rom wet places to dry
places in the soil. Tie a piece of muslin over the end of a tube and
fill with dry soil, tapping it down as much as you can, then stand the
tube in water as in Fig. 30. Fill another with sand {68} and place in
water. Notice that the water at once begins to rise in both tubes and
will go on for a long time, always passing from the wet to the dry
places. It rises higher in the soil than it does in the sand. Enough
water may pass up the tube in this way to supply the needs of a growing
plant. Fill a glass lamp chimney with dry soil, packing it down
tightly, put into water and then sow with wheat. The plants grow very
well. A longer tube may be made from two chimneys fastened together by
means of a tin collar stuck on with Canada balsam or sealing wax (Fig.
31). Our plants grew well in this also, but on a sandier soil, where
the water could not rise so high, it might happen that they would not.

[Illustration: Fig. 31. Wheat growing in soils supplied from below
with water. All the water the plant gets has to travel upwards]

Thus we shall expect great differences in the moisture of various
soils. In some districts there is much more rain than in others, and
therefore the soils get a larger supply of water. Sandy soils allow
water to run through while a loam holds it like a sponge, in a loam
also the water readily moves from wet to dry places. Further, water
runs down hills and collects in low-lying hollows or valleys; here,
therefore, the soil is moister than it is somewhat higher up. What
will be the effect of these moisture differences on plants?

You must find out in two ways. Visit a soil that you know is dry--a
sandy, gravelly or chalky soil in a high situation--and look carefully
at the plants there, then go to some moister, lower ground and see what
the plants show. You cannot be quite certain, however, that anything
you see is simply due to water supply, because there may be other
differences in the soil as well. So you must try the second method,
and that is to find out by experiments what is the effect of varying
{69} quantities of water on the plant growth. Both methods must be
used, but it may be more convenient to start the experiments first, and
while they are going on to collect observations in your rambles.

[Illustration: Fig. 32. Mustard growing in soils supplied with varying
quantities of water. 16 very little water, 3 a nice amount of water,
15 too much water