XIV,2 Witt: Effect of Calcium Sulphate on Cement 229 
of analyses, an explanation was sought on the basis of physical 
changes, or of chemical changes in the constituents, of such a 
nature that they could not be detected by analysis. It is be- 
lieved that the irregularity results from changes that go on 
during the grinding of the clinker, though it has not been proved 
that such are entirely responsible. Slight variations in heat 
treatment in the kiln may partly account for it. However, it 
has been observed in connection with the grinding that such 
factors as the rate at which the clinker is fed to the mill, the 
hardness of the clinker, the relative size of the pieces, and the 
heat generated during the process have their effect. 
With reference to the grinding, the following explanations 
of the irregularity of cement ground at the plant, and between 
the regular product and samples ground in a laboratory mill 
have been offered: 
1. The heat generated during grinding at the plant causes 
the formation of alkali carbonates, which influence the set of 
thé finished cement. (A high temperature is often reached, and 
the clinker is high in alkalies.) 
2. Though the 200-mesh fineness of the plant samples and of 
the laboratory samples is practically the same there may be a 
considerable difference in the fine particles that pass the sieve. 
Neither of these hypotheses has been verified. On the con- 
trary, there is evidence against them. A series of tests re- 
ported in the first paper of this series 't shows that the set is 
only slightly affected by the presence of sodium carbonate. 
Therefore, even if considerable quantities of alkaline carbonates 
were found during grinding, the irregularity of set could not 
be accounted for on that basis. The tensile strength of labor- 
atory samples and of plant samples is approximately the same. 
This would not be likely to be so, if the latter contained an 
appreciably greater percentage of fine particles. 
SUMMARY 
The cement under investigation is unique in its setting quali- 
ties when treated with calcium sulphate. If the clinker is 
ground in a small laboratory mill, its average time of set is 
independent of the amount of sulphate added. When ground in 
* No. 1 was suggested and communicated to me by another investigator. 
" Loc. cit. 
