On the morning of the 31st we sent a canoe with men to search for food in the two or three villages that were visible on the other side.  Four doti purchased just sufficient for four days for our caravan of forty-eight persons.  We then got under weigh, having informed the kirangozi that Urimba was our destination, and bidding him keep as closely as possible to the lake shore, where it was practicable, but if not, to make the best he could of it.  From the debouchement of the Rugufu, the headwaters of which we had crossed on our random route to Ujiji, to Urimba, a distance of six days by water, there are no villages, and consequently no food.  The shore party, however, before leaving Ujiji, had eight days’ rations, and on this morning four days’, distributed to each person, and therefore was in no danger of starvation should the mountain headlands, now unfolding, abrupt and steep, one after another, prevent them from communicating with us.  It must be understood that such a journey as this had never been attempted before by any Arab or Msawahili, and every step taken was in sheer ignorance of where the road would lead the men ashore.  Rounding Kivoe’s steep promontory, whose bearded ridge and rugged slope, wooded down to the water’s edge, whose exquisite coves and quiet recesses, might well have evoked a poetical effusion to one so inclined, we dared the chopping waves of Kivoe’s bay, and stood direct for the next cape, Mizohazy, behind which, owing to wind and wave, we were compelled to halt for the night.

After Mizohazy is the bold cape of Kabogo—­not the terrible Kabogo around whose name mystery has been woven by the superstitious natives—­not the Kabogo whose sullen thunder and awful roar were heard when crossing the Rugufu on our flight from the Wahha—–­but a point in Ukaranga, on whose hard and uninviting rocks many a canoe has been wrecked.  We passed close to its forbidding walls, thankful for the calm of the Tanganika.  Near Kabogo are some very fine mvule trees, well adapted for canoe building, and there are no loud-mouthed natives about to haggle for the privilege of cutting them.

Along the water’s edge, and about three feet above it, was observed very clearly on the smooth face of the rocky slopes of Kabogo the high-water mark of the lake.  This went to show that the Tanganika, during the rainy season, rises about three feet above its dry season level, and that, during the latter season, evaporation reduces it to its normal level.  The number of rivers which we passed on this journey enabled me to observe whether, as I was told, there was any current setting north.  It was apparent to me that, while the south-west, south, or south-east winds blew, the brown flood of the rivers swept north; but it happened that, while passing, once or twice, the mouths of rivers, after a puff from the north-west and north, that the muddied waters were seen southward of the mouths; from which I conclude that there is no current in the Tanganika except such as is caused by the fickle wind.