IX.] CHEMISTRY OF NATURAL WATERS. 105 



regions large quantities of the double chloride of potassium and 

 magnesium, carnallite ; and in others sulphate of soda, sulphate 

 of magnesia, and complex sulphates like bloedite and polyhal- 

 lite. Besides these crystalline salts, the mother-liquors con- 

 taining the more soluble and uncrystallizable compounds may 

 also be supposed to impregnate, in some cases, the sediments 

 of these saliferous formations. The conditions under which 

 these various salts are deposited from sea-water, and their rela- 

 tions to the composition of the ocean in earlier geological 

 periods, are reserved for consideration in 22. Infiltrating 

 waters remove from these saliferous strata their soluble ingre- 

 dients ; which, together with the ancient sea- waters of other 

 sedimentary rocks, give rise to the various neutral saline 

 waters; while the mingling of these in various proportions 

 with the alkaline waters whose origin has been described in 

 1 3, produces intermediate classes of waters of much interest. 



18. I have elsewhere described the results of a series of 

 experiments on the mutual action of the waters of these two 

 classes.* When a dilute solution of bicarbonate of soda is 

 gradually added to a solution which, like sea-water, contains, 

 besides chloride of sodium, the chlorides and sulphates of calcium 

 and magnesium, the greater part of the lime separates as car- 

 bonate, carrying down with it only from one to three hun- 

 dred ths of carbonate of magnesia ; a portion of lime however 

 remaining in solution as bicarbonate. When the chloride of 

 calcium is wholly decomposed, the magnesian salt is attacked 

 in its turn, and there finally results a solution in which the 

 whole of the earthy chlorides are replaced by chloride of sodium. 

 A further addition of the solution of bicarbonate of soda gives 

 them the character of alkaline-saline waters ; which moreover 

 contain an abundance of earthy carbonates. 



19. In the saline waters just considered, chlorides generally 

 predominate, the sulphates being small in amount, and often 

 altogether wanting. Some exceptions to this are however met 

 with ; for apart from waters impregnated with gypsum, whose 

 origin is readily understood, there are others in which sulphate 



* American Journal Science (2), XXVIII. 170. 

 5* 



