172
under ordinary conditions being the spilling of three-fourths of the water, or thereabouts.
But a very effective experiment for those who feeldoubts respecting the moon’s rotation may be con-ducted as follows Let AB (fig. 57, Plate XIII.) bea flat wooden bar of any convenient dimensions (accord-ing to the circumstances under which the experimentis to be conducted). Let fig. 58 present a side viewof the same bar, which, it will be observed, is arrangedto run on casters at A and B, and to turn on a pivotat C. At A let a small circle and arrow be markedon the bar; at C and B let small basins of water beplaced, in which let small wooden rods float,—or pre-ferably, let the rods float in half-filled saucers, them-selves floating in the basins. If now the experimenterwait until the water is Still, the floating rods beingcentral and parallel to the arrow at A, and if he thengently turn the wooden bar round on its pivot at C,the casters rolling on a smooth table or floor, he willsee that the rods floating at B and C both retain adirection almost wholly unchanged throughout themotion; and thus while continuing parallel to eachother and also to any line on the table or floor to whichthey were parallel in the first instance, they no longercontinue parallel to the arrow at A, whose directionchanges throughout the motion. The slight changeof position they undergo is obviously referable tofriction between the vrater and the basins and saucers.Of course the basin C is not essential in this experi-ment, nor the fixed arrow at A. If the basin B were