Mount Roraima. 26 
that it does not form a round pillar. Were we to make 
a ground-plan of it, its diameter would be much greater 
from north to south than from east to west. From east 
to west, as viewed from the south, the mountain is just 
about four miles across ; from north to south it appears 
to be some seven or eight miles. 
Of our actual ascent of the mountain I need not say 
much here; for I have elsewhere told of this in some 
detail. We reached the mountain on its southern side ; 
and on that side, too, we made an ascent. The house 
which we built for ourselves and occupied during our 
three weeks of preliminary exploration, was rather more 
than half-way up the sloping part of the mountain, the 
pediment of the pillar. Up to that point the slope, on 
that side, was grass-covered and swampy ; the cold and 
damp were intense. Above that point, from our house 
to the base of the a6lual cliff, the slope was much more 
steep, and was entirely covered by terrific moraine-like 
masses of huge fragments of rock, over which had 
grown the densest forest of small but gnarled trees, 
matted together by ferns and climbing bamboos. At the 
top of this slope — at the base., that is, of the actual 
cliff — was a curious belt of blackberry bushes — a rare, 
perhaps a unique, instance in the tropics. And imme- 
diately above this, rose, for more than 2,000 feet, the 
sheer cliff. Or rather the cliff was not quite sheer ; for 
its top overhung its base, so that cs I sat at the base of 
the cliff, with my back against it, trying to put roughly 
into papers the too abundant new plants which I had 
gathered on the way up, the water which ever drips 
from every part of the upper edge of the cliff, fell, not 
on to, but beyond me, 
