244 



related plants, but harmless to others. This conception, based 

 on insufficient evidence, as it was, did not gain acceptance, but 

 has smouldered along in agricultural science ever since, flickering 

 up occasionally only to be smothered for lack of experimental 

 proof. 



During the last few years, however, a great deal of new evidence 

 has been presented in favour of DeCandolle's idea. In 1897, 

 and subsequently, investigations at the Woburn Experimental 

 Fruit Farm, in England, have shown that the presence of grass 

 in the soil about apple trees has a marked deleterious effect 

 upon the growth of the trees. It was shown experimentally that 

 this effect could not be due to the removal of nutriment materials 

 nor of water, nor to the exclusion of air, and it was suggested that 

 it must be caused by poisonous bodies emanating from the grass 

 roots. A similar antagonism has been shown to exist between 

 butternut trees and cinquefoil, and between peach trees and 

 several herbaceous plants. In 1904, Livingston published evidence 

 to the effect that bog water exhibits properties of a toxic nature 

 and suggested that the xerophilous character of bog plants may 

 be due to these properties. 



For our knowledge of the presence of toxic substances in 

 agricultural soils we are largely indebted to the work of the U. S. 

 Bureau of Soils. In Bulletin 23 of this bureau, it was shown that 

 the unproductiveness of certain soils examined could not be 

 attributed to any lack of available mineral matter, and that the 

 injurious properties of the soil could be transmitted to its aqueous 

 extract, independent of the salt content. In later publications 

 from the same bureau, evidence was presented in favour of the 

 idea that certain poor soils contain toxic substances which act to 

 retard the growth of roots. Further evidence favoured the 

 conclusion that wheat roots give off substances toxic to themselves, 

 and that this toxicity, as well as that of the soils mentioned 

 above, can be removed from nutrient solutions or soil extracts by 

 the absorbent action of carbon black, ferric hydrate, and other 

 finely divided inert solids. In Bulletin 28 it was shown that when 

 a nutrient solution becomes "exhausted," so that plants grow but 

 poorly in it, it is greatly improved by treatment with carbon 

 black, &c, the suggested explanation being that the roots first 

 grown in the solution gave off substances injurious to themselves, 

 and that these substances were removed by the absorbing action 

 of the finely divided solids. 



Two recent bulletins (Nos. 36 and 40) from the Bureau of Soils 

 have contributed more information on this subject. In Bulletin 36, 

 Livingston and others give more evidence in favour of the 

 existence of toxic bodies in unproductive soils and add certain 

 points as to the nature and origin of such substances. The con- 

 clusion is that toxic material is present in certain unproductive 

 soils, either in very minute quantities or in a very slightly soluble 

 form ; that this material is volatile in some cases and in others 

 non-volatile ; that it is often destroyed by boiling the soil extract 



