488 Loomis, Wedge-tailed Shearwater. [o"t 



of the lower parts more or less obscured with gray or grayish brown. 



A difference exists in the geographic range of the two phases. 

 The light phase is predominant in the Hawaiian Archipelago, 

 dark-breasted birds being of rare occurrence. On San Benedicto 

 Island the dark phase is in the ascendency, greatly outnumbering 

 the intermediates and white-breasted birds. In the Kermadec 

 Islands only the dark phase is represented. On the east coast of 

 Australia and in the Indian Ocean dark birds also prevail; but 

 sporadic white-breasted ones may occur, for Gould figures such a 

 specimen in Volume VII of his ' Birds of Australia.' 



Although there is a difference in their distribution, the phases 

 of this Shearwater do not correlate with climatic conditions after 

 the manner of geographic variation. On Sunday Islet of the 

 Kermadec group, where the light phase is absent, both phases of 

 the Neglected Petrel are abundant, witnessing that there are no 

 climatic barriers exlcuding light phases. In continental dichro- 

 matic species the factor, of island isolation is eliminated and the 

 lack of harmony of phases with environmental conditions is still 

 more apparent. For instance, the light phase of the Red-tailed 

 Hawk prevails in the humid Eastern States, while both the dark 

 and light phases occur in the arid Western States, along with the 

 gray phase of the Screech Owl and the dark and light phases of 

 Swainson's Hawk. Further examples of the distribution of phases, 

 independent of environment, may be found among the Herons l 

 and other groups having dichromatic species. 



Like geographic variation, dichromatic variation has frequently 

 been mistaken for characters of specific rank, giving rise to numer- 

 ous apocryphal species. The light phases of Puffin us chlororhynchus 

 and Pterodroma neglecta have been respectively designated" Puffinus 

 cuncatus" and "(Estrelata leucophrys." 



Dichromatic and individual variations have evolutionary possi- 

 bilities, and there are just as good reasons for treating dichromatic 

 variations (possible mutations) on the subspecies basis as there 

 are for treating geographic variations on that basis. Whatever 

 course is pursued, the fact remains that the subspecies rests on no 

 better foundation than a theory that begs the question; for we do 



i Cf. Bangs, Auk, Vol. XXXII, 1915, pp. 481-484. 



