4^2 THEORIES RESPECTING THE NIGER. 
of which the mouths only are known to us. This 
can rank only as a mere conjecture, which it may 
be difficult to disprove, but which is supported by 
no evidence. The author, indeed, endeavours to 
shew, by a comparative calculation of the waters 
poured into Wangara, and of the extent of its 
lakes, that the former cannot be disposed of by 
mere evaporation from the latter. Without con- 
sidering all the circumstances which must render 
such an estimate precarious, we may observe, that 
he brings into Wangara the great western rivers 
of the Misselad and Wed-el-Gazel, not only with- 
out any proof, but contrary to all the evidence 
that exists. His calculation from the lakes of 
Wangara, too, is built entirely upon the space 
which they occupy in modern maps. But there 
is nothing in the Arabian writers (our only au- 
thority) which can give the remotest idea of 
their magnitude. What he has proceeded upon, 
is the mere arbitrary delineation of modern geo- 
graphers, who, having lakes to delineate, were 
obliged to make them of some size or other. 
Upon the whole, however, if the Niger should 
reach the sea, it does not seem impossible that it 
may be by this channel. 
The next hypothesis is that famous one by 
which the Niger is identified with the great stream 
which passes through the kingdom of Congo. 
