﻿234 MESSRS. A. E. ANDREW AND T. E. G. DAILEY [May I9IO, 



hard gneiss ; and it is true that such rocks occur at the northern 

 end of the lake, with faulted junctions showing a certain parallelism 

 with the general trend of the valley. But, on the one hand, this 

 parallelism is by no means the rule, for in the Ruhuhu area (in 

 German territory) the main Karroo fault runs east and west, and 

 is sharply truncated by the lake-margin ; while, on the other hand, 

 the occurrence of Karroo rocks in the lake-basin is merely an 

 accidental feature of the northern portion of the depression, for by 

 far the greater part of the trough has been formed in solid gneiss 

 or granite. Nor is there a possibility of the lake-valley having 

 been determined by erosion along a soft band in the gneiss, since 

 the strike of the gneiss lies for considerable distances at right 

 angles to the trend of the lake. Again, if the lake-trough were 

 primarily a valley of erosion, one might well expect to find the lateral 

 tributaries of proportionate size : such is not the case. The water- 

 shed separating the rivers draining westwards into the lake from 

 those flowing eastwards to the Indian Ocean nearly coincides with 

 the steep eastern edge of the trough itself. This is remarkable, 

 because the rainfall must be fairly heavy along these eastern 

 slopes. On the western side of the lake the shortness of the lateral 

 tributaries is somewhat less marked : the Bua, for instance, is a 

 fairly large river. In this case, however, there is clear evidence 

 to suggest that the tributary stream is older than the main Nyasa 

 Valley. Taking all the facts into consideration, there seems to be 

 no doubt whatever that the Nyasa trough originated in, and has 

 subsequently ' been developed by, earth-movements. The steep 

 edges of the plateaux facing the lake can be explained as fault 

 features, sometimes dropping to great depths below lake-level and 

 sometimes separated by wide fault-steps, such, for instance, as the 

 coastal plains behind Kota Kota. The general trend of these 

 faults is in a north-and-south direction, but in places they appear 

 to have suffered deflection or to have died out for a space. Along 

 the north-western border a sloping platform sinks graduall}' east- 

 wards into the lake, and appears to dip in this direction until the 

 submarine foot of the Livingstone Iiange is reached. This range 

 towers up to great heights along the north-eastern shore of the 

 lake, and sinks to considerable depths below its surface. It is 

 possible that the fault running along the western side of the Rift 

 Valley is deflected westwards, while the eastern fault continues 

 along the edge of the Livingstone Range, in a way somewhat 

 reminiscent of the older faults limiting the Karroo in the Mwapo 

 and Western Nyika areas. 



Many of the arguments which have been used in support of the 

 tectonic origin of the Nyasa trough can be applied with equal truth 

 to the Shire Valley, which is doubtless a direct continuation of the 

 main Rift Valley. The method in which the Nyasa rift dies out to 

 the south is not clearly understood, but it appears probable that 

 the fault-feature defining the eastern edge of the Shire trough is 

 continued for many miles to the south of its western neighbour. 



The Nyasa depression lies along the axis of a north-and-south 



