260 HOW CROPS FEED. 
were made of the soil and seed before the trial, and of the 
soil and crop afterwards. 
The analyses show that while in all cases the plants 
gained some nitrogen beyond what was originally contain- 
ed in the seed, there was in no instance any loss of nitro- 
gen by the soil, and in three cases the soil contained more 
of this element after than before the trial. Here follow 
the results. 
No. of Exp. Weight of Crop, Quantity of Soil. Gain of Nitrogen. 
the seed taken as 1. —————— 
by plant. by soil. 
1. Lupin,* 3% 130 grms. 0.0042 grms. 0.0672 grms. 
2. Lupin, 4 5B: | ae 0.0047:  D:O05E = 5 
3. Hemp, 5 sy 0.00389: ~ ** <O30600) =* 
4. Bean, 5 2 i 0.0226 -“* -0:0000'>. -* 
5. Lupin,* 3 130 * 0.0217 “ 0.0454 ‘ 
That the gain of nitrogen by the soil was not due to 
direct absorption of nitric acid or ammonia from the at- 
mosphere is demonstrated by the fact that it was largest 
in the two cases (Exps. 1 and 5) where the experiment was 
conducted ina closed vessel, containing throughout the 
whole time the same small volume, about 20 gallons, of 
air. 3 
In Exp. 4, where the soil at the conclusion contained no 
more nitrogen than at the commencement of the trial, it 
is scarcely to be doubted that the considerable gain of ni- 
trogen experienced by the plant came through the soil, 
and would have been found in the latter had it borne no 
crop. 
The experiments show that the quantity of nitrogen 
assimilated from the atmosphere by a given soil is very 
variable, or may even amount to nothing (Exp. 3); but 
they give us no clue to the circumstances or conditions 
which quantitatively influence the result. 1t must be ob- 
served that this fixation of nitrogen took piace here in a 
soil very rich in organic matters, existing m the condition 
of humus, and capable of oxidation, so that the soil itself 
* Experiments made in confined air. 
a 
