5 o ZN STARRY REALMS. 



change its aspect towards us, it has been sometimes 

 argued that it cannot be turning round. The difficulty, 

 such as it is, resolves itself more into a matter of words 

 than of actual fact. Let me try to explain it as follows. 



Suppose a horse is galloping round a circular course, 

 then a spectator at the centre of the course will have one 

 side of the horse turned towards him, while the other side 

 of the horse is invariably turned away. To this extent 

 therefore the movement of the horse round the course is 

 analogous to the movement of the moon around the earth. 

 When the horse has completed his circuit it will be found 

 that he has not only performed a revolution around the 

 course, but he has also rotated once in the same period. 

 It is obvious that the horse must at one moment or another 

 of his journey have faced every part of the horizon suc- 

 cessively. No matter how the course be placed there must 

 have been one point so situated that while passing it the 

 horse was going directly towards the north, while when he 

 has reached the opposite point of the circle he is galloping 

 towards the south. At one intermediate point he is for the 

 moment bound towards the east, but at the point diametri- 

 cally opposite his flight is towards the west. It is clearly 

 impossible that the horse can have faced north, south, east, 

 and west, without performing a movement of rotation. 

 Unless he rotated he would necessarily be always journey- 

 ing towards the same point of the compass. Suppose that, 

 at the moment when he was facing due north, his right 

 side being turned towards the spectator in the middle, we 

 bring him to the opposite side of the course, and again 

 place him facing due north. It is clear that the left side 

 of the animal will now be towards the spectator. In such 



