CALCULATION AND CONTROL OF THE MIX. 389 



In experimenting with various methods for analyzing Portland 

 cement, Hillebrand encountered the question of loss of alkalies during 

 burning, which he discusses * as follows: 



"Long before the last of the sulphur trioxide is expelled alkali begins 

 to volatilize, and it is easy to remove all or nearly all in this manner. 

 The alkali is volatilized as oxide and may be collected in quantity on 

 the under side of the crucible lid. At the intense temperature of the 

 rotary-kiln furnace this action must play an important part, and to it is 

 to be attributed the great loss of alkali noted by me in the cement of 

 1901, as compared with the raw mix from which it was made, an obser- 

 vation which is repeated in the present case and must be general in 

 cement-burning." 



Phosphorus. Phosphorus, combined with lime in the form of lime 

 phosphate, frequently occurs in notable percentages in limestones, par- 

 ticularly in the soft, chalky limestones and " marls" of the Southern 

 States. In analyses this will be reported as phosphoric acid or phos- 

 phorus pentoxide (P205), when it is determined at all. Few com- 

 mercial analysts, however, would look for it in a cement material, and 

 it is therefore rarely reported. 



Late in 1903 samples of a "marl" and clay from a Southern State 

 were sent to a leading testing laboratory to obtain a decision on their 

 value as cement materials. Three different burnings of cement were 

 made from the raw materials in various mixtures, and the. resulting 

 cements gave the tests shown in Table 172, below. In addition to these 

 generally poor results the chemists reported that the cement, for a 

 week or so after setting, was so soft that it could be readily rubbed 

 off by the hand. The various defects in the cements were ascribed 

 by the laboratory experts to the presence in the marls of notable per- 

 centages of phosphoric acid. The matter was referred to me by the 

 Southern company, and at my request Prof. Clifford Richardson exam- 

 ined microscopically several thin sections of the clinker which had 

 been made in the laboratory tests. He reported that the raw mix had 

 been very coarsely ground and the clinker underburned. 



The raw materials, as analyzed at the laboratory, showed the results 

 given in Table 169. Two samples of marl were tested and one of clay. 



Of the three samples of cement made up from these materials and 

 tested as below (Table 170), Cements A and B were made by mixing 

 Marl 1 and clay in different proportions, while Cement C was made 

 from a mixture of Marl 2 and the same clay. 



*Journ. Amer. Chem. Soc., vol. 25, p. 1200. 1903. 



