As to the sample composed of coarse particles reground, it must be considered that although this sample was passed through the No. 100 sieve, yet it was in reality much coarser than sample C, because the particles were harder, and the grinding in the mortar less thorough than the original grinding. Since this sample of reground cement gives so high a strength neat and with one part sand, it appears that the hard particles from which it was made are of excellent quality if ground fine enough, and the relatively lower results with larger proportions of sand must be attributed to imperfect grinding.

The coarse particles retained between sieves 50 and 100 gave a higher strength neat than was expected, but much of this strength may be due to the floury portion of the cement that doubtless adhered to the coarse particles instead of passing through the sieve.

88. The tests in Table 15 were made to determine whether the coarse particles of cement are of greater value in mortar than the same quantity of fine sand. The coarse particles of the cement were sifted out and replaced with sand grains of about the same size. The conclusion drawn from the preceding tests would indicate that some of the coarse particles of cement might be replaced by sand without diminishing the tensile strength; but the tests given in this table indicate that this is not the case when it is a question of substituting sand grains of the same size. Although such a substitution has little effect on the strength of rich mortars, it results in a decreased strength with mortars containing as much as three parts sand to one of cement by weight. (See § 85 in this connection).