400 THE FOOD OF PLANTS 



aerobic root-tubercle bacteria live and are active in the presence of oxygen within 

 the respiring root-tubercle cells ', whereas Clostridium Pasteurianum assimilates 

 nitrogen when living anaerobically. In the latter case hydrogen is evolved, but the 

 root-tubercles show that an evolution of hydrogen is not necessarily connected with 

 the fixation of nitrogen 2 . 



The conditions for the formation of root-tubercles. No further account of the 

 morphology of the tubercles can be given, nor is it necessary to bring forward any 

 proof that they are formed owing to the penetration of certain bacteria into the 

 root-cells 3 . The infection is strictly localized, for the bacteria do not spread 

 through the tissues, and tubercles are formed only upon those parts of the roots 

 which are in contact with an infected soil *. Zinsser has shown that no tubercle- 

 bacteria are present either in the other parts of the root, or in the stem and 

 leaves. Root-tubercle or other bacteria injected into sub-aerial organs die after 

 a time without spreading or increasing in numbers s , hence the roots of seedlings 

 and cuttings only form tubercles when infected from the soil *. 



The positive results obtained with pure cultures by Prazmowski, Beyerinck, 

 and Nobbe, show that a single species of bacterium suffices for the normal forma- 

 tion of root-tubercles, in which however, according to certain authors, two distinct 

 forms may be present. It is also doubtful whether one or many species of tubercle- 

 forming bacteria exist, and whether the so-called ' bacteroids ' are always simply 

 developmental stages of the infecting bacterium, or may also include constituents 

 of the root-tubercle cells 7 . If Bacillus radicicola is the only root-tubercle bac- 

 terium, it must occur in a variety of adaptive forms, for Hellriegel showed that 

 a given soil will not infect every leguminous plant, while Nobbe found that the 

 bacterium from the root-tubercles of Pisum sativum could infect Phaseolus, but not 

 Robinia, TrifoHurn, or Serradella *. 



Hence it is advisable always to infect a plant from the tubercles of others 

 of the same kind. When a particular legume has not been cultivated in a 

 field for some time, it is now a matter of common practice to strew over it soil 

 charged with bacteria from the root-tubercles of the plant in question, or from 

 pure cultures. Infected soil gradually loses its power of inducing the formation of 



1 On the aeriferous system of the tubercles, see Frank, Ber. d. Bot. Ges., 1892, p. 371. 

 8 Cf. Winogradsky, Compt. rend., 1894, T. CXVIII, p. 353. 



3 See the works quoted on the previous page, and also Kionka, Biol. Centralbl., 1891, Bd. XI, 

 p. 282. 



4 Cf. Hellriegel, 1. c. 1888, p. 175; Laurent, Ann. d. 1'Jnst. Pasteur, 1891, T. V, p. 130; 

 Zinsser, Jahrb. f. wiss. Bot., 1897, Bd. XXX, p. 423. 



8 Zinsser, 1. c. On other bacteria, see also Koch, Biol. Centralbl., 1894, Bd. XIV, p. 481 ; 

 Russel, ibid., Bd. LIX, p. 375. 



' Recognized by Hellriegel. Cf. also Prazmowski, Versuchsst., 1891, Bd. xxxvill, p. 58; 

 Nobbe, ibid., Bd. xxxix, p. 350 ; Zinsser, 1. c. 



7 Cf. the works already quoted, and Schneider, Ber. d. Bot. Ges., 1894, p. n ; Gonnermann, 

 Landw. Jahrb., 1894, Bd. XXIII, pp. 649, &c. On the occurrence of different kinds of tubercles, cf. 

 Frank, Ber. d. Bot. Ges., 1892, pp. 170, 293 ; Moller, ibid., p. 243; also Morck, Ober d. Formen d. 

 Bacteroiden, 1891. 



8 Hellriegel, 1. c., 1888, p. 146 ; 1889, p. 142 ; Nobbe, Versuchsst., 1894, Bd. XLV, p. 19 ; 1896, 

 Bd. XLVII, p. 266. 



