298 



NATURE 



[May 6, 1920 



such agency, is worth reconsidering. Certainly 

 the idea that the river has worked back along 

 joints requires some modification, because the 

 depth of the crack, reckoning from the top of 

 the Falls to the bottom of the gorge, is more than 



1 100 ft., and joints do not penetrate so deep. 



Two rivers fan out on the floor of the depression 

 of the greater Ngami, in the same way that the 

 Rusisi does on the plain on the north of Tangan- 

 yika, which has been exposed since the Lukuga 

 tapped the lake and drew off the water ; this last 

 case, according to Arab accounts, has occurred 

 within the last 500 years or less. The Ngami 

 feeders are the Okavango and Chobe rivers. The 

 Chobe flowed south in Chapman's time (1852), but 

 the channel became blocked with reeds and rubbish 

 below the Mababe swamp, and it now goes 

 straight into the Zambezi. The Okavango until 

 quite recently also flowed south into the Ngami 



of Livingstone, but a branch, 



the Selinda, has developed 



which takes the water into the 



Chobe and so into the Zam- 

 bezi, and Ngami is now dry. 



When the Chobe and Oka- 

 vango rivers flowed south to 



Lake Ngami the water over- 

 flowed from that lake into the 



Botletle, which breached the 



eastern wall of the depression, 



and so made its way to the 



Makarikari. This depression 



has an area of 15,000 square 



miles and two "floors"; the 



Soa and Ntwetwe Pans form 



the lowest levels, while around 



are immense grass flats. We 



know fairly definitely that this 



dried up about 1820, thirty 



years before Chapman was 



there, and the Bushmen de- 

 scribed to him how the whole 



expanse was then covered 



with dead hippopotamus and fish. 



Now the Botletle very seldom 



reaches the Makarikari, though the floors may 



fill for a few weeks from drainage from the east. 



When the waters of the upper Zambezi were 

 impounded in the Ngami depression, the water 

 flowed south from the Makarikari into the Letwayo 

 or Okwa, and found its way into the Molopo and 

 so to the Orange River. According to traders 

 who have crossed this part many times, the old 

 channel can still be traced ; certainly the lower 

 Molopo has a bed far greater than would have 

 been cut had it only carried the waters from the 

 tributaries now shown to connect with it. The 

 region between the Makarikari and the bend of 

 the Molopo is the "Great Thirst"; the main 

 routes through the Kalahari are now fairly safe 

 and the Government has put down bore-holes for 

 the accommodation of travellers, but it is still 

 exceedingly difficult to explore away from the main 

 tracks. The natural slope of the plain of the 

 NO. 2636, VOL. 105] 



Kalahari is to the S.S.W., so that there wa;, 

 nothing to prevent the original river taking 

 the course indicated; when the water was 

 diverted to the Zambezi the area became a waste 

 of sand. The French in the western Sahara have 

 similarly shown that the ergs or sand deserts oc- 

 cupy the basins of former river systems. 



Ovamboland is just a great level river plain, 

 the ideal peneplain. Every part of it is covered 

 with shallow depressions, sometimes connected, 

 forming rivers; at others they are in parallel 

 series of disconnected hollows, elongated in the 

 direction of the nearest river. In between are 

 sandy tracts covered with forest. In this there 

 are a number of wide, open tracts, which form 

 the main habitable areas, and each of these open- 

 ings is occupied by a separate tribe of Ovambos. 

 On the north there are the Cunene and Okavango 

 rivers, which have built themselves above the 



Fig. 2.— Rucana Cleft, Cunene River. 



plain, and in flood-time they overflow their banks 

 and send the water down spillways which, in 

 former years, filled up all the depressions and 

 converted the country into a swamp ; something 

 like a third of the plain was then submerged. As 

 the flood subsides, the crops, which are planted on 

 series of little sand-hillocks, rapidly come to 

 maturity in the damp, steaming atmosphere, the 

 palms and morula-trees yield wine, and the lot of 

 the Ovambo was then a pleasant one. 



The spillways from the Okavango still carry 

 water out from the river southwards, but not in 

 sufficient quantity to reach any distance ; and those 

 from the Cunene are quickly diminishing. The 

 rapid lowering of the beds oif the main rivers 

 by erosion has resulted in the desiccation of the 

 country, and at no very distant date Ovamboland 

 will become a land of the "Great Thirst" like 

 the Kalahari. The conversion of an area of 



