142 



colloidal, and that this essential colloid has not been identified. In 

 so far as it does exist, it appears to be (a) a colloidal alumino- silicic 

 acid Avidely distributed through a mass of inert granular material of 

 the same chemical composition or even of an entirely different one, 

 such as sand, the whole being comparable to a freshly-made concrete, 

 but differing from the latter in requiring heat to " set " it. Alter- 

 natively, (6) the colloidal material may be a mixture of colloidal 

 silica and colloidal alumina precipitated simultaneously from the 

 sol state in such a manner as to appear to be a definite chemical 

 compound. This possibihty can only be confirmed or disproved by 

 a large amount of experimental work which is not yet completed. 

 The chief difficulty in accepting the alternative hj^othesis is that, 

 if it were correct, it should be possible to isolate relatively large 

 quantities of colloidal siHca and alumina from such highly plastic 

 and relatively pure clays as the ball clays, but this has not been 

 accomplished. This may be rendered difficult or impossible by the 

 mutually coagulated silica and alumma gels behaving as a compound 

 in which the silica and alumina have so great an affinity for each 

 other that they cannot be separated by means applicable to the 

 isolation of the simpler gels. Finally, (c) as almost any material may, 

 by suitable treatment, be converted into the colloidal state, the 

 characteristic properties of plastic clays may be independent of the 

 chemical comijosition of the colloidal matter present in them. If this 

 were the case, any mineral substance which could be converted into 

 colloids by the natural agencies to which clays had been subject 

 would be possible sources of clay. This explanation has the great 

 advantage of explaining the small projjortion of colloidal matter 

 present in even the most j)lastic clays, as if such colloidal matter 

 were the result of age-long grinding of minute rock particles under 

 water, it is only natural to suppose that the product of such action 

 would be grains of the original rock surrounded by a film of colloidal 

 material. If, on the contrary, clays are produced by mixing colloidal 

 matter (formed separately) with non-j)lastic grains, it is most Hkely 

 that considerable quantities of such wholly colloidal matter would 

 be found in small pockets or fissures in clay beds. This does not 

 appear to be" the case. 



If any or all of these three alternatives were correct they would 

 explain many of the known proi^erties of clays. Both silica and 

 alumina gels readily become irreversible ; even when prepared under 

 the most favourable conditions they are much less " manageable " 

 than many other colloids. Hence, it is only to be expected, that if 

 a complex gel containing both silica and alumina in intimate admixture 

 or even in a state of combination would be extremely difficult to 

 isolate in an approximately pure state. 



The conditions, under which clays are formed from felsjDar and 

 other silicates, are so varied, that they do not throw much light on 

 the nature of clays. China clays in Cornwall appear to have been 

 formed by the action of water and acid vapours on granite at a high 

 temperature, the felspar present being decomposed into soluble 

 potassium silicate which has, presumably, been removed in solution, 

 insoluble quartz and clay, the clay being readily separated from other 



