From the account given by Herodotus[42] of the ceremonials and mysteries connected with the temple of Tyre, it is suggested that the priests used starlight at night for some of their operations, very much in the same way as they might have used sunlight during the day. According to Herodotus, in the temple in question there were two pillars--the one of pure gold, and the other of an emerald stone of such size as to shine by night.[42] Now, there can be little doubt that in the darkened sanctuary of an Egyptian temple the light of α Lyræ, one of the brightest stars in the northern heavens, rising in the clear air of Egypt, would be quite strong enough to throw into an apparent glow such highly-reflecting surfaces as those to which Herodotus refers. Supposing such a ceremonial as this, the less the worshippers--who, reasoning from the analogy of the ceremonial termed the manifestation of Rā,[43] would stand facing the sanctuary, with their backs to the chief door of the temple--knew about the question of a bright star which might probably produce the mystery, the better. Again, the truer the orientation of the temple to the star, and the greater the darkness the priest was kept in, the sooner would he catch the star quivering in the light of either early or late dawn. In the first place, the diaphragms would indicate the true line that he had to watch; he would not have to _search_ for the star which he expected; and obviously the more he was kept in the dark the sooner could he see the star. Is there any additional line of evidence beyond the structural conditions of the temples that the Egyptians used these temples to observe the stars? Here a very interesting question comes in: a temple built at one period to observe a star could not go on for ever serving its purpose, for the reason that the declination of the star must change, as we have seen, by precession. Therefore a temple built with a particular amplitude to observe a particular star at one period would be useless later on. We have here possibly a means of testing whether or not any of these temples were used to observe the stars. In those very early days, 3000 or 4000 years B.C., we must assume that the people who observed the stars had not the slightest idea of these possible precessional changes; they imagined that they were just as safe in directing a temple to a star as they were in directing a temple to the sun. But with a star changing its declination in an average way, the _same_ temple could not be used to observe the _same_ star for more than 200 or 300 years; so that at the end of that time, if they still wished to observe that particular star, they must either change the axis of the old temple, or build a new one. I have mentioned an average time as the change of the star's declination is involved. Now this change of direction is one of the most striking things which have been observed for years past in Egyptian temples.