( 4(i7 ) 



4. Cisticola cisticola (Tumm.) (HI. i). Tt-sl). 

 SeptemljL'r, Wiiiugapn. 



5. Parus atriceps Horsf. (III. p. osJ). 

 September, Waingapo. 



6. Dicaeum wilhelminae Biittik. (III. )i. ."l^l). 



Males and females from ^Vaiiigupo and elsewhere. I Lave (/.'•.) recorded a 

 female from Snmba, which looked like the female of D. mackloti, and of which I 

 conld not suppose that it was the female of D. toilhelminae, as it had a red rump, 

 while the male of T). wilhehninae has no red on the rump. It is, however, the 

 female of D. tmlhelminae, and differs from that of D. mackloti in having a shorter 

 wing, more greyish sides of the breast, and more olive greenish flanks. It is peculiar 

 that the female has a red rump like D. mackloti, while it is absent in the male. 



7. Piprisoma obsoletum (S. Mlill.) (III. p. .581, no. 8). 



A series from Sumba has the wing generallj- about one or two mm. shorter 

 than Timor specimens, but some are quite equal in size to Timor specimens. Other- 

 wise there seem to be no perceptible differences between the specimens from the two 

 islands. 



8. Anthi-eptes malaccensis clilorigaster (Sharpe) (III. p. 580). 



The distribution o'^ Antkreptes malaccensis and allies in the Eastern Archipelago 



is one of the most difficult and interesting chapters to the ornithogeographer. 



When describing Doherty's collections from the Sunda Islands, I called (Vol. III. 



p. 5-15) the Bali bird A. xnalatcensis, separating from it that of Sambawa, which 



I called A. malaccensis chloriyaster, and that of Sumba, which I identified with 



^1. malaccensis celebensis. This has been commented on by Messrs. Meyer & 



Wiglesworth in their great work on the Birds of Celebes, and it is now necessary 



to state that I have been perfectly right in differentiating between the Bali and 



Sambawa birds, the former being bright yellow below, and not, I think, separable 



from xL malaccensis nudaccensis, while the Sambawa males difler at a glance by 



their much more greenish and duller underside. On the other hand, I have erred 



in uniting the Sumba bird with A. mal. celebensis. With much more and better 



material before me from Sumba, Sambawa, and Celebes, it is quite clear that the 



birds from Suml)a, Sambawa, and elsewhere are exactly the same, and less dark 



greenish below than A. mal. celebensis. The question now arises, how to designate 



these birds ? With a large material before me, I find that the birds of the Southern 



Philippines, including the Sulu birds, are true A. mal. chlorigaster, and cannot, 



in my opinion, be separated with satisfaction from those of Sambawa, Sumba 



Flores, Lomblen, Pantar, and Alor. Between the Philip])ine area and that of the 



Lesser Sunda Islands, however, we have the large i.slaud of Borneo ou the one liaud, 



where we find typical A. malaccensis, which is spread over the Malay Peninsida, 



Sumatra, the Natunas, Java, and Bali, and the Celebensian area on the other hand, 



which would very well connect tlie Pliilippines with the Lesser Sunda Islands, but 



we find on Celebes a still darker form, ^I. mal. celebensis! It may be added that 



I'alawan birds (cf. Shelley's Monograph of Sunbirds) are ^i. malaccensis and not 



32 



