country where the grass being only a foot or fifteen inches high, and thickly planted, often looks like a beautiful glade in a gentleman’s park in England.  They are from a quarter of a mile to a mile broad, and from two to ten or more miles long.  The water of the heavy rains soaks into the level forest lands:  one never sees runnels leading it off, unless occasionally a footpath is turned to that use.  The water, descending about eight feet, comes to a stratum of yellow sand, beneath which there is another stratum of fine white sand, which at its bottom cakes, so as to hold the water from sinking further.

It is exactly the same as we found in the Kalahari Desert, in digging sucking places for water for our oxen.  The water, both here and there, is guided by the fine sand stratum into the nearest valley, and here it oozes forth on all sides through the thick mantle of black porous earth, which forms the sponge.  There, in the desert, it appears to damp the surface sands in certain valleys, and the Bushmen, by a peculiar process, suck out a supply.  When we had dug down to the caked sand there years ago, the people begged us not to dig further, as the water would all run away; and we desisted, because we saw that the fluid poured in from the fine sand all round the well, but none came from the bottom or cake.  Two stupid Englishmen afterwards broke through the cake in spite of the entreaties of the natives, and the well and the whole valley dried up hopelessly.  Here the water, oozing forth from the surface of the sponge mantle, collects in the centre of the slightly depressed valley which it occupies, and near the head of the depression forms a sluggish stream; but further down, as it meets with more slope, it works out for itself a deeper channel, with perpendicular banks, with, say, a hundred or more yards of sponge on each side, constantly oozing forth fresh supplies to augment its size.  When it reaches rocky ground it is a perennial burn, with many aquatic plants growing in its bottom.  One peculiarity would strike anyone:  the water never becomes discoloured or muddy.  I have seen only one stream muddied in flood, the Choma, flowing through an alluvial plain in Lopere.  Another peculiarity is very remarkable; it is, that after the rains have entirely ceased, these burns have their largest flow, and cause inundations.  It looks as if towards the end of the rainy season the sponges were lifted up by the water off their beds, and the pores and holes, being enlarged, are all employed to give off fluid.  The waters of inundation run away.  When the sponges are lifted up by superabundance of water, all the pores therein are opened:  as the earthen mantle subsides again, the pores act like natural valves, and are partially closed, and by the weight of earth above them, the water is thus prevented from running away altogether; time also being required to wet all the sand through which the rains soak, the great supply may only find its way to the sponge a month or so after the great rains have fallen.