On the Underground Temperature Gradient. 



357 



Considering the low mean temperature assumed to exist at the 

 surface, there is probably no occasion to alter the above estimate 

 of the limiting thickness. But if we double it, and take it at 

 1200 feet, we shall be well on the safe side ; and there is no doubt 

 that for a very long period the ice-sheet in many places attained 

 a greater thickness than this. Now the melting-point of ice is 

 lowered 1° F. by the pressure of every 2730 feet in the vertical 

 thickness of overlying ice, so that the temperature at the base of the 

 ice-sheet may have been as low as 30° or even 29°, but can never 

 under any circumstances have risen higher than 32°. We cannot 

 err very greatly, therefore, in supposing the surface of the ice- 

 covered ground during the whole Glacial period to have remained 

 at a uniform temperature of 30°-5. The subsequent rise of 

 temperature may thus be estimated at 19° F. 



All geologists are agreed in attributing a duration of several, or 

 many, thousand years to the Glacial period. If this be correct, the 

 temperature at its close must for some miles have increased almost 

 uniformly with the depth. Thus, if in the accompanying figure 

 represent a point on the earth's surface, X a horizontal, 

 and Y a vertical line, the temperature-curve P A near the 

 surface must have been almost a straight line, N P representing 

 the temperature at the depth represented by N, or, rather, the 

 excess of temperature above 30°-5. If N were taken to represent 

 one foot, then N P would represent the temperature gradient. 



Now let us suppose that, at the end of the Glacial period, the 

 mean annual temperature at the surface rose suddenly by 19°; and 

 let the line 3 represent this amount. Then, at some subsequent 

 time, the tempei'ature-curve would be of the form B Q C, where N Q 

 represents the temperature at the depth N, the lines P A, B Q 

 joining at the point where the change of temperature ceases at 



