OF TPTE FLAME TX TTTE EXFLOSIOX OF OASES. 
n49 
and then moves on rapidly from the point where the reflected sound-wave reaches 
the front of the flame. The movement on each side is now seen to he nnsymmetrical; 
the flame to the left does not reach its end of the tube before it is again checked, 
that to the right reaches the end of the tube and sends back a powerful retonation- 
wave. The consecpience of this dissymmetry is the greater intensity and rapidity of 
the wave started from the right over that from the left. Although the reflexions of 
these waves at first run nearly parallel, yet after four journeys to and fro the 
stronger wave catches the weaker and coalesces with it, and the rest of the picture 
shows only repeated reflexions of a single wave. A similar coalescence of two waves 
is shown in flg. 7 of v. Oettixgen and v. Gernet, and this figure closely resemlfles 
a portion of our photograph. It is tlius evident how “ secondary ” waves running 
parallel with the “ primary ” are produced ; there is no need to invoke any “ successive 
partial explosions ” to account for them. 
Fig. 74 shows in outline the path of the flame and the sound-waves, wliich may 
he compared with tiie “ schema ” of fig. 8, according to v. Oettixgen and v. Gerxet. 
In fig. 75 tlie gas was also fired in tlie middle, Init the explosion is more 
symmetrical, and “nearly parallel” waves can 1)e seen to he produced l)y reflexion 
from both ends of the tul^e. 
Our next photograph (7f)) shows the eftect of firing the gas near one end of the 
tul)e. The flame proceeds with increasing velocity to the findher end, where a strong- 
wave is sent hack; this single wave shows less complications than tlie doulde wave 
started in 74. 
In 77 the gases are fired at one-fourtli the lengtli from one end. Very complicated 
reflexions are produced, which only gradually become absorbed. In tliis photograph 
the mode of formation of “ nearly parallel waves” can easily he traced. 
In fig. 78 is seen the effect of lighting the explosive mixture at one end, and at a 
point one-quarter from the further extremity simultaneously. The effect of tlie 
sound-wave proceeding from each jioint of ignition is plainly seen when it reaches the 
other flame. The reflected-wAves, visible in the burning gases, cross the unignited 
mixture, and again become visible in the flames beyond. When the two flames 
coalesce the picture resembles the other photographs. In fig. 79 the gases were 
ignited simultaneously at each end of the tulie. 
[The last three photographs, reproduced in figs. 80, 81, and 82, show the detonation- 
wave set up in a long tube reaching a portion of gas which (having lieen independent!}’' 
ignited) is still in the initial stage of combustion. In fig. 80 the gas was lighted in 
the centre of the tid)e just before the detonation-wave arrived. The detonation is 
slightly damped down on meeting the already ignited gas. In fig. 81 the detonation- 
wave strikes a flame started at the end of the tube. In fig. 82 the initial flame had 
spread considerably before the detonation struck it near the left-hand edge of the 
photograph. In each case the wave is propagated through the already burning gas 
like a retonation-wave.—H. B. D. and L. B., October, 1902.] 
