172 PROFESSOR H. B. DIXON ON THE RATE OF EXPLOSION IN GASES. 
adopted. Tlie strong glass cylinder C, filled with water from a high cistern through 
the tube E, or (when necessary) with mercury from a second movable cylinder, was 
filled with the mixture from the holder by allowing the liquid to run out at the tap D. 
After the cylinder was filled, the three-way tap X was turned so as to send the gas from 
the holder through the drying vessels A,B. B, and through the three-way tap F, into the 
explosion tube. When the explosion tube was full, the tap X was turned so as to put 
Fig. 6. 
it in communication with the cylinder C. The tap at the further end of the explosion 
tube being closed, water or mercury was admitted to C, forcing the gas into the 
explosion tube through the drying vessels. When the manometer K showed the 
desired pressure had been reached, the taps were closed and the connections made 
with the chronograph and firing wires. Just before firing, the taps were opened to 
the manometer K and the pressure in the explosion tube was read off. The taps 
were then shut and the mixture fired. 
For experiments under reduced pressure the explosion tube was filled in the usual 
way. The gas was then sucked from the apparatus by a pump attached to the 
tube K (Fig. 7), a manometer sealed in at M (in place of the wmsh bottle L) giving 
the pressure in the explosion tube. 
Fig. 7. 
For firing the mixture of cyanogen with nitric oxide the following arrangement was 
adopted. The firing piece B was firmly screwed to a stout steel cylinder A. Through 
the other end of this cylinder a tube C was screvved so as to project 2 inches into the 
cylinder. Into the end of this tube a copper detonator holding fulminate of mercury 
was attached, by a caoutchouc stopper with a large hole, the open end of the deto¬ 
nator pointing inwards. The explosive mixture passed from the explosion tube 
through the firing piece and cylinder, escaping at the tube D which was also screwed 
