Species of Pacbycepliala. 95 



Pachycephala neglecta, as its position implies, is somewhat 

 intermediate between the three first and two hist-mentioned 

 species. In the two first the black pectoral collar is almost 

 obsolete ; in P. torquata and in P. vitiensis it is qnite 

 complete and very broad ; in P. neglecta it is not quite 

 complete ; and in P. kandavensis, though it is complete, it is 

 rather narrow. In the two first mentioned species the yellow 

 supraloral patch is very conspicuous ; in P. neglecta there 

 are sometimes traces of it ; but in P. torquata and in the 

 two last I have never found any trace of it. 



The males of these six species are easily determined by the 

 characters above given, but the females and immature of both 

 sexes are somewhat puzzling. The latter may be sorted into 

 two groups — those with uniform chestnut or dark buff under- 

 parts, and those with greyish or brownish-white underparts 

 streaked and barred with grey or brown. There are examples of 

 most species in the uniform plumage moulting direct into the 

 adult male plumage, but among the specimens of P. neglecta 

 in the British Museum there is one in which the streaked 

 and barred underparts appear to be moulting into the uni- 

 form yellow of the adult male plumage. Before the females 

 and young can be correctly diagnosed a much larger series 

 must be obtained. 



Pachycephala graeffii. 



The Viti-Levu Bush Shrike was discovered on Viti Levu 

 by Dr. Grafi'e (Hartlaub, Ibis, 1866, p. 172), and the types 

 were in the Godeffroy Museum in Hamburg. 



There is no example in the Layard Collection, but there 

 are two in the Tristram Collection procured on the Kewa 

 River in Viti Levu by Mr. Victor Williamson, and there are 

 three examples in the British Museum, originally sent by 

 Kleinschmidt to the Godeffroy Museum. 



The other four skins in the British Museum referred to 

 this species by Dr. Gadow obviously belong to other species : 

 d is an example of P. kandavensis, e ought to have been 

 recognized as distinct under the name of P. neglecta, whilst 

 / and g may fairly be regarded as also distinct under the 

 name of P. aurantiiventris, which I propose for them in this 



