The discovery of the curve itself was purely an accident. During the years from 1866 to 1869 the theory was held by many pitchers that the more twist imparted to a pitched ball, the more difficult it would be to hit it straight out. It was thought that even if it were struck fairly, this twist would throw it off at an angle to the swing of the bat. One writer on the game declared strongly against this practice of the pitchers on the ground that, though this twist did do all that was claimed for it, it at the same time caused the ball, when hit, to bound badly, and thus interfered with good fielding. Of course, both of these theories become absurd in the light of the present, but it was doubtless the belief in the former that led to the introduction of the curve. In 1869 Arthur Cummings, pitching for the Star Club, noticed that by giving a certain twist to the ball it was made to describe a rising, outward curve, and his remarkable success with the new delivery soon led to its imitation by other pitchers, and finally to the general introduction of curve pitching. The philosophy of the curve is, in itself, quite simple. A ball is thrown through the air and, at the same time, given a rotary motion upon its own axis, so that the resistance of the air, to its forward motion, is greater upon one point than upon another, and the result is a movement of the ball away from the retarded side. Suppose the ball in the accompanying cut to be moving in the direction of the arrow, B C, at the rate of 100 feet per second. Suppose, also, that it is rotating about its vertical axis, E, in the direction of I to H, so that any point on its circumference, I H D, is moving at the same rate of 100 feet per second. The point I is, therefore, moving forward at the same rate as the ball's centre of gravity, that is, 100 feet per second, plus the rate of its own revolution, which is 100 feet more, or 200 feet per second; but the point D, though moving forward with the ball at the rate of 100 feet per second, is moving backward the rate of rotation, which is 100 feet per second, so that the forward motion of the point D is practically zero. At the point I, therefore, the resistance is to a point moving 200 feet per second, while at D it is zero, and the tendency of the ball being to avoid the greatest resistance, it is deflected in the direction of F. In the Scientific American of August 28th, 1886, a correspondent gave a very explicit demonstration of the theory of the curve, and, as it has the virtue of being more scientific than the one given above, I append it in full.