STATE GEOLOGIST. 209 
The conclusions of these writers are summarized by them as follows: 
“tT, The essential constituents of Portland cement are tri-calcium 
silicate, with varying proportions of di-calcium aluminate. ‘This compo- 
sition may therefore be-expressed by the formula x(3CaO SiO,)—+ 
y(2CaO.Al,O,). From this formula it may be calculated that the cor- 
rect proportion of lime, by weight, in Portland cement is 2.8 times the 
silica plus 1.1 times the alumina. 
“2. Iron oxide combines with lime at a high heat and acts like 
alumina in promoting the combination of silica and lime. For practical 
purposes, however, the presence of iron oxide in a clay need not be 
considered in calculating the proportion of lime required. 
“2. Alkalies, so far as indicated by the behavior of soda, are of 
no value in promoting the combination of lime and silica, and probably 
play no part in the formation of cement. 
“4. Magnesia, though possessing marked hydraulic properties when 
ignited alone, yields no hydraulic products when heated with silica, 
alumina, or clay, and probably plays no part in the formation of cement. 
It is incapable of replacing lime in cement mixtures, the composition 
of which should be calculated on the basis of the lime only, without 
regard to the magnesia present.” 
Professor S. B. Newberry repeated part of these experiments in 
additional work published in 1902* which might be briefly summarized 
as follows: 
He again reports having prepared 3CaOSiO, by mixing the finely 
eround materials dry, and heating the mixture to a white heat. It 
does not, however, harden readily. No direct chemical proof of the 
3CaOSiO, is given. By fusing the mixture over the oxy-hydrogen 
blow pipe the 3CaOSiO, was found to be double-refracting, with rec- 
tangular cleavage, specific gravity 3.022, is constant in volume, hardens 
well. A mixture of calcium carbonate and silica, heated for two hours 
at a red heat, leaves some SiO, uncombined unless at least 21% mole- 
Gules on €a®) to one of Si@s are present. At a white heat the mono, 
di and tri-silicate mixtures become completely combined. The amount 
of water which a given Portland cement will take up chemically is a 
function of the mechanical conditions, porosity, etc. It may be as low 
as 10 or as high as 27 per cent. 
On suspending finely divided calcium silicate in sufficient water to 
dissolve all the lime present, the residues are somewhat indefinite in 
composition, and continue to lose lime on prolonged action; they have, 
however, the following composition: | 
Tri-silicate (not fused), after 29 and 93 days, 3CaO2Si0,3H.O. 
Tri-silicate (fused), 35 and 228 days, 2CaOSiO,H.O. 
The di-silicate (quenched and dusted), did not change in compo- 
sition and did not hydrate. 
*Oement and Engineering News, Nov., 1902. 
14—sS. G. Bull. 3. 
