90 THE FUR SEALS OP THE PRIBILOP ISLANDS. 



B. THE PRESENT CONDITION OF THE HERD. 

 THE CENSUS. 



In tlie work of the present investigation of the fur-seal herd the most important 

 consideration was the making' of as accnrate an ennmeration as possible of the number 

 of animals. This has always formed an important part of every investigation in 

 the past. Bnt, as we have seen, the results have been anything but satisfactory. Tlie 

 great multittule of the animals, when the herd was five times as great as at present, 

 may have left no otiier result possible. At the present time, even with the hei'd so 

 greatly reduced, the task of making a complete census of all the rookeries is by no 

 means an easy one, as the details of our work, which will be found iu the Daily 

 .lonrnaj, will indicate. 



ITS DIFFICULTY. 



Without going into detail here, we may mention among the difliculties of the 

 problem the fringe of idle bulls, savage and immovable, that skirts eacli rookery, the 

 danger of stamijeding the rookeries themselves, the broken and irregular nature of 

 tlie ground, and, linally, the constantly shifting nature of the rookery population. 

 These are some of the merely mechanical difficulties. But more serious for us than 

 any of these was the fact that at the outset the conditions of the problem before us 

 were not at all understood. It had been currently accepted that during a period 

 between the 10th and 2(lth of July the breeding rookeries were at their height and 

 practically aU the animals present. Upon this supposition all previous estimates had 

 been based. With this idea in mind we began our work, only to find as we advanced 

 that the supposition was unfbunded. 



For the smaller rookeries of St. George, and such rookeries as Kitovi, Lagoon, 

 Zapadui Reef, and the cliff portions of Polovina and Tolstoi, it was found possible to 

 make a count of the individual animals by harems. This was accordingly done. On 

 the greater rookeries, as those of Northeast Point, Reef, and Zapadni, no count of 

 individuals was possible, and for these rookeries only harems were enumerated. 



ACTUAL COUNTS. 



Our count of individual cows in ISittJ covered about one-fifth of the rookery space 

 on St. Paul Island, embracing 1,245 harems, with a total population of 1G,(J79 cows, 

 or an average of 13.4 cows to each harem at the height of the season. The average 

 harem of the individual rookeries and parts of rookeries counted ranged from 11 in 

 the lowest to 17.3 in the highest. The lower averages represented thin and scattered 

 portions of I'ocky breeding ground, and none of the counted area contained any of 

 the massed conditions characteristic of portions of the larger rookeries. The highest 

 average belonged to Kitovi rookery, the largest continuous rookery space counted. 

 It contained 3,152 cows in 182 harems, an average of 17.3 cows to the harem. The 

 conditions of this rookery as a whole being more typical of the general conditions 

 l)revailiiig on the larger rookeries, its average was taken as a basis for computing the 

 l)opulation of those rookeries on which only harems could be counted. The appropri- 

 ateness of this average was the more apparent as on North rookery of St. George 

 the 129 harems accurately counted gave an average of 17 cows to the harem. 



