278 



ANNUAL EEPORT 'SMITHSONIAN INSTITUTION, 1908. 



the first instance. I believe that if one must use nitrocellulose, and if, 

 as seems to be the case, cotton is the best material for making it, one 

 ought to use the natural cotton only, and not common yarn, and less 

 still waste, which have both undergone so much forcible mechanical 

 and chemical treatment as to completely alter the character of the 

 cellulose and introduce elements of uncertainty and danger. These 

 should be avoided by the use of ripe raw cotton, which, of course, 

 would have to undergo suitable treatment to eliminate fat, husks, 

 and other impurities, but would not necessitate the whole bleaching 

 operation with its attending defects. 



Formerly the mixture for gun cotton consisted of 1 part of 1.500 

 nitric acid and 3 parts of 1.8-10 sulphuric acid, and each charge was 

 revivified by taking away one-quarter of the waste acid and adding 

 a mixture rich in nitric acid, so as to obtain about the original com- 

 position. The following table shows the result of revivifying the 

 waste acid ten times in a series of oj)erations made in 1886 by Doctor 

 Abelli and the author: 



The original mixture consisted of 1 part of nitric acid to 3 parts of 

 sulphuric acid, both of over 97 per cent monohydrate. Three parts 

 of waste acid were revivified with 1 part of fresh acids. 



It will be seen that the percentage of nitrogen contained in the 

 nitrocellulose reaches a maximum when the percentage of water in 

 the acid mixture is about 9 per cent, and not, as might be supposed, in 

 the stronger acid. 



The majority of factories prepare the nitrating mixture by giving 

 special consideration to the percentage of water in the first instance, 

 because by varying this nitrocellulose of widely different proper- 

 ties can be obtained. I have often said that by varying the concen- 

 tration of the acids, their temperature, and the time of nitration one 

 has three factors, each of Avhich can to a certain extent influence every 

 property of the nitrocellulose obtained. It is the custom in a major- 



