Article X. — A Collecting Bottle especially adapted for the Quan- 

 titative and Qualitative Determination of Dissolved Gases, particidarl\ 

 very Small Quantities of Oxygen* By Edwin B. Powers. 



One of the sources of error in the Winkler method for the de- 

 termination of dissolved oxygen in water, especially where the oxygen 

 content is low, is the diffusion of oxygen into the water before and 

 during the introduction of the chemicals. Another source of error is 

 the mixing of the manganous chloride with the potassium iodide- 

 alkali solution at the surface of the water, the chemicals adhering to 

 the pipettes introducing these reagents having washed off at the top 

 of the bottle, where they react with the oxygen present. In recent 

 work involving the oxygen-free water apparatus described by Shelford 

 in the preceding article of this volume, it was found especially de- 

 sirable to eliminate the above sources of error. This was acconiijlishcd 



d 



a c 



♦This bottle was devised primarily for the study of the oxycjen requirements of 

 crayfishes, with a view to their use as index or,<;;niisiiis. 



577 



