DIPHTHERIA TOXIN AND ANTITOXIN 481 



preparing the toxin under observation was Bacteriologic Peptone, 

 Parke, Davis and Company, the composition and properties of 

 which have been described in a previous article (Davis, 1917). 

 Twenty grams of peptone and 5 grams of sodium chloride were 

 added to every liter of beef infusion prepared as above, dissolved 

 in the cold and then brought to a boil in the steam kettle to insure 

 thorough solution. 



Considerable uncertainty appears to exist as to what is the 

 most satisfactory initial reaction for diphtheria toxin bouillon. 

 Nearly all of the previous investigators, including Roux and 

 Yersin (1888), Spronck (1898), Madsen (1897), Park and Wil- 

 liams (1896), Smith (1899) and Lubenau (1908) have employed 

 either neutralization with litmus, which is crude at best, or "hot 

 titration" with phenolphthalein, a procedure which is admittedly 

 fallacious. The writer (Davis, 1918) in a paper on " Hydrogen 

 ion concentration and toxicogenicity determinations with Bad. 

 diphtheriae" has shown that potent toxin is produced in bouillon 

 by Corynebact. diphtheriae only when the initial reaction falls within 

 a certain zone of alkalinity, included within the hydrogen ion 

 concentration limits of about 7.0 X 10~ 8 (pH = 7.2) to about 5.0 

 X 10 -9 (pH = 8.3). The maximum degree of potency, however, 

 is obtained when the reaction of the broth comes within the nar- 

 row limits of pH = 8.0 to pH = 8.2. This may be readily and 

 consistently obtained by following the procedure given below for 

 adjusting the reaction. 



1. Transfer 10 cc. of the heated broth to a small Erlenmeyer 

 flask and dilute with about 40 cc. of cold, distilled water. Add 

 0.5 cc. of a 1 per cent solution (95 per cent alcohol) of phenol- 

 phthalein as indicator and titrate to a deep pink color against 

 an -& NaOH solution. The latter is preferably prepared when 

 required as an exact 1/100 dilution of a — stock solution. The 

 amount of the strong (if) solution required to neutralize per litre 

 of medium is given directly by the burette reading. Bring to a 

 boil again in the steam kettle, and estimate the hydrogen ion 

 concentration. 



2. While the more accurate hydrogen electrode method is 

 desirable for comparative and standardization purposes, equally 



