496 EEPOBT— 1888. 



rise, after a very small fraction of a second from tlie beginning, in tlie 

 most explosive mixtures to a very fugitive pressure, due to an effect of 

 the ' explosive wave ' discovered by Vieille and Bertbelot ' whereby pres- 

 sure is added during the propagation of the wave of pressure, so that the 

 pressure of a layer distant from the centre of propagation is momentarily 

 greater than of those nearer, and is therefore greater than the pressure be- 

 comes alter the very short interval during which it is equalised ; the very 

 sensitiveness of the Deprez arrangement and the fact that it recorded just 

 the maximum pressure were therefoi'e obstacles against getting a true idea 

 of the state of the gas as a whole during the explosion as to pressure. The 

 explosive wave for oxyhydrogen mixture has the uniform velocity of about 

 2,810 metres per second ; this is called the specific velocity of the explosive 

 wave for this gaseous mixture, and is independent of the initial pressure. 



The Deprez indicator was thei'efore discarded in favour of a mano- 

 meter devised by Bom'don, which indicated the pressures throughout the 

 whole process of explosion and cooling. In place of the eudiometer a 

 cylindrical vessel was used with diameter of its circular base equal to the 

 height of the cylinder, the explosion being made at the centre of the 

 cylinder ; the propagation of the explosion in this vessel was free from 

 the irregularities incident to the eudiometer. This wide cylinder in fact 

 was nearer in its shape and in its action to a sphei'e in which the explo- 

 sion is started at the centre, this being an arrangement the extreme im- 

 portance of which was not fully recognised at first ; the advantage of the 

 spherical arrangement being that, as the propagation of the explosion 

 takes place uniformly in all directions, the effect would reach all parts of the 

 inner surface of the sphere at the same time, the temperature and pressure 

 would be more rapidly equalised throughout the gaseous mixture, and 

 the cooling would be regular from every portion of the outer surface. 



Connected with this manometer was a style for marking on a sheet of 

 paper surrounding a vertical cylinder, which is made to rotate rapidly and 

 uniformly during an experiment, a line representing the rise and fall of 

 pressure throughout the explosion and subsequent cooling. 



The whole duration of an explosion is a small fraction of a second in 

 the case of the most explosive mixtures, and generally one second in these 

 experiments is a large interval of time. 



Mallard and Le Chdtelier. Experiments and Results. 



By observation of the rate of cooling, or strictly of the variations of 

 the pressure during the whole period of cooling, after an explosion, through 

 fhe diminution of pressure as registered by the chronograph for intervals 

 of 001 or a smaller fraction of a second, a curve is got giving the relation 

 between pressure and time, intervals of time being registered by the 

 abscissjB, and the pressures by the ordinates. An equation is then found 

 by theory to express the results of direct observation, and the equation is 

 normally that of a straight line, of which the abscisste and ordinates are 

 known from the conditions and recorded progress of the pressure in each 

 experiment. The form of the equation will be different according as there is 

 or is not condensation, as there is, e.g., when water is one of the products 

 of explosion. Now by these formulae the rate of cooling is represented by 

 means of a straight line, and this straight line in most cases should cor- 



' C. R. t. sciii. p. 21 ; t. xciv. pp. 101, 149, 822 ; t. scv. pp. 151 and 199 ; Ann. de 

 CMm. i5),2S, 1883, p. 289. 



