114 S. A. KOSER AND W. W. SKINNER 



reduced as to give negative results upon plating 1 cc. quantities, 

 larger amounts, 5 cc. and 10 cc, were introduced into broth to 

 determine, insofar as possible, the final disappearance of the 

 organisms in question. This was done by streaking Endo plates 

 from the broth cultures, fishing any typical colonies, and finally 

 applying the usual methods used for the identification of the 

 various members of this group of organisms. 



EXPERIMENTAL 



Since temperature may be expected to exert a marked influ- 

 ence upon the death-rate, experimental samples were held at 

 two different temperatures, namely, in cold storage at 1°C. and 

 at room temperature, 19° to 23°C. Tables 1 and 2 present data 

 showing the viability of Bad. coli in carbonated water at several 

 different pressures and also, for purposes of comparison, in plain 

 tap water. It is evident that carbonation causes a speedy de- 

 struction of the colon bacillus and that this effect is dependent 

 upon the temperature at which the samples are held, being much 

 more pronounced at room temperature than at 1°C. Further- 

 more, the different degrees of pressure of CO2 mentioned in 

 these tables apparently exerted little or no influence upon via- 

 bility, for the organisms were killed as speedily in water saturated 

 with CO2 (at both 20°C. and 1°C.), but under no excess pressure, 

 as they were in the carbonated samples under pressures of 28 

 and 41 pounds per square inch. In fact, where the pressure 

 was released the plate counts frequently were less than those 

 of the samples held under pressure (table 2), a phenomenon 

 which was regularly observed upon several repetitions of the 

 experiment. 



To gain an idea of the hydrogen-ion concentration of carbonated 

 water the indicators brom-phenol-blue and methyl-red were 

 added to different bottles which were then filled with carbonated 

 water at these several pressures. In this way the value was 

 roughly determined as pH 4.0-4.4. Release of the pressure, 

 as indicated in table 2, was followed by very little, if any, im- 

 mediate change in the hydrogen-ion concentration when meas- 



