It is now a recognised principle that there is, so to speak, a wobble of the earth's axis round the pole of the heavens, in consequence of the attraction of the sun on the nearer point of this equatorial ring being greater than on the part of the ring further removed from it. That precessional movement is not quite so simple as it is shown by the model, because what the sun does in this way is done to a very much larger extent by the moon, the moon being so very much nearer to us. In consequence, then, of this luni-solar precession we have a variation of the points of intersection of the planes of the earth's equator and of the ecliptic; in consequence of that we have a difference in the constellations in which the sun is at the time of the solstices and the equinoxes; and, still more important from our present point of view, we have another difference, viz., that the declinations, and therefore the amplitudes, and therefore the places of setting and rising of the stars, change from century to century. Now that we have thus become acquainted with the physical cause of that movement of the earth's axis which gives rise to what is called the precession of the equinoxes, we have next to enter with somewhat greater detail into some of the results of the movement. The change of direction of the axis in space has a cycle of something between 25,000 and 26,000 years. As it is a question of the change of the position of the celestial equator, or rather of the pole of the celestial equator, amongst the stars in relation to the pole of the heavens, of course the declinations of stars will be changed to a very considerable extent; indeed, we have seen that the declination of a star can vary by twice the amount of the obliquity, or say 47°, so that a star at one time may have zero declination--that is, it may lie on the equator--and at another it may have a declination 47° N. or S. Or, again, a star may be the pole star at one particular time, and at another it will be distant from the pole no less than 47°. Although we get this enormous change in one equatorial co-ordinate, there would from this cause alone be practically no change with regard to the corresponding ecliptic co-ordinate--that is to say, the position of the star with reference to the earth's movement round the sun. This movement takes place quite independently of the direction of the axis, so that while we get this tremendous swirl in declination, the latitudes of the stars or their distances from the ecliptic north or south will scarcely change at all. [Illustration: STAR-MAP. REPRESENTING THE PRECESSIONAL MOVEMENT OF THE CELESTIAL POLE FROM THE YEAR 4000 B.C. TO THE YEAR 2000 A.D. (_From Piazzi Smyth._) _Symbols adapted to represent the magnitudes or brightnesses of the stars_, _1^{st}._ ⊙, _2^{nd}._ ⬭, _3^{rd}._ ∆, _4^{th}._ ⊡.]