OF TEMPEKATUBK. 97 



by no means so satisfactory as could be desired. In 

 making these, the nature of the soil in which the 

 thermometers are plunged, should, among other 

 circumstances, be very precisely described; for it 

 is obvious that the result will be essentially aifecled 

 by the peculiar conducting power of the earth. 



119. But although we have no geothermometrical 

 observations which have a direct relation to the con- 

 nexion between terrestrial temperature and vege- 

 tation, yet an approximation to the amount of heat in 

 the earth may perhaps be obtained indirectly. It 

 seems improbable that the surface of the earth should 

 be colder than the mean temperature of the air that 

 rests upon it ; and it seems certain, from the evidence 

 afforded by this country (116), that, in fact, it is 

 at least a degree or two above it ; therefore, in the 

 tropical parts of America, where Humboldt found 

 the mean temperature of the coldest month not to be 

 lower than 79'16° at Cumana, we shall be justified in 

 concluding that the temperature of the earth's surface 

 never falls permanently below that amotmt ; and as 

 the mean summer temperature* of the place was 

 found to be 82 "04°, so it is probable that the earth 

 will have something above that degree of warmth, on 

 an average, in the sammer.f 



* For the warmest month, this great observer gives 84-38° as 

 the mean ; -which corresponds remarkably with the temperature 

 a foot below the svirfaoe in New Grenada, where, according to 

 u correspondent of Mr. Hay, it is 85° during summer, "as a gen- 

 tleman, a planter there, wrote home for his information." {See 

 Zoudon'i Gard. Mag., vi 437.) 



f [The mean temperature of the State of N«w-Tork, for fourteen 



5 



