RECLAIMING LOST NITROGEN. 133 



source of error, they cannot assimilate their nitrogen out of the 

 atmospheric store, but must depend upon the soil ; and that 

 they can only use compounds of nitrogen which are in the soil. 

 Free nitrogen is useless for them. But though ordinary green 

 plants cannot use it, it is perfectly evident that there is some 

 means by which the atmospheric nitrogen can be fixed in the 

 soil in a combined form. Our next problem, then, is to search 

 for the agencies which reclaim this atmospheric nitrogen. 



NITROGEN FIXATION IN FREE SOIL. 



We find in the first place that soil itself if simply allowed to 

 stand untouched for a number of weeks will gain nitrogen 

 without any visible vegetation of plants. A lot of earth placed 

 in a proper vessel and kept free from vegetation will, in time, 

 be found to contain more nitrates than at the outset. Part of 

 these nitrates, doubtless, are due to the process of nitrification 

 already mentioned, by which the nitrogen compounds, \vhich 

 were in the soil, but not in the form of nitrates, are converted 

 into nitric acid by the nitrifying bacteria. But this is not the 

 whole explanation, because analysis of such soil shows that at 

 the end of several weeks there is actually a larger amount of 

 total nitrogen in the soil than there was at the start. If, then, 

 this total nitrogen has been increased, the most probable con- 

 clusion is that it has been in some way derived from the 

 atmosphere. One other suggestion is possible. It may be 

 that such increase in nitrogen is due to the absorption of am- 

 monia gas which is present in the air, at least in small quan- 

 tities. This has been disproved by experiments which have 

 shown that there is a similar absorption of nitrogen, with an 

 increase in nitrates, in soil which is bathed in air practically free 

 from ammonia. No one to-day questions the conclusion that 

 the source of the nitrogen increase in the soil under these 

 experiments is atmospheric free nitrogen. 



Further, it is abundantly demonstrated that this fixation of 



