COLLOIDS 99 



stand, we find at the end of twenty-four hours that there is 

 formed in the liquid which was first uniformly colored and 

 uniformly fluid, four layers distinctly superimposed one 

 above the other. At the top is a layer of clear, highly 

 colored fluid; below this, two layers darker and more turbid 

 and at the bottom a translucent flocculent precipitate. At 

 the end of two or three days there will be in the tube only 

 a sediment and a perfectly transparent slightly colored 

 liquid. This liquid will, even when saturated with salt, be 

 no longer turbid and will traverse dialyzing membranes. 

 There is thus no more colloid and the hianner in which the 

 precipitate is formed and digested in the remainder of the 

 original fluid shows that there were in this fluid granules 

 of different sizes. 



In a solution exactly disodic there will be at the end of 

 twenty-four hours about 50 per cent, of the product in the 

 precipitate and respectively 25, 20 and 5 per cent, in the 

 layers above it. 



The proportions of amino-acids as free molecules or 

 grouped into more or less voluminous granules may be 

 different for different colloids and for the same colloids 

 according to the conditions of the medium in which they 

 occur. But if we judge by the total of reactions between 

 colloids and cells and by the results obtained in submitting 

 colloids to dialysis, we can see that in serum as in egg-white 

 or in a bacterial albumin, or in a toxin, that these fluids con- 

 tain granules of very different volumes as well as amino- 

 acid existing as free molecules, quite as has been established 

 for the arsenobenzenes. 



Thus in recognizing that there are very distinct differences 

 between the formation of a crystal in which the molecules 

 of salt are mechanically deposited one on the other and a 

 colloidal granule in which the molecules are very probably 

 bound together by a single affinity we must note that there 

 are no clear-cut distinctions between the two sorts of sub- 

 stances when considered from the point of view of their 

 physical property of crossing dialyzing membranes. 



There is no membrane which cannot be traversed by the 

 smallest granules of .every .colloid #ndat^is yery important 



