SEAL LIFE ON THE PKIHILOF ISLANDS. 59 



It was in 1873 that. Mr. Klliott estimated the number of seals on the 

 Pribilof Islands at 4,700,000, and he again tells us in 1SS1 that the 

 seals never had been more numerous than they were then; but in 1890 

 lie found them reduced to 959,393 seals, including everything on the 

 islands, or about one-fifth of what the herd had been in 1873. 



Ill 1891 the Treasury agents on the seal islands were instructed to 

 make daily visits to the rookeries during the breeding season for the 

 purpose of noting the peculiar habits of the seals and carefully estimat- 

 ing their numbers at various dates on each rookery, and the highest 

 estimate made, not including the pups, was somewhat less than half a- 

 million. 



1 was one of the agents who did this work in 1891, and I have spent 

 hours and days and weeks, in turn, watching the cows from their first 

 landing. They would often stay away from their offspring for a week 

 at a time. 



I have selected a favorable location on the lleef rookery, where I was- 

 soine 30 feet above the harem and out of danger of being discovered by 

 the seals below, and I have watched one particular pup from its birth 

 until it was a month old; and I found that the cow left it for an hour or 

 two only at first, then for a day, and by the end of the month for four 

 to six days at a time. 



This fact, coupled with another that I observed in 1890, convinced 

 me that the fur seals do not digest their food as rapidly as some other ani- 

 mals, and consequently they can live longer without eating or drinking. 



The other fact referred to is this: In 1890 we killed for the natives 

 on St. Paul Island some 2,3G4 pups, after all the cows had been gone 

 from the island for more than two weeks, and we found the stomachs 

 of all those pups full of pure, undigested milk. 



I walked over all the rookeries on St. Paul Island twice during the 

 season of 1891, beginning at Halfway Point on July 7, and completing 

 the second journey at Northeast Point on July 22, and the highest esti- 

 mate I made of the number of seals on each was as follows: 



K'ookery. Sc;ils. Rookery. Seals. 



This estimate was made on the basis of an average of 40 cows to each 

 bull, and it was assumed that only one half the bulls were in sight at 

 any one time, or, in other words, we could not get close enough to see 

 them without disturbing the seals, so we multiplied the number found: 

 by 2, and the product by 40, in order to obtain, approximately, the num- 

 ber of seals on a rookery. 



It is possible, of course, that the method of computation adopted was 

 not the best and that we probably missed the real number by many 

 thousands, plus or minus, but for all practical purposes of comparison 

 between the condition of the rookeries in 1891 and 1894 it is as good as 

 perfection, for it is enough to show that no matter how many seals were 

 there in 1891, not to exceed one-half of the number were to be found in 

 1894. 



The same is equally true of St. George, where the rookeries, because 

 of their relatively smaller area, show the decrease at a glance to any- 



