1918.] The Plat ycercine Parrots of Australia. 115 



VI. — The Platycercine Parrots of Australia : a Study in 

 Colour-change. By Gregory M. Mathews. 



(Plate III. and Text-figures 1, 2.) 



The Broad-tailed Parrots of Australia, ranged under the 

 genera Platycercus and Barnardius, provide an interesting 

 study in colour-evolution, and close investigation at some 

 later period may lead to valuable conclusions as to the rate 

 and reason of colour-changes. It was necessary to examine 

 the group as a whole and also individually for my ' Birds of 

 Australia/ and in that work I have proposed some theories 

 to account for the facts. I propose to give a resume of my 

 remarks in this place, as this Journal will circulate among a 

 larger and more varied number of readers. I hope that the 

 matter here provided may interest, and perhaps suggest 

 similar criticism of other groups. By means of the co- 

 operation and consideration of independent ideas we may be 

 able to determine some of the factors in bird-evolution, but 

 the complexity of these factors is well shown in the present 

 subject. 



In the first place, the Broad-tails are divisible by means of 

 colour-pattern into two genera, Platycercus and Barnardius. 

 liy acceptance of structural characters alone the two are in- 

 admissible. I hope to prove that the former solution agrees 

 best with the facts. The first genus, Platycercus, is charac- 

 terized by having the feathers of the back bicoloured, a broad 

 edging giving the well-known scalloped appearance to tliat 

 feature. The species of the second genus, Barnardius, have 

 the feathers of the back unicoloured. As the young of the 

 first genus show the scalloping to a greater or less extent, 

 it. is obviously an ancient feature. As it is missing in the 

 immature of the second genus, we have two stocks divei-gent 

 at an early period of their existence, though at the present 

 time their structural characters are to all intents identical. 

 Newer and more accurate modes of examination may 

 perhaps discount this last statement. 



In the genus Platycercus I admit seven species, and in 



I 2 



