236 ANNUAL REPORT 
The cements were allowed to cure for four weeks, but no gypsum was 
added. This, of course, made the high alumina cements quite quick 
setting. Five briquettes wcre made of each cement, I cement: 3 standard 
sand. Two boilirg tests, 6 hours’ boiling, were also made of each 
cement. Beginning with the formula (2.6CaO) S10, (2CaO)AI,O3;, dust- 
ing was observed, and very low and unsatisfactory tests were obtained 
both as regards tensile strength and boiling tests. For this reason these 
have been omitted from the table. 
In considering the results of the tests it must be remembered that 
cements with a silicious clay base harden much more slowly than alumi- 
nous cements, but attain after six months very high strengths, though 
showing up less promisingly in the short time tests. This was found 
to be true in a number of other experiments. 
Conclusions.—1I. [rom these results we can make the assertion that 
for dry ground mixtures the formula (2.8CaO) SiO,, (2CaO) Al,O, is the 
safest. This would correspond to the proportion, in parts by weight, of 1 
part of silica to 2.61-parts of lime, or 1 part of silica to 4.66 parts of 
calcium carbonate. A cement made from a silicious clay and limestone 
would be decidedly unsafe with the theoretical tri-calcium silicate for- 
mula. From the results, however, no deductions can be drawn with 
reference to the theoretical silicates present in the cement. But the 
work seems to support the views of Vogt that we are dealing with a slag 
whose limits of composition are not as narrow as we might suppose. 
But at the same time there are quite definite limits, for with a formula 
of (2.6CaO)S10,, (2CaO) Al,O,, hydraulicity, as we know it in Portland 
cements, practically ceased and “dusting” began. This means that at 
this point hydraulic silicates can not crystallize out any longer. No coal 
ash was added to these cements at all, since they were not in contact 
with coal. The objection might be made that the formula would give 
too low a lime content owing to the introduction of silica and alumina 
in the ash from the coal when burnt in the rotary kiln. There is some 
justice in this objection, as the tendency of the ash will be to lower 
the lime content, and undoubtedly some reaction takes place. 
Yet this naturally depends on the amount and composition of the 
ash and cannot be allowed for offhand, but must be regulated for each 
individual case. But the effect of the ash according to testimony col- 
lected from cement chemists is much less than is frequently supposed. 
That the cement composition is distorted by the presence of ash is, of 
course, obvious, and hence the usual analysis of a cement does not show 
its true composition. This can only be obtained by the ignition of the 
raw mixture without contact with coal. The fixing of the above prac- 
tical formula does not take away the value of the’ Chatelier- Newberry 
formula as a limiting formula. 
