All the temples in Egypt, however, are not oriented in such a way that the sunlight can enter them at this or any other time of the year. They are not therefore solar temples, and they cannot have had this use. The critical amplitude for a temple built at Thebes so that sunlight can enter it at sunrise or sunset is about 26° north and south of east and west, so that any temples facing more northerly or southerly are precluded from having the sunlight enter them at any time in the year. It is imperative to be perfectly definite and clear on the question of the amplitudes above 26° at Thebes. I repeat, therefore, that any amplitude within 26° means that up to that point the sun at sunrise or sunset could be observed some day or days of the year--once only in the year if the amplitude is exactly at the maximum, twice if the maximum is not reached. But in the case of these temples with greater amplitudes than 26°, it is quite clear that they can have had nothing to do with the sun. This being so, we have the problem presented to us whether or not temples were built so that starlight might fall along their axes in exactly the same way that the sunlight could fall along the axes of the solar temples when the sun was rising in the morning or setting in the evening. It is abundantly clear that temples with a greater amplitude than 26° were oriented to stars if they were oriented at all by astronomical considerations. How can this question be studied? What means of investigation are at our disposal? Suppose that the movements of the stars are absolutely regular; that there is no change from year to year, from century to century, from æon to æon; then, of course, the question as to whether or not these temples were pointed to a star, at rising or setting, would be easily and sufficiently settled by going to see; because if the stars did not change their apparent places in the heavens--accurately speaking, their declinations--and, therefore, the amplitudes at which they appear to rise and set, then, of course, a temple consecrated to Sirius ten thousand years ago would view the rising or setting of Sirius now as it did then. But, as a matter of fact, astronomy tells us, as we have seen, that the apparent positions of the stars are liable to change. The change is much greater in the case of the stars than it is in the case of the sun, referred to in Chapters VI. and XI.; but still we have seen that the latter is one which has to be reckoned with the moment it becomes a question of inquiry into any time far removed from the present. Hence, although in the case of the sun, there is, of course, no processional movement, and although a temple once oriented to the sun would remain so for a long time; still, after some thousands of years, the change in the obliquity of the ecliptic would produce a small change in the amplitude at which a solstice is observed.