THE BIHDS OF AUSTRALIA. 
distance as if wounded, then suddenly shooting upwards for a little way, 
which is repeated several times, but gradually getting nearer all the time 
to the surface of the swamp, when this erratic flight ceases, and it perches 
either on the top of the reeds or some branch projecting above them. 
Here the bird rests for some time, then commences its quest for prey 
slowly and stealthily just above the tops of the reeds or rushes, 
carefully scrutinising each small patch of open water or the swamp. On 
one occasion I saw one pounce on a Black Duck {Anas swperciliosa) in 
some open water amongst the reeds. The unfortunate Duck at once 
dived, but the Harrier, expanding its broad wings, effectually prevented 
the Duck from drawing the former under the water, and it could only 
swim about at a certain depth until it was drowned. When this was 
accomplished the Harrier, by gently flapping its wings, steered the prey to 
a small partly submerged log, and then letting go the dead Duck rose 
to the surface. Stretching out one of its long legs, he soon had it on 
the log, and commenced to eat it, but as the fun was now over I 
determined to have the Duck, which I quickly confiscated.” 
There appear to be no notes of this bird from the North-west of 
Australia, which would belong to the race C. a. inexpectatus : all the 
preceding notes are referable to the race C. a. gouldi. 
The racial forms of this bird are not easily determinable. As noted 
above, when Gurney revived Peale’s name he contended his C. wolfi was 
a distinct species. At most it can only be considered as a subspecies, and 
I find Sarasin in Die Vogel Neu-Caledoniens 1913, p. 5, has so classed it. 
That worker, however, unfamiliar with systematic ornithological work and 
literature, has called the New Caledonian bird Circus gouldi wolfi and given 
as distribution of his forms the ranges : 
Circus gouldi typ. ; South and East Australia, New Zealand, Fiji. 
“ Circus gouldi wolfi : New Caledonia, New Hebrides and Lord Howe’s 
Island.” 
The species name must be approximans and the Fijian form must 
bear that name, and therefore Sarasin’s C. gouldi typ. would become 
C. approxirmns typ. ; but if the Fiji and East Australian birds were 
inseparable the New Hebrides one could scarcely bear a different name. 
I do not understand the reference of the Lord Howe Islands bird to Circus 
gouldi wolfi. 
The nomination of the forms would read : 
Circus approximans approxirmns Peale : Fiji Islands. 
Circus approximans wolfi Gurney : New Caledonia, New Hebrides. 
32 
