ON THE ORIGIN AND DISTRIBUTION 



OF THE 



HAWAIIAN AVIFAUNA. 



There is in the Sandwich Islands evidently an older and a more recent avian element. We 

 may safely assume that the most modified forms which are restricted in their distribution to 

 the Hawaiian Group, and form now the so-called "Vrepamclce " or « Drepanidida" are the 

 descendants of the oldest bird-inhabitants. The next in age seem to be the Meliphagida, 

 L e. the genera Moho and Chtetoptila, and the Flycatchers (Chasiempis) ; while those species 

 which are identical with or nearest and still very little removed from American forms are 

 the most recent arrivals. As such close allies of American forms we have to regard 

 Eimantopus knudseni, Fulica alai, Gallinula sandioichemis, and apparently also the Buteo, 

 while Nycticorax nycticorax ncevius, though resident, cannot be separated from the 

 American form, and a constant immigration from America is now going on every 

 autumn (see, for example, Henshaw's remarks under Larus glaucescens, p. 286). 



The limits of the family Drepanida have been very uncertain, and their origin is by 

 no means clear. Dr. Gadow, in 1891, allowed only the genera Drepanis, Vestiaria, 

 Mmatione, Chlorodrepanis, Zoxops, Oreomyza, Remignathus, and Reterorhynchus to belong 

 to that family. Psittirostra (which had been recognized as Brepanine by Cabanis 

 and Sclater) and Loxioides he rejected to the Fringillidce, their tongue being "truly 

 Fringilliner as well as their alimentary canal and the rest of their internal and external 

 characters. Recently (1899) Dr. Gadow believes that all the Finch-like birds of the 

 Hawaiian Islands, i. e., not only Psittirostra and Loxioides, but also CMoridops, Bhoda- 

 canthis (and I may add Telespiza), and Pseudonestor, are Drepanidce. It is thus evident 

 that his former assertions of the Fringilline structure were less valuable than the mere 

 external knowledge of the systematical skin-ornithologist, who, by the wonderful transition 

 (through Pseudonestor\) from the bill of Remignathus to that of Psittirostra is bound 

 to conclude that these forms belong to the same -family." It is, on the other hand, 

 not necessary for the ordinary skin-ornithologist to connect Rhodacanthis and CMoridops 

 with these forms, nor does Dr. Gadow give any reasons for his belief. On the contrary, he 



