OF THE FLAME IN THE EXPLOSION OF GASES. 
323 
temperature and luminosity increase with the velocity; the flames are probably most 
luminous where they are hottest. 
But besides these j)rimary waves there are others to which Oettingen and Gernet 
call special attention; these are secondary waves ” running nearly parallel to the 
primary waves. The photogra|)h given in fig. 5 is referred to as showing very clearly 
four of these waves near together running parallel to the chief wave, which starts 
downwards from d. Other photographs show somewhat similar appearances, fie., of 
weaker waves running nearly parallel with and sometimes coalescing with primary 
waves. Oettingex and Gernet say that they can find no other reason for these 
waves following one another at a short interval, hut that successive explosions have 
taken place from the electrodes exactly as Bunsen imagined. The evidence relied on 
for these successive explosions, even if we accept the general interpretation of the 
photographs given by the authors, appears to me exceedingly slender. My own 
photographs will show how complicated the reflexions are when gases are exploded 
in a short tube, and how readily “ nearly parallel waves ” are produced with a single 
ignition of the gases. In the many photographs of the initial period of the explosion 
taken by me and my fellow-workers there is no indication of any second flame starting 
from the region of the spark. 
PABT II. 
Photographic Analysis of Detonation-Waves and their Pveflexions. 
[In conjunction uith E. H. Strange, B.Sc., cmd E. Graham, B,Sc.) 
In 1895 we were engaged on an investigation^ into the nature of the flame 
produced by the explosion of cyanogen with oxygen, when we made the observation 
which led to the present research. We found that tlie flame could he sharply 
photographed on Eastman’s films witliout tlie addition of any metallic salts, and that 
the films could be rotated very rapidly on a wheel without damage. In the experi¬ 
ments we then made, a mixture of cyanogen with oxygen was fired in a long vertical 
leaden jiipe having a short piece of glass tube let into it to serve as a window. This 
window was focussed on to tlie vertically moving film. When the explosion passed 
through the glass tube an illuminated image of the window was thrown on the 
photographic film, and was drawn out to a length depending on the rate of rotation 
of the film and the time during which the gas in the glass tube was photographically 
luminous. AVe found that cyanogen with its own volume of oxygen (burning to 
carbonic oxide) gave a much brighter and shorter flame than cyanogen burning to 
carbonic acid, a result in accord with the hypothesis that carbon (in carbon 
* “ On the Explosion of Cyanogen,” ‘ Jonrn, Ghem. Soc.,’ 189G. 
2 T 2 
