Hot and Thermal Springs. 281 



the mean temperature of a spring would exactly give the mean 

 temperature of the place of its origin, so long as it be not a 

 mountain spring, which brings down a low temperature from 

 above. Indeed, if this depth were known, a single observa- 

 tion at any season of the year would be sufficient, if the scale 

 of the yearly variations in the temperature of the soil at the 

 same depth were also known. Running springs, which are 

 conducted through pipes for a long distance at a known depth 

 below the surface of the earth, would be very well suited to 

 this purpose, provided they did not descend from steep moun- 

 tains. From this we see how desirable it is, that observations 

 on the temperature of the soil at various depths, at different 

 places, and at different heights above the sea, should be multi- 

 plied as much as possible. 



The continuation of my observations in some of the above 

 springs, after the completion of the one year's series, gives us 

 an idea of the influence of the temperature of the air on that of 

 springs ; for if we find by interpolation the temperatures of 

 these springs on the same days in 1834-1835, on which they 

 were observed in 1835-1836, we have : 



