E. J. Russell and E. H. Richards 549 



and it is therefore not difficult to suppose that ammonia would oxidise 

 on the same model : 



H H H 



I i ■ I 



H— N— H -^ H— N— O— H -^ N— O— H ^ = N— O— H 



o o 



It is known also that nitrates and nitrites are rapidly reduced to 

 gaseous nitrogen under anaerobic conditions. 



This hypothesis is consistent with the general results obtained by 

 sewage workers who, as already stated, have accumulated a vast 

 amount of data on the subject. They have shown that there is no loss 

 of nitrogen under anaerobic conditions, — that it only goes on under 

 aerobic conditions, but that beyond a certain point excess of air reduces 

 the loss. Thus, on the contact bed the conditions, though aerobic, are 

 less so than on the percolating filter, and the loss is greater : the higher 

 amount of dissolved oxygen on the percolating filter reducing the loss. 

 Similarly when the dissolved oxygen is increased in any other way (e.g. 

 by adding tap-water), a reduction of the loss occurs (see p. 527). 



Thus the alternate nitrification and denitrification hypothesis will 

 fit all the facts. But it is by no means definitely established, and it 

 suffers from a rather serious drawback : it involves either a transfer 

 of nitrate from the exterior to the interior of the heap in conditions 

 where actual movement is difficult to imagine (e.g. in heaps under 

 cover or manure stored in bottles), or a formation of nitrate in the 

 interior of the heap where we have never yet been able to observe it. 

 We therefore prefer another hypothesis which equally fits the facts 

 and has the advantages of freedom from any mechanical difficulty, 

 and of being the more general hypothesis of which alternate nitrification 

 and denitrification is only a particular case. 



We have already seen (p. 524) that under anaerobic conditions 

 molecular groupings tend to arise, which become unstable as soon as 

 aerobic conditions set in, and vice versa. These changes have usually 

 been studied in non-nitrogenous compounds, but it is reasonable to 

 suppose that nitrogen compounds may be similarly affected also. If, 

 for example, a derivative of propionic acid formed under anaerobic 

 conditions can on entry of air shorten its chain by eliminating — CII2 — , 

 and becoming a derivative of acetic acid, it is at least likely that some 

 complex nitrogen compound could simplify itself by eliminating nitrogen. 



Assuming the possibility of such a change the mechanical conditions 

 present no difficulty. Where air penetrates into the heap the conditions 



3(^—2 



