404 Experiments on the Combinations of 
formed. I thought that by ‘ascertaining’ the quantity of 
nitrogen evolved by the action of a’given weight of potas 
sium, and comparing this with the quantity of oxygen dis- 
engaged from the oxide. by water, L might be able to de- 
termine its composition accurately. A grain of potassium 
acting in this way, I found produced only +46 of nitrogen ; 
and the red oxide, by its action upon water, produced less 
than half a cubical ineh of oxygen, so that it is probable 
that potash as well as its peroxide is formed in the opera- 
tion. 
Sodium, when brought in contact with fused nitre, pro- 
duced a violent deflagration.' In two’ experiments in which 
T used a grain of the metal, the tube broke with the violence 
of the explosion. I succeeded in obtaining the solid results 
of the defagration of 2 a grain’ of 'sodium; but it‘appeared 
that no peroxide had formed, for the mass gave no oxygen 
by the action of Wafer.” Son Be: : 
When potassium is burnt in’a retort of pure glass, the 
result is partly potash and partly peroxide, and by a long 
continued red heat the péroxide is entirely decomposed. 
A grain of potassium was gently heated in a'small green 
glass retort containing oxygen; it burnt slowly, and with a 
feeble flame ; a quantily Of oxygen was absorbed equal to 
veo Of a cubical’ inch ; by heating the retort to dull red- 
ness, Oxygen was expelled equal to 88, of a cubical inch ; 
the mercury in the thermometer’ in this experiment stood 
at 63° Fahrenheit, and that in the barometer at 30:1 inches. 
In experiments on the electri¢al decomposition of potash 
and soda, when the Voltaic battery einployed contains from 
500 to 1000 Series in’'full action, 'the'metals' burn ‘at the 
moment of their production, and form the peroxides; and 
it is probable, trom the observations of M/ Ritter, that these 
bodies may be proditcéd likewise it’ Voltaic operations on 
potash, at the positive surface. °"' ee ck 
In my early experiments, on potassium and sodium, T re- 
garded the fusible substances appearing at the negative sur- 
face, in the Voltaic circuit, as well as those produced by the 
exposure of the metals to heat and airs as prot-oxides, and 
as similar to the results obtained by heating the metals m 
contact with smail quantities of alkali, : 
T have repeated these, last operations, in which I con- 
ceived that prot-oxides were formed. f 
Potassium and’ sodinm, when heated ‘in’ glass tubes in 
contact wath about half of their weight of potash and soda 
that have been ignited, become first of a bright azure, then 
produce a considerable quantity of hydrogen, and ae 
Le ae orm 
