408 PROCEEDINGS OF THE AMERICAN ACADEMY § 1 3 



K is the "principal ratio" defined above, which may be 

 calculated approximately for a homogeneous substance by 

 the formula, 



K=^ I — 





e§ being the ^^qiiantivale?ice'^ of the atom whose molec- 

 ular weight is m. 



Finally the thickness of the molecular film, /, which is 

 closely connected with the absolute weight of the mole- 

 cule, may be calculated with more or less accuracy by 

 reasonin<2: from its value in a known case to that in an un- 

 known. For this, of course, an approximate knowledge 

 of the atomic grouping is necessary. 



It is evident that two more data must be had in order 

 to obtain a complete solution of the problem, and these 

 are evidently the free path and the strength of the attrac- 

 tion of each atom. That is, we must have a description 

 in full of the nature and distribution of the material with 

 which we have to deal, in order to be able to predict its 

 specific properties under any imposed conditions. 



The determination of the density is evidently equiva- 

 lent to that of the free path, for, under given conditions 

 of pressure, volume and temperature, the free path is 

 determined, by the kinetic theory, when the density and 

 molecular weight are given; but we are yet unable to 

 determine the attractive power of the various elementary 

 atoms, owing to the insufficiency of the data. It is evi- 

 dent, therefore, a priori, that the absolute calculation ot 

 the physical constants will be impossible, until by sure but 

 laborious methods these atomic constants shall have been 

 worked out. 



We have, therefore, tio depend upon six equations to de- 

 termine the ratios between seven unknown quantities, 

 which is clearly possible. These seven quantities are 



