Hot and Thermal Springs. 385 



In opposition to these conclusions it may be remarked, that 

 the water, when cooled at the surface by the air, does not sink 

 with its reduced temperature quite to the bottom, but becomes 

 somewhat warmed again during its descent, at the expense of 

 the warmer layers of water, through which it passes. And this 

 is undoubtedly the case. But since the sinking of the cold 

 currents of water must necessarily be attended with the rising 

 of the warmer water from below, which in its turn is also 

 cooled when it comes to the surface ; it follows, that during 

 the continued cold of winter, the water at the bottom must at 

 one time be reduced to a temperature of 38°. 75. And if this 

 should even only take place once during a severe winter, there 

 being no communication of heat from above, it seems difficult to 

 account for a temperature of 1°.35 to 5°.67 higher than 38o.75, 

 otherwise than by supposing it to be due to heat supplied from 

 the earth. Muncke* ascribes this higher temperature to the 

 action of the sun, as De Saussure's observations were not made 

 till May. But can it be supposed that the rays of the sun pe- 

 netrate to such a considerable depth as that in which the obser- 

 vations were made, and are still capable of having any effect, 

 when we consider how very rapidly the transparency of water 

 decreases with the depth ? It is true that Saussure's observa- 

 tions in the Lake of Geneva^ in the months of February and 

 August, show an increase of -|- 1°.26 ; but this difference might 

 perhaps be overlooked, when we consider the great difference of 

 the depths. 



If the water at the bottom of lakes should become heated by 

 the earth beneath, there will be a stream constantly flowing 

 upwards, which will rise higher the less the temperature at the 

 surface exceeds 38.°7. This stream will reach to that height in 

 which the temperature of water is equal to that of the stream. 

 This stream will continue to rise only so long as the tempera- 

 ture at the surface does not vary. But if a depression take 

 place, there will also be a descending current which will fall to 

 that depth in which the temperature of the water is equal to its 

 own. If the temperature of the lake from its surface to a great 



" Gelilers neues physikal, Worserbuch. viii. 742. in note. 



