xii A MANUAL OF THE BIRDS OF AUSTRALIA, 
The complex persists in the fact that it is sometimes difficult to judge whether 
a species, as it may show little distinction, should be referred to this or the 
earlier migration, and this can only be determined by the relative antiquity 
of the genus considered. This may be gauged to some extent by the distribu- 
tion of the genus and its allies, and it is in connection with these studies that 
genus lumping proves so misleading. We conclude that when this second 
migration took place Australia was connected by land with New Guinea, 
and Torres Straits had not been evolved. Since the division a third irruption 
into Australia has taken place, and the procedure is still going on. This 
irruption has never overrun the continent, but has either run down the east 
coast or along the north coast into the north-west or both, and comprises 
species still living unchanged in New Guinea. It has been suggested that at 
one time a connection of Australia with the Northern Islands existed in the 
north-west, quite distinct and separate from the Torresian connection. There 
exist a few forms which seem to countenance this suggestion, as instance, 
Eremiornis, isolated in the north-west and living on the Montebello Islands, 
a form almost as peculiar as Atrichornis. To crystallise the above migrations 
we may cite a concrete example. The ancestral Moorhen arrived in Australia 
and penetrated into Tasmania ; retaining primitive features upon the separa- 
tion of the latter country, the island form degenerated into a huge semi-flightless 
big-footed bird ; the mainland one, through stress of environment, drought, 
etc., was compelled to retain its flying powers and also its smaller size, but 
still shows primitive appearance. In the north, alteration through necessity 
proceeded, and then the more highly developed bird with the second immigration 
proceeded to colonise Australia, and spread over the continent but not into 
Tasmania. At the third immigration a new development, the Rufous-tailed 
Moorhen, crossed Torres Straits and ranged along the north to the Northern 
Territory and into North Queensland. This sequence can be traced in 
connection with many other birds and is worthy of study. A few words 
must be given in connection with the desert forms; these are apparently 
referable to the first immigration and have been evolved through environmental 
stresses, sometimes so much that their relationship has become obscure, and 
it is always possible that their ancestral relations have become extinct without 
leaving any obvious descendants. It is possible that complete anatomical 
and osteological examination would assist in tracing such to an acceptable 
ancestry, and as a good example for investigation we suggest Peltohyas, the 
Australian Dotterel. 
Though the arrangement in this work is limited to species it must not be 
concluded that we minimise the value of the study of subspecies, but rather 
we have allotted to them their full value, and we appreciate their worth in the 
consideration of zoogeographical problems, especially such as are presented 
by the numerous “islands ”’ indicated on the Australian mainland. These 
have been previously indicated by Mathews, so we will not deal with them 
here, but simply call attention to their presence once again. The study of 
subspecies is now being undertaken by Australian ornithologists, and it is. 
