BY R. GREIG-SMITH. 



185 



The boiled tap-water infusion, "b," was boiled three times 

 upon successive days, as is usual in preparing such culture- 

 media. Tests "a" were made eleven days before tests " b.'' Bv 

 an accident, the "b"' tests were incubated at 26° instead of 22" 

 as in " a," and this should be borne in mind when considering 

 the increase of the unboiled numbers over the start. 



The distilled-water numbers are considered to be normal, for 

 they show a gradual fall as the nutrients are weakened by dilu- 

 tion with water. The tap-water curves rise as the alkalinity is 

 weakened. The numbers of the raw tap-water test indicate 

 that, after all due allowances are made, it is of a toxic nature, 

 which is probably not entirely traceable to the alkalinity. The 

 numbers of the boiled- water test are not so pronounced, and are 

 much the same as liave been obtained in previous experiments 

 with bacterial and protozoon cultures. Tt should, however, not 

 be forgotten that many of these gave normal curves. 



A set of experiments were started when it had become evident 

 that, in all probability, the reaction had more to do with the 

 symptoms of toxic effect than anything else. In the endeavour 

 to get round any individual action of the bacteria, flasks of 

 dilute (0-1%) hay-infusion, made with distilled water to avoid 

 the action of tap- water, were seeded with an amcebse-culture, and 

 with a bacterial culture derived originally from a protozoa-free 



