PROPERTIES OF PORTLAND CEMENT. PART III. 249 



First, part of the clinker was ground while hot. If mixed immediately with 

 water it set in 2 hours, but upon standing 3 days it turned to a quick-setting 

 cement. Second, part of the clinker was cooled before grinding. This also turned 

 to a quick-setting material after aging 3 days, although not so quick as the 

 first. The third sample was soaked while hot in water for 3 hours and then after 

 drying 3 days, was ground. The resulting cement was slow setting and upon 

 aging 3 days more lengthened its time of set very much. These samples were 

 all ground to pass a 100-mesh sieve and to each 2 per cent of plaster was added. 



Conover attributes the slow-setting properties of the quenched clinker to the 

 fact that solidification was brought about so quickly that the excess of aluminate 

 was prevented from separating. 



His results, far from indicating that the more quickly the clinker 

 cooled, the more slowly the cement set, only offer additional corroboratory 

 evidence to the conclusions arrived at by our own experiments. Similar 

 results have been obtained at this laboratory merely by treating seasoned 

 and non-seasoned parts of the same ground cement with a fixed amount 

 of plaster. Conover's experiment should be repeated and the cooling of 

 the clinker should be done in such a manner that alterations in the 

 condition of the free lime would either occur to the same extent in all 

 samples or not occur at all. 



In this connection it may also be stated that H. K. Bamber,*^ after an 

 investigation both on a manufacturing and laboratory scale, failed to obtain other 

 than indefinite results of a negative charcter. However, he found that by grind- 

 ing the clinker at a temperature of about 92° in the presence of a limited amount 

 of live steam, the cement operated upon was made to take up uniformly through- 

 out a small amount of water (about 1.0 per cent) which could not be expelled 

 except by ignition; that this treatment had an enormous influence on the action 

 of the cement when tested for soundness; that the proportion of gypsum required 

 to be added to produce a slow-setting cement was thereby much reduced; and 

 that by this treatment in conjunction with even small percentages of gypsum, 

 the slowing effect of the combination was maintained. 



One per cent of water is capable of hydrating almost 3.0 per cent of 

 lime, and although Bamber failed to give any definite chemical or phys- 

 ical explanation as to the cause of these effects, we do not hesitate to 

 attribute them to the hydration of the lime, as we always obtained similar 

 results by any method of seasoning cement which would convert the cal- 

 cium oxide into its hydrate. 



In Bambei''s method the high temperature at which the reaction takes 

 place induces the free lime to slake very quickly and little, if any, cal- 

 cium carbonate is formed. Consequently, his method is more efficient 

 than the ordinary process of aeration, and therefore the steamed cement 

 maintains its slow-setting properties longer than if the hydration had 

 been brought about by aeration. It is not known whether or not Bam- 

 ber's process of hydrating the free lime by steam damages to any extent 



"-Concrete and Const. Eng. (190), 4, 196. 



