632 CONTRIBUTIONS TO OUR KNOWLEDGE OF SOI L-FERTILITY, Xlll., 



ents into petroleum-insoluble and non-preservative substances 

 upon being boiled with water.* 



The presence of the toxin in the soil depends upon the natural 

 conditions to which the soil has been subjected. Bain, for ex- 

 ample, washes it out, and the extract, instead of being toxic, is 

 nutritive. The decay of the toxin in the soil is accelerated by 

 dry conditions, so that in dry, as in wet weather, we cannot 

 expect to find the extracts toxic. There is a medium between 

 wet and dry conditions which, while probably not influencing 

 the actual formation of the toxin, yet facilitates our being able 

 to demonstrate its presence. 



There is still much to be learned regarding the nature of the 

 toxic solution and of tiie production of toxins. Several matters 

 relating to these have been investigated, and are here recorded. 



The relation oj the quantity of Water used in extractimj the 

 Soil — It is curious that the quantity of water used for extrac- 

 tion should have an influence in enabling the toxic effect to be 

 evident, and although this has been shown to occur,! yet the 

 diflliculty of explaining the phenomenon caused another test to 

 1)6 made, wh^n time should have modified any personal factor. 



A garden-soil, seven days after heavy rain, gave the following 

 result : - 



Bacteria in control-test 

 Bacteria in extract ... 



Cells of Bac. proditjiosus in 

 20 lionrs at 22°. 



100 

 40 



Two days later, when the soil contained 7-05% of moisture, it 

 was shaken up with varying amounts of water and extracts pre- 

 pared. These were seeded with a suspension of Bac. j^fodigiosus, 



* Chapman, Proc. Chein. Soc, xxix. (417), p. 182. 

 t These Proceedings, 1914, pp. 729-731. 



