85 



mixtures^". The viscosity/concentration curves of such mixtures 

 usually show maxima, the positions of which vary with the temperature, 

 and Bramley concludes that the viscosity of liquid mixtures as an inde- 

 pendent test of compound formation in liquid mixtures is unsatis- 

 factory. The writer suggests that the viscosity /concentration curves 

 of mixtures of liquids containing a constant weight of a viscous colloid 

 in solution might give more useful results. The viscosity of the 

 solvent would be masked by the viscosity due to the colloid, dissolved, 

 presumably, in a complex of the two liquids. It is certain that in 

 many cases minima would be found, and it would be interesting to 

 see whether these minima also shifted with change of temperature, 

 and whether their position corresponded with molecular addition 

 compounds. The method could be applied to all mixtures possessing 

 solvent power for a viscous colloid, whether the constituents separately 

 were solvents or not. For instance, mixtures of aliphatic alcohols 

 and benzene hydrocarbons should be examined with nitrocellulose 

 in solution, and mixtures of ab'phatic alcohols and chlorinated paraffins 

 with acetylceUulose in solution, and so on. Although the results 

 obtained would be chiefly valuable as a contribution to the study 

 of binary mixtures, they would also be useful data in the chemistry 

 of emulsoid solutions^^a 



The same question was encountered in a different form some years 

 ago by the writer in a purely technical investigation. It is well known 

 that ethyl alcohol and toluene together form a solvent for nitro- 

 cellulose of low nitration. To fuid the composition of the best solvent, 

 constant weights of nitrocellulose of imperfect solubility were treated 

 with a constant volume of toluene and alcohol in different proportions 

 and the amount of nitrocellulose dissolved by each mixture was 

 estimated. The results showed that the optimum mixture corresponded 

 closely with a mixture of three molecules of alcohol to one of toluene 

 thus suggesting the existence of a complex C7H8, 3C2H5OH. Regarded 

 from the purely colloid point of view, perhaps, the development of 

 solvent power by relatively small additions of other substances to 

 partial or non-solvents is the more interesting owing to its analogy 

 to organised processes such as nutrition. 



The nature of the soUd camphor-nitrocellulose complex is still 

 unknown. Schiipphaus^^ believes that there is some kind of chemical 

 combination between camphor and nitrocellulose, chiefly on the 

 ground that the heat of combustion of celluloid is less than that of its 

 constituents. But this is only the complement of the fact that 

 camphor and nitro-cellulose evolve heat when mixed, which cannot 

 be regarded as evidence of chemical combination. The large changes 

 in surface energy which must accompany gelatinisation would be 

 sufficient to explain the thermal phenomena. 



The optical rotation of an acetone solution of camphor and 

 nitrocellulose is equal to the sum of the separate rotations of the 

 camphor and the nitrocellulose, within the Umits of experimental 

 error^^. 



Dubosc^^ expresses the following views on the constitution of 

 celluloid. Celluloid is a " camphrogel " of nitrocellulose, camphor 

 being the dispersion medium. In the process of gelatinisation 



