404 THE EXPLORATION OF THE NIGER did not altogether regret having avoided coming into contact, just before entering Dendi, with tribes then at war with its people.

The next day, the 17th, the river still wound in much the same way as on the i6th ; in fact, so serpentine was its course, that one of the coolies cried out he did not believe we were on the Niger still, but that we had lost our way. Numerous islands and dense vegetation, with very picturesque views here and there, were the order of the day. Great blocks of red sandstone rose in some places to a height of from about thirty to more than three hundred feet, and at every bend of the stream some new or strange view met our eyes.

We longed to land and seek repose beneath the thick dome of vegetation forming natural arbours, but there was one great drawback about them, the immense number of insects eager to suck our blood. At night the mosquitoes invaded us in vast hordes, and our poor coolies used to roll themselves up in everything they could lay hands on, at the risk of suffocation. As for us, we too suffered terribly, for though when we were asleep our mosquito nets did to some extent protect us, when we were on watch on deck we were nearly bled to death. In the day these pests left us pretty well unmolested, but their place was taken by other persecutors, rather like gadflies, which were able to pierce through our white clothes with a sting as sharp and nearly as long as a needle. I had suffered terribly once before from these horrible diptera when I was on the Tankisso ; in fact they haunt the tropical vegetation of many an African river.

Since we started we had been each day threatened with a tornado, but the storm had not broken after all. On the 1 8th, however, we came in for the tail of one of these meteorological disturbances, and a pretty strong breeze