454 



SCIENCE. 



[N. S. Vol. XXII. No. 563. 



rapid and considerable nitrogen fixation if 

 the necessary conditions, chiefly those of 

 carbohydrate supply, are satisfied. 



But how is the carbohydrate supply to 

 be obtained ? Under the normal conditions 

 of arable land farming there are few pos- 

 sibilities in this direction, the occasional 

 ploughing under of a green crop being 

 the only considerable addition of organic 

 matter other than manure, which is possible 

 in practise. As a matter of experience the 

 plots at Rothamsted, which have been 

 growing crops without manure contin- 

 uously for the last fifty years, indicate but 

 little gain of nitrogen from the atmos- 

 phere. After a rapid fall in production 

 for the first few years, the yield has become 

 so nearly stationary that any further de- 

 cline is not as yet discernible amid the 

 fluctuations due to season. 



of the soil, the material brought down by 

 the rain, and the nitrogen-fixing agencies 

 taken together are just equal to providing 

 the crop with about 17 lbs. of nitrogen per 

 acre per annum in addition to the unknown 

 amounts removed by drainage and in the 

 weeds. The small amount of fixation this 

 indicates and the corresponding low level 

 of production must be set down to the lack 

 of combustible carbohydrate, due to the 

 very complete removal of the various crops 

 from the soil, since the root and stubble 

 left behind after the growth of a cereal 

 crop amount to but a small fraction of the 

 total produce. 



In the case of grass-land the conditions 

 are entirely different, especially when we 

 are dealing with wild prairie or forest, 

 where the annual growth of carbohydrate 

 falls back to the soil and is available for 



TABLE III. 



Average Amounts of Dry Matter and Nitrogen in Total Produce of Various Crops, grown 



loithout Manure at Rothamsted. 



Table III. shows the average yield dur- 

 ing the last five decades of dry matter and 

 of nitrogen from four of the unmanured 

 plots at Rothamsted; it will be seen that 

 the difference in the production during the 

 last as compared with the second period of 

 ten years is no more than would be cov- 

 ered by seasonal variations. In other 

 words, the yield, which, as we learn in 

 other ways, is mainly determined by the 

 amount of available nitrogen, has reached 

 a state of equilibrium when the resources 



^ Carted fallow portion. 

 ^ First and second crops. 



such organisms as the Azotohacter. At 

 Rothamsted two plots of land which were 

 under arable cultivation twenty-five years 

 ago have been allowed to run wild and ac- 

 quire a natural vegetation of grasses and 

 weeds, subject to no disturbance beyond 

 the occasional eradication of scrub and 

 bushes. Samples of the soil taken when 

 the land was still under the plough have 

 been preserved, and the comparison of 

 these with new samples drawn during the 

 last year shows enormous accumulations of 

 nitrogen, even when every allowance has 

 been made for certain inevitable errors in 



