Bulletin of the Museum of Comparative Zoology 
AT HARVARD COLLEGE 
Vol. 128, No. 3 
STUDIES ON AMPHISBAENIDS (AMPHISBAENIA, 
REPTILIA). 2. ON THE AMPHISBAENIDS 
OF THE ANTILLES 
By Carl Gans 
and 
A. Allan Alexander 
Department of Biology 
The University of Buffalo 
Buffalo 14, New York 
With Twelve Plates 
AT HARVARD COLLEGE 
CAMBRIDGE, MASS., U.S.A. 
December, 1962 
OCR text unavailable for this page.Bull. Mus. Comp. Zool., Harvard Univ., 128(3) :65-158, December, 1962. 
No. 3 — Studies on amphisbaenids (Amphisbaenia, Reptilia) 
2. On the amphisbaenids of the Antilles. 
By Carl Gans 
and 
A. Allan Alexander 
TABLE OF CONTENTS 
Page 
Introduction 69 
Material 74 
Procedures 75 
Sex 75 
Annuli and segments 75 
Head sealation 77 
Cloacal region 81 
Body proportions 81 
Appearance of eye 81 
Generic status 82 
Key to the Antillean amphisbaenids 83 
Taxonomic discussions 85 
Genus A mphisbaena 85 
Discrimination of the species -. 85 
Interrelationships of the species 92 
Generic description 94 
Amphisbaena cubana 95 
Geographic variation 95 
A. c. cubana 97 
A. c. barbouri 100 
68 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
Page 
Amphisbacna ivnocens 101 
Geographic variation 101 
Nomenclature 107 
A. i. innocens 107 
A. i. caudalis 110 
A. i. gonavensis Ill 
Amphisbacna manni 113 
Amphisbaena balceri 121 
Amphisbacna caeca 123 
Amphisbacna fcncstrata 131 
Genus Cadea 135 
Discussion 135 
Cadea blanoidcs 136 
Cadea palirostrata 139 
Acknowledgments 142 
Literature cited 142 
Appendix 151 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 69 
INTRODUCTION 
In his third Antillean checklist, Barbour (1937) recognized 
two genera and nine species of amphisbaenids on islands of 
the northern Caribbean. The names of these species and their 
approximate ranges (see Figs. 1-4 for all localities) were : 
Amphisbaena cubana Peters Cuba 
Amphisbaena innocens Weinland Southwestern Hispaniola 
Amphisbaena caudalis Cochran Grande Cayemite Id., Haiti 
Amphisbaena manni Barbour Hispaniola, except the Southwest 
Amphisbaena balceri Stejneger Northwestern Puerto Eico 
Amphisbaena caeca Cuvier Puerto Eico 
Amphisbaena fenestrate (Cope) Virgin Islands 
Cadea blanoides Stejneger Western Cuba 
Cadea palirostrata Dickerson Isla de Pinos 
In 1951 Vanzolini (1951a) resurrected Cope's genus Diphalus, 
originally based on the species fenestratus. Consideration of the 
status of this supposedly monotypic genus, endemic to a small 
group of islands midway in the Caribbean, led to our becoming 
interested in the relations of the other Antillean species of 
Amphisbaena. Besides forming one of a continuing series of 
studies and notes on the Amphisbaenia, this paper was designed 
to supply the West Indies discussion group with a summary of 
yet another group of Caribbean animals. We have included the 
two species of Cadea in keys, maps, and summaries, though the 
information presented rests in part on the work of Zug and 
Schwartz (1958), and on some unpublished data kindly made 
available by the latter. 
While two studies of the variation within single species of 
amphisbaenids have recently been published (Vanzolini, 1951, 
1955), there exists no comparison of the range of variation within 
and between several closely related forms of amphisbaenids. An 
attempt at such analysis might offer a test of the occasionally 
stated and often implicit hypothesis that the variation of one 
common species may serve as a guide to the variability of rarer 
forms. 
No attempt has been made to deal with the broader relation- 
ships of the forms discussed, as it is intended to treat this matter 
jointly for the entire family at a future time. Zoogeographic 
discussion has also been kept to a minimum, derived primarily 
from internal evidence. Extended discussion of this topic is best 
handled for the fauna as a whole, and the zoogeographic scheme 
then synthesized from a maximum of independent analyses of 
individual groups. 
70 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 
71 
> 
03 
«H 
O 
'-3 
.1 a ' 
- 
- 
2 05 <u 
5 ° rt 
2 m o 
33 
3 
^ O * 
v 
t3 
PL, 
,rj o 
CO 
I— I <H 
o 
o 
ft 
>1 
© oo 
« 03 
^J 'r< GO 
00 ^ O 
3 bj a 
Ph 
^ s * 
*g fl CO 
^ a 
^ "3 I 
cS j5 to 
£ ° ^ 
« o I 
,3 
CD 
— o 
a c3 
SsJS 
ft ^ 
do hrt «n 
" o 
03 .3 
a> -<- 
c § 
<x> oo 
o ^ 
OO 03 
oj -i-i 
co 
O 3 
t3 
> 
O 00 t3 .£ 
^ t* *ri 
3 P 03 a> 
72 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
O 
cs 
oo 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 
73 
o 
oo 
^ 
OS 
OS 
3 
cs 
3 
o 
Sm 
O 
w 
as 
— 
OS 
3 
o 
as 
as 
as 
Eh 
o 
o 
■r. 
as 
- 
o 
o 
74 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
MATERIAL 
The specimens for this study came from the following col- 
lections, listed by abbreviations throughout: 
AMNH The American Museum of Natural History, New York, 
New York 
ANSP The Academy of Natural Sciences of Philadelphia, 
Pennsylvania 
BM British Museum (Natural History), London 
CAS California Academy of Sciences, San Francisco, Cali- 
fornia 
CM Carnegie Museum, Pittsburgh, Pennsylvania 
CNHM Chicago Natural History Museum, Illinois 
HM Zoologisches Museum, Hamburg, Germany 
KM Universitetets Zoologiske Museum, Kobenhavn, Den- 
mark 
KUMNH University of Kansas, Museum of Natural History, 
Lawrence, Kansas 
MCZ Museum of Comparative Zoology, Cambridge, Massa- 
chusetts 
MHNP Museum National d'Histoire Naturelle, Paris, France 
MJ Miguel Jaume collection, Museo y Biblioteca Zoologica 
de la Habana, Cuba (data from Schwartz) 
PM Peabody Museum of Natural History, Yale University, 
New Haven, Connecticut 
RMNH Rijksmuseum van Natuurlijke Historie, Leiden, Hol- 
land 
SMF Senckenbergische naturforschende Gessellschaft, 
Frankfurt a. M., Germany 
SU Stanford University Museum, California 
UH Universidad de la Habana (Museo Poey), Cuba, (data 
from Schwartz) 
UMMZ University of Michigan Museum of Zoology, Ann 
Arbor 
UPR University of Puerto Rico, Mayagiiez, Puerto Rico 
USNM United States National Museum, Washington, D.C. 
VM Naturhistorisches Museum zu Wien, Austria 
ZMU Zoologisches Museum der Universitat, Berlin, Germany 
Specimens cited, but not seen, have been preceded by an asterisk 
(*)• 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 75 
PROCEDURES 
For the present study we have spot-checked sizeable samples 
for most characters previously mentioned by authors. We have 
then reviewed the entire series for those characters found useful 
for taxonomic discrimination during the initial analysis. Particu- 
lar attention has been paid to sources of error in data, and dif- 
ferences from values presented in the literature have been 
reviewed though not necessarily mentioned here. Procedures and 
comments on the value of certain of these items are given below. 
The raw data for most characters are given in the appendix. 
Their publication eliminates the need for extensive discussions 
of possible trends in geographic or individual variation. It is 
hoped that they will also facilitate the assignment and analysis 
of material that may be collected or become available in the 
future. 
Sex: Amphisbaenid hemipenes are small, encased in layers of 
fat and connective tissue, and often poorly preserved (Vanzolini, 
1955, p. 691). Accurate determination of sex can only be made 
by dissection of the gonads and associated structures (of which 
the left always lies slightly anterior to the right) . This somewhat 
destructive technique had to be restricted to the two largest 
series (57 A. manni from Los Bracitos, and 59 A. caeca from 
10 km. east of Loiza). These demonstrate no significant sexual 
dimorphism and data from both sexes are lumped here. 
Absence of sexual dimorphism is not surprising as only pre- 
cloacal pore number and size had been shown to vary in other 
species of amphisbaenids (Gans, 1960, 1961). Zug and Schwartz 
(1958) had suggested sexual dimorphism in relative tail length 
and caudal annuli number, but a check of the gonads of specimens 
examined by them showed sufficient error to invalidate their 
results. 
Annuli and segments: All counts were made along the left 
side of the ventral surface, from the first postoral annulus up 
to and including the annulus bearing the precloacal pores (Van- 
zolini, 1951). Irregularities were recorded. 
Graphic analysis of the number of body annuli in samples of 
A. caeca from the vicinity of Loiza suggested that the ratio of 
odd to even counts differed significantly from 1. This did not 
seem to be the case for the few large samples of the other species, 
and the point should be checked when additional specimens be- 
come available. Such asymmetry is not surprising from a theo- 
retical viewpoint, as the dermal annuli lie two to a myotome or 
76 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
vertebra (Smalian, 1884, p. 184). Three sources of variation 
then exist for the counts of dermal body annuli: (1) Variability 
of myotome number. (2) Variability in the number and position 
of intercalated half -annuli (which do not receive dermal innerva- 
tion and may represent supernumerary structures). (3) Varia- 
bility in the position of the first ' ' body ' ' annulus, which may be 
more or less advanced onto the head so that it will be the fourth 
to eighth (?) annulus that corresponds to the first vertebra. 
These factors may explain the rather platycurtic distribution of 
body annuli, and may well introduce some complications in more 
advanced statistical treatment (Klauber, 1941, p. 8 et seq.). 
Some of the published counts for A. innocens differed suffi- 
ciently from ours to warrant a recheck. This check disclosed 
geographic variation in the number of supernumerary dorsal 
half -annuli. This character, the so-called "difference," has been 
plotted in some graphs (but omitted from the appendix). It 
represents the difference between a middorsal and a standard 
mid ventral count of body annuli. 
The caudal annuli were counted from the first complete 
(=ventrally not reduced) post-cloacal up to and including the 
last complete annulus showing regular segments. The half-annuli 
dorsal and lateral to the patch of modified cloacal scales have 
been referred to as "laterals." This nomenclature results in 
counts differing from those occasionally reported in the litera- 
ture, in which laterals seem to have been included with caudals. 
Here, as elsewhere, asymmetric observations are written : left/ 
right. Our separation of the lateral from either body or caudal 
annuli rests upon the observation that the doubling of dermal 
segments per myotome is a characteristic restricted to the body 
(Gans, ms.). The dermal segments of the tail occur one to a 
myotome and the change takes place in the cloacal region. 
In the appendix the annuli counts are given as number of 
body annuli, plus ( + ) number of laterals (left/right), plus 
( + ) number of caudal annuli. When the tail is autotomized the 
number of caudal annuli is recorded in parentheses. The problem 
of caudal autotomy in the amphisbaenids is intended as the sub- 
ject of a subsequent discussion. 
The dermal segmentation of amphisbaenids allows a complete 
shift of the loosely attached skin "tube" both rostrad and 
caudad. The slack is taken up by a curving or narrowing of 
the length of each annulus (Gans, 1960, p. 183). Diametric- 
expansion (thickening of the animal's body) occurs primarily at 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 77 
the lateral grooves or lines. These grooves also seem the pre- 
ferred site of dropping out of extra dorsal or ventral half- 
annuli. A secondary site for this is the general middorsal or 
midventral region, site of grooves in some genera, of a discon- 
tinuous suture only in Cadea. 
The segments of an annulus are far less constant in shape 
than are the annuli ; they vary in proportion both within and 
between successive annuli. Their sutures may or may not align in 
longitudinal series on an animal, the frequency of such alignment 
decreasing from the midventral to the middorsal line. Counts of 
segments were made at midbody, continuously around a single 
annulus selected at random, and recorded as dorsals/ventrals. 
The relative sizes of dorsal and ventral segments have occa- 
sionally been used as species indicators. These relative sizes 
change (with the number of segments) from front to rear of the 
animal. The relative longitudinal curvature of an annulus, or the 
amount of contraction in the region where such measurements 
are made, influences them decidedly, so that refined measurements 
are useless. Only approximate estimates are given in the de- 
scriptions of the species. 
Head scalation : The paucity of landmarks on the amphis- 
baenid head makes it difficult to compare head scales. Only 
rostral, labials, nasals, oculars and mental can easily be defined 
by standard methods. The absence of a parietal organ, foramen 
or eye does not even permit estimation of head scale relation 
from the position of the interparietal. The angle of the gape 
marks the posterior extremity of the large head shields and the 
anterior level of the clearly defined body annuli, which in 4m- 
phisbaena extend onto the occiput. Either of the two pairs of 
large scales following caudad upon the rostral and nasals may 
be considered the "true" frontals (i.e. those scales between 
the orbits), as their relative length and with this their position 
varies intra- and interspecifically within Amphisbaena. Decision 
for either pair makes the other pre"- or postfrontals, respectively. 
Our diagram (Fig. 5) presents a suggested nomenclature that 
will be followed in the subsequent papers of this series. Its use 
does not imply homologies with scales called by the same name in 
other squamate families. 
All our species of Amphisbaena have three supra- and three 
infralabials, and this characteristic has not been tabulated. Con- 
fusion is occasionally caused by the apparent caudad displace- 
ment of the angle of the mouth, because of its coincidence with 
an intersegmental suture of the first body annulus. Labials 
78 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
DO 
Hogl 
Fig. 5. Diagrammatic dorsal, lateral and ventral views of the head scala- 
tion to illustrate nomenclature. 1. Kostral, 2. First supralabial, 3. Second 
supralabial, 4. Third supralabial, 5. Nasal, 6. Prefrontal (frontal of some au- 
thors), 7. Frontal (postfrontal of some authors), 8. Parietal (occipital of some 
authors), 9. Ocular, 10. Postocular (no, or point contact with supralabials), 
11. Temporal (broad contact with second and third supralabials), 12. Mental, 
13. First infralabial, 14. Second infralabial, 15. Third infralabial, 16. Post- 
mental (median chin shield of some authors), 17. Malar, 18. First row of 
postgenials, 19. Second row of postgenials, 20. Third row of postgenials, 21. 
First row of poslmalars (included between last infralabials), 22. Post- 
malar scale (demonstrating a. common kind of asymmetry), 23. First body 
annulus (the cross-hatched segments indicate start and direction of 
counts). 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 79 
should be defined as the scales bordering the actual gape, i.e. 
those scales that separate when the lower jaw is depressed. Seg- 
ments caudad to these are not labials, even when they are en- 
larged. 
The constancy of head scale pattern seems to decrease with 
distance from the rostral. A zone of constant shield arrangement 
covers the snout of the several species. The gap between this zone 
and the first body annulus is filled by scales arranged in a more 
variable pattern which has contributed to nomenclatorial con- 
fusion in the description of the head scales (cf. Grant, 1946, p. 
61). While variation exists, it is the constancy of certain scales 
that deserves particular emphasis. Every specimen has the double 
pair of "frontals" and where large scales exist caudad to these, 
they are restricted to one or two middorsal pairs of parietals 
(only A. caeca from eastern Puerto Rico suggests some deviation 
from this). On the chin there is always a single postmental, 
flanked posteriorly by a single large malar on each side. 
The constancy permits the coding of certain exceptions, useful 
in the discussion of variation. First, there are three items listed 
as "Major" (and coded by the letters in parentheses under this 
rubric in the appendix). Deviations from N, for normal, as illus- 
trated in Figure 5 are : G, for cubana, the fusion of the ocular 
to the second supralabial ; M, for manni, the fusion of the rostral 
to the nasal with a single suture connecting nostril to first 
supralabial ; and F, for fenestrata, the caudad projection of the 
dorsal tip of the rostral which separates the nasals and inserts 
for some slight distance between the prefrontals. 
Analysis suggests that the segmentation back of the region 
of constant head sealation may best be understood in terms of 
body annuli. We have defined the first body annulus as the one 
including the segments back of the angulus oris, even where the 
postlabial segments are enlarged and seem to form a continuation 
of the labial series. Using this definition and the concept that 
accessory dorsal half -annuli may be^present in the nuchal region, 
we can then discuss the several enlarged segments in serial terms. 
The first series would consist of the temporal(s), postoculars, 
and possibly the frontals and would represent an accessory half- 
annulus in the "standard" diagram (Fig. 5). One of the follow- 
ing series would include the first (and another the second) pairs 
of parietals as middorsal segments. Complications due to the 
fusions, segmentations, and other abnormalities of the temporal- 
nuchal region become less disturbing with this approach. Thus 
80 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
the segments of a second dorsal half-annulus, excluded from 
middorsal contact by an anterior shift along the lateral edges 
of the frontals, need not be characterized by the undescriptive 
term of "extra temporals." 
The variation in number of intercalated half-annuli reflects to 
some extent both the length to the back of the frontal and the 
size of the parietals. This is immediately apparent when com- 
paring such species as A. cubana and A. manni (Figs. 13 and 
24). In A. cubana the dorsal halves of the anterior annuli swing 
forward and only the third body annulus lies in a plane normal 
to the long axis of the animal. In A. manni even the first an- 
nulus runs straight dorso-ventrally. Instead of devising a code, 
we have, in the appendix, recorded measurements (read off a 
vernier caliper to the nearest .1 mm, under a dissecting scope, 
and representing the shortest distance between points) from 
snout tip to back of frontals, and from snout tip to back of 
parietals (usually equal to back of first body annulus). These 
measurements are but a few of a somewhat more extensive series 
made during the course of the study. The remaining ones gen- 
erally showed differences already indicated by other characters, 
but did not suggest additional trends. They have been omitted 
from the appendix, though some are used in graphs. While such 
head length measurement should theoretically be plotted against 
snout-vent length minus "head" length, we have omitted this 
nicety as unjustified at this level of comparison. 
It is of interest to note that the species within this group 
agree much more closely in relative length from snout to back of 
first (or second) parietals, than in length of snout to back of 
frontals. It is speculated that this agreement reflects a basic 
functional pattern. What is measured is the distance along which 
the snout is covered with enlarged segments. If this segmental 
enlargement represents an adaptation for friction reduction 
(Vanzolini, ms. ; Gans, 1960), one would expect a correlation of 
the distance-covered with behavior. It may be that the number of 
participating segments is less significant than the matter of 
absolute distance. The several forms would seem to have solved 
the problem by distinct methods. The idea that enlargement of 
the head segments is a secondary or phylogenetically advanced 
development is supported by: (1) the increased size and fusion 
of shields in various amphisbaenid lines, and (2) the observation 
that specimens will often regenerate scarred areas as numbers 
of small segments rather than a single large one. Such evidence 
for a putative ancestral condition seems to have been accepted by 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 81 
Camp (1923) and other authors. 
The chin shield pattern shows variation in relative length of 
infralabials and postmentals-malars. This is best indicated by 
the number of postgenial (between the malars) and postmalar 
(between the third infralabials) scale rows. Four patterns 
(coded in appendix under the heading CHIN) exist: (a) two 
rows of postgenials — no postmalars, (b) two postgenials — one 
postmalar, (c) two postgenials — two postmalars, and (d) three 
postgenials — no postmalars. The number of segments in each 
postgenial and postmalar row have also been listed. The post- 
malar row often shows considerable asymmetry, with one side 
appearing to have been separated from the malar by a suture 
(item 22, Fig. 5) while the other is subdivided into a regular 
series of segments. The medial edges of the malars often show 
a similar cutting off of segments, which explains some of the 
variation in postgenial segment counts. 
Cloacal region: The cloaca is surrounded by a ring of segments 
that project some distance into the proctodeum, and are pro- 
lapsed during defecation. Their number and relative shape 
is subject to considerable variation and asymmetry. A deep 
transverse fold subdivides the cloacal rim into anterior and pos- 
terior portions. The anterior group (precloacal segments) is 
generally larger and more closely jointed, thus forming a shield 
for most of the cloacal opening. The shield is slightly scalloped 
in the center of its posterior margin. The precloacal pores lie on 
the midventral segments of the last body annulus rather than 
on the precloacal segments. The pores show some, possibly onto- 
genetic and sexual, variation in size and pigmentation. Every 
kind or size of pore was counted and the count has been recorded, 
followed by the numbers of pre- and postcloacal segments. Counts 
included all segments regardless of size, though segments lying 
in line radially from the cloaca were counted as one. 
Body proportions: Snout-vent and tail measurements were 
read by adpressing the specimen to" a meter stick. Readings of 
snout-vent length were recorded to the nearest mm, those of tail 
length to .5 mm. 
Appearance of eye: All amphisbaenids have eyes. As with 
other cephalic structures there is a decrease in their relative 
size during life. Observations also suggest a thickening and 
gradual pigmentation of the overlying skin with age. This, plus 
the depigmentation of the retina and reduced translucency of the 
skin due to the vagaries of preservation, is responsible for the 
comments on "invisible'' eyes and "blind" (= caeca) forms 
(cf. Gundlach, 1881). 
82 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
GENERIC STATUS 
Five generic names have been used for one or another of the 
species discussed here. These are : 
Amphisbaena Linne, 1758. Type species : A. fuliginosa Linne 
by first reviser (Fitzinger, 1843). 
Typhloblanus Fitzinger, 1843. Type species: A. coeca (sic) 
Cuvier 1829 by original designation. 
Sarea Gray, 1844. Type species: "A caeca" (=A. ridleyi 
Boulenger, 1890) by original designation. 
Cadea Gray, 1844. Type species: Amphisbaena punctata Bell, 
1827 (names preoccupied, replaced by Cadea blanoides 
Stejneger, 1916) by monotypy. 
Diphalus Cope, 1861. Type species: D. fenestratus Cope by 
monotypy. 
Amphisbaena is the oldest generic name for limbless amphis- 
baenians. The Antillean forms caeca Cuvier, 1829, innocens 
Weinland, 1862, cubana Peters, 1878, bakeri Stejneger, 1904, 
manni Barbour, 1914, and caudalis Cochran, 1928, are customar- 
ily referred here. 
Typhloblanus belongs in the synonymy of Amiihisbaena, even 
though it would be available if it became desirable to distinguish 
the Antillean species by generic or subgeneric status. 
The generic name Sarea does not belong in the present as- 
semblage as Boulenger (1890, p. 481; see also Boulenger, 1890a, 
p. 79 ; Stejneger, 1904, p. 676) has shown that Gray based it upon 
a misidentified specimen of Amphisbaena ridleyi Boulenger, 
1890, from "Porto Bello" (read Fernando Noronha, Brazil). 
The latter form thus becomes the type species and the generic 
name belongs in the synonymy of Amphisbaena. 
Little comment need be made on the genus Cadea, clearly 
distinct from Amphisbaena by the unpaired prefrontal scale, 
by a number of skull characteristics (Vanzolini, 1951a), and by 
the retention of the hypocentrum of the atlas. It includes the 
Caribbean species blanoides Stejneger, 1916 and palirostrata 
Dickerson, 1916. 
The resurrection of Diphalus by Vanzolini (1951a, p. 114) 
made without examining a specimen, stimulated the present 
study. He listed three distinguishing characters: (1) a pointed 
rather than a rounded, or flattened snout, (2) the separation of 
the nasal shields by a narrow process of the rostral, and (3) 
a dental formula (premaxilla, maxilla, dentary) of 5, 4, 7 rather 
than 7, 4-5, 7-9. 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 83 
The first of these characteristics is worthless as the actual dif- 
ference is minor, and there is considerable evidence of onto- 
genetic changes in the head shape. The head of fenestrata is no 
more pointed than is that of innocens. It might also be argued 
that this difference is far less striking than that between the 
Antillean species as a group and such blunt-headed yet con- 
generic ( ?) forms as alba and fuliginosa. 
The separation of the nasal shields is clear and constant, but 
hardly more significant than the fusion of ocular and second 
labial in cub ana or of rostral and nasal in manni. 
Of the characters listed, we should have most confidence in 
the low number of premaxillary teeth, as this number is usually 
remarkably constant in amphisbaenids and equals 7 in all species 
of Amphisbaena checked thus far. Vanzolini apparently obtained 
the counts from Boulenger (1885, p. 449), who must have been 
misled by an aberrant or poorly prepared specimen. The skulls 
of the type and of MCZ 36306 both have 7 teeth on the pre- 
maxilla. 
These two skulls were compared with skulls of each of the 
Antillean species of Amphisbaena (A. c. cub ana, MCZ 10802 ; 
A. c. barbouri, MCZ 12135; A. i innocens, MCZ 27596; A. i. 
caudalis, MCZ 25551 ; A. manni, MCZ 44389, plus 5 cleared 
AMNH specimens; A. bakeri, UPR 11; A. caeca, CM 36377, 
37636, MCZ 36317). The comparison furnished no grounds for 
generic separation. Plates 1-3 show views of skulls of A. fen- 
estrata and A. caeca. Differences similar to those shown may be 
seen between skulls of any pair of the other species. A. fenestrata 
also agrees with the other species of Amphisbaena, all of which 
differ from the two forms of Cadea, in lacking the hypocentrum 
of the atlas. This is a matter to be discussed in a subsequent 
paper. 
KEY TO THE ANTILLEAN AMPHISBAENIDS 
1. All median cephalic shields paired, prefrontals in contact with supra- 
labial, precloaeal pores in continuous row, sharply defined lateral grooves, 
color uniform, possibly with uniform markings on (dorsal) segments 
(Amphisbaena) 2. 
First median post-rostral shield azygous, prefrontals kept from labial 
contact by elongate preocular shield, row of precloaeal pores with central 
gap, poorly marked lateral grooves, segments with irregular spots 
(Cadea) 9. 
84 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
2. Nasal scales separated by posterior process of rostral scale (Fig. 6c) . . 
A. fenestrate 
Nasal scales in contact on midline 3. 
3. Second labial fused to ocular (Fig. 6a) A. oubana 4. 
Second labial distinct from ocular 5. 
4. Body annuli less than 218; caudal annuli 14 or less . . A. cubana oubama 
Body annuli more than 226 ; caudal annuli 14 or more 
A. cubana barboiiri n. ssp. 
5. Nasals fused to rostral (Fig. 6b), the nostril connected to the labials 
by a suture ; sixth caudal annulus generally narrower and of smaller 
diameter so that tail constricted at this point A. manni 
Nasals distinct from rostral, caudal annuli more or less equal in width, 
tail generally lacks constriction 6. 
Fig. 6. Major changes in the head scale arrangement, a, Fusion of 
ocular and second labial, b, Fusion of rostral to the nasal with a single 
suture connecting nostril to the first supralabial. c. Caudad projection of 
the dorsal tip of the rostral, which separates the nasals. 
6. Two, rarely three, rows of postgenials, the first row with three, rarely 
four, equal scales; 14-20 ventral segments per midbody annulus; tail 
with almost no taper, blunt ended 7. 
Two or three rows of postgenials ; if two rows, then the first with two 
large scales, occasionally separated by a small element; 18-24 ventral 
segments per midbody annulus; tail tapering continuously toward the 
tip (A. innocens) 8. 
7. Body annuli 217-236, tail slightly longer (Fig. 9), internasal suture con- 
siderably shorter (Fig. 10) A. caeca 
Body annuli 239-255, tail slightly shorter (Fig. 9), internasal suture 
considerably longer (Fig. 10) A. bal'cri 
8. Two rows of postgenials, the second row with three scales; one row of 
postmalars; caudal annuli 10-13; body annuli 214-225 
A. innocens gonavensis n. ssp. 
Three rows of postgenials, the second row with four or more scales; 
no postmalars; caudal annuli 9-14; body annuli 185-219 
A. innocens innocens 
Three rows of postgenials, the second row with four scales ; no post- 
malars; caudal annuli 18-19; body annuli 200-208 . A. innocens caudalis 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 85 
9. Body annuli 175-218 (counted on ventral surface), caudal annuli 10-14, 
segments in a midbody annulus 25-33, snout rounded, blunt 
C. blanoides 
Body annuli 274-320 (counted on ventral surface), caudal annuli 12-17, 
segments in a midbody annulus 32-39, snout domed, laterally compressed 
C. palirostrata 
TAXONOMIC DISCUSSIONS 
Genus AMPHISBAENA Linne, 1758 1 
Amphisbaena Linne, 1758. Type species: Amphisbaena fuliginosa Linne, 
1758, by first reviser (Fitzinger, 1843). 
Typhloblanus Fitzinger, 1843. Type species: A. coeca (sic) Cuvier, 1829, by 
original designation. 
Sarea Gray, 1844. Type species: "A. caeca' 1 (=A. ridleyi Boulenger, 
1890), by monotypy. 
Diphalus Cope, 1861. Type species: D. fenestratus Cope, 1861, by monotypy. 
Discrimination of the species: The application of the biologi- 
cal species concept (Mayr, ]942) encounters its greatest difficulty 
when one is dealing with populations on adjacent islands, popula- 
tions that may represent either distinct species or races of a 
single polytypic form. Remnants of a morphological concept 
inevitably remain, since it is usually necessary to use the degree 
of phenotypic difference of the allopatric populations as the 
indicator of some sort of "inter-population fertility potential." 
The degree of phenotypic difference may furnish but a poor infer- 
ence of the actual difference between the genotypes (e.g. Zahavi 
and Wahrmann, 1957, p. 354), yet it remains the only indicator 
available for museum material. 
The present analysis has made use of the fact that two of the 
forms here recognized as species were polytypic, one of these 
(A. cubana) on a single land mass. Our species differences 
involve characteristics that remain constant in each of these 
cases and generally a greater number of differences as well. We 
also checked for increased similarity of populations from adja- 
cent parts of adjacent islands. 
Analysis demonstrated that each of the large islands of the 
Antilles (Cuba, Hispaniola and Puerto Rico), and the group 
of small islands known as the Virgin Islands, has a distinct, 
endemic species of Amphisbaena. Hispaniola and Puerto Rico 
each have a second species of more limited distribution, but 
sympatric with the first at least over part of its range. 
i The generic synonymy is complete only for names based on forms from the 
Antilles. 
86 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
The more important reasons for considering the several forms 
distinct at the species or subspecies level are as follows : 
(1) The polytypic species A. cabana is found in the eastern 
two thirds of Cuba and on the Isla de Pinos. From allopatric 
A. i. innocens on the mainland of Haiti and A. i. caudalis on 
Grande Cayemite it differs by fusion of the ocular to the second 
supralabial, by shorter nasal contact (Fig. 7), by possession of 
two rather than three rows of postgenials, by a tendency toward 
a blunt rather than a continuously tapering tail and by the 
presence of caudal autotomy. Adult specimens of A. cubana are 
slightly smaller (Fig. 8), have a smaller range of body annuli, 
and a definitely lower mean number of segments to a midbody 
annulus. The island population of A. i. gonavensis is more simi- 
lar to A. cubana in lacking the third row of postgenials and in 
nasal length (cf. Fig. 7). This suggests that A. cubana may be 
E 
o 
i 
.c 
c 
16 
12 
(0 
V) 
to 
c 
u 
0) 
8 
N. 
o oooo o 
o • o o 
*S o 
D CD 
a 
> eee e» •• • • 
•• • • • • • • 
• A. cubana 
o A. i. innocens 
e A.i.innocens?-PG 
a A. i. caudalis 
a A. i. gonavensis 
10 
14 18 22 26 
Snout -vent Length -cm 
Fig. 7. Scatter diagram of internasal length versus snout-vent length 
of A. cubana and A. innocens. The specimens of A. i. gonavensis and the 
single exemplar from Petit Gonave (PG) appear to be intermediate in this 
characteristic. 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 87 
A fenestrates 
□ a. 
1 1 
1 3 
15 
A 7 
19 
21 
23 
25 
27 
Z E 
LASSES 
Fig. 8. Histogram to show number of specimens of each species of 
Antillean Amphisbaena in the different size classes. The units of the fre- 
quency scale indicate 5 specimens. The numbers on the size class scale 
indicate centimeters, thus class 7 includes snout-vent measurements from 
70 to 79 mm, etc. The solid black rectangles indicate specimens in which 
caudal autotomy has taken place. 
88 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
closely related to A. i. gonavensis, but the constancy of the other 
characteristics and the greater difference of A. i. gonavensis in 
number of midbody annuli, plus its possession of a postmalar row, 
leads us to maintain them as separate species. 
A. cubana is even more distinctly different from the allo- 
patric A. manni, which inhabits northwestern and eastern His- 
paniola. From the latter it differs by fusion of ocular with the 
second supralabial, by normal rostral segmentation, by lack of 
caudal constriction, by a lower range of body annuli, by a clearly 
lower number of caudals, and by a relatively much shorter tail. 
A. cubana also has a lower number of body annuli, a tendency 
toward a higher number of dorsal segments at midbody, a 
greater difference between the number of dorsal and ventral 
segments to a midbody annulus, a tendency to a greater number 
of postgenials and precloacals, and a much lower number of 
precloacal pores. 
(2) A. innocens and A. manni are allopatric over most of 
their ranges (Fig. 3). The only indication of possible sympatry 
is given by the single record of A. manni from the Haitian 
capital of Port-au-Prince, well within the range of A. innocens. 
The latter species seems to be restricted to the southwestern 
peninsula of Hispaniola (and the islands of Gonave, Petite 
Gonave and Grande Cayemite), not extending north much be- 
yond the Cul-de-Sac plain. All Haitian records of A. manni 
come from the northern half of that country. Mainland A. 
innocens is clearly different from A. manni in lack of rostro-nasal 
fusion, in the presence of a third row of postgenials (found in 
only 2 of 140+ specimens of A. manni), in the absence of caudal 
autotomy and the clear caudal constriction exhibited by A. manni, 
by the relatively shorter tail length, by the lower number of body 
annuli (the few apparently overlapping counts come from speci- 
mens collected at opposite ends of the island), and by the lower 
number of caudal annuli. A. innocens also differs by a tendency 
to higher ventral, and clearly higher dorsal (and total) counts 
of segments per midbody annulus, by the presence of mostly 
2 instead of 3 postgenials in the first row, by the presence of 4 
or fewer precloacal pores (only 3 of 140+ specimens of A. manni 
also have 4), and by the tendency toward a higher number of 
pre- and postcloacals. A. manni from northern Haiti differs more 
from the mainland A. innocens than A. manni from the Domini- 
can Republic differs from A. innocens. This is apparent in the 
following characters : number of body annuli, number of caudal 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 
89 
annuli, number of dorsal and of ventral segments per midbody 
annulus, numbers of segments in the first and second postgenial 
rows, number of precloacal pores, and number of precloacal 
shields. No reversals of this trend were encountered. 
(3) Only Amphisbaena manni occurs in eastern Hispaniola and 
this species differs from the allopatric A. caeca on Puerto Rico 
by the presence of rostro-nasal fusion, by a greater range of body 
annuli counts, by a higher number of caudal annuli, by the 
relatively longer tail, and by the higher number of precloacal 
pores. A. manni also has a lower number of both dorsal and ven- 
tral (and total) segments per midbody annulus, but the forms 
agree in the differences between number of segments in dorsal 
and ventral half rings. A. manni tends to have lower numbers 
of pre- and postcloacal segments, a somewhat smaller adult 
size (Fig. 8), and includes only two specimens with a row of 
postmalar segments (only 7 specimens of A. caeca lack this 
row ) . 
E 
O 20 
18 
c 
CO 
I- 
16 
14 
12 
10 
8 
12 
16 20 
Snout - vent 
24 28 
Length - cm 
Fig. 9. Scatter diagram of tail length versus snout-vent length of A. 
caeca and A. bakeri. Western Puerto Eieo refers to the region west of 
Arecibo which includes the zone of possible sympatry. Material of A. caeca 
from this region is more distinct from A. bakeri than eastern A. caeca 
is from A. bakeri. 
00 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
(4) A. innocens and A. caeca are not directly adjacent in 
the sense of this comparison. Their differences are quite as clear 
as those between the preceding pair of forms. 
(5) The most difficult and interesting situation involves the 
status of populations from the northwestern sixth of Puerto 
Rico. We here retain A. bakeri as a good species for the following 
reasons: (a) both A. caeca and A. bakeri have been collected in 
three of the four localities (Fig. 4) from which the latter has been 
£ 
E 
1.5 
XX 
c 
0) 
—J 
_ 1.0 
o 
D 
c 
'A 
• o 
o o 
• • 
Interprefrontal Length 
m m 
Fig. 10. Scatter diagram of internasal length versus interprefrontal 
(contact) length of A. caeca (o, •) and A. bakeri (x). The line, internasal 
length = 3 x interprefrontal length, has been dotted in. As in Figure 
9, solid circles indicate specimens from west of Arecibo. The specimen with 
internasal length equal to zero is the anomalous Lares specimen already 
mentioned by Stejneger (1904), and Grant (1932), and figured by the 
former. 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 91 
seen, indicating geographical (though possibly not ecological) 
sympatry; (b) A. bakeri has clearly higher counts of body 
annuli (range of 239-255, against 217-237 for A. caeca) ; (c) 
A. bakeri has a distinctly shorter tail than have specimens of 
A. caeca from the jointly occupied range (Fig. 9), though the 
range of measurements of eastern A. caeca overlaps them; (This 
suggests that the relatively shorter tail length is not directly 
correlated to the increased number of body annuli.) (d) the 
relatively longer internasal contact length (compared either to 
snout- vent or to interprefrontal contact length — Fig. 10) of 
A. bakeri, which again is clearer when comparison is made to 
specimens from sympatric populations of A. caeca; and (e) the 
relatively larger maximum size of A. bakeri (Fig. 8). 
The presence of such diverse specimens might also be ex- 
plained in other ways, less probable on the basis of present 
evidence. Thus no reason exists to suspect ontogenetic change 
(or ontogenetic selection) for the characters in question. Both 
sexes are present in each sample, eliminating the idea of sexual 
dimorphism. Decision for geographical (= ecological?) polymor- 
phism would require detailed collecting data, and for non- 
geographic polymorphism would require data on broods, both 
of which are lacking. Reference should here be made to the 
peculiar low-count specimens from Aguadilla and Salinas which 
have been discussed in more detail below. 
The increased difference in relative tail length in the jointly 
occupied range might be explained as "character displacement" 
(Brown and Wilson, 1956), but the data are not as clear for 
relative length of internasals, while the data on presence of a 
postmalar row suggest an increased similarity between sympatric 
rather than allopatric populations. Thus, A. bakeri lacks both 
postmalars and a third row of postgenials; more than 210 
specimens of A. caeca have postmalars, but 5 of the 7 specimens 
lacking this row of segments come from western Puerto Rico 
where the frequency of postmalars drops from 99 to 65 per cent. 
(6) The characters differentiating A. bakeri from A. caeca 
make it more, rather than less, different than the latter is from 
both A. innocens and A. manni. 
(7) Only A. caeca is found in eastern Puerto Rico. It differs 
from the allopatric A. fenestrata of the Virgin Islands (St. 
John, St. Thomas, Tortola) by contact between the nasal segments 
(except for one anomalous specimen from Lares, fig. 131 in Stej- 
neger, 1904) , by lower counts of body and higher counts of caudal 
92 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY 
annuli, by somewhat greater numbers of dorsal and ventral, and 
distinctly higher total numbers of segments to a midbody an- 
nulus, and by a relatively shorter tail (Fig. 11). 
In two of these characteristics (higher number of body annuli 
and relatively shorter tail) A. fetiestrata approaches A. bakeri. 
Comparison of Figures 9 and 11 shows different slopes for the 
growth curves of these two species so that the similarity seems 
fortuitous. 
E 
O 20 
I 
18 
O) 
C 16 
0) 
<a 
I- 
14 
12 
10 
o oo o 
o 
OOO OOO O ODD 
odo o oDQDoaao oooooo 
o 
OO O Cfla ' iH- all ■OiU' O ^ 
O O O ODalOCSCaDQDO ODD O OO 
O O O dIDK (DXSOOO 09 
o 
OODOD O 00 
O 
O OO GGOOOO 
OOO 
A fenestrate 
° ° ° • A 
o A c aecc 
8 12 16 20 24 28 
Snout - vent Length - cm 
Fig. 11. Scatter diagram of tail length versus snout-vent length of A. 
fenestrata and A. caeca. 
Interrelationships of the species: The forms under discussion 
show considerable similarity in general scale arrangement, 
habitus, and superficial appearance. Preliminary review of a 
number of internal characteristics did not provide any evidence 
that would modify this statement. 
Granting this similarity, it becomes more difficult to establish 
any simple pattern or sequence of relationships, particularly 
since we are ignorant regarding the genetic basis of the observed 
variation. There is thus no reason for the often implicit assump- 
tion that a difference of X units between the ranges of a character 
in two species implies a greater similarity of genotypes than 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 93 
would a difference of 2X, or of nX. We have chosen to be more 
impressed by numbers of similarities or differences in divers, 
possibly non-associated, characters, than by the extent of differ- 
ences. Trends have been evaluated with this in mind. 
Most noticeable is the loss of the postmalar row as well as of 
the third row of postgenial segments, from east to west (except 
for the A. manni records). Counts of body annuli fall into three 
groups, innocens — c. cubana, c. barbouri — manni — caeca, and 
bakeri — fencstrata, with increasing number. A. innocens and 
A. manni differ from all other forms in having a much larger 
range of body annuli. Caudal annuli drop from c. barbouri to 
c. cubana — i. innocens — i. gonavensis, rise to i. caudalis — 
manni, and drop again to bakeri — caeca, and further to fen- 
estrata. Only manni and some specimens of caeca have a clear 
autotomy constriction ; cubana also shows bobbing of tail but less 
of the tail seems to be lost and occasional specimens may show 
severe caudal scarring without autotomy. Dorsal and ventral 
segment numbers rise from cubana to innocens, drop sharply to 
manni, rise slightly to bakeri — caeca, and drop to fenestrata. 
Only manni differs significantly in number of precloacal pores 
and cloacal segments, and in having a uniform coloration. The 
major segmental abnormalities show no geographic trends. 
The review suggests strongly that 1) A. manni is quite distinct 
from the other forms, and 2) the pattern of variation gives a 
checkerboard assembly of characters, rather than an east -west 
sequence. The differentiation of A. manni may well reflect an 
extensive isolation from and adaptation to a different environ- 
ment than the other forms. Its differences may have been em- 
phasized by the existence of a second form on Hispaniola. It 
seems logical to assume that this second form (A. innocens) 
obtained its initial diversification south of the Cul-de-Sac plain, 
at a time when this was inundated by a rise in sea-level. The 
checkerboard of discrete characteristics might be expected in an 
assemblage of closely related species. Ignorance of the selective 
value of the characters does not permit us to discuss the relative 
importance of selection and genetic drift in the establishment of 
such pattern, nor does the character pattern of itself permit 
conclusions regarding the invasion route (s) by which these 
forms entered the Antillean area. 
Discussion of the affinities of the Antillean forms to other 
amphisbaenids is avoided here as it will form the topic of a 
subsequent paper. The analysis does not furnish evidence against 
94 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY 
the hypothesis that the various species of Amphisbaena represent 
a single invasion and subsequent diversification. Aniphisbaenids 
have been found on too many islands to permit us to retain the 
old idea (Schmidt, 1928, p. 20; Mertens, 1934, p. 31, cf. Rivas, 
1958) that their occurrence necessitates the former existence of 
a land-bridge. Most of the Antillean forms could have arisen 
by crossing of water gaps and subsequent differentiation in 
isolation. The races of A. innocens suggest that this may occur 
rapidly, while the lack of island races in A. fenestrate/, may be 
the result of insufficient time, too small an area, 1 or of the 
ecological homogeneity — or, alternatively, our sample is too 
small. Such isolation does not appear to have existed between 
A. bakeri and A. caeca, a situation deserving additional scrutiny. 
Generic description: The Antillean species of Amphisbaena 
share the following external characteristics : 
An azygous rostral followed by three or more pairs of enlarged 
cephalic shields in contact along the dorsal midline, with the 
nostrils pierced in the first pair (which may be separated by a 
process of the rostral). The second pair (prefrontals) are the 
largest segments of the head. Three supra- and three infra- 
labials, the second by far the largest in each case, and the third 
the smallest. A varying number of postocular and temporal seg- 
ments, lying posteriorly to the ocular and generally much smaller 
than the prefrontal. A large T-shaped mental followed by a 
much larger postmental often flanked posterolateral^ by a pair 
of large malars. 
The snout is pointed, flattened very slightly dorsoventrally, 
and oval in cross-section. The upper jaw protrudes strongly and 
the much shorter lower jaw inserts between the projecting supra- 
labials. The shape of the head is strongly affected by the 
allometric growth pattern of the temporal musculature, which 
protrudes very much more in adults than in juveniles. This pro- 
trusion starts back of the eyes and increases the animal's diameter 
from snout to neck. There is little external emphasis on the level 
of the head joint. A very faint constriction and a narrowing of 
the fourth and fifth body annuli are all that mark this region. 
The lateral grooves are generally well defined, starting gradu- 
ally after the first sixth of the body and ending gradually at the 
1 The Columbia-Lippineott Gazetteer of the World gives the following areas 
(in square miles) for the islands involved : Cuba 43,03(5 ; Isla de Pinos 1,182 ; 
Hispaniola 29,829 ; Cayemite Islands 17 : Grande Gonave 254 : Puerto Rico 3,423 : 
St. Thomas 27; St. John 19: Tortola 21. Other islands of the chain might have 
amphisbaenid populations though none have thus far been reported. 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 
95 
level of the cloaca. The ventral groove is indicated only as a 
gap between aligned segments. The dorsal groove may be sug- 
gested by an alignment alone. Considerable variation exists in 
segment size, but length to width of dorsals varies from 1 (in 
embryos) to 1.5, rarely 2, and of ventral segments from .33 
(in embryos) to .5 to 1 (in adults). 
The precloacal pores lie in a single uninterrupted row of 
normal sized or elongate segments anterior to the precloacal 
shield. The tail is circular to oval in cross section and bears 
no terminal specializations. 
Amphisbaena cubana Gundlach and Peters, 1878 
Geographic variation: The available sample for this species 
numbers 86 specimens, but it is difficult to say much about infra- 
specific variation as 58 of these come from the immediate vicinity 
of Soledad (the Harvard Botanical Garden in Cuba) and 
Cienfuegos, while seven others lack precise locality data. 
Eight specimens pose a problem. Of these, two come from 
Marianao, one from Playa del Chivo, two from Matanzas (city or 
province?), two from Caleta Rosario and one from Soledad. All 
UJ 
o 
=5 
o 
z 
82-83° 
81-82 
200 
c. b a r bo u 
c. cubana 
220 
Body Annul! 
240 
18 14 10 
Caudal Annuli 
Fig. 12. A. cubana. Comparison of counts of body and caudal annuli for 
samples. See discussion in text. 
96 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY 
have significantly high counts of body annuli, 225-240 compared 
to 199-217, and of caudal annuli, 15-18 compared to 10-14 
(omitting the Embarcadero record, of which more below) for 
the rest of the species (Fig. 12). 
With the exception of the Soledad specimen, which may have 
been collected further to the west, the five localities form a 
geographic unit. The high-count specimens may thus represent 
either 1) a distinct allopatric species, or 2) a race, inhabiting 
the western end of the range, or lowland areas below 250 feet. 
The first hypothesis is prejudiced by the similarity of the 
geographically replacing forms (both of which share the ocu- 
lolabial fusion otherwise unique in the group) and the lack of 
evidence for sympatry. The second hypothesis seems on firmer 
ground, though it requires the explanation of two low-count 
specimens labelled "Habana" and Jacksonville, Isla de Pinos, 
respectively. The Habana specimen may have been catalogued 
under the home address of the donor. That from Jacksonville 
represents the only record of A. cubana from the Isla de Pinos. 
It has 6 (instead of 4) precloacal pores and differs in other minor 
ways from mainland specimens. 
Also unclear is the possible restriction of the high-count 
race to lowland areas. Three of the localities are definitely 
coastal, and the inexactly localized other two specimens may well 
come from below the 250-foot and possibly below the 100-foot con- 
tour. In contrast, all but two eastern (low-count) specimens were 
definitely collected at or above 250 feet, and this may even hold 
for the low-count Habana and Jacksonville records. The excep- 
tions are San Juan de los Perros and Embarcadero. The former 
could not be located on a map (it is missing from Gazetteer 30 
of the U.S. Board on Geographic names). The Embarcadero 
specimen differs from the other eastern specimens by a high 
count of caudal, though not of body, annuli and by somewhat 
aberrant head scale proportions. More material from Oriente 
would seem to be desirable. It is interesting that the distribution 
of the low-count specimens coincides with the areas of the two 
eastern islands probably left exposed during various Pleistocene 
inundations, while the distribution of the high-count form repre- 
sents an area flooded during the inter glacial stages (Rivas, 1958, 
maps on p. 316). 
It seems clear that the high-count material represents a sub- 
specifically distinct population, though it remains to be deter- 
mined what factors separate the races. Since the holotype of 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 97 
A. cubana Gundlach and Peters is a low-count specimen, we here 
propose the name barbouri 1 for the high-count western race. 
Amphisbaena cubana cubana Gundlach and Peters, 1878 
Amphisbaena cubana Gundlach and Peters, in Peters, 1878, p. 780. Type 
locality: "Cuba." HOLOTYPE: ZMU 6904. 
Diagnosis: A form of Amphisbaena with the ocular fused to 
the second supralabial. Specimens have 199 to 218 body annuli ; 
10 to 14 (exceptionally 16) caudal annuli; 12 to 16 dorsal, and 
14 to 18 ventral segments per midbody annulus; two postgenial 
and no postmalar rows of chin shields; and 4 (exceptionally 
5 or 6) precloacal pores; caudal autotomy occurs between the 
sixth and ninth, mostly after the seventh postcloacal annulus. 
Description: Meristic characters are summarized in Table 1, 
and Figures 7, 8 and 12, individual data in the appendix. Figure 
13 shows the head scalation, Figure 14 the ventral surface of tail 
and cloacal zone. Plate 4 shows views of the color pattern and 
kinds of caudal damage. 
Preserved specimens are various shades of brown; the ventral 
surface has a pink to violet sheen in life. The color is solid on 
dorsal surface of head and tail ; on the rest of the body the 
rectangular center of each segment is much darker than its 
margin giving the impression of dark spots. The color is darker 
dorsally than ventrally. A few specimens show degrees of darken- 
ing of the segmental margins approaching the uniform dorsal 
color of A. manni. Most specimens have few white segments ran- 
domly scattered over the ventral surface, less than 10 per cent 
of the sample had most of the ventral surface white. The 
existence of pigment on a segment appears to be a yes-or-no 
phenomenon and independent of the general ventral lightening. 
The color drops out in an irregular manner and not at a given 
level. 
The head scalation is characterized by fusion of the ocular and 
the second supralabial. Only two specimens show partial separa- 
tion of these two scales. The first two body annuli correspond 
to the three dorsal half -annuli. The segments of the anterior 
half-annulus are customarily referred to as temporal and post- 
ocular and abut middorsally on the frontals. The next posterior 
half-annulus culminates in the enlarged parietals. The third is 
i Named after Thomas Barbour who collected most of the available specimens 
of these two races. 
98 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
Fig. 13. A. c. cubana. Dorsal, lateral and ventral views of head scales of 
ZMU 6904 (HOLOTYPE). The line equals 1 mm to scale. (V. Cummings, 
del.) 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 
99 
generally of normal size though the postparietals may be en- 
larged. Approximately 10 per cent of the specimens have a 
fourth dorsal half-annulus between the second and third, and 
another 20 per cent show a few, often asymmetric supplementary 
segments in the nuchal region. The anterior edge of the third 
body annulus lies in a plane normal to the long axis of the 
animal. 
Fig. 14. A. c. cubana. Ventral view of cloacal region and tail of MCZ 
10790 from Soledad, Cuba. The line equals 1 mm to scale. (M. Franson, 
del.) 
The chin is characterized by an elongate tear-drop shaped 
postmental flanked posteriorly by a pair of large tear-drop shaped 
segments of the first postgenial row. One or more extremely 
elongate segments fill the narrow gap between the enlarged post- 
genials. These tiny and irregular segments may or may not be 
in contact with the postmental and often break crossways into 
two or more segments. 
The fourth and fifth body annuli are narrower, mark the end 
of the bulging temporal musculature and appear to correspond 
to the position of the atlas. 
General observations: The skull has been figured and discussed 
by Zangerl (1944) and Kesteven (1957). Girdle remnants and 
vertebrae were commented on by Zangerl (1945). The thyroid 
was studied by Lynn and Komorowski (1957). 
Range: Isla de Pinos; Central and eastern Cuba from Cien- 
fuegos to Oriente. 
Locality records (see Fig. 2 for map): ISLA DE PINOS: 
Jacksonville MCZ 46235. CUBA: (Barbour, 1930, 1935, 
1937 ; Boulenger, 1885; Gundlach, 1880; Peters, 1878; Stejneger, 
1904; Strauch, 1881); AMNH 58839; HM 367 A— B ; MHNP 
100 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
3120; RMNH 9978; SU R— 14634; ZMU 6904 (HOLOTYPE), 
9383. Habana Province: Habana? MCZ 35510. Las Villas Prov- 
ince (formerly Santa Clara) : Banos del Ciego, Montero AMNH 
7312. Soledad (Barbour, 1914; Barbour and Ramsden, 1916, 
1919) ; BM 1912-9-4-3; CAS 39292; MCZ 7936 A— N, 10790-811, 
10814, 10817-18, 13525-27, 22722, 32665; USNM 48802-03, 137083. 
Cienfuegos (Stejneger, 1917) ; ANSP 16365; BM 1915-10-22-5— 
8 ; RMNH 4638 ; USNM 58747. Camagiiey Province : San Juan 
de los Perros (North Coast) MCZ 13523. Sierra de Cubitas, Paso 
de Lasca, 5.5 miles northeast of Banao AMNH 77790. Marti 
UMMZ 70923, 70925-26. Three miles west of Cascorro UMMZ 
70924. Santa Cruz Mts. (Cochran, 1934) ; USNM 75861. Oriente 
Province: 8 miles south of Puerto Belie, north of Cabo Cruz 
UMMZ 90720. La Demajagua, near Ramon de las Yaguas (Bar- 
bour and Ramsden, 1916). Cuabitas (Alayo, 1951). Puerto 
Boniato (Alayo, 1951). Embarcadero, Banes Bay (Barbour 
and Ramsden, 1916, 1919) ; MCZ 47897. Guantanamo, San Carlos 
Estate (Barbour and Ramsden, 1916, 1919) ; MCZ 10712, USNM 
63217. 
Habitat: One Soledad specimen was taken under stones near 
a brook (P. J. Darlington, Jr. on field label). Barbour (1914) 
mentions that 26 specimens were obtained following "plows in 
the cane fields near Soledad." He later (Barbour and Ramsden, 
1919) added that this procedure was not uniformly successful 
and failed at a plantation in the Cienaga de Zapata on the Rio 
Hanabana (20 km north of Caleta Rosario). This may have 
reflected local absence from a swampy area. 
Amphisbaena cubana barbouri subsp. nov. 
Diagnosis: A form of Amphisbaena with the ocular fused to 
the second supralabial. Specimens have 226 to 240 body annuli ; 
14 to 18 caudal annuli; 12 to 14 dorsal and 16 to 18 ventral 
segments per midbody annulus ; two postgenial and no postmalar 
rows of chin shields ; and 4 to 6 precloacal pores. None of our 
specimens shows caudal autotomy. 
Holotype: MCZ 12136, a male collected by Thomas Barbour 
at Caleta Rosario on the east shore of the Ensenada de Cochinos, 
Las Villas Prov., Cuba, in "January, 1917 .. . under stones on 
the floor of a heavy lowland forest." For data see appendix. 
Allotype: USNM 26364, a female collected by J. W. Daniel 
at Matanzas, Cuba. 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 101 
Paratypes: MCZ 58788-89, from Marianao; MCZ 12135, from 
Playa del Chivo; USNM 26363, from Matanzas; MCZ 12137, 
from Caleta Rosario ; MCZ 13524 from Soledad. 
Description: Meristic characters are given in Table 1 and 
Figures 7, 8 and 12, individual data in the appendix. Plate 5 
shows views of the holotype. 
The general head scalation and coloration of the form closely 
follows that described for A. c. cubana. All specimens show pairs 
of enlarged parietals and of postparietals. 
Range: Western Cuba from Cienfuegos to Habana. 
Locality records (see Fig. 2 for map) : CUBA. Habana Prov- 
ince ; Marianao MCZ 58788-89 (PARATYPES) ; Playa del Chivo 
(Barbour and Ramsden, 1919), MCZ 12135 (PARATYPE). 
Matanzas Province: Matanzas (Barbour and Ramsden, 1919; 
Stejneger, 1904, 1917), USNM 26363 (PARATYPE), 26364 
(ALLOTYPE). Las Villas Province: Caleta Rosario, Ensenada 
de Cochinos (Barbour and Ramsden, 1919), MCZ 12136 (HOLO- 
TYPE), 12137 (PARATYPE); Soledad MCZ 13524 (PARA- 
TYPE). 
Habitat: ' ' Under stones on the floor of a heavy lowland forest" 
(Barbour and Ramsden, 1919). "Under cardboard on top of 
a cliff; surrounded by ants" (B. B. Collette note on MCZ 58788- 
89). 
Amphisbaena innocens Weinland, 1862 
Geographic variation: The sample numbers 50, including 2 
specimens from the island of Grande Cayemite, 1 from Petite 
Gonave, 7 from Gonave, and 40 from Hispaniola proper. Of the 
latter, 24 come from the vicinity of Jeremie, and the remaining 
mainland specimens from a series of scattered localities. 
While the geographic range of the species is not particularly 
extensive, there is a considerable amount of geographic varia- 
tion. This is probably emphasized by the collecting gaps between 
the several mainland localities from which we have seen speci- 
mens. The variation can best be described by discussing the 
material in terms of six geographically delimited samples, namely 
Jeremie (Moron, Jeremie, Place Negre, Carrefour Sanon, Laye, 
Paroty), Grande Cayemite, Furcy (La Vallee, 1 La Vanneau, 1 
1 Note that these two localities may represent another area. Particularly, the 
data from the La Vanneau specimen representing extreme values (cf. Fig. 15) 
tend to indicate this. It should also be noted that the present groupings are 
pragmatic, artificial, and for convenience of description. Thus, Furcy lies at an 
elevation of 5000' and Diquini is at sea level. 
102 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
10 
CN 
i: 
<i 
o 
CN 
CN 
O 
CN 
O 
O 
CN 
O 
C* 
■v 
1 
E 
E 
D 
O 
c 
D 
L. 
o 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 
103 
10 
CN 
o 
u 
D 
O 
O 
a. 
a 
1~2 
n 
cs 
CM 
oo 
c 
o 
CO 
o 
Q- 
3, ] J- ^ 
1 
c 
B 
> GO 
5 
*o 
— I 1 — I 
1 
CO 
o 
re 21 
I s 
«4-l 
o 
CO 
O 
^ P 
s ^ 
o 
o 
P 
o 
03 
° £ 
«t-i re 
■a 2 
I ^ 
CO £ 
.2 <u 
p 0> 
re ^3 
M - 
o £ 
dp "rt 
CUD 
CO 
re 
"8 
o> 
ox 
CO 
o 
Pi 
«4H 
«t-l ° 
° K 
O 3 
2 re 
E 
8 
O 
U 
• • 
E 1 
? o 
-a *- 
u 
o 
OO 
S 
-o 
> 
o 
c 
o 
o 
Q- 
> 
O 
c 
o 
O 
I— I *-• 
CO 
tC ■£ 
fc | 
bo 
CO 
104 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
Furcy, Port-au-Prince, Diquini), Cul-de-Sac (Thomazeau, 
Manneville, Morne-a-Cabrits), Petite Gonave Island, and Gonave 
Island (Point-a-Roquettes). Some of the data has been graphed 
in Figures 15 to 18. 
The graphs and raw data permit the following conclusions : 
1) the two specimens from Grande Cayemite differ decidedly 
from mainland Hispaniolan specimens; 2) the Gonave specimens 
differ decidedly from the Grande Cayemite specimens and from 
those of the mainland; 3) the specimen from Petite Gonave is 
intermediate between those from Gonave and the mainland 
sample; and 4) the several mainland localities show trends of 
17 
D 
t— 
15 
13 
11 
° J e rem ie 
° Grande Cayemite 
Furcy 
A CuldeSac 
* Petite Gonave 
• Gonave 
o 
o »o 
oo o 
o • 
oo o 
12 16 20 24 
Snout-Vent Length - cm 
Fig. 17. A. innocens. Scatter diagram of tail length versus snout-vent 
length of the geographic samples. 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 
105 
variation from the Cul-de-Sac region to the tip of the Southwest- 
ern Peninsula. 
1. The two specimens from Grande Cayemite Island are dis- 
tinct from the mainland specimens in having the greatest dif- 
ference (Difference, in Fig. 15) between dorsal and ventral 
numbers of body annuli, and for that matter are distinct in 
absolute number of dorsally counted body annuli. They are also 
distinct in number of caudal annuli, in relative tail length (Fig. 
17), and in relative length from snout to back of first parietal 
(Fig. 18). In number of body annuli (Fig. 15), dorsal and 
ventral segment counts (Fig. 16), number of postgenials in the 
second row (Fig. 16), and pre- and postcloacals (Fig. 16), they 
are more distinct from the adjacent Jeremie sample than 
10 
o 

o o 
o 
14 
18 22 26 
Snout- Vent Length - cm 
Fig. 18. A. innocens. Scatter diagram of length from snout to back of 
first parietal versus snout-vent length of the several geographic samples. 
106 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
from the material from the vicinity of Furcy at the base of the 
Peninsula, though they appear intermediate between these two 
samples in several of the characteristics. 
2. The Gonave series is clearly distinct from all mainland 
samples by the presence of a row of postmalars (and absence of 
a third row of postgenials), by a lower number of segments 
in the second postgenial row (Fig. 16), and by differences in 
the proportions of the head scales as indicated by a shorter rela- 
tive distance to back of first parietal (Fig. 18). The several dia- 
grams also suggest differences in body annuli (Fig. 15), dorsal 
and ventral numbers of body segments (Fig. 16), pre- and post- 
cloacals (Fig. 16), and relative tail length (Fig. 17). 
The Gonave series differs from the two specimens from Grande 
Cayemite by higher number of body annuli (Fig. 15), lower 
difference between dorsal and ventral counts (Fig. 15), and lower 
caudal counts (Fig. 15). There are also clear differences in 
numbers of dorsal and ventral segments (Fig. 16), postgenials 
in the second row (Fig. 16), the presence of a postmalar and 
absence of a third postgenial row, and the relatively much 
shorter tail (Fig. 17). The two populations often represent ex- 
tremes within this species. While the limited number of speci- 
mens does not lend too much certainty to any one of the several 
characters, this does appear to be another instance of the 
conspicuous difference of "peripherally" isolated populations 
(Mayr, 1954). 
3. The single specimen from Petite Gonave Island represents 
an interesting case. The island in question is so small (it is 
omitted on most maps, but shown on the plate opposite page 4 
in AVetmore and Swales, 1931) that we suspected that the speci- 
men might have come from Petit-Goave on the mainland. How- 
ever, its character pattern is close to or identical with that of the 
Gonave sample in all of the characters discussed above. Where 
there is some range in the Gonave values, the Petite Gonave 
specimen is at that end of the range closest to the Cul-de-Sac 
sample. 
4. A variety of trends is shown between the three mainland 
samples. It is interesting that these seem continuous for only 
three characters (body annuli, precloacals, and relative length 
from snout to back of first parietal, Figs. 15, 16 and 18). A 
reversing cline is shown for range of difference, ventral segment 
number, and number of postcloacals, while the other characters 
do not exhibit any apparent trends. 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 107 
A matter of possible significance or at least of interest is the 
position of the several island populations relative to these clines. 
The island data repeatedly fall closest to those from the immedi- 
ately adjacent mainland sample (except for the characters that 
clearly define them). It is tempting, but dangerous, to speculate 
that this indicates closeness of relation, i.e., that the islands 
were populated from the adjacent mainland. 
Nomenclature: The name A. innocens was assigned to three 
syntypes from Jeremie by Weinland (1862, pp. 132, 137), with 
the intention of vindicating the beast "von den Indianern . . . 
fur Entsetzlich giftig gehalten." He mentioned that one speci- 
men had been deposited in Berlin (ZMU 1386), and talked of 
one other in his personal collection. The second and third syn- 
types were later acquired by the Museum of Comparative 
Zoology (MCZ 3624-25). The larger of the MCZ specimens (MCZ 
3624) shows the best agreement with Weinland 's original meas- 
urements and is here designated as lectotype of A. i. innocens. 
It seems clear that the Grande Cayemite and Gonave-Petite 
Gonave populations may each well be assigned subspecific status. 
For the first we have Cochran's (1928) name A. caudalis based 
upon just the specimens we have been discussing, and reflecting 
the increased caudal count and tail length. For the other we 
here propose the new name A. i. gonavensis. 
Amphisbaena innocens innocens Weinland, 1862 
Amphisbaena innocens Weinland, 1862, p. 137. Type locality: "in einein 
lichten Schlage von Campeehe-Holz in der Nahe des Hafen-Stadtchens 
Jeremie." Haiti. LECTOTYPE: MCZ 3624 (by present designation). 
PARATYPES: MCZ 3625, ZMU 1386. 
Amphisbaena weinlandi Schmidt, 1928, p. 29. Lapsus for A. innocens Wein- 
land. 
Diagnosis: A form of Amphisbaena without fusion of head 
segments; having 186 to 219 body annuli along the ventral line; 
a difference of dorsal and ventral counts ranging from minus one 
to plus two ; 10 to 15 caudal annuli ; 14 to 17 dorsal and 18 to 
22 ventral segments per midbody annulus; three postgenial and 
no postmalar rows of chin shields; and 4 (occasionally 6) pre- 
cloacal pores. Tail pointed and caudal autonomy absent. 
Description: Meristic characters are summarized in Table 1, 
and Figures 7, 8, 15, 16, 17 and 18 ; individual data in the 
appendix. Figure 19 shows the head scalation, Plate 6 various 
views of Jeremie specimens. 
108 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
Fig. 19. A. i. innocens. Dorsal, lateral and ventral views of the head 
scales of ZMU 1386, from Jeremie, PARATYPE of A. innocens Weinland. 
The line equals 1 mm to scale. (V. Cummings, del.) 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 109 
The preserved specimens are various shades of brown, one 
recently pickled series shows a dark violet-brown head and 
tail, grading to chocolate -brown on the body. The color is solid 
on the dorsal surface of the head and tail; on the rest of the 
body the rectangular center of each segment is much darker than 
its margin, giving the impression of dark spots. The color is 
darker dorsally than ventrally, but there are no white segments 
on which the color has dropped out entirely. One specimen 
(Manneville, MCZ 8748) has the uniform chocolate-brown dorsal 
and ventral color normally characteristic of A. manni. 
The head scalation is characterized by lack of major fusions. 
The first two body annuli correspond to three dorsal half-annuli. 
The segments of the first are commonly referred to as temporal 
and postocular and abut middorsally on the frontals. The next 
posterior half-annulus culminates in the first of two pairs of 
enlarged parietals. There is some geographic variation in the 
point of ventrad penetration of the second half-annulus. This 
extends lowest in Jeremie specimens, while only the parietals 
are doubled in some Cul-de-Sac material, thus yielding two pairs 
of enlarged postfrontal segments. 
The chin is characterized by a quite large second infralabial, 
by a malar segment that lies between this and the third infra- 
labial, often far out of contact with the postmental. The extreme 
lateral segments of the postgenial rows give the impression of 
having been separated from the medial edge of the malars. 
The fourth and fifth body annuli are narrower than the rest. 
The body annuli are generally characterized by extensive irregu- 
larities and asymmetries. 
The tail is short and tends to be pointed and laterally com- 
pressed, particularly in Furcy and Jeremie material. The 
midventral segments of the first postcloacal annulus tend to be 
enlarged, often to one and a half times their normal length. 
General observations: The thyroid of this form has been dis- 
cussed by Lynn and Komorowski {1957). 
Range: Southwestern Hispaniola. 
Locality records (see Fig. 3 for map) : HAITI: — (Barbour, 
1930, 1935, 1937; Gray, 1872; Giinther, 1865; Meerwarth, 1901; 
Stejneger, 1904; Strauch, 1881). Eastern end of San Domingo 
(=Hispaniola) (Cope, 1869). Moron (Cochran, 1924, 1941); 
USNM 60620. Jeremie (Barbour, 1914; Barbour and Loveridge, 
1929; Cochran, 1941; Peters, 1878; Weinland, 1862) ; MCZ 3624 
110 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
(LECTOTYPE), MCZ 3625 (PARATYPE) ; ZMU 1386 (PARA- 
TYPE). Place Negre, near Jeremie, MCZ 63609-20. Carrefour 
Sanon, near Jeremie, MCZ 63621-24. Laye, near Jeremie, MCZ 
63625-27. Paroty, near Jeremie, MCZ 63628-29. La Vallee (Coch- 
ran, 1941) ; RMNH 10008. La Vanneau (Cochran, 1941) ; USNM 
69439. Furcy, MCZ 51417; USNM 117250-51, 118036, 118906. 
Port-au-Prince (Werner, 1910) ; HM 2907. Diquini, USNM 
118903. Thomazeau (Cochran, 1934a, 1941) ; MCZ 37595-97. 
Manneville (Barbour, 1914; Cochran, 1941) ; MCZ 8748, 62511. 
Morne-a-Cabrits (Cochran, 1941) ; USNM 59212. 
Habits and habitat: We have no collecting notes, but offer 
the following items excerpted from a letter dated 10 February 
1944 from Anthony Curtis, in Port-au-Prince, to Doris Cochran, 
and cited with her kind permission: "Amphisbaena comes to 
the surface in the dark, shortly before dawn. I believe they 
always remain underground or under stones except during this 
period, unless they meet the stinging ant Solenopsis geminata, 
when, as I have seen by the observation of other, not captive, 
specimens they are apt to ascend into daylight rather hurriedly. ' ' 
Amphisbaena innocens caudalis Cochran, new comb. 
Amphisbaerui caudalis Cochran, 1928, p. 58. Type locality: "Grande 
Cayemite Island, Haiti." HOLOTYPE: MCZ 25550. PARATYPE: 
MCZ 25551. 
Diagnosis: A form of Amphisbaena with no fusions of head 
scales, having 200 to 208 body annuli in ventral count, a differ- 
ence between dorsal and ventral counts of 14 to 22 ; 18 to 19 
caudal annuli ; 14 dorsal and 20 ventral segments per midbody 
annulus ; 3 postgenial and no postmalar rows of chin shields ; 
and 4 preeloacal pores. Tail cylindrical and caudal autotomy 
absent. 
Description: Meristic characters are summarized in Table 1 
and Figures 15, 16, 17 and 18, individual data in the appendix. 
Plate 6 shows the tail of the holotype. 
The preserved specimens resemble the nominal race in most 
particulars. Their head scalation reminds one of that of Furcy 
and Cul-de-Sac samples. Only the two segments closest to the 
middorsal line are doubled. The midventral segments of the 
first postcloacal annulus are of normal size in the holotype and 
slightly enlarged in the paratype. The tail is cylindrical, of 
circular cross-section, and rounded at the end. 
Range: Grande Cayemite Island, off southwestern Haiti. 
GANS AND ALEXANDER: ANTILLEAN AMPHISBAENIDS 111 
Locality records (see Fig. 3 for map) : HAITI : Grande Caye- 
mite Island (Barbour, 1930, 1935, 3939; Barbour and Loveridge, 
1929; Cochran, 1928, 1941) ; MCZ 25550 (HOLOTYPE), 25551 
(PARATYPE). 
Amphisbaena innocens gonayensis subsp. nov. 
Diagnosis: A form of Amphisbaena with no fusions of head 
scales ; having 214 to 225 body annuli in ventral count ; a differ- 
ence between dorsal and ventral counts between minus one and 
plus two; 10 to 33 caudal annuli; 16 to 18 dorsal and 21 to 
24 ventral segments per midbody annulus; 2 postgenial and 
1 postmalar rows of chin shields; and 4 (exceptionally 3) pre- 
cloacal pores. Tail pointed, of rounded cross section, and caudal 
autonomy absent. 
Holotype: PM 3384 a male collected by Philip S. Humphreys 
at Point-a-Roquettes, Gonave Island, Haiti, February-April 
1959. 
Allotype: PM 3385 a female collected with the holotype. 
Paratypes: PM 3389 a male, PM 3386-88 three females 
collected with the holotype, and USNM 10168 from Gonave; 
MCZ 25549 from Petite Gonave. 
Description: Meristic characters are given in Table 1 and 
Figures 8, 15, 16, 17 and 18, individual data in the appendix. 
Figure 20 shows dorsal, lateral and ventral views of the head 
scalation, Figure 21 a ventral view of the cloaca and tail. Plate 
4 shoAvs various views. 
The color of preserved specimens resembles that of the nominal 
race. The head scalation is characterized by lack of major fusions. 
The first two body annuli correspond to three dorsal half-annuli, 
but two rather than one of these contact the frontal middorsally 
and are excluded from contact. The second pair may occasionally 
achieve point contact just posterior to the f rontals, thus confusing 
the determination of first parietals„ 
The malars are split with their posterior portions forming 
the lateralmost segments of the postmalar rows. 
The body annuli are similar to those of the nominal race, but 
show less irregularity. The tail is rounded in cross-section and 
the midventral segments are not enlarged. The precloacal pores 
are much larger and more clearly expressed in males than in 
females. 
Range: Petite Gonave, and Gonave islands, Haiti. 
112 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
Fig. 20. A. i. gonavensis. Dorsal, lateral and. ventral views of the head 
scales of PM 3388, from Point-a-Roquettes, Gonave, Haiti (PARATYPE). 
The line equals 1 mm to scale. (V. Cummings, del.) 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 
113 
Locality records (see Fig. 3 for map) : HAITI: Petite Gonave 
Island: (Cochran, 1941); MCZ 25549 (PARATYPE). 
Gonave Island: (Cochran, 1941; Cope, 1879; Stejneger, 
1904); USNM 10168 (PARATYPE). Point-a-Roquettes PM 
3386-89 (PARATYPES), 3385 (ALLOTYPE), 3384 (HOLO- 
TYPE). 
Fig. 21. A. i. gonavensis. Ventral view of cloaca and tail of MCZ 25549 
from Petite Gonave Island, Haiti. The line equals 1 mm to scale. (M. 
Franson, del.) 
Amphisbaena manni Barbour, 1914 
Amphisbaena manni Barbour, 1914, p. 318. Type locality: "Cape Haitien, 
Haiti." HOLOTYPE: MCZ 8645. PAEATYPES: MCZ 8646-47, 
USNM 67113. 
Diagnosis: A form of Amphisbaena with the rostral and nasal 
114 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
fused, a suture connecting the nostril to the anterior suture 
denning the first supralabial. Specimens have 209 to 243 body 
annuli, 17 to 22 caudal annuli ; 12 to 16 dorsal and 14 to 20 
ventral segments per midbody annulus; generally (3 per cent 
exceptions) 2 postgenial and no postmalar rows of chin shields; 
and 4 to 9 precloacal pores. The sixth to seventh postcloacal 
annulus is slightly constricted, narrower, and more darkly pig- 
mented. Caudal autotomy occurs here (between the fifth and 
seventh caudal annulus). 
Geographic variation: This species exhibits a certain amount 
of minor geographic variation expressed both as a shift in the 
mean values of several characters between the various samples, 
and secondly in a general divergence of the sample from the 
montane portions of Haiti. 
Thus, the two largest (if poorly preserved) samples from Los 
Bracitos and Samana differ in number of body annuli, dorsal 
segments to a midbody annulus, and precloacal pores (Fig. 22), 
and in various body proportions as well (Fig. 23). These dif- 
ferences are not sufficient to permit assignment of individual 
specimens to either locality. Smaller samples from scattered 
localities (not included in the figures) exhibit a similar pattern 
of variation. What makes this situation different from those 
previously discussed is the extreme range of variation shown 
by the material of the large samples. Table 1 does show that the 
range of variation (of body annuli) is highest in A. manni and 
A. i. innocens among the forms discussed, but this information 
is prejudiced by the fact that the A. i. innocens sample range 
included a clear case of geographical variation. In contrast the 
single Los Bracitos sample of A. manni has a range almost 
equivalent to the observed variation of the entire species, far 
greater than that of any non-Hispaniolan form. 
The sample from several Haitian localities (Bombardopolis, 
L'Atalaye, San Francisque?) shows the only approach to 
recognizable geographic differentiation. This accounts for most 
of the species variation range not overlapped by that of the 
large samples. The illustrations show differences in number of 
body annuli, dorsal and ventral segments to a midbody an- 
nulus, first and second postgenials, and number of precloacal 
pores (Fig. 22), as well as minor differences in relative tail length 
and proportions of head shields (Fig. 23). 
The Cap Hai'tien sample and the single specimen from Port- 
au-Prince do not differ significantly from Dominican material. 
CANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 
115 
azza. 
ci 
ft 
00 
in 
^ J 
1 u 
■1 
:d 
o 
X 
o 
5 
c 
0) 
o 
X 
a 
D 
u 
u 
o 
a 
c 
o 
E 
o 
oo 
116 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
_ 1 - 60 
_ 56 
+■» 
2 " 
O 
<S 
XI 
48 
44 
40 
O OK 
(TOO 
O 
O O 
O O O 
O O O O O 
o o • • o o * 
o o o * * 
O * J» • 
oo«»o • A 
O* • 
oo • • torn* a 
oo •••*••• a 
osa • • • 
• ••• A • • 
o o o 
O • • » 
E 
E 
CO 
h- 
X 
I- 
(D 
z 
lu 
20 
16 
12 
* Mont. Haiti 
^ Cap Ha'i'fien 
o Los Braci tos 
• S a ma na 
A A 
m» no • 
•• ''■*• • oo OA 
anx» o 
o <n>o 
• ••• OOO 
OOO 
OOO 
o oo axo 
o 
10 12 14 16 18 20 
Snout - Vent Length - cm 
Fig. 23. A. manni. Scatter diagrams of tail length and of length of 
snout to back of frontal segments versus snout-vent length for the several 
geographic samples. 
CANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 
117 
Fig. 24. A. manni. Dorsal, lateral and ventral views of the head scales 
of AMNH 49726 from Port-au-Prince, Haiti. The line equals 1 mm to 
scale. (M. Franson, del.) 
The difference would be greater than here indicated if we are 
correct in assuming that the separation between populations is 
primarily altitudinal, because the Bombardopolis material was 
brought in from various points along the slope. 
118 
BULLETIN : MUSEUM OP COMPARATIVE ZOOLOGY 
Description: Meristic characters are summarized in Table 1, 
and Figures 8, 22 and 23, individual data in the appendix. 
Figure 24 shows the head scalation, Figure 25 the ventral 
surface of tail and cloacal zone. Plate 7 presents photographs of 
color pattern. 
Preserved specimens are a uniform brown with only slight 
darkening of the dorsal surfaces and somewhat more of head 
and tail. The tendency toward darkening of the rectangular 
center of each segment may be faintly noted under the micro- 
scope, but it is not otherwise apparent. 
The head scalation is characterized by fusion of rostral and 
nasal. A suture connects the nostril to the first supralabial's an- 
terior border. The temporal and postocular lie anterior to the 
vertically ascending first body annulus. Only in 20 per cent of 
the specimens do they give the impression of representing an 
Fig. 25. A. marmi. Ventral view of cloaca and tail of MCZ 8645 (HOLO- 
TYPE) Cap Haitien, Haiti. The line equals 1 mm to scale. (V. Cummings, 
del.) 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 119 
anterior half-annulus. Here there are often splits in the tem- 
poral, and the several segments form a dorsally widening wedge. 
The vertically ascending first annnlns, which includes the 
slightly enlarged pair of parietals, may have some of its seg- 
ments fused with those in line in the second annulus. The 
segments flanking the parietals are often wider than long, while 
the parietals are longer than wide, their anterior edge lying 
between the medial edges of the postoculars. The suture between 
parietals and frontals is often occupied by a pair of small seg- 
ments that appear as if broken off the posterior edge of the 
frontals (Fig. 24). The last infralabials do not end exactly 
at the angulus oris, but often seem to have fused with or been 
encroached on by segments of the first body annulus. The fourth 
to sixth body annuli may be faintly narrowed and mark the 
turning point of the head on the neck. This rotation seems to be 
compensated for by the ventral halves of the third to fifth 
annuli. 
The eye is very clearly noticeable through the ocular, which 
has a circular bulged out area covering its location. The pupil 
is directed forward and it is possible to see both eyes simul- 
taneously from the front. 
The tail is marked by a clear autotomy constriction at the level 
of the sixth to seventh caudal annulus. Posterior to this it main- 
tains its diameter to within one diameter of the tip, whence it 
reduces conically toward the tip. 
General observations: Two specimens (AMNH 43844 an 
adult, and 41058 a hatchling) showed everted hemipenes, a char- 
acter rarely illustrated for amphisbaenids. Figure 26 shows those 
of the adult. The structures are short, without apparent orna- 
mentation, bilobed with a deeply inserted bifurcate sulcus, and 
with more or less pointed apices (nomenclature modified from 
Dowling and Savage, 1960). 
Lynn and Komorowski (1957) have discussed and figured the 
thyroid. 
Range : Eastern, central and northwestern Hispaniola. 
Locality records (see Fig. 3 for map) : HISPANIOLA: 
(Barbour, 1930, 1935, 1937). HAITI: Port-au-Prince AMNH 
49726. Bombardopolis MCZ 62534-40. L'Atalaye, near San 
Michel du Nord (Cochran, 1941) ; USNM 76656-57. Cap Ha'itien 
(Barbour, 1914; Barbour and Loveridge, 1929: Cochran, 1941) ; 
MCZ 8645 (HOLOTYPE), 8646-47 (PARATYPES), 62532-33; 
USNM 67113 (PARATYPE). San Francisque (Cochran, 1941) ; 
USNM 69175. DOMINICAN REPUBLIC : Paradis, Herrmann's 
120 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
Fig. 26. A. manni. Views of the extruded hemipenes of AMNH 43844 
from Los Braeitos, Dominican Republic. The line equals 1 mm to scale. 
(V. Cummings, del.) 
Finea ("at least 1800', near house") AMNH 51279-82. Rio 
San Juan (Cochran, 1941); RMNH 10010; UMMZ 92180; 
USNM 74680-83, 74685-87. Constanza (Cochran, 1941); MCZ 
44389. Los Braeitos AMNH 41030-36, 41038-84, 41251, 41262- 
63. Una River AMNH 6341. Sanchez (Cochran, 1941; Schmidt, 
1921) ; CAS 55001, MCZ 44399. Samana, Samana AMNH 40421- 
27, 40429, 42286, 50264, 63200-01, 63203. Laguna, Samana 
AMNH 42288-89. Samana District AMNH 40302-05, 43844-52, 
44829-30, 50357-73, MCZ 57147. Boca del Infierno (Cochran, 
1941); USNM 74978-79. Cordillera Central, Villa Altagracia 
(Cochran, 1941) ; MCZ 44390. Santo Domingo Province 
(Schmidt, 1921) ; AMNH 5207. San Pedro de Macoris (Schmidt, 
1921) ; AMNH 7557-58. Isla de las Flechas AMNH 42287. 
Habits: No collecting' data are available; however, the jar con- 
taining the Los Braeitos sample also held a vial with two 
cylindrical eggs measuring 28 x 8.5 and 26.5 x 10.5 mm, re- 
spectively. The shorter and stouter egg held an embryo near 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 121 
term, lying in a twice-doubled position, and connected by its 
umbilical cord to an elongate yolk ( fallantoic) mass (see Plate 
7). The embryo's estimated length of 100+ mm suggests that 
some of the other Los Bracitos specimens had only recently 
hatched. 
Amphisbaena bakeri Stejneger, 1904 
Amphisbaena bakeri Stejneger, 1904, p. 681. Type locality: "Lares, Porto 
Eico." HOLOTYPE: USNM 25541. PAEATYPES: USNM 25537 
(Lares), 27458 (Puerto Eico). 
Diagnosis: A form of Amphisbaena lacking fusions of head 
shields ; with 239 to 255 body annuli ; 14 to 16 caudal annuli ; 14 
to 16 dorsal and 16 to 17 ventral segments per midbody annulus; 
two rows of postgenial and no postmalar chin shields; and 4 
precloacal pores. Caudal autotomy not present, 
Description: Meristic characters are summarized in Table 1, 
and Figures 8, 9 and 10, individual data in the appendix. 
Figure 27 shows dorsal, lateral and ventral views of the head 
scalation, Figure 28 the ventral surface of the tail and cloacal 
zone. Plate 8 shows various views. 
The color of preserved specimens is a light brown, with a 
darker more uniform coloring of the dorsal surface of head and 
tail. The rectangular center of each segment is much darker than 
the margin, giving the impression of a pattern of dark spots. The 
coloration is darker dorsally than ventrally. The chin and vary- 
ing portions of the ventral surface are occupied by light-colored 
segments lacking the dark rectangles, which do not fade but 
rather drop out suddenly along the margins of these light-colored 
regions. 
The head segmentation of the few available specimens is 
characterized by striking variability. While there is no pattern 
of major fusions, there are remarkable differences in the shape 
of the individual segments from one specimen to another. The 
first dorsal half-annulus, comprising generally only temporal and 
postocular, may be followed by a second half-annulus which 
includes the parietals. Occasionally there is an anterior enlarge- 
ment of one of its segments, which then wedges in between tem- 
poral and postocular. The first body annulus then follows, lying 
in a plane normal to the long axis of the body. 
There are two postgenial and no postmalar rows of segments 
under the chin. 
122 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
Fig. 27. A. balceri. Dorsal, lateral and ventral views of the head scales 
of UMMZ 55824 from Mayagiiez, Puerto Rico. The line equals 1 mm to 
scale. (V. Cummings, del.) 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 
123 
The tail is rounded and seems to be slightly flattened dorso- 
ventrally in our material. 
Range: Northwestern Puerto Rico. 
Locality records (see Fig. 4 for map) : PUERTO RICO: 
(Barbour, 1914, 1930, 1935, 1937; Grant, 1932; Stejneger, 1904) ; 
UPR 7; USNM 27458 (PARATYPE). Mayaguez (Danforth, 
1925) ; CNHM 12473; UMMZ 55824; UPR 1. Las Marias UPR 
2. Marieao UPR 10, 11. Lares (Schmidt, 1928; Stejneger, 1904) ; 
USNM 25537 (PARATYPE), 25541 (IIOLOTYPE). 
Fig. 28. A. balceri. Ventral view of cloaca and tail of CNHM 12473 from 
Mayaguez, Puerto Rico. The line equals 1 mm to scale. (V. Cunimings, del.) 
Amphisbaena caeca Cuvier, 1829 
dmphisbaena caeca Cuvier, 1829, p. 73. Type locality: " Martinique," = 
Puerto Rico (cf. Stejneger, 1904, p. 675). LECTOTYPE : MHNP 550 
(Strauch, 1881, p. 80). PARATYPES: MHNP 3114, 3115. 
124 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
Amphisbaena (Typhloblanus) coeca Fitzinger, 1843, p. 22. Emendation. 
Diagnosis: A form of Amphisbaena without fusions of head 
scales ; having 217 to 236 body annuli ; 13 to 18 caudal annuli ; 13 
to 18 dorsal and 14 to 20 ventral segments per midbody annulus ; 
2 rows of postgenial and 1 row of postmalar chin shields (3 and 
0, respectively, in <1 per cent of examined specimens) ; and 
4 (to 6) precloacal pores. The tail is cylindrical and its end 
rounded. A small number of specimens have a visible autotomy 
constriction at the sixth annulus, and autotomized tails show the 
break after the fourth to eighth, mostly after the fifth caudal 
annulus. 
Geographic variation: Most of the observed geographical 
variation seems to involve a seemingly random variation of adja- 
cent populations in one character or another. If this is dis- 
regarded it becomes possible to distinguish four regional group- 
ings, dependent in each case on the existence of adequate samples. 
Arranged in a geographical sequence these are: 1) western 
Puerto Rico (Mayagiiez, Las Mesas, Maricao, Sabana Grande, 
Bucarabones, Lares), 2) west-central Puerto Rico (Arecibo, 
Utuado, Cielitos, Juana Diaz, Salinas), 3) central Puerto Rico 
(Baranquitas, Aibonito, Cayey), and 4) northeastern Puerto 
Rico (Rio Piedras, Loiza, South of Loiza, East of Loiza). 
Figure 29 gives a graphic representation of the confusion of 
trends between these samples. 1 Only two of the trends continue 
in the same direction across the island. In three, the break occurs 
at sample 2, in one at sample 3, while two others show conver- 
gence toward samples 2 and 3 from 1 and 4. None of the present 
differences would justify nomenclatorial recognition of the 
populations. The western Puerto Rican sample seems to be the 
most clearly defined, both in terms of degree of difference and of 
range for the several characters. This was already brought out 
for body proportions in Figures 9 and 10. It is also true for the 
variation of head shields in the temporal region, the segment 
pattern being far more stable in the eastern half of the island. 
This variability of western A. caeca confirms Grant's (1932, p. 
155) remarks, without invalidating the specific distinctness of 
A. bakeri. 
i We have avoided the application of statistical tests because of the composite 
nature of the samples. Changes in the relative number of specimens from any 
of the several here combined localities could in each case affect the steepness, 
though not the direction, of our trends. Furthermore, the vagaries of collectors, 
seasonal differences, etc. have often presented us with obviously non-random 
samples composed either of exceptionally large adults, or groups of recently 
hatched specimens. 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 125 
S A M P L E S 
Body 1 *> 2 <* 3 +> 4 
Z 
< Lateral 1 ^" 2 ^ 3 
4/5 Dorsal 1 fe= 2 
^ Ventral 1 ^ 2 
LU 
co Difference 
1 ^* 2 = 3 
Postmalars ] ^2 = 3^-4 
Precloacals 1 -► 2 ^^ 3 ^i 4 
Postcloacals 1 ^2 — 3 ^1 4 
Fig. 29. A. caeca. Comparison of trends in a number of characters be- 
tween the four geographic samples. Equal signs suggest that there is no 
clear difference. The arrows point from low to high values and their sizes 
indicate whether the trend is weak or strong. Paired arrows suggest an in- 
crease in values as well as in range. A V-shaped equal sign suggests an 
increase in range for equivalent mean. See text for discussion. The diagram 
elegantly suggests the confusion of trends between the samples. 
126 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
A situation that we do not understand at all, but which may 
have definite bearing on the matter of A. bakeri, is posed by two 
specimens from Aguadilla (from which we lack further ma- 
terial) and one from Salinas. When the former were first ex- 
amined, their extremely low body annuli (200, 202), high 
caudal annuli (22, x) counts, odd body proportions, and "uni- 
form" coloration suggested that the locality might be erroneous. 
Dr. F. W. Braestrup of the Kobenhavn Museum kindly con- 
firmed that the collector (Meinert, the well-known Danish en- 
tomologist) had been working at Aguadilla and had been in a 
position to collect amphisbaenids. The Salinas specimen repre- 
sents one of an otherwise normal series of A. caeca, and did 
not reach us until this study had been completed. It agrees with 
the Aguadilla pair in all but color pattern, differing from the 
rest of the Salinas series (and A. caeca in general) in number of 
body and caudal annuli, relative tail length, and possibly in 
arrangement of segments in the occipital region. 
It is unlikely that these three specimens from two quite distant 
localities represent geographic variants. The Aguadilla specimens 
are particularly puzzling, and we here list them as incertae sedis 
in the hope that more material will permit resolution of the 
matter. 
Description: Meristic characters are summarized in Table 
1, and Figures 8, 9, 10, and 11, individual data in the appendix. 
Figure 30 shows the head scalation, Figure 31 the ventral surface 
of tail and cloacal zone. Plates 8 and 9 present photographic 
views. 
Preserved specimens are various shades of brown, solidly so 
on the dorsal surface of head and tail. On the rest of the body 
the rectangular center of each segment is much darker than its 
margin, giving the impression of dark spots. The color is darker 
dorsally than ventrally. Some of the specimens have a white 
ventral stripe, normally ranging from the cloaca anteriorly for 
one-third to one-half the length of the trunk. Some geographic 
variation exists in this character, with Aibonito and Cayey speci- 
mens having little or no white. The stripe is produced by a drop- 
ping out of the pigment on specific segments rather than by a 
fading of the color pattern. 
The head scalation is characterized by lack of major fusions. 
The first three body annuli correspond to four or sometimes five 
dorsal half-annuli, of which the temporal-postocular form the 
first. There may be two, one or no temporals, and asymmetry 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 
127 
often exists. This variation of the temporal region seems more 
pronounced in western Puerto Rico, with variation of the parietal 
segments independent of geography. The dorsal segments of 
the second half-annulus (first one after the temporal) may or 
may not meet at the dorsal midline. All intermediates may be 
observed from full and broad middorsal contact, to point contact 
Fig. 30. A. caeca. Dorsal, lateral and ventral views of USNM 27223 
from Utuado, Puerto Rico. The line equals 1 mm to scale. (V. Cummings, 
del.) 
128 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
of two wedge-shaped segments, to complete exclusion of the seg- 
ments by frontals and parietals. The shift of this second half- 
annulus is ventrad, but not forward (as in A. i. gonavensis). 
The annulus is displaced, or interrupted, but retains its position 
in a plane approximately normal to the long axis of the body. 
Various fusions and irregularities occur along the middorsal 
line, often affecting the segments of the first four to five annuli. 
The chin pattern is adequately shown in Figure 30. Only a 
few western and west-central specimens show variations there- 
from, coupled generally with abnormalities of the pattern of 
dorsal annuli. 
The anterior five to seven body annuli seem narrower, show 
some ventral curvature, and represent the zone of the head 
joint. The trunk annuli show few fusions, dorsal half -annuli and 
the like, except for a zone of irregularity just anterior to the 
cloaca. 
The tail is cylindrical, rounded terminally and bears a poorly 
marked autotomy constriction. This, the sixth or seventh post- 
cloacal annulus is often narrower and more solidly pigmented. 
General observations: The skull of this form was discussed by 
Kesteven (1957), and is figured on Plates 2 and 3. Camp (1923) 
discussed and figured the hyoid apparatus of an embryo. The 
thyroid of this form has been discussed and figured by Lynn 
and Komorowski (1957). Zavattari (1910) comments on hyoid 
muscles. 
Two specimens, a hatchling AMNH 13237 and an adult male 
MCZ 58831, have everted hemipenes. Those of the adult are 
incompletely everted, but show a better preservation of detail 
(PI. 9, fig. E). The general shape is similar to that of A. manni, 
short, without ornamentation except for a loose folding (possibly 
due to the incomplete eversion), bilobed with a deeply inserted 
bifurcate sulcus (its outside edges reinforced by a ridge of skin), 
but with less pointed apices (eversion ?) . The sulci of a hemipenis 
start at the medial surface of each base, swing to its lateral side 
and ascend to the level where the apices separate. From here it 
ascends — the medial base of each apex swinging clockwise 
around one and counterclockwise around the other in a 180° 
circle to terminate at the distal extremity. The folding of the 
free skin of each hemipenis leads to the suspicion that the passage 
of the sulci is straighter in the turgid organ. 
Range: Puerto Rico (with the possible exception of the north- 
western corner of the island). 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 129 
Fig. 31. A. caeca. Ventral view of cloaca and tail of USNM 27223 from 
Utuado, Puerto Eico. The line equals 1 mm to scale. (V. Cummings, del.) 
Locality records (see Fig. 4 for map): PUERTO RICO: 
"Martinique" (Cuvier, 1829; Dumeril and Bibron, 1839; 
Dumeril and Dumeril, 1851 ; Duvernois, 1838 ; Gray, 1831 ; Grif- 
fith and Pidgeon, 1831; Peters, 1878; Reinhardt and Liitken, 
1862; Strauch, 1881); MHNP 550 (LECTOTYPE), 3114 
(PARATYPE) ; RMNH 3563. (Atiles, 1887 in Grant, 1937 ; 
Barbour, 1914, 1930, 1935, 1937; Boulenger, 1890a; Grant, 1932, 
1946; Gundlach, 1881; Peters, 1876; Stahl, 1882 in Grant, 1937; 
Stejneger, 1904) ; KM R-4413 ; TIPR 6, 8; ZMU 8949. Mayagiiez 
(Burt and Myers, 1942; Schmidt, 1928) ; CNHM 102260 (data 
not listed) ; SU 7775 ; UPR 3, 4, 9, 12, A, B. Las Mesas, near 
Mayagiiez MCZ 64133-34. Maricao MCZ 36303 ; UMMZ 73839. 
130 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
Sabana Grande UMMZ 73844. Bucarabones MCZ 61500. Gua- 
nica (Fowler, 1918). Lares (Schmidt, 1928; Stejneger, 1904); 
USNM 25538-40. Arecibo HM 2434. Utuado (Schmidt, 1928; 
Stejneger, 1904) ; USNM 27223. Cialitos UMMZ 73841. 3 miles 
east of Juana Diaz UMMZ 73846. 10 miles east of Juana Diaz 
MCZ 36301. Salinas CM 36276/ CM 37636-39. Barranquitas 
UPR 5. Aibonito (Schmidt, 1920, 1928) ; AMNH 13103, 13132, 
13146-47, 13153, 13237, 13325, 13327-28, 13864-69, 14008; CAS 
54858. Cayey MCZ 36311-24; UMMZ 73832 (+A - M), 73835 
(6433-40). Bayamon (Schmidt, 1920, 1928; Stejneger, 1904); 
AMNH 8591 ; CAS 54861 ; RMNH 9985 ; USNM 27320-22. Camp 
Buchanan UMMZ 73838. Rio Piedras (Schmidt, 1920, 1928) ; 
MCZ 36304 ; UMMZ 73840, 73842. 10 kilometers south of Cano- 
venas (=Loiza) MCZ 36325; UMMZ 73830 (3377-82, 3393-96), 
73831 (+ A - J), 73833 (2155-59), 73836 (3658-66). Near 
Canovenas (=Loiza) UMMZ 73834 (+2702-09, 2711). 10 kilo- 
meters east of Canovenas (=Loiza) MCZ 36326-50 (+1637-38, 
3174-84, 4526, 5046, 6125, 6152, A — H, J — N, P, R — T). 
Humacao MCZ 36302; UMMZ 73837. Punta Santiago MCZ 
58831. El Yunque UMMZ 53253. Luquillo (Ranger's Cabin) 
(Schmidt, 1928; Stejneger, 1904); MCZ 36307-10; UMMZ 
73845 (1091-92); USNM 27005-06. Catalina Plantation, 890 
feet, on El Yunque (Schmidt, 1928; Stejneger, 1904); MCZ 
7902 ; USNM 26879-81. Rio Mameyes, near headwaters UMMZ 
73843. Doubtfully referred specimens: Aguadilla KM R-1414, 
1416. 1 Salinas CM 36277. 1 
Habits and habitat: "All of the specimens were found bur- 
rowing in the ground, most of them uncovered by cultivation. 
One was located about three inches beneath an ant's nest, under 
a log." (Schmidt, 1920, p. 194; 1928, p. 119). 
"Three eggs were found — -one beneath a termite nest, the 
other two under a log wbere the above-mentioned adult was dug 
up. The largest egg measured 42 x 11 mm." (Schmidt, op. cit.). 
The egg in question (AMNH 13237 from Aibonito) contained 
a specimen of hatchling size (78 x 8 mm.) and an elongate 
mass of yolk similar to that described for A. manni. The several 
small Aibonito specimens would then appear to be batchlings or 
pre-term embryos and to mark the minimum size of the species. 
1 Data at end of Appendix. 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 131 
Amphisbaena fenestrata (Cope), 1861 
Diphalus fenestrates Cope, 1861, p. 76. Type locality: St. Thomas, Virgin 
Islands (by present restriction). HOLOTYPE : USNM 11715. 
Amphisbaena antillensis Bernhardt and Liitken, 1862, p. 224. Type locality: 
St. John, Virgin Islands (by present restriction). LECTOTYPE : KM 
E-449. PAEATYPES: KM E-4411 (Virgin Islands), ZMU 4346 (St. 
Thomas). 
Notes on the types: Cope (1861, p. 76) described D. fenestra- 
tes from material discovered by A. H. Riise "in the West Indian 
Islands of St. Thomas and Santa Cruz," but neglected to name 
a type. Stejneger (1904, p. 678) believed the type to be USNM 
11715. This specimen had initially been catalogued as collected 
on St. Thomas by A. H. Kiise. Stejneger crossed off St. Thomas 
in the USNM catalog and wrote ' ' St. John ' ' above this. He cited 
the specimen as from St. John without giving reasons for his 
action. St. Thomas was the home of the collector and is one of 
the islands on which this species occurs. Since the inter-island 
variation of the species is insufficient to allow us to allocate speci- 
mens to islands, it seems best to retain the original locality. 
ST. THOMAS ImI , , LL [J U 
ST. JOHN jj_| lJL| U U 
TORTOLA L-J 
J 
235 245 12 14 
Body Caudal 
A N N U L I 
Fig. 32. A. fenestrata. Diagram of body annuli, caudal annuli, ventral 
segments per midbody annulus, and po^teloacal segments for specimens 
from the three Virgin Islands. 
Grant, to whom we owe much of the material here discussed, 
has presented reasons (1937a, p. 512; 1946, p. 60) for the prob- 
able absence of this species from "Santa Cruz" (=St. Croix). 
We here restrict the type locality to "St. Thomas." 
Amphisbaena antillensis was described by Reinhardt and Liit- 
ken (1862, p. 224) from "St. Thomas' og St. Jan." The 
132 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
Universitetets Zoologiske Museum in Kobenhavn retains two 
specimens (KM R-449 and R-4411) marked " Originalexem- 
plarer" (=Types) in Reinhardt's handwriting. These are the 
remnants of a type series, part of which was exchanged to other 
museums (Braestrup, in litt.). Neither is in perfect agreement 
with the original description, which appears to have been com- 
posite. We select KM R-449 from "St. Jan" as lectotype, because 
it is closest to the original description and bears a definite locality 
label. The restricted type locality is then St. John. 
Diagnosis: A form of Amphisbaena with the nasal scales kept 
from contact by the ascending process of the rostral which 
contacts the frontals. Specimens have 236 to 249 body annuli ; 12 
to 14 caudal annuli ; 13 to 14 dorsal and 14 to 16 ventral segments 
per midbody amiulus ; 2 postgenial and 1 postmalar rows of 
chin shields ; and 4 precloacal pores. Caudal autotomy not 
present. 
Geographic variation : The sample is too small to justify a long 
discussion, but Figure 32 suggests that the Tortola sample dif- 
fers from those collected on St. Thomas and St. John. It is 
perhaps suggestive that there is a tendency for values approach- 
ing those shown by A. caeca with increasing distance from Puerto 
Rico. 
The possible occurrence of this species on Virgin Gorda, Cule- 
bra and Viques has never been confirmed. 
Descriptio?i: Meristic characters are summarized in Table 1, 
and Figures 8, 11 and 32, individual data in the appendix. Fig- 
ure 33 shows the head scalation, Figure 34 the ventral surface 
of the tail and cloacal zone. 
Preserved specimens are various shades of brown, solidly so 
on the dorsal surface of the head and tail. On the rest of the 
body the rectangular center of each segment is much darker than 
its margin, giving the impression of dark spots. The color is 
darker dorsally than ventrally. Most of the specimens have most 
of the midventral region white, due to a dropping out of the 
pigment on a number of the ventral segments, 
The head scalation is characterized by the posterior elongation 
of the rostral which keeps the nasals from contact and inserts for 
some slight distance between the prefrontals. The first body 
amiulus continues dorsad and includes the slightly enlarged 
parietals. Only the postocular and temporal lie anterior to it, but 
occasional specimens show various fusions and complications 
between segments of the first two body annuli. The first amiulus 
lies in a plane normal to the long axis of the animal's trunk. 
The shape of the head and segmentation of the head joint 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 
133 
reminds one of A. caeca, as does the segmentation pattern of the 
trunk. The tail shows no autotomy plane. 
General observations : The skull of this form is illustrated in 
Plates 1 and 3. 
Range: Virgin Islands, from St. Thomas to Tortola. 
tZ7 
4HOi 
<3M 
Fig. 33. A. fenestrata. Dorsal, lateral and ventral views of MCZ 36305 
from Tortola, Virgin Islands. The line equals 1 mm to scale. (V. Cum- 
mings, del.) 
134 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
Locality records (see Fig. 1 for map) : VIRGIN ISLANDS: 
(Barbour, 1914, 1930, 1935, 1937; Cope, 1869; Gray, 1872, 
1873; Giinther, 1865; Stejneger, 1904); KM R-448, R-4411 
(PARATYPE antillensis) . ST. THOMAS: (Boulenger, 
1885, 1890; Cope, 1861; Grant, 1937a, 1946; Gray, 1865, 1872, 
1873; Meerwarth, 1901; Reinhardt and Liitken, 1862; Stej- 
neger, 1904 see NOTES ON THE TYPES; Strauch, 1881) ; BM 
60-4-18-71-72 ; CNHM 51586 (800' altitude, on shady hillside); 
HM 308 ; KM R-4410 ; MHNP 1071 ; USNM 11715 (HOLOTYPE 
fenestrata) ZMU 4346 (PARATYPE antillensis). ST. JOHN: 
(Boulenger, 1885; Reinhardt and Liitken, 1862; Stejneger, 
1904 see above; Strauch, 1881); BM 65-10'2-H ; KM R-449 
(LECTOTYPE antillensis) • VM 12345 1 ; KUMNH 45630 (West 
end Great St, James). Dorothea (Grant, 1946) ; UMMZ 91436. 
TORTOLA: (Grant, 1932a); MCZ 36305-06; SU 14633; 
UMMZ 73847 A-B (+C) ; 80649. 
Habitat: "It lives like Typhlops and is usually found in St. 
Jan under dry leaves of sugar cane which are cut off the cane 
Fig. 34. A. fenestrata. Ventral view of cloaca and tail of MCZ 36305 
from Tortola, Virgin Islands. The line equals 1 mm to scale. (E. Coogle, 
del.) 
i Data in addendum to Appendix. May have been one of syntypes of A. 
antillensis. 
GANS AND ALEXANDER : ANTILLEAN AMPIIISBAENIDS 
135 
during the harvest. These leaves are left covering the fields for 
some time. . . . On St. Thomas, where sugar cane is no longer 
grown, it is found under stones in small holes on moist ground; 
here it is very rare." (Riise, in Reinhardt and Liitken, 1862, 
translation courtesy of F. W. Braestrup.) 
Genus CADEA Gray, 1844 
Cadea Gray, 1844, p. 71. Type species: Amphisbaena punctata Bell, 1827 
(name preoccupied, replaced by Cadea blanoides Stejneger, 1916) by 
monotypy. 
Discussion: Vanzolini (1951a, p. 114) diagnosed this genus 
as follows : 
' ' Snout compressed, pointed, or raised in a small keel. Basip- 
terygoid processes, partes posteriores choanarum and supra- 
temporals present. Quadrate proximally dilated. Meckel's groove 
E 
E 
18 
16 
t 14 
c 12 
- J 10 
~ 8 
* D 
* E 
C. palirostrata 
© • 0©« 
• ©* • oo 
« o 
a 
A 
A 
A 
A A 
Jj& A A 
A 
iA 
C. blancides 
• A 
o B 
a C 
18 
16 
14 
12 
10 
10 
14 
18 
22 
26 
Snout -Vent Length - cm 
Fig. 35. Cadea. Scatter diagram of tail versus snout-vent length demon- 
strating the absence of sexual dimorphism in relative tail length. Solid 
symbols indicate males, open circles and triangles females, open squares 
indeterminate. A-C, C. palirostrata, D-E, C. blanoides. 
136 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
open. Dentition 7 ; 5 ; 7. One unpaired median shield adjacent 
to the rostral. Nasal shields separated from first labials." 
This may be amplified to include certain other characters. 
These are : 
A narrow scale separating the azygous prefrontal from the 
supralabials. Head relatively pointed with quite distinct post- 
cephalic narrowing. Atlas (vertebra) consisting of two neural 
arch halves plus a hypocentrum. Dorsal and ventral folds (of 
the integument) absent, lateral folds expressed as a narrow 
zone in which half -annuli run out, with the division skipping back 
and forth between often non-aligned segments (Gans, 1960, p. 
145, footnote). Dorsal segments rounded instead of rectangular 
(Plates 11 and 12). No caudal autotomy. Tail short, becoming 
wider in the immediate postcloacal region, flattened ventrally 
and domed dorsally. The precloacal pores in two series of 
very small segments running medially from the lateral edge of 
the precloacals, interspersed between the precloacals and the last, 
non-pore bearing, body annulus which often contacts the pre- 
cloacals at the midline. The dorsal surface has a color pattern 
that is more or less independent of the segmental arrangement. 
Cadea blanoides Stejneger, 1916 
Amphisbaena punctata Bell, 1827, p. 236; (? plate 20, fig. 2 — not seen). 
Type locality: "In Ins. Cuba." HOLOTYPE : BM 1946.8.2.20. 
Junior homonym of Amphisbaena punctata Wied, 1825 (= Leposternon 
microcephaltim Wagler, 1824; cf. Wagler, 1833, pi. 16). Also see: 
Bianeoni, 1850. 
Cadea blanoides Stejneger, 1916, p. 85. Nomen novum for Amphisbaena 
punctata. 
Diagnosis: A form of Cadea with 175 to 218 body annuli; 10 
to 14 caudal annuli; 25 to 33 segments to a midbody annulus; 
6 to 8 precloacal segments; and 7 to 9 (occasionally 4) precloacal 
pores in two series. Head with faint lateral compression. Dorsal 
surface of body speckled with brown dots which may coalesce 
into a series of lines on the tail. 
General observations: Zug and Schwartz (1958), some of 
whose data were used for the diagnosis, have discussed the varia- 
tion of this species. They suggested the possible existence of 
an eastern race with low precloacal pore counts. This suggestion 
is not confirmed by the USNM specimen from Matanzas (which 
has 6 pores). The sexual dimorphism in number of caudal an- 
nuli, relative tail length, and smaller size of female specimens 
(Fig. 35) suggested by them could not be confirmed after the 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 137 
entire sample had been sexed by examination of gonads, rather 
than hemipenes. 
Figure 36 shows the head scalation ; Figure 37 is a view of 
the underside of the tail. 
Fig. 36. Cadea blanoides. Dorsal, lateral and ventral views of AMNII 
78397 from the Sierra de las Casas, just west of Nueva Gerona, Isla de 
Pinos. The line equals 1 nun to scale. (V. Cummings, del.) 
138 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
Fig. 37. C. blanoides. Ventral view of cloaca and tail of AMNH 78397 
from the Sierra de las Casas, just west of Nueva Gerona, Isla de Pinos. 
The line equals 1 mm to scale. (V. Cummings, del.) 
The skull and mandible were figured by Vanzolini (1951, figs. 
1-5). The thyroid was studied by Lynn and Komorowski (1957). 
Payne (1906, and in Eigenmann, 1909) described the eye, and 
this description was commented on by Franz (1934) and Bellairs 
and Boyd (1947). 
Range : Cuba, west of Matanzas ; Isla de Pinos. 
Locality records: CUBA: (Barbour, 1930, 1935, 1939; 
Barbour and Ramsden, 1916; Bell, 1827; Bianconi, 1850; Bou- 
lenger, 1885 ; Cocteau and Bibron, 1838, 1843 ; Dickerson, 1916 ; 
Dumeril and Dumeril, 1851 ; Dumeril and Bibron, 1839 ; Gervais, 
1853; Gray, 1831, 1844, 1865, 1872, 1873; Gundlach, 1875, 1880; 
Lichtenstein, 1856 ; Mertens, 1926 ; Payne, 1906 ; Reinhardt and 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 139 
Liitken, 1862; Strauch, 1881; Zug and Schwartz, 1958) ; ANSP 
9670; BM 1946-8-2-20 (HOLOTYPE of A. punctata) HM 5004; 
KM R-4412; MHNP 3116-17, 3117A; RMNH 3577, 3821, 9977. 
UH 545. 1 USNM 5729, 36811-12; ZMU 2635, 4082, 9381, 10496. 
Pinar del Rio Province: Cueva de Santo Toinas, 10 kilometers 
north of Cabezas (Zug and Schwartz, 1958) ; AMNH 77789. 
San Diego de los Banos (Barbour, 1914; Barbour and Ramsden, 
1919; Stejneger, 1917; Zug and Schwartz, 1958); MCZ 7934; 
USNM 27845. Herradura (Barbour, 1914; Barbour and Rams- 
den, 1919; Zug and Schwartz, 1958) ; MCZ 7935. Rangel UH 3. 
Rancho Muridibo, between San Cristobal and Santa Cruz de los 
Pinos MJ 48; UH2. Guanajay (Barbour and Ramsden, 1919; 
Stejneger, 1917 ; Zug and Schwartz, 1958) ; USNM 27846. Sierra 
del Anafe UH 539. Habana Province: Caimito de Guayabal 
(Barbour and Ramsden, 1919; Zug and Schwartz, 1958) ; MCZ 
13571-74. Vedado MJ 22; UH 1. Santiago de las Vegas *SMP 
21537. El Cotorro (Zug and Schwartz, 1958) ; AMNH 46678-79. 
Habana *SMF 26479-80. Matanzas Province: Matanzas USNM 
58746. ISLA DE PINOS: Sierra de las Casas, just west of 
Nueva Gerona (Zug and Schwartz, 1958) ; AMNH 78397. 
Habitat: Barbour (1914) and Barbour and Ramsden (1919) 
mention that specimens were plowed up. Zug and Schwartz 
(1958) collected C. Nanoides from beneath rocks in moist soil, 
while Gundlach (1875, 1880) mentions their occurrence in 
manure piles, below planks and rocks, and comments that they 
are insectivorous. 
Cadea palirostrata Dickerson, 1916 
Cadca palirostrata Dickerson, 1916, p. 659. Type locality: "San Pedro, 
Isle of Pines, Cuba." HOLOTYPE: AMNH 2717. PARATYPES: 
AMNH 2718; MCZ 12052 (ex AMNH 2719) ; USNM 5-4379 (ex AMNH 
2718?). 
Diagnosis: A form of Cadea with 274 to 302 body annuli; 12 
to 17 caudal annuli ; 32 to 39 segments to a midbody annulus ; 6 
to 10 precloacal segments and 6 to 11 precloacal pores in two 
series. Head with strong lateral compression and prominent 
rostral projection. Dorsal surface of body speckled with brown ; 
head with a central light stripe from prefrontal onto nape, often 
diversified by a pair of lateral bars. 
i This specimen bears the astonishing label "Oriente Province, Holgufn. 
Calabazas." Schwartz, who examined it, kindly informed us (in titt.) that this 
collection contains a number of other specimens of west Cuban forms so that it 
is most likely that the citation represents an error. 
140 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
General observations: The remarks on Zug and Schwartz 
(under C. blanoides) apply here as well. 
Figure 38 shows the head scalation, Figure 39 is a view of the 
underside of tail. 
Fig. 38. C. palirostrata. Dorsal, lateral and ventral views of AMNH 
12809 from La Ceiba, Isla de Pinos. The line equals 1 mm to scale. (V. 
Cummings, del.) 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 
141 
Range: Isla de Pinos, Cuba. 
Locality records: CUBA: ISLA DE PINOS 
(Barbour, 
1930, 1935, 1939; Barbour and Kamsden, 1919; Mertens, 1926). 
La Ceiba (Barbour and Ramsden, 1919; Zug and Schwartz, 
1958); AMNH 12809-10; MCZ 12331-33; USNM 61176. San 
Pedro (Barbour and Loveridge, 1929; Dickerson, 1916; Zug 
and Schwartz, 1958) ; AMNH 2717 (HOLOTYPE),2718 (PARA- 
TYPE) ; MCZ 12052 (PARATYPE); USNM 54379 (PARA- 
TYPE?). McKinley (Barbour and Ramsden, 1919; Zug and 
Schwartz, 1958) ; MCZ 12088. Santa Barbara (Barbour and 
Ramsden, 1919; Zug and Schwartz, 1958); MCZ 12448-54; 
*SMF 21539. Nueva Gerona (Zug and Schwartz, 1958) ; CM 
4422 ; MCZ 13506-12, 13514-16 ; *SMF 21535-36 ; ZMU 30768. 
Fig. 39. C. palirostrata. Ventral view of cloaca and tail of AMNH 
12809 from La Ceiba, Isla de Pinos. The line equals 1 mm to scale. (V. 
Cummings, del.) 
142 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
ACKNOWLEDGMENTS 
For the loan of and information on specimens we are indebted 
to Charles M. Bogert and Theresa C. McKnight of the American 
Museum of Natural History, James Bohlke of the Academy of 
Natural Sciences of Philadelphia, J. C. Battersby and Alice 
G. C. Grandison of the British Museum, Alan E. Leviton of the 
California Academy of Sciences, Neil D. Richmond of the Car- 
negie Museum, Robert F. Inger and Hymen Marx of the Chicago 
Natural History Museum, Werner Ladiges of the Zoologischen 
Museums, Hamburg, F. W. Braestrup and H. Volsoe of the 
Kobenhavn Universitetets Museum, William Duellman of the 
University of Kansas, Ernest E. Williams of the Museum of Com- 
parative Zoology, Jean Guibe of the Museum National, Paris, 
Philip S. Humphrey of the Peabody Museum, Yale University, 
M. Boeseman and H. E. Muller of the Rijksmuseum in Leiden, 
Konrad Klemmer of the Senckenberg Museum, George S. Myers 
of the Stanford University Museum, Charles F. Walker and 
T. M. Uzzell of the University of Michigan Museum, Juan A. 
Rivero and Francis J. Rolle of the University of Puerto Rico, 
Joseph Eiselt of the Naturhistorischen Museums, Vienna, Doris 
M. Cochran of the United States National Museum and Heinz 
Wermuth of the Zoologischen Museums der Universitat, Berlin. 
We wish to thank Albert Schwartz for loan of the original data 
sheets on which his study of Cadea was based and F. W. Brae- 
strup for the translation of Reinhardt's original paper. A. S. 
Rand and E. E. Williams helped by discussing the problem at 
various stages of completion. 
The artists have been acknowledged in the captions of figures 
prepared by them. The American Geographical Society assisted 
with a copy of the Index to the Map of Hispanic America. 
This study was supported by National Science Foundation 
Grant NSF G-9054. 
LITERATURE CITED 
(Page references to Antillean forms are given in parentheses.) 
Alayo, D. P. 
1951. Especies herpetologicas halladas en Santiago de Cuba. Bol. Hist. 
Nat. Soc. Habana, vol. 2, no. 7, pp. 100-110 (p. 108). 
Barbour, Thomas 
1914. A contribution to the zoogeography of the West Indies, with 
especial reference to amphibians and reptiles. Mem. Mus. Comp. 
Zool., vol. 44, no. 2, pp. 209-346 (pp. 317-319). 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 143 
1930. A list of Antillean reptiles and amphibians. Zooloyica (New 
York), vol. 11, part 4, pp. 61-116 (pp. 69, 104-105). 
1935. A second list of Antillean reptiles and amphibians. Zoologiea 
(New York), vol. 19, part 3, pp. 77-141 (pp. 85-86, 128-129). 
1937. Third list of Antillean reptiles and amphibians. Bull. Mus. 
Comp. Zool., vol. 82, no. 2, pp. 77-166 (pp. 90, 145-146). 
Barbour, Thomas and Arthur Loveridge 
1929. Typical reptiles and amphibians. Bull. Mus. Comp. Zool., vol. 
69, no. 10, pp. 205-360 (pp. 214, 233). 
Barbour, Thomas and Charles T. Eamsden 
1916. Catalogo de los reptiles y anfibios de la isla de Cuba. Mem. 
Soc. Cubana Hist. Nat. "Felipe Poey, " vol. 2, no. 4, pp. 124- 
143 (pp. 125, 138-139, 143). 
1919. The herpetology of Cuba. Mem. Mus. Comp. Zool., vol. 47, no. 2, 
pp. 71-213 (pp. 181-183). 
Bell, Thomas 
1827. Description of a new species of Anolius, and a new species of 
Amphisbaena; collected by W. S. MacLeay, Esq. in the Island 
of Cuba. Zool. Jour. (London), vol. 3, no. 10, art. 26, pp. 
235-236 (p. 236), tab. supp. 20 (fig. 2 not seen). 
Bellairs, Angus d'Arcy and J. D. Boyd 
1947. The lachrymal apparatus in lizards and snakes. I. The Brille, 
the orbital glands, lachrymal canaliculi and origin of the lachry- 
mal duct. Proc. Zool. Soc. London, vol. 117, part 1, pp. 81-108 
(pp. 91-92, 101). 
Bianconi, J. Joseph (G. Giuseppe) 
1850. Specimina zoologiea Mosambicana. Fasc. 1 Bononiae, 16 pp. 
(p. 9). 
BOULENGER, GEORGE ALBERT 
1885. Catalogue of the lizards in the British Museum (Natural His- 
tory). Second edition. London, vol. 2, xiii -)- 497 pp. (pp. 445- 
446, 449-450). 
1890. Reptilia. In Notes on the zoology of Fernando Noronha. Jour. 
Linn. Soc. London, vol. 20, pp. 481-482 (p. 481). 
1890a. First report on additions to the lizard collection in the British 
Museum (Natural History). Proc. Zool. Soc. London, pp. 77- 
87 (p. 79). 
Brown, W. L., Jr. and Edward O. Wilson 
1956. Character displacement. Syst. Zool., vol. 5, no. 2, pp. 49-64. 
Burt, Charles E. and George Sprague Myers 
1942. Neotropical lizards in the collection of the natural history 
museum of Stanford University. Stanford Univ. Publ. Univ. 
Ser., Biol. Sci., vol. 8, no. 2, pp. 1-52 (p. 49). 
Camp, Charles L. 
1923. Classification of the lizards. Bull. Amer. Mus. Nat. Hist., vol. 
48, art. 11, pp. 289-481 (p. 317, 343). 
144 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
Cochran, Doris Mable 
1924. Notes on the herpetological collections made by Dr. W. L. Abbott 
on the Island of Haiti. Proc. United States Nat. Mus., vol. 66, 
no. 6, pp. 1-15 (p. 11). 
1928. The herpetological collections made in Haiti and its adjoining 
islands by Walter J. Eyerdam. Proc. Biol. Soc. Washington, 
vol. 41, pp. 53-59 (p. 56). 
1934. Herpetological collections from the West Indies made by Dr. 
Paul Bartsch under the Walter Rathbone Bacon Scholarship, 
1928-1930. Smithsonian Misc. Coll., vol. 92, no. 7, pp. 1-48 (p. 
44). 
1934a. Herpetological collections made in Hispaniola by the Utowana 
Expedition, 1934. Occas. Pap. Boston Soc. Nat. Hist., vol. 8, 
pp. 163-188 (p. 184). 
1941. The herpetology of Hispaniola. Bull. United States Nat. 
Mus., no. 177, vii + 398 pp. (pp. 297-303, figs. 83-85). 
Cocteau, Jean Theodore and Gabriel Bibron 
1838? Reptiles. In Ramon de la Sagra, Histolre physique, politique 
et naturelle de l'lle de Cuba. Atlas. Paris, 30 pis. plus index 
(pi. 21, figs. 1-4). 
1843. Reptiles. In Ramon de la Sagra, Histoire physique, politique et 
naturelle de l'lle de Cuba. Paris, xviii + 242 pp. (pp. 194-195). 
Cope, Edward Drinker 
1861. [Remarks defining the following species of Reptilia Squamata.] 
Proc. Acad. Nat. Sci. Philadelphia, pp. 75-77 (pp. 75 76). 
1869. Seventh contribution to the herpetology of Tropical America. 
Proc. Amer. Philos. Soc, vol. 11, pp. 147-169 (p. 164). 
1879. Eleventh contribution to the herpetology of Tropical America. 
Proc. Amer. Philos. Soc, vol. 18, pp. 261-277 (p. 273). 
Cuvier, Georges 
1829. Le regne animal, distribue d'apres son organisation. Nouvelle 
edition. Paris (Deterville), vol. 2, xv + 406 pp. (p. 73). 
Danporth, Stuart T. 
1925. Porto Rican herpetological notes. Copeia, no. 147, pp. 76-79 
(p. 78). 
Dickerson, Mary Cynthia 
1916. Description of a new amphisbaenian collected by the late Dr. 
Charles S. Mead in 1911 on the Isle of Pines, Cuba. Bull. Amer. 
Mus. Nat, Hist., vol. 35, art. 34, pp. 659-662 (all). 
DOWLING, IiERNDON GLENN AND JAY M. SAVAGE 
1960. A guide to the snake hemipenis: A survey of basic structure 
and systematic characteristics. Zoologica (New York), vol. 45, 
part 1, pp. 17-28. 
Duheril, Andre Marie Constant and Gabriel Bibron 
1839. Erpetologie generale ou histoire naturelle complete des reptiles. 
Paris, vol. 5, 855 pp. (p. 492-495). 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 145 
Dumeril, Andre Marie Constant and Auguste Henri Andre Dumeril 
1851. Catalogue methodique de la collection des reptiles du Museum 
d'Histoire naturelle de Paris. Paris, iv + 224 pp. (p. 148). 
Duvernois, Georges-Louis 
1838- Eeptiles. In Georges Cuvier "Le regne animal . . . edition 
1842. accompagne des planches gravees." Paris, 169 pp. (p. 102). 
ElGENMANN, CARL H. 
1909. Cave vertebrates of America. A study in degenerative evolution. 
Carnegie Inst. Washington Publ. no. 104, ix 4- 241 pp. (p. 42- 
47). 
Fitzinger, Leopold 
1843. Systema reptilium. Fasciculus primus. Amblyglossae. Vienna, 
106 + vi + (3) pp. (p. 20). 
Fowler, Henry W. 
1918. Some amphibians and reptiles from Porto Rico and the Virgin 
Islands. Pap. Dept. Marine Biol. Carnegie Inst., vol. 12, pp. 
1-15 (p. 14). 
Franz, Victor 
1934. Hohere Sinnesorgane. 1. Vergleichende Anatomie des Wir- 
beltierauges. In Handbuch der vergleichenden Anatomie der 
Wirbeltiere, Louis Bolk, Ernst Goppert, Erich Kallius and 
Wilhelm Lubosch. Berlin and Vienna, vol. 2, part 2, pp. 989- 
1292 (pp. 1106-1107, 1190). 
Gans, Carl 
1960. Studies on amphisbaenids (Amphisbaenia, Eeptilia). 1. A 
taxonomic revision of the Trogonophidae and a functional in- 
terpretation of the amphisbaenid adaptive pattern. Bull. Amer. 
Mus. Nat. Hist., vol. 119, art. 3, pp. 129-204. 
1961. Notes on amphisbaenids (Amphisbaenia, Reptilia). 2. Amphis- 
baena occidentalis from the Coastal Plain of Northern Peru. 
Postilla, no. 56, pp. 1-17. 
Gervais, Paul 
1853. Recherches sur 1'osteologie de plusieurs especes d 'amphisbenes, 
et remarques sur la classification de ces reptiles. Ann. Sci. Nat., 
Paris, ser. 3, Zool., vol. 20, pp. 293-312 (p. 297). 
Grant, Chapman 
1932. A redescription of Amphisbae-na cneca with a discussion of its 
relationship to A. bakeri. Jour. Dept. Agric. Puerto Rico, vol. 
16, no. 2, pp. 153-155 (all). 
1932a. Herpetology of Tortola ; notes on Anegada and Virgin Gorda, 
British Virgin Islands. Jour. Dept. Agric. Puerto Rico, vol. 16, 
pp. 339-346 (pp. 343-344). 
1937. Two early Puerto Rican herpetologists. I — Dr. Augustin Stahl. 
II. — Dr. Francisco del Valle Atiles. Jour. Dept. Agric. Puerto 
Rico, vol. 21, no. 4, pp. 491-502. 
146 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY 
1937a. Herpetological notes witli new species from the American and 
British Virgin Islands. Jour. Dept. Agric. Puerto Rico, vol. 21, 
no. 4, pp. 503-522 (pp. 512, 519). 
1946. Notes on Amphisbaena. Herpetologica, vol. 3, no. 2, pp. 60-61 
(all). 
Gray, John Edward 
1831. Synopsis reptilium. In Animal Kingdom, by Edward Griffith. 
London, 110 pp. (pp. 66-67). 
1844. Catalogue of the tortoises, crocodiles, and amphisbaenians in the 
collection of the British Museum. London, viii -f- 80 pp. (pp. 
71-72). 
1865. A revision of the genera and species of amphisbaenians with 
the descriptions of some new species now in the collection of 
the British Museum. Proc. Zool. Soc. London, pp. 442-455 (pp. 
442-443, 449-450). 
1872. Catalogue of shield reptiles in the collection of the British 
Museum. Part II. Emydosaurians, rhynchocephalians, and am- 
phisbaenians. London, 41 pp. (pp. 32, 33, 37-38). 
1873. Handlist of specimens of shield reptiles in the British Museum. 
London, iv + 124 pp. (pp. 116-117). 
Griffith, Edward and Edward Pidgeon 
1831. The class Reptilia arranged by the Baron Cuvier, with specific 
descriptions. London, 481 pp. (pp. 246-247). 
Gunther, Carl Albert Ludwig Gotthilf 
1865. Zoological Literature. Reptilia. Zoological Record. (Footnote 
to p. 149.) 
Gundlach, Juan 
1875. Catalogo de los reptiles cubanos. Anal. Soc. Espanola hist, nat., 
vol. 4, pp. 347-368 (pp. 360-361). 
1880. Contribucion a la Erpetologia Cubana. Habana, 99 pp. (p. 65). 
1881. Apuntos para la fauna puerto-riquena. III. Anfibios. Anal. 
Soc. Espanola Hist. Nat., vol. 10, pp. 305-317 (p. 312). 
Kesteven, H. Leighton 
1957. Notes on the skull and cephalic muscles of Amphisbaenia. Proc. 
Linn. Soc. New South Wales, vol. 82, no. 1, pp. 109-116 (all). 
Klauber, Laurence M. 
1941. Four papers on the application of statistical methods to 
herpetological problems. I. The frequency distribution of certain 
herpetological variables. Bull. Zool. Soc. San Diego, no. 17, 
pp. 1-31. 
Lichtenstein, M. H. C. 
1856. Nomenclator reptilium et amphibiorum Musei Zoologici Bero- 
linensis. Namenverzeichniss der in der zoologischen Sammlung 
der koniglichen Universitat zu Berlin aufgestellten Arten von 
Reptilien und Amphibien nach ihren Ordnungen, Familien und 
Gattungen. Berlin, iv -)- 48 pp. 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 147 
Linne, Carl vox 
1758. Systema naturae per regna tria naturae, . . . editio decima, 
reformata. Holmiae, vol. 1, 824 pp. 
Lynn, W. Gardner and Louis A. Komorowski 
1957. The morphology of the thyroid gland in lizards of the families 
Pygopodidae and Amphisbaenidae. Herpetologica, vol. 13, no. 3, 
pp. 163-172 (all). 
Mayr, Ernst 
1942. Systematics and the origin of species from the viewpoint of a 
zoologist. New York, Columbia Univ. Press, xiv + 334 pp. 
1954. Change in genetic environment and evolution. In Evolution as 
a Process. Edited by J. Huxley, A. C. Hardy and E. B. Ford. 
London, pp. 157-180. 
Meerwarth, Hermann 
1901. Die westindischen Eeptilien und Batrachier des naturhistorischen 
Museums in Hamburg. Mitt, naturhist. Mus. Hamburg, vol. 18, 
pp. 1-41 (p. 37). 
Mertens, Robert 
1926. Herpetologisehe Mitteilungen. XL Weitere Bemerkungen iiber 
Mesobaena huebneri Mertens. Senckenbergiana, vol. 8, nos. 3/4, 
pp. 149-150 (all). 
1934. Die Insel-Reptilien, ihre Ausbreitung, Variation und Artbildung. 
Zoologica (Stuttgart), vol. 32, no. 84, pp. 1-209 (p. 31). 
Payne, Fernandus 
1906. The eyes of the blind vertebrates of North America. VII. The 
eyes of Amphisbaena punctata (Bell), a blind lizard from Cuba. 
Biol. Bull., vol. 11, no. 2, pp. 60-70 (all). 
Peters, Wilhelm Carl Hartwig 
1876. L T ber eine von Hrn. Viceconsul L. Krug und Dr. J. Gundlach 
auf der Insel Puertorico gemachte Sammlung von Saugethieren 
und Amphibien, so wie iiber die Entwickelung eines Batrachiers, 
Hylodes martinicensis Dum. Bibr., ohne Metamorphose. Mon- 
atsber. Akad. Wiss. Berlin, pp. 703-714 (p. 708). 
1878. Uber vier neue amerikanische Amphisbaena- Arten. Monatsber. 
Akad. Wiss. Berlin, pp. 778-781 (all). 
Reinhardt, J. and C. F. Lutken 
1862- Bidrag til det vestindiske Origes og navnligen til de dansk- 
(1863). vestindiske Oers Herpetologie. Videns. Meddel. naturhist. Foren- 
ing Kjobenhavn, nos. 10-18, pp. 153-291 (p. 224). 
Rivas, Luis Rene 
1958. The origin, evolution, dispersal, and geographical distribution 
of the Cuban poeciliid fishes of the tribe Girardinini. Proc. 
Amer. Philos. Soc, vol. 102, no. 3, pp. 281-320. 
Schmidt, Karl Patterson 
1920. Contributions to the herpetology of Porto Rico. Ann. New 
York Acad. ScL, vol. 28, pp. 167-200 (p. 194). 
148 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
1921. Notes on the herpetology of Santo Domingo. Bull. Amer. Mus. 
Nat. Hist., vol. 44, art. 2, pp. 7-20 (p. 18). 
1928. Amphibians and reptiles of Porto Eieo with a list of those 
reported from the Virgin Islands. New York Acad. Sci., Scient. 
Survey Porto Rico Virgin Is., vol. 10, pt. 1, pp. 1-160 (pp. 8, 
10, 14, 18-20, 26-27, 29, 117-119). 
Smalian, Carl 
1884. Beitrage zur Anatomie der Amphisbaeniden. Zeitsehr. wiss. 
Zool., vol. 42, pp. 126-202. 
Stejneger, Leonhardt 
1904. The herpetology of Porto Rico. Ann. Rept. United States Nat. 
Mus. (1902), pp. 551-724 (pp. 675-683). 
1916. Notes on amphisbaenian nomenclature. Proc. Biol. Soc. Wash- 
ington, vol. 29, p. 85 (all). 
1917. Cuban amphibians and reptiles collected for the United States 
National Museum from 1899 to 1902. Proc. United States Nat. 
Mus., vol. 53, no. 2205, pp. 259-291 (pp. 277-278). 
Strauch, Alexandre 
1881. 1 Bemerkungen iiber die Eidechsenfamilie der Amphisbaeniden. 
Mel. Biol. Acad. Imp. Sci. St. Petersbourg, vol. 11, pp. 355-479; 
also in Bull. Acad. Imp. Sci. St. Petersbourg, vol. 28, no. 8, 
cols. 45-131 (col. 46, 48 51, 55, 80-81, 84, 87-88). 
Vanzolini, Paulo Emilio 
1951. Amphisbaena fuliginosa. Contributions to the knowledge of the 
Brasilian lizards of the family Amphisbaenidae Gray, 1825. 6. 
On the geographical distribution and differentiation of Amphis- 
baena fuliginosa Linne. Bull. Mus. Comp. Zool., vol. 106, no. 1, 
pp. 1-67. 
1951a. A systematic arrangement of the family Amphisbaenidae (Sau- 
ria). Herpetologica, vol. 7, no. 3, pp. 113-123 (p. 114). 
(MS) Evolution, adaptation and distribution of the amphisbaenid 
lizards (Sauria: Amphisbaenidae). Cambridge, Harvard Uni- 
versity, unpublished thesis, 148 pp., 1951. 
1955. Contribuicoes ao conhecimento dos lagartos brasileiros da familia 
Amphisbaenidae Gray, 1825. 5. Distribuiqao geografica e 
biometria de Amphisbaena alba. Arq. Mus. nation. Rio de 
Janeiro, vol. 42, no. 2, pp. 683-705. 
Wagler, Johannes 
1833. Descriptiones et icoues amphibiorum. J. G. Cottae, Monaccii, 
Stuttgart & Tubingen, 36 pis. plus text (pi. 16, fig. 2, text). 
Weinland, D. F. 
1862. Beschreibuug and Abbildung von drei neuen Sauriern. Abhandl. 
i This paper is variously cited as 1881, 1882, and 1883. Dr. Joseph Eiselt has 
kindly informed one of us that the Vienna Museum separate of this paper is 
marked 1881 in print and seems to have been published in that year. 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 149 
Senckenberg. naturforsch. Ges., vol. 4, no. 2, pp. 131-143 (pp. 
132, 137-140, pi. 5, fig. 2). 
Werner, Franz 
1910. Ueber neue oder seltene Reptilien des naturhistorischen Museums 
in Hamburg. II. Eideehsen. Hamburg, Jahrb. Wiss. Anst., vol. 
27 (1909), suppl. no. 2 (Mitteil. naturhist. Mus. no. 27), pp. 
1-46 (p. 34). 
Wetmore, Alexander and Bradshaw H. Swales 
1931. The birds of Haiti and the Dominican Eepublic. Bull. United 
States Nat. Mus., no. 155, pp. 1-483. 
Wied, Maximilian Prinz von Wied-Neuwied 
1825. Abbildungen zur Naturgeschichte Brasiliens. (15 Lieferungen, 
1822-31) Heft 9 (date per Oken's Isis, vol. 17, p. 922). 
Zahavi, A. and Joseph Wahrman 
1957. The cytotaxonomy, ecology and evolution of the gerbils and 
jirds of Israel (Rodentia: Gerbillinae). Mammalia, vol. 21, no. 
4, pp. 341-380. 
Zangerl, Rainer 
1944. Contributions to the osteology of the skull of the Amphisbae- 
nidae. Amer. Midland Nat., vol. 31, no. 2, pp. 417-454. 
1945. Contribution to the osteology of the post-cranial skeleton of the 
Amphisbaenidae. Amer. Midland Nat., vol. 33, pp. 764-780. 
Zavattari, E. 
1910. Muscoli ioidei dei sauri. In Rapporto con i muscoli ioidei degli 
altri vertebrati. Mem. R. Accad. Sei. Torino (pt. 1), ser. 2, 
vol. 60, pp. 351-392; (pt. 2), ser. 2, vol. 61, pp. 55-134. 
Zug, George R. and Albert Schwartz 
1958. Variation in the species of Cadea (Amphisbaenidae) and a record 
of C. blanoides from the Isla de Pinos. Herpetologica, vol. 14, 
no. 3, pp. 176-179. 
150 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
< 
< 
o 
►J 
O 
00 rJ 
O < 
00 
r-t 
ss 
w 
u 
H 
on 
o 
a. 
u 
o 
s-[eoBO"[o^soj 
SieCBO"[D3JJ 
-■ 1 
PQ 
<! 
33 
H 
o 
z 
u 
l-l 
S3JOJ 
«oa pjTm 
«oj puooag 
S"[esaop 
• S"[BJ;U3A 
s^esaop 
S-[EJ3U3A 
s^esjoa 
-[epneo 
Xpog 
i; e i 
4U3A-4nous 
<l 
<l 
<l 
<l 
<l 
<l 
<l 
<;i 
<i 
Appendix 
MUSEUM NUMBER 
ANNUL I SEGMENTS 
H E A 
CLOACA 
LENGTH 
Dorsal 
V e n t r a L 
O 3 c 3 » 
152 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
MCZ 227 22 
MCZ 32665 
USNM 48B02 
USNM 48803 
USNM 137083 
ANSP 16365 
BM 1915.10.22.5 
BM 1915.10.22.6 
BM 1915.10.22.7 
BM 1915.10.22.8 
2 4 
3 4 
3 5 
4 6 
2 4 
?3?5 
2 4 
2 4 
4 5 
3 5 
2 6 
3 5 
3 4 
3 4 
3 4 
2 4 
2 4 
2 3 
3 4 
3 5 
3 4 
4 5 
3 4 
3 4 
3 5 
3 4 
3 5 
3 5 
3 5 
3 5 
3 5 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 
153 
MCZ 25550 
MCZ 25551 
MCZ 25549 
USNM 10168 
PM A925 
PM A933 
PM A934 
PM A935 
PM A936 
PM A937 
A. manni 
4 7 12 132+13 
4 8 11 176+16 
4 6 12 137+10 
? 7 12 7161+9 
4 6 12 175+12 
4 6 12 193+12 
4 6 12 190+11 
74 6 14 172+10.5 
4 6 712 203*13 
3 6 14 201+12 
AMNH 497 26 
MCZ 62534 
MCZ 62535 
MCZ 62536 
MCZ 62537 
MCZ 62538 
MCZ 62539 
MCZ 62540 
USNM 76656 
USNM 76657 
MCZ 8645 
MCZ 8646 
MCZ 8647 
MCZ 62532 
MCZ 62533 
USNM 67113 
USNM 69175 
AMNH 51279 
AMNH 51280 
AMNH 51281 
AMNH 51282 
RMNH 10010 
UMMZ 92180 
USNM 74680 
USNM 74681 
USNM 74682 
USNM 74683 
USNM 74685 
USNM 74686 
USNM 74687 
MCZ 44389 
AMNH 41030 
AMNH 41031 
AMNH 41032 
AMNH 41033 
AMNH 41034 
AMNH 41035 
AMNH 41036 
AMNH 41038 
AMNH 41039 
AMNH 41040 
AMNH 41041 
AMNH 41042 
AMNH 41043 
AMNH 41044 
AMNH 41045 
AMNH 41046 
AMNH 41047 
AMNH 41048 
AMNH 41049 
AMNH 41050 
AMNH 41051 
AMNH 41052 
AMNH 41053 
AMNH 41054 
AMNH 41055 
AMNH 41056 
AMNH 41057 
AMNH 41058 
AMNH 41059 
9 6 13 159*14 
7 6 12 212*19 
8 6 10 186+20 
7 6 9 208+20 
7 6 8 193*18 
6 11 195+20 
6 11 118+11 
6 11 104+10 
8 10 183+18 
12 
116+13 
8 8 9 174+21 
7 6 11 183+x 
8 7 10 146+17 
6 6 10 137+15.5 
8 8 8 180+19.5 
8 6 7 146+x 
8 6 10 127+13 
8 6 10 180+18 
7 6 11 200+21 
6 6 10 184+19 
8 6 9 148+14 
7 6 11 162+16 
8 6 9 7 91*9 
6 6 10 145*15 
7 6 9 93+10 
6 5 12 140+x 
6 6 9 166+18 
6 6 10 159+17 
76 6 9 118+12 
154 
BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 
155 
156 
BULLETIN : MUSEUM OP COMPARATIVE ZOOLOGY 
GANS AND ALEXANDER : ANTILLEAN AMPHISBAENIDS 
157 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
UMMZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
MCZ 
73B31C 
73831D 
73831E 
73831F 
73831G 
73B31H 
7383LI 
73831J 
73833-2155 
73833-2156 
73833-2157 
73833-2158 
73833-2159 
73836-3658 
73836-3659 
73836-3660 
73836-3661 
73836-3662 
73836-3663 
73836-3664 
73836-3665 
73836-3666 
73834 
73834-2702 
73834-2703 
27 04 
2705 
2706 
2707 
2708 
2709 
2711 
73834- 
73834- 
73834- 
73834- 
73834- 
73834- 
73834- 
36326 
36327 
36328 
36329 
36330 
36331 
36332 
36333 
36334 
36335 
36336 
36337 
36338 
36339 
36340 
36341 
36342 
36343 
36344 
36345 
36346 
36347 
36348 
36349 
36350 
36350-1637 
36350-1638 
36350-3174 
36350-3175 
36350-3176 
36350-3177 
36350-3178 
36350-3179 
36350-3180 
36350-3181 
36350-3182 
36350-3183 
36350-3184 
231* 3 ♦16 16*18 
230+ 4 +14 16+18 
234+ 4 +14 16+17 
226+3/4+14 16+18 
232+3/4+14 16+18 
232+3/4+14 18+18 
232+ 3 +14 16+18 
231+ 4 +(5) 16+18 
229+ 4 +16 16+18 
236+ 4 +16 16+17 
230+ 4 +16 17+18 
234+3/4+16 18+19 
232+ 4 +16 18+20 
232+ 4 +(6) 15+18 
234+ 3 +16 16+18 
N 0.69 0.75 
N 0.65 0.73 
N 0.71 0.74 
N 0.63 0.69 
N 0.71 0.78 
232+ 3 +16 
230+ 3 +16 
234+ 4 +16 
?227+ 4 +15 
228+ 4 +16 
231+ 4 +16 
229+ 3 +17 
231+2/3+15 
225+2/3+17 
228+ 4 +(5) 
230+ 4 +(8) 
228+ 4 +15 
231+ 4 +(4) 
231+ 4 +14 
231+3/4+16 
229+3/4+16 
224+ 3 +15 
233+ 3 +14 
228+ 3 +15 
226+ 4 +15 
228+ 3 +15 
226+ 4 +14 
226+ 3 +15 
231+3/4+14 
223+2/3+15 
230+ 3 +16 
228+3/4+14 
228+3/4+15 
229+2/3+15 
233+2/3+15 
227+3/4+15 
232+3/4+15 
233+3/4+15 
233+ 3 +15 
227+ 3 +15 
233+ 3 +15 
227+ 4 +14 
233+ 3 +15 
232+ 3 +15 
231+ 3 +15 
234+3/4+15 
236+2/3+15 
230+ 4 +15 
225+3/4+14 
232+ 4 +15 
232+3/4+14 
232+2/4+15 
230+ 3 +15 
232+3/4+14 
233+ 3 +15 
16+18 
16 + 18 
16+17 
16+18 
16+18 
16 + 18 
16 + 18 
16+18 
16+18 
16 + 17 
16+17 
16+19 
16+18 
16 + 18 
16+18 
16+17 
15+18 
16+18 
16+18 
16 + 18 
16+18 
16 + 18 
16 + 18 
16+18 
16 + 18 
16+19 
16 + 18 
16+18 
16*18 
16+18 
16+18 
16 + 18 
16+18 
17+19 
16+18 
16+18 
16+18 
16 + 18 
16*18 
16*18 
18*18 
16+18 
16 + 18 
16+18 
18 + 19 
16 + 17 
17 + 18 
18*19 
18*18 
17+20 
0.72 
0.65 
0.68 
0.67 
0.71 
0.62 
0.74 
0.72 
0.79 
0.72 
0.72 
0.80 
0.85 
0.65 
0.82 
0.78 
0.67 0.80 
N 0.66 0.69 
N 0.65 0.67 
N 0.65 0.75 
N 0.6L 0.74 
N 0.59 0.67 
N 0.68 
N 0.65 
N 0.74 
N 0.72 
N 0.73 
0.76 
0.68 
0.75 
0.75 
0.65 
0.74 
0.72 
0.61 
0.64 
0.70 
N 0.59 
N 0.63 
N 0.74 
N 0.71 
N 0.66 
'0.66 
0.58 
0.60 
0.67 
0.67 
0.66 
0.67 
0.69 
0.67 
0.68 
N 0.60 
N 0.71 
N 0.58 
N . 69 
N 0.66 
232+ 4 +14 18+19 
232+2/3+15 16+18 
226+3/4+15 16+19 
228+ 3 +(6) 16+18 
229+3/4+14 16+18 
N 
N 
N 
U 
N 
N 
N 
N 
N 
N 
N 
N 
N 
N 
N 
0.66 
0.74 
0.74 
0.73 
0.73 
0.72 
0.62 
0.60 
0.58 
0.66 
0.64 
0.63 
0.67 
0.70 
0.58 
0.72 
0.78 
0.84 
0.85 
0.75 
0.91 
0.74 
0.88 
0.88 
0.78 
0.89 
0.79 
0.72 
0.68 
0.79 
0.63 
0.66 
0.89 
0.76 
0.71 
0.74 
0.65 
0.70 
0.74 
0.70 
0.71 
0.75 
0.76 
0.74 
0.71 
0.64 
0.81 
0.64 
0.77 
0.74 
0.83 
0.92 
0.90 
0.87 
0.79 
0.76 
0.74 
0.65 
0.70 
0.80 
0.67 
0.67 
0.81 
0.74 
0.67 
3 4 - 9 
3 4 - 8 
3 4 - 7 
3 4 - 9 
3 4 - 9 
3 4- 
3 4- 
3 4- 
3 4- 
3 4- 
3 4 
3 4 
3 4 
3 4 
3 4 
3 4 - 9 
3 4 - 7 
3 4-7 
3 4 - 8 
3 3 
3 4 
3 4 
3 4 
3 4 
3 4 
3 4 
3 4 
3 4 
3 4 
3 4 
B 
b 3 4 - 9 
b 3 4 - 9 
b 3 4 - 8 
b 3 4 - 8 
b 3 4 - 7 
b 3 4 - 8 
b 3 4 - 8 
b 3 4 - 8 
b 3 4 - 8 
b 3 4 - 9 
b 3 4 - 8 
b 3 4 - 7 
b 3 4 - 8 
b 3 4 - 9 
b 3 4 - 9 
b 3 4 - 8 
b 3 4 - 7 
b 3 4 - 9 
b 3 4 - 7 
b 3 4 - 9 
b 3 4 - 9 
b 3 4 - 9 
b 3 4 - 9 
b 3 4 - 7 
b 3 4 - 7 
3 4 - 9 
3 4 - 8 
3 4 - 8 
3 4 - 7 
3 4 - 8 
b 3 4 - 9 
b 3 4 - 8 
b 3 ? - 8 
b 3 4 - 7 
b 3 3 - 7 
3 3-8 
3 4 - 7 
3 4 - 9 
3 4 - 9 
3 4 - 8 
158 
BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY