SAND-DUNES 23 



current. The delta of the Zambezi, if we count 

 it as beginning at the junction of the Quaqua 

 River, is nearly 100 miles long ; but it is rather 

 a remarkable fact that, unlike some other great 

 rivers of Africa, the Nile and Niger, for example, 

 the delta does not project into the sea, but is 

 tucked away in a bight. The same may be 

 said of the Rivers Lycungo, Buzi, Sabi, Lim- 

 popo, and Incomati. On the other hand, at 

 the two great shoulders of Mozambique and In- 

 hambane we find no larger rivers emerging. The 

 explanation of this may be that the curves of 

 the coast-line were at one time very much more 

 pronounced than they are at present, and that 

 the rivers, by the accumulation of silt, are over- 

 taking the flanking convexities, and will eventu- 

 ally succeed in straightening out the inequalities. 

 This explanation might serve to explain one 

 group of phenomena if it stood by itself, but it 

 does not explain the remarkable repetition, nor 

 does it help us to understand how the shoulders 

 come to be there, and why they are both dry. 



The true explanation is, I think, that while 

 the rivers have been busy prolonging their banks, 

 the forelands are being built up even faster. It is 

 a race oceanwards, and the forelands are winning. 



The littoral of the Mozambique coast is in the 

 main of sand-dune formation. The prevailing 



