No abstract

The potential errors in the various tcchniqucs for the Winkler m&hod have been cxamincd and a new tcchniquc dcvclopcd.The accuracy of this technique has been tested by comparison with standards based on dissolving known quantities of oxygen in oxygen-free water. An accuracy of 0.1% was obscrvcc~, so the tcchniquc nppcars suitable for oxygen analysts requiring an accuracy of grcatcr than the 3-50/o obscrvcd with tho common techniques.INl'nODUC'l'ION

An extended series of measurements of oxygen solubility in pure water and dilutions of seawater over the temperature range 0.5 to 35C was made. Nonlinear dependence of oxygen solubility on chlorinity was found. Interpolation tables of the data have been prepared. The accuracy of the techniques was approximately 0.1%, so the tables appear to be more reliable than those in current use. Because many of the existing field data have been collected using techniques for the Winkler method that introduce significant systematic errors due to volatilization and air oxidation errors, the use of the reported solubility values with these field data will produce misleading results.

Chlorinated waters are being introduced into estuarine and coastal areas in increasing quantities. In such systems, the chlorine reacts with the natural bromide and ammonia to produce the highly toxic hypobromous acid, hypobromite ion, and haloamines. Sunlight causes up to 50 percent conversion to bromate ion, which is persistent in natural waters and has an unknown toxicity.