THE AGE OF THE EARTH. 



347 



broken line r v of the chart intersects the gradient and marks tlie locus 

 of stationary viscosity. As above this point teniperatnre relatively to 

 pressure lias augmented more rapidly than the ratio recjuired for con- 

 stant viscosity, it follows that viscosity has been diminished by temper- 

 ature more than it has been raised by pressure. Below the stationary 

 point, on the other hand, an excess of jiressure above the reqnired ratio 

 is available for increase of viscosity. 



For the gradient of 3,900'^ C. excess the transitional depth is indi- 

 cated by the intersection of the broken line T^ V. In both cases the 

 transitional points occupy positions in their respective gradients not 

 far below their full initial temperatures, and pressure having been most 

 stationary the transitional points have moved but little during the 

 whole period of secular cooling, and the earth shells passing through 

 them have divided radius into a lower solid of higher viscosity and a 

 surface couche, partly liquid, partly solid of lower viscosity. So far 

 therefore as viscosity indicates the behavior of conductivity, that also 

 should have been systematically diminished (relatively to the surface 

 value obtained at normal pressure and temperatures and used iu the 

 construction of the gradients) 

 from the surface downward for 

 a small fraction of radius, till 

 at the appropriate depth for 

 each excess and age of cooling 

 it reaches a transitional value 

 and thence increases. 



How this correction, of at 

 present unknown value, aft'ects 

 the coordinates of a given gra- 

 dient qualitatively, is shown 

 by the following figure, in 

 which are given the diabase 

 melting point and pressure R ^ 



line, T) D, gradient b of 3,900° C. excess and 100x10" years' cooling, 

 with the viscosity transitional line V V intersecting it, also a dotted 

 line c, indicating the position of the b gradient corrected for diminished 

 conductivity (viscosity). 



Lagging to the right of the uncorrected gradient, obviously the dotted 

 line would require longer refrigeration to reach the state of solidity, 

 and it is equally important to note that its ])osition requires its emer- 

 gence at the surface with a higher rate than the uncorrected line, and 

 thus extends the time of cooling down to the mean rate which marks 

 for all gradients the present limit of the process. 



c. Liquid-solid conductivity. Closely involved in the above heat- 

 pressure-viscosity correction is the change of conductivity on passing 



