October 13, 1905.] 



SCIENCE. 



455 



sampling the soil (see Table IV.). Of 

 these two fields the Geescroft plots are the 

 more interesting, for though showing the 

 gain of nitrogen is less (45 lbs. per acre per 

 annum against 98 lbs. on Broadbalk), yet 

 continued observation of the herbage that 

 has sprung upon this field has shown the 

 absence of any leguminous plants. Accord- 

 ing to a botanical analysis made in 1903 

 the leguminous plants only constituted 0.4 

 per cent, of the vegetation (as weighed in 

 the dry state) on the Geescroft 'wilder- 

 ness,' whereas the corresponding plot on 

 Broadbalk contained 25 per cent. Now, 

 with no leguminous plants to act as col- 

 lectors of nitrogen the considerable gains 

 of combined nitrogen on this Geescroft land 

 must be set down to the work of Azoto- 

 hacter or kindred organisms which get 

 their necessary supply of carbohydrate 

 from the annual fall of the grassy vegeta- 

 tion. 



TABLE IV. 



Accumulation of Carbon and Nitrogen in Soil of 



Land allowed to run toild for more than 



Twenty Years. 



The fixation of nitrogen must be an 

 oxidizing process, for no other natural rc- 

 * Broadbalk, 1881; Geescroft, 1883. 



action is likely to provide the energy neces- 

 sary to bring the nitrogen into combina- 

 tion. This being so, some light is thrown 

 on the process in nature by an examination 

 of the ratio of carbon to nitrogen in the 

 accumulations referred to above. At 

 starting, the ratio of carbon to nitrogen in 

 the organic matter of the two soils was 

 much the same— a little less than 10 to 1 

 — but the increase of carbon and nitrogen 

 in the Broadbalk field, i. e., the organic 

 matter which has accumulated in the 

 interim, shows a ratio of only 3 to 1, while 

 the corresponding accumulations in Gees- 

 croft field show a ratio not far removed 

 from the original of about 11 to 1. In 

 other words, Avhere there has been the 

 greater accumulation of nitrogen on the 

 Broadbalk field, there has been the greater 

 combustion of carbohydrate, so that the 

 accumulation of carbon is actually as well 

 as relatively smaller. Bacteriological tests 

 seem to show a much greater development 

 of Azotohacter with increased powers of 

 fixation in the soil from the Broadbalk than 

 from the Geescroft wilderness; a fact to be 

 correlated with the presence of a fair pro- 

 portion of carbonate of lime in Broadbalk, 

 but not in Geescroft field. 



Another example may be drawn from 

 the experiments carried on by the late Mr. 

 James Mason at Eynsham Hall, Oxon. 

 He had large cemented tanks filled with 

 burnt clay mixed with appropriate quan- 

 tities of calcium carbonate and phosphate 

 and other nutrient salts, but containing no 

 nitrogen. One of these tanks, after inocu- 

 lation with a trace of ordinary soil, was 

 sown with a mixture of grass-seeds and 

 has carried a weak but purely grassy vege- 

 tation ever since. According to a recent 

 analysis the soil of this tank has in fifteen 

 years accumulated 0.029 per cent, of 

 nitrogen in the surface soil and 0.117 per 

 cent, in the second layer — equivalent to 

 about 870 and 350 lbs. per acre per annum, 



