GREEK (I) Greek calendar, even in its most de veloped forms, differed from all modern European systems in being soli-lunar; i.e., in theory, every year began when the sun was in a certain position (solstice or equinox) and every month began with the new moon. But, as the solar year (solstice to solstice or equinox to equinox) is not even approximately divisible by the lunar month, the result was that some system of adjust ment was necessary. This was found, at an uncertain but early date, in the oktaeteris or eight-year group. Eight solar years are 8 X 365.25 days, roughly, i.e., 2,922 days. Eight lunar years, i.e., eight groups of twelve lunar months each, are 8 x 12 X 29.5o days approximately, i.e., 2,832 days. The difference, 90 days, is about three lunar months. Hence, by inserting, at different times within the oktaeteris, three extra months, approximate agreement with natural phenomena was obtained. But it was at best only approxi mate, for the above lengths of a solar year and a lunar month are not exact (see ASTRONOMY), and to bring the calendar into accu racy would have meant abandoning the principle of beginning each month at new moon ; but this principle was sacred, and could not be abandoned, at least in theory. An important attempt at reform was made by the astronomer Meton of Athens. By his system, the, years, commencing from the end of June, 432 B.C. of our reckoning, were grouped into "great years" of 19 each, with seven intercalary months. This gave, on an average, a month of 29 days, 12 hr., 45 min., 57 sec.,—less than 2 minutes too long. Callippus of Cyzicus later combined four Metonic cycles into one, and finally Hipparchus of Nicaea combined four Callippian periods into one cycle of 304 years, by which a very high degree of accuracy was obtainable. These theoretical constructions, however, affected the official years little, if at all, and intercalation went on in a very lax and unscientific fashion, resulting in civic calendars which were often some months out.
The month was not divided into weeks ; the planetary week is astrological, connected with the dogma that every period of time, including the day, had a planet for its regent. It is not much older than the Christian era, and was never official in classical times (see F. H. Colson, The Week, 1926). There was, however, a di vision into thirds. The first ten days of the month were called µiv icTa 4evos or commencing month; the last third, /Ay 06ivcav, or waning month ; for the middle third, no regular term was in use. Days of a higher number than 20 were either called the first, second, etc., after the eikades (twenties), or numbered backwards; thus at Athens, ivos ErcTf cOLvovros meant the sixth day, counting inclusively, from the end of the month Boedromion. The day was officially computed from sunset to sunset; practically it began at dawn.
Parallel with all this reckoning there ran the farmers' and sailors' calendars, based on observation of the heliacal rising and setting of the constellations (such-and-such work was to be done when the Pleiads rose, navigation was not safe after the setting of Arcturus, etc.), or on such phenomena as the appearance of vari ous flowers, migratory birds, and the like. Many private persons had calendars (parapegmata), on which, by moving a peg from one hole to another, it was possible to see the relative positions of the civil and natural years.
It is clear that the months are mostly named after gods (as Poseideon, cp. the Defian Artemision, the Delphic Heraios, and the Boeotian Hermaios) or after festivals (as Anthesterion, cf. Pamboiotios in Boeotia, Karneios in Laconia). This accords with the probable history of the calendar. It began, not as a scheme of continuous reckoning, but as a series of notable days or groups of days (Eoprai, lepoµriviac), corresponding in some cases at least to seasonal events, such as harvest. The days between these, unless they were days of new moon, or full moon, were not holy, and therefore not reckoned. The Boeotian year began with the winter solstice, the Laconian with the autumn equinox. In some of the later calendars, as those of Achaia and Phocis, the months were not named, but only numbered.
In Hellenistic times, more and more use was made of the official Egyptian year, which was in use under the Ptolemies. It consisted of 365 days, without intercalation, with the result that every four years it lost a day, and in 1,461 years it came right again. The months were called Thoth, Phaophi, Athyr, Choiak, Tybi, Mechir, Phamenoth, Pharmouthi, Pachon, Pauni, Epiphi and Mesori.
In 26 B.C., when Augustus reformed the Egyptian calendar to bring it into line with the Julian, Thoth I was equivalent to Aug. 29. This reformed Egyptian calendar had a long struggle to come into actual use, however. Also important for late times is the native Macedonian calendar, originally twelve lunar months called Dios, Apellaios, Audynaios, Peritios, Dystros, Xanthikos, Artemisios, Daisios, Panemos, Lobos, Gorpiaios, Hyerberetaios. About the beginning of our era this calendar, which was in common use in Asia Minor and Syria, was reformed on Julian lines.