RICHARDS, COLLINS, AND HEIMROD. — COPPER AND SILVER. 129 



Since it is precisely this variation of the highest values for copper 

 which interests us, — for we know that the lower values are certainly 

 too low, — it seemed desirable to obtain further light upon this question. 

 We therefore carried out a similar series to Gray's, using weaker solu- 

 tions, with which the point of inflection in the curve would naturally 

 come at a lower current density. Cathodes and anodes of copper of 

 different sizes were placed in separate portions of a solution of cupric 

 sulphate containing 100 grams to the litre, and the same current was 

 run through the series of cells. The amounts of copper thus deposited 

 are given in the following table. 



The first fact which is at once apparent is the interesting phenomenon 

 of a decrease in the deposit on the smallest plate in experiments c. d, and e. 

 This corresponds to a downward turn in Gray's curve ; and if these 

 results are plotted upon his scale, the following diagram is obtained. 

 (See Figure 3, on the next page.) 



Such a curve as this is only to be produced by the appearance with 

 high current densities of another cause, independent of the cuprous-cupric 

 reaction, but also tending to lower the observed weight of copper. It is 

 not difficult to suggest what this cause may be. When the current 

 increases in strength, the copper ions in the neighborhood of the cathode 

 are no longer able to carry all of the current, and a portion of the load 

 is taken by the only other positive ion present. But the hydrogen which 

 thus plays the electrical role of the copper weighs far less than the equiv- 

 alent weight of the red metal. A small amount of hydrogen, capable 

 of being occluded by the copper without change in the appearance of the 

 surface, would cause an immense deficiency in the gain in weight. When 

 the current strength increases only a little above 0.02 ampere per square 

 centimeter the deposit of the copper becomes powdery, probably because 



VOL. XXXV. — 9 



