FROM THE AIR BY THE LEGUMINOS^ 



29 



becoming empty at the period of maturity. It further rests 

 on the fact that the tubercles are more abundant in a soil 

 poor in nitrogen, while their development is restricted when 

 the soil is rich in that element. 



It is difiticult to grasp the idea of storehouses retaining 

 quantities of nutriment for which the plant is hungering, 

 while they are empty when the nutriment is most plentiful. 



Experiments to show the Influence of Bacteria on the 

 Formation of Nodules. 



Numbers Nitrogen Dry matter 



V supplied in the furnished by 



vessels 



262 

 263 

 264 

 265 

 254 



256 



257 

 270 

 271 



form of 

 nitrate 



the whole 

 plant 



0'007 

 0'007 

 0'007 

 OCO7 

 0-056 

 0-056 

 O'l 12 

 0-II2 

 0-II2 

 0-II2 



0-209 

 0-272 

 0-316 

 0-297 

 2-838 

 2-927 

 6*223 

 6-858 

 6077 

 6-837 



Gain in 



nitrogen 



during 



growth 



— 0-005 



— 0-004 



— o 000 



— Q-OOO 

 -0-028 



— 0-027 



— 0042 



— 0-042 

 -0-042 



— 0-046 



Tubercles formed en the roots 



III slei ilized sand vn'ih addi- 

 Uoii of nitrate. 

 Nil 



In sand provided tvith soil in- 

 fusion and itith addition of 

 nitrate. 



I Numerous tubercles, of which 

 many were old and some al- 

 \ ready empty. 



(As above, but fewer old tu- 

 ( bercles. 



j' A fair number of old tubercles, 

 J but nearly all firm ; a large 

 I number of very small protu- 

 ( berances on the beard. 



From the mass of observations recorded by Hellriegel 

 and Wilfarth there was general agreement in regarding 

 these nodules as means of plant nourishment, and this 

 induced the authors to publish the following statements : — 



(i) The formation of nodules and the plant growth are 

 independent of each other ; plants have been shown to grow 

 in normal fashion without any nodules. 



(2) This formation is independent of nitrogen assimi- 



