ng a combat at night, and what concerns the
details of its course, is a tactical subject; we only examine it here so
far as in its totality it appears as a special strategic means.

Fundamentally every night attack is only a more vehement form of
surprise. Now at the first look of the thing such an attack appears
quite pre-eminently advantageous, for we suppose the enemy to be taken
by surprise, the assailant naturally to be prepared for everything which
can happen. What an inequality! Imagination paints to itself a picture
of the most complete confusion on the one side, and on the other side
the assailant only occupied in reaping the fruits of his advantage.
Hence the constant creation of schemes for night attacks by those who
have not to lead them, and have no responsibility, whilst these attacks
seldom take place in reality.

These ideal schemes are all based on the hypothesis that the assailant
knows the arrangements of the defender because they have been made
and announced beforehand, and could not escape notice in his
reconnaissances, and inquiries; that on the other hand, the measures of
the assailant, being only taken at the moment of execution, cannot be
known to the enemy. But the last of these is not always quite the case,
and still less is the first. If we are not so near the enemy as to have
him completely under our eye, as the Austrians had Frederick the Great
before the battle of Hochkirch (1758), then all that we know of his
position must always be imperfect, as it is obtained by reconnaissances,
patrols, information from prisoners, and spies, sources on which no firm
reliance can be placed because intelligence thus obtained is always
more or less of an old date, and the position of the enemy may have
been altered in the meantime. Moreover, with the tactics and mode of
encampment of former times it was much easier than it is now to examine
the position of the enemy. A line of tents is much easier to distinguish
than a line of huts or a bivouac; and an encampment on a line of front,
fully and regularly drawn out, also easier than one of Divisions formed
in columns, the mode often used at present. We may have the ground on
which a Division bivouacs in that manner completely under our eye, and
yet not be able to arrive at any accurate idea.

But the position again is not all that we want to know the measures
which the defender may take in the course of the combat are just as
important, and do not by any m