308 Van Name and Hill — Solution of Metals 



will serve to illustrate this fact: After an experiment on the 

 rate of solution of tin, the solution remaining in the apparatus, 

 and containing a large excess of ferric sulphate, was always 

 found to give a brown precipitate with hydrogen sulphide, 

 even after standing over night. Again, a mixture of stannous 

 sulphate with a small amount of ferric sulphate, strongly acidi- 

 fied with sulphuric acid and colored red by thiocyanate, re- 

 tained its color for many hours, but bleached rapidly as soon 

 as a little potassium chloride was added, thus showing con- 

 spicuously the much greater velocity in the presence of chlo- 

 ride ion. 



It is certain, therefore, that comparatively little stannic sul- 

 phate can have been formed during the short time occupied by 

 the experiment (about 70 minutes). Nevertheless it is desira- 

 ble to consider what effect the oxidation of the stannous sul- 

 phate by the ferric sulphate would tend to have upon the 

 observed reaction velocity. The important point here is the 

 fact that this reaction produces no change in the titer of 

 the solution toward permanganate. Consequently, the analyses 

 yield us no information concerning the extent to wdiich ferric 

 sulphate has been replaced by stannic sulphate, but give only 

 the sum of the two. Both react witli metallic tin, and with 

 the same ultimate result so far as the yield of products which 

 reduce permanganate is concerned, but the specific rates at 

 wdiich ferric sulphate and stannic sulphate, respectively, react 

 with the metal, may be, and probably are, quite different. We 

 may therefore conclude that the only direct effect on the 

 observed reaction velocity to be expected in a case of this kind, 

 due to the second stage of the reaction, is a downward or an 

 upward trend of the velocity constants, according as the original 

 oxidizer, or the one replacing it, reacts most rapidly with the 

 metal.* 



Experiments 1, 2, and 3, of Table IV show no such trend in 

 the constants, and the averaged values of k may accordingly 

 be assumed to be practically free from error due to the second 

 stage of the oxidation. In Experiments 4 and 5 the constants 

 show a slight rise, so that here the initial reaction velocities, 

 obtained by linear extrapolation in the way described above, 

 furnish the safest basis for comparison with the other metals, 

 and will be so used. Whether the observed rise in the con- 

 stants is due to the second stage of the reaction or to some 

 other cause, no serious doubt attaches to the initial velocities, 

 which are entirely consistent with the values given by the 

 other metals. 



*It is evident that this will in many cases be determined more by rapidity 

 of diffusion than by oxidizing activity. 



