HERPETOLOGY OF PORTO RICO. 677 



publication of their paper on the nerpetology of the West Indian 

 islands, in which case considerable doubt is thrown on the stability of 

 the chief character relied upon for the distinction of the Virgin Islands 

 species, as it is improbable that Reinhardt or Luetken would have 

 parted with the specimen now in Paris if it were unique. The sug- 

 gestion that it is only an abnormal A. fenestrate however, is strength- 

 ened by the number of body rings, which is said to be 247. 



Ampkisbasna innocens from Haiti has been pronounced an unquali- 

 fied synonym of A. caeca by such authorities as Peters" and Straucli/' 

 both of whom had the opportunity of comparing the types of both 

 species. Reluctant as 1 am to question their conclusion without having 

 examined the same material, I would call attention to the small 

 number of body rings in the Haitian specimens, 211-215, and to the 

 occurrence of only two small scutes behind the unpaired postmental 

 instead of three. Moreover, a specimen in the U. S. National Museum 

 (No. 10168) from the Island de la Gonaives, off Port au Prince, Haiti, 

 consequently not very far from the type locality of A. innocens, shows 

 these same characters (211 body rings, 13 caudal rings). In addition 

 it has 18 segments above the lateral line and 21 below it. Another 

 Haitian specimen (Hamburg Museum, No. 1983) is recorded by Meer- 

 warth as having 211 body rings and 11 caudal rings; it has 16 segments 

 above the lateral line and 21 below. Consequently it agrees closely 

 with the others and differs in all these points from true A. cseca. 



The whole question hinges essentially on the amount of individual 

 variation to be found in specimens from the same locality. Unfortu- 

 nately, not only is the number of specimens from each locality very 

 limited, but not all the specimens known to exist have been described 

 individually with sufficient details. The only minutiae given for nearly 

 all the specimens are the number of rings on the body and tail. If 

 we place these figures together it will be seen that in the specimens 

 hitherto recorded there is a certain uniformity in those from each 

 locality, with a perceptible difference according to the various localities, 

 as follows: 



Martinique (?) specimens (A. cseca), 227-230 body rings; 18 on tail. 



Virgin Islands specimens (A.fenestrata), 212-250 body rings; 13-16 

 on tail. 



Haitian specimens {A. innocens), 211-215 body rings; 13-15 on tail. 



That we have not to do with a curious coincidence only, due in a 

 measure to the scantness of the material, seems certain from a consid- 

 eration of the specimens from the Virgin Islands and from Porto 

 Rico. Specimens from the former islands are more common in 

 museums than the other species, about 10 being recorded in some 

 detail, including the one St. Thomas specimen in the Paris Museum 



"Mon. Ber. Berlin Akad. Wish., 1878, \>. 781. 

 &Mel. Biol., XI, 1883, p. 405. 



