112 BROOM ALL : 



ings should therefore be taken of the amount of the gas on 

 each of the three days. Other gas producing bacteria act 

 more slowly. 



4. The gas produced by Bacilhis coli communis is gener- 

 ally assumed to be composed of hydrogen and carbon dioxide 

 in the proportion of H : CO.^ :: 2 : i. To determine this 

 ratio sotlium hydroxide is added to the tube, the open end 

 filled to the top with water, the orifice closed with the thumb, 

 the tube tilted back and forth several times to mix, and then 

 all the remaining gas allowed to collect in the closed limb. 

 The residual gas is assumed to be hydrogen (carbon dioxide 

 having been absorbed by the alkali) and the ratio then calcu- 

 lated. If this ratio of H : CO., approximates 2 : i the reac- 

 tion is considered positive and the water is not above sus- 

 picion. 



5. The liquid in the fermentation tube must be pronoun- 

 cedly acid to indicate the presence of Bacillus coli covimunis. 



6. Bacillus coli communis grown on nutrient gelatine 

 forms non-liquifying, whitish, leafy colonies showing radial 

 lines. 



If the above tests, or most of them, give a positive reac- 

 tion, the water must be regarded with suspicion and the ser- 

 vices of the bacteriologist called in to determine to exactly 

 what extent Bacillus coli coynmunis and other intestinal bac- 

 teria, both harmful and otherwise, may be present. 



The routine application of these tests to a given water 

 supply will soon enable one to become acquainted with the 

 characteristics of the water and to establish a sort of normal. 

 Any sudden or persistent deviation from this normal should 

 be ground for grave suspicion and an investigation should be 

 at once started. 



Interpretation of Results. — The question of the weight to 

 be given to the results of the various chemical and bacterio- 



