DEAR ST. NICHOLAS: Seeing Mr. Stevens’s letter on the “curve” in your April number, I am tempted to reply to it. Unfortunately, the data on which his theory is founded are incorrect. A ball twisting to the left will curve, not to the right, but to the left. May I offer another explanation. The facts are that: 1, the axis of rotation is perpendicular to (or rather the plane of rotation or twist is parallel to) the plane in which the curve lies; and, 2, that the curve is in the _same_ direction as the twist. When a ball is thrown with sufficient velocity, the air is (as Mr. Stevens tells us) “compacted in front of the ball.” When there is little or no twist, the resistance of the air is equal on both sides of the ball, and there is no curve. Now, if the wind be blowing across the path of the ball, the resistance is unequal, and the ball curves away from the wind; so that in practice allowance is made for this curve when throwing a ball in a high wind. When there is no wind, but the ball is thrown with sufficient velocity to create considerable resistance from the air, and at the same time is twisted so as to rotate on its axis, then the resistance offered at a, c (Fig. 1) is greater than that offered at a, b; for a, c is advancing with the velocity of the throw _plus_ the velocity of the twist, while a, b is advancing with the velocity of the throw _minus_ the velocity of the twist. Consequently a greater resisting force being exerted at a, c than at a, b, the ball yields and is forced into the curve B, A′, just as a cross wind would deflect it. The result is a curve—because the forces are constant while the ball is in motion. [Illustration: FIG. 1.] Again, conceive of the ball B (Fig. 2) as at rest and the wind acting on a, b in the direction c, a. The angle of incidence being equal to the angle of deflection, the result will be to force the ball in the direction c′, a—the same result as is produced by the wind acting on a sail. [Illustration: FIG. 2.] On this theory, the relation of the velocity of the twist to the velocity of the throw will determine the nature and degree of the curve, and the point of departure from the straight line. If the ball be thrown too slowly, there is not sufficient resistance to affect its course materially. If it is thrown too swiftly, the velocity of the throw will overcome the tendency of the twist, and there will be no perceptible curve. For this reason the ball, when first thrown, will proceed in an apparently straight line until its initial velocity is so far diminished as to nearly equal the velocity of rotation, when it will begin to curve.