kI':i'L.\cI':.mI':nt of iiii': ANcii'LiXT !■:. ai-rjcan i'()Ki':s'i 515 



l)iU ill \icw (if all I lia\c jusi said, I do not feel that il need be 

 regretted if it should be ])ossibte l<> eoiuert an ijeeasional liill- 

 sidc of it into useful forest. 



'The \alue. as nin-se-trees, of a dr\in,i^- ty])e of \voodiii<i' may 

 well l)e (|tiestioned. Vet the ol)servations I have given suggest 

 that it mav serve, and the ini])ression left with the observer is 

 that it onlv requires the eontinued exelusion of fires and an 

 abundant seed-sowing to extend over the whole s]oi)e I have 

 mentioned the oj^eration that is already taking i)laee in the 

 kloofs, and on part of the hilltoj). Obviously the growth 

 obtained from sueh planting eould not be ex]:)ected to be so even 

 or the returns at all so early as from ploughed ground, and it 

 would be worth attempting only in a case like the above, in whieli 

 not early returns, but the rescue of s])rings at as cheap a cost as 

 ])ossible was the primary consideration. 



Albi::sia or Er\tlirliia or some other useful tree might be used 

 as a shade connecter and wherever natural wooding was absent, 

 being either sown at stake ov put in as live posts; and a 

 later sowing made at stake, at stations roughly hoed in the 

 urass, with the better native fc^rest trees as (in these parts) 

 Khaya and Pxgciim. To get its full moisture-conserving effect 

 from this tyi)e of forest, and to guard it against fire, one would 

 doubtless have to follow nature and include some of the shrubs, 

 if not of the lower-canopy trees, in one's sowing, and later, in 

 order to bind the humus on the steeper slopes and add a shade- 

 layer, to introduce fc;rest-grasses at intervals. They would soon 

 sj/read and meet. Actually, this forest never makes much 

 hunuis, though masses of leaves fall, and there being no white 

 ants and no fire to destroy it, one expects to see accumulation 

 in the course of years. On scraping oft' the top layer of un- 

 decomi)osed leaves, one finds such (juantities of small roots 

 feeding right at the surface that one's immediate im|)ression is 

 that the forest is eating- uji its humus as fast as it deposits it. 

 Pines would quite likely be better if Pin us lialefcnsis, 

 for instance, were found to succeed under the conditions, but 

 the Chirinda type seems very eft'ective in preventing wash and 

 in ])reserving coolness and moisture and a i:)erpetual flow from 

 its springs, though they sometimes get low in a drought. 



If I should be successful, jiersonally. in aiding the re- 

 aft'orestation, begun by nature, of the slope referred to, or the 

 forest-advance itsejf, in the manner I ha\e just suggested, it is 

 l)ossible that I mav yet oft'er the Association a i)aper on " The 

 ]'ie^l)lacement of W(^oded Pasture-land by High Forest!" 



/;/ conclusion, it may be said. 1 think, that this type of 

 forest, while i)rimarily dei:)endent on rainfall, has elastic require- 

 ments even in that respect. Miss Gibbs gives the range 61-158 

 inches for another siii>type of the same general formation, and 

 here we may ])ut the minimum as low. I think, as 40 inches: 

 ])Ossiblv a good deal lower were it not for the more wholesale 

 damage that would then be done by fires. Above a certain 

 figure. var\ing doitlvtless with \ariotis factors, an(l with no 



