444 Mr. D. A. Bannerman on the Distribution of 



isolated in often widely distant localities. Intermediate 

 colonies may be totally wiped out, for it has often been 

 proved that all the Tubinares have a highly-developed 

 hominoj-sense and become strongly attached to the particular 

 breeding-station to which they resort. The birds will return 

 year after year to the same small island no matter to what 

 extent they are subjected to persecution from man, rats, 

 mice, mongooses^ or the other innumerable enemies with 

 which ground-nestirjg birds have to contend. Padre Schmitz" 

 records an unusual agent of destruction when^, in a certain 

 year, all the young P. k. flavirostris were found dried up in 

 their holes on the Salvage Islands suffocated by the sirocco. 



In addition to B. bulweri, we have the interesting case 

 of Oceanodroma castro, which breeds in the north Atlantic 

 Islands and almost certainly in the Hawaiian and Gralapagos 

 Groups in the Pacific. Birds from the Atlantic and Pacific 

 are quite indistinguishable from one another, although there 

 are hundreds of miles between them and no intermediate 

 colonies exist. Both these cases point to a discontinuous 

 distribution. 



A different example is afforded by the White-faced or 

 Frigate- Petrel. The Frigate -Petrel of the north-east 

 Atlantic has become so differentiated from the typical race 

 as to be readily distinguished by the longer bill and the 

 lighter colouring of the upper parts, and should be known 

 as P. m. hypoleuca. 



Two noteworthy examples of curious disti'ibution are the 

 Little Shearwaters, Puffinus assimilis ba7'oli, inhabiting the 

 Azores, Madeira, and Canaries, and Puffinus Iherminieri 

 boydi, which is confined to the Cape Verde Group. These 

 are dealt with more fully on pages 477 and 483. 



It is unfortunate in determining geographical races that 

 it is impossible to fix the parent race. This must not be 

 confused with the typical species, which, as everyone knows, 

 is the term applied to the first member of the species known 

 to have received a name. Necessarily this is not always the 

 most ancient species from which other so-called geographical 

 races have sprung. 



