il; this something we will call "plant food."

Further proof is easily obtained. At a clay or gravel pit little or no
vegetation is to be seen on the sloping sides or on the level at the
bottom, although the surface soil is carrying plants that shed
innumerable seeds. A heap of subsoil thrown up from a newly made well,
or the excavations of a house, lies bare for a long time. The
practical man has long since discovered these facts. A gardener is
most particular to keep the top soil on the top, and not to bury it,
when he is trenching. In levelling a piece of ground for a cricket
pitch or tennis court, it is not enough to lift the turf and make a
level surface; the work has to be done so that at every point there is
sufficient depth of top soil in which the grass roots may grow.

How much plant food is there in the top soil? To answer this question
we must compare soil that has been cropped with soil that has been kept
fallow, i.e. moist but uncropped. Tip out some of the soil that has
been cropped with rye, and examine it. Remove the rye roots, then
replace the soil in the pot and sow with mustard; sow also a fallow pot
with mustard. Keep both pots properly watered. The soil that has
carried a crop is soon seen to be much the poorer of the two. Fig. 22
shows the plants, while their weights in grams were:--

{45}

Green weight After drying

Mustard growing in soil previously
cropped with rye, Pot 1 17.8 62.3

Mustard growing in soil previously
uncropped, Pot 2 3.3 8.6

[Illustration: Fig. 22. Mustard growing in surface soil previously
cropped with rye (Pot 1) and in surface soil previously uncropped (Pot
2)]

{46}

The rye has taken most of the plant food that was in Pot 1 leaving very
little for the second crop. Our soil therefore contained only a little
plant food, not more, in fact, than will properly feed one crop. But
yet it did not seem to have altered in any way, even in weight, in
consequence of the plant food being taken out. In our experiment the
soil was dried and weighed before and after the mustard was grown; the
results were:--

Pot 2 Pot 2_a_

lbs. oz. lbs. oz.

Weight of dried soil before the experiment 6 6 6 7

" " " after "