PROCEEDINGS OF SECTION I. 583 



3 cubic centimeti-es into bottles containing about 50 cubic centi- 

 metres of ordinary broth and gently shaking the mixture. Some 

 advantages undoubtedly we find in the use of bile media, prin- 

 cipally due to the prevention of clotting, although possibly also 

 due to the antibactericidal effect of bile. The cultures are 

 in the great majority of cases pure, and may be directly tested 

 without plating. Rarely with the simple broth culture, and very 

 seldom with the bile cultures, staphylococci from the skin may 

 obscure the issues. This contamination is, however, largely, if 

 not entirely, due to imperfect cleansing of the skin when taking 

 the blood from the vein. 



Urine Cultures. 

 Here, although the urine may, and frequently does, yield a 

 pure culture of B. typhosus, it is advisable to make plates on 

 agar, or better, biTe salt neutral red agar, which, especially for 

 the novice, gives a clearer definition between the lactose ferment- 

 ing and non-lactose fermenting colonies. A number of 

 selected colonies are picked off into litmus lactose peptone water 

 and examined next day, and those not giving acid or acid and 

 gas on this medium are tested further. 



It is well to note that although MacConkey's medium (3) 

 (bile salt neutral red lactose agar) gives a rough separation into 

 lactose fermenters and non-lactose fermenters, certain white colo- 

 nies are found which are really lactose fermenters on the fluid 

 medium . 



Fceces Cultures. 

 Similar methods apply here as in the case of urine cultures. 

 It is advisable, however, to use several plates spread when cool 

 with descending dilutions of the ' fgeces emulsion, and it is also 

 necessary, especially when using ordinary agar, to pick off a large 

 number of colonies, for, apart from the difficulty which exists 

 in exactly separating non-lactose fermenting colonies from simple 

 fgecal coliforms which can be to a large extent overcome by experi- 

 ence, there are in faeces a large number of non-lactose fermenting 

 bacilli whose growth, both on MacConkey's jolates or on the 

 ordinai'y agar, is indistinguishable from those of B. typhosus. 

 B. fcecalis alkaUgenes and certain pseudoparatyphoid bacilli are 

 notable instances in point. 



Testing the Non-lactose Fermenting Colonies on the Blood 



Cultures. 



Having separated from our plates certain non-lactose ferment- 

 ing bacilli it is possible to make a practical diagnosis very quickly 



