PAPERS ON WATER SUPPLY 



89 



Health Association, and has been used to a certain extent, but 

 has the disadvantages of requiring special apparatus and a 

 blank determination in each case. These two methods are prac- 

 tically the only ones in use in this country. A brief description 

 of the chemistry of each method follows. 



In the Levy method the sample is collected in a special 

 pipette, into which is introduced an accurately measured quan- 

 tity of ferrous ammonium sulfate (FeS0 4 (NH 4 ) 2 S0 4 ) ; a so- 

 lution of sodium hydroxide (NaOH) is then added; ferrous 

 hydroxide (Fe(OH) 2 ) is formed and part of this is then 

 oxidized to ferric hydroxide (Fe(OH) 3 ) by the dissolved oxy- 

 gen present. Sulfuric acid (H 2 S0 4 ) is then added and the re- 

 maining ferrous sulfate (FeS0 4 ) is titrated with standard 

 potassium permanganate (KMn0 4 ). An equivalent volume of 

 water is treated with the same volume of reagents, but the 

 sulfuric acid (H 2 S0 4 ) is added first to prevent oxidation. The 

 difference between the titration of the blank and sample will 

 give the amount of ferrous salt oxidized by the dissolved oxy- 

 gen in the water; the dissolved oxygen content in parts per 

 million may be calculated from this. 



In the Winkler method the sample is collected in a glass- 

 stoppered bottle, and solutions of manganous sulfate MnS0 4 , 

 and a mixture of sodium hydroxide, NaOH, and potassium 

 iodide, KI, added. The manganous sulfate and sodium hy- 

 droxide form manganous hydroxide, Mn(OH) 2 ; part of this 

 is oxidized to manganese oxy-hydroxide, MnO(OH) 2 by the 

 dissolved oxygen present in the water. Then sulfuric acid is 

 added; this dissolves the precipitate and liberates free iodine 

 from the potassium iodide according to the following equa- 

 tion : 



MnO(OH) 2 +2'H 2 S0 4 +2KI=MnS0 4 +K 2 S0 4 +3H 3 0+I 2 



The amount of iodine liberated is proportioned to the 

 amount of dissolved oxygen present ; the free iodine is titrated 

 with standard sodium thiosulfate, (Na 2 S 2 3 ), using starch 

 paste as an indicator. Dissolved oxygen is reported in parts 

 per million. 



The Levy and Winkler methods have both been checked with 

 the gasometric method on aerated distilled water. The Levy 

 method has sometimes been condemned because it has seemed 

 to give variable results 1 . On the other hand, it has been rec- 

 ommended because it is claimed that nitrites do not affect its 

 accuracv. 



1. Chlopin; Arch. f. Hygiene, 32, 294-309, 1898. Tiemann and Preusse; Ber., 



12, pp. 1784-5. 



