OF THE FLAME IN THE EXPLOSION OF GASES. 
3.37 
the iinburnt molecules would meet, at the moment of impact, the unburnt molecules 
travelling with an equal velocity in the opposite direction, and chemical combination 
would ensue between molecules l)oth of which were highly heated before they met. 
The difficulty of obtaiiiing an exact measure of the velocity of the waves reflected 
after collision is very considerable. With all the care we could exert, we found it 
impossible to obtain concordant readings for the angles made by the collision-waves 
with the horizontal, but we believe that the difference between the rate of collision- 
waves and waves reflected from the end of the tube is sufflcieiitly large to be fairly 
evident from a number of observations. For instance, six experiments with cyanogen 
and two volumes of oxygen gave as the ratio of the velocity of the detonation-wave (D) 
to the collision-wave (C) :— 
1-581 
sine of angle of C 
sine of angle of D 
1-40 
1-38 
1-19 
1-44 
.1-47 J 
Mean . 
1-41, 
whereas measurements of the reflexion-wave with the same mixture gave :— 
sine of angle of R 
sine of angle of D 
[ 2-111 
I 2-03 I 
2-32 I> Mean 
2-11 
.1-97 
2 - 11 . 
In a similar manner, measurements made on the relative rates of the collision-wave 
and the reflexion-wave in the mixture of cyanogen with one volume of oxygen, 
gave :— 
sine of angle of C sine of angle of R 
sine of angle of D sine of angle of D 
1-91 
1- 79 
2- 13 
1-98 
2 - 22 
2-15 
2-33 
2-08 
2 - 22 
Mean . . 1-95 
Mean . . 2-20 
In this case the numbers are too near for us to draw any certain conclusion, but in 
the first case it is difficult to suppose that the difference of 50 per cent, can be 
entirely experimental error. This inequality is an argument in favour of tlie view 
that there is a difference between a collision with a wall and with an approaching 
wave, but I would not lay stress on it to support my hypothesis of detonation. 
Our photographs have shown tliat the wave of detonation has certain character¬ 
istics by which it may be readily i-ecognised :— 
VOL. CC. —A. U X 
