THE BIRDS OF AUSTRALIA. 
third year the plumage of the old ones, and that the plumage of the 
middle age is so different from both, I consider as yet but as a secondary 
character, which, however, may contribute to make us consider this little 
group to be perfectly characterised.” The details in support of his sub- 
divisions show how closely Kaup examined his birds, and it may be that 
his conclusions will yet be proved to be correct. 
Sharpe, in the Catalogue of the Birds in the British Museum, VoL I.^ 1874, 
lumped Kaup’s subdivisions without explanation and gave diagnostic remarks 
for the recognition of Circus as follows : 
“Outer toe only connected to middle toe by interdigital membrane. 
Tibia and tarsus to all intents equal in length, the difference between 
them not so great as length of hind claw. Accipitrinse. 
Sides of face feathered, the lores always furnished with bristles. 
Hinder aspect of tarsus reticulate. Nostrils oval, with no bony excres- 
cence. Circus. 
“Hinder aspect of tarsus scaled. Nostrils oval with no bony tubercle. 
Astur.^'" 
These definitions are not as good as Kaup’s, and in the case of Astur 
incorrect, as many species allotted by Sharpe to that genus do not show 
the essential (!) character : “ Hinder aspect of tarsus scaled.” 
Gurney’s comments on this {Ibis 1875, p. 221) read : “ In treating of the 
Harriers, Mr. Sharpe comprises all the known species of this group under 
the genus Circus, which is probably a judicious course, as, although that 
genus seems to be naturally divisible into distinct sections (probably four 
in number), it would be difficult to define these satisfactorily without a 
fuller acquaintance than we at present possess, with such variations of 
coloration as are incident to the sexes and successive ages of each species.” 
In the present place, following Sharpe’s “ Handlist,” Circus appears 
at the beginning of the Order, and as this classification is based upon 
evolutionary principles, it would naturally be concluded that Circus was 
the most primitive member of the Order and nearest relative of the 
Pelicans just preceding it. As a matter of fact, this is not so, as when 
Sharpe rearranged the Orders, etc., in the “ Handlist,” he omitted to 
rearrange the genera and species, and Circus is really a specialised form, 
and in the system provided by Pycraft ex Suschkin already detailed, 
and which purposes “ to follow the lines of phylogenetic descent,” it 
appears near the end. 
The Accipitrinec are there the highest, and are preceded by the 
Buteoninse and then the Circinae. No data for this succession is given by 
Pycraft and I have been unable to follow Pycraft’s treatment of this 
14 
