214 cox. 



analysis by excess of liiiic. The best insults are sbown l)y the eifrlit Itranils that 

 agree closely with the theoretical analysis. 



Mr. Aiken has continued his work and taken samples from several Portland 

 cement mills. He" gives results of the tests of these different brands and says 

 "The results herein given certainly emphasize the fact that much cement is of 

 the best and that there is no monopoly of material or proportioning in manu- 

 facture, as shown by these tests of average output from eight different brands, 

 three of them not Lehigh district material." 



Portland cement was thought by Le Chatelier ' to be composed of tricalcium 

 silicate, 3C'aO . SiO;, and tricalcium aluminate, 3CaO . ALO3. In 1897 S. B. and 

 W. B. Newberry ° showed that this idea should be modified to diealcium aluminate 

 2CaO . AI2O3 since the latter showed superior qualities. Richardson ° and other 

 investigators have concurred in the opinion that tricalcium silicate is an essential 

 constitutent of Portland cement ; recently, however Daj' and Sheperd ' have 

 shown rather conclusively that no such compound as tricalcium silicate exists 

 and therefore the so-called tricalcium silicate is simply a mixtui'e of lime and 

 diealcium silicate. The only rational deductions from these facts are those given 

 by R. K. Meade,' namely: "That the lime in Portland cement clinker exists in 

 three forms: First, in eopibination with silica, alumina, iron, etc., to form a 

 magma of orthosilicates, 2CaO . SiO™, orthoaluminates, 2CaO . AUOa, etc. ; second, 

 as the oxide itself in solid solution in this magma of silicates, aluminates, etc., 

 and third, as undissolved oxide; that is lime merely disseminated through the 

 magma forming with the latter simply a mechanical mixture." The fact 

 remains that under the most ideal conditions of fluxing and burning, a cement 

 will carry an amount of lime very nearly corresponding to the Le Chatelier 

 molecular formula as modified by Newberry and Newberry, viz., x [ (3CaO)SiOi'] 

 + y [ (2CaO) Al^Oj]. Overlimed cements are unsound, but up to this point the 

 more lime that is present the greater will be the strength of the cement. Blein- 

 inger, as mentioned in my previous paper ° has shown from his experiments 

 that "for the dry, ground mixtures the formula (2.8 CaO) SiO. (2CaO) AlA is 

 the safest." Even this give a percentage of lime in the burned product higher 

 than that encountered in many cements. 



Magnesium not to exceed three of four per cent of the finished product 

 may replace calcium in Portland cement without deleterious results, 

 and iron in any amount may replace alumina.^" A certain amount of 

 iron acts as a flux, lowers the fusion point of the mixture and pro- 

 motes the combination of the calcium oxide and silica. Several articles 

 have recently appeared in the literature advocating the superiority of 

 a cement for use in sea water, in which j)i'aetically all of the alumina 



'Concrete Jieview 2, No. 11, 2 to 7. 

 'Ann. des Mines (1887), 11, 345. 

 ''Joitrn. 80c. Ohem. Ind. (1897), 16, 887. 



"Address, Assoc. Portland Cement Manufacturers, Atlantic City, N. Y., June 

 15tli, 1904; .Journ. Soc. Chem. Ind. (1905), 24, 733. 

 '' Journ. Am. Chem. Soc. (190G), 28, 1107. 

 ^Chem. Eng. (1907), 5, 344. 

 'This Journal, Sec. A. (1908), 3, 405. 

 ■"Newberry, S. B. and W. B.: Journ. Soo. Chem. Ind. (1897). 16, 801. 



