218 A NATURALIST IN MID-AFRICA. 



chain, Ankole and Karagwe, while to the sonth are 

 the high plateaux south-east of Tanganyika, which 

 are probably inhabited by the same plants as the 

 Stevenson Eoad and Angola. These two areas 

 show a most marked similarity to one another, bnt 

 at the same time there is just the difference one 

 would expect. To prevent confusion it is best to 

 state clearly the apparent main stations along this 

 north to south migration : — 



1. Abyssinia. 7. Victoria Region. 12. Somaliland 



2. Ruwenzori. 8. Kenia. and Tarn Desert. 



3. Central Watershed. 9. Masai highlands. 



4. Stevenson Road. 10. Kilimandjaro. 



5. Angola. 11. Usambara (in part). 



6. Shire highlands. 



If we imagine instead of plants an immigration 

 of human beings, it is much easier to understand 

 in what sense these districts ought to be taken. 

 Supposing some Abyssinian race, which is obliged 

 to remain at a height of at least 3,500 feet for the 

 sake of health, to have colonised Central Africa ; 

 then every one of these 12 sub-colonies would in 

 course of time become different from one another. 



If we suppose that they colonised the country 

 before the Victoria Nyanza was excavated down to 

 its present level, it is easy to see that those parts 

 of the tribe which inhabited its site might con- 



ceivably remain on its shores ; but when things 

 became as they are at present they would have to 

 live in quite a different way to their relations on 



