Empirical or schematic: the Illahun archiveThe practice of determining the lunar month on the basis of empirical
observations of the moon phases can be inferred from a document of the
Illahun archive (c. nineteenth centuryBCE). In this document, an interval
between the‘feast ofAbd’and the feast of half-month (smdt) amounts to 16
days, whereas it should have been 13 (asAbdis normally on day 2, andsmdt
normally on day 15), and another interval, between half-month and feast of
Abd, amounts to 15 days, whereas it should have been 16 or 17. Both intervals
can be interpreted as resulting from the feast of half-month’s occurring two or
three days too late: not on day 15 of the lunar month, but rather on day 17 or
18.^57 This later date of the half-month feast suggests that it was established
empirically, most likely on the basis of observation of the full moon.^58
Furthermore, the lunar month in which this full moon occurred on day 17
or 18 must have begun too early: An early beginning of the lunar month would
have been most likely due to the empirical observation of the disappearance of
the old moon: for if bad weather prematurely prevented the old moon crescent
from being sighted, the beginning of the next month would have been set too
early.^59
Another document from the Illahun archive, however, suggests that lunar
months were determined on the basis of a calculation or schematic calendar.
This document provides a one-year calendar for the allocation of almonds and
honey to six priestly divisions; the overseer of each division is given a 30-day
term, precisely dated (with starting andfinishing dates) according to the civil
calendar. It is evident from these dates that the 30-day terms correspond to
alternate lunar months.^60 The full sequence of the lunar months can be
inferred from the civil calendar dates in the document, and consists of the
(^57) The reverse, that these intervals result from the feast ofAbd(normally on day 2) occurring
two or three days too early, is impossible, because this would pushAbdinto the previous lunar
calendar month.
(^58) Pap. Berlin 10282 recto (Luft 1992: 114–18 no. 49), as interpreted by Luft (pp. 196–8)
and Depuydt (1997) 149–50.
(^59) This is the reverse of calendars (e.g. Babylonian) based on the appearance of the new moon,
where poor atmospheric conditions have the effect ofdelayingthe beginning of the month.
Consequently, if the criterion for beginning the month was invisibility of the old moon, new
moon visibility could not have been used as a secondary criterion, because this would have led to
contradictions (especially in cases of bad weather). In this lunar calendar, therefore, the coinci-
dence ofAbd(day 2) with new moon visibility could only have been notional (Depuydt 1997:
149, but wrongly arguing on this basis that theentirecalendar must have been non-empirical and
schematic).
(^60) Pap. Berlin 10056 verso (Luft 1992: 73–8 (no. 24)); see Parker (1950) 63–4, Clagett
(1989–99) ii. 174–6, 186–7, and Depuydt (1997) 147–51, 178–84. On the question of when
these months began, see above, n. 55.
The Egyptian Calendar 145