7f) THE FUR SEALS OF THE PRIKILOF ISLANDS. 



tlierel'ore be taken, and in a mile of rookery space tlie miniber of (listingnishable 

 marks of this sort is exceedingly small. Where stones exist, there are thousands of 

 them practically iudistinnuishable. On bare slopes, as on Gorbatch, Yostochui, and 

 Polovina, there are no natural landmarks whatever. 



THE ABSENCE OF RELIABLE SURVEYS. 



If perfectly recognizable artificial landmarks could be placed at every angle, turn, 

 and projection of the belt of breeding seals, or if these points could be taken from a 

 distance by instruments and theu reproduced with certainty in the same manner after 

 the animals have left the ground, accurate results might be obtained. Nothing of 

 this kind, however, was done, at least no landmarks remain to show for it. 



THE IRREGULAR NATURE OF THE GROUND. 



But a determination of the rookery borders is not the only diflBculty. Tiie 

 character of the ground is extremely variable. It lies at every conceivable angle 

 and slant. There are narrow, rocky beaches hemmed in by perpendicular cliffs. 

 There are long slopes of jagged bowlders. There are sand Hats and cinder slopes. 

 On the bare places the seals still mass together as closely as they can be crowded, 

 and ou the rocky areas they lie about among the rocks as best they can. Their 

 distribution over the rookeries is as irregular as the nature of the ground. 



It is iu general true that the greater the number of females the more extended are 

 the boundaries of the rookeries; but it is also true that with the decrease of the 

 number of seals the population of the rookeries grows sparse without a corresponding 

 decrease of dimensions. It is probable that when the seals were more numerous 

 they were as evenly distributed over the ground as its nature would permit, and 

 the greater part of each rookery was closely massed; but at the present time their 

 distribution is very irregular, as unequal as the arrangement of the trees in the 

 forest. On some of the rookeries, as ou Tolstoi sands, iu the breeding season the 

 seals lie as thick as swarms of bees. On other rookeries, as the Lagoon, detached 

 harems sprawl over the rocks and individual seals are greatly scattered. Nor are 

 the mechanical imperfection of these estimates all. The counts of live pups made 

 during the seasons of 1896 and 1897 show that at the time when these past estimates 

 were made not inore than half of the cows are present at any one time. 



ACUEiGE MEASUREMENTS CHIEFLY GUESSWORK. 



I n a word, the acreage measurements of the rookeries in the past have been based 

 chierty u|)OU guesswork. More guessing has been done in determining the space to 

 be assigned to individual animals, and finally the rookery population sought to be 

 enumerateil has at best represented only about half the actual number of animals 

 belonging to the herd. The last element of uncertainty was not known until 189(i, it 

 having been assumed up to that time that during the period between the lOtli and 

 20th of July all or practically all the animals belonging to the breeding herd were 

 present upon the rookeries. 



THE MAGNITUDE OF THE PROBLEM. 



It is easy to find in the magnitude of the problem an explanation for the adoption 

 of such ai faulty method. It is not so easy to find an excuse foi- inqilicit reliance 



