INVESTIGATIONS ON THE PRODUCTION OF PLANT FOOD. 199 



all — the argument from analogy. When a bacterial culture is made 

 on an artificial medium in the laboratory the organisms go on de- 

 veloping for a time and then stop, being brought to a standstill by 

 an accumulation of decomposition products with which they cannot 

 deal. It is argued that the same thing must go on in the soil, and 

 that the accumulation here must have a toxic effect just as it does 

 in the artificial culture. 



There is, however, this fundamental distinction between the labo- 

 ratory culture and the soil. In the laboratory culture the medium is 

 made up to deal with one class of organisms only ; as soon, therefore, 

 as substances accumulate which this class cannot decompose the 

 whole action automatically comes to an end. But in the soil there 

 is a great variety of organisms capable of attacking a considerable 

 vaiiety of organic substances : so great indeed that it is difficult 

 to understand how any intermediate decomposition product could 

 accumulate sufficiently to interfere with bacterial activity. The final 

 product of all — the nitrate — is so easily washed out from the soil that 

 it rarely, if ever, accumulates under normal conditions. 



The argument in favour of bacteriotoxins deduced from the analogy 

 of laboratory cultures is therefore devoid of foundation, and we must 

 rely simply on the direct experimental evidence. Of this there is 

 practically none, all sound experiments giving negative results. Nega- 

 tive evidence is notoriously unsatisfying, but until some positive 

 evidence is forthcoming we cannot suppose that bacteriotoxins play 

 any notable part in the soil. 



There is, however, considerable evidence that the growing plant 

 exerts a depressing effect on the soil organisms. Rigid comparisons 

 are not easy, but when the conditions on a fallow plot are made to 

 approximate as closely as is possible to a cropped plot it is found 

 that there is more activity on the fallow plot. 



This has been observed at various places : at Rothamsted, at 

 Ithaca, at Pusa and elsewhere in ordinary arable soils, and it was 

 noticed by Harrison in his studies of paddy soils. It is not clear 

 from the experiments how the action takes place : whether the plant 

 simply exercises some indirect action on the temperature or moisture 

 supply, or whether it directly affects the soil organisms. On the 

 whole the evidence rather tends to indicate a direct action such as 

 might be brought about by some poison given off from the root or 

 left by the plant, or such as would result from the removal by the plant 

 of some substances necessary for the bacteria. 



Further experimental work is in hand on this matter. The effect 

 of the plant on soil bacteria recalls the remarkable effect of one growing 

 crop on another observed by Mr. Pickering. It is obviously im- 

 possible to say that these effects are identical, but we may reasonably 

 hope that the further investigation of each of these problems will 

 throw helpful light on the other. 



