THE BIRDS OF AUSTRALIA. 
all white, whereas in that form they are more or less blue ; the biU is thinner ; 
the long lateral under tail-coverts have the outer web slate, otherwise pure white 
(in the Gough Island bird the under tail-coverts are pure white) ; the under-side 
of the primary- quills shows whitish on the inner web, but the white is not so 
pronounced as in the Gough Island birds, which approach the ilustralian 
birds in this respect. The measurements are: culmen (exp.) (av.) 26 mm., wing 
172-180, tarsus (av.) 35. 
In the West Indian Islands, Bahamas and Bermudas, there breeds another 
form which is obviously the brovm-black bird, when worn having a smoky- 
brown appearance, quite unlilte the blue-black bird at any stage. Though 
Rothschild and Hartert concluded that “ the wing is generally distinctly longer” 
than that of the Mid-Pacific form, I do not find this so, though their other 
separative characters I find constant. The lower half of the lores is white, and 
the bill constantly longer, the tarsus also longer ; my measurements being : 
culmen (exp.) (av.) 30 mm., wing 198-205, tarsus (av.) 39. 
In connection with the preceding must be considered a bird found in the 
Persian Gulf and thereabouts, which is represented in the British Museum by 
three specimens : one from between Gwader and Muscat, the second from the 
Mekran coast, and the third from Aden. 
These have the inner-wing mottled with smoky and the axillaries dark 
smoky-brown ; the lower half of the lores is white ; the sides of the breast do 
not show the brownish patch present in most of the brown -black birds to which 
this must be attached ; the shorter under tail-coverts are white, the longer 
smoky-brown. It will be noted that this form recalls the one inhabiting the 
Galapagos Archipelago, but is easily separated by its longer bill. The first noted 
specimen has the bill 34 mm. long by 10.5 deep, wing (moulting) over 185 mm., 
tarsus 37 ; the second bill (imperfect) 33 mm., the wing 198 mm., tarsus 37 ; 
while the Aden bird has the bill 31 mm. long by 8 deep ; wing 210, and tarsus 
39. These measurements indicate that more than one form may be here confused, 
as these have all been referred to the same species, P. ’persicus Hume. 
It is unfortunate that investigation regarding the names to be applied to the 
preceding birds reveals much that is unsettling, and as I am attempting to 
clear up the confusion surrounding these birds, I am compelled to advocate 
many quite unexpected changes. Firstly, as to the oldest name for either the 
brovTi or blue-black birds, I am confronted with a quite novel proposition. 
I have given Gmelin’s description of his P. ohscura founded upon Latham’s 
Dusky Petrel, and it will be remembered that the latter noted a specimen was 
in the Leverian Museum which measured two inches less. This bird is still 
preserved, whereas the original type of Latham’s species is lost. Through the 
generosity of Dr. Sassi I have been allowed to examine the Leverian Museum 
64 
