lxxii BOTANY OF SOCOTRA. 



specialised type and ancient character of the flora of south Africa, and its 

 relations and similarities with types found at widely separated spots upon the 

 highlands of the east coast of Africa, especially about Angola, the Cameroon 

 mountains, and Fernando Po, in some of the Atlantic islands, in the north- 

 west of Africa and south-west of Europe, as well in the highlands of east 

 tropical Africa and Abyssinia, all point to the hypothesis which was first 

 enunciated by Sir Joseph Hooker, that the south African flora has been 

 continued along the highlands of east Africa and Abyssinia, and that a like 

 connection existed through central Africa, between the western and north-east 

 regions. That at a time when the tropical zone was much cooler than it now 

 is, northern forms of plant-life spread as far as and over south Africa. With 

 the diminution of the cold, these forms were driven back, and retreated north- 

 wards and up the mountains before the advance of a vegetation more tropical 

 in its character, and a few types left on the isolated higher points of land in the 

 regions mentioned, are at the present day the only evidence of the existence 

 of this ancient flora, and of the invasion by what we now call the tropical 

 African flora. 



If we accept this hypothesis, then, in Helichrysum, Babiana, Hamianthus, 

 Dracaena, Begonia, and other Socotran plants with like kinship, we have an 

 outlying fragment of this old African flora, its north-eastern limit, just as in 

 southern latitudes we have its eastern limit in Madagascar and in the Mascarene 

 Islands ; and we learn further, that at a time when this old African flora peopled 

 the land Socotra was yet a part of the continent, and the northward extension of 

 some of the types, for instance, Euryops, into Arabia seems to show that Africa 

 was then also joined in the north-eastern region with Asia. Whether the land- 

 connection with Africa persisted during the time of and the expulsion of the older 

 flora, so that the invading forms could spread directly over the land-surface, 

 which is now Socotra, or a separation of the island took place during the reign 

 of the old flora, it is not easy to decide. I am disposed to believe that there 

 was insulation before the incoming of the new tropical forms, and that only at 

 a later period, when land-connection with Africa was again established, did the 

 general tropical African flora spread its influence before the final separation of 

 the Socotra of to-day. 



But if a land-connection with Africa is necessary for an adequate explana- 

 tion of the African affinities, a like union with Asia must be assumed in order 

 to account for Asian relationships. I have incidentally mentioned that the 

 extension of some of the old African types into Arabia may be attributed to the 

 period of the old African flora, when Socotra was part of the African continent. 

 Any elevation of the land which would unite Socotra with Africa would bridge 

 over the Straits of Bab-el-Mandcb, and a further elevation would bring Socotra 



