18 Trans. Acad. Sci. of St. Louis. 



able shrinkage can go on after the average density of the 

 globe has increased to eight times its present value. For the 

 resistances due to molecular repulsive forces must tend to 

 overcome gravitational pressure, and at length render further 

 contraction impossible. If this state be not completely real- 

 ized when the sun's radius has shrunk to one-half its present 

 value, it must yet be so fully attained in the greater part of 

 the body of the sun that what further shrinkage is possible 

 in the external layers will produce little available energy for 

 maintaining the sun's heat. As to the specific heat of the 

 sun we can only say that, with one unimportant exception, 

 water has the greatest specific heat of all known terrestrial 

 substances ; and it is not probable that the average specific 

 heats of the dense gases comprising the body of the sun can 

 be enormously greater than those of the corresponding gases 

 found upon our earth. Wherefore it is not easy to imagine 

 how our sun can long maintain its radiation after shrinkage 

 has entirely ceased. 



Hitherto we have assumed that k = 1.4, as in common air 

 and most terrestrial gases. Though this value is based upon 

 the study of man}' gases under widely-varying conditions, 

 there are theoretical reasons for supposing a larger value 

 to correspond more closely with the state of things 

 existing in the body of the sun, where it is not improb- 

 able that many of the gases, disassociated by great 

 heat, behave as if monatomic. In such monatomic gases 

 where the energy is applied in the form of transla- 

 tional kinetic energy, and none goes to work done upon 

 the internal structure of the molecules themselves it is known 

 from Clausius' theory of the gases that h attains a maximum 

 value 1.66. In one well-known case this has been experi- 

 mentally confirmed by Professor Kundt, who found for the 

 vapor of Mercury, which on chemical ground is known to be 

 monatomic, the experimental value 1.66. In contrast to this 

 large value of le. some substances of complex molecular 

 structure give experimentally a very small value. Thus in 

 oil-of-turpentine vapor & is only 1.03. But it is not probable 

 that substances of such elaborate structure exist in the body 

 of the sun, where the intense heat necessarily renders com- 



