Batiey.| Chemistry of Gypsum. LSS 
Na,SO, or CaCl,. Winds, floods, or high tides, might cause 
the additional deposit of other substances, such as clay and 
siliceous material. 
‘‘Tn speaking of the occurrence of anhydrite, associated with 
gypsum, Ochsenius thinks it may be due to the fact that hydrous 
calcium sulphate, in solution with potassium salts, may ex- 
change a part of its water for potassium sulphate in the forma- 
tion of such salts as polyhalite, etc., and this change, by 
withdrawal of the water, results in the formation of anhydrite.’’ 
This occurrence of the anhydrite with the gypsum was noticed 
by Prof. G. E. Patrick” in the analysis of the Iowa rock. 
Another method for the formation of gypsum has been sug- 
gested, and that is by the decomposition of calcium carbonate 
by means of iron and copper sulphates, or the materials that 
are found in clay shales. 
Thus: CaCO,+FeSO,=—CaSO,+FeC0O,. 
Even the iron sulphide (pyrite) that is so often found in shales 
would on oxidation yield a sulphate, which in turn might 
decompose the calcium carbonate. 
Another method of formation, and one that can be readily 
seen at the present time, is the action of volcanic gases, con- 
taining sulphur dioxide, in the presence of oxidizing material, 
upon calcium carbonate. 
A method that would account for the formation of the beds 
of ‘‘gypsite,’’ or gypsum dirt, as it is sometimes called, would 
be by the solution of the calcium sulphate from the already ex- 
isting rock by the percolation of water over the beds and its 
subsequent rising to the surface and spreading out as a shallow 
lake. It would be very easy to understand how some sand and 
clay, and even calcium and magnesium carbonate, could be 
mixed with this material, for the winds of the surrounding 
country, which would be probably denuded of vegetation from 
the character of the soil, would readily transport this material 
and deposit it in the lake. So from the solution of a compara- 
tively pure gypsum rock we should have various grades of 
57. Geological Survey of Iowa, Vol. III, p. 292. 
v—10 
