112 U. S. BUREAU OF FISHERIES. 



In order to check the accuracy of these estimates since 1916, it was 

 considered advisable to make another complete pup count, and, be- 

 cause of the increasing size of the seal herd, further delay would 

 have been disadvantageous. 



The pup count began on St. George Island on July 26, four days 

 earlier than in 1921. There was no difficulty from the harem bulls, 

 as the large average harem made it safe to go over the rookeries at 

 such an early date. The count was completed on August 7 on St. 

 Paul Island. Although the harems were not counted on Sivutch 

 rookery, it was possible to make the pup count there. The results 

 were more than satisfactory in every way. 



Up to July 15 the weather on St. Paul Island was the driest that 

 had occurred in recent years. Had the dryness continued through 

 the pup count the fatalities due to suffocation would have been ex- 

 tremely large. It was, indeed, fortunate that only six pups were 

 killed as a result of the counting. To balance this loss there were 

 rescued from crevices and holes during the count 11 cows and about 

 twice that number of pups, all of which would have died had they 

 not been discovered and released. Six albino pups were noticed dur- 

 ing the count. 



Of the total of 185,914 pups in the seal herd in 1922, 95 per cent 

 were actually counted. The remaining 5 per cent were necessarily 

 estimated. On some of the rookeries there are areas made up of 

 large bowlders under which the pups when they are disturbed pack 

 themselves as closely as their strength allows. If the hole is a small 

 one, they can be pulled out and counted, but otherwise there is no 

 way to determine their number accurately. On one rookery on St. 

 Paul Island a section was found with round rocks 2 to 3 feet in 

 diameter. The holes between these rocks were 2 to 4 feet deep and 

 pups falling into them were unable to climb out. To have counted 

 the pups over this section would have caused many of them to fall 

 into the holes where they would have starved to death. During the 

 latter part of the count small groups of pups were on the edge of the 

 water and it was impossible to make a perfect count of them. The 

 number of pups thus estimated, however, was less than 10,000, con- 

 servatively figured. 



C. E. Crompton reports, in the St. George Island journal, under 

 date of November 3, 1921, the occurrence of the birth of a seal pup 

 in the month of October. As practically all seal pups are born be- 

 fore August 7 such a late birth is remarkable. The journal entry 

 follows : 



* * * During the work on the rookeries to-day a very young pup was 

 noticed. The age of this pup was estimated at from 2 to 3 weeks and the writer 

 feels certain that it was not more than 8 weeks of age at the most. It is to 

 be regretted that this birth was not observed, as it is probably the latest one on 

 record. The pup seen to-day could not yet swim and will most certainly have 

 difficulty following the migrating seals in their journey south at tlie end of 

 this month * * *. 



