.OF L O K A R N F. 407 



ilcd by the rope into the \vell ; in the bottom of 

 which it is left for half an hour. With this ap- 

 paratus, 1 obtain what I want: I can now dif- 

 ccrn the pfccife heat of the water by the thermo- 

 meter, without fear of the mercury falling when 

 it is taken out. 



In the cave of the Royal Obfcrvatory at Pa- 

 ris, the mercury of the thermometer Rands 

 through the whole year at the fame degree, cor- 

 rcfponding to the twelfth in our Swcdifh ther- 

 mometer. The waters of Medvi and Lokarnc, 

 therefore, and of other perennial fp rings which 

 have come under my obfcrvation, exceed that 

 temperature only by one degree. The fame 

 temperature is found to prevail in fubterraneous 

 cavities. Now, as the water on the furfacc of 

 the earth is fupplied by lakes and fubterraneous 

 citterns, at Icaft asdittant from the centre of the 

 earth as thcfe fprings ; it follows that the fourcc 

 of thofe wells mufl be extremely deep, whofc 

 temperature is only fix degrees. Bclides, the 

 openings of fuch fprings are ufually inclined to 

 the horizon. In general, however, the heat of 

 water on the furfacc of the earth is different at 

 different feafons in the year. In natural cuvitcs 

 in mountains, although on the fame level with 

 the adjacent plains, water cannot but be uncom- 

 monly cold. But I can fcarcc think there arc 

 many places in which it can become fo cold in 



fu miner 



