Garden," a plot of ground lying between the flower garden and the
spinner's house. But he had experimental plots on most or all of his
plantations, and each day as he made the rounds of his estate on
horseback he would examine how his plants were growing or would start
new experiments.

The record of failures is, of course, much greater than of successes,
but that is the experience of every scientific farmer or horticulturist
who ventures out of the beaten path. Even Burbank, the wizard, has his
failures--and many of them.

One of Washington's successes was what he called a "barrel plough." At
that time all seed, such as corn, wheat and oats had to be sown or
dropped by hand and then covered with a harrow or a hoe or something of
the kind. Washington tried to make a machine that would do the work more
expeditiously and succeeded, though it should be said that his plans
were not altogether original with him, as there was a plan for such a
machine in Duhamel and another was published by Arthur Young about this
time in the _Annals of Agriculture_, which Washington was now perusing
with much attention. Richard Peters also sent yet another plan.

Washington's drill, as we should call it to-day, consisted of a barrel
or hollow cylinder of wood mounted upon a wheeled plow and so arranged
that as the plow moved forward the barrel turned. In the barrel, holes
were cut or burnt through which the corn or other seed could drop into
tubes that ran down to the ground. By decreasing or increasing the
number of holes the grain could be planted thicker or thinner as
desired. To prevent the holes from choking up he found it expedient to
make them larger on the outside than on the inside, and he also found
that the machine worked better if the barrel was not kept too full of
seed. Behind the drills ran a light harrow or drag which covered the
seed, though in rough ground it was necessary to have a man follow after
with a hoe to assist the process. A string was fastened to this harrow
by which it could be lifted around when turning at the ends of the rows,
the drill itself being managed by a pair of handles.

Washington wrote to a friend that the drill would not "work to good
effect in land that is very full either of stumps, stones, or large
clods; but, where the ground is tolerably free from these and in good
tilth, and particularly in light land, I am certain you will find it
equal to your most sanguine expectation, for Indian corn,