GRASS-WARBLER. 
stage into the rufous-headed dress, and there can be little doubt that the 
winter plumage of both sexes is blackish with streaked backs and heads.’ 
My three specimens are in moulting, No. 1 is dropping strongly on back and 
under parts ; No. 2 has the body ready but is moulting in the tail ; No. 3 seems 
to be full plumaged. All have the winter plumage. Apparently their 
breeding-season is, as is usual in Australia, during the summer, and the moult- 
ing for the winter-dress in April and May. Probably a later moulting in Oct. 
and Nov. (or even later ?) takes place.” 
Macgillivray has written from North Queensland : “ Common on the 
grassy flats. One noted carrying building material on the 19th January. 
The nest was fixed in the leaves of a small plant amongst the grass. On the 
23rd January, when making our way homeward across a long grassy flat in 
drenching rain, we flushed another from her nest containing four eggs ; this 
nest was almost round in shape, 2 inches in diameter, the base formed of fine 
grassy fibres and cobwebs, the sides and top a canopy of leaves sewn together 
with fine fibres and cobwebs. They get more numerous on the flats as the 
wet season advances. Common on the Archer in June. They are usually 
absent from the north in the winter.” 
This species was first recorded by Vigors and Horsfield, and when Gould 
received specimens varying in size and coloration he described three as 
distinct species, and certainly he would appear justified were the variation 
unknown. This made four species, but soon it was discovered that two were 
simply seasonal variations, so that three distinct species were accepted until 
the Catalogue of the Birds in the British Museum was written when Sharpe 
determined that all were referable to the same species and, ignoring subspecies, 
allowed Cisticola exilis to range over the whole of Australia, New Guinea, 
the Moluccas, Burmah, the Philippines and China, a remarkable range for a 
sedentary almost flightless bird. 
The whole group of Cisticola has puzzled systematic ornithologists on 
account of the seasonal variation and, as above noted, this species varies accord- 
ing to climatic conditions, so that within a few miles specimens may be secured 
showing different seasonal changes at the same time. Further, while the 
seasonal changes are very pronounced in the more temperate regions they are 
little developed in the tropical zones where they vary less noticeably. How- 
ever locally the variations are limited, that is, constancy is seen when series 
are compared, the different stages of plumage being carefully separated and 
such contrasted. 
Owing to the great variation very long series are necessary to deter- 
mine the subspecies, but when these are available such is comparatively 
easy. 
371 
