On the left of the island we see the water at the bottom, a white rolling mass moving away to the prolongation of the fissure, which branches off near the left bank of the river. A piece of the rock has fallen off a spot on the left of the island, and juts out from the water below, and from it I judged the distance which the water falls to be about 100 feet. The walls of this gigantic crack are perpendicular, and composed of one homogeneous mass of rock. The edge of that side over which the water falls is worn off two or three feet, and pieces have fallen away, so as to give it somewhat of a serrated appearance. That over which the water does not fall is quite straight, except at the left corner, where a rent appears, and a piece seems inclined to fall off. Upon the whole, it is nearly in the state in which it was left at the period of its formation. The rock is dark brown in color, except about ten feet from the bottom, which is discolored by the annual rise of the water to that or a greater height. On the left side of the island we have a good view of the mass of water which causes one of the columns of vapor to ascend, as it leaps quite clear of the rock, and forms a thick unbroken fleece all the way to the bottom. Its whiteness gave the idea of snow, a sight I had not seen for many a day. As it broke into (if I may use the term) pieces of water, all rushing on in the same direction, each gave off several rays of foam, exactly as bits of steel, when burned in oxygen gas, give off rays of sparks. The snow-white sheet seemed like myriads of small comets rushing on in one direction, each of which left behind its nucleus rays of foam. I never saw the appearance referred to noticed elsewhere. It seemed to be the effect of the mass of water leaping at once clear of the rock, and but slowly breaking up into spray.
I have mentioned that we saw five columns of vapor ascending from this strange abyss. They are evidently formed by the compression suffered by the force of the water's own fall into an unyielding wedge-shaped space. Of the five columns, two on the right and one on the left of the island were the largest, and the streams which formed them seemed each to exceed in size the falls of the Clyde at Stonebyres when that river is in flood. This was the period of low water in the Leeambye; but, as far as I could guess, there was a flow of five or six hundred yards of water, which, at the edge of the fall, seemed at least three feet deep. I write in the hope that others, more capable of judging distances than myself, will visit the scene, and I state simply the impressions made on my mind at the time. I thought, and do still think, the river above the falls to be one thousand yards broad; but I am a poor judge of distances on water, for I showed a naval friend what I supposed to be four hundred yards in the Bay of Loanda, and, to my surprise, he pronounced it to be nine hundred. I tried to measure the Leeambye with a strong thread, the only line I had in my possession, but, when the men had gone two or three hundred yards, they got into conversation, and did not hear us shouting that the line had become entangled. By still going on they broke it, and, being carried away down the stream, it was lost on a snag. In vain I tried to bring to my recollection the way I had been taught to measure a river by taking an angle with the sextant. That I once knew it, and that it was easy, were all the lost ideas I could recall, and they only increased my vexation. However, I measured the river farther down by another plan, and then I discovered that the Portuguese had measured it at Tete, and found it a little over one thousand yards. At the falls it is as broad as at Tete, if not more so. Whoever may come after me will not, I trust, find reason to say I have indulged in exaggeration.* With respect to the drawing, it must be borne in mind that it was composed from a rude sketch as viewed from the island, which exhibited the columns of vapor only, and a ground plan. The artist has given a good idea of the scene, but, by way of explanation, he has shown more of the depth of the fissure than is visible except by going close to the edge. The left-hand column, and that farthest off, are the smallest, and all ought to have been a little more tapering at the tops.
* The river is about one mile (1.6 km) wide at the falls, and plunges over 350 feet at the centre. Livingstone greatly underestimated both distances.--A. L., 1997.