IX 



1) The results are permanent. 



2) It is possible to distinguish certain varieties of 

 the coli and typhoid groups, which cannot be differ- 

 entiated by the method, as formerly used. 



3) The results are not dependent to the same extent 

 upon the degree of acidity and alkalinity of the me- 

 dium. 



Table XIV furnishes proof for the last statement. 

 Though in the results, there are slight variations both 

 in the ultimate shade of colour and the time of its 

 production, the reaction never failed to occur whether 

 the bouillon was strongly acid (1.25 H 2 So 4 ), intensely 

 alkaline (1.25 NaOH) or of intermediate degrees of 

 acidity and alkalinity. 



Whatever may have been the colour of the medium 

 before inoculation, whether yellow or red, (i. e. acid, 

 neutral or alkaline reaction) B. typhosus always pro- 

 duced a yellow and B. Coli a magenta or red colour. Thus 

 the colour changes here, are vice versa to those of the 

 old method, where B. coli changed the colour to canary 

 yellow, and B. typhosus left it unaltered, red. 



As it is seen from our experiments (Table XVII and 

 XVIII) the reaction, though it does occur in 4% and 5% 

 lactose bouillon, is observed with the greatest rapidity 

 and most distinctly in 8% lactose bouillon. Thus it 

 is evident, that the latter % lactose is the best, and, as 

 far as our experience goes, the only one to be recom- 

 mended. A lower / than 5% should never be used, as 

 the reaction either does not occur or becomes variable 

 and unreliable as for example in 1% and 2% lactose 

 bouillon (Table XII and XIII). 



- 27 



