BY R. GREIG SMITH. 115 



neutralised the solution with decinormal alkali. Two grms. of 

 the unsifted debris (taken from a depth of four inches) were 

 introduced into the flask containing the solution, and after being 

 shaken at intervals during a day, the solution was filtered and 

 titrated with decinormal acid. The determination showed that 

 there was 1 4% of free lime (CaO) in the debris. 



With regard to the composition of hardened cement there are 

 several theories. It is not a substance of constant composition, 

 and as the percentages of the constituents differ in every l^rand 

 only those that actively play a part in the hardening need be 

 considered. According to Le Chatelier, hardened cement consists 

 of hexagonal plates of crystallised calcium hydrate imbedded in 

 a white mass of interlacing needle-shaped crystals of hj^drated 

 calcium monosilicate. Michaeli considers that the hardening is 

 caused by the formation of a hydrated basic calcium silicate 

 through the combination of free hydrated silicic acid with free 

 calcium hydroxide. Of more recent date are the Newberrys' 

 researches upon the essential constituents of Portland cement 

 which they find to be such that form on the addition of water, 

 tricalcium silicate and varying proportions of dicalcium aluminate. 



It is evident that lime in the hardened cement, capable of 

 being dissolved, is a source of weakness, and it will sooner or 

 later be dissolved by the water. It can, therefore, hardly be 

 doubted that a cement which, after hardening, has become disin- 

 tegrated and porous after exposure to water for a number of 

 years, and which still contains 1 -4% of free lime capable of being 

 dissolved, was not of a nature likely to withstand the action of 

 water. 



In a paper published three years before that already quoted, 

 Stutzer"^ ascribed the disintegration pf the cement to the solvent 

 action of carbon dioxide dissolved in the water. Basinsf his 

 opinions upon the observations of Michaeli, he considered that 

 the addition of an active form of silica such as trass to the 



* Stutzer, Zeit. fiir angew. Cheraie, 1896 (11) 317; Abstract in Jour. Soc 

 Chem. Industry, xv. 595. 



