258 МЕ. Е. Е. IM THURN ON THE PLANTS 
Let us now pass to the consideration of Roraima itself as an area of distinct vegetation ; 
and in so doing afew words must first be said as to the physical features of the mountains. 
Roraima is one (certainly the best known, perhaps really the most remarkable) of a 
group of pillar-like sandstone mountains capped with hard conglomerate, which group 
is, it seems to me, identical in nature and origin with the groups of sandstone pillars, 
capped with conglomerate or hardened mud, of the eppellings already described. In 
short, Roraima and its fellow mountains seem to be an eppelling on a gigantic scale. 
Some notion of how large this scale is may be gathered from the fact that Roraima 
itself, one pillar of the group, is almost exactly four miles wide along its south-eastern 
face, and is apparently seven or eight miles long from south to north, and that its height 
is some 5000 feet above the general level of the plain from which it rises. 
This 5000 feet of height, it must be explained, is made up of a sloping base, the pedi- 
ment of the pillar, of about 3000 feet, which is surmounted by the more strict pillar-like 
portion, 2000 feet in height. The plateau on top of the pillar is a very slightly, 
almost imperceptibly, hollowed basin, four miles wide by some seven or eight long, 
over which are scattered innumerable single rocks and piles of rocks, the largest 
of which are apparently some eighty or ninety feet in height. The sloping basal 
part of the mountain is, everywhere but toward the south-east, covered by dense, but 
not lofty forest; while on the south-east a considerable portion of it (which portion 
does not, however, extend up to the foot of the actual cliff) is treeless and grass- 
covered. Тһе cliff itself is bare, but for a comparatively few mosses, ferns, grasses, and 
trailing plants clinging closely to the rougher parts of its surface, especially where the 
many waterfalls trickle down the rock-face, and for the dwarf shrubs, ever dwarfer and 
more alpine in character toward the top, which have found a lodgment on the few 
transverse ledges which break the evenness of the surface. The hollow basin at the top 
of the pillar is, wherever a little soil has accumulated in the depressions of the bare rock 
which constitutes the greater part of its surface, clothed with a dwarf herb-like vegeta- 
tion of most remarkable appearance, consisting largely of various species of Pepalanthus, 
a Drosera, a few terrestrial orchids (these not very conspicuous in flower), a remarkable | 
low-growing aloe-like Abolboda of which I shall have more to say hereafter, various 
ground-clinging shrubs of alpine Vacciniwm-like character, and of a very few single 
shrubs, all of one species (Bonnetia Roraime, Oliv., n. sp. [No. 330]), of larger growth, 
even though this is but some three feet high. 
Nor in this brief sketch of the physical features of Roraima in their bearing on the 
vegetation is it possible to avoid mention of the great moisture of the atmosphere which 
surrounds the mountain. The shallow basin of the upper plateau always holds much 
water, and probably at times is almost full; the sides of the cliff are ever moistened by 
the innumerable rills and streams poured down from the plateau above on to the 
sloping base; and this basal portion itself is, on the more level undulating parts of its 
exposed surface, a mere spongy swamp, while in its forested parts it is traversed by 
almost innumerable rills hastening down to join the large rivers of the plain below. 
When dealing with the vegetation along our line of march to Roraima I pointed 
out that I could only pretend to speak of the plants actually along that line; in now 
