PIvANS FOR OBTAINING SUBTERRANEAN TEMPERATURES. 26 1 



VALUE AND FEASIBILITY OF A DETERMINATION OF SUBTERRANEAN 

 TEMPERATURE GRADIENT BY MEANS OF A DEEP BORING. 



By G. k. Gilbert. 



SCIENTIFIC NEED OF KNOWLEDGE OF THE NORMAL GRADIENT. 



Theories of the origin of the earth are intimately related to theories 

 of the constitution and condition of its interior. In the field of 

 geophysics there is probably no problem which does not involve the 

 distribution of internal heat. Direct obsen-ation of the nucleus being 

 impossible, inference is depended on, and inferences, so far as they 

 are quantitative, have been and perhaps can be based only on observa- 

 tion of temperature gradient near the surface. For the purpose of 

 testing theories as to the origin of internal heat, it is important to 

 know not only the temperature gradient in the accessible portion of 

 the crust, but also the variation of gradient with depth. If the 

 relations of crust to nucleus have existed so long that the distribu- 

 tion of heat has become systematic, and the heat discharged at the 

 surface is derived from all parts of the sphere, then the gradient in 

 the accessible zone near the surface should be sensibly uniform. If 

 the heat flowing toward the surface is and has been derived from 

 tidal work performed in a subcrustal zone, then also the observed 

 gradient near the surface should be uniform. But if, as assumed by 

 Kelvin and King, the heat of the earth received its general distribu- 

 tion through convection during an initial molten condition, and 

 surface cooling has been in progress only a few million years, then 

 the gradient in the upper portion of the crust should diminish 

 downward. 



NEED OF A NEW DETERMINATION. 



Temperature gradients observed in mines and in wells and other 

 borings present a wide range, and the mean derived from them 

 would probably be found to have a large probable error. But even 

 if its probable error were small, the mean could not claim high pre- 

 cision, because most of the observations heretofore made have been 

 subject to unfavorable conditions. Deep mines exist because of 

 geologic disturbances involving either volcanism or diastrophism, 

 and in either case calculated to disturb for a long time the normal 

 distribution of heat. They exist also because of lack of uniformity 

 of the rocks, and in varied rocks there are usually variations of 

 gradient dependent on variations of conductivity. So there is 



