166 CARNEGIE INSTITUTION OF WASHINGTON. 



ards, which was based upon the equilibrium diagram of the system CaO- 

 Al203-Si02, has disproved this as a general contention. Indeed, one must 

 conclude that the nearer the approach to equilibrium the better the resultant 

 cement, judging from the great stress which is now laid on fine grinding of the 

 raw materials and from the lengthening out of the cement kilns — factors which, 

 by securing more intimate contact of the two components and by increasing 

 the period of heating, obviously make for a nearer approach to equilibrium. 

 Just how close an approach to equilibrium (perfect burning) is desirable or 

 economically possible, can not be stated from the data available; these data 

 can only be obtained, for a given raw mix, by investigation. Such an investi- 

 gation would necessitate careful research to determine the relation between 

 time, temperature of burning, and fineness of grinding of given raw materials, 

 in order to produce good clinker at the lowest possible cost to the manufac- 

 turer. 



Actual cement clinker contains small quantities of MgO and iron oxide, 

 which are advantageous in that their presence during burning promotes the 

 attainment of equilibrium by lowering the temperature at which liquid 

 appears (fluxing) and by increasing the amount of such liquid; but they would 

 seem — to judge from the available reliable evidence — to have little influence 

 on the final main constituents of the clinker. This latter statement, beit 

 noted, refers only to the clinker, and implies nothing as to whether such admix- 

 tures have or have not a beneficial influence on the cementing qualities of the 

 product when mixed with water. 



Now that the constitution of the clinker has been definitely established, it 

 is practicable to attack the problem of the hydration of portland cement; 

 indeed, the results of an investigation along this line have already been pub- 

 lished from the Pittsburgh laboratory of the Bureau of Standards. By ascer- 

 taining precisely what happens on the hydration of each possible constituent 

 of the clinker separately and in all the possible permutations and combina- 

 tions, it should finally prove possible to determine the proportions of the 

 various constituents which should be present in the clinker in order to produce 

 a cement which on setting will possess the most desirable qualities, e. g., the 

 greatest strength; and from a study of the equilibrium diagram one can learn 

 just how to proceed in order to produce a clinker of this optimum composition. 

 In conclusion it may be remarked that the effect of admixture of other mate- 

 rials can only be definitely ascertained by the systematic and somewhat 

 laborious procedure described above — a procedure which, nevertheless, will 

 lead sooner to the discovering of the optimum composition in various cases 

 and for various purposes than the empirical or cut-and-try method which 

 hitherto has been the only method employed. 



(18) Die Bestimmung der Dichte von Mineralien und Gesteinen bei hohen Temperaturen. 



Arthur L. Day, R. B. Sosman, und J. C. Hostetter. Neues Jahr. Min. Geol., 

 40, 119-162 (1915). 



A German translation of "The determination of mineral and rock densities 

 at high temperatures" (Am. J. Sci. (4), 37, 1-39, 1914). Reviewed in Year 

 Book No. 12, p. 144. 



(19) The crystallization of haplobasaltic, liaplodioritic, and related magmas. N. L. Bowen. 



Am. J. Sci. (4), 40, 161-185 (1915). 



Mixtures of diopside with various members of the plagioclase series, referred 

 to as haplobasaltic (from the Greek a7rX6o5 = simple), haplodioritic, and so 

 forth, according to the nature of the plagioclase, are studied by the quenching 

 method of thermal analysis. Equilibrium, determined in this manner, is 



