t or wasted.

Water may be lost then by being pumped up and transpired by weeds. And
this is the way weeds do their greatest injury to crops during dry
weather. The remedy is easily pointed out. Kill the weeds or do not
let them get a start.

There is another way, which we are not apt to notice, by which water
may be lost from the soil. When the soil in the pans in a previous
experiment (page 26) had been wet and set aside a few days it became
very dry. How did the water get out of this soil? That at the surface
of the soil evaporated or was changed into vapor and passed into the
air. Then water from below the surface was pumped up by capillary
force to take its place just as the water was pumped up in the tubes
of soil. This in turn was evaporated and the process repeated till all
of the water in the soil had passed into the air. Now this process is
going on in the field whenever it is not raining or the ground is not
frozen very hard.

Water then may be lost by evaporation.

How can we check this loss?

Suppose we try the experiment of covering the soil with some material
that cannot pump water readily.

=Experiment.=--Take four glass fruit jars, two-quart size, with
straight sides. If you cannot get them with straight sides cut off the
tops with a hot iron just below the shoulder; tin pails will do if the
glass jars cannot be had. Fill these with moist soil from the field or
garden, packing it till it is as hard as the unplowed or unspaded
soil. Leave one of them in this condition; from two of them remove an
inch or two of soil and replace it in the case of one with clean, dry,
coarse sand, and in the case of the other with chaff or straw cut into
half-inch lengths. Stir the soil in the fourth one to a depth of one
inch, leaving it light and crumbly. Now weigh the jars and set them
aside. Weigh each day for several days. The four jars illustrated in
Fig. 30 were prepared in this way and allowed to stand seven days. In
that time they lost the following amounts of water:

Amounts of water lost from jars of prepared soil in seven days.

No. 1 packed soil--lost 5.5 oz. equal to about 75 tons per acre.

No. 2 covered with straw--lost 2 oz. equal to about 27 tons per acre.

No. 3 covered with dry sand--lost 0 oz. equal to about tons per acre.

No. 4 covered with crumbled soil--lost 2.5 oz., equal to about 34 tons
per acre.

Why did not 2, 3 and 4 lose as much water as No. 1?

The soil in jar No. 1 was packe