25 



primary orbit ; but, the disk being fixed to the rod, and not 

 turning on the pivot, the plummet's cord will gradually wind 

 itself round the disk's edge, and when the rod has completed its 

 revolution, the plummet will again be in one with the diametric 

 line NS, but its cord will be round the whole circumference 

 of the disk, a thing which could not possibly happen unless the 

 plummet's rotatory revolution carried it round that edge. 



At each quarter point of the circle on the wall, draw a strong 

 and distinctly marked cross, and mark the points of each that are 

 nearest the centre, «, a, «, a. It is obvious, from the nature of 

 the case, that they cannot turn rounti on their axes ; but if it so 

 happen that the rod AB moves successively past them all in 

 the course of the orbital revolution of its own axis, while its true 

 or mathematical radius-vector preserves invariably the same re- 

 lation to each of the four crosses, and to each of the four straight 

 lines that connect these crosses with the centre, is it not equally 

 obvious that that radius-vector moves forward along the orbit 

 continually, and altogether in the same direction with all its 

 points, without turning round on the axis of the revolving rod 

 any more than the fixed crosses themselves turn round on their 

 axes? They have not rotated on their axes; how comes it then 

 that the radius-vector, rotating as alleged round the moving axis, 

 is always in one with the same unmoved arm of each of them? 

 Again, across the rod AB, at its pivot or axis, place a cross 

 bar of say three inches length, and attach the ends of this cross 

 bar to the two ends of the rod by wires or tight strings, forming 

 altogether two triangles, on the opposite sides of a common base, 

 and having the vertices respectively at the opposite ends of the 

 rod, as in Fig. 3. It is obvious from mere inspection of the dia- 

 gram that if the two ends of the base change their relation to the 

 orbit, by the one going outwards and the other going inwards, that 

 change implies, and can only be the result of, a rotation on the 

 axis. But, if the two ends of the base rotate, the rod, from whose 

 points they are 90° distant on either side, must rotate also, whe- 

 ther that rotation be from west to east, or from east to west. In 

 both cases alike it must diverge from the radius-vector, which 

 can never forsake the centre, or change its relation of perpendi- 

 cularity to the orbit ; but, in diverging and rotating from east 

 to west, contrary to the direction of the orbital revolution of its 

 axis, the rod will necessarily preserve its much vaunted parallel- 

 ism, that parallelism which deceived Galileo, and has deluded all 

 his followers for two centuries. On the other hand, if the di- 

 verging of the cross-stick from the line of motion necessiirily im- 

 plies rotation, the remaining of that cross-stick on the line of 

 motion equally implies wow-rotation; and, when it does remain 

 so, the rod AB itself will remain always in one with the radius- 



