I wish to show next that by studying the orientation of temples erected to watch the stars and sunrise and sunset at times other than the solstices or equinoxes, an immense amount of information may be gained if we endeavour to find the way in which the problem must have been attacked before the year was thoroughly established, and when it was still a question of grass- or corn-kings or gods who had to be propitiated; and we may even be enabled to understand why the particular divisions of the year were chosen. In a solstitial temple the sun makes its appearance only once a year, when it reaches its greatest north or south declination; but in the temples dealing with lower declinations the sun appears twice, once on its journey from the summer to the winter solstice, and again on its return. The first difficulty of the inquiry in the direction I have indicated arises from the fact that the products of different countries vary, and that identical farming operations have to be carried on at different times in these countries. We must, then, begin with some one country, and as the record is fullest for Greece I will begin with it. The first thing we find is that the chief points in the farmer’s year in Greece are about as far from the fixed points in the astronomical year as they well can be. In the Greek information so admirably collated by M. Ruelle in the article on the calendar in Daremberg and Saglio’s monumental “Dictionnaire des Antiquités Grecques et Romaines,” the earlier Gregorian dates on which the seasons were reckoned to commence in ancient Greece were as follows:-- Summer May 6. Autumn (φθινοπωρον) August 11. Winter November 10. Spring February 7. I may also add from the same source that in the calendars of the Latins the dates become:-- Summer May 9. Autumn August 8. Winter November 9. Spring February 7. Now we see at once that these dates are, roughly, half-way between the solstices and equinoxes. This, then, at once brings us back to the orientation problem, which was to fix by means of a temple in the ordinary way dates nearer to these turning-points in the local farmer’s years than those fixed by the solstitial and equinoctial temples. It must be borne in mind that it is not merely a question of stately piles such as Karnak and the Parthenon in populous centres, but of the humblest dolmen or stone circle, in scattered agricultural communities, which was as certainly used for orientation purposes, that is, for recording the lapse of time at night or return of some season important to the tiller of the soil. The advent of the season thus determined could be announced to outlying districts by fire signals at night. I have already pointed out that any temple, dolmen or cromlech oriented to a sunrise or sunset at any dates between the solstices will receive the sunlight twice a year.