341 



materials in tho roots of the fonrtli order wliich spring from the tubercles 

 consist exclusively of metabolic products of the bacteria themselves. 

 The question whether the bacteria take up and work over the free 

 nitro.uen of the air or water directly, or whether the crude nitroficn- 

 containing materials are transmitted to them from the leaves, although 

 the latter seems most probable to the authors, is not further discussed 

 by them. 



Other experiments. — In order that the conditions might be exactly 

 the same for the ditfereut kinds of ])lants, peas, lupine, Robinia, lal)ur- 

 num, and CJleditschia were grown together in the same vessel. Four 

 vessels, each containing one plant of each of the above species, were 

 inoculated with pure cultures of bacteria from Gleditschia soil, and from 

 pea, Kobinia, and lupine tubercles, respectively; the plants in one glass 

 remained sterile. All the plants in the same vessel were treated exactly 

 airivc. In the case of each vessel that plant did best which was inoculated 

 witli the bacteria peculiar tothetuberclesof its own species. Thus,inthe 

 case of the inoculation with bacteria from pea tubercles, the pea plant 

 far surpassed all the other plants in that glass in luxuriance of growth; 

 where Kobinia tubercle bacteria were used the Robinia did better than 

 any other plant, etc. 



Tlie fact that in most cases the root tubercles occurred at a depth 

 eipiivalent to only about one third of that to which the roots extended, 

 seemed to indicate that the spontaneous diffusion of the bacteria in the 

 soil takes place slowl5^ This would, seem to explain the fact that there 

 was no benefit from late inoculation. To test this point the following 

 cx])eriments were made: Six pea plants were placed in each of two 

 sterilized vessels containing sterilized soil; one vessel contained no 

 nitrogen and the other contained nitrogen in the form of calcium nitrate. 

 One ]>lant in each vessel was inoculated June 18 with 7 c. c. of extract 

 of i)ea tubercles, directly on the upper portion of tlie root, by means of 

 a pipette. At the time of the inoculation the plants in the nitrogen- 

 free soil were Just beginning to show signs of starvation. The inocu- 

 lated plant began shortly to show unmistakable evidence of recovery. 

 On the L'Stli of August the inoculated plant in the nitrogen-free soil 

 was about twice as tall and contained nearly twice as man 3' leaves 

 as either of the other five plants. The effect of the inoculation in 

 the vessel containing nitrogen, while less striking, was apparent. 

 The indications are that the ability of these bacteria to diffuse in the 

 soil spontaneously is quite limited. In this trial as well as in the first 

 series of experiments, it was noticed that the pea plants Mhich Avere 

 developed by the aid of inoculation either produced no flowers, or in 

 case they did, did not develop seeds, while those that were not inocu- 

 lated and those manured with nitrogen all bloomed and some few 

 produced se»*ds in spite of their less luxuriant growth. This is stated as 

 being a further indication that tlie action of the bacteria in leguminous 

 plants encourages the vegetable growth at the cost of the development 



