SHORT-TAILED PETREL. 
But it does not seem logical for a moment to suggest that birds met with 
in California in December, could possibly be identical with birds breeding in 
November in southern Australia. The localities negative such an assertion 
as well as the time, for the southern Australian breeding-birds arrive in October, 
and at the present time all the known breeding localities of P. t. hrevicaudus are 
in southern Australia. Stejneger records P. tenuirostris from the Commander 
Islands {Proc. U.S. Nat. Mus. 1887, Vol. X., p. 125, 1888), as “ not common, 
but probably breeding,” and in the British Museum are three specimens from 
Bering Island, all collected by Captain Barrett-Hamilton in May. Examination 
of these specimens points to their being breeding birds. 
All the P. carneifes I have yet seen that have been procured in Japanese 
waters have been obtained in May- June. I conclude this also breeds somewhere 
in the north, and it should be recorded that under P. leucomelas, a species which 
has not been found breeding in the Southern Hemisphere, Godman wrote 
{Monograph of the Petrels, p. 73) : “ It will be noticed that the above records 
refer to the occurrence of P. leucomelas in Corea and the Japanese Islands in the 
months of May and June ... We may infer that it was breeding in these 
northern islands.” If this inference is permissible in this case a coincident 
suggestion must be admitted in the exactly parallel cases (as far as their occur- 
rences in Japanese waters are concerned) of P. tenuirostris and P. carneipes. 
103 
