388 Nichols, Oceanodroma castro. [j^^iy 



It will be seen that the most southern birds are the largest and 

 lightest; the northern smallest and darkest while those from lower 

 British Columbia are intermediate, though closer to the southern 

 than the northern race. The Mackenzie form, being represented 

 by but a single specimen in not very high plumage, cannot be 

 definitely placed in the series though it agrees closely with the 

 Montana and Idaho specimens. No specimens from the type 

 locality of richardsoiii (the northeast corner of Rocky Mountain 

 Park, Lat. 53°, Long. 115°) being obtainable I assume from the 

 above that it ranges up the east side of the Rocky Mountains and 

 through them south of the International Border in unmodified 

 form while the intermountain birds, flemingi, as they distribute 

 northward from the boundary, become smaller and darker. It is 

 possible that on examination, birds from further north of Teslin 

 Lake will exhibit these characters in a still more marked degree. 



AN UNDESCRIBED GALAPAGOS RACE OF OCEANO- 

 DROMA CASTRO. 



BY JOHN TKEADWELL NICHOLS. 



In his Monograph of the Petrels (London 1907-10), Godman 

 says in regard to Oceanodroma castro (Harcourt) : " This species is- 

 a form of 0. leucorrhoa, with which it has often been confounded," 

 etc. To the writer this seems an undervaluation of the species 

 castro. He also thinks that too little emphasis has here been placed 

 on the hair-like black shafts in the white upper tail coverts of 

 Leach's Petrel, — otherwise an excellent comparison of castro and 

 leucorrhoa is given. 



The unfortunately small amount of material at hand for compari- 

 son indicates the existence of three recognizable races of castro.. 

 Pacific birds have a slightly more forked tail, and the Galapagos 



