Gelation in Reversible Colloidal Systems. 



101 



from soluble substances by treatment for months with hot and cold 

 distilled water. The packet was then pressed in a screw press, and the 

 large yield of fluid collected. When the fluid ceased to flow, the 

 solid which remained in the canvas was removed to a stoppered 

 vessel. 



The fluid was found to be a solution of agar. This was proved by 

 evaporating some after it had been thrice filtered to a small bulk, when 

 it was found to set to a typical clear gel. The results of the study of 

 the ternary systems give sufficient grounds for defining the expressed 

 fluid as a solution of agar in water, and the solid which remains in the 

 canvas a solution of water in agar. The eff'ect of the composition of 

 the gel and of temperature upon the distribution of the water and the 

 agar in these two phases was determined. The percentage composition 

 was arrived at by drying a known weight of each, and assuming that 

 the residue was entirely composed of agar. The results which were 

 obtained lie far outside of any error which could have been introduced 

 by this assumption when one considers the pains which were taken to 

 free the gels from foreign bodies. Further, in every case an examina- 

 tion of the dry residue was made in order to prove that it was composed 

 of matter capable of forming with water a typical agar gel. 



Experimental Difficulties, 



The method used to separate the two phases, though at first sight 

 crude, was found to be the most eff'ective. The great error to be avoided 

 is the blocking of the canvas pores by a mass of the solid phase, so 

 that, instead of the true fluid phase, one really expresses fluid which has 

 been forced through a membrane (pressure filtered). Owing to this error, 

 a press, which I had specially made, and in which the piston drove the 

 gel directly down on to a disc of canvas, proved quite useless. Very 

 great force was necessary to express a fluid which was found to be 

 almost pure water. 



To succeed, it is necessary to avoid direct or great pressure. The 

 masses of gel are loosely placed in a long canvas packet, which is then 

 deformed by pressing the ends together. The pressure necessary to 

 yield abundant fluid is now quite small, for the solid framework of the 

 gel is destroyed by being rubbed against the canvas, and is reduced to 

 fine particles, while the fluid easily makes its way through the coarse 

 pores of the canvas. Eaising the pressure always expresses a fluid 

 poor in agar, while with slight to moderate pressure the concentration 

 of the expressed fluid, as tested by determining the solids in the yield 

 at diff'erent stages, remained fairly constant, but always with a slight 

 decrease as time went on. 



The expressed fluid was filtered before the solids were estimated j 

 this was found to lower the amount of solid to a very slight extent. 



