S U M M A R YAt high relative humidity (RH) similar survivals were obtained for storage in oxygen, air or nitrogen. At low RH the survival in nitrogen was much greater than that in air or in a 20 % (v/v) oxygen + 80 % (v/v) nitrogen mixture, in which the survivals were similar. In oxygen alone, survival was even lower than in air. Hence oxygen, or a trace contaminant in it, was responsible for the poorer survival in air than in nitrogen. The mechanism of death caused by oxygen is discussed.
I N T R O D U C T I O NThe aerosol survival of Escherichia c d i has been shown by Cox (19664 b; 1967) to depend upon several variables. Amongst these was the nature of the atmosphere, in that at low relative humidity (RH), but not at high RH, survival was much greater in nitrogen than in air (Cox, 1966a;Cox & Baldwin, 1966). The previous suggestion that this survival difference was owing to the toxic action of oxygen (Cox, 1966a) is now examined. Hess (1965) found that at low RH, but not at high, oxygen caused enhanced death of aerosolized Serratia marcexens and Escherichia coli B.
METHODSThe methods used were as previously reported (Cox, 1966a) except that in all experiments a rotating dnm container (Goldberg, Watkins, Boerk & Chatigny, 1958) was used for storing the aerosol cloud, and only Escherichia coli B sprayed from distilled water was studied. The drum apparatus used for the present work was similar to that used by Cox (1966a). Table I and show that at high RH similar survivals were obtained €or storage in oxygen (> 99*50/0), air or nitrogen (> 9909%). At low RH the survival in nitrogen was much greater than in air or in a 20 % (vlv) oxygen + 80 yo (vlv) nitrogen mixture, in which the survivals were similar. In oxygen alone survival was even lower than in air. Hence oxygen, or possibly a trace contaminant in it, was responsible for the poorer survival in air than in nitrogen. Analytical data of the oxygen did not suggest any obvious contaminants that were likely to have a toxic action at the parts per million concentrations at which they were present. However, such possibilities are not excluded completely since oxidizing agents such as ozone and the oxides of nitrogen at concentrations of a few parts per million might be toxic for Escherichia coli B in aerosols at low RH values. Also, oxygen+other 8-2
RESULTS
Results are given in