154 DE. B. D. STEELE, DE. D. McINTOSH AND DE, E. H. AECHIBALD 



We are led to this conclusion by a consideration of the following facts : 



(1) Large quantities of heat are evolved when conducting solutes are added to 

 either of these solvents. This heat evolution we take to indicate chemical union. 



(2) Compounds containing a varying number of molecules of solvent have been 

 isolated (ARCHIBALD and MC!NTOSH, 'Jour. Chem. Soc.,' 1904, vol. 85, p. 919). 



(3) The ionisation of a compound such as ((CH 3 ) 2 CO), l HBr is much easier to 

 understand than that of a simple polymer such as ((CH 3 ) 2 CO) n . 



In order to apply the foregoing conclusions to a specific case, we will consider a 

 solution of acetone in hydrogen bromide. 



According to our hypothesis, such a solution contains a compound of acetone and 

 hydrogen bromide, the formula of which we will assume to be Ac 3 (HBr) Bl . 



This compound dissociates simultaneously in two different ways, a certain number 

 of molecules being dissociated into acetone and hydrogen bromide, other molecules 

 being dissociated into ions, and the ratio of the number of molecules undergoing each 

 dissociation will be constant. 



Applying the equation ^ = K'a s , we see that the molecular conductivity will 

 increase with increasing concentration of the acetone, the increase neglecting 

 variation of a being proportional to the square of the concentration. Similarly, we 

 see that if a is nearly constant, the specific conductivity will be proportional to the 

 cube of the acetone concentration. If, however, a is not constant, then K/a will be 

 proportional to the cube of the acetone concentration. 



This conception of an intermediate compound which is able to break up in different 

 ways is by no means new to chemists, and the solution of ammonia, which we have 

 already referred to, furnishes an example of such a case, which, in many ways, is 

 analogous to the preceding. 



The compound that is formed in this solution is ammonium hydroxide, and the 

 dissociations are 



(1) NH 4 OH:zNH 3 +H 2 0; 



(2) NH 4 OH - NH 4 + OH. 



The relation between specific conductity and concentration for such a solution 

 has been already developed in equation (1) K = aK'a, which is a special case of 

 equation (2). Here again 



K T7" / 



M---.K', 



so that 



a = ; but a = JL 

 K ^ 



so that K' is simply the molecular conductivity at infinite dilution. 



