QUESTION OF THE SOURCES OP THE NITROGEN OF VEGETATION. 15 
The next point to consider is, whether the continued growth of the various 
leguminous crops has reduced or increased the stock of nitrogen in the surface soils. 
The foregoing Table shows that in 1881 samples were taken from each of the five 
differently mineral-manured plots, but in 1882, 1883, and 1885, when samples were 
taken to considerable depths for the determination of the nitric acid, either plot 4, 
plot 5, or plot 6 was always selected, as on them the growth was better than on either 
plot 2 or plot 3. 
The following Table (VI.) summarises the percentages of nitrogen in the surface 
soils (9 inches deep) of the fallow-wheat land, of the Trifolium repens plots, of the 
Vida sativa plots, of the Melilotus leucantlia plots, and of the Mexlicago sativa plots, 
in the years as indicated. The figures are as before in all cases the means of 2, 3, 
or more determinations on each sample. In each case the results given in the first 
line are the means of determinations made on the individual samples taken from 
different places on the plot, those in the second line are the means of the determina¬ 
tions made on mixtures of the individual samples, and the general means given in the 
third line are the means of the results on the individual and on the mixed samples, 
taken together. It may be further explained that the wheat-fallow samples were 
taken in 1883 from 4, and in 1885 from 3 places on the plot. The Trifolium repens 
samples were taken in 1882 from 2 places on plot 6, in 1883 from 2 places on plot 4, 
and in 1885 from 2 places on plot 5. The Vida sativa soils were taken in 1883 from 
2 places on plot 4, and from 2 on plot 6. The Melilotus leucantlia soils were taken in 
1882 from 2 places on plot 5, and from 2 on plot 6, and in 1885 from 2 on plot 5. 
Lastly, the Medicago sativa soils were taken in 1885 from 2 places on plot 5. The 
determinations were, as before, made by the soda-lime method. 
