THE AGF. OF THE EARTH. 343 



face measures the depth of the oiigiiial conche of fusion. FoUowing 

 the g'radient toward the surface, it is seen (after desci'ibiug its great 

 convexity in the tinid region) to intersect the diabase line a second time 

 and enter a congealed shell or crust formed by cooling a surface portion 

 of the initial fused couclie, and leaving between the nucleus and crust 

 a residual present shell of fusion of 1200 mdes from the top to the bot- 

 tom. The obvious tidal instability of a 20 mile crust resting ui)on 1200 

 miles of truly tinid magma is sufficient basis for the rejection of this 

 particular case of temperature distribution. To fulfill tbe require- 

 ments of rigidity either the time of cooling must be vastly greater to 

 admit of entire congelation, or the initial excess materially less. 



As an illustration of the first of these alternatives, gradient c with 

 the same initial excess as h (S.OOC^ C) has been developed to complete 

 solidity, which on computation proves to have required about GOO x 10'' 

 years, at which time it has but just reached tangeucy wntli the diabase 

 line. Yet we are al)Solutely precluded from accei)ting it as a probable 

 case and assigning 000 x 10'' years as tlie age of the earth, because the 

 temperature values of its emergence at the surface fall below even the 

 75 feet to 1° F. surface rate. Its emergence is at a rate of 0-0081° F. 

 per foot (124 feet per ° F.), which is far less than the (Hallock) rate 

 used in the dotted gradients, itself much less than tlie accepted mean 

 rate of the British Association committee. 



Gradient ^/, 1,950° C. initial excess, and 15x 10'' years secular cooling, 

 falls still some millions of years short of solidity. The initially fused 

 surface conche was about 00 miles in depth, the i)resent crust 33 miles 

 thick, and the present residual tluidity of 33 miles depth from top to 

 bottom. Here again the liquid zone involves tidal instability and 

 requires the rejection of the line. 



Gradient e offers more satisfactory conditions: With an initial excess 

 of 1,750° C'., about the normal melting point of platinum, and an age of 

 20x10*^ years, a condition is reached w^hich throws the convexitj^ of 

 the gradient below the diabase line in complete solidity and fulfills all 

 the conditions. Here then is a possible age for an earth of diabase. 

 Its initial surface couche of fusion would have been about 53 miles, and 

 is now^ wholly cooled into solid crust and united with the original solid 

 nucleus of coin])ression. 



(xradient /■, initial excess 1,230° C. and 10x10'" years secular cooling, 

 would in its first stage have shown only about 5 or miles of surface 

 fusion, w'hich would very shortly have cooled into solidity. 



For those whose interest centers in earths of great age and high 

 temperature, gradient a is given, initial excess 7,800° C and 400x10*' 

 years secular cooling. This has not been projected to the deep, but 

 would not reach solidity until over 1,500 x 10'^ years, a truly uniformi 

 tarian specimen. 



Turning now to the family of three gradients in dotted line, com- 

 puted to conform to the surface rate of 75 feet to 1° F., the first, <;, is 



