1883 J on tTie Island of Socotra and its Becent Bevelations. 313 



Natal to Abyssinia, and in like manner that there was a connec- 

 tion between the western region about Biafra and the Abyssinian 

 district. It would appear probable that at a time when the tropical 

 zone was much cooler than it now is, the northern forms of plant life 

 spread over South Africa, but with the diminution of the cold of the 

 glacial epoch they were driven back and retreated northwards, a few 

 types left on the higher regions being the only evidence of the 

 invasion and the survivors of the hordes extinguished. Thus at the 

 present day on the higher lands of Abyssinia, West Africa, and in 

 South Africa, we have the fragmentary traces of the extension of an old 

 African flora, and of this flora Socotra would appear to exhibit on its 

 hills the most north-easterly limit, just as Madagascar and perhaps 

 the Mascarene Islands show the most easterly extension. 



The African affinities being so explained, what of the Indian and 

 Asiatic ? The higher level of the land necessary to unite Socotra 

 and Madagascar with the African continent, if continued over a 

 slightly wider area, would produce some interesting changes in 

 surface features. Africa would be joined to Arabia, the Persian Gulf 

 would cease to exist, and the Tigris and Euphrates, united in one 

 stream, would pour their waters through a delta occupying much of 

 the Arabian Sea, and through which also the Indus would debouch. 

 Thus a means of transit for the migration of Indo-Malayan types 

 would be afforded. That some such connection as this did formerly 

 exist all evidence conclusively shows, and that by this route migration 

 took place is equally certain, and we have thus an explanation of 

 Indo-Malayan affinities in Socotra and in Africa, without calling in 

 the aid of the now untenable hypothesis of an " Indo-Oceania " or 

 " Lemuria," making a complete land surface over what is now the 

 Indian Ocean. Whatever be the date of the variations in level which 

 brought about the present surface features in this region, it is clear 

 that the separation of Africa from Asia and the formation of the 

 straits of Bab-el-Mandeb took place before the final isolation of 

 Socotra. 



Australian affinities are explained on the same lines as are the 

 Asiatic ; but when we come to deal with the American we touch a 

 matter of which at present we can give no adequate explanation. 

 That the many forms identical generically in the Indian Ocean 

 Islands and in America are sprung from one stock of great antiquity 

 must be admitted, but whether the migrations which have led to their 

 occurrence now in such antipodean regions were directed in an easterly 

 or in a westerly direction, is one of those problems of which we have 

 not yet the materials for a solution. 



How long, then, has Socotra been an island ? It is difficult, nay 

 impossible, to picture the complete history of the island from the 

 earliest geological epochs ; but in brief some, such as this, may be 

 sketched. During the Carboniferous epoch there was in the region 

 of Socotra a shallow sea, in which was deposited on the top of the 

 fundamental gneisses of this spot — which had ere now been certainly 



