636 CONTRIBUTIONS TO OUR KNOWLEDGE OF SOIL-FERTILITY, xiii., 



with a suspension of Bac. prodigiosus, incubated for 20 hours at 

 22°, and counted by the plate-method. 



The water-control was the same dilute potassium sulphate 

 which was used for making the extracts, and, like the extracts, 

 it had been filtered through porcelain to ensure sterility. This 

 is necessary because sterilisation by heat alters the toxic or 

 nutritive effect of the water. 



While the extract from the 5% soil was variable, those from 

 the 15% and 20% showed a steady increase of toxicity. The 

 10% extract was always toxic. 



Having determined that about 10% of moisture, or one quarter 

 of the H.W.C., was best for obtaining a soil furnishing a toxic 

 extract, an attempt was made to determine the optimum temper- 

 ature. For this purpose, the soil was taken soon after rain, 

 when it should have been either feebly toxic or nutritive. 'J'wo 

 hundred gram portions were put into bottles, which were corked 

 and incubated at four different temperatures for varying times. 

 At the start, the soil contained 11-05% of moisture. 



Considering the results generally, it is seen that 15° to 22° 

 give very similar results, and that 28" is the optimum tempera- 



