1(;8 THE FUR SEALS OF THE PRIBILOF ISLANDS. 



The explanation of this conditiou of tliiugs is that on the rookeries of St. George 

 the bhie foxes,' of which there are many, had eaten all the pup carcasses without 

 exception. Tlie final count had to he made simply by skulls, or spinal colums, or such 

 parts of the animals as could be positively identified. 



RECONSTRUCTION OF ST. GEORGE ESTIMATES. 



It is unnecessary to remark that these figures for St. George can not be used as 

 they stand. Some sort of estimate must be made to take their place. No fairer basis 

 for such an estimate exists than to apply to the rookeries of St. George the ratio of 

 dead pups foiuid in October on St. Paul. On this island the ratio of dead pups found 

 in October to the total number estimated to have been born was 11.19 per cent.' This 

 would necessitate the addition of 1,362 pups to make the conditions of St. George 

 comparable with those on St. Paul. This amount, together with 150 pups which were 

 removed from the rookeries of both islands during the breeding season forjjurposes 

 of dissection, makes the total death rate subsequent to the middle of August, and 

 directly chargeable to starvation, aggregate 16,019.^ 



THE DETAILED ESTIMATE. 



This total, as will be seen by the explanations already given, is not entirely 

 satisfactory, but it is a real one so far as it goes, and fully 12,000 of it is an actual count, 



' In Mr. Macoun's report for 1896 this undoubted fact, which he saw with his own eyes, is 

 needlessly ([uestioned. If the foxes, as was the fact, had completely destroyed the 897 carcasses 

 which he coimted, It is beftging the question to assume, as he does, that they, aided by the elements, 

 could not have destroyed the 1,362 additiimal pups necessary to make the conditious of St. George 

 .agree with those of St. Paul. 



-' This percentage is computed on the original census of 1896, not on the revised figures substituted 

 in 1897. 



' A strong effort has been made by Professor Thompson and by Mr. Macoun in their reports of 

 1896 to weaken the force of these definite results regarding the starvation of pups. Not finding it 

 possible to denj- the fact of starvation or its importance, they have endeavored to minimize its effects 

 by insisting that the lauses of early mortality continued into and were at work in the period of 

 starvation. Nothing could be mure misleading than this. The early causes of death are Uncinaria, 

 trampling, starvation through early separation from the mother, drowning, accidents. All these 

 have to do with the wiakness and helples-sness of the very young pups, and must necessarily cease 

 with the close of the breeding season. Tlie pups die from or outgrow the worm before September 1. 

 In fact, this cau.se is practically inoperative after August 20. Pups are only trampled in the first few 

 days after birth, and they are only lost from their mothers, if at all, at this time. They certainly do 

 not drown after they have learned to swim, which they do by the middle of August. The period of 

 death from starvation lasts from August 15 to October 20. The catises of the early and later mortality 

 were absolutely distinct, and that the latter was due practically without exception to starvation 

 needs no demonstration to those who made the count of the dead and dying pups on the rookeries of 

 St I'anl in October, 1896. 



