that a ball
cannot be curved, even when they have ocular demonstration
of the fact. But that has nothing to do with what I have to
say. I have studied the diagram of my anonymous friend, and
am convinced that he is exactly wrong. With the following
diagrams I shall show which way a ball curves with a given
rotation, and give my theory of the curve:

[Illustration]

Suppose, as in the letter published, the ball moves one
hundred feet per second, and revolves so that the equator
moves around at the same rate. Then, in the first diagram,
the friction at B is greatest, and at D is 0. But instead of
curving as my anonymous friend demonstrates, it will curve
in exactly the _opposite_ direction; namely, in the same
direction in which it rotates.

I have appended diagram 2, simply to show the curve where
the friction is 0 at B and greatest at D. Then it will curve
as indicated.

I have a short theory, namely: In the first diagram, the
more rapid movement of B compresses the air on that side,
while at D it is in its normal state. Hence the pressure at
B more than counterbalances that at D, and, as it were,
shoves the ball in the direction of the side D, thus
producing the curve. In the 2d diagram, the letters B and D
interchange in the theory. I would like to hear more about
this subject.

Very respectfully yours,

F. C. J.

* * * * *

BIRMINGHAM, MICH.

DEAR ST. NICHOLAS: I have read with great interest the
articles in the October, December, and February numbers,
about curve-pitching. I have had quite a good deal of
experience in the "one,-two,-three,-and-out" line myself,
and have also, for the last two or three years, been able to
make others have the same experience, by putting them out,
in the same way. Therefore, I venture a reply to the
explanation in the February number, backing my statement by
the experience of many eminent curve-pitchers, and also by
the story in the October number of "How Science Won the
Game."

[Illustration]

The above diagram is the same as your correspondent uses,
and he asserts that the point B is moving faster than D;
consequently, there is more friction at B, whence B is
retarded more than D, and so the