on the Nature of certain Bodies. 57 
the fluid ; but even to the end of the operation, no such ap- 
pearance occurs. 
The crystallized and spongy substance, formed in the first 
part of the process, I am inclined to consider as a combination 
of ammonium and potassium, for it emits a smell of ammonia 
when exposed to air, and is considerably lighter than potas- 
sium, 
I at first thought that a solid compound of hydrogene and 
potassium might be generated in the first part of this opera- 
tion : but experiments on the immediate action of potassium 
and hydrogene did not favour this opinion. Potassium, as I 
ventured to conclude in the Bakerian Lecture for 1807, *is very 
• M. M. Gay Lussac and Thenard seem to be of a different opinion. In the 
Moniteur, to which I have so often referred, it is related, that these distinguished 
chemists, exposing hydrogene to potassium at a high temperature, found that the 
hydrogene was absorbed, and that it formed a compound with the potassium of a 
light gray colour, from which hydrogene was capable of being obtained by the action 
of water or mercury. 
After a number of trials, I have not been able to witness this result. In an expe- 
riment which I made in the presence of Mr. Pepys, and which I have often repeated, 
and twice before a numerous assembly, in retorts of plate glass, four grains of potas- 
sium were heated in fourteen cubical inches of pure hydrogene. At first, white fumes 
arose and precipitated themselves in the neck of the retort. When a considerable 
film of the precipitate had collected, its colour appeared a bright gray, and after the 
first two or three minutes, it ceased to be formed. 
The bottom of the retort was heated to redness, when the potassium began to 
sublime and condense on the sides. 
The process was stopped, and the retort suffered to cool. The absorption was 
not equal to a quarter of a cubical inch. When the retort was broken, the gas in 
passing into the atmosphere, produced an explosion with most vivid light, and white 
fumes. The potassium remaining in the retort, and that which had sublimed, seemed 
unaltered in their properties. 
The grayish substance inflamed by the action of water, but did not seem to be 
combinable with mercury. I am inclined to attribute its formation to the agency of 
MDCCCIX. I 
