INAUGURAL ADDRESS. 13 



become automatically changed or shattered by their own inherent 

 energy. We may also distinguish pseudo-unimolecular actions, 

 which are reallv dissimilar bimolecular actions in which one of the 

 substances concerned is present in such overwhelming quantity as 

 to retain a practically unvarying concentration. Actions of this 

 sort are illustrated by those cases where a small quantity of sub- 

 stance dissolved by a large quantity of water slowly combines with 

 some of it to form new products. Cases in which more than two 

 molecules must simultaneously collide before action can occur are 

 conceivable, but need hardly be considered ; for if all the substances 

 concerned be present in reasonably small concentration the actual 

 chances of such treble or quadruple collisions are negligibly minute. 

 But the same final result is often found to be arrived at in another 

 way. For example, two of the original substances may unite to 

 form an intermediate product, which again reacts with a third 

 original substance to form the final product. The rate of the action 

 here depends not on a single trimolecular, but on two consecutive 

 bimolecular actions, each with its own velocity ; and that of the 

 second is governed partly by the concentration from moment to 

 moment of an intermediate compound, the very existence of which 

 may be unsuspected until the dynamics of the whole action are 

 brought under experimental obser\'ation. Such consecutive actions 

 frequently occur, and both the theory and the practice of their 

 investigation is often full of difficulty. Again, it often happens 

 that the products of a given action tend to undergo a reverse action, 

 so as to reproduce the original substance or substances. In such 

 a case complete action can never occur in either direction, but, 

 whichever end of the system we start from, it always arrives at a 

 condition of mobile equilibrium between the old molecules and 

 the new — at a svstem containing both sorts, mixed in proportions 

 characteristic of the particular case. Here we have really to do 

 with consecutive actions, interdependent and opposed, and both 

 may be unimolecular, or both bimolecular, or one unimolecular and 

 one bimolecular. These reversible changes are of peculiar interest, 

 for they include many of the most important chemical actions 

 known to us, and their study opens up the whole question of 

 chemical equilibrium. 



I have told you already that the molecules of salts and the 

 ions produced from them do not, as a rule, conform to the theoretical 

 law which was expected to govern their equilibrium with one 

 another. This is the law of reversible actions, of which I have just 

 been speaking — a law which well explains the results in other cases, 



