166 Mr Murray? on the Course of the River Niger. ’ 
It may be urged, indeed, as such a separation is not absolute* 
ly impossible, that, if supported by positive testimony, its exis- 
tence cannot be rejected. But we must here remark certain de- 
fects in that mode of evidence upon which Mr Bowdich pro- 
ceeded, so far as it is applied to ascertain the course and direc- 
tion of rivers. It was derived entirely from land travellers, to 
whom the direction of the stream is never an object of any im- 
portance. In tropical countries it is not even very observable, 
unless during the rainy season, when travelling is rarely prac- 
tised. To a man placed at the confluence of two or more 
streams, there is even a tendency to use language directly the re- 
verse of the real fact ; imagination naturally suggesting them 
as branches issuing from a trunk. Thus, a Jenne merchant, quot- 
ed by the editor of Adams, p. 197, says : La separation des 
deux rivieres, est a une demi lieue de Genne, et Genne se 
trouve entre les deux rivieres comme une isle. Une de ces ri- 
vieres court dans le Bambarra, et fautre va k Betoo.” These 
two rivers are evidently the Joliba and Ba Nimma of Park, 
which flow to Jenne and not from it ; and the informant was 
probably aware that they did so, but was led by the analogy 
above mentioned to use this language. Instances of similar forms 
of expression are by no means uncommon, even in European 
writers. 
The mere consideration of these circumstances, seems to ren- 
der it much more probable that these rivers are tributaries fall- 
ing into the Niger, than branches issuing out of it. There is 
not wanting positive testimony to the same effect. The Gamba- 
roo being, according to Mr Bowdich, the river which passes 
close by Tombuctoo, must be Mar Zarah of Adams, who, 
though not quite positive on the point, yet, in opposition to his 
examiners, obstinately stated a preponderating belief,*” that 
this river flowed to the south-west^. The Gambaroo must also 
* Adams’ testimony has been strongly controverted by statements from Ame- 
rica, which undertake to prove that he never could have been at Tombuctoo, (see 
Edinburgh Magazine for October 1818). It is much corroborated, however, by 
Mr Bowdich, who confirms the name of the river Zah mer (evidently the same as 
Mar Zarah), and the reign of Woollo and Fatima or Fatoilma, as king and queen 
©f Tombuctoo. These names were quite unknown in Europe when Adams gave 
his testimony, so that if he never was at Tombuctoo, he must have at least have 
had some good original information respecting that city. 
