81 [Vol. xxxr. 



The Chairman : Before calling upon Mr. Mathews I 

 should like to say one thing and that is, the great trouble 

 has alwaj^s been the inconsistency of the various authors 

 in dealing with the various groups ; and when people find 

 fault with the 'dumping" in one place and the '^ splitting^' in 

 another, it is generally because in the one case the author 

 has arbitrarily used colour or colour-pattern, while in 

 another he has arbitrarily used structure of a kind which was 

 evidently not of generic value. I now call upon Mr. Mathews. 



Mr. G. M. Mathews : I am compelled to side with those 

 who maintain that colour must be utilised in the differentia- 

 tion of generic groups and am confident that this view will 

 latterly prevail universally. I say this with confidence as I 

 was first influenced by the view of the professed adherents 

 to the so-called " structural " school, and my first ^ List of 

 the Birds of Australia ' was prepared with that view as my 

 basis. During its preparation I was being continually 

 impressed with the inadequacy of the structure of a bird as 

 a clue to its generic affinity, and later a monographic study 

 of the Petrels compelled the rejection of that fallacy, as I 

 soon realised that even in the mind of those who counselled 

 the usage of structural characters alone, colour was often 

 the chief factor consulted. 



Study of colour evolution from the nestling to the adult, 

 and the recognition of colour-genera, would certainly obviate 

 many anomalies in the Australian avifauna — as is to be found, 

 for instance, in the genus Pachrjcephala of authors, — if it did 

 not altogether prevent them. The latter result would be 

 achieved, if careful study of the birds was undertaken, and 

 attempts to group them by means of colour were made at 

 the time of the introduction into the genus of each new 

 form. It should always be remembered that the available 

 "structural'-' parts of a bird-skin are, comparatively speaking, 

 trivial and unreliable, as these are more liable to variation 

 by wear and tear than is the colour-pattern of the feathering 

 of a bird. 



To me the great charm of Salvadori's work is that he 

 was a " splitter " on coloration, and it is very easy to 

 follow his work in the 'Catalogue of Birds' and in his 



