HOT WINDS OF THE LLANOS. 141 



mercury in the barometer, the waters of the Apure have 

 only a fall of thirty-four toises from San Fernando to the 

 sea. The fall from the mouths of the Osage and the 

 Missouri to the bar of the Mississippi is not more con- 

 siderable. The savannahs of Lower Louisiana everywhere 

 remind us of the savannahs of the Lower Orinoco. 



During our stay of three days in the little town of San 

 Pernando, we lodged with the Capuchin missionary, who 

 lived much at his ease. We were recommended to him by 

 the bishop of Caracas, and he showed us the most obliging 

 attention. He consulted me on the works that had been 

 undertaken to prevent the flood from undermining the shore 

 on which the town was built. The flowing of the Portuguesa 

 into the Apure gives the latter an impulse towards south- 

 east ; and, instead of procuring a freer course for the river^ 

 attempts were made to confine it bv dykes and piers. It 

 was easy to predict that these would be rapidly destroyed 

 by the sweD of the waters, the shore having been weakened 

 by taking away the earth from behind the dyke to employ 

 it in these hydraulic constructions. 



San Fernando is celebrated for the excessive heat which 

 prevails there the greater part of the year; and before I 

 begin the recital of our long navigation on the rivers, I 

 shall relate some facts calculated to throw light on the 

 meteorology of the tropics. We went, provided with ther- 

 mometers, to the flat shores covered with white sand which 

 border the river Apure. At two in the afternoon I found 

 the sand, wherever it was exposed to the sun, at 52'5. 

 The instrument, raised eighteen inches above the sand, 

 marked 42'8, and at six feet high 38'7. The temperature 

 of the air under the shade of a ceiba was 36'2. These 

 observations were made during a dead calm. As soon as 

 the wind began to blow, the temperature of the air rose 3 

 higher, yet we were not enveloped by a wind of sand, but 

 the strata of air had been in contact with a soil more 

 strongly heated, or through which whirlwinds of sand had 

 passed. This western part of the Llanos is the hottest, 

 because it receives air that has already crossed the rest of 

 the barren steppe. The same difference has been observed 

 between the eastern and western parts of the deserts of 

 Africa, where the trade-winds blow. 



