THE BIRDS OF AUSTRALIA. 
rufous than Indians or Tennaserim ones. . . . Thus we conclude that there 
are neither two nor three species of Kite in India, but two (or three) 
subspecies^ one of which proceeds to further racial variation in other 
localities, Ceylon, Celebes, Australia. . . . For the present it seems best to 
treat of M. migrans as consisting of four subspecies, viz. : 
The typical M. migrans . . Europe, N.W. Africa, Asia east to 
Tomsk, Afghanistan. 
M. m. melanotis T. and Sch. East Siberia, Japan, India, Burmah. 
M. m. govinda Sykes . . Afghanistan, India, Ceylon, Burmah 
to New Guinea. 
M. m. affinis Gould .. Australia. 
They characterized the last as ‘ size small : more rufous.’ ” 
The preceding will sufficiently indicate the interest that has been 
taken in these Kites, and a recent reviewer is Dr. Hartert, who has dealt 
with these in the Vogel palaarkt. Fauna in the part already published, 
but which is not in my hands. 
My own conclusion is that however many races may be admitted 
(and it is certain more than four wiU be), the Austrahan one is quite 
recognisable on account of its more rufous coloration and its very small 
size. It may be that with long series we should have to treat this species 
by the method I have indicated under Haliastur Indus. 
Thus the Celebes birds may be regarded as intermediate between the 
Burmese birds and the Australian ones ; the Ceylon ones intermediate 
between the Indian and the Australian birds ; the North Indian ones 
intermediate between the Japanese and Chinese birds and the Central Indian 
ones, and so on. 
Owing to this perplexing state of affairs and ignorant of Dr. Hartert’s 
published views, I prefer simply to cite the whole of the Australian birds 
under the subspecific affinis ; by no method of classification can affinis 
possibly be regarded as a species. Gould himself noted the very close 
alliance of affinis and migrans, though disputing their absolute identity. 
180 
