NUMBER OF SEALS ON SOUTH ROOKERY. 167 



full view, and Eodger, to use Mr. Barrett Hamilton's own words, "by stampeding 

 tLem added 3 bulls to my list." 



The general proposition, however, that there are still proportionally much fewer 

 males on Bering Island than on Copper Island holds good (even after the killing of a 

 number of them last year on the latter island) and is conceded by all. 



Whether this comparative dearth of males on Bering Island is particularly 

 injurious to the condition of the herd will find a negative answer in the South rookery 

 of Bering Island, as suggested by me on page 64 of my Eussian Fur Seal Islands. 



South Boolcery. — In view of the above suggestion I was ordered to pay special 

 attention to the South Eookery during 1897. 



There being no inhabitable house at this rookery, we were obliged to camp in the 

 neighborhood, and from July 24 to 30 we (Mr. Barrett-Hamilton and myself) visited 

 the rookery two or three times a day. 



It will be remembered that this rookery is very sma.ll and situated under a steep 

 bluff, which makes it possible to count the seals with some degree of accuracy. My 

 visit in 1895 was too short to allow an actual count of the females, but I estimated 

 their number at "about 600," while the maximum number of sikatchi, or bulls, was 

 said by good authority to have been only 5 during that season. In my report upon 

 that visit I urged the advisability of undertaking an exact count of the young ones 

 the following year, in order to ascertain whether these few bulls had been sufficient 

 for the impregnation of the whole number of females. I myself arrived too late in 

 1896 to be able to do it, and nobody else took up the suggestion. It was ascertained, 

 however, that no more than 6 sikatchi had frequented the rookery that year. 



In 1897 Mr. Barrett-Hamilton agreed to undertake the count in common with 

 myself. At first we intended to drive the pups off in a body and count them in that 

 way, but there were various objections to this plan; first, that the driving would 

 materially interfere with the sealing at this rookery, the bachelors hauling up among 

 the females and being culled from among these; second, that so many of the pups 

 were in the water during the day (at low water) (pi. 68) that it would be impossible 

 to gather them all together on land, while in the evening, when they all came ashore, 

 water was high (pi. 69), thus preventing any driving at all. We relinquished this 

 plan the more willingly as we found it quite feasible to make a fairly accurate count 

 from the bluff overlooking the rookery. 



Our general mode of procedure was first to define small separate groups of pups 

 on shore and to count each one of these separately, then compare our figures, and in 

 case of disagreement to count them over and over again until we arrived at nearly 

 the same figures. We then counted those lying among the females nursing, those on 

 separate rocks in the water, and those swimming. After a series of counts we found 

 that only those made in the evening at high water, when all the pups were ashore, were 

 of any value (pi. 69). An average of a selection of our best counts (7) gives 526 pups 

 (minimum, 516; maximum, 533), which may be accepted as nearly exact. 



Similar counts of the females were made regularly. As a matter of course, the 

 figures for the various counts of the females vary very much more than those of the 

 pups for two obvious reasons : First, because the actual number of females present 

 on the rookery varies greatly from day to day, while that of the pups is constant, 

 except for the gradual increase due to new births or slight decrease due to death ; 

 second because there nearly always were a number of females swimming off the 



