﻿Vol. 66.] OX THE GEOLOGY OF KYASALASB. 235 



ridge. This is not an ideal ridge, with sides sloping regularly to 

 the east and west, but consists of a series of high mountain-ranges, 

 isolated mountain-masses, and strips of elevated plateau grouped 

 together along a roughly north-and-south line on each side of 

 the lake. Prof. J. W. Gregory 1 has suggested that the Great 

 Rift Valley of East Africa is due to the falling-in of the keystone of 

 a pre-existing arch. At first sight, objections may be raised to 

 using this theory to account for the position of the Xyasa trough. 

 The Lilongwe River, for instance, is rejuvenated where it reaches the 

 edge of the lake-basin, and is consequently older than this depres- 

 sion ; yet this river and its neighbour, the Bua, flow towards, and not 

 away from, the lake, as might be expected on the arch hypothesis. 

 This objection, however, may be surmounted, if we suppose that the 

 erosion of the rivers was able to keep pace with the uplift — a view 

 which accords with the fact that the Lilongwe and Lintipi have 

 succeeded in crossing a barrier-ridge on their way to the lake. If 

 it is admissible to assume that an anticlinal ridge existed on the 

 present site of Lake Xyasa, one must allow for a very considerable 

 lapse of time between the initiation of this ridge and the formation 

 of the lake- trough. 



Age of the Nyasa Trough. 



The lake-faults cut across post-Karroo faults, and are obviously of 

 post-Karroo age. There is, unfortunately, nothing to fill in the gap 

 between the recent freshwater deposits of the lake-basin and the 

 rocks of Karroo age, so that we are forced to base our conclusions 

 concerning the age of the lake mainly on topographical evidence. 



On the north of the lake, Dr. Rornhardt has described a series of 

 alkaline lavas occupying a continuation of the Rift Valley. These 

 lavas are in part, at least, younger than the rift, for they overlap 

 along its western margin. When the first outbreak of igneous 

 activity took place is doubtful, but the occurrence of volcanic cones 

 shows that vulcanicity has lasted on to recent times. Hot springs 

 at Mount Waller, Kota Kota, near Liwonde in the Shire Valley, and 

 far away to the south at Morambala Mountain probably represent 

 the last phases of vulcanicity. These springs lie along a north-and- 

 south line of weakness roughly following the course of the lake- 

 depression. 



The discovery of recent lacustrine deposits up to 700 feet above 

 the present lake-level opens up a field of investigation to future 

 workers which may prove fertile. 



The view that the lake is of no great age is supported by 

 topographical evidence. The shortness of the lateral tributaries 

 along the eastern border of the lake, despite a comparatively heavy 

 rainfall, is readily explained on the above supposition. Again, 

 the plateau-edges still retain in places a certain sharpness. This 

 is especially the case with the Livingstone Range. The absence of 



1 ' The Great Rift Valley ' London, 1896, p. 231. 



