VIABILITY OF THE COLON-n'PHOrD GROUP 113 



The beverages were prepared and carbonated in the 7-ounce 

 bottles commonly used in the industry. 8ince they were pre- 

 pared as nearly as possible under commercial conditions and no 

 effort was made to sterilize the various ingredients, control 

 examinations of the product were made previous to experimental 

 inoculation to determine the absence of the particular type of 

 organism used in the investigation. In no instance was any 

 difficulty of this kind encountered. Throughout the work 

 commercial CO2 was used for carbonation. As a test for any 

 impurities in the carbon dioxide which might affect the death- 

 rate of the organisms, tap water was carbonated as usual, then 

 heated in the Arnold sterilizer for a short period to expel the 

 CO2 and finally the death-rate of Bad. coli in this water was 

 comjiared to that in parallel samples of the original tap water. 

 No discrepancies other than those which might be attributed to 

 experimental variation were observed. 



Small amounts of a suspension of the various test organisms 

 in sterile tap water were used for inoculation. This was accom- 

 plished in one of the two following ways. The first method 

 consisted of adding equal amounts of bacterial svispension to 

 each bottle just before carbonation. In the second method the 

 samples were prepared, bottled, carbonated, and capped as 

 usual. They were then stored at 1°C. for several days until 

 used, when the bottles were re-opened and inoculated. If opened 

 while still cool, there was little loss of CO2 gas. The first method 

 was used for most of the experiments with Bad. coli. The sec- 

 ond method was necessary when working with Bad. typhosum 

 and Bad. paratyphosum B since by the first method there is 

 more or less spattering of the material during the process of 

 carbonation. 



luMnediatelj^ after inoculation and at definite intervals there- 

 after plate counts were made. To prevent the considerable 

 loss of CO2 upon repeated opening of the same bottles, especially 

 those held at room temperature, a number of bottles were inocu- 

 lated with equal amounts of bacterial suspension and, at 

 each time interval, different sets of two were opened and samples 

 withdrawn for plating. When the numbers had become so 



