﻿228 MESSRS. A. R. ANDREW AND T. E. G. BAILEY [May I9IO, 



and mountains. The river to the south of the Nyika is a mere 

 chain of marshes in the dry season. It gives the general impression 

 of a valley which has been silted up, owing to the failure of its 

 water-supply. Both the Kasitu and the Lower Southern Eukuru 

 have a remarkably low gradient, despite the fact that they traverse a 

 much broken tract of country and lie at an elevation of more than 

 3000 feet. Erosion, even during the rains, must be very slight in 

 the main river-valleys, for the rivers have as much as they can 

 do to keep their courses clear. On the other hand, the lateral 

 tributaries coming from the high ground have, perhaps, a sufficient 

 supply of water, and are of sufficient gradient to erode their 

 channels, and so in time to lower their drainage-areas. Even 

 where river-erosion is almost negligible in the hilly country, 

 weathering is in constant operation, and the detritus from the 

 hills is washed dow r n during the rains into the valleys. As a 

 result the high lands are being lowered, while the low-lying country 

 constituting the larger river-valleys is kept at a general level, 

 or even perhaps raised. If the above operations were allowed to 

 go on undisturbed, a subaerial peneplain at an altitude exceeding 

 3000 feet would at last be produced. 



We have suggested that the Kasitu and the Lower Southern 

 Eukuru Rivers have dwindled in size. This may perhaps be ex- 

 plained by beheading, or be attributed to climatic changes producing 

 desiccation. The former suggestion seems hardly applicable to the 

 Kasitu. The whole subject, however, requires more careful study 

 than we have been able to devote to it. The traveller crossing a 

 waterless tract in the dry season, with perhaps not a trickle or pool 

 to be seen in the stream-courses which he crosses on his journey, 

 might well imagine that the land was always dry and that river- 

 erosion had ceased to be. If, however, just before the rains, he 

 has been tempted by some shade-tree to place his tent close above 

 a dry stream-bed, he may shortly have occasion to change both his 

 camping-ground and his opinions. A dry stream after a few hours 

 of heavy rain may fill to the brim, 8 feet or more, the water, after 

 its first advent, rising with extraordinary rapidity. Another few 

 hours and the flood may be reduced to a muddy rivulet 3 or 4 inches 

 deep, providing, of course, that the neighbouring drainage-area has 

 not been thoroughly saturated by recent rains. 



It is possible that certain valleys which are only marshy during 

 the rains are valleys of intermittent erosion. That is to say, they 

 owe their formation in the first place, and their continuance in the 

 second, to occasional very wet seasons; therefore, the presence of 

 such valleys is not necessarily an argument in favour of the 

 desiccation of a country through climatic changes. 



It is a well-known fact that climate exerts an influence on vegeta- 

 tion and vice versa. By exerting a control over vegetation, man 

 also plays some part in varying climate. It is the custom among 

 the natives to fire the country shortly before the rains set in. 

 A fire, once started, may continue for miles into the neighbouring 

 forests. These fires do but little damage to the forest, -as the bulk 



