

926 BARBOUR: ZOOGEOGRAPHY.

The genus Chamaeleolis, with its single species, is autocthonous in Cuba,
and the reason for its origin is not clear. Its nearest ally certainly seems to be
Anolis equestris (Merrem) but it can hardly be derived from this species. The
only West Indian species of Norops, found commonly on the island, again em-
phasizes the intimacy of its Central American relationship. Still more does
the occurrence of Cricosaura typica Gundlach & Peters. This rare creature,
so far as known, is entirely confined to the small region about Cabo Cruz, and
this closely circumscribed region has conditions very similar to those obtaining
where the genus Xanthusia occurs. A Natrix has been recorded from this
island alone of all the region; but the correctness of the locality of the typical
specimen is open to grave doubt, and it is possible — perhaps even certain —
that it does not occur. Tretanorhinus has already been considered. Arrhyton
is a curious autocthonous genus of which so few specimens are known that it is
impossible to state whether the individuals that have served for the types of
the three species recognized in Boulenger’s Catalogue are really anything more
than the widely varying individuals of a single species. The genus has probably
been derived from some form similar to Contia, or some of its allies, from Central
America. Uvotheca dumerili Bibron is supposed to come from Cuba; but no
specimens have been taken there, so far as I can learn, since the discovery of the
species; and it is probable that this species is Central American, as are all the
others of the genus. The Cuban specimen in this Museum was sent by the Paris
museum, and probably represents one of the original series. Crocodilus rhom-
bifer Cuvier has its ally in Central America, where the species Crocodilus more-
letii A. Duméril is so closely similar as to be almost indistinguishable. All the
other Cuban species of reptiles and amphibians belong to genera which are
found on other islands as well, though their distribution through the group varies.
Hyla occurs in the Bahamas, in Cuba, Haiti, and J amaica, and the Caymans; of
Bufo, there are four species in Cuba, one in Haiti, and one in Porto Rico. Of
these, Bufo empusus (Cope) represents the type which has extended most widely,
being represented in Haiti by Bufo gutterosus Latreille, and in Porto Rico by Bufo
lemur Cope. The Cuban species Bufo longinasus Stejneger, known from the type
alone, is a form whose relationship cannot be made out from the material so far
collected. Its only ally is Bufo ramsdeni described later in this paper. Bufo
peltacephalus Tschudi is obviously of Central American origin.

Five species of Eleutherodactylus known for some time embrace three
peculiar to the island, and two occurring elsewhere,— one, H. recordii Duméril &

Bibron, occurring also in the Bahamas, and H. auriculatus (Cope) in Haiti and

HAITI AND SAN DOMINGO. 227

Porto Rico. A sixth species now described occurs apparently about Guanta-
namo only.

The distribution of the species of Gonatodes, Sphaerodactylus, Anolis, Lio-
cephalus, Celestus, Ameiva, Amphisbaena, Typhlops, Epicrates, Tropidophis,
Alsophis, and Leimadophis may all be found by referring to.the tables of distri-
bution. The distribution of each of these genera includes Cuba, being repre-
sented there by peculiar island species, as is the prevalent mode of occurrence
for these genera throughout that portion of the West Indies in which each occurs.

Tn conclusion, then, we may say that Cuba has sixty-three species of reptiles
and amphibians, of which fifty are peculiar to the island. The proportion is
probably larger than this, since a number of the records of Cuban species in
Haiti will probably be found to represent distinct island races when comparisons

of freshly collected and well-preserved material are possible.

JAMAICA.

No new species have been added to the fauna of Jamaica since I published
my Notes on the herpetology of Jamaica (Bull. M. C. Z., 1910, 52, p. 273-302,
pl. 1-2). Thirty-three species are recorded from the island; and of these, twenty-
six are peculiar. It is not necessary to go further into a special discussion of the
Jamaican fauna, since I have nothing to add to the remarks in the paper referred
to.

Hartt anp San Domtneco.

For the purpose of convenience these islands have been mentioned as Haiti
throughout the paper; and, as the fauna in the two republics seems to be rather
homogeneous, there is no special reason for not using this old Indian word, which
was originally the name for the entire island.

From Haiti alone, of all the Antilles, has been reported a tailed amphibian,—
Spelerpes infuscatus Peters,— and this fact, coupled with the occurrence of Hyla
pulchrilineata Cope, is significant of a relationship with the Mediterranean region,
such as has already been suggested by Scharff. Besides the occurrence of these
two anomalies, the fauna of Haiti is interesting as bearing out the belief that Cuba
and Jamaica have never been joined together, and any similarity of fauna that the
two islands show many be explained by the fact that each of them has at differ-
ent times been connected with both Haiti and Central America. Thus the gen-
era Bufo, Gonatodes, Liocephalus, and Amphisbaena are all found in Cuba and

228 BARBOUR: ZOOGEOGRAPHY.

most of them in Porto Rico, while not a single one occurs in Jamaica. On the
other hand, Aristelliger and Mabuya are found in Jamaica and Haiti, and the
latter in Porto Rico, and neither in Cuba. These remarks refer to entire genera.

The distribution of certain species may be traced out in the same way.
Thus Cuba has but one species of Celestus, while Jamaica has three and Haiti
four, these four being the representatives of both the Cuban and Jamaican spe-
cies. Broadly speaking, then, the Haitian species which have obviously been
derived from the Central American mainland have come by two routes, the one
the Cuba bridge, which undoubtedly endured for a considerable length of time,
and which existed after and perhaps even came into existence after the disap-
pearance of the connection between Jamaica and Haiti, since a number of Cuban
genera and groups of species are found in Haiti and Cuba, but not in Jamaica;
while a very small number of species derived from Haiti have representatives
that passed to Cuba. Alsophis and Leimadophis occur on all three islands,
but all the genera previously mentioned are confined to either pair of “islands,
always including Haiti, and not upon the third. The fauna is a large one, and
embraces at least fifty-five species, for this number takes no account of any of
the possible records for several species which are not absolutely authenticated.
Forty-seven species certainly, and possibly others, are peculiar to the island. It
will be seen from this, that, although a number of the continental genera occurring
upon Cuba are not represented in Haiti, the total number of species found on
the island is almost as oreat as that found upon Cuba, the difference being made
up by those which have been derived through direct connection with Jamaica.

Porto Rico.

Omitting special mention of Mona Island, which is not of great importance
in the present connection, we may pass directly to a brief study regarding the
fauna of Porto Rico. This has been very fully discussed by Stejneger in his
Herpetology of Porto Rico (Rept. U. 8. nat. mus. for 1902, 1904, p., 549-724).

Stejneger says (p. 562) :—

“The herpetological fauna of Porto Rico falls into two groups, namely, the species which
have in all probability originally extended their range from northeastern South America
and those whose ancestors came from the west, primarily from the present mainland of
Central America, and secondarily from the other Great Antilles.”

The fauna of Porto Rico is less m number than that of either Cuba or Haiti;
and is almost exactly comparable to that of Jamaica, there being twenty-six

THE VIRGIN ISLANDS. 229

recorded species, of which two are of accidental occurrence, while fourteen are
exclusively found upon Porto Rico; though a number of others, as Stejneger
has shown, have but a circumscribed range, occurring on Porto Rico, and either
Vieques, Culebra, or Mona Islands. Unlike either Haiti, Cuba, or Jamaica,
Porto Rico has no genus of either reptiles or amphibians that is peculiar to it.
This fact would seem to lend weight to the theory which has already gained
some supporters, that the actual land area of an island has a certain more or less
definite relationship to the species population of that island; and the larger the
island is, the greater the number of genera and species that will be found on it.
Cuba, he largest of the Antilles, and Haiti, the next largest, are in this way
sharply set off from beth Porto Rico and Jamaica, which are far smaller than
either of the other islands, and more or less similar in size.to each other. It is
interesting to note that the same condition of affairs obtains in the East Indies,
where Borneo, the largest island, has a greater number of species than Sumatra,
which is geographically much more closely related to the mainland, and which
might be supposed to have received probably a larger number of mainland types
than any of the other islands.

THe VirGiIn ISLANDS.

Owing to the great depth of water found between this group and the rest
of the Lesser Antilles, it might be supposed that a faunal boundary would be
found to exist here comparable to that which was formerly supposed to exist in
the East Indies between the islands of Bali and Lombok. This, however, is not
exactly the case. Of the Central American genera which do not reach the Lesser
Antilles, Hyla does not get to Porto Rico; Bufo reaches Porto Rico, but no
further; Liocephalus extends to Haiti only; Celestus to Porto Rico. Amphis-
baena, however, is found upon St. John, St. Thomas, and St. Croix; so that, while
none of these genera extend so far as the Lesser Antilles, these islands form the
limit of occurrence for only one of them; whereas the many characteristic
Antillean genera of both reptiles and amphibians such as Eleutherodactylus,
Sphaerodactylus, Anolis, Ameiva, Mabuya, Typhlops, Alsophis, and Leimado-
phis occur throughout the entire chain, without any reference whatever to a
faunal boundary line between these islands. The group consists entirely of
land masses of insignificant area, and with a correspondingly small number of
species upon each island.

230 BARBOUR: ZOOGEOGRAPHY.

Tue Lesser ANTILLES.

Although, generally speaking, the Lesser Antillean fauna may be said to
be a very homogeneous one, certain anomalies of distribution appear which in
some cases are extremely difficult to explain. At the present time the Fer-de-
lance, Lachesis, occurs only upon Martinique and St. Lucia; but early voyagers
wrote that the island of Petit Martinique received its name in their time on
account of the abundance of the Fer-de-lance there. It probably at some time
had a continuous distribution throughout the chain; but for some reason sur-
vived only on the two islands named. Other peculiarities are shown in the
occurrence of Gymnophothalmnus on Martinique and St. Lucia; Constrictor
orophias (Linné) on Dominica and St. Lucia, and possibly on St. Kitts. The
peculiar distribution of Clelia, Pseudoboa, Drymobius, and Herpetodryas
may be seen by referring to the table of distribution. Whether in some cases
accidental introduction has been accountable for these distributions, or whether
they have formerly been of wider occurrence and have died out on other islands,
is a matter upon which it is hardly worth while to speculate since we will never
know. Special mention, however, should be made of the fauna of Grenada,
which is far more like that of Trinidad and South America than that of any of
the other Lesser Antilles. Grenada is nearer Trinidad in point of distance than
any of the other Lesser Antilles, except of course Tobago, which is not Antillean
at all, but purely Trinidadian. This similarity in Grenada is not confined to
those organisms which, by their capacity to stand immersion in salt water, or
for some other reason, may be assumed to have come to the island fortuitously.
It is a fundamental feature which is most evident in those types that we know
are never carried from place to place by chance; and there can be but little
doubt that Grenada has been connected with the mainland of northeastern
South America more recently than have the other islands.

GRENADA.

It is important in this connection to turn to the physical character of the
Antilles. This has been well described in the following terms by Suess (Face
of the earth, 1894, 1, p. 544-545): —

“The Antilles may be divided into several zones.

“The first, innermost zone, which lies within the concavity of the arc, only appears in

the eastern part, in the Lesser Antilles. It is wholly of recent volcanic origin. This zone is
formed by the islands of Saba, St. Eustatius, St. Christopher (St. Kitts), Nevis, Redonda,

GRENADA. 2ol

Montserrat, the western half of the deeply incised Guadeloupe, Dominica, Martinique, St.
Lucia, St. Vincent, the Grenadines, and Grenada. These islands form a continuous circular
are, and some of them include three to four eruptive centres — Martinique, according to
Jonnés, even six.

“The next zone comprises the great and rich mountainous islands of the Greater Antilles
and the narrow but well-characterized border of the Lesser Antilles. The mountains are
formed in complete accordance with the type of the Coast Cordilleras of the south and of the
mountains of Venezuela. The oldest organic remains belong to the lower Cretaceous, and
these are only represented by a few traces; on the other hand, purer and highly fossiliferous
limestone deposits of the Turonian are present in this region, as in Trinidad and in north
Venezuela. The Cretaceous system is followed here, as in Trinidad, by a largely developed
series of Tertiary deposits, some members of which present an astonishing resemblance to
their representatives in Europe. These deposits form the connecting link with the third zone.

To this second group belong Cuba and Pinos Island, Haiti, Porto Rico, the Virginian
Islands with St. Croix, Anguilla, St. Bartholomew, Antigua, the eastern half of Guadeloupe
and a part of Barbados, and further the southern are which runs through Jamaica. That
part which lies in the Lesser Antilles forms a girdle running outside the voleanos.

The third and outermost zone comprises only middle Tertiary or even more recent deposits.
It nowhere rises in mountain ranges which can be properly so called, and is, indeed, as a rule,
quite flat, broader towards the north-west, narrowing away to the south-east. It includes
the Bahamas, all the banks of trifling height which succeed as far as Natividad, then Anegada,
Sombrero, Barbuda, and a part of Barbados. I should, however, be inclined to include in
this zone the whole peninsula of Florida, and perhaps even the level part of Yucatan.

“This regular arrangement of the islands was already clearly recognized in its main
features by Leopold von Buch. ‘Ail the voleanic islands of the Antilles,’ wrote the great
geologist more than sixty years ago, ‘follow one after the other in a continuous chain, uninter-
rupted by any islands which are not voleanic. On the other hand, there appears to the east of
these islands, lying outside them next the great ocean, another series of islands, somewhat
less clearly marked, which exhibits only a few traces of voleanic phenomena, and contains no
active voleanoes. This is an extremely remarkable fact!’ ”’

From the above it will be seen that Grenada belongs to the first zone of
islands, and is the most southern member of that chain. While Suess character-
izes them as “wholly of recent volcanic origin,’ he can not mean that these
islands arose from the sea, each independently, but rather that they represent
in some cases each as many as four or six eruptive centres which have broken
out along the mountain chain which made up the entire Greater and Lesser
Antillean region.

The tilting into the Caribbean sea of Antillea, and the consequent slipping
of the great, thick beds of which it was composed, may explain why each of the
three concentric arcs of islands has a homogeneous structure while it is still
intimately related to the conformation and structure of the other two groups.

The composition of these ares was briefly described above. In the inner
group voleanos have arisen along the great joint plane, and have flowed out so as
to overlie much of the older rocks, which nevertheless exist as the following quo-

232 BARBOUR: ZOOGEOGRAPHY.

tation from Watts makes clear. Thus, their real core has been, in some cases,
almost completely masked. The middle group is the dominant one, and has
become thrust up in many cases to form mountain chains of considerable height,
and to make all of the largest islands; while the outer series has given rise to
only low-lying areas, some of them, however, of great size. Florida and some
of the Bahamas, for instance, have a very considerable area.

Suess does not, however, give many facts as to the actual formation of each
island, and the little information available is scattered far and wide in many
journals. For Grenada there is no paper dealing with the geology which is in
any way comparable to Guppy’s Notes on a visit to Dominica, describing the
geological features of the island (Proc. Sci. assoc. Trinidad, December, 1869,
p. 377-392, 2 pl.). In view of this, the following quotation from an address
published in the London times by Dr. Francis Watts, Imperial Commissioner
of Agriculture for the West Indies, is of interest and importance :—

“A glance at the map of the West Indies almost certainly gives rise to the impression
that the island represent the tops of a submerged land area, at one time connecting North and
South America. Geologists support this popular impression, and refer to this lost country
as Antillia, believing that it had an actual existence; but they are by no means agreed as to
the period at which it was in being.

“Most of the islands bridging the gap across the Caribbean Sea show marked volcanic
characters. Thus if we consider the range from St. Thomas southward to Trinidad, we find
the Virgin Island group, Saba, St. Eustatius, St. Kitts, Nevis, Montserrat, Guadeloupe
(western half), Dominica, Martinique, St. Lucia, St. Vincent, the Grenadines and Grenada
all considered as volcanic, while St. Croix, Anguilla, Barbuda, Antigua, Guadeloupe (eastern
half), and Barbados are recognized as being built up extensively or entirely of sedimentary
and coral rocks. The study of these latter has afforded grounds for assigning geological dates,
more or less vague and tentative, to the various formations there found.

“Tt is generally assumed, though not perhaps very definitely stated, that the volcanic
islands contribute little which will enable one to learn much of the past history of the chain
of which they constitute so large a part — a view, which, coupled with a misunderstanding
concerning the true nature of some of the rocks classed as volcanic, has impedded the advance
of knowledge in no inconsiderable degree.

“Observations carried on for some time in the Island of Antigua have led to the conclu-
sion that the so-called volcanic rocks of that island — rocks to which previous writers have
referred as the oldest series, or the old igneous basement on which the latter series have been
superimposed — have not the character accorded to them, but that they are highly altered
and displaced portions of the sedimentary rocks which occur in an unaltered condition in the
middle of the island.

“These unaltered sedimentary rocks, which have usually hitherto been wrongly described
as tuffs, contain fossils in places, though these are not abundant. They consist of sandstones
(often so fine-grained as to be mud stones) and conglomerates. Associated with these rocks
are considerable masses of fossil wood completely silicified and in most instances so well pre-
served that the minutest structure is capable of examination in microscopic sections. In
addition to these woods, there are associated with these rocks silicious masses containing

GRENADA. 233

shells of fresh and brackish water types in which, as in the case of the woods, the structure is
often well preserved. These fossils should prove of value in determining the age of rocks,
and it does not appear that up to the present they have received sufficient attention in this
connection.

“Tt seems highly probable that these rocks are of similar age to the so-called “Scotland
Series” of Barbados, and that they are of considerable geological age. As they underlie
limestones, which geologists refer to Miocene times, they may well be of early Miocene or even
Eocene age.

“The point now to be considered is that in the southwestern portion of Antigua these sedi-
mentary rocks can be seen to pass into the series commonly described as volcanic, and to
give rise to metamorphosed rocks connected with the sedimentary on the one hand, and the
apparently voleanic on the other, the cause of this alteration being the intense activity which
in later geological times, obliterated the traces of the continental sedimentary areas and sub-
stituted the volcanic masses of such islands as St. Kitts, Nevis, Montserrat, and Dominica.

“With the clue afforded by the study of the Antigua rocks it seems possible to detect
traces of the old continental sedimentary areas in several of the islands hitherto regarded as
purely volcanic. In this manner, rocks which are probably of sedimentary origin may be
found in St. Lucia, where it is interesting to note that small quantities of fossil wood, appar-
ently similar to some of the fossil wood of Antigua, has been found in association with them.

“Again in Grenada, an island hitherto described as entirely volcanic, more or less altered
rocks of an apparently sedimentary character have been seen, and it is said that small quanti-
ties of limestone are associated with these. If this latter fact is verified, it may link up this
series with those of Antigua.

“The geologically unexplored Grenadine islets, lying between Grenada and St. Vincent,
may afford evidence of these hitherto unsuspected stratified rocks and may add to the story
of the old Antillean continent. The forms assumed by their weathering suggest in places
sedimentary rather than volcanic rocks, as do also some of the outlines of parts of Grenada
itself.

“These now nearly obliterated sedimentary rocks occurring as the Scotland Series of
Barbados, the so-called tuffs of Antigua and in a few other parts of the West Indies between
Trinidad and the Virgin Islands were probably derived from, or may have constituted part of,
the old Antillean continent. A careful study of this formation, coupled with the recognition
of its wider distribution than has been hitherto recognized, may be useful in determining the
age of that continent, which will probably be found to belong to the Eocene period.

“The object of these notes is to draw attention to the fact that there exists in the West
Indies many plainly indicated, but imperfectly studied, geological problems offering attrac-
tions to geologists and promising fair harvests of results. Much has been done by various
observers to indicate the general lines of research, but now the details require to be filled in.
It would seem that Antigua offers, perhaps, the best starting point for studies such as are
here indicated; after working out the problems of Antigua, those of Guadeloupe and St. Croix
would probably be readily solved. After this, it would seem possible to unravel the difficul-
ties presented by the volcanic islands, and for this purpose a detailed study of Grenada, the
Grenadines, St. Vincent and St. Lucia would doubtless indicate the line of attack. It is
thus evident that, to geological students wishing to combine pleasant travel with interesting
work, these islands offer many attractions and some opportunities of gaining recognition for
discoveries of interest.

“Such students, should they come, will find in most places plenty to engage their atten-
tion, and kindred spirits willing to direct their inquiries into useful channels without loss of
time, and one would view with pleasure and appreciation the formation of a small expedition

234 BARBOUR: ZOOGEOGRAPHY.

composed of students qualified for field work, under a competent senior, who would visit the
West Indies for a course of post-graduate study, and would look forward to the acquisition
of results of considerable scientific interest and value.”

That the general process of island formation has been extremely slow, and
that the evolution from a great land mass of almost continental size and perhaps
Eocene age, to the present condition of an extensive archipelago took place
evenly and without any great regions disappearing below the sea to reappear
later, is shown by the distribution of the land animals. These do not appear
now to fall into series corresponding in their distribution to the island ares, but
taken altogether form a single, remarkably homogeneous fauna, many character-
istic features of which may be traced from Cuba or Jamaica to Grenada. Natur-
ally types derived from Central America predominate in Cuba and in Jamaica
(Barbour, Bull. M. C. Z.,.1910, 52, p. 276), as naturally types derived by the
connection with northern South America predominate in Grenada; while in
Porto Rico, for example, the fauna is mixed as one would expect it to be. In
times past, great stress has been laid on the significance of the great deep in
the Virgin group which separates these islands from St. Croix by 2,000 fms.
Nevertheless, the same rocks reappear on both sides of the cleft, and the distri-
bution of the fauna shows that this cleft has not always existed. Such a deep of
limited area between two land masses lying unaffected by it and yet so near at
hand, has, by the fact of its depth alone, but little significance. The phenome-
non of islands with closely related faunae being separated by a small area of
very deep water is met with frequently in all parts of the world. In the East
Indies the Ké Islands are connected on the one hand with Ceram by a series of
shallows, reefs, and islands; while, so far as is known, only deep water exists
between the Ke and Aru groups and Papua. The fauna in Ké shows many
Papuan types as well as types derived from Ceram. Such deeps as these are in
no wise to be confounded with extensive clefts like that between Jamaica and
Cuba, or between Borneo and Celebes. These are very real zodlogic boundaries
which have existed since the dawn of land life on the earth. Their importance
is pointed out especially by the fact that, although they are narrow, yet no inter-
change of fauna whatever has taken place across them by “flotsam and jetsam”
methods of dispersal. Yet animals should float across these straits as easily as
across great stretches of oceans; and this, some writers persist in believing, con-
stantly occurs. Eduard Suess in 1893 contributed a very enlightening article
to Natural science (1893, 2, p. 180-187) having the title Are great ocean depths

permanent?

GRENADA. 235

In 1910 when I published my Notes on the herpetology of Jamaica, I quoted
Dr. R. T. Hill as believing that the West Indies ‘“‘have persisted without Conti-
nental connection or union with each other since their origin.’”’ He wrote this
after the publication of his Notes on the geology of the island of Cuba (Bull.
M.C. Z., 1895, 16). I find however by re-studying his Geology of Jamaica
that after his more prolonged examination of the geology of Jamaica he changed
his opinion very radically, for here (Bull. M. C. Z., 1899, 34, p. 224) he says: —

“Tn late Oligocene or Miocene time there was a tremendous orogenic movement which
resulted in uplift, whereby many of the islands were connected with each other, and possibly
an insular southern portion of Florida, but not establishing land connection with the North
and South American continents. .... In Miocene or early Pliocene time the islands were
severed by submergence into their present outlines and membership, which they have since
retained with only secondary modification.”

It will be seen that these views agree more closely with the conclusions
which are forced by an interpretation of the facts presented by the distribution
of existing animals. The Lesser Antilles have probably been subject to more
violent disturbances than the Greater. Nevertheless the existing species show
that nearly all of the islands if not every one of them, has never been completely
submerged since their original breaking away from the land which now forms
northeastern South America, or more probably that part of it which is repre-
sented in remnants by the great flat-topped mountains such as Roraima in
British Guiana. Grenada and possibly some of the other more southerly Lesser
Antilles have probably been in temporary union with the mainland once or
twice since the first connection was severed.

Regarding the origin and affinity of the Grenadian fauna, then, we may
summarize by saying that in its salient features it is a typical Antillean island
fauna. It shows closer affinity with South America than the other islands to
the northward. It has evidently been connected with South America after it
became separated from St. Vincent and the other islands north cf it; and this
separation may have been caused by subsidence, accompanied by a slow folding
which began north and progressed to the southward, separating the islands suc-
cessively from one another. This is suggested by the progressive diminution of
typical South American forms from south to north. The geological structure
of many of the islands of the inner Antillean are is so masked by the products
of extensive and recent volcanic activity that the zodlogist can look for but little
assistance from the geological investigation of these islands. Their past history
will only be definitely known when the land fauna of each island is completely

investigated.

236 BARBOUR: ZOOGEOGRAPHY.

The introduction of the mongoose has, in Grenada as elsewhere, resulted
in marked changes affecting the relative abundance of many animals, and none
have suffered more than the reptiles. In writing of several species, I have in-
cluded notes furnished me by Dr. Allen regarding mongoose ravages. I have
previously had something to say regarding the conditions in Jamaica, and there
as in Grenada, the same genera seem to have suffered most from the introduction
of this pest (Barbour, Bull. M. C. Z., 1910, 52, p. 270-301).

CONCLUSIONS.

The fauna of the West Indian region as a whole is homogeneous; that is to
say, certain genera, as for instance Ameiva, probably derived from northeastern
South America, have penetrated through the chain with a local species upon
practically every island, and reached as far as Cuba and Jamaica. Naturally,
the types derived from South America are most abundant in the islands which
were last in connection with South America, as in Grenada; and disappear
gradually, dropping out one by one, until but comparatively few reach the
Greater Antilles. Conversely, types of Central American origin predominate
in Cuba and Jamaica; Central American types, derived through both of these
islands, occur on Haiti and Porto Rico. Other genera which may be called
purely Antillean are found throughout the entire group of islands; and the most
characteristic of these are Alsophis and Leimadophis, whose distributions parallel
one another in a most remarkable manner, as does also the distribution of Eleu-
therodactylus. There is no use in expanding upon the limits of extent of the
other genera, as these may easily be seen by referring to the tables. The point,
however, which is most worthy of emphasis is that this homogeneous fauna is
characteristic not only of reptiles and amphibians, but of birds and mammals,
as well as of groups of terrestrial invertebrates, Peripatus affording an excellent
example. When this fact is taken into consideration with the regularity with
which the important genera occur on practically every island that has been scien-
tifically explored, with a species peculiar to each, it becomes evident at once that
these groups of individuals on each island have been isolated for a long time, and
that evolution by isolation would be impossible if individuals were being carried
from one island to another by flotsam and jetsam dispersal. The inherent
tendency of any isolated group of individuals to vary in some definite direction

and to fix peculiar characters within the group would be impossible if a constant

CONCLUSIONS. 237

exchange of individuals from one island to another were taking place. If this
has taken place in the past, it should be occurring still; and the fact that it does
not occur now is good proof that this method of dispersal has never played a
part of any importance in the past. A very few species have been carried about
fortuitously,— whether by human agency or otherwise, it is impossible to know,
— but this has been the exception and not the rule. The West Indian fauna is
obviously one which has descended from species which formerly ranged over great
areas of land which, in connection with each other at various times, as well as
with the mainland of Central and South America, have gone to make up the area
we may call Antillea.

ANNOTATED LIST OF THE SPECIES.

AMPHIBIA.

Spelerpes: infuscatus PrETeErs.

Peters, Monatsb. Akad. wiss. Berl., 1879, p. 778. BouLEenceER, Cat. Batr. Grad. Brit. mus., 1882, p. 72.

In speaking of the type, Peters says: ‘‘Auch von dieser Art besitzt das
Museum nur ein einziges Exemplar (No. 6556), durch das Museum Godeffroy aus
Hayti erhalten.”

No discovery has ever been made in the West Indian region that surpassed
in zodgeographic importance the finding of a salamander upon Haiti. Taken
in connection with the distribution of Spelerpes fuscus (Bonaparte) in Italy and
southern France, it is one of the suggestions that there may have been an ancient
connection between southern Europe and Antillea. Mr. Mann who collected
in Haiti (November, 1912—February, 1913) did not succeed in finding this species,
although he searched especially for it.

Hyla septentrionalis BouLeNncErR.

- BouLEencER, Cat. Batr. Sal. Brit. mus., 1882, p. 368.

Stejneger (The Bahama Islands, 1905, p. 330) states the reasons for crediting
this species ‘to Boulenger. It is found commonly upon Cuba and among the
northern Bahama and Cayman Islands. I have examined the type of Hyla
wrightit Cope in the U. 8. national museum, and can state positively that it is a
synonym of this species. The type of Hyla insulsa Cope has been lost. From
my variable series collected near Cienfuegos, Madruga, Matanzas, Havana,
Herradura, San Diego de los Bafios, Pinar del Rio and Guantanamo, I have no
doubt but that this species is a synonym also.

Hyla vasta Corr.

Cork, Proc. Acad. nat. sci. Phila., 1871, p. 219. Bouncer, Cat. Batr. Sal. Brit. mus., 1882, p. 351.

This enormous species, one of the very largest in the genus, was among those
found by Gabb while he was in charge of the Geological survey of San Domingo.
It is apparently very rare, since its great size would render it noticeable if it were
at all common, to say nothing of its probably sonorous voice, which must be far-

sounding. Cope remarks that in structure it appears to resemble Hyla lichenosa

AMPHIBIA. 239

Giinther of Mexico. Boulenger, however, says that this species is a composite
one which may be resolved into H. venulosa (Laurenti) and H. nigropunctata
Boulenger; although just why Giinther’s name is not available for this last
species is not quite evident (Cat. Batr. Sal. Brit. mus., 1882, p. 364, 366).

Hyla dominicensis (Tscuupt).

Tscuupt, Class. Batr., 1838, p. 30. Bounmnesr, Cat. Batr. Sal. Brit. mus., 1882, p. 370 (part).

In my Notes on the herpetology of Jamaica (Bull. M. C. Z., 1910, 52, p. 289)
I showed how this species differed from Hyla brunnea Gosse of Jamaica, and also
remarked that a comparison of a large series from Haiti and San Domingo with
the type of Cope’s H. ovata (M. C. Z., No. 1,518) showed them to be one and
the same species. ‘The species is confined to Haiti, whence the Museum has

many specimens from all parts of the island.

Hyla brunnea Gosser.

Goss, Naturalist’s sojourn in Jamaica, 1851, p. 361. Barsour, Bull. M.C. Z., 1910, 52, p. 288.

Although a number of species of the genus Hyla have at one time or another
been either described or recorded from Jamaica, there can be no doubt but that
this form and Hyla lichenata (Gosse) are the only ones that really occur there.
This species is common, is confined to the island, and takes there the place of
H. dominicensis (Tschudi) in Haiti and H. septentrionalis Boulenger in Cuba.

Hyla lichenata (Gosse).

Gossg, Naturalist’s sojourn in Jamaica, 1851, p. 362, pl.7. Barsour, Bull. M.C. Z., 1910, 52, p. 289.

A curious and apparently very rare species, confined to Jamaica.

Hyla pulchrilineata Corr.

Corr, Proc. Amer. philos. soc., 1869, 11, p. 163. Bounrnasr, Cat. Batr. Sal. Brit. mus., 1882, p. 378.

The beautiful little tree-frog first discovered by Gabb in San Domingo is
represented in the collection by a fine specimen taken at Puerto Plata, San
Domingo, by Mr. M. Abbott Frazar. The species is confined to the island, and
is the only West Indian Hyla that is not referable to the section Trachycephalus.
Its origin and relationships provide a fruitful field for speculation, and many be
said to provide a parallel to the very unexpected discovery of Spelerpes in Haiti,
since it is a not very distant relative of the widespread Huropean Hyla arborea
(Linné).

240 BARBOUR: ZOOGEOGRAPHY.

Bufo longinasus STEJNEGER.

STEINEGER, Proc. U.S. nat. mus., 1905, 28, p. 765.

The type alone is known (U.S. N. M., No. 27, 419) of this remarkable little
toad, quite unrelated to the other Antillean members of the genus. It was cap-
tured at dusk on a rock in a mountain stream near El Guamda, in the Province
of Pinar del Rio, western Cuba, by Messrs. William Palmer and J. H. Riley,
during their extensive collecting trip of 1900. Naturally nothing is known of
its habits, or distribution. The fully webbed hind feet indicate that its habits
are probably more strictly aquatic than the others of the genus in Cuba. Thanks
to a note of Mr. William Palmer I was able to collect at about the exact type
locality, but unfortunately I was unable to find a single specimen.

Bufo ramsdeni, sp. nov.

Type: — No. 3,213, M. C. Z., one specimen, Los Hondones, Monte Libano,
Guantanamo, Cuba. June, 1913, C. T. Ramsden, collector.

This beautiful species requires comparison with Bufo longinasus Stejneger
only. It differs in having the toes but little webbed, the fingers longer, in having
more numerous dorsal tubercles and a very strikingly different coloration.

Head and body depressed; snout concave along median line, projecting
and rather pointed; canthus rostralis rounded; interorbital space much wider
than upper eyelid; tympanum scarcely visible; first finger shorter than second;
tips of fingers and toes not dilated and hardly swollen at all; fingers slender,
longer than in B. longinasus; toes slightly webbed; subarticular tubercles large,
single; both metatarsal tubercles well developed; a tarsal fold; tibiotarsal artic-
ulation of adpressed hind limb reaches midway between insertion of fore limb
and tympanic region; skin itself smooth, with small wrinkles and many tubercles
arranged in more or less regular longitudinal series; paratoids greatly developed
as in Bufo longinasus. Color (in alcohol) above almost black, a conspicuous
light stripe from snout to vent, this stripe is extended laterally on the eyelids
and thus widens to a rhomb-like figure on the head. The light stripe is bordered
on each side by a narrow band of deep velvety black. Below, the throat is
mahogany-brown, the fore part of chest and angles of the jaws deep brown; the
belly is white profusely spotted and streaked with dark brown, as are also the
lips and upper surfaces of the limbs.

The type and only known specimen of Bufo longinasus is a male; B. rams-

deni while almost the same size is a female with the sexual organs decidedly

AMPHIBIA. 241

immature. It is, however, hardly conceivable that the marked differences which
they exhibit can be explained on the basis of sexual dimorphism. If this should
turn out to be the case it would be unique in the genus. I prefer to believe that
there are two related species representing this remarkable section of the genus,
probably as in so many other species, one confined to the western region and the
other to the eastern. The differences having arisen in these cases probably not
so much through isolation as through the extremely different environmental
conditions.

I have named this remarkable species for its discoverer, my hospitable
friend Mr. C. T. Ramsden, of San Carlos, Guantanamo.

The stomach of this specimen was filled with partially digested ants and
small beetles. The locality whence it came is at a considerable altitude above

sea level. The type of B. longinasus came from a lowland stream.

Bufo peltacephalus Tscuupt.

Tscuupr, Class. Batr., 1838, p. 52. Bounencsr, Cat. Batr. Sal. Brit. mus., 1882, p. 325.

Boulenger cites p. 89 of Tschudi’s work as the original description. This
refers to a nomen nudum only; the short description, which is, nevertheless,
perfectly recognizable, is with many others intercalated in the general discussion
of the Bufonidae.

This is the most widely distributed of the four species of toads known from
Cuba. The Museum has specimens from Pinar del Rio, San Diego de los Bafios,
Havana, Guantanamo, and Santiago de Cuba. All four of the species are con-
fined to Cuba.

The theory that the West Indian toads are a disappearing group is not
entirely tenable. This species and Bufo empusus exist in large numbers, although
their appearance is extremely erratic, almost comparable to that of Scaphiopus
in the eastern United States. Once their mode of hiding becomes known, they
may be obtained easily. About Pinar del Rio this species excavates a deep bur-
row under a loose rock, and by turning rocks they may often be found. In
other places, they live about houses, especially in the mouths of the rain-water
drains. They are often seen moving about at night, their sallies being apparently
far more frequent than those of Bufo empusus.

Bufo gutturosus LATREILLE.

LATREILLE, Hist. nat. Rept., 1802, 2, p.135. Bounenaer, Cat. Batr. Sal. Brit. mus., 1882, p. 324.

Upon comparing an adult toad from Haiti with a Bufo lemur Cope, the

249 BARBOUR: ZOOGEOGRAPHY.

characters that Stejneger (Rept. U. 8. nat. mus. for 1912, 1904, p. 570) pointed
out as likely to be diagnostic are found, just as he suspected. Stejneger had no
Haitian specimen; but his conclusion, based on the evidence of previous descrip-
tions and figures, is confirmed by the material available here.

It is said to be common locally but to be very erratic in its appearance. Mr.
Mann collected specimens at Cape Haitien, Diquini, Momance, Manneville,
and Grand Riviere. I have seen no examples from the San Domingo end of the
island.

This species, then is confined to Haiti, being replaced on Porto Rico by

its ally Bufo lemur Cope.

Bufo lemur Corr.

Corr, Proc. Acad. nat. sci. Phila., 1868, p. 311. Sresnecur, Rept. U.S. nat. mus. for 1902, 1904, p.
570, fig. 1-5.

As Stejneger surmised, this toad is quite distinct from the Bufo of Haiti,
and hence is confined to Porto Rico, where it is very rare. Stejneger heard
reports of its occurrence on Vieques Island, but was unable to obtain any speci-
mens. August Stahl, the venerable Porto Rican naturalist, told Stejneger that
he himself had never once found the ‘‘sapo concha,” as it is called by the Spanish-
speaking natives. In forty years only six specimens had come into his posses-
sion, brought in by peasants who found them by chance. Of these, two are now
in the collection of the Museum (M. C. Z., No. 2,180).

I can not do better than to quote verbatim Stejneger’s account of how he
finally succeeded in procuring specimens near Arecibo :—

“We made diligent inquiries in the neighborhood through our young friend Mr. Enrique
L. Brascoechea, and finally, as we had almost given up hope, he found an old beggar who
remembered a place where as a boy he had seen the sapos. He was promised a substantial

reward, and on the last evening of our stay in Arecibo brought in five fine specimens captured
about 3 miles south of the town. He said he had found them in holes under the roots of

”

palm trees at the border of a fresh-water pond, where they kept in hiding during the day.

Bufo empusus (Corr).

Cork, Proc. Acad. nat. sci. Phila., 1862, p. 8344. Gunp.acu, Erpet. Cubana, 1880, p.87. BouLENGER,
Cat. Batr. Sal. Brit. mus., 1882, p. 326.

Gundlach found this species, so rare in collections, rather common locally
in the Cienago de Zapata, in the jurisdiction of Colon, in the savannas near
Los Ciegos, and Los Palacios in the Vuelta Abajo region of the Province of Pinar
del Rio. I was able to obtain a splendid series in some plains about five miles
northeast of Herradura, also in the Province of Pinar del Rio. I should

AMPHIBIA. 243

never have learned of their existence had it not been for the aid given me by an
intelligent young Cuban, Sr. Lucio Alfonso. In response to my inquiry regarding
the occurrence of the ‘Sapo de Concha,” he took me to an upland clayey pasture
near by, which had recently been burned over. Scattered thickly over this
whole area were the mouths of a great number of small burrows. These were
evidently of two sorts, for the openings of some were carefully rimmed with
smoothly patted clay, while the others were rough and looked unfinished. Those
with the rims each contained a Bufo empusus. The tube-like burrow was per-
feetly cylindrical and from seven to ten inches deep. The toad which always
looked larger than the diameter of the burrow, was to be found in a small chamber
at the bottom, its curious hornlike, shelly, head forming an operculum which
closed the burrow accurately. To accomplish this closing most advantageously,
the toad was sometimes found resting on its side or back. The small individuals
were frequently observed near the surface, their little heads just filling the
mouth of the tubes. Only one or two adults were observed near the surface.
The other burrows spoken of were inhabited by Tarantulas, which were
excessively common.

Gundlach says, and he is confirmed by what the country folk told me, that
these toads come forth to sing in unison on warm nights after a rain. There is
no apparent rule to guide their appearance and they are not heard during many
apparently favorable nights. The people told me that they had never met them
outside their burrows, probably because they had not visited this particular
field by night. From what I have heard, I am sure this species always occurs

in these colonies in areas where the soil is suitable for making their burrows.

Eleutherodactylus ricordii (Dumfrit ET BrBron).

Dumérit et Brpron, Erpét. gén., 1841, 8, p.623. Boureneanr, Cat. Batr. Sal. Brit. mus., 1882, p. 213.

A common species throughout Cuba and the northern Bahama Islands.
It has been taken in Florida, and seems to be spreading northward (Barbour,
Proe. Biol. soe. Wash., 1910, 23, p. 100).

In 1887 Cope admitted that his Hylodes planirostris was synonymous
with this species (Proc. U. 8. nat. mus., 1887, 11, p. 406). There seems, how-
ever, to be a tendency for Cuban examples to have a shorter fourth toe than
Bahaman individuals. This difference does not appear to be an invariably
constant character.

244 BARBOUR: ZOOGEOGRAPHY.

Eleutherodactylus dimidiatus (Cops).

Corr, Proce. Acad. nat. sci. Phila., 1862, p. 151.

This distinct species from eastern Cuba was overlooked by Boulenger when
he was preparing the Catalogue of Batrachia Salientia in the British museum,
1882. There are several specimens in the U. 8S. national museum, which I have
examined. While in 1913 I found it at a point near Pozo Prieto, Barrio of Los
Negros, Jurisdiction of Jiguani and Mr. V. J. Rodriguez secured it at the Pan
de Azucar also in the Sierra Maestra.

Eleutherodactylus cuneatus (Cope).

Cope, Proc. Acad. nat. sci. Phila., 1862, p. 152. _BouLmenamr, Cat. Batr. Sal. Brit. mus., 1882, p. 209.

This little frog was among those sent to the Museum by Major Wirt Robin-
son, U.S. A., from Santiago, Cuba. The species was first described from eastern
Cuba. In February, 1912, I found a single typical example near Cienfuegos
The U.S. national museum has specimens from several localities in Pinar del

Rio, collected by Palmer and Riley.

Eleutherodactylus plicatus, sp. nov.

Type: — No. 3,056, M. C. Z., an adult, La Union, Monte Libano, Guan-
tanamo, eastern Cuba, March 23, 1913, C. T. Ramsden, collector.

The type of this peculiar species, the sixth of its genus to be observed in
Cuba was sent to the Museum with specimens of H. dimidiatus and E. ricordwi
from the same locality.

In its habit it reminds one of #. cuneatus but it differs strikingly in the shape
of the head, which is much more rounded in front, in the remarkable dorsal
rugosity, in the position of internal nares and vomerine teeth and in the absence
of the characteristic black inguinal markings.

Tongue medium, oral, entire behind; vomerine teeth in two long series
which converge posteriorly and almost meet at the median line, extending
laterally but very little beyond the choanae which are situated nearer the median
line than in the other species; head broad and rounded; snout only slightly
declivitous; nostrils much nearer tip of snout than eye; upper eyelids slightly
less wide than interorbital space; tympanum medium sized and round about
one half the diameter of the eye; its distance from the latter equal to its own
diameter; fingers with exceedingly small discs, the first and second fingers are
equal in length; first toe much shorter than second; subarticular tubercles well

AMPHIBIA. 245

developed; no marked plantar tubercles; a rather small inner and outer meta-
tarsal tubercle; a feebly developed tarsal fold, which may be due to shrinkage;
hind limbs being bent forward along body the heel reaches beyond the tip of
the snout, bent vertically to the axis of the body they overlap considerably;
skin above extremely rugose and plicate; throat, chest, and anterior aspect of
thighs perfectly smooth, belly and posterior femoral aspects finely granular.
Color of freshly preserved example. Uniform olive-brown above, below all
surfaces a beautiful roseate pink, front of thighs more brilliant pink, throat with a
few dark dots.
I am very loath to add a species to this multitudinous genus on the basis
of a single specimen. Nevertheless this specimen is so peculiar in many ways
that I am unable to include it in any known species. At first I thought that the
dorsal rugosity was an artifact of preservation, as the other specimens taken at
the same time and place, show no signs of this peculiar condition. It is a very
slender and elegant form, beautifully colored and probably, localized*in eastern
Cuba, as, for instance H. unicolor Stejneger is confined to El Yunque in Porto
Rico.

Eleutherodactylus varians (GunpLacu & PETERS).

GunpuacH «& Peters, Monatsb. akad. wiss. Berlin, 1864, p. 390. BouLencur, Cat. Batr. Sal. Brit.
mus., 1882, p. 215.

A Cuban species which is next to H. ricordiz, the most abundant and widely

distributed species in central Cuba. We collected a large series about Soledad,

near Cienfuegos during February, 1912. I have never found this species in

Oriente.
Eleutherodactylus luteolus (Goss).

Gossz, Naturalist’s sojourn in Jamaica, 1851, p. 366, pl. 7. BouLenemr, Cat. Batr. Sal. Brit. mus.,
1882, p. 208. Barsour, Bull. M. C. Z., 1910, 52, p. 286, pl. 2, fig. 1.
A very interesting and abundant species, confined to Jamaica. The meta-
morphosis from egg to adult is direct, the frothy egg-mass being laid in a de-
pression in moist earth; the period of incubation is very short.

‘

Eleutherodactylus jamaicensis BarBour.
Barsour, Bull. M. C. Z., 1910, 52, p. 287, pl. 2, fig. 2.
A rather rare species, which I had supposed was confined to the highlands

about Mandeville, Jamaica. Recently, however, Mr. C. T. Brues, has brought
back a beautifully preserved specimen from Newton, Jamaica.

246 BARBOUR: ZOOGEOGRAPHY.

Eleutherodactylus weinlandi, sp. nov.

Type:— No. 2,050, M. C. Z., an adult, Puerto Plata, San Domingo, Decem-
ber, 1881, M. Abbott Frazar, collector.

The type of this distinct new species is the one that Garman recorded
from Puerto Plata as Hylodes lentus (Bull. Essex inst., 1887, 19, p. 14). It is
related to EH. richmondi Stejneger of Porto Rico, and through it to EZ. lentus of
St. Thomas. In some characters it suggests H. jamaicensis; and these species
together probably form a natural series through the Greater Antilles as ZH.
auriculatus, H. martinicensis, and EH. gohnstonei do in both the Greater and the
Lesser Islands. The migration of the first group took place after, and of the
second group before, the opening up of the deep passage through the Virgin
Islands.

Tongue rather narrow, slightly emarginate behind; vomerine teeth groups
meeting to form a continuous ~—~—-shaped series behind the choanae, there is
no space between the series and the choanae, but the series extends laterally
beyond the choanae almost to a line with the margin of the jaws; nostril much
nearer tip of snout than eye, the distance from the eye being equal to its diameter.
Tympanum slightly less than one half the diameter of the eye, its distance from
the latter about equal to its diameter; discs of fingers small; first toe short,
only reaching the subarticular tubercle of the second; subarticular tubercles
well developed; two well-developed metatarsal tubercles; no plantar tubercles;
no tarsal fold; the bent limbs being pressed along the sides, the knee and elbow
meet; hind limb being extended along the side, the heel reaches the centre of the
eye; hind limbs being placed vertically to the body-axis, they overlap; skin
above and on the flanks very finely granular; underside smooth; posterior
border of thigh areolate.

Color, spirit specimen: dorsum dark brown, a light line on each side begin-
ning at the nostril, running through the eye, and ending in a light area which
covers the sacral region. Limbs, except hinder side of thigh, and belly light
and immaculate. The sacrum and hind limbs appear to have been red in life.
Garman states that they were so, but evidently by inference also.

This species is named to commemorate the work of Dr. D. F. Weinland,
whose indefatigable collecting on this island greatly increased our knowledge
of Antillean herpetology.

In the collection made by Mr. W. M. Mann there is an immature example
from Diquini, Haiti, which seems to be referable to this species.

AMPHIBIA. 247

Eleutherodactylus richmondi STesJNEGER.

STEJNEGER, Rept. U.S. nat. mus. for 1902, 1904, p. 593, fig. 25-29.

A species related to E. lentus (Cope), but confined to the region near the
summit of El Yunque Mountain, Porto Rico.

Eleutherodactylus lentus (Corr).

Cops, Proc. Acad. nat. sci. Phila., 1862, p. 151. BounLenenr, Cat. Batr. Sal. Brit. mus., 1882, p. 207.

This species is confined to St. Thomas. The specimen from Puerto Plata,
San Domingo, recorded by Garman, (Bull. Essex inst., 1887, 19, p. 14) does
not belong here, but to H. weinlandi now described as new. Boulenger also
includes San Domingan examples under this species, but these in all proba-
bility may also be referred to EH. weinlandi Barbour.

Eleutherodactylus unicolor STEJNEGER.

STEJNEGER, Rept. U.S. nat. mus. for 1902, 1904, p. 597, fig. 35-39.

A very peculiar species, confined to El Yunque Mountain, Porto Rico.
Stejneger remarks that it has no near relation in all the Antilles.

Eleutherodactylus monensis (MEERWARTH).

MeerwanrtH, Mitth. naturh. mus. Hamburg., 1901, 18, p. 39, pl. 1, fig. 11; pl. 2, fig. 4-5. Sresnracsr,
Rept. U.S. nat. mus. for 1902, 1904, p. 595, fig. 80-34.

A species confined to Mona Island.

Eleutherodactylus auriculatus (Corr).

Cops, Proc. Acad. nat. sei. Phila., 1862, p. 152. Sresnecrr, Rept. U. 8. nat. mus. for 1902, 1904, p.
583, fig. 15-19.

Having specimens of EH. martinicensis (Tschudi) from several of the Lesser
Antilles among them St. Kitts, Martinique, Monserrat, and Guadeloupe,
I can confirm Stejneger’s statement that this Greater Antillean form well merits
separation. It is known from eastern Cuba, the type locality, San Domingo,
and Porto Rico. Mann’s single example from Grand Riviere, Haiti, appears to
have the tympanum slightly larger than is shown in Stejneger’s figure.

Eleutherodactylus antillensis (ReiInHARDT & LUTKEN).

» Reruarpt & Lirken, Vid. Meddel. nat. foren. Kjébenh., for 1862, 1863, p. 209. SrryneGcEr, Rept.
U.S. nat. mus. for 1902, 1904, p. 591, fig. 20-24.

A species which Stejneger has shown to be perfectly distinct from E. auri-
culatus, with which Boulenger had previously united it. It is definitely known

248 BARBOUR: ZOOGEOGRAPHY.

only from St. Thomas and Vieques. It may occur upon St. Croix and St. John,
but there is no definite record for these islands.

Eleutherodactylus martinicensis (Tscuup1).

Tscuup!, Class. Batr., 1838, p. 77. Boutenesr, Cat. Batr. Sal. Brit. mus., 1882, p. 214 (part).

This famous little frog, so easily and so often carried from place to place,
is now very widespread in the Antillean region. Although this species develops
without metamorphosis by skipping a free living larval stage, it was really the
Porto Rican H. auriculatus (Cope) in which this characteristic was first noticed,
although all the Lesser Antillean and many of the Greater Antillean species have
been at one time or another included under this name.

In Jamaica its occurrence in Hope Gardens is due to its introduction by
Lady Blake (Barbour, Bull. M. C. Z., 1910, 52, p. 288). The records for San
Domingo (Boulenger, loc. cit.) and Porto Rico (auct.) refer to EL. auriculatus
(Cope). Fielden says that it was introduced into Barbados from whence in
1882 there were specimens in the British museum. From Barbados it is said
to have been introduced into Grenada; but the species from Grenada is a differ-
ent species, named HL. johnstoner. This may be taken to indicate that the south-
ern Lesser Antilles support a species quite distinct from true H. martinicensis,
and which has been introduced perhaps from St. Vincent to Barbados and
Grenada. Most unfortunately no St. Vincent frogs are in the collection, and
Fielden (Zoologist, 1889, p. 298) gives no clew as to whence the Barbadian frogs ~
came. Since records from practically all the West Indies have been published
for ‘‘Hylodes martinicensis,” it is best to admit here only such as undoubtedly
belong to this species in its restricted sense. The Museum has several hundred
from Martinique, the type locality, but whence, Stejneger states, it was said to
have been brought from Guadeloupe. From this island also there is a very large
series in the collection. Others are preserved from St. Kitts, and neither Stej-
neger nor I have been able to find that they were in the least different. I can
say the same for specimens from Saba, and Monserrat. It has been recorded
by Cope from St. Eustatius and St. Martins. Reinhardt and Liitken (Vid.

_Middel. nat. foren. Kjobenh. for 1862, 1863, p. 159) record it from Barbados;
and as this is more than twenty years earlier than Fielden wrote, it does not lend
any confirmation to his statement, made in 1899, that it was introduced into
Barbados twenty years before that date, which would be 1869. The specimens
which Reinhardt and Liitken record must have been taken prior to 1862. There

AMPHIBIA. 249

are other records for small frogs in the West Indies; but without specimens it is
impossible to refer them to their correct position, and for that reason their records

are not included in this paper.

Eleutherodactylus johnstonei, sp. nov.

Plate, figs. 1, 2, 5, 7, 9.

Types:— No. 2,759, M. C. Z., two figured specimens, St. Georges, Grenada,
G. M. Allen and C. T. Brues, collectors. Besides the types there is an enormous
series of specimens, young and adults, from various localities in the coastal low-
lands of the island.

Dr. Allen’s notes present the following brief history of this species:— ‘‘ Mr.
Septimus Wells tells me that this small piping frog was introduced about 25
years ago [t. e. about 1885] from Barbados, and has since spread up country.
It is, of course, absent from the forests of the high region about Grand Etang.
I heard the last one near the side of the highway about three miles above St.
George’s.”

I am very loath to designate as types of a new species individuals from a
locality of which the species is not truly a native. Since, however, no specimens
are available from Barbados, and the species was also possibly introduced there,
there is no alternative but to use the material in hand.

This is a short-limbed, heavy-bodied species belonging to that section of the
genus having a granular belly, and strong oblique groups of vomerine teeth
posterior to the internal nares, and converging posteriorly toward the mesial
line.

Tongue rather narrow, oval, free, and unnicked behind; vomerine teeth in
two strong oblique series, arising immediately behind the posterior inner margin
of the choanae, not extending laterally beyond the latter, and strongly converging
backward, the interval between the series slightly less than their distance from
the nares; nostrils much nearer tip of snout than eyes, their distance from the
eyes about equal to or slightly larger than the diameter of one of the latter;
upper eyelid less wide than interorbital space; tympanum almost exactly half
the diameter of the eye, its distance from the eye not over two thirds its diame-
ter; fingers with well-developed discs, first very slightly shorter than second;
dises of toes smaller than those of fingers; tip of first toe not quite reaching the
base of the dise of the second; two metatarsal tubercles, moderate in size, and
with the outer larger than the inner; soles with rather abundant, feebly devel-
oped tubercles; no tarsal fold; the bent limbs being pressed along the side, knee

250 BARBOUR: ZOOGEOGRAPHY.

and bowel meet; hind limb being extended along the side, heel almost reaches
eye; hind limbs being placed vertically to the axis of the body, the heels just
overlap, very slightly; skin above densely granular; a fairly well-defined glan-
dular ridge along back from snout to vent; belly and hinder aspect of thighs
strongly glandular; throat, upper portion of chest, and anterior aspect of thighs
smooth; no differentiated adhesive disc area.

The variation in color is very great, and may be gauged by the figures of the
typical specimens. Other variation is seen in the dorsal granulation, which is
always well developed, however, and in the fact that in the largest individuals
the heels barely meet posteriorly, when the limbs are laid vertically to the body.
The extreme adults, then, have very short limbs.

The species belongs near H. auriculatus (Cope) from the Greater Antilles,
and E. martinicensis (Peters) from Guadeloupe, Martinique, St. Kitts, Jamaica,
and elsewhere. The latter species was, according to Stejneger (Rept. U.S. nat.
mus. for 1902, 1904, p. 584), introduced from Guadeloupe to Martinique; and
still more recently it has been carried to Jamaica (Barbour, Bull. M. C. Z.,
1910, 52, p. 288). H. johnstonei can be separated from FE. auriculatus Cope by
its vomerine teeth groups, which are less extended laterally, by its larger
tympanum, which is nearer the eye than in the other species, by its shorter
hind limbs, and the more considerable development of tubercles on back and
soles of feet. In EH. martinicensis (Peters), the exposed tympanum is still less
in size. FH. antillensis Reinhardt and Liitken, from St. Thomas, is easily
separable by its small feet. HH. johnstonet grows to a larger size than any
specimen of H. martinicensis which I have been able to examine. The largest
Grenadian specimen is 32 mm. from snout to vent. Porto Rican examples
(E. auriculatus) reach a size of 43 mm., according to Stejneger, and there are
several of almost this size in the collection here.

In the Zoologist (1889, p. 298), Col. Fielden of Barbados makes the follow-
ing note regarding what he calls Hylodes martinicensis: —

“The little Tree-frog Hylodes martinicensis, Tschudi, is undoubtedly a very
recent introduction; twenty years ago, as I am credibly informed, it was quite
unknown. It is now spread over the entire island... .”

This is substantiated by Bellin, who wrote in 1758, in his Description geo-
graphique des Isles Antilles possedes par les Anglais (Paris, 1758, p. 61), that
upon the island of Barbados ‘‘Il n’y a ni Crapauds ni Grenouilles.”’

This species is named for His Honor Robert S. Johnstone, Chief Justice
of Grenada and the Grenadines, who very kindly aided the members of the
Museum’s expedition while at Grenada.

AMPHIBIA. 251

Eleutherodactylus urichi (BorrTcEr).
Plate, fig. 3, 4, 6, 8, 10.
BorrrerEr, Journ. Trinidad field. nat. club, 1894, 2, p. 88.

Without specimens from Trinidad, it is impossible to decide certainly
whether the specimens which Dr. Allen brought from Grenada are absolutely
the same as this species or not. They certainly agree closely with the original
description, and must be very closely related, if not identical.

-The large series are from the virgin forest about the Grand Htang, a lake
high up in the central part of the island. To this region this species is absolutely
confined. Not a single specimen was taken in the lowlands. It oceurs com-
monly under stones, rotten logs, and in decaying leaves and other fallen vegeta-

tion. Its note is ‘

‘a series of eight or nine rapid ‘criks,’ the last two or three
less rapid, and the whole not louder than a cricket.”

The two figures show how varying the color pattern may be; and the
coloring itself of the alcoholic individuals is also very various. In general it
agrees with Boettger’s description, but no two specimens are alike. About

one hundred were preserved.
Leptodactylus pentadactylus (LauRrEnt1).
LAURENTI, Syn. Rept., 1768, p. 32. BovuLmnenr, Cat. Batr. Sal. Brit. mus., 1882, p. 241.

This species, with an enormous distribution over the South American main-

land, has been recorded from but three West Indian islands,— Dominica, St.

Kitts, and St. Lucia. This distribution strongly suggests an artificial introduc-

tion as an article of food. It may, however, have been extirpated upon other
islands where it once occurred for this very reason.

As argument against their artificial introduction, we may cite Father Labat,
who, in his accurate and engrossing narrative entitled ‘“‘Nouveau voyage aux
iles d’Amerique,” &c., (La Haye, 6 vols., 12mo, 1724, and Paris, 8 vols., 12mo,
1742), informs us that ‘On trouve a la Martinique & en quelques-autres Isles,
les. plus belles Grenoiilles du monde, on les apelle Crapauds parce qu’elles sont
vétues comme les crapauds d’Europe, c’est-a-dire de gris avec des taches ou
rayes Jaunes & noires; elles ne se tiennent pas dans l’eau, mais dans les bois ott
elles croassent tres-fort, sur tout la nuit....leur chair est blanche, tendre &
délicate on ne jette que la téte.”

Upon Martinique they have perhaps been now exterminated on account
of their culinary value.

bo
On
bho

BARBOUR: ZOOGEOGRAPHY.

Leptodactylus inoptatus, sp. nov.

Type:— No. 3,087, M. C. Z., one male, Diquini, Haiti, W. M. Mann,
collector.

Vomerine teeth in two curved series behind the choanae; tongue large,
slightly indented behind; nostrils very near tip of snout; tympanum circular,
over two thirds the diameter of the eye; interorbital space over twice width of
upper eyelid; first and second fingers equal in length; toes but very slightly
webbed at base, third and fifth of equal length; subarticular tubercles very
greatly developed, soles almost smooth; two metatarsal tubercles, the inner
near the origin of a vestigial dermal fold; tarsometatarsal joint reaching nostril
when the hind limb is carried forward along the body; skin smooth on middorsal
area, slightly granular on sides, coarsely granular on belly; two dorsolateral
glandular folds; a pronounced fold from eye over tympanum to shoulder;
ventral dise feebly marked; figures and toes with slightly expanded tips; fingers
and toes with well-developed dermal emarginal fingers; a spine-like tubercle
very prominent on posterior region of each eyelid.

Color, middorsal area lilac or purplish brown; sides heavily marbled with
dusky olive; lower surfaces dirty white. A white spot near the shoulders on
each side. No white stripe on lips.

In speaking of L. albilabris from Porto Rico, Stejneger says (Rept. U. S.
nat. mus. for 1902, 1904, p. 561) :—

“The status of the frog, Leptodactylus albilabris, is quite different from that of the other
species of the fauna. In the Antilles it is restricted to the Virgin Islands, St. Croix, Vieques
and Porto Rico. There is apparently no indigenous species of the genus in the Caribbean
chain, the other two which occur in various islands being evidently introduced by man and
belonging to widely different South American species. No Leptodactylus or related form is
found in Haiti, Cuba, or Jamaica. But the most curious feature is that a frog which neither
Dr. Boulenger nor I can distinguish from L. albilabris, is a native of southern Mexico, State
of Vera Cruz, and the Isthmus of Tehuantepec. I know of absolutely no parallel to this
extraordinary range, which is inexplicable on ordinary distributional grounds, for certainly
it would transgress the boundaries of the probable to suppose either that this species had once
covered the whole country between Tehuantepec and the Virgin Islands and become extinct
in the intermediate territory, or that there had at any time been a direct connection between
the localities mentioned to the exclusion of the large Antilles. Nor can it for a moment be
supposed that the species exists in the latter without having attracted attention. No doubt
there are many species yet to be discovered in these islands, but L. albilabris is not likely to
be one of them, for it is one of the commonest, most obtrusive, and most easily caught ba-
trachians wherever it occurs. In suggesting accidental introduction by man I am fully aware
that this explanation does not at first appear plausible, as there does not seem to be or to
have been any direct route of communication between southern Mexico or Yucatan and

AMPHIBIA. } 253

Porto Rico or the Virgin Islands, but I offer it as the only possibility I can think of. The
wrecking of a vessel with a cargo of logwood or mahogany a hundred years or more ago might
account for this remarkable distribution.”

Now considering that L. albilabris occurs upon more than one island, it is
necessary to postulate a series of unlikely events to explain its accidental occur-
rence. It is more probable that the Porto Rican and mainland individuals have
by some fortuitous parallelism become indistinguishable if they originally be-
longed to different stocks or if both were derived from the same stock then they
have preserved their identity in parts of their range and varied with isolation in
other parts. The occurrence in the East Indies of Rana macrodon in the greater
Sunda Islands and New Guinea, and of Rana modesta, a derivative of R. macrodon
in Celebes is a case of similar distribution. The Lesser Antillean species are
almost certainly not introduced, since Labat informs us that frogs (L. penta-
dactylus) were a common article of food in Martinique and other islands as early
as 1724. That they probably belonged to this species is shown by his stating
that they lived in the woods and not in the water and by his description of their
colors. The occurrence of a peculiar Leptodactylus upon Haiti, and there is
evidence that a species occurs upon Cuba also, is not as astonishing as appears
at first.

It is quite possible that other species remain undiscovered, while in the
Lesser Antilles at least it is entirely probable that they may have been extermi-
nated upon some islands because of their edible qualities. That the genus was
once generally distributed throughout the area I have no doubt whatever.

In February, 1912, while collecting with Mr. R. M. Grey upon his CoLonia
of the Soledad estate near Cienfuegos, Cuba, we both saw a large frog leap a
number of times over a grass plat and take refuge in a small pond. Mr. Grey
said it was a “‘bull frog” and that thereabouts such frogs were rare and shy. He
has been trying, so far without success, to secure one, and I also searched for it
in vain. I recognized at once that it was far larger than any Eleutherodactylus
which I had ever seen and at the time concluded that it was a large Hyla which
was perhaps approaching the water for breeding purposes. In the light of the
finding of this species upon Haiti, I have but little doubt that it was an un-
known Leptodactylus.

Leptodactylus validus Garman.
GarMAN, Bull. Essex inst., 1887, 19, p. 14.

This species was described from St. Vincent (Types M. C. Z., No. 2,183).

Later, however, Boulenger called specimens from the same island L. caliginosus

254 BARBOUR: ZOOGEOGRAPHY.

Girard. The latter species was described (Proc. Acad. nat. sci. Phila., 1853, p.
422) from specimens taken at Rio de Janeiro by the members of the U. S. explor-
ing expedition. Boulenger (Cat. Batr. Sal. Brit. mus., 1882, p. 247) makes L.
caliginosus embrace Hallowell’s species, malanonotus, from Nicaragua, which is
certainly a valid form, as specimens in the Museum from Nicaragua show.
They differ markedly from the figure and description of Brochi’s L. echinatus
(Miss. sci. Mex. Batr., 1882, p. 18, pl. 5, fig. 4), and also from a specimen from
Merida, Venezuela, obtained through Rosenberg of London. This frog was from
material collected by Sr. Bricefio y Gabaldon, which was studied by Boulenger;
so that it represents exactly Boulenger’s L. caliginosus. Upon comparing this
specimen with a large series from St. Vincent and Grenada, I find that they are
the same in every respect. The only obstacle which now remains in the way of
reducing L. validus to a synonym is the fact that we have no topotypes of L.
caliginosus for comparison. I did not find this species when collecting about
Rio, but only L. poecilochilus Cope. A species of amphibian covering a range
which extended from Rio de Janeiro to St. Vincent in the West Indies and to the
Isthmus of Tehuantepec would be foreign to everything which we know of
amphibian distributions. For this reason I keep the species separate, believing
that a comparison with examples from Brazil will show this to be justified. L.
caliginosus Girard will thus be the name for Brazilian examples, L. validus
Garman for those from St. Vincent, Grenada, probably Trinidad, and certainly
part of Venezuela; and L. melanonotus (Hallowell) for those from Central
America, probably from Tehuantepec to Panama. I showed (Barbour, Bull.
M.C. Z., 1906, 46, p. 228) that some very large specimens from San Miguel and
Saboga Islands, in the Bay of Panama, were a distinct species, and gave them
the name L. insularwm.

From Dr. Allen’s field notes I glean that there is no evidence that L. validus
has been introduced in Grenada, and the fact that it is a species characteristic
of the fauna of the small area of primeval forest still remaining in the interior of
the island would make it appear certainly to be indigenous. Dr. Allen says that
what he took to be this frog pipes like our common Hyla pickeringii. Specimens
seen near the shore of the Grand Etang when disturbed, dove into the water,
and hid among the dead leaves on the bottom. They occurred, however, in the
sodden woods as well.

AMPHIBIA. 255

Leptodactylus albilabris (GiinrueEr).
Ginrtuer, Ann. mag. nat. hist., 1859, ser. 3,4, p. 217. Sresnucnr, Rept. U.S. nat. mus. for 1902, 1904,
p- 561, fig. 6-10.

The history and status of this species, which has been considered one fortui-
tously introduced is discussed among the remarks upon L. inoptatus, sp. nov.
from Haiti.

Phyllobates limbatus Corr.

Core, Proc. Acad. nat. sci. Phila., 1862, p. 154. BouLencmr, Cat. Batr. Sal. Brit. mus., 1882, p. 195.

The types of this species were six specimens now in a very bad state of
preservation which were sent to the U. 8. national museum from the region of
Guantanamo, Cuba, by Mr. Charles Wright, a very accomplished botanical and
zodlogical collector. I can not learn that the species has been taken since until
1912. Gundlach in all his trips over the island never collected it. I found it
at two neighbouring stations. One a seborucal near Colonia Limones, another
in an area of dense woods near Colonia Guabairo, both on the Soledad planta-
tion near Cienfuegos.

The seborucal, as any densely wooded area upon a region of limestone rocks
is called, was like an island in a sea of cane fields and the jungle was so dense and
thorny that it could only be entered by following up the bed of a small stream.
In company with Mr. R. M. Grey I found one of these tiny frogs under a stone
on the bank of the stream. Another was seen but not secured. Greatly dis-
appointed, for I realized at once the identity and interest of the find, we revisited
the locality several times but were unsuccessful.

Some days later while enjoying the hospitality of Captain Beal of Guabairo,
Mrs. Barbour and I were collecting along an old wood trail in the early morning
when we came to a stony place in the road, evidently a sort of stream bed flowed
over only during a heavy rain. By diligent searching under these stones we
obtained a good series all of exactly the same minute size. The species is an
extremely active and delicate one and to obtain uninjured specimens means
quick and accurate work with the forceps.

The colors in life are as Cope has described them and he was certainly
correct in adding that “This is perhaps the smallest frog known, and is a very
prettily marked one.”

Believing it probable that this species might in reality not be a true Phyllo-
bates at all I cleared an entire specimen with chemical reagents and was thus
able to study the sternum critically. The examination showed that the species

was in every way a typical Phyllobates.

BARBOUR: ZOOGEQGRAPHY.

bo
or
lop)

The finding of one member of this genus in Cuba is the only reason for still
retaining Phyllobates bicolor in the list, which is I think unjustifiable.

Since the foregoing was written Mr. C. T. Ramsden has found a few speci-
mens of this species at Monte Libano, near Guantanamo. These are approxi-
mate topotypes and thanks to Mr. Ramsden’s generosity I have been able to
make direct comparison of the specimens from these widely separated localities.

They are exactly similar in every respect.

REPTILIA.

Gymnodactylus fasciatus Dumérin et Bisron.

Duménrit et Brsron, Erpét. gén., 1836, 3, p. 420. BouLrnenr, Cat. lizards Brit. mus., 1885, 1, p. 31.

This the only Antillean species of the genus occurs upon Martinique. It
has only been taken a few times, and of its habits, or of its possibly wider dis-

tribution, nothing is known.

Gonatodes albogularis Dumérit et Breron.

Dumerit et Brpron, Erpét. gén., 1836, 3, p. 415. Barsour, Bull. M. C. Z., 1910, 52, p. 289.

This species, formerly supposed to be confined to Cuba, has been found in
Jamaica, where it has undoubtedly been introduced by the direct carrying trade
from Santiago de Cuba and Kingston. In Jamaica it has only been captured in
and about the latter city. In Cuba it is common only in the region about the
cities of Havana and Santiago de Cuba. I found it in houses in the latter city
in 1909, and a fine adult was given me by Sr. Victor J. Rodrignez, caught in the
buildings of the National University at Havana in February, 1912. Gundlach
found it also in these two localities only. This distribution looks at once like a
case of fortuitous introduction. It has, however, been found in the country as
well, and there is a specimen in the British museum from the San Juan River
district (Boulenger, Cat. lizards Brit. mus., 1885, 1, p. 59). The principal Rio
San Juan is near Matanzas and this year I saw what I am convinced was this
species near Matanzas. A fine adult pair caught by Mr. Charles T. Ramsden

near Guantanamo is in the Museum the gift of Mr. Ramsden.

Gonatodes notatus (Remsarpt & LUTKEN).

Reruarpt & LurKen, Vid. meddel. nat. foren. Kjébenb. for 1862, 1863, p. 280.

Although Boulenger (Cat. lizards Brit. mus., 1885, 1, p. 59) places this
species in the synonymy of G. albogularis, with a query attached, it seems wisest

REPTILIA. 257

to keep them separate until specimens from Haiti can be compared with others
from Cuba. The description seems to indicate a valid species. The type
locality is Acquin, Haiti.

Phyllodactylus spatulatus Corr.

Copr, Proc. Acad. nat. sci. Phila., 1862, p.176. Boutmncer, Cat. lizards Brit. mus., 1885, 1, p. 81.

This is apparently an excessively rare species, the types of which were
caught at Barbados by Theodore Gill, who spent some time collecting upon that
island. This case is more or less similar to that of T’arentala cubana, a species
which is probably common enough, and yet which, on account of its crepuscular
or nocturnal habits, very seldom finds its way into the hands of the general
collector, and still remains very rare in museum collections.

Recently a specimen has come to hand which was collected in Curagao.
It has been compared with the types of this species by both Dr. Stejneger and
myself, and there can be no doubt as to its identity.

Thecadactylus rapicaudus (Hourruyn).
Hovrttuyn, Verh. Zeeuwsch. genoot. wet. Vlissingen, 1782, 9, p. 323, pl. 3, fig. 1. BouLenamr, Cat.
lizards Brit. mus., 1885, 1, p. 111.

This is another species of gekko which occurs in both Central and South
America, and upon many of the West Indies,— in some cases possibly distributed
fortuitously. There are specimens in the Museum from Saba, St. Bartholomew,
St. Lucia, Dominica, Anguilla, Guadeloupe, and Grenada. Other records are
St. Thomas, St. Croix, Nevis, and Antigua.

The only other species in the genus is 7’. australis Giinther, which occurs on
the islands in Torres Straits. Such a distribution as this can only be explained
by assuming that the genus is one of extraordinary antiquity,— an assumption
that we can not prove,— or that the species are in reality unrelated, and that the
similar characters are the result of the operation of some law of convergence.
The species may really represent separate genera, which are possibly very dis-
tantly related to each other.

Three examples were secured by Dr. Allen in Grenada. The species is rare
locally, although common on many other of the Antilles. ‘‘Wood slave,” is
native name according to Dr. Allen. In Jamaica this name is sometimes used
for the Mabuya, and very often for Sphaerodactylus gilvitorques. For the latter

‘

genus in general, the name is “‘water-lizard,” or “pawley lizard.” Curiously

enough it is never used for Aristelliger, which should suggest Thecadactylus so

258 BARBOUR: ZOOGEOGRAPHY.

strongly to the native observer. The remarkable distribution of these wide-
spread native names, and their curious application to different local species,
would provide a most absorbing problem for some student of folk-lore and
linguistics. The same sort of a series of stock-names for littoral fishes appears
spread from Bermuda and Florida to the Lesser Antilles, and might well be in-
cluded in the same study.

Aristelliger praesignis (HALLOWELL).

Ha uowE Lt, Proc. Acad. nat. sci. Phila., 1856, p. 222. Boutenesr, Cat. lizards Brit. mus., 1885, 1,
p. 146. Barsour, Bull. M. C. Z., 1910, 52, p. 289.

The Museum’s specimens are from Jamaica, Grand Cayman, and Cayman
Brac. This species needs comparison with A. irregularis Cope (Proc. Amer.
philos. soc., 1885, 22, p. 387) from Cozumel Island, Yucatan, which we have
not seen. The question as to whether the various mainland records for Belise
and Nicaragua really refer to A. praesignis is one of prime importance. Do they
represent the mainland inhabiting ancestors from which the Antillean individuals
have been derived, or are they waifs which have been carried to the mainland
from Jamaica by the considerable schooner traffic? If the former possibility is
true, then we might expect A. irregularis Cope to stand valid. It is probably
better to assume that it does (Cat. lizards Brit. mus., 1887, 3, p. 485). If not,
Boulenger’s contention that it is identical is doubtless correct.

Aristelliger lar Corr.

Core, Proc. Acad. nat. sci. Phila., 1861, p. 497. BouLEencEr, Cat. lizards Brit. mus., 1885, 1, p. 147.

The type of this species (M. C. Z., No. 3,607) is a very large individual, over
a foot in total length. It was captured by Weinland near Jeremie, Haiti, and is
the only one that the Museum has ever received.

Aristelliger nelsoni, sp. nov.

Type: —No. 7,891, M. C. Z., an adult from Swan Islands, Caribbean Sea,
March, 1912, George Nelson, collector. Seven paratypes.

This species may be distinguished from its close relative A. praesignis by its
larger average size, several specimens are 235 mm. in length and by its shorter
snout which is but slightly longer than the distance from eye to ear. Besides
these characters which are not very striking, there is a marked difference in the
dorsal scales. In the Jamaican examples these are rough, shagreen-like granules,

while in those from Swan Island the skin is much more finely granular and much

2

REPTILIA. 259

smoother. The type has but seven upper and five lower labials to below the
posterior border of the orbit. In the series of paratypes eight upper and six
lower labials are found but this is evidently not the typical condition. The
ventral scales of the preanal region are larger than those of elsewhere, which
does not seem to be the case in A. praesignis. Lamellae under fourth toe 15 in
number in the type, which seems to be a constant character.

The large size, shorter and less depressed snout, the lesser number of labials,
finer dorsal granular scales and rather greater number of subdigital lamellae
serve to distinguish this from A. praesignis Hallowell. The smaller number of
subdigital lamellae and smaller size separate it at once from A. lar Cope. Its
very prominent palpebral spine separate it from A. georgeensis (Bocourt) from
St. George Island, and its small number of labials again serve to distinguish it
from A. irregularis Cope from Cozumel Island. These four are the only species
previously described. The genus is recorded from Belize by Boulenger (Cat.
lizards Brit. mus., 1885, 1, p. 147).

The following field notes are important. Mr. Nelson states that the species
is extremely common on both Swan Islands. It is frequently heard croaking
at dusk both among the cocoanut palms, in the houses (on Swan Island), and
in the woods. On Little Swan Island the species occurs only in the jungle since
this island is and always has been uninhabited and is without any cleared land.
The island (one and one half miles by one half mile in size) rises in sheer cliffs
from the ocean on all sides. There is no introduced vegetation and it is here
that the peculiar mammal Capromys thoracatus True is found abundantly.
There is thus no reason whatever to suppose that this lizard has come to the
island fortuitously. It is ninety-eight miles from the nearest point of the Hon-
duran coast and one hundred and thirty-five miles from the Cayman group.

This species is named for Mr. George Nelson of the Museum staff, the first
to explore thoroughly these interesting islands.

Tarentola cubana GunpLAcH & PETERS.

Gunpuacu & PrTrers, Monatsb. Akad. wiss. Bérlin, 1864, p. 384. Bovu.Eencer, Cat. lizards Brit. mus.,
1, 1885, p. 195; 3, 1887, p. 490.

This rare gekko is confined to Cuba. There are specimens in the Museum
which Major Wirt Robinson, U.S. A., sent from Santiago; and the writer once
had a living specimen which came from Cienfuegos and which most unfortunately
was lost. Gundlach only found two specimens during his life in Cuba, one near
San Diego de los Bafios, in the Sierra de Rangel, and the other at Cabo Cruz.
In February, 1913, I got a fine example under a stone at this same locality.

260 BARBOUR: ZOOGEOGRAPHY.

This is the only Antillean representative, in fact the only American repre-
sentative, of a genus the other species of which inhabit southern Europe, north-
ern and western Africa, the Cape Verde Islands, and Madeira. The fact that
the Cuban species is so rare in collections would make it seem possible that the
species may exist on Haiti; it is not likely to occur on Porto Rico or Jamaica,
or on the Lesser Antilles. It is curious that its existence in the West Indies
should be confined to Cuba, although we observe too many similarities in West
Indian and West African generic ranges for the general habitat of Tarentola to

seem very peculiar.

Sphaerodactylus torrei, sp. nov.

Sphaerodactylus sputator BOULENGER, Cat. lizards Brit. mus., 1885, 1, p. 219 (non Sparrman).

Anderson (Bihang. K. Svensk. vet-akad. Handl., 1900, 26, afd. 4, no. 1, p.
27) has shown that the name Sphaerodactylus sputator (Sparrman) must be
transferred to S. fantasticus Duméril & Bibron, since he has identified the original
specimens from St. Eustatius in the Stockholm museum with that species. It
now becomes necessary to rename the Cuban species, which I have done in honor
of the distinguished Cuban naturalist, Carlos de la Torre.

This is apparently a common species in eastern Cuba. It is supposed to
occur also upon San Domingo; but no comparison from the two localities can be
made here, nor did Boulenger, in 1885 have any from Haiti.

Type: — No. 6,916, M.C.Z. An adult, Santiago de Cuba, Wirt Robinson,
collector. Two paratypes.

Snout rather pointed, longer than distance from eye to ear opening; ear
aperture vertical, ovoid, rather small. Rostral large, with median cleft above,
nostril between rostral, first labial and one to three small nasals; five upper and
five lower labials; mental large, truncate posteriorly; no regular chin shields,
but many polygonal scales, which pass into the throat granules. Upper eyelid
with small spine-like scale above the middle of the eye. Upper surfaces with
very small minute granule-like scales, largest on the snout; scales of abdomen
imbricate and larger. Tail cylindrical, with verticals of scales and a row of
enlarged scutes below. Dirty brown or whitish, upper surfaces with five dark
brown bands between the end of the snout and sacrum, the three anteriors often
extending almost or quite across neck and chest. A varying number of complete
bars on tail. Two longitudinal whitish stripes on the snout.

Upon my collecting trip in 1913 I got two fine large specimens from Cabo
Cruz and Mr. C. T. Ramsden gave me a beautifully marked individual from

Guantanamo.

REPTILIA. 261

Sphaerodactylus elegans REINHARDT & LUTKEN.
ReinHArRpDT & Livrken, Vid. Meddel. nat. foren Kjobenh. for 1862, 1863, p. 275. BouLmNncmrR, Cat.
lizards Brit. mus., 1885, 1, p. 220.

A very small species represented by individuals from San Diego de los Bafios
Caibarien, Remedios, and Havana, Cuba. These little lizards look very like
what might be the young of S. torret, of which species we have no specimens of a
similar size. The measurements given by Boulenger for the largest S. elegans are
much smaller than those which he gives for S. torre:. The type of coloration is,
however, entirely different, and adults of but little more than an inch long were
not uncommon at San Diego de los Bafios. This was the only locality where we
ourselves met the species, although we often heard elsewhere of the ‘‘Salaman-
quita de la Virgen,” as it is called by the people. Recently four examples from
Momance, Haiti, have been received from Mr. W. M. Mann.

Spharodactylus cinereus CocTEAUu.

CoctTrau Sagra’s Hist. Cuba. Rept., 1834, p. 166, pl. 17. Gray, Cat. lizards Brit. mus., 1845, p. 168.
BouLENGER, Cat. lizards, Brit. mus., 1885, 1, p. 220.

The only specimens which I have seen and which agree with this form are
from Cienfuegos, Santiago de Cuba and Los Negros in Jiguani, Oriente. Boulen-
ger records it from San Domingo, Martinique, and Caracas, Venezuela. The
Martinique specimen, which was received from the Paris museum, probably
came from San Domingo, as did many other specimens in the Paris museum
which are labeled Martinique. The specimen from Caracas was probably either
carried there by mistake, or else belongs to another species. S. cinereus is proba-
bly confined to Cuba and Haiti. Mr. Mann collected an adult at Momance,
Haiti.

Sphaerodactylus nigropunctatus Gray.

Gray, Cat. lizards, Brit. mus., 1845, p. 168. BouLenceEr, Cat. lizards Brit. mus., 1885, 1, p. 220, pl. 18,
fig. 2.

Of this species, the type locality of which was unknown, the Museum has
specimens, which Garman has identified, from Santiago de Cuba, collection of
Major Wirt Robinson. Recently Mr. C. T. Ramsden has sent a specimen from
Guantanamo which seems to be referable to this species. It is, however, almost
black in general color with two white dots side by side upon the nuchal region,

262 BARBOUR: ZOOGEOGRAPHY.

Sphaerodactylus decoratus Garman.
GarMAN, Bull. Essex inst., 1888, 20, p. 111. Barsour, Bull. M. C. Z., 1904, 46, p.57. Rostn, Lunds
univ. arsskrift, 1911, 7, no. 5, p. 27.

The type (M. C. Z., No. 6,220) isfrom Rum Cay. Itis common upon Andros
Island and rare on New Providence. Specimens from the last two localities
were collected in the Bahamas during the summer of 1904 by Dr. Glover M. Allen,
Owen Bryant, and myself.

_ This species, which in its lepidosis is so very like S. punctatissimus Gray and

S. flavicaudus Barbour, is well marked by its strikingly conspicuous coloration,
different from any other species in the section of the genus having the minute
granular dorsal scales.

The thought occurs at once that S. punctatissimus, S. flavicaudus, and S.
decoratus may in reality be merely color phases within a species which has a
lepidosis varying slightly in connection with each coloration. But this fact, and
the added one that no intergrades in coloration have ever been found make
it seem probable that they are really distinct species. Field observations based
upon the hatching of broods of young would settle this point, as we know both
young and full grown individuals of most of the forms. The coloration does not
seem to change during growth. The young of S. decoratus looks extremely like
S. elegans from Cuba; while the young of S. cinereus and S. flavicaudus are very
alike.

Sphaerodactylus flavicaudus Barsour.

Barpour Bull. M. C. Z., 1904, 46, p. 56. Rosin, Lunds univ. arsskrift, 1911 7, no. 5, p. 27.

The type series of fourteen specimens (M. C. Z., No. 6,953) came from
Mangrove Cay, Andros Island. Rosén found it common at Mastic Point on the
same island. So far as known it does not occur elsewhere.

This species is very closely related to S. nigropunctatus Gray, if our speci-
mens from Santiago really represent Gray’s species. The Andros examples may
be distinguished by the scales of the back and flanks being more nearly the same
size. The coloration is different. Major Robinson’s labels state that his speci-
mens were brown, with mustard-colored tails, while all the types of the Andros
species were pale cream color in life, with brilliant yellow tails. One of Robin-
son’s specimens shows fine dark dots thickly scattered over the dorsal surfaces.
A specimen marked in this way doubtless suggested the name nigropunctatus to
Gray. Boulenger mentions that the spots have faded, and the type is now
brown, with tail and lower surfaces lighter.

REPTILIA. 263

Sphaerodactylus alopex Corn.
Cops, Proc. Acad. nat. sci. Phila., 1861, p. 499. BouLenerr, Cat. lizards Brit. mus., 1885, 2, p. 222.
Garman, Bull. Essex inst., 1887, 19, p. 18.

The types, five specimens which Dr. D. F. Weinland took at Jeremie, Haiti,
are the only ones in the Museum (M. C. Z., No. 3,343). They are in excellent
preservation.

The species seems somewhat allied to S. oxyrrhinus Gosse, of Jamaica, but
is apparently confined to Haiti and San Domingo. The specimens of S. oxyr-
rhinus that have been recorded from Haiti probably belong to this species.

Sphaerodactylus oxyrrhinus Gosss.
Gossg, Ann. mag. nat. hist., 1850, ser. 2,6, p.347. BouLenasr, Cat. lizards Brit. mus., 1885, 1, p. 222.
Barsour, Bull. M. C. Z., 1910, 52, p. 291.
Known from but few specimens, all from Jamaica. Fischer (Bericht
Naturh. mus. Hamburg for 1887, 1888, p. 23) records this species from Haiti;
but the specimens should be re-examined. They are very likely S. alopex Cope.

Sphaerodactylus dacnicolor Barzsour.

Barzour, Bull. M. C. Z., 1910, 52, p. 292, pl. 1.

The two type specimens (M. C. Z., No. 7,276) are the only ones known
belonging to this species. They were from Port Antonio, Jamaica.

Sphaerodactylus argus Gosse.

Gossg, Ann. mag. nat. hist., 1850, ser. 2,6, p. 347. Barsour, Bull. M.C. Z., 1910, 52, p. 292.

The commonest species in Jamaica, where it occurs all over the island.
What Gundlach (Erpet. Cubana, 1880, p. 59) called this species, was undoubtedly
a peculiar phase of coloration of the young S. cinereus. A specimen of this
species which I collected near Cienfuegos, and several from Jiguani appear as
far as their general type of color pattern is concerned, to be typical S. argus.
This, however, they certainly are not.

Sphaerodactylus argivus GarRMaNn.

GarRMAN, Bull. Essex inst., 1888, 20, p. 3.

Boulenger, in the Zoological record for 1888, remarked that this was a
synonym of S. argus, of Jamaica. I have examined the types — eight specimens
(M. C. Z., No. 6,223) from Cayman Brac — and I find, upon comparing them
with a series of over one hundred Jamaican examples, that they differ so con-

264 BARBOUR: ZOOGEOGRAPHY.

stantly in color that they can not be considered the same as the Jamaican species,
which doubtless gave rise to the Cayman Brac form. Yellow, dark-rimmed
ocelli are always present in Jamaican examples, sometimes only a very few, quite
scattered, and again plentiful in numbers and arranged in regular series. In
many individuals these have a tendency to become confluent and to give rise to
lines upon head and neck. These do not occur in S. argivus, the color of which
Garman has adequately described. On the mainland these individuals would
be considered as representing geographic races or subspecies, and where their
ranges overlapped intergrades would be expected and found. But when the
individuals are confined upon islands, no intergrading on account of the geo-
graphic discontinuity of their ranges can take place. Under these conditions
it is not unreasonable to consider these very similar forms ‘‘species.’’? Werner
has described S. argus continentalis from Honduras, a fact of the greatest zoégeo-
graphical interest. Its importance is emphasized by this trinomial designation;
yet for the sake of consistency, we would suggest that the species be known as
S. continentalis Werner (Verh. Zool. bot. gesellsch. Wien., 1896, 46, p. 345).

Sphaerodactylus notatus Bairp.

Bairp, Proc. Acad. nat. sci. Phila., 1858, p. 254. Bounenarr, Cat. lizards Brit. mus., 1885, p. 226.
Barsour, Bull. M. C. Z., 1904, 46, p. 57. Sresnecur, Bahama Islands, 1905, p. 331. Rosin,
Lunds univ. arsskrift, 1911, 7, no. 5, p. 27.

There are specimens in the Museum from Key West, Florida, where the
species was doubtless accidentally introduced; Havana, Cienfuegos, San Diego de
los Bafios, Jiguani, and Guantanamo, Cuba, Little Abaco, New Providence, and
Stranger’s Cay, Bahamas. It has also been reported from San Domingo and
Great Abaco. The San Domingan specimens are undoubtedly S. difficilis Bar-
bour.

Sphaerodactylus exsul, sp. nov.

Type: — No. 7,894, M. C. Z. an adult. Little Swan Island, Caribbean Sea,
March, 1912, George Nelson, collector. A fine series of paratypes.

Closely related to S. notatus Baird from which it has apparently been derived.
Similar to this species in all the general features of its squamation but differing
markedly in having slightly smaller and very much more weakly keeled dorsal
scales. Its coloration is similar in type but much brighter and more varied. A
conspicuous spectacle-like marking with two white spots is often seen on the
posterior neck region. I have not observed this in S. notatus.

Mr. Nelson says that these little lizards are very abundant in the accumula-

REPTILIA. 265

tions of humus and fallen leaves in the cavities and depressions so very common

in the sharply eroded aeolian limestone of the island.

Sphaerodactylus difficilis, sp. nov.

A new species, belonging among those having large, strongly keeled, and
imbricate scales above, being without granular scales along the vertebral line,
and having the breast scales smooth. Thus it is evidently related to S. notatus
Baird and its allies. Its nearest ally is indeed S. notatus; and from this species
it may be separated at once by its much greater adult size, longer tail, entirely
different type of coloration, and by the somewhat larger dorsal scales, which at
first sight recall those of S. grandisquamis and S. macrolepis. This similarity
was so striking that two of the four specimens in hand were labeled by Garman
S. macrolepis, although they had smooth breast scales, not keeled, and came from
Puerto Plata, San Domingo, instead of the Virgin Islands. He identified other
similar Haitian specimens as S. nigropunctatus.

The typical series consists of four specimens, two collected by Mr. M. A.
Frazar, in December 1881, at Puetro Plata, San Domingo; the other two from
Santiago de la Vega, San Domingo, caught by Mr. A. H. Verrill in 1908.

Type: — No. 7,834, M.C.Z. Santiago dela Vega, San Domingo. A. Hyatt
Verrill, collector.

Snout rather rounded, not conspicuously pointed; distance of eye from tip
of snout and ear opening almost equal; rostral moderate, with conspicuous
median groove, nostril between rostral, first supralabial, two small postnasals
and a larger supranasal, which is separated from its fellow of the other side by
two small median seales; three large supralabials, followed by one much smaller
(the same on both sides), a prominent supraciliary spine over the middle of the
eye; head above and on sides covered with small, elongate, juxtaposed, keeled
scales, those on snout enlarged; scales on back large, imbricate, keeled, about
12 to 18, equaling in size the distance from snout to ear; mental large, nearly
same size as rostral; two very large infralabials followed by a small one to below
centre of eye; two small, squarish chin-shields behind mental, followed by flat
smaller scales, which extend all over throat, becoming quite small, but always
flat and not granular; larger and smooth on lower neck and chest, scales of
abdomen larger, but seldom exceeding in size those of back; limbs with smaller,
keeled, imbricate scales; similar on upper surface of tail, but smooth below, and
with some widened to form rather irregular transverse plates.

Length of body snout to vent, 28 mm. Length of tail, 31 mm. Total
length, 59 mm.

266 BARBOUR: ZOOGEOGRAPHY.

Color, in alcohol, light gray-brown, with many darker spots scattered
irregularly over the whole dorsal area; some cover a single scale; others are
composed of several dark scales juxtaposed. On the tail the spots average
larger than on the body. Below, minutely speckled with dark brown upon a
lighter field, throat and upper neck heavily spotted with dark brown marking
but little smaller than those of the dorsal area. Under surface of tail much more
heavily speckled than belly.

The two Puerto Plata specimens have darker spots merged into lines above
and also lighter spots. On the shoulders of one are two enlarged white spots
with a black mark between, as often seen in S. grandisquamis. In the other, no
other such mark is evident; but the three dark lines run forward, one to each
eye and one to the nape region. These two specimens also vary slightly from
the type and from one another in the size of the dorsal scales. In one they are a
little larger than in the type; in the other, slightly smaller; this is the one with
the white and black shoulder mark. These two are from Puerto Plata, San
Domingo (M. C. Z., No. 5,444). The fourth specimen has a narrow head, more
narrow than that of any of the others. At first I thought it represented a dis-
tinct species, but artificial distortion due to capture or preservation may be
responsible for the difference noted. In its other features it is the same, except
that it has smaller dorsal scales, like the second specimen mentioned from Puerto
Plata. It resembles this example also somewhat in coloration,— immaculate
below; faint longitudinal lines on the back, three passing forward to the eyes and
neck; a spectacle-like marking on the shoulders and another faint one on the
sacral region (M. C. Z., No. 7,835).

More material is needed before it can be definitely settled whether all these
specimens represent one species; but I think they do, and that this species is
simply unusually variable in coloration and in the size of the dorsal seales.
Almost, if not quite, as great variation in coloration, and that strikingly similar,
too, has been described by Stejneger for S. grandisquamis of Porto Rico. This
new species seems more or less to connect S. grandisquamis and its allies with
the group to which S. notatus is related. It also stands between these groups in
geographical position.

Mr. Mann’s recent collection contains an example from Grand Riviere,

Haiti which I place with this species.

REPTILIA. 267

Sphaerodactylus corticolus Garman.

GarMAN, Bull. Hssex inst., 1888, 20, p. 111. Sresnecer, The Bahama Islands, 1905, p. 331. Rosin,
Lunds univ. arsskrift, 1911, 7, no. 5, p. 27.

The type locality is Rum Cay (Types M. C. Z., No. 6,219). It occurs also
on Watlings Island. It has not been reported from any other locality.

Though closely related to both S. notatus and S. argus with their allies, it is
apparently a species easily distinguishable.

Sphaerodactylus goniorhynchus Copr.

Cop, Proce. Acad. nat. sci. Phila. for 1894, 1895, p.440. Barsour, Bull. M.C. Z., 1910, 52, p. 291.

I can add nothing to my previous remarks on this species (loc. cit.). It is
confined to Jamaica. It seems to be very distinct, and has a type of coloration
which is quite unique in a genus where there is very remarkable variation along
certain more or less definite lines among the different species and individuals
within a species.

Sphaerodactylus gilvitorques Corr.

Cork, Proc. Acad. nat. sci. Phila., 1861, p. 500. Boutencerr, Cat. lizards Brit. mus., 1885, 1, p. 227.
Barsour, Bull. M. C. Z., 1910, 52, p. 291.

The type said by Cope to be in the Museum of the Academy of natural
sciences at Philadelphia is from Jamaica, to which island the species is confined.

This lizard is really not uncommon, although, on account of its burrowing
habits, it is seldom seen; it has always been rare in collections of Jamaican rep-
tiles. I found many specimens in various parts of the island, in leafy mould
under logs and stones, especially while collecting Peripatus on the hills about
Bath.

Sphaerodactylus copii SrernDACHNER.
STEINDACHNER, Voy. Novara. Rept. 1869, p. 18, pl. 1, fig. 5. BouLenamr, Cat. lizards, Brit. mus.,
1885, 1, p. 225. Ginruer, Ann. mag. nat. hist., 1888, ser. 6, 2, p. 363.
The type specimen was said to have come from South America, with no more
definite clue to its origin. It has now been found on Dominica, if Gunther’s
identification is correct, and this may be the true type locality; at any rate, it

does not seem to have been reported definitely from anywhere else.

Sphaerodactylus microlepis Reinnarpt & LUTKEN.

Reruarpt & LirKken, Vid. Meddel. nat. foren Kjobenh. for 1862, 1863, p. 278. BouLmncrr, Cat.
lizards Brit. mus., 1885, 1, p. 224. Bounenemr, Proc. Zool. soc. London, 1891, p. 351.

Boulenger points out how this species differs from its ally, S. copii. He
remarks that it is one of the most widely distributed West Indian member of the

268 BARBOUR: ZOOGEOGRAPHY.

genus. Individuals that he considers to belong here have been collected on St.
Lucia, St. Croix, and Dominica. I do not know the species from autopsy.

Sphaerodactylus pictus Garman.

GarMaAN, Bull. Essex inst., 1887, 19, p. 20.

This species, which to judge by descriptions is well worthy of being con-
sidered separate from its Lesser Antillean neighbours, S. fantasticus, S. copii, S.
vincenti, and S. microlepis, is represented in the collection by the types only
(M. C. Z., No. 6,071), of which there are four, from the island of St. Kitts or St.
Christopher.

Sphaerodactylus vincenti BouLENGER.

BouLenceER, Proc. Zool. soc. London, 1891, p. 354.

No specimens of this apparently perfectly valid species have come under my
observation. So far as known it is confined to St. Vincent.

Sphaerodactylus melanospilus Bocourt.
Bocourt, Miss. sci. Mex. Reptil., 1873, p. 44.

This species, the type locality of which is St. Lucia, is not represented in the

collection.
Sphaerodactylus richardsonii Gray.

Gray, Cat. lizards Brit. mus., 1845, p. 168. BouLenaer, Cat. lizards Brit. mus., 1885, 1, p. 227, pl. 18,
fig.6. Barsour, Bull. M. C. Z., 1910, 52, p. 290.

This, the finest and most conspicuous member of the genus, is confined to
Jamaica, where it seems to be very rare. Gosse (Naturalists’ sojourn in Jamaica,
1851, p. 254) met with only a single example, that near Montego Bay in the
western part of the island; while those taken by the writer were caught near '
Kingston. They came from the palm-leaf thatch of old roofs. During the
winter of 1911-1912 Dr. H. L. Clark of the Museum staff found it again at

Montego Bay.
Sphaerodactylus picturatus Garman.

GarMAN, Bull. Essex inst., 1887, 19, p. 19.

This species, which is very closely related to S. anthracinus Cope, of Mexico,
is represented in the Museum by the type series, comprising three specimens
(M. C. Z., No. 3,342) from western Haiti, collected by Dr. D. F. Weinland.

A comparison of these examples with Boulenger’s description (Cat. lizards
Brit. mus., 1885, 1, p. 225) of what he calls S. anthracinus Cope, based on a
specimen from San Domingo, shows that they are beyond doubt identical.
Turning, however, to Cope’s original description (Proc. Acad. nat. sci. Phila.,

REPTILIA. 269

1861, p. 500), we note that he expressly states that his S. anthracinus from Mexico
has four upper labials and five lower, ‘‘two large and three small.” The Haitian
examples, both ours and Boulenger’s, have four above and below. This char-
acter also seems to be one which is very stable in this genus.

The color described by Boulenger agrees quite well with that of our types
in its important features of head lines, and spots on back limbs and tail. But
Cope says, ‘“‘Color black, the large dorsal scales tinged with blue.”’ This colora-
tion which is unique among Sphaerodactyli is coupled with the different number
of labials mentioned in Cope’s very inadequate description, and the fact that the
type came from Mexico, while S. picturatus comes from Haiti, a province not
closely related zodgeographically with Mexico, should afford sufficient reason for
keeping the two species distinct, and for concluding that S. anthracinus is con-
fined to Mexico.

Since the above was written, I have found a single specimen, also from Haiti,
which Garman has identified with his species, and which I believe also is abso-
lutely typical in squamation. It is uniform steely gray in color, and shows that
Haitian specimens have a distinct dichromatism. Cope’s S. anthracinus may
really have come from Haiti, and it may have been inaccurately described by
him. If, then, the species is not to be found in Mexico, and is really the same in
the matter of the labials, Cope’s name has precedence, and this species must be
known as S. anthracinus.

This species is, moreover, very closely related to S. grandisquamis Stejneger,
from Porto Rico. It may be distinguished by the more elongate and flatter
snout, and by the slightly larger dorsal scales.

Recently a Sphaerodactylus, which, as far as one may decide from a single
specimen, belongs to this species, has been received in exchange from the Ameri-
can museum of natural history in New York. It was collected in the Sierra de
Jalibonico in Cuba. Similar individuals were called by Gundlach (Erpet.
Cubana, 1880, p. 61) S. fantasticus. I have seen specimens identified by him
in the ‘‘Museo Gundlach” of the Instituto de Segunda Ensafanza, in Havana.
While still more recently several enormous examples have been collected by Mr.
W. M. Mann at Diquini, Haiti. .

Sphaerodactylus asper GARMAN.
Garman, Bull. Essex inst., 1888, 20, p.113. Rosfn, Lunds univ. arsskrift, 1911, 7, no. 5, p. 27.

The types, two specimens from Middle Bight, Andros Island, Bahamas
(M. C. Z., No. 6,222) are the only specimens in the collection. Neither Rosén,

270 BARBOUR: ZOOGEOGRAPHY.

nor Bryant, who collected in 1904 in northern Andros and about Mangrove Cay,
found this species. It is large and conspicuous among Sphaerodactyli.

Garman suggests that this species may be related to his S. picturatus. It is
indeed very closely related, differing only in having five labials above and below,
and in not having the median head scales enlarged as they are in S. picturatus.

Sphaerodactylus grandisquamis STEsJNEGER.

SresNnEGER, Rept. U.S. nat. mus. for 1902, 1904, p. 602, fig. 51-52.

This species, which had hitherto been confounded with S. macrolepis
Giinther, the type locality of which was St. Thomas, has been shown by Stejneger
to be perfectly distinct, yet to have close affinity with Giinther’s species. Stejne-
ger’s species may easily be separated by its much larger dorsal scales. In its
distribution it is confined to Porto Rico and Vieques.

Sphaerodactylus monensis MrerrwarTuH.

Mererwarta, Mitth. Naturh. mus. Hamburg, 1901, 18, p. 20. Sresnucur, Rept. U.S. nat. mus. for
1902, 1904, p. 607.

A species related to S. macrolepis Ginther and S. grandisquamis Stejneger,
and confined to Mona Island.

Sphaerodactylus macrolepis Reinnarpt & LUTKEN.

Reruarpt & LirKken, Vid. meddel. nat. foren Kjobenh. for 1862, 1863, p. 174, 279.

Originally described from St. Croix, St. Thomas, and Vieques, it was after-
ward confused with S. grandisquamis Stejneger. It is, so far as known, confined
to St. Croix and St. Thomas, since Stejneger found that Vieques individuals
belong with his Porto Rican species. The Museum has it from St. Thomas only.

Sphaerodactylus sputator (SpaRRMAN).

SpaRRMAN, K. vet. akad. Handl., 1784, 5, p. 164, pl. 4, fig. 1.
Sphaerodactylus fantasticus Dumértt et Bipron, Erpét. gén., 1836, 3, p. 406, pl. 32, fig. 2. BovuLencrr,
Cat. lizards Brit. mus., 1885, 1, p. 223. :

As I said when describing S. torrez, the name S. sputator (Sparrman) must be
applied to a Lesser Antillean species. Boulenger’s only Antillean specimen that
could belong with this species came from the island of Antigua. Its occurrence
which he reports in Mexico and Venezuela would be completely out of accord
with everything known regarding the distribution of the many species of this
genus. Yet Reinhardt and Liitken (Vid. Middel. nat. foren. Kjobenh. for 1862,
1863, p. 277) record it from Haiti as well as from Venezuela. Individuals of
Sphaerodactylus, like other gekkoids, may be carried about accidentally. The
type locality was St. Eustatius for S. sputator and Martinique for S. fantasticus,

REPTILIA. 271

and very probably the species are each confined to a single island. Boulenger’s
example from Antigua needs comparison with a topotype before its identification
can really be considered settled.

Chamaeleolis chamaeoleontides (Dumérin et Bipron).

Dumérit et Brsron, Erpét. gén., 1837, 4, p. 168. BouLenaerr, Cat. lizards Brit. mus., 1885, 2, p. 7.

This species, which in many respects is the most remarkable development
within the fauna of the whole West Indian area, is found only upon Cuba. It is
far from common; and attempts to secure it usually bring in specimens of Anolis
equestris, a species which is not dissimilar in general form. Dr. de la Torre told
me that the only specimens he had ever seen were those collected by Gundlach.
The latter unfortunately gives little information as to its occurrence, saying
only, ‘‘Esta especie es rara y la he encontrado solamente en bosques.” This
was one of the species which we were most anxious to secure, and this year after
many ineffectual searches we secured four specimens, three from the forests of
the Sierra Maestra in the Jurisdiction of Jiguani and one from La Maya also in
Oriente which was kindly given me by Mr. C. T. Ramsden. The Museum has
also received a specimen in exchange from the U. S. national museum taken at
Santiago de los Vegas by Messrs. Palmer and Riley. Still more recently a
specimen from near the Rio Hanabana collected by Don Francisco Morales
and received through Dr. de la Torre permit me to make careful comparison
between specimens from the east and west sections of the island. They
appear to be the same. Cope’s C. porcus very probably came from Wright’s
collection made near Guantanamo, although it was without locality. It is
apparently a synonym of this species pure and simple.

This species is frequently mentioned under the name Chamaeleolis fernandina
Cocteau, in Sagra’s Hist. Cuba, Rept. (French ed.) 1848, p. 145, pl. 15. Both
Mr. Garman’s copy and my own of Cocteau and Bibron are the French edition,
and not having the Spanish edition available, I wrote to Dr. Stejneger for any
notes he might furnish me from his copy. With his usual kindness he wrote as
follows: —‘‘I regret to say that the information which I have relative to the dates
of Sagra’s Histoire, etc. is very fragmentary, as yet .... The work appeared
in two editions, Spanish and French, but the Spanish apparently slightly ahead of
the French. Both were issued in parts. Parts 1 and 2 of the Spanish edition
were published in 1838 and contained the text by Cocteau up to the genus
Typhlops. Turtles and lizards therefore date from 1838. Then Cocteau died
and Bibron wrote the snakes and the rest. The snakes appeared in 1840, of the

272 BARBOUR: ZOOGEOGRAPHY.

French edition I know less, only that pls. 4 and 18 were published in 1838.—
Therefore Chamaeleolis Cocteau 1838; Anolis chamaeleoides Dum. & Bibr.
1837; Chamaeleolis fernandina Cocteau 1838.”

Xiphocencus valenciennesii (Dumsérit et Brpron).
Dume&rit et Bipron, Erpét. gén., 1837, 4, p. 131. Bounenemr, Cat. lizards Brit. mus., 1885, 2, p. 9.
Barsoour, Bull. M. C. Z., 1910, 52, p. 293. ,

This conspicuous lizard is uncommon upon Jamaica, the only island where it
- occurs. The genus is not represented elsewhere in the West Indies. The only
other species known occurs about Bogotd, in Columbia. It also is very rare in
collections, so that I have not been able to assure myself as to their really belong-
ing to one genus. Such a discontinuous distribution can only mean that the
genus is a very old one, which once ranged through Central America and to
Jamaica while the latter was joined to that area, and that it has died out every-
where except about Bogoté and in Jamaica; or it may be that these similar spe-
cies have come to look alike by some process of convergent evolution, and that
they have no actual relationship with one another. Central America may not
have extended to Columbia while it was still one with Jamaica. There would
seem to be no special reason why such a genus should die out in Central America;
and yet among birds Sycalis jamaicae Sharpe is confined to Jamaica, and except
for a closely circumscribed species peculiar to western Mexico, all the other mem-

bers of the genus occur in South America from Columbia southward.

Anolis equestris Mrrrem.

MerreM, Syst. Amphib., 1820, p.45. Bovu.encrr, Cat. lizards, Brit. mus., 1885, 2, p. 21.

The largest and most conspicuous member of the genus. It occurs only in
Cuba, although it has been recorded from Jamaica. Within the island, its range
is general, since in the Museum there are specimens from Havana and Santiago
de Cuba, as well as from other localities. The specimens which Garman re-
corded from Bahia (Bull. Essex inst., 1887, 19, p. 27) came, of course, from
Bahia Honda, Cuba, not Bahia, Brazil. This is not made at all clear in the
context, where he suggests that it may represent another species.

Upon my trip to Cuba in 1912 I made special search for this lizard, which I
found far from common. I shot a half grown individual in a tree in the
mountains near San Diego de los Bafios, procured a fine adult, living, from a tree
in the lowlands near Herradura, and saw two, only one of which was caught by

the aid of Dr. J. L. Bremer, near Madruga. No other specimens were observed,

REPTILIA. 273

although it is known everywhere to the people as ‘‘Camaleon verde.”’ On my
trip in 1913 I found it rarely about Guantanamo and in the Sierra Maestra, but

as we were forced to be sparing of aleohol no examples were preserved.

Anolis garmani STEJNEGER.

SrrsNeGER, Amer. nat., 1899, 33, p.602. Barsour, Bull. M.C. Z., 1910, 52, p. 294.

A strictly arboreal species, corresponding to the A. ricordw of Haiti, A.
cuviert of Porto Rico, and A. equestris of Cuba. It is quite widely distributed
through the wooded regions of Jamaica.

Anolis ricordii Dumérin et Bisron.

Dumérit et Brsron, Erpét. gén., 1837, 4, p. 167. BouLencer, Cat. lizards Brit. mus., 1885, 2, p. 22.

This, the giant Anolis which is confined to San Domingo, was said by
Duméril and Bibron to be quite common. They had but one specimen, sent to
the Paris museum by Alexander Ricord, whose notes they undoubtedly quoted.
As a matter of fact, the species is a rare one in collections, and is probably also
uncommon in a wild state, as are its close relatives on the other Greater Antilles.
Mann’s Haitian collection contains two from Diquini. Other specimens in the

Museum are from Santiago in San Domingo, collected by Mr. A. Hyatt Verrill.

Anolis cuvieri Mprrem.
MerreM, Syst. Amphib., 1820, p.45. Srrsnecer, Rept. U.S. nat. mus. for 1902, 1904, p. 627, fig. 81—
84, 87.

Stejneger has shown that this lizard is confined to Porto Rico, Vieques, and
Tortola and that the early history as to its identity and habitat was obscure and
confused.

Anolis gundlachi Prrers.
Prrers, Monatsb. Akad. wiss. Berlin, 1876, p. 705. Stnsnecer, Rept. U.S. nat. mus. for 1902, 1904,
p. 633, fig. 89-91.

Stejneger says that this species seems to be ‘‘confined to the coffee belt and

the high mountain tops above it” in the island of Porto Rico.

Anolis monensis STEJNEGER.

Stresnecer, Rept. U.S. nat. mus. for 1902, 1904, p. 646, fig. 98-101.

Related to A. gundlachi and A. cristatellus, and confined to Mona Island.

BARBOUR: ZOOGEOGRAPHY.

bo
“I
TS

Anolis cristatellus Dumérit et Brpron.

Dumérit et Brpron, Erpét. gén., 1837, 4, p. 148. Sresnecrer, Rept. U. 8. nat. mus. for 1902, 1904,
p. 638, fig. 92-94.
Anolis scriptus GARMAN, Bull. Essex inst., 1887, 19, p. 23.

This Anolis is widely distributed, and extremely abundant where it occurs.
There are several hundred specimens in the collection from Haiti, Porto Rico,
St. Thomas, and Morant Island of the Virgin group. <A comparison of this great
series with the types of Garman’s A. scriptus shows them to be identical. The
original description of this species concluded with the sentence ‘‘Readily dis-
tinguished from A. cristatellus, which it closely resembles, by the greater size of
the two vertebral rows.”’ I can not see, however, that these are at all enlarged;
and there is no other character in which they vary from true A. cristatellus.
They are simply specimens of this species for which the locality label got in some
way confused. They were said to have been collected by Louis Agassiz on Silver
and Lena Keys, Florida. I can not learn of the location of these keys; and the
lizards doubtless came from Porto Rico or St. Thomas, where the species is
excessively common. I doubt the correctness of the locality label on the Haitian
specimens, as I do those on the Haitian specimens of A. pulchellus and A. stratu-
lus. Meerwarth, however, records specimens from the same island (Mitth.
Naturh. mus. Hamburg, 1901, 18, p. 21).

Anolis stratulus Cope.

Cops, Proce. Acad. nat. sci. Phila., 1861, p. 209. Sresnecer, Rept. U.S. nat. mus. for 1902, 1904, p. 651,
fig. 105-107.

This species is a distinct and handsomely marked one from St. Thomas,
Tortola, Just van Dyke, Vieques, and Porto Rico. The Museum, besides hay-
ing specimens from the last mentioned island and St. Thomas, has a couple
which are said to have come from Port au Prince, Haiti. Stejneger remarks
that it is desirable that this locality record be verified. For my part, I think
this is simply a result of a mixing of labels such as has also occurred in the case
of Anolis pulchellus, which is credited to Port au Prince also, doubtless incorrectly.
and in the case of A. cristatellus, which has been described as A. scriptus from the
Florida Keys.

Anolis homolechis (Corr).
Copr, Proc. acad. nat. sci. Phila., 1864, p. 169. BouLencrr, Cat. lizards, Brit. mus., 1885, 2, p. 28.

This widely distributed but quite uncommon woodland species may easily
be distinguished from every other Anolis on Cuba by its ivory-white dewlap.

REPTILIA. 275

We found it at Soledad, near Cienfuegos and in the forest about Madruga, south
of Matanzas. We also have it from Matanzas itself collected in 1912 by V. J.
Rodriguez and brought to the Museum by Dr. C. de la Torre. It is also not
rare in the heavy forests of the foothills of the Sierra Maestra and in 1913 we
found it in the woods near Pozo Prieto, back of Los Negros, Jurisdiction of
Jiguani, and also at Cabo Cruz. It is confined to Cuba.

Anolis gingivinus Corr.

Cork, Proc. Acad. nat. sci. Phila., 1864, p.170. BouLrnaer, Cat. lizards Brit. mus., 1885, 2, p. 28, pl. 2,
fig. 1. :
Anolis virgatus GARMAN, Bull. Essex inst., 1887, 19, p. 41.

Boulenger, in the Zoological record for 1887, was the first to place Garman’s
species from St. Bartholomew in the synonymy of Cope’s species from Anguilla.
This I can confirm definitely, as I have compared a topotype from Anguilla
(M. C. Z., No. 6,157) with the type series of thirty-five specimens which F.
Lagois collected on St. Bartholomew and described by Garman (M. C. Z., No.
6,165). A comparison of the text of Cope’s and Garman’s descriptions, or
of Boulenger’s description of Cope’s species, will show that they are all obviously
describing the same form. It is strange that Garman should not have noticed
this, since he had this same Anguilla specimen for comparison, but separated
by many pages in his list from the description of A. wrgatus. In the jar with
Garman’s types I find a label which identifies them with Daudin’s Anolis lineatus.
According to Boulenger this species has keeled ventral scales, which A. gingivinus
has not. Another label suggests that this species may be A. alliaceus Cope, also
from ‘Anguilla Rock near Trinidad.’”’ Cope, however, described that species
from specimens having no locality, and Gunther places A. alliaceus as identical
with specimens from Dominica.

In 1869, Cope (Proc. Amer. phil. soc., 11, p. 159) recorded this species from
the island of St. Martins.

In 1871 Cope noticed it from St. Eustatius, saying also, “‘This species was
described by me from specimens in the British Museum said to be from ‘ Anguilla
Rock, near Trinidad.’ It appears from Dr. van Rijgerma’s investigations,
(Proc. Acad. nat. sci. Phila., 1871, p. 220) that the island of Anguilla, far north of
Trinidad, was meant.”

Anolis alliaceus Corr.

Cop#, Proc. Acad. nat. sci. Phila., 1864, p.175. Gtnrurr, Ann. mag. nat. hist., 1888, ser. 6, 2, p. 363.

This species was first described from specimens without any data as to
locality. Gimther declares that specimens from Dominica are indistinguishable

276 BARBOUR: ZOOGEOGRAPHY.

from the type, so that Anolis oculatus Cope is a straight synonym of this species.
The only unfortunate feature of Giinther’s notice is the fact that he claims
Garman’s A. lividus, from Montserrat, and A. sabanus, from Saba, are the same
as Cope’s A. alliaceus. This is not the case; and such a statement makes one
doubt whether the identification of A. oculatus and A. alliaceus may not perhaps
be equally faulty. The descriptions, however, encourage the belief that this
settlement of the status of the Dominican lizard may be correct.

Anolis asper GARMAN.

GARMAN, Bull. Essex inst., 1887, 19, p. 31.

In regard to this and other species, the validity of which has been doubted
by Boulenger in the Zoological record for 1887, I can only say that the types
agree with Garman’s descriptions; and a careful reading of these will in every
case, I think, point out the characters by which the species may be distinguished
inter se. This species was included with A. nubilus, A. sabanus, A. speciosus,
and A. lividus, in what Boulenger called A. leachii (Zool. rec. Rept., 1887).
Ginther split this species in 1888, and retained A. lividus and A. sabanus in
what he called A. alliaceus. Both of these courses are incorrect, and an exami-
nation of the large typical series of each species show that all of them may easily
be distinguished. This species occurs upon Marie Galante, whence came the
large series of types (M. C. Z., No. 6,162).

Anolis speciosus GarRMAN.
GARMAN, Bull. Essex inst., 1887, 19, p. 42.

This species also occurs on Marie Galante, and is easily distinguishable from
the other species on that island (A. asper) by the characters given in Garman’s
original description (M. C. Z., No. 6,172).

Anolis lividus Garman.
GARMAN, Bull. Essex inst., 1887, 19, p. 48.
A distinct species from Montserrat, represented in the collection by almost
fifty specimens, the types (M. C. Z., No. 6,176, 6,177).
Anolis sabanus GARMAN.
GARMAN, Bull. Essex inst., 1887, 19, p. 39.
This striking species, with a bold and entirely unique type of coloration, is

found upon the Dutch Island of Saba, whence the types were sent to the Museum
by F. Lagois (M. C. Z., No. 6,161).

REPTILIA. 277

Anolis nubilus Garman.

GarRMAN, Bull. Essex inst., 1887, 19, p. 32.

A distinct species, so far known only from the island of Redonda, where
Richardson captured the types (M. C. Z., No. 6,181).

Anolis leachii Dumfrit et Breron.

Dumeérit et Brsron, Erpét. gén., 1837, 4, p. 152. BouLenerr, Cat. lizards Brit. mus., 1885, 2, p. 29

(part).

Duméril and Bibron did not know whence came the type of this species,
except that it was from the Antilles. The identity of this species with Anolis
ferreus Cope, from Guadeloupe, fixes its locality. Boulenger’s concept of this
species embraced individuals from Guadeloupe, Dominica, Nevis, and he stated
that the lizards from St. Eustatius were the same. Giinther recognizes the
distinctness of the Dominica specimens as A. alliaceus Cope (q. v.); and un-
fortunately no material is available here from the other islands, so that the
status of Nevis and St. Eustatius specimens must remain in doubt. Boulenger
has also recorded the species from Antigua, but specimens from this island should
also be compared with those from Guadeloupe (Ann. mag. nat. hist., 1894, ser. 6,
14, p. 375).

Anolis cepedii Mrrrem.

MerreM, Syst. Amphib., 1820, p. 44. Garman, Bull. Essex inst., 1887, 19, p. 34.

Garman identifies an enormous series of specimens from Martinique with
this poorly characterized species. There seems, however, to be no reason why
this should not be done; and the name thus takes precedence over A. alligator
Duméril & Bibron, which has been used by Boulenger and many others. This
species seems to be confined to Martinique, and is closely related to A. trinitatis
Reinhardt & Liitken (q. v.).

Anolis griseus GARMAN.

GARMAN, Bull. Essex inst., 1887, 19, p. 36.

Although Boulenger (Proc. Zool. soc. London, 1891, p. 355) has considered
this species a synonym of Anolis richardit Duméril and Bibron, nevertheless it is
safer to consider these species distinct. Boulenger had no topotypes of A.
richardii, which are said to have come from Tortola; and until such specimens are

compared directly with this species from St. Vincent, an opinion as to its validity

278 BARBOUR: ZOOGEOGRAPHY.

carries no weight. Duméril and Bibron’s descriptions, while often verbose,
generally omit mention of the characters which are now known to be most
diagnostic in Anoles. The types of this species from near Kingston, St. Vincent,
are M. C. Z., No. 6,184.

Anolis vincentii Garman.

GarMan, Bull. Essex inst., 1887, 19, p. 46.

This lizard has by Boulenger been called a variety of ‘‘ Anolis alligator” (Proc.
Zool. soc. London, 1891, p. 255); and if it existed on the mainland where the
ranges of the forms might overlap such an expression would fit the conditions.
But insular faunae remain perfectly isolated; and as no transition zones exist
about them, the groups of individuals differentiated upon each island must be
considered as species, unless they are absolutely identical with those on another
island. The types of this species are M. C. Z., No. 6,178-6,179, and came from
St. Vincent.

Anolis luciae Garman.

GarMAN, Bull. Essex inst., 1887, 19, p. 44.

This species is perfectly distinct, and has been adequately described by
Garman. There is no occasion to combine it with A. extremus Garman, or with
what Boulenger has called A. alligator in remarking that these, with A. gentilis
Garman and A. cinereus Garman, were all one species. The peculiar character
of the dorsal lepidosis, noted by Garman, serves at once to separate this species
from the others. It is known from the island of St. Lucia, whence Garman
collected the types (M. C. Z., No. 6,173).

Anolis extremus Garman.

GARMAN, Bull. Essex inst., 1887, 19, p. 35.

I can not agree with Boulenger (Zool. rec. Rept. for 1887) in considering
this species a synonym of what he calls Anolis alligator Duméril & Bibron. Nor
can I agree with Garman in considering it a variety of A. cepedit Merrem, which
is the name he prefers for A. alligator. A comparison of the types (M. C. Z.,
No. 6,183) with Garman’s original description and with the other species leads
me to the conclusion that the description is perfectly accurate, and shows well
how different this species is from its relatives. This form is found upon Bar-
bados.

REPTILIA. 279

Anolis marmoratus Dumérit et BrBron.

Dumérit et Brpron, Erpét. gén., 1837, 4, p. 139.

Having no Guadeloupe specimens, I am unable to test the question of the
distinctness of this species from A. leachit. Boulenger (Cat. lizards, Brit. mus.,
1885, 2, p. 31) says:— “‘I am not able to detect in the description a single char-
acter of sufficient importance to justify a separation of this form from A. leachii.”’
Since that was written Giinther has admitted A. alliaceus Cope from Dominica
to be different from A. leachii; and it would be very strange to find the same
lizard upon Guadeloupe and Martinique, and a different one upon Dominica.
So that, upon this inferential evidence alone, there is but little doubt as to the
validity of the species. Garman has identified a large series from the island
of Desirade as belonging here; they should, however, also be compared with
specimens from Martinique.

Anolis bimaculatus (SparRMAN).

SparrMan, K. vet. akad. Handl., 1784, 5, p. 179, pl. 4, fig. 4. Garman, Bull. Essex inst., 1887, 19,
p. 29.

With this ancient and much disputed species Garman has identified speci-
mens from St. Kitts and Nevis. This seems an excellent solution of the question.
There can be no doubt of the distinctness of lizards of these islands from those
of Guadeloupe (A. leachii) and Dominica (A. alliaceus) with which this species
was associated by Boulenger (Cat. lizards Brit. mus., 1885, 2, p. 29), who in-
cluded it in part with what he called Anolis alligator, another composite species.
In a recent letter, Dr. Stejneger calls my attention to what Andersson has to
say regarding this species (Bihang. Svensk vet.-akad. Handl., 1900, 26, 4, 1,
p. 27). He has the types from St. Eustatius in hand, and believes that the
name should replace A. leachii; but the species are better kept apart until
examples from Guadeloupe can be compared with others from these three islands.

Anolis richardii Dumérit et Brsron.

Dumerit et Brsron, Erpét. gén., 1837, 4, p. 141. Boutrneer, Cat. lizards Brit. mus., 1885, 2, p. 37.

In spite of the direct statement of O’Shaughnessy (Ann. mag. nat. hist., 1875,
ser. 4, 15, p. 275), that Gray’s A. occipitalis and A. stenodactylus belong to this
species, which Duméril and Bibron described from the island of Tortola, we can
but emphasize the enormous discrepancies between the descriptions of all these
species. Boulenger has followed O’Shaughnessy; but by examining the list of
specimens in the British museum given by Boulenger, it may be observed that

280 BARBOUR: ZOOGEOGRAPHY.

no topotypes were available for comparison. So that these opinions as to the
status of Gray’s species carry no weight whatever, since with Anoles such matters
can be settled only by the comparison of well authenticated material from a
definite locality. Unfortunately the case was complicated by the fact that
Gray’s A. occvpitalis had no definite data regarding origin; and his type of A.
stenodactylus was said to come from Jamaica, where the species does not occur.
Until these types can be critically compared with Tortola specimens, their status

must remain unsettled.

Anolis trossulus Garman.

GARMAN, Bull. Essex inst., 1887, 19, p. 14.

This lizard, which Garman found abundant in Grenada in 1887 is still by far
the most abundant reptile on the island; and Dr. Allen obtained a very large
series, and brought back as well living examples, two of which successfully passed
through a New England winter at ordinary room temperature. They feed
voraciously on flies and meal worms, and make most interesting pets. The
following are Allen’s field notes regarding the color of the gular appendage in
wild living examples: —

‘““Gular sac dark brown on anterior edge, and on the anterior part of the
scale rows; posterior ventral edge paler brown; scale-rows on posterior half, or
two-thirds of the sides of the sac, are pale greenish yellow, the skin between the
rows pale.tea-color.”’

Allen records many wild specimens as being “‘nearly blackish above with
greenish mottlings; below chin green, angles of jaw white with black mottlings,
belly green; inside of legs green.”’ In our living specimens here in captivity,
the most common condition is a uniform or slightly mottled or clouded rich
brown above, on each side a lateral band of light green. The belly is light green,
limbs barred or mottled with green and brown. This color may change under
excitement to uniform brownish or greenish, or to a condition where dark cross-
bars are present on the dorsum. A short, very light green stripe near the axilla,
and white lines on the lips, are frequently present. The green lateral band is
seldom lost, and that only under conditions of the greatest fright or excitement.
The color phases appear to be very clearly controlled by the emotions, as is
common with other color-changing reptiles. While this is not the sole factor
underlying metachrosis, it is certainly far more important than many experi-
mentalists are wont to believe. I know of no field observer of living lizards who
has not been impressed this same way.

The types of this species are M. C. Z., No. 6,181.

REPTILIA.

)
(
a

Anolis trinitatis Remmuarpr & LUTKEN.

ReinHarpt & LirKen, Vid. Meddel. nat foren Kjobenh. for 1862, 1863, p. 269.
Anolis gentilis GARMAN, Bull. Essex inst., 1887, 49, p. 34.
Anolis cinereus GARMAN, loc. cit., p. 35.

The latter synonymous form was described as a subspecies, and was founded
by Garman upon color differences, which he states were evident in comparison
with specimens of A. trinitatis Reinhardt & Liitken. The color was said to be
“intermediate between that of the light-grayish varities, from Trinidad and
Petit Martinique [A. gentilis| and the brownish from Barbadoes,” [A. extremus].
It is described as being “‘olive or bluish; the legs show more of the lilac color.”
According to Allen’s careful and extensive field notes, made from the living
specimens, they are described as being invariably light grayish in life, and as
being found entirely confined to the shore formation of vegetation. They are
almost always found on the bark of the cocoloba and manchioneel trees; indeed,
Allen found them nowhere else, and never took one in the interior of the island.
Allen reports that the belly is grayish in life, the upper side often spotted, and
the gular appendage pale gray, or perhaps better, greenish gray. He was re-
quested especially to investigate this latter character, as the pouch color is a very
valuable diagnostic feature. Until comparison can be made directly with the
allied forms in their living state, it is impossible to separate specimens from
Trinidad, Grenada, and Petit Martinique.

‘ Boulenger (Cat. lizards Brit. mus., 1885, 2, p. 31) considers A. trinitatis a
synonym of A. alligator so-called. A comparison of numbers of specimens from
Trinidad and Martinique shows that this is not justified, and the species should
stand alone as quite distinct from A. extremus from Barbados, on the one hand,
and from A. cepedw from Martinique, on the other. On zoédgeographical grounds
this course is reasonable, since the fauna of Trinidad, Grenada, and the Grena-
dines is more homogeneous than is the fauna of Grenada and Barbados, or

Grenada and St, Lucia or Martinique.

Anolis grahamii Gray.
Gray, Cat. lizards Brit. mus., 1845, p. 247. Barsour, Bull. M. C. Z., 1910, 52, p. 296.
A Jamaican lizard easily confused with the following, but perfectly distinct.

It occurs only in eastern Jamaica, so far as my collecting shows. The specimens

which I recorded having taken at Mandeville were really from Mooretown.

282 BARBOUR: ZOOGEOGRAPHY.

Anolis opalinus Gosse.

Gossg, Ann. mag. nat. hist., 1850, ser. 2, 6, p. 345. Barsour, Bull. M. C. Z., 1910, 52, p. 296.

A rare Jamaican species now. It is confined to the western part of the island,
where Gosse reported it as common; but as this region is not very densely popu-
lated, the mongoose has made greater ravages here than elsewhere among the
lizards. Near villages the mongoose is often hunted and killed as a pest, and near
these villages the lizards persist most abundantly.

Anolis iodurus Gosst.

Gossg, Ann. mag. nat. hist., 1850, sem, 2,6,p.344. Barsour, Bull. M.C. Z., 1910, 52, p. 295.

A species which has been confused with A. grahami Gray. It is quite
widespread, but nowhere very common in Jamaica.

Cope (Proc. Acad. nat. sci. Phila., 1861, p. 210) says: — ‘“‘the ventral
plates of this species are usually smooth; they are sometimes carinate. Speci-
mens exhibiting the latter structure have been described by Dr. Hallowell as
A. punctatissimus.” It is thus evident that Cope confused A. ioduwrus Gosse,
with smooth ventrals, and A. grahamit Gray, with keeled ventrals, as so many
herpetologists have. I have noted the distinctness of both these species and A.
opalinus in my Notes on the herpetology of Jamaica (Bull. M. C. Z., 1910, 52,
p. 295).

Anolis conspersus GarRMaN.

Garman, Proc. Amer. philos. soc., 1887, p. 273.

Garman, in describing this species, compared it with the Jamaican A.
grahamu. Thisisnot apropos; the species need comparison only with A. zodurus,
since it has perfectly smooth ventrals. Boulenger (Zoél. rec. Rept. for 1887,
p. 10) says that this species is the same as A. grahamii, which is doubly incorrect.
The species is a distinct and easily recognizable one. The extremely sharp,
knife-like canthal ridge, and the very small occipital shield, as well as other
characters, distinguish this species at’once from its Jamaican ally, A. codurus.
It is confined to Grand Cayman Island, whence we have a type series of eighty-
seven examples (M. C. Z., No. 6,021).

Anolis distichus Corer.

Core, Proce. Acad. nat. sci. Phila, 1861, p.208. Bounenenr, Cat. lizards Brit. mus., 1885, 2, p. 33.

This species comes from New Providence, Great Abaco, and Cat Island in
the Bahamas; and, according to Boulenger, is the same as the A. dominicensis

REPTILIA. 283

of Reinhardt & Liitken. I have compared specimens from the Bahaman Islands
with Cope’s type (M. C. Z. No. 3,346), from Jeremie, Haiti, and find that they
differ very slightly. In Bahaman specimens the canthus is not so steep and
sharply angled, nor is the internasal region quite so prominent. They are,
except for these trifling differences, which are only evident in the average, and
which are subject to individual variation, absolutely alike. If we knew the
color in life of the dewlap of Haitian specimens, we might,— I may almost say
we should,— find it different from Bahaman. Rosén has separated the Andros
Island individuals largely on this excellent character.

Anolis distichoides Rosén.

Rosén, Lunds univ. arsskrift, 1911, 7, no. 5, p. 29, fig. 12, a-d.

A species very closely allied to A. distichus Cope, and confined to Andros
Island. In life it has a gular appendage, yellowish red in color instead of light
yellow. There are other small and apparently variable differences in squama-

tion.
Anolis cybotes Core. o

Corr, Proc. Acad. nat. sci. Phila., 1862, p.177. BovuLmnamr, Cat. lizards Brit. mus., 1885, 2, p. 34, pl. 1,
fig. 5.

This species was among the many from Haiti which the indefatigable Wein-
land collected at Jeremie, and of which the types are in the Museum (M. C. Z.,
No. 3,619). “Other specimens have come to hand from Samand, San Domingo,
so that the species evidently has a wide range over the island. Mr. Mann also
collected many at Diquini, Haiti.

Anolis haetianus Garman.

GarmaN, Bull. Essex inst., 1888, 20, p. 42.

The types of this species (M. C. Z., No. 6,191), which are three in number,
certainly differ from specimens of A. cybotes Cope in having keeled ventral
scales. The other differences of which Garman speaks are well marked. There
is another important character, however, which he overlooked, and that is
that the size of the plate-like scales on the frontal region are much larger in
this species than in A. cybotes.

Known from Haiti only.

Anolis citrinellus Cope.

Corr, Proc. Acad. nat. sci. Phila., 1864, p. 170, pl. 1, fig. 3. BouLenenr, Cat. lizards Brit. mus., 1885,
2, p. 35.

A beautiful little species, confined to Haiti. The specimens in the Museum
are from the vicinity of Port au Prince (M. C. Z., No. 1,326).

BARBOUR: ZOOGEOGRAPHY.

bo
oa)
Se

Anolis poncensis STEJNEGER.

SrpsNEGER, Rept. U.S. nat. mus. for 1902, 1904, p. 665, fig. 117-120.

A species closely circumscribed by the bounds of a small area near the city
of Ponce, Porto Rico.
Anolis evermanni STrsJNEGER.

SresNEGER, Rept. U.S. nat. mus. for 1902, 1904, p. 647, fig. 102-104.

Confined to the same regions in Porto Rico as A. gundlachi.

Anolis krugi PrtTers.

Perers, Monatsb. Akad. wiss. Berlin, 1876, p. 707.

Known only from Porto Rico, closely related to A. poncensis and A. pul-
chellus.

Anolis acutus HaLLowe.u.

HALLOWELL, Proc. Acad. nat. sci. Phila., 1856, p. 228. BouLEncEr, Cat. lizards Brit. mus., 1885, 2,
p. 36.
A species not very distantly related to Anolis krugi of Porto Rico, but con-
fined to St. Croix in the Virgin group.

_ Anolis isolepis Corr.

Cork, Proc. Acad. nat. sci. Phila., 1861, p. 214. BouLencer, Cat. lizards Brit. mus., 1885, 2, p. 46.

A very rare Cuban species. In the collection here there is an excellently -
preserved adult which was received from the American museum of natural
history by exchange. It was collected by Mr. Barnum Brown, probably in
the Sierra de Jatibonico, Cuba. More recently during February, 1913, we
procured a specimen in the heavy forest which lies beyond Los Negros in the
Sierra Maestra, Jurisdiction of Jiguani. Since then I have examined a specimen
taken by Mr. C. T. Ramsden near Guantanamo; he kindly informs me that the
color of the dewlap in life is ‘‘apricot yellow.”

Anolis lucius Dumfrit et Brsron.
Douménrit et Brsron, Erpét. gén., 1837, 4, p. 105. BouLrenasr, Cat. lizards Brit. mus., 1885, 2, p. 45
(part).
This species both Dr. Stejneger and myself believed to be distinct from the
following on the basis of Cocteau’s beautiful plate in Sagra’s work, compared
with Cope’s description and specimens referred to A. argenteolus from the

REPTILIA. 285

Province of Oriente, Cuba. Now a comparison can be made of specimens taken
by us at Madruga, in the province of Matanzas, and at Matanzas itself, and
others taken by Dr. de la Torre at Matanzas and still others taken in central
Cuba by Messrs. F. M. Chapman and Barnum Brown of the American museum
in New York. These examples show that the species A. lucius is found in
central Cuba, viz., at Matanzas, Sierra de Jalibonico, Sancti Spiritus, Trinidad
and Madruga, and is very distinct from A. argenteolus, the type of which came
from Monte Verde, near Guantanamo in eastern Cuba. A notice of the dis-
tinctive character is given under the discussion of the next species.

Anolis argenteolus Copr.

Cops, Proc. Acad. nat. sci. Phila., 1861, p. 213.

There are specimens of this species in the collection from about the city of
Santiago de Cuba, collected by Major Wirt Robinson and by myself in 1908.
In 1913 we found that this species has a wide distribution in eastern Cuba, and
collected it in Cabo Cruz, Los Negros in Jiguani, Bueycito near Bayamo, and
from Monte Libano near Guantanamo.

It may be distinguished at once from specimens of, or from Cocteau’s
figure of, A. lucius by its colorationand form. It is a much more slender species,
having a very narrow head almost twice as long (from tympanic orifice to snout)
as broad, whereas A. luciuws has a head barely once and a half times as long as
broad. In the former species the head, and especially the snout, is very depressed,
almost spatulate; this is not the case with A. lucius, in which the profile line
from top of head to tip of snout is simply a gently inclined plane surface. The
color of A. argenteolus is a more or less uniform ashy gray, with mottlings of
slightly darker gray. The dewlap is greenish white, the throat being pure white
in the male, and mottled with darker in the female. A. lucius is conspicuously
banded on the back with generally five distinct dark gray bars across an ashy
gray field. A broad white band is always present encircling the entire head.
This is very evident in life and passes along the lips and around the back of the
neck. There is a white spot invariably present on the occipital region, which is
not found in A. argenteolus. The two species are so different in habit and color-
ation as to perfectly easily distinguish them from a considerable distance when
they are observed in a wild state.

It is a noteworthy fact that this and the preceding species are entirely
confined to cliffs of limestone rock. They are sometimes found upon trees
growing on or near the ‘‘paredones”’ but far more often upon the rock itself.

286 BARBOUR: ZOOGEOGRAPHY.

Both this species and A. luciws are confined to the island of Cuba, one
being confined to the central region and the other replacing it in the eastern

province of Oriente.

Anolis wattsii BouLENGER.

BouLEeNGER, Ann. mag. nat. hist., 1894, ser. 6, 14, p. 375.

This distinct and interesting species is known from the island of Antigua
only. Boulenger notes its very close relationship to Anolis krugi of Porto Rico.

Anolis lineatopus Gray.
Gray, Ann. mag. nat. hist., 1840, 5, p. 113. Boutrnenr, Cat. lizards Brit. mus., 1885, 2, p. 39, pl. 1,
figs. 1-2. Barsour, Bull. M.C. Z., 1910, 52, p. 294.
A very common species, apparently confined to the hot, rather dry coastal
plain region about Kingston and Spanish Town, Jamaica.

Anolis sagrei Dum#rit et Brsron.

Doumérit et Brsron, Erpét. gén., 1837, 4, p. 149.

This species, which is so closely allied to the one following, is very common in
Cuba, where I have taken it about Havana, Pinar del Rio, Puerto Principe,
Matanzas, Guantanamo, Santiago, and in fact at every locality where I have
ever collected on the island; and there are specimens in the Museum from other
places, showing it to be ubiquitous. It occurs upon Jamaica, whence Stejneger
has received specimens, some of which are now in the M. C. Z. I was unable to.
find it there. Boulenger (Cat. lizards Brit. mus., 1885, 2, p. 40) records it also
from Belise and the surrounding region, as well as Caracas. I have seen no
individuals from the mainland, but feel very skeptical as to whether any species
of Anolis has such an improbably discontinuous range.

Anolis ordinatus Corr.

Corr, Proc. Acad. nat. sci. Phila., 1864, p. 175.

I follow Dr. Stejneger in considering the Bahaman lizards so closely allied
to A. sagrei as really being distinct, in spite of the more recent assurance of
Rosén that they are not. The difference in dewlap color of which Stejneger
speaks is a character which I also have learned in the field to be of very great
taxonomic importance. The lizard is recorded from most of the northern
Bahamas. M. C. Z., No. 3,424 is said to come from Double Headed Shot Key,
Florida. This record is in need of confirmation. —

REPTILIA. 287

Anolis nelsoni, sp. nov.

Type: — No. 7,892, M. C. Z. an adult male from Swan Islands, Caribbean.
Sea, March, 1912, George Nelson, collector. A large series of paratypes.

This species belongs with those having in the male a tail strongly compressed
especially along its upper edge; ventral scales keeled; supraorbital semicircles
separated by one row of scales. Thus its nearest relatives are A. sagret Dum. &
Bibr. and A. lineatopus Gray. It is more similar to the former than to the latter
species.

It is so conspicuously different from both species that it is only necessary to
point out the distinguishing characters. It differs from A. lineatopus Gray in
having a slightly more pointed snout; in having all the scales of the head and
snout much larger. The frontal ridges enclose but two or three instead of three,
four or even five series of small scales. The ridges themselves consist of but
three much elongated scales instead of five — six much less prominent scales. The
seales of the whole snout region are much larger and more strongly keeled than
in the Jamaican species. The dorsal scales are similar in size and character.
The ventrals of A. nelsoni are keeled and have the keels forming continuous
ridges as in A. sagret. The nuchal and dorsal fold as well as the crest of the tail
is much more strongly developed in the Swan Island species than in either of its
relatives. From A. sagret it may be distinguished at once by its dewlap color,
deep olive-gray instead of flaming orange almost pink. The ventrals are also
smaller and less strongly keeled. The head scales are similar in arrangement
but larger and much more strongly keeled. The dorsal scales are not feebly
keeled and while very small are nevertheless imbricate and strongly keeled.
There can be no question as to the close relationship of these two species yet the
Swan Island form is strongly differentiated.

Mr. Nelson’s field notes show that this lizard is excessively abundant in all
situations on both islands. He remarks that they change color very freely. <A
few young examples especially preserved to show their brilliant marking are
noteworthy in having a brilliant lemon-yellow head, besides the usual gray
rhombs in a blackish area, as is seen in A. sagret.

Anolis greyi, sp. nov.

Type: — No. 7,890, M. C. Z. an adult male, Camaguey (Puerto Principe),
central Cuba. March, 1909, Thomas Barbour, collector. Three cotypes.
Other specimens from the original series in the U. S. national museum.

When I identified the collection made in Cuba in 1909, I mistook this species

288 BARBOUR: ZOOGEOGRAPHY.

for Anolis homolechis, a species having a pure white dewlap in life and smooth
ventrals. On our recent visit to the island we found the true A. homolechis
near Cienfuegos and Madruga. A comparison at once showed that the series of
specimens from Camaguey really represented an undescribed form more closely
related to Anolis sagrei than to any other species. Its relationship is so evi-
dently with this well-known and widespread form that it is less confusing to
point out the differences between these two species than to present a prolix
diagnosis.

Comparing adult males, the frontal ridges of A. grayt are much more nearly
parallel than in A. sagrei. In the former species they are separated usually by
but a single row of scales or perhaps by two small scales lying side by side. In
the latter, however, there are either two or three distinct rows filling in the
ovoid area limited by the ridges. This is the most striking difference. In A.
grayt the head is more depressed, giving a very much more flat profile, not so
declivous as in A. sagrei. In this species also the ventrals are heavily keeled,
while in the new one the keels, while distinct, are far less strongly developed.
A less conspicuous distinction is the possession of larger and flatter plates about
the occipital scale than are found in A. sagret.

The color is similar — the same in alcohol. I remember very distinctly
that when these specimens were collected, there was a prominent. difference in
the dewlap color in life. Most unfortunately I did not realize at that time how
valuable a diagnostic feature this color character is.

I have named this species for my friend, Mr. Robert M. Grey, who is in
charge of the Harvard Botanical Station maintained at the Soledad estate near
Cienfuegos by the generosity of Mr. Edwin F. Atkins of Boston.

Anolis bremeri, sp. nov.

Type:— No. 7,889, M.C. Z. a single adult male, Herredura, Cuba, February,
1912, Thomas Barbour, collector.

This species belongs near A. sagrez, and its allies having keeled ventral scales,
but differs in squamation and in having a very large dewlap which is rich deep
maroon-brown in life, while in A. sagrez it is smaller and orange-red, flecked with
gray.

Adult male. Top of head with two curved frontal ridges which converge
anteriorly and inclose a shallow frontal area. Head scales including those of the
supraocular disc all strongly keeled. Supraocular semicircles separated by a

single row of scales, three of which are elongate and elevated to form a short

a

REPTILIA. 289

but distinct median ridge. Occipital about half the size of the ear opening
and separated from the supraocular semicircle by two or three rows of scales.
Supraocular dise composed of eight polygonal keeled scales, separated from the
semicircle by a single row of very small granular, polygonal or elongate scales.
Canthus rostralis composed of few scales, one extremely elongated in the middle
and flanked by others less elongate on either side. Superciliary ridge consisting
of one very long anterior shield, followed by a number of much smaller scales.
Loreal rows five; subocular semicircle strongly keeled; supralabials six, suture
between fourth and fifth being about under the centre of the eye. Temporal
seales small and granular, bordered above by a rather irregular zone of enlarged
scales, which extends from the posterior border of the orbit to a pomt some dis-
tance above the ear opening. Dorsals small, flat, smooth, showing no tendency
to imbricate, a median zone consisting of from three to four series of enlarged
more or less distinctly keeled scales. This zone commences at about the region
of the fore limbs. Ventrals much larger than dorsals, imbricate and distinctly
but not strongly keeled. Fore limbs above with rows of very strongly keeled and
enlarged scales, similar scales being also present on the anterior aspect of the
thigh. Subdigital expansions feebly developed, consisting of about eighteen
transverse lamellae under phalanges II and III of fourth toe. Tail strongly
compressed with about fifteen evident verticils along its proximal region. Hach
of these is separated by about seven vertical rows of keeled scales. Upper edge
feebly serrated, each section with four small spines. Dewlap very large and
capable of great dilation, extending along the midventral line to a point consid-
erably posterior to the line of the axillae. A very few feebly enlarged postanal
scales. No dorsal fold.

Color of living animal, grayish and reddish brown, variously marbled with
spots, blotches and wavy bands of dark brown or black. Belly pale ashy or
greenish. Dewlap fine rich lustrous maroon (blackish in alcohol).

A single specimen only found resting upon the thatched roof of a native
house. I was told that it was a species which was only rarely found. A most
observant young Cuban, Lucio Alfonso, who was with me at the time, said that
it was a lizard, very partial to thatched roofs. In general habit it resembles A.
sagret, but head, body, and limbs are much more slender.

Named for my friend, Dr. John Lewis Bremer, of the Harvard Medical
School, who has often most kindly aided me while upon collecting trips.

290 : BARBOUR: ZOOGEOGRAPHY.

Anolis luteosignifer Garman.

GARMAN, Bull. Essex inst., 1888, 20, p. 4.

Boulenger, in the Zoological record, Reptiles, 1888, p. 9, questions the dis-
tinctness of this species from A. sagret. Of the validity of the species there can
be no possible doubt. The two forms may be recognized at once. <A. luteo-
signifer has a snout almost flat instead of swollen in the internasal area, and
hence a quite different profile from A. sagrev. The granular scales are not only
much more uniform in size than in A. sagrei, but they also average smaller in size
The types of A. luteosignifer (M. C. Z., No. 6,228) are in excellent preservation,
and have been compared with a well-preserved series of A. sagrei which I col-
lected near Santiago de Cuba.

The species is known only from Cayman Brac; but Maynard’s notes show
that this species, or some close ally, occurs also on Little Cayman. This is also
said to be the only lizard on Cayman Brac.

Anolis vermiculatus Dumérit et Bipron.

Dumértt et Brsron, Erpét. gén., 1837, 4, p. 128. BouLenerr, Cat. lizards Brit. mus., 1885, 2, p. 41.

A large and striking lizard found only upon Cuba. Professor de la Torre
gave me a young example from the Sierra de Guane, in Pinar del Rio. Messrs.
Palmer and Riley found it at various localities in Pinar del Rio. Gundlach found
it only near the source of the Rio Taco Taco, and remarked that when pursued,
it took to the water. Thus it is evidently confined to the province of Pinar del
Rio.

Anolis loysiana CoctTrauv.
Coctrau, Comptes rend Acad. roy. sci., 1836, 3, p. 226. BouLmnemr, Cat. lizards Brit. mus., 1885, 2,
p. 42.

One of the most peculiar species of the genus. It is confined to Cuba.
I was until this year entirely unable to find this extremely well-marked species
which Gundlach says has a wide distribution. He did not consider it rare
and remarked that he always found it ensconced in crevices on the surface of
the trunks of rough barked trees. I have recently secured a specimen from
San Diego de los Bafios from the U.S. national museum. In February, 1913,
we collected a fine specimen in the village of Los Negros, Jurisdiction of
Jiguani, Oriente, Cuba. It was on the rock of a small cliff beside the stream
which has its source near the village.

REPTILIA. 291

Anolis argillaceus Corr.

Cops, Proc. Acad. nat. sci. Phila., 1862, p. 176.

I have never succeeded in finding this species myself. Boulenger placed it
in the synonymy of A. loysiana. Gundlach (Erpet. Cubana, 1880, p. 52), how-
ever, recognized it as a distinct species. After seeing the specimen which is
preserved in the Gundlach collection in Havana I was strongly convinced that
Gundlach was correct although as the cases are sealed the specimen could not
be examined. On April 6 while collecting at Alto de la Union on Monte Libano
near Guantanamo, Mr. C. T. Ramsden, found and sent me a lizard which proves
to belong to this species and a comparison with A. loysiana makes it evident that
they are not even very nearly related although the head scales are similar in
general character. A. argillaceus is brilliantly marbled with yellow and dark
brown about the head, has a rosy white dewlap, longitudinal dark brown bands
on a speckled light brown field along the body, which is entirely without evidence
of spines or tubercles. A. logsiana is perfectly lichen-like in coloration, very
pale, ashy gray in general tone with very irregular and only slightly darker
markings. A. loysiana is strongly depressed in habit, especially in life, while
A. argillaceus is strongly compressed though stoutly built. Its habit recalls
that of A. solepis.

Anolis semilineatus Corr.

Cork, Proce. Acad. nat. sci. Phila., 1864, p.171. BouLrenacer, Cat. lizards Brit. mus., 1885, 2, p. 68, pl. 5,
fig. 1.

This species is confined to Haiti; the only specimens in the collection are
from the eastern extremity of the island across from Samand, San Domingo.
Boulenger records specimens in the British museum from San Domingo only,
so that the species appeared to be localized in this region, until the receipt of
ten examples from Diquini, Haiti collected by Mr. W. M. Mann.

Anolis spectrum PETERS.
Prerers, Monatsh. Akad. wiss Berlin, 1863, p. 136. BouLencer, Cat. lizards, Brit. mus., 1885, 2, p. 69.
A rare species confined to Cuba. I know of it from the literature only.

Gundlach, who has collected most if not all of the known specimens found it
only near Matanzas and Cardenas.

292 BARBOUR: ZOOGEOGRAPHY.

Anolis cyanopleurus Corr.

Cops, Proc. Acad. nat. sci., Phila., 1861, p.211. BouLmnaer, Cat. lizards Brit. mus., 1885, 2, p. 69.

Another of the rare species confined to Cuba, and of which I have never been

able to find a specimen.

Anolis alutaceus Corn.

Corr, Proc. Acad. nat. sei. Phila., 1861, p. 212. Boutrnamr, Cat. lizards Brit. mus., 1885, 2, p. 71.

Apparently a rare species; it was not represented in the Museum until on
our trip in 1912 when we-secured a specimen at Madruga and one near Cienfue-
gos. In 1913 we got a fine series in the Sierra Maestra forests near Los Negros
in the jurisdiction of Jiguani and a few specimens at Monte Libano near Guan-
tanamo. The British museum did not possess a specimen at the time the second
volume of the Catalogue of lizards was written. Gundlach (Erpet. Cubana,
1880; p. 49) also remarks that he has found but few examples of this species
during his years of collecting. He adds that its range is not confined to any one
part of the island and that it is a woodland species.

It is confined to Cuba.

Anolis angusticeps HaLLoweEt...
HALiLowe tt, Proc. Acad. nat. sei. Phila., 1856, p. 228.

I have identified with this very imperfectly described species a large series
of lizards some of which I caught at Los Negros, Baire, San Luis, Monte Libano,
Pozo Prieto and other stations in the Province of Oriente, and many other
specimens which Mr. C. T. Ramsden has sent me for study from a number of
points about Guantanamo. I had at first named the adult males as a new spe-
cies but the large material which I have studied now convinces me that all are
referable to this species, first described from Cienfuegos and so poorly charac-
terized that Boulenger was unable to guess at its validity or even to place it in
the British museum Catalogue of lizards. When the first young examples were
received from Mr: Ramsden I thought that possibly I had to deal with A.
cyanopleurus, a species which I have never seen but this species has keeled
ventral scales and the figure of the head given by Bocourt (Miss. sci. Mex.
Reptil. pl. 16, fig. 29) indicates a very different species. The greatly enlarged
scales on the ventral aspect of the tail and other characters agree well with
Hallowell’s diagnosis. The colors fall into the range which Gundlach described

from two living specimens (Erpet. Cubana, 1880, p. 45).

REPTILIA. 293

The species is confined to Cuba and the following notes will serve to make
more clear its relationships.

It is closely related to A. homolechis, and has similar habits. It may be
recognized at once in the field by its enormously developed gular sac, which in A.
homolechis is pale white, while in this species it is carmine in the centre with a
broad marginal band of brilliant yellow. Besides this, it may be distinguished
by having fewer scales between the occipital shield and the superciliaries, and in
having the scales which surround the occipital for some distance upon the upper
temporal regions of a much smaller size. The canthus rostralis is separated from
the frontal ridge by two or three rows of scales.

Altogether this species is very similar to A. homolechis in details of squama-
tion. Without examining the type of A. homolechis, which is, according to Cope,
in the British museum without locality, it is impossible to be perfectly sure that
the species which I am naming now is not A. homolechis itself and the one with
the white dewlap a new species. Cope’s description is ambiguous, and the two
forms with which I am dealing are obviously variable in many characters.
Nevertheless, the dewlap color, which I have learned to regard as a very excellent
one, divides them sharply into two groups. Cope mentions the dewlaps as being
large. It is rather large in homolechis, but much larger in angusticeps. How-
ever, his note of the canthus as separated from the frontal ridge by but one row
of seales, of the occipital as separated from the superciliaries by,many rows of
scales, and by his mention of the metallic shades of the ventral surface would
seem to fasten the name homolechis to the group of individuals having the white
dewlap. This conclusion was reached last year by Dr. Stejneger and myself
before I had this related form. That we are dealing with two distinct forms I
have no doubt; but without an adequate description or a re-examination of the
type, there is no use attempting to do more than call attention to the facts, and
leave a more detailed discussion of status until the original type of homolechis

can be examined.

Anolis porcatus Gray.

Gray, Ann. mag. nat. hist., 1840, 5, p. 112.

This extremely distinct species, which was figured by Cocteau in Sagra’s
Historia de Cuba, Reptiles, pl. 9, is confined to Cuba. It has very frequently
been confounded with the North American A. carolinensis and the Bahaman
A. brunneus. Even Boulenger, in the Catalogue of lizards, 1885, 2, p. 43, in-
cluded it in the synonymy of A. carolinensis.

294 BARBOUR: ZOOGEOGRAPHY.

Gray (loc. cit.) recorded the locality whence the species was supposed to
have come by saying that it ‘‘Inhab. Cuba and Texas.” There is no doubt,
however, but that the Cuban species was the one described.

It is abundant all over the island, occurring most commonly in gardens.
It is especially fond of sitting on the plants which have long leaves, species of

Pandanus, various bromeliacious plants and particularly upon agaves and aloes.

Anolis maynardi Garman.

Garman, Bull. Essex inst., 1888, 20, p. 7.

This species, which is confined to Little Cayman Island, differs markedly
from A. porcatus of Cuba, its closest ally. Garman has mentioned the very
long narrow snout, and the rather weak longitudinal rugae of the head, which
separate the species from the shorter and more rugosely headed Cuban form.
Of the two types (M. C. Z., No. 6,227), only one is in useful condition, the other
having apparently been very badly dried and eaten by insects.

Anolis brunneus Corr.

Cork, Proc. Acad. nat. sci. Phila., 1894, p. 482, pl. 10, fig.3. Barsour, Proc. Biol. soc. Wash., 1911, 23,
p. 49.

Even though it is not absolutely certain that the common green Anolis of
the several Bahamas does not differ constantly upon the various islands, still
at present it seems best for reasons already given (Barbour, loc. cit.) to consider
these lizards together under this name, and so to distinguish them from their
allies in Cuba, A. porcatus, and on the American mainland, A. carolinensis.
Rosén, (Lunds univ. arsskrift, 1911, 7, no. 5, p. 30) although he had no Cuban
material for comparison, nevertheless quite independently arrived at a somewhat
similar solution of the vexed question regarding the status of these Bahaman
lizards.

Anolis oligaspis Corr.

CoprE, Proce. Acad. nat. sci. Phila., 1844, p. 480, pl. 11, fig. 5.

This species was described by Cope from an imperfect female taken on New
Providence Island. He compares it with A. krugi of Porto Rico, but remarks
that he has not seen the latter. On my trips to New Providence I have never
got this species, and have often doubted whether the type was not a displaced
specimen, wrongly credited to the Bahaman fauna. When, also, however,
Rosén’s account of what he determined as Anolis porcatus appeared, it seemed at
once probable that what he termed the “‘rare form”’ of this species was nothing

REPTILIA. 295

more or less than Cope’s A. oligaspis. A comparison of Rosén’s poor figure
(Lunds univ. arrskrift, 1911, 7, no. 5, p. 30, pl. 1—, fig. 8) with Cope’s makes it
appear as certain as possible under the circumstances that this species is an ex-
tremely rare one which occurs very sparingly on New Providence, whence the
type came, and Andros Island, whence Rosén brought two specimens. These
he confused with A. porcatus, which, in the Bahamas, should be called A. brun-
neus.

Anolis leucophaeus Garman.

Garman, Bull. Essex inst., 1888, 20, p. 109.

This species, of which the type (M.C. Z., No. 6,226) comes from Great.
Inagua, is by common consent considered the same as Cope’s A. cinnamomeus
(Proc. Acad. nat. sci. Phila., 1894, p. 435); and there seems to be no reason for
not including his A. moorei also in the synonymy (loc. cit., p. 433).

Anolis chlorocyanus Dum“érit et BIBRON.

Dumsmarit et Bisron, Erpét. gén., 1837, 4, p. 117. BotxLeneer, Cat. lizards Brit. mus., 1885, 2, p. 44.

Garman (Bull. Essex inst., 1887; 19, p. 24) retained this species as distinct
from A. coelestinus Cope (Proc. Acad. nat. sci. Phila., 1862, p. 177). An exami-
nation of his material, which consisted of two fine adults from Samana, San
Domingo, the types of A. coelestinus (M. C. Z., No. 2,347), and three others which
he collected at Tiburon, Haiti, forces the conclusion that these are really one and
the same species, although it must be confessed that the head scales indicate
that the specimens fall into two classes. There are, however, intermediates,
and for this reason I do not feel justified in keeping the species distinct. Bou-
lenger had already combined them. The form occurs upon Haiti only. Mr.
Mann collected one example at Diquini, Haiti.

Anolis pulchellus Dumsérit et Brsron.

Dumerit et Bisron, Erpét. gén., 1837, 4, p. 97. Sresecer, Rept. U.S. nat. mus. for 1902, 1904, p.
660, fig. 112-116.

Stejneger has shown that this species does not really come from Martinique,
the type locality mentioned by Duméril and Bibron, since Plée, who sent the
specimens to the Paris museum, sent them from Martinique, but collected in St.
Thomas and San Domingo. The former was undoubtedly the type locality.

The species is known from Porto Rico, St. Thomas, St. Croix, Tortola,
Just van Dyke, and Vieques. Garman records specimens collected by Acker-
mann at Port au Prince, Haiti; but Stejneger surmises, undoubtedly quite

296 BARBOUR: ZOOGEOGRAPHY.

correctly, that these may have been shipped from Haiti, but collected elsewhere.
There is no other Haitian record for the species. There is also a specimen in the
U.S. national museum labeled ‘“‘ Nassau, New Providence”; but Stejneger has
already noted that this was, of course, impossible, and that some mistake had
occurred in the labeling.

Norops ophiolepis (Corer).

Cop, Proc. Acad. nat. sci. Phila., 1861, p. 211. Bounmenesr, Cat. lizards Brit. mus., 1885, 2, p. 26.

A not uncommon species, confined to Cuba. Several other species occur
on the mainland of tropical America, but this is the only Antillean member of
the genus. Its presence in Cuba serves to emphasize the fact that this island is
far more intimately related to the mainland than any of the others.

My observations agree with those of Gundlach that this species is never
arboreal, but is found only among grass or rushes, its non-climbing habits beg
correlated with the absence of subdigital lamellae. At Soledad, near Cienfuegos,
this species was quite abundant in the tall grass of pasture land. It was difficult
to catch owing to its habit of dropping to the ground and hiding quickly. It was
also seen about Pinar del Rio, Madruga, and near Havana, where Mr. Samuel H.
Scudder procured the only specimen the Museum had prior to our visits.

Iguana rhinolopha WirGMaAnn.

Wreemann, Herpet. Mexicana, 1834, p. 44. Gray, Cat. lizards Brit. mus., 1845, p. 186. BouLENGcER,
Cat. lizards Brit. mus., 1885, 2, p. 190.
Iguanas were recorded from Grenada by Garman (Bull. Essex inst., 1887,
19, p. 50) as being intermediate between this species and J. tuberculata (Laurenti).
A careful revision of these two species, made with the aid of extensive collections
from many localities, will be necessary before their exact status can be settled.
That they are really distinct I have no doubt whatever, but as yet their ranges
can not be accurately defined. Stejneger suspects that if intermediates really
do exist, they may be explained by the fact that the species have been carried
about by human agency ‘‘in innumerable instances,” and that the intermediates
may be ‘“‘hybrids from introduced stock, or because of their geographic distribu-
tion” (ex litt.). I favor the latter explanation, as apparently the accidental
introduction of vertebrates by human agency is a far rarer phenomenon than is
often realized.
Boulenger gives the distribution of what he calls Iguana tuberculata var.
rhinolopha as Central America and West Indies. It was formerly widespread

REPTILIA. 297

among the Antilles, but it has suffered great diminution in numbers on account
of the ravages of the mongoose. It is so rare now in Grenada that neither Allen
nor Brues were able to get a single specimen, although they offered rewards
and searched for it diligently.

Reinhardt and Liitken record this species from St. Kitts and St. Lucia;
Garman, from Grenada, Saba, and St. Thomas; and Boulenger, from St. Lucia.

Iguana delicatissima LauRENTI.

LAURENTI, Syn. Rept., 1768, p. 48. BouLenerr, Cat. lizards Brit. mus., 1885, 2, p. 191.

This species has a curious, more or less haphazard distribution, since it occurs
upon Swan Island, Nevis, St. Bartholomew, and St. Martins; and Boulenger
includes ‘‘ Brazil”? within its range, based doubtless upon some wrongly labeled

specimen. Reinhardt and Liitken record it from Martinique and Guadeloupe.

Cyclura cyclura (Cuvier).

Cuvier, Reg. anim., 1829, 2, p.45. Cocrrau, Sagra’s Hist. Cuba. Rept., 1838, p. 96, pl. VI.

This species is confined to Cuba, the Isle of Pines and the neighboring Cays.
It has been confused by Boulenger with several well-marked forms. He has
included C. carinata Harlan, from Turks Island, C. baeolopha Cope, of Andros
Island, and C. collet Gray from Jamaica, in what, apparently for the sake of
avoiding the repetition of the generic name, he calls C. carinata (Cat. lizards
Brit. mus., 1885, 2, p. 193-195). The species from Turks, Andros Islands, and
Watlings Island I know from examination to be well-defined species, and differ-
ent from C. cyclura. This was pointed out by Stejneger when he described the
Watlings Island Iguana (C. rileyi Proc. Biol. soc. Wash., 1903, 16, p. 129). At
the time of writing the catalogue, Boulenger had only mounted specimens of
the Jamaican Iguana; and lizards stuffed after the manner of half a century ago
certainly lose their specific characters, as many stuffed reptiles in this Museum
prove. But to judge from Gosse’s description and beautiful figure, there can
be no doubt but that the Jamaican species is as valid as the rest, which certainly
are very distinet inter se. Dr. Stejneger recently told me that there is a Jamai-
can example in the United States national museum and that his examination
of this specimen had shown it to be quite distinct from the others.

I have learned that this species is very rare upon the mainland of Cuba
being only found in the most unfrequented localities. It has been exterminated
over most of the island because of its excellence as an article of food. It is said
still to be common on many of the Cayos lying off the coast of Cuba, especially

298 BARBOUR: ZOOGEOGRAPHY.

Cayo Romano, and some cayos near Manzanillo. It also occurs about the Pan
de Guajaibon, in the Sierra de Rangel, near Cabo San Antonio, Cabo Maisi,
and Cabo Cruz. The Museum has a fine specimen from the region of Guanta-
namo given by my friend Mr. C. T. Ramsden and another from the hills near
Santiago de Cuba collected by Capt. now Col. Wirt Robinson, U.S. A.

In 1888 Garman (Bull. Essex inst., 20, p. 105) recorded Cyclura nubila
Gray 1831, from the Caymans. He simply quoted Maynard’s field notes which
read ‘‘The Iguana occurs commonly in the cliffs of both this island (Cayman
Brac) and Little Cayman.” In 1911 the Museum received from Mr. W. W.
Brown the skin in alcohol of an Iguana from the Caymans with no data regarding
the island whence it came. The head is badly shot but it appears to be a speci-
men of Cyclura cyclura. It is not fully adult. Unless the type is still well
preserved Gray’s C. nubila must be disregarded, the description is vague and no
type locality is given.

Cyclura collei Gray.

Gray, Cat. lizards Brit. mus., 1845, p. 190. Barsour, Bull. M. C. Z., 1910, 52, p. 293.

Gosse (Proc. Zool. soc. London, 1848, p. 99, pl. 1) has contributed a beauti-
ful plate and good diagnosis of this species, now so rare. His long and well-
written account of its habits (loc. cit., and Naturalists’ sojourn in Jamaica, 1851,
p. 76-77) is as absorbing today as it was when first penned!

Cyclura baeolopha Corr.

Corr, Proc. Acad. nat. sci. Phila., 1861, p. 123. Barsour, Bull. M. C. Z., 1904, 46, p.58. SresNEGER,
The Bahama Islands, 1905, p. 334. Rosin, Lunds univ. arsskrift, 1911, 7, no. 5, p. 37.

This and the various other species of this characteristic West Indian genus
are probably all confined to the islands whence they have been described and
perhaps to neighbouring islets. The range of each species is narrowly circum-
seribed; and the various species may easily be distinguished inter se by a careful
reading of the descriptions, or better still, by comparisons of specimens. Bou-
lenger unfortunately relegated to the synonymy of what he called Cyclura
carinata Harlan (Cat. lizards Brit. mus., 1885, 2, p. 193) this as well as other valid
species; whereas he applied Harlan’s name to his Cuban and Jamaican specimens
when in reality the name belonged to the species peculiar to Turk’s Island.

Cyclura baeolopha is common about the central region of Andros Island,
whence Bryant obtained an excellent series in 1904. Rosén did not find it in

the northern part of the island.

REPTILIA. 299

Bryant reported that the natives about the village at Mangrove Cay hunted
the ‘‘guana”’ regularly with dogs as an esteemed article of food. I have, in fact,
seen specimens brought to the market at Nassau on several occasions,— always
said to be from Mangrove Cay. The natives twist loose one of the animals
claws, and drawing out the sinew attached to it, are thus able to tie the animal’s
feet together over its back, so that it is helpless. In the markets at Panama
heaps of helpless individuals of the common edible Iguana (Iguana iguana) may
be seen any day tied up in this way, and they find a ready sale at from five to
ten cents each (U.S. currency). The females full of eggs being most esteemed.

Cyclura rileyi SresnecER.

STEIJNEGER, Proc. Biol. soc. Wash., 1903, 16, p. 129. Rosén, Lunds univ. arsskrift, 1911, 7, no. 5, p. 36.

Known only from Watling’s Island. The field notes quoted by Stejneger
show that C. V. Riley, who discovered this species, found it common on two
small cays in the large saltwater lagoon on Watling’s Island. The larger island
of these two is known locally as Iguana Cay. This Museum’s collection now

contains one of Riley’s original series.

.

Cyclura cornuta (BONNATERRE).
BoNNATERRE, T'abl. encye. erpét., 1789, p. 40, pl. 4, fig. 4. Sresnecer, Rept. U.S. nat. mus. for 1902,
1904, p. 670, fig. 122-126.

The Rhinoceros Iguana is found upon Haiti, Mona, and Navassa Islands.
Stejneger has discussed the status of this species so fully that there is no object
in going into it here, since no new material has come to hand. The question as
to whether this name embraces more than a single species must be left until many

additional specimens can be studied.

Cyclura carinata Haran.

Haran, Journ. Acad. nat. sci. Phila., 1824, 4, p. 242, 250, pl. 15. Copn, Proc. U.S. nat. mus., 1887, 10,
p. 437.

A very distinct species, confined to Turk’s Island. Stejneger and Rosén
agree as to the status of this species, which was first pointed out by Cope with
relation to its congeners. There is a topotype in the Museum collected by Prof.
A. §. Bickmore, so that I have been able to make an actual comparison with

some of the other species.

300 BARBOUR: ZOOGEOGRAPHY.

Leiocephalus carinatus Gray.

Gray, Philos. mag., 1827, 2, p. 208. BouLencmr, Cat. lizards Brit. mus., 1885,

2, p. 165. Bargour,
Bull. M. C. Z., 1904, 46, p.58. Rosin, Lunds univ. arsskrift, 1911, 7, no. 5, p. :

A rather wide ranging form. Cuban and Bahaman individuals seem to be
indistinguishable. Known from Cuba, Great Bahama, Great Abaco, Stranger
Cay, New Providence, Andros, Cat, and Crooked Islands.

In Cuba, as in the Bahamas, this is a species characteristic of the zone of
vegetation bordering the ocean beach. It is very common, living in burrows in
the masonry of the walls of Morro Castle and Cabafias Fortress at Havana, and
at the Morro of Santiago de Cuba, as well as at the forts on the foreshore near
Matanzas.

Leiocephalus virescens STEJNEGER.

SrrsNeGER, Proc. U.S. nat. mus., 1900, 23, p.471. Rosin, Lunds univ. arsskrift, 1911, 7, no. 5, p. 35.

A local race confined to Green Cay, south of New Providence Island, de-

rived from L. carinatus.

Leiocephalus varius GarMAN.

GarMAn, Proc. Amer. phil. soc., 1887, 24, p. 274.

The types of this species are in the collection, well preserved (M. C. Z.,
No. 6,023). It is confined to Grand Cayman. On Cayman Brac Garman
recorded L. carinatus, but these specimens are not quite typical of that species.
No member of the genus has been taken on Little Cayman.

Leiocephalus loxogrammus Core.

Corr, Proc. U.S. nat. mus., 1887, 10, p. 437.

The type of this species came from Rum Cay, collected by the naturalists
of the U.S. F.C. 8. AtBatross. It was later taken at the same island by May-
nard and recorded by Garman. (Bull. Hssex inst., 1888, 20, p. 13). More
recently still Stejneger has reported it from Watling’s Island. It-is not known
elsewhere, but may be looked for upon others of the Bahamas as yet herpeto-
logically unexplored. etal
Leiocephalus-vittatus (Hattowsi1),

HAtiLowE .u, Proc. Acad. nat. sci. Phila., 1856, p. 151. BoutLmnamr, Cat. lizards Brit. mus., 1885, 2,
p. 163-164.

In 1909 I found this species very abundantly about Camaguey (Puerto

Principe) Cuba, and a series was obtained there. The Museum had other

REPTILIA. 301

specimens from Matanzas and Havana, but most unfortunately none from San
Domingo. I am unable to make comparison of Cuban and San Domingan
examples, which Boulenger groups together. It is confined to these two islands,
if the San Domingan records are not based on the young of L. personatus, which
look quite similar to this species.

In 1912 I found the species much more widespread than I had imagined and
specimens were secured from Havana, several localities in the Province of Pinar
del Rio, about Cienfuegos, near Madruga, Bayamo, Guantanamo and elsewhere.
The species was observed almost everywhere except in the immediate vicinity
of the seacoast where its place is taken by L. carinatus. Its most characteristic

- habitat is perhaps in the guarda rayas of the cane fields.

Leiocephalus macropus Cope.

Cork, Proc. Acad. nat. sci. Phila., 1862, p. 184. Bounencsr, Cat. lizards Brit. mus., 1885, 2, p. 163.

This seems to be a rare species confined to eastern Cuba. It was not repre-
sented in the collection until Capt. Wirt Robinson presented two specimens
taken by him near Santiago after the American occupation.

Cope (Proce. U. 8. nat. mus., 1887, 10, p. 438) says, ‘‘Boulenger, in the Vol.
II of the Catalogue of the lizards in the British Museum, regards L. raviceps as
the same as L. macropus. They are, however, different species, belonging to
different sections of the genus.’”’ Boulenger did not place L. raviceps as a syno-
nym of L. macropus, but of L. vittatus, a course apparently justified.

Leiocephalus schreibersii (GRAVENHORST).

GravennHorst, Nova acta Acad. caes. leop.-carol., 1836, 18, p. 739, pl. 54, fig. 15-16. BouLEnamr, Cat.
lizards Birt. mus., 1885, 2, p. 162.

A very well-marked species known from Haiti and Great Inagua. There
are specimens in the Museum from both localities. Mann’s recent series adds

seven from Manneville, Haiti.

Leiocephalus melanochlorus Corr.

Cop, Proc. Acad. nat. sci. Phila., 1862, p. 184. Bounenemr, Cat. lizards Brit. mus., 1885, 2, p. 164.

The typical series, about twenty examples collected near Jeremie, Haiti,
by Dr. Weinland (M. C. Z., No. 3,598), is still well preserved. Other specimens
have come from Puerto Plata, San Domingo, collected by Mr. M. A. Frazar,
and from Furey, Haiti, collected by Mr. W. M. Mann. Known only from this

island.

302 BARBOUR: ZOOGEOGRAPHY.

Leiocephalus personatus Corr.

Cork, Proc. Acad. nat. sci. Phila., 1862, p. 182. BouLmncmr, Cat. lizards Brit. mus., 1885, 2, p. 164.

The striate head scales and other characters, as well as the absence of the
white lateral band or series of white spots, make one feel justified in keeping
this species separate from L. vittatus, although Boulenger remarked (loc. cit.)
that they were perhaps not specifically distinct. The solution to be expected
would be that this species replaced L. vittatus on San Domingo; one having no
specimens of L. mttatus from San Domingo would form this view. Boulenger,
however, recorded such specimens in the British museum; but it is possible
that these are really referable to L. personatus. At the time of writing the Cata-
logue, he had only one Cuban ZL. wttatus for comparison.

The types of this species (M. C. Z., No. 3,615) are two examples in fair
preservation. More recently we have received it from Puerto Plata, San Do-
mingo, and from Furey, Haiti. Cope based his species L. trigeminatus on the
young of this species, as he himself admitted (Proc. U.S. nat. mus., 1887, 10,
p. 488), after Boulenger had placed L. trigeminatus in the synonymy of L. perso-
natus. Although both species were described from Weinland’s material, and the
types of both were said by Cope to be in the Museum of Comparative Zodélogy,
nevertheless the types of L. personatus only are to be found.

Leiocephalus eremitus Cope.

Coreg, Proc. Acad. nat. sci. Phila., 1868, p. 122. Boutmnesr, Cat. lizards Brit. mus., 1885, 2, p. 165.

A species confined to the small island of Navassa, near Haiti.

Leiocephalus herminieri (Duménrit et Brsron).

Dumérit ET Brsron, Erpét. gén., 1837, 4, p. 261. BouLenamr, Cat. lizards Brit. mus., 1885, 2, p. 166.

This species was said by Duméril and Bibron to have been sent to the Paris
museum from Trinidad by l Herminier, and from Martinique by Plée and Guyon.
Boulenger had only a skeleton from Martinique. Whether this creature has
been extirpated by the mongoose or not it is impossible to say. It is not repre-
sented in Garman’s collection from Martinique, nor does it appear to have been
recorded from there in recent years. It is not known from any other of the
Lesser Antilles, which would make it seem possible that it was a Trinidad species
that had been erroneously credited to the fauna of Martinique.

REPTILIA. 303

Celestus sagrae CocTEau.
Cocrnau, Sagra’s Hist. Cuba, Rept., 1838, p. 180, p. 20. BouLenamr, Cat. lizards Brit. mus., 1885, 2,
p. 293.

A species of which little is known. It is confined to Cuba. The only
examples which I have seen have been two, one obtained by exchange from Dr.
Stejneger of the U.S. national museum, which came from eastern Cuba. This
was the locality indicated for most of the collection of Mr. Charles Wright which
was made in the Sierra de Yateras, near Guantanamo. The second specimen
was given me by my friend, the Cuban botanist, Sr. Jesus Valdivia, whom I
have to thank for many favors. It was caught near the city of Havana. Mr.
William Palmer wrote me that he had met the species once in the Vuelta Abajo
region near San Cristobal. Gundlach mentions that the species is rare and
unknown to many natives. Those who know it call it ‘‘Culebrita de quatro
patas,”’ the four-footed snake.

Celestus sepoides (Gray).

Gray, Ann. mag. nat. hist., 1852, ser. 2,10, p.281. BouLmnamr, Cat. lizards Brit. mus., 1885, 2, p. 294.

This species, the only Celestus having four toes, has twice been named as
the type of a genus peculiar to Haiti. Once it was called Sauresia by Gray
(loc. cit.); and ten years later it was redescribed by Weinland as Enbryopus
habichit. (Abhandl. Senck. naturf. gesellsch., 1862, 4, p. 132, pl. 5, fig. 1).

« If this species occurred upon the mainland, the possession of four toes
would be quite sufficient for the erection of a genus to contain it; but for the
Haitian lizard, such a course only masks its relationships. It is in all other
respects a Celestus pure and simple, and represents the modified descendants of
individuals, left isolated in Haiti, whose relatives in Cuba have become Celestus
~ sagrae, and in Porto Rico, C. pleei. It should stand in the genus between these
two species.

Not recognizing this genus, I may be considered as inconsistent after what
I have said regarding keeping separate C. crusculus and C. costatus. To me it
seems that letting this distinction stand, we are going to just as unreasonable
an extreme as when several valid species are thrown together simply because
they look very much alike. Here we are over-emphasizing an isolation which is
just what Giinther has warned against when by merging species we may easily
over-emphasize the relationships of the islands represented by the material
being studied. Giinther and Boulenger,— the latter in handling Cyclura, for

304 BARBOUR: ZOOGEOGRAPHY.

. example,— have perhaps quite unconsciously over-emphasized the closeness of
the relationship of the Antilles with one another.

The name Sauresia may be retained as a subgeneric heading, if it is desirable
to point out the fact that the Haitian individuals of this group of individuals
have become more modified in structure from the generic type than have the
individuals on Cuba and Porto Rico.

Celestus pleii (Dumérit et Bisron).

Dumérit ET Brsron, Erpét. gén., 1839, 5, p. 605. Strrsnecer, Rept. U.S. nat. mus. for 1902, 1904,
p. 622.

A species rare, like all the others in the genus, because its ground-inhabiting
nature makes it an easy prey for the marauding mongoose. Stejneger has shown
that the type locality was beyond doubt erroneous (it was given as Martinique),
and that the species is confined to Porto Rico.

Celestus impressus Corr.

Cope, Proc. Acad. nat. sci. Phila., 1868, p.127. Barsour, Bull. M. C. Z., 1910, 52, p. 298.

A species which is still fairly abundant in the hilly country about Mandeville
in Jamaica. It is confined to that island.

Celestus occiduus (SHAW).

Suaw, Zoology, 1802, 3, p. 288. Barsour, Bull. M. C. Z., 1910, 52, p. 294.

In my Notes on the herpetology of Jamaica, I included within this species
C. striatus (Gray), and C’. hewardii (Gray). These species, probably on account
of very deficient material, had been considered distinct since their description,
and confined to Jamaica.

Garman (Bull. Essex inst., 1887, 19, p. 21) says that a “‘careful study of the
type of Cope’s D. stenurus convinces me that Dr. Boulenger is right in placing
it in D. striatus.” The type of Cope’s Diploglossus stenurus, from Jeremie,
Haiti (M. C. Z., No. 3,612), is a very large individual, for a long time the only
Haitian specimen of the species in the collection. It agrees closely with the
description given by Boulenger (Cat. lizards Brit. mus., 1885, 2, p. 289, pl. 16,
fig. 1) for Diploglossus striatus, which I believe to represent only a variation of
Celestus (Diploglossus) occiduus (Shaw). If, however, the small auditory open-
ing of which Boulenger speaks in his description of C. (D). striatus is really con-
stant, then the individuals having it may be entitled to specific rank. The ear
opening is small in the type of C. stenwrus (Cope); but this character shows itself

REPTILIA. 305

to be somewhat variable in the species, which is represented in the collection by —
good series of examples. The dorsal scales in Cope’s type have a slight tendency
toward mucronate condition. This may be due to old age; but if it proves to
be constant, and if it never occurs in senescent Jamaican individuals, the Haitian
form is really distinct. Unfortunately no such old Jamaican example is at hand.
Until the accumulation of much additional Haitian material it is better to keep
the forms separate. Mann’s two recent examples from Diquini are not old
individuals and do not settle the question of the identity or distinctness of these
two species.

Celestus rugosus Corr.

Cops, Proc. Amer. philos. soc., 1879, 18, p. 272. BouLmncer, Cat. lizards Brit. mus., 1885, 2, p. 288.

This species seems to represent C. occidwus (Shaw), of Jamaica, upon San
Domingo; as C. costatus Cope represents C. crusculus Garman. Boulenger,
judging from Cope’s description,— a course which I also have to follow,— allowed
C’. rugosus Cope to stand as a distinct species. There is no specimen in the col-
lection here; but the form seems to be a perfectly valid one, confined to San
Domingo.

Celestus crusculus (GARMAN).

Garman, Bull. Essex inst., 1887, 19, p.22. Bargour, Bull. M. C. Z., 1910, 52, p. 295.

Another beautiful species (Type M. C. Z., No. 6,051) which is apparently
disappearing fast. It is found only upon Jamaica.

In the Zoological record for 1888 Boulenger raised the question as to whether
this species was not Diploglossus phoxinus Cope (a species which should really
stand as Celestus costatus (Cope). As stated under the caption of this species,
I think pending the accumulation of more material these species should be kept
separate. Gunther (Ann. mag. nat. hist., 1888, ser. 6, 2, p. 364-65) believes
we should be most careful about admitting that the West Indian Islands are
populated by related but peculiar species. On commenting on Garman’s work
in naming many of these island forms, he says: ‘‘In this case any one studying
the distribution of reptiles over the West Indies would, by relying upon state-
ments such as are propounded by Mr. Garman in his recent [1887-1888] publica-
tions on West Indian reptiles be misled into the view of a more complete isola-
tion and specialization of the faunas of the various islands than obtains in reality.”’
It is in just this point of view that Giimther errs deeply. It is this very close
relation of the insular faunae that we must be most cautious not to exaggerate.

The presence of the same species upon neighboring islands is very significant

306 BARBOUR: ZOOGEOGRAPHY.

when true. The presence of allied species upon neighboring islands is what ex-
perience with West Indian reptiles teaches one to expect. To say that these
_ Celesti upon Jamaica and Haiti are the same when one can not be certain of
this from the material at hand is misrepresenting the state of affairs far more
probably than to hold them separate, when we find so few species really the same
upon the two islands, and so many types, if one may call them such, represented
on each of the two islands by related species. We are erring more when we over-
emphasize the similarity of the fauna of Jamaica and of Haiti; for the relation-
ship shows in its truest light when the differences between the species on the
islands are pointed out and held at their true value.

I pointed out in 1910 (loc. cit.). that Boulenger’s Diploglossus bakeri was
apparently this species. For the use of Celestus instead of Diploglossus, con-
sult the same paper.

Celestus maculatus (Garman).

Garman, Bull. Essex inst., 1888, 20, p. 105.

Boulenger, in the Zoological record for 1888, erroneously noted that this
species was a synonym of C. striatus (Gray) (= C. occidwus Shaw). Garman
correctly pointed out its relationship to his C. crusculus. It needs comparison
with no other species. It may be separated at once from C. crusculus; and plenty
of specimens, including the type, of this species are available for comparison.
The second and third digits are, as Garman pointed out, much elongated. They
are, in fact, nearly twice as long as those in any specimen of C. crusculus of the
same size.

The type is unique in the collection (M. C. Z., No. 6,231). It was collected
on the island of Cayman Brac. It is possible that the species might occur upon

other islands of the Cayman group.

Celestus costatus (Cops).

Corr, Proc. Acad. nat. sci. Phila., 1861, p. 494. Garman, Bull. Essex inst., 1887, 19, p. 23.

Garman (loc. cit.) was the first to point out that the type of Cope’s Pana-
lopus costatus was in reality a mutilated lizard of the same species, a perfect
specimen of which Cope later characterized as Celestus phoxinus. The earlier
type (M. C. Z., No. 3,606) was described as representing a strange genus having
legs without toes. I believe that a partly healed up groove around the lizard’s
body offers the key to why it was so mutilated. Some young Haitian caught the
lizard and tied a string tightly about its body, hence the groove; then he pro-
ceeded to snip off the fingers and toes — a form of amusement, or perhaps experi-
mentation not unknown to more civilized youth. The lizard then escaped, and

REPTILIA. 307

the limb stumps healed over. Finally, Dr. D. F. Weinland collected the lizard,
in due time it became the type of a supposed new genus.

This imperfect specimen, however great its historic value, is almost worth-
less for purposes of comparison with C. crusculus Garman, from Jamaica, its
very close ally. I believe it most unwise to conclude that these two species
are identical until a comparison can be made, based on more extensive collections
from Haiti. I have given reasons for wariness in making such conclusions under
the heading of C. crusculus Garman (p. 305).

If the toes had been cut off the type of C. maculatus, it would be quite indis-
tinguishable from the type of C. crusculus; but with the feet perfect, the species
may be recognized at once. Some such character may exist in C. costatus.

Cope said, when describing Celéstus phoxinus, (Proc. Acad. nat. sci. Phila.,
1868, p. 125) that “‘This elegant species was found by Dr. D. F. Weinland, near
Jeremie, Hayti, and was placed by him in the Museum of Comparative Zoology,
Cambridge, Mass., in care of Prof. Agassiz.” I have been quite unable to locate
the type, nor do I know whether Garman had it for comparison with C. costatus

when he announced the identity of the two species.

Cricosaura typica GunpLacH & PETERS.

GunpuacH & Peters, Monatsb. Akad. wiss. Berlin, 1863, p. 362, pl., fig. 1-6. Bocourt, Miss. sci.
Mex. Reptil., p. 313, pl. 20, fig. 14-18. Boutencsr, Cat. lizards Brit. mus., 1885, 2, p. 329.

This remarkable archaic monotypic genus has for herpetologists an interest
equivalent to that which is held for Solenodon among students of mammals.
Its excessive rarity, there does not seem to be a single example preserved in
an American museum, outside of Cuba, makes it the most-to-be-desired booty
of any naturalist who may collect in Cuba.

On the continent, its allies of the genus Xantusia were almost equally rare
until it was found that X. wgilis Baird lived by day among the drooping dead
leaves of the great tree yuccas of our southwestern deserts (Van Denburgh,
Occ. papers Cal. acad. sct., 1897, 5, p. 123-128). Xantusia henshawi Stejneger
lives among granite boulders, and can be found for a few moments before dark
only, as it does not come out from its hiding place until after sundown.

Lepidophyma, the only other genus of the Xantusiidae is also semi-nocturnal.
Sumichrast is quoted by Bocourt (Miss. Sci. Mex. Reptil., p. 311-312) as saying
that it lives in the darkest humid forests. Its habits are somewhat like those of a
gekko, living in the decayed galleries, or in the crevices of old tree trunks, much
more rarely among rocks. It is never seen abroad by day; but nevertheless is
not really rare, once one knows where and how to search for it.

308 BARBOUR: ZOOGEOGRAPHY.

There can be no doubt but that the habits of Cricosaura are somewhat
similar to those of Xantusia and Lepidophyma.

It has been customary to consider the members of this family as types
reminiscent of a fauna which has almost passed. The rarity of the individuals
of the various species has been brought forward as an argument that they are in
fact dying out. In view of the real abundance of the creatures when once their
hiding places become known, this argument ceases to be weighty. The scat-
tered ranges of the various forms, each narrowly circumscribed, and the structure
of the animals show that the group is, to be sure, an archaic one. The surviving
species seem, however, to have adapted themselves to a mode of life successful
under present conditions. Gundlach (Erpet. Cubana, 1880, p. 63) found all the
specimens known under stones near Cape Cruz, the southernmost point of the
island. This is all that is known of the habits of the species.

All of the foregoing was written before I visited Cuba again during 1913.
This trip included a special, and somewhat arduous, excursion to Cabo Cruz.
Prof. de la Torre and I went together and he had with him Gundlach’s manu-
script field notes which give in minute detail the exact places at each locality
visited where Gundlach’s actual collecting was carried on. By means of this
I was able to go to what I believe was the exact spot where Gundlach got his
specimens and to find again this most interesting form. I only got one example
under a large stone about one hundred yards a little north of east from the light-
house. Here there is a small area which is sunken and where red earth occurs
with scattered fragments of the country rock of limestone. The surroundings |
are sandy toward the point of the Cape and “‘diente perro” to the North and
East. Cacti and plants which are excessively thorny recall the habitat of Xan-
tusia on the mainland.

The species is more active than I had expected to find it and two specimens
were unfortunately lost by Cubans who were helping look for it, as they escaped
into crevices in the outcropping limestone here eroded into spear points and knife
blade edges which ring like metal, as one walks over them and which makes this
““diente-perro”’ country anathema to all Cubans.

Centropyx intermedius Gray.

Gray, Griffiths Cuvier’s Animal kingdom, 1830, 9, Syn. reptil., p. 31. BouLEenceEr, Cat. lizards Brit.
mus., 1885, 2, p. 340.

Cope first recorded this species from Barbados; and I cannot find the slight-
est difference between specimens from Barbados and the description of main-

REPTILIA. 309

land specimens,— I have none for actual comparison. Fielden (Zoologist, 1889,
p. 297) contributes an excellent account of this species, its habits, ete. It occurs
also in the Guianan subregion of South America.

Garman’s name, C. copii, based upon three specimens M. C. Z., No. 6,076

from Barbados, has no standing in nomenclature, since it is a nomen nudum.

Am>iva auberi Cocteau.

Coctrau, Sagra’s Hist. Cuba. Rept., 1843, p. 74, pl. 6. Boutmnarr, Cat. lizards Brit. mus., 1885, 2,
p. 356.

This species is widespread in Cuba, to which island it is confined. There
are specimens in the Museum from Santiago, Bahia Honda, and Puerto Principe.
Near the latter town I found it very common, running about over the sun-baked
ballast of the railroad line. Like its congeners, it prefers open, dry plains, with
plenty of exposed rock for basking. On my last trip to the island I found it
common near the beach between Morro Castle and Cojimar near Havana, and
met with it sparingly at San Diego de los Banos, Herradura, Pinar del Rio,
and Madruga.

Ameiva dorsalis Gray.

Gray, Ann. mag. nat. hist., 1838

,1,p BouLENGER, Cat. lizards Brit. mus., 1885, 2, p. 357. Bar-
BourR, Bull. M. C. Z., 1910, 52, p.

277.
298.

An interesting species, confined to the island of Jamaica, where it has
grown rare and difficult to obtain owing to the reptile-devouring habits of the
introduced mongoose.

Ameiva thoracica Cope.

Cops, Proe: Acad. nat. sci. Phila., 1862, p. 64. BovuLENGER, Cat. lizards, Brit. mus., 1885, 2, p. 356.
Rossin, Lunds univ. arsskrift, 1911, 7, no. 5, p. 37.

A common species throughout its range, which includes the Bahaman
Islands of New Providence, Eleuthera, and Andros. Cope (Proc. U. S. nat.
mus., 1887, 10, p. 438) records a specimen taken by Townsend (Cruise of the
ALBATROSS), at Great Abaco. Its occurrence on this island is nevertheless prob-
lematical. In 1904, Allen, Bryant, and I searched diligently for it at various
stations on the island, without result.

Ameiva chrysolaema Corr.
Cork, Proc. Acad. nat. sci. Phila., 1868, p. 127. Bouncer, Cat. lizards Brit. mus., 1885, 2, p. 355.

Confined to Haiti; a distinct form, related to the Bahaman A. thoracica.

Mr. W. M. Mann obtained many examples near Manneville.

310 BARBOUR: ZOOGEOGRAPHY.

Ameiva taeniura Corr.

Cope, Proc. Acad. nat. sci. Phila., 1862, p.63. Boutmnesr, Cat. lizards Brit. mus., 1885, 2, p. 350.

The types, three fine specimens (M. C. Z., No. 3,614), were the only examples
in the collection until the receipt of many from Diquini and Manneville collected
by Mr. W. M. Mann. ‘The species is confined to Haiti.

Ameiva vittipunctata Copn.

Corr, Proc. Acad. nat. sci. Phila., 1871, p. 220. Bounmneer, Cat. lizards Brit. mus., 1885, 2, p. 355.

The types of this species came from near the city of San Domingo. Re-
cently, however, the Museum has received a fine suite of specimens from Mo-
mance, Diquini, and Manneville, collected by Mr. W. M. Mann.

Ameiva maynardi Garman.

Garman, Bull. Essex inst., 1888, 20, p. 110.

This form, derived from Haiti, occurs upon Great Inagua in the southern
Bahamas. The types are M. C. Z., No. 6,225.

Ameiva exul Cope.

Corr, Proc. Acad. nat. sci. Phila., 1862, p.66. StrsnecEr, Rept. U.S. nat. mus. for 1902, 1904, p. 612,
fig. 59-66.

This ground lizard is rather more wide-ranging than many of the species
of the genus. Various students, notably Stejneger, have compared specimens
from the different islands; and they find them apparently indistinguishable
upon Porto Rico, Vieques, St. Thomas, St. John, St. Croix, and Water Island.
Nevertheless, Garman (Bull. Essex inst., 1887, 19, p. 11) says, ‘‘Porto Rico
specimens appear to be a little lighter in color, more red on the back and head,
and have the whitish frecklings on the hinder part of the body and the base of
the tail, as also the dark spots along the flank, less numerous and more distinct
than those from St. Thomas. The latter have the colors a trifle darker, more
olive, and the white spots and the black spots less faded.”’ I have re-examined
the specimens, now more than twenty years older than when Garman studied
them, and I find the difference to be really striking. The series are probably
very different in life, and probably represent groups of individuals strongly
tending toward the formation of distinct species by isolation, by the constantly
increasing tendency of an isolated group of individuals to vary according to

latent peculiarities which become in succeeding generations increasingly patent.

REPTILIA. dll

Stejneger, however, with an enormous series of Porto Rican specimens, empha-

sizes the great variability in coloring which he finds.

Ameiva alboguttata BouLeNcErR.
BoutEnGeER, Jahresb. Naturw. verein Magdeburg, fiir. 1894-1896, 1896, p. 112. Sresnecer, Rept.
U.S. nat. mus. for 1902, 1904, p. 618, fig. 67-72.
A species evolved through the isolation on Mona Island of individuals simi-
lar to those which gave rise, on Porto Rico, to A. exul.

Ameiva lineolata Duménrit et Breron.

Dumérit Et Bisron, Erpét. gén., 1839, 5, p. 119. Bourmnerr, Cat. lizards Brit. mus., 1885, 2, p. 349.

A beautiful species, confined to Haiti, where good series were collected by
Dr. Weinland, near Jeremie, and by Mr. Frazar, near Puerto Plata, at the oppo-
site extreme of the island. This would suggest that it occurs widely spread.

Ameiva polops Cope.
Cops, Proc. Acad. nat. sci. Phila., 1862, p.66. BouLenecmr, Cat. lizards Brit. mus., 1885, 2, p. 350.

A well-defined species, confined to the island of St. Croix.

Ameiva wetmorei STEJNEGER.

STEIJNEGER, Proc. Biol. soc. Wash., 1913, 26, p. 69.

This species so recently discovered stands directly with the lineolata-polops
group and serves to bridge the long supposed discontinuity of the range of these
allied species. The finding of this important lizard as Stejneger says ‘‘illustrates
once more the fortuitous nature of reptile collecting.”’ Porto Rico, the island
whence this species comes was of all the Antilles presumably the best known.

Ameiva corvina Corr.

Corr, Proc. Acad. nat. sci. Phila., 1861, p. 312. BouLencerr, Cat. lizards Brit. mus., 1885, 2, p. 358.

The type of this species is a single specimen from the island of Sombrero
(M. C. Z., No. 5,532). It has since been taken upon Anguilla Island, whence
the British museum had specimens (Boulenger, loc. cit.). Garman (Bull. Hssex
inst., 1887, 19, p. 10) mentions that between the type and specimens ‘“‘from
Haiti there is apparently little difference.” This is perfectly true; but there
can be no doubt that M.C. Z., No. 3,616, four specimens said to have been
taken at Jeremie, Haiti, by Weinland, are really not from Haiti. The catalogue
refers to their having had an “‘old number’”’; and probably in the process of re-

312 BARBOUR: ZOOGEOGRAPHY.

cataloguing, or while sent to Cope for identification, they may have been mixed
with Weinland’s collection, and finally labeled ‘‘ Haiti.”

Ameiva atrata Garman.

Garman, Bull. Essex inst., 1887, 19, p. 8.

This is a distinct species, related to A. corvina, but confined, so far as known,
to the island of Redonda. The type is M. C. Z., No. 6,084.

° D ANA
Ameiva analifera Corr.

Corr, Proc. Amer. philos. soc., 1869, 11, p. 158.

This species is allied to A. pleei, of Martinique, and through it to A. suri-
namensis. The types came from St. Martins and St. Bartholomew. The
Museum has a number from the latter island.

Ameiva garmani, sp. nov.

Type: — No. 6,141, M.C. Z., Anguilla, F. Lagois, collector.

In 1887 Garman noted, but did not describe, this specimen from Anguilla
as a variety of A. analifera Cope. A re-examination of the specimens of Cope’s
A. analifera, from St. Bartholomew, shows that A. garmani differs from the
other examples in several characters which the series from St. Bartholomew
show to be constant. These differences, taken in connection with the very
markedly different coloration, prove that the Anguilla form constitutes a dis-
tinctly valid species. In species of Ameiva, the unvarying color pattern and
color of individuals of the same age is an excellent diagnostic character.

Ameiva garmani is similar to A. analifera Cope, but differs in the following
characters:— median gulars larger on the average in A. garmanz than in A.
analifera, and differently arranged, in that large and small scales occur irregu-
larly, while in the St. Bartholomew specimens the scales are evenly graded in size,
and are regularly arranged. In A. garmani, the largest of the outer tibials is very
different in size from that of the other species, being not only larger, but having
its vertical and horizontal dimensions almost equal, while in the other specimens
this scale is much longer than high. There are corresponding differences in
size and position of the other enlarged tibials owing to their filling an area about
a scale of such different character. As Garman (Bull. Essex inst., 1887, 19,
p. 10) said, this species is ‘‘easily distinguished” from its ally “by the color;
lighter brown anteriorly, with large light gray or olive spots posteriorly, which
gives the hind legs the appearance of being gray reticulated with brown,” in-

REPTILIA. 313

stead of dark brown with gray fleckings as seen in the other species. The tail
is light brown, much blotched with gray, and with scattered black dots covering
a scale each, instead of dark brown with quite widely scattered light specks.

Ameiva erythrops Cope.

Corr, Proce. Acad. nat. sci. Phila., 1871, p. 221.

Garman remarks that this form ‘‘is not entitled to more than varietal dis-
tinction”’; nevertheless, owing to the fact that this is an insular race, and that
there can never be any geographic continuity of ranges, a binomial is employed
for its designation. It is confined to St. Eustatius. Boulenger (Cat. lizards
Brit. mus., 1885, 2, p. 359) considers this a synonym of Gray’s A. punctata and
A. major, for he considers these both one species. The types of these forms came
from the Guianas, and those examined by Duméril and Bibron (Erpét. gén.,
1839, 5, p. 117) came from Cayenne and Trinidad; so that this species may be
at once erased from the Antillean list, as specimens from Guiana and St. Husta-
tius might reasonably be surmised to be specifically distinct.

Ameiva erythrocephala (Daupin).

Daupw, Hist. nat. rept., 1802, 3, p. 22. Garman, Bull. Essex inst., 1887, 19, p. 9.

Daudin expressly states that his description is based on specimens from the
island of St. Christopher, whence came the large series upon which Garman
reported. Boulenger in the Zoological record for 1887, says that this is what
he called A. punctata. I have discussed this mainland species under A. erythrops
Cope (ante). If this lizard from St. Eustatius should prove to be the same as
that from St. Christopher, then Daudin’s name is applicable for both, a fact
which Boulenger overlooked when writing the catalogue. There are no speci-
mens from St. Eustatius in the collection; but from description, I judge the
species found there to be distinct, though very closely related.

Ameiva pleii Dumérit et Brsron.
Doumfrit et Brsron, Erpét. gén., 1839, 5, p. 114. Bourenaer, Cat. lizards Brit. mus., 1885, 2, p. 354.
A species confined to Martinique. It is related to the South American A.

surinamensis.

Ameiva pluvianotata GarMAN.

GarMan, Bull. Essex inst., 1887, 19, p. 6.

A beautiful species, from Montserrat. The types are M. C. Z., No. 6,086.

314 BARBOUR: ZOOGEOGRAPHY.

Ameiva fuscata GARMAN.

GarMAN, Bull. Essex inst., 1887, 19, p. 5.

This fine lizard is confined to the island of Dominica. The types are M. C. Z.,
No. 6,087. Giinther (Ann. mag. nat. hist., 1888, ser. 6, 2, p. 363) says of this
species, ‘Closely allied to A. surinamensis; the Dominica form may be kept dis-
tinct on account of the modified scutellation of the fore and hind feet. The large
scutes of the forearm are separated from the rows of scutella of the toes by broad
areas of very minute scales.’??’ From this, it will be seen that the species is a
very distinct one:

Ameiva aquilina Garman.

GarMaN, Bull. Essex inst., 1887, 19, p. 3.

This species is really close to A. swrinamensis, as would be expected from the
situation of the islands which this species inhabits, Grenada and St. Vincent.
The greater number of femoral pores and a number of other characters, some of
which, though perhaps not all, may be found diagnostic in each individual, so
that the specimens may always be separated quite easily. The types are
M. C. Z., No. 6,088 and 6,089.

Dr. Allen procured eight specimens, mostly about St. George, Grenada.
He was told by local observers that some years ago the species almost disappeared
on account of the mongoose, but that at the present time it appeared to be gain-
ing considerably in numbers. This is very interesting, since I learned in Jamaica
that exactly the same thing had occurred there, with Ameiva dorsalis Gray.
This would indicate that some sort of a balance is being reached, compensating
for the introduction of the mongoose. The lizards may have gradually changed
their habits, though this does not appear from field observations; or they may
have grown shyer. They are certainly very wary now, and can usually be got
only by shooting.

Tiaporus fuliginosus Corr.

Corr, Proc. Amer. philos. soc., 1892, 30, p. 132, pl.

This lizard, which Cope states stands in the same relation to Ameiva as
Monoplocus does to Centropyx, is said to have been found only on Swan Island.
The monotypic genus is thus very narrowly restricted. I had the opportunity
to study the types at the U.S. national museum, by the courtesy of Dr. Stejneger.

Since this was written Mr. George Nelson has made several collecting trips
to Swan Islands. He reports that no such lizard as this Ameiva-like Tiaporus
exists on the Islands. The Jamaicans and Cayman Islanders living at Swan

a...

REPTILIA. 315

Island do not know of any such species, though they are thoroughly familiar
with the local fauna. This ground inhabiting species has probably been exter-
minated by the introduced cats. No cats are found feral upon Little Swan
Island, only upon Swan Island, but the lizard may have occurred upon this
island only.

Scolecosarus alleni, sp. noy.

Plate, fig. 11-15.

The fact that this curious burrowing teid genus occurred on Grenada was
first made known by Garman (Bull. Essex inst., 1887, 19, p. 12). He obtained
a single specimen there, and identified it with Scolecosaurus cuviert Fitz. As a
matter of fact, this name should really stand as Scolecosaurus cuviert (Dum. and
Bibr.), all previous references to the species being nomina nuda, or unrecogniz-
able descriptions. A careful examination of the seven specimens obtained by
Allen and Brues shows that, as might well be expected, the Grenadian examples
really are distinct from those which Duméril and Bibron described; and which
came from Colombia. It is said to occur in Venezuela, and more recently it
has been recorded from Trinidad by Mole and Urich (Journ. Trinidad field. nat.
club, 1894, 2, p. 82). Specimens collected recently by Dr. Thaxter enable me
to compare the species from Grenada and Trinidad.

Type:— No. 7,793, M. C. Z., an adult, St. George’s, Grenada, G. M. Allen,
collector. :

Upper head shields consisting of a pentagonal internasal, a pair of prefrontals
forming a short suture, a frontal considerably longer than broad, heptagonal,

a pair of large parietals, a narrow quadrangular interparietal of equal length,

™ and four supraoculars. Nasal more than twice as broad as high; loreal about

“one half the size of nasal; the longest infraorbital in contact with the third,

fourth, and fifth supralabials; eight temporals; six upper labials, third and last
largest, and of almost equal size; six lower labials; mental equal in size to one
of the first pair of labials next to it; first pair of chin shields posterior to the large
postmental, forming a wide suture; collar shields enlarged, eight in number.
Twenty-eight scales around middle of body, and forty from occiput to base of tail;
ventrals somewhat longer than broad, forming eight longitudinal series, and
twenty-six transverse series from the two enlarged pectorals to the preanals;
these are four in number, and the median posterior is the smallest. Tail long,
rounded at the end; covered with imbricate elongate hexagonal scales, the upper
smooth, the lower carinated. Reddish brown above, lighter beneath. Varia-

316 BARBOUR: ZOOGEOGRAPHY.

tions:— Some specimens show traces of a light, dark-bordered lateral line on the
body, in several this is fairly distinct on the tail. There is considerable variation
in the supralabials; in one, the fifth and sixth are of about equal size and slightly
larger than the third. This specimen has but five infralabials.

This species may be’distinguished from S. cuwert (Dum. & Bibr.) by the
possession of any one or more of the following characters (they may not all be
unvarying or equally diagnostic in the same individual) :— The frontal is usually
distinctly longer than broad; the last supralabial is not the largest; there are
more than four infralabials; the dark-bordered light line is almost never present
on the body, but it may be evident on the tail. Duméril and Bibron (Erpét.
gén., 1839, 5, p. 455) speak of seven upper labials of which the fifth is a scalene
triangle; in S. alleni the fifth is pentagonal. Boulenger mentions but six upper
labials; it is somewhat difficult to determine just where labials end and contour
scales begin. Again, Duméril and Bibron characterize the Colombian speci-
mens as having twenty-six rows of scales, and six rows of ventrals. Boulenger,
Garman, and IJ all agree on twenty-eight rows about the middle of the body.

Except Trinidad, Grenada is the only island in the West Indies where this
genus is found.

Scolecosaurus trinitatis, sp. nov.

Type: — No. 8,947, M. C. Z., an adult, Caparo, Trinidad, A. B. Carr,
collector. Found under fallen cocao pods. Eight paratypes bear the same
data.

The collection given recently by Professor Thaxter contains nine of these
lizards from Trinidad and ten from Grenada. For the first time it is possible to
compare a series from these islands. They represent a new species closely re-
lated, both seeming to differ in the same manner from the Columbian S. cuviert.
In describing the previous species I have pointed out these different characters.
In Grenadian specimens the prefrontals, form a short suture, this is the condi-
tion which the published records also indicate for the mainland specimens.
The specimens from Trinidad, on the other hand, seem to have the prefrontals
invariably well separated, and not forming a suture. The series of nine all show
this condition. Thus, even if it should not prove to be absolutely constant it is
at least the characteristic condition and as the opposite seems equally constant
in the Grenadian specimens, the Trinidad form is entitled to specific rank.

REPTILIA. 317

Gymnophthalmus pleii Bocourt.

Bocourt, Miss. sci. Mex. Reptil., 1881, p. 473, pl. 22, H, fig. 3. Bounmnecerr, Cat. lizards Brit. mus.
19, 1885, 2, p. 429.

Garman (Bull. Essex inst., 1887, 19, p. 12) records the capture of this species
at Martinique, the type locality, and at St. Lucia, whence G. lwetkenii is supposed
to have come. He says, ‘‘Bocourt gives St. Lucia as the locality for G. litkenz;
we failed to secure a specimen in all our collecting.”

Boulenger (Proc. Zool. soc. London, 1891, p. 353) writes: “Like Mr. Gar-
man I find the St. Lucia specimens to be referable to G. pleii, described by Bocourt
as from Martinique, and not to G. luetkeniz, of the same author, from St. Lucia.”

It is thus evident that this species, G. plew, is apparently the only one found
on both islands, in which case the locality for the type of the other species was
undoubtedly incorrect.

Amphisbaena punctata Br.

BELL, Zool. journ., 1828, 3, p. 236. Boutrnerr, Cat. lizards Brit. mus., 1885, 2, p. 450, pl. 23, fig. 3.

This little-known species is recorded only from Cuba. I succeeded in pro-
curing two specimens, one at San Diego de los Bafios, and the other near Herra-
dura. Both were turned out in ploughing. This species grows to be larger and
is much more light colored in life than the following. The few specimens of this
genus which I saw in the Museums in Havana all belonged to this species. It
may be confined to the region from the Province of Matanzas, west to Pinar
del Rio.

Amphisbaena cubana PETERs.

Perers, Monatsb. Akad. wiss. Berlin, 1878, p. 780, pl. —, fig.4. Boutenerr, Cat. lizards Brit. mus.,
1885, 2, p. 446.

Stejneger has discussed the question as to the distinctness of this species
under the heading of A. caeca in Porto Rico (Rept. U. 8. nat. mus. for 1902,
1904, p. 676). There can be but little doubt of its distinctness. It is confined
to Cuba.

Since the above was written I have been able to examine twenty-six speci-
mens obtained by following plows in the cane fields of the Soledad estate, near
Cienfuegos. All of these have the ocular fused with the second labial, so that
there can be no doubt as to the stability of this character. Boulenger was in-
clined to regard this condition as anomalous and to unite this species with A.
caeca. ‘There is, however, no reason whatever for doing this. In the region
about Cienfuegos, where these specimens were found I did not find a single

318 BARBOUR: ZOOGEOGRAPHY.

example of A. punctata. The people, however, led me to believe that another
“Culebrita ciega” occurred which was not T'yphlops lwmbricalis. The latter is
often called ‘‘vibora.”’

Amphisbaena innocens WEINLAND.

WEINLAND, Abh. Senck. naturf. gesellsch., 1862, 4, p. 137, pl. 5, fig. 2.

Weinland states that he captured three specimens near Jeremie, Haiti; and
at the end of the description says that one was sent to the Berlin Museum,—
whether before or after the description was written does not appear. At any
rate it is quite certain that M.C. Z., No. 3,624 and 3,625 are the other two speci-
mens of the three captured, since they bear the label ‘‘Jeremie, Haiti,-collected
by Dr. D. F. Weinland.” They were received when the Weinland collection
was purchased during Louis Agassiz’s administration.

Regarding the distinctness of this species from A. caeca I can only say that
both specimens have two instead of three scales behind the unpaired postmental,
and that their number of body rings is low, 212 and 213. Since these characters
fall within the range of what Stejneger (Rept. U. S. nat. mus. for 1902, 1904,
p. 677) has suggested is probably diagnostic for the species, there can be no
doubt but that there is now sufficient evidence as to the stability of these char-_
acters to make it quite certain that this is a valid species.

In the Zoological record for 1865 (p. 149) Giinther, after reviewing coe S
Catalogue of the Amphisbaenidae, which appeared that year, makes the note
that ‘“we understand that the A. innocens (Weinland) has been previously de-
seribed.”’ Gray (Cat. shield rept., 1872, 2, p. 38), although admitting this

cc

species to his list, repeats that it is “‘understood to be previously described.”
It is difficult to understand how this apparently erroneous idea can have arisen,
unless some obscure publication has escaped all recent herpetologists.

Since all the above was written Mann’s collection has brought a typical
example from Manneville, Haiti. This shows all the peculiar characters of the
types.

Amphisbaena manni, sp. nov.

Type: — No. 8,645, M. C. Z., Cape Haitien, Haiti, W. M. Mann, collector.
Three paratypes from the same locality.

This is the Haitian representative of A. caeca of Porto Rico. It resembles
this species in having a similar arrangement of head and chin shields. The series
shows that this condition is quite unstable but probably equally so in A. caeca.

The number of dorsal and ventral segments of a single annulus of scutes is the

REPTILIA. 319

same in both species being 16 above and 18 below. The species, however, really
differ widely, in that A. manni has eight femoral pores and A. caeca only four.
The new Haitian form has a greater number of rings on the tail, 24, against an
average of 17. The number of body rings is high, 227. This is exceeded by an
occasional specimen from Porto Rico.

The paratypes show that the number of femoral pores is regularly high.
Two having eight and the other (abnormal in other respects) but seven. The
tail rings number twenty-three to twenty-five except in this abnormal example
in which the tail is much reduced there being only seven rings. It shows no sign
of injury and is probably simply malformed! It is hard to imagine how an
Amphisbaena could have had its tail accidentally removed.

The apparent anomaly in distribution presented by the fact that there were
two species of Amphisbaena on Cuba and Porto Rico and but one on Haiti is
now cleared up. Before Mr. Mann left for Haiti I suggested that diligent
search would reveal a new Amphisbaena, and it is fitting that it should bear his
name.

Amphisbaena caeca CuvirErR.
Cuvier, Reg. anim., 1829, 2, p. 73. Strmsraur, Rept. U.S. nat. mus. for 1902, 1904, p. 676, fig. 129-
132.

The habits, habitat, and relationships of this species have been fully dis-
cussed by Stejneger. The species is confined to Porto Rico.

Amphisbaena bakeri STeJNEGER.

SrEJNEGER, Rept. U.S. nat. mus. for 1902, 1904, p. 681, fig. 183.

A well-defined species, confined to Porto Rico where its distribution seems
to be very local. The’only specimens known which bear a definite locality
record are from Lares. This is in sharp contradistinction to the distribution of

A. caeca, which occurs widespread over the entire island.

Amphisbaena fenestrata Cope.

Cops, Proc. Acad. nat. sci. Phila., 1861, p. 76. Boutenenr, Cat. lizards Brit. mus., 1885, 2, p. 449.

This distinct species is confined to the Virgin Islands. The types came
from St. Thomas and St. Croix. It has since been recorded from St. John, which
with St. Thomas formed the type locality of the synonymous species, A. antil-
lensis Reinhardt & Liitken.

320 BARBOUR: ZOOGEOGRAPHY.

Mabuya spilonota WiraMann.

Wiremann, Arch. f. naturg., 1837, p. 135.

This Jamaican species should bear this name, as has been shown by Stejneger,
who agrees with Garman as to its validity. (Stejneger, Rept. U.S. nat. mus. for
1902, 1904, p. 608). This is the species of which I (Bull. M. C. Z., 1910, 52,
p. 299) gave a short account regarding its habits, abundance, and distribution,
unfortunately using the name M. sloanii, inadvertently following Garman instead
of Stejneger, although the latter is obviously correct in the application of these
names.

Mabuya sloanii (Daupin).

Daupiy, Hist. nat. rept., 1803, 4, p. 287, pl. 55, fig.2. Srmemrcrr, Rept. U.S nat. mus. for 1902, 1904,
p. 608, fig. 56-58.

Stejneger (loc. cit.) has given excellent reasons for using this name for the
Porto Rican Mabuya, and also for considering this form distinct from that
upon Jamaica. This species, in common with its other generic relatives, is
one of the very first species to succumb to the ravages of the mongoose. In
Porto Rico it was so rare that Stejneger and the other naturalists with him did
not find a single example. This has been the experience of many other collectors
upon other islands. A regrettable fact since large series of specimens are particu-
larly desirable to settle definitely the many questions regarding variation within
the species of these variable forms, and to determine how distinct the various
insular races really are.

The range of Mabuya sloanwi comprises the islands of Vieques, Porto Rico,
Mona, St. Thomas, St. John, Just van Dyke, St. Croix, and besides this the
limited material indicates that Haitian specimens should be referred to the
same species. .

Mabuya mabouia Dumé@rm er Breron.

Dumérit et Brsron, Erpét. gén., 1839, 5, p. 646. Garman, Bull. Essex inst., 1887, 19, p. 52.

The specimens in the Museum are from Martinique. The describers state
that they have received it also from Guadeloupe. It is not probable, however,
that the scines upon these two islands should be alike, and those upon Dominica
different, when the latter island lies exactly between them.

Mabuya lanceolata Corr.

CorE, Proce. Acad. nat. sci. Phila., 1862, p. 187.

This species was based upon Barbadian specimens collected by Theodore
Gill. I have seen no specimens and cannot comment upon the validity of the
species.

REPTILIA. 321

Mabuya dominicana Garman.

GarMAN, Bull. Essex inst., 1887, 19, p. 52.

“Mr. Garman describes specimens from Dominica as a distinct species,
said to be distinguished from M. agilis by having the supranasals separate from
each other, and possessing from 68 to 72 scales in a series between the chin and
vent. His M. agilis is stated to have the supranasals in contact with each other
and only 54 or 56 scales between chin and vent.

The eight adult specimens from Dominica before me vary in both these
respects; some have the supranasals in contact, others not. Between chin and
vent there are 60 scales in two, 62 in one, 63 in two, and 64, 65, and 67 severally
in single specimens. Therefore the characters on which Mr. Garman based his
distinction are in these specimens so obviously variable that no herpetologist
will place any reliance upon them.

But to prove the variability of these characters in this species beyond further
dispute I took from a gravid female six embryos, all fully developed and about
half the length of the mother. The mother had the supranasals in contact with
each other and 62 scales between chin and vent. Of her progeny two had
the supranasals as in the mother, in three they were separate from each other,
whilst one might be assigned to either category. The scales on the abdomen
are in 57 rows in two of these embryos and in 61, 62, 65, and 66 rows severally
in their brothers.

Specific distinctions in these days are held to be, and often may be, matters
of individual opinion and, as a rule, I abstain from entering into any discussion
about them; but they sometimes have a direct bearing upon wider and more
important questions. In this case any one studying the distribution of reptiles
over the West Indies would, by relying on statements such as are propounded
by Mr. Garman in his recent publications on West-Indian reptiles, be misled
into the view of a more complete isolation and specialization of the faunas of
the various islands than obtains in reality. He states in fact in this instance
that the widely distributed Mabuia agilis has been sufficiently differentiated in
Dominica to form a distinct species or whatever it may be called, whilst the
examination of even a small number of examples disproves this statement.
Distinctive characters, no matter how trivial they appear to be, become impor-
tant enough to the systematic zoologist if they be found constant in a number of
specimens and correlated to some other point of the life of an animal; but unless
this has been ascertained to be the fact, their indiscriminate use impedes rather
than advances zoology.”

bo

BARBOUR: ZOOGEOGRAPHY.

(oh)
bo

Ginther’s remarks, which I have quoted at length, apply far better to his
own attitude than to that of Garman; he has not shown at all that the range
of variation for M. dominica, which he had increased, touches the range for
that of M. agilis; while the average condition seen in the two categories of
individuals show how distinct they are. Invidious remarks regarding the sense
of specific discrimination are always dangerous.

The types are M. C. Z., No. 6,049.

Mabuya luciae Garman.

GarMaN, Bull. Essex inst., 1887, 19, p.51. Boutencrr, Proc. Zool. soe. London, 1891, p. 353.

While considering this species as a variety of M. agilis, Boulenger admits
its subspecific rank. For the sake of consistency it is here considered a full
species; there being, however, obviously no question as to its origin from this

mainland species. It occurs upon St. Lucia. The types are M. C. Z., No. 6,046.

Mabuya aenea Gray.
Gray, Griffiths Cuvier’s Animal kingdom, 1831, 9, Syn. reptil., p. 70. Garman, Bull. Essex inst.,
1887, 19, p. 53.

Garman considers specimens from St. Vincent, Grenada, and Trinidad dis-
tinct from M. aurata Schneider, from southern Central America to Brazil.
Boulenger (Proc. Zool. soc. London, 1891, p. 355) considers the Antillean speci-
mens referable to the latter species. If, when it becomes possible to study large
series, the characters which Garman points out are found to be constant, then
each of the two groups of individuals will be obviously entitled to hold specific
rank.

Typhlops lumbricalis Linns.

Linné, Syst. nat. ed. 10, 1758, 1, p.228. Srersnecer, Rept. U.S. nat. mus. for 1902, 1904, p. 684.

This species extends from the region of the Guianas to both the Lesser and
Greater Antilles. It has long been known from many islands, among them
Dominica, St. Kitts, Antigua, the Virgin Islands, Guadeloupe, Martinique,
Mona, Porto Rico, Haiti, Jamaica, Cuba, and some of the Bahamas. In the
latter group it was first recorded by Cope (Proce. U.S. nat. mus., 1887, 10, p. 439)
from Great Abaco. In 1904 I found it again on the same island (Bull. M. C. Z.,
1904, 46, p. 59). Curiously enough it was not found on either New Providence
or Andros Island until Rosén collected specimens at both localities (Rosén,
Lunds univ. arsskrift, 1911, 7, no. 5, p. 37). In Cuba and Jamaica its distribu-
tion is very wide, and I have found it at practically every station visited.

REPTILIA.

(Je)
bo
ww

Typhlops pusillus, sp. nov.

Type: — No. 8,719, M. C. Z., Cape Haitien, Haiti, W. M. Mann, collector.
Four paratypes, two from Grand Riviere, one from Cape Haitien and one from
Ennery,: Haiti.

Similar to 7. lwmbricalis in color and number of rows of scales, which are
twenty in number, but widely different from this and the other West Indian
species in details of cephalic squamation.

Snout depressed and somewhat projecting; nostrils below lateral horizontal
edge; rostral a little less than two sevenths the
width of the head, not extending backward to the

line of the eyes; details of head shields as shown

in drawings. Scales in twenty rows; about 370
scales on midventral line from chin to vent, and _—‘Top and side view of head of type
nineteen under tail, which ends in a spine. Color ae ep
brown; edges of scales darker than their centres; ventral surfaces cream color.

Size of type, about five inches. '

Typhlops tenuis Savin.
Savin, Proe. Toole eae London, 1860, p. 454. BouLenarr, Cat. snakes Brit. mus., 1893, 1, p. 28.
Rosén, Lunds univ. arsskrift, 1911, 7, no. 5, p. 37.

Rosén found a single specimen which he identified with this species at Mastic
Point, Andros Island. It had, however, twenty-two rows of scales about the
body instead of eighteen; and as this character is not variable among many
Typhlopidae, 7. tenwis among them, there is ground for reasonable doubt as to
the correct identification of the species. Rosén does not mention that he had
other specimens for comparison. The species is one which ranges through
southern Mexico and Guatemala.

Typhlops rostellatus SresNEGER.

Srresnucer, Rept. U.S. nat. mus. for 1902, 1904, p. 686, figs. 146-147.

A species peculiar to Porto Rico, which upon that island takes the place of
its close relation, 7. platycephalus Duméril & Bibron from Martinique, and 7’.
dominicana Stejneger from Dominica. There yet remain many islands from
which representatives of this group of species have not yet been described, but
upon which they undoubtedly exist.

324 BARBOUR: ZOOGEOGRAPHY.

Typhlops platycephalus Dumérit rr Bisron.
Dumeérit et Brpron, Erpét. gén., 1844, 6, p. 293.

A species sent from Martinique by Plée to the Paris museum. It is con-
fined to the island, and is very rare in collections; of its abundance on the island
nothing is known.

Typhlops dominicana STEJNEGER.

STEINEGER, Rept. U.S. nat. mus. for 1902, 1904, p. 687.

Stejneger has shown that Boulenger’s description (Cat. snakes Brit. mus.,
1893, 1, p. 30) of 7. platycephalus Duméril & Bibron, based on specimens from
Dominica, differs in important characters from the original description based on
specimens from Martinique. He then proposes for the species apparently con-
fined to Dominica, of which in 1893 there were four specimens in the British
museum, the name T'yphlops dominicana.

Leptotyphlops albifrons (WacLER).
WAGLER, Spix’s Serp. Braz., 1824, p. 68, pl. 25, fig. 3. BouLencr, Cat. snakes Brit. mus., 1893, 1,
p. 63.

This very widely distributed species, besides being recorded as distributed
on the Tropical American mainland, has been recorded in the Antilles from
Watlings Island, Grenada, and Antigua. It doubtless ranges widely through
the Lesser Antillean chain. Recently Mr. Nelson has brought back a series of
eight specimens from Swan Islands. He reports it common there and found in
the leaf mould of the forest, it is often seen in broad day light crawling about in
paths and clearings.

Leptotyphlops bilineata (Schlegel).
ScuiEeceL, Abbild. Amphib., 1844, p. 36, pl. 32, fig. 5-6. BounLmencnr, Cat. snakes Brit. mus.,1898, 1,

p. 70.

Recorded originally by Duméril and Bibron (Erpét. gén., 1844, 6, pp. 331)
from Martinique and Guadeloupe. There was also a specimen from Barbados
in the British museum. The snake is apparently rare, and is seldom found by
collectors.

Epicrates subflavus STEJNEGER.
STESJNEGER, Proc. U.S. nat. mus., 1901, 23, p. 469-470.
The Jamaican yellow boa was only separated in 1901, after very many

years of confusion with the Porto Rican species, H. inornatus. It has been al-
most exterminated by the ravages of the mongoose (Herpestes birmanicus

REPTILIA. 320

(Thomas)), and while excessively rare on Jamaica, is said to still be quite abun-
dant on Goat Island, not far from Old Harbor. It probably still occurs along
with the Iguana (Cyclura collei Gray) in the Heathshire Hills, in the southern
part of the island. ;

Epicrates striatus (FiscHHR).

Fiscuer, Abhandl. Naturh. ver. Hamburg, 1856, 3, p. 102, pl. 2, fig. 2. BounLenerr, Cat. snakes Brit.
mus., 1893, 1, p. 96-97. SrrynecER, The Bahama Islands, 1905, p. 336.

This species occurs on San Domingo, New Providence, and Andros. In
1904 I mentioned Bahaman specimens under the name of LE. strigilatus Cope;
since that time I have been able to compare much additional material, and now
subscribe to the views of Boulenger, Stejneger, and, more recently, Rosén, that
the San Domingan and Bahaman individuals are conspecific (Barbour, Amer.
nat., 1906, 40, p. 230).

Epicrates monensis ZENNECK.

ZENNECK, Zeit. wiss. zool., 1898, 64, p. 64, pl. 3, fig. 58-62. Strsnecer, Rept. U.S. nat. mus., for 1902,
1904, p. 692.

This boa, confined to Mona Island is one of a number of small species whose

status and relationships are as yet more or less enigmatical.

Epicrates chrysogaster Corr.

Corr, Proc. Amer. philos. soc., 1871, 11, p.557. Strsnecrr, The Bahama Islands, 1905, p. 335.

Stejneger has pointed out that this species should be kept separate from its
ally, H. fordit (Giinther), from Haiti, on account of having forty-three rows of
seales, and fifty-four dorsal spots, whereas H. fordii has 33-87 scale rows, and
69-78 dorsal spots. It is confined to Turk’s Island, and is also very closely
related to the preceding.

Epicrates fordii (GiinTHER).

GtnrTuER, Proc. Zool. soc. London, 1861, p. 142, pl. 23. BouLmncrr, Cat. snakes Brit. mus., 1893, 1,
p. 98.

Boulenger unites this species with the preceding, but probably this form
from Haiti is not really the same as the Turk’s Island species. More material
is badly needed of all these boas.

Epicrates inornatus (REINHARDT).

RernHarpt, Danske vid. selsk. Afhandl., 1843, 10, p. 253. Srrsnucpr, Rept. U.S. nat. mus. for 1902,
1904, p. 688.

This boa, closely related to the Jamaican yellow boa, and but recently
recognized as distinct, is confined to Porto Rico. It has grown very rare, and
recent collectors have had great difficulty in procuring specimens.

326 BARBOUR: ZOOGEOGRAPHY.

Epicrates gracilis (FiscHER).
Fiscuer, Jahrb. Hamburg, wiss. anst., 1888, 5, p. 35, pl. 3, fig. 8. BouLENGER, Cat. snakes Brit. mus.,
1893, 1, p. 98.

This boa, which is said to be ‘‘blackish gray, somewhat lighter on the
belly,” and having “‘six longitudinal series of small black spots,” has evidently
a type of coloration very different from other species in the genus. Boulenger
suggests that it is “‘closely allied” to EL. fordiz; but to judge from the descrip-
tion, they seem to be very unlike. This is evidently a rare species, since up to
1893 no specimens had found their way to the British museum. It was described
from San Domingo, yet none of the collections from that island in the Museum

contain specimens.
Epicrates anguHfer Breron.

BIBRON, Sagra’s Hist. Cuba. Rept., 1843, p. 215, pl.25. Bounmnesr, Cat. snakes Brit. mus., 1893, 1,
p- 96.

The ‘‘majase,” or ‘‘maja,” as the Cuban boa is called locally, is still common
in many regions of the island. Miller, quoting Palmer’s field notes (Proc. U. 8.
nat. mus., 1904, 27, p. 346) says, in speaking of the great bat caves about Bara-
coa and Guanajay, ‘‘The people of the neighborhood assured me that the majas
(the Cuban boa, Epicrates angulifer) coil themselves among these roots [of the
trees near the mouth of the cave], and grab at the bats as they fly out. I was
told that a snake frequently secures a bat in this manner.” I have also heard
this tale.

This is far the largest member of the genus. It grows commonly to be
twelve feet long and to be as large in circumference as any boa, or as a Python
of far greater length. Gundlach records individuals of five and seven yards
length. Such specimens would be excessively rare now owing to the persecution
which the species has undergone with the increasing population and cultivation
of the island. I got two living specimens near Havana, which were singularly
intractable; and the species is characterized by its very bad temper. I have

seen other specimens elsewhere. It is wide spread and common.

: |

Boa hortulana Linna.
Linn&, Syst. nat. ed. 10, 1758, 1, p. 215. BouLmnesr, Cat. snakes Brit. mus., 1893, 1, p. 101-102.
Dr. Allen brought back three examples of this arboreal species, all of them

taken near St. George’s. Garman (Proc. Amer. philos. soc., 1887, 24, p. 279)
records specimens from Trinidad, Grenada, St. Vincent, and the Grenadines.

)

REPTILIA. 327

He shows how excessively these vary in color, his three examples from Grenada
differing much from one another. The same is true of Allen’s specimens. There
is, then, no evidence, either in squamation, or coloration, of the formation of a
local race of this species upon Grenada. Boulenger (Proe. Zool. soc. London,
1891, p. 355-356) records the finding of specimens on St. Vincent and Becquia.
No mention, however, was made of these specimens in the Catalogue of snakes
in the British museum, the last volume of which, with addenda, did not appear
until 1896.

~ Boa grenadensis, sp. nov.

Type:— No. 7,991, M. C. Z., St. George’s, Grenada, August 29, 1910, G. M.
Allen, collector. Another example, taken August 20, No. 7,990, M.C. Z., is a
cotype.

Boulenger (Cat. snakes Brit. mus., 1893, 1, p. 100) records four specimens
of what he calls Corallus cookii var. C. from Grenada. These alone, among
all his series, are characterized by being yellowish brown above, uniform or with
mere traces of dark markings; belly yellow, without spots. Allen brought back
two examples, both from St. George’s, Grenada; and as these agree in coloration
with what Boulenger has described, there can hardly be reasonable doubt but
that a local race, with characteristic color, has become differentiated upon this :
island. The scale counts for Boulenger’s specimens follow :—

268 270 264

4lio1, ASi09) 4lios; Age

270

The type, M. C. Z., No. 7,991, counts 45744, while the other example has 43
scale rows, but was damaged so that the other counts can not be made. The

253-270 : . .
Toca: Lhis does not vary far from a simi-

range for these six is, then, 41-45
lar series from other localities. Garman did not find this snake on Grenada.

According to Allen, it is known locally as ‘

‘sarpint”’ in contradistinction to
“snake,” this latter name being confined to the Racer. For the use of Boa

instead of Corallus consult Stejneger, Proc. U. 8. nat. mus., 1902, 24, p. 184.

Tropidophis melanura ScHLEGEL.

ScHLEGEL, Essai phys. serp., 1837, 2, p.399. BouLrencrr, Cat. snakes Brit. mus., 1893, 1, p. 111.

Stejneger (The Bahama Islands, 1905, p. 336) states that Tropidophis takes
precedence over Ungualia, since the former was based on Schlegel’s Boa melanura
by Bibron in 1840, and thus antedates Gray’s Ungualia by two years.

This species is not uncommon, and I found it and heard of its regular occur-
rence in just one lurking place. A number of specimens were found in various
localities on Soledad estate, near Cienfuegos, Cuba, always hidden in the rolled

328 BARBOUR: ZOOGEOGRAPHY.

up ‘yaguas,’ expanded bases of the leaves of royal and cocoanut palms. Where
there is a grove of these trees in an orchard or some other wooded locality, the
curled up fallen leaves will often be found to contain this snake. I never, how-
ever, came across a specimen where the leaves fell so as to be exposed to the full
force of the sun. The species is strictly nocturnal and evidently does not prowl
about in the open, unwooded country.

The Museum had several specimens from ‘‘Cuba”’ before the receipt of the
series from Soledad in 1912. In 1913.I had a fine series of well-preserved ex-
amples given me by Mr. C. T. Ramdsen who took them in the region about
Guantanamo. In addition to the bluish and reddish phases of coloration two of
these are clear unmarked light buff. This species is evidently the most variable
in coloration of any in the genus.

Tropidophis pardalis (GuNDLACcH).

Gunptacu, Archiv. naturg, 1840, 1, p. 359. Bounencrr, Cat. snakes Brit. mus., 1893, 1, p. 118.
StEesnEGER, the Bahama Islands, 1905, p. 336.

The Bahaman “ thunder snake”’ has been referred to 7. maculata by Cope,
but incorrectly, as I have identified specimens with this species (Bull. M. C. Z.,
1904, 46, p. 59; Amer. nat., 1906, 40, p. 230) from New Providence; and Stej-
neger has since agreed to this identification.

The species has been found on Cuba, New Providence, Andros, and Eleu-
thera Islands.

A specimen No. 6,114, M. C. Z. from Cuba, is labeled in Garman’s hand as
Ungalia curta, and probably served him as the type of that species. He gave
Cuba as the locality, and this is the only Cuban specimen having such a label.

The species is a synonym of the above.

I did not meet with this species during my collecting trips in Cuba.

Tropidophis maculata (Bipron).

Bipron, Sagra’s Hist. Cuba. Rept., 1848, p. 212, pl. 24. BouLencer, Cat. snakes Brit. mus., 1893,
1, p. 112.

This species may possibly be a composite one. Cope held that Haitian
individuals were distinct from others, and proposed for them the name of 7.
haetiana. They have not been shown to be really different. Nevertheless, at
present only a few specimens have been recorded from the islands where the
species occurs, and large series will unfortunately be increasingly difficult to
obtain. In Cuba the species is far from common, while on Jamaica and Haiti
the mongoose is already omnipresent. The snake is also recorded from the

REPTILIA. 329

island of Navassa, where the mongoose has probably not been introduced. In
Jamaica I found it extremely rare, and got only one specimen, the mongoose
being undoubtedly responsible for its: scarcity. From Haiti the Museum has
but two specimens. From Cuba there are three, two from Havana and one
from Soledad, near Cienfuegos. While the specimens from each of these locali-
ties show certain peculiar characteristics in coloration, their squamation does
not vary definitely. The series from Jamaica, Haiti, and Cuba seems to belong
to the same species.

Mr. Mann’s recent collection contains three specimens from Cape Haitien
and one from Diquini, Haiti. In only one is the interparietal present as Cope
supposed is always the case with Haitian examples. In three examples the scale
rows number 27 and in one 25, in none 29. This is the number also declared a
character of 7’. haetiana. Nevertheless much more material is needed to show
that the place modes do not differ with series from the several islands. I may
be quite wrong and inconsistent in not allowing 7. haetiana to stand separately
but there is not much proof of its distinctness.

Tropidophis semicincta (GunpLAcH & PETERS).

GunpuAcH & Peters, Monatsb. Akad. wiss. Berlin, 1864, p. 388. BouLENGER, Cat. snakes Brit. mus.,
1893, 1, p. 113.

Apparently a very rare species, confined to Cuba. Unfortunately I was
unable to obtain a single specimen of this species. There is a fine example in
the Gundlach collection in the Instituto de Segunda Ensafanza at Havana,
and two in the U. S. national museum at Washington.

Tropidophis conjuncta FiscHer.

Fiscuer, Jahresb. Hamb. wiss. anst., 1888, 5, p. 31, pl. 3, fig. 5. BouLencrr, Cat. snakes Brit. mus.,
1893, 1, p. 113.

Another obscure and little-known species, confined to San Domingo.
Tropidophis cana (Corr).
Cops, Proce. Acad. nat. sci. Phila., 1868, p. 129. BouLrmncmr, Cat. snakes Brit. mus., 1893, 1, p. 114.

A very distinct species confined to the island of Inagua in the southern
Bahamas.

Constrictor orophias (Linn&).

Linné&, Syst. nat. ed. 10,1758, 1, p.215. BovuLeneerr, Cat. snakes Brit. mus., 1893, 1, p. 118.

Stejneger, in showing that for these large tree snakes the generic name
Constrictor should be used instead of the long-established Boa, remarked at the

b

330 BARBOUR: ZOOGEOGRAPHY.

same time that Linné’s Boa orophias was the same species as that named Con-
strictor divinilogquus by Laurenti (1768). The latter name, variously spelled,
has usually been used for the Lesser Antillean Constrictor. Boulenger had, in
1893, specimens in the British museum from St. Lucia, Dominica, and Trinidad.
The Philadelphia Zoological garden has had one said to have been taken upon
St. Kitts.

Andersson (Bihang K. Svensk. vet.-akad. Handl., 1899, 24, afd. 4, no. 6,
p. 33) also shows that the type of Linné’s Boa orophias is the same as the later
named diviniloqua.

Tretanorhinus variabilis Dumérit nt Bipron.

Dumérit et Brsron, Hrpét. gén., 1854, 7, p. 349, pl. 80, fig. 4. Bounenerr, Cat. snakes Brit. mus.,
1893, 1, p. 282.

This species is the Cuban representative of a small genus, the other two
species of which occur in Central America. It is a strictly aquatic snake which
never leaves the water. It is not uncommon about ponds and rivers, both
rushing and sluggish. It does not often sun itself, but remains for the most
part hidden in dense aquatic vegetation or among and under loose stones. The
natives all say that it is nocturnal and that they often meet it in rivers swimming
about when they are fording the streams by night. It is a difficult species to
find, but is undoubtedly rather common and widespread. I have specimens
from near Cienfuegos and San Diego de los Bafios while I saw others near Pinar
del Rio city, and near Herradura. It is called “‘catibo,” and I often heard of it
by this name in other parts of the island. In 1913 I caught it in the Rio Tana,
Ingenio ‘“‘San Ramon,” not far from Manzanillo.

Drymobius boddaertii (SrenrTzEn).

SENTZEN, Meyer’s Zool. arch., 1796, 2, p.59. BouLenanr, Cat. snakes Brit. mus., 1894, 2, p. 11.

Recorded by Boulenger from St. Vincent. A species common and wide-
spread on the mainland (Proce. Zool. soc. London, 1891, p. 355). In the Cata-
logue (loc. cit.) Boulenger mentions a specimen from Grenada sent by this
Museum to the British museum. Garman collected examples only on St. Vincent,
to judge by his paper on West Indian snakes (Proc. Amer. philos. soc., 1887,
24, p. 284). Ican find no Grenadian examples in the Museum now, and Garman
would not have been likely to have sent the only one from this locality to London.
Neither Dr. Allen nor Mr. Brues, in their recent exploration of Grenada, saw or
heard of this species, so that in all probability the British museum’s snake came
from Kingston, St. Vincent, apparently the only island where Garman found the

REPTILIA. dol

species, and perhaps the only island where it occurs. It would certainly be
natural to expect this species upon Grenada, and perhaps its absence there may
be laid to the account of the mongoose, as may also be the case with other species

which might be expected upon this island.

Herpetodryas carinatus (LINNs).

Linn&, Syst. nat., ed. 10, 1758, 1, p. 223. Bouimncrr, Cat. snakes Brit. mus., 1894, 2, p. 73.

This mainland species has been recorded from the island of Guadeloupe,
and Boulenger says that he has compared this specimen from Guadeloupe in the
British museum with mainland examples, and can find no difference between
them. It does not appear to be recorded from others of the Antilles, and may
possibly have been brought to Guadeloupe as a rat exterminator. It would be
expected to occur upon other islands if its distribution had not been by artificial

agency. Of course it is not improbable that the locality may not be correct.

Herpetodryas vincenti BouLENGER.
BouLENGER, Proc. Zool. soc. London, 1891, p. 355.

This species, which was originally described as a variety, differs in having a
greater number of ventrals and subcaudals, and in being more slender than the
mainland form. The types, in the British museum, came from St. Vincent.

The two specimens in the Museum from this island also show these characters.

Uromacer catesbyi (ScHLEGEL).

ScHLEGEL, Essai phys. Serp., 1837, 2, p. 226. BouLmnamr, Cat. snakes Brit. mus., 1894, 2, p. 115.

A rare member of this peculiar genus, which is wholly confined to Haiti.
Its habit, which is so strikingly like that of Dryiophis in the Old World and
Oxybelis in the New, is a marvellous adaptation to arboreal life, since this snake
is hardly more related to either of these genera than they are to each other.
The Museum has several specimens, including three which Mr. W. M. Mann
caught at Momance, Haiti.

Uromacer frenatus (GiNnTHER).

Ginruer, Ann. mag. nat. hist., 1865, ser. 3, 5, p. 116. Bounmenemr, Cat. snakes Brit. mus., 1894, 2,
p. 116.
Another species confined to Haiti. I had some difficulty in locating the
types of Garman’s Uromacer inornatus, which Boulenger rightly placed in the
synonymy of this species. Finally, however, I found four examples which

agreed exactly with Garman’s description as to scale counts, so that there can

332 BARBOUR: ZOOGEOGRAPHY.

be no possible question as to the fact that these are the actual types. They had
been relabeled by Garman as Uromacer oxyrhyncus, but they really belong with
this distinct species. The specimens are No. 3,345, M. C. Z., two examples from
Haiti, and No. 3,610 M. Z. C., two examples from Jeremie. Garman in his
description, undoubtedly by a lapsus calami, credited all four specimens to
Jeremie, Haiti (Proc. Amer. philos. soc., 1887, 24, p. 284). In Mr. Mann’s
collection there are three from Diquini, Haiti and one from Manneville.

Uromacer oxyrhynchus Dumirit er Brsron.

Doummrit et Brsron, Erpét. gén., 1854, 7, p. 722, pl. 83, fig. 1. BouLenerr, Cat. snakes Brit. mus.
1894, 2, p. 116.

The third and last member of this genus is, of course, also from Haiti only.
The specimens in the Museum are from Santiago de la Vega, collected by Mr.
A. H. Verrill; Samana, San Domingo, collected by Mr. M. A: Frazar, and
Diquini from Mr. W. M. Mann. Other examples are simply labeled Haiti.
There is no evidence from the material preserved here to show that the species
of Uromacer are confined to particular ranges within the area of the island.

Hypsirhynchus ferox GinrHer.

Gtnruer, Cat. snakes Brit. mus., 1858, p. 48. BouLEnerr, Cat. snakes Brit. mus., 1894, 2, p. 117,
pl. 6, fig. 1.

This species was originally described as coming from Barbados, and became
established in the literature as a Barbadian anomaly. ‘Thus Wallace mentions
it in the Geographical distribution of animals (2, p. 72). Boulenger, however,
has shown that the type was purchased for the British museum from Hugh
Cuming, and that all his collection contains only species characteristic of Haiti
or San Domingo. The genus is, then, a monotypic one, confined to the island.
Hypsirhynchus scalaris Cope, the type of which is No. 1,517, M. C. Z. was de-
scribed when H. ferox was supposed to come from Barbados; and the distance
of that island from Haiti may have suggested the supposed distinctness of the
species. Cope’s type appears to be sightly abnormal in that it lacks a loreal.
This would not seem sufficient to establish a species when only one specimen is
known which is characterized in this way.

Mr. Mann’s collection contains an example from Momance, and another
from Manneville. Both perfectly typical.

REPTILIA. 333

Alsophis angulifer Brsron.

Brsron, Hist. Cuba. Rept., p. 222, pl. 27. BouLenerr, Cat. snakes Brit. mus., 1894, 2, p. 120, (part).

A very common Cuban species. No. 2,195, M. C. Z., came from the Paris
museum in an old exchange and is marked as a type by A. Aug. Duméril.
Boulenger has considered the specimens from the Bahamas and the Cayman
Islands all identical with the Cuban examples. They are separated here.

This species has been met with on my Cuban trips at every locality visited
and specimens have been preserved from Pinar del Rio, Herradura, San Diego
de los Bafios, Havana, Manzanillo, and Cienfuegos. It is said to reach a length

of over two meters but none was seen larger than four feet.

Alsophis caymanus GARMAN.

GarMaAn, Proc. Amer. philos. soc., 1887, 24, p. 276.

This species replaces the Cuban A. angulifer upon the island of Grand
Cayman. The type is No. 6,020, M.C.Z. It is of doubtful validity.

Alsophis fuscicauda Garman.

GARMAN, Bull. Essex inst., 1888, 20, p. 106.

Similar in its origin to the preceding, but confined to Cayman Brac. Until
more material is available, it is impossible to settle definitely the status of
these two species from the Cayman Islands. The type is No. 6,235, M. C. Z.
This seems also little more than a color phase.

Alsophis brooksi, sp. nov.

Types:— No. 7,893, M. C. Z., two specimens (lacking tails). Little Swan
Island, Caribbean Sea, March, 1912, George Nelson, collector.

This form which is simply an island race of the Cuban A. angulifer Bibron, ’
comparable to A. fuscicauda Garman from Cayman Brae and A. caymanus
Garman from Grand Cayman differs from all of these in coloration. The entire
upper surfaces in the fully adult become completely suffused with dark brown,
which becomes a solid color along the middorsal region. All of the scales of both
half grown and adult specimens are heavily punctulated with dark brown. In
the other races they are free from this characteristic spotting. The ventral
scale counts for the two specimens, 182 and 183 show that the average condition
for this form is evidently above that of the other races, their highest scale count,

rarely observed, is 180.

BARBOUR: ZOOGEOGRAPHY.

(S%)
oo
Ss

This snake is not found upon Swan Island having been probably extermi-
nated there by either the early settlers or the cats which they brought. Named
for Dr. William A. Brooks, President of the Swan Island Commerical Company
who aided Mr. Nelson throughout his trip.

Alsophis vudii Cops.

Cops, Proc. Acad. nat. sci. Phila., 1862, p. 74. Barsour, Bull. M. C. Z., 1904, 46, p.60. Amer. nat.
1907, 40, p. 231. Rosin, Lunds univ. arsskrift, 1911, 7, no. 5, p. 40.

This Bahaman ground racer, so closely allied to A. angulifer of Cuba, has
been found so far only upon the islands of Eleuthera, New Providence, Long
Island, and Green Cay. This erratic distribution suggests that it probably
exists undiscovered upon many others. Thanks to Mr. Witmer Stone I have
examined the type No. 3,472, Acad. nat. sci., Phila. of Cope’s Diadophis rubescens;
it belongs to this species.

Alsophis rufiventris (Dumérit ET Brsron).

Dumérit et Brsron, Erpét. gén., 1854, 7, p. 668.

Of the types of this species one was said to have come from Brazil, and the
other has no locality record. Garman refers specimens from St. Kitts, Saba,
and Nevis, to this species while Boulenger (Cat. snakes Brit. mus., 1894, 2,
p. 125) adds St. Eustatius to the list of localities. He includes Alsophis cinereus
Garman and A. rijersmaet Cope, as synonyms.

Alsophis cinereus GarMaAn.

Garman, Proc. Amer. philos. soc., 1887, 24, p. 282.

This form, which is closely related to A. rufiventris, but easily distinguished
by having twenty-one rows of scales instead of twenty-three is found upon the
islands of St. Bartholomew and Anguilla. The types are No. 6,139, M. C. Z.

Alsophis rijgersmaei Corr.

CoprE, Proc. Amer. philos. soc., 1869, 11, p. 154.

Cope says, ‘‘So far as yet known, this distinct species is confined to the small
island of St. Martins.” There are no specimens in the collection here, so that
I can not verify Boulenger’s placement of this species as a synonym of A. rufi-
ventris (Duméril & Bibron).

REPTILIA. 339

Alsophis porto-ricensis Rninnarpt & LUTKEN.

ReinHaArpt & Lirken, Vid. Meddel. nat. foren Kjobenh. for 1862, 1863, p. 221. SresnecrrR, Rept.
U.S. nat. mus. for 1902, 1904, p. 700, fig. 170.
Closely related to A. antillensis. These species parallel in their distribu-
tion Leimadophis stahli and L. exiguus, since, like the former, this snake is con-
fined to Porto Rico and Mona Island. JL. stahli has not been found upon Mona,

but doubtless occurs there.

Alsophis melanichnus Corr.

Cork, Proe. Acad. nat. sci., Phila., 1862, p. 76.

I have identified with this poorly described species a fine large male taken at
Santiago de la Vega, San Domingo, by Mr. A. H. Verrill. After reading the
following in Stejneger’s Herpetology of Porto Rico (Rept. U. 8. nat. mus. for
1902, 1904, p. 701), I compared it carefully with his description of A. portoricen-
sis. Stejneger writes :—

“The question whether Cope’s A. melanichnus, from Haiti, is identical with
the Porto Rican species can not be said to be settled yet. The scale formula
seems to be the same, but it is not certain that there may not be other distin-
guishing features. .... We have such startling proof of the effect of isolation
working with a variable material to produce separate forms of these snakes
that it is quite unwarrantable to lump the various names without incontrover-
tible proof.” 4

Stejneger says the scale formula ‘“‘seem to be the same”’ in this species and
A. portoricensis. Cope, however, only mentioned the one fact that it had seven-
teen scale rows. Cope mentions No. 1,522, M. C. Z., from Jeremie, Haiti, as the
type; but the specimen can not be found. Upon comparing our only specimen
with Stejneger’s description, or with a specimen of A. portoricensis, I find that
the Haitian snake agrees in all characters, except in a very elongate loreal, much
lower than high, and almost a rough triangle in outline, the base being anterior.
It has two first temporals on each side, but these have the appearance of being
the elements derived from an abnormally divided single scale. The ventral
scales are 189 in number, and subcaudals 108, the anal is divided. These counts
will be seen to vary from the range and average of Stejneger’s forty-four speci-
mens from Porto Rico, which was 17 rows, 169-183 (average 178) ventrals, sub-
caudals 112-129. This combination of characters makes it seem almost certain
that the Haitian species is perfectly distinct from, but very closely related to,
the Cuban A. angulifer, as well as A. portoricensis.

1336 BARBOUR: ZOOGEOGRAPHY.

I am inclined to believe that Werner’s Dromicus w-nigrun belongs with
this species.’ It was described in 1909. (Hamb. jahrb. wiss. anst., 1909, 26,
beih. 2, p. 222).
; Alsophis antillensis (ScHLEGEL).

. ScHIEGEL, Hssai phys. Serp., 1837, 2, p. 214. Srmsnecer, Rept. U.S. nat. mus. for 1902, 1904, p. 704,
fig. 171-174.

Like Leimadophis exiguus, this form is found upon.St. Thomas, Stas ohn,
and Culebra. The specimens which Garman recorded from Haiti (Proc. Amer.
philos. soc., 1887, 24, p. 282) really belong to this species; and hence it must be
concluded that in some way they have gotten their locality label misplaced.
The species is, of course, in all probability not found on Haiti.

Alsophis sanctae-crucis Corr.

Cork, Proc. Acad. nat. sci. Phila., 1862, p. 76.
This species replaces A. antillensis, of St. Thomas, upon St. Croix; to which

island, so far as known, it is confined.

Alsophis anomalus (PETERs).

Peters, Monatsb. Akad. wiss. Berlin, 1863, p. 282. BouLmenerer, Cat. snakes Brit. mus., 1894, 2,
p. 125.

A very distinct species, confined to the island of Haiti.

Alsophis leucomelas (Dumérit ET BriBRon).

Dumérit et Brsron, Erpét. gén., 1854, 7, p. 666. BouLencerr, Cat. snakes Brit. mus., 1894, 2, p. 123
(part).

Originally described from Marie Galante and Guadeloupe. Boulenger
adds the localities Antigua and Montserrat. No material representing this
species is in the collection, so that it is impossible to say whether or not the snakes
are the same on the different islands. The great variability which Boulenger
speaks of is very likely due to the differences in color in specimens from the vari-
ous islands. This character is often very diagnostic and unvarying among speci-
mens from a single island.

Alsophis sibonius Cope.

Core, Proc. Amer. philos. soc., 1879, 18, p. 275.

Apparently confined to Dominica, this snake appears to warrant separation
from leucomelas of Duméril and Bibron, with which Boulenger Cat. snakes Brit.
mus., 1894, 2, p. 124 has united it. The single specimen in the Museum agrees

closely with Cope’s description.

a ; REPTILIA. Hanes oot
ye x ‘ 2a AA \ J ONAL?

—

(/ (= Alsophis bruesi, sp. nov. )}

Type: — No. 7,792, M. C. Z., adult, near St. George’s, Grenada, G. M. Allen
and C. T. Brues, collectors. Six other specimens from the same source in the
collection. -

By a curious coincidence the only ground snake that either Brues or Allen
obtained upon Grenada was a true Alsophis having scales in seventeen rows and
with two distinct pores in each one. This species they found rather abundant,
and they brought back seven fairly well-preserved examples. Neither Garman
nor any other collector, apparently, has hitherto found this genus upon Grenada.
Undoubtedly some peculiarity of habit or habitat has enabled it not only to
survive in spite of the presence of the mongoose, but to become, apparently,
the most abundant ophidian on the island.

Rostral much broader than high, scarcely visible from above; internasal
suture a little shorter than prefrontal suture; frontal equal to its distance from
end of snout, shorter than parietals, separated from preocular by only a short
interspace; supraocular at its posterior margin almost equal in width to the
anterior margin of frontal; nasal divided, shorter than its distance from eye;
loreal a little more than twice as long as high, an almost perfect rectangle; a
large preocular; two postoculars, lower smaller than upper; a large anterior
temporal, with almost always a small scale intercalated above it, anteriorly.
In the type this small scale has a lower position on one side so that one small
temporal seems to be followed by the larger one. Two large temporals follow;
nine supralabials (in all specimens), second and third in contact with-labial-~
fourth with preocular, and slightly entering orbit; five and six in orbit; nine or
ten lower labials, five in contact with anterior chin shields; anterior chin shields
scarcely more than half as long as posterior; scales smooth, in seventeen rows,
each with two conspicuous apical pits; ventrals 198, anal divided; subcaudals
in 121 pairs. Length of body 700 mm., tail 310 mm. Color (in alcohol) above
uniform olive-drab or slightly darker, below whitish or light slatey gray, with
two lateral white lines on the ventrals, which may have slightly darker margins.
Throat usually yellowish, the fourth scale row on each side light in color; a dark
‘loreal band generally extending for some distance behind the eye.

Variations: — A young specimen about 400 mm. in total length has each
side margined with blackish, and light lateral line covering the fourth and half
of the fifth row of seales. The characteristic squamation and coloration is
remarkably stable. The largest specimen, which is not as well preserved as the
type, measures 760 mm. in total length, and 320 in length of tail. The range of

338 BARBOUR: ZOOGEOGRAPHY.

variation in the scale counts is shown by the following numbers taken from the

cotypes: — ventrals 201, subcaudal pairs (tail imperfect); ventrals 199, sub-
caudal pairs 127; ventrals 194, subcaudal pairs 125; ventrals 205, subcaudal
pairs 120; ventrals 205, subeaudal pairs (tail imperfect); ventrals 193, subcaudal
pairs 121. In every case the anal is divided. The range then may be expressed

as v. 193-205; sube. 120-127 pairs.

Leimadophis ater (Gosse).

Gossg, Naturalist’s sojourn in Jamaica, 1851, p. 228. Bargour, Bull. M. C. Z., 1910, 52, p. 300.

A snake which has been practically exterminated by the mongoose. It was
formerly abundant upon Jamaica, but has not been found at all by recent col-
lectors.

Leimadophis callilaemus (Gosse).

Goss, Naturalist’s sojourn in Jamaica, 1851, p. 384. Barsour, Bull. M. C. Z., 1910, 52; p. 300.

Another species which, though it has become rare, has not yet been quite
extirpated by the mongoose. It is confined to Jamaica.

Leimadophis andreae Rrernnarpt & Lirnpn.

Reinnarpt & LirKken, Vid. meddel. nat. foren Kjébenh. for 1862, 1863, p. 214. BouLencer, Cat.
snakes Brit. mus., 1894, 2, p. 140.

The rather complicated synonymy of this species seems to have been settled
satisfactorily by Boulenger. The species is confined to Cuba, and is represented
in the collection here by the specimens which were collected by F. Poey and which’
served Garman for the types of his synonymous Dromicus cubensis (Proc.
Amer. philos. soc., 1887, 24, p. 280). Besides these we secured a series by fol-
lowing ploughs on the Soledad sugar estate, near Cienfuegos. This species

seems to spend much of its time burrowing under ground or hiding under stones.

Leimadophis temporalis (Corr).

Corr, Proc. Acad. nat. sci. Phila., 1860, p. 370. Boutenanr, Cat. snakes Brit. mus., 1894, 2, p. 148.

A well defined and apparently very uncommon Cuban species. Garman
(Proc. Amer. philos. soc., 1887, 24, p. 281) believes that this is the species figured
by Jan as Dromicus nuntius. Boulenger, however, (loc. cit., p. 181) identifies the
latter species with Urotheca lateristriga, from Central America. Cope states that
the type is in this Museum. The only specimen is No. 297. This is beyond
doubt the typical specimen, although it bears no contemporary type label.

I have never found this species in the parts of Cuba where I have collected.

REPTILIA. 309

Leimadophis parvifrons (Corr).
Cops, Proc. Acad. nat. sci. Phila., 1862, p.79. Boutenerr, Cat. snakes Brit. mus., 1894, 2, p. 141.

This species is confined to Haiti. The types are No. 3,602, M.C.Z. It was
among the collection made by Mr. A. H. Verrill at Santiago de la Vega, San
Domingo, and in Mr. W. M. Mann’s collection there are many specimens from

Diquini, St. Marc, Cape Haitien, Manneville, and Grand Riviere, Haiti.

Leimadophis cursor (LAchPEDE).

Lacrepéps#, Hist. nat. quad. ovip. et serpens, 1789, 2, p. 96, 281, pl. 14. BouLmncsr, Cat. snakes Brit.
mus., 1894, 2, p. 139 (part).
Apparently common upon Martinique, whence Garman recorded many
specimens (Proc. Amer. philos. soc., 1887, 24, p. 280). Boulenger includes two
specimens from Guadeloupe, which quite likely represent a different race.

Leimadophis boulengeri, nom. nov.

This is the Dromicus ornatus of Garman (Proc. Amer. philos. soc., 1887, 24,
p- 281). His specific name is apparently preoccupied by Coluber ornatus Shaw
(Zool., 1802, 3, p. 477), which evidently belongs to this genus.

The new name is given in honor of Dr. G. A. Boulenger, of the British mu-
seum, to whom I owe a vast deal for kindly hospitality and assistance.

This St. Lucia species is easily distinguishable from that of Martinique
(A. cursor) by having fewer pairs of subcaudal scales, 85-91 as against 101-107.
The types are No. 6,135, 6,136, 6,137, M. C. Z.

Leimadophis stahli Sresnecer.

SresneGER, Rept. U. 8. nat. mus. for 1902, 1904, p. 695, fig. 161-166.

This species, which has formerly been confused with A. parvifrons (Cope)
from Haiti, is related to that species, but perfectly distinct, and confined to
Porto Rico.

Leimadophis exiguus (Cope).
Cops, Proc. Acad. nat. sci. Phila., 1862, p.79. Sresnecrer, Rept. U.S. nat. mus. for 1902, 1904, p. 698,
fig. 167-169.

This snake, closely related to the preceding, replaces it upon the islands of
St. John and St. Thomas. It has recently been found upon Culebra; and, as
Stejneger suggests, it probably occurs upon Vieques as well.

340 BARBOUR: ZOOGEOGRAPHY.

Leimadophis juliae (Corr).

Corr, Proc. Amer. philos. soc., 1879, 18, p. 274. Boutencerr, Cat. snakes Brit. mus., 1894, 2, p. 139.

A fine species, which has been found abundantly upon Dominica.

The difference in color which Garman has mentioned (Proc. Amer. phil.
soc., 1887, 24, p. 281) is very conspicuous in specimens from this island and
Marie Galante, and there seems to be no reason why, if this is found to be the
case in large series, the Marie Galante species should not be considered a valid

race. It may be called

Leimadophis mariae, sp. nov.

The types are No. 6,138, M.-C. Z. The characters have already been
given by Garman.
Leimadophis perfuscus (Corr).

CopE, Proc. Acad. nat. sci. Phila., 1862, p.77. BouLEenerr, Cat. snakes Brit. mus., 1894, 2, p. 133.

A well-defined species confined to Barbados.

Leimadophis melanotus (SHaw).
Suaw, Zoology, 1802, 3, p.534. BouLencer, Cat. snakes Brit. mus., 1894, 2, p. 134.

This mainland species has been recorded from Grenada by Garman (Proc.
Amer. philos. soc., 1887, 24, p. 282).

Allen did not obtain a single specimen, and it has probably gone the way of
most of the ground-inhabiting ophidians of the island.

Boulenger includes within this species specimens from Colombia, Venezuela,
Trinidad, and Tobago.

Urotheca dumerilii Bisron.

Brsron, Sagra’s Hist. Cuba. Rept., 1843, p. 218, pl. 24. Bovu.encrr, Cat. snakes Brit. mus., 1894, 2,
p. 181.

This snake is of peculiar interest if it is really the only Antillean species
existing on the Central American mainland. Neither Gundlach, Charles
Wright, nor any recent collector has succeeded in finding this species upon
Cuba. As with so many of the Cuban species sent to Paris by Sagra, it may
really be confined to the Central American mainland. The only specimen of
the species in this Museum was one sent by A. Aug. Dumeril from the Paris
museum. It was said to be a ‘‘type”’ in the loose sense in which the word was
then employed and may really represent one of the original series which probably
did not come from Cuba.

REPTILIA. 341

Arrhyton taeniatum GinTHER.
GunrTueEr, Cat. snakes Brit. mus., 1858, p. 244. Boutmencrr, Cat. snakes Brit. mus., 1894, 2, p. 252.
This was the first known species of this remarkable genus peculiar to Cuba.
In 1894, only the type was preserved in the British museum, and it still seems
to be extremely rare. In January, 1913, I procured a fine example from near

Cojimar, not far from Havana.

Arrhyton vittatum (GuNDLAcH AND PETERS).

Gunpuacu & PETERS, Monatsb. Akad. wiss. Berlin, 1861, p. 1002. BouLrencrr, Cat. snakes Brit. mus.,

1894, 2, p. 252.

Another rare Cuban species. The collection contains a typical example
from Soledad, near Cienfuegos, Cuba. It was given me by Mr. R. M. Grey,
who caught it in the Harvard Botanical Experiment Station. Mr. Barnum
Brown caught another example near the Rio Analla, Cuba, which is now in this
Museum.

Arrhyton redimitum (Cops).
Cork, Proc. Acad. nat. sci. Phila., 1862, p.81. Bounencer, Cat. snakes Brit. mus., 1894, 2, p. 252.

Apparently equally as rare as the preceding species. Also Cuban. I am
rather inclined to believe that these three species may in reality all represent
one variable form and be referable to a single species. Much more material
is needed to settle this point.

Clelia cloelia (Daup1n).

Daun1n, Hist. nat. rept., 1803, 6, p. 330, pl. 78. Boutznerr, Cat. snakes Brit. mus., 1896, 3, p. 108.
Oxyrhopus plumbeus WaGLER. GUNTHER. Garman, Proc. Amer. philos. soc., 1887, 24, p. 285.

The specimens upon which Garman based his record for Oxyrhopus plum-
beus Wagl., Giinther are in excellent preservation. No special comment upon
them is necessary. Allen obtained no specimens, and did not even hear of the
species, from which we may assume that it has either grown very rare or become
extinct. Garman’s example seems to be the only one known from Grenada;
he had two others, also fully adult, from St. Lucia. Boulenger (Cat. snakes
Brit. mus., 1896, 3, p. 109) records specimens in the British museum from St.
Lucia and Dominica. As is well known, this is a wide-ranging species, occurring
through tropical America from about the City of Mexico to southern Brazil.
In first recording these specimens from Dominica Boulenger (Proc. Zool. soc.
London, 1891, p. 352), remarked that they were “‘hitherto recorded only from
Trinidad in the West Indies.”” Garman’s West Indian records were published
some years earlier.

342 BARBOUR: ZOOGEOGRAPHY.

Fitzinger’s name Clelia, based on this very species of Daudin, antedates
the name Oxyrhopus of Wazgler, in common use, by four years.

Pseudoboa neuweidii (Dumérit ET Breron).

Duméritet Breron, Erpét. gén., 1854, 7, p. 1001. BouLencrr, Cat. snakes Brit. mus., 1896, 3, p. 112.
Scytale coronatum SCHNEIDER. Dumiérinet Brsron. Garman, Proc. Amer. philos. soc., 1887, 24, p. 285.

I have followed Boulenger (loc. cit.) in using this name for the three small
snakes (No. 4511, M. C. Z.), from Grenada which Garman collected and recorded.
They agree perfectly with Boulenger’s description of newweidii, while they may
be distinguished at once from Schneider’s species, coronatus, in that they have
each nineteen and not seventeen rows of scales. The original neuweidii of Dumé-
ril and Bibron was apparently a composite species, and it has been divided by
Boulenger into the above form and O. guerini Duméril and Bibron. There is no
material at hand by which to verify this separation, but from Boulenger’s descrip-
tions this solution of the question would seem to be excellent.

The species has been taken only this once in Grenada, so far as I can find
published record. It is probably the same species as that recorded from Trini-
dad by Mole and Urich (Journ. Trinidad field nat. club, 1894, 2, p. 86) as Scytale
coronatum Schneider. It seems to be confined to these islands, besides being
widespread on the mainland. Its occurrence is to be expected on Tobago. The
fact that neither Allen nor Brues found or heard of it in Grenada last year would
suggest that it has probably been exterminated by the mongoose.

I believe with Stejneger that the possession of undivided subcaudals is a
character of sufficient importance to warrant our separating these species generi-
eally from Clelia (Oxyrhopus auct.). Stejneger has shown that Pseudoboa is an
available name for these species (Proc. U. 8. nat. mus., 1902, 24, p. 187-188).

Ialtris dorsalis GiinrHEr.

GitnTHER, Cat. snakes Brit. mus., 1858, p. 126. BouLmncsr, Cat. snakes Brit. mus., 1896, 3, p. 137.

The only specimens in the collection are the two from Jeremie, Haiti, col-
lected by Dr. D. F. Weinland, which served Cope as types for his J. vultuosa
(Proc. Acad. nat. sci. Phila., 1862, p. 73). They are No. 3,600, M. C. Z.,
although Cope in the original description gave them as No. 1,519, M. C. Z. This
was the number from the old catalogue, which was long since discontinued,
unfortunately giving rise to confusion of numbers, and making the identification
of some of the early types a matter of difficulty.

REPTILIA. 343

Lachesis lanceolatus (LactrEpE).

Lactprpe, Hist. nat. quad. ovip. et serpens., 1789, 2, p. 98, 119, pl. 4, fig. 1.

There has long been question as to the real distribution and occurrence of
the ‘‘Fer-de-Lance” in the West Indies. It is probably at present confined to
the islands of Martinique and St. Lucia. Labat, in his Nouveau voyage aux
Isles de ? Amerique (La Haye, 1722, 1, p. 429) makes it quite clear that he knew
this snake well. He says, after remarking on the venomous character of a
snake which he killed in his hen-roost one morning :— ‘‘ Pour peu qu’on soit fait
au pays on distingue aisément la couleuvre d’avec le serpent, parce que la cou-
leuvre a la téte longue et ronde comme une aiguille, & que le serpent l’a plate,
& presque triangular.” This quotation, showing that the good Father was
quite sound in his identification, was followed later in his narrative (4, p. 105)
by the following interesting statement :— ‘‘On n’en voit (des serpents viperes)
dans toutes les Antilles qu’a la Martinique, Sainte Alonsie ot Lucie & 4 Bequia,
qui est un des Grenadins, qu’on appelle a cause de cela, la petite Martinique... .
On ne voit dans les autres Isles que des couleuvres qui ne sont point venimeuses,
& qui méme sont utiles, en ce qu’elles font la guerre aux rats. Wiles [sic] sont
rares 4 la Guadeloupe, et méme fort petites.”’ In connection with these quota-
tions, we may well observe also what Symington Grieve says in his Notes upon
the island of Dominica (London, 1906, p. 63-64) : — “‘Of snakes there are at least
three kinds, one of which — the tetche — grows to a large size; and I was told
of specimens of from fourteen to twenty feet in length. But the usual size, of
which I saw several, and killed one, was about six or seven feet. The other two
varieties I saw were quite small, and none of the snakes on Dominica are believed
to be poisonous.”’ So much then for the occurrence of this extremely well-known
and conspicuous species upon Guadeloupe and Dominica. Yet Boulenger
(loc. cit.) in his Catalogue records specimens in the British museum from both
these islands. Were they like so many other specimens recorded from this
region, caught upon one island, but shipped to the Museum from another? As
for the occurrence in the Grenadines which Labat mentions, it can only be said
that the species does not apparently exist there now; while the names of Bequia
and Petit Martinique are used for two different islands.

Regarding the identity of the West Indian Fer-de-Lance with that of the
mainland, the characters which Garman gives for his L. caribbaeus (Garman)
from St. Lucia do not seem distinctive. Yet there can be no possible doubt but
that St. Lucia specimens (Types No. 4,812, M. C. Z., and in British museum)
range lower in scale rows — twenty-five to twenty-seven rows in five specimens —

344 BARBOUR: ZOOGEOGRAPHY.

than do those from Martinique, twenty-nine to thirty-three rows in six specimens.
Thus apparently the Dominican specimen recorded by Boulenger probably came
from St. Lucia (having twenty-seven rows), and the Guadeloupe specimen from
Martinique (having thirty-three rows), which is exactly what one would expect
from the fact that the pairs of islands are respectively British and French.

While it is quite possible that the study of a large series of St. Lucian ex-
amples would firmly establish the fact that there was a peculiar species upon the
island for which Garman’s name is available, it is quite impossible to recognize
this race, and then to identify the snakes from Martinique with those of the main-
land. This would convey an entirely false impression, as to their origin and
relationship. The individuals on both islands are all derived from the lanceo-
latus stock, which is widespread upon the mainland. Until more material is
available, it is impossible to attack this problem satisfactorily. We can not,
however, recognize a peculiar form which divides the range of the widespread
species. This would be contrary to what we hold to be a zoégeographic law.

Since the preceding lines were written, I find that Mr. A. E. Brown has
recorded the scale counts for two specimens which were once brought alive to
the Philadelphia Zoological gardens, and which were said to have come from
Martinique. In these the scale rows were twenty-five in number. This fact,
coupled with the long known and singularly frequent inaccuracy of locality data
accompanying living specimens, which are shipped about from place to place
until they reach their final destination, is enough in itself to make it appear very
probable that these were really St. Lucian specimens.

Chrysemys palustris (GMELIN).

GMELIN, Syst. nat., 1788, 1, p. 1041.
Chrysemys var. rugosa (SHAW). BouLENGER, Cat. chelonians, etc., Brit. mus., 1889, p. 79.
Chrysemys scripta palustris GM. SrEBENRocK, Zool. jahrb. Suppl. 1909, 10, p. 464.

This somewhat variable form is the only fresh water tortoise occurring in
the West Indian area. I have followed Stejneger (Rept. U. 8. nat. mus. for 1902,
1904, p. 710) in ranking this form as a full species. There are specimens at hand
from Jamaica (type locality), Cuba, Haiti, and Porto Rico. They do not, how-
ever, vary in such a way as to warrant the designation of varieties confined to
each island. Stejneger andI (Bull. M. C. Z., 1910, 52, p. 301) had both expected
that such subspecies might be defined. As yet, however, the material available
for study is very meagre. The species is very shy, and is usually confined to
lowland muddy ponds.

Siebenrock (loc. cit.) follows Boulenger (loc. cit.) in recording the species

REPTILIA. 345

from both Guadeloupe and Martinique. There is no reason whatever for sup-
posing that the species occurs on either of these islands. This tortoise was re-
ported by Maynard as occurring on Grand Cayman Island, probably by importa-
tion from Cuba. It has not been found by more recent collectors. It is common
in most fresh water ponds all over Cuba and Jamaica. It frequently finds its
way into the Cuban markets.

Crocodilus rhombifer Cuvier.

Cuvier, Ann. Mus. hist. nat., 1807, 10, p.51. Bounmnemr, Cat. chelonians Brit. mus., 1889, p. 237.

This species is confined to Cuba, whence the Museum has several specimens.
It is locally known as ‘‘Cocodrilo.”” It is very closely related to Crocodilus
moreletiit, A. Dumeril of Guatemala and Honduras,— a rare species from which
the Cuban form has doubtless been derived. Quoting from the field notes of
Mr. C. J. Maynard, Garman (Bull. Essex inst., 1888, 20, p. 108) records the
following in writing of the herpetology of Little Cayman Island: — ‘Two
species of crocodile have been taken on this island, and one on Cayman Brace.
I saw but a portion of one specimen. The natives assured me the species were
similar to those found in Cuba.”’ The species referred to are doubtless this and
the succeeding. Whether they are resident in the Cayman group, or whether
they only reach the islands occasionally by swimming would be well worth as-
certaining definitely. As yet no answer to this question can be given.

This crocodile is apparently entirely confined to the submerged regions of
the Cienaga de Zapata and the Cienaga of the Islade Pinos. I found it abundant
in 1913 about the Laguna de Punta Gorda not far from where the Rio Hanabana
enters the great swamp. The three which I killed were from five to seven feet
long, but it grows much larger. In the Museum of the Institute of Secondary
Education in Havana there is an enormously massive skull taken by Gundlach
from a specimen nearly seventeen feet long. This species is very aquatic living
in the floating vegetation and half floating reed beds of the ‘‘tembladera”’ in the
Cienaga. In the region which I visited the swamp had a hard firm substratum
upon which was overlaid from two to five feet of very soft liquid mud and on the
surface of this half afloat were the great reed beds with more open areas of stagnant
water and lily pads of several sorts. This region supported an enormous popula-
tion of water fowl and waders as well as the crocodiles, which hide among the
roots of the cat tails when disturbed, or sun themselves upon the floating beds
of matted lily pads and water plants.

346 BARBOUR: ZOOGEOGRAPHY ;

Sa

Crocodilus americanus LAURENTI-

LAURENTI, Syn. rept., 1768, p.54. Boutenaer, Cat. chelonians, Brit. mus., 1889, p. 281.

This is the most common crocodilian of the West Indian region. It was
formerly abundant in Jamaica, though now it is largely restricted to the swampy
deltas about the mouths of the Milk River and Black River. It occurs in a few
other favorable localities, but is rare everywhere. It is generally called the
“alligator,” but the name of ‘“‘crocodile” is also quite widespread. In Cuba it
is more common, and occurs widely distributed in suitable localities. It is
generally known to the natives as Caiman, in contradistinction to ‘“‘Cocodrilo,”
a name used for the other Cuban crocodile (C. rhombifer Cuvier). In Haiti it
is said to be common, but of its occurrence in either Haiti or San Domingo
practically nothing is known. Outside of the Greater Antillean region it ranges
through Mexico, Central America to the Pacific coast region of Ecuador. In
Florida, whence it was first made known by Jeffries Wyman (Amer. journ. sci.
arts., 1870, 49, p. 105-106), it is now very rare. Its persecution by local hunters
has practically exterminated it about Biscayne Bay, and left but few individuals
about Cape Sable and among the northern Keys.

There are specimens in the Museum from Florida, Cuba, Haiti, Jamaica,
Mexico, Honduras, Nicaragua, and from both sides of the Isthmus of Panama.

.

a

& oa : ae on ihe otha =e crocodta & hare Capa 4 iad

Pie « ate opeanate in’ Sip een ne “Fowtta,
ae: coos date. Caapise ant from both aides vf i

349

TABLE OF DISTRIBUTION.

‘sopeqieg |

“epeuady) |

NOdeLIeO) |

‘onbrysnyy | ;

~“emboog |

"queout, “4g |

wroNy “9g; |

ronbrayaeyy |

“BoTUTUIOC, |

‘oVURpED) OLTe TY

‘apentsoq: | :

-adnojapens) ;

“VBII9S}UO Ty

“epuopoy
‘ensyUy
“epnqieg

“STAON

SHOT “IS
“equs
--mowopoujaeg “1g |

“ayaa

UTE “9g
— soroaquiog |
x10ID “9g |
‘SBMOWT, “3S
“uyOr “3g
—Bpoys0, |

“BUOT,

oe
é

BARBOUR: ZOUGEOGRAPHY. TABLE OF DISTRIBUTION. 349

Great Inagua.
Cuba.

Watling’s Island.

Reprivta

Great Bahama.
Great Abaco.
Little Abaco.

Eleuthera.
| New Providence,

St. Lucia.

Porto Rico.

Grand Cayman.
Redonda.
Montserrat. yl

Andros Island.
Cat Island.

! Long Island,
Crooked Island.
“Little Cayman.
Cayman Brac.
Dominica.
Mustique.
Carriacou.

Aristelliger praesignis (Hallowell) ny | [oad rg
Aristelliger lar Cope ; E ® |
Aristelliger nelsoni Barbour ® |

Tarentola cubana Gundlach & Peters +

Sphaerodactylus torrei Barbour |

Sphaerodactylus elegans Reinhardt & Lotken eel

Sphuerodactylus cinereus Cocteau o||

Sphacrodactylus nigropunctatus Gray Bott ae

Sphaerodactylus decoratus Garman ans 3 +i + e

Sphaerodactylus flavicaudus Barbour : 6

Sphnerodactylus alopex Cope a iaalltte ita : ®

Sphaerodactylus oxyrrhinus Gosse : 5 :

Sphaerodactylus dacnicolor Barbour 5 cal le

Sphaerodactylus argus Goase .| :

Sphaerodactylus argivus Garman GH hes » : ® | |

Sphaerodactylus notatus Baird 5 || a5) 5 |/ae : 5 + | inal

Sphaerodactylus exsul Barbour ® |

Sphaerodactylus difficilis Barbour : : Sah : ® | Tate

Sphaerodactylus corticolus Garman a : -|6 | | | |

‘Sphaerodactylus goniorhynchus Cope he VE hic = 1

Sphaerodactylus gilvitorques Cope 5 é Aig . a |

Sphucrodactylus copii Steindachner : ; : a ’ 7 |

Sphaerodactylus microlepis Reinhardt & Ihatken : 2 . -|- alee: = j= |

Sphaerodactylus pictus Garman ate whe ale E ot hee ls : ite | |

Sphaerodactylus vincenti Boulenger a4 he ‘ : 5 ; aie 5) fic al ie Gs : = fee lhe , : ae |

Sphaerodactylus melanospilus Bocourt oi 8 on bonts ; 5 5 =| gies é é “pF Real fend Ue = Ll |
|
|

+++0

+O+

Sphaerodactylus richardsonii Gray : on Oe ; 5 4
Sphserodactylus picturatus Garman 2 ' mbes C eo

Sphaerodactylus asper Garman : ®

Sphaerodactylus grandisquamis Stejneger P 6 oa) EA Ae i : . i i

Sphaerodactylus monensis Meerwarth : aha ; : ale |
Sphacrodactylus macrolepis Reinhardt « Litken a : abe ela a A : a) ote = | }
Bpliacrodaoty{usenutator (jarrnian): Ate 5 i Hote 2) LN he abe ote A= Veleohe Foal ea al | |

Chamaeoleolis chamaeleontides (Duméril & Bibron)| . | - 5 cae b
Xiphocercus valenciennesii (Duméril « Bibron) A Fal ee anced fe PY fer (ead cnt Pee

Anolis equestris Merrem a es Als 8\ ton na’ te ma Pelee : |
Anolis garmani Stejneger Oe We) Py EEN Ta Be AA bg ray Sars, Wied" hacey Wi) FS |

Anolis ricordii Duméril & Bibron EAR ace Veh mel na! aay |e 2a) bo fea feel Pea aly | | |

Anolis cuvieri Merrem Bl Pa ah Ren (ot ba a | eet ‘ + | |
Anolis gundlachi Peters ale 5 % yh Lae (ecb oat -of Potulets |
Anolis monensis Stejneger ME pe : A hy : Silat a + |
Anolis cristatellus Duméril & Bibron silts o ay hie re Pn Ue + oot eit esnl vais | i

Anolis stratulus Cope ad ba ey oe Uf A) lacey We] Weed BO Bus | Oye |
Anolis homolechis Cope Ne Sethe puesfs aa es

Anolis gingivinus Cope Siisalie eae lts yi |
Anolis alliaceus Cope |

SETS :

' a i :

351

sopra |
“epeuaisy) |

“ noaBinre,) |

‘anbysnyy |

“emboag |

“gueoul)\ “40 | 2 z ise) @
“wion’y “98 | - -@

_enbranzeyy | +

“eorULMo(y

ae “eyuepen ae yl © ®
—peaisoqy

-adnoyopensy fogs |

“7BIIAS} UO --@
“epuopay Seg Des iesc)

TABLE OF DISTRIBUTION.

“Bnsyuy

“upnqueg.

|

|

|

‘ent aysngy “4G

“Bqeg

|

-eymuy

“orauquiog |
“x10I9 “48
‘SBUIOUL “IS
—-aNor “3g
“BlOWIOL,
“epeaouy

‘sonbatA | * -

“BUOTAL |

oe eee eee

| ++

Repriuia

‘Anolis asper Garman

Anolis speciosus Garman
Anolis lividus Garman

Anolis sabanus Garman

Anolis nubilus Garman

Anolis leachii Duméril & Bibron
Anolis cepedii Merrem

Anolis griseus Garman

Anolis yincentii Garman

Anolis luciae Garman

Anolis extremus Garman
Anolis marmoratus Duméril & Bibron
Anolis bimaculatus (Sparrman)
Anolis richardii Duméril & Bibron
Anolis trossulus Garman
Anolis trinitatis Reinhardt & Litken
Anolis grahamii Gray

Anolis opalinus Goase

Anolis iodurus Gosse

Anolis conspersus Garman.
Anolis distichus Cope

Anolis distichoides Rosén
Anolis cybotes Cope

Anolis haetianus Garman
Anolis citrinellus Cope

Anolis poncensis Stejneger
‘Anolis evermanni Stejneger
Anolis krugi Peters

Anolis acutus Hallowell

Anolis isolepis Cope

Anolis lucius Duméril & Bibron
Anolis argenteolus Cope

Anolis wattsi Boulenger

Anolis lineatopus Gray

Anolis sagrei Duméril & Bibron
Anolis ordinatus Cope

Anolis nelsoni Barbour

Anolis greyi Barbour

Anolis bremeri Barbour

Anolis luteosignifer Garman
Anolis yermiculatus Duméril & Bibron
Anolis loysiana Cocteau

Anolis argillaceus Cope

Anolis semilineatus Cope
Anolis spectrum Peters

Anolis cyanopleurus Cope
Anolis alutaceus Cope

Anolis augusticeps Hallowell
Anolis porcatus Gray

Anolis maynardi Garman
Anolis brunneus Cope

BARBOUR: ZOOGEOGRAPHY. TABLE OF DISTRIBUTION. 351
Cee Lt eee) Pen Praie TH
| a)
- STA TSH) oH oles Wh cai a
|| 8 \s\al_|d| aldlale jedi) (4 |
a|4 HlElS|El=le\ als) Bia 3| ig Ate! s| le aie >
aa 22/3 |z|2/ele\z|= 2) lSielalalal lalaldlalale e/a|e)ai2lalalé
SV EVEUSIRIElalaig ie \z/eizlelaleletelal aie glglelsiclc aise elalalalg
_ja\s 8/2 |4/5/E |e )5|0)6|/6/S)0)4 Sl alms ie lel/elela a| 23/2 /2|2| |e
: 8
{3}
2 B
®
+
®
| S
|. | | c)
®
} + +
a
| 3)
4 ea
Hebe lieth | “
+
3 | (:)
- + - cS)
+
@
| @
| +
+
| +
Ni =
he +
| &
| | +
| | + +
: Sean] Voor) aus nl oe +
c}
{::)
| rs a
fh
+
| +
| a5
| +
GAA é: ci

t
ia

te
i i
ii

, fon flav- aplllarn g ee ee 2 =
gel, When thenol May “> achfrecied, Aaah renels Vs eu Mute Corner ; wes
lant as ete, Lenn VMs Abueh; ution dent Le as Antal

ma: Neat hody , 9F mor Mud bagthe, (2 may Hawker fre,
- cwatthey baal 1 mm, i fedagn, (F Wem - oe Lend
garry |
j -- age —atunf orto? FT <p jore fa ; @ coal brain |
ER
Pe OO heags ates eh ee Hie
ott aes ne ae | ka a he
fs ; a. ee, NI sei
Soar 1 Sippy icon
eat... Cos ehh Ai oreh facemore, Sh le Re | Peanening Hme., "

il
{
L
{

\

N Vat.
Afar 5 nfutata ?

GY : besa lrertiens
Fi

UA ; :
(Lex stig Leas le

}
Ad Ie Of. a non
Ken. anadee eee eo
AU Oe bee We

oh
ganged Annes de
Qnete Spall we

oT . LA Mprulas ee

a HA! Crs Lae of A
if Be pel. Aa
el Aug

ie Louth “ee

va

(p bed A. Tia man )

2p Pi,
hantionat chow AawrKte

ot, Vomells: hw oli: itp a Atnetry A LANFRO -QUARB 0
ee tale eh CHS. paste

: 2/2, if.

oS pe OV DX 4 a)...

Rolo. amos. Z)
(PK. motrue’

a

it kb AM gam)

ees zara ey

lade

ee oR genttes.
7

pat ntee Hyg bir hfe

(eo bi (asf pest ee

; = ay eS - m7

) <. oa

ee ste
oie ce

q om

f
tH WE uGAe

ar ae

a 7

Wages: Luuadbrufa
Aatesladi: lays: Janka
ie Thee, peop casedee Laubal Dances

th. Cans LAr tele |

7 pli ern ||

. an the APA

res ip uc Nbupherd a

Ath, : AM romreeneg Cita,

~ Or itebede APL cs

(ee alee Mec (6,2 Dy Rater set

aglfalend
( Lirvobeiy

Eee.
Co AManr® os

£. RBG, Df

Qe,
lay bcere ae ‘

A

VO 4 yy re.

|
i
|
|
j
{
i]
(

REPTILIA

BARBOUR: ZOOGEOGRAPHY.

| Swan Island. ,

| Great Bahama.

Great Abaco.
Little Abaco.

Eleuthera.

New Providence.
Andros Island.
Cat Island.

Grand Cayman.
Little Cayman.

Turk’s Island.
Great Inagua.

Rum Cay.

(Choaitedl Island.

Watling’s Island. cS
| Long Island.

Cayman Brac.

Anolis oligaspis Cope

Anolis leucophaeus Garman

Anolis chlorocyanus Duméril & Bibron ©
Anolis pulchellus Duméril & Bibron

Norops ophiolepis (Cope)

Iguana rhinolopha Wiegmann
Iguana delicatissima Laurenti

Cyclura eyelura (Cuvier)
Cyclura collei Gray

Cyclura baeolopha Cope
Cyclura rileyi Stejneger
Cyclura cornuta (Bonnaterre)
Cyclura carinata Harlan

Leiocephalus carinatus Gray
Leiocephalus virescens Stejneger !
Leiocephalus varius Garman
Leiocephalus loxogrammus Cope
Leiocephalus vittatus (Hallowell)
Leiocephalus macropus Cope
Leiocephalus schreibersii (Gravenhorst)
Leiocephalus melanochlorus Cope
Leiocephalus personatus Cope
Leiocephalus eremitus Cope?
Leiocephalus herminieri (Duméril & Bibron)

Celestus sagrae Cocteau
Celestus sepoides (Gray)
Celestus pleii Duméril & Bibron
Celestus impressus Cope
Celestus occiduus (Shaw)
Celestus rugosus Cope

Celestus crusculus (Garman)
Celestus maculatus (Garman)
Celestus costatus (Cope)

Cricosaura typica Gundlach & Peters
Centropyx intermedius Gray

Ameiva auberi Cocteau
Ameiva dorsalis Gray
Ameiva thoracica Cope
Ameiva chrysolaema Cope
Ameiva taeniura Cope
Ameiva vittipunctata Cope
Ameiva maynardi Garman

@ |

Lele |

1Found on Green Cay, south of New Pr;

BARBOUR: ZOUGEOGRAPHY. TABLE OF DISTRIBUTION. 558

~| Swan Island,

Little Abaco.
| Eleuthera,

Reprint

Crooked Island.

Bartholo)

Great Abaco.
Long Island.

Rum Cay.

| New Providence,
| Andros Island.

Anolis oligaspis Cope | : Sy es

Anolis leucophacus Garman 184 Ve (ios a be are a) |
Anolis chlorocyanus Duméril & Bibron Ke ; ea per van eae +
Anolis pulchellus Duméril & Bibron ?

Norops ophiolepis (Cope)

Iguana rhinolopha Wiegmann
Iguana delicatissima Laurenti

Cyclura cyclura (Cuvier)
Cyclura collei Gray

Cyclura bacolopha Cope
Cyclura rileyi Stejneger
Cyclura cornuta (Bonnaterre)
Cyclura carinata Harlan |:

Leiocephalus carinatus Gray Het
Leiocephalus virescens Stejneger ' |
Leiocephalus varius Garman Wes |p
Leiocephalus loxogrammus Cope alts
Leiocephalus vittatus (Hallowell) Nees
Leiocephalus macropus Cope 5
Leiocephalus schreibersii (Grayenhorst)

Leiocephalus melanochlorus Cope Hees ees thertiect eaten
Leiocephalus personatus Cope Heals: Watt Peoul call
Leiocephalus eremitus Cope? |

Leiocephalus herminieri (Duméril & Bibron) |

ae | | |
: +} 2

} bacmmnedevrtinies i.

Celestus sagrae Cocteau
Celestus sepoides (Gray)
Celestus pleii Duméril & Bibron
Celestus impressus Cope
Celestus occiduus (Shaw)
Celestus rugosus Cope

Celestus crusculus (Garman)
Celestus maculatus (Garman)
Celestus costatus (Cope)

Cricosaura typica Gundlach & Peters +|

Centropyx intermedius Gray

|
Ameiva auberi Cocteau
Ameiva dorsalis Gray
Ameiva thoracica Cope
Ameiva chrysolaema Cope
Ameiva taeniura Cope
Ameiva yittipunctata Cope
Ameiva maynardi Garman

|-
| | |
aya esatich | sit eta] | | | | | | | | ) | |

HAAR ILC UTTELLCEL ERIE TIL IAe1 1 | [elehalalahl ania

*Found on Green Cay, south of New Pr -: * Confined to Navassa, near Huiti.

Nai
Heya

'
:

-

‘sopEqaeEg, ; . ae

“epelialy de

TOOVIEIT, |

vonbysnyyy

“emboog

qUODUTA “49

“SIONT 4S |

onbrury.ae yy - | i

“BOLUTULO Cy |

“oUUBeY) OLE W

“Opealsoy

‘adnojapensy |

*YBITOSITO FAT SE ae te :

“eupuopary

“eNSTUy

“epnq.reg, i

‘STAD NY o ||

SOT TS |

veqng | ; +

“MouLOpoy reg, 4g | o ||

-eymsuy

UE “3S |

TABLE OF DISTRIBUTION.

‘OLOIQUIOG |

"s1019 “49 |

‘SBMOUL “39 + os

“uYoe 4g)

“G[O}10,{,

“epesouy

2 Confined to Navassa, near Haiti.

‘sonbar | |

“BUOY, : apo 4 |

“OSUIUMOT ULY 2) 1]1B TT a a ion oe ; =~ too ; hesitant! cate @

Teoreue a! RTT | PTR, i tact ine ; yy Steels a.

yidence.

‘She Fie f r

305

TABLE OF DISTRIBUTION.

‘sopeqreg
“epeuery

“NODBLLIED, |

5 ‘onbusnyy |

v“ernboog |
|

“yUadUIA “4g

“eION'T “99

on brmiy.1e yy

“BOTUTULO(T

‘oqURpeD OLE

‘apeasad

‘adnojepens)

“enSIUWy
~epnqieg

‘SIAON

SHIM “3S
“snyRysngy 4g
“eqeg |
“mauTOfOILeEg “4¢

-opmauy |

UTE “48
“OLa.IqUIOg |

“x01 “3g |

‘seMIOUL, 4S
“uyor “4g

“BOHOL, |
-epegouy |
= “‘sonborA.

-oory ojzog |
“BuO | ae

Patent |p eo St en ee

rs
e+ cae
° | | |
-@ | |
| rol
| |
J “8

ical
: E

BARBOUR: ZOOGEOGRAPHY. TABLE OF DISTRIBUTION.

e1
| | Pa Tea pte 3) oh
| et 3 lo | | } J | 'é) s
|) léla) (3\ | Isls al lala gl 8 E
Repriisa SIE AE EEE fl Bll |e] |g s| |g | |as
Sie | 2/2) 8) e/S/a\s/el4lsi= Ola) alZ) lel algl als =| 13 |ale lg
ee) sala) eleif(SlSi2\2\2) alelS) 2) a2) 4ls|slelaia al |g le 4/2
2/e/ ElE|Z/Elzlsle el elalclelctelgielele is elelelels zla\s Ae Ale
4\6/5|8|8|2)4/)2|8|5| JE/SSlOlRjOps|Rlalele | <lala n\a\a 25 als
Ameiva exul Cope ijl= mi (=
Ameiva alboguttata Boulenger an eke BS 3 +
Ameiya lincolata Duméril & Bibron i oy! (ee fees 5 +r
Ameiva polops Cope See pedal : e
‘Ameiva wetmorei Stejneger a5 EI 5 3 =
‘Ameiva corvina Cope as Ves Bs) ts fe (Oa : ‘eae
Ameiva atrats Garman 5 kel fs : ‘ Ke 7 ale Alc
Ameiva analifera Cope F Fd ieee lied WOM Sal ta We bes a Pr
Ameiva garmani Barbour 34 cat ke be ate i 4 les 6
Ameiva erythrops Cope ml (Re) fies Pn a O53" ‘ é P ce
Ameiva erythrocephala (Daudin) ates thats Pa lis ae, 7

Ameiya pleii Duméril & Bibron
Ameiva pluvianotata Garman 3) fa So Vay es oy Mes bed bo : Aye ie
Ameiva fuscata Garman 3 SM eal cole breatie : ’ : BW) Wea Pree ees FS (ay bead Ves bent a
Ameiya aquilina Garman

®

‘Tisporus fuliginosus Cope 1 |

Scolecosaurus alleni Barbour
Scolecosaurus trinitatis Barbour!

Gymnophthalmus pleii Bocourt of liad alllts Atte ll re ex sien Dee . - : ot ie! lio cil 2) feces Leos 5 Leet jas + +
Amphisbaena punctata Bell Bl staal collie then ean lee alba +

Amphisbaena cubana Peters Hee | Feyt [eae ect pane oy al be wha Pol pe

Amphisbaena innocens Weinland NN ateatcotaeel ese ieution keicoubre C)

Amphisbaena manni Barbour lie il (ier 3 jen beet Neg foc | end ae Mt Boal Uc ®

Amphisbaena cacca Cuvier al pot) fie bo Cin fect te eval oelicg +

Amphisbaena bakeri Stejneger 5 a let te) bel | wes -
Amphisbaena fenestrata Cope b ee fre! Na A 54 es ===

Mabuya spilonota Wiegmann ah) ASH as hoy SA aM fies | ce Peel ada Ba] Pe +

Mabuya sloanii (Daudin) : 24 bea] es 5 es Oe he 72) eS) re ay

Mabuya maboia Duméril & Bibron 3 Z 5 | Veal bead Us a ea . : Th. be 4
Mabuya lanceolata Cope F ate : : ; al Bs

Mabuya dominicans Garman 5 : 4 A # A Ss le

Mabuya luciae Garman oN ie i . = ? od al e}e
Mabuya aenea Gray My 3 pak Nan be “ rd ' = é _ a) ee . rb
‘Typhlops lumbricalis Linné Abs! faite =|= 445 4) See | + = = ahs
Typhlops pusillus Barbour S

‘Typhlops tenuis Salvin pelt atte is
‘Typhlops rostellatus Stejneger en eh aH ea Se P
‘Typhlops platycephalus Duméril & Bibron : ; Aehe 3 : y F : a=

‘Typhlops dominicana Stejneger ral US 5 al 3 IG - . =
Leptotyphlops albifrons (Wagler) + eA pe Meal ter] ('G - -|. ait:
Leptotyphlops bilineata (Schlegel) Ph gresd feck) he 5 : 2 ome |e E . a - E -
Epicrates subflavus Stejneger feted doped pebe dele feted. os

Epicrates striatus (Fischer) Be ee ea ee ee |

)

BARBOUR: ZOOGEOGRAPHY.

Crooked Island.

Great Banana

AMPHIBIA.

New Providence.
Andros Island.
Turk’s Island.
Great Inagua.
Grand Cayman.

Swan Island.
Eleuthera.

| Long Island.

nm =

o

Little Cayman.
Cayman Brac.

Jamaica.

Spelerpes infuscatus Peters

Hyla septentrionalis Boulenger ea epe tere titi there (ect Th cet is ells ant eel ee ten I toe ae Weta Rote
Hyla vasta Cope shite Salis 3 [Ne
Hyla dominicensis (Tschudi)
Hyla brunnea Gosse

Hyla lichenata (Gosse)

Hyla pulchrilineata Cope

Bufo longinasus Stejneger We] coo pee |i onl Secu sale onl) sol (Meme i etallcentis ula
Bufo ramsdeni Barbour Sia [eevee bed icray leer Nee yal Pencee | cual devel fossa I
Bufo peltacephalus Tschudi Healt eat a ove Pet eat et Uaioe aD oe dl Hes bt
Bufo gutterosus Latreille Vcc lite esos ec sre ence |
Bufo lemur Cope |

Bufo empusus (Cope) | Nee, Bec ele [h ea Heetnl EAR al bte e a ad L a tT

T

Eleutherodactylus ricordii (Duméril & Bibron) Heston eee iste lasts Weis lati lect eels
Eleutherodactylus dimidiatus (Cope) SUAS ie Ray (hoes, Haken Peer eM Reece a eset Bee | ma ceil a
Eleutherodactylus cuneatus (Cope) Dei aco ean Sobol hata ete eral volar el eas (ahs
Eleutherodactylus plicatus Barbour | | een |
Eleutherodactylus varians (Gundlach & Peters)

Eleutherodactylus luteolus (Gosse)

Eleutherodactylus jamaicensis Barbour

Eleutherodactylus weinlandi Barbour

Eleutherodactylus richmondi Stejneger Steed oe pesca Ml Ee iiesso Ns Slit oe
Eleutherodactylus lentus (Cope) epee eet es card et dhe Keke fs silt | Uae
Eleutherodactylus unicolor Stejneger

Eleutherodactylus monensis (Meerwarth) al ecras sp ae zed Petal Abe] feet eg Cae lies co pad
Eleutherodactylus auriculatus (Cope) as ea Ye a tela! Wonoss Nee ae ira Da loa ok hee peel ae [a Ect
. Eleutherodactylus antillensis (Reinhardt & Littken))

Eleutherodactylus martinicensis (Tschudi)

Eleutherodactylus johnstonei Barbour

Eleutherodactylus urichii (Boettger)

+

Leptodactylus pentadactylus (Laurenti)
Leptodactylus inoptatus Barbour
Leptodactylus validus Garman
Leptodactylus albilabris (Gimther)
Phyllobates limbatus Cope Sib Lofton tel smal eed cee Ne ce el el tel te
REpvILIA.

Gymnodactylus fasciatus Duméril & Bibron

Gonatodes albogularis Duméril & Bibron eval eet a eas tea ape eng eal SING a th bec ae
Gonatodes notatus (Reinhardt & Liitken) freer kes eae Piece oem [

Phyllodactylus spatulatus Cope Haga costed Ipeser ie ( [ema Me ae Pee Rey ic ee [cel Sra Nee

Thecadactylus rapicaudus (Houttuyn)

® +

— Authentic record

BARBOUR: ZOOGEOGRAPHY. TABLE OF DISTRIBUTION. 347

| Swan Ialand.

AMPHIBIA.

justatius,

Eleuthera,

af

St. Bartholomew.

Little Cayman.

Little Abaco.
‘Andros Island.
Cat Island.
Watling’s Island.
Crooked Island.
‘Turk’s Island,
“Cayman Brac,
St. Thomas,

St. Martin.

St. Croix
Redonda.

Anguilla.
Saba.
| Barbados.

Spelerpes infuscatus Peters

Hyla septentrionalis Boulenger S++ +). ++
Hyla vasta Cope : é
Hyla dominicensis (Tschudi) al ale ofr : :

Hyla brunnea Gosse : ae : +
Hyla lichenata (Gosse) BP rf Be Nese Ws) Kes | he) ke -
Hyla pulchrilineata Cope tia elke Pa es tes +

Bufo longinasus Stejneger wipro fis , i , 7 Ay
Bufo ramsdeni Barbour ep efede fps fede o} pe} -] 18

Bufo peltacephalus Tschudi sfPesfes | Oy (hag) heal bien +

Bufo gutterosus Latreille | Ae Piles lect scuba ties ss - +

Bufo lemur Cope rH es IF BUH SN opel Pca (as +
Bufo empusus (Cope) lees on iad Ut Oe 2 i + :
Eleutherodactylus ricordii (Duméril & Bibron) Fn eat en a oe

Elecutherodactylus dimidiatus (Cope) . 3 A
Eleutherodactylus cuneatus (Cope)

Eleutherodactylus plicatus Barbour

Eleutherodactylus varians (Gundlach & Peters) c Poa bed (ot) rc Yen UE od Mies
Eleutherodactylus luteolus (Gosse) ehebejatepef-efeahed =]. c +

Eleutherodactylus jamaicensis Bai bour . Si es Md fen hay) Wal rd ee) Bp é C3)

Eleutherodactylus weinlandi Barbour Bey poe ec) eal ee fg foe fio eel Nea ed Aj Heil ie C7)

Eleutherodactylus richmondi Stejneger . 4 | ror] Pera He fao.! Vc he Bis iia lueetes =

Eleutherodactylus lentus (Cope) Te eater beets Sales s : +

Eleutherodactylus unicolor Stejneger = . ol. eves dics By 24 bc =

Eleutherodactylus monensis (Meerwarth) 5 Br) boy bo) aod bes! co! Bel ty bay fe =

Eleutherodactylus auriculatus (Cope) mb he bral ale + '

Eleutherodactylus antillensis (Reinhardt&Litken)). . . . 5 alle : t -

Eleutherodactylus martinicensis (Tschudi) BP West ced Gc) a foal | fa leno. rs fs + - -— + O + 4+ - +
Eleutherodactylus johnstonei Barbour > ce
Eleutherodactylus urichii (Boettger)

+O+++

+@

Leptodactylus pentadactylus (Laurenti) . ay Loa Peal he al Use tS : + ‘ .[. +/-/=
Leptodactylus inoptatus Barbour PA lest Pion eels weed a (es , @
Leptodactylus validus Garman A leodesclan|reckcatben fe hadesstt ; : ; : a fe é ® Sy Pee
Leptodactylus albilabris (Ganther) ihe, betel oko key Marit: Peay 5 += ==
Phyllobates limbatus Cope rd fed Area pe 2 V3 HB boss Be ec ; +

Reprint.
Gymnodactylus fasciatus Duméril & Bibron Be bee) sl peel “Oy Hee Pee bed eal eben tee 4 ‘ ; ol=

Gonatodes albogularis Dumdéril & Bibron Huss scl he Pe fies Fal rls + a
Gonatodes notatus (Reinhardt & Ltitken) TE) aes ty sdl alla esis. fh at ese : Bye 2

Phyllodactylus spatulatus Cope alias pel eek BH ES 68 ic
‘Thecadactylus rapicaudus (Houttuyn) AEA ee, frend ie Jaa}. To
i i __ {dain

pelts) | RS Pe

= Authentie record + Coll. M. 6. Z. © Type Coll M. C. Z,

347

TABLE OF DISTRIBUTION.

‘sopuqtec |

“NoowEABD |
anne

“emboogy

“juoour, “48 |

“BION “4g

onbruyie yy

“VOTUTLO(T

-apUuTeE) OV

‘opeasocy
~ -adnopepeng,

*PBITISZUO TAT

“epnq.ieg

“SIAON
ar
‘sniyeysng “49g |
agus |

Swe
UT 4g |

‘OIaTquIOg |

-‘SBUNOYT, 49
uYyOf 38

~epedouy

‘sonbat, |

‘OOT"Y oylog
* “BUOT, |

‘ofutmiocr uve» rarer |

“B[O}10 J, ie

ate,

alt

@t 4

® Type Coll M. C. Z.

M.C. Z.

357

TABLE OF DISTRIBUTION.

‘sopequeg |

“epeuery |

NodevLI109) |

-onbysnyy |

“embooq |

“quaout, “4g |

“won “3g |

onbranreyy |

“BOTUTMOCy |

‘aqueey onepy |

“apB.isacy |

‘odnojapens)

“PBITOS} UO Ay |

“epuoposy

“ensyuy

“epnqieg

“STAO NT

“SHIM “48

‘snyeysnq “49 | Ss

“eqeg |

“MOTEOTOYW TE “49 |

eymauy |

“UnIeTN 3S |

“O1OIQUIOG |

x1019) 4g |

“‘Sseuloyy, “49 =

aw
lise Ri
Aa ioe

ed

OS thisreias

—— a

——

2 RFE Soot =

ee

a Ey
Sueur ye

357

TABLE OF DISTRIBUTION.

BARBOUR: ZOUGEOGRAPHY.

“sopqaugy

“upwuarp ae : : 2 ®

“NOOWLOD,

|
| _onbasnyy | iol
aa Mb 2 Tar
cy oe Tae ra
wong 4g : !
| vanbranan py a 2 5 = a
-morUTULOGy : + 5 5 +:
“UND ony : 5 i 5 1

ae eee ee

“opearsac]
semi. | Cae Se oan ;

qoasyuO WY | : : anti — !

“upuopayy
omnginuy = 0 i i 1
‘upnqmg, _
“SIAONT

“MouTOpOy Aug

| sojqniuy | x eae : : oe -®

Sa) : ; : = aa ae
myo 4g | 5 a es = = z Py: =

i : = = ' Se Se
+4 a + | : htt + eG as puke
+ ++
i)
(amo omer | = E a
sumarkg pus 4 ®
([ emcnrsy ee +4++1 ras + nore y
| sungiuuy quot F 1
[ pasmpeamg | 1 St aoe oe 3 re
pars p70015 Gees) Aa EE Ee ca a ied ae el ears |
| ~spuvpsy Suoy | : a aoe AGT ea i el
{ “keg wm Al
“PUNT 8,FuT TAL
spuesravp | an | hopin ec ane, Ene : a
“=puvysy soxpuy | hie oe Roem aie a a
OUaDIAORE, MON Le ee ae el
agnor. | a 7 Molo hie ae Resort ae 1
omay onaer | a rae ; ohe ears 6c |
~ en = — MEAG
wormyug nor |. Le ji em ane oe ae
rg | ee
| | E z E
Z 2 i as
| z PAS 5 f2¢ E eee Se rf
: | ED Bafa: g 2 BF apt = ag 3 feee.8 g
& | 5 = z = =) S i ea
| SeeeG" gh 522822 5 2 £ 53 S25 — BOCES aoe eee &
| e538 28 gese52 2 6 2 €: Bak § SASK SHO ge ge saCe
| RAW bible: LAE ee
| figges 23 222222 2 2 3 gai ¢ ggibsEseeesaiede 22
| 288888 aE Sessa 2 5 3 el one 2 Bee2R5 2522228252 32
eeeges 25 g23922 5 3 2 gn 215 2 2222 SEeeEGEEGEE TI
lsanana 22 G22222 8 2 § 2 G2) 2 PEP! a

+

REPTILIA

BARBOUR: ZOOGEOGRAPHY.

Epicrates monensis Zenneck
Epicrates chrysogaster Cope
Epicrates fordii Giinther
Epicrates inornatus (Reinhardt)
Epicrates gracilis (Fischer)
Epicrates angulifer Bibron

Boa hortulana Linné
Boa grenadensis Barbour

Tropidophis melanura Schlegel

Tropidophis pardalis (Gundlach)
Tropidophis maculata (Bibron)

Tropidophis semicincta (Gundlach & Peters)
Tropidophis conjuncta Fischer
Trophidophis cana (Cope)

Constrictor orophias (Linné)
Tretanorhinus variabilis Duméril & Bibron
Drymobius boddaertii (Sentzen)

Herpetodryas carinatus (Linné)
Herpetodryas vincenti Boulenger

Uromacer catesbyi (Schlegel)
Uromacer frenatus (Giinther)
Uromacer oxyrhynchus Duméril & Bibron

Hypsirhynchus ferox Giinther

Alsophis angulifer Bibron

Alsophis caymanus Garman

Alsophis fuscicauda Garman

Alsophis brooksi Barbour

Alsophis vudii Cope

Alsophis rufiventris (Duméril & Bibron)
Alsophis cinereus Garman

Alsophis rijgersmaei Cope

Alsophis porto-ricensis Reinhardt & Liitken
Alsophis melachichnus Cope

Alsophis antillensis (Schlegel)

Alsophis sanctae-crucis Cope

Alsophis anomalus (Peters)

Alsophis leucomelas (Duméril & Bibron)
Alsophis sibonius Cope

Alsophis bruesi Barbour

Leimadophis ater (Gosse)
Leimadophis callilaemus (Gosse)

:
ie
| :

| Great Bahama.
| Great Abaco.

| Little Abaco.

| New Providence.
| Andros Island.

| Eleuthera.

| Cat Island.

Tat | ean 1 >
{ | 2)
eet Fog i =p
Ieee feat tae Ohara | | ta Az
5 ae erate: | |
= oie aanel cal ied yl ge Pe es
Ss Sid! a) 1; Si) a1] ae)
oa | Fipetet erst Pociapes eles 3 | =
La Zia| 3| &) | Be lpeel eat
Jost By eats sy Bs] >| A) } ol
Pe ate | ah || es hia [eal ay at ees | 2 |@
oo] Sia} oi=)8) 101d] ais]
Fp OT? ert ire a | 2|.2|/3
lg] lore! @| ei aiS) 8) 2\s
S| 5/8/215| 8/31 2\2| 2| ela
ial = a
BIS|/AIOle|/S/01S6/3/0/5/ a
| ee ee | | eae
|
Let
| |
| i
| |
| ad
|
pe | —
| eGBs [| Bae
| ers |, eel af aoe
| | ee as | ee,
| | | |
peas) areal heat ens 10
| | Testes
Peed inks eee Pu) aaa
| ie al [eo eal
hates ae tee. ee 2)
nl a pitetee|| i+
| {icxent late Gal ate
| | |
| ieee
Co Pa fe [it a
| Peal Mae Ls) Jee (ee) |
| eee he |
fee at aaa
| | eel atest
eee faces |
| if
1S]
[eee sl rae acide
i | 1 iy
| {| | Hee SR
| Be faae eer |
i Ses athe) |
(Pee eters = |
| ee jaa
eon, pee liste
| | Heed erie
| ts 5 ait
| eres [ae j
Hera | js si
{iol emcee [22 ST ait |
| [in seat | SS er oe
[ao eal | eat | |
Pees Boss ie rast } ‘Tacs
ieee 2B betel liso. coll ete oat | }
laetsllaersh ,. t+ | |
eee iets he Gil Peacoat
eS We ae eee
| ak se ay 7)
| ese a |
eewellepoee|
lice laa jd
fa er pee es | is)
Meshes limes Py | | |
bc ae el en
bia | | =
he ieee
| |
| | prado, ec FG eee
aA b | ee oH Seal ls 4
| race irk) part's |
| . PAP a bites *
| isa Risk Final
| | 144i 1 eal
| |
Jo ae | har eal hee |
| Saal alee] eel ea [bs
| (eee oll oa aa ey

“NOOBIEIGD |

anbysn yy |
“ernboag: |

“qaaout, “4g |
SELON
anbrunjie yy

“GOTULUIO(T

“ayUR[Uy one],

-apeaisaqy

-adnopapensy

“quAIOSPUOTNY
“gpuopexy
“ensyuy
“epngieg

“staan |

‘SHIT 4S |

“snQeISNG “AG |

“aque |
“MOUIO[OUILEE “39

“eyMsay |

unreyy 49 |

“o1oAquIog |

S
2
5
Z
z
=)
6
sl
B
&

“xO “4g |
‘SLO, 4S) |
ayor ag |
“BIO}IOJ, |
“upesouy |
-sanbor\, |

Cov 03104

359

TABLE OF DISTRIBUTION,

BARBOUR: ZOUGEOGRAPHY.

squvout A 4S

ques} HoT

“upuopayy

“SN WySNgE IS,
auqug
AOUIOLOUIAEL IS

*OIDIGUIOS

suyor 4g '
| “TIOMO, i

} = " |

eaeeoET a Ea ae os

puns 307 |
Fug eam |
“puupsy salam |
purus 90) |
~spumpsy soxpay |

*gouaPLAOA ANON,

oonqy ONT
“oouqy Wat)
~murTuEy ANE)

=puUpsy UWA

REPTILIA
temporalis (Cope)
odilus americanus Laurenti

hyton vittatum (Gundlach & Peters)
udoboa neuweidii (Duméril & Bibron)

theca dumerilii Bibron
hyton ridimidum (Cope)
lelia cloelia (Daudin)
tris dorsalis Ginther
cheais lanceolatus (Lacépéde)

Leimadophis andreaeae Reinhardt & Litken

4rhyton tacniatum Ginther
Srocodilus rhombifer Cuvier

f=

prysenys palustris (Gmelin)

Leimodophis

ay
AR (P'S
Nee

PLATE.

ELEUTHERODACTYLUS JOHNSTONE! Barbour. Page. 249.

. 1— Dorsal view.
Fig.
Fig.
. 7— Ventral view of foot.
. 92— Lateral view of head.

2.— Dorsal view.
5.— Ventral view of hand.

ELEUTHERODACTYLUS URICHI Boettger. Page 251.

. 3.— Dorsal view.
. 4— Dorsal view.
. 6— Ventral view of hand.

8.— Ventral view of foot.

. 10.— Lateral view of head.

SCOLECOSAURUS ALLENI Barbour. Page 315.

. 11.— Dorsal view of head.
. 12.— Lateral view of head.
. 13.— Ventral view of head.
. 14— Dorsal view of hand.
. 15.— Dorsal view of foot.

Herpetology of the West Indies.

Helotype €o.Boston

\ oped

PUBLICATIONS

OF THE

MUSEUM OF COMPARATIVE ZOOLOGY

AT HARVARD COLLEGE.

There have been published of the Buttetin Vols. I. to LIV.;
of the Mrmorrs, Vols. I. to XXIV., and also Vols. XXVL. to XXIX.,
XXXI. to XXXIV., XXXVI. to XXXVIIL, XLI., and XLIV.

Vols. LY. to LVIII. of the BuLLettn, and Vols. XXV., XXX.,
XXXV., KXKIX) ML OUI MI MOL VN to LV Lor the
Menmorrs, are now in course of publication.

A price list of the publications of the Museum will be sent on appli- a
cation to the Director of the Museum of Comparative Zoology, Cambridge,
Mass. a

Barbour, Thomas
AUTHOR

A Contribution to the Zodgeography
Of"the West Indies, with special
reference to amphibians §& reptilac

SMITHSONIAN INSTITUTION LIBRARIES

iii

9088 00054

|

958 44