Providing that wet clay is not heated to a temperature higher than that of boiling water it appears to undergo no chemical change and on cooling it will again take up water[15] and be restored to its original condition except in so far as its colloidal nature may have been affected by the heating. If, however, the temperature is raised to about 500° C. a decomposition of the material commences and water is evolved. This water--which is commonly termed 'combined water'--is apparently an essential part of the clay-molecule and when once it has been removed the most important characteristics of the clay are destroyed and cannot be restored. The reactions which occur when clay is heated are complex and are rendered still more difficult to study by the apparent polymerization of the alumina formed. Mellor and Holdcroft (29) have recently investigated the heat reactions of the purest china clay obtainable and confirm Le Chatelier's view (10) that on heating to temperatures above 500° C. clay substance decomposes into free silica, free alumina and water, the two former undergoing a partial re-combination with formation of sillimanite (Al2O3SiO2) if a temperature of 1200° C. is reached. Mellor and Holdcroft point out that there is no critical point of decomposition for clay substance obtained from china clay, as it appears to lose water at all temperatures, though its decomposition proceeds at so slow a rate below 400° C. as to be scarcely appreciable. [Footnote 15: Some clays are highly hygroscopic and absorb moisture readily from the atmosphere. According to Seger (7) this hygroscopicity distinguishes true clay from silt and dust.] After the whole of the 'combined water' has been driven off, if the temperature continues to rise, it is found that at a temperature of 900° C. an evolution of heat occurs. This exothermal point, together with the endothermal one occurring at the temperature at which the decomposition of the clay seems to be most rapid, has been found by Le Chatelier, confirmed by Mellor and Holdcroft, to be characteristic of clay substance derived from kaolin and china clay, and the two last-named investigators state that it serves as a means of distinguishing kaolinite or clayite from other alumino-silicates of similar composition. These thermal reactions have not, as yet, been fully studied in connection with plastic clays; with china clay, as already noted, they probably indicate a polymerization of the alumina set free by the decomposition of the clay substance, as pure alumina from a variety of sources has been found by Mellor and Holdcroft to behave similarly.