226 SIR ANDEEW NOBLE: RESEARCHES ON EXPLOSIVES. 



with which I have experimented justify the conclusion at which Sir F. ABEL and I 

 arrived with respect to gunpowder, viz., that any attempt to define by a chemical 

 equation the nature of the metamorphosis which one explosive may be considered to 

 undergo would only be calculated to convey an erroneous impression as to the definite 

 nature of the chemical results and their uniformity under different conditions. 



The apparatus shown in Plate 4 was employed for two purposes, (l) to determine 

 the time for the complete ignition of various explosives, or for various forms and 

 thicknesses of the same explosives ; (2) to determine the rate at which the exploded 

 gases part with their heat to the walls of the vessel in which they are enclosed. 



The high and low speeds that can be given to the drum permit these two 

 observations to be made by a single experiment. Thus, in Plate 10, I show the 

 commencement and part of the curves of two experiments, the one (fig. 1) fired at a 

 pressure of a little over 12 tons per sq. inch (1829 atmospheres), the other (fig. 2) 

 nearly 18 tons (2744 atmospheres). At the point "A" the charge is fired, and it will 

 be noted that the circumference of the drum is travelling at about 40 inches per 

 second. From fig. 2 it will be seen that at 2 seconds after firing the speed has, in 

 this experiment, been reduced to about an inch per second. 



The times required for the completion of ignition are given in Plate 11, and are 

 obtained from the curves shown on Plate 10 and from two similar curves. The 

 vertical scale in Plate 11 for the three last densities is doubled to make them accord 

 with density O'l, the spring employed in that experiment being half the strength of 

 that used for the last three. 



I may point out that when fired in close vessels the rate of combustion of the 

 explosives, even in the cord form, appears to be very constant, the increase of pressure 

 apparently nearly compensates for the reduction of surface, the differences in time 

 of burning being due to want of uniformity in the lighting, which in many cases is 

 very variable. This is illustrated by comparing the times of ignition of densities O'l 5 

 and 0'2 in Plate 11, where the total time from firing to complete ignition is less for 

 density O'l 5 than for density 0'2. In reality, however, after complete lighting the 

 latter is burning quicker, as may be seen by comparing the angles made by the 

 curves with the axis of abscissae. 



But this question is too large to enter into fully in the present paper. I therefore 

 only give the times of approximate complete combustion of cordite and M.D. cordite 

 of different diameters when exploded at a pressure of about 9 tons on the square inch. 



