144 The Philippine Journal of Science 1914 



With the crystalHne limestone, a hydraulic index of 2.04 gave 

 satisfactory results although some difficulty was experienced in 

 obtaining a sufficiently high temperature. However, when the 

 index was lowered to 1.87, the clinker dusted and the cement 

 became extremely quick setting, and when the index was raised to 

 2.17 the mixture became too refractory. 



With the coralline limestones, no difficulty was experienced in 

 obtaining good cement from mixtures having hydraulic indices 

 of 1.90, 2.11, and 2.31, and even the last was easier to sinter 

 than 5a, the least refractory of the mixtures, prepared with the 

 crystalline limestone. 



In this connection, it is only fair to note that results were 

 obtained with limestone 50 and clay 48 (mixtures 1 and la). 

 However, this clay contains 12.05 per cent of iron oxide, 15.04 of 

 alumina, 1.18 of alkalies, and only 52.05 of soluble silica. It is 

 very fusible, and if utilized for the rotary process is apt to cause 

 trouble similar to that described in the following extract of an 

 article by J. G. Dean:" 



* * * the clay used in manufacturing was low in silica and high in 

 iron oxide and alumina. The Silica-Alumina Ratio 



per cent SiOz 



per cent AljOs+per cent FcjOa 



would average a trifle less than 2. 



When pure limestone was used with this clay it was nearly impossible 

 to produce a cement that could be depended on for setting time and tensile 

 strength, and if the lime content of the mix was high enough to burn 

 properly in the kilns the cement would seldom pass the "boiling test." 



In order to overcome this defect in the clay^ it was necessary to mix the 

 limestone from the upper strata of the deposit, which was siliceous in itself 

 and carried in addition the sand and silt that filtered into it, to the purer 

 stone from the bottom of the quarry. By this haphazard method we were 

 able to keep the Silica-Alumina Ratio high enough to produce a high grade 

 cement. 



This method required very careful mixing and grinding, as with cements 

 having a low Silica-Alumina Ratio the limits of variation are extremely 

 narrow. If the lime content is lowered the cement becomes erratic in 

 setting qualities and other peculiarities of over clayed cement. If the lime 

 content is raised the cement will fail on constancy of volume tests and will 

 require "air slaking." 



These peculiarities become very complicated with such materials when 

 they have been properly proportioned but improperly ground before 

 burning. * * * The burner will complain of its being "soft" or over 

 clayed. The boiling test on the cement will reveal "free lime" and the 

 tests for setting time and strength will indicate an excess of clay, while the 



'Chem. Eng. (1909), 10, 52. 



