Annals of the Transvaal Museum 
207 
Falconidae 
Polyboroides typicus — Gymnogenys typicus! (cf. Richmond, Auk, 1902, p. 92). 
I cannot find myself in agreement with the present arrangement of the 
Harriers, as they appear to me to be an ancient group, if we are to judge 
by the isolation of the species. Mathews has reduced the numerous generic 
names that have been applied in the past to four which can be utilised, 
the rest falling to the synonymy of these four (cf, B. Australia, v. 10, 1915), 
as follows : 
Circus type C. aeruginosus. 
Spizacircus type C. macropterus. 
Spilocircus type C. assimilis. 
Pygargus type C. pygargus. 
The first three appear to me to be allied, the second and third repre- 
senting subgenera of the first in South America and Australia respectively, 
while the fourth represents a distinct genus. The emargination of the 
primaries must be an anciently developed character, since we find it quite 
as constant in the immature as in the adults. Circus aeruginosus is a 
migratory species, characterised by its long and stout tarsus and the 
emargination of the second outermost primary so far back that it is almost 
hidden basally by the upper primary coverts; C. ranivorus of South Africa 
is a resident ally of aeruginosus, and, like C. macropterus and C. assimilis, 
has developed local differences. Pygargus pygargus is also a migrant to 
South Africa from the northern hemisphere, but differs from C. aeruginosus 
in having a much shorter and more slender tarsus and the emargination 
of the second primary situate a considerable distance beyond the tips of 
the upper primary coverts. C. macrourus (Gmelin) is similar to Pygargus 
in not having the fifth outermost primary emarginate, but differs in having 
the emargination of the second outermost primary nearer* the base as in 
Circus', in the tarsus it is intermediate between Circus and Pygargus. 
Having regard to these differences, not to mention those of colour, which 
are important, I feel it is necessary to give the species a new generic name. 
Pseudocircus, appropriate as indicating its position, Accipiter macrourus 
Gmelin the type. The last species on the South African list is C. maurus 
(Temminck), a resident species, which is perhaps allied to Pseudocircus, 
though differing therefrom in its black coloration and in having the fifth 
primary emarginate on the outer web; for this species I propose the new 
generic name of Melanocircus. 
Gurney {Ibis, 1875, pp. 353-370), in reviewing the classification of the 
Accipitres, has shown that Astur polyzonoides dcn.^ A. tachiro should not 
be referred to Astur, and places them under Scelospiza, type A. francesii 
A. Smith ; but with this arrangement I cannot concur, although agreeing 
that they should be removed from Astur. Apart from superficial resem- 
blances between some members of the genus as commonly associated, the 
scaling of the tarsus indicates several well-defined groups, or natural genera. 
In the typical genus Astur, type A. palumbarius = gentilis, the tarsi and 
toes are very short and stout and the scales on the former are broad, trans- 
verse scutes anteriorly and posteriorly, separated laterally by a number 
of small, reticulate scales. In A. polyzonoides, well as in H. tachiro, the 
tarsi and toes are long and slender, and the reticulate scales are replaced 
