102 REPORTS ON THE STATE OF SCIENCE.—1917. 
The reason for the neglect of this branch of work is undoubtedly due 
in large measure to the nature of the raw material cellulose. This sub- 
stance, not occurring naturally in a pure state, has to undergo drastic 
treatment in the way of bleaching, cleaning, and so forth before it is 
suitable for the manufacture of nitro-cellulose, and all treatment of the 
cellulose has a large influence on the character of the finished product. 
Furthermore the material largely employed in the explosive nitro-cellulose 
industry, namely cotton waste, is in itself waste product from the spinning 
mills. As most of the investigations are carried out by the chemists 
of the manufacturing firms, much of the work, probably the most valuable 
part, is never published. 
The colloidal solutions which have been investigated fail roughly 
into two classes, those made with volatile solvents such as ether-alcohol, 
acetone, ethyl acetate, &c., and those made with comparatively non- 
volatile solvents, of which nitro-glycerin and camphor are the chief. 
The principal work in the former class is that of Baker, who has investi- 
gated the viscosity of three important manufactured varieties of nitro- 
cellulose in various solvents, namely, acetone, ethyl formate, methyl, 
ethyl, propyl and amyl acetates, ethyl butyrate, aceto-ethyl-toluidide, 
ethyl tolyl ethyl carbamate, ethyl phthalate, ether-alcohol, ether-methyl 
alcohol, &c. He finds that the viscosity follows the law N=N, (1+AC)*, 
where N and Nj are the viscosities of solution and solvent respectively, 
C is the concentration, and A and R are constants depending both on 
solvent and solute. 
In general the viscosities of nitro-cellulose solutions are diminished 
by all treatment which the substance or its raw material undergoes, 
such as bleaching and cleaning the cellulose, stabilising the nitro-cellulose 
by boiling, heating the nitrating mixture or the finished product, and 
heating or exposing to light the solutions under investigation. The 
presence of traces of impurities in all stages of the work also adversely 
affects the viscosity. These causes, as well &s fundamental differences 
in the cellulose itself, account for the varying and discordant results 
obtained by different workers. 
In the system nitro-cellulose-nitro-glycerin, 7.e., blasting gelatine, 
the problem is to obtain a stiff gelatinous colloid which will not exude 
nitro-glycerin and will transmit the explosive wave with certainty. The 
preparation of a suitable nitro-cellulose, which in practice forms about 
7-8 per cent. of the mixture, is a matter of great delicacy both in 
selection of raw material and method of procedure, and practically no 
information is available on the subject. 
Hargreaves explains the behaviour of blasting gelatine by assuming 
a webbed structure of nitro-glycerin and nitro-cellulose, through the 
meshes of which free nitro-glycerin passes. As the colloid ages, more 
and more of the free liquid is absorbed into the gel, thus producing the 
state of insensitiveness to detonation which is one of the great troubles of 
the manufacturer, Comey (‘ Seventh Int. Cong. App. Ch.’) has, however, 
shown that liquid nitro-glycerin when confined in narrow tubes is but 
a poor transmitter of the detonating wave. Experiments by G. W. 
MacDonald and the writer have shown that the problem is still more 
complicated, and that, although the colloidal condition of the system 
has a large influence on the detonation of the explosive, its behaviour 
can be totally altered in a few hours merely by exhausting the air which 
