ly so much use to a cavalry man,
whose slashing cut through shield or helmet renders _weight_ an absolute
necessity. The light blade might be brought to bear with all the speed
and force of the strongest man, but would be of no avail in those cases
where hard, dense, and heavy substances have to be cut through.

A fly may dash against a pane of plate-glass with the utmost speed and
yet fail to break the glass; but a cricket-ball thrown with a tenth part
of the velocity will smash the window to pieces. This is only an
analagous case, which indicates very fully the existence of the two
factors in the _vis-viva_ necessary to produce a certain result.

If you get your blade too light it will not be serviceable for
heavy-cutting work, whatever the speed of the cut; and if you get the
blade too heavy, it will be impossible to use it effectively on account
of its weight.

Everything depends upon what a sword is expected to do; and in selecting
a blade this cannot be too carefully borne in mind.

The Easterns have not, and indeed never had, any idea of using the
point; but they are far and away our superiors at edge work, and their
curved scimitars are admirably adapted for effective cutting, because
the edge, meeting the object aimed at on the slant, has great cutting or
slicing power.

This brings us to the most important matter in connection with cutting
weapons--the "draw."

If you take a razor in one hand and _hit_ the palm of the other hand a
smart _blow_ with the edge, no harm will be done; but if you vary this
hit, by making it lighter and putting the slightest possible _draw_ into
it, a cut will be the result, and blood will flow freely. That is to
say, anything like _drawing_ the edge along the skin will produce a cut.

Turn to the case of the scimitar. It will be seen that the curved form
of the blade _from hilt to point_ renders it impossible for a sweeping
cut, given with the arm extended its full length and with the shoulder
as centre of the circle, which the hand traces out in making the cut, to
be other than a "draw," because the edge _must_ meet the object to be
severed on the slant.

[Illustration: Fig. 18.--The scimitar.]

Excellent examples of this kind of cutting are to be found in the
circular saw and the chaff-cutting machine.

But this is not the case with a nearly straight-bladed broad-sword,
which requires what may be termed an artificial draw, either backward or
forward, in order that