Again, seeding from adjacent timber can often be relied upon. This
is a question of economy in logging operations, lay of the ground,
prevailing wind direction, fertility of the stand and other local
considerations. A valley with healthy fir woods on either side is
likely to seed up promptly even if a half mile wide. So is a flat
at the leeward foot of a hill timbered on the summit where the
wind strikes. A cutting on a ridge is correspondingly unlikely to
restock. Theoretically if a tract of timber were large enough, it
could be opened up by logging operations which, instead of proceeding
steadily from one edge, might skip every other landing or so until
the most remote portion was reached after a few years, and then
work back again, cleaning up the neglected portions after they
had seeded the first openings. The same effect sometimes results
from actual accidental practice.

It is apparent that rules cannot be laid down for general application.
Generally speaking, a logger interested in fir reforestation should
study his ground to see if naturally, or, with inexpensive aid, the
cut-over area will not reseed from the sides and from the cull trees
he will leave uncut. If not, he may leave a few merchantable seed
bearing trees provided the soil is such as to make them deep-rooted
and wind-firm. Groups are better than single trees because less
likely to be blown down and easier to protect from the slashing
fire. More should be left toward the windward edge. But before
tieing up any considerable sum in merchantable trees he should
consider the cost and safety of supplementing any shortage of natural
supply by artificial seeding.

WESTERN HEMLOCK (_Tsuga heterophylla_)