. Now only
the top inch is touched by the hoe, and so it appears that the layer
thus loosened shields the rest of the soil from the sun's heat. If
this is the case we ought to find that any other loose material would
act in just the same way. We must, therefore, set out a fourth plot
alongside the others, cover it with straw or cut grass (a cover like
this is called a mulch), and take the temperature there. Some of the
results were as follows:--

Air
Date temperature Soil temperature

Hoed plot Mulched plot

1910
Sept. 24th 15 1/2 inch 17.5 12.25
3 inches 12.5 11.75
6 inches 12.25 11.5

Oct. 5th 17 1/2 inch 17 15.5
3 inches 16.7 15
6 inches 15.5 14.5

Remarks. Sept. 24th: Warm day after a rather cold spell. Oct. 5th:
After a long spell of dry, warm weather.

The untouched plot had become smothered in weeds and could no longer be
used for this experiment. The mulched soil is, however, cooler even
than the hoed soil, and our expectation that mulching would keep the
soil cool has turned out to be correct.

{88}

[Illustration: Fig. 42. Soil sampler. (See p. 82 for description)]

It may be expected that the hotter soil--the unhoed plot--will also be
drier than the others, and this can be found out by a simple
experiment. Take a sample by making a hole six indies deep with
straight and not with sloping sides: this is best done by driving a
tube two inches wide into the soil (Fig. 42): if you have not got such
a tool you may use a trowel, but you will have to be very quick and
very careful. Weigh the soil--or a part of it if you have got a great
deal--then set it to dry in a warm place for three or four days. Weigh
again when it is dry: the difference gives the loss of water: find what
it would be in a hundred parts. Our results were:--

{89}

Date Percentage of water in the

Untouched soil Soil hoed once Soil hoed three
weekly times weekly

1910
June 4th 21.1 19.5 17.9

June 20th 14.7 16.0 16.0