Hot and Thermal Springs. 371 



Before we arrive at this knowledge, it may be ascertained 

 whether probably correct results will be obtained, if we sup. 

 pose the mean decrease of temperature in the soil, found from 

 Boussingault's observations between the tropics up to the snowy 

 regions, and which so nearly coincides with the result of my 

 observations up to the height of 1173 feet, to be the same for 

 the whole height up to the snowy regions on the Alps in our 

 latitudes. In the following table I have calculated the height, 

 at which the mean temperature will be 32°, from the known 

 mean temperatures of several elevated spots, for the most part 

 in the vicinity of the Alps, always supposing 2°.25 difference 

 of temperature to correspond to 677 feet difference of height. 



With the exception of the Lowenberg (which lies 3° to S" 

 further to the north than the rest of the places) and .S"^ Bern- 



northern side of the chain to lise to a height of 15,900 feet. However, I hope 

 to prove further on, that the law of the decrease of temperature in the soil 

 is not necessarily disturbed by differences in the height of the limits of peren- 

 nial snow. 



I am at present engaged in making maximum and minimum thermometers 

 which shall not be subject to the accidents of the usual ones. Should this 

 succeed, and the experiments yield favourable results, these thermometers 

 would considerably facilitate observations in uninhabited heights and such as 

 are difficult of access. If, for example, they were to be buried four feet deep 

 and left for a whole year, the yearly maximum and minimum would be suffi- 

 cient to find the mean with the required exactness. In this manner it would 

 even be possible to ascertain the mean temperature on the toj) oi Mont lilanc 

 by bur^ring maxifuynpi and minimum thermometers deep in the perpetual 

 snow. '-- '^ 



T 2 



