216 ACROSS AFRICA. [Chap. 



May, and only give us a little, and make ns work for it." His pro- 



1874. posal was tliat when I returned I should pay the peoj)le who 



worked daily, and then they would understand. He said he 



wished a trade-road passed by his village, to bring traders there. 



After pulling an hour and a half, the breeze freshened up 

 almost in our teeth; so I put into a convenient little inlet, 

 which I discovered to be part of the other river. It was all 

 swamp, marsh, or low, flat plains inside a long bank, with some 

 small openings ; deep water in places, shoals, sand-banks, long- 

 grass, etc. 



I suppose the drift matter of the lake, W'hich gravitates to- 

 ward this outlet, forms the banks and morass, owing to the 

 want of a passage for it. A fair instance of this was given dur- 

 ing the seven or eight hours we were here, a large quantity of 

 drift-wood having come in and worked away into the grass, 

 without leaving any sign of its passage. The inlet in which we 

 lay was only a break in the bank, and the water works through 

 the grass into the Lukuga. 



I entertained strong hopes of being enabled to undertake the 

 work I so much desired, of tracing the course of the Lukuga. 

 But at Ujiji not a guide or interpreter could be obtained for 

 that route, and not a man would follow me alone. And w^hen 

 I began to estimate the cost of cutting a channel through the 

 > grass and of buying canoes, I found the necessary expenditure 



*so heavy that I confess I did not feel myself justified in incur- 

 ring it. For I firmly believed that the stream was too consid- 

 erable to be lost in marshes, or be merely a backwater. I had 

 also the word of the chief, who accompanied me on entering the 

 river, that his people had traveled for months along its banks. 



The entrance is situated in the only break in the hills that 

 surround the lake, the mountains of Ugoma ending abruptly 

 ten or twelve miles north of Kasenge, while those that en- 

 circle the southern end trend away to the Avestward from Has 

 Mirrumbi, leaving a large undulating valley between the t\vo 

 ranges. 



When passing doAvn south on the eastern shore of the lake, 

 near Kas Kungwe, the guides pointed to this gap in the mount- 

 ains, and asserted that there the outlet of the lake was to be 

 found. At various points on my journey afterward I obtained 



