REPLACEMENT OF THE ANCIENT E. AFRICAN FOREST 5OI 



luwer-rainfall plateau area (at Salisbur}- ) have i)roduced 

 tlourishing- trees. Alt the same time, I abundantly realize that 

 my personal observations have been carried out from a centre 

 that is not only blessed with o^ood summer rains, but is excep- 

 tionally favoured through the relative cool that accompanies 

 a considerable elevation, and through the usual possession of 

 light winter rains in addition. So I will not generalize. 



On the other hand, a reduced rainfall — and even a relatively 

 slightly reduced rainfall — might well have produced a powerful 

 indirect effect. After a good rainy season the heavy ureen 

 fringe of Hypoestes and other tall herbs and shrUbs that sur- 

 roiuids our dense forests lets little fire in, and it cannot travel 

 if it ^ets in. In a very dry year the opposite tends to happen, 

 and there can be little doubt that dry cycles must see a \astly 

 greater destruction of forest than wet ones. Since grass-veld 

 became its neighbour and grass-fires became the order of the 

 vear, the Chirinda type of plant formation has depended for its 

 continued existence on the adecjuate co-operation of herb and 

 shrub and tree. The herbs that belong to it cannot flourish and 

 keep back fire without the coolth and relative moisture that is 

 retained, i^'licn the rainfall cjh'cs it, by the canopy of the trees 

 and shrubs and lianas, and if the trees are deprived of luscious 

 greenness in their herb-fringe and carpet, they are also depri\ed 

 of their protection against fire. It is even i)ossible that the 

 trees of the formation might inhabit a far wider general area than 

 they now do were it not for this indirect result of a smaller rain- 

 fall, and the fact, already referred to, that at least one of them 

 flourishes under the latter condition in the Gardens at Salisbury 

 (I do not know, however, what the cultivation has been) may 

 suj)port this hy}X)thesis. 



T have already suggested that this indirect factor will be 

 specially powerful where the soil is relatively poor or sandy, 

 and that this may be one reason why forest tends to disai)]>ear 

 from soil of this kind first. It may just be worth suggesting 

 further that another indirect factor working in the same direc- 

 tion — a less important one— ;;/a3' have been — elephants. At any 

 rate, mv recollection of their work in the Sitatonp'a forest, in 

 which they were somewhat ])lentiful at the time of my visit, 

 leads me to suspect that the rather poor forest that tends to 

 occur on the poorer sandy soil in Chirinda would be more liable 

 to such damage by elephants as might tend to let fire in. 

 Whether, when they were here in the great numbers described 

 by old natives, their destruction of the " green 'fringe " actually 

 did more than counterbalance the lessened ferocity of the fires 

 that might ha\e resulted from the general trampling (yet greater 

 drying) of the grass-veld round the forests, can best be settled 

 by observation of forests in which they are still abundant, but 

 I am inclined to think that they may have acted as a slightly 

 accelerating factor in the destruction of forest by fire. A more 

 important accelerating factor will have been the late fires that 

 were in vogue during the native regime. According to all 



