DISAPPEARANCE OF DEAD PLPS. 



167 



MANY EARLY DEAD PUPS DISAPPEARED. 



In considering this count, however, several facts must be noted. It was found 

 that a large number of the pups originally counted in August had wholly disappeared 

 or been reduced to loose bones. This was particularly true on the sandy areas and 

 where the dead lay in the route of the living seals as they passed back and forth to 

 the sea. An eftbrt was made to establish a correction for this loss, and 20 per cent' 

 of the original number counted in August was tixed upon as likely to cover as nearly 

 as could then be determined bodies which had so disappeared. Making allowance 

 lor this correction the total number of starved pups, as shown by the enumeration, 

 would be increased to 12,392. 



STARVING PUPS. 



It was further evident from the outset that all the pups about to die of starvation 

 had not yet died. It was impossible, however, to delay the count longer, and so, to 

 overcome this new dififlculty, an ett'ort was made to count the starving pups among 

 the healthy ones as they were driven from the rookery preparatory to the counting of 

 the dead. The best estimate, which was a very unsatisfactory one, gave 1,540 as the 

 number of pups whose appeaiauce indicated that they were certain to starve. This 

 number was doubtless very inadequate, as later observations on the rookeries in the 

 same season showed that pups continued to starve until as late as the 20th of October, 

 twenty days after the count was completed on St. Paul Island. Making, however, the 

 addition of the dying pups just mentioned, as shown by the actual count, we have the 

 number of deaths justly chargeable to starvation increased to 13,930. 



THE WORK OF THE FOXES. 



Another fact in connection with the actual count must be noted. On the Island 

 of St. (leorge, as already indicated, the total number of dead pups found in October 

 was 897. In August the number found was 735. On some of the rookeries of this 

 island, however, a much smaller number of dead was found in October than was 

 found in August. The condition of the St. George rookeries was so peculiar that we 

 may give here in detail both counts. They are as follows: 



DEAD PUPS. ST. GEORGE ISLAXD. 



Rookery. 



North , 



Staraya Artel 



Zapadni 



East 



Little East . . 



Total . . 



Dead pnps. Total dead 



August, ' pups, Octo- 



1896. ber 6, 1896. 



' The investigations of 1897 show clearly that this estimate of 20 per cent to cover the disap- 

 pearance of dead bodies between Augnst and October is far below the facts. Our experiences with 

 Kitovi and Lukanin in 1897 would indicate that 50 per cent would not have been an excessive 

 estimate. Many of the carcasses are washed away in the storms. Every one im which the skin is 

 broken is <inickly reduced to bones by the gulls and foxes. Furthermore, many bodies are necessarily 

 overlooked as they lie hidden among the rocks. 



