THE ATMOSPHERE AS AN ANVIL. 223 



on the surface of charcoal-grains ; the first layer of molecules must 

 be consumed before the second can be reached, and so on. Hence the 

 process, although very rapid, must take a sensible time. In the nitro- 

 glycerine, on the other hand, the two sets of atoms, so far from being 

 in different grains, are in one and the same molecule, and the internal 

 combustion is essentially instantaneous. Now, this element of time 

 will explain a great part of the difference in the effect of the two ex- 

 plosions, but a part is also due to the fact that nitro-glycerine yields 

 fully nine hundred times its volume of gas, while with gunpowder the 

 volume is only about three hundred times that of the solid grains. 

 There is a further difference in favor of the nitro-glycerine in the 

 amount of energy liberated, but this we will leave out of account, al- 

 though it is worthy of notice that energy may be developed by internal 

 molecular combustion as well as in the ordinary processes of burning. 

 " The conditions, then, are these : With gunpowder we have a vol- 

 ume of gas, which would normally occupy a space three hundred times 

 as great as the grains used, liberated rapidly, but still in a perceptible 

 interval. With nitro-glycerine a volume of gas, nine hundred times 

 that of the liquid used, is set free, all but instantaneously. I^Tow, in 

 order to appreciate the difference of effect which would follow this 

 difference of condition, you must remember that all our experiments 

 are made in air, and that this air presses with an enormous weight on 

 every surface. If a volume of gas is suddenly liberated, it must lift 

 this whole weight, which, therefore, acts as so much tamping mate- 

 rial. This weight, moreover, cannot be lifted without the expenditure 

 of a large amount of work. Let us make a rough estimate of the 

 amount in the case of nitro-glycerine. We will assume that in the 

 experiment at Newport the quantity exploded yielded a cubic yard 

 of gas. Had the air given way instead of the rock, the liberation of 

 this volume of gas must have lifted the pressure on one square yard 

 (about nine tons) one yard high, an amount of work which, using these 

 large units, we will call nine yard-tons, or about sixty thousand foot- 

 pounds. Moreover, this work must have been done during the ex- 

 cessively brief duration of the explosion, and, it being less work to 

 split the rock, it was the rock that yielded, and not the atmosphere. 

 Compare now, the case of gunpowder. The same weight of powder 

 would yield only about one-third of the volume of gas, and would, 

 therefore, raise the same weight to only one-third of the height ; 

 doing, therefore, but one-third of the amount of work, say twenty 

 thousand foot-pounds. Moreover, the duration of the explosion being 

 at least one hundred times longer than before, the work to be done in 

 lifting the atmosphere during the same exceedingly short interval 

 would be only -^^-^ of twenty thousand foot-pounds, or two hundred 

 foot-pounds, and, under these circumstances, you can conceive that it 



might be easier to lift the air than to break the rock.* 



1 We here omit, for want of space, the beautiful explanation, afiforded by the " New 



