364 SEAL LIFE OX THE PRIHILOF 'ISLANDS. 



every individual breeding male or bull seal who bad a harem, noting 

 and counting very carefully, too, every idle bull, or, in other words r 

 every bull whose youth, strength, and vigor fitted and qualified him for 

 a harem had there been cows to be found in sufficient numbers to sup- 

 ply them, which, unfortunately, there were not. So carefully and so sys- 

 tematically was the counting done that I feel I can recommend the 

 figures as being as nearly correct and reliable as it is possible to get 

 them. 



Under the head of bachelors, or young males, are included all the 

 seals on the islands other than those on the breeding rookeries, many 

 of them being young females, too young to go on to the breeding 

 grounds. 



The bachelors have been estimated by me in the usual manner of 

 estimating a bunch of seals, and they may very possibly run a thousand 

 or two more or less than the figures given. 



The number of breeding females or cows is based on an arbitrary 

 average of 40 to the harem, or 40 cows to every breeding bull, as was- 

 adopted in and followed since 1891, though 1 am of the opinion it was 

 an overestimate and that the harems never did and do not now contain 

 an average of 40 cows each. 



Having adopted that number, however, and having used it so long 

 in our estimates, it was necessary to use it in the present instance for 

 the sake of making fair comparisons when considering the steady 

 annual decrease of the seal herd and the shrinkage of the rookery area* 



Admitting the average number of cows in a harem to be less than 

 40 and I believe all who know anything about seal life on the rook- 

 eries will admit it is then the total number of seals in the herd, as- 

 estimated by me, will be that much less in proportion. 



By way of explanation I will say that when we first attempted to 

 count the bulls, in 1891, for the purpose of getting, approximately, at 

 the number of seals on the islands, it was deemed best to run the risk 

 of overestimating the herds, lest Great Britain should object to our 

 figures and insist on a recounting and, possibly, discover an error upon 

 which to base an argument against us for the purpose of showing our 

 anxiety to prove the wicked wastefulness of pelagic sealing. 



As the seals were at that time too numerous and the harems too com- 

 pact to admit of our going through and among them as we can now, we 

 simply aimed to count every bull we could see and multiply the num- 

 ber found by two. on the ground that it was not possible to penetrate 

 the mass far enough to see more than one-half of them. 



And, lest that was not enough, we allowed an average of 40 cows to 

 each harem, although we were quite certain it was too high. 



I have gone over the rookeries every year, in season, since 1891, and 

 I have noted the steady decrease of the herd from 500,000 then to 

 237,800 in 1895, when, because of the decrease, I was able to go in 

 among the herd at the height of the season and count every bull on the 

 islands. 



Whether we erred in our estimates in our first crude efforts to get at 

 the facts is of no consequence now, for the fact remains that, no matter 

 what the actual numbers were in 1891, more than one-half of the whole 

 herd has been exterminated since then. 



Taking it for granted that the estimates were wrong, the proportion 

 is still correct for all practical purposes, so that if we take the 500,000 

 of 1891 against the 237,800 of 1895, we find an average annual decrease 

 of 52,440 for the five years beginning with 1891 and ending with 1895. 



That the average annual loss has been greater than this can be- 



