328 Transactions of the South African Philosophical Society. 



development of species. But these are, according to him, only- 

 external resemblances ; there are in reality very great differences. 

 The true Cape Eegion is, on the whole, without trees, while in West 

 Australia portions of the Eremsea, which climatically correspond to 

 the Karroo, have big trees and scrub, about the height of a man. 

 Besides, the Cape Eegion possesses many succulents; these are 

 wanting in West Australia, with the exception of a few that belong 

 to cosmopolitan orders. The Cape exceeds to a tremendous extent 

 West Australia in the number of bulbous and tuberous plants; 

 further, the annual plants which play an important part in the 

 composition of the Cape Flora, do not find their counterpart in West 

 Australia, except in a few places of the Eremsea. 



These statements by Diels, if unchallenged, might tend to make 

 the pendulum swing too much in the opposite direction. I am 

 afraid Diels has not sufficiently taken account of the fundamental 

 difference between the Flora of the Karroo and the Flora of the South- 

 western Eegion. If we stick strictly to the latter there are certainly 

 such close connections that we cannot avoid the conclusion that 

 they are the result of a former land-connection south of the Equator. 

 Hooker refers to Proteacece, Compositce, Iridece, HcemodoracecB, 

 Polygalece, BestiacecE, EricacecB (in South Africa), corresponding to 

 Epacridece (in Australia), Papilioncaece, Butacem, Thymelece, Santa- 

 lacecB, and some others which, by their abundance in both countries, 

 point to parallelism of development from a common origin. Ben- 

 tham (1873) has worked out this point with reference to the 

 Composites. C. B. Clarke has emphasised it with reference to 

 certain -CyperacecB. Schindler (1904) is of opinion that the 

 Halorhagace^ have had a Southern origin. Stapf (1904) has 

 pointed out that the endemic forms of Graminacece of his *' Cape 

 Province " have their allies not with the Tropical African grass 

 Flora but with the Temperate types of Eurasia and Australasia. 

 He clearly recognises the difficulty of a Northern connection, 

 although not a few of the genera of grasses in question are 

 represented on the mountains of Tropical Africa by allied forms, 

 which, as it were, form a bridge to the large Northern centres of 

 development. To these belong Avenastrum, Agrostis, Meiica, Poa, 

 Festuca, Bromus, Brachyjjodium. Others pinch out without passing 

 into the Tropics, such as Pentaschistis and Pseudohromus, or are 

 poorly represented beyond it. But when one wants to picture to 

 oneself in detail how this distribution has been brought about, 

 the difficulties seem insurmountable. The genus DantJionia has 

 another centre of development in Australia and New Zealand, where 

 it radiates to the North to Temperate Asia. This process is repeated 



