264 Bulletin American Museum of Natural History. [Vol. XVI, 



% ments are given below (p. 277). These specimens are pro- 

 visionally referred to etensis. 



Didelphis marsupialis battyi Thomas. 



Didelphis marsupialis battyi THOMAS, Novit. Zool. IX, 137, April, 1902. 



Type locality, Coiba Island, Panama, Colombia. 



Geographical Distribution. Coiba Island. 



"Closely allied to D. m. caucae Allen, and agreeing with it in most 

 details; but the face uniformly dark, with distinct white spots, about 

 \ in. in diameter, round the roots of the supraorbital and malar tufts 

 of bristles. These white spots are clearly the remnants of the usual 

 light frontal and cheek patches, the lower cheeks and lips being in 

 the present animal no lighter than the rest of the head. No light 

 dorsal bristles present in any of the specimens. Tail white for rather 

 less than half its length, its basal fifth being like the body. 



"Skull rather narrow, the muzzle long and the zygomata little 

 expanded. 



"Dimensions of the type, measured in the skin: Head and body 

 430 mm.; tail 390; white part of tail, 180; hind foot (wet), s.u. 57, 

 c.u. 63; ear (wet), 50; skull, greatest [length in middle line, 108; 

 basal length, 100; greatest breadth, 52.5; combined length of three 

 anterior upper molariform teeth, 18.4. 



" Type : Old female. Original number, 106. Collected 6 May, 1901. 

 Four specimens. 



"The four specimens are all exactly similar in size and colour, and 

 no doubt represent an insular form of the Colombian D. m. caucae." 

 THOMAS, /. c. 



Compared with a large series of specimens from the main- 

 land, D. m. battyi seems to represent a small insular race, as 

 shown by several topotypes kindly presented by the col- 

 lector, Mr. J. H. Batty, to this Museum. I am also indebted 

 to Mr. Batty's kindness for a transcript from his note-book 

 of the measurements of the specimens taken before skin- 

 ning. I am thus able to supplement Mr. Thomas's descrip- 

 tion with the flesh measurements of not only his type, but 

 also of 7 additional specimens, given in the subjoined table 

 (see p. 278). The four females, rather strangely, happen to 

 range rather larger than the four males, doubtless owing to 

 the fact that the females had reached a greater maturity 

 than the males. If the females of the Coiba Island series 

 and the females of the Boqueron and Boquete series be taken 





