ROOT-NODULES OF LEGUMINOSAE. 1 17 



It does not seem impossible to Hiltner, that the virulence may be increased beyond that natu- 

 rally reached, by cultivating the bacteria for a long time on root extracts of increasing concentration, 

 since it has been already demonstrated by Nobbe and Hiltner that repeated culture on gelatin con- 

 taining extracts of the green parts of the plant increases the virulence of the bacteria materially. 



The nodule bacteria are probably drawn to the root-hairs by chemically attractive excretions 

 whose nature is not yet determined. They appear tobespecific for each species of legume. In general 

 they seem to be organic acids and acid phosphates. These are the substances which according to the 

 convincing experiments of Stutzer cause the transformation of bacteria into bacteroids. This trans- 

 formation takes place within the nodules as soon as the enveloping slime is dissolved by these acids, 

 leaving the bacteria unable to penetrate further. Only when pea plants grown in a nitrogen-free 

 solution become weakened, and secrete less protective substances, bacteria of poor virulence can 

 penetrate the root-hairs, but these produce inactive nodules. 



Nodule bacteria which have penetrated into the roots may be absorbed by the plant. Plants 

 inoculated with lupin bacteria obtained from Hochst failed to produce nodules, although sections of 

 the roots a few days after inoculation contained groups of bacteria. Further investigation showed 

 that these had completely broken down into tiny microsome-like bodies, when they could be demon- 

 strated at all. Although obtained from lupin nodules these were probably not genuine lupin-bacteria. 

 Under similar conditions Hiltner's own isolations readily induced nodules and behaved differently 

 on gelatin, i. e., grew less vigorously. 



Conditions of nutrition affect nodule formation. Solutions or soils poor or lacking in nitrogen 

 are more favorable than those containing this element. Yet it only needs bacteria of a higher degree 

 of virulence, adapted to the species, to infect even the best nourished plants. 



Saltpeter prevents the formation of nodules. That this is not due to the abundance of nitrogen 

 supplied as was formerly supposed, but to a directly injurious effect on the bacteria, is shown by the 

 following experiments: 



Inoculated pea plants remained free from nodules when grown in a liter of nutrient solution 

 containing 5 mg. nitrogen in the form of saltpeter, while in a similar solution free from nitrogen 

 abundant nodules were formed. The same results were obtained with Robinia and Alnus. It seems 

 impossible that such a small amount of nitrogen should completely repress nodule formation by 

 stimulating growth of the plant. The saltpeter must act, therefore, not by helping the plant, but by 

 injuring the bacteria. Small doses of saltpeter are soon absorbed by the plant and then nodule 

 formation sets in. Only when plants are transferred every 2 or 3 days into fresh solutions containing 

 saltpeter do they remain permanently immune. 



In quartz-sand cultures the effect of saltpeter was more marked than in mixtures of sand and 

 soil though not nearly so great as in solutions, but 50 mgs. of nitrogen in the form of KNO 3 to i liter 

 of sand did not completely suppress nodule formation. The nodules remained small but bacteroid 

 formation proceeded very rapidly. Saltpeter increases bacteroid formation extraordinarily, not only 

 in the nodules but also in fluid cultures of the bacteria. 



Within certain bounds, the more sandy the soil the greater is the injury caused by fertilization 

 with saltpeter. As the most vigorous plants were found in rich humus soils where nodule formation 

 was little suppressed by the saltpeter, it is evident that the suppression of root-nodules is not due 

 simply to the invigoration of the plants by saltpeter. 



Other nitrogenous compounds, such as ammonium sulphate had much less effect on nodule 

 formation and extract of horse manure was entirely harmless. 



The activity of the nodules begins to be visible in the parts above ground only when the avail- 

 able nitrogen in the soil begins to be exhausted. In nitrogen-free soil, seedlings must draw their 

 supply from the seed. As this supply is exhausted before the activity of the nodules begins, a period 

 of hunger intervenes, during which the leaves become strikingly yellow, and the growth of the plant 

 almost ceases. Hellriegel and Wilfarth observed this and Hiltner has repeatedly confirmed it. A 

 small amount of nitrogen placed in the soil, especially stable manure, will tide the plants over this 

 stage, but a dose larger than that actually required for the development of the nodules hinders nitro- 

 gen assimilation. For the same reason on soils containing nitrogen a mixed growth of leguminous and 

 non-leguminous plants is believed to be better than unmixed growths since the non-legumes will 

 quickly use up the nitrogen, and the legumes under such conditions will come into activity sooner. 



Hiltner thinks from the foregoing that the Leguminosae strongly influence the process of nitrifi- 

 cation in the soil. 



In saltpeter solutions, the root-hairs do not curl, showing that the bacteria are not able to pene- 

 trate. Yet the bacteria are not killed, for they succeed in penetrating after the saltpeter has been 

 absorbed by the plant. 



The immunity of the plant afforded by already active nodules appears to be due to an excretion 

 from the bacteroids. 



