BY R. GREIG-SMITH. 633 



and incubated at 22°. Next day, plates were prepared and the 

 bacteria counted. 



The results are in keeping with previous tests, in which it was 

 shown that the general tendency was for the extract to be toxic 

 when comparatively small quantities of water relatively to the 

 soil were used ; with larger quantities, the extracts became 

 nutritive, but again became toxic at about the 1 :1 ratio. 



It is difficult to use less than 30 c.c. of water per 100 grams 

 of soil in preparing an extract. In order, therefore, to obtain 

 information regarding the possibility of toxic action in soil-water, 

 the test-bacteria were added to soil directh' with varying quan- 

 tities of water. The test-bacteria are thus in competition with 

 the soil-bacteria, and the conditions are not the same as when 

 extracts are used. At the same time, it has been shown that 

 the presence of soil- and test-bacteria in extracts do not affect 

 the general trend of the toxic or nutritive effect, as indicated by 

 the decrease or increase of the test- bacteria. So far as the test- 

 bacteria are concerned, the soil-bacteria are relatively inert, 

 during the time of the experiment. 



Twenty-gram portions of soil were put into petri-dishes, and 

 1 c.c. of a suspension of Bac. prodigiosus was added to each along 

 with varying quantities of water. The soils were thoroughly 

 mixed with the water, and the dishes were put into a damp- 

 chamber and incubated at 28° for 20 hours. They were then 

 transferred to flasks of sterile water, shaken, diluted, and ^qC.c. 

 smeared on agar-plates, dried at 37°, incubated, and counted. 



For further information regarding the soils, an extract was 

 made, and the numbers E/C were obtained by dividing the in- 



