THE PHOTOGRAPHIC ANALYSIS OF EXPLOSION FLAMES. 541 
immediately afterwards ‘ detonation ’is set up. Indeed, during the ‘ pre-detonation ’ 
phase a series of successive such ‘ignitions ahead’ may occur, detonation being set 
up immediately after the last one. 
Photographs will also be shown analysing the phenomenon of ‘ spin ’ in detonation 
which was first observed five years ago in detonations of moist 2CO+O, mixtures 
by C. Campbell and D. W. Woodhead, working in the late H. B. Dixon’s laboratory, 
and later confirmed in our laboratories at South Kensington. It has also been 
observed in detonations of methane-oxygen, pentane-oxygen, undiluted acetylene and 
other media, and apparently is caused by the helical rotation of a luminous ‘ head ’ 
of detonation in the flame front together with a long luminous ‘ tail’ behind it. The 
pitch (L) and frequency (f) of such rotation in detonations of a given medium varies 
with the internal diameter (D) of the tube, so that, while the ratio L/D for diameter 
up to about 1 inch is constant, and nearly equal to 3-0, the helical velocity of the 
rotating ‘ head ’ of detonation is approximately of the same order, irrespective of the 
medium or of the tube diameter. Thus, in the case of a moist 2CO+0O, medium, 
which exhibits the phenomenon most markedly, f=44,000 per sec., and p=3-95 cms. 
in a tube of 1-3 cms. internal diameter, the velocity of the ‘ head ’ of detonation along 
its helical path being 2,500 metres per second. 
These discoveries have necessitated a revision of the old classic conception of 
‘ detonation ’ and opened up new lines of inquiry concerning it. 
Within recent years the photographic method has also been used for the purpose 
of analysing (i) ignition phenomena, including the ‘induction’ period of explosions, 
(ii) the influence of moisture upon explosions of carbonic oxide-oxygen mixtures, and 
(iii) the influence of strong electrical fields upon gaseous explosions, and some results 
of such analyses were shown on the screen. 
Of the thirty photographs shown on the screen during the lecture, the following 
eight are reproduced here by kind permission of the Royal Society showing the scope 
and power of the method in analysing the phenomena concerned in gaseous explosions. 
The thin black vertical lines in the photographs are due to ‘ reference marks ’ on the 
outside of the tube at regular intervals of 20 cms. from the firing spark; a ‘ time 
scale ’ in millisecs. is shown in white on the right-hand thereof. 
Phil. Trans. Roy. Soc., A 228 (1929), Plate 9, No. 35.—No. 1 shows the effects of a 
series of superimposed ‘ shock waves’ at regular intervals of 1/400 sec. upon the flame 
movement during the initial stage of a moist 2CO+ 0, explosion initiated near the 
open end of a tube 2 metres long and 1-25 cms. internal diameter, the other end of 
which was closed. The flame, starting off from the igniting spark with an initial 
uniform velocity of circa 2 metres per sec. only, was struck and accelerated by five 
successive ‘ shock ’ waves, each of which abruptly raised its speed to a higher uniform 
level. In this way five successive uniform flame speeds of 9:2, 71:5, 76-4, 101-1 and 
122-0 metres per sec. respectively, were imposed upon it during the first 0°65 metre 
ofits travel. The short white arrows indicate where the four last shock waves overtook 
the advancing flame. 
Phil. Trans. Roy. Soc., A 228 (1929), Plate 10, No. 37.—No. 2 shows the flame of 
such a moist 2CO-+ O, explosion in the * pre-detonation ’ stage traversing the second 
metre section of the tube after having been accelerated (as in No. 1) by a series of 
“shock ’ waves. On entering it the flame was rapidly accelerated, but soon attained 
a nearly uniform velocity of 1,575 metres per second, at which it continued for a 
distance of about 20 cms. Its velocity then decelerated somewhat to about 
1,275 metres per second, and several successive ‘ ignitions-ahead ’ occurred, as the 
flame rapidly caught up with the ‘ shock waves’ which previously had passed through 
it, the last one being as much as 3 cms. ahead. As the flame spread from this last 
and extremely well-defined ignition point, a sudden and very intensive local combus- 
tion occurred, giving rise to both ‘ detonation’ and ‘ retonation’ waves. The speed 
of the ‘ detonation’ thus set up was 1,920 metres per second. 
Forthcoming Phil. Trans. Paper, 1931, Photograph No. 4.—No. 3. In this case a 
moist 2CO-+ O, medium has been ignited by the minimum requisite high tension spark 
at a point 2:5 metres from the closed end of the explosion tube, a ‘ detonator’ being 
simultaneously fired right up against the closed end, the space intervening between 
the ‘detonator’ and the explosive medium being filled with nitrogen. By such 
device a powerful ‘shock wave’ was set up by the detonator behind the explosive 
medium at the moment of its ignition. A feebly luminous flame started off from the 
igniting spark with a uniform velocity of 38°6 metres per sec. ; after about 4 millisecs. 
it was overtaken, and its velocity abruptly accelerated to 308 metres per sec., by the 
