196 Dr. E. F. Armstrong and Mr. K. J. Caldwell. [Aug. 26, 



The strength of the acid selected was N/500 hydrogen chloride; the 

 two sugar solutions used contained 171 and 342 grammes sucrose per 

 litre. The polarimeter tube had been in use for a considerable time 

 with acid solutions. The solutions of sugar and acid were accurately 

 measured out from burettes into a small flask into which steam had 

 been passed, according to Ostwald's directions, to free it from alkali. 

 The mixture was then rapidly filtered into the polarimeter tube. The 

 first reading was taken 10 or 15 minutes after mixing, when the 

 temperature had attained to that of the thermostat. 



As under these conditions the rotatory power of the solution fell 

 quickly, the polarimeter readings had to be taken very rapidly, so 

 that the possible error in the tabulated values (although amounting to 

 only O'l per cent.) is about 0'03, making a possible error of 3 units 

 in the third column of differences. 



By taking a very large number of points, such minor errors are 

 eliminated and when the velocity constant K is calculated according 

 to the usual logarithmic law a very steady series of values is 

 obtained. 



From Tables I IV it will at once be obvious that the value of the 

 velocity coefficient K steadily increases during about the first 4 or 

 5 hours of the change and then remains constant, the figures obtained 

 being in striking contrast with those of Ostwald and also with those 

 given later in this paper, Tables VI, VII, for experiments in which 

 stronger solutions of acid were used : in all these experiments the 

 successive values of K vary up and down on either side of a mean value. 



The fact that the value of K rises is an indication that change 

 proceeds faster than the mass-action law requires. On the other hand, 

 the figures do not offer definite evidence that the change proceeds at 

 a strictly linear rate, although in all cases, especially that recorded in 

 Table II, the approximation to a straight line law during the first 

 2| hours is very close : indeed, if the possible error in the difference be 

 taken into account, the first 10 or 15 per cent, of the change is 

 practically linear. When the values are plotted on rectangular 

 co-ordinates, the curve obtained falls between the straight line and the 

 mass-action curve. 



Furthermore, Tables III and IV indicate that the approximately 

 linear portion of the change persists the longer the larger the propor- 

 tion of sugar to acid. 



It is obvious from these results that the analogy between acid and 

 enzyme action is complete. In both cases, when the proportion of 

 hydrolyst is relatively small the change is at first approximately a 

 linear function of the time and subsequently a logarithmic function ; 

 whilst when a larger proportion of hydrolyst is present, the change is 

 from the first a logarithmic function which may become modified by 

 secondary causes. The association theory of hydrolysis put forward in 



