S/iips, 4'c. against Caimon Balls. 409 



fig. 3 at G, the respective balls would, after sliding or rollino- 

 up the boards T8 fig. 1, and IC fig. 3, strike each of the 

 curves in such a direction as to follow their shapes and re- 

 turn, without any reflection taking place ; and in iig. 3, those 

 balls which entered at G would return at X, and vice versa. 



Case 2. — If the force of the ball K, fig. 1, should only be 

 so far evaded by the inclination of the side, as to penetrate to 

 about half the depth of its own size or less (shown large at 

 xy, fig. 2) ; and if there were no elasticity in the substances ; 

 there would, it seems, then arise a great force to repel the ball 

 beyond what is immediately caused by the inclination of the 

 side, on account of the rotary motion the ball would have 

 acquired by its action against the inside of the indentation : 

 thus suppose BAHQ, fig. 2, be a section of the ball going 

 nearly in a parallel direction CB, and suppose IKLQ be the 

 indentation, in which place we will fancy the substance of 

 the side to be so hard as not to give way any more, the effect, 

 it seems, would then be, that the centre of the ball B would 

 begin to describe part ojf a circle BN, about the centre I (the 

 point where the indentation and the remainder of the side 

 rneet, and of the size of the ball itself). Then if the indenta- 

 tion should be deep, and the velocity great, the ball would be 

 forced completely out of it, and fly far above the top of the 

 ship, because die part of the circle BN, which the centre B 

 of the ball would begin to describe, would be nearly perpen- 

 dicular to the side SD ; and as there would be nothing to 

 change the direction of the ball after it has once acquired this 

 new motion, it would fly off" in the direction of the most per- 

 pendicular part of the circle BN, and continue in this direc- 

 tion, though not of the continued circle BN, but in a straight 

 line BR: itj however, the indentation should be small, the line 

 BR would be more nearly parallel to the side SD, in which 

 case the motion of the ball woidd not be caused to differ so 

 much from the direction of the side, but that it might strike 

 the flat board TS, fig. 1, in a direction KV, which would 

 prevent it from flying away and direct it to the side again, so 

 that it followed the return part and flew back again, after 

 having been reflected backwards and forwards, not bv means 

 of any elasticity, but by the reaction of the inside surface of 

 the indentation against the ball (as before described); and as 

 there would be a loss of force at every blow, each indentation 

 would be less than the preceding one, and each angle of re- 

 flection would be more obtuse, as is shown in fig. 3, till the 

 ball arrived at the return part WA, so as tt) follow the shape 

 of it, ceasing sensibly to rebound when the indentation ceased 

 sensibly to take place : but as the indentation and point I 

 Vol.'62. No. 308. Dec. 1823. 3 F would 



