1. The. addition of purified acetylmethylcarbinol to sterile skimmilk, adjusted to an acidity and a temperature satisfactory for the rapid production of the carbinol when the citric acid-fermenting streptococci are present, did not result in the formation of appreciable amounts of diacetyl in 48 to 72 hours. The results were the same when carbon dioxide, hydrogen, nitrogen, or oxygen was bubbled through the milk as when no gas was used.

2. The production of diacetyl in acidified skimmilk cultures of the citric acid-fermenting streptococci was definitely influenced by bubbling various gases through the freshly acidified cultures. Oxygen regularly gave a higher yield of diacetyl than the control through which no gas was bubbled, while carbon dioxide, hydrogen or nitrogen gave lower yields. With all of the gases there was commonly a greater production of diacetyl, as well as acetylmethylcarbinol plus diacetyl, when the cultures were acidified with a mixture of citric and sulfuric acids than when acidified with sulfuric acid alone.

3. Various gases had the same general effect on the production of diacetyl in butter cultures as in pure cultures of the citric acid-fermenting streptococci, but the actual quantities of diacetyl formed appeared to be smaller with the butter cultures than with the pure cultures. When oxygen was bubbIed through butter cultures prepared with various amounts of added citric acid, the yields of both diacetyl and acetylmethylcarbinol plus diacetyl were roughly proportional to the amount of citric acid added.

4. When acetylmethylcarbinol was added to milk and the milk inoculated with S. lactis, analyses after incubation showed no evidence of an oxidation of the carbinol to diacetyl.

5. The data obtained indicate that the oxidation of acetylmethylcarbinol to diacetyl in a butter culture is due to the activity of the citric acid-fermenting streptococci rather than to a direct chemical oxidation.