66 



fixation. It is probable that under field conditions, as Koch and SeydeP 

 have shown, the removal of nitrogen by crops is necessary to further fix- 

 ation. The results obtained from the above tests do not indicate that the 

 presence of nitrogen is of material influence on fixation. 



As previously stated an effort was made to observe the influence of soil 

 texture on fixation. It would be interesting to note the intensity of the de- 

 velopment of the Azotobacter flora in soils of widely different types and to 

 investigate the marked differences where they do appear. The widespread 

 occurrence of Azotobacter is not suggestive of its susceptibility to difference 

 in soil types. Von Feilizen, however, has reported the very slight occurrence 

 of Azotobacter in peat 2 and moor 3 soils, even after long periods of cultiva- 

 tion, and, though probably without much justification, the idea has grown up 

 that fixation is more marked in soils of light open texture than in the 

 heavier soils. Since cultivation obviously affects the process and the culti- 

 vated soils considered in this paper received cultivation varying widely in 

 kind and intensity, it would be illogical to compare results from them as 

 evidence in support of fixation by any special type. It is only, therefore, 

 the virgin samples which are considered in this relationship. A general 

 classification of five different types was found to include all of the 44 virgin 

 samples, and, while the number of soils was not the same in every classifica- 

 tion, an average of the nitrogen fixed by each class would reveal any note- 

 worthy difference between them. In Table III is given the soils with their 

 numbers which were thus classified and the average in milligrams of nitro- 

 gen which each type fixed. 



TABLE III. Average Nitrogen Fixed ~by a Number of Virgin Soils of 



Various Texture. 



The comparison would perhaps be more fair if the number of soils in 

 each type were the same, but since the average is taken it seems justifiable 

 to report on all the soils. Collected as the soils were with reference only to 

 soil type and having received no treatment whatever, the results have a 

 considerable degree of interest. The nitrogen fixing flora has presumably 



and Seydel. Cent. f. Bakt. Abt. II, Bd. 31. 1911. p. 570. 

 Feilitzen von H., Svenska Mosskulturfor. Tdskr. 25, 1911, No. 1, pn. 53-57. 

 Feilitzen von H., Fuhling's Landw. Ztg. 59. 1910. No. 14, pp. 489-492. 



