RICHARDS, COLLINS, AND HEIMROD. COPPER AND SILVER. 149 



amount corresponding to the ampere in computing Griffiths's E. Such 

 an inaccuracy may have come about iu two ways : either the electro- 

 magnetic methods for determining the energy of an ampere may have 

 caused the original investigators of the ampere equivalent to overestimate 

 the strength of their current, or else they may have used a form of volta- 

 meter which gave a smaller deposit of silver than the form used by 

 Glazebrook in standardizing Griffiths's cells. This last supposition is the 

 only one with which the voltameter is concerned, for it is clear that 

 a constant error in the amount of silver, occurring both in the original 

 comparison of the ampere value, as well as in the standardizing of 

 Griffiths's cells, could have no effect, — the weight of silver would be 

 eliminated from the result. The only safe mode of comparing results of 

 this kind is to reduce all figures to the standard of some one form of 

 silver voltameter, capable of giving constant values, as has been done 

 above. ■Patterson and Guthe, in their otherwise valuable paper, do not 

 seem to realize this fact, and only add to the confusion by appearing to 

 suppose that the results with their plethoric voltameter are directly appli- 

 cable to Griffiths's equivalent. (See line 4, p. 281 of their paper in the 

 Physical Review, Vol. 7.) 



This paper is by no means a final statement of the matters which it 

 concerns. It is in the nature of a preliminary contribution, and we in- 

 tend now to test more rigidly the accuracy of the "porous cup volta- 

 meter," as well as to study from a physico-chemical standpoint the cause 

 of the anomalies which are rendei'ed harmless by its use. Some light 

 has already been obtained, and more seems not far off. 



Summary. 



In conclusion, a brief summary of the chief points treated in the 

 paper may serve as an index to the more intricate details. 



1. Metallic copper dissolves in cupric sulphate with the formation of 

 cuprous sulphate, and unless acid is present cuprous oxide or hydroxide 

 is formed by hydrolysis ; in short, most of Foerster and Seidel's results 

 are confirmed. 



2. On the other hand, the saturation of the solution with cuprous 

 salts is shown to cause too high results in the copper voltameter. 



3. Vanni's method of correcting the copper voltameter by regulating 

 the amount of acid present is shown to be unsound, although its results 

 are not very erroneous. 



4. Before the cupric solution is made so dilute that its solvent action 



