Vol. 70.] MIOCENE OF THE VICTORIA XYA>'ZA, ETC. 157 



A thin crystalline crust is deposited round the shore, and has 

 essentially the same composition as the water. It is gathered by 

 the natives in baskets for mixing with their food, and for treating: 

 their tobacco. I noticed a similar efflorescence of soda on the 

 borders of the Kimera swamp, north of Homa Mountain. The 

 water is gradually rising, as shown by some half -immersed, dead 

 candelabra euphorbias, already standing 10 feet away from the 

 shore. 



West of the mouth of the Awach Eiver, the ground rises to form 

 a terrace running east-south-eastwards and west-north-westwards, 

 about a mile wide, between the Nyanza and the Kimera swamp ; 

 and it faces the Nyanza in a cliff about 30 feet high, flattening out 

 again on the west. The lake has now retreated for about a quarter 

 of a mile from the base of the cliff, which is cloaked at its foot 

 by talus. The downward succession of the beds (dipping 5° north- 

 north-eastwards), as revealed in the cliff, consists of : — 



Thickness in feet. 



(1) Grey or pale-buff argillaceous sandstone, with small 



flakes of biotite and larger plates of greenish talc, 

 also markings of dendritic manganese ; current- 

 bedded and coarser-grained than the lower beds. It 

 contains pebbles of quartz up to three-eighths of an 

 inch long. Calcite occurs along the bedding-planes. 5 



(2) Brown, shaly, sandy clay 3 



(3) Hard argillaceous sandstone, in thin beds alternating 



with shales 3 



(4) Brown shaly clay, often wedging out 3 



(5) (Base not seen.) Argillaceous current-bedded sand- 



stone. Coarse beds, alternating with finer 10 



A large surface of bare rock is exposed on the summit of the 

 broad stony terrace, and the beds, instead of showing a plane 

 surface, appear in gentle undulations, giving rise to a number 

 of shallow basins. No ripple-marks are visible anywhere, and not 

 a trace of any fossils. On the terrace the dip changes in the 

 southern part to 5° south-westwards, whence we niay infer that 

 the terrace itself is a low anticline. 



Farther west, only Beds 2 and 3 are visible in the cliff, which 

 flattens out and finally becomes quite grass-grown. After some 

 interval, another group of strata appear, evidently at a slightly 

 higher horizon than the above-described series, but the actual 

 junction was not visible. The cliff shows : — 



Thickness in feet inches. 



Recent soil 1 



Gravelly grey clay 2 



White calcareous travertine 6 



Greenish-grey clay 1 



(Base not seen.) White calcareous travertine in 

 undulating beds, with tusks of elephant, bones of 



zebra, antelope, etc 3 



Still farther west, near a very big sycamore fig-tree, which forms 



