EECENT RESEARCHES IN THE TESTING OF CEMENT. 261 



This effect varies greatly according to the time of setting and 

 temperature of the air, and in some instances may show a greater 

 strength after three days than after seven days. 



When neat cement is dried there is in nearly every case a'strain 

 exerted due to an internal expansion, with cement mortar in the 

 proportion of three to one, the reverse happens ; the voids between 

 the grains of sand, being filled with the expansible matters, the 

 mass becomes more solid, and the strength is increased. Taking 

 an average of many good samples of cement, and by taking the 

 briquettes out of water for twenty-four hours, and drying them 

 thoroughly before breaking on the twenty-eight day, the neat 

 cement has about two-thirds the strength, and the mortar three 

 to one, about one-third more strength than if kept continuously 

 under water. 



Another important point in the testing of cement is the quality 

 of sand used. The greatest care is necessary to maintain a uniform 

 standard from year to year. The strength of cement when mixed 

 with sand is affected greatly by the size and shape of the grains ; 

 the dosage, or amount of cement required to fill in between the 

 grains, and also (in some instances where the sand used is not 

 pure quartz) by the strength of the sand itself. The usual method 

 of preparing standard sand is to sift it through a sieve of four 

 hundred meshes to the square inch, and then through one of nine 

 hundred meshes, retaining that left on the nine hundred sieve as 

 standard, but even then the strength obtained will vary according 

 to the quality of the sand, even when taken from the same quarry. 

 Diagram 3 represents the variations which occur ; No. 1 is the 

 coefficient of strength of standard sand obtained from a Pyrmont 

 quarry ; No. 2 from Arncliffe ; No. 3 from the same quarry at 

 Pyrmont, but of a softer quality. No. 1 is 15% over and No. 3 

 10% under the proper standard. 



The variation in strength obtained by using different kinds of 

 sand, in concrete work, is necessarily very much greater ; in 

 Diagram 3 is represented the coefficient of strength of various 



