342 THE AGE OF THE EARTH. 



tlio inciting teni]>eratnre for the sanie deptli, and henre in fnsion. Con- 

 versely, any jiomt below the diabase line, being' below the melting tem- 

 perature for that pressure and depth, falls into solidity. Thus the chart 

 IS divided into two areas by the bne, that above it representing Huidity, 

 and that below, solidity. For a diabase earth to have been wholly 

 melted an initial excess of 70,200''' C. would be rcipiired. Obviously 

 any earth having an initial excess of less than the surtace melting point 

 1,170° C. wonhl have been always completely solid. Any initial excess 

 above that ligure and below 70,200^' C. requires, at the moment before 

 refrigeration began, a solid nucleus and a fused zone above it extend- 

 ing to the surface, and the lower the initial excess the larger the solid 

 nucleus of compression and shallower the initial couche of surface 

 fusion. Knowing the law of the rise of the fusion-point it is a simple 

 matter of coinputatioii to determine, for any assumed initial excess, 

 exactly the radial value of the original solid nucleus and of the original 

 supernatant liuid couche. For the region covered by the chart these 

 values may be directly scaled off. 



Fourier's equation enables us to go further and assuming that refrig- 

 eration is the result of conduction alone, to determine the temperature 

 distribution for any given period of refrigeration, and what is of particu- 

 lar geological interest — the rate at whicli the fused couche is encroached 

 upon by an overlying superficial crust of congelation, and the exist- 

 ence, depth, temperature, and pressure differences of any residual fluid 

 couche between the upper and lower solids. The relation therefore of 

 any temperature gradient to the diabase line oft'ers an immediate test 

 of its admissibility as a probable case. Any temperature gradient 

 that in passing across the area of the chart from below to the surface 

 intersects the diabase line must in reaching the low temperatures of 

 the surface intersect it again, and the zone included between the pair 

 of points of intersection being above the line, and hence for that inter- 

 val of radius above the fusion temperatures, must be a melted shell, 

 and as on the criterion of solidity the existence of any considerable 

 fusion is precluded, such a case of temperature distribution may be 

 rejected. 



I will now trace the several temperature distributions upon the chart, 

 and note their relations to time and solidity, beginning with the family 

 delineated in continuous lines (surface rate 50.0 feet to 1*^ F.). Line b, 

 the gradient of Kelvin, 3,900° 0. initial excess, 100 x 10'^ years secular 

 cooling, is seen to enter the chart from the lower regions, maintaining 

 even to the shallow depth of 220 miles from the surface, practically, 

 its original maxiinuin temperature. From the center of the earth up 

 to this point it has remained in the initial solid of compression. At 

 the depth noted it intersects the diabase line and jiasses into fusion. 

 Since almost the full initial temperature is maintained up to its inter- 

 section, it follows that that depth nearly marks the original surface of 

 the solid nucleus and that the distance of 220 miles thence to the sur- 



