( 435 ) 



of a rufous tinge. One has the tinder tail-coverts grey with rufous tips and bars, 

 the other mfons but grey towards the base. 



This seems to be a fairly distinct subspecies of assimilis. In onr large series 

 of assimilis, as enumerated .above, there are only two or three specimens which are 

 uniformly grey underneath, and in those the grey is paler than in the ivebsteri. 

 Moreover, the under tail-coverts in the former are brighter and always rufous. 

 It would be a very strange thing if all three adult New Hanover specimens were 

 aberrations ! Moreover, the young specimen from New Hanover is darker than 

 any of onr young G. a. assimilis. 



So far we are quite at ease ; but there are also Cacomantis of this group 

 on New Ireland, New Britain, and on the d'Entrecasteanx Islands. About these 

 we are not at all certain. Years ago we have discussed the specimens from 

 Fergusson Islands, and we still find that tbey have longer bills than nearly 

 all assimilis from other places. The few skins before us from New Ireland 

 (three adult ones) are grey underneath, almost without a rufous wash, and their 

 tails are very long. This has already been pointed out by Finsch. Their under 

 tail-coverts are uniform cinnamon rufous. One from Duke of York Island (coll. 

 by Kleinschrnidt) is more rnfons underneath. Few of the tails are complete, 

 and we consider this material too meagre to deduct any conclusions from. The 

 questions are : 



1. Are the birds from New Ireland, Duke of York, and New Britain the 

 same as C. a. websteri? 



2. Are they a closely allied form, different from websteri, and still more 

 so from C. a. assimilis ? 



3. Is the form from the d'Entrecasteanx Islands another uew subspecies, standing 

 somewhat between websteri and assimilis, differing from the former by its larger 

 bill and mostly more rufous undersurface, from the latter by its larger bill ? 



We are inclined to think that the Fergusson form requires a name, and that 

 the one from New Ireland, etc., may perhaps be united with websteri. 



Cacomantis castaneiventris Gould. 



Cacomantis castaneiventris Gould, Ann. & Mag. J\'at. Hist. (3), xx. p. 2G9 (18G7 : Cape York, 



Australia). 

 Cacomantis infauslus (nou Cabanis & Heine !) Finsch, Notes Leyden Museum, rxii. p. 81. 



1 <J ad. Cooktown, May 1900. Olive coll. (No. A 75.) 



1 ad. Cooktown, without date. Olive coll. 



1 ad. Mountains of British New Guinea, no exact locality. E. Weiske coll. 



1 ad. " Between Rivers Laroki aud Vauapa," Owen Stanley Mountains. 

 E. 'Weiske coll. 



1 " ? " ad. Kotoi district, 4000 ft., 13. viii. 1898. A. S. Anthony col!. 



3 c? ad. Avera, Upper Aroa River, January, February 1903. (Nos. A 177, 207, 

 313.) A. S. Meek coll. 



1 cT, 2 ¥ ¥ , 1 unsexed, all ad., 1 £ juv., Owgarra, Angabnnga River, November 

 and December 1904, January 1905. (Nos. A 1874, 1884, 1916, 1977, 2105.) A. S. 

 Meek coll. 



3 cJ, 2 ¥ ad., 1 S juv. Bihagi, head of Mambare River, north side of Owen 

 Stanley Mountains, February to April 1900. (Nos. A 2379, 2402, 2537, 2614, 2665, 

 2727.) A. S. Meek coll. 



