vii.] LEGUMINOS^E AND NITROGEN FIXATION 205 



subject came with the publication, in 1886, of the 

 researches of Hellriegel and Wilfarth. These investi- 

 gators found that when plants were grown in sand 

 and fed with nutrient solutions, the Gramineae, the 

 Cruciferae, the Chenopodiaceae, the Polygonaceae, grew 

 almost proportionally to the amount of combined 

 nitrogen supplied ; and, if this were absent, nitrogen 

 starvation set in as soon as the nitrogen of the seed 

 was exhausted. With the Leguminosae, however, a 

 plant was observed sometimes to recover from the 

 stage of nitrogen starvation, and begin a luxurious 

 growth which lasted until maturity, though no com- 

 bined nitrogen was supplied. In such cases the root 

 of the plant was always found to be set with the little 

 nodules characteristic of the roots of leguminous plants 

 when growing under natural conditions. Further experi- 

 ments were made in which the plants were grown in 

 sterile sand, but as soon as the stage of nitrogen hunger 

 was reached, a small portion of a watery extract of 

 ordinary cultivated soil was added ; whereupon, the 

 plants receiving the extract recovered from their nitrogen 

 starvation and grew to maturity, assimilating consider- 

 able quantities of nitrogen. The renewed growth and 

 the assimilation of nitrogen were always found to be 

 attendant upon the production of nodules on the roots. 

 The nodules were found to be full of bacteria, to which 

 the name of Pseudomonas radicicola has been given. 

 They could only be produced by previous infection either 

 by an extract of the crushed nodules or of a cultivated 

 soil ; in some cases (lupins, serradella) only by soil which 

 had previously carried the same crop. 



These results, though not at first accepted by Lawes 

 and Gilbert, led to a repetition of the experiments, 

 which brought out the fact that in their earlier 

 trials with leguminous plants the necessary inocula- 



