1G COLLECTIONS FR03I MELANESIA. 



a. Upper tail-coverts olive-yellow like the back ; 



outer aspect of secondaries distinctly grey; 



pectoral collar joined to ear-coverts melanura. 



b. Upper tail-coverts black ; outer aspect of secon- 



daries olive-yellow, with which they are mar- 

 gined or washed externally. 



a'. Pectoral collar joined to ear-coverts clio. 



b'. Pectoral collar separated from ear-coverts. 

 a". Upper surface olive-greenish ; wings ex- 

 ternally washed with greenish grey .... macrorhyncha. 

 b". Upper surface golden olive ; wings ex- 

 ternally washed with the same colour . . obiensis. 



After having gone over the series in the British Museum, which 

 likewise served as the basis of Dr. Gadow's studies, I regret that I 

 must entirely disagree with him. He appears to me to have argued 

 from immature specimens when he tries to show the variation of 

 the species and attempts to prove that they run one into another. 

 If it were possible to find in the same island examples of these Pa- 

 chycephalm with the black breast-band united to the ear-coverts 

 and others with this disunited, all being fully adult birds, then Dr, 

 Gadow would have proved his point ; but this is exactly what does 

 not take place, the difference in coloration being accompanied by a 

 different habitat. It is not right to compare immature birds of one 

 form with adults of another, because in their young stages all these 

 species are unquestionably very difficult to distinguish apart ; but if 

 fully adult birds are compared, I do not think there ought to be 

 any difficulty in determining four distinct species. 



Again, with regard to his observations that the colouring of the 

 upper tail-coverts " is of no specific importance," some attention 

 must be given to the age and condition of the specimens. In 

 the group with the upper tail-coverts black, it will be found that 

 the basal ones are always more or less tipped with olive, and that 

 it is the long ones which are black. If, therefore, • the latter are 

 shot away, a superficial observer would jump at once to the conclu- 

 sion that the upper tail-coverts are olive-yellow. Even in this case 

 the yellow-washed quills (instead of grey) would enable one to dis- 

 tinguish the P.-macrorhynclia group from P. melanura. Then, 

 again, it would appear from moulting specimens that the upper tail- 

 coverts when first grown have more olive-yellow on the margins 

 than in the fully adult bird. Therefore in these PacJu/cephala', as 

 in other birds, only fully plumaged specimens should be compared, 

 if one is to understand the relations of the various species. 



There can be no doubt that the locality " Celebes " attached to 

 the specimen from the Gould Collection is erroneous ; and Mr. 

 Wallace has remarked on the absence of the genus in the above 

 island, in the essay quoted by Dr. Gadow. 



