RECENT RESEARCHES IN THE TESTING OF CEMENT. 259 



Chemical analysis may show the several constituents to be in 

 the proper proportion, but nevertheless not properly combined to 

 form a true Portland cement ; and the accidents which have 

 occurred through the failure of concrete, especially in sea water, 

 have generally been traced, not to an improper choice of material, 

 b)ut either to over, or under burning. It is therefore of the greatest 

 importance in large engineering works, where cement concrete 

 occupies such a prominent position, to ascertain before hand, 

 whether the cement is sufficiently reliable to withstand the strain 

 it is calculated for, and that a reasonable time will not change its 

 nature to such a degree as to render it dangerous. 



The usual methods of testing cement are well known, but there 

 are a few points concerning the behaviour of cement while under 

 test, which are important to users of cement. In order always to 

 obtain a uniform crushing and tensile strain, entirely independent 

 of the operator, there are two points necessary, first the percentage 

 of water used for making the cement briquettes should in every 

 case be exactly the same for the same class of cement, and 

 secondly, the pressure exerted in filling the moulds should be the 

 same in every case. The practice of guessing the quantity of 

 ■water, which varies for each quality of cement, has been universally 

 practised, and consequently the operator could, according to his 

 own ideas, or inclination, vary the results of the crushing and 

 tensile strength very considerably, and in so doing, good cement 

 might be condemned, and inferior cement could be made to stand 

 the required tests of strength. The pressure exerted in filling the 

 moulds, as well as the percentage of water, has also generally been 

 left to the operator to decide, and it affects the strength no less 

 than the percentage of water used. 



The amount of pressure exerted may be regulated approximately 

 T^y having the moulds kept together with springs, which open out 

 with too great a pressure, and if the percentage of water used is 

 ^correct, the greatest pressure the springs will stand without open- 

 ing out, must be exerted before the water will come to the surface ; 

 therefore the only difficulty to overcome was to discover definitely 



