245 



must have prepared his media in such a way that only pigeon 

 blood agar supported growth. 



Among other conditions concerning media it may be men- 

 tioned that Pfeiffer's bacillus according to Olsen (2) can tolerate 

 variations in the concentration of (added) salt from 0— 2<>/o, 

 while it is very sensitive to high concentrations of agar. If 

 the latter exceeded 3 <y the growth was greatly checked. 



According to Jordan and Rosenow it is important that 

 the air surrounding the culture plates be kept moist. Most 

 authors give no special information on this point. I myself 

 have not taken precautions to keep the air moist although 

 cultivation, as mentioned on p. 160, took place in an incubator 

 where the air was rather dry. All that was noticed was 

 that the growth seemed to be not so good once or twice during 

 frosty and windy weather and consequent increased dryness 

 and motion of the air in the incubator (but without altering 

 of its temperature). Even on partially dried up plates I have 

 also obtained good growth. 



Practically all the methods applicable to solid media can 

 also be used for liquid media for which reason the latter need 

 not be dealt with further. Fildes' method however was un- 

 suitable for broth in my experience (see also pp. 170 — 171). 



A special method of isolating Pfeiffer's bacillus is the ino- 

 culation into animals. Carpano (2) got constant positive re- 

 sults by injecting into the pleural cavity of a rabbit (after 

 preliminary cultivation in blood broth). The animal died in a 

 couple of days, when cultivations were made from the pleu- 

 ral exudate. Opie & others as previously mentioned (p. 37), 

 obtained more positive results in influenza by injecting sputum 

 into a mouse than by various other methods. 



That inoculation into animals would be suitable as a general 

 method for the primary isolation of Pfeiffer's bacillus does 

 not however seem probable, since a number of different in- 

 determinable factors would probably render the result uncer- 

 tain. In addition to this the quantitative relations of the 

 different bacteria in the original material could not be deter- 

 mined by this procedure. 



We have only spoken of the treatment of the material. The 

 question of the choice of material naturally entirely depends 



