1884.] Construction of a Chapter in the History of the Earth. 195 



likewise are found in tlie Newcastle series of New Soutli Wales, but in 

 Europe only in Secondary beds. Allowing that some of these genera 

 are purely artificial, and that the species grouped under them may not 

 really be allied in every case, it is on the other hand probable that 

 some forms placed under distinct genera should properly be united with 

 some of those grouped under the genera above mentioned, and, making 

 the most liberal deduction for the value or want of value of negative 

 evidence, I think that there is still a very considerable weight of pro- 

 bability, on this count alone, in favour of a newer type of vegetation 

 having originated in Australia in Palosozoic times and in the Permian 

 period commenced to spread over the rest of the world. 



The explanation seems to be that, on the advent of the Glacial 

 period, the flora, which had supplanted the older types in Australia, was 

 driven towards the Equator. As the climate ameliorated, it did not 

 again retreat towards Australia, either because its place was taken by 

 newer species, or, more probably, because, owing to changes in the distri- 

 bution of land and water, it could no longer do so, but to the north — or 

 what for convenience we may provisionally call the north, — of the 

 Equator it lived on in what is now India and, gradually spreading over 

 the hemisphere, produced a profound modification in the pre-existing 

 floras of what we now know as the Old World. 



The flora of the Wianamatta beds, as I have explained, shews a 

 certain relationship with that of the Damudas, but none with that of the 

 Newcastle beds as far as species go ; of the genera, however, three out of 

 the six, or, if we include the Hawksbury beds, four out of seven are also 

 found in the Newcastle beds. The beds newer than the Wianamattas 

 have yielded a flora consisting of nine species belonging to seven genera, 

 of which, if we except the PhyUotheca australis, only one species is allied 

 to an Indian form, viz., Pecopteris australis, Morr. allied to P. indica, 

 Oldh. and Morr. from the Rajmahals. We have here a distinct decline in 

 the closeness of relationship between the Indian and Australian floras, 

 and, though, of course, this might be due to the imperfection of the 

 record, the probabilities are against its being entirely due to that cause, 

 and we may safely conclude that some barrier separated the two areas, 

 by which the floras of India and Australia were kept apart and followed 

 separate and consequently diverging lines of descent. 



Turning now for a while to South Africa, I must commence by declar- 

 ing it as my opinion that the relationships between the Indian and African 

 floras of the periods I am discussing are with difficulty explicable, unless 

 it is granted that there was in those days a continent, or at any rate a conti- 

 nuous chain of islands, stretching from South Africa towards India. I 

 am aware that Mr. A. R. Wallace has declared such to be uncalled for and 

 26 



