6o2 Journal of Agricultural Research voi. xv, no. n 



In the nitrogen-fixing experiments 50 cc. of a sterile culture solution 

 in 350-cc. Erlenmeyer Jena flasks were inoculated with 2 or 5 gm. of 

 soil and incubated, usually for three weeks. A medium consisting of the 

 following compounds was used : 



Potassium phosphate (Kn2HP04) o. 2 gm. 



Magnesium sulphate (MgS04) 2 gm. 



Sodium chlorid (NaCL) 5 gm. 



Mannit 20. o gm. 



Ferric chlorid (FeClj) Trace. 



Water i, 000 cc. 



This solution was made slightly alkaline to phenolphthalein with 

 sodium hydroxid. In other experiments this medium has never failed 

 to give the characteristic Azotobacter film if the organisms were present 

 in sufficient numbers and other conditions were favorable. The figures 

 reported are in all cases actual gains in nitrogen. 



Inoculations were made in duplicate. During incubation the growth 

 of Azotobacter was noted, and at the end of the incubation period the 

 total nitrogen was determined. Only the average of duplicate nitrogen 

 determinations are reported. 



In the tables giving results the Azotobacter film has been reported as 

 typical if the characteristic brown or black uniform growth covered the 

 surface. A question mark (?) has been placed opposite those samples 

 from which only one duplicate gave a film or when there were Azoto- 

 bacter present, but which failed to give the characteristic growth. 



Nitrate nitrogen was determined in duplicate by the phenol-disul- 

 phonic-acid method. Qualitative tests for ammonia were made with 

 Nessler's reagent. Where the aqueous extract of the soil gave only a 

 slight yellow color, it was reported as a trace. When the color was some- 

 what more pronounced, it was reported as slight; and where a heavy 

 brick-red precipitate was formed when the reagent was added, it was 

 reported as abundant. In some instances quantitative determinations 

 of ammonia nitrogen were made by distillation with magnesium oxid. 

 A number of determinations upon samples reported as a trace showed 

 from 0.3 to 0.5 mgm. of nitrogen per 100 gra. of soil. Variations from 

 these methods were sometimes made, but they are mentioned in con- 

 nection with the individual experiment to which they apply. 



EXPERIMENTAL DATA 



The soil for these experiments was furnished through the kindness of 

 Prof. Walter G. Sackett, of the Colorado Agricultural Experiment 

 Station. It came "from a locality where the niter trouble has been very 

 severe and where nitrification has evidently been very active in the 

 past if not at present. " This soil was used because of the exceptionally 

 vigorous nitrogen-fixing (Azotobacter) flora. 



