Theory of Explosions. 173 



initial decomposition determining the kind of reaction which 

 ultimately takes place. If the temperature of the whole mass 

 be previously raised, then the " critical" velocity will be- 

 come less. Berthelot considers that a specific change takes 

 place in the stability of an explosive as its temperature is 

 raised. This is doubtless true ; but if a minimum tempera- 

 ture of any part be the necessary and sufficient condition for 

 the production of a detonation, then the ease with which it 

 can be obtained, when the mass starts with a high tempera- 

 ture, will cceteris paribus be greater than if the original 

 temperature is low. If, therefore, we find that nitro-glyce- 

 rine is more liable to detonation the higher its initial tempe- 

 rature, we shall not be required to make any assumption as 

 to " increased sensitiveness," since we see that the minimum 

 temperature will be more easily reached, and that therefore 

 the "critical" velocity of initial decomposition may be 

 smaller. In other words, supposing we try to detonate 

 nitro-glycerine by an explosive which just fails at ordinary 

 temperatures, we should expect its chances of success to in- 

 crease as the temperature rises ; and this does in fact occur. 



The sensitiveness of an explosive to detonation will also, as 

 Berthelot points out, depend on its state of mechanical aggre- 

 gation. The critical velocity required to produce detonation 

 will cceteris paribus depend on the nature and value of the 

 elastic constants of the explosive as well as of the medium in 

 which it is to be exploded. We should, in fact, expect a 

 change in the critical velocity of detonation if we exchanged 

 the viscous resistance of liquid nitro-glycerine for the elastic 

 resistance of the same substance when frozen. Again, it 

 seems possible, as a result of this theory, that less powerful 

 detonation might be required to explode a given substance in 

 water than in air ; I am not, however, aware of any experi- 

 ments on this point. And so in other cases, though the 

 critical velocity of detonation must necessarily be a very 

 complex function, and difficult to predict, I see no reason on 

 that account to minimise its importance. On the other hand, 

 it seems to me to be in complete harmony with Abel's experi- 

 ments, and substantially embodies the view set forth by Dixon 

 in his paper " On Conditions of Chemical Change in Gases/'' 

 Phil. Trans. 1884. As far as I know, there is not a single 

 experiment which offers any evidence against it. What is 

 required by the theory for the formation of a detonation is 

 that a small part of the mass should be raised above a given 

 temperature, and not that a large portion should be raised to 

 a temperature below it. 



This leads at once to the consideration of the second point, 



