234 
the Sahhara be a repetition on a great scale qf this same 
phenomenon, as there is every reason to suppose; far 
from being composed of the humus depauperatus of 
Linneus, it is only a surface of sand abandoned by the 
sea, like that of Mogador and Tangier, and which 
never has been fit for vegetation. 
This conjecture is almost converted into a certainty, 
when we consider the small elevation of the Sahhara 
above the level of the sea. We see several rivers, as 
the Wad Drah, the Wad Taffilet, and others, come 
from the southern part of mount Atlas, and lose them- 
selves in the desert without being able to proceed for 
want of declivity. 
The Senegal and Gambia rivers precipitate themselves 
from the neighbouring mountains of Kong, in a north- 
erly and north-westerly direction; the first running to- 
wards the borders of the Sahhara, and the latter into 
another large plain. They make here a sudden devia- 
tion towards the west, and after a thousand windings, 
like those of the Meander in Asia Minor, they reach 
the sea by an unpereeivable declivity, forming number- 
less small islands in their course; because the falling of 
a tree, or any obstruction is sufficient to divert or divide 
their feeble current. i 
These circumstances seem to prove, that when the 
mountains of Kong formed an island, these large rivers 
precipitated themselves into the sea of the Sahhara, and 
that since this sea has been filled up by the sands gradual- 
ly accumulating; those rivers have directed their course 
towards the ocean, as the amassing sand successively 
forced them to deviate from their first direction. The 
current being weak, the least obstacle was sufficient to 
turn them aside, as in ourtim^s happens to the Stnegaj 
