54 



THE FUR SEALS OF THE PBIBILOF ISLANDS. 



FLUCTUATIONS OF POPULATION. 



The daily counts of the breeding season of 1897 may here again be cited to give 

 an idea of the real condition of the rookeries at their maximum. The following figures 

 are for that part of Kitovi rookery called the Amphitheater, which contained . 

 according to the count of pups made on August 3. about 1,1*45 breeding females lor 

 the season : 



Amphitheater of Kitovi. 



Such is the height of the season. The actual count shows a difference of 6 per 

 cent between its beginning and its maximum and a difference of 38 per cent between 

 the maximum and its close, while between two individual days of the period there is 

 as great a difference as 20 per cent. 



INCREASE OF FAMILIES. 



Nor is the fluctuation in individuals all that is to be noted in this consideration of 

 the height of the season. The following count of harems on this same breeding 

 ground shows equally important results: 



Amphitheatei- of Kitori, 1897. 



It is thus apparent that during this time of supposed rookery stability the 

 number of harems underwent quite as marked a change as did the number of indi- 

 vidual animals composing them. The daily observations of this breeding ground and 

 frequent photographs of its area show, moreover, that the extent of ground occupied 

 grew steadily from day to day. 



WHAT THE HEIGHT OF THE SEASON MEANS. 



What the height of the breeding season really means, therefore, is a time in 

 rookery development when the stream of incoming cows about equals the stream of 

 outgoing ones. It is the time when the greatest number of cows are actually present 

 at one time. It marks the maximum of rookery development, which probably covers 



