liv ITINBEAEY. 



was large anfl rather acid, not being quite ripe. On tlie wall were 

 some pretty little golden-flowered Melastomas; and on trees and wall 

 and old stumps were numbers of the bi'illiant crimson Vlricularia 

 CampheUiana. There were none of the white species, U. alpina, to be 

 seen, of which we had found a single specimen here on our first trip. 



The vegetation on the ledge is in genfial much moie dwarf and 

 scattered than in the upper part of the foi'est-belt, though there 

 are, at the lower part especially, very close clumps of a small bamboo, 

 which is very difficult to peneti-ate. Upwards it is much more exposed 

 and wind-swept, and the scanty soil easily becomes washed away — 

 except in some more sheltered corners. The ledge is very clearly 

 distinguishable in the picture in Vol. I., facing p. xxii, Itineraiy. It 

 is seen crossing the wall obliquely, from the right-hand side above the 

 middle of the view, to the top of the left side by the rounded outer 

 eminence ; and from below it presents an irregular but continuous 

 green surface. The grassy area in the fiont is a pait of the slopes 

 immediately below the forest-belt, and is just above the El Dorado 

 swamp, some 5400 feet elevation. 



At the highest point, before reaching the deep gully, where the bluff 

 is most prominent, the trail runs on almost bare broken rock with very 

 dwarf gi-owths at the edge, being here sheltei'ed by the pi-ojecting 

 portions of a sort of irregular ledge above. The unobscured view of the 

 rough and bare face of the walls, the irregular slopes of the forest-belt 

 and the rock-strewn savannahs below, with the village and the winding 

 Kukenaam river, and the whole stretch of country away to the horizon, 

 with mountains in the blue distance, is entiancing and unsurpassable 

 of its kind. From the greater heights of the summit,, everything below 

 is seen depressed into a more uniform level : here they stood out more 

 boldly, and present an altogether indescribable picture. One was loth 

 to turn away froni it — it was such a feast for the eyes and ima.gination. 

 At various points of the ascent there are fine >iews obtainable of the 

 surroundings, but this one from its range and comprehensiveness is 

 unique. 



The stieam at the bottom of the deep enlly was very small, the 

 water from above falling like heavy rain, splashing the bare rock of the 

 wall as well as the few bushes against it on the further side, by which 

 one can get some little assistance in the ascent. Its size depends on 

 the rainfall above. It swells quickly after a heavy shower, the rocks 

 being already saturated ; and in the wet season there must be a very 

 consiilerable fall. Roraima-paru would be a general name for all these 

 placep, though a particular one bears the name Kamaiva — hence the 

 name of the village, Kamaivawong. 



The disintegration 8.nd breaking away of the upper part of the 

 wall, which have left this uneven ledge, must have largely been 

 caused by this stream and the numerous others that soak the upper 

 edges, in the wet season especially. Even now a rapid process of 

 alteration is going on, here particularly, and on the western slope down 

 which must rush a great accumulation of water from above, very largely 

 through the wide channel by which one enters on to the plateau. The 

 outer part of this slope is swept clean below of any kind of vegetation. 

 The sand.stone rock all over the mountain is very variable in its 



