THE BIRDS OF AUSTRALIA. 
concluded that the white form was larger than the dark one, but was confounded 
by finding that in some instances, for which he had no explanation, this law 
did not hold good. This confusion is relieved when the matter is treated in 
the way I propose. 
There seems to be no doubt whatever, that these birds range northwards 
from their breeding-places in the “off season.” It appears, moreover, to be 
due to this cause that so much confusing material exists in museums; instead 
of series we find abnormal specimens, or else birds from abnormal localities, and 
these are most misleading. This bird is so big, so common, and well known 
that it is only when it appears as an aberration or in some unexpected place, 
that it suffers death in the cause of science ! 
Carefully considering the nature of the bird that is figured and described 
by Latham, I would conclude it was a northern wanderer from the Antarctic, 
probably breeding at Graham’s Land, which is directly south of the extremity 
of South America. From lack of measurements we may conclude that Forster’s 
specimen was the young of this, as “ Corpus omne ferrugineo-fuliginosum ” 
would agree with such; as Eagle Clarke notes a like coloration at the South 
Orkneys (see 'post). 
With the Graham Land breeding bird, due to lack of specimens, I would 
associate that breeding at the South Orkneys of which Eagle Clarke wrote 
{Ibis, 1906, p. 173) : “ The proportion of birds in pure white plumage in the 
rookeries was not more, perhaps less, than 2 per cent. The colour of the birds 
ranged from very dark brown through all shades of chocolate, and from grey 
through light grey and mottled white to white.” 
From photographs of birds nesting at Cape Geddes, South Orkneys, repro- 
duced on pis. XXXI. and xxxii. of the Scottish Antarctic Exp., the pale birds 
seem most frequent. Unfortunately no measurements of the birds procured 
are given, so that the absolute sizes are not known to me. According to Wilson’s 
reasoning, which I consider good, series would enable us to differentiate 
between the South Orkneys and the Graham’s Land breeding birds. I have 
found that these birds are peculiarly constant in measurements when careful 
comparisons are made. 
I therefore conclude that Staaten Land be accepted as the type-locality of 
Gmelin’s Procellaria gigantea, and that the typical subspecies, to be known as — 
Macronectes giganteus giganteus (Gmelin), 
be regarded as the Graham’s Land breeding bird, and that name be used to 
include the South Orkney breeding form, pro tern. 
At the Falkland Islands there breeds a uniformly coloured dark bird, 
almost black, as Wilson puts it — which is smaller than the southern bird. Its 
186 
