MOLECULAR ENERGY 295 



of N involved Brownian rotations 1 of spherules of 

 mastic. The equation which he used to interpret 

 his observations was correct only if there was on the 

 average an equipartition of energy between rotation 

 and translation for these spheres. Of course, the latter 

 had to be large enough so that he could observe any 

 markings on them and thus detect and measure the 

 angular displacements of such markings. He used 

 spheres of diameter 13/* (13X10~ 3 mm.), which were 

 therefore about 100,000 times as large in diameter as 

 a molecule. For particles at least as small as this, 

 therefore, equipartition holds. 



From the values of the specific heat of monatomic 

 gases it appears that the degrees of freedom of rota- 

 tion, which we should expect the molecules to have, 

 are effectively ankylosed 2 at all temperatures to which 

 they have yet been subjected. It further appears 

 from the values for other gases that as the temperature 

 increases there is an increasing number of molecules 

 for which there is no longer ankylosis. There is also 

 reason to believe that collisions of the molecules 

 result in the acquisition of energy in these degrees of 

 freedom only if a definite quantity of energy is made 

 available by the impact. If, at the temperature rep- 

 resented by the average molecule, the velocity of 

 translation of the fastest molecules is too small for 

 a collision to result in this quantum of energy, then 

 none of the k.e. of translation is absorbed by other 

 degrees of freedom. 



1 Cf. Table II, p. 282. 



2 The term "ankylosis" entered the language of statistical 

 mechanics from medicine, where it represents a stiffening or fixation 

 of a joint. 



