51 



COUNT OF STARVED PUPS. 



At the time of tlie first count of dead pups in Augnat, it was thought that in mak- 

 ing the subsequent count of starved pups it would be possible to distinguish between 

 the earlier and later deaths. l-?ut as the victims of starvation began to be added to 

 the earlier dead pvips it became evident that no such distinction could be made. It 

 was therefore necessary in the second count to include all recognizable dead carcasses 

 to he found on the rookeries. From this number the total of the earlier count could 

 be deducted, and thus the deaths due to starvation found. 



It was desired to jiostpone the second count as long as possible to give the pups 

 time to starve. As the storms of .September, however, began to be more violent the 

 earlier carcasses began to disintegrate very rapidly, ami accordingly the count was 

 begun on September 28 instead of October 1 as at first intended. 



As tli«re evidently still remained a considerable number of starving pups, an effort 

 was made to estimate them that they might be included in the coiint. Accordingly, 

 Mr. Barrett-Hamilton and Mr. Murray, with native assistance, cleared the rookeries 

 of seals in advance of the observers, driving them slowly into the water to afford an 

 opportunity of counting the weak and dying ])nps. This count was necessarily 

 nnsatisfactory, but Mr. Murray and Mr. Barrett-Hamilton agree that it is to be con- 

 sidered an underestimate ratlier than an overestimate, and the figures are therefore 

 not open to dispute. 



The count of starved pups was made. jointly by Mr. Clark and Mr. Macoun. With 

 practically no excepfion every pup carcass was seen by one or the other of the two 

 observers, and for the most ))art by both. Where the rookeries were wide, lines were 

 stretched across dividing them in narow strips, which were traversed from eiul to 

 end in making the count. Where the l)reediiig grounds were narrow, the space was 

 divided lengthwise between the two men, and with natives to helj) patrol the shore 

 and outer edges the bodies were counted. The natives were not allowed to count 

 but simply to point out the location of carcasses. Mr. .). B. Crowley, chief Treasury 

 agent on St. Paul, was present during the entire count, and assisted in pointing out 

 pups and directing the natives. The work of counting was completed on Octolier 1. 



It was evident, from the beginning, that some carcasses included in the count of 

 August had disappeared through being drifted over by sand, washed away by the 

 surf, or worn to pieces by the passing to and fro of the seals. It was agreed between 

 Mr. Macoun and Mr. Clark that the nuitter should be kept in mind during the 

 progress of the count, and if possible an agreement reached at the conclusion as to 

 the necessary correction for the loss. It was mutually agreed that 20 per cent would 

 cover the loss of early pups included in the first but not recognizable in the second 

 count. 



The total count of dead pups in the second enumeration was found to be 20,3.31. 

 To this 20 per cent of the former count of 10,309, or 2,061, is to be added. The total 

 number of starving pups counted was 1,527. A further addition of 150 must be 

 made for dead bodies removed from the rookeries for dissection or other purposes 

 after the first count but before the second. Adding these together, we have a total 

 of 24,069. Deducting 10,309, the August count, the total number of starved pups 

 for St. Paul Island in 1896 is found to be 13,760. 



The following is the count in detail by rookeries: 



St. Paul Island. 



Rookerv. 



October 1, 1896. 



Kitovi 



Lukanin 



Lagoon 



Tolstoi 



Zapartiii 



Little Zapadni. 

 Zapadni Keef. . 



Gorbatch 



Ardiguen 



Reef 



Sivutcli Rock . . 



Polovina 



Little Polovina 



Vostocbiii 



Morjovi 



Total . . . . 



20, 331 



1,527 



