143 



detritus by a slowly moving stream of ■\\'ater. This explanation — 

 though widely accepted — is by no means satisfactory, and if it is 

 assumed to account sufficiently for the low degree of plasticity of 

 china clay it does not explain the high plasticity of Devonshire ball 

 clays not many miles away, unless the latter are presumed to 

 have no connection with the Cornish clays. Rohland^, on the 

 contrary, attributed the low plasticity of some kaolins to the colloidal 

 clay having been largely removed in the sol state. 



So far as can be ascertained at present, the colloidal material to 

 which clays appear to owe their characteristic properties has been 

 produced by the very prolonged action of water on rocks, chiefly 

 those composed of one or more alumino-silicic acids, the precise 

 nature and origin of which is still uncertain, though rej)resented 

 roughly by the formula H4Al2Si209 in the case of Cornish china clays, 

 Dorset and Devonshire ball clays, and possibly of other less pure 

 clays. This colloidal constituent of clays appears to be a very finely 

 divided, solid cellular substance* which can absorb water like a sponge 

 and thereby form a kind of jelly which retains the water by capillary 

 attraction and only permits it to evaporate very slowly at the ordinary 

 atmospheric temperature. Consequently, the projiortion of water 

 present in a clay paste of given consistency is a rough measure of the 

 colloidal matter present and when precisely similar clays are compared, 

 it may also be a measure of the plasticity of the paste ; the latter 

 property is partly due, however, to the size of the grains coated by 

 the colloidal material and by the thickness and other ph3\sical 

 characteristics of the colloidal coating and the extent to which it 

 penetrates any pores or interstices in the granular material. If such 

 a colloidal material were isolated, it would apparently be peptised by 

 a dilute solution of alkali, or by lime water, and recoagulated by strong 

 acids. On heating, it would first contract greatly and simultaneously 

 l^art with a considerable part of its absorbed water. On further 

 heating up to 500° C. or above, it would be decomposed with the 

 evolution of water and the formation of an irreversible material 

 largely colloidal in character, though different in many ways from 

 most well-known colloids. On further heating,it might undergo other 

 changes, the nature of which can only be summarised as including 

 polymerisation, but one product which under favourable conditions 

 may be expected to be formed is crystalline sillimanite, AljOgSi Oo. 

 When heated with caustic alkaUes, bases, and most metallic oxides, 

 the colloid would probably form mixtures of the corresponding silicates 

 and aluminates. The colloid alone would probably be highly resistant 

 to heat, but in the presence of a metallic oxide (other than alumina) 

 it would fuse more readily. 



It is comparatively easy to obtain colloidal sUica from calcined, 

 and therefore irreversible, or from the crystalline forms of silica, 

 but the process is so drastic (including the fusion of the material with 

 sodium carbonate) that it appears inapplicable to clay, as the latter 

 is completely decomposed during the heating with the flux and the 

 final colloidal product is merely a mixture of colloidal alumma and 



* It need not, of course, consist cf any one chemical compound; even if it 

 consisted chiefly of one such compound, it would seldom, if ever, be pure. 



