FLORA OP THE SAVANNAHS. 461 



arriving at a field like this, so much unlike anything he has ever 

 before seen. Here are full-blowing large orchids, with red, white, 

 and yellow flowers ; and, among the grasses, smaller ones of great 

 variety, and as great scientific interest ; melastomaceous plants 

 of various genera (Arthrostemma and Osbeckia) utricularias, 

 droseras, rare and various grasses, and cyperoids of small sizes 

 and fine kinds, with a species of Cassytlia ; in the water, Oerato- 

 pliyllum and bog-mosses. Such a variety of forms and colours is 

 nowhere else to be seen, or met with, in the island. This scenery 

 is enlivened by groves of moriche and cabbage palms growing, 

 here and there, in great luxuriance in the more inundated 

 spots. 



The transition from this kind of savannah to the dry savannah 

 of the plain, or plateau, is exemplified in those known by the 

 name of O'Mara and Piarco. They do not exhibit that variety 

 of plants which adorn those already mentioned, but they still 

 retain the same striking features, and the soil is undoubtedly of 

 the same composition. The grasses are higher and coarser, and 

 the savannahs themselves interspersed with two small trees the 

 Chaparro (Curaiella) and a Bunclwsia ; here and there also a 

 shrub of Miconia, or Vismia, with byttnerias, ruellias, and 

 osbeckias. In the water-courses are to be found two plants, viz., 

 an Eriocaulon and Tonina, which are, perhaps, to be found also 

 at Aripo, together with the following : Xyris, Mayaca, Hydrolea, 

 and, bordering on the high wood, Rapatea. 



Somewhat resembling the above, but more destitute of similar 

 shrubs, are the savannahs of Icacos, and those of Savanetta and 

 Couva. The driest, however, as also the least interesting of our 

 " natural savannahs," is that situated near Arima. It is over- 

 grown with chaparros, under whose scanty shade grow coarse 

 scrophularacese, principally Beyrichia, with a smaller quota of 

 grasses and cyperoids ; and, in moist spots, an abundance of 

 Heliconia Psittacorum. 



I would not, without much hesitation, class the savannahs on 

 the mountains from St. Joseph to Arima among natural 

 savannahs. They may possibly owe their origin to the destruction 

 of forests by fire, in parts where the layer of soil was too thin to 

 nourish a fresh generation of trees. They offer little interest to 

 the botanist, and yet their exploration is rather difficult, the high 

 rank grass they produce Pennisetum, Sttaria, Andropoyon, &c. 



