J74 THE DESICCATION OF AFRICA. 



tell me that there are still great forest trees standing along 

 the course of this river, though they look as if they had been 

 dead for centuries, and at one place there is a great debris heap 

 of boulders, sand and trunks of trees in the bed of this old 

 river, which may have been the prime cause of its diversion. 



Before the Zambesi drained the Ngami basin, the Kafue 

 was not diverted into its present course, but made straight 

 for the Ngami depression, instead of, as at present, turning 

 abruptly at right angles and joining the Zambesi below the 

 Falls. 



The wear and tear of the rock-lip of the Victoria Falls is 

 so tremendous that every year the level is sinking and more 

 water is being drawn from the Chobe-Okavango system. 

 Livingstone noticed evident deepening of the river bed above 

 Sesheke ("Missionary Travels," p. 216). The Chobe Swamp 

 will soon be drained, though at the present moment (May, 191 8) 

 owing to unprecedented rains it is a vast inland sea, the old 

 channels to the Ngami are sanded up and the water cannot 

 escape fast enough over the Falls. What is to be feared is 

 that, after such a flood as has occurred this year, the water 

 from the Okavango will find a straight course by the Selinda 

 branch to the Zambesi and the Ngami Lake will disappear as 

 completely as the Kumado Lake in the Makarikari. 



To complete the preservation of what is left of the river 

 system of the Kalahari and to prevent desert conditions from 

 spreading beyond the Kalahari, as they obviously are at the 

 present day, it will be necessary to weir up the Chobe River, 

 between the swamps and its confluence with the Zambesi, to 

 prevent the water that is turned into the system by the pro- 

 posed weir on the Cunene River from escaping down the 

 Zambesi, and increasing, perhaps, the beauty of Mosioatunya, 

 as the natives call the Victoria Falls, but doing no good to the 

 country. The site for this weir is about thirty miles from the 

 junction of the Chobe with the Zambesi, three miles east of the 

 village of Ngoma; here there is high ground on both sides of 

 the river. Lower down, at the Sebuba Rapids, seven miles 

 from the confluence, there is a bar of hard rock crossing the 

 river, but there is only high ground on the south; on the north 

 there is a dead level plain, flooded by the Zambesi. The same 

 bar of rock crosses the Zambesi, and forms the Alambova 

 Rapids. The idea of the weir is not to form an absolute bar- 

 rier to the outflow of the waters, but to provide an impediment, 

 which will make the water pound up against it, and allow the 

 water from the Okavango, coming straight to the Ngami, to 

 clear the old channels of the sand which now blocks their beds. 

 The weir, in such tremendous floods as have occurred this year, 

 would be entirely submerged, but gradually, as the old chan- 

 nels are scoured out, the water would flow into the Makarikari, 



