PROPERTIES OF PORTLAND CEMENT. PART III. 237 



it is entirely regauged during the process of mixing, the heat generated 

 is so great that the tester becomes aware of the rise in the temperature 

 of the paste even though liis hands are protected by rubber gloves. The 

 amount of water required and the stiffness of the paste are also abnormal. 

 Consequently, it seldom occurs that a cement is entirely regauged 

 without the fact being noted. On the other hand, when the cement is 

 such that regauging occurs only to a small extent, this fact is not made 

 apparent because such cements may heat but slightly and require less 

 than 24 per cent of water, whereas some cement pastes which set normally 

 require more than this amount and also heat considerably during the 

 process of mixing. Therefore, although the contents of this paper have 

 made it possible to identify partial regauging, cement testers seldom 

 notice or record the phenomenon when it only occurs to a slight extent. 

 In fact, no attention has heretofore been given to the subject. As a 

 result, the setting properties of many cements are reported upon favorably, 

 whereas if the true facts were known, the report would be very un- 

 favorable. 



THE CONTROL OF THE SETTING PEOPBKTIES OF POETLAND CEMENT. 



The primary object of the preceding work was to find a means 

 whereby the changes in the compounds in Portland cement might be so 

 controlled that serious alterations in the rate of setting could not occur. 



Fortunately, it is not essential that the rate of setting remain absolutely 

 fixed and unchanged in order successfully to solve this problem. In 

 ordinary construction work, changes in the set are not of serious im- 

 portance, unless the cement becomes abnormally quick or slow setting. 

 When an abnormally quick setting cement is desired and specified for 

 tidal, or other special purposes, it is only essential that the cement 

 remain abnormally quick setting. Therefore, the set may change, but 

 in so doing it must constantly remain within specified, desired limits. 

 An analysis of the cement before it is packed, such as is outlined by our 

 treatment of the selected material, will not only inform the manu- 

 facturer if the product of his kiln is capable of being put under practical 

 control, but also fix the minimum amount of retarder required to do so. 



Cement 0; can be taken as a specific example. We found that when the free 

 lime was all present as calcium oxide, it required 1.0 per cent of plaster to 

 prevent regauging and 1.5 per cent to produce a normal initial set (not less than 

 30 minutes ) . After the cement had seasoned until the microscopic test showed 

 only slaked lime, it required but 0.5 per cent of plaster to prevent regauging and 

 produce a normal set. (See Table XXIX.) 



Therefore, it would be predicted of this cement that if less than 1.0 per cent 

 of plaster is added it could season so that four radical changes would occur in 

 its setting properties, and that the set at times would become abnormally quick. 

 Two changes could be brought about by seasoning the cement so that the free 

 lime would gradually slake. The first effect of this would be to eliminate the 

 regauging. As soon as this was accomplished an abnormally quick-setting paste 



