366 Kr eider — Oxygen in Air and in Aqueous Solution. 



Ko correction was found necessary for the blank determina- 

 tions, since when boiled water was used the solution was onlj 

 faintly colored with iodine, which requires only a drop of 

 arsenic solution to bleach it. As is evident from the table, the 

 determinations according to this method are not reliable beyond 

 0'05 per cent, but for practical purposes this is sufficiently ac- 

 curate. For the sake of comparison two determinations by the 

 pyrogallic acid method were made upon a portion of the same 

 air used in the last experiments, the results being 20*93 and 

 20'88 per cent respectively. While the pyrogallic acid method 

 is capable of much greater accuracy when applied in HempePs 

 improved apparatus, in ordinary burettes it will probably not 

 yield more closely agreeing results than the above method. 



Determination of dissolved Oxygen. — A determination of 

 oxygen dissolved in water can be completed by the above 

 method in about ten minutes by means of the apparatus illus- 

 trated by the accompanying figure. 



The apparatus consisted of a flask of about 300^°"* 

 capacity, into the bottom of which was sealed a stop- 

 cock with a long exit tube. Upon the neck was cut 

 the fiducial circle c and immediately above this stop- 

 cock e was sealed as shown. The neck of the flask 

 was drawn out and sealed to stop-cock d and the 

 bulb, a^ of about 30^°^' capacity blown in it. The 

 capacity of the apparatus between stop-cock, J, and 

 the fiducial mark, c, was carefully determined. 



The manipulation for the determination of dis- 

 solved oxygen was as follows : The flask was held 

 in the position shown by a clamp fastened to a mov- 

 able support. Stop-cock h being closed, the water 

 was admitted through e and the air allowed to 

 escape through d until the level of water was that indicated 

 by the line/*. (When the water to be examined is not satu- 

 rated with air, the flask must first be filled with carbon dioxide 

 and the water entered by replacement of that gas.) With d 

 closed, sufficient water was allowed to escape through h to 

 bring the surface to ^, which was then closed. The nitric oxide 

 generator was then attached to </, and by opening h the gas was 

 allowed to replace the water until the meniscus coincided with 

 c, when d was closed and the generator disconnected. Two 

 cubic centimeters of strong hydrochloric acid were introduced 

 through e by expelling nitric oxide through d^ in which a drop 

 of water formed an effective trap to prevent the entrance of 

 air. Then the potassium iodide was admitted in the same way. 

 The solution of iodide for this purpose was free of oxygen and 

 contained one gram in three cubic centimeters. It was kept 

 under pressure of carbon dioxide in the bottle previously 



