5 ° 
THE TANGANYIKA PROBLEM. 
of high and ancient plateaux. Each so-called rift, how- 
ever, seems far more correctly conceived as a plait-valley, 
or what I shall call in future a “ eurycolpic fold,” the word, 
from the Greek evpvKo\7ro<i, simply suggesting the fact that 
valleys having the peculiar characters of “ graben ’ may 
be caused by folding due to lateral pressure ; and, as we 
have seen, this sort of folding has occurred all along the 
main central “graben ” series in Africa. Throughout this 
series the valleys seem invariably to have arisen as by- 
products of such folding, and not through the vertical 
subsidence of strips of an ancient plateau. 
Having thus discussed the structural features of the 
main eurycolpic depression, we may now endeavour to 
ascertain the number, extent and inter-relationships ot 
these remarkable folds. In attempting to do this we are 
greatly indebted to a valuable contribution to the subject 
by Dr. Kohlschiitter, who has recently returned from the 
German Gravitational Survey in East Central Africa. 
During the course of the investigations of this survey it 
was found that the great Nyassa valley, after passing to 
the north of the group of small volcanic cones in its floor, 
divides, giving off first a branch to the north-east which 
becomes wider, and finally appears to open out and 
disappear among the hills of Ussangu and Uhehe. The 
more westerly branch of the valley of Nyassa is directly 
continuous with the valley of Lake Rukwa, and in the 
vicinity of this lake the valley sends off another branch 
which runs to the south and appears actually to join the 
Nyassa valley near the region of Mount Waller. North- 
ward the Rukwa valley has been found to run at an acute 
angle into the great depression of Tanganyika itself, and 
it appears extremely probable that it actually crosses the 
Tanganyika trough in the region of Karama, appearing 
