EICHARDS, COLLINS, AND HEIMROD. — COPPER AND SILVER. 145 



With this cell were compared the following modifications of the silver 

 voltameter: first, the large crucible voltameter at 20", with the anode 

 wrapped in filter paper, such as was used in the preceding experiments 

 upon copper ; secondly, a voltameter at 20° prepared according to Lord 

 Rayleigh's standard prescription (a platinum bowl, 10 centimeters in 

 diameter, containing a 15 per cent solution of argentic nitrate and a large 

 silver plate enclosed in filter paper) ; thirdly, the same arrangement at 

 0° ; fourthly, an exact imitation of Patterson and Guthe's method, for 

 which the electrolytes used in Lord Rayleigh's method were digested with 

 argentic oxide and filtered ; fifthly, the " porous cup method " at 60° 

 instead of at 20° ; sixthly, the same with a double porous cup enclosing 

 argentic oxide in the annular space ; and finally, the " porous cup method " 

 at 0°. The table opposite explains itself. 



Discussion of Results. 



The comparison of these results is highly instructive. The most 

 noticeable fact is that our "standard porous cup method" gives the low- 

 est result of any of the methods tried, since all the figures in the fifth 

 column are greater than unity. In the next place, we find that Patter- 

 son and Guthe's method gives a result 0.12 per cent higher than Lord 

 Rayleigh's ; a figure which corresponds closely with their own estimate, 

 0.11 per cent, while Lord Rayleigh's method gives results 0.082 per 

 cent higher than ours. Evidently the method used in the first three ex- 

 periments is essentially similar to Lord Rayleigh's, and may be averaged 

 with it. Again, it is apparent that the interposition of argentic oxide in 

 the hot determinations had a varying and unsatisfactory effect, as if it 

 introduced a new source of error. Finally, it appears that change of 

 temperature causes a somewhat smaller variation in the results from the 

 " porous cup method " than in those from Lord Rayleigh's old standard : — 



The obvious inference from these facts is that the porous cup is really 

 effectual in protecting the cathode from a part at least of the disturbing 

 influences under consideration, especially at low temperatures. The 

 apparent gain of 0.016 per cent exhibited by the method at 0° was un- 

 doubtedly due to the difficulty in washing the very finely divided spongy 

 metal which is deposited from a freezing solution ; but Lord Rayleigh's 



VOL. XXXV. — 10 



