INFLUENCE OF WATER 277 



varying moisture content to stand for some time and then inoculated 

 it into mannite solutions and obtained maximum fixation in the soils 

 containing fairly small quantities of water. Later, however, he 

 decided that soil should be damp but not wet and well aerated 

 for maximum nitrogen-fixation. The water requirements vary with 

 different soils. As a general rule, the higher the humus content of 

 the soil, the more water will be required for optimum nitrogen- 

 fixation. The quantity of water present may, however, become so 

 great that it may kill all Azotobacter in addition to stopping nitrogen- 

 fixation. 



An insufficient supply of moisture checks both nitrification and 

 nitrogen-fixation. This occurs in some soils when the water content 

 has been reduced to 16.5 per cent. This again varies with the soil, 

 for Schlosing found bacterial activity less in fine-grained soils than 

 in lighter, coarse-grained soils. A difference in moisture content 

 of 1 per cent, according to Dafert and Bollinger, is sufficient to pro- 

 duce a marked change in the oxidation going on in the soil. 



The moisture requirement of the nitrogen-fixing bacteria, accord- 

 ing to Lipman and Sharp, is more nearly that of the ammonifying 

 than of nitrifying organisms. In a sandy loam it was found to vary 

 between 20 and 24 per cent, the anaerobic nitrogen-fixers are most 

 active, but the action of the aerobes is slightly depressed. Thus, 

 in many soils two maxima of nitrogen-fixation occur, depending upon 

 whether the conditions are favorable for the anaerobic or aerobic 

 organisms. 



Traaen's results differ from Lipman's in showing only the one 

 maximum, as is seen from the following, which gives the milligrams 

 of nitrogen fixed in 100 gm. of soil. 



5 per cent. 10 per cent. 17.5 per cent. 25 per cent. 30 per cent. 

 Temperature. H S O. H S O. H 2 O. H 2 O. H 2 0. 



13C 0.1 1.5 11.2 13.4 5.4 



25C 1.9 1.9 13.2 16.6 15.5 



He used a loam soil with a maximum water capacity of 27.4 per 

 cent. It is quite evident from his statement that anaerobic organ- 

 isms played a prominent part in the fixation at the higher moisture 

 contents. 



Since the carbohydrates disappeared much more rapidly in the 

 soils containing the greater quantities of water, it is quite possible 

 that greater quantities of nitrogen per gram of carbohydrate con- 

 sumed are fixed where the smaller quantities of water are applied. 

 This, together with the different methods used by the several 

 investigators, would explain the apparent discrepancy in their 

 results. 



In a series of pot experiments in which a calcareous loam receiving 

 various amounts of water was used, the author found the moisture 

 content for maximum nitrogen-fixation to lie between 15 and 22 per 

 cent. These results also bring out the two maxima which were first 



