86 Studies of Diffusion through Rubber Membranes [Sept 



Repeated removals of the diffusate in the independent scarlet R 

 experiment, and replacements with fresh solvent for a period of 

 about a weck, led to the Separation from the original pigment- 

 product of all its red diffusible matter. The bag, at that stage of 

 the treatment, contained considerable brownish-red, indifTusible 

 material, which evidently was not scarlet R. This result, and simi- 

 lar observations with other pigments, emphasized Dr. Gies' opinion 

 that it might be possible to purify pigment preparations in this way 

 and that their value as coloration agents, for histological staining 

 especially, might thus be considerably enhanced. 



It will be noted that those pigments which diffused most rapidly 

 were the so-called " fat colors," i. e., those soluble in, or staining, 

 the common fats and oils. Again, with these pigments the dif- 

 fusion is the most rapid, and therefore the most satisfactory, when 

 the solvents are those which, in the State of vapor, soften rubber. 

 It will thus be seen that apparently the membrane, as well as the 

 solvent, exert selective action. This is true to a far greater extent 

 in experiments of this kind than in the ordinary dialyses in aqueous 

 media. 



When we arrived at this point in these experiments, to which 

 we could give but a few hours weekly, our period of residence at 

 Columbia University was about to close and, after completing some 

 repetitions of previous observations in this connection, we were 

 obliged to discontinue the work. It is Dr. Gies' intention to pro- 

 ceed along lines suggested by the results already obtained in this 

 preliminary investigation. 



