[miller] idea in chemical MECHANICS 247 



for the temperature, to indicate, not that it was very hot, but that that 

 factor too was kept constant (by the use of a thermostat). 



The concentration of the permanganate, of course, changed during 

 the experiment, we saw it fade away ; the rate too changed ; but as the 

 change in the rate was caused only by change in the concentration of 

 one of the reagents, it was a relatively easy matter for Harcourt and 

 Esson to discover the relation between the rate and the concentration of 

 that one substance; the effect of the concentrations of the others was 

 determined like that of the temperature, by comparing different experi- 

 ments in which these concentrations were different, while remaining 

 constant throughout each. 



This method of working is the " idea " of which I spoke. It may, 

 perhaps, seem a very obvious idea, and hardly worth while making a fuss 

 about; and, indeed, Harcourt and Esson seem to have considered it so. 

 But it is true, although to those who are not familiar with this branch, 

 it may seem difficult to believe, that even now — forty years after the 

 publication of Harcourfs experiments — this method is not in general 

 MBB, and that for thirty-six years it was never used at all and practically 

 all experiments on the rates of chemical reactions were carried out in 

 solutions in which the concentrations of all the chemicals varied greatly 

 during the experiments. 



In reply to the question " How is it possible to find the connection 

 between rate and concentrations from experiments so badly planned as 

 these," the answer is " By the method of guess and try." Make some 

 plausible assrmiption, express it mathematically, and compare the ex- 

 periments with the methematical deductions from the assumption, or 

 guess. If they agree, well and good. If they don't, guess again. The 

 trouble is, that in case of a bad guess the experiments themselves don't 

 give much help in making a better. 



This method, then, I shall call "the Method of Guess and Try," 

 in contradistinction to Harcourt and Esson's method of " Systematic 

 Exploration." 



Of course, there are connecting links. Two of the concentrations, 

 for instance, may be small, and the others large ; the effect of the former 

 being arrived at by guess and try, and that of the latter by systematic 

 exploration. Hood i used such a half-and-half method in 1878. 



With regard to the method of guess and try; it is hard to teach 

 people how to guess, — that is '' chemical instinct ;" but van't Hoff, 

 in his celebrated " Études "^ or " Studies on Chemical Dynamics," 



I Phil. Mag. (5)'6 371 (1878). 



* Etudes de dynamique chimique, 1884, p. 87. 



