105 



agar, with great regularity. A slight growth was noted on 

 this medium 1 in a single case (I 71) only. 



Later on my attention was directed to the advisability 

 of controlling the ascitic agar more accurately, and by degrees 

 I came to insist on the following requirements for an 

 ascitic fluid suitable for the verification of 

 Pfeiffer's bacillus: (1) The ascitic fluid should not in 

 any case give a marked reaction with the benzidin test. This 

 test is carried out as follows: To 1 vol. of a freshly prepared 

 6— 10o/o solution of benzidin in glacial acetic acid, is added 

 1 vol. of oxydol (or water + perhydrol), and after mixture, 

 1 vol. of ascitic fluid. (2) The ascitic agar when finished 

 should be neutral or weakly alkaline to azolitmin paper,* 

 as it has been shown that this reaction is most favourable for 

 the growth of Pfeiffer's bacillus. (Ordinary agar may con- 

 veniently react rather more alkaline to azolitmin paper). To 

 adjust the reaction, the marked alkalinity of the ascitic fluid 

 is usually reduced by adding an appropriate amount of hydro- 

 chloric acid. (3) Agar plates, with and without ascitic fluid, 

 are inoculated with various cultures of the whooping-cough 

 bacillus, Gonococcus, and Meningococcus, preferably such as 

 have only been grown for a few generations in vitro. A much 

 richer growth should appear on the ascitic agar than on that 

 without ascitic fluid (in so far as any growth takes place at 

 all, which will never be the case with the whooping-cough 

 bacillus when it has only been subcultured a few times, but 

 in the case of Gonococcus and especially of Meningococcus 

 growth may occur on ordinary agar). 



Lastly as a fourth control it can be insisted that Pfeiffer's 

 bacillus is unable to grow on the medium. As a number of cul- 

 tures are usually examined together they will to a certain 

 extent serve as controls to one another. 



The three first requirements were only regularly satisfied 

 from July 1920, an ascitic fluid (denoted, for short, by ,,1,023") 

 sp.gr. 1,023, from that date onwards being most frequently 

 used, which only gave a very slight blood reaction. The reac- 

 tion was adjusted as described, every time a batch of ascitic 



* For the preparation of this, and for other matters relating to 

 the reaction of media see pp. 225—232. 



