506 REPLACEMENT OF THE ANCIENT i:. AFRICAN FOREST 



Sylz'aticitm (Euphorbiacece ), and Ccltis dioica ( Uricacccie ) . all 

 big trees of Chirinda, are flourishing under the unwonted condi- 

 tions in spite oif a somewhat dry series of years. The same may 

 be said of IViddringtoiiki IVIiytei of the mountain forests, and I 

 have been told that this tree, amongst others, comes u]j, and 

 grows up, outside the forests in unlnu'nt ])ortions of (iermaa 

 East Africa. My trees named abo\e. Hke those of the luu-o- 

 pean forests, show clearly their independence c^f the massed 

 condition for such protection as they may need from excessive 

 sun — and wind — ^by readily protecting themselves when neces- 

 sary through lower branching, a stockier growth, and, in one oi 

 two cases, denser foliage. The firmer, smaller " sunlight ' 

 leaves were naturally assumed earlw in ready adajjtation to the 

 early need for reducing transpiration. 



The moisture that canopy — and especiall\-, as in Chirinda, 

 layers of canopy — directly and indirectly conserves is the real 

 essential : not to the big trees, once started, for they are shade 

 endurers and shade makers rather than shade l(jvers and sun- 

 light (fearers : but to the formation as a whole, in keeping green 

 its defence against fires. Yet neither the advantage in relation 

 to fire, nor delfence against drought, can be regarded, I think, 

 as the forest's raisoii d'etre, though they may exi)lain its long 

 survival here as against possibly less efficient sub-tyijes of the 

 same formation that may have gone first in the drier area? ; 

 for what is essentially the saiue general type is found elsewhere 

 in the world with fires absent and a rainfall approaohinu- 20C- 

 inches. The fact is that, with vegetable reproduction what it 

 is, it is not crowding, but the converse that most needs accounting 

 for. and the latter is probably in mo-;t cases due to thinning 

 agencies or conditions of hardship. The sharp demarcation 

 'between veld and forest — the " solid wall " of Roose\'elt and 

 Heller, " definite line " of Belt, and " abrupt cliff " of my own 

 description (Jour. Linn. Soc. Bot. xl.) — is, in the places in 

 W'hich we see it to-day. man's handiwork, direct or indirect : but 

 the difference is in any case between two plant societies, each 

 of which normally supplies conditions that are intolerable to 

 the other, and make invasion and commingling impossible excep: 

 with outside assistance. 



Duration- — The various complicating factors I have re- 

 ferred to earlier make it futile to try to estimate the probable 

 duration of the period of destruction. A very crude estimate, 

 based on fire alone and on a higher average rate of destruction 

 than is observable here to-day in the case of unprotected forest- 

 patches, gave the figure of 105,000 years. Even if we should 

 admit the systematic destruction of forest for culti\'ation dur- 

 ing many centuries, the figure would still have to be great. Yet 

 man's association with extinct animals at Broken Hill suggests 

 that this need not, in itself, be a formidable objection to the 

 view that primary forest of the Chirinda type once covered the 

 stri]> from the mountains to the sea, and was removed chieflv 

 by the two agencies 1 ha\e suggested — especiall\- by, and. for 



