480 EIGHTH PACIFIC SCIENCE CONGRESS 



No petrel breeds in these waters, not even Puffiniis pacificns and 

 P. Iherminieri. The absence of the former is of some moment as the 

 distribution pattern certainly suggests that the Indian Ocean popula- 

 tions must have formerly been connected with those of the Pacific 

 Ocean through this area. At present the nearest breeding colonies, 

 east and west, are separated by some 1,500 nautical miles of sea way. 

 Furthermore, the Indian Ocean birds are sub-specifically inseparable 

 from the bulk of the Pacific Ocean birds. In the Pacific Ocean a dis- 

 tinctive race is recognisable from the region of the Kermadecs (Murphy, 

 1951, p. 15) though it exists in close proximity, relatively speaking, 

 to other Pacific populations. 



The absence of resident pan-tropical petrels in this area is rendered 

 even more curious by the presence in it, as a wintering migrant, of the 

 storm-petrel, Oceanites oceanicus, breeding in the Antarctic Region of 

 the far south (Serventy, 1952a, p. 107). 



In the shallow seas of this water gap there appear to be no re- 

 ports of breeding stations of Boobies, Tropic-birds, Frigate-birds, and 

 such terns as Sterna juscata, Anous stolidus, Anous minutus and Gygis 

 alba. Goenoeng Api in the Banda Sea (Van Bemmel and Hoogerwerf, 

 1940, p. 421) is a singular exception. Foraging and wandering indi- 

 viduals and flocks of some of these species, however, do occur, par- 

 ticularly of Sula leucogaster. 



Along the northern Australian shoreline, from Torres Strait to the 

 Buccaneer Archipelago in northwestern Australia, there is an inex- 

 plicable scarcity of Larus novae-hollandiae, so plentiful on the remain- 

 ing parts of the Australian coastline. 



It seems that factors are now in control of the situation, account- 

 ing for these distributional breaks, which might not have been operat- 

 ing in the immediate or sub-recent past. A superficial search for pos- 

 sible causes suggests that the surface water temperatures might be, at 

 least in part, responsible. According to the data in Schott (1935, pi. 

 20-23), an extensive band of high water-temperatures exists around 

 New Guinea, northern Australia and out into the Indian Ocean. This 

 is maintained above 28°C. from about November to February and the 

 position of the band correlates very closely with that section of the north 

 Australian area in which Puffinus pacificus is absent as a breeding 

 species. According to Dr. J. Gentilli, of the School of Geogiaphy of the 

 University of Western Australia, the waters of northern Australia, in- 

 cluding the Gulf of Carpentaria, are among the warmest open sea sur- 

 faces in the world, approaching that of the Red Sea. 



