TEANSACTIONS OF SECTION B. 399 



where K = observed velocity constant, and n is usually = 1. Now the rate of hydra- 

 tion of acid anhydrides in acid solution is independent of the acidity, but in alkaline 

 solution it increases, roughly, in p)oportion to the hydroxyl ion. If the rate was 

 always proportional to the hydroxyl it would continue to fall as the solution became 

 increasingly acid, since this involves a continuous fall in the concentration of the 

 hydroxyl ; but as a fact the velocity remains constant after the concentration of 

 hydroxyl has fallen to 10~'. This suggests that the reaction is taking place in two 

 ways — (1) by interaction with the catalyst and (2) without its co-operation : and that 

 the equation should be written 



K = a + b X CoH- 



Then when Cqh falls to a certain value, depending on the relative magnitudes of 

 a and b, the second term will become negligible, and the velocity will be constant : 

 whereas when Coh i^ large the first term will vanish and the velocity will be pro- 

 portional to Cqh- If this is true in this instance it may be true generally : only a 

 careful examination of such reactions as admit of accurate measurement can decide. 



It may be doubted whether there are any chemical reactions which are not cata- 

 lysed. As Van 't Hoff has pointed out, the influence of the solvent on the velocity is 

 the resultant of two factors, one of which is purely catalytic. When a reaction is 

 said not to be catalysed, this really means no more than that the catalytic influences 

 have remained constant under the conditions of the actual experiments. 



It follows from Van 't Hoff's theory that in order to determine the real catalytic 

 influence of a solvent, we must refer the amount of change not to the molar concen- 

 tration of the reacting substance, but to its solubility : if K is the rate of change, and 

 S the solubility, the catalytic power of the solvent is proportional to K x S. This 

 magnitude has been determined by Dimroth' for one of his triazol derivatives in 

 various solvents. He finds that while the rate of change and the solubility vary in 

 the ratio of about 100 to 1, the product varies only in the ratio of 3 to 1. He 

 concludes that this value is approximately constant for all solvents. He omits the 

 case of water, but from his data an approximate value for water can be calculated, 

 and is found to be less than 1/10,000 of that for any organic solvent. This seems to 

 dispose of the view that K x S is constant. But, apart from this, it is evident that 

 what the product really represents is the catalytic influence of the solvent. It is 

 certainly remarkable that the value of this should vary so little for so wide a range of 

 organic solvents (including methyl alcohol, chloroform, and nitrobenzene), and even 

 more remarkable that these values should be of quite a different order from that for 

 water. It is very much to be wished that further data could be accumulated on this 

 point. 



In view of Prof. Lewis's discussion of the temperature coefficient, I may draw 

 attention to a remarkable fact pointed out by Von Halban." He shows that not only 

 are the temperature coefKcients of isomeric changes, and of monomolecular changes 

 in general, abnormally high, but also those of numerous reactions which are only 

 apparently monomolecular : reactions, that is, which are really polymolecular, but 

 appear to be of the first order because all the reacting substances but one are present 

 in large excess, such as the inversion of cane sugar. 



The abnormally high temperature coeflficients of monomolecular reactions may be 

 explained thus. A rise of temperature will increase the translational velocity of the 

 molecules, and in general also their internal energy. The second of these effects 

 will tend to increase the rate of change, but in polymolecular reactions the first may 

 be expected to diminish it, since the more rapidly moving molecules will remain 

 within reach of one another's attraction for a shorter time. The ordinary positive 

 temperature coefficients must, therefore, be due to the second effect predominating 

 over the first. If we could find a reacting substance whose internal energy was not 

 increased by a rise of temperature, its rate of reaction should fall as the temperature 

 rose. This is only possible for a monatomic gas ; and it is remarkable that the only 

 two reactions whose rates, properly considered, are found to be less at higher tem- 

 peratures, are reactions of monatomic gases ; the recombination of the atoms of 

 active nitrogen, and that of gaseous ions ; in the latter instance the positive ion is no 

 doubt usually polyatomic, but only one of its atoms really takes part in the change, 

 and the internal energy of this atom cannot be increased by a rise of temperature. 



' Ann., 377, 131. 



^ Z. Phys. Chem., 67, 174, 1909. 



