times gone by these collisions may have been more serious and more frequent. But I do not want to trouble you about the collisions so much as the general movements to and fro. That very phrase "to and fro" reminds us of one most important feature of the movements, namely that they are sometimes "to" the centre and sometimes "fro" or from it. At any given moment, therefore, we can divide the stars into two lots, one lot going to the centre and the other lot coming from it. Let us think of the first lot alone. If we watch their movements and represent them by arrows, then all these arrows will point to the centre: they will converge towards a point just like the arrows in the picture of Mr. Boss's cluster (Fig. 91). Now, in the case of this cluster, we said that the stars were not really moving towards a point, but in parallel lines: we knew that because we had the second method of watching the velocities to help the first. But if we had not had that second method to help us, we could not really have told whether the movement was in parallel lines or to a point: the perspective appearance would be just the same. And we might easily make a mistake in concluding, when we see arrows representing star movements directed to a point, that the movements are really parallel. They may really be to a point. Lots of boys christened Henry are called Harry "for short," but you might easily make a mistake if you concluded that every Harry had been really christened Henry, for some of them are christened Harry direct. And I think that astronomers are liable to make just this very mistake about the
