360 Transactions of the South African Philosojjhical Society. 



remains of the oldest African Flora which finds its southern limit in 

 South Africa, its eastern limits in Madagascar and the Mascarene 

 Islands, and its north-eastern limits in Socotra. Though I must 

 presuppose a still older element, namely, the typical Cape Flora to 

 which the above-mentioned genera cannot be said to belong strictly 

 speaking, I think the general reasoning of Balfour with the facts 

 before him are perfectly sound. 



Of course the obvious explanation of these facts is a former land- 

 connection with Africa — an explanation adopted by Balfour. He 

 also assumes a former land-connection to explain the presence of 

 Indo-Malayan types in Socotra, Madagascar, and the Mascarene 

 Islands. Incidentally this would explain the presence of a number 

 of such types in Tropical Africa and South Africa. We gather 

 (p. Ixxiv) further from him that in early and middle Tertiary times 

 Socotra was in great part submerged, and that since then it has not 

 been connected with the East. If these views are correct, the 

 invasion of the Indo-Malayan types must have taken place in 

 pre- Tertiary times — a conclusion which is in agreement with the 

 views as to the carving out of the Indian Ocean at which we 

 previously arrived. 



There remains now only to be considered the origin of the 

 Mediterranean and other Northern extra-tropical elements which are 

 found in the Flora of South Africa. We have altogether left aside 

 Ettingshausen's views (1875, p. 624) of the former existence of a 

 universal Tertiary Flora, of which a good deal is supposed to have 

 been left stranded in South Africa. There seems, however, little 

 doubt that many types, now widely spread, have originated in the 

 Northern Hemisphere. There can be very little doubt that a fair 

 number of these have reached our parts, and have even pros- 

 pered in the "Cape Province." Perhaps the most conspicuous 

 example is the genus Bhus, of which no less than eleven species 

 occur on the Cape Peninsula alone, yet the close analysis to which 

 the sect Gerontogece, Engl., which includes our own species of Bhus, 

 has been subjected by Diels (1898) leads him, and I think rightly, 

 to the conclusion (p. 642) that this section has branched from the 

 type (Stamm) of the genus presumably during the early Tertiary in 

 the southern portion of the eastern part of the Northern Hemi- 

 sphere. Originally it probably included forms, the organisation of 

 which was adapted to moderately dry and sunny localities. When 

 in the course of the Neogen the geological revolutions in Western 

 Asia and Europe caused migrations on a large scale, and at the same 

 time more intimate relations were established between East Africa 

 and the Indian region, Bhus took part in the general invasion of 



