ROOT-NODULES OP LEGUMINOSAE. 131 



experiments on high moorland at Bremen. In this case the strikingly large nodules formed on the 

 yellow lupin from spontaneous infection were completely inactive, because of the poor degree of 

 virulence [word here used with wider meaning] of the invading bacteria. 



Experiments with soy-beans in pots confirmed this result and also explains how this increase in 

 virulence can be brought about. Pots of Dahlem earth and sand, half of which were inoculated with 

 Japanese soy-bean earth, were planted with soy-beans. Only late in the season did the difference 

 appear between the inoculated and uninoculated plants. The greatest difference, however, was 

 observable at maturity. The leaves of the uninoculated plants were then almost white, having lost 

 almost completely their xanthophyll and their roots were free from nodules while the leaves of the 

 inoculated plants were deep yellow and their lateral roots full of nodules as large as peas. The amount 

 of dry substance from the inoculated plants was 33 per cent greater, and of seed about 74 per cent 

 greater than that from the uninoculated plants. The next year plants in these same pots of soil with- 

 out further inoculation showed an even greater difference. Those in the uninoculated soil remained 

 completely nodule-free while the others showed extraordinarily strong nodule formation, so that every 

 root on the upper third was closely covered with nodules the size of peas. Moreover the activity of 

 the nodules the second year was remarkably increased. Even if it is very probable that the increased 

 number of bacteria in the soil plays a r61e here, and that because of the great lack of nitrogen in 

 the soil the nodules became active earlier and hence had a longer season in which to work, nevertheless 

 the most important factor must be sought in the increased virulence, for while the total amount of 

 nodule formation was considerably greater the second year it was not so much so as to go parallel 

 with the increase in activity. Increased virulence here, as in organisms pathogenic to animals, is 

 due to repeated passage through the organism in question. The dry substance of the uninoculated 

 the first year being taken as 100, we have the following yields: 



Year. Uninoculated. Inoculated. 



IQOO IOO 



IQOI 66.4 1458 



Hiltner took this into consideration in obtaining virulent cultures for inoculation for all import- 

 ant species of legumes, and with few exceptions his cultures were passed several times through the 

 appropriate host-plant. The better results of 1901 and 1902 in the open field he thinks is largely 

 referable to this fact. 



The chain of evidence for the greater activity of virulent bacteria was completed by making 

 simultaneous inoculation experiments with bacteria of differing virulence, mostly on peas. The 

 results entirely confirmed the preceding observations. 



The effect of difference in virulence is also shown in the case where soil from Zehringer in Anhalt 

 in which peas were known to form active nodules only on the deeper roots was used for inoculation 

 in comparison with a third generation of pure cultures. The result was that the bacteria from the 

 pure cultures penetrated the roots at once and formed numerous nodules on all the upper parts of 

 the roots while those from the Zehringer earth produced nodules only on the deeper side roots and 

 sparingly at that, i. e., as in the fields. Only virulence can explain why the bacteria in the Zehringer 

 soil did not penetrate the roots at once. Hiltner thinks that a moderate degree of virulence on the 

 part of the bacteria is sufficient to cause leguminous plants to attain their maximum size. He thinks 

 it also unquestionable that virulence may be increased in other ways than by repeated passage 

 through the plant, i. e., as Remy showed by feeding the inoculated plant with saltpeter. 



It is an important question for agriculture whether one can not obtain too great a degree of 

 virulence. It has been shown that at first the bacteria act toward the plant as real parasites. If 

 this is correct, as appears on more grounds than one, we have to reckon with the possibility that the 

 fight will be the more severe for the plant as the bacteria become more virulent. Indeed Nobbe and 

 Hiltner record an experiment in which highly virulent pea bacteria caused direct injury to the plants 

 growing in pots of nitrogen-free sand. In this case the bacteria penetrated the roots in such numbers 

 that almost every root-hair was infected, and everywhere numerous nodules appeared in which, 

 however, bacteroid formation was suppressed owing to the excessive growth of the bacteria. As a 

 result no nitrogen assimilation occurred and the aerial parts of the plants were plainly weakened. 

 Another experiment in 1902 at Uahlem on pot plants with especially virulent bacteria resulted in a 

 decrease in the crop: Inoculated plants gave a dry substance amounting to 34.1 g. per pot.* while 

 uninoculated plants gave 47.1 g. 



*Possibly in some of these cases Bad. tumefaciens may have been the cause of the nodules. See Crown-gall : Its 

 cause and remedy. Bull. 2i3,B. P. Ind.,U. S. Dept. Agric., 1911, plate xvr, za. 



