116 



Eragrostis), and it is for each farmer to decide whether it is 

 worth his while to leave the grass unburnt so as to give it a 

 chance to progress towards the Anthistiria stage of the suc- 

 cession, which is undoubtedly of a superior type, and at the 

 same time to lessen soil erosion, or to continue to make use- 

 of the harder grasses, by continually burning them to obtain 

 young shoots. 



Stable grassland. Climax stages. 



In the subsequent stages of development a large number 

 of sj)ecies take part, the majority belonging to the tribe 

 Andropogoneae, which includes Anthistiria iinberbis (inSinde 

 or Red Grass), the species which is usually dominant. When 

 we compare these subsequent species with the pioneers, we 

 notice many points of difference. The chief characters of the 

 latter have already been considered. The former are not so 

 xerophytic, nor so deep rooted. They form a closer mat over 

 the surface of the soil, and thus prevent the run-off of water 

 to a large extent. Their shallow spreading root systems obtain 

 moisture first, before it is able to soak down to the lower 

 levels, and this is partly why they are able to oust the- 

 pioneers. Light, however, is another factor of considerable 

 importance. The seedlings of Anthistiria, for instance, prefer 

 to germinate in the shade of other herbage. It is particularly 

 interesting to notice in mixed transitional grassland, how 

 young plants of Anthistiria almost invariably come up through 

 the middle of a tuft of Aristida, which gradually becomes 

 smothered as the Anthistiria grows taller and shades it, and 

 also forms a mass of roots immediately above those of Aristida. 

 Aristida seedlings themselves, on the other hand, do not seem' 

 to tolerate shade. They usually germinate in the open spaces. 

 The pioneer species therefore prepare the way — and it is a* 

 necessary preparation- — for the establishment of the climax 

 types. 



Over by far the largest areas of Eastern grassland Anthis- 

 tiria imherbis (inSinde) is the dominant species. It covers 

 enormous areas in the Eastern Province of the Cape, Natal, 

 the Northern and Eastern Free State, and the Transvaal. 

 There are several varieties, which are not always easy to dis- 

 tinguish. A closer study of the species would repay attention, 

 and ought to be undertaken. Stapf has distinguished three- 

 varieties, viz., mollico'tna with leaves and involucral spikelets 

 densely hairy, argentea with leaves densely hairy, but invo- 

 lucral spikelets glabrous or scantily hairy, and hvrchellii withe 



