684 T - c - CHAM BERLIN 



which rises at a very low rate near the surface and is followed by 

 a slowly increasing rate for about one-third the distance to the 

 center, beyond which it rises at a decreasing rate to the center; 

 or, if traced from the center outward, this computed curve of 

 temperature declines faster and faster at every step for about two- 

 thirds of the distance and then declines less and less rapidly to a 

 vanishing-point near the surface. Hence if conductivity be 

 assumed to be the same at all depths, the outward flow of heat on 

 such a gradient would increase in rate from the center to the two- 

 thirds point and then grow slower toward the surface, from which 

 it follows that, on these assumptions of uniform compressibility 

 and uniform conductivity taken by themselves, the internal heat 

 should have been progressively lowered in the deep interior and 

 raised in the more superficial parts. The conductivity of rocks is 

 so very slow, however, that its effects at the surface under the con- 

 ditions named cannot have been large up to the present unless 

 the earth is much older than even radioactivity seems to imply. 



This first approximation to a theoretical curve of heat, even 

 when modified by conduction, has not been supposed to represent 

 the actual distribution of heat at the present time, for reasons that 

 follow. 



There is ground to think that compressibility falls off as 

 increased degrees of compactness are attained. In working out 

 the curve which was published in Geology, I, 566 (Chamberlin 

 and Salisbury), Dr. Lunn used as a guide the Laplacian law of 

 density which postulates that density varies as the square root 

 of the pressure. This distribution of density harmonizes fairly 

 well with such astronomical tests as are available and gives a mean 

 density for the earth which is near that required by the earth's 

 total weight. The assumption that the increased density of the 

 interior is all due to compression, however, makes no allowance 

 for the probable transfer of lighter matter to or toward the sur- 

 face by extrusive action which would tend to increase the mean 

 specific gravity of the residue. The curve of Dr. Lunn may be 

 regarded as a second approximation. 1 But this, as noted, does 



1 Year Book No. 3, Carnegie Institution of Washington, 1904, p. 156; also "Geo- 

 physical Theory under the Planetesimal Hypothesis," Section II of "Tidal and Other 



