306 BARBOUR: ZOOGEOGRAPHY. 



when true. The presence of alUed species upon neighboring islands is what ex- 

 perience with West Indian reptiles teaches one to expect. To say that these 

 Celesti upon Jamaica and Haiti are the same when one can not be certain of 

 this from the material at hand is misrepresenting the state of affairs far more 

 probably than to hold them separate, when we find so few species really the same 

 upon the two islands, and so many types, if one may call them such, represented 

 on each of the two islands by related species. We are erring more when we over- 

 emphasize the similarity of the fauna of Jamaica and of Haiti; for the relation- 

 ship shows in its truest light when the differences between the species on the 

 islands are pointed out and held at their true value. 



I pointed out in 1910 (loc. cit.) that Boulenger's Diploglossus bakeri was 

 apparently this species. For the use of Celestus instead of Diploglossus, con- 

 sult the same paper. 



Celestus maculatus (Garman). 

 Garman, Bull. Esspx inst., 1S88, 20, p. 105. 



Boulenger, in the Zoological record for 1888, erroneously noted that this 

 species was a synonym of C. striatus (Gray) ( = C occiduus Shaw) . Garman 

 correctly pointed out its relationship to his C. crusculus. It needs comparison 

 with no other species. It may be separated at once from C. crusculus; and plenty 

 of specimens, including the type, of this species are available for comparison. 

 The second and third digits are, as Garman pointed out, much elongated. They 

 are, in fact, nearly twice as long as those in any specimen of C. crusculus of the 

 same size. 



The type is unique in the collection (M. C. Z., No. 6,231). It was collected 

 on the island of Cayman Brae. It is possible that the species might occur upon 

 other islands of the Cayman group. 



Celestus costatus (Copk). 

 Coi-E, I'roc. .\ciul. nai. .sci. Pliila., 1S()1, p. 494. Garman, Bull. Esspx inst., 1887, 19, p. 23. 



Garman (loc. cit.) was the first to point out that the type of Cope's Pana- 

 lopus costatus was in reality a mutilated lizard of the same species, a perfect 

 specimen of which Cope later characterized as Celestus phoxinus. The earlier 

 type (M. C. Z., No. 3,606) was described as representing a strange genus having 

 legs without toes. I believe that a partly healed up groove around the lizard's 

 body offers the key to why it was so mutilated. Some young Haitian caught the 

 lizard and tied a string tightly about its body, hence the groove; then he pro- 

 ceeded to snip off the fingers and toes — a form of amusement, or perhaps experi- 

 mentation,[not unlcnown to more civilized youth. The lizard then escaped, and 



