$9^ GEOGRAPHICAL SYSTEMS* 
reached that destination without having encoun- 
tered any portion of the great African desert ; 
the extent Of which, it is evident, was wholly un 
known to Ptolemy. His informants could fur- 
nish no astronomical observations, nor accurate 
data of any kind, by which to fix the position of 
the countries through which they travelled. In 
such cases, the ancient geographers, to gratify the 
natural love of completeness, had recourse often 
to very arbitrary delineations. Instances have 
occurred, and will again occur, in which very re- 
mote objects were brought into contact. It ap- 
pears to me then, that Ptolemy has linked toge- 
ther the geographical features of central and 
northern Africa ; that in the Gir he linked the 
head of the Adjidi to the head of the Misselad ; 
and, though no part of the Niger can be referred 
to the north, its tributaries from that quarter are 
probably the rivers of Sigilmessa, hypothetically 
united to it. I doubt, however, if the Daradus 
and Staichir can be viewed as the Gambia and 
Senegal. Every remote space actually traversed 
is exaggerated, instead of being so remarkably 
diminished, as, in the present instance, that be- 
tween Mount Mandrus and the sea. Besides, 
from the Fortunate Islands, or Canaries, being 
placed opposite to these rivers, it seems clear^ 
however far Carthaginian discovery may have 
penetrated, that Ptolemy's knowledge of the 
