Theory of Gunpowder . 505 



analysis, is commonly used as a control. In our experiments 

 this check is inapplicable, as the quantity of residue was not 

 weighed. But, on the other hand, we have an equally certain 

 verification in the fact that the quantities of K, N, S, C, con- 

 tained in 1 grm. powder must be found in almost unchanged 

 quantity in the products of combustion. Such a conipai'ison 

 gives for our analysis, — 



In the unexploded powder, — 

 K 0-3050; N = 01096; S = 0-0984; C = 0'0769; = 0-4057. 



In the exploded powder, — 

 K 0-3050; N = 0-1096; 8 = 00989; = 00780; = 0-3936. 



The agreement in the quantities of potash and nitrogen, from 

 which the equations for the calculation of the above scheme are 

 taken, serve as proof of the accuracy of the calculation ; the 

 agreement of the three quantities of sulphur, charcoal, and oxygen, 

 as so many proofs of the accuracy of the experiments. 



It follows from this equation that 1 grm. of powder in ex- 

 ploding yields 1931 cubic centims. of gas. The quantity which, 

 according to the previous theory, ought to be found, amounts to 

 3309 cubic centims., or above one-third more. 



After having established the transformation which occurs in 

 the combustion of gunpowder, we only need a determination of 

 the temperature of its flame, to have all the elements from which 

 may be deduced a theoretical measure of the work performed in 

 the combustion of powder. We have therefore to occupy our- 

 selves with the fifth question proposed at the commencement of 

 our i-esearch, respecting the temperature of the combustion of 

 powder. In order to obtain a certain basis for the solution of 

 this question, it is necessary to form an accurate representation 

 of the nature of the flame of powder. If we suppose 1 grm. 

 of powder simultaneously burnt throughout its entire mass, 

 w units of heat will be liberated ; the products of decomposition 



w 



of the exploded powder will be thereby heated — degrees at the 



moment of their formation, in which quotient s is the specific 

 heat of the products of combustion as compared with water. 

 This quotient is therefore the temperature of the flame, mea- 

 sured by the thermometric unit of the heat of combustion. But 

 the temperature of the flame experiences in fact a steady lowering, 

 caused by radiation and conduction of heat. As the tempera- 

 ture — only remains constant for an infinitely short space of 



time, it cannot be measured by the usual thermometric means. 

 The phaenomena which occur in the continuously burning scin- 

 tillating flame of the pressed powder are entirely similar. This 

 Phil. May. S. 4. No. 103. f^upjd. Vol. 15. 2 L 



