
EXPERIMENTAL METHODS QI 
by contact of organic matter with oxygen or any 
other gas—that it might, at times, be dependent 
upon the inherent instability of the organic matter 
itself. Independently of the fact, therefore, that 
the sealing of the flask after all the air had been 
expelled, and during ebullition of the fluid, was a 
much simpler process than having to admit calcined 
air and sealing the flask after it had cooled, it 
seemed possible that the existence of a compara- 
tive vacuum might, for other reasons, sometimes 
prove an advantage. 
The method, therefore, adopted by me for a long 
time was this. After each flask, of one to two 
ounces capacity, had been thoroughly cleaned with 
hot water, three-fourths of it was filled with the fluid 
that was to be made the subject of experiment. 
The neck of the flask or retort was then drawn 
out in the blow-pipe flame, two or three inches 
from the bulb, till it was about a line in diameter. 
The neck having been cut across in this situa- 
tion, the fluid within the vessel was boiled con- 
tinuously for from ten to twenty minutes. At first, 
ebullition was allowed to take place rapidly (till 
some of the fluid itself frothed over) so as to 
procure the more thorough expulsion of the air 
and bring all parts of the flask in contact with 
the boiling water; then the boiling was main- 
tained for a time at medium violence over the flame 
of a spirit-lamp, and the neck of the flask was 
subsequently sealed in the blow-pipe flame during 
ebullition. 
After a little practice one soon became able to 
