172 RELATION BETWEEN CLIMATE AND VEGETATION 



which abounded on both banks of the upper part of the river, 

 we saw vast barren plains or immense forests of the Pahn Ca- 

 randa (^Copernicia cerifera), the trunk of which either bristles 

 with the persistent petioles of thtJ bygone leaves, or is slender, 

 naked, and terminating in a roundish head of dove-tailed leaves, 

 something like our own Chatnarops. From the bastions of the 

 fortress of Paraguay may be seen vast tracts of country covered 

 witli the Caranda, the glaucous foliage of which produces a most 

 remarkable and extremely fine eftect. This plain is the Gran- 

 Chaco, inhabited by the Guaycourous and Tobas Indians, and 

 bounded by the Rio Paraguay on the one side, but extending 

 towards the west to the foot of tlie Andes, south of Bolivia, 

 where it is lost among the Pampas of the Argentine Republic. 



During our stay at Fort Bourbon I made daily excursions in 

 the neighbourhood, and gathered a large number of interesting 

 plants either in the plains, or the woods on the river side, or on 

 some of the conical hills covered with a large Cereus, and which 

 are scattered all over the environs. A few days before our 

 departure the grass of the prairie accidentally took fire, and it 

 spread with such rapidity that by the next morning the beautiful 

 verdure we had so much admired was converted into a dismal 

 black. The Carandas alone escaped, and their blueish heads 

 were brought still more into relief by the black ground with 

 which they were contrasted. For forty-eight hours the eye 

 could track the course of the destroying element, but after that 

 it disappeared in the distance. 



The temperature of Fort Bourbon is, as ascertained by M. 

 Boussingault's method, 28° C.,* which would make it one of the 

 hottest jdaces in the world ; but this number is evidently too 

 high. In the interior of the fort the thermometer marked in the 

 shade at 2 or 3 o'clock in the afternoon nearly 40° C, and in 

 the sun as much as 50° C. The temperature of the water was 

 constantly, during our stay, about 29° or 30° C. 



On our return from Albuquerque we found nothing of any 

 interest. 



Our next expedition was to Miranda, a Brazilian settlement 



* The fort being upou a small hillock, it is easy to understand that the 

 soil is heated to a greater depth than in the plains. The water found at the 

 bottom of a cavern which we visited near Nova Coimbra was 24° C, which 

 is probably the mean temperature of this district. 



Although there are no doubt accidental circumstances which invalidate 

 the results obtained by M. Boussingault's method, its general utility cannot 

 be denied. My friend Mr. Pentland, to whom I entrusted the observations 

 to be made with the thermometer in both Bolivia and Peru, is of opinion 

 that the results obtained by examining the higher parts of those countries 

 are less correct than those similarly obtained in other parts of the Equi- 

 noctial Zone. I am of the same opinion. 



