390 SCIENCE PROGRESS. 



weight. These two values agree, however, if the sub- 

 stance when gaseous and when liquid consists of simple 

 molecules. 



It is also of interest to compare the molecular weights 

 obtained from associated solutions with those determined bv 

 independent methods for the liquid molecular weights of the 

 pure dissolved substances. The only property which has to 

 any extent been used to give an idea of the degree of molec- 

 ular aggregation is surface energy. From measurements of 

 this property Ramsay and Shields (1894) conclude that at 20 

 the average liquid molecular weights of methyl alcohol, ethyl 

 alcohol, and acetic acid are respectively (CH 3 OH) 232 = 74, 

 (CH 5 OH) r65 = 76. and (CH 3 COOH) 2 . I3 = 128. On refer- 

 ring to the table on p. 385, it will be seen that although in 

 benzene the solution molecular weights of acetic acid are 

 not far removed from the above value, even in moderately 

 dilute solutions the molecular weights found for the alcohols 

 are much larger than those given for the pure liquids by 

 surface-energy observations. A more extended comparison 

 of the two sets of numbers shows that, in general, the re- 

 sults are not in harmony ; as a rule the solution molecular 

 weights are the larger, and hence if the liquid molecular 

 weights deduced by Ramsay and Shields are correct, in 

 most cases the solvent cannot be regarded as merely 

 breaking up the complex molecules of the pure liquids 

 into simpler molecules, but as promoting further aggrega- 

 tion. In a few cases the opposite effect is exerted. Thus, 

 according to surface-energy observations, acetone, nitro- 

 ethane, and propionitrile are associated liquids, but in ben- 

 zene they give normal molecular weights. Of course the 

 agreement between the results of the two methods could 

 not be otherwise than imperfect in the present state of the 

 questions. Because, on the one hand, the values of the 

 molecular weight given by strong solutions are not the 

 result of molecular aggregation alone, but are affected by 

 the attractive forces already considered ; and, on the other 

 hand, the molecular weights derived from surface-energy 

 observations can hardly be regarded as anything but ap- 

 proximations, since the method of calculation is not free 



