THE BIRDS OF AUSTRALIA. 
it to specific rank, differentiating between geographical subspecies and 
geographical species, and selected King George Sound as the type locality, 
and then synonymised Grant’s Z. shortridgii. This may be incorrect, and 
from the data I would conclude that the types came from Perth; in which 
case Z. shortridgii would be available for the Albany birds, if these prove 
separable from the Swan River form. 
I have named 
Zosterops gouldi warreni. 
“ Differs from Z. g. gouldi Bonaparte in being darker, especially below. 
Warren River, South-west Australia.” 
Warren River district. 
Then, as Ashby has recently confirmed my conclusion that the Carnarvon 
form was separable, 'writing : “ The birds at Geraldton are smaller and brighter 
in tint than birds observed at Perth.” I would admit tentatively four forms : 
Zosterops gouldi gouldi Bonaparte. 
Perth district, West Australia. 
Zosterops gouldi warreni Mathews. 
Warren River district, South-west Australia. 
This darker form probably continues through the very damp south-west 
corner district. 
Zosterops gouldi shortridgii Grant. 
King George’s Sound, probably to Stirling 
Ranges and inland to Broome Hill, etc. 
Nearer the typical form than the Warren River form. 
Zosterops gouldi edwini Mathews. 
Carnarvon. 
While it is possible that the Dirk Hartog Island form may prove separable, 
although Carter wrote : “ The Green-backed White-eye was numerous both 
on Dirk Hartog and the Peron, and some were seen on Edel Land. A com¬ 
parison with skins from the south-west showed no differences.” 
160 
