130 HEAT 0E EARTHS OUTER CRUST, 



any given place. Let a thermometer, with a suf- 

 ficiently long stem, have its bulb sunk between 

 3 and 6 feet deep in the earth, so that the gra- 

 duated scale projects above the ground, and that 

 consequently the oscillations of the column of 

 mercury may be observed without withdrawing 

 the instrument. One observation in the day is 

 then perfectly sufficient, since all accidental ex- 

 ternal variations, and even the regular ones during 

 the day disappear at tliis depth. Indeed, it 

 appears that at a depth of 6 feet, four observa- 

 tions during the month are sufficient to enable 

 you to deduce the true mean for the month, and, 

 from twelve such monthly data, the mean of the 

 year. 



We may obtain information concerning the 

 temperature of the ground at any place by ex- 

 amining the springs, but our results are less cer- 

 tain, unless we are perfectly sure that the springs 

 have their sources at very moderate depths. 

 Springs, which rise from a depth of 30 feet or 

 more, give, as you will readily understand, too 

 high a temperature ; and that this is the case with 

 them you may infer with the greater likelihood of 

 truth, the less are the variations to which they are 

 subject in the course of the year. Springs of an 

 invariable, or nearly invariable temperature, have 

 their origin undoubtedly at considerable depths, 

 and are allied in their nature to hot wells. 



