NITROGEN FIXATION 239 



been obtained. Such being the state of the case, it is of interest to note that 

 direct evidence of the fact has been advanced by Schlossing and Laurent 

 (1890) who have calculated how many milligrams of nitrogen a pea takes 

 from the air during the several months of its vegetative growth, checking 

 this result by estimating the increase in nitrogen in the soil and in the crop. 

 The following summary shows that the agreement between these calculations 

 is almost perfect : — 



Atmospheric nitrogen introduced into the culture vessel 2681-2 ccm. 

 „ ,, withdrawn from „ „ 2653.1 ,, 



Amount of nitrogen assimilated 29-1 ccm. 



= 365 mg. 



Nitrogen in soil and seed 326 mg. 

 „ », ,, crop 73-a „ 



Nitrogen assimilated 406 mg. 



Another point of interest is that demonstrated by Nobbe and Hiltner 

 (1899 b), that the atmospheric nitrogen is combined in the nodules and not at 

 all in the leaves of the nodule-bearing plant. It was previously thought that the 

 leguminous plant in presence of Bacterium radicicola became altered in some 

 way so that it acquired the power of fixing free nitrogen. Nobbe and Hiltner 

 infected plants of Rohinia with Bacterium radicicola, growing them in a culture 

 fluid from which nitrogen had been excluded, and observed the formation of 

 nodules under water ; but so long as these remained immersed they were found 

 to be useless to the plant, and it was only when they were brought into the air 

 that they commenced to assimilate the nitrogen. These facts prove that the 

 nitrogen must enter the nodules themselves if it is to undergo fixation. 



Finally, we may note that many authors have been successful in seeing 

 the evolution of the Bacteria into bacteroids in artificial nutritive solu- 

 tions (Beijerinck, 1888 ; Hiltner, 1900 ; Stutzer, 1901). Nevertheless 

 it has not as yet been explained why only certain and not all Bacteria are trans- 

 formed into bacteroids in the plant, a phenomenon which is of the utmost 

 importance in relation to the conservation of Bacterium radicicola in nature 

 (p. 240). 



If we describe the association of Leguminosae with nodule bacteria as a 

 case of symbiosis, then, looking backwards, we may also term the relationship 

 of Clostridium pasteurianum to the two associated bacteria as symbiosis also, 

 and we may take this opportunity of drawing attention to some other cases 

 of the same kind. Beginning with Elaeagnus and the alder, we find that both 

 these trees produce on their roots nodules which remind us of those of the 

 Leguminosae, and which seem to carry out similar functions. Hiltner (1896) 

 has shown in the case of the alder that when the nodules are absent it can develop 

 only if nitrogenous compounds are provided, but that after the nodules have been 

 formed the nitrogen of the air is sufficient. [Hiltner gives a very instructive 

 illustration of alders, which have been grown in sand destitute of nitrogen, one 

 with and one without nodules (Hiltner, 1904, p. 63).] Similar conditions 

 obtain in Elaeagnus, but research is required as to the mode of development 

 of the nodule-forming organisms in both cases ; recently Shibata (1902) has 

 supplied us with an interesting account of their structure. 



Seeing that the use of free nitrogen by higher plants is not confined to the 

 Leguminosae, we are lead to believe that Frank's (1890) view is the correct 

 one, and that this power may be possessed by all plants to a greater or lesser 

 degree. It will be sufficient to say on the other hand that careful experiments 

 made on the majority of Phanerogams, e. g. on Gramineae and Cruciferae (Aeby, 

 1896, Pfeiffer and Franke, 1897) have given only negative results. In other 

 cases, however, and especially in such as exhibit a symbiotic union between 



