242 BACTERIAL ORIGIN OF GUMS OF ARABIN GROUP, 



products such as molasses or the liquor from potato starch, I tried 

 the eifect of adding molasses to agar. The sample of molasses 

 contained the following constituents: — cane sugar 38%, fruit sugar 

 9%, other organic matters 16%, soluble ash 10%, and water 27%. 

 The most favourable yield of slime was about half of what 

 would have been obtained had saccharose and asparagine been 

 employed. Either the dextrose or the excessive saline matter 

 prevented a greater production, but it is also possible that the 

 nitrogen was deficient. Molasses contains from 0'013 to 0-027% 

 of nitrogen, which, being calculated to asparagine, would mean 

 0-07 to 0"144% ; the medium with 5% of molasses would thus 

 contain the equivalent of from 0-0035 to 0-0072% of asparagine, 

 which is undoubtedly too small. Experiments were made with 

 urea on account of its cheapness industrially, and these showed 

 that nothing was to be gained by adding it to the molasses. The 

 addition of potato-juice in varying quantities was also inopera- 

 tive. The presumable deficiency of nitrogen could not therefore 

 play a part in the lessened yield, and the other constituents of 

 the molasses must be considered as being the active agents. 

 With small quantities of molasses (2% and under) an ordinary 

 infection as by smearing the plates in the manner already de- 

 scribed was sufiicient, but with larger quantities a mass-infection 

 was absolutely necessary to obtain a yield of slime. This was 

 done by smearing a small loopful of culture (made upon saccha- 

 rose-potato-agar at 30°) over an area of about a quarter of an 

 inch diameter in the middle of the plate. In 2-i hours the 

 slime which had formed was spread over a wider area of about 

 an inch diameter. In another 24 hours the area was increased 

 to 2 inches, and in another day the slime was sufficient to enable 

 the whole plate to be smeared. The slimes were scraped from 

 the plates on the eighth day and weighed. In the following table 

 four experiments are tabulated :— 



