458 Sir Andrew Nolle [Jan. l<s, 



Immediately after the explosion, if the vessel be quite tight, the 

 valve at " B " is slightly opened, and the gas allowed to pass into the 

 tubes containing pumice stone and concentrated sulphuric acid to 

 absorb the aqueous vapour, and then into the gasometer, where the 

 volume is measured and the temperature of the gas at the moment of 

 measurement and the barometric height is taken. 



When it is quite certain that all the air is removed from the con- 

 ducting tubes, the gas is alloAved to flow into one of the tubes, and 

 shortly afterwards, or at fixed intervals, into the other two tubes for 

 the analysis of the permanent gases. 



When the whole of the gas has been transferred to the gasometer, 

 the explosion vessel is opened. There is always a considerable quan- 

 tity of water on the cylinder surface, generally smelling slightly of 

 ammonia. 



As much as possible is removed by a weighed sponge, and placed 

 in a weighed vessel closed by a ground-glass plate.. After all that it 

 is possible to absorb has been removed, a weighed vessel of calcium 

 chloride is placed in the chamber, and left for one or two days, when 

 the same procedure is repeated with a second vessel, when the vessel 

 is generally found to be quite dry. 



The analyses of the permanent gases produced were chiefly 

 determined by Dr. Sodeau with his gas analysis apparatus, which is 

 admitted to be the most convenient that has yet been devised. The 

 reconciliation of the analyses showed how accurate were the deter- 

 minations — I may say much closer than I expected. 



Several of the earlier analyses were made by the kindness of Sir 

 James Dewar, at Cambridge, and a few analyses as checks were made 

 at the National Physical Laboratory. 



Having found that there was a considerable difference in the 

 products of explosion when the explosives Avere fired under different 

 densities, I considered it necessary to determine the heat for all 

 points between the densities of '05 and "5. For this purpose I 

 used a calorimeter arranged as is here shown, and it is practically, 

 with slight variations, the same as that given by Ostwald in his 

 " Manual of Pliysico-Chemical Measurements," the surfaces of the 

 various vessels being nickel-plated and carefully polished. 



The heat capacity of the equipage and the exploding vessel being 

 carefully determined, for some hours before the experiment the 

 calorimeter is kept in a room maintained at as even a temperature as 

 possible, the explosion vessel, with the charge to be exploded, being 

 in the water, so that the whole system may be at the same 

 temperature. 



The rise in temperature by the explosion being approximately 

 known, the water in tlie outer cylinder is, before firing, kept at a 

 temperature approximately half-way between the initial and final 

 temperatures. 



The thermometers employed were calorimetric, used only for 



