482 LEWIS DAVIS 



satisfactory results for routine production may be obtained by 

 the colorimetric method. The simple "comparator" of Hurwitz, 

 Meyer, and Ostenberg (1916) is recommended with standardized 

 boric acid-potassium chloride-sodium hydroxide mixtures of pH 

 = 8.0 and pH = 8.2, prepared as directed by Clark and Lubs. 

 Exactly 10 cc. of the neutralized bouillon are transferred to a 

 "comparison" tube, diluted with 10 cc. of distilled water, and 

 mixed well. Ten cubic centimeters of the mixture are now re- 

 moved to another tube and 0.5 cc. of an 0.02 per cent solution of 

 phenolsulphonphthalein in 50 per cent alcohol next added. Two 

 other tubes are prepared containing 10 cc. respectively of the 

 standardized pH = 8.0 and pH = 8.2 mixtures, with 0.5 cc. of 

 the phenolsulphonphthalein solution in each tube. The tech- 

 nique for comparison is as given by Clark and Lubs (1917). 

 Usually, the value will be found to very closely approximate 

 pH = 8.2. If, as may sometimes happen, the color in the tube 

 containing medium plus indicator is lighter than the compensated 

 pH = 8.0 standard, £> NaOH can be added directly to the former 

 tube until the desired tint is reached. The burette reading, mul- 

 tiplied by two (since the equivalent of 5 cc. of the medium is 

 used) gives the amount of ^ NaOH required to correct each 

 liter of broth. 



Experience has repeatedly confirmed the observation that 

 toxin of greater potency is obtained from broth contained in large 

 flasks than in small ones. Other conditions being equal, we can 

 expect to find diphtheria toxin of higher strength after growth in 

 a six liter flask containing three liters of broth than in a liter 

 flask containing 500 cc. Three liters of broth, dispensed into 6- 

 liter, Florence type flasks have been employed in producing the 

 toxin under observation. 



It has undoubtedly been noted in every laboratory engaged in 

 diphtheria toxin production, that, if for some reason, it becomes 

 necessary to resterilize media, there is a resultant diminution in 

 the strength of the final toxin. From experimentation in prog- 

 ress, to be reported upon in a later publication, it appears that 

 food accessory factors, possibly of a vitamine character, are con- 

 cerned in the production of diphtheria toxin. This makes it 



