10 METHODS OF J\Vi:ST/a.\TIO\ 



will then have to ]h: cinijloycd. Suitable methods, some 

 of them requiring- only a few mm." of gas, have been 

 developed, especially by Krogh (1926). The new micro 

 gas analyzer described recently by Scholander (1942) 

 should also be mentioned as a very promising apparatus. 

 The taking of samples of soil gases uncontaminated with 

 atmospheric air presents signal difficulties, but suitable 

 methods have been described by Leather (1915). 



B. Aquatic hahitafs. The most widely used method 

 for determining the oxygen content in water is Wink- 

 ler's (1888) method. Its principle is to absorb the oxy- 

 gen present in the sample on manganous hydroxide. 

 Upon acidification, in presence of iodide, an amount of 

 iodine equivalent to the absorbed oxygen is liberated and 

 this amount is determined by titration with thiosulfate. 

 The method is rapid and very reliable provided that the 

 water is pure. This limitation is well known, but unfor- 

 tunately has not been taken into account by the authors 

 of some of the papers to be reviewed in the present 

 survey. Serious errors result if the water is impure, 

 but several modifications have been proposed that may 

 be employed when this is the case. Rideal-Stewart's 

 (1901) procedure is convenient when larger amounts of 

 nitrates or ferrous salts are present ; a preliminary treat- 

 ment eliminates the interfering substances by means of 

 permanganate. For water containing organic matter, 

 hypochlorites, or sulfite wastes, the modification by The- 

 riault and McNamee (1932) may be used in which the 

 impurities are oxidized prior to the actual oxygen de- 

 termination, through an alkaline solution of sodium hy- 

 pochlorite. Both the original method and the two modi- 

 fied procedures just mentioned are described in detail in 

 a publication of the American Public Health Associa- 

 tion (1936). Another excellent modification for use in 

 the presence of organic contaminants has been developed 

 by Alsterberg (1926). In this case the undesirable con- 

 stituents of the sample are oxidized with bromide. 



