528 Profs. Percy Frankland and Marshall Ward. 



The above table clearly shows 



(1.) That^e uninfected unsterilised Thames water contained no bac- 

 teria which could be mistaken for typhoid bacilli by the methods 

 of investigation employed. The phenol broth-tubes which 

 became turbid on inoculation with this water were always 

 found by plate cultivation to yield colonies which have all the 

 appearance of the B. liquidus formerly described by me 

 (' Zeitschrift fur Hygiene,' vol. 6, 1889; 'Roy. Soc. Proc.,' 

 vol. 53, 1893, p. 186). In all these experiments in which the 

 method of phenol broth-culture has been employed, I have 

 repeatedly found that this same organism succeeds in develop- 

 ing in the broth to which three drops of the phenol solution 

 per 10 c.c. of broth are added. 



(2.) That the typhoid-infected steam-sterilised Thames water con- 

 tained living typhoid bacilli throughout the entire period of 

 forty-eight days (January 16th March 5th, 1894) over which 

 this series of experiments extended. 



(3.) That in the typhoid-infected unsterilised Thames water, the 

 typhoid bacilli were still demonstrable on 5.2.1894, or twenty 

 days after their introduction ; but seven days later, on 

 12.2.1894, they were no longer discoverable. 



(4.) That in the typhoid-infected steam-sterilised TJtames water, 

 which had l>een furnished with the Thames water bacteria by in- 

 oculating with a few drops of unsterilised Thames water, the 

 typhoid bacilli were still demonstrable on 19.2.1894, or thirty- 

 four days after their introduction, whilst seven days later, on 

 26.2.1894, they were no longer discoverable. 



(5.) Thus, not only was the longevity of the typhoid bacilli far 

 greater, as usual, in the sterilised than in the unsterile waters, 

 but of the two unsterile waters, the one naturally so, and the 

 other rendered unsterile by the inoculation of a few drops of 

 unsterile Thames water, the naturally unsterile one proved to 

 be decidedly more antagonistic to the vitality of the typhoid 

 bacillus, than the water rendered artificially unsterile, as we 

 may call it, by inoculation with unsterile Thames water. 



(6.) This result is the more significant and important, inasmuch as 

 it was shown (see p. 521) that in the naturally unsterile water 

 no multiplication of the water bacteria took place, whilst, in 

 what we may call the artificially unsterile water, it was shown 

 (see p. 523) that a very large multiplication of the introduced 

 water bacteria certainly did take place. 



(7.) In the previous series of experiments (see p. 516) it was 

 equally clearly shown that the typhoid bacilli enjoyed a 

 greater longevity in the unsterile deep well water than iu 



