75 [Vol. XXXV. 



the structure of the beak. The group which originally 

 was included under the genus Himatione consists of two 

 very large sections, one with a curved beak more like the 

 original Drej)anine bill and a series of birds with a 

 perfectly straight bill. These were originally separated 

 into two genera^ Himatione and Oreomyza. It has since 

 been proved that Oreomyza cannot be used^ as it was 

 employed for something else, and therefore the name 

 Oreoniystes has been coined for these straight-billed birds. 

 In both Himatione, the curved-billed group^ and OreomysteSi, 

 the straight-billed group^ each includes one red species and 

 a number of green species— the difference is this : in the 

 the curved-billed group the red species has developed quite 

 a different type of feathering on the throat and on the head 

 and neck, while the straight-billed red bird has developed 

 no difference except that of colour. 



In this case I quite agree with Mr. Perkins, the explorer 

 of the Sandwich Islands, in separating the curved-billed 

 green birds from the curved-billed red birds under the name 

 of Chlorodrepanis ; but I do not agree with those who wish 

 to separate the straight-billed red bird from Oreomystes^ 

 because 1 cannot find any other character besides that 

 of the scarlet colour by which to separate it. However^ 

 if I may be permitted to say so, I cannot agree that the 

 question of genera is of such importance from the point 

 of view of ornithology as many of us would make out. 

 I quite admit that Species as we see them to-day are the 

 work of evolution and of Nature, but the idea of Genera 

 is a purely human invention, like that of nomenclature 

 generally, to facilitate the study of birds and the communi- 

 cation of our ideas about them with our fellow-students. 



The classification ought to be carried out as far as 

 possible on phylogenetic and evolutionary lines, as sug- 

 gested by Mr. Pycraft, but I think genera ought to be 

 subordinated to usefulness, as they were originally intended 

 to be, and therefore multiplication of genera on characters 

 of very small value should be very highly deprecated — in 

 fact, I think what the late Dr. P. L. Sclater said is 



