We now come to the application of high explosives to warfare in the shape of bursting charges for shells. This is the latest phase of the problem, and it is undoubtedly fraught with the most important consequences to both attack and defense. Difficult as it has been to obtain an exact estimate of the force of different explosives under water, the problem is far greater out of the water and under the ordinary conditions of shell fire; the principal obstacle being in the fact that it is physically impossible to control the force of large quantities in order to measure it, and small quantities give irregular results. Theoretically, the matter has been accomplished by Berthelot, the head of the French government "Commission of Explosives," by calculating the volume of gas produced, heat developed, etc.; and this method is excellent for obtaining a fair idea of the specific pressure of any new explosive that may be brought forward, and determining whether it is worth while to investigate it further; but the explosives differ so much from each other in point of sensitiveness, weight, physical condition, velocity of explosive wave, influence of temperature and humidity, that we cannot determine from mere theoretical considerations all that we would like to know. Various methods of arriving at comparative values have been tried, but the figures are very variable, as will be seen by the following tables. Berthelot's commission, some ten years ago, exploded ten to thirty grammes of each in 300 pound blocks of lead and measured the increased size of the hole thus made. The relative result was: No. 1 dynamite 1.0 Dry gun-cotton 1.17 Nitro-glycerine 1.20 Powder blew out and could not be measured. Mr. R.C. Williams, at the Boston Institute of Technology, in the winter of 1888 and 1889, tried the same method, but used six grammes in forty-five pound blocks of lead. He obtained a relative result of-- No. 1 dynamite 1.0 Dry gun-cotton 1.37 Nitro-glycerine 2.51 Explosive gelatine 2.57 Forcite gelatine 2.7 Warm nitro-glycerine 2.7 Gunpowder 0.1 The powder gave great trouble in this case, also, by blowing out. M. Chalon, a French engineer, obtained some years ago, with a small mortar, firing a projectile of thirty kilos and using a charge of ten grammes of each explosives, the following ranges: Meters. Blasting powder 2.6 No. 1 dynamite 31.4 Forcite of 75 per cent. N.G. 43.6 Blasting gelatine 45.0 Roux and Sarran obtained by experiments in bursting small bomb shells the following comparative strengths of ranges: Powder 1.0 Gun-cotton 6.5 Nitro-glycerine 10.0