,}•• 



1000 

 853 



772 



732 



306 



and a new Secondary Battery. 55 



in this paper, was 1050, that of water being 1000. The follow- 

 ing results were obtained* : — 



Table II. 



Figures of merit. 

 Daniell 



Cu 

 Sod. carb. 



pt "V 



Dil. sulph. ac. J 



Pd 1 



Dil. sulph. ac. J 



Pb 1 



Dil. sulph. ac. J 



6. When the copper plates were clean and polished, the figure 

 of merit was far below that mentioned above ; but the more fre- 

 quently the plates were excited, and the more corroded they 

 became, the stronger was the current produced. Copper plate 

 as it comes from the rollers will, under the influence of powerful 

 electric currents, very quickly acquire a condition of surface fa- 

 vourable to the development of secondary currents; and that 

 condition, once acquired, is apparently always retained. Indeed 

 the highest result obtained with any experimental couple was 

 given by a pair of copper plates whose surfaces appeared to be 

 completely covered with crystals of carbonate of copper, the so- 

 lution in which they had been placed having evaporated. 



7. Other alkaline solutions were then tried, viz. of the hydrates 

 of sodium and potassium and the carbonate of potassium. The 

 results obtained by using copper, platinum, and palladium in 

 solutions of the carbonates and hydrates of sodium and potas- 

 sium are given in Table III. : — 





Table III. 









Sod. carb. Pot. carb. 



Sod. hvd. 



Pot. hyd. 



Copper . . 



. 853 883 



867 



896 



Platinum . 



. 831 827 



772 



756 



Palladium . 



. 740 758 



777 



645 



8. Lead, silver, zinc, brass, and tin plate were tried in these 

 same solutions, but they all (with the exception of brass in car- 

 bonate of potassium, which had a figure of merit 820) gave re- 

 sults far inferior to those recorded in Table III. The diacid 

 carbonates were found to be far less energetic than the neutral 

 carbonates. Increasing the specific gravity of the sod.-carb. 



* In this and all subsequent Tables the highest uniform results given by 

 good experiments are recorded, as all the errors of observation, whether in- 

 strumental or functional, tended to diminish rather than to increase the 

 figures of merit. 



