of the B. M. Catalogue of Birds. 397 



seems to have had before him only a solitary example of the 

 genus from Madagascar ; but he had a pair of the New 

 Caledonian species. His description is unintelligible, unless 

 "outer," inlineCjbe aslipofthepen for ^^inner;" andspeaking 

 from a careful examination of eight specimens now before me, I 

 can say that it would be scarcely possible for the whitish ter- 

 minal tip on the outer central pair of tail-feathers to wear off 

 until the tail was worn to a stump. There is no authority for 

 Dr. Gadow^s statement that this bird is a native of the New 

 Hebrides. The fact is that it is a typical Myiolestes, a peculiar 

 Pacific-Ocean family, from which there is no pretext for dis- 

 sociating it ; it is intermediate between M. vitiensis and 

 M. maxima, while M. nigrogularis from Fiji makes a far 

 nearer approach to Xenopirostris. 



But it is to the classification of the Pachycephalinse that 

 we take most exception. We have certainly not to complain 

 here of races and subspecies, but of wholesale '^'^lumping^' 

 without any reason assigned, in defiance both of literature 

 and of specimens, even when the latter are in the British 

 Museum. There is no class of birds in which the variations 

 in colour are more interesting or important — not only every 

 Pacific group, but in many cases each island of a group 

 having its own localized form. Be they species or not, they 

 are recognizable, and the catalogue which does not notice 

 these modifications with a reference to their bearing on 

 geographical distribution is valueless. 



Under the first species given, Pachycephala melanura, we 

 find included P. macrorhyncha and P. clio. We should 

 hesitate to reject Salvadori's very clear diagnosis of P. ma- 

 crorhyncha because sometimes a few black feathers may 

 occur on the side of the neck. There may certainly be a 

 tendency to inosculate in all these closely allied forms. The 

 Sula and Bouru form, P. clio of Wallace, is rejected because 

 " the colour of the upper tail-coverts is of no specihc im- 

 portance.^^ But Wallace gives other distinctive characters, 

 size and broader pectoral band, which hold good in our 

 specimens ; besides which we find the black of the tail more 

 intense, and the rectrices edged with deep yellow instead of 



SER. v. VOL. II. 3fi 



