FIXATION OF ATMOSPHERIC NITROGEN 407 



However, he later modified his belief in the origin of tubercles as due 

 to outside infection, and accepted the interpretation of his pupil 

 Brunchhorst who claimed that the bacteria-like bodies in the tubercles 

 were merely reserve food materials. Because of their resemblance to 

 bacteria Brunchhorst named them bacteroids. 



The studies of Marshall Ward, published in 1887, proved not merely 

 that tubercle formation is due to outside infection, but that such infec- 

 tion may be brought about at will by placing the roots of young plants 

 in contact with pieces of old tubercles. Hellriegel in his preliminary 

 communication of 1886 also showed that outside infection is necessary 

 for the production of tubercles, and called attention to the true func- 



FIG. 133. Ps. radidcoia. i, From Melilotus alba; 2 and 3, from Medicago saliva. 

 4, from Vicia mllosa. (After Harrison and Barlow from Lipman.) 



tion of the latter as laboratories wherein nitrogen compounds are 

 manufactured out of elementary nitrogen. The true worth of Hell- 

 riegel's investigations was brought out more clearly in another paper 

 that he published jointly with Wilfarth in 1888. The authors showed 

 that in sterilized soils legumes behaved precisely like non-legumes, and 

 died ultimately of nitrogen hunger when not provided with nitrates or 

 other suitable nitrogen compounds. On the other hand, when the 

 sterilized soil was later infected with a few drops of leachings from fresh 

 soil that had supported a normal growth of legumes, the starving plants 

 recovered and grew vigorously. Under the same conditions non- 

 legumes did not recover. The recovery of the starving legumes was 

 found to be coincident with the formation of tubercles. 



