BY R. GREIG SMITH. 445 



of galactosazone together with traces of a vitreous yellow body. 

 In the dilute alcoholic filtrate from the bulk of the galactosazone 

 a small quantity of galactosazone was also obtained. No other 

 osazone was present, and thus galactose is the sole product of the 

 hydrolysis of the gum, which is therefore a galactan. 



When first obtained from the tissues of the plant, the bacte- 

 rium grew as brittle, translucent white colonies which upon sub- 

 cultivation became transparent and gelatinous and appeared, 

 microscopically, to consist of an aggregate of gelatinous clumps. 

 To make sure that the culture was pure, a tube of glucose- 

 gelatine was infected and incubated at 30° for three or four 

 hours to obtain a distribution of the bacteria throughout the 

 medium before dilutions were made and plates were prepared. 

 The plates revealed the presence of another organism which pro- 

 duced 3^ellow slimy colonies on glucose-gelatine. To thoroughly 

 eliminate this impurity, a gelatinous colony was inserted into 

 nutrient bouillon which was incubated at 30° for 24 hours. 

 Then dilutions were made in glucose-gelatine and plates were 

 prepared. The colonies that developed were all of the gelatinous 

 type, and a pure culture had apparently been obtained. After 

 some time I noticed the impurity again appearmg in my cultures, 

 but a passage through bouillon enabled the culture to be again 

 purified. Once more, after the lapse of a month or so, the 

 impurity again appeared. On considering the matter, it ap- 

 peared to be extremely probable that the bacteria were one and 

 the same and that the difierent appearance of the colonies was 

 due to an alteration in the solubility of the gum. Experience 

 had indicated that the change could be brought about quickly by 

 incubating a glucose-gelatine culture at 30'. 



To prove the matter, the following process w^as adopted. The 

 culture was used for infecting nitrate-bouillon, which was, after 

 a period of incubation, employed for preparing dilutions in 

 glucose-gelatine. From the glucose-gelatine a pure colony of the 

 gelatinous bacterium was abstracted, and again nitrate-bouillon 

 was infected and glucose-gelatine plates were prepared. The 

 colonies were entirely of the gelatinous type. Portions of one 



