chap. xvm.-J SIR R. MURCHISON'S THEORIES CONFIRMED. 483 
Tchad : thus the Victoria and the Albert lakes, being 
the two great reservoirs or sources of the Nile, may 
be the first of a system of African equatorial lakes 
fed by the northern and southern drainage of the 
mountain range, and supplying all the principal rivers 
of Africa from the great equatorial rainfall. The fact 
of the centre of Africa at the Nile sources being 
about 4,000 feet above the ocean, independently of 
high mountains rising from that level, suggests that the 
drainage of the Equator from the central and elevated 
portion must find its way to the lower level and reach 
the sea. Wherever high mountain ranges exist, there 
must also be depressions; these situated in an equa¬ 
torial rainfall must receive the drainage from the high 
lands and become lakes, the overflow of which must 
form the sources of rivers, precisely as exemplified in 
the sources of the Nile from the Victoria and the 
Albert lakes. 
The fact that Sir Roderick Murchison, as a geologist, 
laid down a theory of the existence of a chain of lakes 
upon an elevated plateau in Central Africa, which 
theory has been now, in great measure confirmed by 
actual inspection, induces me to quote an extract from 
his address at the anniversary meeting of the Royal 
Geographical Society, 23d May, 1864. In that address, 
he expressed opinions upon the geological structure, and 
the races of Central Africa, which preceded those that 
I formed when at the Albert lake. It is with intense 
interest that I have read the following extract since my 
return to England :— 
“ In former addresses, I suggested that the interior 
mass and central portions of Africa constituting a great 
plateau occupied by lakes and marshes, from which 
the waters escaped by cracks or depressions in the 
subtending older rocks, had been in that condition 
during an enormously long period. I have recently been 
enabled, through the apposite discovery of Dr. Kirk, 
the companion of Livingstone, not only to fortify my 
conjecture of 1852, but greatly to extend the inferences 
i I 2 
