164 University Geological Survey of Kansas. 
present also as carbonate, as it frequently is in limestones, 
then the amount of free calcium and magnesium would be still 
higher. 
If this series of determinations is analyzed, it is seen, first, in 
reference to the silica, that it increases in the finished sample 
about in proportion to the loss of water. In the dust, which 
contains more water than the finished sample, the amount of 
silica is, as would be expected, less. The tailings are made up 
of the hardest and most impure portion, and this portion would 
be expected to contain more silica; in fact it contains a very 
large amount. The set sample, since it contains about the same 
quantity of water as the original cement, would be expected to 
contain about as much silica, which it does. The calcium sul- 
phate is highest in the dust, as though this represented the purest 
part of the material, which was blown off with the steam on ac- 
count of its lower specific gravity. It is no doubt true that a 
partial separation of the siliceous material from the pure gyp- 
sum could be made by the process of winnowing, as it may be 
called. The tailings also contain more of the limestone than 
any of the othersamples. In the light of these facts, it is doubt- 
ful whether there is any advantage in grinding and reheating 
the tailings, as they contain so much more silica and limestone 
than the average of the bed. Allusion has already been made 
to the varying amounts of water. That in the original sample 
is only 0.29 per cent. above the theoretical amount. After boil- 
ing it is 4.91 per cent., as compared with the theoretical amount, 
4.38 percent. This is an excellent agreement, when we con- 
sider that the work is done on a commercial scale and it is 
practically impossible to dehydrate every particle of the gypsite. 
When the plaster is set it takes up a sufficient quantity of 
water, so that there is only 0.81 per cent. more than in the 
original sample. This amount is only 0.22 per cent. above the 
theoretical amount. The material was partially dehydrated, 
losing three-fourths of its water, and then as soon as an oppor- 
tunity was offered the crystals again formed with the original 
quantity of water, and the interlacing crystals made a firm, hard 
substance. 
