THE ARGON FAMILY OF GASES 659 



molecule is monatomic. The generalisation affords a powerful 

 means of checking the chemist's value of the atomic weights, and 

 it is at least very satisfactory that no discrepancies have so far 

 been encountered. 



1 should be inclined to give the specific heat evidence rather 

 less weight than Sir Oliver Lodge does, not on account of its 

 relative importance, but, rather, because it involves other and 

 more important issues which cannot yet be held to be satisfac- 

 torily decided. At bottom, the first question at issue is whether, 

 during changes of temperature, heat energy is communicated to 

 the structure of the atom or not. Has the atom a structure, that 

 is, that can absorb thermal energy from without and be put into 

 vibration or caused to " squirm " with thermal energy ? If it has, 

 then one never ought to get the S/3rds ratio for the specific 

 heats. If it has not, this value should be given by monatomic 

 gases, provided, but only provided, that one further and rather 

 unexpected assumption is made. 



The first assumption that the atom, in addition to being the 

 chemical unit of matter, is also the physical unit as regards the 

 degradation or deco-ordination of the energy of motion of matter, 

 may now be generally admitted. It certainly was not admitted 

 for all temperatures, I remember, by Prof. Schuster in a dis- 

 cussion at the Manchester University Physical Society, which 

 I had the honour to take part in some years ago ; and whether 

 it does apply for temperatures at which the atom radiates its 

 characteristic spectrum is a matter which still may be con- 

 sidered open to discussion. But for ordinary ranges of 

 temperature a closely allied assumption is made in the kinetic 

 theory of gases in the form that the molecules are perfectly 

 elastic. The assumption is really a far more important one 

 than the thesis of the monatomicity of certain molecules, and 

 though one might employ the latter in discussing the former, 

 the opposite process is hardly safe. 



On the second point, I have sought light and leading from 

 physical friends in vain. The 5/3rds ratio involves the further 

 assumption that the thermal energy of rotation of a monatomic 

 molecule is zero. The monatomic molecule, in fact, possesses 

 one degree of freedom less than it ought to have if it were 

 merely a molecule and nothing more. For the rotation of 

 spherical granules under Brownian movement has been ob- 

 served by Perrin, and the equipartition of energy for this 



