GEOPHYSICAL LABORATORY. 135 



as nearly as may be with what we may imagine the conditions of natural 

 rock formation to have been. The character of the product will often 

 determine whether the conditions were really those of nature or not, 

 and thus we obtain our first direct clue to the conditions which gov- 

 erned the formation of natural igneous rocks. These simple problems 

 of two minerals in different relations have been found to admit of 

 reasonably complete solutions in a great variety of cases which have 

 been reviewed from time to time in these annual reports and have 

 been published in full detail in appropriate scientific journals. 



Pursuing the same reasoning, the next step of importance was to 

 bring together three of these primary components in all the combina- 

 tions which could reasonably be expected to have significance in rock 

 formation, and to study these under the same variety of (measured) 

 conditions to which the binary mixtures have been subjected. Obvi- 

 ously, this was to be a task of enormously greater magnitude than those 

 which had preceded it, and so indeed it has proved, but it has now been 

 successfully accomplished — the first strictly quantitative study of its 

 kind, so far as we are aware. 



THE MINERAL SYSTEM LIME-ALUMINA-SILICA. 



Speaking technically, this problem is the determination of the tem- 

 perature-concentration equilibria in the three-component system, lime, 

 alumina, silica. It is important from the purely scientific standpoint 

 as a study of the physico-chemical behavior of a complex silicate 

 system at high temperatures in which the results obtained find direct 

 and fundamental application in the investigation of rock magmas and 

 lavas; it is important for the technical world because of its bearing on 

 the constitution of portland-cement clinker. Commercial portland 

 cement is simply this clinker finely ground and moistened with water, 

 after which it sets and hardens in a short time to a rock-like mass. 

 Experience has shown that the qualities of portland cement are directly 

 dependent on the mineral composition and the physical properties of 

 this original clinker. Chemists have determined the variations per- 

 missible in the cheinical composition of the clinker, and cement experts 

 have devised suitable methods of preparing the clinker to obtain a 

 sound commercial cement. Much has also been written on its prob- 

 able mineral composition and on the processes involved in the setting 

 of cement, but the present investigation fixes the mineral composition 

 of the clinker for the first time, and thus furnishes a definite basis for 

 the definition of the factors involved in the more strictly commercial 

 aspects of the problem. 



The present investigation has extended over a period of nearly eight 

 years. When it was started, the possibility of resolving such a complex 

 silicate system was uncertain because suitable methods of attack Vv^ere 

 not available for the purpose. The study was therefore begun with 



