282 
THE LUMBWA AND ELGON CAVES 
Both Mr. Dobbs and Mr. Knight-Bruce appear to be of the 
opinion that these caves are the work of the Lumbwa natives 
excavating in search of salt earth for the use of their stock. 
Until further evidence is available this view must stand, for 
the reason that the natives are digging in the caves and carrying 
away this earth at the present time. I only recently had an 
opportunity of examining a number of these caves, and I am 
able to support the view that they are of artificial origin. 
The only other series of caves of such magnitude occurs on 
Mount Elgon, the inhabitants of which are the Lako, El Gonyi, 
and Savei, and are, by the way, of the same race as the Lumbwa 
or Kipsikis. I visited a number of these caves many years 
ago and tentatively formed the opinion that some of them 
were the work of wave-action operating upon soft layers 
of volcanic ash during a period when the waters of an inland 
sea attacked the flanks of the mountain, the mountain mass 
being later on raised through tectonic folding of the earth’s 
crust. There is little doubt that the waters of the inland sea 
now known as Lake Victoria once covered a greater extent 
than at present ; in fact, there is clear evidence that they ex- 
tended up the Nyando valley certainly as far as Muhoroni, 
but that only premises a comparatively small rise in level. 
Now we know that in Miocene times the land comparatively 
near to the present shore of the lake was the habitat of beasts 
such as the Dinotherium, and the bones found in the beds 
near Karungu could only have been deposited in swampy 
flats of no great depth, and the greater extent of the lake at 
that period corresponds quite well with its extension up the 
Nyando valley and over the Kitosh plain. Dr. Oswald has 
recorded beds of Miocene age near Karungu about 140 feet 
thick. 
In addition to the facts just recorded, there is a theory 
that at a later date than Miocene times the lake extended far 
away to the north-east past Mount Elgon, and certainly an 
examination of the map produces the impression that originally 
the waters of the lake may have included Lake Kioga and 
also the chain of lakes called after Salisbury and Gedge in 
Kumama country ; if this is so, a careful examination of 
the area would probably disclose traces of old lake-terraces, 
