1871.] Mercurial Colloids and certain Fatty Ketones. 433 



dicate that the chief compound produced in the reaction above referred to 

 might be regarded as a colloid body. I therefore took advantage of the 

 late Professor Graham's beautiful dialytic method* for effecting its puri- 

 fication from crystalloids, and have met with complete success. 



a. Preparation of Colloid Liquid. — As the preparation of a strong 

 solution of the pure acetone mercuric compound suitable for dialysis is 

 attended with some difficulty, I may now describe in detail the mode of 

 operating proved by experience to afford the most satisfactory results. 



Forty grammes of pure mercuric chloride are to be dissolved in about 

 500 cub. centims. of hot water and the solution then allowed to cool, even 

 though crystals of the salt separate. Twenty-nine grammes of potassium 

 hydrate are next dissolved in about 300 cub. centims. of water : 15-20 cub. 

 centims. of acetone should now be placed in a capacious glass balloon, and 

 diluted with 250 cub. centims. of water. The reaction is then to be 

 managed as follows : — About 150 cub. centims. of the alkaline solution 

 should be added to the aqueous acetone, and then 250 cub. centims. of 

 the mercuric chloride gradually poured in. Re-solution of the mercuric 

 oxide first thrown down proceeds slowly at the outset, if the mixture be 

 not warmed. After a time the oxide quickly redissolves, if the contents 

 of the balloon are briskly agitated. When the first half of the mercuric 

 solution has been added, the remaining 150 cub. centims. of potassium 

 hydrate are to be cautiously poured in and the residual mercuric chloride 

 then mixed, with the precautions already stated. 



The solution prepared in the manner described is usually turbid, but 

 can be easily filtered clear from the small amount of mechanically sus- 

 pended matter. The filtrate should next be placed on a large hoop 

 dialyzer, covered as usual with carefully prepared parchment-paper, and 

 the vessel floated on a considerable volume of distilled water. After two 

 days' action the diffusate will be found to contain a large quantity of potas- 

 sium chloride, some potassium hydrate, and but a very small amount of 

 mercury. The process of diffusion is to be continued, the diffusate being 

 replaced by pure water twice each day, until the liquid on which the 

 dialyzer floats no longer affords a cloud when treated with a solution of 

 silver nitrate. The process may then be considered terminated, and the 

 pure colloidal liquid obtained. The contents of the dialyzer can be now 

 removed, and should be free from all odour of acetone. A few drops, when 

 evaporated to dryness on platinum-foil and the residue ignited, should 

 volatilize completely. 



The mode of operating just described affords the strongest colloidal 

 liquid that can be conveniently prepared directly in a pure state ; but 

 where degree of concentration is of no importance, I find that it is better 

 to dilute the alkaline mercurial solution with its own volume of pure water 

 just before dialyzing. 



The properties of the colloidal liquid will be described most conveniently 

 * " Liquid Diffusion applied to Analysis," Philosophical Transactions, 1861. 



