( 122 ) 



from Savn are darker above than thoso from Flores, Djampea, and Kalao; tbose 

 from Timor and Alor looking: above almost like a Kestrel, the feathers rusty red 

 with broad blackish cross-markings, those from Savu being darker, the blackish 

 brown on the back more extended, so that the feathers might be described as 

 blackish brown with rufous edges. These latter are very much like young Flores 

 birds; but all those from Flores, and still more so those from Djampea and Kalao, 

 are paler above. It is, however, evident that the latter are in older plumage and 

 more or less abraded and faded, being before their moult, and I tliiuk that all their 

 differences can be accounted for in that way, and that they are not of snbspecific or 

 specific value, especially as one of the young Savn birds agrees entirely with those 

 from Djampea and not with the other Savu birds. 



Unfortunately we did not receive any specimens from Lombok, but the British 

 iluseum possesses one, the type of Sharj)e's Astiir wallacei. This is still more 

 rufous than the more western forms before me (from Flores), but does not differ 

 otherwise. It has been described as a male, but is doubtless a female. The young 

 bird from Lombok, described by Dr. Sharjjc in a footnote on p. 128 of Vol. 1. of the 

 Cat. B. Brit, ^fu.■i., seems certainly to Ixdong to the old bird from that place, while 

 the Buru specimens will be found to belong to another form. As it is, we must, I 

 think, distinguish between, at least, an eastern and a western siibspecies, or perhaps 

 three, namely: — 



1. A very rufous form (Lombok). 



2. A less rufous darker form (Flores, Djampea, Kalao). 



3. A less rufous paler form (Timor, Savu). 



I think, however, that possibly the Flores, Djampea, and Kalao birds may 

 have to be united with the Lombok form, and if this my surmise is correct, the 

 latter would jirobalily have to be named A. tortjuatu.s wullocei (Sharpe). More 

 material of adult birds from Lombok and Sambawa is required to solve this 

 jiroblem. 



With regard to other allied forms I wish to offer the following remarks: — 



In my first article on the birds collected in Sumba by Mr. Dolierty, I mentioned 

 ■A ilawk from Sumba of this group as Astur torqnatu.'i. Afterwards 1 found that 

 Dr. A. B. Meyer had described an Urospizias sumbaiiims from that island, and 

 therefore I referred (Nov. Zool. IV. p. 270) my bird to U. sumhai'nsis. I have now 

 several young birds and a fine old pair sent by Mr. Everett. The young ones are 

 exactly like tliose from Flores, Djampea, and Kalao. The old birds agree perfectly 

 with the majority of the Flores birds, except that the cross-bars below are more 

 greyish and less reddish. As this character varies to a certain extent in specimens 

 from the same places, it must remain doubtful whether it is here a specific or an 

 individual character. The characters stated by me as distinguishing it from the 

 true A. torquatus are useless, as they vary in other individuals, and Dr. Meyer's 

 characters have to be disregarded for the purpose, as he compared it with a more 

 distinct ally. I expect that the Sumba bird can either be united with the Flores 

 one, or may be subspecifically distinct, but nothing more. 



I cannot bring myself to unite with ,1. torqitattm the Australian A. rruentii.t, 

 which seems to differ in being more regularly barred below and in having a longer 

 tarsus. The birds from New Guinea are probably not separable from the latter. 



The change from the longitudinally marked plumage of the immature bird to 

 the cross-barred jthmiage of the aihilt bird is effected through a moult, as shown by 

 some of our skins. 



