GREAT SQUARE CASTLES OF ROCK 
any changes or movement. As soon as the sun appeared, 
rendering the lower sky of a golden yellow and of vivid 
Indian red above, the northern part of the lagoon was 
enveloped in mist, which rose in angular blocks, vertical 
on the south side, slanting at a sharp angle on the north. 
These pointed peaks of mist remained immobile, as if they 
had been solid, until the sun was well up in the sky. 
I went once more to gaze at the glorious panorama. 
In the morning light new and important details were 
revealed, such as a strange series of dykes of a prismatic 
shape, of which I could count as many as seven. Great 
transverse depressions or grooves from south-southeast to 
north-northwest, with a dip south-southeast could in that 
light be now plainly detected, and this time two great 
square castles of rock, instead of one, were disclosed upon 
the third range of undulations. 
The high ridge to the southwest displayed a subsidence 
on a large scale in its central portion, where bare, vertical, 
red walls had been left standing on each side. 
Then there were various other curious, concave depres¬ 
sions or gateways formed in the great tableland, which had 
for its marked characteristic concave curves on all its 
slopes. 
On leaving camp — nearly at noon, after a serious 
quarrel and fight among my men, which left me worried 
to death by their petty nonsense and incessant grumbling 
— we journeyed at an elevation of 2,300 feet, finding 
shortly after an almost circular cuvette of deep grey 
cinders, 100 feet deep (elevation at the bottom 2,200 feet). 
Twelve kilometres farther on we came upon another 
great depression extending from east to west, with an 
enormous belt of grassy land. There was the usual cluster 
of trees and palms in the centre, hut larger than usual. 
To the south were campos, lovely prairies, with sparse and 
stunted trees — chiefly Goma arabica or acacias. 
The elevation of the upper edge of the cuvette was 
vol. I. — 19 289 
