Information respecting Botanical Travellers, 195 



The great object of the expedition determined upon during the 

 favourable season of 1838-39 was to connect my researches with 

 those of the celebrated Humboldt at Esmeralda, on the Uf)per Ori- 

 noco. This was a task not without its dangers, but such that, should 

 I be able to reach my aim, the success would prove its own reward, 

 and another chasm in the geography of South America would be 

 filled up. I succeeded, and reached Esmeralda from the east on Fe- 

 bruary 22d, 1839, after a period of nearly five months from the 

 time of our departure from the Brazilian boundary Fort San Joaquim, 

 and after enduring the greatest hardships and deprivations. But while 

 the details and geographical discoveries made during this expedition, 

 will doubtless appear in the Journal of the Royal Geographical So- 

 ciety, under whose auspices I travelled, I may here give a general 

 view of the natural productions which distinguished the regions over 

 which our steps, and canoes carried us. 



San Joaquim do Rio Branco is situated at the junction of the river 

 Takutu with the river Parima of the natives, or Branco of the Portu- 

 guese. We may almost consider it, as being situated in the middle 

 cf those extensive savannahs, bounded by the Carawaimi, Canucu, 

 Pacaraima, and Parima mountains, no doubt once the basin, and the 

 origin of the tradition of that mysterious lake which figured for cen- 

 turies on our maps, and which even the extensive geographical re- 

 searches of a Humboldt, and his proofs of its present non-existence, 

 have not been able to erase entirely from some recent maps. 



These savannahs are covered with Gramineee, and Cypei^acece, 

 chiefly of the tribes Panisecs, Chloridece, Agrostidece, interspersed 

 with the genera Cyperus, Car ex, Scirpus, &c. And while the former 

 afford food to numerous herds of cattle, the sedges give the appear- 

 ance of herbage to regions which otherwise would appear waste, and 

 barren. The savannahs consist of undulating ground ; this is just 

 as we might expect, admitting the opinion, that they formed once 

 the bed of a vast inland lake. Many parts are marshy ; even the 

 lower declivities of those gradually rising hills, exhibit frequently a 

 greater moisture than the parched appearance of the environs would 

 warrant, and to the explanation of which the severe dews, the cha- 

 racteristic meteorological phaenomenon under the tropics, and the 

 mechanical process of percolation, give a clue. These moist places 

 are distinguished by a livelier vegetation, consisting chiefly of spe- 

 cies of the genera Utricularia, Eriocaulon, Xyris, Pontederia, Drosera, 

 Marica, &c., high above which the ma-'jestic Mauritia flexuosa towers 

 its fan-shaped leaves, and gives that peculiar aspect to the landscape, 

 which stamps it at once with undeniable features as tropical. Accord »- 



p2 



