30 THE AMERICAN NATURALIST. [VoL. XXXIV. 
1. The Torresian subregion. This is the Papuan subregion 
of Professor Hedley, but Professor Spencer prefers the name 
Torresian, as less liable to lead to confusion, and as indicating © 
the former land connection. It includes Papua and north and 
northeastern Australia as far south as the Clarence River. 
2. The Bassian subregion, which comprises the eastern and 
southeastern coast region, and Tasmania. The name indicated . 
the route by which “the South American contingent must have 
passed." 
3. The Eyrian urbcs including all the interior, south- 
ern, and western part of the continent. It is separated by the 
coastal ranges from the Torresian region on the northeast, and 
from the Bassian region on the southeast. 
The theory of a former connection between Australia and 
South America by means of a great southern continental area 
now seems thoroughly established. Mr. Wallace thought the 
relationship between the marsupials of the two countries, and 
between several families of birds and reptiles, too general to be 
accepted as direct evidence. He admitted, however, that cystig- 
nathous frogs, several genera of fresh-water fishes, and one 
genus of snail were undoubtedly transported from one region to 
the other, but thinks this transportation was by means of float- 
ing ice. Mr. Lydekker, in his Geographical History of Mammals 
(1896), still maintained that the evidence was too scanty to show 
in what latitude the connection occurred. He admits that the 
discovery of dasyuroid marsupials in the Tertiary of Patagonia 
indicates a former communication, and he points out the impor- 
tance of the determination of the rocks recently brought back 
by the Axtarctic as indicating a continental area; but he thinks 
the presence of Australian types of rats in the Philippines “is 
of the utmost importance in respect to Australia receiving its 
mammalian fauna from southeastern Asia"; and he believes 
that this fauna entered the island by way of New Guinea. 
Professor Spencer calls attention to the fact that polyproto- 
dont forms decrease in importance toward the north and north- 
east in Australia (which is the opposite of what should be the 
case if Professor Lydekker's theory is the true one), and point- 
edly remarks that Australian rats in the Philippines do not prove 
