Cross Iiwculadoii of ilw Root Tiihn-ch- BmicrUi. 191) 



the root nodule bacteria from ataeia, tlie nodules were derived 

 from clover bacteria adherinji: to the surface of the small ch.ver 

 seeds and not renioved in the jirocess of sterilisation. With this 

 exception all the other results showed that imi direct cross inocula- 

 tion is possible between native and introduced le<ruminous plants. 



Discussion. 



The reason for this might l.e that the l)acteria. when taken 

 directly frt>m a particular root nodide. were biologically adapted 

 to the special nutritive conditions peculiar to the given 

 plant, and thus might be unable to directly adapt themselves to 

 a new host pknt. If, however, the bacteria were ti-ansfened co 

 sterilised soil and allowed to live there, or grown on artificial non- 

 living media as a saprophyte feu- some months, the special nutii- 

 tive adaptation in the course of repeated generations miglit gradu- 

 ally merge into a moi'e generalised condition, so that it could 

 inoculate a plant further and further away either systematically 

 or in physiological condition, the longer it remained in the soil. 

 Ultimately, the soil might contain bacteria capable of infectin;/ 

 any or all leguminous plants, from whatever ho.st tlu.'y might havf 

 been derived. 



In the above case, the l)acteiial contents were used to inoculate 

 the cultivated legumes, and so their specialised nature prevented 

 inoculation with plants presenting dissimilar l)io]ogical and 

 nutritive conditions. According to Buhlert^. all nodule organisms 

 are forms of Bacillus radicicola, but the bacteria best adapted lijpi 

 inoculating are those from the same host plant, and e()nse(iuently. 

 unless the nodule bacteria have become generalised in tlie soil, tiiey 

 are too specialised to infect other plants when used directly from 

 the nodule. 



That this is the correct explanation is indicated by the fact that 

 in some cases difiictdty ua-; experienced in inoculating red clovei' 

 and lucerne seedlings with bacteiia taken directlv fi-om root 

 tubercles of the same plants, namely, red clover and lucerne 

 respectively. Apparently loot tubercle Ijacteria taken dii'ectly 

 from the living tu1)ei-cle are a)it to die out in sterilised soil or in 

 water cultures, and will usually do so befoi-e they have become 

 sufficiently generalised to be cai)al)l(' of infecting a foi-eign ho.st 

 plant. 



On the nthei- hand, when the l)acteria fi'om one ]»lant Mi-e ^rown 

 on nutrient gelatine media, they appear to i-apidly develo]) more 



1 H. P.uhleit. (entralbl. nakt,. .Jt^ia .\ht. 2, l», 1-2I.2 (pp. 148-ln;?. •2-26--24ii. 27:^2S.^, S92-895). 



