163 



horse blood corpuscles, l<y glucose, Vioo % decolorised China blue, 

 and V500 % phenol red were added. The mixture was filtered through 

 a Berkefeld filter and run into test tubes which were inoculated 

 with the strains on p. 110. The tubes were kept under observation 

 in the incubator for a week. (For the first three days they were 

 kept approximately horizontal, and afterwards, vertical). The average 

 growth was rather weak. The only strains that decidedly changed 

 the colour of the medium were H 274 and H 323, which produced 

 weak acid reactions. The same result was obtained when the ex- 

 periment was repealed. Since the acid reaction appeared however 

 when the medium was prepared with broth fermented with Bacil- 

 lus coli without the addition of glucose (in which medium 2 strains 

 of Bacillus haemoglobinophilus canis with strong fermentative pro- 

 perties, did not give any change of colour), there can hardly be 

 any question of glucose fermentation. 



The strains Me 2 and 3, which were later on tested separa- 

 tely for glucose fermentation in a liquid medium, gave a weak acid 

 reaction (while a strain of Bacillus haemoglobinophilus canis which 

 was inoculated for comparison, gave a very strong reaction). Whether 

 it was a true fermentation reaction in this case, was not investigated. 



On the whole therefore no certain and even moderately 

 marked power of fermentation could be demonstrated either 

 on solid or in liquid media. 



As mentioned on pp. 49 — 51 other authors have found a di- 

 stinct though rather „capricious" fermentative power by using 

 suitable substances, without being able to determine precisely 

 what properties a medium' must possess in order that Pfeiffer's 

 bacillus may exert its power of fermentation. There is absolu- 

 tely no reason for concluding from the negative result of my 

 fermentation experiments, that my strains of Pfeiffer's bacillus 

 were of a different nature from those of the authors cited. 



According to Gosio (2) lack of the power to attack ar- 

 butin is characteristic of Pfeiffer's bacillus as opposed to 

 bacteria which it might be confused with. I therefore (March 

 1921) prepared Fildes agar containing 1% arbutin (in auto- 

 claved watery solution) and inoculated these plates with all 

 the stock strains of Pfeiffer's bacillus, — practically the' same 

 collection as that given on p. 110, together with Me 2 and 3. 

 In no case did Pfeiffer's bacillus produce the slightest change 

 of colour. For the sake of comparison many other species 

 of bacteria were inoculated on the same medium. Most of 

 these gave a positive reaction, which showed itself as a marked 



11* 



