of preparing and employing Gun-cotton. 543 



cotton by oxygen, carbonic acid and nitrogen in the proportion of 



Carbonic acid corresponding to C 24 . 48 volumes 

 Nitrogen corresponding to N 5 ... 10 „ 



or, for 100 volumes of the total gas, 



Carbonic acid 82*8 volumes 



Nitrogen 17'2 „ 



The German formula should give 



Carbonic acid 48 volumes 



Nitrogen *..... 12 „ 



or, for 100 volumes of the total gas, 



Carbonic acid , 80 volumes 



Nitrogen 20 „ 



The difference in the two proportions of nitrogen, 17 per cent, in 

 the case of our formula, 20 per cent, in the case of the German 

 formula, is sufficient to permit a choice between the two. 



The following are the conditions under which we have worked. 

 Gun-cotton mixed with about twenty times its weight of pure 

 oxide of copper was placed in a green glass tube, and tapped so 

 that it occupied a length of about 20 centimetres. Above it 

 was placed, first, about an equal length of coarse oxide of copper, 

 and lastly of pure reduced copper. The gaseous products of 

 combustions were collected over mercury in graduated tubes, 

 when a sufficient volume of gas had first been rejected. The 

 total volume of gas suitably cooled was first noted, then the 

 carbonic acid was absorbed by potash, and the residue calculated 

 as nitrogen. 



This combustion requires great care. There are in fact two 

 dangers — that of not completely reducing by the copper the 

 nitrous vapours arising from the decomposition of the gun-cotton, 

 and that of not completely changing into carbonic acid the car- 

 bonic oxide produced in the same decomposition. To avoid 

 them the two columns of coarse and of fine oxide must be 

 kept at redness throughout the duration of the experiment, and 

 the progress of the decomposition must be so regulated that 

 the gases produced traverse these two columns with extreme 

 slowness. 



In a great number of analyses, made with our different speci- 

 mens, we found the numbers between 17'2 and 17*5. 



We have moreover verified our formula by weighing the nitro- 

 gen as gas. The weight of gas is 14*14 per cent, in the Ger- 

 man formula, and 12*15 per cent, in ours. This difference, 

 again, is sufficient to permit a choice between the two formulas. 



The analysis was made in the following manner. At the bot- 



