228 



factor}', till finally it was impossible to obtain a good growth 

 on it. This was nol a transient phenomenon; on repeating 

 the test in 1921 and 1922 with the blood of a number of diffe- 

 rent horses only a weak growth appeared. This happened 

 whether the inoculation was made from cultures which had 

 been grown for a long time (on Fildes agar) or whether it 

 was made directly from the throat of a person (March 1922) 

 who had been a Pfeiffer's bacillus carrier for about two months 

 in consequence of an attack of influenza. The blood was taken 

 from the „serum horses" of the Institute (principally diphtheria). 

 The failure of the blood as a nutritive medium for Pfeiffer's 

 bacillus might be attributed to the experiments on the increase 

 of antitoxin by the injection of manganese salts which had 

 been gradually started on a large scale during the first half 

 of 1920. But in the last test in March 1922 blood was taken 

 from 3 horses which had not received such injections (they 

 had for some time been immunised against Staphylococcus, 

 Streptococcus, and Pneumococcus respectively), without it pro- 

 ving more active. 



I have thus not succeeded in probing the mystery of this 

 peculiar phenomenon that apparently exactly the same medium 

 which for a long time proved to be ideal, subsequently per- 

 sistently failed, but it is none the less important to report 

 the experience because it is of great significance in judging 

 the technique for the cultivation of Pfeiffer's bacillus. Several 

 authors maintain for instance that in criticising the work on 

 this bacterium, — particularly when there are negative fin- 

 dings — it is of great importance that the investigator gives 

 accurate information about how the media were prepared. This 

 is naturally always desirable, but the important point is not 

 whether one or other of the large number of media detailed 

 below, each useful in itself, was employed. The actual proof 

 that Pfeiffer's bacillus can always grow on the medium used 

 is by far the most important point. 



It was chiefly the „V"-function of the blood that failed 

 because the latter had consistently been quite suitable for the 

 symbiosis reaction (p. 120). It seemed however, as if the same 

 effect could be obtained at first with 1 /4~~ 1 /2°/oo dissolved blood 

 corpuscles, as later on with 1— 2% . 



All the horse and sheep blood I have tested was completely 



