46 The Non-persistence of Bacterlo-toxhis in the Soil 



toxin hypothesis involves some difficult and not altogether consistent 

 assumptions as to their properties. It requires the assumption amongst 

 others not only that a number of toxins actually exist in the soil, but 

 that while some are decomposed at temperatures below those which 

 hitherto have been found necessary for toxin destruction, others 

 are sufficiently stable to withstand the relatively high temperatures 

 ordinarily adopted in partial sterilisation. 



Furthermore the toxins are supposed to be so unstable as to be 

 destroyed by aeration or exposure to light but at the same time they 

 are stable enough to resist decomposition in the soil. Their properties 

 vary in moist and in air-dry soils and while they are sufficiently potent 

 to inhibit growth in untreated soils they are no longer able to do so in 

 those treated with volatile antiseptics, i.e. w^here the "agricere" is 

 distributed. It becomes necessary to assume that hydrogen sulphide, 

 phenol, etc.. destroy them and that the process of salting out in the 

 soil either decomposes them or renders them permanently insoluble. 

 Apart from the various assumptions which the toxin hypothesis neces- 

 sitates, it is still clear that the extracts of certain soils prepared as 

 suggested by the above investigators produce effects on introduced 

 organisms which are indicative of injurious bodies. Whether or not 

 these bodies belong to the true toxins cannot be decided until more 

 work has been done but the results appeared to be sufficiently striking 

 to warrant a repetition of the work with English soils, not only from 

 its relation to the growth of soil bacteria, but also in regard to the 

 changes induced by partial sterilisation. 



I. We have accordingly studied the rate of growth of bacteria in 

 the extracts of a number of different soils and compared it with the 

 rate of growth in physiological salt solution. In some cases a temporary 

 depression occurred, but this was followed by a definite recovery wnthin 

 72 hours from the time of inoculation. From two of the poorest soils 

 extracts were obtained which showed marked depression of the test 

 organisms and resemble to a certain extent the effects noted by Greig 

 Smith. Comparison with the saline controls shows however that even 

 these extracts are somewhat better and certainly not worse than the 

 controls which were devoid of toxins. These are shown in Curve I. 



II. We next attempted to ascertain whether heating the soil extract, 

 which is said by Greig Smith to decompose the " toxin," did as a matter 

 of fact improve the extract for bacterial growth. In all the normal 

 soils tested it did not; on the contrary, it had the opposite effect and 

 led to still lower bacterial numbers (Curve II). 



