ASPECTS OF AN OLD AGRICULTURAL QUESTION. 263 



— to say nothing of the nitrifying organisms now known to 

 be especially effective — is continually being oxidised to 

 nitric acid and its salts. 



Against these possible gains of nitrogen to the soil we 

 have to set a series of sources of loss. Although the 

 upper layers of soil can hold certain salts of nitrogen with 

 great tenacity, it is well established that the drainage water 

 carries off large quantities to the depths of the subsoil, and 

 into the drains, rivers, etc., to the sea, and we know that 

 de-nitrifying organisms cause the freedom of certain quanti- 

 ties of free nitrogen from the decomposition of nitrates and 

 other compounds of that element. 



It was impossible to trace in detail this complex circu- 

 lation of nitrogen in Nature, and Hellriegel and Wilfarth's 

 experiments brought the question to this crux. 



If leguminosse and gramineae are grown side by side 

 in pots, in nitrogen-free soil (or soil in which the known 

 nitrogen contents are small), the gramineae die of nitrogen 

 starvation : so do the leguminosas, unless their roots are 

 infected and develop plenty of nodules containing the sym- 

 biotic 'organism. In the latter case the closed system of pot, 

 soil, and plant gains in combined nitrogen to such a large 

 extent that there is no doubt the excess of nitrogen comes 

 ultimately from the free nitrogen of the air. 



Of the truth of this conclusion I convinced myself by a 

 series of experiments at Cooper's Hill, and it has been 

 abundantly confirmed by many other observers. 



The controversies which raged about this time, and in 

 which Frank, Berthelot, Beyerinck, Breal, Lawes and Gil- 

 bert, Koch, Nobbe and Hiltner, and Prazmowsky, were 

 especially prominent (22), led to extremely useful informa- 

 tion and ideas which would carry us too far if examined in 

 detail. The new knowledge did not alter the above posi- 

 tion, and the results can best be made clear by shortly sum- 

 marising the results of experiments by Laurent and Schloes- 

 ing, Koch and Kossowitsch and one or two others. 



Laurent and Schloesing (24) repeated the old experi- 

 ments of Boussingault and Lawes, Gilbert and Pugh, under 

 the new conditions necessitated by the new knowledge. 



