312 Professor Bayley Balfour [April 20, 



Citliarexijlon. And the Geraniaceous Dirachma has its nearest affinities 

 in the Wendtiece and Vivianiece of Chili and Peru. 



To summarise the features presented by the Pheenogamic flora of 

 the Island of Socotra, we may say : — 



1. It is that of a continental island and presents features of great 

 antiquity. 



2. Kelative proportion of orders to genera and of these to species 

 is large. 



3. There are few annuals. 



4. It possesses much individuality and further exhibits three 

 distinct elements, (a) of a dry parched region, (6) of a moister tropical 

 region, (c) of a cooler and more temperate region. 



5. Its affinities are essentially Tropical African and Asian, but the 

 African element predominates, and in the African element we find in 

 great force the features of the flora of the mountainous region of 

 Abyssinia, West Tropical Africa and South Africa, and also of 

 Madagascar. This element of the flora too is that of the higher regions 

 of the island (c of the last paragraph). 



6. The flora of the dry region is the typical Arabo-Saharan. 



7. The flora of the moister tropical region is that of the Old 

 World tropics generally. 



8. There are a few Indian and American types. 



It may not be out of place to conclude these introductory remarks 

 with a reference to what may be learned from the biological and 

 physical features of the past history of the island and the changes it 

 has undergone, and the way ii^ which the features as we now see them 

 have come to be. 



The position of the island would a priori lead one to expect that it 

 had formerly been a part of Africa. This it undoubtedly has been, 

 but its separation from the mainland is of some antiquity. It is 

 indeed, with Madagascar, to be regarded as the remains of a greatly 

 advanced African coast-line at a remote period. This connection 

 with the mainland explains the general African affinities of its fauna 

 and flora ; but there remains the problem of the South and West 

 African relations for elucidation. The botanical features of Africa 

 have long excited investigation and speculation. The Cape flora, as is 

 well known, is one of a very old and highly specialised type, and in 

 Northern Africa and Western Europe there are many plants in the 

 flora which, as Mr. Bentham has pointed out, are very nearly allied 

 to corresponding Cape species. Again, on the east side of Africa 

 one finds on the Abyssinian hills and the range stretching south, 

 a vegetation distinctly South African. Further, on the Cameroon 

 Mountains and Fernando Po there are not many characteristic 

 Cape types, but what there are, are identical with sj)ecies of the 

 Abyssinian hills. The relations and similarities of the floras of 

 these different regions would all seem to point to the conclusion 

 first promulgated by Sir Joseph Hooker, that the South African 

 flora has been continued along the highlands of East Africa from 



