The national media have reported that the chlorination of water during treatment is responsible for the formation of potentially harmful chlorinated organic materials—notably chloroform—in the nation's water supplies. The following report by three research scientists from the Natl. Envir. Res. Ctr. of EPA describes that agency's research concerning these organohalides. The report concludes that the number of organohalides formed during the chlorination process does not constitute any immediate threat to the public health or welfare, but that more research into possible long‐term effects is warranted.

Presented here is a method for quantitative recovery of volatile organic compounds followed by a description of apparatus and procedures employed to detect 0.5 μ.g/1 of the substances.

The persistence of 28 common pesticides in raw river water was studied over an eight-week period. Twelve organochlorine, nine organophosphorus, and seven carbamate pesticides were studied at a concentration of 10 ^g/liter. No measurable degradation or chemical change was observed for the following organochlorine compounds: bhc, heptachlor epoxide, dieldrin, dde, ddt, ddd, and endrin. Azodrin was the only organophosphorus compound that was stable throughout the study. All carbamate compounds were significantly changed after one week, and all but Baygon were completely lost after eight weeks. Where possible, the degradation or chemical conversion products of the pesticides were identified.

Methyl mercury is extracted as the bromide salt from fish and sediment and as the chloride salt from water samples. All extracts are treated with a common cleanup procedure that results in the conversion of methyl mercury to the iodide salt for electron capture gas chromatographic analysis. Recoveries ranged from an average of 88.5% for water samples to averages of 95.5 and 96.3% for perch and sediments. Methods for controlling contaminants and interferences are discussed for all phases of the method. Particular problems encountered were column poisoning and detector poisoning. When the method was applied to sediment samples collected from a polluted river, a correlation could be established between total mercury and methyl mercury when the concentration of total mercury was in the 0–10 μg/g region. For samples of very high inorganic mercury, the correlation failed.

No abstract