THEORIES RESPECTING THE NIGER* 401 
continent ; and "whither it goes is quite uncertain. 
It would be difficult to express, in more accurate 
terms, the actual state of our knowledge in respect 
to the Niger. Mela mentions then as a plausible 
conjecture, that this river may end in being the 
Nile ; but he positively rejects the hypothesis of 
its sinking under ground, very reasonably ascribing 
such a report to the long tract of unknown terri- 
tory through which it flows, and where, not being 
seen, it is supposed not to exist. 
The Roman writers appear, from the sources 
above mentioned, to have derived somewhat more 
extended ideas with regard to Western Africa, than 
were attained by the Greeks. Ptolemy, however, 
generally speaking, obtained a much more correct 
view of the courses of the African rivers, than any 
of his predecessors. He, first among the ancients, 
rejects wholly the idea of the identity of the Nile 
and Niger, lays down the sources of the former in 
their true position, and exhibits the latter as a dis- 
tinct and separate stream. If we are asked, how- 
ever, in what direction he represents it as flowing, 
the question is exceedingly difficult to answer. The 
following extract includes all he communicates on 
the subject. After remarking, that " in the inte- 
" rior flow mighty rivers, the Gir and Niger," he 
proceeds to describe the latter by his usual mathe- 
matical mode of enumeration. 
" And the Niger, which joins together the 
mountains Mandrus and Thala. It also forms the 
VOL. II. c c 
