AMPHIBIAX MILLIONS. 307 



Undoubtedly some abnormal birth-shapes must make their ap- 

 pearance occasionally ; but at no time while I was there, searching 

 keenly for any such manifestation of malformation on the rookeries, 

 did I see a single example. The morphological symmetry of the 

 fur-seal is one of the most salient of its characteristics, viewed as it 

 rallies here in such vast numbers ; but the osteological differentia- 

 tion and asymmetry of this animal is equally sur2:)rising. 



It is perfectly plain that a large percentage of this immense 

 number of seals must die every year from natural limitation of life. 

 They do not die on these islands ; that much I am certain of. Not 

 one dying a natural death could I find or hear of on the grounds. 

 They evidently lose their lives at sea, preferring to sink with the 

 rigor mortis into that cold, blue depth of the great Pacific, or be- 

 neath the green waves of Bering Sea, rather than to encumber and 

 disfigure their summer haunts on the Pribylov Islands. 



Prior to the year 1835, no native on the islands seemed to have 

 any direct knowledge, or was even acquainted with a legendary tradi- 

 tion, in relation to the seals, concerning their area and distribution 

 on the land here ; but they all chimed in after that date with great 

 unanimity, saying that the winter preceding this season (1835-36) 

 was one of frightful severity ; that many of their ancestors who had 

 lived on these islands in large barraboras just back of the Black 

 Bluffs, near the present village, and at Polaviua, then perished 

 miserably. 



They say that the cold continued far into the summer ; that im- 

 mense masses of clearer and stronger ice-floes than had ever been 

 known to the waters about the islands, or were ever seen since, 

 were brought down and shoved high up on to all the rookery mar- 

 gins, forming an icy wall completely around the island, and loomed 

 twenty to thirty feet above the surf. They further state that this 

 frigid cordon did not melt or in any way disappear until the mid- 

 dle or end of August, 1836. 



They affirm that for this reason the fur-seals, when they at- 

 tempted to land, according to their habit and their necessity, 

 during June and July, were unable to do so in any considerable 

 numbers. The females were compelled to bring forth their young 

 in the water and at the wet, storm-beaten svirf -mar gins, which 

 caused multitudes of mothers and all of the young to perish. In 

 short, the result was a virtual annihilation of the breeding-seals. 

 Hence, at the following season, only a si)ectral, a shadowy imitation 



