98 Mr R. Adie’s Account of Electrical Experiments. 
the latter, water is abundantly decomposed, to supply oxygen for 
the composition of a metallic oxide. 
Pure water free from air, or air perfectly dry, are well known not 
to act on the oxidizable metals zine and iron at ordinary tempera- 
tures. It is also well known that water, as soon as it absorbs a 
small quantity of air, immediately begins to oxidize them. In order 
to ascertain if this was the case when these metals were combined 
with platina, silver, and copper, after the usual manner of galvanic 
pairs, I prepared various couple’ and placed them in glass tubes, 
filled with recently boiled pure water, and hermetically sealed. These 
remained for weeks without shewing any oxidation. One tube con- 
tained 6 zinc and copper couples, and there was no trace of action. 
A tube was next filled with recently boiled salt and water, and a 
silver and iron couple; still no change: yet a similar couple placed 
in distilled water within an hour from the time of its distillation, 
but which had not been boiled previous to filling like the other tubes, 
for the first six hours gave distinct evidence of the composition of an 
oxide. A silver and iron couple was then placed in a tube filled 
half with water, half with air; for the first week there was an abun- 
dant deposit of the hydrated peroxide of iron, then the dusky green 
protoxide began to form. After three weeks the tube was opened 
under water, when there appeared to be about one-sixth of the air 
absorbed. Had any water been decomposed by this couple, the evolved 
hydrogen would have produced a pressure in the sealed tube. A 
copper and iron couple was placed in a well-stoppered phial, and 
the fitness of the phial for the experiment tested, by filling it with 
pure boiled water to see that there was no action; it was then open- 
ed under recently boiled water, and half filled with pure nitrogen ; 
this shewed no change. A similar experiment was performed with 
oxygen, which immediately commenced the rapid formation of the 
hydrated peroxide of iron. In all these experiments the oxidation 
was strictly limited by the supply of oxygen in the tubes, and when 
great care was taken to exclude it there was no action, not even 
when the tubes were exposed for several days to a bright sun. 
The development of an electrical current, when measured by the 
deposit from the oxidizable metal, was in the above experiments 
entirely dependent on the presence of the oxygen of the atmosphere, 
but I felt desirous of proving this by the measurement of the elec- 
trical current itself. A small zinc and platina couple was connected 
with the galvanometer, and two phials prepared, one containing re- 
cently collected rain-water, the other, the same water well boiled, 
and the air excluded while cooling. On dipping the couple into the 
unboiled rain-water, the galvanometer indicated three and a half 
degrees ; then performing the same experiment with the boiled water 
the deflection was only from half a degree to a degree. 
The next experiment was to prepare a battery which could be 
