ORNITHOLOaX OP S.E. NEW GUINEA. 4S9 



grey, mottled witli irregular whitish cross bars ; rest of under 

 surface of body light tawny red, transversely barred with greyish 

 white, these white bars much broader on the abdomen and under 

 tail-coverts, which are white barred with rufous ; the sides of the 

 breast nearly uniform rufous : under wing-coverts and axillaries 

 pale tawny like the breast, and narrowly barred with white, 

 the outer coverts and the edge of the wing white with a few 

 rufous bars ; quills brown below, pale creamy rufous towards the 

 base of the feathers, and narrowly barred with dusky brown on 

 the inner webs. Total length 16 inches, culmen I'l, wing 10"4, 

 tail 7-7, tarsus 2-7. 



This Goshawk is not A. approximans, as it is so much clearer 

 grey, so much lighter below, and, moreover, has a broad red collar 

 round the hind neck. The latter feature separates it from A. 

 wallacii, which, moreover, has the throat and thighs reddish. 

 Its nearest ally proves, therefore, to be A. torquatus of Timor, 

 with which Count Salvadori now identities it ; but in addition to 

 the larger proportion of rufous on the under parts, the New- 

 Gruinea bird has the under tail-coverts barred with rufous, whereas 

 in A. torquatus they are entirely white, the thighs also being nearly 

 all white. These may, however, be merely individual differences. 



Count Salvadori has recorded a Groshawk from Yule Island, 

 which he is inclined to refer to A. cruentus (Gould). I think 

 he makes out a good case against my A. cruentus of the ' Cata- 

 logue of Birds ' (p. 127) ; for I fancy the bird described by me 

 must be only A. approximans -. this I have thought for some time, 

 as I have watched a specimen living in the Zoological Gardens 

 which got whiter below as it grew older till it much resem- 

 bled the specimen described. At the same time the Tule-Island 

 Goshfawk, if it is the same as the bird now brought by Mr. Stone, 

 can hardly be the true A. cruentus of Gould, according to the 

 figure of the hind neck ; but the description agrees in stating 

 that there is a chestnut collar round the latter. It is a great 

 drawback that Mr. Gould gives no measurements ; but sooner 

 than create a new species on certain slender differences, I shall 

 follow Count Salvadori in considering this to be the true A. cru- 

 entus until such time as an examination of the typical bird from 

 "West Australia proves us right or wrong. The measurements of 

 the present bird approach very closely to those given by Count 

 Salvadori. 



Dr. James procured a young Goshawk in South-east New 



LINN. JOUEN. ZOOLOGY, VOL. XIII. 40 



