[21] EGGS OF THE PLAICE, FLOUNDER, AND COD. 447 



strongly favor the idea that the eggs scatter evenly over a considera- 

 ble area. I have not found any data relative to the mechanical distri- 

 bution of such objects by shaking; nevertheless it remains an undoubted 

 empiric fact that the process of shaking and stirring causes an even dis- 

 tribution of hard bodies, such as grains of corn or seed, and of hard 

 bodies in fluids {e. g., in emulsions). 



The reason why irregnlar pushing motions, made in every direction 

 and of a certain duration, applied to a large number of bodies, produce 

 a tolerably even distribution of such bodies, must be found in the fol- 

 lowing circumstances: 



Bodies like eggs receive such pushes either in a direction perpendicu- 

 lar to their radius or in any other direction. In the latter case the push- 

 ing motion becomes divided, one motion turning the eggs, the other sim- 

 ply pushing them; but as in the water the entire surface of the egg is 

 almost invariably struck simultaneously in one and the same direction, 

 rotation sets in but very seldom. 



If thrusts of equal strength struck the globe at all its radii at one and 

 the same time, for whose endless number of radii we will assume a 

 certain large fixed number, the globe would not move; and if all these 

 thrusts were made in quick succession, the globe would, after a certain 

 number of thrusts, assume its original position. The intervals in which 

 these thrusts are made may vary ; it is also possible that each radius 

 must receive a three or four fold number of thrusts before the cycle of 

 motions is completed ; but the globe of the egg would invariably have 

 to return to the same spot. If the motions, however, become entirely 

 irregular it becomes highly improbable that a cycle of motions will be 

 completed after a small number of such thrusts ; only after an illimit- 

 able number of thrusts, therefore after an illimitable period has elapsed, 

 there will be a probability that such a series of cycles has been com- 

 pleted, and that, thereby, the body has been brought back to its old 

 place or within its neighborhood. 



What applies to one egg, applies to all. There remains, therefore, 

 only the possibility that all eggs move from their starting-point in one 

 and the same direction, and that, consequently, they do not scatter. 

 This becomes all the more improbable the larger the number of the eggs ; 

 for this would presuppose that all the thrusts which strike the eggs are 

 made absolutely parallel with each other. This may occasionally be 

 the case in currents ; but as soon as the thrusts are made irregularly in 

 different directions, the eggs will scatter. Every radius of each indi- 

 vidual egg runs the ^ame chance of being struck, and as the thrusts 

 are made in different directions, the individual eggs will receive them 

 in different ways. The more the eggs are scattered, all the more — and 

 in proportion to the cube of the distances — will the probability disap- 

 pear that they will meet again at any time. 



It will not be necessary to discuss the question in what manner an 

 even distribution through space is finally brought about, because such 



