IX. J CHEMISTRY OF NATURAL WATERS. 107 



21. I find, however, that hydrated carbonate of magnesia 

 readily and completely decomposes a solution of gypsum when 

 agitated with it, with formation of carbonate of lime and 

 sulphate of magnesia ; and the same result is produced by the 

 native hydrate of magnesia when mingled with a solution of 

 gypsum in presence of carbonic acid. Now there may be 

 dolomites which contain an admixture of hydro-carbonate of 

 magnesia, as there certainly are others which, like predazzite, 

 are penetrated with hydrate of magnesia.* The reaction 

 between solutions of gypsum and such magnesian limestones 

 (with the intervention, in the case of predazzite, of atmospheric 

 carbonic acid) would suffice to explain the results obtained by 

 Mitscherlich, and the appearance in certain cases of sulphate 

 of magnesia as an efflorescence on dolomites. In the exper- 

 iments above described, the nearly pure crystalline dolomites 

 from the Guelph and Niagara formations were made use of. 



22. When sea-water is exposed to spontaneous evapora- 

 tion, the lime which it contains separates in the form of sul- 

 phate, gypsum being but sparingly soluble in a concentrated 

 brine, and the greater portion of the chloride of sodium crystal- 

 lizes out in a nearly pure state. The mother-liquor of specific 

 gravity 1.24, having lost about four fifths of its chloride of 

 sodium, still contains dissolved a large proportion of sulphate 

 of magnesia. If the evaporation is continued at the ordinary 

 temperature till a density of 1.32 is attained, about one half 

 of the magnesian sulphate separates, mixed with common salt ; 

 and by reducing the temperature to 6 C., a large portion of 

 pure sulphate of magnesia now crystallizes out. The further 

 evaporation of the remaining liquor by the heat of summer 

 causes the potassium-salt to separate in the form of a hydrous 

 double chloride of potassium and magnesium, an artificial car- 

 iiallite.f 



[* In subsequent experiments it was found that certain dolomites contain a 

 little hydrous carbonate of magnesia capable of decomposing a limited amount 

 of solution of gypsum. See the author, On Lime and Magnesia Salts, Ameri- 

 can Journal of Science for July, 1866, 96-101.] 



t The hydrous dovible chloride of potassium and magnesium, to which the 

 name of carnallite has been given, occurs in large quantities in the upper 



