VIII, B, 6 Barber: Dysentery Bacilli 549 



maltose used contained a trace of glucose, a third series was 

 carried out, using the same media as in the second but in 

 addition a broth and an agar made with a maltose shown by 

 chemical test to be pure. Ten daily transfers, and 3 at 

 intervals of three days, failed to alter the character of the race. 



In addition, tests were made of the above cultures after 

 incubating four, seven, fifteen, and nineteen days. Here a ten- 

 dency to an increase of acid producers was noted in the maltose 

 media, while in the sugar-free media the strain remained un- 

 changed. 



In summary, it may be said that there is evidence that in 

 this race of dysentery bacillus the presence of maltose does not 

 materially increase the proportion of maltose fermenters except 

 after long contact with the sugar in the same tube. 



To compare the effect of substances other than maltose, a 

 typical red and a typical blue race were sown on slants of the 

 following litmus agars: Glucose, levulose, lactose, saccharose, 

 raffinose, glycerine, dextrin, inulin, salacin, erythrin, and dulcite. 

 All tubes were alkaline after three days' incubation except 

 glucose, levulose, and glj'^cerine, which were cherry red in both 

 the red and the blue races. After seven days' growth, each 

 culture was transferred to a new agar of the same kind. The 

 colors remained the same as after the first transfer. On 

 transfer back to maltose, the originally red strain gave red 

 cultures in all cases and the originally blue strain gave blue. 

 Apparently growth on other substances does not affect the fer- 

 mentative character of the two races with respect to maltose; 

 and, in a short time, at least, neither race acquires the power 

 of attacking a new substance. 



For confirmation, a red and a blue strain, neither of which 

 had been grown on a medium containing maltose or other sugar, 

 were sown on the media used in the first test, and, in addition, 

 on mannite, galactose, and amygdalin-litmus agars. Both races 

 gave the same reaction on all media. 



Some further experiments were made to determine the com- 

 position of the secondary colonies. These colonies have a large 

 proportion of irregular cells, many of which are coccoid or 

 resemble in form small yeast cells. Many of these cells were 

 isolated and grown to determine their nature. One protocol maj' 

 serve to illustrate the nature of these experiments: 



From the top of a large secondary colony eight days old, 10 

 normal cells were isolated, of which 3 grew; and 14 coccoid 

 forms, of which 6 grew. Grown in hanging drops of ordinary 



