150 dr. f. Oswald on the [June 1 9 14, 



directions west and east, north-west and south-east, and north-north- 

 east and south-south-west. Crossing the plateau (here reduced b} r 

 erosion to a width of only 6 furlongs) from the escarpment east- 

 wards until I overlooked the upper course of the Kuja, I descended 

 near Kenin, returning to the Kuja Gorge. The angle of the slope 

 immediately below the nearly-vertical cliff is as high as 65°. 

 Near the junction with the dolerite the buff-coloured quartzitic 

 sandstone, veined and blotched purple with haematite, becomes 

 changed to a snow-white quartzite, and the dolerite is here highly 

 amygdaloidal with amygdales of quartz (as already described), 

 and decomposes into a thick mantle of fertile red soil. The cone 

 of Kinsunsi on the right bank of the Kuja, near the western exit 

 of the gorge, is an apophysis of the sill or laccolite, and is 

 surrounded by an aureole of white quartz. 



It seems natural to assume that these quartzites belong to the 

 same series as the quartzites of the Nandi Hills on the north side 

 of the Kavirondo Gulf ; but they are far more uniform in 

 character than the Nandi Quartzites described by Mr. G. F. Scott- 

 Elliot & Prof. J. W. Gregory l as belonging to the Karagwe 

 Series, which is so much more extensively developed on the western 

 side of the Victoria Nyanza, where it reaches a thickness of over 

 1200 feet. The quartzite belongs to the uppermost division of the 

 Series, which has been traced to the southern end of Tanganyika 

 and is probably of the same age as the Waterberg Series of the 

 Transvaal : that is, presumably Devonian. 



The quartzitic sandstones of the Kisii Highlands have not been 

 subjected to the pressure which has tilted up the Karagwe Beds to 

 a high angle, and crushed them until they are highly contorted. If 

 it were not for the intrusion of the underlying dolerite, these Kisii 

 sandstones would doubtless have still preserved their original 

 horizontality. 



Crossing the Kuja as I proceeded on my march to Kisii Boma, 

 I found the ford in the centre of the gorge to lie at an altitude 

 of 4894 feet. Comparing this with the altitude of the ford 

 outside and below the gorge (that is, 4441 feet), we find a fall of 

 453 feet in 2| miles, or a gradient of 1 in 29, accounting for the 

 numerous rapids of the Kuja in its rocky bed through the gorge. 

 This gradient stands in marked contrast with the lower and succeed- 

 ing stretch to the Sakwa bridge (only 94 feet in 6| miles). 



In addition to the evidence of the rejuvenation of the river, as 

 shown by the very frequent rapids and waterfalls, the angles of 

 slope of its valley yield further support to this conclusion. From 

 the 65° of the quartzite immediately below the precipices of the 

 edge of the plateau the angle rapidly changes to a long slope of 10° 

 over the easily-decomposed dolerite, and, on approaching the river, 

 it suddenly alters to a short slope of 30° (cut also in the same 

 dolerite) down to the present bed of the river. 



If one keeps a northerly direction over the nearly-level summit 



1 ' The Geology of Mount Euwenzori ' Q. J. G. S. vol. li (1895) pp. 677, 678. 



