162 THE SAVANNAHS OF GUYANA. [Sess. Luxx. 
man: the grasses are there, but seldom in dense tufts; or 
the grasses are not there, because the savannah is not always 
grassland. Very often, as on the savannahs to the east of 
the Surinam river, we find only here and there small patches 
of grass and large parts of the savannah are without a 
sprig of grass. The description of the savannahs on high 
plateaux corresponds a little more to the savannahs of 
Guyana, which, however, are not on high plateaux. For 
although. according to my opinion, not very much of a 
European meadow is to be found in the savannahs of Guyana, 
nothing is more characteristic of them than the scattered 
trees, stunted, gnarled, dwarf, with occasionally lofty indi- 
viduals, which belong to species not present in the forest. 
True is it always that palms occur in the savannah, 
especially the Mauritia flexuosa, perhaps one of the most 
grandiose palm-trees in the world. It is worth noting here 
that these scattered dwarf-trees differ again in species from 
east to west. In the savannahs of the Commewyne river, 
the stunted Psidium gujave, the poisonous Euphorbia cotint- 
Jolia, and the gregarious Clibadium surinamense prevail. In 
the savannahs of the Surinam river the Huphorbia cotinifolia 
disappears, the Psidium and Clibadiwm diminish, whereas the 
Melicocea bijuga,a Sapindaceous plant, and the Anacardium 
ocudentale predominate. In the Lara district, however, the 
Melicocca disappears nearly entirely, while the other above- 
named plants are equally distributed amongst Apocynacez, 
Clusiacez, Rubiaceew, Myrtaceex, and are there in consider- 
able numbers. On these savannahs we find, but not very 
often, a curious plant with a delicious fruit, the Bati-bati, of 
which, up till now, the scientific name is not known, and no 
description could be found in the notes of any one of the few 
botanists who have visited the Guyanas or studied their 
flora. Of the western districts I have few recollections; 
but, so far as I remember, grasses and grass-like Liliaceous 
plants prevail, Melastomacee frequently occur, Cactuses 
here and there appear on elevations of the undulating soil. 
If we consider now the descriptions given for savannahs in 
America, that given by Schomburgk, especially for Guyana, is 
vague, more metaphorical than satisfactory ; the descriptions 
of the savannahs of Minas Geraes and other parts of Brazil 
can hardly be said to apply entirely to Guyana. What can 
