Notes on Plants at Roraima. 169 
as it were, a terrace about midway up the slope. The 
upper level of this terrace, which lies at a height of about 
5400 feet, is almost everywhere swampy, though here 
and there a few rocks crop out. This is the place so 
enthusiastically described by Dr. SCHOMBURGK, on ac- 
count of the extraordinary richness of its vegetation, as 
a ' botanical El Dorado' ; and it was here too, just within 
the forest which edges this swamp that we built our 
home and made our head-quarters. It is to this point 
too that the open savannah extends, for above, all is more 
or less densely forested. Behind this swamp, which caps, 
as it were, a terrace, half-way up the face of the moun- 
tain is a ravine ; and again beyond this ravine, 
in which it must be remembered that the forest begins, 
the mountain slopes up very abruptly to a height of about 
6,500 feet, to the base, that is : of the actual cliff. In 
the accompanying diagram (p. 170) all up to the ravine 
is distinguished as the savannah slope ; all above, 
to the base of the cliff, as the forest slope. It should be 
noted that the forest slope is not uniformly clad with 
trees. The lower part is densely wooded, covered as it 
were, by dense jungle ; next comes a belt of bush, 
rather than of jungle ; while still higher, just under the 
cliff, the masses of rock which have fallen from above, 
lie like a moraine, on which are scattered, however, 
sparse trees, the low, wide-spreading branches of which 
interlock in a remarkable way.* The actual face of 
the cliff is, of course, bare ; but wherever ledges run up 
for any distance these are often tree or bush clad ; and 
* This moraine-like part of the slope is curiously like the well-knowu 
1 Wistman's Wood' on Dartmoor. 
Y 
