THEORIES RESPECTING THE NIGER. 407 
truncated, and carried up only as high as they 
had been ascended by Europeans. He also re- 
moves Tombuctoo eastward to its true distance 
from the sea. He retains the lake of Guardia, 
and endeavours to form a most extraordinary al- 
liance between it and the system of the Arabian 
geographers. He makes it stand for all the lakes 
described by them in the different parts of Nigri- 
tia. On the east side of it he places the city of 
Gana, and round it, all the cities of Wangara, 
which became thus west of Gana, instead of seve- 
ral hundred miles east. But in the map of the 
world, on a polar projection, published in 1714, 
though composed with a different object, he alters 
entirely his construction of this part of Africa. 
The Niger and Senegal are there represented as 
separate rivers ; are made to arise from two lakes 
near to each other, and to flow, the one east and 
the other west. The lake of Guardia is obliterat- 
ed, and the eastern part of Nigritia delineated ac- 
cording to the data of Edrisi. 
This reform was followed up by D' Anville, who, 
in 17 05 , communicated to the French Academy a 
treatise "On the Rivers in the Interior of Africa."* 
Here, instead of the single stream of the Niger 
rolling across nearly the whole breadth of Africa, 
he distinguishes three rivers. 1st, The Senegal, 
* Academie des Inscriptions, Vol. XXVI. 
