954 Journal of Agricultural Research vol. vi.no. 24 



ported a limited number of experiments following 30 years of continuous 

 treatment. 



It was with the ultimate object of attempting a study of the funda- 

 mental causes for variations in certain changes in soil nitrogen that this 

 work was undertaken. Before it could be begun, however, it was first 

 necessary to determine the existence of differences under our conditions. 

 Furthermore, it seemed desirable to compare the relative effect of long- 

 continued treatment with that of the shorter periods reported by others. 



The Missouri Experiment Station possesses a series of fertility plots 

 that offered exceptional opportunity for making the above-mentioned 

 studies. The plots cover a rather wide range and have just passed their 

 twenty-fifth year of continuous treatment as outlined in the original 

 project. Some very valuable data as to the effect upon fertility as 

 measured by crop-producing power have been obtained. It would 

 seem as if 25 years would have so materially changed the micro- 

 organic life therein that such could readily be detected, provided such 

 differences were actually brought about. With such material upon which 

 to work it would seem that the verity of marked differences in similarly 

 treated soil reported by others could be established. 



The data herein reported have been obtained from some of the plots 

 that have given the most marked differences in yield, as it was believed 

 such plots would offer the best material upon which to work. The work, 

 however, is concerned only with demonstrating the existence of differ- 

 ences and offers only one or two suggestions as to the actual cause of 

 such differences. We hope to be able to throw more light upon this 

 particular field at a later date. Furthermore, the particular data here 

 reported have to do only with bacterial numbers and with ammonia and 

 nitrate-forming abilities. 



Certain facts, which, we believe, we have very clearly demonstrated, 



have been of value to us in directing further work. It is with the hope 



that the data may be of similar value to others that we present them in 



this paper. 



PLOTS STUDIED 



The fertility plots of the Missouri Experiment Station are located on 

 the soil type classed as Putnam silt loam. They were first planted to 

 the present system in 1889 and, with few irregularities, have received the 

 same treatment as outlined. Each plot consists of one-tenth acre and 

 is surrounded by an alley 3 feet wide. In selecting from the large num- 

 ber of plots the few that could be handled in our work, an effort was 

 made to include as wide a variation of treatment as possible and at the 

 same time to avoid inherent soil differences in order that the work might 

 be comparative. 



The plots studied are, with the exception of No. 29 and 30, located 

 just at the crest and on the eastern slope of a gentle rise. Plots 29 and 



