270 SCIENCE PROGRESS 



by adding to the reaction medium some neutral substance, 

 which increases its viscosity while taking no part in the chemical 

 change. This progresses more slowly than before, and, by 

 reason of the internal friction, the Brownian movement of the 

 resulting particles is diminished, so that they more readily form 

 centres of crystallisation, while their smaller sphere of movement 

 limits them to a lesser size. A third method for the prepara- 

 tion of sols consists in the employment of reacting solutions of 

 a high degree of aggregation. The greater the degree of aggrega- 

 tion, the less the Brownian movement, and the smaller also the 

 number of particles which will have to combine together to 

 form the slow-moving centres of crystallisation, which will be 

 proportionately increased in number. Thus a sol of the 

 metal barium, prepared by Bredig's method, may be treated 

 with dilute sulphuric acid with the formation of a sol of BaS0 4 

 and escape of hydrogen. 



Aluminium hydroxide, as it usually occurs, is a typical 

 representative of the colloid state. This is due to its great 

 insolubility and to the production of associated basic salts as 

 intermediate products during its formation. Small macro- 

 scopic crystals of this substance may be obtained by the slow 

 union of extraordinarily dilute reagents, or by artificially in- 

 creasing the solubility. The former method has long been 

 known to mineralogists. In the case of a difficultly soluble 

 substance like barium sulphate, of which the solubility is about 

 io -4 , we have seen that macro-crystals are produced from 

 the mixing of two litres of very dilute solutions after some 

 years. In the case of a body whose solubility is only about 

 io -10 , the number of crystallisation centres will be so much 

 greater that there will be insufficient nutrient material for 

 them to grow into crystals. So that, to obtain crystals, very 

 large quantities of extraordinarily dilute solutions must be 

 used. When however micro-crystals have been obtained, if they 

 are put into a super-saturated solution obtained in any way 

 they will grow to macroscopic size. This is the basis of the 

 first method. The second method was employed by v. Wei- 

 marn, who made use of the greater solubility of aluminium 

 hydroxide at the boiling point. Half a litre of N/2,500 A1C1 3 . 

 6H 2 was poured into the same volume of N/2,500 NH 4 OH 

 solution, and the mixture boiled until an opalescence was 

 clearly visible. The heating was then discontinued and the 



