222 



Messrs. W. Macnab and E. Eistori. 



explosives), reference was made to certain experiments we had then 

 begim for the purpose of determining the actual maximum tempera- 

 ture reached diu*ing explosion. "We have now made a long series of 

 experiments in this direction, and propose to communicate some of the 

 results so far obtained, although the research is not yet complete. 



Some experiments have already been made by others for the pur- 

 pose of determining the temperature during explosion by placing 

 strips of metal of different melting point in a closed bomb and 

 observing the result after firing. Noble and Abel in their well-known 

 communication on explosives foimd in this way that the temperature 

 produced by the explosion of black gunpowder was slightly above the 

 melting point of platinum. 



There have been also several communications, on the same subject, 

 made by others, who have deduced the temperature during explosion 

 from theoretical considerations, but these calculations involved assiunp- 

 tions which as yet do not rest on an experimental basis. It appeared 

 to us desirable, therefore, to endeavour to determine experimentally, 

 and with greater accuracy than has hitherto been done, the actual 

 temperature developed when an explosive is fired in a closed vessel. 



The practical solution of this problem is, however, beset by several 

 difficulties : amongst others, the intensity of the temperature, the 

 extreme shortness of duration of the maxinumi temperature, and the 

 necessity of conducting the explosive reactions in a closed vessel. 



We were led to try a modification of the pyrometric method 

 developed hy Sir W. C. Eoberts-Austen, by observing that a thin 

 platinum vdre used for firing the explosive in the vessel by electricity 

 was often melted by the heat produced by the explosion, while thicker 

 platinum wires, which serA'ed to support the capsule containing the 

 explosive, were unaffected. 



This showed that the temperature reached was above the melting 

 point of platinum, and also that the duration of the maximum tem- 

 perature was ver}^ short. In the case of the thin wire, the small mass 

 of the metal allows the heat to penetrate it with sufficient rapidity to 

 raise it to the melting point before the period of maximum temperature 

 is past, while with the thick wire the time does not suffice for the 

 larger mass to be heated to the same extent. 



These considerations led us to argue that if rhodium-platinum 

 couples of wires of different diameters, sufficiently thick not to be 

 melted during explosion, were used in a bomb, the deflections of the 

 galvanometer indicated would vary inversely with the sizes of the 

 wires forming the couples ; that in this way we might get data which 

 would enable us to calculate the deflection of an infinitely thin couple 

 which could ])e capal>le of taking up the heat in an infinitely short 

 time, and that this deflection, expressed in degrees, would repre- 

 sent the actual maximum temperature reached. We also expected 



