684 CONTRIBUTIONS TO OUR KNOWLEDGE OF SOIL-FERTILITY, iv., 



From these experiments, it is evident that the soil-bacterio- 

 toxin is not destroyed, and is probably unaffected by chloro- 

 form or ether. As it is not dissolved by these solvents, it 

 follows that the numerical increase, which follows the treat- 

 ment of the soil with these disinfectants, is not caused by any 

 alteration in the bacteriotoxin already present in the soil. 



The Effect of Reaction of the Medium upon the Toxic Action 

 of Soil-Extracts. — Two sets of experiments, with gradually 

 increasing amounts of lactic acid and of sodium carbonate, 

 showed that a neutral or faintly acid solution favoured the 

 growth of Bar. ])rodif/iosus in filtered soil-extracts, but that 

 the reaction had no influence upon the toxin, was shown by 

 the control-experiments in which no toxin was contained, 

 behaving in a precisely similar manner. 



The Soil -Bacteriotoxin is not Volatile. — Beyond the fact 

 that the soil-bacteriotoxin is toxic to bacteria, is thermola- 

 bile, is destroyed by sunlight, and is insoluble in chloroform 

 and ether, we know little about it. It appears to be non- 

 volatile, for when air-dried soil was heated at 75° C in a 

 slow current of air under diminished pressure, the few drops 

 of condensed water that were collected possessed no toxic 

 properties. 



Action, of Rai/i. — In the soil, the bacterial growth is 

 weakened by the toxin and strengthened by the soil-nutri- 

 ents, and the fertility, so far as these factors are concerned, 

 will depend upon the equilibrium established between the 

 two. One of the effects of rain is to dissolve the toxin, and 

 carry it into the subsoil, where it may decay after the manner 

 of the toxin in filtered extracts. The Hawkesbury soil. No. 

 4, was undoubtedly less toxic after heavy rains in January, 

 1911, than it was in October, 1910, and the earlier part of 

 the same year. 



This observation led to an experiment being tried to test 

 the distribvition of the toxin after moistening the soil with 

 water, similar to what would occur with a heavy shower of 

 rain. A kilogram of soil was placed in a wide glass jar (the 



