Origin of the Flora of South Africa. 107 



that they must have had an Antarctic origin, a view which is 

 frequently attributed to him and was even held by Charles Darwin. 

 If my views as to the former climatic conditions of South Africa are 

 correct, then there was no possibility for an extensive development 

 of phanerogamic plants in the Antarctic regions ; further, if these 

 views are correct there was no connection between Antarctic regions 

 and South Africa. 



The relationships of the Flora of South Africa to that of 

 Australia have sometimes been exaggerated. On the other hand, the 

 more one studies these relationships, the more one comes to the 

 conclusion that they are indeed very great. They do not only refer 

 to the orders which are usually quoted, namely, the Proteacese, Com- 

 positae, Iridese, Haemodoracese, Buettneriacese, Polygalaceas, Res- 

 tiaceae, Ericaceae, Epacridaceae, certain tribes of Papilionaceae, 

 Rutacese, Thymelseaceae, Santalaceae, and anthospermous Rubi- 

 aceae, but a close relationship is also found in certain tribes of 

 Cyperaceae, in certain grasses which have no northern direct allies, 

 in certain tribes of Liliaceae. and in other natural orders. In all 

 these cases we find that the relations of the two floras are plainest 

 if we compare the flora of that little corner in south-western Cape 

 Colony with the flora of temperate Australia. There seems to be 

 no escape from the conclusion that at one time there was a direct 

 land connection approximately in the latitude of South Africa between 

 South Africa and Australia, and that this land connection disap- 

 peared and left traces of its flora both in South Africa and 

 Australia. Now it is generally admitted by geologists that such a 

 land connection existed in Permian times, a theory which was first 

 brought forward by the late Mr. H. F. Blanford, in the year 1875. 

 A direct land connection between South Africa, Madagascar, and 

 India through Ceylon is admitted to have persisted right into the 

 middle of the Tertiarv period ; it is further admitted, especially on 

 zoo-geographical evidence, that by that time the direct connection 

 between South Africa and Australia had disappeared. Since the 

 causes which led to the excavation of the Indian Ocean must have 

 acted gradually, and probably did not act evenly, I see nothing in 

 the way of assuming that in Cretaceous times, and perhaps even 

 somewhat later, there were still such considerable remnants of the 

 ancient Austro- African Continent, as to allow the interchange of 

 plant types between Australasia and South Africa. 



As the land receded more and more in the Indian Ocean, it was 

 more and more restricted to tropical conditions, and consequently 

 those "plants which could not accommodate themselves to these 

 conditions died out ; remnants of them only being left in those parts 

 which retained a temperate climate. It is a curious fact that Austra- 

 lasia further shares a number of types of plants with the extreme 

 southern end of America, and to a certain extent with its western 

 coast lands as far as Chili. Amongst these there are a few of those 

 types which we found to be common between Australia and South 

 Africa. On the other hand, we do not find in South Africa any 



