342 ANNUAL REPORT 
best example of catalytic action we have in the addition of gypsum to the 
ground clinker. 
It has been found by experience that hard water. tends to retard the 
setting, which also applies to sea water. ; 
In general it may be said that slow setting cements result as a rule 
in stronger and more consistent cements than quick setting cements. The 
“same statement also applies to the hardening. Cements which assume 
their maximum hardness in a short time are liable to go back in their 
strength, as has been shown by many tensile and crushing tests. 
TENSILE AND CRUSHING STRENGTH. 
In use, cement is largely subjected to crushing strains, and hence 
crushing tests would seem the legitimate mode of testing cement, but 
owing to the expensive apparatus required in making these tests, the 
tensile strength of a cement is normally taken as the criterion of its 
strength. Roughly speaking, the crushing strength is from Io to 12 
times the value of the tensile strength. 
The tensile strength test is carried out by filling brass molds of 
the well-known figure 8 shape with a paste consisting of the cement to 
be tested, either mixed with sand of certain definite quality. The 
molds are filled either by hand or by means of a machine. The briquettes 
thus made are kept in moist air for 12 to 24 hours, when they are im- 
mersed in water, in which they are kept till it 1s desired to test them, 
usually from 7 to 28 days. New cements in regard to which but little 
is known are tested frequently for longer periods—a year or more. 
Testing Machine.—The briquette, whose area at the narrowest point 
is one square inch, after hardening, is put into the tensile strength testing 
machine and broken at the weakest point. In American practice the type 
of machine, like the Fairbanks, Fig 65, in which shot is allowed to flow 
into a balanced bucket until the pressure, multiplied by levers, is sufficient 
to break the test piece commonly employed. The flow of shot is stopped 
automatically on the breaking of the briquette, and by weighing the 
charge of shot run in the pressure on the briquette may be read off. 
Other machines like the Riehle depend on the gradual movement of a 
weight towards the end of a lever till the briquette is broken when the 
weight has arrived at a certain point. Again other machines employ 
hydraulic pressure, but the one most commonly employed is the Fairbanks 
type. 
In reporting results of tensile strength tests all values of poorly frac- 
tured briquettes should be rejected. In practice the tensile strength results 
vary as much as 18 to 20 per cent above or below the average and are 
much less consistent than the crushing strength tests. The effect of the 
kind of sand upon the tensile strength 1s exceedingly marked, and varia- 
tions in the results are greater than in the crushing tests, though, of 
