altogether evanescent. Now it is by the re-action of the air against
_one_ side of the impinging plane, that the progressive motion is
determined in the opposite direction, which re-action is proportioned
to the _rate_ of impact, the angle remaining the same. If then
we suppose a re-action corresponding to the _greatest rate_ of
revolution, which is that due to the _outermost_ portion of the
impinging surface (that most removed from the axis of rotation) we
shall have a _progressive_ motion in the whole apparatus greater
than the rate of impact of the _innermost_ or more central
portions of the revolving plane; and accordingly the re-action will be
thereabouts transferred from the back to the front of the propulsive
apparatus, and tend to retard instead of advancing the progress of
the machine to which it is attached. This inconvenience is felt and
acknowledged by all those who have employed this principle to obtain a
progressive motion, and accordingly a provision has been made against
it in the _removal_ or _reduction_ of the central portion
of the revolving vanes, with a view to let the air escape or pass
through as the instrument advances; a provision which is certainly
effectual to that end, but at the cost of the _surface_, which is
the ultimate source of the required re-action. All this is avoided
in the use of the perfect screw. There, the rate of rotation and the
angle of impact mutually corresponding, may be said to play into each
other's hands; the spiral becoming more extended as the impact becomes
less forcible, that is as it approaches the centre, where both
altogether vanish or disappear; thus obviating the possibility of any
interruption to the course of the machine from the contrarious impact
of the air, however quick or however slow the motions, either of the
screw itself or of the machine which is propelled by its operation. In
attestation of this fact and as showing the immunity of the perfect
screw from the disparaging effects experienced by the other modes of