502 REPLACEMENT OF THE ANCIENT E. AFRICAN FOREST 



accounts, most of the burning' was annually postponed till Seji- 

 tember and October, when the grass is very dry, and when the 

 forest fringe and carpet is also at its driest, owin^' both to the 

 hot sun and to the strong, dry winds that are blowing con- 

 tinually from Aug-ust onwards. These late fires are often 

 exceptionally fierce and destructive. It is also possible that 

 throughout the hot, low veld the rate of destruction may have 

 been comparatively high. 



A rcfardiuf/ factor may well have been the grass-eaters of 

 our once great ungulate population, and the known facts in this 

 connection e\en constitute a strong argument against the view 

 that fire has been an important factor at all. We know from 

 old hunters that formerly in the Free State and elsewhere these 

 grass-eaters used to eat the veld bare, and it is the fact, appar- 

 ently, that in British East .\frica to-day. and at any rate in con- 

 siderable portions of German East Africa, they eat the grass 

 so close that fires do not take place. Our own older natives 

 state that this has happened over particular areas even in the 

 section with which I am more particularly dealing, but that it 

 has not been a general phenomenon, annual fires having taken 

 ]5lace through the rest of the country throughout their life-times. 



This shows that even with only a native population in the 

 land it has been possible to have periods of grass-fires. I have 

 already referred to a considerable destruction of forest fhat took 

 place through fire 30 years or more 'before the white man came. 

 Selous records the grass-fires as occurring annually in his hunt- 

 ing days in the Transvaal. Bechuanaland. Matabeleland, and 

 Mashonaland ; they 'have been, apparently, as regular a phe- 

 nomenon in Northern Rhodesia with buck still plentiful. I do 

 not know, however, whether this applies back to the time at 

 which the whites entered that country. I well remember refer- 

 ences to the severity of the annual fires in Nyasaland in early 

 numbers of the B.C. A. Times : and I will shortly quote a passage 

 frcmi an account of Alexander Whyte's botanical exploration of 

 Mount Milanji 26 years ago ( Trans. Linn. Soc.. 1894. vol. iv. 

 Pt. I, p. 3), that shows them to have been occurring annually 

 before that time with " de])lorable " and '' devastating " effects 

 on the forest. All this suggests strongly that such a superabun- 

 dance of antelopes, etc., as will completely stop grass-fires, has 

 been a local and temporary phenomenon, alternating with periods 

 of fire ; and. of course, so long as fires take place at all. they 

 will, in dry seasons, destroy forest. 



But the three strongest arguments in favour of fire as a 

 factor are indirect. The first is, the result of protecting forest 

 from fire. I will describe this in Section 6. The second is this : 

 The grasses which covered this country when the treks came in 

 still cover it wherever fires have continued annually, and grazing 

 has not been excessive. Where, on the other hand, grazing has 

 been very close and burning has for long stopped, new gras.ses 

 are replacing the old ones. The third is the strongest argument 

 of all. It is afforded bv the existence to-dav. over all this vast 



