THE LUMBWA AND ELGON CAYES 
285 
of Kisumu due east to opposite Muhoroni, and this produced 
a gigantic fault which was the cause of the Nandi Plateau. 
At a later date a north and south fracture' occurred which was 
the origin of the fault scarp which runs from Kabras to Kibos. 
The reason for the occurrence of this fault on an east 
to west line is obscure : the folds of the continent are from 
east to west, and therefore most of the fault scarps run north 
and south as in Taita, the Ukamba country, the Rift Valley, 
&c., and one cannot but suspect that it is in some wry con- 
nected with the intense volcanic activity which is centred 
in Tinderet. This problem will, however, not be clearly 
solved until a detailed geological survey * of the region is 
made. There, however, appears to be little doubt that the 
caves on the Nandi Escarpment are due to the falling together 
of masses of the granitic dykes, fractured and dislocated, 
during the formation of the fault scarp. A certain amount 
of water erosion has doubtless increased the cavities, for we 
hear of the occurrence of small streams in some of them. 
On the south side of the Nyando valley we have what is 
probably a similar fault scarp, but it is not so sharply defined, 
and the upthrow on the face is not so great ; in fact, it appears 
to have been masked to some extent by volcanic deposits, but 
the foci of eruption have not been identified. It is, however, 
probable that the Lumbwa highlands have been elevated 
by faulting in Tertiary times, but the information regarding 
the area south of the Nyando valley is at present too scanty 
to warrant any dogmatic conclusions. 
To refer again to the east and west fault scarp parallel with 
the railway between Muhoroni and Kibos (as a matter of fact, 
the bearing or strike of this fault is about W.N.W. and E.S.E.), 
The highest portion of the Nandi Plateau is close to the fault 
plane, and the terrain of the Nandi country falls steadily to 
the north until the valley of the Nzoia is reached ; north of 
the Nzoia valley the country gradually rises again towards 
Mount Elgon. It would therefore not appear that any causes 
which produced the elevation of the Nandi and Lumbwa 
plateaux would have appreciably affected Mount Elgon and 
raised that mountain to a similar extent. 
As previously mentioned, there is a theory of considerable 
