278 The Twenty-Ninth General Meeting. 



with certain precautions; if, for instance, a dilute solution of so- 

 called " water-glass," which is an alkaline silicate, be poured into 

 an excess of acid ; the silica does not all fall as a jelly, but a certain 

 amount of it remains in the liquid as dissolved silica. It was with 

 silica so dissolved that the late Professor Graham made some of his 

 remarkable observations on diffusion, that is to say, on the greater 

 or less facility with which two liquids can permeate, each the other ; 

 an enquiry founded on the principle that the velocities with which 

 the molecules of one substance will spontaneously interpenetrate 

 between those of another substance, in the process of mutual trans- 

 fusion, are very different. This process plays an important part in 

 the chemistry and physics of Nature, and I must ask your attention 

 to it in so far as it affects the problem we are discussing. If the 

 liquid containing silica separated by an acid, from an alkaline silicate, 

 and retained dissolved in the liquid, be placed in a vessel with a 

 bottom or a side formed of some porous material, such as a bit of 

 parchment paper, and dipped in a beaker of water, the other ingre- 

 dients will pass away through the porous diaphragm in the outward 

 direction, leaving behind the silica dissolved in water, for some of 

 the water will pass inwards into the inner vessel from the beaker. 

 The reason of this is, that substances of the type which Graham 

 termed colloids, move through the pores of such a diaphragm, or 

 indeed along any tube of fine bore, with extraordinary sluggishness, 

 as compared with the facile transit with which other substances, 

 belonging to what he called the crystalloid type, will permeate the 

 diaphragm or traverse the narrow tube. 



The acids and the salts in the solution belong to the latter type, 

 and are thus dialysed more rapidly into the outer beaker than is the 

 case witb the silica solution, which being of the colloid or gelatinising 

 type, is left almost entirely behind. 



The particles of colloid bodies such as gelatinised silica have, 

 furthermore, a remarkable tendency to adhere together, and to shrink 

 to a compact mass ; and they shew a preferential tendency in one 

 colloid body to aggregate to itself the particles of any other colloid 

 body not necessarily of a similar nature with itself, and to 

 cohere with it : while towards a crystalloid body, on the other 



