PELAGIC BIRD FAUNAS OF THE INDO-PACIFIC OCEANS 481 



The rich breeding station o£ Goenoeng Api, a volcanic cone rising 

 directly from the ocean floor in the Banda Sea (Van Bemmel and Hoo- 

 gerwerf, 1940) suggests that possibly safe nesting sites may be a con- 

 tributing factor (see Gibson-Hill, 1949, p. 232, and Serventy, 1952 for 

 data on human interference with sea-birds in these seas). 



2. ON A SUB-SPECIATION CENTRE IN THE SOUTHWEST PACIFIC. 



A Study of the distribution of the sea-bird species in the Pacific, 

 particularly of the petrels, indicates various island groups, isolated from 

 each other, where differentiation has gone on to produce fairly distinc- 

 tive forms. It is not my purpose to make an exhaustive analysis of all 

 of these, but to draw attention, very briefly to one such area in the 

 Southwest Pacific. The idea is put forward with a view to stimulating 

 discussion on the subject and to find out whether the thesis has a wider 

 application than among sea-birds. 



Its existence has been emphasised by the recent taxonomic work 

 of Murphy (1951) and Murphy and Irving (1951), on Puffinus pacificiis 

 and Pelagodroma marina, respectively. 



The widespread P. pacificiis, though it has geographically separated 

 populations in the Indian and Pacific Oceans, has not differentiated 

 subspecifically in strict correspondence with these geogiaphical di- 

 visions. But there has, however, developed a distinctive race breeding 

 in the Kermadec Islands. 



Pelagodroma marina in the Southwest Pacific occurs in the 

 southern Australian and New Zealand regions and in the Kermadecs. 

 The Kermadecs area is occupied by the most strongly marked sub- 

 species though it "is isolated by only short distances from populations 

 characterized by what might be called the world-wide pattern" (Murphy 

 and Irving). 



The Kermadecs region has evidently provided conditions which 

 have isolated populations of sea-birds both of more northern and more 

 southern distribution to produce demonstrably different forms. What 

 these isolating mechanisms were is unknown to me, but must un- 

 doubtedly be related to the hydrology of the area. 



In this general area, though not at the Kermadecs themselves, 

 other local forms have developed. They may possibly be the result of 

 the same set of isolating factors. The present distribution of parent 

 and daughter forms need not necessarily be the same for each species, 

 and the biological fronts between them may have shifted fairly con- 



