GORGES AND RAVINES. 167 



again comes to daylight, blocks and fragments of 

 sandstone have been observed. 



The previously described lakes of Clare, which have 

 subterranean outlets, act somewhat similarly ; for if 

 rain comes on after a drought, they rise at first 

 rapidly, while subsequently, although more water 

 may be flowing into them, their surface will be lower, 

 showing that at the first the passage was stopped by a 

 deposition of mud or marl, which the rush of water 

 had first to dislodge, and then carry away, before ob- 

 taining a free passage for itself to run off. 



Similar reasoning may be applied to account for 

 the present condition of Lake Ashaugi, as all detritus 

 washed into the lake may be disturbed and carried 

 away during sudden floods. It may be objected, that 

 if detritus can be thus carried away through sub- 

 terranean passages, it is quite unnecessary that the 

 gorges and lake-basins of Abyssinia should be 

 supposed to be due to surface shrinkage, as all the 

 detritus denuded out of them may have been thus 

 disposed of. But let us not slight the significant 

 facts that the basin lies at the junction of a number 

 of valleys; that as "no ice ever existed in the 

 country,'' 1 there was no denudant capable of exca-' 

 vating out the lake-basin ; that as " there is a sub- 

 terranean passage from the lake," 2 there must be a 

 fault or break connected with its basin; and that 



1 Blanford. 2 Ibid. 



