Iviii PROCEEDINGS OF THE GEOLOGICAL SOCIETY. 



speaking of them with reference to the earth's surface, one was in- 

 ternal, another external, and the third superficial. No attempt, 

 however, had been made to examine the efficiency of these different 

 causes to account for all the phaenomena which may be referable to 

 them. It was to remedy this defect that I undertook the investiga- 

 tions contained in the paper of which I am speaking. 



Assuming the primitive temperature of the globe to have been very 

 much greater than at present, there is manifestly no difficulty in ac- 

 counting for any higher superficial temperature than the present, at 

 past epochs, provided those epochs be sufficiently remote. They 

 must, however, be exceedinglj'^ remote to enable us thus to account 

 for a variation of temperature which should sensibly affect the climatal 

 conditions in any part of the earth. The terrestrial temperature, to 

 the depth of about 70 feet, varies with the progress of the seasons, 

 the variation becoming less as the depth is greater, until, at about 

 the depth just mentioned, it is no longer sensible, so that a ther- 

 mometer placed there would indicate a constant temperature during 

 the whole year. A second thermometer at a greater depth would 

 also indicate a constant temperature throughout the year, but higher 

 than that indicated by the preceding one. If this second thermometer 

 were placed at a still greater depth, it would indicate a still higher 

 constant temperature ; and the increase of temperature between the 

 two thermometers would be proportional to the distance between 

 them, i. e. the temperature in descending below the first thermometer 

 would increase at a constant uniform rate. 



Again, if the cooling of the earth were to continue for an indefinite 

 period of time, assuming the temperature of external space, the sun, 

 and the earth's atmosphere to remain as at present, the superficial 

 temperature would approximate indefinitely near to a certain limit. 

 The difference between that limit and the earth's present superficial 

 temperature is the effect due to the remains of the primitive heat. 

 Now theory gives us a simple relation between the amount of this 

 effect and the rate of increase above-mentioned as we descend below 

 the earth's surface*. Consequently, knowing the one, we can im- 

 mediately determine the other, and thus, having ascertained the above 

 rate of increase, we know the amount of superficial temperature which 

 is now due to the earth's primseval heat, assuming always that heat 

 to be the cause of the existing internal temperature of the globe. 

 This amount is thus proved not to exceed about the ^^^th of a centesi- 

 mal degree, so nearly has the earth's superficial temperature approxi- 

 mated to that ultimate limit beyond which it could never descend, 

 supposing external conditions to remain the same. It was calculated 

 by Poisson that, to reduce the superficial temperature by one half of 

 the above amount, or g\yth of a centesimal degree, it would require 

 the enormous period of one hundred thousand millions of years. It 



* l{f denote the excess of the present superficial temperature above the final 

 limit to which the temperature would descend in an indefinite period of time, and 



f 

 g the rate of increase of temperature mentioned in the text, we have — = h, where 



h is nearly equal to unity^ 



