QUALITATIVE BACTERIOLOGICAL ANALYSIS. 199 



acid and 0*002 per cent, of methyl-violet added. The colonies 

 of B. typhosus grow in a characteristic manner and remain 

 distinctly bluer than the surrounding medium. Stoddart 

 recommended a medium containing 0'5 per cent, agar and 5 per 

 cent, gelatine. A sufficient quantity of this medium is poured 

 into a dish or flat flask, so as to fill it to a depth of 5 mm. ; 

 the dish or flask is then sterilised and allowed to cool gradually 

 (preferably in the steriliser over night). Drops of moisture 

 which may have condensed on the cover are drained off, and the 

 centre of the medium touched with a platinum needle charged 

 with the material to be examined. In order to prevent con- 

 densation the dish is enclosed in a larger one and incubated at 

 35 C. for twenty-four hours. If B. typhosus is present, about 

 two-thirds of the dish will be found occupied by a circular 

 opalescence, which later extends throughout the whole dish. 

 Under similar conditions the B. coli produces the usual colony 

 of limited extent. If the inoculation is made from a mixture 

 of both organisms, there results a central disc of coli bacilli, 

 surrounded by an opalescent halo of typhoid growth, which 

 yields perfectly pure sub-cultures. Unfortunately many of the 

 varieties of B. coli, which are motile, produce exactly the same 

 halo appearance as the B. typhosus ; so the test cannot be relied 

 upon to distinguish between these two microbes. 



Hankin has published a method by which he states he has 

 been able to isolate the B. typhosus from water supplies in 

 India. The method is as follows : 



^ Five tubes are taken, each containing 10 c.c. of neutral 

 bouillon. To the first tube no addition of Parietti's solution 

 is made. It merely serves as a control of the capacity of the 

 bouillon used to permit the growth of microbes. To the 

 remaining tubes are added one, two, three and four drops of 

 Parietti's solution respectively. Each tube is then infected with 

 a few drops of the water or other liquid to be tested. The 

 tubes are covered with india-rubber caps and placed in the 

 incubator at 37 C. for twenty-four hours. On the following- 

 day a variable number of the above tubes will be found to be 

 turbid. One of the series has now to be chosen for further use. 

 The tube containing the highest number of drops of Pariettfs 

 solution that is yet turbid should be discarded. Usually the 



