dly_, by bringing up a soil which has
not been exhausted by previous cropping--in fact a virgin soil.

The success which attends the operation of subsoiling or deep ploughing
must manifestly be greatly dependent on the character of the subsoil,
and good effects can only be obtained when its chemical composition is
such as to supply in increased quantity the essential constituents of
the plant; and it is no doubt owing to this that the opinions
entertained by practical men, each of whom speaks from the results of
his own experience, are so varied. The effects produced by deep
ploughing on the estates of the Marquis of Tweeddale, are familiarly
known to most Scottish agriculturists, and they are at once explained by
the analyses of the soil and subsoil here given, which show that the
latter, though poor in some important constituents, contains more than
twice as much potash as the soil.

Soil. Subsoil.

Insoluble silicates 87.623 82.72
Soluble silica 0.393 0.12
Alumina and peroxide of iron 4.129 8.60
Lime 0.341 0.18
Magnesia 0.290 0.24
Sulphuric acid 0.027 0.03
Phosphoric acid 0.240 trace
Potash 0.052 0.12
Soda 0.050 0.04
Water 1.956 3.26
Organic matter 5.220 4.02
------ -----
100.321 99.33

In addition to the difference in the amount of potash, something is
probably due to the large proportion of alumina and oxide of iron in the
subsoil, which for this reason must be more tenacious than the soil
itself, which appears to be rather light. In other instances, the use of
the subsoil plough has occasioned much disappointment, and has led to
its being decried by many practical men; but of late years its use
having become better understood, its merits are more generally
admitted. We believe, that in all cases in which the soil is deep, more
or less marked good effects must be produced by its use, but of course
there must be cases in which, from the defective composition of the
subsoil or other causes, it must fail. It may sometimes be possible _a
priori_ to detect these cases, but in a large majority of them our
knowledge is still too limited to admit of satisfactory conclusions
being a