52 RESULTS OF EXPERIMENTS AT ROTHAMSTED, 



otherwise remain unexplained. To this, and other points involved, I 

 shall refer again presently. 



Lastly, I have to summarise those of the results and conclusions of 

 Boussingault, which bear upon the present aspect of the question of 

 the sources of the nitrogen of vegetation. In his earlier experiments, 

 as in those at Eothamsted, sterilised materials had been used as soils ; 

 but in 1858 he commenced a series in which more or less of a rich 

 garden-soil was mixed with sand and quartz. In some cases the plants 

 were grown in free air, and in others in closed vessels with confined 

 air. In several cases there was more or less gain of nitrogen ; but the 

 greatest gain was in an experiment with a Lupin grown in a closed 

 vessel. Boussingault points out that it was the soil and not the plant 

 that had fixed the nitrogen. The result was so marked that he 

 repeated the experiment in 1859, when he obtained almost identically 

 the same amount of gain as in 1858. He also put 120 grams of the 

 rich soil into a shallow dish, moistened it with distilled water, and 

 exposed it to the air as an experiment on Fallow. The results showed 

 a small gain of nitrogen. 



Boussingault further found, that mycodermic vegetation went on in 

 rich soil, and he considered the gains of organic nitrogen represented 

 the remains of such vegetation ; whilst the Fallow experiment indicated 

 that the experimental plants had little to do with the action. His 

 general conclusion was, that from the numerical results it must be 

 believed, that the soil had fixed nitrogen ; and he considered that, if 

 there were no absolute proof, there was strong presumption, that the 

 nitrogen of the air takes part in nitrification. 



In the next year, 1860, he put into one large glass balloon a 

 mixture of rich soil and sand, and into another a similar mixture, with 

 cellulose in addition ; each was moistened with distilled water, and the 

 vessels were then closed up for 1 1 years. During this period, without 

 cellulose rather more, and with cellulose rather less, than one-third of 

 the nitrogen of the soil was nitrified ; but in neither case was there 

 any gain of total combined nitrogen. There was, indeed, in both 

 cases, a slight loss of nitrogen indicated. Boussingault concluded that 

 free nitrogen had not contributed to the formation of nitric acid. 



The later results of Boussingault did not, therefore, confirm those 

 he obtained in 1858 and 1859; and in answer to a letter from me he 

 wrote in 1876, that he was not aware of any irreproachable observation 

 which established the reality of the fixation of free nitrogen by the 

 soil. He further stated his belief, that neither the higher plants, nor 

 mycoderms, nor fungi (champignons), fix free nitrogen. He also 

 maintained the same view in conversation in 1883. 



SUMMARY AND GENERAL CONSIDERATIONS ON THE SOURCES 

 OF THE NITROGEN OF OUR CROPS. 



It did indeed seem that, in Boussingault's results of 1858 and 

 1859, there was the germ of the germ theory of the fixation of free 



