THE BIRDS OF AUSTRALIA. 
would tend to accuracy were I to fix absolutely some particular place. 
I propose, therefore, that Sorong be considered the restricted locality for Psittacus 
griseus Bechstein, Psittacus goliath Kuhl and Microglossus aterrimus var. major 
Salvadori. As a fourth synonym, I doubtfully cite Microglossus salvadori Meyer, 
the description of which reads : “ Foem ? Forma M. atcrrimi Gm., sed 
prhnulaceo-flavus : subtus anguste, supra late nigro fascia tiis ; alis et cauda 
reliqua plus minusve nigro irroratis ; capite cristaque vividi-atris, irregulariter 
fasciatim pallidissime flavo signatis et maculatis ; maxilla fusca, apice 
pallido, mandibula albiscenti ; pedibus nigris. Long. tot. c. 550 millim., 
al. 370, caud. 275, crist. 130, rostr. culm. c. 75. max. alt. 40, mand. alt. 36, 
max. lat. ad bas 26, mand. lat. ad bas 36, tars. 20.” 
From a consideration of the preceding it is somewhat obvious that 
more than three subspecies are indicated, and it is probable that many will 
be hereafter allowed. The configuration of New Guinea suggests confirmation 
of this surmise, and more especially our knowledge of New Guinea avian 
forms. The extent of New Guinea when viewed in comparison with the 
collections available prohibits lumping such as has been so casually employed. 
Thus, I would accept six subspecies at the present time in the following 
manner. 
Probosciger aterrimus aterrimus (Gmelin). 
Type-locality : Salwatty. 
Western Papuan Islands, Salwatty, Waigiou, Misol, Gemein. 
Stresemann has selected for alecto Lesson the restricted type-locality 
Waigiou, so that if Waigiou birds were separable from those of Salwatty 
they would be called Probosciger aterrimus alecto (Lesson). Further, Van 
Oort has stated that Misol birds appear smaller than those from Salwatty, 
Waigiou and Gemein. These birds are small throughout. 
Probosciger aterrimus intermedins (Schlegel). 
Type-locality : Aru Islands. 
The race inhabiting the Aru group is smaller than the preceding, as can 
be seen from a criticism of the numerous measurements already cited. It 
would be of no value for me to write down a further long list of measurements, 
as no one accepts the figures prepared by any other worker. The only 
method of sure working is to make your own calculations and confirm or 
disagree with what has been already done. In the present case it is obvious 
that different methods of measurement have been pursued ; as an instance the 
culmen of Aru birds given by Salvadori as S3 120-100, $$ 90-85, while Van 
Oort writes, culmen dd 69-95 mm., $$ 73-92 mm., and so on. Some workers 
have given themselves the trouble to criticise such figures, but this is useless, 
92 
