7g 



ROPINESS IN WATTLK BARK IMFDSIONS, 



a wavy structure as it' the tlatteued colony lia<l an undulating' surfaci'. No further 

 change hail occurred by the sixteenth day. 



The diffuse phase, Bl, was grown in a glycerin phosphate medium to see in 

 which dircclion an alteration would occur. On the 17th day, H'J was present, but 

 it had disappeared l)y the 23rd. In another test tiie bacteria on the Otli day 

 were all transition forms of B2, and on the 12th there were 75 % of Bl, 15 % 

 of B2, and 10 % of the transition form noted on the Gth. 



The results seem to show that dextrose tends to maintain the phase 112 and 

 glycerin the phase Bl, but this is influenced by the nature of the salt. 



Ordinary bouillon maintains the phase B2 and that is why the existence of 

 a phase lil was not suspected for a loiii;' time. II is ilit'lVvent with A2 which is 

 slowly changed to Al in bouillon. 



After finding that levulose favoured the production of the ropy substimce 

 more than dextrose, the experiment with the various salts was repeated upon two 

 occasions using levulose as the carbohydrate. The jihases Al, A2 and B2 were 

 tested, and, for the sake of com]iaiison. a further set of salts with dextrose was 

 used for B2. The experiments duplicated one another, and the salient points of 

 both are embodied in the table. Both sets of Al (levulose) and B2 (dextrose) 

 are omitted because they were almost entirely negative; the second set of A2 

 was the same as the first set. 



After sterilisation, the dextrose medium in the first experiment bad a reaction 

 of — 1.5° to methyl-red and -(-3° to phenolphthalein, and this was sufficiently 

 {•Ikaline to give precipitates in the tubes containing the salts of lime. In the 

 course of the experiment, the reaction was tested on the seventh day in tests 4 



Table xix. — Levulose ivith Various Salts. 



