BY R. GREIG-SMITH. 



737 



Fifty days later, the soils were again tested. No rain had fallen 

 for nearly a month, and the soil was consequently very dry. 

 Experiment xiv. 



The soil was nutritive, but, as shown by the effect of heat, it con- 

 tained thermolabile toxins. The subsoil was toxic, and the toxicity 

 was increased by boiling the extracts. One expected to find the soil 

 strongly toxic after the spell of dry weather, and the failure to 

 realise this, shows that an accumulation of toxic substances does 

 not occur in dry soils. The condition is similar to that which takes 

 place in the laboratory, with soils that are air-dried and stored. 

 They rapidly lose their toxicity and become nutritive. Inferen- 

 tially, a certain percentage of moisture is necessary for the forma- 

 tion of toxins in soils. 



From these experiments, it is seen that the demonstration of 

 toxins in soils depends upon obtaining a soil in which the toxins 

 preponderate over the nutrients, and in using an appropriate dilu- 

 tion in making the extracts. Equal parts of soil and water gener- 

 ally yield the most toxic extract. The toxins of the soil are thermo- 

 labile, while those of the subsoil used were thermostable. The 

 existence of two kinds of soil-bacteriotoxins are thus indicated. 



ix. The Formation of Toxins in the Soil. 

 The leaching out of the thermolabile toxins from soil by rain, the 

 occurrence of thermostable toxins in the subsoil, and the reappear- 

 ance of thermolabile toxins in the soil, make it appear probable 

 that thermolabile toxins are produced entirely in the soil, and do 



53 



