the JVaters of the Si. Lawrence and Ottawa Rivers. 241 



chlorate of potash. By evaporation a portion of silica separated, 

 and the solution gave with ammonia a colourless precipitate, 

 which was in great part soluble in potash, and contained alumina; 

 redissolved in hydrochloric acid, however, it gave with a sulpho- 

 cyanide evidence of the presence of peroxide of iron, and with 

 molybdate of ammonia an abundant yellow precipitate, indicating 

 phosphoric acid. Heated on silver foil with caustic potash, the 

 aluminous precipitate gave a distinct manganese reaction. 



When the Ottawa water is evaporated to dryness in a platinum 

 capsule with an excess of hydrochloric acid, and the residuum 

 treated with dilute acid, there remains a large amount of silica, 

 coloured brown by organic matter. It becomes white by ignition, 

 and is then perfectly pure, and equals one-third of the whole 

 solid matters present. A portion of the water was evaporated 

 to one-fortieth and filtered; when further evaporated to one- 

 fourth, it deposited on the platinum vessel an opake film, which 

 was but partially soluble in hydrochloric acid. The liquid was 

 now dark brown, and reddened turmeric-paper. Evaporated to 

 dryness and ignited, the portion soluble in water was strongly 

 alkaline to test-papers, and perceptibly so to the taste. The in- 

 soluble portion was decomposed by hydrochloric acid without 

 effervescence, pulverulent silica separated, while the acid retained 

 in solution a portion of lime, but no magnesia. 



The residue from the evaporated water is deep brown in colour ; 

 when ignited, it diffuses an agreeable vegetable odour, and a 

 little carbon remains. The water was not examined for nitrates ; 

 but the absence of any deflagration during the ignition showed, 

 that, if present, they were in small amount. The season at which 

 the water was collected, at the close of a long winter, could more- 

 over scarcely be supposed 'to favour the presence of nitrates. 

 The following numbers are deduced from two or more concord- 

 ant determinations made upon two and four litres of water, and 

 calculated for ten litres, or 10,000 grms. : — 



Carbonate of lime .... '2480 grm. 



Carbonate of magnesia . . . 'OGOG ... 



Chlorine -0076 ... 



Sulphuric acid '0161 ,.. 



Silica -2060 ... 



Chloride of sodium .... 'OGO? ... 



Chloride of potassium . , , '0293 ... 



Residue dried at 300° F. . . '6975 ... 



llcsiduc ignited '5340 ... 



The amount of silica remaining dissolved in the water, evapo- 

 rated to one-twentieth, was found in two experiments to be 0*019 



Phil. Mag. S. 4. Vol. 13. No. 86. April 1857. S 



