and the Constituents of Well- and River-waters. 61 



in all probability to the volatilization of the chlorides of mag- 

 nesium and sodium* and decomposition of nitrates. 



As the spontaneous development of vegetable and animal life 

 in Thames water is one of the surest indications of the presence of 

 decaying organic matter, I allowed two separate gallons of the 

 water which had been treated with iron, to remain exposed to the 

 air and light during two months in the summer. At the end of 

 that time, not the slightest signs of either vegetable or animal life 

 were observable; but the bottles containing two gallons of the 

 ordinary water, similarly exposed, were filled with a green vege- 

 tation, which afforded a refuge for innumerable animalculse. This 

 fact appears to offer, perhaps, the best proof of the partial or 

 entire destruction, by iron, of the putrescible organic matter 

 contained in Thames water. As the influence of iron upon the 

 organic matter and the sulphurets contained in water appears 

 to vary very considerably according to the temperature of the 

 atmosphere and the season of the year, I have made some addi- 

 tional experiments on the action of a much larger surface of iron 

 upon a given quantity of water. Although iron in the form of 

 wire exposes a less surface to the water than it would in many 

 other forms, I have nevertheless found it more convenient to use 

 the wire. 



Two analyses of the Grand Junction Company's water were 

 made on the 20th and 21st of October, and the results of the 

 first analysis were, — 



Solid residue (dried at 248° F.) . . 20-1247 grs. per gal. 

 Including matter volatile on ignition . 1'3456 „ 



In the second analysis, — ■ 



Solid residue (dried at 248° F.) . . 20-2530 grs. per gal. 

 Including matter volatile on ignition . 1-4871 „ 



Half a gallon of the water taken on the 20th of October was 

 put into a bottle containing 2~ ounces of iron wire and allowed 

 to remain for forty-eight hours. A copious brown deposit was 

 formed. The water was filtered and evaporated down in an open 

 porcelain dish to about four fluid ounces. The deposit formed 

 on boiling was perfectly white. It was transferred to a platinum 

 dish, evaporated and dried in the usual manner, and then ignited. 

 The analysis gave, — 



Total solid residue 11'4391 grs. per gal. 



Including matter volatile on ignition. 0-20G3 „ 



Hardness of the water by Clarke's scale, 6°-2 



* Tlu; intense yellow colour imparted to the gas-flame during the igni- 

 tion of the residue of Thames water, is a proof that the chloride of sodium 

 is partly volatilized. 



