398 Canon Tristram on the Eighth Volume 



silver grey. We presume that the writer must have ex- 

 amined these birds by gas- or candle-light. 



We observe throughout the volume that every species 

 described by Mr. E. P. Ramsay is invariably accepted with- 

 out hesitation or question, and consequently the one sub- 

 species admitted in this genus is P. kandavensis, the descrip- 

 tion being copied, but no specimen examined ; for the three 

 specimens in the British Museum from Kandavu, Dr. Gadow 

 very rightly relegates to P. vitieusis. It seems impossible, 

 from the description, to distinguish the two. But unfortu- 

 nately Mr. Layard^s observations have been quite overlooked ; 

 they point out that P. vitiensis is the special form of 

 Kandavu, and it is scarcely probable that a second very 

 closely allied, if not indistinguishable, race should occur in 

 the same very small island. Next, in describing P. gutturalis, 

 the author says, " I have copied Ramsay^s description of 

 P. occidentalis, although the specimens from Western Aus- 

 tralia in the Museum do not agree with his diagnosis " ! 

 We can only add, neither do our own three from Western 

 Australia. Yet the next species is P. occidentalis, without a 

 specimen to justify it. On P. littayei, adult female, is the 

 remark, ^'^ Canon Tristram says 'the female is little less 

 brilliant in colour than her mate.' " Then the writer pro- 

 ceeds to state, as if in contradiction to this statement, the 

 differences, which are taken word for word from the descrip- 

 tion by Tristram. This seems scarcely a fair way of treating 

 a describer. But P. astrolabi is perhaps one of the least 

 excusable of all the mistakes of the volume. Having assumed 

 that P. chi'istojjhori is identical with P. astrolabi, Dr. Gadow 

 proceeds, " The diagnosis given by Tristram for his P. chris- 

 tophori {^) agrees with an immature male of P. astrolabi." 

 But just above he has described the young male, stating it 

 is one of the types of P. christojjhori. We have seldom met 

 with a more amusing instance of begging the question. The 

 very measurements given might have created some suspicion ; 

 they are 7'7, 0-9, 4-2, 3, and 1, as against 6, 0-75, 3-4, 2*5, 

 1. More correctly the last three measurements should have 

 been 3-1, 2*4, and 09. Not only is there this material 



