30 THK SLIME OF THE HOUSEHOLD BATH-SPONGE, 



original colonies were by no means pure. Some contained Radio- 

 bacter, which grows best in combination with other l:)acteria. 

 ■Others contained Sareinpe and short rods. A second smearing, 

 followed by the usual dilution-method of plating bacteria, enahled 

 pure cultures to be obtained. There were several kinds of 

 Rhizobia, some stouter some thinner, but all produced what 

 appeared to be a similar kind of slime. One of the organisms 

 was taken for further work. A bacterium forming a tough 

 slime was taken as a second microbe because it was probable 

 that this was the cause of the slime being so adhesive to the 

 sponge. The latter micro-organism appeared to lie midway 

 between Rhizobium and Bac. alatus Greig-Smith. 



The two bacteria were smeared upon a synthetic agar-medium 

 containing phosphate and aspatagin in combination with various 

 sugars, etc., in order to see which sugar to use in preparing a 

 quantity of the slime. The tough slime-bacterium refused to 

 grow, while the other (Bac.ii.) produced luxuriant slimes from 

 saccharose, levulose, and dextrose, and no slime from lactose, 

 glycerin, dextrin, maltose or mannit. Another test was made, 

 using bean-extract as a basis, with the same result as far as the 

 loose slime-bacterium was concerned. The tough slime-bacterium 

 (Bac. i.) produced a luxuriant slime in the presence of saccharose, 

 levulose, glycerin, mannit and maltose, but failed to respond to 

 lactose and dextrin. 



These slimes were used to prepare films for microscopical 

 observation. Both bacteria were Gram-negative and varied ver}' 

 much in size and shape, not only with regard to different sugars, 

 but even with each sugar; and, upon the same film, great varia- 

 tions were seen. 



Witli Bacillus i., a smaller or larger oval shape predominated. 

 From glj'cerin, the cells were comparatively small and stained 

 deeply; from the other media, they stained diffusely. From 

 maltose and dextrose, types similar to Bac. alatus* with winged 

 xjapsules were common. Levulose and mannit favoured the 



These Proceedings, 1905, p.570. 



