The ancient Egyptian calendar was a solar calendar with a 365-day year. The year consisted of three seasons of 120 days each, plus an intercalary month of 5 epagomenal days treated as outside of the year proper. Each season was divided into four months of 30 days. These twelve months were initially numbered within each season but came to also be known by the names of their principal festivals. Each month was divided into three 10-day periods known as decans or decades. It has been suggested that during the Nineteenth Dynasty and the Twentieth Dynasty the last two days of each decan were usually treated as a kind of weekend for the royal craftsmen, with royal artisans free from work.
Because this calendrical year was nearly a quarter of a day shorter than the solar year, the Egyptian calendar lost about one day every four years relative to the Gregorian calendar. It is therefore sometimes referred to as the wandering year (Latin: annus vagus), as its months rotated about one day through the solar year every 4 years. Ptolemy III's Canopus Decree attempted to correct this through the introduction of a sixth epagomenal day every four years but the proposal was resisted by the Egyptian priests and people and abandoned until the establishment of the Alexandrian or Coptic calendar by Augustus. The introduction of a leap day to the Egyptian calendar made it equivalent to the reformed Julian calendar, although by extension it continues to diverge from the Gregorian calendar at the turn of most centuries.
This civil calendar ran concurrently with an Egyptian lunar calendar which was used for some religious rituals and festivals. Some Egyptologists have described it as lunisolar, with an intercalary month supposedly added every two or three years to maintain its consistency with the solar year, but no evidence of such intercalation before the 4th century BC has yet been discovered.
Current knowledge of the earliest development of the Egyptian calendar remains speculative. A tablet from the reign of the First-DynastypharaohDjer (c. 3000BC) was once thought to indicate that the Egyptians had already established a link between the heliacal rising of Sirius (Ancient Egyptian: Spdt or Sopdet, "Triangle"; Greek: Σῶθις, Sôthis) and the beginning of their year, but more recent analysis has questioned whether the tablet's picture refers to Sirius at all. Similarly, based on the Palermo Stone, Scharff proposed that the Old Kingdom observed a 320-day year but his theory has not become widely accepted. Some evidence suggests the early civil calendar had 360 days, although it might merely reflect the unusual status of the five epagomenal days as days "added on" to the proper year.
The first lasted from roughly June to September, the second from roughly October to January, and the last from roughly February to May. As early as the reign of Djer (c. 3000BC, Dynasty I), yearly records were being kept of the flood's high-water mark.Neugebauer noted that a 365-day year can be established by averaging a few decades of accurate observations of the Nile flood without any need for astronomical observations, although the great irregularity of the flood from year to year[a] and the difficulty of maintaining a sufficiently accurate Nilometer and record in prehistoric Egypt has caused other scholars to doubt that it formed the basis for the Egyptian calendar.
The Egyptians appear to have used a purely lunar calendar prior to the establishment of the solar civil calendar in which each month began on the morning when the waning crescent moon could no longer be seen. Until the closing of Egypt's pagan temples under the Byzantines, the lunar calendar continued to be used as the liturgical year of various cults. The month may have been divided into four "weeks" of 7 or 8 days, reflecting each quarter of the lunar phases. Because the exact time of morning considered to begin the Egyptian day remains uncertain and there is no evidence that any method other than observation was used to determine the beginnings of the lunar months prior to the 4th century BC, there is no sure way to reconstruct exact dates in the lunar calendar from its known dates. The difference between beginning the day at the first light of dawn or at sunrise accounts for an 11–14 year shift in dated observations of the lunar cycle. It remains unknown how the Egyptians dealt with obscurement by clouds when they occurred and the best current algorithms have been shown to differ from actual observation of the waning crescent moon in about one-in-five cases.
A second lunar calendar is attested by a demotic astronomical papyrus dating to sometime after AD144 which outlines a lunisolar calendar operating in accordance with the Egyptian civil calendar according to a 25-year cycle. The calendar seems to show its month beginning with the first visibility of the waxing crescent moon, but Parker displayed an error in the cycle of about a day in 500 years, using it to show the cycle was developed to correspond with the new moon around 357BC. This date places it prior to the Ptolemaic period and within the native Egyptian Dynasty XXX. Egypt's 1st Persian occupation, however, seems likely to have been its inspiration. This lunisolar calendar's calculations apparently continued to be used without correction into the Roman period, even when they no longer precisely matched the observable lunar phases.
The days of the lunar month—known to the Egyptians as a "temple month"—were individually named and celebrated as stages in the life of the moon god, variously Thoth in the Middle Kingdom or Khonsu in the Ptolemaic era: "He... is conceived... on Psḏntyw; he is born on Ꜣbd; he grows old after Smdt".
The civil year comprised exactly 365 days,[q] divided into 12 months of 30 days each and an intercalary month of 5 days, were celebrated as the birthdays of the gods Osiris, Horus, Set, Isis, and Nephthys. The regular months were grouped into Egypt's three seasons, which gave them their original names, and divided into three 10-day periods known as decans or decades. In later sources, these were distinguished as "first", "middle", and "last". It has been suggested that during the Nineteenth Dynasty and the Twentieth Dynasty the last two days of each decan were usually treated as a kind of weekend for the royal craftsmen, with royal artisans free from work. Dates were typically expressed in a YMD format, with a pharaoh's regnal year followed by the month followed by the day of the month. For example, the New Year occurred on I Akhet 1.
Following Censorinus and Meyer, the standard understanding was that, four years from the calendar's inception, Sirius would have no longer reappeared on the Egyptian New Year but on the next day (I Akhet 2); four years later, it would have reappeared on the day after that; and so on through the entire calendar until its rise finally returned to I Akhet 1 1460 years after the calendar's inception,[r] an event known as "apocatastasis". Owing to the event's extreme regularity, Egyptian recordings of the calendrical date of the rise of Sirius have been used by Egyptologists to fix its calendar and other events dated to it, at least to the level of the four-Egyptian-year periods which share the same date for Sirius's return, known as "tetraëterides" or "quadrennia". For example, an account that Sothis rose on III Peret 1—the 181st day of the year—should show that somewhere 720, 721, 722, or 723 years have passed since the last apocatastasis. Following such a scheme, the record of Sirius rising on II Shemu 1 in 239BC implies apocatastases on 1319 and 2779BC ±3 years.[s]Censorinus's placement of an apocatastasis on 21July AD139[t] permitted the calculation of its predecessors to 1322, 2782, and 4242BC. The last is sometimes described as "the first exactly dated year in history" but, since the calendar is attested before Dynasty XVIII and the last date is now known to far predate early Egyptian civilization, it is typically credited to Dynasty II around the middle date.[u]
The classic understanding of the Sothic cycle relies, however, on several potentially erroneous assumptions. Following Scaliger, Censorinus's date is usually emended to 20July[w] but ancient authorities give a variety of 'fixed' dates for the rise of Sirius.[x] His use of the year 139 seems questionable, as 136 seems to have been the start of the tetraëteris and the later date chosen to flatter the birthday of Censorinus's patron. Perfect observation of Sirius's actual behavior during the cycle—including its minor shift relative to the solar year—would produce a period of 1457 years; observational difficulties produce a further margin of error of about two decades. Although it is certain the Egyptian day began in the morning, another four years are shifted depending on whether the precise start occurred at the first light of dawn or at sunrise. It has been noted that there is no recognition in surviving records that Sirius's minor irregularities sometimes produce a triëteris or penteteris (three- or five-year periods of agreement with an Egyptian date) rather than the usual four-year periods and, given that the expected discrepancy is no more than 8 years in 1460, the cycle may have been applied schematically according to the civil years by Egyptians and the Julian year by the Greeks and Romans. The occurrence of the apocatastasis in the 2nd millennium BC so close to the great political and sun-based religious reforms of Amenhotep IV/Akhenaton also leaves open the possibility that the cycle's strict application was occasionally subject to political interference. The record and celebration of Sirius's rising would also vary by several days (equating to decades of the cycle) in eras when the official site of observation was moved from near Cairo.[y] The return of Sirius to the night sky varies by about a day per degree of latitude, causing it to be seen 8–10 days earlier at Aswan than at Alexandria, a difference which causes Krauss to propose dating much of Egyptian history decades later than the present consensus.
Egyptian scholars were involved with the establishment of Julius Caesar's reform of the Roman calendar, although the Roman priests initially misapplied its formula and—by counting inclusively—added leap days every three years instead of every four. The mistake was corrected by Augustus through omitting leap years for a number of cycles until AD4. As the personal ruler of Egypt, he also imposed a reform of its calendar in 26 or 25BC, possibly to correspond with the beginning of a new Callipic cycle, with the first leap day occurring on 6 Epag. in the year 22BC. This "Alexandrian calendar" corresponds almost exactly to the Julian, causing 1Thoth to remain at 29August except during the year before a Julian leap year, when it occurs on 30August instead. The calendars then resume their correspondence after 4Phamenoth/ 29February of the next year.
For much of Egyptian history, the months were not referred to by individual names, but were rather numbered within the three seasons. As early as the Middle Kingdom, however, each month had its own name. These finally evolved into the New Kingdom months, which in turn gave rise to the Hellenized names that were used for chronology by Ptolemy in his Almagest and by others. Copernicus constructed his tables for the motion of the planets based on the Egyptian year because of its mathematical regularity. A convention of modern Egyptologists is to number the months consecutively using Roman numerals.
A persistent problem of Egyptology has been that the festivals which give their names to the months occur in the next month. Gardiner proposed that an original calendar governed by the priests of Ra was supplanted by an improvement developed by the partisans of Thoth. Parker connected the discrepancy to his theories concerning the lunar calendar. Sethe, Weill, and Clagett proposed that the names expressed the idea that each month culminated in the festival beginning the next.
^In the 30 years prior to the completion of the Aswan Low Dam in 1902, the period between Egypt's "annual" floods varied from 335 to 415 days, with the first rise starting as early as 15 April and as late as 23 June.
are shrunk and fit under the two sides of the standard
^Other possibilities for the original basis of the calendar include comparison of a detailed record of lunar dates against the rising of Sirius over a 40 year span, discounted by Neugebauer as likely to produce a calendar more accurate than the actual one; his own theory (discussed above) that the timing of successive floods were averaged over a few decades; and the theory that the position of the solar rising was recorded over a number of years, permitting comparison of the timing of the solstices over the years. A predynasticpetroglyph discovered by the University of South Carolina's expedition at Nekhen in 1986 may preserve such a record, if it had been moved about 10° from its original position prior to discovery.
^1460 Julian years (exactly) or Gregorian years (roughly) in modern calculations, equivalent to 1461 Egyptian civil years, but apparently reckoned as 1460 civil years (1459 Julian years) by the ancient Egyptians themselves.
^Per O'Mara, actually ±16 years when including the other factors affecting the calculated Sothic year.
^Using Roman dating, he said of the relevant New Year that "when the emperor Antoninus Pius was consul of Rome for a second time with Bruttius Prasens this same day coincided with the 13th day before the calends of August" (Latin: cum... imperatore quinque hoc anno fuit Antonino Pio II Bruttio Praesente Romae consulibus idem dies fuerit ante diem XII kal. Aug.).
^Meyer himself accepted the earliest date, though before the Middle Chronology was shown to be more likely than the short or long chronologies of the Middle East. Parker argued for its introduction ahead of apocatastasis on the middle date based on his understanding of its development from a Sothic-based lunar calendar. He placed its introduction within the range c. 2937 – c. 2821BC, noting it was more likely in the Dynasty II part of the range.
^Specifically, the calculations are for 30°N with no adjustment for clouds and an averaged amount of aerosols for the region. In practice, clouds or other obscurement and observational error may have shifted any of these calculated values by a few days.
^This seems to be the case, for example, with astronomical records of the XVIII Dynasty and its successors, including the Ebers Papyrus, which seem to have been made at Thebes rather than Heliopolis.
Krauss, Rolf; et al., eds. (2006), Ancient Egyptian Chronology, Handbook of Oriental Studies, Sect. 1, Vol. 83, Leiden: Brill.
Luft, Ulrich (2006), "Absolute Chronology in Egypt in the First Quarter of the Second Millennium BC", Egypt and the Levant, Vol. XVI, Austrian Academy of Sciences Press, pp. 309--316.
Neugebauer, Otto Eduard (1939), "Die Bedeutungslosigkeit der 'Sothisperiode' für die Älteste Ägyptische Chronologie", Acta Orientalia, No. 16, pp. 169 ff. (in German)
O'Mara, Patrick F. (January 2003), "Censorinus, the Sothic Cycle, and Calendar Year One in Ancient Egypt: The Epistemological Problem", Journal of Near Eastern Studies, Vol. LXII, No. 1, Chicago: University of Chicago Press, pp. 17–26.