2 12 Journal of Travel and Natural History 



overflowing streams, meeting here, flood the Mahabe flats for many 

 miles eastward, forming a sheet of water nearly twenty miles in 

 breadth, where at other seasons not a drop can be found. At 

 such times the river is navigable from the lake to Sebetoanes, and 

 one might then travel in a canoe from Chapos at the terminus of 

 the Botletlie to the mouth of the Zambesi, on the east coast, or for 

 several hundred miles northwest of the lake from a very long way 

 beyond Lebebes in the same direction. And if, as the natives 

 assert, the Teouge branches off from another river beyond Lebebes 

 flowing to the west coast, then it would appear that the continent 

 of Africa is probably navigable for boats right across from east to 

 west — (vol. i. p. 184). 



A similar difticulty in defining the exact line of the water shed 

 occurs in other parts of Africa, as between Lake Tchad and the 

 Atlantic, and probably also more on the eastern side of the continent. 



With regard to the numerous salt pans which mark the hollows 

 in the country around Lake N 'Garni we learn that some of them 

 are of great size — that of Ntwetwe is 18 miles broad and upwards 

 of ICO miles in length, and when in the middle of it the effect was 

 that of being surrounded by a broad expanse of a calm and white 

 ocean. To what cause are we to attribute these saline crusts or 

 efflorescence ? Does the salt proceed from beds of salt in older 

 deposits ? Are they the last dried parts of a salt sea ? Do they pro- 

 ceed from the evaporation of fresh-water lakes, or are they the pro- 

 duct of salt springs. Mr Chapman's description rather points to 

 the latter two causes. He says : 



" The underlying mud of these pans is an unctuous, tenacious, substance very 

 like cement, and a hard greenish honey-combed cavernulous or vermiculated 

 sandstone (?) lies scattered at intervals. In some of the smaller outside pans a 

 hard white crust of limestone is found on the surface of the soil, which, having 

 been broken by the hoofs of game, lies scattered aroimd like flat pieces of ivory. 



The springs on the north side of these pans have generally a bank of tuft 

 which those south of the Botletlie, when they have a distinct bank, have it on 

 the south side. Some of the springs are no more than little pits dug out of the 

 bottom of sloping limestone hollows or ponds by the aid of a bushman's spade 

 and a sharpened stick. Some of these ponds are broad and shallow, without 

 any bank, and the surface is covered with loose shingle, while others are an 

 irregular or more often a rounded fissure in limestone tufa, with two or three 

 successive layers in the bank underneath. The pits or wells are generally filled 

 with small rounded shingle, while outside is more generally a slope by which 

 men and animals descend to the water. 1 do not think that the game has 

 broken the banks to that extent, but they would naluially aiiproacli the water 

 on the most accessible side" — (vol. ii., p. 61). 



