450 
Birds of Celebes: Dicaeidae. 
Five males in the Sarasin Collection and one in the Dresden Museum have 
the wing 48 — 51 mm; two females in the former collection; wing 48. 
Distribution. Celebes — Minahassa (Wallace a I, a 11\ Meyer a 9, etc.); Gorontalo Distr. 
(Meyer a .9); Lembeh Id. and Banka Id. (Nat. Coll.); S. E. Peninsula — Kandari 
(Beccari a S); W. Celebes (Doberty 5); S. Celebes (P. & F. Sarasin 3, Everett 4, 
Doherty 5). 
This little Flower-pecker was discovered by Wallace in the mountains of 
the Minahassa, where it haunts low bushes and shrubs and feeds on fruit (al). 
Meyer, Guillemard, and Platen, like Wallace, met with it at an elevation 
of upwards of 2000 feet in the same neighbourhood; but the first-named also found 
it at Limhotto, which is low land, and the Sarasins and Everett found it at 
Macassar as well as on the hills of the south. It is most nearly related to 
P. sangirensis of Great Sangi, from which it may be distinguished by its yellow 
sides and flanks, not grey tinged with yellow, and by the brighter, yellower 
tint of its olive-green upper surface. 'Ihe only other known species which seems 
at all nearly related to them is Mr. Biittikofer’s A. amiae of Flores (1). 
A. vincens (Sclat.) of Ceylon is dull steel-blue above, where the forms of the 
Celebesian area are olive-green, and it has the three lateral tail-feathers tipped 
with white*. A. quadricoior of Panaon, Philippines, is black above with a red 
back and olive-yellow rump. 
The genus Acmonorhgnchus was set up by Mr. Oates (Faun. Brit. Ind. II 
1890, 381), for A. vincens of Ceylon, in consequence of its having 9 primaries, 
instead of 10 as in Prionochilus and Pachgglossa. Mr. Biittikofer shows (1) 
that A. quadricoior, A. aureolimbatus, A. sangirensis and A. annae belong to the 
same genus. It is most nearly allied to Piprisoma, w'hich also has 9 primaries 
and which Mr. Oates only distinguishes by its bill which “viewed from above 
is nearly an equilateral triangle with the two sides sinuated”, for the square tail 
and nostrils free from hairs are points not sufficient to separate it from the 
Celebesian species. Mr. Biittikofer is of opinion that Piprisoma modestum of 
Burmah and I'enasserim may really belong to Acmonorhgnchus, but, if this be 
true, Acmonorhgnchus as a genus would be broken down, and Acmonorhgnchus 
aureolimbatus (a large name for so small a ’bird!) would have to be called Pipri- 
soma aureolimbatum. 
There are one or two facts known in connection with Acmonorhgnchus and 
the allied genera, Piprisoma, Pachgglossa and Prionochilus, which seem to be 
instructive. Mr. Oates shows (1. c. 375) that in Prionochilus there is a small 
first primary about equal in length to the tarsus; in Pachgglossa it is reduced, 
being not longer than the hind toe; in Piprisoma and Acmonorhgnchus it is gone. 
As the normal number of primaries is ten, it follows that Piprisoma and Acmono- 
rhgnchus have lost a quill and represent, therefore, in respect of the wing a 
recent modification of structure. Further, though we are unable to speak of 
Piprisoma, Acmonorhgnchus seems to display recent modifications of colour. Ac- 
monorhgnchus vincens of Ceylon is so like Pachgglossa, of which only one species 
