COLLOIDS 235 



Master of the Mint) we owe the first and very important researches 

 concerning the diffusion of dissolved substances. The first 

 memoir on the subject is the Bakerian Lecture to the Koyal 

 Society for 1849. In the course of these experiments Graham 

 discovered not only that dissolved salts and other compounds 

 diffuse, that is they move from place to place in the liquid, with 

 very various degrees of rapidity, but that the same compounds 

 are divisible into two main classes according as they possess or 

 do not possess the power of passing when in solution through 

 animal membranes or parchment paper. These two classes 

 Graham called " crystalloids " and " colloids." The former in- 

 cludes not only substances like salt and sugar which are capable 

 of crystallising, but compounds such as hydrogen chloride (hydro- 

 chloric acid) which are not known in the crystalline state. These 

 substances diffuse through membranes with various degrees of 

 rapidity. The other class, colloids, are typified by gelatine or 

 glue (/coXXa glue). They are substances which swell [up in- 

 definitely when soaked in water, they show no sign of crystallisa- 

 tion, but when dry break with a conchoidal fracture like resin or 

 glass. They are known to have a large molecular weight, and 

 they pass, when in quasi-solution, very slowly through mem- 

 branes. Many colloids, however, are not of organic origin like 

 glue, but are obtained by a great variety of processes from salts, 

 oxides, and even metals, which under ordinary conditions are 

 practically insoluble in water. An example or two will make the 

 matter clear. Silica Si0 2 is known in the form of rock-crystal, 

 quartz rock, agate, carnelian, opal, cairngorm, flint, and other 

 stones, sometimes nearly pure as in rock crystal, more usually 

 coloured by the presence of small quantities of ferric and other 

 oxides. These agree in being practically insoluble in water, but 

 if melted with caustic soda or merely boiled for some hours with 

 a solution of the same the silica is rendered soluble, and a moder- 

 ately strong solution of the product in water constitutes the 

 " water-glass " of the shops. If a diluted solution of water-glass 

 is poured into an excess of dilute hydrochloric acid no precipi- 

 tate is formed and the solution remains clear. It contains 

 silicic acid, a colloid, together with the sodium chloride which 

 has been formed and the excess of hydrochloric acid, both 

 crystalloids. By pouring the solution into a bag of parchment 

 paper and suspending it in water the latter substances diffuse 

 away, and if fresh water is supplied several times in place of the 



