348 THE AGE OF THE EARTH. 



isothermally from solid to licjuid. llere again the results of Dr. Hanis* 

 throw important light. 



The relatively higher thermometric conductivity of the solid over the 

 liquid of equal temperature indicates au additio-nal plus correction for 

 time values. 



Both the minus correction due to convection, and the plus corrections 

 based upon conductivity diminished below the Everett tigures, sink in 

 importance as we i^ass from earths of higher to those of lower initial 

 excess, so that until some approximate quantitative values can be given 

 them we have no warrant for extending the earth's age beyond 

 24,000,000 years. 



That the application of the criterion of solidity here made to Kel- 

 vin's method is o])en to the objection of being based on the physical 

 relations of an extremely superficial fraction of radius is obviously 

 true. Ignorance of the deeper interior distribution of specific materials 

 and of their relations to the degree of heat and the range of pressure 

 to which they are subjected forbids the construction of a generalized 

 line of melting temperatures for the whole of radius. 



It might therefore be contended that a reversal of the diabase con- 

 ditions is possible, and the deeper materials may possess the property 

 of ice fusion, their melting temperatures suffering depression in>>stead 

 of elevation. The high densities required in lower earth depths have 

 constantly suggested the concentration there of heavy metals and the 

 examples of meteorites has further influenced the idea of a metallic 

 nucleus chiefly of iron. And as iron at normal pressure unmistakably 

 exhibits ice fusion, any great iron mass at the center might be sup- 

 posed to exist as a licpiid in spite of the enormous pressure there 

 exerted. 



The distribution of materials and of '' state" under this assumption 

 involves a metallic (iron) nucleus, liquid from ice fusion, overlaid by 

 less dense couches which at some unassignable depth pass into sili- 

 cates of the diabase type, solid from compression under the law shown 

 for diabase, and solid to the surface as required by tidal effective rigid- 

 ity. 



Ice fusion however is au exceptional phenomenon, nor have we any 

 but the most limited data as to its range as regards temperature and 

 l)ressuie. Iron is conceded to contract in the act of fusion, but cold 

 iron is more dense than the substance either just above or just below 

 the fusion point. It is not beyond the range of probability that exces- 

 sive pressures might bring about the same density in iron that cooling 

 does, and thus isothermally convert ice fusion into the normal type and 

 produce a solid nucleus. However that maybe, tidal effective rigidity 

 excludes fusion of either type for at least 0-2 of radius. 



*The change of beat conductivity on pas.siug isothermally from solid to liquid. 

 C. Barns, Jm. Jour. Sci., July, 1892. 



