CLINKER COOLING, GRINDING, AND STORAGE. 



537 



Nihoul and Dufossez, in the course of the experiments described 

 on page 540, tested the comparative effect of calcium sulphate in four 

 different forms i.e., as crude gypsum, as calcined plaster, as anhydrous 

 plaster and as chemically precipitated calcium sulphate. Their con- 

 clusions were: (1) that with the precipitated calcium sulphate and cal- 

 cined plaster the retardation of set is proportional to the amount of 

 sulphate added: and, (2) that with crude gypsum this is true only 

 when less than 2 per cent of gypsum is employed, larger percentages 

 causing acceleration rather than retardation of set. 



Lewis has carried out a short series of experiments on the influence 

 of calcium sulphate on the strength of the cement, applying the cal- 

 cium sulphate in three different forms gypsum, plaster of Paris, and 

 anhydrous plaster. The results, given in the table below, are not de- 

 cisive: but seem to show a somewhat greater regularity of effect when 

 plaster of Paris or anhydrous plaster are used than when gypsum is 

 employed. 



TABLE 197. 

 [EFFECT OF FORM OF SULPHATE USED. (LEWIS.) 



To summarize the matter: The active retarding agent is the sul- 

 phur trioxide present in the gypsum or plaster. As anhydrous plaster 

 and plaster of Paris both contain somewhat higher percentages of SOs 

 than gypsum, they will exercise a proportionally greater retarding 

 effect, weight for weight, than will gypsum. But for ordinary practice 

 this slight advantage is immensely counterbalanced by the fact that 

 gypsum costs usually less than half as much as either of the plasters: 

 and for ordinary practice, therefore, gypsum is the only form of cal- 

 cium sulphate that can be considered available. In certain plants, 

 however, where the sulphate is added after the cement has been ground, 

 it is necessary to use plaster of Paris; because gypsum as bought is 

 ground too coarsely to add to a finely pulverized cement. 



