PHYSICAL OEOGBAPHT OF ABYSSINIA. 159 



terranean outlet, for two reasons. The first is, that 

 although the army was on its banks in April, at the 

 close of the driest season in Abyssinia, there was no 

 appearance of the lake having been at any period of the 

 year more than a few feet higher than it then was. 

 Situated as it is, almost at the edge of the steep descent 

 to the eastward, with high hills to the west and north- 

 west, the rainfall must be very heavy ; and it is difficult 

 to conceive that in the months of July, August, and 

 September it can be less than eighty inches, which is 

 sufficient to increase the waters of the lake by thirty or 

 forty feet.'- So far as it was possible to judge, there was 

 no evidence of the water having been more than a third 

 of that amount above the lowest level. It is extremely 

 improbable that the rainfall could be evaporated within 

 the year from the surface of the lake. 



The second reason is the perfect freshness of the water. 

 This alone appears conclusive as to the existence of an 

 outlet, for otherwise the water must gradually become 

 charged with salts. A similar case occurs in Berar, in 

 India, at the little lake of Lonar. This lake is a little 

 smaller, but its drainage area is very much less than that 

 of Ashangi, while the ground drained consists of pre- 

 cisely similar rocks. Lonar Lake, however, instead of 



1 The rainfall in places similarly situated in India varies from about 100 

 to 200 inches. Taking the drainage area of the lake as a circle with a radius 

 of 5 miles, or nearly 80 square mUes, and estimating the area of the lake 

 itself as 12 square miles, we have an increase in water for a rainfall of 80 

 inches of sfi-i-so = 533 inches, or 44 feet 5 inches. Of this, if 10 feet, 

 an excessive estimate at this elevation, and in so damp a climate, be allowed 

 for evaporation, the amount of increase would still be more than 34 feet. I 

 believe this estimate to be below the truth. 



