Hot and Thermal Springs. 347 



means alone that springs are enabled to take their rise under 

 the glaciers, and to flow uninterruptedly throughout the winter. 

 Only in such places where these high glacier-bearing longitu- 

 dinal valleys of the Alps are intersected by cross valleys, or 

 where high glacier-bearing mountains stretch down into deep 

 valleys, will the towering glacier masses be foi'ced down the 

 steep declivities, and urged forwards by their own weight into 

 the valleys situated below the limits of perennial snow, where 

 they will continue uninterruptedly to melt away from under- 

 neath, whilst their upper surface will only be melted during the 

 warmer seasons." 



The melting of the glaciers from underneath can only take 

 place where the mean temperature of the soil is above 32". For 

 at such elevations, where it is equal to or below 32°, and where 

 the glacier prevents the access of the warm summer air to the 

 soil, it will no longer be possible for the ice to be melted on its 

 under surface. Let us suppose an alpine valley, the mean tem- 

 perature of which is 32°, to be covered with a glacier at a time 

 when the temperature at the surface is also 32°, and that the 

 masses of snow have the same temperature, the snow will then 

 neither receive heat from, nor give heat to, the soil beneath. If 

 the temperature of the soil be above 32°, this excess of tempera- 

 ture will be expended in melting a part of the snow, until the 

 equilibrium be restored ; in the contrary case, the soil will re- 

 ceive heat from the snow. The law of the decrease of tempera- 

 ture from the centre of the earth to its surface is, therefore, not 

 permanently deranged by a covering of snow, where the mean 

 temperature of the soil is 32°, and it is consequently impossi- 

 ble to conceive, that a greater quantity of heat should at any 

 lime be emitted from the interior, by which the snow should be 

 melted. The covering of snow will produce no other effect, 

 except that the variations of the temperature of the soil, occa- 

 sioned by the variations of the seasons, will be confined within 

 very narrow limits, or if the covering of snow be very thick, 

 that they will entirely disappear. However, as the temperature 

 of the surface is only the consequence of the cooling of the 

 earth, caused by the radiation of heat from its surface, which 

 cooling is not compensated in all climates by the radiation from 

 the sun, it is possible ihat, benealli a glacier where no radiation 



