Notes on Plants at Roraima. 181 
Just where the rock and tree belt meets the base of the 
cliff is a very narrow strip of quite distinct vegetation, — 
so distinct indeed that we might almost regard it as a 
distinct belt, which we might call the bramble belt. 
The ground there is covered by a dense thicket of 
bramble bushes Rubus guyanensis, Focke, [No. 106], in 
general appearance altogether like our English black- 
berry bushes. Among this were large masses of the 
South American form, appearing very similar to our 
English form, of the common bracken {Pteris aquilind). 
There, too, were many little bushes of Marcetia taxifolia 
very strongly suggestive of English heath. There, too, 
was a flowering Laurestinus {Viburnum ^labratum 
H.B.K. No. 220) curiously like the familiar plant of our 
gardens. To me, after my long stay in the tropics, the 
whole scene suddenly seemed very home-like and 
pleasant. But the next minute as I turned in another 
direction, the illusion was dispelled by the sight of great 
thickets of palms {Geonoma Appuniana) and a few 
singly standing and very stately tree ferns. 
Up from the bramble-belt, passing obliquely up the 
cliff face, ran the ledge by which we ascended to the top 
of Roraima. The lower part of the ledge, for perhaps 
two-thirds of its length, is wide, much broken and very 
uneven of surface. This part is somewhat irregularly 
bush-covered. Then the continuity of the ledge is sud- 
denly almost broken by a deep ravine, a part of the 
rock having been worn away by a stream which falls on 
to it from the cliff above. The ravine thus made is 
almost bare of vegetation. Above, the ledge slopes 
somewhat steeply but evenly from the point where it 
re-commences to the top ; and this part of it is cov- 
