ut the periphery of the ball, a current similar to that caused by
the revolving flywheel of a steam engine.

"If, now, at a point on the face of the ball we let the arrow, R,
represent the direction and intensity of this rotary current of air, and
if at the same point we let the arrow, A, represent the direction and
intensity of the retarding action of the air, then we will find by
constructing a parallelogram of forces that the resultant or combined
effect of these two currents acts in the direction indicated by the
dotted arrow, T. In other words, we have a sort of compression, or force
of air, acting on the face of the ball in the direction indicated by the
arrow, T. This force, as we can readily see, tends, when combined with
the original impetus given to the ball, to deflect or cause time ball to
curve in the direction of the dotted line, B P, instead of maintaining
its right line direction, B K. If the ball rotate about its vert axis in
the opposite direction, the curve, B N, will be the result."

To the above demonstrations it is only necessary to add an explanation
of one other feature. The question has arisen why it is that the ball
apparently goes a part of its course in a straight line and then turns
off abruptly. One might suppose at first thought that the greater speed
at the beginning would create the greater resistance and consequently
cause the greatest deflection. This, however, is not true. The
difference between the resistance upon opposite points of the ball in
the circumference of its rotation always remains the same, no matter how
great the force of propulsion, and therefore the increased force of the
latter at the beginning has no effect on the curve. But while the force
of the twist itself is not affected by the rate of the forward movement,
its effect upon the ball is greatly nullified. The force of propulsion
being so great at first, drives the ball through the air and prevents it
from being influenced by the unequal resistance. It is only when the two
forces approach one another in strength that the latter begins to have a
perceptible effect. As soon, however, as it does, and the course of the
ball begins to change, the direction of the dotted arrow, T, begins to
change likewise. It follows the course of the ball around, and the more
it curves the more this resultant force tends to make it curve, and this
continues until the ball has lost either its twist or its forward
motion.

Having estab