614 THE GUM FERMENTATION OF SUGAR CANE JUICE, 



same kind. Fortunately the alteration or reversion to the 

 normal type occurred in the course of a few crops or transfers 

 (erroneously called generations) with one of the races which was 

 considered as being typical of group jBjB, and the identity of the 

 groups was made evident. While with this race the change 

 occurred by the 8th transfer, two other similar /3/3 races had 

 not changed by the 40th transfer. From a glistening lobular 

 yellow raised stroke, it altered to a dull white amoeboid 

 expansion spreading over the greater portion of the agar 

 surface. Concomitant with this alteration, the granular or 

 spongy structure of the bacterial protoplasm changed, and when 

 stained the rods appeared narrow and homogeneous. The nature 

 of the growth in bouillon, and on potato, also changed. In 

 short, the race had become identical with those of group ^. The 

 reversion is shown in a somewhat diagrammatic form in the 

 plate which accompanies this paper. 



Why /3/3 should be so different from /3 can only be explained 

 by assuming that it has been caused by the physical conditions 

 that lla^■e obtained during the process of manufacture of the 

 sugar. In the various samples of sugar the three groups of 

 bacteria were generally found in varying proportions. This 

 might point to the sugar having been infected at different periods 

 of the manufacturing process, for if it were otherwise, only one 

 group of bacteria would be found in each lot of sugar. As a 

 rule, however, the prevailing bacteria were of the ^ kind. 



The cultural variation of the race does not have much effect 

 upon the production of gum. With one race of the /i^ type, 

 which maintained its character through 24 transfers, the amount 

 of gum formed in 12 days at 37° was found to be 26-7 grms. for 

 100 o-rms. saccharose originally taken, the culture medium being 

 identical with that already described. 



Bacillus levaniformans, n.sp. 



Sha2Je, &c. — The films prepared from cultures on nutrient agar 

 fchow the organism as a rod with rounded ends ; the individuals 

 occur singly and in chains. The dimensions of the rod vary con- 



