92 



THE FUR SEALS OF THE PRIBILOF ISLANDS. 



THIS ESTIMATE ALSO UNSATISFACTORY. 



In the light of our experience of the present season, however, this estimate for 

 St. George still proves unsatisfactory. It was found that in the first hasty view of 

 Little East rookery on July 9, 1896, a considerable portion of the breeding ground was 

 hidden from view by the clift's, because the most advantageous observation point was 

 not selected. The proportion of killable seals furnished by St. George in 1890 proved 

 wholly misleading because in 1897 the island furnished only one-ninth of the quota. 

 The observations of the present season also show that a count of harems after the 

 25th of July gives no idea of the real condition of the rookeries at the height of the 

 season. On rookery ground under inspection during 1897 for this 'purpose it was 

 found that between the 13th and 25th of July, while the number of cows diminished 

 one-third the number of families had been increased one-seventh through the ingress 

 of young and idle bulls following the breaking up of the harem system. 



But this information was not at hand when the census of 1896 was prepared on 

 August 1, and the estimate seemed to represent as nearly as possible the actual 

 condition of the breeding herd at the time known as the height of the season. 



THE GREAT EXCESS OF PUPS. 



As our observations on St. Paul continued, and especially when we came to enter 

 the rookeries to count the dead pups, our attention was attracted to the fact that the 

 pups seemed to be much more numerous than the estimates of cows in the height of 

 the season would warrant. This was particularly noticeable on the sand flat of 

 Tolstoi. 



To test the matter a careful count of the live pups on Kitovi rookery was made on 

 August 15. This rookery had been taken as the typical one in making up the census 

 and the most accurate count of cows was made upon it. A total of 5,940 live pups 

 were found. To this number 109 dead pups counted a week before were to be added ? 

 making a total of 6,049 pups for a rookery where 3,152 cows had been counted on the 

 13th of July at the supposed maximum of its expansion. 



COUNT OF PUPS. 



The counting of live pups was continued on all parts of rookeries where cows 

 had originally been counted. In some cases the counts could not be made accurately 

 because the pups were beginning to swim well and took to the water. Where the 

 counts were most satisfactory the proportion of live pups to cows was about 1.90 to 1. 

 The proportion on Kitovi was 1.91, and here the count was made under favorable 

 conditions because a heavy surf kept the pups from going to sea. Where the lowest 

 percentages were found the least accurate counts had been made. 



The following is a detailed statement of the count of live pups: 



Count of live pups, 1S9G, St. Paul Island. 



