Mr. P, L. Sclater on Macgregor's Paradise-bird. 415 



named it Xanthomelus macgregori (Ibis, 1890, p. 153), be- 

 lieving it to be most nearly allied to Xanthomelus aureus. 

 Mr. De Vis [l. s. c.) calls the species a " very distinct kind of 

 Bower-bird,^' and states his opinion that its proper syste- 

 matic position is between Amblyornis and Xanthomelus. I 

 will therefore compare it with the latter form, of which a 

 fine example belonging to the Leyden Museum has been 

 kindly lent to me by Dr. Jentink. 



There is certainly a general resemblance in colour and 

 shape between Cnemophilus and Xanthomelus, and the feet in 

 both forms are large and strong, although this feature is 

 carried to a much greater extent in Xanthomelus, which has 

 the tarsi much stronger and rather longer than Cnemophilus. 

 In Xanthomelus, moreover, the scutellations of the front of 

 the tarsus are well marked, whereas in Cnemophilus the 

 scutella are fused into one nearly uniform plate. The wings 

 of Cnemophilus are much shorter and more rounded than 

 those of Xanthomelus. But it is in the bill of these two 

 forms that the greatest divergence is observable. 



In Xanthomelus the bill is long and strong, the loral 

 plumes are short, and the base of the bill, nostrils, and 

 culminal ridge are quite bare. In Cnemophilus the bill is 

 •shorter and not so thick, the loral plumes are elongated, pro- 

 jecting forwards, and covering the base of the bill so far as to 

 partially cover the nostrils. Besides this the frontal plumes 

 are elongated and elevated into a compressed ridge, which is 

 carried forward over the culmen and backward to the base 

 of the very singular thin crest, composed of five or six 

 lengthened feathers, which springs up immediately behind 

 the front. 



In these last characters Cnemophilus is quite distinct from 

 other birds, but obviously approaches Diphyllodes. I should 

 be disposed, therefore, to place Cnemophilus along with the 

 Paradise-birds rather than along with the Bower-birds, if 

 these two groups are to be kept apart. But there can be 

 no doubt that the Bower-birds are closely allied to the Para- 

 dise-birds, and several well-known recent authorities have 

 united them into one family. 



2f2 



