Vol. X. 

 1910 



1 Iredale, Bird Life on the Kermadec Islands. 13 



all over the island, but in no case are these of any extent, and 

 none, as far as was known, exist on the outlying islets. The 

 account given by Cheeseman of the breeding habits of this bird 

 requires modification. That account was drawn up very shortly 

 after the bird was discovered. As stated above, it is not solitary; 

 the colonies are always placed on the ridges, and it would be a 

 very exceptional bird that burrowed at the base of a tree-fern. 

 It does not arrive until October, and leaves in May. A downy 

 young one was examined on 29th March ; it was of a beautiful 

 dove-grey, with white on the breast. Fully-fledged young were 

 noted as early as 3rd May. 



The second burrowing CEstrelata was formerly recorded as 

 CE. cookii, but, though a close ally, is perfectly distinct, and was 

 separated by Rothschild under the name nignpeunis. It breeds 

 more numerously than the preceding species on Sunday Island, 

 whilst colonies exist on the outlying islets, as well as Macauley 

 and Curtis Islands. This is a most lovely and gentle little 

 bird, never biting hard when handled. On 4th January a 

 colony was examined, and fresh eggs were numerous, very {q\w 

 being hard-set, whilst in many holes birds, but no eggs, were 

 found. On 29th February downy young were noted, and these 

 were miniatures of those of cervicalis, though the adult plumage 

 of these two species is very different. 



Most interest was evinced in the species of (Estrelata reported 

 from the Kermadecs as breeding on the surface of the ground, 

 and not burrowing, as is usual with members of that genus. 



Cheeseman, in 1890, recorded three species — mollis, Gould, as 

 breeding in the summer on Sunday Island ; sp. indet., as breed- 

 ing in the winter on Meyer Island ; and neglecta, Schlegel, as 

 having been previously recorded. 



In 1893 Hutton wrote on the Petrels of the Kermadecs, and 

 then allowed three species. He showed that the bird Cheese- 

 man recorded as mollis, Gould, on Buller's identification was 

 neglecta, Schlegel ; added philippi, Gray, and introduced a new 

 species, lencopJirys. Their habits he was not certain about, and, 

 endeavouring to obtain fresh information, obtained such that in 

 1904 he reduced the three species to varieties. However, to his 

 three varieties he accorded such distinct habits as to make 

 doubtful the reduction. Consequently BuUer, in 1906, in the 

 " Supplement to the Birds of New Zealand," still maintained as 

 valid three species, but with this difference : he retained mollis, 

 philippi, and neglecta, including leucophrys as a synonym of the 

 last-named. In the work quoted Buller strangely introduces 

 under the three names habits at variance with his authorities, 

 thus quoting und^i- philippi Button's account of 1 893, although 

 in 1904 Hutton, from later information, had retracted that 

 account. 



It will be seen from the foregoing rough sketch that the 



