[Auk 

 July 



Each genus was ordinarily represented by but a single species 

 in a place, so that the number of species and the number of 

 genera in any one locality were nearly equal. This resulted 

 from the fact that in a large number of genera the islands pos- 

 sessed but a single widely distributed species of each, and that 

 in a large number of other genera, though each genus existed in 

 the islands in several species, each of these species occupied a 

 limited area of its own made up of one or more adjacent islands. 

 In this limited area it existed by itself, sharply separated from 

 the other species of the same genus. 



This reappearance of genera in new specific forms in each 

 distinct area was so frequent with a large number of them that 

 the members of the party learned to expect and to look for local 

 species of these whenever a new island was reached. 



On making a study of the distribution of the genera and 

 species of birds collected by the party in these islands on our 

 return to the United States, the facts above noticed were found 

 to be further reaching and of more importance than was at first 

 supposed, and to point to a law of distribution which, if estab- 

 lished, must have great influence upon the theories for the crea- 

 tion of species. 



In this study only the collections made by the members of the 

 party have been used. Great care was taken with these in 

 noting the exact place of collection, sex, color of eyes, etc., 

 while with many collections formerly made from the islands it 

 was supposed sufficient to label them as from the Philippines. 



About five thousand specimens of birds were collected by the 

 party, belonging to nearly or quite four hundred species. These 

 were collected on seventeen distinct islands of the group, which 

 were chosen from their size and location as representative of the 

 whole. 



These collections, while not comprising all species known 

 from the islands, are so nearly complete that any just conclusion 

 drawn from their study must be accepted as truth, which further 

 exploration will only strengthen. 



The general arrangement followed has been that of Lord 

 Walden and of Prof. R. B. Sharpe, in their published lists of 

 Philippine birds. The names of genera and species have been 

 made to agree generally with those given in the 'Catalogue of 



