154 



of Cuba* to 22° or 23°. This increase is near- 

 ly in proportion to that of the mean tempera- 

 ture of the atmosphere, from the latitude of 48° 

 to the tropics. 



We have just seen, that in the Cueva del 

 Guacharo, the water of the river is nearly 2° 

 colder than the ambient air of the cavern. 

 The water, whether in filtering through the 

 rocks, or in running on stony beds, assumes 

 without doubt the temperature of these beds. 

 The air contained in the grotto, on the contrary, 

 is not in repose g it communicates with the at- 

 mosphere without. Though under the torrid 

 zone, the changes of the exterior temperature 

 are exceedingly small ; currents are formed, 

 which modify periodically the internal air. It 

 is consequently the temperature of the waters, 

 that of 16 8°, which we might look upon as the 

 temperature of the earth in those mountains, if 

 we were sure, that these waters do not descend 

 with rapidity from more elevated neighbouring 

 mountains. 



It follows from these observations, that, when 

 we cannot obtain results perfectly exact, we 

 find at least under each zone certain numbers, 

 which indicate the maximum and minimum. At 

 Caripe, in the equinoctial zone, at 500 toises of 



* Obs. Astr. Tom. i, p. 134. The mean temperature of 

 the air at the Ha van n ah, according to Mr. Ferrer, is 25 6°. 



