processes are fixed and others movable is not something that engages his interest;
certainly, it is not an instinctive reflex to conceive of the civil operations and the
natural operations as a duet. Varro is still in precisely the same situation as Cato a
century before, for in the De AgriculturaCato “uses the civil calendar only for the
dates of business affairs, such as contracts. For other purposes he reckons mainly
by the stars.”^149 For Varro, as the case of the Vinalia shows, the opportunity to
make a meaningful connection between the state ’s calendar and the operations of
the natural cycle of the year is not taken up. From this perspective, Suetonius’s
famous comment on the disarray in the Republican calendar, quoted above, reveals
itself as coming from a post-Julian point of view: it is hard to believe that the man
who saw no difficulty in discussing a vintage festival in April was very concerned
“that the harvest festivals did not coincide with the summer nor the vintage festi-
vals with the autumn” (Suet. Jul.40.1).^150
We turn now to Varro’s De Re Rustica,which was published after the Julian
reform, in 37 b.c.e.^151 When Varro talks in the first book of this later work about
the divisions of the year and their relationship with agriculture and nature we are,
all of a sudden, in a completely different world (1.28.1). First he gives the natural
divisions of the year, linking the beginning of the various seasons to the time when
the sun is in various constellations: dies primus est ueris in aquario, aestatis in tauro,
autumni in leone, hiemis in scorpione(“The first day of spring is [when the sun is] in
Aquarius, that of summer in Taurus, of autumn in Leo, of winter in Scorpio,”
1.28.1). He then gives the lengths of these natural seasons (spring contains ninety-
one days, summer ninety-four, and so on). Now he does something he could not
possibly have conceived of ten years earlier, and he marks the novelty of what he
is about to do with a telling phrase, referring to “these figures, correlated to our
civil days, the ones that are now in existence,” quae redacta ad dies ciuiles nostros, qui
nunc sunt^152 — at which point he proceeds to the extraordinary step of giving
Roman calendrical dates for these seasonal phenomena, in a way that would have
been simply and literally impossible before 1 January 45 b.c.e.: spring now begins
on 7 February, summer on 9 May, autumn on 11 August, and winter on 10 Novem-
ber.^153 Not ten years have gone by since the calendar’s reform, and the reorienta-
tion that Varro has performed is profound. In the first of these two works, the fes-
tivals are in chronological order but for the most part undated and not related to
natural processes that they might be thought to track; in the second, the Roman
calendar is capable of capturing the cyclical predictability of nature itself. This is
the first documented step in a process that will eventually lead to the calendrical
dates ousting stellar phases from Latin parapegmata as the index to meteorologi-
- Years, Months, Days II: Grids of the Fasti