220 



cox. 



tity of lime. A mixture wliicli would more nearly correspoiid to the 

 theoretical analysis of Lehigli Portland cement given above is as follows ; 



Table VII. 

 [The numbers give parts of the materials by weight.] 





Total. 



Individual constituents. 



Silica 

 (SiO.). 



Alu- 

 mina 

 (AloOs) 



Iron 

 oxide 



(FeoOs) 



^(S 

 (CaO). '■^^eo). 



1 



Volatile 1 



(CO.., 



H,0, 



etc.). 



Clay (No. 1) 



100 

 292 



60.17 

 l.OS 



22. 65 

 0.5 



4.68 



0.3 1.85 

 162.6 



6.35 

 127.65 



Limestone (No. 7) 



Unburned 



392 

 134 



61.22 



23.15 4.66 



162.9 1.85 



134. 



Volatile 



Burned 



' 1 



258 



23.7 



8.97 



1.80 



63.1 i 0.72 



1 









While the limiting formula is of great assistance in the pro^JOrtioning 

 of cements, it leaves undetermined the silica-alumina ratio, which is 

 of utmost importance. There is every reason to believe that clay number 

 1 would produce a better cement if silica in some shape were added to 

 the clay base. With the substitution of a certain amount of silica for 

 alumina or iron oxide, molecule for molecule, in the limiting formula a 

 safer cement would be produced in that the overburning temperature 

 would be increased.-" The state of subdivision of the free silica in the 

 raw materials used for this jjurpose is an extremely important considera- 

 tion. The sandstones of the region, although high in silica, are probably 

 not suited for this purpose in that they are composed of oxides of iron 

 and alumina and quartz grains so large as to be unsuited to cement manu- 

 facture. 



This sandstone, which contains a25proximately 73 per cent silica, 

 when pulverized to the degree necessary for rock analysis and one gram 

 digested in 200 cubic centimeters of 5 per cent sodium hydroxide for 45 

 minutes on a water bath, allowed 7.6 per cent of the silica to pass into 

 solution.-^ It is unnecessary to resort to physical means to determine 

 that the greater part of the silica is crystalline, for the quartz grains 

 are visible to the naked eye and veiy distinct with a lens. The grains 

 are very loosely held together and can be broken apart Ijy rubbing be- 



-° The ideal mixture can not be overburned. 



-'Quartz is not absolutely insoluble in caustic alkalies (Lunge, G. and Mill- 

 berg, G. : Ztschr. f. angew. Ghem. (1S97), 10, 3fl3, 4-25 ), as is sometimes supposed 

 to be the case and the use of a dilute solution of the latter does not give a correct 

 separation of quartz and amorphous silica, however the error can generally be 

 neglected. Hillebrand {Bull. U. S. Geol. Sur. (1907), 305, lOG), recommends as 

 the most satisfactory the use of a dilute solution of sodium hydrate in which the 

 solution of the amorphus silica is almost immediate. 



