liEPLACEMENt ()!• THE ANClENT E. AEKKAN EOi< ol 513 



Bersaiiia — they are entering into successful warfare with the 

 l)yrophytes of the lower slopes, while the big climbers Landol- 

 pliia, Clioi'isfxlis, Caiitliiiitu, Klioicissiis, Uvaria, Secamonc, and 

 Lygodiiiiii siibalatitiii are flinging themselves out over Bracliy- 

 stcyias and Uapacas, and already smothering some of them. 

 That the pyrophobes should be capable of growing up under 

 trees so exclusive and hard on the ground as the last two genera 

 has interested me immensely, and jjrompted the suggestion I 

 shall state below ; 'but our forest^trees are obviously good 

 fighters. It is possible that few trees have so drying an effect 

 on the ground as Eucalypts — they have quite dried out a small 

 swamp that used to provide me with Anopheles. Yet, in my 

 close-])lanted gum shelter-belts, forest trees and shrubs, as well 

 as several importations, have come up everywhere and ,uro\vn 

 freely. They include the fine trees Trieliilia chirindensis, 

 Eckehcrgia Meveri. Pygeitin africanum, Maba niiialala, Croton 

 sxlvaficitni, Sapiiuii Mannianiiin, and Celiis dioica. 



The prompt advance of forest that follows the cessation of 

 fires is an argument in favour of the latter as the agent of its 

 destruction. The circumstances under which the seeds of forest- 

 trees germinate and succeed outside the forest enajjlc one to 

 gauge their actual needs. I have seen seedlings that have 

 si)rung up in numbers on the outskirts and flourished in broad 

 sunshine, while the rain lasted, dry up and disajjpear as the dry 

 season i)rogressed. Llere the ground was bare of fallen leaves, 

 or practically so, and baked dry. On the other hand Kliaya, 

 Trichilia, Pvgeiiui, Raitwolfia, Teelea. Croton, EJiretia, Schefflero- 

 dcndron, and Landolphia sown in ploughed ground survived 

 their first dry season well without shade, the ifact that they 

 were covered up, together with the moisture retained by the 

 ]3loughed ground, and its ])enetrability, evidently fulfilling all 

 needs. To-day they form a well-grown grove. Albizaia 

 chiriiidensis, again, gives very little shade when one would ex- 

 pect it to be much wanted, for it drops its leaves in June and 

 early Jiflv. Yet it seems a successful nurse tree. There is always 

 a fair amount of good humus layer under it, and this, apparently, 

 is able alone to carry a fair ])ro])()rtion of the seedlings it shelters 

 through their first dry season. In general, I should say that one 

 never finds a forest-seedling surviving the latter when some 

 slight layer is not present, unless it finds tilled ground. The 

 seeds slip or are washed down Ijetween the undecomposed leaves, 

 and are sheltered by these ; and the layer generally, heli)ed by 

 shade when present, tends to retain much moisture between the 

 falls of rain, and also provides a jjenetrable bed for the seed- 

 ling's roots, with the result that it has begun to be established 

 by the time the dry season comes. If, now, the humus is such 

 as to ensure the continuance of a sufticient moisture supply, the 

 l)lant does not fear sunliglu. If not, it burns unless sliaded. 

 Tims moisture, ratlier than sliade, is the real necessity in their 

 first season, for, at any rate, a great many of the finer and com- 

 moner trees of this tyi)e of forest, though it is seldom that the 



