in Ferric Salts and in Chromic Acid. 313 



It is evident that this substitution of cupric for ferric 

 chloride would tend to cause a progressive change in the reac- 

 tion velocity during the experiment. A careful consideration 

 of the probable effect of the various factors influencing the 

 result seems to show that such changes as those observed in 

 Experiments 9 to 12, in which the direction of the change 

 depends upon the acidity, are not incompatible with the point 

 of view which we have adopted. We shall, however, make use 

 of only the extrapolated initial velocities, so that the cause 

 of the changes in velocity is for our purpose of minor 

 importance. 



Discussion of Results Obtained with Ferric Salts. 



Table VIII contains a summary of the results given in the 

 previous tables, each velocity constant being the averaged 

 result of the various experiments performed under the con- 

 ditions specified. These values have been obtained by one or 

 the other of the two following methods : (a) For groups of 

 comparable experiments in which the results were normal they 

 are the averages, for all members of the group, of the average 

 reaction velocity in each single experiment, (b) For groups 

 of experiments in which the constants showed an unmistakable 

 trend upward or downward they are the averages of the initial 

 velocities as found by linear extrapolation. Although where 

 method (b) is used the constants so calculated must be given the 

 preference as more truly representative, the values calculated 

 by method (a) are placed beside them, enclosed in parentheses, 

 for comparison. 



Considering first the upper section of the table, containing 

 the ferric sulphate constants, we observe that in the presence 

 of 5 molar sulphuric acid five metals give practically the same 

 velocity, which, from the point of view of the diffusion theory, 

 would be interpreted to mean that here the rate of diffusion of 

 the ferric sulphate is determining the velocity. With decreas- 

 ing acidity, however, the agreement becomes poorer, the metals 

 tending to draw apart and to show individual velocities whose 

 order is that of the electromotive series, the more positive 

 metal dissolving more rapidly. It is evident here that the 

 specific nature of the metal has a distinct influence on the reac- 

 tion velocity. 



Such results as these are seemingly in direct contradiction to 

 JSlernst's hypothesis of infinitely high reaction velocity at the 

 boundary surface between two phases. Nernst's hypothesis, 

 as we have already seen, demands that in all normal cases, that 

 is, in all cases not affected by secondary disturbing influences, 

 the observed reaction velocity shall be determined by the rate 



