86 THE FUR SEALS OF THE PRIBILOF ISLANDS. 



to iiulicate that tlie spafe unit for massed breeding: sronnds should be smaller. But 

 for the underestimate which may therefore be involved on this account we can otter 

 no correction. For the underestimate due to the early date at which the count was 

 made we can make ii rough estimate. Tiie daily count on Lukanin rookery for the 

 season of 1897 shows that between the Sth and 15th of -luly tliere was an increase of 1.5 

 jier cent in rookery population. This would increase the ligures for 1805 as orijjinally 

 given to about 80,000, and after doubling for absentees the corrected total would 

 be about 100,000 breeding females. The inclusion of the yearlings and 2-year-old 

 females does not aft'ect the total, as they were not present, and no allowance need be 

 made for them. 



THE ESTIMATE REVISED. 



This total of 100,000 females, or giving to each female a pup and adding the 

 estimated number of breeding bulls, making 325,000 "breeding seals and young," is 

 probably within 10,000 of the facts for the season of 1895. That it comes thus near 

 the truth, however, is the result of accident rather than good management. The 

 corrections which, in the light of subsequent experience, we have been able to make, 

 are vital to its truth and change the results radically. The original results could not 

 have been trusted alone, and were wholly misleading. 



THE IMPORTANT FEATURE OF THE ESTIMATE. 



The really important feature in the estimate of 1895 is the count of cows in which 

 it originates. This was a distinct step in advance, in that it approached a rational 

 basis. In the application of the unit of si)ace to the rookeries not counted the method 

 was unfortunate. The area of breeding ground was taken from inaps in themselves 

 imperfect, on which the rookery outlines were sketched by the aid of the eye. The 

 rookery boundaries, as we have shown, are constantly changing as the season 

 advances, and there being no definite landmarks to guide the observer, it is impos- 

 sible that the outlines should be correctly located. The enumeration is therefore 

 carried into the region of pure speculation and has only the value of the individual 

 judgment of the person tracing the ma])s. 



It is fair to say, however, that nothing definite and exact was claimed for the 

 census of 1895, as Mr. True's own words, in commenting upon it, will show. He says: 

 "I do not think that any estimate can be made wiiicli will approximate the truth 

 more than remotely," and he continues to say that the chief use of such calculations is 

 "the elimination of fanciful estimates of the number of seals." 



MR. CROWLEY'S ESTIMATES FOR 1805. 



In leaving the estimates of 1895 it is necessary to refer to two other calculations 

 of rookery population made for the same year on a dift'erent basis. One of these is 

 by Mr. J. r>. Growlej', chief agent in charge of these islands, lie finds, as he says, 

 by actual count, a total of 99,930 breeding cows and 5,552 breeding bulls. When we 

 make the necessary doubling of this estimate of cows and add the pups we have a 

 total of about 305,552 "breeding seals and young." Of the methods or details of this 



