238 PLANT SUCCESSION. 



V. Migration from the Valleys. 



Migration from the bases of colonisation appears to have taken 

 place principally in an eastward direction. This is particularly 

 clear in regard to migration from the "central valley," since, as 

 previously mentioned, the veld west of this is practically devoid 

 of trees except for a few clumps near the head of the valley. With 

 regard to these it is quite feasible to assume that they have been 

 formed by eastward migration from "valley A" rather than from 

 the "central valley" itself. That migration from "valley B" has 

 also been in an eastward direction is shown by the facts that though 

 clumps of trees occur on both sides of the valley they are larger, 

 more numerous, and extend to a much greater distance on to the 

 hillside on the east. 



The reason for this eastward migration is somewhat difficult 

 to determine. That it should be due to prevalent winds appears 

 improbable since most of the species which occur in the open, and 

 notably Ciissonia spicata, have their seeds distributed by birds. 

 The principal exception in this respect is Gombretum kraussii, 

 whose seeds are wind distributed. A closer study of the migration 

 from the "central valley" seems to throw some light on the 

 problem. 



On the eastern side two fairly well-marked depressions pass- 

 into this valley, which itself branches into two just above the 

 spring, one branch passing towards the east, the other towards 

 the west. Surveying the whole area from a vantage point near 

 the hill top, it is clear that migration has proceeded along these 

 tributary depressions and through the branches at the head of the 

 vallev. Three avenues of clumps can be traced eastwards from 

 the valley, one along each of the depressions, and a third through 

 the eastern branch at the head. A short line of clumps has also 

 been formed along the western branch. These features of the 

 migration are illustrated by the accompanying map. A line of 

 clumps has also been formed through the head of "valley A," and 

 passes above those from the "central valley." The absence of any 

 other lines of migration from this "valley A," which contains the 

 same trees as the central valley though fewer of them, and the 

 absence of any tributary depressions passing into it is suggestive. 



Migration from "valley B" has not proceeded along quite such 

 definite lines as from the "central valley," probably because there 

 are no clearly marked depressions on the eastern side. This side, 

 however, is not nearly so steep as, and is much moister than, the 

 other. In fact, the eastern side is quite swampy in parts and a 

 small stream flows down it into the main stream. 



These observations suggest an explanation along the following 

 lines: — The particular type of Tree Veld under consideration is 

 much more mesophytic than the commoner Thorn Veld (as indeed 

 is shown by its composition), and is less able to withstand grass 

 fires. At the same time the trees composing it are strong light 

 demanders. Ordinarily they are confined to shallow valleys, where 

 they obtain plenty of moisture, and are sheltered from the full 



