2 T. C. CHAM BERLIN 



discovered that certain of the heaviest known atoms were spon- 

 taneously giving out heat as an incident of their own disintegration. 

 It was found that if radioactive substances pervade the whole body 

 of the earth as richly as they do the accessible portions, the heat 

 generated by them would be much greater than the amount the 

 earth is now discharging at its surface. The certainty that the 

 earth had been cooling at once disappeared; it seemed quite as 

 likely that the temperature of the earth was rising as falling and, 

 so far as heat is concerned, the volume of earth quite as likely 

 sweUing as shrinking. This cut at the very roots of the former 

 tenet that shrinkage was chiefly due to the lowering of the tempera- 

 ture. The possible potentialities of the new source of heat were not 

 only embarrassing in themselves, but they made the great heat 

 h3rpothetically inherited from the white-hot earth a superadded 

 burden of embarrassment instead of the facile explanation of 

 shrinkage and deformation it had once been supposed to be. 



Nor was this all: It was obviously necessary to devise some 

 special hypothesis to obviate the surplus of heat the radioactive 

 substances would give if they had a uniform distribution throughout 

 the interior of the earth. Since no pressure or other physical 

 condition is known to reduce appreciably the thermal output of 

 radioactivity, a restriction of the radioactive substances themselves 

 to a shallow surface shell seemed the only hypothesis available. 

 But here the tenet of a molten globe arose as a new form of embar- 

 rassment. The radioactive substances are exceptionally heavy, 

 and in a liquid mass they should naturally concentrate toward the 

 center, not toward the surface. Convection, of course, if it were 

 sufflciently active, might be supposed to prevent much concentra- 

 tion toward the center, but convection carries down as well as up 

 and tends to give a more or less uniform distribution — just the 

 distribution the hypothesis is seeking to avoid. 



Nor is this the limit of the embarrassments attending the old 

 view. A molten state implies that the larger part of the potential 

 resources of shrinkage were exhausted before the formation of a 

 crust made a record of shrinkage possible, at least so far as shrinkage 

 depends on arrangements and combinations of material. A molten 

 state offers nearly ideal conditions for the physical adjustment of 



