NO. 9 MAMMALS FROM CAVES IN HAITI MILLER 1 5 



impossible, the .smaller size of the Haitian specimens is so constant as 

 compared with material from Porto Rico and the Virgin Islands that 

 the existence of two members of the genus appears to be established. 

 The circumstance must not be overlooked that the Haitian food refuse 

 was accumulated by owls, while that formed elsewhere was chiefly 

 if not entirely deposited by men. Tt is possible therefore that the 

 difference in size may be partly due to selection of the rodents used 

 as food — the owls tending to capture smaller, more easily devoured 

 individuals, the men preferring the larger ones. That the owls were 

 able to eat animals as large as the largest Porto Rican Isolohodon is 

 shown by the frequent presence in the deposits of Aphcctrciis jaws of 

 equally large size. Whatever bearing the possibility of selection may 

 have, the facts are as follows : 



Among more than 600 Haitian mandibles the eight largest have 

 toothrows of the following lengths, 16.2, 16.2, 16.2, 16.4, 16.6, 17.0, 

 17.2 and 17.2 mm., while the extremes of Anthony's measurements 

 of individuals selected from a series of 200 Porto Rican specimens are 

 17.6 and 19.2 mm. The three longest maxillary toothrov/s among the 

 Haitian specimens selected for large size are 16.2, 16.2 and 17.0 mm. ; 

 Anthony gives 17.2 to 19.3 mm. as the range of variation among 

 adults in his series of 17 skulls. The interorbital breadth can be 

 measured accurately or approximately in seven of the Haitian skulls. 

 It ranges from 15 mm. to about 18 mm. ; Anthony's extremes are 

 19.8 and 23.5 mm. in six skulls from Porto Rico. In two Haitian 

 specimens the length of the frontal bone along the median suture is 

 18.6 mm. and 20.0 mm. ; the extremes of eight from Porto Rico 

 are given as 22 mm. and 30 mm., with only three specimens less than 

 24.5 mm. The breadth of rostrum at premaxillary suture does not 

 exceed 11 mm. in any of 15 Haitian specimens (some of them ob- 

 viously immature), while in seven from Porto Rico it ranges from 

 13 mm. to 14.5 mm. Under these circumstances it seems necessary 

 to recognize the Haitian Isolohodon as a distinct form. 



The status of the Isolohodon whose bones have been found in 

 kitchen middens in the Dominican Repulilic is a matter of special 

 interest now that it becomes impossible to regard the Haitian member 

 of the genus as identical with 7. portoricensis. I once said that there 

 appears to be no way to distinguish between Dominican, Porto Rican 

 and Virgin Island specimens ; ' and after going over the ground again 

 in the present connection I am of the same opinion. A palate from 

 San Pedro de Macoris, Dominican Republic, is broken in such a 



' Proc. U. S. Nat. Mus., Vol. 54, p. 508, October 15, 1918. 



