ce, or to prevent a night attack by destroyers,
no sure means has yet been found except examination before dark of
a very large area around the fleet that is sought; but the area
is too great for a search rigid enough to give complete security,
and will probably be so until swift aircraft can scout over long
distances at sea. Accepting for the minute the convention that
the main body of each side goes at the cruising speed of 10 knots,
and that darkness lasts 12 hours, each side will go 120 miles in
darkness; and if the two main bodies happen to be going directly
toward each other they will approach 240 miles in the darkness of
one night. Therefore, a coming fleet, in order to feel entirely
safe, would in daylight have to inspect by its scouts a circle
of 120 miles radius. To insure safety against destroyer attack,
the area would have to be much greater on account of the greater
speed of destroyers.

[Illustration]

Unless our defending fleet knew with reasonable sureness, however,
the location, speed, and direction of motion of the coming fleet,
so that it could make its dispositions for attack, it would hardly
desire to meet the enemy at night, unless it were confident that
it would meet the train and not the main fleet or the destroyers.
Night attacks, both on sea and land, are desirable, if the attacker
can inflict surprise on the attacked, and not be surprised himself.
In the darkness a flotilla of destroyers may make an attack on
the various vulnerable colliers and supply vessels of a fleet,
or even on the main body, and achieve a marked success, because
that is the role they are trained to play. But the tremendous power
and accuracy of battleships cannot be utilized or made available
in darkness; and therefore a commander-in-chief, anxious to defeat
by superior skill a coming fleet larger than his own, would hardly
throw away all chance of using skill by risking his main body in a
night encounter. Every operation planned by strategy is supposed
to result from the "decision" which follows the estimate of the
situation; even if in some simple or urgent cases, the decision
is not laboriously worked out, but is almost unconscious and even
automatic. Now, it is hardly conceivable that any estimate of the
situation would be followed by a decision to go ahead and trust to
luck, except in very desperate circumstances. In such circumstances,
when hope is almost gone, a desperate blow, even in the dark, may
save a situ