1907.] AND CONTRACTION OF THE EARTH. 229 



there is both a uniform temperature and an independent gradient 

 equivalent to the effects of 10 million years of cooling. 

 § 16. Various Methods for Determining the Depth to Which Cooling 

 has Extended. 



Owing to the causes already indicated it seems reasonably certain 

 that the cooling of the earth has not extended to any great depth. 

 Moreover the original temperature gradient increasing with the depth 

 still exists, but the flow of heat has changed the form of the curve. 

 It will doubtless strike many as a somewhat remarkable fact that 

 Professor Milne finds from earthquake observations that a fairly 

 abrupt transition in the constitution of the crust occurs at a depth 

 of about 30 miles (Bakerian Lecture, Proe. Roy. Soc., 1906, p. 369). 

 A depth of 45 miles was fixed upon for this transition by the Hon. 

 R. J. Strutt, from certain considerations arising in his researches on 

 radium (Proc. Roy. Soc., Vol. yy, 1906, p. 483). Mr. R. D. Oldham 

 finds that below this layer of crust the material of the globe seems to 

 be fairly uniform until we reach great depth, about 0.4 of the radius 

 from the center, where the change in the rate of propagation of 

 earthquake waves shows that some discontinuity intervenes. Strutt 

 remarks that the matter of the interior can scarcely consist mainly 

 of iron, as has been commonly supposed, because with a thin crust of 

 rock this would make the earth's mean density too great. 



The ascertained depth of earthquake shocks and the observed 

 rate of propagation of seismic waves both indicate a thin crust, but 

 the phenomena of wave propagation make the thickness of the crust 

 somewhat greater than that derived from the observed depth of 

 earthquake disturbances. This suggests that earthquakes do not 

 originate entirely beneath the crust, but chiefly in its lower layers. 

 From Milne's results we seem justified in concluding that the cooling 

 has not extended much if any below 40 miles, or i/iooth of the 

 radius. The material below that depth acquires its properties mainly 

 from the pressure to which it is subjected. 



We have already considered the extension of Fourier's method 

 which would enable us to approximate the true conditions near the 

 surface ; and it only remains to add that at greater depths there is a 

 transition to the elliptical law, which holds approximately throughout 

 the nucleus as a whole. There is obviouslv some uncertaintv about 



