44 DUNN — SOME HAITIAN SNAKES rv^fviP 



catesbyi from Gonaives (M. C. Z., no. 12,869, ?), I am unable 

 to separate from those from Haiti. Its scale formula is Sc. 17, 

 V. 174, C. ? 



I do not know whether oxyrhynchus occurs on Gonaives. 

 The similarity in form of dorsalis to oxyrhynchus might sug- 

 gest that it takes the place of the latter on that island. On 

 Tortuga catesbyi is represented by scandax. A specimen of 

 frenatus from Tortuga (U. S. N. M., no. 59,920, d") has a lower 

 ventral count than the Haitian specimens whose formulae 

 Boulcnger gives, i.e., 171 against 182-193. But with only one 

 speeunen, and no other differences, I hesitate to separate it. 

 Its formula is V. 171, C. 194, Sc. 17-11. 



The Tortuga oxyrhynchus is apparently not separable from the 

 Haitian form. Seven specimens (U. S. N. M., nos. 59,923-24, 

 59,456-60), afford the following figures: ventrals, 196-209 

 (average, 202.6); caudals, males, 199-213 (average, 205.3); 

 caudals, females, 197-209 (average, 202). 



In view of the fact that I have here dealt, however briefly, 

 with the entire genus Uromacer, a key to the forms recognized 

 may not be out of place. Also it is of interest to remark that in 

 catesbyi the dorsal scale rows in both sexes are 17-11 ; in frenatus 

 males have 17-11, and females 17-13; and in oxyrhynchus males 

 have 19-11, and females 19-13. The types of Garman's V. 

 inornatus are identical with specunens of frenatus. I cannot 

 distinguish any difference between specmiens of oxyrhynchus 

 from Haiti and from San Domingo. 



Snout twice as long as the eye, rostral twice as broad 

 as deep. 

 Scales, 17-11, V. 163-176, catesbyi 



Scales, 19-11, V. 181, scandax 



Snout two and a half times as large as eye, rostral as 

 broad as deep. 

 Scales, 17-13 or 11, V. 171-193, frenatus 



Snout three times as long as eye, rostral deeper than 

 broad. 

 Scales, 17-13, V. 205, dorsalis 



Scales 19-13 or 11, V. 196-210, oxyrhynchus 



