H, Rose on a new mode of preparing Aluminium. 165 
sodium and then applied heat. I however obtained no satisfac- 
tory results. Rammelsberg also followed closely the method o 
Deville but obtained no good results; moreover he was seldom 
able to prevent the bursting of the glass-tubes with which he ex- 
perimented in consequence of the action of the vapors of sodium 
upon the chlorid of aluminium. It appeared to me that consid- 
erable time, labor, expense and long experience would be neces- 
sary in order to obtain even small quantities of this remarkable 
metal. 
The application of chlorid of aluminium and its compounds 
with the alkaline chlorids, is particularly objectionable for the 
reason that they are volatile, attract moisture very readily, and 
therefore the access of air must be prevented when treating them 
with sodium. 
I therefore early thought of substituting for the chlorid of 
aluminium the flnorid of aluminium, or rather the compounds of 
the latter with alkaline fluorids and which are known to us from 
the investigations of Berzelius. Berzelius pointed out the strong 
affinity of fluorid of aluminium for fluorid of potassium and 
fluorid of sodium, and that the eryolite occurring in nature is 
a ptire combination of the fluorid of aluminium with the fluorid 
of sodium. 
A short time since, I returned to them, on having obtainec 
through Mr. Krantz in Bonn, a considerable quantity of the 
