234 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



It was at one time thought that the bacteria invaded the whole plant, 

 but, as Zinsser * has shown, no nodule bacteria are present either in 

 any parts of the root besides the nodule or in the stem or the leaves, 

 and if the bacteria are inoculated into these parts they do not increase 

 in numbers and die after a time ; hence the roots of seedlings and cuttings 

 form nodules only when infected from the soil and not after infection 

 through any other part of the plant. 



Conditions under which Nodules are Formed. — The conditions under 

 which nodules are formed on the roots appear to be : (a) the roots must 

 be in contact with infected soil, since only the parts of the root in such 

 contact become infected ; (b) nitrates must not be present in any great 

 quantity, for Nobbe and Hiltner t showed that saltpetre prevented the 

 formation of nodules to a greater or less extent ; (c) the presence of lime, 

 phosphates, and potash in sufficient quantity ; and (d) the presence of 

 a go^d supply of air in the soil. In this connection it is noticeable that 

 the nodules are very rarely produced upon roots of plants in water cultures, 

 and leguminous plants are very seldom found in wet soils. 



Is the same Bacterium capable of forming Nodules on all Species 

 of Leguminous Plants ? — This question is of practical importance, since 

 if all legumes can be infected by one and the same organism it is evident 

 that one inoculation of the soil would be sufficient for any and all 

 leguminous plants. Much study has been given to this point, but it 

 cannot be said to be settled yet. Hellriegel found that he obtained 

 an increased yield of peas, clover, vetches, and beans when he watered 

 them with washings from the soil in which peas and clover had grown, 

 but not when he used washings from soil in which seradella (Ornithopus 

 sativus) and lupines had grown ; and he therefore concluded that there 

 were essential differences between the bacteria from these different 

 legumes. Nobbe, Hiltner, and others came to a similar conclusion, but 

 they did not consider that the differences warranted the view that different 

 species of bacteria were concerned. They practically held the same view 

 as has been expressed by Moore, t that the organisms of the different 

 leguminous plants were all forms of the same species, Psmdomonas 

 radicicola, but that by long-continued growth upon one plant they had 

 become so adapted to that particular partner that they form nodules 

 upon others only with difficulty at the outset, but that after continued 

 multiplication upon a neutral medium they are capable of infecting any 

 leguminous plant. Moore states that, after growing the bacteria obtained 

 from pea nodules for two weeks upon a medium free from nitrogen, he 

 was able t > infect various clovers (Trifolium pratense, T. incarnatum, 

 T. repens, T. alexandrinum, and T. hybridum), melilot (Mclilotus alba), 

 cowpja (Vigna catjang), lucerne (Medicago sativa), broad bean (Vicia 

 Faba), common bean (Phaseolus vulgaris), fenugreek (Trigonella foenum- 

 graecum), and vetches (Vicia lutca, V. villosa, and V. fulgens). Only 

 lupines proved refractory. Maze § considered that there were two groups 

 of nodule-forming organisms — those adapted to live in an acid soil and 



* Jnhr. wtSf. Bat., 1R97, Bd. xxx. p. 423. 



t lV/-.s,/r/,.s.s/., iH'.i;}, IM. xlii. p. 477, and Hiltner. 1896, Bd. xlvi. p. 161. 



1 U.S.A. Dep. Ac,?:, Bur. PL hid., Bull. 71, pp. 25. 26. 



I I tut, ii VIn$t. Pasteur, xi. 1897, pp. 145-155. 



