Soil Organisms concerned in Nitrogen Cycle. 515 



comparable results, but they attributed the beneficial effect of the humus 

 solely to the iron which it contains. In view of this statement, it must be 

 pointed out that both the purified sodium humate and the calcium humate 

 used in the above experiments contained appreciable quantities of iron, 

 which was presumably adsorbed by the colloidal humus, and was by no 

 means entirely removed by the subsequent solution and re-precipitation. 

 Further, while the sodium humate must contain quite as much iron as does 

 the water extract of bacterised peat, since both have been prepared from 

 the .same kind of raw peat, yet the effect of the bacterised peat is markedly 

 greater than that of the chemical preparations from raw peat. The 

 inference seems to be that during the bacterisation some soluble 

 organic substances are produced, besides the soluble humates, which have 

 the effect of greatly increasing the nitrogen fixation by Azotobader. These 

 substances appear to be formed in comparatively small quantities during 

 treatment of the peat with weak alkalies, since such a preparation increases 

 fixation to some extent, but they also appear to be lost during 'further 

 purification of the crude humate. Such substances are not present in raw 

 peat, at least in a water-soluble condition, since a water extract of such 

 peat depresses the rate of nitrogen fixation. That the beneficial effect 

 of the crude humus is not due to any physical action of the colloidal 

 extract is shown by the results with artificial humus, which appears to 

 have practically no effect. This is in accordance with the results of 

 Krzemieniewski.* 



The addition of natural humus to the culture medium of Azotobader, in 

 the hands of practically all investigators, has proved to be beneficial to 

 nitrogen fixation, but the degree of benefit obtained differs very widely, 

 rising in Krzemieniewski's* researches from 2 - 32 mgrm. nitrogen fixed 

 without humus to 21*52 mgrm. with sodium humate from soil. This 

 divergence is very probably due to the varying degree of bacterial decom- 

 position which has taken place in the humus before extraction ; the greater 

 the decomposition, the better being the result ; and it is most probable that in 

 bacterised peat the bacterial action has taken place under the circumstances 

 most favourable for the production of the essential organic substances. 

 This conclusion receives support from the work of Lohnis and Greenf , who 

 found that the humus from fresh stable manure increased fixation by 

 9 - 8 mgrm., while that from similar manure which had been "humified" 

 resulted in an increase of 14"4 mgrm. A similar experiment, which they 

 carried out with peat, showed no difference between the effect of fresh and 



* Krzemieniewski, ' Bull. Acad. Sci. Cracovie,' No. 9, pp. 929-1050 (1908). 

 t Lohnis and Green, ' Centr. Bakt. Par.,' Abt. II, vol. 40, pp. 52-60 (1914). 



2x2 



