154 REIBLING AND REYES. 
the fact that unsound cements may possibly, through subsequent 
exposure to the air before being used, or because of mixing with 
sand or other aggregate, produce durable masonry, has led to 
a somewhat general belief that failure to pass the tests for 
soundness, especially the hot test, is not a positive proof of 
inferiority. 
On the other hand, further observations and experience have 
strengthened the early, published work of this laboratory, which 
led to the conclusion that perfect soundness was especially 
important to the efficiency of Portland cement used in the 
Tropics. 
THE EFFECTS OF SEASONING. 
Underburned Portland cement.—The data given in Table 
XLII show the general nature of the sede produced from 
decidedly underburned clinker. 
If such material is used before it has seasoned it will act 
like sample 1, in which the disruptive force was sufficient to 
cause the complete disintegration of both neat and sand mortars 
soon after the cement had set. The first step toward a sound 
product is shown by sample 2. Here while the neat mortar 
also disintegrates, the sand mortar remains sound; although, as 
shown by the 3- and 7-day strength values, the expansion is 
sufficient to make its effect upon the strength apparent. The 
further seasoning of the cement soon eliminates this drop in 
strength. The neat mortar undergoes similar changes until 
eventually a product like sample 3 is obtained, which con- 
tinues to harden and to show a slow but steady increase in 
strength for apparently an indefinite length of time. 
The points which we wish to emphasize about soft-burned 
cements are that the clinker seasons very readily, that the hard- 
ening process is slow, and that ordinarily the expansion due to 
free lime occurs very soon after the set. The last is of special 
importance, since the expansion of the free lime in the indurated 
mortar develops so quickly that it is safe to assume the ultimate 
soundness if the neat mortar does not disintegrate after being 
submerged in normal water for the usual twenty-eight days. 
The similarity between such underburned Portland and natural 
or Roman cements is very close. However, the presence of 
harder burned material in the latter makes its behavior less 
certain, for as is shown in Table XLIII the harder burned clinker 
possesses characteristics which are entirely different from those 
just described. 
