Proposed Substitutes for Gunpowder. 89 
Lenk’s process; the conclusions of those chemists being 
founded partly upon some analytical results, and partly 
upon the increase of weight which they found cotton to 
sustain when submitted to treatment with the mixed acids. 
They found the greatest increase in weight to be 78 per 
cent.,a number slightly in excess of that which would 
correspond to the requirements of the formula which they 
adopt. ; 
An experimental inquiry into the composition of gun- 
cotton, as obtained by Von Lenk’s process, has been insti- 
tuted by Mr. Abel; and the very numerous analytical and 
synthetical results which he has obtained confirm the 
correctness of the formula assigned by Crum and Hadow 
to the most explosive gun-cotton, and demonstrate satis- 
factorily that the products obtained by following strictly 
the instructions given by Von Lenk are invariably trinitro- 
cellulose, in a condition as nearly approaching purity as a 
manufacturing operation can be expected to furnish. 
The most explosive gun-cotton is perfectly insoluble in 
mixtures of ether and alcohol; but by varying the pro- 
portions and strength of the acids employed for the con- 
version of cotton, products of less explosive character are 
obtained, which are more or less freely soluble in ether and 
alcohol (furnishing the well-known material collodion). If, 
therefore, in manufacturing gun-cotton, the, conditions 
essential to the production of insoluble pyroxylin are not 
strictly fulfilled, the uniformity of the product will suffer. 
The ordinary products of manufacture are never alto- 
gether free from soluble gun-eotton; but the proportion 
present is small and very uniform, amounting to about 1°5 
per cent. They contain, besides, a small quantity (about 
O'5 per cent.) of matter soluble in alcohol alone, and 
possessed of acid characters, which is evidently produced 
by the action of nitric acid upon such small quantities of 
resinous or other matters foreign to pure cellulose, as are 
not completely removed from the cotton fibre by the puri- 
fication which it receives. 
There appears good reason to believe that this impurity 
in gun-cotton is of comparatively unstable character, and 
that the great proneness to spontaneous decomposition 
which has been observed by Pelouze and Maury, De Luca, 
and others, in some specimens of gun-cotton, is to be as- 
_cribed in great measure to the existence in those specimens 
of comparatively large proportions of those unstable bye 
products. 
