1362 
MESSES. J. B. LAWES, J. H. GILBERT, AND M. T. MASTERS, 
mineral manure, a given amount of nitrogen as nitrate of soda yielded more hay, more 
nitrogen, and more mineral matter, and that there was a less reduction in the yield 
of each in the later years, than when the same amount was applied as ammonia-salts. 
Now, when each is applied in conjunction with mineral manure, there is again more 
hay produced, and more mineral matter taken up, but rather less nitrogen, with the 
nitrate. But, whilst there is a decrease in each item in the later years on plot 9, with 
the mineral manure and ammonia-salts, there is an increase of each on plot 14, with 
the mineral manure and nitrate of soda. 
Referring to Part I. for a fuller consideration of these points, and to Part III. (to 
follow) for still more detailed evidence, it must suffice here to state very briefly some 
of the conditions of these differences of result on the two plots. The nitrogen of the 
nitrate distributes more rapidly both in the upper and in the lower layers of the soil 
than does that of the ammonia-salts. Accordingly, both free-surface-feeders and more 
deeply-rooting species are encouraged on the nitrate plot, and hence the collective 
herbage on that plot is less susceptible to the adverse influence of drought than that 
on plot 9. A striking instance of this occurred in 1870, and the conditions are 
somewhat fully described in Part I. already referred to ; but it may be stated briefly 
that whilst in that year of very unusual drought there was a deficiency compared with 
the average of about 23 cwts. of hay on the mineral and ammonia plot 9, there was a 
deficiency of less than '1-J cwt. on the mineral and nitrate plot 14. On this plot, the 
deep and wiry-rooted Bromus mollis contributed nearly half the herbage in that year ; 
and examinations and analyses of the soils and subsoils of the plots showed much the 
more extensive root-development in the lower layers on plot 14, and very much less 
moisture there remaining. It is obvious that the deeper-rooting species had drawn up 
much of their needed water from the subsoil; partly, perhaps, directly, and partly by 
virtue of increased capillary action induced by the pumping out of the upper layers. 
It was also found that, probably in part due to the action of the roots, and in part to 
chemical reactions, the subsoil of the nitrate plot was more disintegrated and amelio¬ 
rated than that of either plot 9 or the unmanured plot 3. 
With these few general remarks, indicating some of the characteristic differences of 
result on the two plots, we now turn to a detailed consideration of their comparative 
botany. 
The following Table (LXXV., pp. 1364-5) gives, in the usual form, a record of the 
facts relating to the plot under consideration ; it also shows the amounts yielded by 
the different orders and species compared with those on the unmanured plot 3, and on 
plot 9 with the mineral manure and the nitrogen supplied as ammonia-salts. 
In the first place there were, on the average, two fewer Graminese, two fewer 
Leguminosse, and 16 fewer Miscellaneae, than without manure, the average number 
found in the four separations in the samples from plot 14 being 29. Compared with 
plot 9 there was but little difference in number; in fact, only one more species. 
There is not very much difference in the average percentages of total Graminese, 
