176 THE DESICCATION OF AFRICA. 



The Loangwe River, for instance, which was once on the Ngami 

 level, now enters the Zambesi below the Victoria Falls, at an 

 elevation of 774 feet, whereas Lake Ngami is 3,117 feet and 

 the Soa Pan 2,954 feet above sea-level. Kasungula, at the 

 confluence of the Chobe and Zambesi, is about 30 feet below 

 Ngami ; the site I have indicated as suitable for a weir cannot 

 be much more than 10 feet below Ngami; in fact, Chapman 

 found the Chobe flowing up the Talmalukan towards Ngami in 

 1853.* As we have taken the Soa Pan to have been on the 

 course of the Central South African river, the Proto-Orange, 

 the Zambesi must have cut downwards 2,161 feet since it cap- 

 tured the Loangwe. This seems a vast amount, but when one 

 regards the gorges of double and treble the depth, cut by our 

 small rivers in Cape Colony, through the coastal rampart, it 

 docs not seem so great after all. 



South of Andersson's Vley, all trace of a river system 

 becomes lost, and what rain does fall collects in isolated, 

 shallow pans. We And, however, on the south-east the Kaap 

 Plateau draining into the Kalahari, by stream-beds of con- 

 siderable size, though now dry. All these are directed to the 

 north-west and some of them connect with the central portion 

 of the Molopo River. The present commencement of the 

 Molopo near Mafeking is a stream that often flows, but the 

 water does not get many miles west of the town; the river, 

 after proceeding in a westerly course, turns north-west, then 

 due west, and then due south. At this last bend, marked 

 Mokopon, I have placed the junction of the Proto-Orange with 

 the present bend of the Molopo, as the most probable place, 

 making, thus, the Molopo a tributary of this river above the 

 bend ; below this, the Molopo would be the bed of the original 

 principal stream. According to the relief map in L. Schultze's 

 book on Namaland and the Kalahari (1907), the main river 

 may have joined the Nosob first, the Molopo entering it lower 

 down. At any rate, the natural inclination of the land would 

 lead the water in the direction I have stated, and the Makari- 

 kari Lake, if refilled, would not flow north or east through the 

 Kwai or Macloutsie Poorts. The Kalahari sand has obliterated 

 the more northerly portion, between Mokopon and the Makari- 

 kari, just as the Saharan sand has obliterated the middle 

 portion of the Igharghar. 



On the west of the Kalahari, we have the highlands of the 

 coastal rampart, and these drain by several very large, dry 

 rivers into a central channel, the Molopo; they all run in a 

 south-easterly direction, and the chief one is the Nosob. 



^Between the site of the weir and Kasungala there are the Sebuba. 

 rapids (Chapman's "Travels," 1868, vol. I., p. 184). 



