An Investigation of a CW of gradual Chemical Change. 417 



we come to consider more particularly its action and the effect of its 

 variation on the rate of decomposition. 



Variations in Potassium Iodide. 



It has been already pointed out that it is essential to the uniformity 

 of the rate of change that there should be potassium iodide present, 

 for in our preliminary experiments we showed that the reaction was 

 stopped if the oxdising material was allowed to accumulate in the 

 liquid instead of being removed by its reaction, with potassium iodide. 

 At the same time there is no evidence to prove that potassium iodide 

 takes part in the primary reaction. If we examine a few cases in 

 which the amount of iodide was varied we shall soon see what sort 

 of influence it exercises. The quantity used in all our experiments 

 was as a rule 0'76 millionth-gram-molecule per c.c., a very small 

 quantity in proportion to the other ingredients. 



In the following table, in the initial experiment the amount used 

 was very much smaller, 0*00946 m.g.m., and similar quantities were 

 added one by one and their effect upon the rate ascertained. 



Table XI. 



The rate in the same mixture in the ordinary experiments when the 

 usual quantity (0'7G m.g.m.) of potassium iodide was introduced was 

 0-00554. 



It thus appears that when the quantity of the substance present 

 initially is very small, doubling the amount produces a marked in- 

 crease of the rate, but after a certain amount has been added, farther 

 small quantities produce no marked result. Such a series then does not 

 correspond to the ordinary form of variation with quantity of neutral 

 salt, but one would be led to infer that if we call 0*0064 the normal 

 rate we shall only get this rate when the amount of iodide present is 



2 E 3 



