1893.] PETRELS FROM THE KEEMADEC ISLANDS. 755 



variations ; for it seems to be certain that no individuals of tlie 

 variety occur on Meyer Island when the winter Mutton-bird is 

 breeding. 



7 (EsTEELATA PHiLLiPi, Grray. 



Norfolk Island Petrel, Phillip's Voyage to Botany Bay, p. 161 

 (London, 1789). 



Procellaria jihillijiii, G. E. Grray, Ibis, 1862, p. 246. 



(E. mollis, dark variety, BuUer, Trans. N. Z. Inst. vol. xxiv. 

 p. 85. 



Three specimens from Sunday Island. Length 16 inches, 

 wing 11*5, tail 4*75, bill 1*2, tarsus 1*5, mid toe 1*8. 



Upper surface sooty black, the feathers on the forehead and on 

 the back (in one specimen only) margined with brown. Under 

 surface grey, washed with brown on the abdomen. Sometimes 

 some hght feathers on the lores and chin, and a dark mark in 

 front of the eye. Bases of all the contour feathers white. 

 Primaries white on the inner webs for the greater part of their 

 length, the white terminating bluntly and reaching the shaft 

 some distance outside of the tips of the lower wing-coverts. 

 Outer tad-feathers brown, with white at the inner bases only. 

 Bill black. Tarsi, first joint of inner and middle toes, and the 

 web between them, brownish ; the distal parts of the foot black. 

 The wings when folded exceed the tail by more than an inch. 



I have little doubt but that this bird is the same species as the 

 Norfolk Island Petrel of Grovernor Phillips. It approaches the 

 last species, but can be distinguished by the shape of the white 

 on the inner web of the primaries as well as by its colours. 

 Having examined three specimens which, although varying slightly 

 in colour, are constant in this x'espect, I have no hesitation in 

 admitting it as distinct. 



It is this bird, probably, when flying on the New Zealand coast, 

 that I formerly mistook for Pterodroma atlantica (Gould) \ of 

 which there is no authentic record of its having been taken near 

 New Zealand. 



INIi'. Cheeseman informs me that he did not see this s])ecies at 

 the Kermadecs and knows nothing about its breeding-habits ; but 

 that Mr. Bell sent him specimens with the summer Mutton-birds, 

 so that it probably breeds with them from September to November. 

 The Norfolk Island Petrel is said to form burrows in the sand. 



It is a remarkable fact that in the genus (Estrelata there are 

 three bicolour species each closely related to a unicolour species — 

 viz. (E. neglecta to CE. solandri, (E. arminyomana to (E. trini- 

 iatis, and (E. mollis to (E. brevirostris — the two forms appearing 

 in all three cases to breed near together on the same islands. I 

 am not aware of the same thing occurring in any other genus of 



' This speciea has been identified with Procellaria fidii/lnosa, Forster, but it 

 is not tlie (Edrclata fulujlnosa of BuUor'is ' Bii-ds of New Zealand,' which appears 

 to be the larger species Pterodroma macroptera (Smith). 



