Theory of Explosions. Ill 



fired from its geometrical centre. Then by Vieille's experi- 

 ments (Comptes Rendus, 1882 and 1883) on the time of ex- 

 plosion, it seems likely that the outer portions of fulminate 

 will be decomposed before they are removed to any appreci- 

 able distance from their original positions. We shall there- 

 fore have a sudden expansion in all directions, caused by the 

 increase in volume of the explosive substance during the 

 explosion. There seems no reason under perfectly symmetri- 

 cal conditions why the expansion should not go on as it began, 

 until the cooling of the sphere of hot gases becomes so 

 marked as to prevent further expansion. 



If the conditions, however, are not such as to allow of 

 symmetrical expansion — as always occurs in practice — then 

 we shall have the bounding surface of the explosion-gases 

 more curved in some places than in others — that is, the strain 

 will be greater at some parts than at others, and in fact may 

 become so great at points of great curvature as to lead to 

 what I venture to call a state of " break-down.''' In other 

 words, the compressed gases will in this case escape, not by 

 gradual expansion but by jets, from points whose position is 

 fixed by the conditions of explosion, In these jets we should 

 have the necessary and sufficient conditions for the establish- 

 ment of vortex motion. If vortex motion were set up, then 

 it seems likely that much greater effects might be transmitted 

 in some directions than in others, though at considerable 

 distances the effects would tend to become uniform in all 

 directions. This view of the actions of explosions will, I 

 think, enable us to explain several difficulties which I men- 

 tioned as occurring in the interpretation of Abel's experiments. 

 These experiments are fully described by Abel in various papers 

 in the ' Philosophical Transactions/ and in the Annates de 

 Chimie et de Physique, vol. 165. They have been discussed 

 at some length by M. Berthelot in his work on Explosives, 

 and though much light has been thrown on many points, I 

 venture to think that the complete elucidation of others is still to 

 be sought. Among these I would place the three following : — 



The want of correspondence between the explosive actions, 

 as measured by the effect produced on copper plates, and the 

 effects produced in causing other explosions. The apparent 

 capriciousness of explosions of the more violent kinds; and 

 the production of explosions by influence. 



With respect to the explosion of gunpowder, M. Berthelot 

 has nothing new to add to the generally received theory that 

 it differs in nothing from ordinary combustion, except that it 

 is more rapid. 



The theory of detonation, however, is more fully treated, 



