Page:Africa by Élisée Reclus, Volume 4.djvu/434

 350 SOUTH AND EAST AFRICA. a natural nor supernatural origin. They must have been excavated by the hand of man. That such prodigious excavations in extremely solid rock, extending away into complete darkness, branching out in various directions, and from 12 to lo feet from floor to coiling, were fonncd as dwelling-places, or even as strong- holds, is simply absurd. For natives such as those of the present day to have cut out even one cave would have been a sheer impossibility, with the tools they possess. IJut there are not merely one or two excavations. There are surprising numbers of them — sufficient indeed to house a whole tribe, as I am informed that they extend all round the mountain. " There is one point of great interest as tending to throw some light on the sub- ject. The caves all occupy a certain horizon or level of the mountain, and all occur in the compact agglomerate, none in the level beds immediately overhead. a very remote era some ver}' powerful race, considerably advanced in arts and civilisation, excavated these great caves in their search for precious stones, or possibly some ])recious metal. However improbable this theory may seem, it is the only one that suggests itself to me after months of cogitation. Unfortunately, though I was from the first without a doubt about their being of artificial origin, this idea never crossed my brain while I was at Elgon, and I consequently made no special examination for evidence of precious stones or metals. Are we to suppose that the Egyptians really got so far south ? If not what other race could have cut these extraordinary recesses P " * The Paxoani and Tana Rivers. The Pangani River, whose lower course forms the boundary between the British and German protectorates, receives its first supplies from Mounts Moeru and Kilima-Njaro. Of all these mountain torrents the easternmost is the Lumi, which rises at the foot of Kimawenzi, and flows at first southward to the elonsrated ba>^in of Lake Jipe at the foot of the Ugono eo^arpments. This basin stands at an elevation of no more than 2,400 feet, nevertheless the plain stretching south of the Kilima-Xjaro terraces lies at a still lower level, for an emissary flowing from the lake immediately to the west of the Lumi influent trends north-westwards in the direction of the foot of the mountain itself. After its confluence with the Ruvu (Ru-Yu'i, and several other torrents, this emissary from the lake is already a river of considerable volume. Fischer found that it was nearly 350 feet broad, with a depth of over 3 feet. Farther down the stream, which here receives scarcely any tributaries, flows southwards and then to the south-east, forcing its way over a series of fulls and rapids through the rocky barriers which run trans- versely to its course. It continues to he obstructed by other falls down to the lower reaches, so that the Pangani becomes navigable only within some 24 miles of the coast, where its channel is confined by two elevated terraces of coralline formation.
 * ' Looking at evervthinff, I can come to but one conclusion, and that is, that in
 * Through Maaailand, p. 510. «