At Mashue--where we found a never-failing supply of pure water in a sandstone rocky hollow--we left the road to the Bamangwato hills, and struck away to the north into the Desert. Having watered the cattle at a well called Lobotani, about N.W. of Bamangwato, we next proceeded to a real Kalahari fountain, called Serotli. The country around is covered with bushes and trees of a kind of leguminosae, with lilac flowers. The soil is soft white sand, very trying to the strength of the oxen, as the wheels sink into it over the felloes and drag heavily. At Serotli we found only a few hollows like those made by the buffalo and rhinoceros when they roll themselves in the mud. In a corner of one of these there appeared water, which would have been quickly lapped up by our dogs, had we not driven them away. And yet this was all the apparent supply for some eighty oxen, twenty horses, and about a score of men. Our guide, Ramotobi, who had spent his youth in the Desert, declared that, though appearances were against us, there was plenty of water at hand. We had our misgivings, for the spades were soon produced; but our guides, despising such new-fangled aid, began in good earnest to scrape out the sand with their hands. The only water we had any promise of for the next seventy miles--that is, for a journey of three days with the wagons--was to be got here. By the aid of both spades and fingers two of the holes were cleared out, so as to form pits six feet deep and about as many broad. Our guides were especially earnest in their injunctions to us not to break through the hard stratum of sand at the bottom, because they knew, if it were broken through, "the water would go away." They are quite correct, for the water seems to lie on this flooring of incipient sandstone. The value of the advice was proved in the case of an Englishman whose wits were none of the brightest, who, disregarding it, dug through the sandy stratum in the wells at Mohotluani: the water immediately flowed away downward, and the well became useless. When we came to the stratum, we found that the water flowed in on all sides close to the line where the soft sand came in contact with it. Allowing it to collect, we had enough for the horses that evening; but as there was not sufficient for the oxen, we sent them back to Lobotani, where, after thirsting four full days (ninety-six hours), they got a good supply. The horses were kept by us as necessary to procure game for the sustenance of our numerous party. Next morning we found the water had flowed in faster than at first, as it invariably does in these reservoirs, owing to the passages widening by the flow. Large quantities of the sand come into the well with the water, and in the course of a few days the supply, which may be equal to the wants of a few men only, becomes sufficient for oxen as well. In these sucking-places the Bakalahari get their supplies; and as they are generally in the hollows