OBSERVED DURING THE RORAIMA EXPEDITION. 269 



The chief constituents of this turf-like vegetation were vast quantities of a new species 

 of Papal an thus (P. Roraima}, Oliv., n. sp. [No. 294]), and great masses of S/ihoguum-liku 

 mosses. In the latter grew, in such ahundance as to redden the ground, the pretty little 

 Sundew (Drosera communis, A. St.-H. [No. 313]). Groups of very luxuriant Pitcher- 

 plants (Ileliamphora) were there also. Great quantities of tiny shrubs, of alpine character, 

 interwove their branches with each other and with the mosses ; among these were 

 We'mmannia guianensis, Klotzsch [No. 327], Marcetia juniperina, DC. [No. 319], Psy- 

 chotria concinna, Oliv., n. sp. Baccliaris [No. 241], Ledolhamuus [No. 308], Befaria 

 [No. 310], Vacciniiim [Nos. 326, 329], Pernettya [No. 333, ex parte], and Gaultheria 

 [No. 332]. The small Epidendra, as on the ledge, were here too, as was also the tiny 

 Misseltoe (Phoradendron [No. 323]) and the Fritillary-like Lisianthits [No. 306]. 



A beautiful Tofieldia (T. Schomburgkiana, Oliv., n. sp. [No. 297]) and the somewhat 

 similar Nietneria corymbosa, Kl. & Sch. [No. 298], with large yellow flowers, were 

 conspicuous. 



In the crevices of the rocks the vegetation was different. There was a very beautiful 

 TJtricularia (TJ. montcma? Jacq. [No. 293]), larger and deeper in colour, but slightly less 

 graceful, than TJ. CampbeUiana, and there were three species of fern. One of these 

 latter was a very stunted form of Lindsay a striata, Dryand. [No. 301], which, in its 

 ordinary form, is common in many parts of Guiana. The other two were absolutely new 

 — one a Eymeiwphyllum, which Mr. Baker has named H. defectum, Baker, n. sp., [No. 

 318] ; the other a Gymnogramma ( G. cyclophylla, Baker, n. sp. [No. 295], a second species 

 of the same group of this genus to which belongs G. elapltoglossoides, Baker, n. sp., 

 [Nos. 101, 215], found on the lower slopes of Roraima. Only one other species of this 

 very distinct group is known, and that has been found in the Amazon valley. 



I have now briefly noticed the most striking plants which we met with on Roraima ; 

 but, before closing this paper, there are one or two points which I wish, finally, to set 

 down in order. 



First, as to Brocchinia cordylinoides, Baker ; this is only known to occur on the 

 Kaieteur savannah and on Roraima, but in the latter place apparently only above a 

 height of 5500 feet. There is a remarkable difference of vigour in the habit of the plant 

 at these two places respectively. After seeing a large number of individuals of the plant 

 at both places, it is obvious that at the Kaieteur it attains a much greater size and forms 

 a much taller stem ; and, if I may judge from the comparative abundance or scarcity of 

 flower-stalks, it seems to flower much more freely at the Kaieteur than on Roraima. 

 A possible explanation of some of these facts seems to be that the position and the 

 circumstances that it finds on Roraima, are beneficial to the plant ; that the most impor- 

 tant of these circumstances of its existence is an atmosphere, like that of Roraima or of 

 the Kaieteur, so saturated with moisture as to effect the constant replenishment of the 

 large quantity of water retained in the leaf -axils of the plant ; and that the plant having 

 found its way to the Kaieteur (which, though much below the proper level, is atmospheri- 

 cally so peculiarly suited for it), it has taken root there and, in its new surroundings of 

 higher temperature, has there developed a new vigour. Lastly, as regards this plant, I 

 cannot refrain from once more alluding to its possible, even probable, distribution in the 

 other widely scattered distinct areas already enumerated. 



Closely connected with the Brocchinia is TJtricularia Humboldt ii. Like the Brocchinia, 



