The effect of the attraction of the sun and moon upon the equatorial protuberance of the earth is to draw it round from west to east by a very slow motion, and make the ecliptic cross the equator each year about one minute of arc to the east of where it crossed it the year before. So, then, the sidereal year, or interval between the times at which the sun is in a certain position amongst the stars, is longer than the solar year, or interval between the times at which the sun crosses the equator at the vernal equinox. Now the sun's position in the zodiac refers to the former, his appearance at the equinox to the latter kind of year. Each solar year then--and these are the years we usually reckon by--the equinox is at a point fifty seconds of arc to the east on the zodiac, an effect which is known by the name of the precession of the equinoxes. Now it is plain that if it keeps moving continuously to the east it will at last come round to the same point again, and the whole period of its revolution can easily be calculated from the distance it moves each year. The result of such a calculation shows that the whole revolution is completed in 25,870 years, after which time all will be again as it is now in this respect. [Illustration: FIG. 11.] If we draw a figure of the zodiac, as below, and know that at this time the vernal equinox takes place when the sun is in the Fishes, then, the constellation of the Ram being to the west of this, the date at which the equinox was there must be before our present date, while at some time in the future it will be in Aquarius. Now if in any old description we find that the equinox is referred to as being in the Ram or in the Bull, it tells us at once how long ago such a description was a true one, and, therefore, when it was written. This is the way in which the Zodiac carries with it an intimation of its date. Thus in the example lately referred to of the Persians and their four stars, it must have been about 5,000 years ago, according to the above calculation, that these were in the positions assigned, which is therefore the date of this part of Persian astronomy, if we have rightly conjectured the stars referred to.