24 
SIR J. B. LAWBS AND PROFESSOR J. H. GILBERT ON THE 
Indeed, so variable is the amount of nitrogen in samples of our Rothamsted subsoils 
taken on one and the same plot, dependent on the varying proportions of clay, sand, 
gravel, chalk, &c., that as we have fully illustrated in former papers, no estimates of 
the difference in the amounts of total nitrogen, either in the subsoil of the same plot at 
different periods, or in the subsoils of plots differently manured, or differently cropped, 
can be relied upon. 
Again, as the amount of the “original” or already existing nitric-nitrogen is, as has 
been very fully shown, greatly dependent on the description, and on the amount of 
crop that has been grown, and other circumstances, it is not to be expected that it would 
bear any direct relation to the richness or poverty of the subsoil in total nitrogen. 
Referrinsf to the amounts of nitric-nitrogen formed in the different subsoils, and 
within the different periods of exposure, there is, as is to be expected in the case of 
an action depending on the development and activity of an organism of the habits, 
requirements, and mode of action of which we know but little, considerable irregularity, 
both from period to period with the same sample, and within each period with the 
different samples. 
Confining attention in the first place to the results relating to the wheat-fallow, 
and the leguminous crop subsoils, recorded in the upper two divisions of the Table, 
it will be seen that the first, second, third, and fourth periods of exposure each 
comprised 28 days; whilst the fifth period extended over 246 days, or about 35 weeks, 
during a considerable portion of which, however, the soils had doubtless become too 
dry for activity and nitrification. It is^ probable that a period of 28 days is too short 
to insure the active development of the organisms, and consequent energetic nitrifica¬ 
tion. Then again, the extraction of the soils by water under pressure must, it is to 
be supposed, remove some of the organisms, instead of allowing of themAiatural 
multiplication. Comparison of the results from period, to period must, therefore, be 
made with some reservation. 
But apart from any irregularities in the case of individual samples, or at individual 
periods, if we compare for each period the mean results for the three wheat-fallow 
samples, with the means for the leguminous crop subsoils, it is seen that during four 
of the five jJeriods, the leguminous crop subsoils show considerably more nitrification 
than the wheat-fallow ones; and whilst over the total period the wheat-fallow subsoils 
show an average of 0‘791 nitrogen nitrified per million of soil, the leguminous crop 
subsoils show 1'056 per million. 
Ag ain, the figures in the bottom division of the Table, relating to the wheat-fallow 
and the rotation clover land samples, collected in the spring of 1886, show in all cases 
comparable as to length of period, more nitrification in the clover land than in the 
wheat-fallow land subsoil; and this is so both with the unseeded and the seeded 
samples. 
Thus, then, these results with the raw, and mostly clay, Rothamsted subsoils, 
containing not more than 6 or 8 parts carbon to one of nitrogen, confirm those 
