464 
SIR ANDREW NOBLE: RESEARCHES ON EXPLOSIVES. 
It will be observed that at the higher densities and pressures there is generally a 
very tolerable accordance in the temperatures obtained from the two formulae, but as 
the density and pressure diminish the divergence becomes, in all cases, considerable, 
hut very greatly more with the explosives which develop very high temperatures and 
which give rise to large percentages of carbonic anhydride. 
The only construction I am able to put upon the tolerably close approximation of 
temperature given by the two formulae at high densities and pressures, and the wide 
differences which exist in some of the explosives at low densities, is that, as I think it 
reasonable to expect, at high densities dissociation of the carbonic anhydride is 
prevented by the very high pressure, and that the great difference between, for 
instance, Italian ballistite and nitrocellulose R.H. at, say, the density of O'l, is due, 
firstly, to the difference of the temperature at which the nascent gases are generated, 
and, secondly, to the proportion of C0 2 which is subject to dissociation, 
Formula (I) gives for Italian ballistite at d = O'l a temperature of nearly 5000° C., 
and for this explosive the temperature given by units of heat/by specific heat is 
nearly constant, while the percentage of C0 2 is 38'2. The same formula gives for the 
nitrocellulose at the same density a temperature of formation of 3200° C., while the 
percentage of C0 2 is only 19'4 5. 
I have pointed out that, under atmospheric pressure, the dissociation of C0 2 
commences at about 1300° C., and the very much higher temperatures of formation 
of the gases of the Italian ballistite, combined with its double percentage of C0 2 , 
appear to me to be sufficient to explain the results obtained with this explosive. 
If reference be made to Plate 16, it will be seen that while at the density of O'l 
there is, with Italian ballistite, a difference of about 1800° C. between the two 
formulae, there is with the nitrocellulose a difference only of under 800° C. 
The theory I venture to submit is as follows :— 
The nascent gases are generated at temperatures approximately as given by 
equation (l) and by the red curves in Plates 16, 17, and 18. 
Under the low densities and pressures at the very high temperatures with which 
we are concerned, the C0 2 and possibly some H 2 0 are partially dissociated, giving 
rise to the fall in temperature exhibited by the results obtained from equation (2) at 
low densities. At high densities, as already pointed out, the two equations give, in 
some cases, accordant results, in all cases tolerable agreement; it therefore appears 
to me to be reasonable to suppose that the facts I have recorded are due to partial 
dissociation at low densities and pressures, which dissociation is prevented by the 
very high pressures ruling at densities of 0'40, 0'45, and 0'50. 
As no free oxygen is ever found in the analysis, in cooling down any free oxygen 
due to dissociation must have combined and the heat lost by dissociation regained. 
The re-combination must, however, be very gradual, as no discontinuity is observed 
in the cooling curves. 
A certain amount of confirmation is given to the view I have taken by the fact 
