ALASKA FUR-SEAL ROOKERIES, 1910. 9 



During the height of the season counts Mere made on the following 

 rookeries: 



Counts of Cows on some St. Paul Rookeries during Height of Season, 



1897, 1909, and 1910. 



i Counts of Mr. George A. Clark. 



COUNTS OF PUPS. 



Owing to the fact that all the cows are never present on the rook- 

 eries at a given time, it is obvious that the only approach to an accu- 

 rate census of the breeding females is to be made by counting all the 

 pups on all the rookeries. Such a procedure is not only arduous but 

 wasteful, since the cows in early August, when the counting is usually 

 done, are readily driven into the sea and a portion must inevitably fall 

 a prey to the pelagic sealer. Accordingly it was the custom, for 

 several years prior to 1906, to count the pups on a number of rookeries, 

 and with such data estimate the entire herd. In more recent times 

 the number" of such pup counts has become gradually lessened until 

 this year Kitovi was the only rookery examined, with the following 

 result: Total number of pups, 1,966; dead, 62. 



The implication that Kitovi is a typical average rookery must rest 

 upon the assumption that it stands between those in which the decline 

 is great and those in which it is at a minimum. As a matter of fact, an 

 examination of the counts of Kitovi during the past four years shows 

 that in reality it has been remarkably constant so far as the cows are 

 concerned. Commencing with 1907 the number of pups each year is 

 1,959, 1,960, 1,979, and this year there are 1,966. 



Last year there were 55 active bulls on Kitovi and 1,979 pups; this 

 year there are 62 bulls and 1,966 pups. The average harem last year 

 was 36; this year, 31.7; a difference due almost wholly to the increased 

 number of active bulls. And, furthermore, this slight difference is of 

 far-reaching importance when we come to consider the application of 

 these data to the estimate of the entire herd. With 1,381 harems, 

 each numbering 36 cows, the estimate would be 49,716; if each com- 

 prised 31.7 cows there are then 43,777 in the breeding herd, a differ- 

 ence of 5,939, or 1 1,878 when the pups are included in the count, due 

 solely to the presence of 7 active, extra bulls. 



