of Tanganyenka, allowed the central waters to drain eastward.  All the African lakes hitherto discovered are shallow, in consequence of being the mere ‘residua’ of very much larger ancient bodies of water.  There can be no doubt that this continent was, in former times, very much more copiously supplied with water than at present, but a natural process of drainage has been going on for ages.  Deep fissures are made, probably by the elevation of the land, proofs of which are seen in modern shells imbedded in marly tufa all round the coast-line.  Whether this process of desiccation is as rapid throughout the continent as, in a letter to the late Dean Buckland, in 1843, I showed to have been the case in the Bechuana country, it is not for me to say; but, though there is a slight tradition of the waters having burst through the low hills south of the Barotse, there is none of a sudden upheaval accompanied by an earthquake.  The formation of the crack of Mosioatunya is perhaps too ancient for that; yet, although information of any remarkable event is often transmitted in the native names, and they even retain a tradition which looks like the story of Solomon and the harlots, there is not a name like Tom Earthquake or Sam Shake-the-ground in the whole country.  They have a tradition which may refer to the building of the Tower of Babel, but it ends in the bold builders getting their crowns cracked by the fall of the scaffolding; and that they came out of a cave called “Loey” (Noe?) in company with the beasts, and all point to it in one direction, viz., the N.N.E.  Loey, too, is an exception in the language, as they use masculine instead of neuter pronouns to it.

If we take a glance back at the great valley, the form the rivers have taken imparts the idea of a lake slowly drained out, for they have cut out for themselves beds exactly like what we may see in the soft mud of a shallow pool of rain-water, when that is let off by a furrow.  This idea would probably not strike a person on coming first into the country, but more extensive acquaintance with the river system certainly would convey the impression.  None of the rivers in the valley of the Leeambye have slopes down to their beds.  Indeed, many parts are much like the Thames at the Isle of Dogs, only the Leeambye has to rise twenty or thirty feet before it can overflow some of its meadows.  The rivers have each a bed of low water—­a simple furrow cut sharply out of the calcareous tufa which lined the channel of the ancient lake—­and another of inundation.  When the beds of inundation are filled, they assume the appearance of chains of lakes.  When the Clyde fills the holms ("haughs”) above Bothwell Bridge and retires again into its channel, it resembles the river we are speaking of, only here there are no high lands sloping down toward the bed of inundation, for the greater part of the region is not elevated fifty feet above them.  Even the rocky banks of the Leeambye below Gonye, and the ridges bounding the Barotse