464 EIGHTH PACIFIC SCIENCE CONGRESS 



cold ocean currents or there may have been other causes. And the 

 more millions of years have been available for development, the greater 

 is the chance for exceptional and favourable combinations of factors 

 to occur; this may possibly also have played its part." 



It would appear (as will be seen later from the distribution of the 

 skuas, Pacific albatrosses, the fulmars, and the petrels of the Cape 

 Verde and Madeira group of islands) that the main highway for these 

 faunal transgi'essions was alongside the wester?! coasts of the adjacent 

 continents, in other words in the eastern Pacific and in the eastern 

 Atlantic (cf. Murphy and Penneyer, 1952, p. 35). Ekman (1953, p. 

 248) presents similar arguments for the invertebrate and fish faunas 

 of the shelf, stating that "of the two main longitudinal routes for the 

 benthal fauna, the West African and the West American, the latter is 

 by far the more important." The surface temperature charts provided 

 by Schott (1926, 1935) demonstrate the latitudinal narrowness, even 

 at the present time, of the warm-water tropical belt in the eastern 

 Atlantic and the eastern Pacific. 



No evidence of a transgression exists in the Indian Ocean; if it 

 occurred no southern elements could have survived in the absence of 

 cool-water refuges in the geographically limited northern Indian Ocean 

 (cf. also Ekman, ibid. p. 72, who believes that no climatic change of 

 importance took place in the Indo-Malayan region either in the Ter- 

 tiary or Quaternary Periods and "the tropical fauna of the Cretaceous 

 and Eocene could maintain itself in all its tropical abundance and 

 developed until our own time.") 



The possible existence of other, mid-oceanic, routes for the passage 

 of the cool-water faunas across the tropics in Glacial Periods and in 

 the later Tertiary is suggested by the occun-ence of islands in the equa- 

 torial Pacific and Indian Oceans with immense stores of ancient 

 guanos accumulated in the Pleistocene, and quite unassociated at the 

 present time with any bird colonies capable of producing them. Hut- 

 chinson (1950) has done a lasting service in marshalling all known 

 facts on the subject. He assumes considerable climatic and hydrological 

 changes in the central Pacific to account for the accumulation of these 

 deposits, the like of which is now known only from the Humboldt 

 Current region of South America. As C. A. Fleming has pointed out 

 to me (in litt) these equatorial guano islands may be significant as 

 pointers to the existence of flourishing bird haunts in the Pleistocene, 

 when the equatorial upwelling divergence belts were more intensive 

 than they are now. As Hutchinson has shown for the Peru and Chile 

 coasts such areas are very sensitive to changes in world climate and the 

 post-Pleistocene period has seen radically altered conditions in the 

 equatorial Pacific and Indian Oceans. This subject will provide re- 



