EXPERIMENTS ON SOLID AND GASEOUS EXPLOSIVES. 367 



the consequences of an accident disastrous, it was decided not to exceed half this limit. 

 The second enclosure was alone used for higher pressures, it being, as we shall see, of 

 stronger construction. 



Apart from the effect of actual pressure, that of the sudden impact or blow given 

 by the more rapid explosives has to be considered. As will be seen below, some 

 mixtures of compressed coal gas and oxygen develop their full pressure in something 

 like one ten-thousandth of a second and, in fact, occasionally detonate. It is difficult 

 to estimate the actual strain produced by a force so suddenly applied.* When we 

 consider that the present work comprised the repeated explosion of such mixtures, it 

 will be seen that exact calculation becomes impossible. In all probability, during the 

 course of the first few explosions of this kind the part of the material nearest the 

 inner surface is strained to beyond its limit of elasticity, and therefore yields. In the 

 case of steel, like the present, having a fair elongation, the first effect is actually to 

 strengthen the enclosure ; the inner layers of the steel having been thus permanently 

 elongated are under an initial compression which will help them in resisting further 

 deformation. Aided, however, by the extremely rapid variations of temperature, this 

 effect will in time cause surface cracks. Under successive strains the cracks will 

 deepen to an extent that may become dangerous. Being on the inner surface of the 

 chamber, the extent of the damage cannot be clearly ascertained. In the present 

 work this danger was guarded against by a method which, though perhaps some- 

 what crude, is at least easily carried out and, faute de mieux, may be considered 

 satisfactory. On the outer surface of the enclosure a ring was accurately turned ; the 

 plane through the centre of this ring passes through the centre of the sphere and 

 through the gas and mercury inlets : it therefore encircles the weakest portion of the 

 enclosure. A large micrometer gauge was made, by means of which the diameter of 

 this ring was from time to time measured. This micrometer will clearly show an 

 increase of one three-thousandth of an inch on the 8-inch diameter, or a change of 

 about one two-hundredth of one per cent. 



Up to the present no variation of diameter has been detected, and it is reasonable 

 to infer that the apparatus has not been strained to a dangerous extent. 



A sectional drawing of the enclosure is given in fig. 5. 



The recording gauge screws into A, the steel ring (D, fig. 3) pressing on to the 

 ledge a and thus forming a joint. The end of the gauge fits closely into the neck 

 b and protects the joint from contact with the heated gases. The firing plug fits 

 into the aperture B. 



When gaseous mixtures are to be tested, the two valves which screw into C and D 

 are brought into use. The cavity is first filled with mercury through C and the gas 

 is then forced in through D. As soon as the mercury has been driven out, the valve 



* It is usual to take an instantaneous load as equivalent to twice the same statical load. In the present 

 case, however, we have to deal with the momentum of the gas itself, which is travelling at an enormous 

 speed. 



