( 2r,r ) 



they vary a great deal in the latter and are by no means always round. In fact, 

 there are specimens of Ch. carbonaria before me in which they are narrower and 

 longer than in the specimens from J'ergusson Island. Therefore I have no hesitation 

 in considering Cli. propinqua merely a .synonym of Ch. carbonaria. This latter is 

 fairly distinct from Oh. atrocaerulea from the Moluccas, and must, I think, stand as 

 a species, tiot merely a .subspecies. On the other hand it seems .sometimes very 

 difficult to distinguish Ch. laemosticta .Scl. from New Britain and New Ireland, and 

 I am inclined to think that the latter should stand as a subspecies of Ch. carbonaria. 



I feel uneasy about the genus under which to classify these species. They have 

 often been included in Dicrurus, but .Sharpe unites them with Chibia and Salvadori 

 calls them Dicruropsis. I do not see much of generic characters in either of these 

 supposed genera. 



Nests, with three eggs each, were found in October and December. The eggs 

 are of two principal sorts of varieties. One has the shell pure white, without gloss, 

 covered with small deep purplish brown and pale purplish grey spots and dots- 

 These measure 29-6 : 21-3, 30 : 21, and 2S)-8 : 21-6 mm. The otlier is of a creamy 

 ground-colour, and spotted with larger patches of a kind of brownish brick-red and 

 the same pale i)ur])lish grey patches, liut mostly larger. They are a little shoi-ter, 

 measuring 29 : 22, 29 : 21-5, and 29-1 : 22 mm. E. H. 



12. Melilestes fergussonis sp. nov. 



Melilestes speciebus M. iliolopihus et M. affinis dictis similis, sed multo major. 

 Al. c? 71—72 mm., ? 63—64; culm, c? 25, ? 21 mm. 



Hub. Fergusson Island. 



This species closely resembles in colour i/. iliolophus and M. ajjinis, if the two 

 are more than subspecies of one species, and is probably only subspecifically distinct. 

 The differences in size, especially in the length of the bill, and the separate locality 

 whence we have received it, however, are remarkable. The iris is dark hazel ; bill 

 black ; about basal half of mandible whitish. 



The sexes differ in size, as they also do in M. iliolophus and M. affinis, liut it 

 seems that this was not noticed before. 



Gadow, Cat B. IX., has placed the above-named species in the genus Arachno- 

 thera, together with M. novaeguineae, while he allowed M. megarhijnchus to remain 

 among the MeHphatjidae and put it in the genus Ptilotis. Without wishing to 

 enter into a discussion on the genera of the Meliphagidue, which, I believe, are on 

 the whole di\ided very reasonably in the Catalogue of Birds, I cannot agree to that, 

 as I believe that M. megarhynchxts, 2f. uovaegaineae, ^^. iliolopihus, and allies are 

 all congeneric, and differ widely from Arachnothera in the form of the bill and 

 nostrils. On the other hand they differ from true rtiloiis, and one might for the 

 present accept without hesitation Salvadori's generic name Melilestes, with M. mega- 

 rhynchus as the " type," and including M. novaeguineae, iliolophus, affinis, fergus- 

 sonis, poliopierus, and probabl}- also Melilestes celebensis and subspecies (see antea, 

 p. 153), which latter is certainly not an Arachnothera. It differs also from the 

 Papuan Mdilestes in its short tarsus and toes, its naked ring round the eye, and a 

 narrower bill. Another question which I have often asked myself and which I cannot 

 answer to my own satisfaction is whether Arachnothera is in its right ]>lace among 

 the Kectariniidae, and whether it is not a lloney-lviter. Oates has already allowed 

 it the rank of a subfamily, 



A nest was found in December. It is liislencd to some leaves and a thin twig. 



