purposes without constituting in themselves a calendar, and without the
weeks being either numbered or named. Just like weekdays, which are effective
especially if combined with other dating schemes (e.g. Julian dates), the days of
the lunar month could have functioned as an effective method of dating—
especially aslunadates always appear, in the inscriptions, together with a
Julian date.^83 In this respect, thelunaformula could have functioned as an
integral part of the inscriptions’dates.
The inclusion of thelunaformula within the dates of the inscriptions is,
indeed, a sufficient reason to interpret it primarily as a form of dating. As we
have seen, it is attested as such in the Ferentium inscription of 67BCE,ina
period when the local calendar itself may still have been lunar, as were other
Italian calendars in this period (see Chapter 4 n. 192). The persistence of this
formula well into late Antiquity could well have been a survival of local Italian
dating methods and calendars, which in the Republican period had been
mostly lunar.
The survival of lunar calendar reckoning in the culture of the LatinWest is
also evident in other ways, not least in the well-attested lunarparapegmata.As
noted above, the availability of lunarparapegmataexplains how the full range
of lunar days could have been used inlunainscriptions, and suggest that a
continuous count of lunar days must have been commonly reckoned. It is
unlikely that the use of lunarparapegmataserved only astrological purposes
and not the more general function of time-reckoning.Parapegmatain the
LatinWest are known to have been used as calendars, at least those that
include the Julian calendar; it is entirely plausible, therefore, that lunarpara-
pegmatawere similarly used as lunar calendars.^84
(^83) A similar phenomenon has been observed in Ch. 3 in relation to Egyptian lunar dates:
although lunar months were generally not identified by name or number, lunar dates were still
used, in conjunction with civil calendar dates, as a method of dating.
(^84) I disagree with Lehoux’s theory, in his otherwise excellent study of Greek and Roman
parapegmata, thatparapegmatawere distinct from calendars and designed to track only‘extra-
calendrical’phenomena, i.e. phenomena not represented in calendars, which would explain why
lunar days are absent in Greekparapegmata(since Greek calendars were anyway lunar), whereas
in the Roman tradition, where the calendar was not lunar, lunar days were sometimes included
(Lehoux 2007: 31, 54, 86, 97). This theory is over-neat, and ignores a considerable number of
Latinparapegmatathat include the Julian calendar and must necessarily have served calendrical
purposes: the Pompeii calendar inscription (above, n. 56), theFastiGuidizzolenses(which is
actually aparapegma), the Nîmesparapegma(both above, n. 28), the Dura-Europosparapegma
(n. 58), and the CapuaFastiparapegma (Lehoux 2007: 194). To these may be added the Julian
calendarparapegmatathat were built into clepsydras, although their prime purpose would have
been to track the seasons (rather than the Julian date) for the calibration of seasonal hours: such
clepsydras withparapegmatahave been found in Salzburg and in Grand (Vosges), see references
in Neugebauer (1975) ii. 870 nn. 5–6; in the river Rhine (AE2003: 1279; above, n. 33); and in
Vindolanda (M. Lewis 2009; I am grateful to Andrew Birley for this reference). A passage of
Cicero (Ad Atticum5. 14, cited by Lehoux 2007: 200–1) is further evidence thatparapegmata
were not extra-calendrical and could serve to track the days of the Roman calendar. On this basis
I am arguing that lunar days inparapegmata, which appear in fact in some of the Julian
324 Calendars in Antiquity