OF THE ANDAMAN ISLANDS. 281 
a narrow central throat stripe appears; most of the feathers of the 
breast and abdomen exhibit large (though not solarge as in the fully 
adult), somewhat wedge-shaped, terminal brown shaft spots, and 
the tibize and vent feathers acquire a fulvous brown tint. In this 
stage the whole lower surface is sometimes tinged with a dirty 
pale fulvous brown. The third stage, that of the fully adult, has 
been already described in detail, Vol. II, page 142. 
I do not exactly understand the changes that take place in 
the barrings of the tail. In the quite young bird, above describ- 
ed, there are six bars besides the sub-terminal one ; in the two 
in Colonel Tytler’s collection which I examined, aud which were 
also, as I now know, young and not adult birds, there were four 
bars only, besides the sub-terminal one. In the old adult described, 
Vol. II, page 142, there are five. In the specimen described by 
Captain Beavan there were apparently six, besides the terminal 
one. As arule in the Spizaeti, the bars in the young are more 
numerous and more or less equidistant; as the bird grows older, 
some of the bars get absorbed, the terminal bar becomes much 
broader than the rest, the succeeding interspace becomes broader 
than any of the other interspaces, and the remaining bars 
often reduced to half the number that exist in the young bird, 
also grow broader and more pronounced. But then, as in many 
other of the Raptores, the changes of plumage of the lower 
surface and of the tail are not always synchronous, and speci- 
mens will be found with a perfectly adult tail which have not 
yet assumed the perfect adult plumage below, and again others in 
which the chin, throat, and whole lower parts, are those of the 
perfect adult, while the tail has not yet completed its changes. 
39 quat.—Spilornis Davisoni, Hume. 
This species appears to be quite as common at the Anda- 
mans as Lilginit. I have examined now fully fifty specimens, and 
all I previously said of this species (Vol. II, pp. 147-8) is correct. 
Call it a local race or sub-species, or species, its coarse stout 
tarsi, with their coarse conspicuous scutation, and wings only 
varying from about 14 to 154 inches, separate it at once from the 
other races of cheela, melanotis, Jerd., and Rutherfordi, Swinh., 
of both of which as of Davisoni, I have now very large series. 
As J said long ago, this race seems to me to come nearest 
pallidus, Wald., but it differs (at rate if Mr. Sharpe’s figure and 
description, Cat. I, p. 290, pl. IX, are correct) from this latter 
in two important particulars at least. In the first place instead 
of having a comparatively short though full crest, it has a very 
long one, completely covering when not raised the whole back 
of the nape; in a fine male, measured from the forehead where 
it may be said to commence, it extends 5-2 inches backwards. 
In the second place the tail is differently marked. In padlidus the 
tail is all dark brown, tipped of course paler and with one broad 
