PHYSICAL AND CHEMICAL PROPERTIES. PARTS IV, V. 167 
These bars were submerged in running water and changes 
in their length were detected and measured with a Bauschinger 
apparatus. The results obtained are plotted in fig. 26. 
The diagram shows that it took from three to seven months’ 
intmersion in water to slake the hard-burned free lime in these 
commercial cements. The free lime in the cement clinker man- 
ufactured by Campbell and White existed in particles so 
coarse that they remained on the 20-mesh sieve, and the sub- 
sequent regrinding of this clinker to a fineness passing the 
100-mesh sieve left all of the free lime in a very fine and un- 
protected state. In commercial cements the free lime is less 
finely ground and more or less protected by coatings of slag 
which accounts in a measure for its slower hydration. Sample 
4 which was reground to pass the 100-mesh sieve, showed 
no decrease in volume until after seven months thus proving 
that the free lime in the hard-burned commercial product slaked 
much less readily than that prepared in the small experimental 
kiln. The slower slaking of the free lime in commercial prod- 
ucts is readily accounted for by the longer time during which 
the raw-meal is confined in the clinking zone of the kilns. 
These results show our general experience that abnormal 
expansion, like the free lime on which it depends, is not elimi- 
nated entirely by a prolonged aération of either the ground 
cement or clinker. 
Expansion bars made of 1:3 Ottawa sand mortars changed 
in volume more than was indicated by observations on the neat 
mortars. Usually the sand mortars changed very little in volume 
during the early period of hardening, the tendency being to 
contract unless considerable free lime was present; but after 
two or three months, expansion took place, due probably to the 
filling of the voids in the sand mortar. After six or seven 
months the characteristic contraction was very well defined. 
Usually, also, the total change in the length of the bars of sand 
mortar was at least half that of neat mortar. An instance is 
given in fig. 27 in which the change in volume was even more 
marked in the sand than in the neat mortar. 
The irregularities in the curves of expansion of cement mor- 
tars can be attributed to differences in the physical condition 
of the particles of free lime and to the impermeable nature 
of indurated neat mortar. The much greater permeability of 
a 1:3 sand mortar permits the hydration to progress more uni- 
formly than in neat mortars, and, therefore, the results of hydra- 
tion are more clearly defined in the sand mortars. For the same 
reason aération tends to produce more sharply designated 
changes in volume. 
1122323 
