327 



length of 2.S1, aud the six females an average length of 2.48, showing a 

 considerable sexnal variation in size. Yet the smallest males (2.G4 and 

 2.63) fall below the largest female (2.68), if the skulls are all correctly 

 marked. Kone of the other females, however, exceed 2.55, and only 

 three of the males fall below 2.70. In the New York series, the sex is 

 not indicated ; but, judging from the proportion of the small to the large 

 skulls, the sexes are about equally represented in the two series, but in 

 the New York series there is a very gradual decline from the largest to 

 the smallest. The northern series of eighteen is selected from a series 

 of twenty-three 5 the New York series of thirteen from a series of thirty. 

 In each case only very old skulls were chosen, the immature specimens 

 in each case being thrown out in order to have a fair basis for compari- 

 son. The immature and middle-aged specimens greatly predominate in 

 the New York series, owing, doubtless, to the species being more closely 

 hunted there than in the more unsettled districts of the far north. 



Taking these two series as a basis for a general comparison, there is 

 indicated a considerable decrease in size from the north southward, 

 amounting to 0.26 in length and 0.24 in width, or about one-tenth of the 

 average size of the New York series. A single specimen, marked 

 " Brookhaven, Miss.", aud another marked " Tuscaloosa, Ala.", how- 

 ever, have a length respectivelyof 2.60 and 2,80, the former equaling the 

 largest New York specimens, and the latter nearly equaling the average 

 size of the males of the northern series, while a single male skull from 

 Fort Eandall, D. T., 2.90 in length, is the second in size of the whole 

 series; one Fort Yukon specimen only being larger? Other specimens 

 from the Upper Missouri region, however, are much smaller, as are 

 other specimens from Prairie Mer Eouge, La., indicating that the speci- 

 mens above mentioned are much above the average for their respective 

 localities. 



No. 4 4 



