356 MR. R. I. POCOCK ON THE [ Apr. 21, 
present time in this area, the former in Java and the Nicobars, 
the latter all over the Indo- and Austro-Malayan Islands and 
Australia. Selenocosmia, which must be regarded as a direct 
descendant of Phlogiellus, must itself be looked upon as the anaestor 
of the Malayan Coremiocnemis, the Assamese Lyrognathus, the 
Burmese, Indian, and Ceylonese Chilobrachys, and of the two 
Australian genera Selenotypus and Selenotholus, all of which, in 
the absence of evidence to the contrary, may be held to have 
originated within the areas of their present distribution. The 
affinities of the aberrant genus, Pacilotheria, which is specialised 
both in structure and habits, are more doubtful; but there is no 
evidence against the hypothesis that it has been evolved in India 
itself. 
The presence of Pwcilotheria and Chilobrachys in Ceylon and 
India attests their occupation of the latter area before the severance 
of Ceylon; and the extension of Selenocosmia into Australia but 
not into New Zealand, similarly attests a southern migration into 
the former country before its separation from South-eastern 
Asia, but after the isolation of New Zealand. It seems probable, 
indeed, that this southern immigration into Australia synchronised 
with that of the ancestors of the Australian marsupial mammals, 
and in this connection it is significant to note that the latter are 
believed to have originated in South-eastern Asia* and to have 
entered Australia in Eocene times. The great difference, however, 
in distribution between the Marsupials and Selenocosmiine may 
be explained by the survival of the latter, and the extinction of 
the former, in the area of their birth. 
The Thrigmopeeine, which are confined, so far as is known, to 
India, appear to have been developed from the Aviculariine stock 
ata late date, namely after the depression of the connecting land 
with Ceylon; and the Ornithoctoninz, which range from Siam 
and Burma to the Moluceas, seem similarly to have put in an 
appearance after the separation of Australia from the continent 
to the north of it. 
The Eumenophorine must have entered or been developed in the 
Ethiopian Region at an early date, antecedent to the separation of 
South Arabia and Sokotra from what is now Somaliland, and 
before the formation of the Mozambique Channel divided Mada- 
gascar from Hast Africa. 
The Harpactivine, on the contrary, which range from Somali- 
land to Cape Colony but are unknown in Madagascar, seem to 
have originated independently within the Ethiopian Region 
subsequent to the Miocene period. The least specialised of the 
genera (Harpactirella and Pterinochilus) have not departed far 
from the type of structure found in the Aviculariine. 
There remains the South-American fauna to be accounted for. 
Owing to the general similarity that obtains between the — 
genera of this area and those of other parts of the world, the 
! Lydekker, ‘ Geogr, Hist, Mammals,’ p, 5 (1896). 
