HARPER F. ZOLLER AND W. MANSFIELD CLARK 327 



flasks while evacuating. After pumping out the bulk of the air 

 with the Geryke pump till the Plucker tube discharge showed a 

 declining spark, the mercury-vapor pump was started and it was 

 allowed to pump for 20 minutes more or until the Plucker tube dis- 

 charge showed a continued decline in brilliancy. The flasks were 

 sealed off while the pumps were in operation. 



Separation of the Volatile Acids from the Cultures. — 10 per cent 

 sulfuric acid was added to the cultures until a drop of thymol blue 

 showed a distinct red; this represented a pH of 1.0 or less. Steam 

 formed by boiling distilled water in the presence of barium hydroxide 

 was led into the flasks holding the cultures. The flasks in turn were 

 connected to vertical condensers through which ice-cooled water 

 circulated. The condensers were arranged vertically to facilitate 

 the delivery of the condensed fatty acids. A small piece of paraffin 

 was placed in the cultures to overcome frothing during distillation. 

 The volume of liquid in the culture flasks was held constant by means 

 of a screened flame. Fractions of 500 cc. each were caught and 

 titrated with 0.1 N NaOH to phenolphthalein. The distillation was 

 continued until the last fraction required no more than 2 cc. of 0.1 n 

 NaOH. Usually ten to fourteen fractions were collected. These 

 were evaporated in toto and made up to a definite volume. 



Distillation was conducted upon a like volume of sterile medium 

 to determine the factor for carbon dioxide, etc. 



Quantitative Analysis of the Volatile Fatty Acids. — The modified 

 Duclaux method published by Gillespie and Walters^ was followed 

 in the main in the determinations. It was found advisable to deter- 

 mine new constants for formic, acetic, propionic, and normal butyric 

 acids with the distillation apparatus employed. The stifl was of 

 Jena glass throughout and contained a small, closed end, inverted 

 tube in the solution to insure against excessive spattering and to 

 produce a steady flow of vapor. The constants found did not depart 

 radically from those obtained by Gillespie and Walters. The stiU 

 flask was heated on the sides and bottom with electrically heated 

 coils. 



An aliquot of the total volatile acid concentrate was employed 

 which contained from 60 to 80 cc. of 0.1 n volatile acids. 



1 Gillespie, L. J., and Walters, E. H., /. Am. Chem. Soc, 1917, xxxix, 2027. 



