oO The Non-persistence of Bacterio-toxins in the Soil 



IV. Soils that have been treated with toluene yield extracts capable 

 of carrying far higher numbers of bacteria than the extracts of un- 

 treated soils, and this has been assumed to be due to the destruction of 

 toxins. Experiments show, however, that these extracts contain a 

 larger amount of soluble nitrogenous organic matter, to which increased 

 bacterial development can with high probability be ascribed. They 

 differ only slightly if at all from untreated extracts or even saline solu- 

 tions to which a similar amount of material has been added (Curve IV). 

 The higher nutritive value of extracts of soils which have been treated 

 with volatile antiseptics has been attributed, without proof, however, to 

 an increased extractibility of the soil owing to the localisation of the 

 "agricere"; there is strong presumptive evidence that the soil flora and 

 fauna which succumb to treatment, are contributory sources. 



V. The investigation, therefore, gives no support to the view that 

 bacterio-toxins persist in, or are normal constituents of, uncropped soils. 

 Hence the suggestion cannot be accepted that partial sterilisation effects 

 are due wholly or in part to their destruction. 



Experimental. 



The experiments recorded below were arrived at with the following 

 seven soils, all of which have previously been under investigation either 

 in the laboratory, pot-culture house, or the field. 

 Allotment Soil. Similar to the garden soil used in partial steriUsation 



experiments. 

 Broadbalk Soils. ([<i] Manured and [h\ unmanured): under continuous 



wheat experiments since 1843. 

 Woburn Soil. (Unmanured) : imder continuous wheat experiments 



since 1876. 

 Millbrook Soil. Light, sandy field soil. 

 Chelsea Soil. From Chelsea Physic Garden. 



Harpenden Common Soil. Heath soil, distinctly acid in reaction. 



In the examination of these soils two deviations from Greig Smith's 

 method have been introduced. In the first place we have adopted the 

 use of standard physiological salt solution (0-8 % pure sodium chloride 

 in distilled water), instead of 0-2-0-5 % salt in distilled or tap water; 

 secondly, and in order to permit of the growth of the test organisms 

 being more closely followed, enumerations of bacteria were made directly 

 after inoculation of the extracts as well as after 4, 8, 24, 48, and 72 hours, 

 instead of the single count frequently recorded by Greig Smith. With 



