the Study of the Drepanididse. 575 



Tims Oreomijza is represented hj a distinct species on each 

 of six islands^ as are also Hemignathus, Heterorhynchus, and 

 Loxops on each of four. On the other hand, Himatione, 

 Vestiaria, and Psittacirostra range over the whole main 

 group, each with a single unmodified species. Chloro- 

 drepanis occupies an intermediate position with two very- 

 distinct forms on Kauai, and another form, sometimes con- 

 sidered divisible into three or more species, ranging over the 

 remaining islands. Of these, however, the distinguishing 

 characters are so slight that it is questionable whether they 

 are worthy even of subspecific rank, and in any case such 

 characters are by no means to be considered equivalent to 

 those which separate the different species of Oreomyza. 

 The latter are clearly the results of isolation, one island 

 having been colonized by a species from another, which 

 has subsequently acquired peculiar characters. It might be 

 supposed that the birds in the other class which shew no 

 change on the various islands are in some way less susceptible 

 to the effects of isolation. Probably this is not the case, and 

 the true explanation is to be found in considering the habits 

 of the members of the different genera. 



Himatione, Vestiaria, and Psittacii'ostra are all birds 

 which take extensive flights, often at a great height in the 

 air, and frequently form small companies in these flights. 

 If we stand on the main ridge of some of the islands the 

 birds may be seen passing high overhead from leeward to 

 windward or vice versa. All freely traverse open country, 

 in passing from one feeding-ground to another. Consequently 

 when storms arise they are extremely likely to be carried 

 across the channels between the islands, and no doubt this 

 often happens. The birds of the other class, such as Hemi- 

 gnathus, Oreomyza, Heterorhynchus, &c., do not take these 

 extensive flights, but keep closely to the forest, very rarely — 

 and most of them never — venturing into the open. Very 

 seldom would they be likely to get blown across from one island 

 to another. In short there is little doubt but that individuals 

 of Vestiaria and its class are transferred from one island to 

 another sufficiently often to prevent any true isolation, which 



2q2 



