EASTEBFIELD and Kobehtson. — Vapour Densities. 499 



The numbers obtained by the members of a large class of 

 elementary students varied from 17 to 21 for the molecular 

 weight of water. Still better results were obtained for the 

 molecular weight of methyl alcohol. It is instructive to allow 

 the students to perform a series of experiments at different 

 concentrations. In the case of water in phenol the observed 

 molecular weight increases very rapidly with the concentra- 

 tion (molecular association). Scarcely any such effect is 

 noticed with methyl alcohol in phenol. 



Art. XLVIII. — The Vapour Densities of the Fatty Acids. 



By Professor Easterfield and P. W. Kobertson. 



[Read before the Wellington Philosophical Society, 11th February, 1902.] 



It is well known that a large number of substances have 

 vapour densities at their boiling-points which are a little 

 above those calculated from their molecular weights. This 

 may in many cases be explained by the fact that the gaseous 

 laws which are used in the calculations are not rigorously 

 true at the point of liquefaction. In other cases, however, 

 the abnormality is undoubtedly due to the fact that asso- 

 ciation of the molecules takes place at temperatures in the 

 neighbourhood of the boiling-point. 



The first substance to attract the attention of chemists 

 was acetic acid. That the abnormality in this case is reallv 

 due to the formation of molecular complexes is shown, first, 

 by the fact that the normal vapour density is not reached 

 till 110° above the boiling-point. Secondly, the value for the 

 expression MW/T' (where M is the molecular weight, W the 

 latent heat of vaporization) is 15, while for liquids of normal 

 molecular weight a constant value of about 21 is obtained. 

 This low value can only be explained on the assumption that 

 the molecules are associated in the gaseous state. 



Similarly, it was found that normal butyric and isovaleric 

 acids were associated, although to a less extent. In general 

 it may be said that this is true of all the lower fatty acids and 

 their derivatives, which do not decompose on heating. This 

 is quite analogous to their behaviour in solution. In benzene 

 and naphthalene most hydroxyl compounds, and especially 

 acids, associate.* This is also true for the solvents bromo- 

 form, nitrobenzene, and parabromtoluene. Even in phenol, 



* Auwers, Zeit. Phys. Chem., 1893, &c. 



