20 



turn round on its axis ; or, what is the same thing, unless each 

 of the points describe an orbit of rotation round M's axis, which 

 remains invariably at the same distance from E's centre? Let 

 this question be answered if it can consistently with the theory of 

 lunar rotation. 



The real rotation of the revolving body, in the case of so-call- 

 ed parallelism, may be shown by this experiment. Make the disk 

 M of pasteboard or stiff paper, mark it as in the diagram, and fix 

 it by a pin through its centre to the outer end of a material vincu- 

 lum or radius- vector, which shall revolve by its other end round 

 E's centre. Above the disk M, also on the pin, place a smaller 

 disk, just large enough to show distinctly the same marked points 

 of circumference ; and then place the radius with its two disks 

 on the line NS. In revolving, both of the disks on the pin 

 would keep their places, with their points 1, 1, on the radius all 

 round ; but while the smaller disk, which is virtually nothing 

 more than M's axis itself expanded, for the sake of distinct per- 

 ception of the changes it might undergo, keeps steadily with 1 to 

 the radius, make M turn round on the pin continually, precisely 

 in proportion to the velocity of the orbital revolution of its axis, 

 degree for degree, and minute for minute, and it will preserve 

 its parallelism throughout the revolution ; but that it has not the 

 less rotated is unquestionable ; for the experimenter's own fin- 

 gers have pushed it round, carrying its successive points of cir- 

 cumference continually past the radius-vector, and away from 

 E's centre, to the extent of 90° or a quarter rotation, in each 

 quadrant of the orbit ; while, on the other hand, the small 

 disk certainly has not turned round upon its axis, the pin ; the 

 pin has not turned round in its place on the radius ; and the ra- 

 dius has not turned hack at E's centre, or changed in the slight- 

 est degree its proper perpendicular relations as the shortest 

 straight line between the centres of E and M ; or, in other 

 words, it has not rotated or turned round about on M's axis, 

 any more than E's centre itself, with which such a rotation or a 

 change of position in respect of any point fixed or moveable on 

 the circumference, is impossible. What remains then but to 

 admit the fact, that the so-called parallelism is produced by M's 

 retro-rotation from east to west, while its axis revolves from west 

 to east, with all the parts and molecules of the mass moving along 

 with the same degree of velocity under the influence of the centri- 

 fugal power acting upon them through the axis ? On the other 

 hand, were both the disks alike to keep the same points 1, 1, on 

 the radius, it follows certainly from the same premises that neither 

 of them rotates on the pin axis, but that the mass of each simply 

 revolves from west to east with diff^erent degrees of velocity pro- 

 portional to the diametric or radial distances of its particles from 



