Ivi ITINERARY. 



cotyledonous plants held the pegs securely for the ropes. AVe cut 

 away the spiny leaves of some Abolhoda and Connellia and some hard 

 stems of tStcgolepis, and as the ground was very wet, covered it with a 

 thick layer of Bonnetia tops hefore laying down the waterproof sheet. 

 The site was a low ridge overlooking the gorge hy which we entered 

 the plateau, and the distance for the men Avas thus very short. As it 

 happened, it was by no means the best situation, for just a day or two 

 before going down, I came upon a deep, wide, and perfectly dry cave 

 under a great pile of rocks, that would have been a splendid working 

 room for drying the plants, and a refuge for all of us, being protected 

 in the front by high slabs thnt allowed a good light to pass in over 

 them. It was only a little farther in towards the East, on the left of 

 a wide open valley, but ajiproached by a rather nnrrow passage which 

 had served to hide it. 



The seven illustrations of the summit in Vol. I., from the photo- 

 graphs taken by McConnell, show very clearly the genei-al configuration 

 of the plateau and the prevailing character of its vegetation. Another 

 photograph gives a very good picture of a sheltered site, where trees of 

 three new species of Bidi/mopanax, /Sciadophj/llum, and Stifftia were 

 growing. They were mostly very thick-stemmed, spreading and ixm- 

 brageous, the tallest of about 15 feet or more in height; and Jacob 

 was sent up to stand in the outer branches to give an indication of the 

 size of one of them, and that one not the largest, being on the outside, 

 as seen in the photograph. 



It mny be said at once, there is no forest anywhere on the top of 

 Roraima. Wherever trees of the genera mentioned or of Bonnetia, 

 Vv'liich ni.ay be much taller, are found, the clumjis are always small, 

 situated where some sort of soil has accumulated, a\\ay from any great 

 wash of water, and not exposed to the full blast of the north-eastern 

 winds. The usual position is in some curve or angle of the valleys, 

 facing south or west ; and in one such I came upon an unusual growth 

 of a tree that at a distance looked from its habit like a palm. It must 

 have been some 25 feet or more in height, with a clean stem, un- 

 branched except at the top, where there was a thick head of leaves, 

 from the apparent size of which I concluded it must be of one of the 

 three genera mentioned above. It must h.ave been draAvn up in its 

 effort to reach fuller light by the crags against which it grew, and it 

 had the appearance of a tree which had grown among others and then 

 been isolated. Its site was not very accessible; and as it was not in 

 flower, nor- could I have cut it dou n at the time, no serious effort was 

 made to get under it. 



The most favourable situation for the tallest Bonnetias Avas in deep 

 narrow guilies hollowed out in the wider valleys or open parts of the 

 plateau. The stronger sunlight above had in places drawn up the 

 plants to what I estimated as being fi-om 35-40 feet. Here also were 

 the most luxuriant growths of such plants as the myrtles, cyrillas, 

 hemp-agrimony, melastomes, tree-ferns, sedges, and grasses, and mis- 

 cellaneous others. 



It will thus be realised that in a general glance round the summit, 

 there is nothing to relieA^e the desolation of the disintegrated sand- 

 stone of which it is composed ; and the first impression is as if some 



