the ground and light they use would be denied to the seedlings,
but undoubtedly the yield would be greater than could be secured
if they were destroyed.

This means that under similar conditions we may go still further
and actually apply the selection system, especially if the original
stand is nearly pure hemlock. So far we have discussed areas left
by present-day logging methods. Suppose, however, the owner of a
good tract of hemlock, having decided that conditions do not warrant
trying to get fir, is willing to modify his methods for the sake of
better hemlock returns at some future cutting. He would probably
do best to take out only the mature trees, leaving everything which
is still growing with fair rapidity. Greater light will stimulate
these immensely as well as encourage further seeding of the ground.
The few merchantable trees he spares, together with those now
unmerchantable, will, in perhaps twenty years, make another excellent
crop. By leaving a fairly dense stand he prevents the windfall
danger which threatens the survivors of too vigorous cutting, and
also prevents them from assuming the branchy form of trees which
receive too much side light. The fire danger is much reduced by
resultant shading of the ground and slightly by the lesser cover of
debris. In short, he makes the most economical use of the ground,
and the capital represented by the trees he spares is well invested.

To sum up, hemlock lends itself to almost every form of management.
Determination as to which is most advisable is governed by its
extremely variable manner of occurrence and by the local promise
offered by associate species. The foregoing discussion can only
serve as suggestive when considering given conditions.