394 NATHAN BERMAN AND LEO F. RETTGER 



The methyl red test of Clark (1915) was employed because it 

 gives an index of the 'true reaction' of the test media. This test 

 has proven itself to be of much value. 



The behavior of Bacillus coli in the media containing the 

 glucose and lactose is of particular interest. The biuret and 

 Sorensen tests gave essentially the same figures for the control 

 medium (plain peptone broth) as in previous experiments. The 

 indol appeared on the third day, and remained throughout the 

 course of the experiment. The final reaction was practically 

 neutral to phenolphthalein. In the media containing glucose 

 the results were quite different. The hydrogen ion concentra- 

 tion was markedly increased, both glucose media (0.2 per cent 

 and 0.4 per cent) becoming methyl red positive. In the tubes 

 containing 0.2 per cent of glucose the sugar was completely 

 fermented by the third day. Little metabolism seemed to take 

 place after this. The titratable acidity remained high. The 

 Sorensen values were almost the same as for the controls. No 

 indol tests were obtained in the glucose tubes even after four 

 weeks. Although 0.2 per cent glucose was used up in three days, 

 0.5 per cent was not completely fermented in a period of four 

 weeks. The biuret tests indicated that no "peptone" digestion 

 had taken place in the sugar-containing tubes. 



The behavior of Bacillus coli in the lactose media was about 

 the same as in the glucose broth. In the tubes containing 0.2 

 per cent lactose the methyl red test suddenly changed from posi- 

 tive to negative. The indol test became positive with the 

 decrease in hydrogen ion concentration. It appeared as if 

 conditions gradually became favorable for growth, and that 

 the rate of bacterial development increased as the acidity was 

 neutralized. The coordination of the different tests is of special 

 significance. 



Fermentable sugar exerts the same influence on the metab- 

 olism of Bacillus vulgaris as on that of Bacillus coli. In the 

 medium containing no carbohydrate the protein is completely 

 broken down by the former organism, however. Within twenty- 

 four hours .there was appreciable indol formation. Indol was 

 not produced, on the other hand, in any of the glucose-bouillon 



