FOOD-POISONING BACILLI 381 



The methods for the isolation of the members of the group 

 vary with the nature of the infected material to' be examined. 

 In the case of abscesses caused by the paratyphoid bacillus, 

 the organism is usually accidentally discovered during the 

 application of ordinary methods. When deliberate search for a 

 member of the group is required, usually either the faeces or the 

 blood constitutes the material to be examined. In the former 

 case, advantage is taken of the fact that the food-poisoning 

 bacilli do not ferment lactose. Thus, if McConkey bile-salt 

 lactose-agar plates (p. 50) be used, the organisms sought for 

 will appear as colourless colonies which can be picked off for 

 systematic investigation. In the case of blood, ordinary methods 

 will prove sufficient. 



Capacity for fermenting sugars has been largely applied in 

 work on this group. All the members produce practically the 

 same reactions. They originate acid and gas in glucose, lajvu- 

 lose, sorbite, mannite, dextrin, maltose, dulcite, galactose and 

 arabinose, like b. coli, but produce no change in lactose, cane- 

 sugar, salicin or inulin. Although differences in fermenting 

 capacity have been noted in different strains, the existence of 

 such cannot be relied upon for differentiating members of the 

 group from one another. The sugar reactions are only of use 

 in demarcating the lines between the food-poisoning group and 

 b. coli on the one hand, and b. typhosus on the other. The 

 differentiation of members of the group can only be effected by 

 applying the agglutination tests to the serum of animals suffer- 

 ing from natural or artificial infection. The chief point here is 

 that in such infections, the occurrence of group agglutinins in 

 the serum is much in evidence. Herein lies the necessity for 

 having at hand the historic strains of the organisms referred to 

 above. In dealing with an organism, it is first of all advisable 

 to take the serum of the ' infected individual, estimate the 

 highest dilution with which it clumps the strain isolated, and 

 compare the result obtained with the effect of the serum on the 

 historic strains. The unknown strain is most likely to be allied 

 to that strain which is agglutinated by a similar dilution of the 

 serum used. Frequently, in the investigation of an organism, 

 it is necessary to inject it into an animal and study the 

 agglutinating properties of its serum on the infecting strain 

 and upon allied organisms. Here considerable information may 

 be obtained by the use of the absorption method. If from such 

 a serum, for instance, an unknown organism has absorptive 

 qualities similar to that of a historic Gaertner, its being named 

 a Gaertner bacillus would be justified. It is customary in any 



