Before we begin to consider the question of stars at all, we must be able to describe them--to speak of them in a way that shall define exactly which star is meant. We can in these days define a star according to its constellation, or its equatorial or ecliptic co-ordinates, but all these means of reference were unknown to the earliest observers. Still we may assume that the Egyptians could define some of the stars in some fashion; and it is evident that we here approach a matter of the very highest importance for our subject, to which I shall have to return in a subsequent chapter. So far as we have been dealing with the sun and the observations of the sun at rising and setting, we have taken for granted that the amplitude of the sun at the solstices does not change; the amplitude of 26° at Thebes for the solstices is practically, though as we have seen not absolutely, invariable for a thousand years; but one of the results of astronomical work is that the _stars_ are known to behave quite differently. In consequence of what is called _precession_ the stars change their place with regard to the pole of the equator; and further, in consequence of this movement, the position of the sun among the stars at the solstices and equinoxes changes also. In reference to the sun's path we considered what are called the ecliptic and the equatorial co-ordinates. The ecliptic defines the plane in which the earth moves round the sun, and 90° from that plane we have the pole of the heavens; celestial latitude we found reckoned from the plane of the ecliptic north and south up to the pole of the heavens, and celestial longitude was reckoned along the plane of the ecliptic from the first point of Aries. We had also declination reckoned from the equator of the earth prolonged to the stars, and right ascension reckoned along the equator from the first point of Aries. The pole of the heavens or of the ecliptic, then, we must regard as practically, but not absolutely, fixed; but the pole of the earth's equator is not fixed, it slowly moves round the pole of the heavens. _In consequence of that movement there is a change of declination in a star's place._ Going back to the diagram (p. 49), we find that the amplitude of a body rising or setting at Thebes or anywhere else depends upon its declination; so that if from any cause the declination of a star changes, its amplitude must change. That is the first point where we meet with difficulty, because if the amplitude changes it is the same as saying that the place of star-rising or star-setting changes; that is, a star which rises in the east in a certain amplitude this year will change its amplitude at some future time.