the Decomposition of the Earths, &c. 337 
without much success. To produce the result at all, required 
a charge by the action of nitric acid, which the state of the 
batteries did not permit me often to employ ;* and the metal 
was generated only in very minute films, which could not be 
detached by fusion, and which were instantly destroyed by 
exposure to air. 
I had found in my researches upon potassium, that when a 
mixture of potash and the oxides of mercury, tin, or lead, was 
electrified in the Voltaic circuit, the decomposition was very 
rapid, and an amalgam, or an alloy of potassium was obtained ; 
the attraction between the common metals and the potassium 
apparently accelerating the separation of the oxygene. 
The idea that a similar kind of action might assist the de- 
composition of the alkaline earths, induced me to electrify 
mixtures of these bodies and the oxide of tin, of iron, of lead, 
of silver, and of mercury ; and these operations were far more 
satisfactory than any of the others. 
A mixture of two-thirds of barytes and one-third of oxide 
* The power of this combination, though it consisted of one hundred plates of 
copper and zinc of six inches, and one hundred and fifty of four inches, at this time 
was not more than equal to that of a newly constructed apparatus of one hundred and 
fifty, of four inches. It had been made for the demonstrations in the Theatre of the 
Royal Institution in 1803 ; and since that time had been constantly employed 
in the annual courses of Lectures, and had served in different parts, for the numerous 
experiments on the decomposition of bodies by electricity, detailed in the Bakerian 
Lectures for 1806 and 1807, and a number of the plates were destroyed by corrosion. 
I mention these circumstances, because many chemists have been deterred from pur- 
suing experiments on the decomposition of the alkalies and the earths, under the idea 
that a very powerful combination was required for the effect. This, however, is far 
from being the case ; all the experiments detailed in the text may be repeated by 
means of a Voltaic battery, containing from one hundred to one hundred and fifty 
plates of four or six inches. 
MDCCCVIII. X X 
