XIV, 2 Witt: Effect of Calciwm Sulphate on Cement 223 
included. In Table I are shown some tests on laboratory sam- 
ples which have been made from time to time. For convenience 
in comparison, they have been arranged in the order of their 
sulphuric anhydride content. 
Commercial cements clinkered in rotary kilns require the 
addition of some substance subsequent to burning. I have 
never found a cement either in the United States or in the East 
that did not contain calcium sulphate, though of course the 
amount varies greatly with conditions. ‘This cement, however, 
is an exception, when ground in a laboratory mill. Only one 
of several samples ground without the addition of gypsum was 
quick. setting, and that one required only a small addition of 
gypsum to give a normal set. Further, the addition of gypsum 
has little effect on the setting time. The results obtained with 
laboratory samples are irregular when considered individually, 
but if those for each half per cent sulphuric anhydride are aver- 
aged (omitting the samples having a flash set, and calculating 
to round numbers) we have the following, which show that 
the average setting time is apparently independent of the sul- 
phuric anhydride content: 
Sulphuric anhydride. Time of set. 
Per cent. Minutes. 
0.50 to 1.00 200 
1.01 to 1.50 250 
1.51 to 2.00 220 
2.01 to 2.50 230 
Above 2.50 240 
Table II shows some of the results obtained with plant sam- 
ples, likewise arranged with reference to their sulphuric anhy- 
dride content. Averaging these results, we have 
Sulphuric anhydride. Time of set. 
Per cent. Minutes 
Less than 1.80 33 
1.80 to 2.00 161 
Above 2.00 171 
Many plant samples below 1.80 per cent sulphuric anhydride 
are quick setting. With percentages greater than 1.80, the aver- 
age sample has a normal set, though the period is not so long 
as for the laboratory sample containing no retarder. The sam- 
ples shown in Table II were taken from the conveyors at the 
plant in the course of control work extending over several 
months. Samples were taken regularly every half hour, and 
when one was found to be quick setting, it was put aside for 
analysis and further examination. 
