Mr. F. D. Brown on Molecular Attraction. 259 



In applying these arguments we must remember that the 

 molecules are in constant motion about their centres of gra- 

 vity. In the case of C0 2 , for example, the direction of the 

 line joining the centres of the carbon atom and one of the 

 oxygen atoms varies continuously; hence the molecule of 

 C0 2 , as regards the action on other molecules, may be regarded 

 as practically in the condition of a centre of carbon surrounded 

 by a shell of oxygen. 



Among inorganic substances there are but few which take 

 the form of volatile liquids, and these are mostly isolated in- 

 stances ill-suited for comparisons of this kind. It is therefore 

 to the homologues and isomers of organic chemistry that we 

 must chiefly look for a confirmation of these views. In the 

 alcohols of the ethyl series, for example, the larger the mole- 

 cules the greater must be the attraction between them, and 

 consequently the higher the boiling-point; this, as is well 

 known, is in accordance with fact. In the case of isomeric 

 alcohols, the influence of the position of the atoms comes con- 

 spicuously to the fore. It is clear that if the atoms of carbon 

 of two different molecules cannot approach each other so nearly 

 in the case of one isomer as in another, the attraction between 

 the molecules will be less, and the boiling-point consequently 

 lower. Now in secondary arid tertiary alcohols the carbon 

 atoms are more sheltered by each other, are, as it were, more 

 removed from the exterior of the molecule than in primary 

 alcohols; at the same time the boiling-points are lower, which 

 is in accordance with our hypothesis. 



If we replace two atoms of hydrogen in an alcohol by 

 one of oxygen, we increase the attraction of the molecules, since 

 we substitute a certain number of attractions (Ji o) and (c o) 

 for the relatively small attractions (Ji li) and (c h) : the increase 

 of boiling-point which we should expect is confirmed by expe- 

 riment. 



If, on the other hand, we replace the hydroxylic hydrogen 

 of an acid by methyl or ethyl, we lower the boiling-point, 

 although we increase the number of attractions exercised 

 between the molecules. But we must remember that by this 

 substitution we combine a considerable number of atoms with 

 that carbon to which is annexed the oxygen of the molecule, 

 thus causing a more even balance of the molecule about that 

 carbon, and throwing it therefore more into the centre of the 

 system. The attraction (ho) to which the total molecular 

 attraction is so largely due will therefore be exerted over 

 greatly increased distances, and the boiling-point will be to a 

 like extent diminished. That this argument is founded on 

 fact is, I think, shown by a host of other instances: thus, 



