differing much from 15°C), we can calculate the mole- cular volume -= of a substance by simply adding together
the atomic volumes and the co- volume. It is, of course, necessary to know the molecular weight of the substance and its constitution and make the necessary deductions, if any, for double linkage and so on. Generally the number thus obtained does not differ much from that found by experiment, provided the following conditions be satisfied : (1.) The molecular weight taken must be exact. Sup- pose for a moment that this has erroneously been taken
2m twice too large, then the molecular volume would be — =-
a (m being the correct molecular weight).
To find the same value by the theoretical method,* we should have to take the sum :
2 2 atomic volumes + twice 25*9.
The error in the molecular weight would lead us to the sum :
2 2 atomic volumes + once 25*9,
i.e. the theoretical number for the molecular volume would be too small by one co-volume.
These considerations show, on the one hand, that the co- volume has the value assigned by Traube and, on the other hand, that the experimental determination of the
molecular volume -j of a substance gives us a means of
checking the exactitude of the molecular weight provisionally accepted. The value of m must satisfy the equation,
= S atomic volumes + 25*9. d
(2.) The substance must not be associated, otherwise the normal co- volume would only have to be counted once