velvety sward, still wet with dew, directly northward, till I was attracted by a dense clump of tree-like plants. These were indeed remarkable. Their naked stems, several feet in circumference, at last branched in two, while at the ends of these branches were long grass-like broad leaves. In the absence of flowers and fruits it was impossible to determine whether these remarkable plants belonged to the Pandanacea or Vellozieæ. These strange plants rose straight up into the air from the sandstone rubble, which was covered by an Eriocaulon and a curious grey-black grass. My brother had seen a group of these plants in 1838 when he first ascended this sandstone range; but on that occasion also they were not in flower or fruit. On reaching the declivity, a breeze from the north came loaded with a delicious scent, and our astonished eyes were attracted by innumerable stems of white, violet, and purple flowers which waved about the surrounding bush. These were groups of superb Sobralias; and amongst them S. Elizabetha rose tallest of all. I found flowering stems of from five to six feet high. But not only these orchids, but the shrubs and the low trees, still dripping with heavy dew, were unknown to me. Every shrub, herb, and tree was new to me, if not as to its family, yet as to species. I stood on the border of an unknown plant zone full of wondrous forms, which lay, as if by magic, before me. I once again felt the same delighted surprise which had overpowered me when I first landed on the South American continent; but I now seemed to be transported to a new quarter of the globe, amongst the Proteacea of Africa and New Holland and the Melaleucea of the East Indies and Australia. The leathery, stiff leaves, the curiously coiled branches, the strange large flowers of various forms, the dazzling colour of these-all were essentially different in character from all vegetation that I had before seen. I did not know whether to look first at the wax-like gay flowers of certain species of Thibaudia, Befaria, and Archytaea, or at the large, camellia-like flowers of a Bonnetia, or whether to fasten my eyes on the flower-loaded plants of various kinds of

HOW TO REACH RORAIMA.

85

Melastoma, Abolboda, Vochysia, Ternstromia, Andromeda, Clusia, Kielmeyera, or on the various new forms of Sobralia, Oncidium, Cattleya, Odontoglossum, and Epidendron, which covered the blocks of soft sandstone-and there were very many plants not at the time in flower. . . . . Every step revealed something new.'

To the ethnologist also the district will prove interesting; for it is so remote and unexplored that the Arecuna Indians, who chiefly inhabit it, are in a very unusually primitive condition-for instance, they alone still sometimes use stone, instead of iron, girdles for baking purposes; and, moreover, the strangeness of Roraima seems to have made deep impressions on the minds of these Indians, and to have filled their thoughts with folk-lore to an unusual extent.

In short, there is a great reward in store for the traveller, whether he be botanist or ethnologist, who, having sufficient pecuniary means, will first gain experience of the ways of travelling in that part of the world, so that his knowledge of ordering an expedition may be as precise as possible, and will then go to Roraima prepared, at all costs, to spend as long a time in the district as may enable him to make his way slowly round the mountain; and his labours will possibly result, as no other means can, in the discovery of a way even to the top of Roraima.

A few words will not be out of place as to the best way of approaching the mountain. Schomburgk, Brown, and Eddington visited it by going up the Essequibo and Roopoonooni to the neighbourhood of Pirara, a route which I have already described, and then making their way northward to the mountain; McTurk and Boddam-Wetham went up the Mazeruni and then walked southward to the mountain. But there is a way, as yet untried, which I am convinced will prove far more practicable. This is up to Potaro and from there westward across the savannah. I have already described the journey up the Potaro as far as the Kaieteur fall; and it is evident that there need be no great difficulty in taking boats of considerable size up to that point.

There a depôt should be formed. Small boats might either be carried past the fall and launched on the Potaro above the Kaieteur, or these might be procured from the Ackawoi Indians who live above, but at some distance from, the fall; and thus the expedition might proceed by water yet further in the direction of Roraima. The walk across the savannah from the point where it may be necessary to leave the river would be shorter and almost certainly less laborious than the corresponding walk by either of the other routes.

But, as my last word on this subject, I must strongly warn any against approaching Roraima without first fully weighing the difficulty and the cost.

CHAPTER IV.

ASPECTS OF PLANT-LIFE.

Common Misconception of Tropical Scenery-The Special Case of Guiana -General Type of Foliage like that of Temperate Climates-Colouring of the Foliage-Colour of Flowers in Mass-Beauty of Individual Flowers-Scent-Chief Types of Guiana Vegetation-A Scene in the Forest-Palm Forests-River-side Vegetation-The Cokerite PalmSavannah Scenery-Water Plants.

THE appearance of a country which has been little modified by the hand of man, depends, in very great measure, upon its vegetation. Much has lately been written on the real, as opposed to the commonly conceived, appearance of tropical vegetation. But men in temperate regions are still apt to think that tropical plant-life blazes with gorgeous colour, and is composed almost exclusively of quaint forms. Two fallacies, as to colour and as to form, are involved in this conception. The spread of the colour-fallacy is due to the fact that it is the more gorgeous plants which, being selected from an infinitely greater number of less brilliant hue, are grouped together in our glass-houses. The form-fallacy has arisen partly from a similar cause, but chiefly from the fancy sketches of tropical scenery made by artists. This latter source of error may be well studied, for example, in certain pictures of Guiana scenery by a German named Carl Appun, a botanist and a draughtsman of some merit, who lived for some years in the interior, and who has furnished almost the only attainable pictures, drawn on the spot, of that scenery. In these pictures palms and other plants of forms strange to temperate regions, occupy the whole scene. Appun knew how to draw plants, so that even in his most crowded compositions it is

possible to recognise the species of each individual plant; but his pictures are no true records of the scenery, because in them, much as the gardener does in his hothouse, he grouped only the most striking plants, and entirely omitted all such as are but little distinguished in character or size of foliage from the plants of temperate regions. In correction of the false views thus spread Mr. Wallace's careful analysis of tropical scenery in general, in his admirable essay on Tropical Nature, is of great value. The purpose of the present chapter is to supplement, as far as may be, that general account by representing the most characteristic aspects of the special plant-life of Guiana.

The forests and woods of Guiana, which, it must always be remembered, are situated at a very low level above the sea, are mainly composed of trees and shrubs of much the same general type, as regards both form of growth and of foliage, as our own Spanish chestnuts, oaks, acacias, and laurels. Three things must, however, be remembered in thus transferring in imagination our own forms of vegetation to the tropics; and these are, in the first place, that in the tropics, the trees and plants of all sorts are generally on a much more gigantic scale, and that this rank growth and, especially at this low level, the absence of small neat-growing plants, such as elsewhere carpet the ground and fill up the spaces, gives an impression of weediness; secondly, that the light being much more intense, the spaces within the gigantic outlines of the scenery are seen in even exaggerated bareness and nakedness; and thirdly, that scattered among these familiar forms a large number of novel forms

occur.

Starting with this general idea of the vegetation, it will be convenient, first, to consider the three special points, as to the occurrence of colour, of novel and striking forms, and of scents, and then to draw, as far as may be done in words, a few typical scenes of vegetable life in the forest and on the savannah.

The general colour of the forest is due rather to the