[HOME]

Egyptian Dates

Intro page      |      How to Read the Tables      |      The Egyptian Calendars       |       Sources      |      Analysis

Civil dates: (Excel) (HTML) (CSV)                                                   Lunar cycle: (Excel) (HTML) (CSV)

This page gives access to a set of conversion tables for determining the Julian equivalent of Egyptian civil and lunar dates in the Ptolemaic era. Two tables are provided: a table converting civil dates to Julian dates, and a table notionally converting lunar dates to civil dates according to the lunar cycle of pCarlsberg 9.

The Egyptian Lunar Calendar                                                                                     The Egyptian Civil Calendar

This account is largely taken from R. A. Parker, The Calendars of Ancient Egypt and L. Depuydt, Civil Calendar and Lunar Calendar in Ancient Egypt.

The Existence of the Lunar Calendar

Lunar dates are extremely rare in Egyptian records. In most cases where they do occur, only a lunar day number (or name) is given. However, the existence of an actual lunar calendar, in which lunations are organised into a named and repeated sequence of months, is absolutely certain, having being established by H. Brugsch (ZÄS 10 (1872) 1) from two double dates in the temple of Edfu that explicitly named both civil and lunar months:

• Ptolemy VIII Year 28, IV Shomu 18 (civil) = III Shomu 23 (lunar) (= 10 September 142)
  

• Ptolemy VIII Year 30, II Shomu 9 (civil) = Hb jnt 6 (= 2 July 140)

Since 1872 only one additional double date has been published, from pLouvre 7848 (R. A. Parker, MDAIK 15 (1957) 208):

• Amasis year 12, II Shomu 13 (civil) = I Shomu 15 (lunar) (=19 October 559)

This double date has been important for establishing precise Saite chronology, and also for resolving the management of intercalation in the lunar calendar, as discussed below.

The Lunar Month

The Egyptian lunar month began (at least notionally) on the day of lunar invisibility -- functionally, if not completely accurately astronomically, the new moon. Each month had 29 or 30 days. The days of the month had individual names as well as day numbers, as follows (from R. A. Parker, The Calendars of Ancient Egypt, 11):

It is not known which day was omitted in a 29 day lunar month.

The Lunar Year

The lunar year consisted of 12 or 13 lunar months. The first lunar month began with the (nominal) lunar invisibility occurring in the first civil month.

As with the civil calendar, the months had two names. One set gave each of the months individual names:

Wpt rnpt is documented as both the first and the last month of the year in different lists, and, as noted, some months have different names in different lists. As yet, there is no evidence that the lunar months were called by the monthly names of the corresponding civil months (Thoth, Phaophi etc), although the two lists clearly have some names in common, and although the civil names appear to have their origins in a lunar series of names related to the set listed above.

The other way to identify a lunar month was to give it a seasonal name (I Akhet etc). L. Depuydt, Civil Calendar and Lunar Calendar in Ancient Egypt 187ff., showed that the surviving lunar double dates are consistent with the theory that each lunar month was ordinarily given the seasonal name of the civil month in which it started. Thus, the first month of the lunar year, lunar I Akhet, began on the day of the first new moon in civil I Akhet (Thoth).

The name of the 13th lunar month of a year, when such a month occurred, is unknown. This month could occur in two ways: either by starting in one of the epagomenal civil days, or by starting in the same civil month as the preceding lunar month (e.g. two lunar months might start on I Akhet 1 and I Akhet 30). This last phenomenon is the so-called blue moon.

While the name of the 13th month is unknown, we are able to place it in the sequence of lunar months by examining the seasonally-based lunar names, in which lunar I Akhet is always the month starting in civil I Akhet. Depuydt pointed out that there are, in theory, two possible approaches. The first ("yearly pairing") is that the months ran through a 12-name sequence regardless of the occurence of a blue moon, and the 13th month, if it occurred, was given a name that was always at the end of sequence. This would certainly happen for 13th months starting in the epagomenal days in any case. The second possibility ("monthly pairing") is that the sequence of 12 month names was interrupted by the blue moon whenever it occurred, so that the blue moon had its own name.

Failing an actual attestation of a blue moon, the use of one technique or the other can be detected by examining the relationship between civil and lunar day numbers after the occurrence of one. Ordinarily, since a lunar month starts after the beginning of the civil month from which it takes its name, the day numbers in the first part of the lunar month will lag behind those of the civil month of the same name, but the day numbers of the last part of the lunar month will be in advance of the those of the first part of the next civil month. That is, the following relationships will apply in double dates:

First part of Lunar month X: Lunar Month X day Y = Civil month X day Z, Y < Z
Last part of Lunar month X: Lunar Month X day Y = Civil Month X+1 day Z, Y > Z

With monthly pairing, these relationships will also hold for months occurring after a blue moon. However, with yearly pairing, the blue moon itself will take the name of the next civil month, and will begin before the civil month, causing all subsequent months to begin in advance of the month. This reverses the relationships between civil and lunar day numbers:

First part of Lunar month X: Lunar Month X day Y = Civil month X-1 day Z, Y < Z
Last part of Lunar month X: Lunar Month X day Y = Civil Month X day Z, Y > Z

Depuydt pointed out that the Saite double date given above, II Shomu 13 year 12 of Amasis = (lunar) I Shomu 15, allows this test to be applied. This double date implies that I Shomu 1 (lunar) = I Shomu 29 (civil), a condition that can only be met if there had been a blue moon earlier in the year. But in this case Y=1 < Z=29, which is the result expected for monthly pairing. Hence we may conclude that a blue moon month had its own name. However, we do not know what it was.

Regulation of the Lunar Year

It is uncertain whether the length of the lunar month was determined by a formula or by observation. However, 25 wandering civil years = 9125 days is almost exactly 309 lunar months, to an accuracy of about 1 hour. This allows a notional distribution of 29 or 30 day months to be made over a period of 25 wandering years, creating calendar months which are synchronised to actual lunar months to an accuracy of about a day. Such a table will be reasonably accurate for approximately 300 years (= 12 cycles of 25 years each), assuming it is optimally accurate at the time it is compiled.

Two statements of such a formulaic cycle are known: pRyl. 589 and pdem Carlsberg 9. pRyl. 589 dates from the reign of Ptolemy VI. pdem Carlsberg 9 dates from the reign of Antoninus Pius (in or after A.D. 144), but refers to four earlier 25 year cycles starting in A.D. 44, under Claudius (named as "Tiberius"). The pRyl. 589 cycle survives in a very incomplete state. That of pdem Carlsberg 9 is essentially complete, but in a form that is much more concise than pRyl. 589. It is therefore subject to alternate interpretations. Two significant interpretations have appeared in the standard literature, by Parker and Depuydt.

The introduction to the pRyl. 589 table gives an clear and explicit statement of the theory that 309 lunar months occupies exactly 25 Egyptian years. In its original form the table was a complete and explicit listing of the lunar month dates and month lengths of an entire 25-year cycle; it also gave a statement of the zodiacal sign of the Sun for each month.

Year 1 of the cycle is given as year 1 of Ptolemy VI, which corresponds to year 2 of the proleptic pdem Carlsberg 9 cycle. The surviving fragments give the following sequence:

Thoth 20; [days: 29]; Phaophi 19; d[ays: 30]; Hathyr 19; d[ays: 30]; Choiak 19 [days: 30].....

The month lengths for what appears to be the last 6 months of year 3 and the first three months of year 4 in pRyl 589 also survive:

(3) .... 29, 30, 30, 29, 30, 29
(4) 30, 29, 30.....

It is generally agreed that pdem Carlsberg 9 gives the starting date of lunar months in at least every second civil month. Parker, assuming that the Carlsberg cycle represented actual usage, filled in the gaps on the basis of the Ptolemaic and Roman date equations known to him. However L. Depuydt, In Mem. Quaegebeur II 1277, 177, argued that the papyrus had been interpreted incorrectly and that it must be understood as giving all the start dates of lunar months, except for the 13th month (blue moon or epagomenal) in years of 13 months. Depuydt's interpretation is reflected in the table provided here. In Depuydt's view, I think correctly, the table does not necessarily reflect actual practice, but rather gives a balanced solution to the problem of packing 309 lunar months into 25 wandering years which gives the simplest possible relationship between civil and lunar years.

Depuydt's reconstruction is more systematic than Parker's, but differs from the recorded lunar months used by Parker in several places. This is not a problem unless the table specified the actual sequence of lunar months in the cycle, as Parker had supposed it did. Against Parker's view, one may note that lunar data from Medinet Habu from the reigns of Ptolemy XII and Cleopatra VII, which has come to light since Parker's work, is usually one day out of alignment with Parker's reconstruction.

To the extent that the lunar calendar was governed by any such cycle, pRyl. 589 would probably be the more accurate representation of the Ptolemaic lunar cycle. Enough of the pRyl. 589 cycle survives to show that it is not the same as the pCarlsberg 9 cycle: pRyl 589 places the start of lunar II Akhet at 19 Phaophi, pdem Carlsberg 9 gives 20 Phaophi for the same cycle year, and the surviving entries for years 3 and 4 of the pRyl. 589 cycle are also not consistent with the pdem Carlsberg 9 cycle. However, not enough survives to allow us to reconstruct the pRyl. 589 cycle. For this reason, pdem Carlsberg 9 is used as the basis of the cycle calendar provided here, as it has been in all discussions of this problem since Parker. Actual lunar synchronisms show that this cycle is not historically accurate for the Ptolemaic period, but it may be used to approximate lunar months to within a couple of days, which is usually good enough for chronological purposes.

Website © Chris Bennett, 2001-2011 -- All rights reserved