256 MR. E. F. IM THURN ON THE PLANTS 



present, in most remarkable degree, the appearance of the well-marked designs laid out by 

 a landscape-gardener ; the whole effect is bike that of an artificial garden, with regular 

 groups of shrubs separated by wide paths and roads of clean bare rock. Moreover, it is 

 not only in the fissures that plants grow on this savannah. As on the eppellings, so here 

 too, a certain number of plants find sufficient foothold in the vegetable accumulations in 

 the slight depressions in the conglomerate sheet before these have^been engraved deeply 

 enough to leave the sandstone exposed and to make regular fissures. 



But not only is the arrangement of the vegetation of the savannah thus very remarkable ; 

 the plants composing this vegetation are also individually of great interest. As might 

 be expected, very few of them occur in the forest which everywhere, and for a great dis- 

 tance, surrounds this strange open space. Much more remarkable is it that very few of 

 these plants occur on the nearest savannah, nor, indeed, on the general savannah-land of 

 the interior. And, most noteworthy of all is it, a very large number of these peculiar 

 plants of this isolated savannah occur, often with slight but interesting differences, on 

 Roraima. 



By far the most striking, as it is also the most abundant, plant on the Kaieteur savannah 

 is a huge aloe-like Bromeliaceous plant, Brocchinia cordylinoides, Baker, which was 

 gathered there by Mr. Jenman and myself some years ago, but which was, until the Roraima 

 expedition, unknown elsewhere. This gigantic plant, so striking as to compel notice 

 even from the most unobservant traveller, is ranged in enormous numbers on the Kaie- 

 teur savannah, and indeed makes, to a large extent, the strangeness of that strange scene. 

 There the height of a full-grown specimen, under favourable circumstances, is about 14. 

 feet, and, in the older specimens at least, the crown of leaves is supported on a tall bare 

 stem. It seems also there to flower abundantly. We shall see that the plant occurs, but 

 with slightly different characters, on Roraima. Moreover, at the Kaieteur, in the axils 

 of the leaves of this Brocchinia, and only in that position, grows a very remarkable and 

 beautiful TJtricularia ( TJ. Humboldtii, Schombk.), with flower-stems 3 or 4 feet long, sup- 

 porting its many splendidly large violet flowers. This plant too we found on Roraima, 

 and with slightly different characters from those which it exhibits at the Kaieteur. 

 Another remarkable and distinct plant on the Kaieteur savannah is a low-growing 

 Brocchinia (B. reducta, Baker), also previously known only from there, and may be 

 roughly described as resembling three or four sheets of yellowish-grey foolscap paper rolled 

 loosely one round the other, the whole standing on one end of the roll. This plant I did 

 not observe on Roraima, though I feel convinced that it will one day be found there ; but 

 I did see it, in very considerable quantity, in one small district about halfway between the 

 Kaieteur and Roraima. Only one other plant common, but with a difference of form, 

 to the two districts can be mentioned here. Mr. Jenman found at the Kaieteur a very 

 striking new Iforonobea (31. Jenmani, Engl.) ; and I found on Roraima another very 

 remarkable congener (II. intermedia, Engl., No. 337), of which its describer says that it 

 is intermediate between M. riparia and M. Jenmani. 



In short, the Kaieteur savannah and Roraima may be regarded as two isolated areas 

 marked by a very peculiar vegetation, which vegetation is, however, to a noteworthy 

 extent, common to the two. 



