146 Royal Society: — 



obtained from Hirtenberg and Stowmarket. Tbe general analytical 

 results are as follows : — 



Air-dry gun-cotton contains very uniformly about two per cent, 

 of water, which proportion it reabsorbs rapidly from the atmo- 

 sphere after desiccation. If exposed to a moist confined atmosphere, 

 it will gradually absorb as much as six per cent, of water ; but it 

 rarely retains more than two per cent, upon re-exposure to open air. 



The mineral constituents of gun-cotton vary according to the 

 quality of the water employed in its purification. The average pro- 

 portion of ash furnished by gun-cotton prepared at Waltham Abbey, 

 where the water used is hard, amounts to one per cent. It should 

 be observed that the process of " silicating " the gun-cotton, which 

 is prescribed by Von Lenk, but the value of which is not admitted, 

 has been applied at Waltham Abbey only in special experimental 

 operations. Its use naturally adds to the mineral constituents con- 

 tained in the finished products. 



The proportions of matters soluble in alcohol alone, and in mix- 

 tures of alcohol and ether, were found to be remarkably uniform 

 in products of manufacture obtained by strictly following Von Lenk's 

 directions. In the ordinary products from Waltham Abbey, the 

 matter extractable by alcohol amounted to between 0'75 and 1 per 

 cent., and consisted of a yellowish nitrogenized substance possessed 

 of acid characters, and evidently produced from matters foreign to 

 cellulose (which are retained by cotton fibre after its purification), 

 and the products of oxidation which escape complete removal 

 when the gun-cotton is submitted to purification in an alkaline 

 bath. The average proportion of matter extractable by ether and 

 alcohol after the alcoholic treatment is from 1 to 1*5 percent. This 

 consists of one or more of the lower products obtained by the action 

 of nitric acid upon cotton-wool, the existence of which was esta- 

 blished by Hadow. The causes of the invariable production of small 

 proportions of these substances in the ordinary manufacturing ope- 

 rations, and of their existence in larger quantities in exceptional 

 instances, have been carefully examined into. Their absolute re- 

 moval from specimens of gun-cotton, purified for analytical purposes, 

 was found to be almost impossible. 



The methods employed for determining the proportions of carbon, 

 hydrogen, and nitrogen in gun-cotton, and the relative proportions of 

 carbonic acid and nitrogen furnished by its combustion, have been very 

 carefully tested. Four different methods of determining the carbon 

 were employed, and forty-nine successful estimations of that element 

 have been accomplished in a variety of products of manufacture. A 

 number of very concordant hydrogen-determinations, and eighteen 

 direct estimations of the volumes of nitrogen furnished by the 

 complete oxidation of gun-cotton, have been made. The indi- 

 vidual as well as the mean results obtained in these analytical 

 experiments correspond much more closely to the requirements 



of the formula € 6 H 7 N 3 O u = € 6 1 „ N q \ & 5 , trinitro-cellulose, or 



€ 12 H 14 7 , 3N 2 O-, trinitric cellulose, than to the formula recently 

 assigned for gun-cotton by Pelouze and Maury, € 24 H 36 ld , 5 N 2 O s . 



