BY R. GREICi SMITH. / ;> 



nor did they include any other active organism than in the tests 

 tabulated above. The method employed for separating the 

 bacteria was the ordinary one of infecting molten agar with a 

 few crystals of the sugar and pouring the agar into a Petri dish 

 after the sugar had dissolved. From several plates thus prepared^ 

 one containing from 15 to 30 colonies was taken, and every 

 colony was infected into nutrient agar and into saccharose peptone,, 

 prepared as described in a preceding paper (antea, p. 592). When 

 in about three or four days, the saccharose peptone in some of 

 the tubes had become white and opaque, all the saccharose peptone 

 cultures were tested for gum by precipitation with alcohol, and for 

 reducing sugar with Fehling's solution. The active bacteria 

 were thus indicated, and the further identification of the bacteria 

 was made from the agar culture. 



With regard to the bacillus, both were types of the organism 

 already described as the gum bacillus, Bac. levaniformans, which 

 has been shown to rapidly cause the inversion of saccharose in 

 solution. The inert bacteria had no action upon sugar, and were 

 not investigated further. The streptothrix slowly inverted 

 saccharose. The aspergillus was a race of Aspei-gillus glaucus^ 

 which readily formed aerial ascospores. The inverting action of 

 Aspey'gillus is well known. 



The second sample was a soft-grained refined but not quite 

 white sugar. The colonies that developed on the infected agar 

 plates were found to consist of the gum bacillus alone. One 

 portion contained the derived type, another contained a majority 

 of the derived type, with a few colonies of the normal bacillus. 



The third sample was similar to the second, but of a slightly 

 darker shade. Compared with the second sample, it was nearly 

 sterile, but the few colonies that developed consisted of the 

 derived and the normal type of Bac. ievaniformans in the ratio of 

 about 1 to 6. 



A fourth sample of white crystalline refined sugar contained 

 the normal bacillus ; only one colony of the derived type was 

 .obtained from one of the portions. 



A sample of raw sugar solution of the consistency and appear- 

 ance of thin molasses was, when received, slowly fermenting, that 



