(2) and (4) were performed under similar conditions. The distance 

 between the poles was 2.2 cm. The current was generated by a battery 

 of Bunsen cells. Even with a current of 5 c.c. gas per minute zinc was 

 found in the solution in traces, while considerable quantities of cobalt and 

 nickel separated as a coating upon the cadmium. (1) and (3) were also 

 failures. A separation was not obtained even approximately. 



Summary. 



As a result of the foregoing experiments, it was found that the amount 

 of copper, cadmium or zinc deposited in a given time was proportional to 

 the strength of the solution, and that the presence of free acid in moderate 

 quantity did not materially affect the result. 



Increasing the distance between the poles resulted in diminishing the 

 amount of metal deposited, but the rate of decrease diminished as the 

 distance between the electrodes increased. 



Elevation of temperature caused an increase in the amount of metal 

 deposited, the rate of increase being greatest at 80° in neutral and acid 

 copper solutions, and at 60° in cadmium solutions containing free acid. 

 On the other hand, the amount of zinc deposited in solutions, to which 

 free acid had been added, diminished as the temperature rose, nothing 

 being deposited at 80°. 



Attempts to secure compact and adherent deposits of cadmium and zinc 

 in neutral solutions were failures. 



In acid solution copper and cadmium separated completely and satis- 

 factorily. The zinc deposits were spongy, but the precipitation was com- 

 plete. 



Lead was mainly deposited on the negative pole, both in neutral and 

 acid solutions. Manganese was precipitated on both poles, but the 

 amount of peroxide separating on the kathode was reduced to mere 

 traces by the presence of free acid. 



The following separations were satisfactorily accomplished : copper 

 from zinc, cobalt and nickel and cadmium from zinc and manganese. 



Attempts to deposit copper in the presence of iron and cadmium, and 

 zinc in the presence of iron, cobalt and nickel, were successful. Nor was 

 it possible in the presence of the last three metals named to estimate 

 cadmium. 



In conclusion, I wish to express my obligations to Prof. Edgar F. 

 Smith, at whose suggestion the work was undertaken. To his super- 

 vision and advice is largely due whatever value may attach to these 

 results. 



