IN CENTRAL AFRICA. 343 
time may^ indeed^ have effected actual changes in the 
importance and population of towns ; we must not, for 
instance, be surprized at the difference between the de- 
scription of Leo and that given at the present day.* 
Who was there but believed, only a short time since, 
that Timbuctoo was equal in extent to one of our large 
cities, and contained within its walls a population of a 
* The king of Tombuto, according to Leo, had three thousand 
horsemen under his command, and an almost innumerable multitude of 
archers. One very remarkable circumstance is that fire destroyed, 
according to this author, (if we read oppidi pars) almost half the city 
in the space of Jive hours. Leo's recital is the most important, and 
perhaps the most authentic, of any we have hitherto possessed. These 
motives induce me to place it entire before the reader, in order that he 
may compare it with the new narrative. I shall quote the Latin 
version made from the Italian of Leo (2nd edition), though it is not 
quite so accurate as might be wished : it is well known that Leo wrote 
it originally in Arabic. See Joannis Leonis Africani de totius Africse 
Descriptione lib^ Tiguri, 1559. 431 and following pages. 
TUMBUTUM REGNUM. 
Hujus regni nomen nostris fere temporibus ab ejus- 
dem nominis oppido desumptum volunt, cujus conditorem 
fuisse dicunt quemdam Mense-Suleiman, hegirfe anno sex- 
centesimo decimo : in duodecimo milliario a quodam 
fluviolo situm fuit, quod e Nigro flumine effluebat; cujus 
domus omnes in tuguriola cretaceaf stramineis tectis sunt 
t These words are not a translation of capamefatte dipali, coperte 
di creta, in the Italian text. 
