442 Tubercles on the Roots of Leguminous Plants. [Nov. 21, 



tubercle provides continual successions of new cells for the bacteria to 

 accumulate in while it absorbs the older ones behind. Between the 

 bacteroid tissue and the cork it provides vascular bundles (1) to feed 

 the bacteria and convey the carbohydrates necessary to produce 

 proteids, and (2) to take up the dissolved substance of the bacteroids 

 as required. The thin cell walls of the bacteroid tissue conduce to 

 the same end. 



I think it will be admitted by all who study the literature of this 

 subject, that the only real point at issue between Prazmowski and 

 myself is the nature of the bacteroids and their origin from the 

 filaments. I interpreted them as extremely minute budding "gem- 

 males," and not bacteria ; Prazmowski, with Beyerinck, regards them 

 as true Schizomycetes. We have all alike failed to actually see the 

 process of budding or fission, a fact which will surprise no one who 

 has examined these extremely minute bodies, which are, as Beyerinck 

 rightly puts it, among the smallest of living beings. 



The fact of infection, and the mode of infection, by means of a 

 hypha-like filament passing down the root-hair were definitely estab- 

 lished by myself in 1887, and it is satisfactory to find it confirmed in 

 every essential detail by Prazmowski. Our views as to the symbiosis, 

 the struggle between the protoplasm and the " gemmules " (or 

 "bacteroids") are the same: though Prazmowski and Beyerinck 

 carry the matter a step further in definitely inferring the absorption 

 of the conquered bodies of the latter, a point in part supported by 

 some of my experiments. 



As to the occurrence, origin, and structure of the tubercles, 

 Prazmowski' s account is simply in accordance with my own ; and it 

 is interesting to note how many points of detail — the distortions of 

 the root-hairs, the relations of the branching filaments to the nuclei 

 and cell-contents, and those of the incipient tubercle to the end of the 

 filament, for example — are confirmed by him. 



There is one point of extreme importance between Beyerinck and 

 Prazmowski on the one hand, however, and myself on the other ; 

 they are positive on the subject of the cultivation of the "bacterium" 

 in nutritive media outside the host-plant — or rather the other 

 symbiont — whereas I feel too little confidence in my cultures to 

 assert that the " germ " is definitely isolated and recognised. It is 

 true, I have obtained colonies in the cultures which may be those 

 referred to by these writers, and I may remark that so long ago as 

 1887 I wrote that certain flocculent clouds in my cultures may be 

 colonies of the organism in question, as I obtain similar clouds of 

 multiplying " germs " on the root-hairs of my water-cultures. More- 

 over, in some cases I have clear proof that among the colonies in my 

 cultivations the germ in question existed, because I infected peas and 

 beans with them; but it would be going further than the facts 



