428 C. L. Speyers — Molecular Weights of Liquids. 



A word now in regard to association. 



Let us consider water. Its density in the gaseous state leads 

 to a formula for its molecule of H 2 0. When metals react 

 with it, we get hydroxides and oxides of the general formula 

 M(OH) a; and M x O y . In all chemical respects water behaves 

 as if it had the formula H 2 0, two replaceable hydrogen atoms 

 and one replaceable oxygen atom, no more, no less. This 

 formula, therefore, has a chemical, a scientific, significance. It 

 states that in chemical reactions two atoms of hydrogen and 

 one of oxygen go together. But in some physical relations, in 

 those relations which involve water as a liquid and in which it 

 stays a liquid, that formula does not account for things. Water 

 is now not so active as corresponds to H 2 0. To bring its 

 activity up to what experience has determined to be the 

 normal, the standard, activity, we must take two, three, or 

 more times the mass corresponding to H 2 0. We are, there- 

 fore, tempted to say water is associated, and to write (H 2 0) a 

 where a runs up to three or four for water and reaches much 

 higher figures with some other liquids. But in so doing we use 

 an unsuitable formula, because the liquid does not behave chem- 

 ically as this expression suggests. So it is very undesirable to 

 speak of associated molecules. More definite, and far more 

 in accord with what we know, to say that the activity of the 

 water is depressed below the normal than to say its molecule 

 has been associated. The weak notion of associated molecule 

 is replaced by the virile notion of activity. We shall call the 

 above quantity denoted by a, the activity factor. The activity 

 factor a denotes the number of normal grammolecules which 

 must he taken to give an activity equal to the normal, or stand- 

 ard, activity, the standard activity being determined by 

 experience. We are not to consider a as constant for any 

 liquid but to vary with the conditions as well as with the 

 liquid, and so the objection made by Young and Fortey* that 

 (1) gave values for n which meant a^>\ in certain mixtures of 

 liquids for which in the pure state a = 1 is not valid. 



The increased activity beyond the normal, which means «<1, 

 called electrolytic dissociation or ionization, is another matter. 

 There are regularities here which enable us to cast the processes 

 into chemical equations. This we cannot do when a>l. When 

 a<^l we find a chemical activity beyond the normal, so when 

 dy>l we may by analogy look for a depressed chemical activity. 

 This idea has not yet been put to the test. 



Consider two volatile liquids, 1 and 2, not miscible in all pro- 

 portions within certain limits of temperature and without 

 chemical action the one upon the other. Let n x , n 2 , denote the 



*1. c, 46. 



