PROPERTIES OF PORTLAND CEMENT. PART III. 225 



Although the normal consistency pastes of these cements set very 

 slowly, the cements themselves are extremely quick-setting; so much so, 

 in fact, that they set as soon as they come in contact with the water. 

 To prove this we packed the materials in beakers and then covered them 

 with water. When this was done they set so quickly as to be capable of 

 supporting the set needle almost immediately. However, the degree 

 of hardness attained by the wetter, undisturbed cement was so slight 

 that no difficulty was experienced in troweling or kneading the set 

 mortar thus formed into a plastic paste. 



In other words, it was found that these non-plastered, non-seasoned cements 

 set so quickly that they became regauged (tempered) by any process of manipu- 

 lating them which disturbed the relative positions of their particles after water 

 had been added. Consequently, the use of such cement in ordinary construction 

 work would result in extremely slow setting concrete, and the setting properties 

 shown by Table XXIV and figure 16 are those of freshly set, regauged cements. 

 The work of S. Kasai " illustrates this fact, since it shows that freshly set 

 regauged Portland cements harden and soften in the erratic manner characteristic 

 of the set curve recorded in figure 16. 



Another characteristic phenomenon which occurred when these cements 

 were gauged with water, was the almost immediate generation of a con- 

 siderable quantity of heat. For instance, 100 grams of Oj cement, 

 added to 100 cubic centimeters of water in a 250 cubic centimeter 

 asbestos-covered flask, gave a rise in temperature of 8°. 2 in 4 minutes. 

 Both the soundness and microscopic test of these cements showed that 

 they contained free, unslaked lime. Therefore, it might be assumed 

 that the early generation of heat was due entirely to the hydration of 

 the free lime. However, it required about 7.0 grams of freshly ignited 

 lime to produce the same heating effect as the 100 grams of 0^ cement. 

 We doubted the existence of 7 per cent of uncombined calcium oxide 

 in this material and therefore attributed part of the heat generated to 

 the setting of the cement. Later developments proved this conclusion 

 to be correct. 



THE EFFECT OF PLASTER ON NON-SEASONED PORTLAND CEMENT. 



Having ascertained these facts about the cements in their natural, 

 non-seasoned state, we next investigated the effects produced by adding 

 various amounts of plaster of Paris. The use of either anhydrous 

 calcium sulphate (CaS04) or gypsmn (CaSO^-^HjO) in an equally 

 fine state of subdivision would have produced the same results, provided 

 that weight for weight correspondingly greater or smaller quantities had 

 been employed according to the amount of water which they contained. 



Generally speaking, the sulphate of calcium when added in quantities up to 

 2 to 3 per cent retards the natural set of Portland cement. • Used in larger 



"Das Abbinden der Portlandzemente. Onoda, Japan (1908). 



