346 FIXATION OF FREE NITROGEN BY BACTERIA. 



and o. 1 8 p broad, and therefore belong to the smallest of known 

 micro-organisms. Even the Chamberland filter cannot restrain 

 them, and they escape through its pores. As the name implies, 

 they are endowed with motile power, which is frequently so strong 

 that individual rovers are able to escape from the parent colony, 

 make their way across the gelatin plate, and found a daughter 

 colony at a distance. The rod cells are not invariably of the 

 ordinary cylindrical shape ; on the contrary, variously bulged or 

 lobed forms appear in larger or smaller number according to cir- 

 cumstances, and a forked branching, resembling the Greek y, is 

 very frequent. This peculiarity, be it remarked en passant, is also 

 shared by other bacterial species, e.g. the Pasteuria ramosa already 

 mentioned in an earlier paragraph. Bacillus radidcola does not 

 produce any enzyme capable of dissolving gelatin, starch, or cellu- 

 lose, or of inverting saccharose ; neither has spore formation been 

 detected, a circumstance harmonising with the fact that a tem- 

 perature of 6o-7o C. suffices to destroy this fission fungus. On 

 the other hand, the organism appears to support drought and frost 

 without sustaining any injury. 



A few words must be devoted to the varieties exhibited by the 

 nodule bacteria. Hellriegel established (though without employ- 

 ing pure cultures) that the nodule bacteria of peas cannot develop 

 nodules on lupins and Seradella (Ornithopus sativus). This obser- 

 vation, which was challenged by A. B. Frank, was confirmed by 

 NOBBE, working with pure cultures, in conjunction with SCHMID, 

 HILTNER, and HOTTER (L). Whether the species should be divided 

 merely into races or varieties, as advocated by the observers just 

 named ; or whether we should here speak of different species in 

 the sense used by naturalists, and consequently express them by 

 different specific names, as was attempted, e.g. by A. SCHNEIDER 

 (L), is a point of remote importance. BEYERINCK (XV.) has also- 

 become a convert to this view, having been unsuccessful in inducing 

 the formation of nodules in Vicia Faba by means of bacterial cul- 

 tures from those of Ornithopus. On this point an interesting dis- 

 covery was made by NOBBE, HILTNER, and SCHMID (L), according 

 to whom the bacteria from any given species of Leguminosve 

 produce the most plentiful development of root nodules in the 

 shortest time when applied to other plants of the same species, the 

 potency diminishing in the case of plants of merely allied species, 

 and finally becoming nil when a greater specific difference exists 

 between the original plant and the one inoculated. The various- 

 separate species (or varieties) of the nodule bacteria can, to a certain 

 extent, therefore replace one another. Thus, those of the pea are 

 also efficacious for all the examined species of the genera Vicia 

 and Phaseolus, but, on the other hand, without effect on Robinia, 

 Ornithopus, red clover, kidney vetch, and other clovers. Those 

 of Robinia form nodules only on Phaseolus and a few species 

 of the genus Trifolium. Finally, according to the researches of 



