WATER. 337 



the soil, or constitute lakes. Rain-water although nearly pure is 

 not absolutely exempt from extraneous matters. The air, especial- 

 ly after continued drought, always holds dust in suspension ; this 

 yields to the rain by which it is precipitated, whatever soluble mat- 

 ter it may contain. 



It is further ascertained by the experiments of Cavendish and 

 Seguin, that whenever the electric spark traverses a humid mixture 

 of oxygen and azote, nitric acid and nitrate of ammonia are produced. 

 Now this frequently happens; and according to Professoi Liebig 

 storm-rain always contains nitric acid associated with lime or ammo- 

 nia. Common rain seldom contains nitrates, merely faint traces of 

 common salt.* 



In river and spring- water there necessarily exists a larger amount 

 of dissolved substances derived from the strata they pass through, 

 varying in nature according to the geological structure of the locali- 

 ty. From old crystalline rocks, like granite, water issues sometimes 

 so little impregnated with salts, as to be almost identical with dis- 

 tilled water ; that, on the contrary, which rises from a calcareous or 

 gypseous bed is always contaminated with salts of lime. Notwith- 

 standing the minute quantity of saline or earthy ingredients in spring 

 and river-waters, they are drinkable, and considered good when 

 they are limpid, without odor, capable of dissolving soap, and fitted 

 for vegetable cookery. These two last characters are essential, 

 inasmuch as proving that the waters contain only infinitesimal quan- 

 tities of soluble salts of lime. 



The action of tests readily indicates the nature of the dissolved 

 salts. 



Water contains : sulphates or carbonates, if nitrate of barytes 

 causes a precipitate ; a sulphate, when the precipitate is not redis- 

 solved by the addition of nitric acid ; 



Chlorides, if it give with nitrate of silver a curdy precipitate, in- 

 soluble upon addition of nitric acid; 



Lime, when rendered turbid by oxalate of ammonia ; 



Magnesia, if when mixed with pure ammonia, and preserved in a 

 closely stopped vial, a white flocculent deposite ensues. This test, 

 however, is only applicable to water that has been boiled sufficiently 

 long to expel all the carbonic acid in solution, and which would tend 

 to hold any carbonate of lime dissolved. Carbonate of lime is sepa- 

 rated from water by ammonia, after some hours, in the form of gran- 

 ular crystals, which adhere to the sides of the vessel. 



In order to render the operation of tests more sensible, the bulk 

 of the water may be reduced to a half or a fourth by evaporation. 



Besides fixed salts, river- water always contains those of ammo- 

 nia, particularly the carbonate ; this fact was first ascertained, re- 

 lative to the Seine water, by M. Chevreul.f Subsequently, Pro- 

 fessor Liebig has discovered the same ammoniacal salt in rain-wa- 

 ter ; and M. Hunefeld has proved, that spring-water likewise con- 



* Annales de Chimie, t. xxxv. 2e s6rie. 



t Chevreul, Annales de Chimie, t. Ixxxii. p. 5(1 



