( 55 ) 



NOTES ON PAPUAN BIRDS. 



By the HON. WALTER ROTHSCHILD, Ph.D., and ERNST HARTERT. 



(Plates II„ in.) 



INTRODUCTION. 



WE have, at the Tring Mnsenm, within the last years, received a number of 

 more or less important collections from different places in New Guinea and 

 from the adjacent islands, and we intend to work them ont in families, beginning 

 with the Pittirlai' and Parrots. In some cases we shall review the whole geinis, 

 when of special interest, and when enough material is to hand. Where we seem to 

 know enough of a genus we shall try to work out the actual affinities and relations 

 of the various forms to each other. In doing so we shall often be obliged to reduce to 

 subspeciKc rank many forms hitherto named binomially as " good species." When 

 certain allied forms replace each other geogra])hically, and when the characters of 

 any of them are indicated in any of the others, they must, in our opinion, be treated 

 as subspecies, even if all connecting links are not seen in our scanty material. 

 The largest material is generally very scanty in comparison to the wide area 

 inhabited by the birds and to the number of individuals in existence, and we 

 cannot, therefore, expect too much of what represents a species or subspecies in 

 our drawers. Also when the diiferences are no other than a larger bill, or slightly 

 larger size, snch forms must be called by three names, even if the differences 

 are not bridged over. We often find in ornithological works notes saying that 

 specimens from, say the Aru Islands, ar^ smaller (wing 3 or 4 mm. shorter) than 

 those from, say New Guinea, but that they do not constitute a separate species. 

 On the other hand, nobody hesitated to separate specifically a bird which is twice 

 the size of another. But this is theorising most arliitrarily. Our principal object 

 is, or at least should be, to recognise the differences seen iu individuals from 

 varions localities, uot to attach names to the specimens before us I If, however, 

 we find that the specimens from a certain country differ from those of another 

 country, we must give them a name, in order to be able to speak about them and 

 to call attention to their existence. It does not in the least matter whether we 

 are able to classify one or two single individnals, as long as we discover the fact 

 that there are two fairly constant forms of a certain species. It is of much greater 

 advantage to show the affinity of the birds of a genus by uniting into species 

 the closely allied representative forms — for example, to have only 9 species with 

 19 subspecies of the genus Cyclopsitta, than to have 19 species. The mere list of 

 names shows that there are 19 groups of allied forms, while those in one gronp 

 are more or less alike; but the list of 19 binomially named " species " gives no 

 idea of their affinities at all. We believe, therefore, that our nomenclature will 

 advance our knowledge, tliough it will not be accepted or applauded by many of 

 our ornithological friends. 



The maps of distribution will not only show our present knowledge, but, 

 what is more important, what a vast area of New Guinea is still unexplored. 



