406 ARTICLE BY DR. J. A. ALLEN. 



near tlic const, the route varying in different years, being' evidently 

 governed by tlio runs of fish and the position of the various fishing banks. 

 They move leisurely as compared with the males, which have preceded 

 thein, the females being heavy with young, and pausing often to feed 

 and sleep, but landing nowhere till they reach their sole and only breed 

 ing grounds on the Pribilof Islands. 



Tlie Pribilof herd has thus had its own exclusive home, with fixed 

 and definite lines of migration along the western coast of North 

 America. 



2. The Commander Islands herd is evidently distinct and sei^arate 



from the Pribilof Islands herd. Its home is the Com- 

 Commaii. cr s aui s. jj^j^^i^p oi-oup of ishiuds ou the wcsteru side of Bering 

 Sea,and itslineof migration is westward iind soutliward alongthe Asiatic 

 coast. To suppose that the two herds mingle, and that the same ani- 

 mal may at one time be a member of one herd and at another time of 

 the other, is contrary to what is known of the habits of migrating ani- 

 mals in general. Bt^sides, while the two herds are classified by nat- 

 uralists as belonging to one and the same species, namely, the Callo- 

 rhinm ursiwift, they yet present slight physical differences, as in the 

 shape of the body and in the character of tlie hair and fur, as regaids 

 both color and texture, sufficient not only to enable exjwrts in the fur trade 

 to recognize to which herd a given skin belongs, but sufficient to affect 

 its commercial value. As yet, expert naturalists have been unable to 

 make a direct comparison of the two animals, but the differences alleged 

 by furriers as distinguishing the representatives of the two herds ])oint 

 to their being separable as subspecies, in other words, as well-marked 

 geographic phases, and thus necessarily distinct in habitat and migra- 

 tion. 



3. Since fur-seal breeding rookeries are reported to have formerly 



existed on some of the small islands off" southern Cal- 

 isiandsoflf southern jfomia, it has becu assujued that they were a portion 



of the Pribih)f herd, which sometimes remain south to 

 breed. Such an assumption is entirely opposed to what is known of 

 the habits and distribution of marine life and to Avell-grounded princii)les 

 of geographic distribution, namely, that a fur-seal breeding on an 

 arctic island, which it annually travels thousands of miles to reach, 

 would also choose for a breeding station an island in subtropical lati- 

 tudes. Portunately the rebuttal of this assumption does not depend 

 ujjon the generalizations of the naturalist, since si)e(amens have been 



recently obtained from (iuadalupe Island which show 

 Gun.ininpo Island. ^^^^^^ ^^^^.^^^ ^ fur-scal formerly occurrcd there, and is 



still found there in small mnnbers, it is not only not the Pribilof s])ecies, 

 but a seal belonging to a distinct genus, hitherto only known as an in- 

 habitant of the southern hemis])here. This Guadalui)e Island fur-seal, 

 of which I have had the opi)ortunity of exannning, in conjunction Avith 

 Dr. O. Hart Merriam, a series of four skulls, ])roves to be a species of 

 the genus Arctooej^lialiis, and is api>arently closely allied to the fur- 

 seal of the Galapagos Islands, the previously most northern known 

 limit of the genus. 



4. There is not only no evidence to show that the fur-seal of the Prib- 



ilof Islands ever lands upon any ])art of the shore or on 

 Fin-Ss "^ ^^''''^■^" any part of the islands of the"w(>stern coast of North 



America south of the Piibilof Islands, but there is also 

 no evidence that it ever brings forth its young at sea. either in the 

 Avater or on floating beds of kelj). Such a method of breeding is obvi- 

 ously a i)hysieal impossibility, Avheu the character of the animal, and 

 particularly the condition of the young at birth, is duly considered. The 



