IN CENTRAL AFRICA. '509 
Suppose^ however, it were determined, notwith- 
standing all these reasons to withdraw Timbuctoo farther 
to the east by a degree or more, it would then become 
necessary greatly to increase the journeys from Kakondy 
to Time. In making them three English miles and 
three fifths an hour, the true measure has probably been 
somewhat exceeded, but the situation of Timbo com- 
pelled this 3 while, to place Timbuctoo on the third 
degree of west longitude, would be to suppose a march 
of more than four miles and a half an hour. 
Another circumstance here comes to our assistance : 
it is that by carrying it farther towards the east, we must 
suppose the traveller greatly and constantly deceived in 
a western direction on his journey from Time to Tim- 
buctoo, and, on the contrary, in going from Timbuctoo to 
Fez, he must have made a similar and not less conside- 
derable error in all his bearings towards the east. How 
should we account for this singularity ? 
From all these considerations, could I risk changing 
the result of M. Caillie's itinerary and altering the con- 
struction of his route ? could I, in short, without sufficient 
motive, give the preference to any one map over another ? 
Some judgment may be formed upon this question from 
the diversity of situations assigned to Timbuctoo by 
different authorities : 
