5CX) REPLACEMENT OF THE ANCIENT E. AFRICAN FOREST 



dense forest-patches, of the high plateau country, the backbone' 

 of which forms .the water-shed between the basins of the Congo 

 and Zambezi rivers, that " they are commonly on the sovirce or 

 banks of a stream." Many of our kloofs — with streams or ])ro- 

 bable underground moisture — carry dense bush that is ])racti- 

 cally a sub-community of our high forest formation, and often 

 contains a considerable sprinkling of true forest forms, as 

 Khaya, Plptadenia, the shrub Tn'calysia, and others I might 

 name. The forest-patches on dolerite. as I have said, often 

 occupy the hill tops, but the red loam in any case probably con- 

 serves moisture better than the sandstone. Of the little patches 

 on the latter formation that I have referred to more particularly 

 earlier, it is interesting to note that the elongated one, nearest 

 Chirinda, now no longer carrying forest, follows roughly a 

 shallow kloof, in which a spring has sometimes ai)peared after 

 great downpours in our older, heavier, rainy seasons, while those 

 nearest Chipete were connected this last season, after a 31.4-inch 

 rain in four days, by quite a stream — on a convex surface with 

 no trace of a previous suriface-flow excepting from the lowest 

 of the patches. The survival of one or two other \ery small 

 patches has obviously been aided by the presence of low, wide- 

 spread ant-heaps of Termes bcllicosus, full, of course, of col- 

 lected humus. So that, on the drier soils, it is in the damper 

 places or above probable subterranean waters that 'forest most 

 tends to survive. Again, it is the rule — which has its notable 

 exceptions — that the forests have eastern and south-eastern 

 aspects. 



Actually, all this is explicable also in relation to fire, and 

 against the other interpretation we nmst place the present non- 

 afiforestation of the va.st bulk of the doleritic area with the aspect 

 described, and the ease and success with which forest-trees can 

 be re-established on the drier soils, continuing to take care of 

 themselves if looked after for the first season or two, and then 

 merely protected from fire. Also the fact that considerable 

 pieces of Chirinda are on the sandy soil, yet carr\- flourishing 

 forest — lower and slightl\- less dense, it is true, on one of them, 

 yet apparently thoroughly healthy. Chipete, it is true, has 

 latterly rather the appearance of decay, but it is on the dolerite. 

 and I will show below that this a])pearance is probablv due to 

 another reason than dr-ying up. It must be remembered, for the 

 piece of country with which I am s))ecially dealing, that the rain- 

 fall is even now heavy right from the coast to the mountains, ex- 

 cepting within particular circumscribed areas. If the inner 

 plateau lands of lighter rainifall, which line the great cur\c of 

 dotted forest and comprise much of .South .Vfrica and most of 

 Rhodesia, ever boasted high forest, and this were destroyed 

 before the advent of fire-using man, one might be disposed to 

 attribute it to a general decay accompanying a falling rainfall, 

 I doubt whether our coast rainfall has yet fallen so low as to 

 act thus as a directly destructive factor, and it is interesting 

 that seeds of forest-trees, sent bv mvself manv vears ago to the 



