BY J. BROWNLIE HENDERSON, AND P. W. JONES. RS 
of course also at 100 degrees C., thus preventing loss 
by condensation. To prevent back currents into the 
colder end of the wide tube, a section of cork about 
5 mm. thick that just fits snugly into the wide tube is fitted 
over the inner end of the narrow tube, and the tube pushed 
in till it nearly touches the explosive mixture. There is 
thus no “dead air”’ in the large tube and the air sucked 
through it does the maximum quantity of work in sweeping 
through the vapour of mercuric chloride. When the 
apparatus is fitted together, the bath is filled with water, 
and the water boiled. The “silver”? end of the tube is 
connected to a small wash bottle containing water to control 
the rate of flow by observing the air bubbles, and the wash 
bottle connected toa graduated aspirator so that the volume 
of air drawn through can be measured. We found by experi- 
ments that no mercury escaped the silver foil when the 
rate of suction was not greater than eight litres per hour, 
and this rate was adhered to in all our experiments. 
b] 
In testing the method, an explosive was first taken 
which, from its normal heat test (17 minutes) and British 
origin was presumably free from mercury. The tube con- 
taining the silver was weighed before starting the experiment. 
The explosive mixed with the chalk was heated for two 
hours and 16 litres of air drawn through. On withdrawing 
the tube containing the silver, drops of liquid which proved 
to be nitro-glycerine were noticed in this (and in every 
other experiment) on the inside of the outer portion of the 
tube. The tube was therefore washed out with ether 
(care being taken that no silver was removed) dried, 
and weighed. In two experiments with this explosive 
there was no alteration in the weight of the tube and no 
mercury was recovered from the silver on heating in a com- 
bustion tube. The silver was, however, slightly discoloured 
in each case, probably due to a trace of oxidation, but as 
there was no increase in weight, the oxidation must have 
been very slight. 
100 grammes of this same explosive were then ground 
up with 100 grammes of French chalk which had been 
thoroughly mixed with two ccs. of a solution containing 
1 milhgram of mercuric chloride per cc. This was 
equivalent to 1 part of mercuric chloride in 50,000 parts 
