FROM THE AIR BY THE LEGUMINOS^ 23 



Check experiments confirmed ihe original results, and 

 these scientists concluded that the Gramine^e, being unable 

 to live in soil devoid of nitrogen, did not have the property 

 of fixing atmospheric nitrogen. 



Totalh- different was the case of the Leguminos^e, whicli, 

 planted in a soil completel}' devoid of nitrogen, throve 

 well, while peas planted in this same soil the year after 

 followed their normal development. Hellriegel, however, 

 was induced tiirough a fev/ failures to put a stop to his 

 experiments and to studv, till 1873, the influence exerted on 

 plant development by each of the various known factors of 

 growth, such as the cjualitv of the seed, the volume of soil, 

 its mechanical structure, tlie period of sowing, light, heat, 

 air and, finally, moisture. 



It was onl\' in 1883, after the founding of the Agricul- 

 tural Station at Bernburg, that Hellriegel and ^^"ilfarth were 

 able to recommence their researches. The first three }'ears, 

 1883-1885, were devoted to fresh experiments on the 

 Gramineaj and Leguminos^e. Once more a negative con- 

 clusion was obtained as to the power of the Gramine^e to 

 fix atmospheric nitrogen. 



Peas showed irregular development, although, as every 

 precaution had been taken as to the nature of the soil and 

 general conditions, the causes of this irregularity could only 

 have been accidental. 



These negative results led the scientists to believe that 

 bacteria certainly must play a part in the growth of 

 Leguminos^e, and induced them, in 1886, to undertake 

 experiments in that direction. The results being so con- 

 clusive, in 1887 experiments were renewed, and the opinion 

 they had formed was still furtlier confirmed. 



Further, Hellriegel's memoir gives us details as to all 

 the experiments conducted on oats and barlev, with 

 numerous illustrations of the various methods emploved. 

 The value of these experiments is still more evident when 

 it is seen what pains were taken to avoid sources of error. 



