/ ' > • m Subterranean Temperature. 41 



Mimed that the verifications have been made, there is no cer- 

 tainty with respect to them. 



Second Consuleraticm. — The increase of the subterranean 

 temperature is calculated by comparing the results of experi- 

 ments made in deep places either with the results obtained at a 

 higher level, or with the mean temperature of the country, 

 which is preferable. In the latter case, the accuracy of the com- 

 parison is only perfect when the mean temperature in question 

 is precisely known. Now, with the exception of the experi- 

 ments of the Paris Observatory, I do not believe that the means 

 which I have employed above, can be depended upon as correct 

 to within half a degree on either side. 



Third Consideration. — Doubts may be entertained regarding 

 the absolute depth of the points at which most of the observa- 

 tions have been made. It appears probable, that almost all 

 the observers have referred this depth to the plane in which 

 the entrance of the shaft nearest the stations is situated. If 

 they had taken pains to determine, by rigorous operations, the 

 distance comprised between each station and the point of the 

 upper surface of the rock situated in the same vertical line, they 

 would not have failed to mention it. Now, as the shafts are sel- 

 dom opened on heights, it is to be presumed that most of the 

 depths assigned are too small, and that, consequently, taking 

 this cause alone into consideration, the augmentations of tempera- 

 ture which might be deduced from them would be too great. 



The influence of this cause of inaccuracy is so much the less, 

 the deeper the places are in which experiments are made ; and 

 this observation is equally applicable to the two other causes 

 which we have examined. 



Fourth Consideration, — All the statements collected respect- 

 ing the temperature of fresh water springs, of artificial spouting 

 fountains, of currents of water sufficiently large to turn machi- 

 nery at their origin, such as the river Isle, near Vaucluse, and 

 the Touve, near Angouleme, concur to prove the increase of 

 subterranean heat. The only exception which I know is, when 

 the springs are overlooked by high mountains, on which a great 

 quantity of snow annually melts. It is to this case of except 

 tion that the two following facts refer, which were observed by 

 Saussure, the first in the month of August 1789, the other on 



