86 Mr. Davy’s Lecture on some new analytical Researches 
If the gas is examined at this time, its volume is found to 
be a little increased, by the addition of a small quantity of 
hydrogene. 
No new fumes are produced by a second application of a 
low heat ; but when the temperature is raised nearly to the 
point of sublimation of potassium, the metal rises through the 
crust, becomes first of a copper colour and then of a bluish 
black, and soon after inflames and burns with a most brilliant 
red light. 
After this combustion, either the whole or a part of the 
fluoric acid, according as the quantity of potassium is great or 
small, is found to be destroyed or absorbed. A mass of a 
chocolate colour remains in the bottom of the retort; and a 
sublimate, in some parts chocolate, and in others yellow, is 
found round the sides, and at the top of the retort. 
When the residual gas afforded by this operation, is washed 
with water, and exposed to the action of an electrical spark 
mixed with oxygene gas, it detonates and affords a diminu- 
tion, such as might be expected from hydrogene gas. 
The proportional quantity of this elastic fluid, differs a little 
in different operations. When the fluoric acid has not been 
artificially dried, it amounts to one sixth or one seventh of the 
volume of the acid gas used ; but when the fluoric acid has 
been long exposed to calcined sulphate of soda, it seldom 
amounts to one tenth. 
I have endeavoured to collect large quantities of the choco- 
late coloured substance for minute examination ; but some 
difficulties occurred. 
When I used from eighteen to twenty grains of potassium, 
in a retort containing from twenty to thirty cubical inches of 
