62 EEPORT OF BRITISH COMMISSIONERS. 



observations or ill-founded conjectures sucli as the resident agents on 

 the breeding islands have been able to make with their limited opjjor- 

 tunities. The circumstances in 1891 were, however, exceptionally 

 favourable for acquiring information of a comparable kind on the ques- 

 tion of distribution. 



212. The observations at command for 1891 practically cover pretty 

 thoroughly the period of about two months during which seals are ordi- 

 narily taken by pelagic hunters in Behring Sea, extending from the 

 middle of July to the middle of Sejitember, and they are nmch more 

 complete for the eastern than for the western part of the Behring Sea. 



213. On consideration of the material to be dealt with, it was decided 

 that it might be most advantageously divided into two periods of about 

 a month each, the first including all dates from the 15th July to the 15th 

 August, and the second those between the 15th August and the 15th 

 September. All the lines cruized over in the first of these periods were 

 plotted on one set of maps, and those in the second period on another. 

 The parts of these tracks run over during the night, and in which seals 

 therefore could not well be observed, were indicated on the maps in a 

 dilferent manner from the day tracks, as far as possible; and with the 

 assistance of the logs, the numbers of seals seen in certain intervals 

 were then entered along the various routes in a graphic manner. The 

 places in which pelagic sealers had reported seals to be abundant or 

 otherwise, as well as those in which sealing- vessels were found at work 

 by the cruizers, and other facts obtained from various sources, were also 

 indicated on the maps. 



214. Without attempting to enter into further details here as to the 

 methods employed, the general results arrived at may now be briefly 

 described: 



It is evident, in the first place, that the seals are most abundant in 

 the water in the immediate vicinity of the shores of the breeding islands, 

 this abundance of seals extending often not more than half-a-mile from 

 the fronts of the breeding grounds, and seldom for 3 or 4 miles in such 

 a way as to be at all notable. In the case of the Pribyloflf Islands, it 

 is also observed that seals were numerous in both the monthly periods 

 in the tract included in a general way between St. Paul and St. George 

 Islands, though they diflered much in this respect even at nearly 

 approximate dates. It is further clearly shown that the Pribyloff and 

 Commander groups form the main centres of abundance of seals in 

 Behring Sea during the summer; but that while this is undoubtedly the 

 case, the seals are not found to decrease in numbers with any approx- 

 imation to regularity in zones concentric with the islands, — always 

 excluding the seals in the immediate neighbourhood of the shores. 



215. It is therefore not possible to outline a series of zones in which 

 the number of seals present will bear an inverse ratio to the distance 

 from the islands. It is, however, possible to draw an approximate limit 

 lor a region about the Pribylofit group, which will roughly define the 

 area of abundant seals at sea during each of the two monthly periods 

 chosen. In the case of the region about the Commander Islands, data, 

 though almost wanting for the first monthly period, and but scanty for 

 the second, are sufficient to indicate a general mode of distribution 

 similar to that demonstrable in the first case. Within the areas of 

 abundant seals, these animals are, however, by no means regularly 

 distributed, even at any particular fixed date, but are scattered in 

 irregular patches in the diffuse character already described, and are 

 very often thickest locally towards the outer limits of the area. 



210. Beyond these areas, seals are found more or less sparsely scat- 

 tered over a great part of Behring Sea, which in the first period extends, 



