104 FEOM THE NIGER TO THE NILE 



The fact that a passage, used by the French in 1902 

 along the east coast to N'gigni on the north, was impassable 

 two years later when I was there, might be taken as evidence 

 that the Lake is drying up. But so swift a shrinkage as this 

 would imply is not possible to believe, and it is most prob- 

 able that there was more water than usual in the Lake in 

 1902 owing to flood. 



Another fact that has perhaps created the impression 

 that the Lake is decreasing is that chains of islands, that 

 once were separate, are now more or less joined together by 

 marsh. But I think that this may very likely be due to the 

 silting up of mud and sand against the obstruction of the 

 islands by the opposing influence of the Yo and Shari, the 

 two rivers that feed the Lake. In fact, my observations 

 go to show that the Lake is practically two lakes, divided 

 by the fifteen miles or so of marsh and maria bush that I 

 attempted to cut through, thus forming the separate basins 

 of the two rivers. Moreover, the Buduma Kachella told me 

 that there was no communication between the two parts, 

 and I found that the people on the different sides knew little 

 of each other. 



This theory is further borne out by the very marked 

 difference in the character of the scenery and of the people. 

 On the north, the shores are flat and bare and the surface of 

 the water, which is nowhere more than 4 ft. deep, is broken up 

 by innumerable small uninhabited islands that are little 

 more than sand-flats. The people are neither numerous nor 

 flourishing and lead a more or less piratical, lawless existence. 



But in the south, or Shari basin, everything^has a more 

 flourishing appearance. The depth of the water is from 



