THE BAHAMA ISLANDS 337 



from that island of a specimen possessing interparietals. That a certain 

 amount of differentiation has taken place, there can be no doubt; how much, 

 can only be settled by the accumulation of much more material. 



Tropidophis cana (Cope). 

 Ungalia cana Cope, 1868, Proc. Acad. Nat. Sci. Phila., p. 129. 



The low number of ventrals alone is sufficient to distinguish this species. 



Alsophis angulifer vudii (Cope). 

 Alsophis vudii Cope, 1862, Proc. Acad. Nat. Sci. Phila., p. 74. 



Boulenger has denied this form recognition by placing it as a synonym of 

 the Cuban A. angulifer (Bibron). I find, however, that the Bahama speci- 

 mens, on the average, have fewer ventrals. Thus in ten specimens from New 

 Providence, Eleuthera and Long Island, the ventrals vary between 158 and 171, 

 averaging 164, while in twenty-one Cuban specimens the average is 171, the 

 extremes being 164 and 180. There is also an average smaller number of sub- 

 caudals in the Bahama form, viz., 110 as against 115 in the Cuban. Under 

 these circumstances it seems best to recognize the Bahama form by name, and 

 as the two forms intergrade, a trinominal appellation is here applied to it. 



A single specimen from Eleuthera, collected by Mr. Kiley during the 

 Baltimore Geographic Society Expedition, has 19 scale rows, while all the other 

 specimens examined have 17. I can discover no other differences, and with 

 only one specimen it is impossible to say whether this deviation is individual 

 or not. 



LeIMADOPHIS (?) RUBESCENS (Copo). 

 Diadophis rubescens Cope, Proc. Amer. Philos. Soc, xxii, 1885, p. 403. 



Boulenger refers Cope's Diadophis rubescens to Alsophis angulifer (vudii) 

 as a synonym. Apparently he bases this indentification upon the complete 

 agreement of the number of scales and scutes constituting the regular scale 

 formula. I have not had an opportunity to examine the unique type specimen, 

 but I would call attention to the fact that Cope describes his species as having 

 only one pore to each scale. This, if correct, would preclude its being an 

 Alsophis, while on the other hand I can see no good reason why it may not be 

 a form of Leimadophis more or less closely allied to L. andrew (Eeinhardt and 

 Luetken ) . 

 22 



