208 REIBLING AND REYES. 



either failing to recognize the indefinite composition of the product, or 

 neglecting entirely properly to ascertain or consider the nature and 

 significance of changes in physical and chemical properties which dif- 

 ferent methods of seasoning induce. A few observations will demon- 

 strate our reasons for this belief and suggest as well certain facts a 

 careful consideration of which, it is thought, would lead to more definite 

 results. 



The fact that the calcareous hydraulic cements do not represent well 

 defined compounds, but mixtures of calcium aluminates, calcium fer- 

 rates and calcium silicates, with varying amounts of uncombined calcium 

 oxide, must be regarded as having been conclusively proved from the 

 laborious and thorough synthetic work of Dr. 0. Schott *° and the 

 analytical studies recorded in Parts I and II *^ of this paper. Further- 

 more, we have shown that in non-seasoned cement uncombined lime may 

 have different physical properties according to the duration and degree 

 of heat to which it has been subjected during the process of burning 

 and that in the seasoned commercial cement the uncombined lime may 

 exist partially or wholly in the sintered, non-sintered, slaked and ciys- 

 talline, and slaked and amorphous states, and combined as inert calcium 

 carbonate. Since synthetic work has shovm that caleimn oxide, hydrox- 

 ide and carbonate, each effect the physical properties of Portland cement 

 in a different manner, it is evident that properly to study the changes 

 produced during the seasoning of Portland cement, it is essential to 

 note those which occur in the condition of the free lime which may be 

 present. 



With these facts in mind we have endeavored to select from the mass 

 of literature on this subject conclusions which appear to be the most 

 reasonable, or which otherwise require considera'tion. 



The exact nature of the phenomena involved and of the compounds 

 causing the setting of Portland cement are still matters more or less 

 of theory. According to Le Chatelier the early hardening of Portland 

 cement is due to a deposit of interlocking crystals of hydrated calcium 

 aluminates separating from a super-saturated solution. He further 

 states that only those varieties of hydrated calcium aluminate which are 

 permanent in the presence of an excess of lime-water need be considered 

 in connection vnth the set of cement. 



Whether we accept this theory or not, we must admit that the set 

 is caused by the presence of certain compounds in Portland cement the 

 quantity, composition and activity of which are not revealed by chem- 



" The Calcium Silicates and Calcium Aluminates contained in Portland Cement. 

 (1906). 



*' This Journal, Sec. A (I9\(y) , 5, 3Q7-418. 



