Ancient Chronologies

By

Elias Bickerman

Chapters 1-3 and Tables 1-4 from

Bickerman's Chronology of the Ancient World
(N ew York, 1982 ), pp. 1- 125 .

This material is presented solely for non-commercial educational/research purposes

 CONTENTS
LIST OF FIGURES 6

PREFACE 7

INTRODUCTION 9

I THE CALENDAR 13
The Day 13 The Moon and the Month 16
The Lunisolar Year 22
Greek Calendars 27 The Athenian Calendar 34
The Macedonian Calendar in Egypt 38
The Egyptian Year 40
The Roman Calendar 43 The Julian Year 47
The Natural Year 51 The Zodiac 56 The Week 58

II CHRONOGRAPHY 62
Relative Chronology 62
Naming the Year 63 The Eponymous Year 67
The Eras 70 Indiction 78

III APPLIED CHRONOLOGY 80

Principles o f Reduction 81 Chronography 87
Practical Suggestions 89

ABBREVIATIONS 93
 CONTENTS

NOTES 96

THE TABLES 109

I Astronomical Canon 109
I I Rising and Setting of Stars 112
III Olympic Years, etc. 115
I V Roman Julian Calendar 125
V Lists o f Rulers 126
V I Athenian Archons 138
V II Roman Consuls 140
V III Roman Emperors 163
IX Comparative Table for Early Rome 165
X Chronological Tables 167

in d e x 219

LIST OF FIGURES
1 T h e lunar cycle, p. 17
2 List o f months, p. 20
3 Some local Julian calendars, p. 48
4 Th e earth’s orbit, p. 52
5 T he (geocentric) path o f the sun in different seasons, p. 53
6 T he solar and sidereal day, p. 55
7 T he order o f the planets, p. 59
8 T he week, p. 60
 PREFACE

T his book was originally written at the suggestion o f Eduard

Norden. I was young at that time, and did not realize the
difficulty o f m y task: knowledge is required to prepare a w ork
o f scholarship, but only ignorance gives the courage to publish it.
Nevertheless, m y very imperfect w ork seems to have been o f
some use: it has appeared twice in German, also in Italian, and
most recently in Russian. Its aim is to answer a simple question:
h ow are w e able to date the events o f ancient history? For instance,
w e say that Caesar was assassinated on 15 March 44 b c . H ow do
w e know it? T o answer this kind o f question, w e have to under­
stand the calendar systems used b y the ancients and their time
reckoning.
T he English edition was undertaken at the invitation o f
Professor H. H. Scullard. A t his wish a supplement containing
various chronological tables has been added. T he text itself has
been com pletely revised and often changed to make it clearer or
to adapt it to m y present views. O n this occasion many slips o f
previous editions have been tacitly corrected. It is a pity that the
reviewers o f m y book preferred to praise it instead o f pointing
to its faults. A book and its author profit from blame and not
from approval. W e all ow e a never-to-be-repaid debt o f gratitude
to Jeanne and Louis Robert for their censorious Bulletin
épigraphique.
Professor L. Bonfante (N ew Y ork University) competently
rendered m y Italian text into English; m y students M r Albert
Baumgarten, M r D . Graves and M r Alan Koenigsberg checked
references to sources and made the index. Professors O . Neuge-
bauer and R. A . Parker kindly answered questions o f the author,
w h o was also privileged to use the typescript o f Professor
8 C H R O N O L O G Y OF THE A N C I E N T W O R L D

Neugebauer’s lectures on Astronomical C h ron ology delivered at

B row n University in 1941-2.
I am writing this preface at the end o f m y last year as Professor
o f Ancient History at Colum bia University. It seems fitting to
dedicate the English edition o f m y book to the m em ory o f
W illiam Linn Westermann, m y dear friend and predecessor in
the same Chair, and to M orton Smith, m y dear friend and
successor: et, quasi cursores, vital lampada tradmit.

Colum bia University, e. j . b .

March 1966
 IN T R O D U C T I O N

T ime is the proper dimension of history . A fact is historical

when it has to be defined not only in space but also in time. A fact
is placed in the fourth dimension, that o f Tim e, by measuring
its distance from the present. Chronology, an auxiliary o f
history, enables us to state this time-interval between a historical
fact and ourselves by converting the chronological indications
o f our sources into units o f our ow n time reckoning.
If, for example, it is said that Horace ‘died on the fifth day
before the Kalends o f December when C . Marcius Censorinus
and C . Asinius Gallus were consuls’ (decessit V Kal. Dec. C.
Marcio Censorino et C . Asinio Gallo cotisulibus: Suet. De viris ill. 40)
chronology translates this Roman date into one o f our dating
system: 27 N ovem ber 8 b c , and thus expresses our time-distance
from Horace’s death.
W e count by year-units (1967, 1968, and so on) which do
not recur, and by months and days, which recur every year.
Each complete date, therefore, consists o f tw o parts, the calendar
date, which repeats itself periodically (i.e. 27 November), and the
chronographic date, w hich occurs only once (i.e. 8 bc ).
Such is the division o f this volume, which is meant to be
neither a shortened handbook nor merely a guide for converting
dates, but rather an introduction to the basic elements and
problems o f ancient chronology. The plan o f the book is therefore
to explain the structure o f the ancient calendar, the principles
followed in antiquity in com puting the years, and the rules which
w e can derive from these principles in relating ancient dates to
our ow n time reckoning.
O ur time reckoning uses three standard units: the day, the year
and the month. A (solar) day is the time taken by the earth to
revolve once on its ow n axis. A (tropical) year is the length o f the
10 C H R O N O L O G Y OF TH E A N C I E N T W O R L D

period which the earth takes to complete its revolution around

the sun. This solar year equals 365 days, 5 hours, 48 minutes and
almost 46 seconds. O ur month, on the other hand, does not
depend on natural phenomena. It consists o f a varying number
o f days (28 and 29, 30, 31), the sum o f which amounts to 365
and, thus, equals the number o f days in the year. This irrational
arrangement is a relic o f the Roman calendar which our calendar
continues w ith only one modification.
For practical reasons, a calendar year must be made up o f
integral days. In reforming the traditional Roman calendar, C .
Julius Caesar established a year o f 365 days w ith an added
‘bissextile’ day (now 29 February) every four years to account
for the difference between the solar and the common civil year.
Thus, four Julian years equal 1,461 solar days. Therefore, the
Julian calendar advances by c. 44 minutes every four years with
reference to the sun. A t the end o f the sixteenth century ad the
accumulated difference between the Julian calendar and the solar
year amounted to about ten days. Accordingly, Pope Gregory
XIII omitted ten days in the year 1582 (so that 5 October became
15 October) and suggested that three intercalary days be omitted
every four hundred years, so that the years 1600 and 2000, but
not 1700, 1800 and 1900, are leap years. Except for this correction,
our ‘Gregorian’ calendar is still the Roman calendar as reformed
by Caesar (see p. 47). This is w h y historians use the Julian
calendar for the dates before ad 1582. A year contains 365 days
and begins on 1 January; an additional day (29 February) is
added every four years (1, 5, 9, etc. b c ; ad 4, 8, 12, etc.).
Th e purpose o f chronology is therefore to convert the chrono­
logical references o f our sources into the Julian dates o f our era
(bc , or ad ). The device o f counting backward from the (supposed)
date o f the birth o f Christ was first used by D . Pctavius (in 1627)
and has been in regular use from the end o f the eighteenth
century.1
W e should remember, however, to use the true (Gregorian)
reckoning in calculating the dates o f seasonal events for remote
periods such as the barley harvest in Babylonia c. 1700 bc (c f
S. Langdon and J. K . Fotheringham, The Venus Tablets o f
 INTRODUCTION II

Ammizaduga (1928), 69) or the inundation o f the Nile. For

instance, c. 4200 BC the Julian dates would have been 34 days in
advance o f the Gregorian calendar (and o f the sun): the solstice
24 June (G reg.)=28 July (Jul.). Cf. Ed. Meyer, A P A 1904, 43.
This volum e deals w ith the dates furnished by the ancients
themselves. W e do not take into consideration the methods o f
relative dating developed in archaeology nor the methods o f
direct dating established by modern science.
T he typological method o f archaeologists, for example, dates
a Greek vase according to its style, that is, finds its relative position
within a certain stylistic development. Th e typological evolution
must then be related to some ancient time-scale in order to
obtain an absolute date. Th e relative dating o f archaic Greek vases
is based on finds made in Italy and depends on the dates o f the
Greek colonies in Italy. Th e relative chronology o f these colonies
is given by Thucydides (VI, 1): Gela was founded forty-five
years after Syracuse, and so on. Thucydides’ relative chronology,
in turn, is related to our reckoning w ith the help o f the tables
o f Eusebius (sec p. 88), where, for instance, the founding o f
Syracuse is recorded under a date which corresponds to 733 b c .2
O n the other hand, methods o f natural science allow us, in
certain circumstances, to determine the age o f artifacts directly.
For instance, the remains o f ancient organic matter, such as wood,
w oo l and bones, that once was alive, can be dated by the Carbon-
14 technique; trees, by tree-rings; magnetic measurements and
the thermoluminescence technique serve for dating pottery, and
so on. O f course, the historian must use his own judgm ent in
evaluating the information furnished b y a laboratory. For
instance the age o f a log given by its tree-rings or by the radio­
carbon method refers to the time when the tree was cut down.
T he log in question may have been used in a building a long time
after the construction o f the latter, for instance for repair.3
 CH APTER I

THE CALENDAR

O ur calendar takes into account the revolution o f the sun, which

produces the ‘day’ and the ‘year’. O ur ‘month’ is a conventional
unit. Ancient peoples, however, with the exception o f the
Egyptians and the Romans, based the civil calendar on the phases
o f the moon as well as on the movement o f the sun.4

THE DAY
The regular alternation o f day and night constitutes the first
measure o f time. T he Celts and the Germans counted b y ‘nights*
(Caes. B .G . VI, 18; Tac. Germ, n ) ; Hom er reckoned time
according to ‘dawns’.
Th e w orking day, in practice, coincided with the daylight
hours because o f the insufficiency o f artificial means o f lighting.
T he period o f darkness did not count. Th e w ord ημέρα (hetnera:
‘day’) is used in tw o senses: (i) for the time from the sun’s rising
to its setting; (2) for the time from the sun’s rising to its rising
again (Geminus, Ekmenta astronotniae 6).5 Th e same is true for
the Latin w ord dies, for our w ord ‘day’, and so on. (The com po­
site word νυχθημερόν for ‘a night and a day’, used, e.g., in Paul 2
Cor. 11, 25, is not attested before the first century a d .) Thus, the
day was everywhere considered to begin in the morning. This
was true in Greece and Rome, in Babylonia and Egypt, as it is
true for our ow n usage. Pliny (N.H. II, 188) w rote: ‘the actual
period o f a day has been kept differently by different people . . .
b y the com m on people everywhere from dawn to dark’ (ipsum
diem alii aliter observare . . . vulgus omne a luce ad tenebras).
O n the other hand, the complete day, for the purpose o f the
calendar, is generally reckoned in conformity w ith the respective
calendar systems. Th e peoples w h o use lunations as the basic time-
measurement (p. 16), for instance the Athenians (Varro, ap.
 14 C H R O N O L O G Y OF THE A N C I E N T W O R L D

Gell. Noct. Att. Ill, 2), the Gauls (Caes. B .G . VI, 18), the Germans
(Tac. Germ. 11), the Hebrews, and others, counted the complete,
twenty-four hour, day from evening to evening. W e, too, still
speak o f a ‘fortnight*. W here, as in Egypt, the calendar dis­
regarded the moon, the official day began at dawn. The Zoro-
astrians, w h o condemned the lunar reckoning as false, insisted
that the day was a period between tw o sunrises (cf. H. S. N yberg,
T exte zum Mazdayanischen Kalender, Uppsala Uttiv. Arsskrifi
1934, 11). Again, the Babylonian astronomers used the midnight
epoch for lunar computations (O. Ncugebauer, PA P hS 107
(1963), 529).
For some reason, which was already unknown to the Romans
themselves, the Roman dies civilis (cf. Thes. Ling. Lat. Ill, 1214,
60) also began at midnight (Plut. Quaest. Rom. 84).
Th e different periods o f the natural day were distinguished
according to the movement o f the sun (e.g. ‘morning’) and to
man’s use o f the day-time (eg. ‘dinner-time’). The corresponding
Greek expressions are collected in Pollux 1, 68; the Latin in
Censorinus 24 (cf W . Sontheimer, R E IV A , 2011). T he require­
ments o f war led to the division o f day and night into watches
(φυλακαί, vigiliae). Th e Babylonians, the O ld Testament and
Homer (II. X , 253; Od. XII, 312) had three watches during the
day and three more during the night, while the Greeks and the
Romans later adopted the Egyptian system o f four watches
(Eurip. Rites. $), which was also widely used in civil life to
indicate parts o f the night (cf. e.g. Asclep. Anth. Pal. V , 150).
The division into hours is first attested in Egypt. As early as
c. 2100 b c , the Egyptian priests were using the system o f twenty-
four hours: ten daylight hours, tw o twilight hours, and twelve
night hours. This arrangement, based on the decimal method o f
counting, gave w ay c. 1300 bc to a simpler system w hich allotted
12 hours to the day and 12 hours to the night. The Babylonians
similarly divided the day and the night b y 12. T he Greeks,
according to Herodotus (II, 109), learned this arrangement from
the Babylonians. T he Greek term ώρα, from which, via Latin
hora, w e get our w ord ‘hour’, originally referred to a season,
then to the fitting or appointed time (e.g. Arist. Ath. Pol. 30, 6;
 TH E CALENDAR 15
Sappho, ap. Hcphacst. D e re tnetr. n , 3 = D . L. Page, Poetac Melici
Graeci (1962) fr. 976, for a lovers’ assignation). The sense o f
‘hour’ is first attested in the second h alf o f the fourth century
bc (Pytheas in Geminus, Elem. Astro. 6, 9; Arist./r. 161). A t the
same time the expression a ‘half-hour’ appears in our sources
(Menander).
Th e hour o f the ancients, however, was not, as it is for us,
part o f the w hole (astronomical) day, but χ^ part o f the
actual length o f the time from sunrise to sundown and, again,
from sundown to sunrise. Thus, the length o f an hour varied
according to the latitude and the season.6 These seasonal hours
equalled between £ and f o f our hour (for a table o f correspon­
dences see Ginzel II, 166; Kubitschek, 182). T he hours were
reckoned from the rising o f the sun or, at night, from the com ing
o f darkness. Thus, the seventh hour roughly corresponded to our
midday (or midnight)7 and marked the end o f business hours.
*Εξ ωραι μόχθοις Ικανώτα rat, ai δε μετ' αύτάς γράμμασι δεικνυμεναι
Ζ Η Θ Ι λεγουσι βροτοίς (Anth. Pal. X, 43)· ‘Six hours are most
suitable for toil, and the four that come after, when shown in
letters, say to men “ Live” .’ (The Greeks used letters o f the
alphabet as figures: thus 7, 8, 9 and 10= Ζ Η Θ Ι = Live.) The
ninth hour, dinner-time in Imperial Rom e (Mart. IV, 8), varied
from 1.30 to 2.30 p.m . (Ideler, Lehrbuch, 260).
As Xenophon (Mem. IV, 3, 4) says, the sun during the day, the
stars during the night, showed the time. T he length o f a man’s
shadow indicated the progress o f the day (Aristoph. Eccles. 652).®
V ery primitive hand-tables gave the approximate relation
between the length o f the human shadow and the (seasonal) hour
o f the day. For the nightly offices in the temples, Egyptian priests
as early as c. 1800 b c used the so-called star-clock. (The apparition
o f a certain star in the proper decade o f a month signalled the
hour.) Sundials and water-clocks made possible a more precise
measurement o f time.9 The earliest preserved water-clock (e. 1600)
and shadow-clock (c. 1450) have been found in Egypt. According
to Herodotus (II, 109) the Greeks learned to use the sundial from
the Babylonians. A later tradition (Favorinus, ap. D iog. L. II, 1)
ascribed the construction o f the first Greek sundial to Anaxi-
16 C H R O N O L O G Y OF THE A N C I E N T W O R L D

mander o f Miletus (c. 550) or to Anaximenes, his disciple (Plin.

N .H . II, 187). In Rome, the first sundial was constructed in
293 bc (Plin. N .H . VII, 213).
O ur hours o f equal and constant length were invented and
used b y savants such as astronomers and writers on cosmography
(cf. Strabo II, 5, 36, p. 133). There were tw o systems o f counting,
w hich divided the complete day into tw elve equal parts, as the
Babylonian priests did, or into twenty-four constant units, as the
Egyptian priests reckoned. Th e Hellenistic astronomers adopted
the Egyptian division o f the calendar day but, follow ing the
Babylonian counting system, they divided the Egyptian hour
into sixty equal parts. T h ey used water-clocks in w hich a pre­
determined quantity o f water would always pass in the same
period o f time. Medieval astronomers follow ed the same arrange­
ment, and mechanical time-keepers were scaled accordingly, so
that w e still count sixty minutes to one hour. The use o f the
variable hour, however, was retained in everyday life, and per­
sisted in some parts o f the Mediterranean world w ell into the
nineteenth century.10

THE MOON AND THE MONTH

As constant as the alternation o f day and night is the w axing and
w aning o f the moon which is repeated (on the average) every
29-53 days. T he moon has no light o f its own, but ‘the sun places
the brightness in the m oon’, as Anaxagoras said (Plut. De facie
929 b), to w hom Plato (Cratyl. 409 A) attributed the discovery
that the moon receives its light from the sun. Because its period
o f rotation on its axis is about the same as the period o f its circling
the earth, the same side o f the moon always comes into our view.
But when the moon, the sun and the earth are in a line so that
the moon comes between the sun and the observer on the earth,
the sun illuminates the back o f the moon, and the satellite is
invisible to us {conjunction synodos). As the moon continues to move
eastward (that is counter-clockwise) from the sun, it reappears
from one to three days later at twilight, in the western sky, as
the new crescent. Th e illuminated (right) part o f the lunar hemi­
sphere waxes every night. About fourteen days later, when the
 THE CALENDAR 17

Fig. 1. The lunar cycle

m oon is in opposition to the sun, so that the observer is between

the tw o celestial bodies, the whole face turned toward us is
illuminated in the light o f the full moon (<dichomenia). Afterwards,
the moon again approaches the sun; only the left side o f the
lunar hemisphere shines, then the moon disappears at dawn in
the eastern sky, and the lunar cycle begins anew (Fig. 1).
Alm ost every people o f the earth have used the lunar phases
for measuring time: lutta régit menses (O vid, Fasti III, 883). The
Greek w ord μ ψ [men) and our term m onth’ equally point to the
moon. Th e same is true for the terminology o f the Semitic
languages. For instance, in H ebrew the w ord yerah means ‘m oon’
and ‘month’, and the other w ord for ‘month’, hodesh, properly
signifies the ‘new ’ crescent.
A s a matter o f fact, almost all the peoples o f the Mediterranean
w orld, the Celts (Plin. N .H . xvi, 44), the Germans (Tac. Germ.
11), as well as the Hebrews and the Babylonians, began the
month at the apparition o f the young crescent, as the Islamic
peoples still do today for their religious calendar: the new moon
signals the longed-for end o f the fast month (Ramadan). Th e
beginning o f the month was sometimes publicly announced (for
Greece cf. Nilsson, Kalender 29). In early Rome, the pontifex
minor observed the sky and announced the new moon and,
l 8 C H R O N O L O G Y OF THE A N C I E N T WO R LD

consequently, the new month to the king (Macr. Sat. I, 15, 9).
N o t even the rationalization o f the Greek calendar (sec below,
p. 19) could separate the beginning o f the month from the new
moon: ‘D o yo u not see, how a slender-horned moon in the
western sky marks the beginning o f the new month?’ (Arat.
Phaeti. 733).
In principle the lunar months o f all the ancient peoples run
parallel. ‘The (Doric) month Karncios is what the Athenians call
Metageitnion’ (Plut. Nie. 28). Th e Athenian Pyanepsion, the
Macedonian Dios, the Babylonian Tashritu, and so on, were
different labels o f the same lunation.
Figure 2 shows the correspondence o f names o f the month in
several calendars. Y e t the observation o f the crescent could be
hampered by the local atmospheric conditions, and the beginning
o f the new month at a given place could be accordingly delayed.
For instance, Ashurbanipal (668-626) received a report as follows:
O n the 29th w e made an observation. O n account o f the appear­
ance o f clouds w e did not see the m oon/11
O n the other hand, neither is the length o f the lunation con­
stant (it varies from 29*26 to 29*80 days) nor is the interval
between the conjunction and the visibility o f the new crescent
always the same. Several variable factors, such as the distance o f
the m oon from the sun at the time o f conjunction, determine the
visibility o f the new moon, and the computation o f these factors
became the main problem o f Babylonian astronomy in the
Hellenistic age. Last but not least: sighting the new crescent also
depends on the longitude and latitude o f the observer. Points in
the west have a later sunset than points in the east. O n the other
hand, i f the interval between the conjunction and the apparent
new m oon varies between 16 hours 30 minutes (in March) and
42 hours (September) in Babylon (latitude 32*5°, longitude 45e)
it oscillates between 23 and 69 hours in Athens (latitude 38°,
longitude 23e).12 For Greece, Geminus (9, 14) gives a general
rule: ‘Th e new moon is visible at the earliest one day, at the
latest three days after the conjunction.’ Therefore, as based on
the sighting o f the new moon, tw o or three months o f 30 days
(or o f 29 days) could occur in a ro w .13 O n the other hand, the
 THE CALENDAR 19

government sometimes antedated the beginning o f the month.

A court astronomer could write to Esarhaddon (681-668): O n
the thirtieth I saw the moon, it was in a high position for the
thirtieth day. The king should wait for the report from the city
o f Ashur, and then m ay determine the first day o f the month.’ 14
A s long as the beginning o f the month was determined by
observation o f the new crescent, the months o f all Mediterranean
peoples ran parallel. But the lunation is an awkward instrument
for measuring time. It is the movement o f the sun w hich deter­
mines the succession o f seasons and, thus, the rhythm o f man’s
life. T he lunation, however, is not an even divisor o f the solar
year. Th e earth makes a complete circle around the sun in 365^
days. Therefore, the solar year is longer than twelve lunations by
about 11 days (29 ^ x12= 354 ) and about 18 days shorter than
13 lunar months. Each lunar month falls behind 11 days in the
twelve-m onth solar year and, within the cycle o f 32^ years, passes
through all the four seasons. This is what happens in the M oham ­
medan calendar. Therefore, as Geminus (8) says, the ancients had
before them the problem o f reckoning the months by the moon,
but the years b y the sun. The evolution o f the calendar, thus,
follows three logically and also historically successive stages: 1.
Separation o f the beginning o f the month from the sighting o f
the new moon. 2. T he empirical adjustment o f the lunar count
to the course o f seasons, that is, practically to the solar year.
3. T he cyclic calculation o f lunar months. Th e first stage is reached
by most peoples. Th ough the Mohammedan month in principle
begins with the crescent, the beginning o f the fast month Ramadan
was fixed in Turkey by calculation from the date o f the latest
observed new m oon (cf. Ideler, Lehrbuch, p. 501). Th e Greeks
never went and never wanted to go beyond the second stage.
Th e Babylonians mastered the third problem. The Egyptian
official calendar did not take the moon into account; and die
Romans, in historical time at least, disregarded lunation as a
time-measure. Accordingly, w e have to deal separately w ith the
Greeks, the Babylonians, whose time-reckoning was follow ed in
the w hole Levant, the Egyptians, and the Romans, w ho, in the
end, created our ow n calendar system.
20 C H R O N O L O G Y OF THE A N C I E N T W O R L D

Athens Delos Miletus Delphi

X. ' Εκατομβαιών 'Εκατομβαιών Πάνημος I. *Απελλαΐος
Μεταγειτνιών Μεταγειτνιών Μεταγειτνιών Β ουκά π ος
Βοηδρομιών Βουφονιών Βοηδρομιών Βοάθοος
Πυανεφιών Ά ηατουριω ν Πυανοφιών Ή ρ α ΐο ς
Μαιμακτη ριών Ά ρη σ ιώ ν Ά π α τ ουριών Δηδοφόριος
Ποσειδεών 1 Ποσειδεών Π οσειδεών Π οιτρόπιος 1
Γαμήλιων I. Ληναιων Ληναιων Ά μ ά λ ιο ς
Άνθεστηριών Ιε ρ ό ς *Ανθεστη ριών Βύσιος
Έλαφηβολιών Γαλαξιών I. Αρτεμισιώ ν Θε οξένιος
Μουνυχιών Α ρτεμισιώ ν Ταυρεών ' Ενδυση οιτρόπιος
Θαργηλιών θαργηλιών Θαργηλιών Ηράκλειος
Σ κιρ οφηρυίιν Πάναμος 1 Κ α λ α μ α ιώ ν *ΙλαΤης

Actoli^ Thessaly Boeoda Rhodes

Λαφραΐος Φυλλικός Ί π ποδρόμιος Πάναμος
Πάναμος ι.Ίτώνιος Πάναμος Καρνεΐος
I. Προκόκλιος Πάνημος Π αμβοιώ τιος Δ ά λιος
Ά θα ναΐος θεμίστιος Δαμάτριος I. θεσμοφόριας
Βουκάτιος Ά γ α γ ύ λ ιο ς Ά λαλκομόνιος 1 Διόσθυος
Δ ιο ς 1 Έ ρμ α ΐος I. Βουκάτιος θενδαίσιος
Εύσαΐος Α π ο λλώ ν ιο ς 1 ‘Ερμαΐος Πεδαγείτνιος
*Ομολώιος Λεσχανόριος Προστατήριος Βαδρόμιος
* Ερμαιο ς *Αφριος Ά γριώνιος Σμίνθιος
Διονύσιος θ υ ΐο ς θιού ιο ς Ά ρ τ α μ ίτιο ς
Ά γο νη ο ς *Ομολιός *Ομολώιος *Α γριάνιος
Ίπ ποδρόμιος Ίπ ποδρόμιος θειλούθιος 'Υακίνθιος

Epidauros Cos Macedonia Babylonia (Jews)3

ι. Ά ζ ό σ ιο ς Πάναμος Λώιος Duzu (Tammuz)
Καρνεΐος Δάλιος Γορπ αΐος Abu (Ab)
Π ροράτιος Ά λ σ ε ίο ς *Υπερβερεταΐος Ululu (Elul)1
' Ερμαιο ς X. Καρνεΐος I- Δ ιο ς Tashritu (Tishri)
Γάμος θευδαίσιος * Αιτελλαΐος Arahsamnu
(Marheshvan)
Τόλεος Πεταγείτνυος Αύδναΐος Kislimu (Kislev)
Ποσιδαΰος Καφίσιος Πε ρίτιος Tebetu (Tebeth)
Ά ρ τ α μ ίτιο ς Βαδρόμιος Δύστρος Shabatu (Shebat)
Ά γ ρ ιά ν ιο ς Γεράστιος Εανδικός Addaru (Adar)1
Π άναμος Ά ρ τ α μ ίτιο ς ' Αρτεμίσιας I. Nisanu (Nisan)
Κύκλιος Ά γ ρ ιά ν ιο ς Δ αίσιος Aiaru (Iyyar)
Ά η ελλα ΐο ς *Υακίνθιος Πάνεμος Simanu (Sivan)

Fig. 2. List of months

 THE CALENDAR 21

NOTE
1 They are the normal leap months, though other months could also be
intercalated. For Athens c f W . K. Pritchett, C P h 1968, 53. The order o f the
months in this Table follows the Attic calendar, in which Hckatombaion
usually fell in high summer. The succession o f the months in other calendars,
however, is not always certain, and the correlation w ith the Athenian calendar
is often hypothetical.
Our knowledge o f Greek calendars is very limited. For instance, w e do not
know all the months o f Argos and Sparta, and cannot fill up the gaps by
conjecture (cf. W . K. Pritchett, A ] A 1946, 358). The calendar o f the Thessalian
League was not followed, for example, in die Thessalian city o f Scotussa (cf
J. Pouilloux, B C H 1952, 449). The Greek months were generally named after
festivals, and the festivals o f the same name could bc celebrated at different
times in different cities. The same name could also bc pronounced differently
in another city: the Macedonian month Loos was called Olaios in the (Mace­
donian) city o f Thessalonikc and in the East o f Parthia (cf L. Robert, R P h 1974,
193, n. 7). Again, a festival and a month name could be peculiar to a specific
city, e.g. Bosporius to Byzantium ( c f L. Robert, R P h 1959, 230). Furthermore,
the months’ names were changed for political reasons—for instance, to honour
a king (cf K. Scott, Y C S 1931, 199; L. Robert, in Melanges Isidore L év y (1953)1
560, and in M onnaies antiques en Troade (1966), 15.
O n Greek calendars see Samuel ch. Ill (and Index o f months, 284) with the
indispensable addenda and corrigenda o f Robert (1973), 77· For Istros c f D . M.
Pippidoi, Epigraphische Beitràge z u r Geschichtc Istrias (1962), 57; for Samothrace
c f L. Robert, Gnomon 1962, 56. Foreign groups in the Hellenistic Age sometimes
used the native calendar: see e.g. P. Roussel, L es Égyptiens à D élos (1916), 204.
2 For the Sumerian months see Y. Rosengartcn, L e concept sumérien de consom­
mation (i960), 408, and A. Falkenstcin, Festschrift fu r J . Friedrich (1959), 148. On
calendars in Ebla in the third millennium bc cf. G. Pettinato, Orietis Antiquus
( 1977)» 157· The names o f the Babylonian months given above originated in
Nippur and became widespread after c. 2000 bc (S. Langdon, Babylonian
Menologies ( i 935))· O n Babylonian months before the introduction o f the
Nippur calendar c f D. O. Edzard, A B A 72 (1970), 140. Calendar o f Mari: J. R.
Kupper, in Symbola . . . F. M . T h. de Liagrc B oh l dedicatae (1973), 260. Baby­
lonian month names at Ugarit: Ch. Virolleaud, L e palais royal d ’ Ugarit, II
(1957), no. 162. The Hebrews adopted the Babylonian calendar after 587 BC
under Babylonian dominion.
22 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

THE LUNISOLAR YEAR

Experience shows that on the average a lunation lasts no more

than thirty days. This makes it possible to regulate the length o f
a month, w ithout abandoning its relation to the moon. The
Sumerians, then the Babylonians, and the peoples following
the Babylonian system, eg . the Assyrians, limited the length o f
a month to a maximum o f thirty days. Th e first appearance o f
the new crescent on the eve o f the thirtieth day o f a month
marked the beginning o f a new month. If, however, the new
crescent, for whatever reason, was invisible, die next month
began anyw ay, on the eve o f the thirty-first day o f the current
month. Months o f 29 and 30 days therefore alternated in
irregular sequence.15 T he adjustment o f the lunations to the
solar year was more difficult A s a matter o f fact, many
primitive peoples paid no attention to this problem. They
did not care h o w many lunar months follow ed one another
between tw o crops.
Th e fiscal needs o f the government, however, demanded a
certain stability in the calendar. For instance, it was convenient
for the central administration that a certain tax should be paid
in a certain month in the whole territory o f the state. The
Sumerian bureaucracy, as early as c. 2500, advanced to the practice
o f exact and detailed daily, monthly and yearly accounting [cf.
M . Lambert, R H i960, 23). The lunisolar year, that is, the
agricultural year o f tw elve lunations, was probably an accounting
device. Sumerian records from c. 2400 give evidence for the
practice o f inserting months from time to time in order to keep
the traditional month o f the barley harvest, the Nisanu o f the
Babylonians, in the harvest season.
T he intercalation was ordered by the government. For instance,
the Babylonian king Hammurabi, c. 1700 b c , decreed:16 ‘Since
the year has a deficiency, let the month which is beginning be
know n as the second Ululu, but the tribute due in Babylon on
the 25 th o f the month Tashritu, let it arrive in Babylon on the
25th day o f U lulu II.’ In other words, the month following
Ululu, which usually was called Tashritu, was to be U lulu II,
 THE CALENDAR 23

so that the month o f Tashritu was moved ahead thirty days. B y

means o f such additional months which were inserted irregularly,
on occasion tw o or three times during an agricultural year, and
at varying intervals, the Babylonians and the peoples o f Western
Asia generally regulated their calendar dow n to the sixth century
b c . Trade in agricultural commodities, as early as c. 1900 and as
late as c. 525, was often stipulated in terms o f the ideal calendar.
Th e dates w ere to be delivered in the month o f ‘Tishri’, though
in a given year the time o f picking dates could fall in a month
w ith a different name according to the official calendar.17 It is
probable that the farmer and the merchant relied on the stellar
calendar (p. 51) which was independent o f the vagaries o f the
official time-reckoning.
Ptolem y (Almag. Ill, 7 p. 254, ed. Heiberg) tells us that the
ancient observations o f heavenly phenomena were preserved
almost com pletely from the reign o f the Assyrian king Nabonassar
(747-733) onwards. Some reports o f court astronomers dating
from the first h alf o f the seventh century bc have been discovered.
The lunar eclipses w ere systematically observed and recorded
from c. 730 (cf A . J. Sachs, Late Babylonian Astronomical Texts
(1955) p. xxxi). Th e numerical relation between the length o f
lunar months and that o f solar years could have been established
as early as the seventh century. Yet, as late as the third quarter o f
the sixth century, and perhaps for a long time afterwards, official
letters continued to inform the local officials that the current year
should be embolismic. O n the other hand, cuneiform documents
show that from c. 600 the intercalations follow ed certain norms.
Between C11 and 387, that is, for 224 years, w e know o f 78 leap-
years.18 Since the quality o f m any years is still unknown, it
is possible that the court astronomers follow ed the simple rule o f
3 intercalations for each 8 years. It is also possible that from the
second part o f the sixth century on, they follow ed the schema o f
7 intercalations for every 19 years, though the choice o f inter­
calated years m ay have been decided from case to case. As
Geminus put it: ‘It is a matter o f indifference if, while preserving
the same disposition o f intercalary months, yo u put them in
other years.’ In any case, the Babylonian astronomers succeeded
24 C H R O N O L O G Y OF THE A N C I E N T W O R L D

in limiting the variations o f the N ew Year’s date. Thus, under

Cyrus, between 538 and 520, 1 Nisanu never fell before 12 March
or later than 18 April. (Easter now falls between 22 March and
26 April.) In other words the first month always coincided with
the early spring season, while the beginning o f every month
agreed w ith the course o f the moon.
Th e prestige o f the Babylonian civilization was such that its
lunisolar calendar, imperfect as it was at that time, was adopted
c. 1100 by the Assyrians.1** Later the Babylonian kings, like the
Egyptians before them (O. Tufncll, Lachish (1958), 133), propa­
gated their reckoning system in the conquered territories (cf.
R. Dhorm e, RAss 1928, 54), as in the case o f the Jews.
T he pre-Babylonian time reckoning o f the Hebrews is virtually
unknown. It is certain that the calendar was lunisolar. Th e names
o f some months are known and seem to refer to agricultural
seasons. For instance ‘Abib' (Ex. 13, 4) is the time o f ripening
barley. T he months were also numbered. In 586, after the
annexation o f Jerusalem by Nebuchadnezzar, the Jews began to
reckon by the regnal years o f the kings o f Babylon (e.g. II Kings
24, 12) and to use the imperial calendar. A s the ancient Rabbis
already noted, the Jews had also adopted the Babylonian month
names: Nisan is Nisanu, and so on.20
T he Persian kings, after the conquest o f Babylon in 539,
adopted the Babylonian calendar. In the reign o f Artaxerxes II
(c. 380) the court astronomers switched definitely to the 19-year
cycle, which became standardized in 367: from n ow on, the
month Addaru II was intercalated in the years 3, 6, 8, 11, 14 and
19, and the month U lulu II in the year 17 o f every cycle. In this
w ay, the variations o f 1 Nisanu were reduced to 27 days and the
difference between the 19 solar years and 235 lunar months
brought down to c. 2 hours. A s a result, the corresponding years
o f each cycle were practically identical: in 367, in 348, in 329,
and so on, 1 Nisanu coincided with 2T March.
Like their predecessors on the throne o f Babylon, the Achae-
menids made the Babylonian calendar official in the whole
Persian empire. This is shown b y the documents found at
Elephantine in Egypt. Since these records happen to com e from
 THE CALENDAR 25

a Jewish military colony, modern scholars erroneously speak o f

a ‘Jewish’ calendar at Elephantine.21 N ew ly discovered papyri
prove that this calendar was used by Gentiles and that it was
the official calendar o f the Persian empire to the end o f the
Achaemenids (cf. E. J. Bickcrman, ArchOr 1967, 205).
After the fall o f the Persian empire, Sclcucus I continued the
practice o f the Achaemenids. He ordered that the ‘Syrian’
(Babylonian) months receive Macedonian names (Malalas, p. 257,
O xon.). For the Seleucid court and the Greek settlers Nisanu
became Artemisios, and so on. Later, the Parthian kings followed
the Seleucid arrangement.22
W e do not know whether the Selcucids regulated the inter­
calation in the calendars o f the subject cities. W hen the Greek
cities became independent, they were free to rearrange their time
reckoning as they wished. As a result, at the time when the
Roman emperors imposed the use o f the Julian calendar (e.g.
see p. 50), i Dios o f Ascalon corresponded to 1 Apcllaios o f
nearby Gaza, and fell nine days behind 1 Dios o f T yre (cf.
Fig· 3)· O n ^ other hand, tw o horoscopes from Dura-
Europos, coins minted at Seleucia on the Tigris, the usage o f
Josephus, w h o equates Nisan with Xanthikos, Dios with Mar-
chesvah (Bab. Arah-samna) and so on, and last but not least the
fact that in the Julian calendar o f Antioch the first month o f the
year was Hypcrbcretaios which corresponded to October— all this
evidence proves that from the first century ad on, the Mace­
donian months were one month behind the Babylonian calendar:
D ios n ow corresponded to the eighth and not to the seventh
month o f the Babylonian reckoning. W e do not know when,
how , and for what reason this happened. In the Parthian empire
the change occurred between ad 17 and 31, as it seems. A single
excessive intercalation, ordered, for whatever reason, b y the
Parthian king, w ould suffice to disturb the series o f Macedonian
months. But neither the Jews in Palestine nor the city o f Antioch
in Syria were subjects o f the Arsacids.2*
Th e aforementioned vagaries o f local calendars were sometimes
caused by arbitrary intercalations. But the fasti were also a part
o f the given religious system. For instance, the Mosaic law bound
20 C H R O N O L O G Y OF THE A N C I E N T W O R L D

the beginning o f the new month to the new crescent and the
liturgical year o f Jerusalem depended on the time o f barley
ripening (Lev. 23, 10; cf. Ex. 12, 2). Th e arbitrary or precalculated
calendation o f Babylon must have disagreed again and again with
the sighting o f the new moon in Jerusalem and the grow th o f
crops in Judaea. Thus, the religious calendar o f Jerusalem became
separated from civil reckoning. Months and days (cf. S. Gandz,
J Q R 1949, 264) were inserted at convenience, though the science
o f the ‘calculators o f the calendar’ was not disregarded. As late
as the second century ad the Jewish authorities ordered the inter­
calation when the need arose. ‘The doves being still young, the
lambs still weak, and the (barley) grain not yet ripened . . . I have
decided to add thirty days to the year/
W e do not know when and h ow the new system was estab­
lished. Th e schismatics o f the Dead Sea Scrolls community
refused to accept it, and used dicir own schematic calendar for
‘the proper reckoning o f the time’ o f festivals.24 Thus, the
manipulated, ‘pontifical’ calendar o f the Tem ple was already in
use in the first century bc . Therefore, it is impossible to deduce
the date o f Christ’s last Passover and o f the Crucifixion from
any scheme o f fixed calendars (in fact, there is no calendar date—
day and month, or even just a month name— in the whole N ew
Testament). Later, but not before the fourth century, the Jewish
authorities accepted the principle o f precalculated calendation for
the liturgical year and, for this purpose, adopted the same Baby­
lonian cyclical scheme which regulated the civil calendar.25
Thus, the Jewish religious calendar o f today, w ith its Baby­
lonian month names and the Babylonian arrangement of inter­
calations, is still the Babylonian 19-year scheme, albeit with some
minor modifications. The great ‘elegance’ o f this reckoning was
praised by J. Scaliger, the founder o f chronology as science
(De emend, temper. (1583), 294). For similar religious reasons, the
lunisolar calendar continued to be in use in the Orient despite
the introduction o f the Julian calendar (see p. 50). In fact, it was
not the solar year o f the Caesars but the Islamic, purely lunar,
calendar which ended the use o f the cyclical (Babylonian,
Seleucid) time-measurement in the Near East.
 THE CALENDAR 27
GREEK CALENDARS
Th e Greeks went their ow n way. The early history o f the Greek
calendar is virtually unknown. T he reading o f some month names
in Mycenaean and Knossos texts, written before c. 1200 b c , is
uncertain, and, were it certain, w ould not help the chronologist
much. Th e word meno w ould indicate, it seems, that these months
were lunar. K hm er is reticent about any calendar. W e learn from
him that the apparition o f the new moon (Od. X IX , 306) was a
festive occasion (Od. X X , 156), but he mentions no month names,
and does not number the months within the year, though he
counts months (lunations) o f pregnancy (//. X IX , 117. Cf. Hyntn.
Merc. 11). A Homeric year seems to be seasonal: the year
goes wheeling around and the same seasons return (Od. X I, 294;
c f Hes. Th. 58; Op. 561). Th e Homeric Hymns and Hesiod speak
o f the same primitive calendar. Hesiod numbers the days in the
period o f a ‘w axing’ and o f a ‘waning’ month, but he can also
number the days consecutively through (the ‘29’ : τρισανάδα,
Op., 814), and he speaks o f the ‘middle’ days o f the month.26
W hen and h ow the later calendar system o f a lunisolar year
began, w ith months named after festivals and divided into
decades, w e do not know . T he hypothesis27 that the reform
originated at Delphi in the eighth (?) century cannot be either
disproved or proved. Its force is weakened b y the observation
that the sources do not mention this activity o f the Delphic
oracle.
Th e names o f the months were generally derived from a
festival which was celebrated in the given month. For instance,
Lcnacon was the month in which the Dionysiac festival o f the
Lenaea was held, and so on. The months within the year and the
days within the month were not counted, except for some
Hellenistic calendars (cf, e.g., L. Robert, La Carie Π (1954), 194 ;
E. L. Hicks, W . K . Paton, Inscriptions o f Cos (1891) Index V ;
P. Herrmann, D IV A 80 (1962) 8).
A month was rather divided into three decades, and the days
were then counted within the decade.28 T he origin o f this
tripartite division, w hich was already used b y Hesiod, is unknown
(cf Ginzel II, 319; E. Gjerstad, OpuscAa Atheniensia T (1953)» 187).
28 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

T he problem o f the irregular length o f the visible lunation was

solved in Greece as follows: ‘For business and social life’ (προς
τήν πολιτικήν αγωγήν) the length o f the m onthly period was
rounded o ff to 29J days, so that tw o months came to 59 days.
For this reason the civil months (oi κατά πόλιν) were considered
alternately full (πλήρςς), consisting that is o f 30 days, and hollow
(κοίλοι), o f 29 days (Geminus, 8, 3). The synchronization with
the moon was therefore lost, so that the Greeks had to distinguish
between the civil ‘new m oon’ (νουμηνία), that is, the first day o f
the month, and the actual new moon, νουμηνία κατά σελήνην
(cf. Thuc. II, 28). N othing illustrates the religion o f the polis
better than the fact that the festivals o f gods were celebrated
according to the civil calendar (cf. p. 36). But the Greeks had
no priestly caste which could have opposed this rationalization
o f the fasti. W e do not know when the Greeks limited the length
o f the year to twelve lunations. Homer, o f course, knows that
there is a sun year (e.g. Od. X IX , 306), but neither he nor Hesiod
indicates whether a fixed number o f months corresponded to the
sun’s course. Again, w e do not know whether the Greeks origin­
ally used the haphazard intercalation o f the Babylonians (cf
p. 23). T he earliest method o f intercalation know n to Geminus
(8, 6) is very primitive, yet it is already rational: ‘The ancients
added the intercalated month every other year.’29 This parallels
the alternation o f full and hollow months. T w o lunisolar years
o f this kind contain 25 months, that is, c. 737 days as against the
73o|* days o f tw o solar years. Nevertheless, Greek cities (Herod.
II, 4; Censor. 18) and the Romans as well (sec p. 43) were satisfied
w ith this device. The Macedonians brought the same biennial
scheme into Egypt, and held to it in the age o f Eratosthenes and
Archimedes (p. 38).
After speaking o f the biennial cycle in Greece Geminus (8)
continues: ‘As the days and the months did not agree w ith the
moon, nor did the years keep pace w ith the sun, they sought for
a period which should, as regards the years, agree with the sun,
and, as regards the months and the days, w ith the m oon.’ In fact,
both the lunar months and the solar year are reducible to the
same time unit: the day. A given intercalary cycle attempts to
 THE CALENDAR 29

make the number o f days the same for the sun years and for the
lunar months within a given period o f time. The proportion
is easy to calculate: 365*25:29-30 = 1 :1 2 ; 2:25; 3:37; 8:99;
11:13 6 ; 19:235. As Gcminus tells us: ‘The first period they con­
structed was the octactcris (or eight-year cycle) which contains
2,922 days, 99 months (o f which the years 3, 5 and 8 are inter­
calary), and 8 years.’ Yet, as Geminus (8) again informs us, while
the eight years contain 2,922 days, 99 lunar months contain
2,9232 days. Thus, in 16 years, the octaeteris will be behind by
3 days in comparison with the moon. Accordingly, a new schema
was put forward: a 19-ycar cycle o f 235 months, including seven
embolismic months, and 6,940 days. The 19-ycar cycle was pro­
posed in 432 bc by the mathematician M eton, lampooned by
Aristophanes (Aves, 995). The scheme then was improved by
Callippus in 330 and by Hipparchus about 125 b c . Th e astrono­
mers used these cycles for their calculations (B. L. van der
Waerden, J H S i960, 169), and M cton’s cycle was o f great
practical importance for the construction o f popular almanacs
w hich offered weather forecasts. W hen Aratus (750) refers to
M eton, he says nothing about the calendar use o f M eton’s cycle,
but speaks o f the true message which the stars beam to men,
particularly to mariners, w ith regard to weather-changes. In this
sense, as Diodorus (XII, 36) says, to his own day a great number
o f the Greeks used M eton’s period (cf. Samuel II).
Influenced by Geminus’ report o f the progress o f cyclic systems,
and by the parallel account o f Censorinus, modem scholars for
a long time believed, and some o f them continue to believe, that
Greek cities docilely and steadily followed the rules o f inter­
calation which were put forward by astronomers. But Geminus,
w ho elsewhere speaks o f a ‘civil’ calendar, nowhere says that
8-ycar, 16-year and other such cycles were used by the cities.
T he simple fact that the Greeks often lengthened the year by
adding fractions o f a month, day or days, and sometimes shor­
tened the year in the same w ay (p. 31), excludes the idea that the
polis ever adopted any astronomical system o f intercalation. The
magistrates charged w ith bringing the lunar months into approxi­
mate correspondence w ith the seasons may have used the cycles
30 C H R O N O L O G Y OF THE A N C I E N T W O R L D

devised b y astronomers as standards by which the calendar

variations could bc adjusted.
As late as the middle o f the third century ad the rather primitive
octaeteris was normal for the Greeks, the Jews and the Church
(Africanus ap. Hieron. Ad Daniel. 9, 24 = P L X X V , 524; Eus.
H.E. VII, 20; M . Richard, Muséon 1974» 307). The Alexandrian
church c. 277 adopted the 19-year cycle. Accepted b y Rom e in
525, the latter has remained in force until today for calculation
o f Easter dates (cf. Ed. Schwartz, Z N T W 1906, 64). C f also A.
Strobcl, Ursprung und Geschichte des friihchristlichen Osterkalenders
( i 977).
W ith or w ithout astronomical advice, the magistrates o f
Greek cities, just as did their counterparts in Rom e (p. 45) or
Babylon (p. 22), ordered intercalations according to the need
o f the moment. In the third century b c , at Samos, a year had
four ‘embolismic’ months (Ch. Michel, Recueil d'inscr. grecques
(1899) no. 899). Censorinus, writing in 238, when the Julian
time-reckoning had already been accepted by the majority o f
Greek cities, explains the disarray o f pre-Julian lunisolar calendars
by the uncertainty concerning the actual duration o f the solar
year. In fact Hipparchus (c. 125 bc ) still had to oppose the opinion
o f those astronomers w ho believed that the length o f time in
which the sun passes from a solstice to the same solstice again is
exactly 365^ days (Ptol. Almag. Ill, 3). Hipparchus him self was
able to give the almost exact value o f the length o f the year
(3654 o f a w hole day) which is less than 7 minutes in
excess over the true mean year (cf T . Heath, Aristarchus o f Samos
(1913), 297). Y et he acknowledges the possibility o f error in the
observations, which according to him could amount up to
I day for the time o f a solstice and up to 6 hours for the time o f
an equinox. Thus too Ptolemacus, w h o quotes Hipparchus,
was not so sure o f ascertaining the length o f the solar year (Ptol.
Almag. Ill, 1, 1), and the astrologer Vettius Valens (IX, 11) c.
a d 15$ (cf O . Neugebauer, H T R 1954, 65) still quoted several
values exceeding 365*^ days (cf O . Neugebauer, Rivista degli studi
orientali 1949, 92).
Igitur cum tanta inter viros doctissimos fuerit dissensio, quid mirum
 THE CALENDAR 31
si am i civiles, quos diversae civitates rudes etiam turn sibi quaequae
statuebant, tam inter sc discrepent quant cum illo naturali non con-
truant (Censor. 19, 4). In consequence, as Censorinus says, the
relationship between what were in principle the same months
o f different cities was disturbed by haphazard intercalations and
by the renaming o f months which are often attested (e.g., at
Argos: Thuc. V , 54; Xen. Hell. IV, 7, 2; V , 1, 29; at Sparta:
Plut. Agis, 16; in Macedonia: Plut. Alex. 16). T he absence o f a
fixed calendar is also evidenced by the contract clause: ‘i f a
month should bc intercalated', eg., IG R R IV, 949 (Chios),
A B S A X X II (1916-18) 196 (Mylasa) (cf. also W . K . Pritchett,
CPh 1947, 235; B C H 1957, 277). The Thessalian month o f
Thyos at one time coincided with the Delphic Enduspottropios
(G D I, 17200), at another time w ith the Delphic Bysios (Fouilles de
Delphes, G. Colin, Inscr. du trésor des Athéniens, no. 213). In a
document (forged in the Hellenistic period) in Dem. XVIII, 157,
the Macedonian month o f Loos is equated w ith the Athenian
Bocdromion, though, in principle, it corresponded to the
Hckatombaion.
O n the other hand, as Censorinus states, the calendar often
did not keep pace with the natural year. T w elve lunations are
longer than twelve Greek months (which comprised 6 x 30 +
6 x 29 = 354 days) by 0-36707 days, so that, in order to have the
lunar months in agreement w ith the moon's phases, it was
necessary to insert three days every eight years. This in turn
disturbed the agreement o f the calendar w ith the sun.
According to Cicero ‘it is the custom o f the Sicilians and all
Greeks, since they wish their days and months to agree with the
movements o f the sun and moon, to remove an occasional dis­
crepancy by shortening a month by one day or at most tw o
days . . . they also sometimes lengthen a month by one day or
tw o’. Est consuetudo Siculorutn ceterorumque Graecorum, quod suos
dies mensesque congruere volunt cum solis lunaeque ratione, ut non
numquam, si quid discrepet, eximant unum aliquem diem aut summum
biduum ex mense . . . item non numquam uno die longiorem mensem
faciunt aut biduo (Cic. Verr. II, 2, 129).
Th e resulting confusion can be illustrated by some statements
32 C H R O N O L O G Y OF THE A N C I E N T W O R L D

o f Greek authors. Aristoxenus, a disciple o f Aristotle, in order

to explain the disagreement o f theoreticians concerning the
musical scales, compares it to the state o f Greek calendars: ‘The
tenth day o f the month for the Corinthians is the fifth for the
Athenians, and the eighth somewhere else’ (Elem. harm. II, 37).
Three centuries later, Diodorus (I, 50) explains to his readers
that the Thebans in Egypt do not intercalate months or suppress
days in the year as most o f the Greeks do. T w o centuries after­
wards, Plutarch (Arist. 19) observes that the beginning and the
end o f months in various Greek cities did not coincide. A wit
could say that in Abdera, the proverbial city o f fools, every one
had his ow n crier proclaiming a new moon for his master alone
(Athen. VIII, 41, p. 349 b., cf. Corpus Paraemiogr. Grace. I, App. 2,
no. 61, and Crates, ap. Athen. Ill, 117 b (on Ceos)).
The actual sequence o f the calendar in different Greek cities
remains unknown. Authors, naturally, mention only exceptional
facts (e.g. Alexander set back the calendar by one day: Plut. Alex.
25), and the double dates come down to us by chance and in a
haphazard manner. Th e Boeotian month o f Panamos in principle
corresponded to the Athenian Metageitnion. The battle o f
Plataea (479 bc ) took place on 27 Panamos according to the
Boeotian calendar, but on 4 Boedromion according to the
Athenian calendar; at that time the beginning o f the Athenian
month came seven days later than the Boeotian (Plut. Arist. 19;
Camill. 19; cf. Μ . P. Nilsson, De Dionysiis Aiticis, Diss. Lund
1900, 7). In 423 bc , 14 Elaphebolion, in Athens, corresponded to
12 Gcrastios in Sparta; in 421 bc 25 Elaphebolion corresponded
to 27 Artamitios, which preceded Gerastios (Thuc. IV, 119 ;
V , 19). A Spartan month could be nine days behind the moon
(Herod. VI, 106; cf Pritchett, B C H 1957, 278). It happened,
rarely, that tw o cities agreed to begin the months on the same
day (Knossos and Tylissus c. 450; T o d I, no. 33). The confusion
remained the same in the Hellenistic period. In the collection o f
letters ascribed to Themistocles (Ep. 7, 1) it is said that the last
day o f the Athenian Boedromion ‘is the same day’ as 10 Panemos
in Corinth; there is, therefore, a difference o f ten days. In the
second century b c , in Tanagra, the first day o f the month o f
 THE CALENDAR 33

Thiouios, was, in the lunar calendar, κατά δε τον θεόν, the

eleventh day o f the follow ing month, Homoloios (IG VII, 517).
A law o f Stymphalia dating from the third century bc sets a
final possible date for a trial ‘until the tenth day (o f the month)
according to the moon (κατά σελ [άναι/])* (IG V , 2, 357). Each o f
the cities o f the Eubocan league, around 290 b c , had its own
calendar (e.g. : μηνος Ληναίώνος ώς Χαλκιδεΐς αγουσι). The
League decided that the months should be o f equal length, but
at the same time it allowed each city to add as many as three
days (IG XII, 9, 207, Suppl., 178). In the Cretan confederacy,
the 20th at Knossos once corresponded to the 4th in Gortyna
(IG XII, 3, 254). O n the other hand, the months ran parallel at
Knossos, Latona and Olus in 116 bc (Syll. 712). The same was
true for Ephesus and Smyrna c. 100 bc (O G IS 438, 90). The
calendar difference between Miletus and Magnesia in 196 bc
was o f one day only (Syll. 588). The confusion o f the Greek
calendar appears strange to us; but a calendar is a conventional
device just like weights and measures. Each polis had its own mode
o f time reckoning as it had its ow n month names and numerals.
For instance, the Athenians preserved the acrophonic notation
(where Δ was 10) until c. 100 bc (cf Μ . N . Tod, A B S A 1950,
126). There was no more reason for an Athenian to get worried
about the disagreement of his calendar with that, say, o f Sparta,
than for a Frenchman to be preoccupied w ith the fact that the
clocks everywhere in France indicate the hour o f Paris, which
(since 1911) has been Greenwich Mean Tim e: so that, for instance,
in Besançon on 1 N ovem ber the legal time is 40 minutes behind
the sun (cf. P. Couderc, Le Calendrier (1961), 125).
Moses (Gen. 1, 14) and Plato (Timaeus 38 c) were in agree­
ment that God placed the luminaries in the firmament as measures
o f time. Accordingly, as Geminus (8) says, the principle that
the sacrifices should be offered after the manner o f the forefathers
was understood by all Greeks as meaning that ‘they should keep
the years in agreement w ith the sun, and the days and months
w ith the m oon’. In this w ay, the same sacrifices will be offered
from year to year in the same season when they fall due. Th e
Greeks (Plato, Leg. VII, 809 d) and Jews agreed on this point.
34 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

THE ATHENIAN CALENDAR

Th e functioning o f the Athenian calendar in the Classical age is

rather better know n to us than the time-measurements in the
other Greek cities.30 Firstly, the documentary evidence is more
abundant. Secondly, the list o f archons makes it possible to
determine the Julian year o f documents. Thirdly, the Athenians
used tw o official dating systems simultaneously: the civil lunisolar
calendar and the schematic Prytany reckoning. The Prytanis was
the w orking committee o f the Council (Boule) which governed
Athens for a certain fraction o f the year. T h e Council ot Five
Hundred is elected by lot, fifty from each tribe. T he members
from each tribe function as the prytanis in turn, the order being
determined by lot. T he first four serve for thirty-six days each,
the last six for thirty-five. [This makes 354 days for ten prytanies.]
For they reckon the yearly period [o f the Council] according to
the m oon.’ This statement o f Aristotle (Ath. Pol. 43. 2) is valid for
his time and for the period after c. 408 (Mener, 215: cj. W . K.
Pritchett, B C H 1964, 473). The length o f service o f the Council
before this date is uncertain. From an accounting record IG I,
324 = T o d I, 64) it was inferred that in 426 5-423 2 tour prytany
years amounted to 1,464 days (cf. IG I-, 155: Ginzel II. So), but
some figures are not preserved on the stone, and tire restorations
are doubtful.31
The Boule probably took office, together with the archon. on
i Hekatombaion, but in 411 its mandate ended on 13 Skirophorion
(Arist. Ath. Pol. 32), that is, some fifteen days before the term o f
the archon’s year 412/411.
After Aristotle's time, from 307/6 to 224 3 there were twelve
tribes. During this period prytanies and the months ot the civil
year probably run parallel (cf. Pollux VIII. 115: Mérite. 135).
There were thirteen tribes from 223/2 to 202 1, eleven in 201 1,
and again twelve from 200 bc until the time ot Hadrian.
In Classical Athens the count o f prytanies served as the working
calendar o f the government. For instance, the armistice ot 423
was accepted by the popular assembly when the tribe ot Acmanris
 TH E CALENDAR 35

held the prytany (Thuc. IV, 119). The revalidation o f the extant
laws had to bc voted by the popular assembly on the n t h day
o f the first prytany (Dem. 24, 25), and so on. Th e financial records
o f the government (Arist. Ath. Pol. 47), including mining leases
(M. Crosby, Hesp. 1950, 192), were reckoned on the same
time-standard. A public debtor who had not paid by the ninth
prytany lost his civic rights (Dem. 24, 87). T he civil calendar, the
twelve months o f the archon’s year, was used for general indica­
tions o f time. Thus, for instance, a marriage w'as concluded when
Polyzelus was archon (367 bc) in Skirophorion, and the divorce
document written when Timocratcs was archon (364 bc ) in
Poseidcon (Dem. 30, 15).32
The extant evidence shows that the Athenians did not use
M eton’s cycle or some other regular system o f intercalation for
adjusting the official calendar,33 though as Petavius supposed
(Idelcr I, 318), the cyclic calculation might help the magistrates in
adjusting the calendar to the course o f the sun (cf. p. 30). In
Athens, as in Sicily (p. 31), months were added as needed. For
instance, c. 420, the people decreed that the archon o f the coming
year should intercalate the month o f Hekatombaion (IG I, 76).
As late as the second century b c the intercalation was handled so
haphazardly that tw o successive years could have extra months
(Margaret Thompson, The New Style Silver Coinage o f Athens
(1961), 612). The Athenians may have adopted the principle
o f alternating full and hollow months (p. 28, and cf Meritt, 84).
In practice, days could be suppressed (cf W . K . Pritchett, B C H
1964, 460, 473) or inserted (W . K . Pritchett, B C H 1957, 276) at
will. The essential reason for such adjustment was that the dates
o f most religious acts w ere fixed in the official calendar. The
temple o f Dionysus in Limnae could be opened only once in a
year, namely on 12 Anthesterion (Dem. 30, 15), and so on.
Th e fasti, first published by Solon (Plut. Solon, 25; Nilsson,
Kalender, 68), were inscribed on stones. Thus, everyone could
read that, for instance, the sacrifice to the Kourotrophos which
was to be offered by the deme o f Erchia had to be offered on 3
Skirophorion.34
It w ould have been an offence against the gods i f these fixed
30 C H R O N O L O G Y OF THE A N C I E N T W O R L D

dates were disregarded. But it was possible to change the position

o f the given fixed date in respect to the movement o f the heavenly
bodies. For instance, the theatrical representation at the Great
Dionysia was to be held on ic Elaphebolion (cf. L. Dcubner,
Attische Feste (1932), 142). In 270, for some reason, the perform­
ance was to be postponed. Accordingly, the four days following
9 Elaphebolion were counted as the second, third and fourth
inserted 9 Elaphebolion. ’ Ελαφηβολιώνο[ς\ [cjvarct ίοταμένου
τ€τάρτ€ί έμβολίμωι (W . B. Dinsmoor, Hesp. 1954, 299). Again, the
Athenians could rename the month M ounichion first Anthesterion
and then Boedromion, to allow Demetrius Poliorcctes to be
initiated in the lesser mysteries ofEieusis (celebrated in Anthesterion)
and in the greater (celebrated in Boedromion) during his short stay
in their city (Plut. Demetr. 26). On the other hand, as the popular
assembly did not meet on feasts and unlucky days (Busolt-Swoboda,
II, 988), this tampering w ith the calendar could also play into
the hands o f the politicians. I f the gods, as Aristophanes supposed,
lived according to the true time, the Athenian calendar often
made them ‘g o to bed without their supper’ (.Nubes, 618).
Therefore, the length o f a given civil year or month must
be established empirically, and the proposed schemes o f the
Athenian civil year can only bc tentative. T he prytany year some­
times can help in this respect. Athenian documents often bear
double dates: the date o f the civil and o f the prytany calendar.
Th e prytanies were numbered from 394, and the day within the
prytany was stated from 346 on. For instance, a decree gives the
equation: 23rd day o f the IX Prytany = 11 Thargelion (Syll. 287).
In this w a y w e can fmd out whether the given civil year was
intercalated. (The length o f each prytany in the intercalated year
was extended.) For instance, in 333/2 the 29th day o f the I
Prytany fell on the 9th day o f the second month (Metageitnion) o f
the civil year. Thus, the first prytany had a length o f 39 days (at
least), and the year 333/2 was intercalary (IG Π, 338). O n the
other hand, in 332/1, the 19th Elaphebolion, that is to say, the
255th day o f the standard civil year, corresponded to the 7th day
o f the V in Prytany (IG II, 345). Thus, the Prytanies I—VII con­
tained 248 days, and the w hole year included (248:7) x 10 = 354
 THE CALENDAR 37
(or 355) days. In other words, the year 332/1 was a common
year. O f course one o f the double dates m ay be lost, or even not
recorded at all on stone [cf., e.g., IG II, 337 o f 333/2). Th ough in
the fourth century the prytany year and the archon’s year were
coterminous, w e do not know the Julian dates for 1 Hekatom-
baion. Th e hypothesis, deduced from Plato, Leg. VI, 767 and
Arist. H. Anitn. 543 b (Ginzel II, 380; Samuel, 64), and repeated
b y recent scholars, that the beginning o f the year coincided with
the summer solstice m oon remains unproven and improbable. It
postulates that the Athenians tried to balance the number o f
inserted and suppressed days in every official year. T h ey were
probably more casual. The parapegma (p. 58) made it easy to
overlook the vagaries o f the official time-measurement. The
equation o f an Athenian w ith a Julian date is possible only in
exceptional cases (cf. W . K . Pritchett, CPh. 1947, 235). This is
true even for astronomical dates (e.g. Ptol. Alrnag. IV, 11: the
eclipse o f 23 December 383 occurred when Phanostratus was
archon, in the month o f Poseideon), since the astronomers used the
Athenian calendar names for their theoretical calendar o f mean
lunations35 (cf. B. L. van der Waerdcn, Museum Helveticum 1958,
106). As to the pry tallies, their lengths may have been tampered
w ith, but no evidence o f this practice has yet been found. The,
at least relative, stability o f the prytany calendar m ay depend on
its use in financial records. The use o f a schematic year o f 12 x 30
days months in business (A. Mommsen, Chronologie (1883), 48;
Sontheimer, RE X V I, 16) again palliated the inconvenience o f
the civil year.
T he most important corrective was furnished by the direct
observation o f the moon. Neglecting the official reckoning, the
man in the street started to count the days o f the new month at
sighting o f the new crescent. As Aristophanes (Nubes, 626) sen-
tentiously advises Athenian politicians to regulate days o f their
life according to the moon: κατά aeXrjwjv ώς ayew χρη τοΰ βίου τάς
ημέρας, from the beginning o f the second century the official
dating included the reference to the true course o f the moon. The
date within the year was given ‘according to the archon’ and
‘according to the deity’, i.e. Selene, the M oon (cf, e.g., IG
38 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

II, 967: *Ελαφηβολιώνος ivarei μετ' είκάδας κατ' άρχοντα κατά

Otov 6e Μουνιχιωνος ScoSe/caric). Hence, in this year o f Achaios’
archonship, when the official calendar recorded the date as 19
Elaphebolion, the m oon was already in the 12th day o f the
follow ing lunation (Mounichion). But in the year o f Euergetcs,
19 Elaphebolion o f the archon was only tw o days behind the
moon (cf. B. D . Meritt, Hesp. 1957, 73; Pritchett-Neugcbauer,
15; Pritchett, 330; J. Pouilloux, R E A 1964, 211).

THE MACEDONIAN CALENDAR IN EGYPT

Alexander the Great brought the Macedonian lunisolar calendar
to Egypt, and the Ptolemies held to it for a long time. Th e
months had 29 and 30 days, the days o f the month were numbered
successively, and the ‘29’ was omitted in a hollow month
(P. Cornelly 1) so that the last day o f the month was always
counted as ‘30*. A month was intercalated from time to time
(Plut. Alex, 16; FrGrH 257 a, 3\ cf. P. Oxyr. XVII, 2082). H ow
the calendar was handled outside Egypt remains unknown. For
the Scleucids (cf. p. 25) Alexander the Great died on 29 Airu o f
the Babylonians, that is, in the evening o f 10 June 323 (A. E.
Samuel, Ptolemaic Chronology (1962), 47). According to Alexander’s
Ephemerids the king died on the last day o f the month Daisios
(Plut. Alex. 75-76). Thus, at this time, the Macedonian calendar
agreed w ith the moon. (O r did Alexander, as later Scleucus I,
use the Babylonian cyclic system?) Th e date o f Alexander’s death
in Pseudo-Callisthenes (Historia Alexandri Magtii 146, ed. G.
Kroll: 13 Pharmuthi = i3 June) is erroneous.
T he Greek documents o f Ptolemaic Egypt offer numerous
equations between the Macedonian and the Egyptian dates. The
latter are easily converted into Julian dates (sec p. 40). T he evid­
ence shows that until c. 240 the Macedonian months agreed with
the moon. It appears that the calendar was regulated by the
Egyptian 25-year cycle. As in the old Macedonian calendar, an
intercalary month was inserted every other year (cf. p. 28),
though the cycle required only nine intercalations (1,309
lunar m onths= 2 5 Egyptian ycars= 9 ,i2 5 days). But because the
calendar was regulated by the solar year, it did not become con­
 THE CALENDAR 39

fused by superfluous intercalations. O n ly the order o f the names

o f the months in the solar year was affected.
T he Macedonian calendar survived in Egypt chiefly for cult
purposes. Even the feasts o f the Egyptian gods were set up by the
Alexandrian court according to the Macedonian calendar. The
relationship o f months to seasons, however, was affected by the
adjustment to the solar year and b y the extra intercalations. The
ist o f D ios varied in position from 25 August in the beginning,
to 15 January at the end o f the reign o f Ptolem y II. O n the other
hand the Egyptian calendar, which was simpler, was more con­
venient for everyday affairs (see a list o f Egyptian months on a
stone in Samos, a Ptolemaic possession: L. Robert, Etudes épi­
graphiques (193^), t i 8 and cf. P. Roussel, Les cultes égyptiens à Délos
(1916), 204). Already in the middle o f the third century b c , the
Greeks in E gypt calculated according to the Egyptian calendar;
so that, for example, a Greek in 257 bc asked on what Egyptian
date o f that year the birthday o f the king would fall— the king’s
birthday was fixed according to the Macedonian calendar (P.
Cair. Zen. IV, 59541). Th e Macedonian calendar was used for all
official acts. So in M orocco today, the solar calendar is used in
business life, while officially, the dating system runs according to
the Islamic lunar year (cf. E. Westermarck, Ritual and Belief in
Morocco II (1926), 150). But the Egyptian year, in turn, being
mobile, the Greeks consulted the stellar calendar (see p. 54)
arranged according to the course o f the Egyptian year. In this
w ay, the festivals could be celebrated year after year at the same
time o f the true solar year (P. Hibeh, 27; P. Paris, 1; F. Blass,
Ars Eudoxi, 1887).
Under the rule o f Ptolemy III the synchronization o f the
Macedonian calendar with the moon was neglected. For example,
i Gorpiaios in 232 bc fell five days after the full moon. A t the end
o f the third century b c (cf. P. Tebt. Ill, 820) the Macedonian
calendar was adjusted to agree with the Egyptian, so that the
names o f the Macedonian months were only different denomina­
tions for the Egyptian months. A t first the equation was Dystros
= Thot, and so on. Ptolem y VII Philometor then re-established
the Macedonian calendar in 163 BC (U. W ilcken, Urkunden
40 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

Ptolemderzeit I (1927), 496), but his action was repealed after

his death in 145 b c . A new equation o f Macedonian and Egyptian
months (this time Dios = Thot) came into use later (the first
evidence is in 199 b c ), although the preceding system existed
until the end o f the second century a d . As in the area ruled by
the Seleucids, in Egypt, too, the Greek calendar was replaced by
the local calendar. H owever, while in the case o f the Seleucids
the lunisolar calendar was merely corrected, the Ptolemies in effect
com pletely abolished the lunisolar reckoning o f time.36

THE EGYPTIAN YEAR

‘Th e Egyptians, alone and always, had a year o f definite length.

Other peoples varied it b y different but equally erroneous
reckonings’ (Macrob. Sat. I, 12, 2). T he Egyptians had ami
certus modus because their year was composed o f days only.
These 365 days were schematically grouped into four seasons,
and twelve months o f thirty days plus five supplementary
days (έπαγόμεναή.*7 Th e days within the month were counted
successively. T h e months were counted, from the first to the
fourth, within each o f the three agricultural seasons: ‘Inundation’
(when the N ile overflowed the fields), ‘Going out’ (from the Nile
waters; time o f agricultural work) and ‘Deficiency’ (the season
o f low water). In later popular usage, the Egyptian months were
named after festivals. W e transcribe these names (inherited by the
Copts) according to their Greek form : Thot, Phaophi, Athyr,
Choiak, T yb i, Mecheir, Phanemoth, Pharmuthi, Pachon, Payni,
Epeiph, Mesore (plus five cpagomenal days). C f T . G. H. James,
The Hekanakhte Papers (1962), 3.
T he resulting year, which was \ day shorter than the actual
solar year, could have been corrected b y means o f intercalation,
but this was not done in Egypt. Therefore every four years the
beginning o f the year (1st o f Thot) was delayed by one day
in respect to the solar year. Every month— in the course o f
the cycle o f 1,461 Egyptian years (=1,460 Julian)— thus rotated
through all seasons o f the solar year. ( C f Sethe, G G N 1920, 30;
R. A . Parker, Revue d'Egyptologie 1957, 85.) It should be noted,
 THE CALENDAR 41

however, that the length o f a Sirius cycle is somewhat variable

(1cf Μ. Γ. Ingham, J E A 1969, 36).
The decree o f Canopus (OG/S, 56) says: ‘It came about that
the festivals which were celebrated in winter fell in the summer,
and that those celebrated in summer were instead in the w inter.’
Th e priests, however, blocked the reform proposed in 238 bc
b y Ptolem y III to correct the annas vagus.
In effect, alongside the official year, there was the popular lunar
calendar o f alternating months o f 29 and 30 days which is attested
from c. 1900 on. It was basic in everyday life and used for cult
purposes (cf D . Bonneau, Revue d'Egyptologie 1971, 57)· A t some
time (before 235 bc ), the Egyptians devised a 25-year cycle o f 309
months which indicated the dates o f the civil calendar on which
the lunar months w ere to begin (cf. R. A. Parker, J N E S 1957, 39;
1970, 217; Neugebauer, 90). The Egyptian lunar month began in
the early morning (cf R. A . Parker, JN E S 1970, 217).
T he Egyptian mobile year was independent o f both sun and
moon, but, as often among primitive peoples (Μ. P. Nilsson,
Acta Orientalia 1941, 1 =Opusc. Selecta II, 54), it was related to a
fixed star. Th e sighting o f Sirius in the east horizon at sunrise,
on 19 July, after the star had been invisible for about 70 days, fell
close to the beginning o f the flood o f the Nile, and thus to the
Egyptian N e w Year, the first day o f the first month o f the
Inundation season, that is, 1 Thot. The Egyptians praised the star
as ‘ the Bringer o f the N ile’ and ‘the Renewer o f the Year’ (Plut.
De Isid. 38; cf Th. Hopfncr, Plutarch iiber Isis and Osiris II (1941),
174). The sighting o f the star was officially announced: ‘Sirius
rises on 16 Pharmuthi’, and so on (see p. 83). T he Sirius year,
however, corresponds to the solar year (which is by c. 12 minutes
shorter). Thus, the rising o f Sirius fell on 1 T h ot once every 1,460
Julian years. Censorinus tells us that it happened on 21 July (in
fact 20 July), ad 139.
In agreement w ith his contemporaries— coins were issued at
Alexandria representing the fabulous bird, the phoenix, the
symbol o f renewal, w ith the legend A IO N (]. V ogt, Die Alexan-
drinischen Miinzen (1924), 115)— Censorinus spoke o f the return
o f annus canicularis (Sirius being also called Canis maior). From his
42 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

words modern scholars inferred, without any warrant, that the

Egyptian annus vagus must have started on the day when the
heliac rising o f Sirius fell on i T ’iot. From Fréret (1758) on, they
hesitated between 1322 and 2782 as the starting years o f the
Egyptian calendar (Idclcr I, 126). Ed. M eyer went back to 4241.
B ut the dispute is futile. A calendar is a tool which cannot be
justified b y either logic or astronomy. T he Egyptian calendar
took account o f the N ile and not o f Sirius*8 (O . Ncugcbaucr,
Acta Orientalia 1938, 169; JN E S 1942, 396). Furthermore, there
is no inherent necessity to start a new calendar on its first day.
England changed from the Julian calendar and the year beginning
on 25 March, to the Gregorian style and the adoption o f 1 January
as the N e w Year, on 2 September 1752.
In fact originally the Egyptians, together w ith many primitive
peoples, did not count by years, but by agricultural seasons
(Diod. I, 26, 5). A ll conjectures about the date o f the introduction
o f the annus vagus are premature. W e can only state that there is
evidence o f the use o f the variable year from the V Dynasty on,
that the rising o f Sirius was observed as early as c. 1900, and
that the celebration o f this event was, from the Middle Kingdom,
a changeable date in the civil year.
Th e ‘day* and the (lunar) ‘month’, as their hieroglyphic signs
show, w ere related to the sun and the moon respectively (K.
Sethe, Vom Bilde zutn Buchstaben (1939), 23). T he Egyptian word
for ‘year’ does not have any astronomical connotation; it means
‘renewal’ and each year was a beginning (E. Edel, Altaegyptische
Grammatik (1955), 179; id .J N E S 1949, 35; cf. Gardiner, 70).
Th e variable year probably was introduced for administrative
purposes. T he Egyptians had the schematic financial year o f
12 x 30 = 360 days. T he atinus vagus originated when the financial
year, by the addition of five epagomenal days, was equated with
the mean agricultural year, which in Egypt happened to have
the same length as the solar year: the regularity o f the flood o f
the N ile was conditioned by the snow thaw in mountains o f
Ethiopia (cf D . Bonneau, La crue du N il (1964), 29).
T he divergence o f the Egyptian year from the course o f the
sun is almost imperceptible in one lifetime: the difference in forty
 THE CALENDAR 43

years amounts only to ten days. For Herodotus (II, 4) the Egyptian
year agreed with die cycle o f the seasons. Th e advantages o f the
Egyptian calendar— its simplicity and regularity— are so obvious
that astronomers, from Hellenistic times to Copernicus, used it.
For the same practical reason, the schematic year o f 12 x 30 + 5
days, probably based on the Babylonian business year o f 12 x 30
days (cf. O . Neugebauer, J N E S 1942, 400), became the official
system o f time-reckoning in Persia under the Sassanids as well as
in Armenia and Cappadocia.39 (According to Arabian astronomers
the Sassanian year was adjusted to the succession o f seasons by
the intercalation o f one month every 120 years.) C f A. Christen­
sen, L'Iran des Sassanides (1944), 168; E. S. Kennedy and B. L. van
der W aerden, J A O S 1963, 315; E. J. Bickerman, ArchOr 1967,
197; id. in Cambridge History o f Iran III.

THE ROMAN CALENDAR

T he Roman calendar, at the time o f Caesar, consisted o f 12

months; 4 o f 31 days (Martius, Maius, Q uintilis= July, October),
7 o f 29 (Ianuarius, Aprilis, Iunius, Sextilis = August, September,
N ovem ber, December) and 1 o f 28 (Februarius), totalling 355
days in a year.40
Every other year (in the even years bc ) 22 or 23 days were
intercalated. T h e intercalation took place in February, after the
feast o f Terminalia (23 February) ; while the 5 remaining days o f
February were added at the end o f the intercalary month (Inter-
calaris), so that this month consisted o f 27 or 28 days.
Th e first day o f the month was called Kalendae, the 5 th (or the
7th in a month o f 31 days) Nonae, the 13 th (or the 15 th o f the
months w hich contained 31 days) Idus. Counting backwards
from these established dates, one calculated the days o f the
month. T he calculation was inclusive; that is to say, the day
from which one counted and the day to be designated were both
included. Thus 2 January was: ante diem I V Non. Jan.; 2 M arch:
ante diem V I Non. Mart. The day before the last day to bc counted
was called pridie (Table IV , p. 125). Th e last days o f February,
after the Ides, were counted in the intercalary year back from the
beginning o f the inserted month: ante (diem) V Kal. Intercaiaris
44 C H R O N O L O G Y OF THE A N C I E N T W O R L D

= 20 February (Cic. Pro Quinct. 79). Sometimes as late as

14 February it was unknown whether the pontifices w ould inter­
calate a month: in this case 14 February became ante diem X
Terminalia (Dessau, 6302 = Dcgrassi, 719).
Counting the days in succession, although possible, appears
rarely (A. Gagner, in Festskrift Per Perssoti (1922), 202).41
N o account o f the moon was taken in this system; on the
contrary, the biennial insertion o f 22 (23) days must have destroyed
all agreement w ith the lunations. Y et the days within the month
were numbered from the com ing moon phases backwards. A
pontifex announced the new crescent (p. 17) and according to
its form and position told h o w many days were to be counted
until the Nones, that is, the first quarter. A t the Nones it was
again proclaimed h ow m any days there were until the Ides (the
full moon), and on w hich days the festivals were to be celebrated
(Ginzel II, 173).
T he Roman calendar, however, w ith its m any peculiar features
(the length o f the months, the intercalation system) was rather a
conscious effort at ‘synchronizing the civil and solar years’
(Censor. 20, 6). Y et, the Roman quadrennial cycle consisted o f
355 + 3 7 8 + 3 5 5 + 377 days. It looks as i f its author tried to adjust
the lunar year to the path o f the sun, that is, to the agrarian year.
(The months o f 29, 30, and 31 days also occur in Greek meteoro­
logical calendars.) But the Roman 4-year cycle amounted to
1,465 days, that is, it was four days longer than four solar years.
Thus the calendar was behind four days every quadrennium in
respect to the seasons. In fact, it was not easy to establish the
true length o f the solar year (p. 30). Herodotus (I, 32) erred in
this respect; the great engineer Harpalus (c. 480) believed that the
revolution o f the sun takes 365 days and 13 hours and, as late as
c. 190, Ennius spoke o f 366 days o f the solar year. W e cannot be
surprised at the mistakes made by the Roman peasants c. 500:
‘Y our knowledge, Romulus, o f weapons was better than o f
stars’ (Scilicet arma tnagis quam sidera, Romule, noras, O vid , Fasti
I, 28).
T he Rom an pseudo-solar cycle was probably a modification
o f the ‘year o f Romulus’, that is, o f the purely agricultural ten-
 THE CALENDAR 45

month year which ran from March to December, the ‘tenth’

month. Prim itive peoples often take into account only the period
o f agricultural activity and neglect the rest o f the natural year.
T he whole annual period, from one spring to the next, is divided
into fractions o f irregular length. Such ‘months’, up to 39 days
each, are attested for the Romulean year (Plut. Nurna 18; Lydus,
De metis. I, 16) and for ancient Italy generally (Censor. 22, 6;
19, 6). Th e month names from Martius to Junius seem to refer
to the stages o f grow th o f crops and cattle (cf. J. G. Frazer, The
Fasti o f Ovid II (1929), 8; J. Bayet, Histoire de la religion romaine
(1957), 89). The ancients usually attributed the introduction o f
this calendar to Num a (and its defects to later changes. Cf. Cic.
De leg. II, 12, 29). M odern authors mostly attribute the system to
the Decemviri (mid-fifth century bc ). But the latter formulated,
rather, an intercalation law (Macrob. I, 13, 21). O n the other
hand, the calendar presupposes the Capitoline cult: the Kalends
are dedicated to Juno, and the Ides to Jupiter.42
The divergence o f this calendar from the sun’s course was so
patent that c. 450 the Decemviri already tried to correct the
system. There was also an intercalation law o f M \ Acilius Glabrio,
in 191 bc . But these reforms did not help. M any other projects
for adjusting the calendar to the solar year were made (Macrob.
I , 13; Liv. I, 19; cf. Idelci II, 69; Ginzel II, 253) but apparently
never accepted by the Romans (Mommsen, 44). Rather, at some
unknown time, they abandoned the rule o f schematic intercalation
and, just like Athens and other Greek cities (sec p. 31), practised
intercalation according to their needs. From the Second Punic
W ar to Caesar’s reform in 45 bc , the pontifees adjusted the
calendar at will. As in Greece the standard was, or was to be, that
the same sacrifices should be performed at the same seasons
(Quod ad tempus ut sacrificiorum libamenta serventurfetusquepecum . . .
diligenter habenda ratio intercalandi est, Cic. D e leg. II, 121, 29). In
fact intercalation became a tool o f politicians in their struggles
for power, and it was often handled arbitrarily and without
regard to the seasons.
Th e result o f such arbitrary intercalation was that the formula
o f the contracts in Cato (De agr. 150) contained the clause si
46 C H R O N O L O G Y OF THE A N C I E N T W O R L D

intercalation erit. In 50 b c Cicero, on 13 February, still did not know

whether or not there w ould bc an intercalation on the 23 rd
(Ad Att. V , 21, 14), and in 70 bc he had explained to his listeners as
a peculiarity o f the Greeks their preoccupation with making their
calendar correspond to the sun (Verr. II, 2, 129). In fact the Roman
calendar did not correspond to either the sun or the moon, but
4ging vielmehr ganzlich ins Wilde’ (Mommsen).
It follows that all the attempts to establish fixed intercalary
cycles for this calendar are in vain (cf. Ideler, Lehrbuch, 309 and
Mommsen, 44). The incidental documentation in our hands, as
already mentioned, allows us only to draw the general conclusion
that the Roman calendar, from the beginning o f the First Punic
W ar (264 bc ) until the beginning o f the Second (218 bc ), corre­
sponded more or less to the Julian calendar, perhaps running
behind the latter by a few weeks; that during the Hannibalic
W ar intercalation was neglected so that in 190 bc the Roman
calendar was ahead by 117 days; and that this difference had
declined to 72 days in 168 bc , so that in the intervening 22 years
the calendar must have been intercalated 12 times. It can bc sup­
posed that the calendar at the time o f the Gracchi was almost in
correspondence w ith the seasons, as is shown by the dates for
military campaigns during the period approximately from 140
to 7 0 b c . In Caesar’s time, however, intercalation was again
abandoned : in 46 bc there was a lag o f 90 days.
T he above summary o f the use o f the Roman calendar in the
second century b c follows G. De Sanctis.43 W e must emphasize
that our information is insufficient for generalizations. There are
only tw o astronomical equations: the solar eclipse o f 14 March,
190 was sighted in Rom e on 11 July (Roman) (Liv. X X X V II,
4, 4), and the lunar eclipse o f 21 June, 168 was seen on 4 Septem­
ber (Roman) (Liv. X LIV , 37, 8). O n the other hand, under the
terms o f some contracts o f the same period (Cato, De agr. 146)
the grain and the olives arc harvested in the due months, at the
end o f M ay and in November respectively. Th e contracts
probably indicated the ‘ideal’ month, which was independent o f
the vagaries o f the official calendar (cf p. 23). But the battle o f
the Cam pi Raudii, near Vercelli, on 30 (Roman) July, 101 bc
 THE CALENDAR 47
actually was fought in midsummer (Plut. Marius 26). T he
numerous dates which have com e down to us from the time o f
Caesar cannot be converted into Julian dates w ith certainty (cf.
Ginzcl II, 273; J. Carcopino, César (1936), 696). See n ow J.
Beaujcu, in Mélanges offerts à J. Heurgon (1976), 13.

THE JULIAN YEAR

Caesar did not reform the Roman calendar, but abandoned it
and instituted the solar calendar o f 365J days which was stable
and agreed w ith the seasons. He could well have said, with
reference to Greek calendar schemes: ‘the Julian year shall not
be outdone b y the calendar o f Eudoxus' (Nec meus Eudoxi vincetur
fastibus annus, Lucan X , 187). First it was necessary to insert
90 days in 46 b c in order to bring the months back to their right
seasons (cf. A . Rehm, R E III A , c. 1153). From 1 January 45
(T. Rice Holmes, Roman Republic I (1923), 339) a com m on year
o f 365 days, and the months o f their present length were in use.
T he ten extra days over the former 355-day year were placed at
the end o f different months so that the usual dates o f feasts
remained undisturbed (cf. A. Rehm, in Epitymbion fur H. Swoboda
(1927), 225). For instance, the feast which was celebrated on
21 December was not moved, though the notation o f its day
changed from X Kal. Jan. to XII Kal. Jan. since the month now
had 31 and no longer 29 days.
Every fourth year an extra day was inserted after VI Kal. Mart.
(= 2 4 February), and this added day was called bis sextum Kal.
Mart. In late imperial times, the intercalary year was accordingly
called annus bissextus, from which comes our ‘bissextile'.
After Caesar’s death, the pontifices erroneously inserted the
extra day every three years, so that Augustus in 9 bc had to
omit the intercalation for 16 years. O n ly from a d 8 on did the
Julian calendar function w ith regularity (Macrob. Sat. I, 14, 4; cf.
M . Hoffmann, Caesars Kalender (1934); G. Radke, R h M i960,
17 8 ;J. Beaujeu, Mélanges . . .J . Heurgon (1976), 13-
In the W est the Julian year was introduced without modifica­
tion, but in the Eastern provinces, where Greek was the official
language o f Roman administration, the new reckoning was
48 C H R O N O L O G Y OF THE A N C I E N T W O R L D

Antioch Lycia & Sidon Tyre*·*

Hyperberetaios1 = Oct. = LoosJ·4 i Dios = i 8. i i ( = N o v )
Dios = Nov. = Gorpaios i ApcUaios = 18.12
ApcUaios = Dcc. =Hyperberetaios i Audynaios = 17.1
Audynaios =Jan. =Dio$ i Pcritios = 16.2
Pcritios =»Fcb. =ApeIIaio$ i Dystros = 18.3
Dystros =Mar. = Audynaios i Xanthikos = 18.4
Xanthikos = Apr. = Pcritios i Artemisios = 19-5
Artemisios = May = Dystros i Daisios = 19.6
Daisios =June = Xanthikos i Panemos = 20.7
Pancmos =July = Artemisios i Loos = 19.8
Loos = Aug. = Daisios i Gorpaios = 17.9
Gorpaios = Sept. = Pancmos i Hypcrberetaios;= I 9.I0

Alexandrias G azai Ascalon*

29. 8= 1 Thot = i Gorpaios = 1 Loos
28. 9 = 1 Phaophi = i Hypcrberetaios = i Gorpaios
28.ί ο * i Hathyr = 1 Dios1 = i Hyperberetaios
27.11 = 1 Choiak = i Apellaios = 1 Dios
27.12 = 1 Tybi = 1 Audynaios = 1 Apellaios
26. ι = i Mecheir = 1 Peritios = i Audynaios
25. 2=1 Phamenoth = i Dystros = i Pcritios
27. 3= 1 Pharmuthi = i Xanthikos = 1 Dystros
26. 4=1 Pachon = i Artemisios = i Xanthikos
26. 5= 1 Payni = i Daisios = i Artemisios
25. 6 = 1 Epeiph = 1 Panemos = 1 Daisios
25. 7 = i Mcsore = 1 Loos = 1 Panemos
2 4 . 8-28.8: s Epagomenai

Asia Smyrna6 Bithynia Paphos7

31 Days 23. 9 = 1 Kaisarios = 1 Kaisarios = 1 Heraios = i Aphrodisios*
30 24.10=1 ApeUaios = 1 Tiberios = i Hermaios = i Apogonikos
31 23.11 = 1 Audynaios = 1 Apaturios = i Mctroos = i Alnikeios
31 24.12=1 Peritios = i Poseidon = 1 Dionysios = i Julios
28 24. ι = i Dystros = i Lenaios = i Hcraklcios = i Kaisarios
3i 21. 2 = 1 Xanthikos = 1 Hierosebastos = i Dios = i Sebastos
30 24. 3= 1 Artemisios = 1 Artemisios = 1 Bcndidaios = i Autokratikos
31 23. 4 = 1 Daisios = i Euangdios = i Stratios = 1 Demarchexasios
30 24. 5 —1 Panemos = 1 Stratonikos = 1 Pcriepios = i Pleisthypatos
31 23. 6=1 Loos = i Hekatombaios = 1 Areios = i Archicrios
31 24. 7 = 1 Gorpaios = 1 Antiochcios = i Aphrodisios;= i Hestios
30 23. 8 = 1 Hyper­
beretaios = 1 Laodikos ~ i Demetrios = i Loos

Fig. j . Some local Julian calendars1

 THE CALENDAR 49

generally adapted to local taste, as to the beginning o f the year

and the names and lengths o f months. For example, 6 January
in Rom e was equated elsewhere with n Tybi, 6 Audnaios, 14
Julos, and so on (Epiphan. Panar. 51, 24; cf. Mommsen, R StR
HI, 755).
The imperial government introduced the solar year slowly
and, as it seems, in agreement with the local authorities. Salamis
(Cyprus) was, probably, the first Greek city to accept Caesar's
reform in 46 bc (cf G. Jerphanion, L'Atitiq. class. 1932, 21). In
Egypt, Augustus, in 26 b c , reformed the Egyptian variable year
by adding the sixth epagomenal day every four years (ad 3, 7, 11,
and so on). From this time on, the ‘Alexandrian’ year, as it was
called, always began on 29 August. In the province o f Asia the
Julian year was adopted c. 9 bc and the N ew Year was to coincide
w ith Augustus’ birthday on 23 September (D. Magie, Roman
Rule in Asia Minor (1950), 1343 ; U . Lafti, S C O 1966, 1). The non-
Julian calendars, however, partly survived in the W est (Kubit-
schek, 136) and in many parts o f the East. The Roman government
NOTE
1 Local forms o f the Julian calendar have been preserved in three medieval
manuscripts which give synchronistic tables, day for day, for the calendars o f
eighteen provinces and cities. W . Kubitschek, DWA LVII, 3 (1915)- N ine o f
these hemerologia arc reprinted in H. Lietzmann, Zcitrcchmmg . . . für die
Jahre 1-2000 nach Christos (1934), 106. Samuel, 173 gives a (updated) list o f
months for sixteen provincial forms o f the Julian year. The indigenous popula­
tion sometimes used old names for Julian months, e.g. the Julian Hypcrberctaios
(18 S cp t.-i7 Oct.) was also called Thesre(Tishri) in Roman Arabia (F. Preisigke-
E. Kiessling, Sammelhuch griechischeti Urkuttden ans Àgypten X , 10288.
2 New Year. Between ad 458 and 483 the N ew Year day o f Antioch was
shifted to i September (E. Honigmann, Byzantion 1945, 338, and cf. Grumcl,
195). The N ew Year day o f the Lycian calendar is uncertain.
3 Cf. E. Schwartz, G G N 1906, 343. The calendar o f Caesarea (Palestine) was
o f the same type: cf J.-P. Rey-Coquais, Analecta Bollandiana (1978), 55·
4 In some cities, the first day o f the month was called ‘Sebaste’, and the
second day numbered as the ‘first’. Cf W . Kubitschek, ib. 81.
5 Calendars o f Alexandrian type; all months have 30 days.
6 C f L. Robert, REA 1936, 23.
7 Cf K. Scott, YCS 1931, 214; C. Bosh, Kleinasiatische Miinzen der romischett
Kaiserzeit II, 1 (1935). 132.
8 N ew Year on the birthday o f Augustus.
50 C H R O N O L O G Y OF THE A N C I E N T W O R L D

did not impose the official calendar in the provinces. Galen c. 160
has to explain the Julian year to his readers and states that numer­
ous Greek cities ‘and the inhabitants o f Palestine' continued to use
the time-reckoning ‘according to the m oon’ (In Hippocr. Epid.
XVIII, i, p. 23, ed. Kuhn). Such cities as Ephesus and Miletus
still clung to the old calendar in the age o f the Automnes (cf.
Magie, ib. 1343, n. 40), and Rhodes as late as ad 244-48 (cf J.
Oates, JE A 1969, 206). A t Sardis in ad 459 a document was dated:
V Kal. M ai (27 IV), 4 Daisios (cf Samuel, 187). Such double
dating was also used in Macedonia (cf Robert, 400). For centuries
after the introduction o f the Julian (Alexandrian) year in Egypt,
people continued to date according to the ‘old style’ (κατ' αρχαίους),
that is, according to the variable calendar (cj. e.g. U . W ilcken,
Chrestomathie (1912) no. 497). A late sixth- or seventh-century
papyrus gives the approximate equations between Roman (Julian)
and Alexandrian (Julian) months: September = Thot, and so on
(H. Gundel, A P E 1956, 13). Nicopolis ad Istrum, a Roman
municipium in Bulgaria, followed the Julian calendar (cf G.
M ihailov, Inscr. Graecae in Bulgaria repertae II (1958), 669). The
free city o f Thessalonica also used the Julian year (Ginzcl III, 7),
but Odcssus (Varna), a provincial city o f Mocsia Inferior, as late
as January 215, held to the lunisolar calendar which, at this date
at least, was in accord w ith the moon (L. Robert, RPh 1959,
210). N or was the Julian calendar adopted in the Bosporan
kingdom (cf. Corp. Inscr. Regni Bosporani (1965), 845). In most
cities, however, the moon calendar was disarranged. For instance,
at Tyras, on the northern shore o f the Black Sea, 30 Artemision
corresponded to 27 April in 182 (the conjunction fell on 1 May),
and 8 Lenaios corresponded to 17 February in 201 (conjunction:
22 January) (Inscr. Ponti Euxini I (1916), nos. 2 and 4). A t Gerasa
(Palestine) the Macedonian lunisolar year was also in confusion
(C. B. Welles in C . Kraeling, Gerasa (1938), 476). Sometimes the
disagreement with the sun year is so wide that it becomes
puzzling. In a Palestinian document o f 124, 19 October is given
as the equivalent o f 15 Dystros. Y et Dystros should have approxi­
mated to the Julian March (cf P. Benoit in Discoveries in Judean
Desert II (1961), no. 115).
 THE CALENDAR 51

A history o f the diffusion o f the Julian year has not yet been
written, but the solar year made the cult o f the sun popular
(Μ. P. Nilsson, A R I V 1932, 1 6 6 = id. Opuscula I, 462; S. W ein­
stock, J R S 1948, 37). The importance o f diis religion in the later
Roman Empire is evidenced by the fact that the Church trans­
ferred the date o f Christ’s birth to the birthday o f the uncon­
quered sun (dies natalis solis invicti: cf. B. Botte, Les origines de la
Noël (1932)). O n the representations o f Julian months in the
arts, cf. H. Stern, Journal des savants 1965, 122.

THE NATURAL YEAR

A ll the ancient calendars before the Julian year (except for the
late Babylonian 19-year cycle) were inadequate. They diverged
from the sun, disagreed w ith the moon, and always differed one
from another. But the heavens and the earth offered standards o f
time-reckoning which w ere independent o f the official calendars
and common to all: the succession o f seasons and the changing
aspects o f stars.
Th e geocentric path o f the sun (‘ecliptic’) is a circle, the plane
o f which is inclined to the plane o f earth’s equator at an angle o f
about 230 27'. This tilt causes the change o f seasons. A ll life on
the earth depends on the sun, and the amount o f light and heat
received from the sun mainly depends on the angle at w hich its
rays fall on the earth’s surface. There are four ‘turning points’
(tropai) o f the sun : tw o solstices when it reaches its farthest posi­
tions from the earth in the ecliptic, and tw o ‘equinoctial’ points
at the intersection o f the ecliptic and the equator o f the earth.
W hen the sun, m oving on its inclined orbit northwards, arrives
at the equinoctial point, it equally irradiates the north and south
poles, and the duration o f day and night at this time is equal and
the same over the w hole globe. Crossing the equatorial line from
south to norch (vernal equinox), the sun irradiates more and more
the northern atmosphere: the length o f the day and the intensity
o f the sun’s rays, w hich fall more directly on the surface o f
the northern hemisphere, increase and reach maximum at solstice,
when the sun stands directly over the tropic o f Cancer (230 5' N).
52 C H R O N O L O G Y OF THE A N C I E N T W O R L D

Afterwards, the sun begins its southward movement, crosses the

equator again at the autumnal equinox point, days get shorter
and shorter, and the rays striking our hemisphere are more and
more slanted and thus less and less efficacious. The minimum is
reached at the winter solstice, after which the sun begins its
northward course again. At the latitude o f Rome (410 54' N)
the insolation at summer solstice is more than three times greater
than the amount o f irradiation at winter solstice. (The maximal
changes in the distance o f the earth and sun during the yearly
revolution ot the sun are insignificant, about 1 per cent o f the
mean distance in each direction, and thus they do not influence
our climate.) Th e sun’s revolution in the sky determines the
repetition o f the same seasons at about the same time, and, there­
fore, the rhythm o f vegetation, and o f animal and human life.
This is the natural year (Figs 4, 5).
Local conditions fit the main periods o f the natural year: for
instance, inundation, sowing and harvesting time after the inun­
dation, and the period o f low water were the three seasons in
the Nile Valley. For the Sumerians, the winds divided the year
into a hot and a cold period (B. Landsbcrgcr, J N E S 1949» 248).
Similarly, the Greek and the Roman natural year was originally
subdivided into tw o parts. Four seasons are first mentioned by
Aleman (apud Athen. X , 416 d), and their limits remained
fluctuating. W h at counted in the farmer’s (or mariner’s, soldier's,
etc.) year were not these abstract concepts, but the phenomena o f
the natural year. The departure o f the cranes signalled the time
 THE CALENDAR 53

co sow (Aristoph. Aves 709) ; the com ing o f the twittering swallow
announced that the spring had begun (Hesiod, Op. 566). W hen
the tender branch o f the fig tree put forth leaves, men realized
that the summer was near (St Matth. 24, 32). Χειμών, hiems, for
military historians included autumn, that is, the whole bad
season (cf. M. Holleaux, R E A 1923, 352). ’Οπώρα is the height o f
the summer, but also the time o f gathering the fruit.
The stars, however, arc more reliable than the fig-tree in
marking the progress o f the year. From time immemorial man
had observed that fixed stars (which retain the same position with
respect to one another) change nightly in the sky. The light o f
these self-luminous bodies is effaced in the overwhelming bright­
ness o f the sun, and a given star is visible only when it is suffi­
ciently remote from the sun. T he sun advances among the stars
on its annual eastward course along the ecliptic. T he celestial
dome, however, performs its diurnal motion in a contrary
direction: both sun and stars rise in the east and sink below the
western horizon. Therefore, the sun, which reaches the same
point o f ecliptic every 365 days, must travel 1:365 portion o f its
path to come back to the same fixed star. As 2 4 x 6 0 = 1,4 4 0

Fig. 5. The (geocentric) path of the sun in different seasons

54 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

minutes: 365 gives the quotient 4, the sun lags about 4 minutes
behmd the stars in its daily course. Th e true and uniform period
o f the earth’s rotation with respect to stars is c. 23 hours 56
minutes. The (mean) solar day is 24 hours. W hen the sun pro­
gresses far enough from a given star, the latter appears above the
eastern horizon just before sunrise (the ‘heliacal’ rising). From
now on the star gains about 4 minutes daily on the sun and
rises earlier and earlier every night until it catches up with the
evening sun and is again lost in its proximity. The schedule for
the setting o f the same star in the west is similar. These four
epochs (the first and the last apparition in the east; and the first,
before the sunrise, and the last, just after sunset, descent under
the western horizon) occur only once during a solar year, and, for
a given latitude, on the same dates, which can be regarded as
constant for historical purposes. For instance, the respective dates
for Sirius in Athens (38° N) in 43 bc were: the heliacal rising:
28 July; the last visible ascent: 31 December; settings, on 5 M ay
and 26 Novem ber. Thus, the star was invisible between 5 May
and 28 July {cf. F. Boll, R E VI, 2427; Gundel, ib. IIIA, 339).
A s early as the beginning o f the second millennium the
Egyptian priests computed the daily delay o f stars. In the Hellenistic
age, Greek navigators used a sort o f computing instrument for
the same purpose (Fig. 6).44
A natural and reliable standard o f time-measurement, the
stars appear in the farmer’s almanac o f Hesiod, besides the voice
o f the crane {Op. 448), to point the propitious time for agricul­
tural w ork: harvest when ‘the Pleiads, daughters o f Atlas’ rise,
and sow when they set {Op. 383 \cf Aratus, 266). Th e shepherds in
Sophocles’ Oedipus Tyrannus (1137) describe the period o f pasturage
as six lunations ‘from spring to Arcturus’. The ancient mariners
depended on the stars for navigation and time-measurements:
‘then the sailor numbered the stars and gave them names’
{navita turn stellis numéros et nomina fecit, Virg. Georg. I, 137). In
Athenian contracts o f bottomry, the charged interest went up
from 22-5 per cent to 30 per cent after Arcturus (Dem. 35, 10).
Scholars {e.g. Hippocrates) again used time references o f the
natural year (K. Dcichgraeber, A P A 1933, 29). For Aristotle
 THE CALENDAR 55

(Hist. Anitn. VI, 569 b) the rising o f Arcturus ends the summer,
and the migration o f cuckoos occurs between the rising o f Sirius
and the spring (ib. IX, 633 a). Thucydides’ ‘divisions o f time’
(V, 20: κατά τούς χρόνους) were again seasons: summer’, that
is, the period o f military operations, and ‘winter’. His readers
probably knew the chronological meaning o f these terms (cf.
W . K . Pritchett, B. L. van der Waerden, B C H 1961, 29).
W ithin the season, Thucydides used the subdivisions o f the
farmer’s year: ‘when the grain comes into ear’ (IV, 1), ‘before the
grain was ripe’ (IV, 2) and so on, and sometimes celestial pheno­
mena, such as winter solstices (VII, 16, 2; VIII, 39, 1; c f A . W .
Gomme, Commentary on Thucydides III (1956) 699, 716). He
regarded his chronological system as more exact than the reckon­
ing in civil years (V, 20). Four centuries after Thucydides, in
Diodorus’ annalistic w ork, the years arc named after the Athenian
archons, but within the year the time is indicated in seasonal
terms: thus, for instance, Agathoclcs o f Syracuse began his
African campaign in the fruiting season (X IX , 65) and returned
to Sicily ‘at the time o f the setting o f the Pleiads’ (X X , 69).
T he art o f reading the signs written in the sky, those o f the
night, o f the month and year (Xen. Mem. IV, 7), this gift o f
56 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

Prometheus (Aesch. Prom. 457), before the Imperial age was a

part o f basic education. This fact explains the great success o f the
didactic book in verse about the stars written by Aratus (died in
240 bc ). In Rome too, daily conversation included the morning
rising o f the Lyre just as today w e talk o f the weather (Plut.
Coes. 59), and the aspects o f the stars constituted the basis for
meteorological forecasting (Cic. Ven. II, V , 27). So Polybius
(IX, 14) claimed that generals too should be able to tell the
length o f the day and night by the stars, as well as recognize the
solstices and equinoxes, and be capable o f constant observation o f
celestial bodies. As Copernicus (paraphrasing Plato, Epin. 987a)
once said: ‘the ancients were favoured with a clear sky, because
the Nile, so they say, never gave o ff vapours like those o f the
Vistula*.45
O f course, neither the given agricultural work nor the setting
o f the Pleiads occurs everywhere and always at the same Julian
date. The setting o f the Pleiads happened for Agathocles on
6 April and in Diodorus’ time on 8 April. But the stellar or
agricultural reference gave a universal and undisputed, though
approximate, indication o f time within the year. Th e situation
changed only with the introduction o f the Julian year. W hen he
w rote (c. 36 bc ) his treatise on agriculture, Varro could refer to
the Julian dates o f the agricultural calendar, ad dies civiles nostros
qui nunc sunt (De re rust. I, 28, 1). Nevertheless farmers’ calendars
continued to juxtapose the stellar and the Julian dates, since the
stars were regarded as harbingers and even as originators o f
weather changes (cf. G. Boll, Griechische Kalender, SBH A 1910,
1911, 1913, 1914, 1920; A . Rchm RE, Suppl. VII, 175). For
Virgil (Georg. I, 218) and for Petronius (55) the spring began not
on 22 April, but tinder the sign o f Taurus, and it was indicated
by the arrival o f the stork, titulus tepidi tewporis. A mosaic o f
St Romain-en-Gal (Museum o f St Germair.-en-Laye) well ill­
ustrates this popular clim atology (cf G. Lafayc, R / 1 1892, 322).

THE ZODIAC
T he accord o f the natural calendar, regulated by the stars, with
the sun and w ith the civil reckoning was established by dividing
 THE CALENDAR 57

the yearly path o f the sun through the fixed stars into twelve
equal sections, according to the number o f lunations in a solar
year. This is the Zodiac. The Babylonians were, probably, the
first to trace it and divide it in signs o f 30 degrees each. The twelve
signs were named after relevant constellations which, however, as
Geminus (1) warns his reader, do not exactly fit the allotted
portions o f the sky.

Signifef inde subest, bis sex et sidéra comptent

Hunc: Aries, Taurus, Gemini, Cancer, Leo, Virgo,
Libra, Scorpius, Arquitenens, Capricornus et urnam
Qui tenet et Pisces . . .

(PoetaeLatin. Minor, ed. W . Bachrens V (1883), 352). Th e Zodiacal

year began with Aries, that is, the sign o f the vernal equinox.
W hen the sun entered the sign o f Cancer, it was summer; Libra
corresponded to the autumn; and Capricorn marked the
winter. Thus, the position o f the sun in the Zodiac was o f the
greatest importance for the course o f the seasons.
For us, the longest day (the summer solstice) is 22 June
(Gregor.). For the ancients it was the time when the sun was in
the first degree o f the Cancer. This zodiacal clock was simpler to
use than the star-clock o f the natural year (p. 54). For instance,
V a n o dated the Roman feast o f Robigalia as follows: ‘ W hen the
sun reaches the tenth degree o f Taurus’ (Plin. N H XVIII, 286).
T he Babylonian astronomers, however, as early as c. iro o related
the official lunisolar calendar to the rising o f the given stars in a
given month. A Babylonian astronomical w ork written before
700 b c used the schematic year o f 12 x 30 montas to correlate
the star calendar and the official time-reckoning (B. L. van der
Waerden, JN E S 1949, 6; Pritchett and van der Waerden, B C H
1961, 41). In Greece, Mcton was probably the first who, in 432,
publicly displayed the stellar calendar which, using the zodiacal
division, indicated the daily progress o f the sun. For instance, an
almanac o f this kind announced: ‘30 days (o f Aquarius). The
1st : The sun in Aquarius. Th e 2nd: The Lion begins to set in the
evening. Setting o f Lyra. T he 5th: the evening setting o f the
Cygnus begins/ and so on (cf. A. Rehm, S P A IV 1904, 97). N o w ,
58 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

it was easy to relate this star calendar to the official reckoning.

Suppose the sun entered the sign o f Aquarius on N -day o f the
civil calendar. Th e evening setting o f Cygnus w ould, then,
happen on the civil day N + 4, and so on. These tables were con­
strued with regard to the astronomical cycles (p. 29) and also
gave weather prognostics. An ingenious device (parapegma) made
it possible to mark the days o f the given calendar month by
movable pegs inserted in holes beside the stellar references. For
instance, in the almanac o f Euctcmus the appearance o f the
swallow was fixed on the 2nd day o f Pisces. Th e parapegma
allowed the conversion o f this indication into a date o f the local
calendar. In a similar w ay, the zodiacal year was used in modern
Persia. Until the introduction o f the Gregorian calendar, in 1925,
the Persian financial year ran from one spring equinox to the
next, and its twelve months were named after the zodiacal signs
(cf. S. H. Taqizadeh, B S O A S 10 (1939-42), 132).
W c may note on this occasion that for the ancients the chart
o f the sky differed somewhat from ours. The change was deter­
mined by the phenomenon o f the precession o f the equinoxes,
discovered by Hipparchus (O. Ncugcbauer, J A O S 1950, 1). The
vernal point moves westward, along the zodiac. (The causes o f
this retrograde movement arc a part o f the general law o f gravita­
tion and the precession, in turn, confirms N ew ton’s theory.)
Consequently, the distance o f the stars from the equinoctial
points changes. Hipparchus could calculate that eg. the distance
o f Spica from the autumnal equinoctial point in his ow n time
was 6 ° but in the time o f the astronomer Timocharis (c. 300 b c ),
8° (Ptol. Almag. VII, 2); accordingly, today Aries is in the ancient
sign o f Taurus. In other words, today at the spring equinox the
sun enters the sign o f Pisces, between c. 1000 b c and a d 1000
the venial point was in Aries, between 3000 and 1000 b c in
Taurus, etc.46

THE WEEK

The Venerable Bede, that famous chronologist o f the Middle

Ages, said that the division o f time natura aut consuetudine aut
auctoritate decurrit (PL, X C , 279). The year is timed by nature.
 THE CALENDAR 59

Fig. 7. The order of the planets

the unequal lengths o f the months by tradition, and the week by

authority.
The artificial time-units o f three, five, seven, etc., days occur
among many peoples (cf. Nilsson, Time-reckoning, 324). For
instance, a seven-day period o f time is often mentioned in
Sumerian and Babylonian texts (cf. Langdon, Menologies, 89;
H. and J. Lewy, H U C A X V II (1942-3), 6). T he Romans used the
market week (as do many primitive peoples, Nilsson, ib. 325) o f
eight days, known also to the Etruscans (Macrob. Sat. I, 15, 13).
These nundinae were indicated in the calendar by the letters A to
H : seven days o f w ork, the eighth day for the market (Macrob.
I, 16, 32).
T he rural population came to the city at the Nundinae (Varro,
De re rust. II, Praef). Thus public auctions and the like were also
held on the Nundinae, which became a day o f festivity. Varro
implies that the countryman only shaved for market-days when
he went to town. (Quoties prisons homo ac rusticus Romanus inter
nundinum barbam radebat? Varro ap. Nonius, p. 214, 28.) O ur
6o C H R O N O L O G Y OF T H E A N C I E N T W O R L D

1Γ
0
u
1
u
CO
1
u
rt
CO
8
>*
A(nnus)

0 1 2 3 4 5 6 7 8 9 Month M(ensis)

0 0 10 3 0 2 1 0 6 4 3 2 1 6 5 January I 2
I 1 II 4 10 4 3 1 0 6 5 3 2 1 0 February 5 6
2 2 12 5 20 5 4 3 2 0 6 5 4 2 1 March 5
3 3 13 6 30 0 6 4 3 2 1 6 5 4 3 April 2
4 4 14 0 40 1 0 6 5 3 2 1 0 5 4 May 0
5 5 15 I 50 3 2 0 6 5 4 2 1 0 6 June 4
6 6 l6 2 60 4 3 2 1 6 5 4 3 1 0 July 2
7 0 17 0 70 6 5 3 2 1 0 5 4 3 2 August 6
8 I l8 2 80 0 6 5 4 2 1 0 6 4 3 September 3
9 2 19 4 90 2 1 6 5 4 3 1065 October I
November 5
December 3

Fig. 8. The week

The Sunday is found by addition o f the numbers under S(aeculum), A(nnus)

and M(cnsis). For seventeenth-nineteenth centuries the dates arc Gregorian.
For January and February in bissextile years [cf. p. 47) use the last column.
For example, a funerary inscription is dated as follows: post consulatunt [of
Arcadius and Rufinus] die Lunae, IX Cal. Iun. (E. Diehl, Inscriptions Latinae
Christianae I, no. 582). The date corresponds to 24 May ad 393 and the day
is said to be Monday. Let us check this statement. 393 = 300 + 93. According
to our table, 300 corresponds to the number 3 in the column S ; 93 corresponds
to the figure 5 in the column A. May corresponds to the number o in the
column M. W e add: 3 + 5 + 0 = 8 . Accordingly, 8 May was a Sunday in
ad 393. Therefore, 22 May was also a Sunday, and 24 May was a Tuesday.
24 May was, however, Monday, in ad 392. The author o f the inscription
probably confused the year o f the consulate of Arcadius and Rufinus ( ad 392)
and the year post consulatum, that is ad 393 (Mommsen). The figures in the
example quoted here are in bold type in the table.
 THE CALENDAR 61
week, o f which Bede spoke, goes back to the authority o f the
Bible and Jewish practice. Tow ard the end o f the first century,
Flavius Josephus could state (Contra Apionem II, 39, 282) : ‘There
is no city, Greek or barbarian, not a people, to whom our custom
o f abstaining from w ork on the seventh day has not spread.’
The origin o f this septenary time unit (Hebrew shabua; cf.
Hebrew sheba— ‘seven’) is unknown. The days o f the Hebrew
w eek arc counted as they still are in the Greek Orient and b y the
O rthodox Church (and therefore among the Slavs). In Western
Europe, on the other hand, the days arc named after planets:
M oon, Mars, Mercury, Jupiter, Venus, Saturn, Sun. O ur week,
in fact, has its secondary origin in the planetary, astronomical
w eek o f the Imperial age, which gave each day its ruler, that is
to say, the planet which governed the first hour o f each day.
T he Jewish week began on the Sunday. Thus, for instance,
St Matth. 28:1: ‘after the Sabbath, toward the dawn o f the first
day o f the w eek’. In the planetary week, the sequence o f days
corresponded to the order o f planets according to their distance
from the earth (Fig. 7): Saturnus, Jupiter, Mars, Sun, Venus,
Mercurius, Moon. Thus, the first day was Saturday. But the
planets also ruled the 24 hours o f each day. T he first (and conse­
quently the 8th, the 15th, the 22nd) hour o f Saturday were
allotted to Saturn, the 23rd to Jupiter, the 24th to Mars; and the
first hour o f the next day to the Sun, which thus ruled Sunday.
Therefore, in the planetary (and our own) week, dies Solis
follows dies Saturni (cf F. Boll, RE VII, 2558). From this comes
the custom, introduced in the third century a d from East to
W est, o f indicating the most important dates according to the
weekdays as well (e.g. C IL III, 1051 : X K . hm. Inn. XVIII die lovis
= 23 M ay, a d 205).47
T he planetary week, which according to Celsus (ap. O rig. c.
Cels. VI, 22) was a part o f ‘Persian theology’ (cf. F. Cum ont,
R H R CIII, 1931, 54), penetrated into the W est under Augustus
(cf Tib. U, 3, 18). Constantine, in 321, sanctified the astrological
w eek by ordering that ornnes judices . . . et artium officia cunctarum
venerabili die solis quiescant (Codex Just. Ill, 12, 2). The farmers
were expressly excused from observing this ordinance.48
 C H A P T E R II

CHRONOGRAPHY

T again and again and are always the same: one

im e - u n it s r e t u r n

day is like another. O n ly events— birth and death, a good

harvest, a bad harvest— singuiarize time-units by making them
unequal in value and thus memorable. Chronography, the
method o f establishing time-intervals between events and
between them and the present, is thus different from calendar-
iography, w hich deals with standard elements o f time
measurement.

RELATIVE CHRONOLOGY

Th e simplest and most ancient method o f dating is the relative

time-reference which does not require any chronological devices
(eg. the Epidamnians exiled the aristocrats ‘before the w ar’ : Thuc.
I, 24). Except for savants, men have little interest in absolute time
notations; they use, instead, relative time-references. Primitive
peoples usually do not know h ow old a man is, but only w h o is
the oldest in a group (Nilsson, 98).
The counting o f generations is the simplest device o f chrono­
graphy. In order to measure the length (otherwise unknown) o f
the IV and V Egyptian dynasties, scholars add up life-ages o f
successive court officials. In the same way, the earliest Greek
historians set up a chronological fram ework for their narratives
by counting generations. Th e first Greek historical w ork by
Hccatacus o f Miletus was entitled ‘Th e Genealogies’. The A lex­
andrian scholars used the same device in order to establish
synchronisms: ‘Hecataeus lived at the time o f Darius I and was
older than Herodotus.’
Relative chronology hinges on some know n time-point.
Thucydides dates events w hich led to the Peloponnesian W ar
indirectly, making the attack on Plataea his reference point
 CHRONOGRAPHY <$3
(F. Jacoby, G G N 1928, 1). In order to establish the date o f the
sack o f Rome by the Gauls, a date which was fundamental for
his Roman chronology, Polybius (I, 6, 1) states that the event was
contemporaneous w ith the peace o f Antalcidas and the siege
o f Rhegium by Dionysius, and that it happened 19 years after
the battle o f Aegospotami and 16 years before the battle ofLeuctra.
W ith the aid o f a scries o f such fundamental datings, w hich he
used as points o f reference, Eratosthenes (c. 250 b c ) composed the
first scientific chronology.
Every dating, however, is useful only i f its distance from the
present is know n; each dating system must be related to the
present. An inscription, know n from the name o f the place in
which it was found as the Parian Chronicle (Martnor Parium),
enumerates events o f the past according to the distance from its
ow n date (264/3 BC) : ‘From the time when Cecrops became king
o f Athens— 1,218 years/ This device, which (in a short time)
made the Parian list useless, shows the inherent inadequacy o f
relative chronology, which is intelligible only in connection with
an absolute date.49
O n the other hand the elements o f absolute chronology are not
isolated dates but uniform time-units, an uninterrupted series o f
which leads to the present. Absolute chronology borrows the
concept o f ‘year’ from the calendar, but the chronological year
is an historical unit, that is, a link in a series o f years, whether they
be numbered or otherwise individualized. This labelling dis­
tinguishes the chronological year from the calendar unit.

NAMING THE YEAR

W e have a uniform standard o f time, our civil year. (The fiscal,

the ecclesiastical, the school year, etc., serve specific purposes
only.) This civil year is the Julian year, w hich did not exist before
45 b c . O ur N ew Year also comes from the Julian year. Th e Greek
language distinguished between the yearly cycle o f seasons
(1eniautos) and the civil year (etos) (Ad. W ilhelm, S W A IV
C X LII (1900), 4). An eniautos lasted from any chosen time-point
in the natural year to its recurrence. Greek calendars based on the
natural year could begin w ith Aries as well as w ith Cancer, and
64 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

so on. Th e etos was a conventional time-unit. Th e new moon was

one o f the most important Greek festivals, while the new year,
or ‘new new-m oon’ had no importance (OGIS, 458, 21: via
νουμηνία). In Rome, wartime operations began at the Kalendae
Martiac, but this day was in no w ay distinguished from the other
first days o f the months.
The Greeks celebrated birthdays every month (W . Schmidt,
R E VII, 1136). The annual renewal o f treaties was performed at
the same festival, not on the same calendar date (Thuc. V , 23, 4);
the treasurers rendered accounts at the Panathenaea. For Polybius
a year is a variable quantity: its beginning and length change in
the course o f his w ork, according to his sources and his organiza­
tion o f material (cf., however, P. Pedech, La méthode historique de
Polybe (1964), 449). Th e fluctuating value o f the etos came to be
stabilized for administrative or religious reasons. In both Egypt
and Babylonia, important festivals o f the ‘beginning of the
Year’ were celebrated from the most ancient times. The period
between tw o consecutive N ew Year festivals became the earliest
chronographic unit, marked by year-names such as ‘year o f
counting o f cattle’, ‘year o f the victory over the Nubians’, and
so on. A scries o f years thus described constituted the earliest
chronological tables. Such a table, written at the end o f the V
Egyptian dynasty, has been preserved on the ‘Palermo Stone*
(cf. note 66 and K . Sethc, G G N (1919)» 303). Another w ay o f
defining a civil year was to begin it at the fixed date, when
some major magistrate took office: Hekatombaion for the
(eponymous) archon in Athens; 15 March for the consuls in
Rome, from 222 to 153 b c (Mommsen RS:R I, 599); 13 Aiaru
for the ‘limm u’ o f the city o f Ashur, and so on. This eponymous
year became a chronological unit (o f variable length by reason
o f intercalation), but w ith a definite beginning. The office-year,
however, was not the same for all magistrates. The prytany year
in Athens did not have to coincide with the archon’s year (p. 34) ;
similarly, the Roman consuls took office on 15 March and (from
153 b c ) on r January, but the tribunes began their year on 10
December. Th e Roman emperors numbered the years o f their
tribunician power, which was renewed annually. From Augustus
 CHRONOGRAPHY <>5

to Trajan, the tribunician years w ere reckoned from the accession

day; but from Trajan until the Severi, they were numbered from
10 December. 5°
T he stability o f the office-year made possible its use as a
chronographic unit. T he years were indicated by the name o f the
eponymous magistrate (archon, ephor, etc.): ‘o faithful jar o f
wine, born with me in the consulship o f Manlius' (0 nata tnecutn
consule Manila . . . pia testa, Hor. Odes III, 21). I f the eponymous
magistrate served for six months, die civil year had the same
length; it was έξάμψ ος (cf. Busolt-Swoboda II, 457; IG XII, 5,
881 (Tenos); R. Herzog, A P A 1928, 50 (Cos)). Likewise, in
Babylonia the ‘year’ originally comprised six months as there
were tw o ‘Akitu’ festivals, one in the month o f Tashritu and
another in the month o f Nisanu (F. Thureau-Dangin, Rituels
Accadiens (1921), 87; S. A. Pallis, The Babylonian Akitu Festival
(1926)). In early Sumerian texts, the dating by the term o f office
o f a magistrate corresponded to years o f varying length; W . W .
Hallo, J C S i960, 189. C f H. Tadmor, J C S 1958, 26.
Thus, the calendar year was identical with the office-year o f
the eponymous magistrate. In a document commemorating the
introduction o f the Julian year in the province o f Asia, the N ew
Year was described in Latin as tenipns anni novi initiumque magis-
tratuum, and in Greek as ‘die beginning o f the term in office’
(O G IS 458, 14). Similarly the Pracnestine Fasti note under
i January: Annus novus incipit quia eo die magistratus ineunt.
Accordingly, the pre-Cacsarian Roman calendar year already
began with January since, from 153 b c on, the consuls entered
office on i January (cf Dcgrassi, Fasti Antiati, no. 9). Th e offering
o f the strena was similarly advanced from 1 March to 1 January
(L. Deubner, Glotta (1912), 34). O n ly under the Caesars, under
the influence o f astrology, did the N ew Year as such acquire the
value o f a time mark, and thus gave rise to our civil year, and
our N ew Year holiday (cf Μ . P. Nilsson, A R W 1916, 66, and
M . Messlin, La fête des calendes de Janvier (1970)).
W here the iteration o f the eponymous magistracy was per­
missible (as in Rome), the repeated magistracies could bc counted.
The consul-year o f 44 b c was Caesare V et Antonio (‘Caesar for
66 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

the fifth time and Antoni us’). Royal years as well could be
numbered. Th e regnal years naturally were counted from die
accession day. Thus the regnal year, like the eponymous year,
determined the beginning and the end o f the civil year or, at
least, ran independently from the latter. Such was the case o f the
Ptolemies in Egypt during the entire time diat they used the
Macedonian calendar, and o f the Selcucids (cf p. 38).
The same is true for the regnal year o f other Greek and Mace­
donian kings. But in Egypt and Babylonia naming o f years
preceded the reckoning by the numbered regnal years. The latter
system became standard in Babylonia only in the Kassite period,
that is, from seventeenth century b c on, according to the now
usually accepted chronology (see p. 84). Thus, the regnal year
had to be adjusted to the standard civil calendar. The Egyptians
reckoned the period from the accession to the next N ew Year
(I Thot) as the first year o f the reign. Th e next full calendar year
was counted as the second year o f reign, and so on. O n ly under
the eighteenth through the twentieth dynasties did the regnal
year run from the accession day to its anniversary. O n the other
hand, in Babylonia the period from the accession to the next
N e w Year (1 Nisanu) was called ‘the beginning o f the reign’,
and the next full calendar year was numbered as the first year o f
the new king.51
The Roman emperors did not count their years o f reign but
their tribunates; yet dating b y the regnal years o f the Caesars
was w idely used in Palestine (cf Luke 3, 1), Syria, Arabia,
Bithynia, Pontus, Cyprus and Egypt (cf J. Goldstein, J N E S 1966,
8). T he counting o f Imperial years was adapted to the local styles
o f reckoning. In Syria, for example, the second year o f the new
emperor began on the next 1 O ctober after his accession, that is,
at the next N ew Year o f the calendar o f Antioch (cf C . Cichorius,
Z N T W 1923, 18). In Egypt, the second regnal year began on
29 August after the accession, that is, the Alexandrian N ew Year
(see p. 50). For Byzantine dating, see F. Doelger, Byz. Zeitschr.
1932, 275; id. SB A 1949, no. 1.
Th e chronographcrs, in order to be able to use the years o f
reign as chronological units, had to relate them to a standardized
 CHRONOGRAPHY 67
year in order to make them uniform. T he year in which a
sovereign came to the throne was accordingly aîtribuced some­
times to his predecessor (antedating), and sometimes to his
successor (dating in advance). For example, while the last year
o f the reign o f Alexander the Great (d. 10 June 323 bc ) was
usually counted as the first year o f Philip Arrhidacus, in some lists
the w hole year was assigned to Alexander (S. Smith, RAss 1925,
186). For the same chronographic reason a Babylonian list
attributes to Alexander only seven years o f reign in order to
make his years follow the reign o f Darius III, which ended in
330 bc . Babylonian documents naturally count Alexander’s years
from his ascent to the Macedonian throne, in 336 bc (cf Ed.
M eyer, Forschungcn II, 457).

THE EPONYMOUS YEAR

The main bulk o f datings given in our sources from the ancient
Near East, Greece, and Rome, refer to the eponymous years.
Therefore, in order to understand these chronological references,
w e must be able to ascertain the distance o f the given eponymous
year from the present. First, w e have to determine the relative
chronology o f the eponymous year in question, that is, its place
in the succession o f eponymous magistrates o f the given city,
and secondly, w e must link the list o f eponymous magistrates
to our absolute chronology.
Th e latter problem can bc solved as soon as w e obtain a
synchronism for the list in question. Thus, the whole series o f the
eponyms o f the city o f Ashur from 893 to 6 6 6 b c is dated, thanks
to the mention o f the solar eclipse o f 15 June 763, in the year o f
one o f these eponyms. Alexander the Great was the stephanephoros
o f Miletus, probably in 333 b c . His name in the list o f these
stephatiephorci, which begins in 525 b c , dates the whole series
(A. Rehm, Milet III; Delphitiion (1913)» no. 122).
Th e estab.ishment o f the relative chronology o f eponyms is
rarely possible unless w e have the ancient lists o f them; otherwise
the names float in time. The catalogue o f Athenian archons from
the Persian W ar to 302 bc has been preserved in the Books
68 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

X I - X X o f Diodorus, w h o in his annalistic narrative mentions the

Athenian archon o f each year from 480 b c on. Dionysius o f
Halicarnassus (Dinarc. 9) enumerates the archons until and
including 293/2 (cf Dinsmoor, 39). For the later period w e have
only fragmentary and disconnected lists on stone. From 356/5
on (with some interruptions such as under the oligarchic regime
from 321/0 to 308/7) the annual secretaries (grammateis) followed
each other in a regular sequence according to the tribes from
which diey came: the grammateus o f the Erechtheis was follow ed
by the grammateus from the Aeges, and so on. Thus, the tribe
o f the grammateus indicates the place o f the corresponding archon
in the tribal cycle (W . S. Ferguson, Athenian Archons (1899)).
Y et, there were also disturbances within the cycle (cf. Pritchett,
385). Thus the number o f Athenian archons before 480 and after
292 whose Julian year is certain remains very small, e.g. Phainippos
in 490 (battle at Marathon). O n ly five archons o f the third century
(after 292) are dated w ith certainty by synchronisms (Dinsmoor,
45; Samuel, 210). O n the date o f die archon Arrheneides, cf.
Pritchett, 288.
T he case o f the archon Polyeuctus, whose date is crucial for
Delphic chronology, illustrates the difficulty o f dating the
Athenian archons o f the Hellenistic Age. T w o synchronisms show
that he exercised his functions at the time o f Antigonus Gonatas
(263-240) and Seleucus II (246-225) (cf L. Robert, R E A 1936, 5).
His year in office must thus be placed between 246 and 240. His
probable date w ould be 246/5 (cf G. Nachtergael, Historia 1976,
62). Y et the date 251/0 (E. Manni, Fasti Ellenistici e Romani (1961),
82) or the date 249/8 is still supported b y competent scholars (cf
Meritt, 234; Samuel, 214; Meritt, Historia 1977, 168).
Thus all proposed lists o f the Athenian archons o f the Hellen­
istic A ge differ and all are equally uncertain (cf Manni, op. cit. ;
Samuel, 212; Meritt, Historia 1977, 168).sz O n the archons
between ad 96 and 267 cf. S. Follet, Athènes au Ilème et Même
siècles (1976).
O ur reconstruction o f a series o f eponyms generally depends
upon the existence (and availability) o f corresponding ancient
records. A t some date a city decided to write down a list o f its
 CHRONOGRAPHY 69
past magistrates and to continue it each year. For instance, a list
o f priests at Cos which begins in 30 b c was published in 18 b c .
T he aforementioned list o f the stephanephoroi o f Miletus covers
the period from 525/4 to 314/13 (ib. nos. 123-8 name eponyms
from 3 £3/12 to AD 2/3). Th e list was engraved in 334/3, and
afterwards the name o f the eponym was added every year. The
question for the chronologist is h ow far back such a record is
reliable. In the time o f Plato (Hipp. Maj. 285 e), the Athenians
believed that the list o f archons, starting from Solon (594-3) was
reliable. Yet, compilers could easily tamper with the list or simply
invent the eponyms or kings o f hoary antiquity. W c reject as
impossible the figures given in a cuneiform list for the twenty-
three kings o f the First Dynasty o f Kish w ho allegedly reigned
for 24,510 years. W e disbelieve the list o f archons for life and o f
decennalian archons o f Athens for 1068-684 b c , but we may also
question whether the first annual archon was Creon w h o exer­
cised his office in 683 b c , as the Manner Parium tells us.”
The Romans dated by consuls until a d 537 when Justinian
(Novell. 47) introduced the dating according to the regnal years
o f the emperors. From 534 in the W est and after 541 in the East,
only the emperors held the consulship. Yet, the dating by
consuls continued to bc used in Egypt until 611. Accordingly,
w e have the complete list o f consuls from Brutus and Collatinus,
the founders o f the Roman Republic in 509 b c , to Basilius in
a d 541: 1,050 years.54

From c. 300 b c on, the fasti are reliable, as the Greek historical
tradition and contemporary documents show. It is probable that
the original list was composed by the pontifices c. 300 b c . The
question is how far the list for 509-c. 300 is trustworthy. Follow­
ing Mommsen, modern historians generally accept the list except
for the first years o f the Republic. T he Julian years o f early consul­
ship, however, remain uncertain because o f the disagreement
am ong sources. The cornerstone o f ancient Roman chronology
was the capture o f Rom e by the Gauls, since this event was the
earliest fact o f Roman history mentioned and dated by contem­
porary Greek authors. Th e date corresponded to 387/6 b c (see
p. 63; c f F. W . Walbank, Commentary on Polybius I (1957), 46;
70 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

P. Pedech, La méthode historique de Polybe (1964), 438). Yet, the

Roman consular list indicated 382 b c . In order to use the Greek
synchronism, Diodorus twice gives the names o f the same
Roman eponyms, to wit, for O lym p. 96, 3-97, 3 and O l. 98,
3-99, 3 (cf Ed. Schwartz, R E V, 695). L iv y reaches the date 387/6
by inserting a quinquennium o f anarchy without the magistrates
(VI, 35, 10: solitudo magistratuum). The Fasti Capitolini insert four
years o f dictators sine cotisule and in this w ay arrive at 391/0 as the
date o f the Gallic sack o f Rom e (cf Mommsen, 114; id. R S tR II,
i, 160).
As a matter o f fact, before 222 there was no fixed date for
taking office. A consul could start and end his consulship at any
date within the seasonal year (see Mommsen, ib. I, 597). O n the
other hand, the length o f the seasonal year was also variable (cf.
p. 44). Thus the number o f consulships was hardly the same as
the number o f Julian years between the foundation o f the
Republic and the redaction o f the consular list c. 300 bc . 55
W hat has been said concerning the eponyms is also true o f the
royal lists. W e are able to ascertain the succession o f kings and
their dates only on the basis o f corresponding lists compiled by
ancient historians (see e.g. the list o f the rulers o f Pcrgam um in
Strabo, 624 C ; cf. W . Kubitschek, R E XI, 996). W here such lists
are lacking, as for example for Parthia or Pontus, a new dis­
covery might at any time change the accepted order o f the
kings and their chronology. This has already happened more than
once (cf e.g. Th. Reinach, Histoire par les monnaies (1902), 167;
E. J. Bickerman, B O 1966, 15).

THE ERAS

T he datings by eponyms or regnal years are isolated items which

must be grouped in a series continued to the present. T he era
(that is, ‘number’ : cf A. Ernout, A . Meillet, Dictionn. étymolog.
de la langue latine4 (1959), s.v. aéra) numbers the years. It is enough
to know its point o f departure for converting its datings into
Julian years. A church council took place in T yre on 16 September
643 o f Tyrian reckoning. W e know that the Tyrian era began in
the autumn o f 126 bc and that the Tyrian year (in the Roman
 CHRONOGRAPHY 71

period) started on 18 October. Therefore the aforementioned

Tyrian date corresponds to 16 September a d 518. ( O f course this
conversion rule is inapplicable to purely lunar, or even lunisolar,
dates, where w e must also know the character o f the year and
month in question, eg . whether the year was intercalated.)
This convenient method o f dating came into public use only
in the Hellenistic Age. Indeed, the era postulates a uniform year
as its basic unit. Such a year (leaving out the Egyptian mobile
year) was first achieved, thanks to the 19-year cycle, in Babylonia
(p. 24). Th e first ‘era’ came into being there also when Seleucus I
began to count his regnal years according to the Babylonian
calendar and Antiochus I continued the counting o f his father’s
years. His successors, in turn, followed his example and in this
w a y the earliest dynastic reckoning was adopted in the whole
Seleucid empire, as ‘the years o f the Greek domination’, to use
the name given to this era by the Jews and the Syrians.
T he epoch from which the Seleucid years were counted was
the Julian year 312 /n . After reconquering Babylon in August o f
312, Seleucus I, in the next royal year (the 7th year o f Alexander,
son o f Alexander the Great), began to count his satrapal years (cf
S. Smith, KAss 1925, 190). In this he followed the example o f
Antigonus and other satraps in Babylon (cf Ed. Meyer, For-
schungen II, 458). According to the Babylonian calendar, the 7th
year o f Alexander IV began on 2 April 311. The Macedonians,
however, counted the years o f Alexander IV from the death o f
Philip Arrhidacus in the autumn o f 317. Thus, for them the 2nd
year o f the satrap Seleucus began in the autumn o f 311, while
for the Babylonians the same year began on 22 April 310. As
king, Seleucus continued the reckoning o f his satrapal years
(E. J. Bickerman, Berytus 1944, 73; A. Aym ard, REA 1955, 105).
For the court, the Seleucid year began between 1 Loos and 1 Dios,
that is, in the late summer or early autumn (cf C . B. W elles,
Royal Correspondence in the Hellenistic Period (1934), 18; id. The
Parchments and Papyri (1959), 10).
Th e beginning o f the Seleucid year could vary according to
the calendar o f the city. In the Julian calendar o f Antioch, the
year began on r Dios (1 October) (cf p. 25). The same epoch
72 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

was later used by Arab astronomers. T he Arabs called the Sclcucid

reckoning the era o f Alexander, though al Biruni recognized
this error.56
The Selcucid era remained in use in some parts o f the Near East
until modern times (Ginzel I, 263), and it was imitated by several
Oriental dynasties. Th e Arsacids in Parthia counted their years
from the spring o f 247 bc (Arsacid era, cf Kuglcr, II, 444),
though the Greek cities in the kingdom used the epoch o f the
autumn o f 248, but they also employed ‘the old style’ (ώς Se
7τρότ€ρον) o f the Seleucids.57 T he era o f the kings o f Pontus and
Bithynia (297/6) was also used in the Bosporan kingdom (cf
R. Fruin, Acta Orientalia 1934. 29; W . H. Bonnet, Historia 1961,
460). According to Syncellus the era began in 283/2 (cf G.
Vitucci, II regno di Bitinia (1953), l i)· Pharnaces I o f Pontus
counted the years from 337/6, the accession date o f Mithridates
o f Cius, the founder o f the dynasty, but his successor Mithridates II
changed to the era o f 297/6. E. Diehl, R E X IX , 1850; cf L.
Robert, Etudes anatolienttes 1937, 231.58
T he era o f Diocletian (erovs Δι,οκλητι,ανοΰ) can also be classed
among the dynastic reckonings. Diocletian introduced into
Egypt the dating according to the consular year, beginning on
i January. T he reform inconvenienced the astronomers, since
all astronomical observations were noted according to the
Egyptian mobile year. Thus, the astronomers continued, even
after Diocletian’s abdication, the fictitious numbering o f the years
o f Iiis reign, from 29 August 284-59 This era appears in horoscopes
(cf O . Neugcbauer and B. v. Hocsen, Memoirs o f the American
Philosophical Society 48 (1959))· From Egypt, the era came to the
W est thanks to its use for Easter calculations (cj. Ginzel I, 231;
Ambrosius, P L X V I, 1050); but its more general use remained
limited to Egypt from the sixth century a d . The Coptic church
still uses this reckoning.
The cities which w on independence from the Seleucids or
other monarchs started to use their own eras, which generally
commenced w ith the year o f liberation. Thus, a list o f officials
o f the city o f A m yzon has the heading: ol γεγονότος άφ' ου
Rapes ηλ€υθ*ρώθησαv (167 b c ) (cf L. Robert, La Carie II (i 954-)>
 CHRONOGRAPHY 73

309). For example, in 126/5 Tyre announced to other cities her

new independence (S E C II, 330), began her own era, and dis­
played her new status by issuing a new coinage. The earliest
examples o f such freedom eras are those o f Tyre from 275/4,
which probably celebrates the end o f the local dynasty (W . Ruge,
R E V IIA I. 1896), and o f Aradus in 259, which refers to indepen­
dence from the Seleucids (H. Scyrig, Syria 195r, 192). Th e so-
called Pompeian era (64 or 63 b c ) again refers to the liberation o f
the city in question from the Seleucids or the Maccabees. At
Antioch the ‘Pompeian’ era began in 66 b c (H. Seyrig, Syria
1954, 73 » 1959» 7°)· Sometimes cities agreed on a com m on era
(H. Seyrig, R N 1964, 37). O n the use o f the city eras o f Berytus,
Sidon and Tyre in Byzantine times, see H. Seyrig, Syria 1962, 42.
O n the era o f Edessa sec A. Maricq, Syria 1955, 278.
Th e so-called provincial eras, such as that o f Macedonia (148 b c ),
o f Achaca (146 b c ) , the ‘Sullan’ era (85/4) in Asia Minor, etc.,
counted the years o f Roman rule in the province or city in
question: (έτους) ά ‘Ρώμης (H. Seyrig, Syria 1959, 71)· hi Egypt,
Octavian’s conquest (κράτησις Καίσαρος θεοΰ υίοΰ) marked an
epoch (from 1 August 30 b c ) which lasted until the first years o f
Tiberius (U. W ilckcn, J R S 1937, 138; J. Bingen, C E 1964, 174).
Similar arc the eras which are counted from the date o f a
victory and were used by the Greeks o f Greece, Asia M inor and
Syria. Thus the eras o f Pharsalus (June 48 bc ) and Actium
(2 September 31 bc ) refer to the transfer o f domination from
Pom pey to Caesar and from A ntony to Augustus, respectively.
Com pare for example an inscription from Lydia which says:
έτους είκοστου καί πρώτον της Καίσαρος τοΰ πρεσβυτερου
αύτοκράτορος θεοΰ νείκης (=Pharsalus 48 B e ), τετάρτου Se της
Καίσαρος του νεωτερου αύτοκράτορος θεοΰ υίοΰ (= Actium 31
B e ), στεφανηφόρου δε καί ίερεως της ‘Ρώμης Άπολλωνίδου τοΰ
Αίσχρίωνος μηνός Δαισίου δωδεκάτηι.60
A ll the sacred eras belong in the same category, and ulti­
mately they all have as their model the era o f Actium.
The Jewish era from the creation o f the world starts on
6 October 3761 bc (cf Finegan, 126). Th e Byzantine creation era
began on 21 March 5508 bc , and later on 1 September 5509 bc .
74 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

After the attempts o f Hippolytus, Clement o f Alexandria and

others, the so-called Alexandrian computation o f the date o f
creation was w orked out: 25 March 5493 bc . Later in the seventh
century, the creation was placed in the year 5508 bc . The Eastern
church avoided the use o f the Christian era since the date o f
Christ’s birth was debated in Constantinople as late as the four­
teenth century (Grumel, 62).
The commemorative eras number the years tromsome historical
event. For example, in Paphlagonia years were reckoned ‘from
the twelfth consulate’ o f Augustus, that is, 5 bc (or, as the Paphla-
gonians reckoned, from 6/5 bc ). Likewise, for some time the
Athenians dated ‘from the visit o f the Emperor Hadrian’ in a d
126. The Manicheans reckoned from M aui’s birth (or death).61
T he NeoplatonistS computed the years from the accession o f
Julian the Apostate ( a d 361; cf Marinus, Vita Procli). The
Christian era o f incarnation, invented in a d 532 (cf p. 81) and
the Islamic era from Muhammad’s flight to Medina (from 15
June 622) are o f the same class.
It should bc noted, however, that many eras deduced by
modern scholars from the dates on coins are imaginary. Though
numismatists continue to develop ingenious theories about the
supposed Alexander’s eras in Phoenicia (cf e.g. I. L. Merker,
Americ. Numism. Soc. Notes X I (1964), 15), the dates so interpreted
on the coins o f A cco refer to the local rulers (E. T . N ewell, The
Alexander Coinage o f Acco and Sidon (1916), 59; G. Kleiner, Abh.
dcr Deutsch. Akad. Berlin (1947), 24). In Sidon and Aradus the
letters o f the alphabet were used to mark successive (annual?)
issues o f coins. After the twenty-fourth series, the counting began
again. In Sidon, the legend o f the scries ‘N ’ (= 13) changes from
‘Alexander’ to ‘Philip’. This change assigns this group to the year
324/323 (R. Dussaud, R N 1908, 450). In the scries ‘ 18' the name
‘Alexander’ is substituted for that o f Philip: Sidon fell into the
hands o f Ptolem y w h o did not issue coins in the name o f Philip
Arrhidaeus (cf E. T . Newell, ib. 36, whose chronology o f coins is,
however, incorrect). Again, the numbers on coins o f T yre (for
which numismatists invented imaginary eras) are misread (H.
Seyrig, Syria 1957, 93). Again, numismatists imagine that the coins
 CHRONOGRAPHY 75

o f Alexandria Troas bearing the dates from ‘ 137’ to ‘235’ attest

a city era from the renaming o f the city by Lysimachus, c. 300 b c .
H. v . Fritze, Nomisrna 19 Π , 27; A. R. Bellinger, Coins (Troy:
Supplementary Monographs 2) (1961), 93. In fact, we do not know
when the city received the name o f Alexandria (Strabo 13, 593),
and it is difficult to believe that she would count the years not
from the foundation by Antigonus but from the renaming date.
A n yw ay, an era ah urhe condita would be without any parallel
in antiquity (cf. p. 77). Alexandria Troas became a part o f the
Seleucid empire in 280, and thus it is probable that the city con­
tinued to use the Seleucid reckoning when she became indepen­
dent (cf p. 71). O n her coinage now c f L. Robert, Monnaies
antiques en Troade (1966). Similarly, an era o f Eumcnes II from
188 b c never existed; L. Robert, Villes de FAsie Mineure2 (1962),
253 ·
A third group o f eras was invented by scholars and mainly used
by historians. The disagreement between local calendars and
eponyms made it desirable to find a method o f dating w hich would
be understandable everywhere. The periodic Panhellcnic festivals
offered such a com m on time standard (cf Thuc. Ill, 8, 1 ; V , 49, 1 ;
and A. W . Gomme, Commentary on Thucydides II (1956), 258).
An inscription dates the appearance o f Artemis in Magnesia by
reference to die O lym pic year (140 O lym piad, first year), to the
Pythian games and to the Athenian archon o f the year (Syll. 557).
Greek writers, such as Pausanias, often use the reckoning accord­
ing to the Olympiads in order to date some event. This implies
the existence and use o f lists o f O lym pic victors. The first list was
published by the sophist Hippias. It was then kept up to date and
often re-edited. T he list for the Olympiads 1-249 has been
preserved in Eusebius’ Chronicle (ed. J. Karst, 89). Fragments o f
earlier catalogues are collected in FrGrH 414 ff.
T he numbering o f Olympiads was introduced by Timaeus or
by Eratosthenes. Other Panhellenic games, such as the Pythian,
were also sometimes numbered. Th e trustworthiness o f the earlier
part o f the list o f O lym pic victors, which begins in 776 b c , is
doubtful.62
From Eratosthenes on, all Greek chronology was based on the
76 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

Olympiads. A ll other datings were synchronized w ith the

Olympiads (cf. e.g. the dating o f Moses in Eusebius Pr. ev. x , 9).
The Byzantine chronographers continued to refer to the O ly m ­
piads. The documents, however, arc only rarely dated according
to this chronographic standard (cf e.g. Inschriften von Olympia,
530; A. Rehm, Didyma II, 214).
The counting o f the years within an Olym piad goes back to
Eratosthenes (FrGrH, Com m entary II, 707), but an O lym pic*
year per se did not exist: the games were held every four years
(776, 772 b c ; ad i , 5, etc.), alternatively after 49 and 50 months,
in midsummer at a full month (Samuel, 191). A more precise
date is not possible (cf. Ginzel II, 304; B. R. Sealcy, Class. Rev.
i960, 185). Chronologists equated each year o f the O lym pic
quadrennium w ith the corresponding Attic year, which also
began in the summer. It seems that the author o f the Parian
Chronicle in 264 b c already used this device (cf. FrGrH C om ­
mentary II, 670). N o one, o f course, had to count years o f an
Olym piad in conform ity w ith the Athenian calendar. Many
scholars used the Macedonian year which began in the autumn.
It seems that follow ing his sources Porphyry used now the Athen­
ian, now the Macedonian year (cf. FrGrFl ib. 855). Polybius’
flexible O lym pic year (p. 64) coincided roughly with the autum­
nal Achaean year: cf. F. W . Walbank, Historical Commentary on
Polybius l (1957), 3$; Samuel, 194.
Th e use ot the O lym pic years in chronography posed the
problem o f their equations with years expressed in some other
system o f datings. Thus, a Roman consular year, which from
153 b c began on 1 January, corresponds to parts o f tw o O lym pic
years. Thus, OI.180, 1 = 60/59 b c is equated in Diodorus with the
consular year 59 b c , in Dionysius o f Halicarnassus with the con­
sular year 60 b c . T he first method, which was also used by
Polybius, gives 775 b c as the epoch o f the Olympiads, while the
second, which w e generally follow, gives 776 b c as the starting
point o f the reckoning (cf. FrGrH II, 664; Ed. M eyer, Kleine
Schriften II (1924)* 288). Again, the use o f the Macedonian year
leads to the epoch 777 b c (cf. G. Unger, SBA 1895, 300; Ed.
Meyer, Forschungen II (1899), 446).
 CHRONOGRAPHY 77

Similarly, the conversion o f Athenian dates to Roman dates

and vice versa could bc done in tw o ways. Diodorus, for instance,
ends his chronographic list with the consular year 59 bc (cj. p. 91)
which for him corresponds to the archonship o f Herodus in
60/59 bc . On the other hand, Castor ended his work w ith the
year 61, yet he equated it w ith the archonship ofTheophcm us in
61/60 (cf. Leuze, 74; W . Kolbc, A M 1912, 107).
An era ah urbe condita (from the founding o f the city o f Rome)
did not, in reality, exist in the ancient w orld, and the use o f
reckoning the years in this w ay is modern. T he Romans used this
epoch only to measure time distance from it to some subsequent
event: for example, L ivy IV, 7 says that the consular tribunes
came 310 years after the founding o f the city (cf III, 30, 7;
VII, 18, 1). Similarly, an inscription states that Nerva restored
liberty ‘848 years after the founding o f the city’ (Dessau, 274).
This mode o f relative dating was already used in the Roman
Republic. For instance, an inscription o f Puteoli (Dessau, 518) is
dated ‘90 years ah colonia deducta (that is, 105 b c ) (cf Dessau, 157;
genio municipii anno post Interanuiani conditam 704). Relative datings
o f this kind are incorrectly called ‘eras’. Consequently, modern
scholars speak o f the ‘era o f Tanis’, referring to an Egyptian
inscription with the mention o f ‘400 years o f the city o f Tanis’
(K. Sethe, A Z 1930, 85; R. Stadelmann, C E 1965, 46). J. v.
Beckcrath, Untersuclnmgen zur politischen Gcschichte der zweiten
Zwischenzeit in Aegypten (1965), 153. H. Goedicke, C E 1966, 23.
C f Numb. 13, 22: Hebron built seven years before Zoan (Tanis).
Th e principal reason for not using the system ab urbe condita
was that the age o f the city was disputed: est enim inter scriptores
de numéro annorum controversia (Cic. Brut. 18, 72). Th e date o f
the founding in Roman historiography— excluding the more
extreme opinions (for example Cincius Aiimentus in Dion. Hal.
I, 74: 729/8 bc )— oscillates between 759 and 748 bc . For a long
time the Polybian date o f 751/0 served as a norm for Cicero, Livy
and Diodorus (cf. Perl, 20); then Atticus in his Liber annalis
m oved the founding back to 753 bc (Cic. Brut. 18, 72). This
date was taken up and popularized by Varro. The list o f the
magistrates o f the Republic compiled under Augustus (Fasti
78 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

Capitolini) indicates the years ab urbe condita, w hich arc, how ­

ever, counted from 752 b c . Tradition established the festival
o f Parilia on 21 April as the birthday o f Rome. Thus a year ab
urbe condita which ran from 2 r to 20 April corresponded to
parts o f tw o consular years, and its identification w ith one o f
them depended on the chosen system o f conversion (cf. Leuze,
2 5 2 ).

INDICTION

The number o f an indiction shows the position o f the year within

a cycle o f 15 years: a d 312-326, etc. The cycles themselves are
not numbered, so that the number o f the indiction is usually
used only to relate to another dating system. This kind o f time­
reckoning was introduced in a d 312 (Chronicon Paschale) and
became obligatory for the dating o f documents from a d 537
(Justinian Novel. 47).
Indiction ( = ‘declaration’, ίνδικτ&ον, ίπινέμησις) originally
referred to the announcement (indictio) o f the compulsory
delivery o f foodstuffs to the government (annona), an obligation
which under Diocletian became the cornerstone o f the Roman
fiscal system. A t first the term was used only w ith reference to
taxation (cf. U . W iicken, A P F 1911, 256). Thus, e.g., in a d 368
a village had to pay 44,617 denarii, κατά τον τύπον της ια ίνδικ-
[τίοωνο?] (Wiicken, Chrest. 281). Th e population knew the tax
year better than the official consular date. Accordingly, from the
second half o f the fourth century on, the indiction appears in all
kinds o f documents, for instance in a petition to offer to rent
3 arourai ‘for sowing them in the 10th year o f this prosperous
indiction’ (W iicken, ib., 380). Th e indictions, however, were not
numbered. For Julian equivalents o f the years within an indiction,
from a d 312 on, see R E I, 666.
The origin o f the indiction cycle and its meaning remain
unknown. In Egypt the fiscal period o f 15 years was in use from
a d 297 (cf. W iicken, A P F XI, 313; Grumel, 192).

The year o f indiction generally began on 1 September, but in

Egypt it varied according to the date o f the tax announcement
in the summer. (For the table, sec F. Hohmann, Zur Chronologie
 CHRONOGRAPHY 79

der Papyrusurkunden (1911), 40.) Thus, June o f the 14th indiction

in Egypt fell in the 15th indiction o f Constantinople (P. M .
Meyer, Juristische Papyri (1920), no. 52 (o f a d 551)). In the West,
the inclusion o f indictions in the Easter Table o f Dionysius
Exiguus (1cf. p. 81) made this time reference popular. I11 the chaos
o f medieval datings this one was at least stable (cf. J. E. W . Wallis,
English Regnal Years and Tables (1921), 9)· Reckoning by indiction
continued to be used by the Supreme Tribunal o f the H oly
Roman Empire until the dissolution o f the latter in 1806, and is
still carried on in some modern calendar tables, for instance in
H. Lietzmann’s Zeitrechnung (1934), w h o gives indictions from
a d 298 until a d 2000.

T he conversion rule for an indiction number is to add 3 to the

year number o f the Christian era and divide the sum by 15. The
remainder gives the indiction number o f the year. Th e Byzantine
dates from the Creation arc to be divided by 15 (O . Seeck, RE
IX, 1330; Ginzel II, 148).
 C H A P T E R III

APPLIED C H R O N O L O G Y

T he know ledge o f ancient calendars and dating systems must

in principle enable us to convert the dates o f our sources into
units o f our reckoning. This is generally possible for the ancient
datings expressed in terms o f the Julian year. According to our
sources, Caesar was murdered on the Ides o f March in the year
when Caesar was consul for the fifth time and Antony was his
colleague. According to the consular list the year C . Caesare V
et M . Antonio consulibus corresponded to 44 bc . The Ides o f
March corresponded to 15 March. Caesar, thus, was killed on
15 March 44 bc .
Th e same, or almost the same, certainty can be obtained for
dates o f the Babylonian cyclical calendar (see p. 24), and for
Egyptian calendar dates— i f the Julian year is know n (cf. p. 40).
For instance, a letter dated 2 Mcsorc, year 29 (o f Ptolemy II), was
written on 22 September 257 b c (P. Cairo Zen. $9096). For Greek
history and Roman pre-Julian dates, except for some particular
cases (for instance, the astronomically fixed dates), w e must be
satisfied w ith establishing the Julian year and the approximate
season o f the event in question.
For the Near East, the margin o f error rapidly increases when
w e go back beyond c. 900 b c . Until the fourteenth century, in
the most favourable cases, the margin w ill be about ten years
and more; until the seventeenth century, about fifty years, and
still earlier, about a hundred years. For the pre-literate period we
have no historical dates, but must rely on the archaeological
chronology (see p. 11).
Approxim ate as our knowledge may be, we must know how
it is obtained. H o w do we get the equation between the ancient
and our ow n datings? T o answer this question w e have first to
understand the origin o f our tim e reckoning.
 APPLIED C H RO NOL OGY 8ï

PRINCIPLES OF REDUCTION

T he Church required Easter to fall on the first Sunday after the

spring full moon, that is, the first full moon after 21 March. This
necessitated computation o f the Easter cycles and tables. In ad
525, Dionysius Exiguus was asked by Pope John I to compile a
new table. He used the table o f the church o f Alexandria which
employed the era o f Diocletian (see p. 72), but being unwilling
to reckon from the reign ‘o f an impious persecutor’, he chose ‘to
note the years’ from the Incarnation. In his table, the year 532
ab incarnatione followed the year 247 o f Diocletian (PL LXVII,
493). Accepted by the Sec o f Rome, Dionysius’ table was revised
again and again, for instance by Bede in 725 (PL X C , 859), and
served the Roman Catholic church up to the introduction o f the
Gregorian calendar in 1582. W ith Dionysius’ Easter computa­
tion, the West also adopted his era. For instance, the era o f
Incarnation was already used by the author o f the Computatio
Paschalis compiled in ad 502.63 Thus, our reckoning simply
continues a Roman one. Therefore, all ancient datings which
directly or indirectly can bc related to the counting o f the years
o f Diocletian can also be converted into Julian dates.
Secondly, the dating according to the Roman consuls was still
used in the fifth century, and Dionysius himself wrote his w ork
consulatu Probi iunioris (ad 525). The aforementioned Computatio
Paschalis gives the equation a d 562= year 21 post consulatum
Basilii. As we have the complete fasti o f the Roman consuls for
1,050 years from Brutus and Collatinus to the aforementioned
Basilius, we can easily assign Julian years to each o f them, pro­
vided that die ancient dates arc trustworthy (cf p. 69).
Third, w e have the so-called ‘Ptolemaic’ canon, the list o f
kings preserved in Thcon’s commentary on Ptolemy’s astro­
nomical work. Com posed by Alexandrian astronomers for their
ow n calculations, this list, based on the Egyptian mobile year,
begins with the accession o f the Babylonian king Nabonassar
on 27 February 747 bc . It gives astronomically exact dates o f
successive reigns (Babylonian, Persian, Ptolemies, the Roman and
Byzantine emperors), and in some manuscripts the list is con­
82 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

tinued until the fall o f Constantinople in 1453. Here again, modern

chronology is linked directly to an ancient system o f reckoning.64
If, for example, w e want to know which was the first year o f
Diocletian’s rule (which in itself does not have to be identical
w ith the beginning o f the era o f Diocletian), the Chronicon
Paschale tells us that he was proclaimed emperor on Γ7 September
under the consulate o f Carinus II and Numerianus. From the
fasti consulares w e get the corresponding Christian year, a d 284.
Pctavius proceeded in this manner in 1627. Idclcr, instead, made
use o f an astronomical observation which is dated synchronisti­
cally: 81 years from D iocletian= 1,112 years from Nabonassar
(that is, from the beginning o f the Canon o f the Kings); the
equation gives a d 284 as the first year o f Diocletian’s rule. In
order to fix the first year o f the emperor, Scaligcr (De emendatione
temporum, V) in 1582 established that the Coptic church, in con­
tinuing to calculate the era o f Diocletian, equated a d 1582 (from
29 August) with the 1299th year o f Diocletian. In other words,
all Roman dates, i f they are complete and reliable, can be directly
expressed in Julian years. A ll the other datings o f ancient chron­
ology arc linked to our reckoning by direct or indirect syn­
chronisms w ith Roman dates. For instance, the Egyptian
chronology is based on the list o f the Pharaohs, made by Manetho
under Ptolemy II (FrGrHy no. 609). His list contains the reigns
o f Persian kings, beginning with Cambyses, who ruled in Egypt
and w ho also appear in the Royal Canon. In this w ay a correspon­
dence with Roman chronology is obtained. Ancient Indian
chronology depends on the date o f K ing Asoka, in whose edict
five Hellenistic kings are mentioned (Antigonus Gonatas, etc.).
W e can date these kings, thanks to Roman synchronisms.
Accordingly, the approximate date o f Asoka can bc established
(P. H. L. Eggermont, Chronology o f the Reign o f Asoka (1956)).
W here the link to Roman chronology is broken, we grope
vainly for certitude. Take, for example, Egyptian chronology.
The aforementioned king-list o f Manetho has been preserved
only in Christian summaries (FrCrH, 609). As w e have seen,
the mention o f Persian rulers allows us to connect his list with
Roman reckoning. T he references to later Pharaohs in Babylon-
 APPLIED C H RO N O L O G Y 83
ian texts and astronomical data in Egyptian documents confirm
the general reliability o f Manctho’s list for the N ew Kingdom
and later dynasties up to the sixteenth century b c (M. Alliot,
JN E S 1950, 2 11; R. A . Parker, Revue d'Egyptologie 1952, 101).
Y et the exact datings before c. 800 are rarely obtainable. The
accession o f Ramesses II is dated by various cgyptologists to 1304,
1292, or 1279 b c .65
Manetho’s figures for the period o f anarchy between the
M iddle and the N ew Kingdom (c. 160 years) and for the first
intermediate period between the O ld and the Middle Kingdoms
(c. 900 years) are, however, unreliable. Thus, the link with
Roman chronology is twice broken. A papyrus letter states that
Sirius will rise on 16. VIII o f the year 7. T he king in question is,
in all probability, Scsostris III, or it may bc Amenemmes III, his
successor (XII Dynasty). Secondly, the rise o f Sirius is not observed
but predicted— that is, calculated— 21 days in advance. W c do
not, however, know how. The Julian date o f the event is c. 1880.
Thus, we know that the XII Dynasty reigned from c. 2000 to
c. 1800. T he royal canon preserved in a Turin papyrus (Gardiner,
47) gives a total figure o f 995 years for the O ld Kingdom until
the end o f the VI Dynasty. Assuming that the figure is exact, w e
still do not know the length o f the interval between the VI and
the XII Dynasty. According to Manciho, the first Pharaoh,
Menés, ruled from 4242 (V. Struve, Vestnik Drevnei Istorii 1946,
fasc. 4, 9). The most recent estimates vary between 3100 and
2800. Yet, according to the same astronomical and historical
dates, Mènes was also placed toward the end o f the fifth millen­
nium (1cf. L. Borchardt, Qiiellen und Forschungen zur Zeitbestim-
tming der àçyptischen Geschiclite II (1935), 117). W e cannot disprove
this hypothesis. W e can only say that archaeological considera­
tions suggest that it is best to accept the shorter chronology and
not to throw the X II Dynasty back to the fourth millennium
(Gardiner, 66).66
Assyro-Babylonian chronology is based on the Royal Canon
which begins with the Babylonian king Nabonassar. Th e king-
lists which go down to Nabonassar would in principle allow us to
convert all Assyro-Babylonian royal datings into Julian ones; but
84 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

these lists are often unreliable. Th e Assyrian scribes, for instance,

suppressed some kings w ho w ere later considered usurpers (B.
Landsberger, J C S 1954, 101). T he compilers also made successive
some dynasties which were contemporary with one another. The
regnal years were already counted c. 2500 in the Sumerian city
o f Lagash (M. Lambert, R H i960, 24; for Larsa, cf F. B. Kraus,
Z A 1959, 136). But this dating system came into com m on use
only under the Kassitc dynasty. Before this time, all years received
official names which referred to some event marking the year. If,
for example, we say that Rimsin o f Larsa was defeated in the year
31 o f Hammurabi, this means that the date-formula ‘year in
which Hammurabi destroyed Rimsin* received the 31st place in
the Babylonian list o f year names in the reign o f Hammurabi.
The Assyrians dated by annual eponyms. For instance, an original
document o f K ing Esarhaddon, found in his palace, is dated by
the magistrate (limmu) o f the year (= 6 76 b c ). But elsewhere in
the second and first millennia the time-reckoning b y regnal
years prevailed.
The fixing in time o f the famous Babylonian legislator,
Hammurabi, on whose dating many others depend [cf D.
Edzard, Die zweite Zwischenzeit Babyloniens (1957), 15), illustrates
the inherent difficulty o f w orking with king-lists. Hammurabi
was a king o f the I Dynasty. A Babylonian king-list goes down
from the I Dynasty to Kandalanu o f the Royal Canon (647-626).
Thus, we have here a link to Roman chronology. Th ough the list
is damaged and includes the II Dynasty (o f the Sealand on the
Persian G ulf), w hich apparently never reigned over Babylon, it is
possible, by using the dates o f this list, to place Hammurabi in
the second half o f the twentieth century b c [cf. Ed. M eyer, Die
atteste Chronologie Babyloniens, Assyriens und Àgyptens (1931), 1).
Y et, recently discovered documents prove that Hammurabi was
contemporary w ith Shamshi-Adad I o f Assyria, who, according
to the Assyrian list, reigned in the second half o f the eighteenth
century. Should w e bring Hammurabi down or m ove Shamshi-
Adad up? T he rather fluid chronology o f the Pharaohs and the
Hittites and vague archaeological inferences led recent scholars
to suggest 1792-1750 or 1728-1686 as the most probable dates o f
 APPLIED C H RO NOL OGY 85
Hammurabi. O ther scholars prefer to place him in 1848 or even
c. 1900. As a matter o f fact, the Assyrian kings themselves disagree
w ith each other and w ith the information supplied by the royal
list when they state the interval between a given king and some
predecessor.67
T he Royal Canon is also basic for Greek chronology, together
w ith a chronographic fragment from Eratosthenes (FrGrH, 241
F 1), in which arc given the intervals between the main events o f
Greek history until the death o f Alexander (dated in the Canon
o f Kings): ‘From the fall o f T ro y to the return o f the Heraclids
80 years, from here to the Ionian colonization (Ionian migration),
60 years, then until the guardianship ofLycurgus, 159 years, from
here to the beginning o f the Olympiads, 108 years; from the
1st Olym piad to the campaign o f Xerxes, 297 years; from here to
the beginning o f the Peloponnesian Wars, 48 years, and until the
end o f these wars and o f the Athenian hegemony, 27 years, and
until the battle ofLeuctra, 34 years; from this time to the death
o f Philip, 35 years, and, finally, until the death o f Alexander,
12 years.’
In this w ay it is possible to say that the beginning o f the
Peloponnesian W ar was in 431 b c . Furthermore, Thucydides
mentions the O lym pic games (for instance) in the twelfth year
o f the war (V, 49). Because the distance o f the Peloponnesian W ar
from the first O lym piad is also established by Eratosthenes, the
date o f the Olym piad 1/1 is 776 b c ; this is confirmed by Cen-
sorinus, w ho equates the consular year Ulpii et Pontiniani ( a d

238) w ith the 266th Olym piad.

Let us now take another example. Diodorus (XI, 1, 2) places
the expedition o f Xerxes in the first year o f the 75 th Olym piad,
when Calliadcs was archon in Athens and Sp. Cassius and P.
Verginius consuls in Rome. T he consular date seems to give a
direct link to Roman chronology. But according to the Roman
fasti, Sp. Cassius and Verginius were consuls in 486 BC. This
disagrees wich the Greek dating. In fact, the name o f the Athenian
archon Calliades is already given b y Herodotus (VIII, 51), w h o
also states that the battles were fought in the time o f the O lym pic
games (VII, 206). OI.75, 1 is 480/79 b c , the same year o f the
86 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

archon Calliades. Diodorus made a mistake in his Roman

synchronism (cf Perl, τοό).
In this way, by means o f reciprocal controls o f synchronization
and with the help o f astronomy, the founders o f modern chron­
ology, J. Scaliger (1540-1609) and D . Pecavius (1583-1652),
calculated the fundamental dates, which, in turn, permitted the
conversion o f other dates. Petavius, in Rationarium Temporum II,
presents the material which justifies the currently accepted equa­
tions between ancient datings and the Julian years.
The references to celestial phenomena, particularly the eclipses,
allow us to control the systems o f ancient chronology since their
dates can bc calculated astronomically, and thus, independently
o f the said system. Th e solar eclipse, which occurs during the
period o f the new moon, is observable only from that part o f
the earth on which the m oon’s shadow falls. The lunar eclipse,
which can occur only at full moon, is visible everywhere. The
eclipses recur in the same sequence within the period o f 233 lunar
revolutions, that is, every 18 years and 11 days (F. Boll, R E VI,
2338). Thus, the approximate date o f the observation must be
know n in order to identify the phenomenon w ith an eclipse o f
the astronomers. Therefore, it is not possible to date w ith certainty
the solar eclipse seen by Archilochus (frag. 74 D ), generally
thought to bc that o f 6 April 648 b c . Th e observations o f Venus
made under K ing Ammizaduga o f the first Babylonian dynasty
have been preserved. But since the same phenomena recur every
56 years on approximately the same dates in a lunar calendar, the
observations can as w ell agree w ith the dates 1977 or 1581 b c ,
for the first year o f Ammizaduga. (C f J. D . W eir, Venus Tablets
o f Ammizaduga (1971), 12; E. Huber, B O 1974, 86; R. Reiner and
D . Pingree, Venus Tablet o f Ammizaduga (1975).) Again, the Julian
dates o f Sirius (p. 41) would differ by several years according to
the place o f observation; c f E. Hornung, A Z 1965, 38. O nly
historical evidence allows us to choose the right historical date.68
However, as soon as the cyclic period to which an observation
belongs is known, astronomy can date the phenomenon with
absolute precision and therefore establish with certainty a whole
series o f dates. Thus, for example, Assyrian chronology is pinned
 APPLIED C H RO NOL OGY 87

down by the mention o f the solar eclipse which occurred on

15 June 763 bc in the list o f the eponyms o f Assur. Th e disputed
dates o f the scientist Heron o f Alexandria (c. a d 62), o f the
astrologer Vettius Valens (c. a d 152-162) and o f the astronomer
Clcom edes (c. a d 370) were established by modern recalculation
o f celestial phenomena mentioned by these writers (O. N euge-
baucr, 178; id. H T R 1954, 66; id. AJPh 1964, 418).69 T he
beginning o f the Peloponnesian W ar in 431 is confirmed by
Thucydides’ reference (II, 28, 1) to an eclipse which actually
occurred on 3 August 431 b c . Mithridates V I o f Pontus died in
63 b c , as a Roman synchronism (Pompey’s march to Petra)
shows. According to our sources, he reigned fifty-six years and
was thirteen years old at accession. This gives 133 and 120 b c
respectively as the dates o f his birth and accession (Plin. X X V , r,
6; Memnon, 32). According to Justinus (X X X V II, 2) brilliant
comets shone in the year he was begotten (134 b c ) , and in the
year he became king (120 b c ) . In fact, Chinese sources record the
appearance o f comets in 134 and 120 b c (cf. Fincgan, 242). T he
Julian year o f the battle at Thermopylae is fixed by the reference
to the O lym pic and the archon year. Polyaenus (1, 32, 2) mentions
‘the rising o f a star’ before Leonidas’ battle. I f he means the hero
o f Thermopylae, and i f this star is Sirius, the battle must have
been fought c. 1 August (J. Labarbc, B C H 1954, 1; id. Revue
Belge de Philologie 1959, 69). Th e seasonal occurrence o f the
flooding o f the N ile can help to establish the date o f Pom pey’s
death (D. Bonneau, R E L 1961, 105).

CHRONOGRAPHY

Hellanicus o f Lesbos was the first who, in the time o f the

Peloponnesian W ar, attempted to adjust various systems o f
chronological references to a com m on standard, namely to the
years o f the priestesses o f Hera in Argos. Following his example,
later Greek savants prepared synchronistic tables. Since Timaeus
and Eratosthenes, these tables were generally based on the reckon­
ing o f the Olympiads. Castor o f Rhodes (c. 60 bc ) added Roman
and Oriental datings. Using the w ork o f their predecessors, the
Christian ch/onographers put secular chronography into the
88 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

service o f sacred history. A w ork o f this kind, the ‘Canon’ in the

second part o f Eusebius’ Chronicle, composed c. a d 300, was
translated by Jerome and continued until 378. Jerome’s compila­
tion became the standard o f chronological knowledge in the
West. J. Scaliger, the founder o f modern chronological science,
aimed at reconstructing the w ork o f Eusebius.
The Canon gives a continuous series o f synchronisms. T he years
after Abraham (1 A b r.= 2 0 i6 b c ) , w ith whom for Eusebius all
reliable chronology began, arc equated with the royal years,
the Olympiads, etc., and events are mentioned under their
respective dates. For instance, the birth o f Christ is mentioned
under the year 2015 o f Abraham, which was also the 25th year
o f the reign o f Augustus and fell into the 184th Olym piad, that
is, incidentally, the year 2 b c according to our reckoning, which
goes back to Dionysius Exiguus (see p. 81).70
T he datings o f Eusebius, often transmitted incorrectly in manu­
scripts, are o f little use to us today, except in a few cases where
no better information is available {cf. ρ. τι). However, a modern
‘Eusebius’, a w ork which w ould adequately summarize the
present state o f applied chronology, is still lacking.71 W e must
realize that w e cannot establish our ow n handy chronological
tables except on the basis o f tables, lists, and so on, prepared by
the ancients themselves, w ho in turn were handicapped b y the
absence o f the standard time-reckoning. Under 45 b c , a con­
temporary chronicler notes: annus or[ditiatione Caesaris] mutatus
{‘Fasti Ostienses\ ed. L. Vidman, Rozpravy o f the Czechoslovak
Academy LX V II, 6 (1957))· Y et the introduction o f the Julian
year alone could not standardize chronology, particularly since
the Julian year itself began at different dates in each country. In
England and its American colonics, the year began on 25 March
until 1752. T w o examples m ay illustrate the difficulties which
confronted a chronologist even after the introduction o f the
Julian calendar. For instance, Porphyry v/as a specialist in
chronological research; yet in his biography o f Plotinus he had
to use the regnal years o f the Roman emperors. Thus, Porphyry’s
reader would have needed some handy tables o f chronology to
understand his datings. Y et the imperial years were not identical
 APPLIED CH RONOLOGY 89

with the Julian years, and the reader w ould not know which
form o f the Imperial year was used by the author (c f R. W altz,
R E A 1949, 41; M. J. Boyde, CPh 1937, 241). Errors were un­
avoidable. Jerome, a chronologist himself, writing after a d 374
congratulates a certain Paul on his hundredth birthday (Ep. ad
Paulum). Yet elsewhere (De viris ill. Ill, 53) he states that Paul
knew personally Cyprian o f Carthage w ho had died in a d 259.
M ani used the Babylonian form o f the Seleucid era (from 311 bc ),
and we have information com ing from various sources about his
life and death. Y et these sources disagree about his chronology,
though he lived in the third century a d . This lack o f certainty
in the matter o f chronology made it possible for the Sassanid
traditions to reduce the period from Alexander to the Sassanids
from 557 to 226 years. Th e Jews also allotted only 52 years to the
Persian period o f their history, though 206 years separate Cyrus
from Alexander.72
Ancient historians often had to use different systems o f dating
concurrently since they were unable to unify the references they
had found in their sources. See e.g. W . den Boer, Mnemosyne
1967, 30 on Herodotus; O . M orkholm , Antiochus I V o f Syria
(1966), 196; J. Goldstein, I Maccabees (1976) 24.

PRACTICAL SUGGESTIONS

In ancient (and medieval) chronology w e use the Julian calendar

and not the Gregorian one which is used now . Both coincide
c. a d 300; but then the Julian dates run behind the Gregorian
calendar by three days every four hundred years. I11 the reverse
direction, from c. 100 b c , the Julian year is in advance o f the
Gregorian calendar by three days every four hundred years, so
that e.g. 29 December 102 b c (Gregorian) was already 1 January
101 b c Julian (cf. p. 10).
In using ancient datings given in era or regnal years, w e must
take into account tw o possible pitfalls. First, the beginning o f the
year was not standardized but left to local choice. For instance,
the A ctium era began on 23 September 31 bc at Philadelphia, but
in 32 b c at Amisus (Μ. N. Tod, A B S A X X III (1918-19), 212).
90 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

Similar were the variations for the Macedonian and Actium eras
in Greece (F. Papazoglou, B C H 1963, 517).
T he regnal years o f the Achaemcnids began in the spring for
the Babylonians, in the autumn for the Egyptians, and were
probably counted from the accession day by the Persian court (cf
Thuc. VIII, 58). Further, each city in the same realm for various
reasons could count the regnal years differently from one another
and from the court reckoning (H. Seyrig, R N 1964, 58).
Again, the numbering o f regnal years docs not need to agree
with history. Charles II o f England actually became king on
29 M ay 1660, but his regnal years were counted from the death o f
Charles I on 30 January 1649. Ancient rulers, too, could for various
reasons antedate the beginning o f their reigns (cf. E. J. Bickcrman,
Berytus 1944, 77). O n the other hand, a disputed succession could
confuse the scribes. T w elve years after the death o f Philip
Arrhidaeus, in 305 b c , a cuneiform document was dated: King
Philip, year 19’ (Isid. Levy, Jourtt. Asiat. 1952, 269).
W e use the standard Julian years and reckon them backward
‘before Christ’. This reckoning postulates a zero year between
the dates ‘ b c ’ and ‘a d ’. But such a year is lacking in our compu­
tation. This point is to bc kept in mind when calculating the
intervals between events before and after Christ. The simplest
method is to use the astronomical convention: 1 BC=year 0;
2 b c = i , and so on. For example w e ask how old Augustus was
when he died in a d 14. He was born in 63 bc . Thus the equation
is: 6 3 - 1 = 6 2 ; 62 + 14 = 76. In fact, Augustus died 35 days
before reaching his 76th birthday (Suet. Aug. 100).
Th e lack o f the zero era in Christian reckoning also explains
the conversion rule for the era years. For instance, the first year
o f the Seleucid era (o f Macedonian style) is 312/n b c . This means
that the zero year for this era is 313. Thus, to obtain the Julian
year corresponding to a Seleucid year for the pre-Christian period,
w c have to subtract the number o f the Seleucid year from 313.
For instance, year 200 Sel. = 313 -2 0 0 = 113 bc and year 312 Sel.
is 313 - 3 1 2 = 1 b c . But year 313 Sel. is ad i . Accordingly, for the
post-Christian years o f the Seleucid era, the number o f the
Julian year o f the epoch (312) is to be subtracted from the number
 APPLIED CH RONOL OGY 91
o f the Seleucid year. Thus, 522 Sel.=522 - 3 I 2 = ad 210 or rather
i October 2 1 0 -3 0 September ad 21 i .
T he lack o f the zero year also explains the rules for the con­
version o f the number o f an Olym piad. For the period b c , that
is, up through OI.194, the number o f the Olym piad is reduced
b y one, multiplied by four, and the product is subtracted from
77 6. Th e result gives the Julian year b c i n which the games were
held, that is, the first year o f the O lym piad in question. For
example, wliat is the Julian year o f the 180th Olym piad? The
operation is as follows: 1 8 0 - 1 = 179; 1 7 9 x 4 = 7 1 6 ; 7 7 6 - 7 1 6
= 6 0 b c , or, more precisely, 60/59. This is die first Julian year
o f the 180th Olym piad.
O n the other hand, for the period a d , that is from the 195th Ol.
on, the number o f the given Olym piad is again to be reduced by
one, the result multiplied by four, and 775 to be deducted from the
product. For instance, Eusebius’ Chronicle names the O lym pic
victors up to the 249th O l. inclusively. N o w , 249 - 1 =248 ; 248 x 4
= 992; 9 9 2 -7 7 5 = 2 1 7 . Julius Africanus gave a catalogue o f the
winners in O lym pic games until his time, that is a d 217. Eusebius,
w ithout saying so, a century later reproduced Africanus’ list (cf.
Ed. Schwartz, R E VI, 1378). But using ancient datings expressed
in terms o f O lym pic years, we should not forget the possible
variations in synchronization : the source m ay have equated
O l. 180, i, not w ith 60/59 b c , but w ith 61/60 b c , and so on (see
p. 76). T o put it bluntly: anyone trying to convert an ancient
dating into one expressed in terms o f our reckoning should
remember the legal m axim : caveat emptor.
 ABBREVI ATIONS

JOU RN A LS AND COLLECTIONS

ΑΒΑ Abhandhingen dcr Baycrischcn Akademie

ABSA Annual of the British School at Athens
AFO Archiufiir Orientforseining
AGGG Abhandhingen der Gottinger Gelchrten Gescllschaft (Akademie)
AJA American Journal of Archaeology
AJPh American Journal of Philology
AM Mittcilungen des Deutschen Archaeologischen Instituts, Athenischc
Abteilung
APAW Abhandhingen der Preussischen Akadetnie der Wisserschaften
APF Archiufiir Papyrusforschung
A RIV Archiufiir Religionswissenschaft
ArchOr Archiu Orientalt
ASAA Anmiario della Scuola Archcologica d'Atene
Ath Athenaeum
AZ Zeitschriftfur Aegyptische Sprache
BASOR Bulletin of the American Schools of Oriental Research in Jerusalem
and Baghdad
BCH Bulletin de Correspondance Hellénique
B IF A O Bulletin de Γ Institutfrançais d'archéologie orientale
BO Bibliotheca Orientalis
BSL Bulletin de la Société de Linguistique de Paris
BSLL Bulletin de la Société des Lettres de Lund
BSOAS Bulletin of the School of Oriental and African Studies
CAH Cambridge Ancient History
CAH2 Cambridge Ancient History /-//, New Edition
CE Chronique d'Égypte
CIL Corpus Inscriptionum Latinarum
CPh Classical Philology
CR Comptes Rendus de l'Académie des Inscriptions
DWA Denkschriften der Wiener Akademie
FrGrH F. Jacoby, Fragmente dcr griechischen Ilistoriker
GDI Sammhing der griechischen Dialekt-Inschriften
GGA Gottingische Gelehrte Anzeiger
GGN Nachrichten der Gottingischen Gelchrten Gesellschaft
Hesp. Hesperia
94 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

HSCPh Harvard Studies in Classical Philology

HTR Harvard Theological Review
HUCA Hebrew Union College Annual
IE] Israel Exploration Journal
IG Inscriptiones Graecae
JAOS Journal of the American Oriental Society
JBL Journal of Biblical Literature
JCS Journal of Cuneiform Studies
JEA Journal of Egyptian Archaeology
JHS Journal of Hellenic Studies
JNES Journal of Near Eastern Studies
JO AI Jahreshefte des Ôsterreichischen Archaeologischen Instituts
JQR Jewish Quarterly Review
1RS Journal of Roman Studies
MOI Mitteilungen des Orientalischen Instituts (Academy o f Berlin)
OGIS Orientis Graeci Inscriptiones Selectae ed. W . Dittcnberger (1903-5)
OLZ rOrientalistische Literaturzeitung
PAAJR Proceedings of the American Academy of Jewish Research
PAPhS Proceedings of the American Philosophical Society
Phil Philologus
PL Patrologia Latina ed. Migne
RA Revue archéologique
RAss Revue d'Assyrologie
RE Pauly-Wissowa, Real-Encyclopadie der classischen Altertums-
wissenschaft
REA Revue des études anciennes
REG Revue des études grecques
REJ Revue des étudesjuives
REL Revue des études latines
RH Revue Historique
RH R Revue de Γ histoire des religions
RLA Reallexikon der Assyriologie
RLAC Reallexikonfur Antike und Christentum
RN Revue Numismatique
RhM Rheinisches Museum
RPh Revue de Philologie
SBA Sitzungsberichte der Bayerischen Akademie
SCO Studi classici e orietitali
SEG Supplementum Epigraphicum Graecum
SH A W Sitzungsberichte der Heidelberger Akademie
SOAW Sitzungsberichte der ôsterreichischen Akademie ( Wien)
SPA W Sitzungsberichte der preussischen Akademie der Wissenschaften
Syll Sylloge Inscriptionum Graecarum, 3rd ed (1915-24) ed. W.
Dittenbcrger
 ABBREVIATIONS 95

TAPhA Transactions of the American Philological Association

YCS Yale Classical Studies
ZA Zeitschriftfiir Assyriologie
ZDMG Zeitschrift der Dcutschen Morgenlàndischen Gesellschaft
ZN TIV Zeitschriftfiir die neutestamentliche Wissenschaft
ZPE Zeitschriftfiir Papyrologie und Epigraphik

BOOKS
Busolt- G. Busolt and H. Swoboda, Griechische Staaiskunde I—II
Swoboda (1920-26)
Deg rassi A. Degrassi, Inscriptiones Latinae Liberae Reipublicae (1957—<53)
Dessau H. Dessau, Inscriptiones Latinae Selectae (1892-1916)
Dinsmoor W . B. Dinsmoor, The Archons of Athens in the Hellenistic
Age (1931)
Fincgan J. Finegan, Handbook oj Biblical Chronology (1964)
Gardiner A. Gardiner, Egypt of the Pharaohs (1961)
Ginzel F. K. Ginzel, Handbuch der Chronologie I—III (1906-14)
Grumcl V. Grumcl, La Chronologie (1958)
Idclcr L. Idcler, Handbuch der Chronologie I-II (1825)
Idelcr, L. Ideler, Lehrbuch der Chronologie (1831)
Lchrhuch
Jacoby F. Jacoby, Atthis (1949)
Kubitschek W . Kubitschek, Grundriss der antiken Zeitrechnung (1928)
Kuglcr F. X. Kuglcr, Stemkunde und Sterndienst in Babel I-II and
Suppl. I-III (1907-35)
Langdon S. Langdon, Semitic Menologies (1935)
Lcuze O . Leuze, Romische Jahrzahlung (1909)
Mcritt B. D. Mcritt, The Athenian Year (1961)
Meyer Ed. Meyer, Forschungen zur Alien Geschichte ( 1892-9)
Mommsen Th. Mommsen, Romische Chronologie (1859)
Mommsen, Th. Mommsen, Romisches Staatsrecht (1887)
RStR
Ncugebaucr O . Ncugebaucr, The Exact Sciences in Antiquity (1957)
Nilsson Μ . P. Nilsson, Primitive Time-Reckoning (1920)
Nilsson, Μ . P. Nilsson, Die Enstehung und religiose Bedeutung des
Kalender griechischen Kalenders (1918)
Perl G. Perl, Kritische Outersuchungen zu Diodors romischer
Jahrzahlung (1957)
Pritchett W . K. Pritchett, Ancient Athenian Calendars on Stone (1963)
Pritchett- W . K. Pritchett and O . Ncugebauer, The Calendars of
Neugcbauer Athens (1948)
Robert J. and L. Robert, Bulletin épigraphique (REG)
Samuel A. E. Samuel, Greek and Roman Chronology (1972)
Tod Μ . N. Tod, Selection of Greek Historical Inscriptions (1946)
 NOTES
1 There is no adequate, full-scale treatment o f ancient chronology. L. Ideler,
Handbuch der Chronologie I—II (i 825—6) and his shorter Lehrbuch der Chronologie
(183:), though outdated, offer even today die best over-all picture. F. K.
Ginzel, Handbuch der Chronologie I—III (1906-14), useful as a collection of
material, though often at second hand, is also antiquated. For Greece and
Rome sec A. E. Samuel, Greek and Roman Chronology. Calendars and Years
in Classical Antiquity (1972). For comparative chronology see Μ . P. Nilsson,
Primitive Time-Reckoning (Skrifter of the Humanistika Vetettskapssamfunder i
Lund, 1920). For current bibliography cf. L'Année Philologique s.v. Calendaria,
and for Greece sec J. and L. Robert, Bulletin épigraphique in REG. For Egypt
sec J. Janssen, Annual Egyptian Bibliography, 1947 ff. Yearly bibliography on
the Near Eastern chronology can be found in the journal Orientalia.
2 R. van Compcmollc, Études de chronologie et d'historiographie siciliotes.
Institut historique belge de Rome. Études . . . d'histoire ancienne V (i960);
J. Boardman, JHS 1965, 5; Molly Miller, The Sicilian Colony Dates (1970).
O n the uncertainty o f typological dating cf. e.g. J. Moreau, Die Welt der
Kelten (1958), 132.
3 D. R. Brothwell, E. S. Higgs, G. Clark (ed.), Science in Archaeology (2nd cd.
1970); S. Fleming, Dating in Archaeology (1977)· The radio-carbon dating is
particularly important for prehistory, but for various reasons, e.g. the vari­
ations o f the disintegration rate o f C-14, the radio-carbon date may widely
disagree with the true date. CJ. Trevor Watkins (cd.), Radiocarbon Calibra­
tion and Prehistory (1976) and CAH I, 1, s.v. Radiocarbon. For current
information about dating techniques in archaeology, consult relevant
articles in Antiquity. For recent estimates of prehistoric chronology cf. G.
Clark, World Prehistory (2nd ed. 1969) and CAH I, 1 (1970).
4 O n our own calendar sec, e.g., P. Couderc, Le Calendrier (1961). For Baby­
lonia, our sources (in addition to information from ancient historians, which
is incorporated in the works o f Ideler and Ginzel, and documents) also
include astronomical records. The following are basic works: F. X. Kugler,
Sternkunde und Sterndienst in Babel, I—II (1907-24) and Suppl. I—III (1913—35) ;
O. Neugebaucr, Astronomical Cuneiform Texts [1955); A. Sachs, Late
Babylonian Astronomical Texts (1955). C f O. Ncugebauer, The Exact Sciences
in Antiquity (1957), 97, and JNES 1945, 1. For Egypt cf. p. 40.
Among other ancient peoples, those of Western Asia generally followed
the Babylonian system (p. 24); the calendars of the western lands (Gaul,
Spain and Germany) are not known well. On the Celtic calendar, cf.
P. M. Duval, La vieprivée en Gaule (1952), 342. Id. Mélanges Carcopino (1966),
295. On Germans cf. Ginzel III, 55.
 NOTES 97

5 Here and often elsewhere, Gcminus is quoted in the English translation of

Sir Thomas L. Heath, Greek Astronomy (1932).
6 The Egyptian hours: K. Sethe, GGN 1920, 106; L. Borchardt, Aegyptische
Zeitmessung (1920) ; Neugebauer, 82; J. Lauer, B1FA i960, 171. For Baby­
lonia cf. F. Thureau-Dangin, RAss 1930, 123; 1932, 133; 1933, 151; id.
Osiris 1 9 3 9 , 112; B. L. van der Waerden, JNES 1949, 18; 1951, 25. For
Greece and Rome cf G. Bilfinger, Die antiken Stundenangaben (1888); id.
Der biirgerliche Tag (1888). On clocks and sundials cf A. Rehm, RE VIII,
2416; M. C. Schmidt, Antike Wasseruhren (1912), H. Diels, Antike Technik
(1924), 157. Waterclocks in Egypt: S. Schoch, Abhandl. Akad. Mainz 1950,
no. 10, 908. For Babylonia cf S. Smith, Iraq 1969, 77. On sundials cf S.
Gibbs, Greek and Roman Sundials (1976)· On portable and multiple sundials
cf E. Büchner, Chiron 1971, 457, R. J. Bull, BASOR 1975, 29. On the use of
minutes cf P. Tannery, RA 1895» 359- On the survival of variable hours cf
G. C. Lewis, Historical Survey of the Astronomy of the Ancients (1862), 242.
On the introduction o f sundials in Greece cf D. R. Dicks, JHS 1966, 29.
7 C f e.g. F. Hiller von Gartringcn, Inschriftcn von Priene (1906) no. 112, line 60:
Ζ Θ η κ ^ ν Be to dAei/ χ / χ α ά π ό α ν α τ ο λ ή ς ή λ ι ο υ 8d η μ έ ρ α ς μ έ χ ρ ι π ρ ώ τ η ς τ η ς
νυκτός ώ ρ α ς . . .
8 Ο . Schlicsscl, Hermes 1936, p. 104; Ο . Ncugebaucr, SOAIV 240, 2 (1962),
p. 27.
9 L. Ideler, Über astronomische Beobachtungen der Alien (1806), 20; O. Neuge­
bauer and H. B. Van Hosen, Greek Horoscopes, Memoirs of Amer. Philos.
Soc. 48 (1959), 95. The same kind o f instrument was used in Athenian courts
in order to give the same amount o f time to the accuser and the defendant.
Cf. Busolt-Swoboda II, 1161.
10 0 n Egyptian equal hours cf. O. Neugebauer, Egyptian Aslronomical Texts I
(i960), 119; Neugebauer, 81, 86; on Babylonian counting of hours cf the
papers o f F. Thureau-Dangin quoted above, note 6.
11 R. Pfeiffer, State Letters of Assyria (Amer. Orient. Series VI, 1935). 298; H.
Sauren, Actes de la XVIIème Rencontre Assyrologique (1970), 13. On direct
observation o f the moon cf B. Z. Wacholder and D. B. Weinberg, HUCA
1971, 136.
12 C. Schoch, in S. Langdon and J. K. Fotheringham, The Venus Tablets of
Ammizaduga (1928), 97. For Athens see Ginzel I, 93.
13 Kuglcr, Suppl. III (1935)» 2 5 5 ·
14 R. Pfeiffer (note 11), no. 303.
15 C f Kugler II, 301; II, 232; Suppl. I, 136, 175, 186.
16 C f L. W . King, Letters and Inscriptions of Hammurabi III (1898), 12; A. L.
Oppcnheim, Lettersfrom Mesopotamia (1967), 100.
17 N. Schneider, Zeitbestimmung der Wirtschaftsurkunden der III Dynastic von Ur
(1936). Cf. F. Thureau-Dangin, RAss 1927, 181. The calendar was an
instrument o f State economy. The Sumerian administration began the
fiscal year after the delivery o f new barley to granaries and the settlement o f
98 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

relevant accounts, i.e., about two months after harvest. For other purposes
the year began before or after harvest (cf. Kuglcr ÏI, 3 0 Γ; Y. Rosengarten,
Le concept sumérien de consommation (i960), 410). For Mari cf. M. Birot,
Archives royales de Mari XII, 2, p. 20. Consequently, the same month could
have several names in the same city; e.g. it might bc called the month of
sheep-shearing, when the account concerned sheep (cf B. Landsbergcr,
JNES 1949, 262, 273; Rosengarten, op. cit., 423). C f Nilsson, Kalcnder, 73.
18 R. A. Parker and W . H. Dubbcrstein, Babylonian Chronology 626 bc- ad 75
(Brown University Studies XIX; 1956). As R. A. Parker kindly informs me,
his diagram o f intercalated years has to bc corrected as follows: not 492-1
but 500-499 was intercalated. C f G. Cameron, JNES 1965, 181. Cf. also
D. Sidersky, Étude sur la chronologie assyro-babylonientte, Mémoires présentées à
l'Acad. des Inscriptions 13 (1920), 115; id. RAss 1933, 68 (the Julian dates o f 1
Nisanu). On the 8-year cycle and, from 499 bc . the 19-year cycle in
Babylonia cf. B. L. van dcr Wacrden, AFO 1963, 97. But the latter cycle
was followed without deviation only from 380 bc: on (Ncugebauer, 140).
This cycle ‘is quite accurate; only after 310 Julian years do the cyclically
computed mean new months fall one day earlier than they should’ (Neugc-
bauer, 7). C f also note 20 and T. Heath, Aristarchus of Samos (1913)» 293.
19 R. Labat, Hémérologies et ménologies assyriennes (1939), 25. C f id. MIO 1957,
229. The nature o f the pre-Babylonian calendar of the Assyrians is uncertain.
The problem o f the Assyrian calendar is still insoluble (cf Μ . B. Rowton,
CAHl, i, 229). O n the Assyrian calendar in Cappadocia cf. N. B. Jankowska,
ArchOr 1967, 524. O n the Elamite calendar cf R. Reiner, AFO 1973, 97.
20 E. Mahler, Handbuch der jiidischen Chronologie (1916) is out o f date. On
Biblical time-reckoning cf. R. de Vaux. Ancient Israel (1961), 178 ; Fincgan.
Cf. my review BO 1965, 184; J. van Goudever, Fêtes et calendriers bibliques3
(1967); H. N. Smith, TheJewish New Year Festival (1947); A. Caquot, RHR
191,1 (determination o f the new moon). The names o f four Hebrew months
are recorded in Scripture (cf. A. Lemaire, Vêtus Testamentum (1973), 243).
O n the Gczcr calendar cf S. Talmon, JAOS 1963. 177; John C. L. Gibson,
Textbook of Syrian Semitic Inscriptions I (1971)· 11-
On the modem Jewish calendar sec Maimonidcs, Sanctifications of the New
Moon (Yale Judaica Series XI, 1956); B. Zuckermann, Materialen zur alten
jiidischen Zcitrechnung. Jahresbericht der jüdisch-theoJogischen Seminars in
Breslau, 1882; D. Sidersky, Étude sur l’origine astronomique de la chrono­
logie juive, Mémoires près, par divers savants à FAcad, des Inscr. XII, 2 (1916);
id. Études sur la chronologie assyro-babylonienne, ib. XIII (1916), 140. It is a pity
that none o f later writers on Jewish chronology Hiscusses. or even knows,
the material collected and interpreted in Ed. Schwartz, Christliche und
Jiidischc Ostcrtafeln, ^4 GGG N.F. VIII, 6 (1905), 121. In the present
Jewish calendar the 19-year cycle is longer by about two hours than 19
solar yc9iTs(Jewish Encyclopaedia III, 501). Accordingly, the Jewish New Year
now disagrees by roughly one week with the sun (W. M. Feldman, Rabbinic
 NOTES 99

Mathematics and Astronomy (1931), 207). Sec also S. Powels, Der Kalender
der Samaritaner (1977), 25.
21 On the calendar used in the Elephantine documents cf. D. Sidcrsky, Revue
des étudesjuives 1926, 59; L. Borchardt, Monatsschr. fiir Geschichte desJuden-
tums 1932, 299; R. A. Parker, JNES 1955, 71. Cf. also M. Lidzbarski,
Ephemerisfiir Semitische Epigraphie II, 221 : an Iranian uses the same calendar
in Persian Egypt. On the same calendar used by the Persian administration
at Pcrsepolis sec R. T. Hallock, Persepolis Fortification Tablets (1969), 74.
Here the intercalation doubled the sixth month (cf E. J. Bickennan, ArchOr
1967, 197). For equations o f Babylonian and Egyptian months in late
Egyptian texts: F. Hintzc, MOI 3 (1955) Η 9 ί R· A. Parker, A Vienna
Demotic Papyrus (1959), 30.
22 On the Seleucids cf E. J. Bickerman, Institutions des Séleucides (1938) 144 and
205. The existence o f the official calendar did not prevent cities from
inserting names o f particular months: cf e.g. OGIS, 233 (‘Pantheon’ at
Antioch in Persia); L. Robert, RPh 1936,126 (‘Antiocheion in Stratonicea).
For the Parthians, cf W . W . Tam, CAH IX, 650; G. Lc Rider, ‘Susc sous
les Seleucids et les Parthes’, Mémoires de la mission archéologique en Iran
XXXVIII (1965), 35. C f E. J. Bickerman, BO 1966, 328.
23 C f J. Johnson, Dura Studies, Thesis, U. o f Pennsylvania, 1932; Dura-
Europos, Preliminary Reports VII-IX (1939), 309; C. B. Welles, Eos 1957,
469; Samuel, 143.
24 On the calendar o f the people o f the Dead Sea Scrolls cf]. M. Baumgarten,
JBL 1958, 249; S. Talmon, Revue de Qumran i960, 474; J. A. Sanders, The
Psalm Scroll of the Cave II (1965), 91; M. Limbeck, Die Ordnung des Heils
(1971), 134. On the schematic calendar in the Book o f Enoch cf. O. Neuge-
bauer, Orientalia i960, 60. The calendar quarrels berween the Jews and the
Karaites arc very instructive for the understanding of the similar disagree­
ments. C f Z. Ankori, Karaites in Byzantium (1959).
25 Biruni, Chronology oj the Ancient Nations, tr. E. Sachau (1879), 68, states that
the Jews began to use the precalculated calendar about two hundred years
after Alexander (that is, c. year 20 0 o f the Seleucid era, or c 1 10 bc ). This bit
of information cannot be disproved or proved. It is possible that the calendar
schemes were changed several times in Jerusalem, but it is also possible that
Biruni reproduces an argument used in the polemics between the Jews and
Karaites.
26 Μ . P. Nilsson, Die Entstehung und die religiose Bedeutung des gricchischen
Kalendcrs, in Lunds Univers. Arsskrift, N.F. XIV (1918); 2nd ed. in Scripta
Minora o f the K. Humanistika Vctenskapssamfundet i Lund, 1960/61. On dates
in pre-Homenc documents cj. J. Chadwick, The Mycenean World(1976), 9 7 *
191; Samuel, 64. For Homer cf. E. Buchholtz, Die homerischen Realien j
(1871), 33. Hesiod’s calendar is entirely seasonal, that is, agricultural ( Theog.
58), and the change o f seasons is marked by rising and setting o f stars. The
mention o f die month Lenaion (v. 504) is interpolated (cf. Samuel, 66; D. R.
100 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

Dicks, Early Greek Astronomy (1970), 25). On the subdivision o f the month
in Homeric hymns and Hesiod cf. T. W . Allen, W . R. Halliday, E. E. Sikes,
The Homeric Hymns2 (1936) ad H. Merc. 19; H. L. Lorimcr, BSA 1951, 806.
27 Μ . P. Nilsson, Geschichte der griechischen Religion I2 (1955), 644. C f F.
Jacoby, Atthis (1949), 287. The arguments adduced for the very early use of
the 8-year cycle (Ideler, Lehrbuch, 116; Nilsson, RE 17, 2387), namely, the
celebration o f the Olympic games alternately in 49 and 50 months, and of
the Pythian games every eight years from 656 until 583 (Sch. Horn. IL X,
252; Sch. Pind. Ol. Ill, 33) are o f little value. C f J. L. Fothcringham, JHS
1919, 176. According to Censorinus, the octaeteris was devised by Cleo-
stratus o f Tencdos, who lived after Anaximander (Plin. N.H. II, 8, 31), that
is, after c. 550, C f D. R. Dicks, JHS 1966, 26.
28 The counting o f days within a decade could vary. For example, in Argos,
the ninth day o f a month was called ή ν ά τ α π ρ ά τ α , the seventeenth ε β δ ε μ ά τ α
μ έ σ α , the twenty-sixth έ κ τ α δ ε υ τ ά τ α (A. Boethius, Dcr Argivische [Calender
(1922), 64, and cf Samuel, 91). In Athens, the first day was called ν ο υ μ η ν ί α ,
the days from the second to the tenth Scvrepa ( τ ρ ί τ η , etc.) ί σ τ α μ έ ν ο υ , the
days o f the second decade π ρ ώ τ η (etc.) έ π ι δ έ κ α , the twentieth elκ ά ς , and
the last day o f the month έ ν η κ α ί ν έ α (‘old’ and ‘new’). For the last decade,
progressive numeration was used in documents from the time of Alexander
the Great on: the twenty-first was δ ξ κ ά τ η υ σ τ έ ρ α , the twenty-second
δ ε ύ τ ε ρ α μ ε τ ' έ ι κ ά δ α ς , and so on. On the other hand, until the end ot the
fourth century bc , retrogressive numeration (φ θ ί ν ο ν τ ο ς ) was common. Cf.
e.g. Aristoph., Nubes, 1131 and 1134: π έ μ π τ η , τ ε τ ρ ά ς , τ ρ ί τ η , μ ε τ ά τ α ύ τ η ν
δ ε ύ τ ε ρ α , . . . ε ύ θ υ ς μ ε τ ά τ α ύ τ η ν ε σ θ ' έ ν η τ ε κ α ί ν έ α . Thus, in a full month
we have to subtract the number o f the given Greek days from 31 to End the
date o f our notation. As to the hollow month, the position ol the leap day is
still debated. It was dekate phthinontos, that is, the ‘21’ according to Meritt,
38; id. Historia 1962, 441; id. Hesperia 1964, 1, who refers to Schol. Arist.,
Nubes, 1131. Cf. Samuel, 60; B. D. Meritt,AJPh 1974, 264; W. K. Pritchett,
California Studies in Classical Philology 1976, 181. Curious was the notation
o f days for the last decade in Rhodes, at least in the second century ad
(IG XII, i, 4): the last day o f the month was always called triakas. The day
before the last, the pro(triakas), was omitted in the hollow month. Then days
from 28 to 22 were counted backward, from 30th, so that our 22nd day was
*29’, our 28th day *23’, but our 21st day was *21*.
29 On the term ε μ β ό λ ι μ ο ς cf W . Vollgraf, Mnemosyne 1916, 49; Meritt,
TAPhA 1964, 200 ff.
30 W . K. Pritchett and O. Neugebauer, The Calendars of Athens (1948); B. D.
Meritt, The Athenian Year(1961); W . K. Pritchett, Ancient Athenian Calendars
on Stone (1963); id. The Choiseul Marble (1970); Meritt, PAPhS 115 (1971)»
97 offers a new reconstruction o f the Athenian calendar from 432 to 401,
which is inevitably as uncertain as were the previous attempts.
 NOTES ΙΟ Ι

31 Cf. B. Keil, Hermes 1894, 61; Mcritt, 60; W . K. Pritchett, AJPh 1964, 40.
O n IG I, 304 b, cf id. BCH 1964, 4 5 5 ; id. Hesp. 1965. 131.
32 C f D cm . 3, 4; 19, 57; 21, 86; 24, 26; 37, 6; 42, 5; 49, 6; 49, 22. See A.
Mommsen, Chronologische Untersuchungen (1883), 143.
33 J. K. Fotheringham (JHS 1919, 172) was probably the first scholar to state
that Ge minus refers to the cycles propounded by astronomers which were
never adopted by the cities. As a matter of fact, the Athenians did not even
have a fixed leap month. C f W . K. Pritchett, CPh 1968, 53.
34 G. Daux, BCH 1963, 603. C f M. Jameson and S. Dow, ib. 1964, 154, 180;
S. Dow, Historia i960, 270; S. Dow and R. F. Healey, Sacred Calendars of
Eleusis (1965); J. D. Mikalson, The Sacred and Civilian Calendar of the
Athenian Year (1975).
35 The equations o f the summer solstice o f 27 June 432 bc and of 26 June 106
bc with 13 and 14 Skirophorion respectively given in the Milesian parapegma
(sec p. 58) probably concern the same ‘ideal’ astronomical calendar. B. L.
van der Waerden, JHS i960, 170 and 180.
36 A. E. Samuel, Ptolemaic Chronology (1962). C f also Samuel, 145. Julian dates
o f the Ptolemies: T. C. Skeat, The Reigns of the Ptolemies (1954); id. JEA
i960, 91; 1962, 100; A. E. Samuel, Études de Papyrologie IX (1964), 73;
P. W . Pcstman, Chronologie égyptienne d'après les textes démotiques (1967).
For the reign o f Ptolemy II cf L. Kocncn, Eine agonistische Inschrift aus
Àgypten (1976). O n the financial year see J. Bingen, CE 1975, 239.
37 R. A. Parker, The Calendars of Ancient Egypt (1950); Ed. Meyer, Àgyptische
Chronologie, in APAW , 1904, and 1907; A Z 1907, 115; Ed. Meyer, Chrono-
logie égyptienne (1912). K. Sethe, GGN 1919, 287-319; ib. 1920, 28-55 and
97-141; S. Schott, Aegyptische Festdaten, Ahhand. der Mainzer Akademie,
1950. For the conversion o f Egyptian dates, E. Lundsgaard, Aegyptischer
Kaletuler derJahre 3000-200 v. Chr. (Copenhagen, 1942). For the conversion
o f the Egyptian dates into Egyptian Julian dates (cf p. 50) cf B. L. van dcr
Waerden, Isis 1956, 387; M. Chaîne, La chronologie des temps chrétiens de
l'Égypte et de l'Ethiopie (1923).
38 We do not even know to what level the waters o f the Nile had to rise in the
third millennium bc before the Egyptians considered the flood as having
begun. Furthermore, the visibility o f the rising o f Sirius is uncertain.
L. Borchardt and P. W . Neugebauer, OLZ 1924, 370.
39 On the Sassanian calendar cf S. H. Taqizadeh, Old Iranian Calendars (1938) ;
M. Boyce, BSOAS 1970, 513; id. in J. de Menascc, Troisième livre du
Denkart (1972), 262; V. Lifshitz, in Russian translation of the present work
(1975), 320; and Bickerman’s chapter on Chronology in Cambridge History
of Iran III (forthcoming). O n the Armenian calendar cf Ginzel I, 314. The
Chorczmian calendar: V. Lifshitz, Acta Antiqua 1968, 4 3 5 · The Cappadocian
calendar is known only in its Julian form (cf p. 50), and its functioning
remains uncertain. C f Ginzel, RE X, 1917; K. Hanncll, BSLL 1931/2, 22.
40 Our knowledge o f the Roman calendar comes from two different sources:
102 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

from rhc living tradition and from ancient writers and documents. We still
follow the Caesarian calendar, and the system o f Roman dating (Nones and
Kalends) was used until the sixteenth century (Ginzel III, 115). Among the
basic sources arc Macrobius, Sat. (I, 13) and Ccnsorinus (De die natali,
written in ad 238). In addition (excluding numerous lesser passages in
different writers, etc.) we have stone calendars, among them one of the
pre-Julian year [Fasti Antiatcs vctcrcs: A. Dcgrassi, Inscriptions Latinae
liberae reipublicae (1957) no. 9); id. Inscriptioties latinae XIII, 2 (1963); F.
Maggi in / 1/fi Pontifie. Accademia di archeologia, Ser. Ill, vol. IX, 1 (1972).
Among modem studies o f the Roman calendar, Mommsen’s Romische
Chronologie2 (1859) remains basic and unsurpassed More recent surveys:
A. K. Michels, The Calendar of the Roman Republic (1967) and Samuel, ch. V.
Cf. also F. Della Corte, Antico calendario dei Romani (1969).
41 Ginzel II. 243; G. Wissowa, Hermes 1923, 392; L. van Johnson, AJPh 1959,
133; A. Magdelain, REL 1962, 201; A. K. Michels, Hommages à Albert
Grenier (1962), 1174. O n the linguistic aspect of dating, Ginzel II, 175 ;
A. H. Salonius, Zur romischen Danerung, in Annales Acad. Scient. Fenicae,
Ser. B, XV (1922). In the Republican period the inclusive calculation was
not used for counting the years: J. Bcaujeu, REL 1976, 329. Cumbersome
as was the Roman counting o f the days, it was sometimes used by Romans
even in Greek cities: cf. L. Robert in Laodicea du Lykos (ed. J. des Gagniers)
(1969). 325·
42 Sec Μ . P. Nilsson, in Festskrift Per Persson (1922) \} = Opuscula II (1951),
979; H. J. Rose, Primitive Culture in Italy (1926), 88. For further conjectures
about the pre-history o f the Roman calendar cf K. Hancll, Das Altromische
eponytn Amt (1946), 9 9 ; J· Hu beaux, Rome et Veies (Bibl. Fac. Phil, et Lettres,
Univ. Liège CXLV, 1958), 66; L. V. Johnson, TAPhA i960, ι ο ί ; id. AJPh
1963, 28; E. Gjerstadt, Acta Archaeologica 1961, 193; G. Radke, RhM 1963,
313; R. Werner, Der Beginn der romischen Republih (1963); Michels (supra
note 40), 121 ; Samuel, 165. On the Etruscan calendar cf K. Olszscha. Glotta
I 9 5 4 < 71; J· Heurgon, JRS 1966. 1. On other calendars in Italy cf. J. W.
Whatmotigh, HSCPh 1932.
43 G. Dc Sanctis, Storia dei Romani III (1916), Index s.v. Calendario, and IV, I
(1923). 368. Cf. also M. Hollcaux, Û.tudes d'épigraphie IV (1952), 336 ; V
(1957). 24; P. Meloni, Latomus 1954, 533. For some recent suggestions on
Julian equivalents o f the Roman pre-Julian calendar, cf. e.g. R. Dcrov,
Phoenix 1973, 348, ib. 1976, 265; Antiquité classique 1976, 265 (covering the
period 290-168 bc); M. Morgan, Chiron 1977, 89 (First Punic War); P.
Marchetti, Antiq. Class. 1977, 473 (the years 203-196); id. BCH 1976, 411
(168 bc); M.-Th. Rapsact-Charlicr, Historia 1974, 278 (59-45 bc).
4 4 O n the limits o f autumnus sec Ph. Fabia, REA 1931, 122. On three and four
seasons in Greece, cf G. M. A. Hanfmann, The Season Sarcophagus in
Dumbarton Oahs (1951). On observing the movement of stars cf K. Sethe,
CGN 1919, 291; R. W . Slolcy, JEA 1931» 166; Ncugebauer, 84; id. in
 NOTES 103

Hypsikles, ed. V. de Falco, M. Krause, ^4 GGG LXII (1965). O n the Greek

computers’ cf. D. de Solia Price, Gears from the Greeks, PAPhS 64 (1974)» 7 ·
45 The natural year: Nilsson, Kalender 21. The seasons: Ginzel II, 182, 308, and
passim. Observation o f the stars and meteorological forecasts: A. Rehm,
RE, Suppl. VII, coll. 175-198. For the question o f how much the sky was
really observed, cf H. Vogt, SHAW, Abh. 1 (1920), 54; R. Boeker, RE,
Suppl. IX, 1610 if. Further cf Aristotle, Hist. Animal., ed. A. L. Peck, II,
p. 383 (Loeb Classics). On ‘seasons’ in Thucydides cf O. Lushnat, RE,
Suppl. XIV, 1134; D. P. Orsi, Quadcrni di storia (1975)» H 7 · Plato, too,
seems to know only two seasons: cf. A. D. Nock, Gnomon 1934, 290. On
the Roman natural year cf. J. E. Skydsgaard, Varro th Scholar (Analecta
Romana Inst. Danici, Suppl. IV, 1968), 45. On the natural year in Egypt
and Mesopotamia cf. A. M. Bakir, The Cairo Calendar (1974)» P· Labat, Le
calendrier babylonien des travaux, des signes et des mois (1965). For the dates of
the most important phases o f the stars for antiquity see Ginzel II, 517 and
Table 11 below.
46 The zodiac: F. Cumont in Dictionnaire des Antiquités V, 1046; B. L. van der
Waerden, AFO 1953, 216; Neugebauer, 140; H. F. Gundel, R E X , 462. A
cuneiform text o f about 1500-1000 bc mentions some zodiacal signs: E. F.
Weidner, Syria 1956, 180. The division o f the ecliptic into twelve equal
signs by Babylonian astronomers is already attested in the early fifth century:
B. C. A. Aabe and A. Sachs, Centaurus 1969, 1. On the symbolism o f the
zodiac cf. W. Hartner, JNES 1965, 1. Parapegmata: A. Rehm, RE XVIII 4,
col. 1295; Pritchett and van der Waerden, BCH 1961, 31; A. Wilhelm,
Epitymbion H. Swoboda (1927), 144 ·
47 Cf. Mommsen, 309; E. Diehl, Inscriptiones Latinae Christianae (1928), III, 3II.
48 Latest survey and bibliography: E. Lohse, in Theologisches Worterbuch zum
Neueti Testament VII (i960), 1; F. H. Colson, The Week (1926); S. Eriksson,
Wochentagsgotter, Mond und Tierkreis (1956). On the planetary week and its
spread, see also Nilsson, Geschichte der griechischen Religion II (1950)» 4 6 7 ;
H. Gundel, RE XX, 2143 ; F. X. Doelger, Antike und Christentum VI (1941),
252; S. Gandz, PAAJR XVIII (1948-9), 213; H. Ingholt, ‘Parthian Sculp­
tures’, Memoirs of Connecticut Academy XII (1954), 40; A. Degrassi, Atti del III
Congresso Intemazionale di Epigrafia (1959), 104. The Jewish week: H. and J.
Lcwy, HUCA XVII (1942/3)» 1. O n the Nundinae cf W. Lintott, Classical
Quarterly 1968, 189. O n the market-days in the Roman empire cf. R.
McMullen, Phoenix 1974, 3 3 3 · O n die Roman agricultural week, cf J.
Heurgon, REL 1947, 236.
49 For the Marmor Parium and similar texts, see FrGrH, nos. 239 and 252.
Jacoby, 160; cf. D. W . Prakkcn, Studies in Greek Genealogical Chronology
(i 9 4 3 ).
50 Mommsen, RStR, 896; H. Mattingly, JRS 1930» 78; R- P. Longen,JRS
1931,131 ; M. Hammond, The Antoninc Monarchy (1959), 72. ForJulian day-
dates o f accession, etc., o f the emperors cf. L. Holzapfel, Klio 1912, 1913,
104 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

1918, 1921 (partly out o f date). O n the terms etos and eniautos cf. Μ . P.
Nilsson, Eranos 1957» 115. C f the terms chronos and tempus, meaning ‘a
year’: E. Lofstcdt, Late Latin (1959), 117. For the naming of a year cf
Archives royales de Mari XIII (1964), no. 47: a year was first called ‘King
Zimri-Lim dedicated a throne to god Dagon.’ But when the throne was not
ready, another name for the year had to be found. On the accession dates of
Roman emperors cf Mommsen, RStR, II, 2, 796; F. dc Martino, Storia della
costituzione romana IV, 2 (1974)» 171; J. Béranger, Recherches sur l'aspect
idéologique du Principal· (1953), 102.
51 Sec A. Gardiner, JEA 1945, 11 ; W . Helck, Analecta Biblica 1959, 113 ;
Gardiner, 71 ; J. Ôzemy,JEA 1964, 58. For Babylonia sec note 67. On regnal
years o f Hebrew kings: Finegan, I94;jepscn, R. Hahnhardt, Untersuchungen
zur israelitisch-jiidischen Chronologic (1964)· Bar Kochba’s years were counted
from 1st Nisan: cf B. Kanacl, IEJ 1971, 411.
52 Some (often hypothetical) lists o f eponyms of Greek cities outside Athens
may be cited here. Alexandria (under the Ptolemies): W . Pcremans and E.
Van’t Dick, Prosopographia ptolemaica III (1956) ; J. Ijscwijn, De sacerdotibus
Alexandra . . . et Lagidarum eponymis (Verhandelingen von de K. Vlaamse
Academie XLII, 1961). Bocotia: P. Roesh, Thespieset la confédération béotienne
(1961), 84; R. Etienne and D. Knocpf, Hyettos de Béotie (1976), 349 (for the
period250-171). Delphi: G. Daux, Chronologiedelphique (1943); E. Manni,
Ath 1950, 88. Delos: F. Durrbach, Inscriptions de Délos II (1929). 327·
Miletus: A. Rchm, Didyma II, 380. Rhodes: F. Hiller v. Gartringen, RE,
Suppl. V, 835; Chr. Blinkenberg, Lindos II (1941); L. Morricone, Annuario
della scuola archeologica in Atene (1952), 27, 351. Sparta: Samuel, 238.
Thessaly: A. M. Babakos ( Μ π α π ά κ ο ς ) Praxeis koines diatheseos . . . kata to
dikaion tes archaias Thessalias (1962), 255. See also W . SchonBeder, Stadt- und
Bundesbeamten des griechischen Festlandes, Diss. Leipzig (1917); R. Munstcr-
berg, Beamtennamen aufgriechischen Miinzen (1917) = Wiener Numismatische
Zeitschrift 1911 ff.
53 O n the Athenian archon lists cf Jacoby, 169. On the archons before 480 bc
cf. T. J. Cadoux JHS 1948, 70; Samuel, 195; R. Meiggs and D. Lewis, A
Selection of Greek Historical Inscriptions (1969), no. 6.
54 See T. R. S. Broughton, The Magistrates of the Roman Republic I— II and
Supplement, 1951. A. Degrassi, I Fasti Consolari dcll'Impero Romano dal 30
a.C. al 613 d.C. (1952). The consular fasti of the Republic have come down
to us in three editions o f the Augustan Age. (a) The Fasti Capitolini set up in
the Forum between 36 and 30 bc . The text has been partly preserved; its
gaps can bc filled up with help o f later sources, such as the Chronographcr
o f the Year ad 354, the so-called Fasti Hydatani, compiled in ad 468 and the
Paschal Chronicle compiled in Greek in ad 630. (b) Livy, and for the lost
parts o f his work, a list o f consuls in Cassiodorus* Chronicle, published in
ad 519. (c) The Roman eponyms for 486-302 bc in Diodorus XI-XX. Cf
Ed. Meyer, Kleine Schriften II (1924)» 288. The consular lists of the afore­
 notes ro5
mentioned chronographers are in Chronica Minora Ι - Ι Π , ed. Th. Mommsen
(1892-8).
55 Sec in general Mommsen; id. Romische Forschungen II (1879), p. 151; O.
Lcuzc, Romische Jahrzahlung (1909); E. Pais, Ricerche sulla storia del diritto
pubblico di Roma II (1916); K. J. Bcloch, Romische Geschichte (1926); K.
Hancll, Dûs altromische eponyme Amt (1946); A. Degrassi, Fasti et Elogia
(1947); id. Fasti Capitolini (1954); G. Perl, Kritische Untersuchungen zu
Diodors romischerJahrzahlung (1957).
56 See L. Idclcr, Über astronomische Beobachtungen der Alien (1806), 256. S. H.
Taqizadeh, BSOAS X (1942), 129.
57 On the Arsacid calendar cf. p. 25 and note 22. O n the Arsacid era in Baby­
lonian documents cf. J. Oelsner, in Altorientalische Forschungen, Π Ι 1976, 25;
in Hatra: J. Tcixidor, Syria 1966, 93 ; 1973, 414 · The Seleucid era continued
to bc used in Babylonia, particularly by astronomers and thus for the dating
o f important events. For instance, the Manicheans stated that Mani was
boni on 8th Nisan (14 April) o f the Seleucid year 527 (ad 216). His first
revelation is similarly dated to 1 April 228 and the second one to 19 April
240. Cf. L. Koenen, ZPE 1972, 249. The Seleucid era continued to be used
in Christian Syria: cf L. Bernhard, Die Chronologie der Syrer, SBWA
263, 3 (1969). n o .
58 E. Minns, Scythians and Greeks (1913). no. 646, note 17. Cf. G. Perl in
Studien zur Geschichte und Philosophie des Altertums (ed. J. Harmatta) 1968,
299 (era o f Bithynia, Pontus, and kingdom o f Bosphorus.)
59 As a matter o f fact, Diocletian’s era antedates his accession. He was pro­
claimed emperor on 20 November 284. T. C. Skeat, Papyrifrom Panopolis
(1964), 145. But the Julian year began in Egypt on 29 August. Thus, the
second year o f Diocletian started on 29 August 285, and in this way the
years o f his reign came to be counted in Egypt from 29 August ad 284.
60 Sec P. Herrmann, DIVA LXXVII, 1959, 8. C f S. Accame, II dominio
romatio in Grecia (1946), 11 ; Μ . N. Tod, ABSA XXIII (1918/19), 212; id. in
Studies presented to D. M. Robinson II (1953)» 383; H. Seyrig, Syria 1950, 6;
J. Bingen, CE 1964, 14. On the era ‘o f Caesar’ (that is, ‘o f Pharsalus’) cf.
Robert, 1972, 388. On the Actium era in Cyrcne cf. L. Robert, Hellenica
XI-XII (i960), 533 ; G. Perl, Klio 1970, 320. Tw o ‘provincial’ eras co-cxisted
in Macedonia: that o f the organization o f the Roman province (148-7) and
that from the 3rd year o f Augustus which began 116 years later (cf. Robert,
1976, 3 5 9 )· Further recent bibliography about local eras: Samuel, 246.
61 Paphlagonia: see e.g. OGIS 532= Dessau, 8781. Cf. H. Dessau, Zeitschr.
fiir Numism. 1906, 335; W . Rugc, RE XVIII, 2532. Athens: P. Graindor,
Athènes sous Hadrien (1934)· Manichccs: W . B. Henning, Asia Major 1952,
198. C f the era from ad io / i i in Thessaly: A. H. Kramolish, Chiron 1975,
337. G. Le Riddcr, RN 1969, 280 suggests that the letters on the coins of
Aradus (p. 74) refer to monetary magistrates.
62 Numbering o f Olympiads: see Trucsdcll S. Brown, TimaeusofTauromenium
ιοό C H R O N O L O G Y OF T H E A N C I E N T W O R L D

(1958), 10; o f the games, L. Robert, RPh 1930, 39. List of Olympic victors:
L. Mcretti, Memorie deU'Accad. dei Lincei VII ser. II (1957). Trustworthiness
o f this list: Th. Lenschau, Phil. 1936, 391 ; F. Jacoby, Atthis (1949), 58.
63 See P. Lehmann, Phil. 1912, 297. Cf. also Ed. Schwartz, Christliche und
Jüdische Ostertafeln, AGGG N.F. VIII, 6 (1905); A. van dcr Vyer, Revue
d'histoire ecclésiastique 1957, 197; G. Ogg, Vigiliae Christianae 1962, 2. On
Dionysius cf. B. Krush, APA 1937, 57.
64 The royal canon: Chronica Minora, cd. Th. Mommsen, Π Ι , 359. Cf. Ginzel I,
139; Kubitschek, 61. Similar lists: e.g.y FrGrH Nos. 255 f., Pap. Oxyrh. 31,
2551, with a commentary by P. Sattler, Studicn ans dem Gebiete der alien
Geschichte (1962), 29; C. Corteman, CE 1956, 385.
65 An Egyptian papyrus records a moon observation in the 52nd year of
Ramcsscs II. But as lunar dates arc repetitive, the observation could refer to
the year 1253, 1250 bc, etc. Thus, its place within the range o f possible dates
depends on synchronisms which can be found'only in Mesopotamian
chronology: cf. R. Parker, JNES 1957, 42. Accordingly, the recent estima­
tions of the accession date o f the Pharaoh are: Jon D. Schmidt, Ramesses II
(1972): 1290 b c ; W . C. Hayes, CAH I: 1173; and R. O. O. Faulkner, ib. Π ,
i, 225: 1304; M. L. Bierbicr, The Late New Kingdom in Egypt (1975), 109:
1279.
66 On Manctho, see W . Hclck, Untersuchungen zu Manetho und den aegyptischen
Kiinigslisten (1956). The recent reconstruction of the list o f the Pharaohs:
Gardiner, 429. The most recent surveys o f chronological questions: Et.
Drioton, J. Vandier, L'Egypte (1962); W . C. Hayes (supra n. 65) with
addenda, ib. I, 2, 949; II, 1, 729, 760. On the XVIIIth dynasty cf. also J. G.
Read,JNES 1970, 1; D. B. Rcdford, BASOR (1973) 211; 49. On the later
period cf K. A. Kitchen, The Third Intermediate Period in Egypt, 1100-650
bc (1973) and E. Went c,JNES 1976, 269.
67 M. B. Rowton, CAH I, 1 (1970), 193 (in fact, originally published in 1962);
P. Garelli, Le Proche-Orient asiatique I— II (1969-74). See also chronological
tables for third and second millennium in CAH I, 2; II, 1-2 and in Garelli
(also for the first millennium bc). The chronology of the third millennium
hinges on the still unknown length of the interval between the last dynasty
o f Akkad and the 3rd dynasty o f Ur. C f Rowton, ib. 219 and W . W . Hallo,
RLA III, 713. The essential work on the chronology o f the ancient Near
East (Egypt included) in the second millennium is H. Tadmor, in The
World History of theJewish People, First Series II (ed. B. Mazar, 1970), ch. V,
with clironological tables from c. 1900 to c. 900 bc . O n Assyrian and
Babylonian lists o f kings cf. F. Kraus, in Mededelingen o f the Netherlands
Academy, N .R. 28, no. 2 (1965) and A. K. Grayson, Assyrian and Babylonian
Chronicles (1975)· Further cf J. J. Finkelstein, JCS 1966, 65 (royal gene­
alogies); R. Hachmann, Zeitschr. des Deutschen Palàstina-Vereins 1977, 97
(Assyrian royal dates). The Hittite chronology remains obscure: cf A.
Kammenhuber, Orientalia 1970, 278. For late Babylonian kings see J. A.
 NOTES 107
Brinkman, Political History o f the post-Kassite Babylonia 1138-722 bc (1968)
and id. BO 1970, 301. O n neo-Babylonian rulers cf. R. Borger, JCS 1965,
74; J. Oates, Iraq 1965, 135.
68 Eclipses: for the period between 4200 and 900 bc : P. W . Ncugebaucr, Spez.
Kanon der Sonnenflnsternisse fiir Vorderasien und Aegyptcn (Astrononiische
Abhandlungen VIII, 4 (1931), id. Spez. Kanon der Mondfinsternisscfiir Vorder­
asien undAegyptcn, 3430-1 v. Chr. (Astr. Abh. IX, 2 (1934), Kiel); M. Kudlek,
Solar and Lunar Eclipses in the Ancient Near East (1971). Lunar eclipses from
1400 to 10 3 b c : H. Dubbs, JNES 1947, 124. For the Greco-Roman age:
F. K. Ginzel, Spez. Kanon der Finsternisse (1899). Eclipses recorded in ancient
sources: Boll, RE VI, 2355. Solar eclipses in the Bible: F. R. Stephenson,
Palestine Exploration Quarterly 1975, 107. Comets: Gundcl, RE XI, 1183.
Eclipses, comets and earthquakes in the Byzantine age (after ad 285):
Grumel, 458 and 476. Instructions for converting astronomical dates, with
tables: P. W. Neugebauer, Astrononiische Chronologie I-II (1929), and TaJeln
zur astronomischen Chronologie I-II (1912 if.); R. Schramm, Kalenderiograph-
ischer und chronologischer Tafeln (1908); U. Bachr, Tafeln zur Behandlung
chronologischer Probletnen (1955) (Veroffentlichungen des astronomischen Rcchnen-
Instituts zu Heidelberg III, 1-3); B. Tuckermann, Planetary, Lunar and Solar
Positions (for 601 bc- ad 1649): (Memoirs of the American Philosophical Society
LVI, LIX, 1962, 1964); W . D. Stahlman, O. Gcngerich, Solar and Planetary
Longitudes (for 2500 bc- ad 2000) (1963); H. Goldstine, New and Full Moons
1001 bc -ad 1630 (Memoirs of the American Philosophical Society, 90, 1973).
69 Likewise, the horoscope o f the philosopher Proclus (Mannus, V. Procli, 35)
establishes his birth-date: 8 Feb. 412. C f J. M. Dillon, Classical Review
1969, 274.
70 On ancient chronographers, cf Ed. Schwartz, 'Die Kônigslisten des Eratos­
thenes und Kastor, AGGG XL (1894): FrGrH, 239-261. On Christian
chronographers, cf. H. Gelzer, Sextus Iulius Africanus I-II, 1 (1880-5).
Except for some fragments, the Chronicle o f Eusebius has been preserved
only in Armenian (German translation o f J. Karst, 1911) and in Jerome’s
Latin compilation, which was re-edited by J. K. Fotheringham (1923) and
R. Helm (1924-6, reprinted in 1956). The first part of Eusebius’ Chronicle,
dealing with the chronology o f the various nations, was omitted by Jerome.
The Euscbian origin o f the Canon tables has been doubted; cf Ed. Schwartz,
RE VI, 1383; D. S. Wallace-Hadrell, Eusebios (i960), 155. C f A. Momig-
liano, in The Conflict between Paganism and Christianity (ed. A. Momigliano,
1963), 82; J. Sirinclli, Les vues historiques d’Eusèbe de Chare (19Ô1), 31.
71 The Fasti Graeci and the Fasti Romani by H. Clinton (1841; 1850) arc anti­
quated but not yet replaced. The same is true for the shorter work o f Carl
Peter, Chronological Tables of Greek History (1882). The tables o f dates in
CAH and similar works do not indicate the essential point: how the Julian
date has been fixed. For Athens cf. p. 68 and Samuel, 195. Abundant
material for local history can be found in the Fasti given in the new volumes
ιο8 C H R O N O L O G Y OF T H E A N C I E N T W O R L D

o f Inscriptions Graecae, e.g. for Epidaurus (IV, i), for Arcadia (V, 2), and for
Actolia (IX, 1). Cf. also note 52. For the Ptolemies cf. note 36, for the
Seleucids cf R. A. Parker and W . H. Dubberstein, Babylonian Chronology
626 bc-a d 75 (1956). For Julian day-dates of accession, etc., of the Roman
emperors cf L. Holzapfel, Klio 1912, 1013; 1918, 1921 (partly out of date);
R. O. Fink (et alii), Feriale Duranum, YCS VII (1940); P· Buresh, Les
titulatures impériales dans les papyrus (1964). Chronological lists of high
Roman officials can often be o f help in dating documents. Cf., for instance,
prefects o f Egypt: O. W . Reinmuth, Bulletin of the American Society of
Papyrologists 1967 and 1968, 11 (partly outdated); G. Bastianini, ZPE, 17
(1975), 263; Governors o f Judaea (70-134) and Macedonia (57-IIIrd c.):
H. G. Pflaum IE] 1969, 227; G. Alfoeldi, Fasti Hispanienses (from Augustus
to Diocletian) (1969); J. Winkler, Die Reichsbeamten von Noricum (i960);
H. G. Pflaum, Les carrières procuratoriennes sous le Haut-Empire romain
(1960-1) ; A. Chastagnol, Fastes de la préfecture de Rome au Bas-Empire (1962) ;
W . Meyers, L*administration de la province romaine de Belgique (1964); A.
Jagenteufel, Die Statthalter . . . Dalmatia (Schriften der Balkankommission,
Antiquar. Abt. o f the Austrian Academy XII, 1958); D. Magie, Roman Rule
in Asia Minor (1950)» 1579 ; H. K. Sherk, The Legates of Galatia (Johns
Hopkins University Studies in History 69, no. 2, 1951)'» B. E. Thomasson,
Die Statthalter . . . Nord-afrikas (Acta Inst. Romani Regni Sueciae DC, i960).
Governors o f Coele-Syria: J. F. Gilliam, AJPh 1958, 225. Governors of
Arabia: H. G. Pflaum, Syria 1957, 128. For the chronology o f the period
between the Severi and Diocletianus cf the papers of X. Loroit (ad 235-49)
and o f M. Christole (ad 252-68) in Aufsteig und Niedergang der romischen
Welt(ed. H. Temporini, Second Series II, 1958) andj. P. Rea, Pap. Oxyrh.
XL (1972), 15 (for Egypt). For ad 294-313 cf C. H. W . Sutherland and
R. A. G. Carson, The Roman Imperial Coinage VI (1967). The dating in
Egypt under Diocletian and the other tetrarchs: J. D. Thomas, CE 1971,173.
72 G. F. Moore, Judaism I (1927), 6; R. N. Frye, The Heritage of Persia (1963),
171.
 T H E TABLES
TABLE I

The Astronomical Canon

The Canon, reproduced here after C. Wachsmuth, Einleitung in das Studium
der Alien Geschichte (1895), 3 ° 5 >has been preserved in astronomical manuscripts
which generally continue the list up to the time of the scribe, e.g. until ad 911.
The Canon was established by astronomers of Alexandria as a chronological
basis for their computations. It goes back to the Babylonian king Nabonassar
since the continuous astronomical observations began under his reign. The
astronomers of Alexandria, who used the Egyptian mobile year, reduced the
dates o f their sources to the same reckoning. For instance, Nabopolossar died
on 8 Abu o f his 21 regnal year, that is on 15 August 605. But the Canon ends
his reign at the date o f the last day o f the Egyptian year 605/4, that is on
20 January 604. The names o f Babylonian kings became corrupted in Greek
translation and transmission. According to Babylonian documents they are as
follows:

Nabonassar
Nabunadinzri
Ukinzir and Pulu (=Tiglathpileser III; cf. II Kings 15, 19)
Ululas = Shalmaneser IV
Mardukbaliddin
Arkcanos= Sargon II
‘Kingless*, that is the period o f local pretenders, Mardukzakirshum and
Mardukbaliddin, whose legitimacy was denied by the Babylonian author of
the list
Belibni
Ashumadinshum
Nergalushezib
Mushczib Marduk
‘Kinglcss’ (from the destruction o f Babylon by Sennacherib to the restoration
by Esarhaddon)
Esarhaddon
Shamashshumkin
Kandalanu=
109
 no C H R O N O L O G Y OF T H E A N C I E N T W O R L D
Nabopolossar
Nebuchadrezzar
Amel-Marduk (Evil-Mcrodach)
Ncriglissar
Nabonidus

Βασιλέων έτων έπισυναγωγή

Ναβονασσάρου ώ ι8 27 Feb. 7 4 7 - -22 Feb. 733

Ναδίου β 23 Feb. 7 3 3 - -21 Feb. 731
Χινζήρος καί Π ώ ρ ο υ € κα 22 Feb. 7 3 1 - -20 Feb. 726
Ίλουλαίου € κϊ 21 Feb. 726--19 Feb. 721
Μ α ρ δ οκεμπ άδ ου Φ λη 20 Feb. 721--16 Feb. 709
Άρκ€ ανοΰ € μγ 17 Feb. 709—14 Feb. 704
αβασίλευτα β με 15 Feb. 704--14 Feb. 702
Βιλίβου Ύ μη 15 Feb. 702— 13 Feb. 699
9Α π α ρ α ν α δ ί ο υ ς νδ 14 Feb. 699—12 Feb. 693
χΡ η γ ε β η λ ο υ a νε 13 Feb. 693-- I I Feb. 692
Μεσησιμορδάκου 8 νθ 12 Feb. 692--10 Feb. 688
αβασίλευτα V a II Feb. 688-- 8 Feb. 680
9Α σ α ρ α δ ί ν ο υ ιγ 7Γ 9 Feb. 680— 5 Feb. 667
Σαοσδουχίνου κ Ρ 6 Feb. 667--31 Jan. 647
Κινηλαδάνου κβ ρκβ I Feb. 647--26 Jan. 625
Ναβοπολασσάρου κα ρμγ 27 Jan. 625--20 Jan. 604
Ναβοκολασσάρου μγ ρττζ 21 Jan. 604--10 Jan. 561
Ίλλοαρουδάμου β ρπη II Jan. 561-- 9 Jan. 559
Ν η ρ ι γ α σ ο λ α σ σ ά ρ ου 8 Ρ? β 10 Jan. 559 — 8 jan. 555
Ναβοναδίου Φ σθ 9 Jan. 555- - 4 Jan. 538

ΙΙε ρσ ών βασιλείς

Κ υ ρ ου θ σιη 5 Jan. 538— 2 Jan. 529

Καμβύσου V σκ$ 3 Jan. 529—31 Dec. 522
Δαρείου πρώτου λς σξβ 1 Jan. 521—22 Dec. 486
Ξέρξου κα σπγ 23 Dec. 486—16 Dec. 465
9Α ρ τ α ξ έ ρ ξ ο υ π ρ ώ τ ο υ μα τκδ 17 Dec. 465— 8 Dec. 424
Δαρείου δευτέρου ίθ τμγ 7 Dec. 424— i Dec. 405
9Α ρ τ α ζ έ ρ ξ ο υ δ ε υ τ έ ρ ο υ μς τπδ 2 Dec. 405—20 Nov. 359
Ώ χ ο υ κα υι 21 Nov. 359—15 Nov. 338
9Α ρ ω γ ο ύ β υιβ 16 Nov. 338—14 Nov. 336
Δαρείου τρίτου 8 υιϊ 15 Nov. 336—13 Nov. 332
9Α λ ε ξ ά ν δ ρ ο υ Μ α κ ε δ ό ν ο ς V υκ8 Η Nov. 332—11 Nov. 324
 T H E TABLES III

βασιλείς Μ α κ ε δ ο ν ω ν

Φίλιππον τον μετ*

'Αλέξανδρον τον
κτίστην ζ υλαζ 12 Ν ον. 324- - 9 N ov. 317
’Α λ ε ξ ά ν δ ρ ο υ ε τ έ ρ ο υ Φ υμγιθ 10 Ν ον. 317- - 6 N ov. 305
Πτολεμαίου Λάγου κ νξγλθ 7 Ν ον. 305 - - I N ov. 285
Φίλαδέλφου λη φα οζ 2 Ν ον. 285- -23 O ct. 247
Ευεργέτου κε φκςρβ 24 O ct. 247 “ -17 O ct. 222
Φιλοπάτορος φ φμγριθ 18 O ct. 222--12 O ct. 205
'Επιφάνους κδ Φ Ηρμγ 13 O ct. 205-- 6 O ct. l 80
Φίλομητορος λε χβ ροη 7 O ct. 180— 28 Sept. I46
Ευεργέτου δευτέρου Χ θ χλα σζ 29 Sept. 146—20 Sept. II7
Σωτηρος λς χ α σμγ 21 Sept. 117- - I l Sept. 8l
Δ ιο νύ σο υ νέου κ θ X?* σ ο β 12 Sept. 8 1 - - 4 Sept. 52
Κλεοπάτρας χβ Φ ιν σ Ϋ δ 5 Sept. 52 - -30 Aug. 30

Examples of Babylonian Royal Lists

(XVIII Vorlauf. Bericht über die Ausgrabungen von Uruk-Warka, cd. H. J.
Lcnzen, 53 and 55.)
A
Nebuchadrezzar
Amcl-Marduk (Evil-Merodach) 2 years
Neriglissar 2 years 8 months
Labashi-Marduk 3 months
Nabonidus 5 years

B
Alexander 7 (?) years
Philip 6 years
Antigonus 6 years
Seleucus 31 years
Antiochus (I) 22 years
Antiochus (II) 15 years
Seleucus (II) 20 years
 T A BL E II

Rising and Setting oj Stars

Heliacal phenomena: near sunrise. Acronical and cosmical phenomena: near
sunset. The dates indicate the Julian day and the fraction o f the day (Greenwich
meantime) counted from Midday. For instance. May 17.20 means 17 May,
4 hours 48 minutes p .m . The difference between Greenwich time and the
hour in Rome and Athens is 50 minutes and 1 hour 3$ minutes respectively.
(Adapted from Ginzel, pp. 520 If.)

Year
LATITUDE —500 —300 -1 0 0 + IOO + 300
η Tauri (Pleiades)
Heliacal Risings

34 ° May 17.20 May 18.20 May 19.20 May 20.17 May 2 1 .II
38 »* 20.71 »* 21.65 .. 22.57 It 23.46 »» 24.32
42 »» 25-97 1» 26.81 „ 27.60 »» 28.36 »* 29.10
46 June 3.88 June 4.50 June 5.07 June 5-59 June 6.04

Heliacal Settings

34 April 5-33 April 6.60 April 7.88 April 9-15 April 10.39
38 ·* 5.29 ·· 6-55 » 780 1· 9.06 1· 10.28
42 ·» 5.16 *» 6-39 .. 763 1» 8.85 »> 10.05
46 ·* 4.89 •1 6.12 „ 7-33 ■1 8.52 ·· 9.70

Acronical Risings

34 Sept. 29.38 Sept. 30.83 Oct. 2.28 Oct. 3.80 Oct. 5-34
38 tt 25.85 *» 27.36 Sept. 28.86 Sept. 30.42 It 2.03
42 *» 21.08 II 22.62 „ 24.16 >» 25-78 Sept. 27-47
46 »» 14.11 » 15-70 » 1736 ·» 19.07 It 20.85

Cosmical Settings

34 Nov. 3.46 Nov. 4.83 Nov. 6.21 Nov. 7.60 Nov. 9.02
38 „ 3-96 ·> 5.34 „ 6.74 II 8.16 »» 9.58
42 *» 453 *» 5-94 .. 7.3 6 ·· 8.80 ·· 10.26
46 »» 5*21 »» 6.64 „ 8.09 It 9-57 »· 11.05

1 12
 Year
LATITUDE 500 - 300 - IOO + IOD + 300
<a Orionis (Bctclgeuse)
Heliacal Risings
34 ° June 25.27 June 25.71 June 26.15 June 26.60 June 27.06
38 ·· 29.04 »* 29-35 » 29.66 »* 29.99 30.34
42 July 3-44 July 3-59 July 3-75 July 3.92 Juiy 4-13
46 II 8.67 II 8.61 „ 8.57 »* 8-57 »» 8.60

Heliacal Settings
34 May 3.11 May 4.00 May 4.85 May 5.69 May 6.49
38 *» I.42 »» 2.26 „ 3.08 ·· 3.88 If 4.62
42 April 29-57 April 30.37 .. 1.14 *» I.89 II 2.59
46 •1 27.53 ·· 28.27 April 29.00 April 29.69 April 30.31

Acronical Risings
34 Nov. 27.84 Nov. 28.76 N ov. 29.67 Nov. 3Ο.58 Dec. 1.46
38 *» 29-55 >> 30.41 Dec. 1.27 Dec. 2.14 •1 2.99
42 Dee. 1.46 Dec. 2.28 j· 308 »» 3-89 •I 4.69
46 ·· 3.67 *> 442 .. S-i 6 ·· 5.9Ο •1 6.6 2

Cosmical Settings
34 Nov. 22.12 Nov. 23.19 Nov. 24.23 Nov. 25.26 Nov. 26.27
38 M 21.04 ** 22.09 „ 23.13 »» 24.I6 *» 25.17
42 »* *9-93 »> 20.97 .. 21.99 *» 23.OI ·» 24.OI
46 •1 18.76 ·· 19.80 ,1 20.80 ·» 2 I. 8 I 22.79

Year
LATTTUDB - 500 - 300 - IOO +100 + 300
a Canis major (Sirius)
Heliacal Risings
34 ° July 23.61 July 23.69 July 23-77 July 23.87 July 23.99
38 It 28.13 •1 28.11 28.10 ?» 28.13 •I 28.17
42 Aug. 2.01 Aug. 1.89 Aug. 1.79 Aug. 1-73 Aug. I.7O
46 1» 7-25 II 7.04 It 6.84 •I
6.68 1» 6.54

Heliacal Settings
34 May 6.91 May 7.24 May 7-54 May 7.82 May 8.06
38 »» 3-31 II 3 -6 o •1 3-86 II
4.09 •1 4.28
42 April 29.49 April 29.74 April 29.94 April 30.10 April 30.23
46 M 25-38 •1 25-57 •I 25.72 II
25.82 II
25.88

Acronical Risings
34 Dec. 29.54 Dec. 29.89 Dec. 30.25 Dec. 30.63 Dec. 31.00
38 Jan. 2.11 Jan. 2.40 Jan. 2.69 Jan- 303 Jan. 3-34
42 •1 5-99 II 6.19 •I 6.42 II
6.68 I» 6.93
46 ·» 10.21 •1 10.33 It 10.47 ·»
10.63 I·
iu.8i

Cosmical Settings
34 Nov. 25.83 Nov. 26.37 Nov. 26.88 Nov. 27.38 Nov. 27.83
38 ·· 22.92 11 23-43 1, 2391 I» 24.36 •I 24.79
42 »» 19.84 II 20.33 •120 .7 7 II
21.18 •1
21.58
46 »·
16.58 II 17.03 1117-43 •1
17.80 II 18.14

II3
 Year
LATITUDE - 5° ° - 3 0 0 1 0 0 + 1 0 0 + 3 0 0

a Bootis (Arcturus)
Heliacal Risings
34 ° Sept. 21-73 Sept. 23.29 Sept. 24.79 Sept. 26.24 Sept. 27.67
38 •> 18.85 » 20.49 »» 22.07 I» 23.60 *· 25.07
42 »» 15.69 »» 17.48 *> 19.18 »» 20.80 ·· 22.36
46 ·· 12.15 » 14.11 tt 15.98 *» 17.76 n 19.44

Heliacal Settings
34 O c t. 2 5 .9 6 O c t.
25-97 O c t. 2 6 .0 0 O c t. 2 6 .0 8 O c t. 2 6 .1 6
38 N o v . 2 .2 7 N o v . 1 .9 2 N o v . 1 .6 1 N o v . 1.36 N o v . 1.17
4 2
*t 1 1 .0 8 t) 1 0 .2 3 9 .4 6
>1 8.8Ο »* 8 .2 3
4 6
a 2 1 .8 0 tt 2 0 .2 7 >1 1 8 .9 2
•1 17.72 II 1 6 .6 3

Acronical Risings
34 Feb. 29.56 March 2.14 March 3-67 March 5 ·ϊ 5 March 6.56
38 • I 26.23 Feb. 27.94 Feb. 29.58 II 2.14 •13.61
42 II 22.51 » 24.39 26.17 11 Feb. 27-87 Feb. 29.47
46 II 18.17 .. 20.30 22.30 II II 24.18 • 25-93
I

Cosmical Settings
34 May 25-59 May 25.23 May 24.90 May 24.60 May 24.31
38 June 4.18 June 3-43 June 2.73 June 2.07 June 1.47
42 II 15.11 II 13.90 •1 12.76 tt ii - 7 i tt 10.73
46 •1 27.40 •I 25.65 1» 24.02 »t 22.52 tt 21.13

Year
LATITUDE - 500 - 300 - 100 +100 + 300
a Lyrae
Heliacal Risings
34 ° Nov. 16.04 Nov. 16.23 Nov. 16.34 Nov. 16.46 Nov. Ι6.5Ι
38 •I 10.16 10.35
•I 10.47 •I »t 10.58 II 10.62
42 II 3-51 it3-71 3-83 II tt 3-93 •1 3.98
46 Oct. 26.32 Oct. 26.60 Oct. 26.76 Oct. 26.89 Oct. 26.95

Heliacal Settings
34 Jan. 16.48 Jan. 16.20 Jan. 15.96 Jan 15.75 Jan. 15.56
38 II 22.98 tt 22.61 22.27
•I 21.98 21-73
42 •1 30.36 tt 29.88 11 29.44 29.06 28.73
46 Feb. 8-43 Feb. 7-76 Feb. 7-14 Feb. 6.62 Feb. 6.15

Acronical Risings
34 April 27.00 April 27.01 April 26.98 April 26.90 April 26.79
38 II 20.47 20.48 20.46 II •1 20.39 »» 20.28
42 • I 13.09 II13.13 11 13.12 •I 13.06 tt 12.95
46 tt 4-07 II 4.22 4.28 II II 4.25 tt 4.16

Cosmical Settings
34 Aug. 945 Aug. 9.02 Aug. 8.59 Aug. 8.24 Aug. 7.90
38 •1 16.42 tt 15.91 1» 15.40 •1 14.98 11 14.58
42 II 24.24 tt 23.62 II 23.04 tt 22.55 •I 22.10
46 Sept. 2.66 Sept. i.88 Sept. 1.19 tt 31-57 »· 31.01
II4
 T A B L E III

Synchronistic Table

Olympic years, years ab urbe condita (according to Varro) and Egyptian mobile
years. (After Table V in Ginzel.)
a.

Olymp.
Year Varr. J. I Year Varr. Year Varr.
1 ·

a*
BC a.u.c. O Thoth BC a.u.c. BC a.u.c. O Thoth

77 6 I,I Feb. Feb.

775 2 735 19 11,2 23 694 60 21.3 >3
774 3 734 20 3 23 693 61 4 13
773 4 733 21 4 23 692 62 22,1 12
772 2,1 732 22 12,1 22 691 63 2 12
771 2 731 23 2 22 690 64 3 12
770 3 730 24 3 22 689 65 4 12
769 4 729 25 4 22 688 66 23,1 II
768 3,1 728 26 I 3 .I 21 687 67 2 IX
767 2 727 27 2 21 686 68 3 IX
7 66 3 726 28 3 21 685 69 4 II
765 4 725 29 4 21 684 70 24,1 10
764 4 ,1 724 30 14,1 20 683 71 2 10
763 2 723 31 2 20 682 72 3 10
762 3 722 32 3 20 681 73 4 10
761 4 721 33 4 20 680 74 2 5 .1 9
760 5.1 720 34 I 5 .I 19 679 75 2 9
759 2 719 35 2 19 678 76 3 9
75« 3 718 3(3 3 677 77 4 9
757 4 717 37 4 19 676 78 26,1 8
756 6,1 716 38 I6.I l8 675 79 2 8
755 2 715 39 2 I8 674 80 3 8
754 3 714 40 3 I8 673 81 4 8
753 X 4 713 41 4 I8 672 82 2 7 .1 7
752 2 7 .1 712 42 I 7 .I 17 671 83 2 7
751 3 2 711 43 2 17 670 84 3 7
750 4 3 710 44 3 17 669 85 4 7
749 5 4 709 45 4 17 668 86 28,1 6
748 6 8,1 708 46 18,1 16 667 87 2 6
Feb. 707 47 2 16 666 88 3 6
747 7 2 26 706 48 3 16 665 89 4 6
746 8 3 26 705 49 4 16 664 90 2 9 .1 5
745 9 4 26 704 50 I 9 .I 15 663 91 2 5
744 10 9 .1 25 703 51 2 *5 662 92 3 S
743 II 2 25 702 52 3 15 661 93 4 5
742 12 3 25 701 53 4 15 660 94 3 0 ,1 4
741 13 4 25 700 54 20,1 14 659 95 2 4
740 14 IO,I 24 699 55 2 14 658 96 3 4
739 15 2 24 698 56 3 14 657 97 4 4
738 16 3 24 697 57 4 14 656 98 3M 3
737 17 4 24 696 58 21,1 13 655 99 2 3
736 18 II,I 23 695 59 2 u 654 IOO 3 3

II5
 d.

Olymp.
Year Varr. I Year Varr. a X Year Varr. a>» X
BC a.u.c. Thoth BC a.u.c. 0 Thoth BC a.u.c. 3
Thoth

Feb. Jan. Jan.

653 131 3M 3 603 151 2 21 551 203 57,2 8
652 102 32,1 2 602 152 44.3 21 550 204 3 8
651 103 2 2 601 153 4 21 549 205 4 8
650 104 3 2 600 154 45.1 20 sis 206 58.1 7
649 105 4 2 599 155 2 20 547 207 2 7
648 106 3 3 .1 X 598 156 3 20 546 208 3 7
647 107 2 X 597 157 4 20 545 209 4 7
646 108 3 I 596 158 46,1 19 544 210 59.1 6
645 109 4 X 595 159 2 19 543 211 2 6
594 160 3 19 542 212 3 6
Jan. 593 161 4 19 541 213 4 6
644 no 3 4 .1 31 592 162 47.1 18 540 214 60,1 5
643 III 2 31 591 163 2 18 539 215 2 5
<$42 112 3 31 590 164 3 18 5’8 216 3 5
641 II 3 4 31 589 165 4 18 537 217 4 5
640 II 4 3 5 .1 30 588 166 48,1 17 536 218 61,1 4
639 «5 2 30 587 167 2 17 535 219 2 4
638 Il6 3 30 586 168 3 17 534 220 3 4
637 II 7 4 30 585 169 4 17 533 221 4 4
636 i:8 3<
5.i 29 584 170 49.1 16 532 222 62,1 3
635 II 9 2 29 583 171 2 16 531 223 2 3
634 120 3 29 582 172 3 16 530 224 3 3
633 121 4 29 58l 173 4 16 529 225 4 3
632 122 3 7 .1 28 580 174 50,1 Ï5 528 226 63.1 2
631 123 2 28 579 175 2 15 527 227 2 2
630 124 3 28 578 176 3 15 526 228 3 2
629 125 4 28 577 177 4 15 525 229 4 2
628 126 38.1 27 576 178 5M 14 524 230 64,1 X
6*7 127 2 27 575 179 a 14 5^3 231 a z
626 128 3 27 574 180 3 14 522 232 3 X
625 129 4 27 573 181 4 14 Dec.
624 ISO 3 9 .1 26 572 182 5 2 ,1 13 521 233 4 31
623 IÎI 2 26 571 183 2 13 520 234 65.1 31
622 132 3 26 570 184 3 13 519 235 2 31
621 133 4 26 569 185 4 U 518 236 3 31
620 1 -4 40,1 25 568 186 5 3 ,1 12 517 237 4 30
619 135 2 25 567 187 2 12 516 238 66,1 30
618 I36 3 25 566 188 3 12 515 239 2 30
617 137 4 25 5<55 189 4 12 5H 240 3 30
616 138 4M 24 564 190 5 4 ,1 XX 513 241 4 29
615 139 2 24 563 191 2 XX 512 242 67,1 29
614 I40 3 24 562 192 3 II 5” 243 2 29
613 I4 I 4 24 561 193 4 IX 510 244 3 29
6l2 142 4 2 .1 23 560 194 5 5 .1 10 509 245 4 28
6x1 143 2 23 559 195 2 10 508 246 68,1 28
610 144 3 23 558 196 3 10 507 247 2 28
609 145 4 23 557 197 4 10 506 248 3 28
608 I46 4 3 .1 22 556 198 56.1 9 505 249 4 27
607 147 2 22 555 199 2 9 504 250 69.1 27
606 I48 3 22 554 200 3 9 503 251 2 27
60s 149 4 22 553 201 4 9 502 252 3 27
604 ISO 4 4 .1 2X 552 202 5 7 ,î 8 501 253 4 26
il 6
 Olymp.
Q. Q.
Year Varr. 1 « Year Varr. 1 . Year Varr. I
BC a.u.c. 3 Thoth BC a.u.c. 0 Thoth BC a.u.c. Thoth

Dec. Dec. Dec.

500 254 70.1 26 450 304 82,3 X4 400 354 95.1 I
499 255 2 26 449 305 4 *3 399 355 2 I
498 256 3 26 448 306 83.1 *3 398 356 3 I
497 257 4 25 447 307 2 «3 Nov.
496 258 71.1 25 446 308 3 13 397 357 4 30
495 259 2 25 445 309 4 12 39<5 358 96,1 30
494 260 3 25 444 310 84.1 12 395 359 2 30
493 261 4 24 443 311 2 12 394 360 3 30
492 262 72,1 24 442 312 3 12 393 361 4 29
491 263 2 24 441 313 4 II 392 362 97.1 29
490 264 3 24 440 314 85.1 II 391 363 2 29
489 265 4 23 439 315 2 II 390 364 3 29
488 266 73.1 23 438 316 3 II 389 365 4 28
487 267 a 23 437 3T7 4 10 388 366 98,1 28
486 268 3 23 436 318 86,1 10 387 367 2 28
485 269 4 22 435 319 2 10 386 368 3 28
484 270 74.1 22 434 320 3 10 385 369 4 27
483 271 2 22 433 321 4 9 384 370 9 9 .x 27
482 272 3 22 432 322 87.1 9 383 371 2 27
481 273 4 21 431 323 2 9 382 372 3 27
480 274 75.1 21 430 324 3 9 38i 373 4 26
479 275 2 21 429 325 4 8 380 374 IOO,I 26
478 276 3 21 428 326 88.1 8 379 375 2 26
477 277 4 20 427 327 2 8 378 376 3 26
476 278 76.1 20 426 328 3 8 377 377 4 25
475 279 2 20 425 329 4 7 376 378 101,1 25
474 280 3 20 424 330 89.1 7 375 379 2 25
473 281 4 19 423 331 2 7 374 380 3 25
472 282 77.1 19 422 332 3 7 373 38i 4 24
471 283 2 19 421 333 4 6 372 382 102,1 24
470 284 3 19 420 334 90,1 6 371 383 2 24
469 285 4 I8 419 335 2 6 370 384 3 24
468 286 78.1 I8 4x8 336 3 6 369 385 4 23
467 287 2 l8 417 337 4 5 368 386 103.I 23
466 288 3 l8 416 338 9M 5 367 387 2 23
465 289 4 17 415 339 2 5 366 388 3 23
464 290 79 .x 17 414 340 3 5 365 389 4 22
463 291 2 17 413 34 X 4 4 364 390 104.1 22
462 292 3 17 412 342 92.1 4 3<$3 391 2 22
461 293 4 l6 411 343 2 4 362 392 3 22
460 294 80,1 l6 410 344 3 4 361 393 4 2X
459 295 2 l6 409 345 4 3 360 394 105,1 21
458 296 3 l6 408 346 93 .x 3 359 395 2 21
457 297 4 15 407 347 2 3 358 396 3 21
456 298 8I.I 15 406 348 3 3 357 397 4 20
455 299 2 15 405 349 4 2 356 398 106,1 20
454 300 3 15 404 350 94.1 2 355 399 2 20
453 301 4 14 403 351 2 2 354 400 3 20
452 302 82,1 14 402 352 3 2 353 401 4 19
451 303 2 14 401 353 4 I 352 402 107.1 19
351 403 2 19

117
 u> <*» te* te* U * te l u* w te» te l W te» U» te* U> U* u» ω V-J t e l U * u * Vte* (te* Vtel u * (tel Vte* te* (te* te* U * te* te* te* te* te* te* te* t e l te* te* te* te * te* t e l fri £
Ο υ o p O o M hl M KJ M N KJ ► -> KJ te* u * te* te* te* U* te* te* 4 - 4k 4· 4> 4 k 4k ' y . Q
*« 4k te» O +* M te* 4k «y. O» k j CO O O ►J Kl te* 4 k Vy» O v v | OO Ό O ►j KJ te* 4k vy> O* k j OO Ό «1
Μ 1- 8 . 3 00 « o ·- £ te* £ te» OO v o O

W
<
4k 4* 4 k 4k 4k 4k 4 > 4k 4 k 4k 4k 4 * 4k 4k 4k 4 k 4 k 4k 4k 4k 4k 4k 4k 4 k 4k 4k 4k 4* 4 k 4k 4k C
νΛ «y. te» vy. — t £ £ £ £ £ £ £ £ £ te» W Vkj t e l te* te» VO u * u* £ £ £ K* KJ KJ £ KJ £ £ -* ·-< -* « •te ■te -V ►te >-· *
Qtt - 1 4k s
te* ►J O Ό Λ K» Ο Ό o o -k l o . vy. 4 - «-* K* M O Ό OO -k l O» v t 4k te* K* - * U Ό OO k j O . vy. 4k te* U M O vS S 3 I I 4k

►4 >-« *■« ►J ►H -* i-v t-t k* ►te M >-*

M — — — « *- -1 Q o O
Ό j C\ 4k <*> U H Π « . 1
- v _ iiy m p .
4^ te* *4 — 4k U* K» k* 4k V** U »-· ■K W K> ■-· 4* u* S» « 4k te* Kl H 4k o* KJ 4k te* KJ M 4k te* Kl k* 4k te* Kl 4k U· KJ •te 4 k te* KJ M 4k w

? M
Ο* "J kj >*3 Ό OO OO 0 0 oo Ό Ό Ό Ό O O o O - M M M Kl K> Kl M w o* t e l (te* 4 k 4k 4- 4^ v y . vy. vy. vy. Ov Ov Ov Ov k j Ό Ό kj OO o ô co co «o <
Thoth

κ> κ> KJ KJ W S» M K) w U K> Μ KJ K» t o K» K> M K> Kl Kl Kl Kl Kl M Kl KJ KJ Kl Kl KJ Kl M K* KJ Kl Kl KJ Kl M Ki Kl Kl Kl te* « !

vy» tev Vy» 'y» te» t e . te» «y. vy. 8 , 8 Os O Ov O v Ov -J -k l - j OO OO 8 OO CO OO CO OO 0 0 o o Ό Ό eu
Ό « Ό «O
►* KJ te* 4 - te» O» s J OO Ό M » te* 1 vy» O . ^ 3 Oo Ό 3 3 £ te* S vy. CO Ό o ■8 M KJ te* £ te» 00 «o O
» M KJ (te* 4 k vy. Ov *•3 CO Ό * 3 8 R
w
<
•-Λ '-Λ ■y. U 4k 4 - -U 4. 4 - 4> 4* 4k 4* 4* 4k 4 . 4k 4 k -U 4 « 4 « 4k J - 4* 4* 4k 4k 4 k 4k 4k 4k 4*· 4 « 4k 4». 4k 4k 4* 4k 4 * 4k 4 k 4k 4k 4* 4 k 4k 4 k 4k 4k îj
c
Ο Ο Ο g Ό O Ό «O O c OO OO y- o o c o OO 0 0 c c c c M kJ "J k j -J kj kj » O > O . Ov Ol b CK vy. v y , vy, tev «y. vy. 5
μ Μ ►* o « OO 3 8 . vy» 4 « V- K> 3 Ό o o ' - J O v vy» 4 k Vkj M >-» υ Ό 00 3 O» vy» 4k te* KJ ^ ci Ό CO k j 8 «y. 4 - te* K* ►te 8 Ό OO k j O v te» 4k O
Μ

ΟΟ ** »-« »-* M -< (te ·-

ν-* teJ te l W Kl K> KJ K* KJ
ΚΙ ·-* p e OO '-l 8 vy. 4k te* £ M 8
kj Μ 4k w KJ *-« 4 k teJ M 4k te* s» M 4* w KJ ~ 4k v*> M M * U1 Kl M 4». (te* K· H 4k (tel Kl H. 4k Vte* W ►te 4k te* KJ C* 4k te* K> M 4k teJ K· H
Olymp.

kj kj
te»
Κ» *4
tev «y» o
KJ
O
KJ
O
KJ KJ W K> *4
O k j ■kj k j k j
K* M
OO CO o o
KJ U to
oo Ό
M
Ό
KJ KJ t e l te*
VO Ό O O
t e l te* (te* u *
O O ►" «
(te*
•H « f* M >- *■« k* KJ K· Kl KJ te* te* «te* te* 4k 4k 4k 4k vy» t e . t e . te» Ov O v O v
? c»
S -

Ν W w KJ M ►J K) K» K» KJ >4 t4 K> K> K* tJ K> K» Kl K> K1 Kl KJ KJ KJ K» Kl Kl KJ Kl Kl Kl Kl KJ M KJ Kl K* KJ KJ Kl Kl KJ Kl KJ KJ ><

Ο o O K· te* te* te* te* te* te* te* ES
Ο — υ M — — — •H it f' N Kl t-i Kl Kl KJ lo (* j te* te* 4k 4 k 4k £ £ 4 k 4k g
►< KJ (M 4*> te» I I OO « O »-* w te» 4k vy. O» -k l or. VO O K» o » 4k vy. O* - J OO VO O M U te* ·* . «y» O» k l OU XJ O ** T j te* £ vy. Ov kJ S n
K

w
<
te» te» ' y te» 'y , y . 'y> «y. 'y . vy» Vy, Vy. 'y . v y . vy, U i v y . vy. vy. vy. ■y. y* 'y . M v y v y . <y. (y» vy. (y» 'y> «y» Vy. ■y. vy. >y. vy. «y. vy. y- vy. 'y . vy. 'y t e . Vy. t e . te» vy» C
'y . 'y . «y» te» 4k 4» 4 - 4 - 4- 4- 4- 4- te* te* te* te* te» te l k- te l v u M Kl K) KJ [O H £ f-* KJ Ki — *— — H — — — —
-* Ν H U 'C OO v J vy» 4 - £ Ο Ό a k J O» · ' · 4 - '-j (y 'C OO k j C' v y 4 k te* - O O OO k l O Vy. 4k te* KJ ►- O 3
— o N O a « i. £

-4 ►- » -· •H ►- kte M •te
4k 4- 4k 4* te* te* te* te* te* te *
•*k (M £ »-* O to 'Λ 1 'y , 4. t*
v _ /iy in p .
W (4 M 4- U * JJ « 4k U* K» M 4k U * to »-< 4k te * k> ~ 4k te * Kl 4k u* N 4k te * KJ k* 4k te* Kl H 4k te* Kl •te 4k te* K1 ►te 4k tel Kl « 4k te*

o H
KJ KJ KJ »J KJ KJ KJ U KJ Kl Kl KJ to K) KJ KJ M tJ KJ o
W O , 'y - 1-0 4k 4* 4* •y . -y . 'y 'y . k j k j OO cr M o* te. <**
KJ lKJ
**
Μ *- O» o » Ov Ο . Ό -1 oo OO Ό Ό Ό Ό O o o o >-* M — Kl Kl Kl KJ 4 > 4k o
c r
 Ô.
Year Varr.
Year Varr. t . Year Varr. 1 . 1 1
BC a.u.c. 0 Thoth BC a.u.c. O Thoth BC aux. O Thoth

Oct. Sept. Sept.

200 554 I 45 .I 12 I50 604 157.3 30 IOO 654 170,1 17
199 555 2 12 149 60$ 4 29 99 655 2 17
198 556 3 12 I48 606 158,1 29 98 656 3 17
197 557 4 II 147 607 2 29 97 657 4 16
196 558 146,1 11 I46 608 3 29 96 658 171,1 16
559 2 II 145 609 4 28 95 659 2 16
195
194 560 3 II 144 6 10 I 59 .I 28 94 660 3 16
193 561 4 10 143 611 2 28 93 66l 4 15
192 562 I 47 .I 10 142 612 3 28 92 66 2 172,1 15
191 503 2 10 I4 I 613 4 27 91 663 2 15
190 564 3 10 I4O 614 160,1 27 90 664 3 15
189 565 4 9 139 615 2 27 89 665 4 14
188 566 148,1 9 138 616 3 27 88 666 I 7 3 .I H
187 567 2 9 137 617 4 26 87 667 2 Μ
186 568 3 9 136 618 161,1 26 86 068 3 14
185 569 4 8 135 619 2 26 85 669 4 13
184 570 149 . i 8 134 620 3 26 84 670 174,1 13
183 571 2 8 133 621 4 25 83 671 2 13
182 572 3 8 132 622 162,1 25 82 672 3 13
181 573 4 7 I 3I 623 2 25 8l 673 4 12
180 574 150,1 7 I30 624 3 25 80 674 175,1 12
179 575 2 7 129 625 4 24 79 675 2 12
178 576 3 7 128 626 163,1 24 78 676 3 12
177 577 4 6 127 627 2 24 77 677 4 II
176 578 151,1 6 126 628 3 24 76 678 176,1 II
579 2 6 125 629 4 23 75 679 2 II
175
174 580 3 6 124 630 164,1 23 74 680 3 II
173 581 4 5 123 631 2 23 73 681 4 10
172 582 152.1 5 122 632 3 23 72 682 177,1 10
171 583 2 5 121 633 4 22 7* 683 z 10
170 584 3 5 120 634 165,1 22 70 684 3 10
169 585 4 4 II9 635 2 22 69 685 4 9
168 586 153,1 4 Il8 636 3 22 68 686 178,1 9
167 587 2 4 II7 637 4 21 67 687 2 9
166 588 3 4 Il6 638 166,1 21 66 688 3 9
165 589 4 3 II5 639 2 21 65 689 4 8
164 590 154,1 3 II4 640 3 21 64 690 179.1 8
163 591 2 3 113 641 4 20 63 691 2 8
162 592 3 3 112 642 167,1 20 62 692 3 8
161 593 4 2 III 643 2 20 61 693 4 7
160 594 I 55 .I 2 no 644 3 20 60 694 180,1 7
595 2 2 109 645 4 19 59 695 2 7
159
158 596 3 2 I08 646 168,1 19 58 696 3 7
157 597 4 I 107 647 2 19 57 697 4 6
156 598 156,1 I 106 648 3 19 56 698 181,1 6
155 599 2 I 105 649 4 I8 55 699 2 6
154 600 3 I 104 650 169,1 I8 54 700 3 6
Sept. 103 651 2 l8 53 701 4 5
153 601 4 30 102 652 3 18 52 702 182,1 5
152 602 I 57 .I 30 IOI 653 4 17 51 703 2 5
151 603 2 30
 M M M M « M M — *-< SJ SJ SJ SJ s» SJ S J SJ CO C o Co co co co co 4* 4* 4* 4- 4* 4 * 4 - 4* CA
a
M w 4» u 00 Ό CO 4*· VA co 00 co M co VA GO Ό Ci
M O» O SJ Ov •Ί CO * o O *- SI £ VA CA ■ 'J a o ►H SJ VA O n -J oo Ό O t t O n •vJ o
t*
"J vJ Ό Ό -J vJ "J ^3 -J 'J 'J "4 *o Ό v J 'J 'J *•4 ^3 •Ί ~*3 c
V» ά va 4- a S i lo (* » ·*· CJ w w «-» •o SJ 'ï •2 •s *-< X *
g U h
Co O O % £ Ά 4*. co t >-« Ό oo " 4 O n VA 4>- co si •H o Ό CO ' J O n VA 4*· CO *SJ2 •o O Ό OO N j O n va 4* co SJ »-« O £ oo * O n va 4»

►* ·- »-* *-« *-· *·* M M

Ό Ό ce 00 OS oo OO ce oo oe
'f u i
f* · SJ ►H )P oo 'J o VA co
P t“ * O l
4k Co SJ *-« 4* «*» N *· 4*- CO s> *-« 4k Co SJ ►H 4 * · co SJ -4 4 *- co SJ M 4*· CO S> ·— 4^ co SJ — 4*. Co S· co SJ H co SJ co SJ « 4*. co
« 4^ 4^ —
i 4*.
> C
O/ï H
►J S> SJ S) SJ SJ SJ SJ SJ SJ SJ SJ SJ SJ SJ S J SJ SJ S J SJ SJ SJ co co
CO coCO co co C sf
co S £ 4*· 4* Ά VA VA VA On o* & O n "J 'J '-J 'J OO OO oo CO O O o Ό O O o o W s» SJ Co CO c o co 4» 4» 4* 4» VA
Όer
o

M *-« M «O M*-« M M M *-« fc

o 00 ^4 O » CA 4 k vo S. M O O oo " 4 O n Ca 4k CO SJ M 0

w
Ό " 4 ''J -4 Ό Ό >4 C
s ^ On On On O n
S •R On CA 'CJA •kVAj 'CJA > 4 O
SJ •H 32 oo >g S A 2 Co SJ ·-« S nO oo - J O VA 2 n

120
ο ο ο
M ο ο ο 8 ο ο 'C Ό Ό
C* y» 4k vo v μ Ο Ό
voco ~ -l P» va
co SJ ·- 4k co si ►H 4* CO S) H 4k CO SJ •H Olymp.
>
SJ SJ SJ SJ SJ SJ SJ SJ SJ SJ SJ SJ
O o O O « ►-* SJ« SJ s> SJ SJ co co Cs*
Ow
5- _
co SJ
Ά VA VAVA VA •H e ?
00 Vj
1COAn4O*n ON CA ON S' C Anc
ΌAVA Ά CO a 4k CO s»

ce co » co oc co co co oc OC co OO CO 03 CO oo oo 00 CO CO OO 00 00 OO OO o o oo co oo oo co oo oo O0 0 0 ce co 00
4- 4* vo Vo ·— Vo Vo Co vo Co Vo Ù l ÎJ SJ t- *J SJ 6 — M — — H — — -«
S* SJ fj o î?
t H υ Ό CO " J ON VA 4 - Co SJ M O Ό oo -J CA VA 4 k CO s* k* O Ό 00 'J On ά 4k CO SJ ·— Va 4k h Ü
o a
oOO *sjo R

N K SJ S* SJ SJ SJ SJ SJ SJ SJ SJ SJ
*■* « ►H ·- H ·- H- o
o o
Ό - ) Os Ά •1- Vo M o co
*0-»
0 fi <o - 1
«· - Olymp.
M 4k co SJ M 4^ Vo SJ M 4*- c o SJ M 4k Co S· H 4- CO SJ •H 4k co SJ « 4k CO SJ M 4k CO SJ M 4k Co s» ►j 4k C i

>
c
W W W W U U U M £.
6

« 0 0 0 0 « * - M « v < SJ SJ SJ SJ CO vo Co vo 4* 4k 4k CA VA VA VA On O n On ^4 00 oo oo OO Ό O Ό
O
Q
7
7
7

*o o o o O *■
5·
 1ZI
ζτ z iooi 052
τζ VLiz 2001 6*2 * * £56 002 Li 2 £06 o5i
ζτ t IOOI 8*2 5 £ 256 66l Li I*2£2 206 6*1
ÎZ £ OOOI Ltz 5 2 i 56 861 Li * 106 8*1
tz z 666 9*2 5 l ‘**2 o 56 L6l 8i £ 006 L ti
iz i'9iz 866 5*2 5 * 6*6 96I 8i 2 668 9*1
iz t L66 ttz 9 £ 8*6 56i 8i 1*1£2 868 5* i
tz £ 966 itz 9 z Lt6 *61 8i * L6% ttl
tz z 566 2*2 9 V£tz 9*6 £61 61 £ 968 £*1
tz v iiz *66 1*2 9 t 5*6 261 61 2 568 2*1
tz t £66 0*2 L £ **6 161 61 I*o£2 *68 1*1
iz £ 266 6£2 L z £*6 061 61 * £68 0*1
iz z 166 8£2 L 1*2*2 2*6 6gi 02 £ 268 6£i
iz l'tiz 066 L£z L * 1*6 881 02 2 168 8£i
iz t 6g6 9£z 8 £ 0*6 ^•81 02 1*622 068 L£ 1
9Z £ 886 5 £2 8 2 6£ 6 981 02 * 688 9 £i
9Z z L26 *£2 8 1*1*2 8£6 Î8 i 12 £ 888 5 £1
9Z VZiz 986 ££2 8 * L£6 *8i 12 2 £88 *£1
9Z t 5 g6 2£2 6 £ 9Î6 £8i 12 1*822 988 ££1
Lz £ *86 l£2 6 z i£6 281 12 * 588 z£i
Lz z £86 o£2 6 1*0*2 *£6 m 22 £ *88 i £i
Lz Vziz 286 622 6 * ££6 081 22 2 £88 o£i
Lz t 136 822 01 £ 2£6 6L1 22 VLzz 288 621
8* £ 08 6 Lzz 01 z i£6 8^1 22 t 188 8£i
82 z 6L6 922 01 V6£z o£6 LLi £2 £ 088 Lzi
82 V i iz 8^6 522 01 t 626 9L1 £2 z 6L% 921
82 t LL6 *22 II £ 8^6 iLi £z 1*922 8^8 521
6z £ 9L6 £22 II z Lz6 tLi £2 * LL8 *21
6z z iL6 222 II i*8 £* 926 £Li tz £ 9^8 £21
6z vo i z tL6 122 II t 526 zLi tz 2 5^8 221
6z t £L6 022 21 £ *20 1L1 tz 1*522 *^8 121
o£ £ zL6 6l2 21 z £26 0L1 tz * 021
0£ z 1L6 81 z 21 VL£z 226 691 iz £ zZ.8 611
o£ V6tz 0 L6 Liz 21 t 126 89i iz 2 1^8 8II
o£ t 696 912 £1 £ 026 L91 iz 1**22 0Z.8 Lu
aunf £1 z 6l6 991 iz * 698 911
1 £ 896 5l 2 £1 i ‘9£2 8i6 59i 92 £ 898 5n
I z L96 *12 £i * Z.16 *91 92 2 ^98 *11
I 1*8*2 996 £12 *1 £ 916 £91 92 i ‘£zz 998 £11
I * 596 212 tl 2 5 i6 291 92 * 598 211
z £ *96 112 *1 i ' 5£2 *16 191 Lz £ *98 III
z z £96 012 *1 * £16 091 Lz 2 £98 ou
z VLtz 296 602 5i £ 216 651 Lz 1*222 ^98 601
z t 196 802 5i 2 116 85i Lz * 198 801
£ £ 096 Loz 5i I**£2 016 Lii 8* £ 098 Ζ,οι
£ z 656 902 5i * 606 951 8* 2 658 90i
£ I ‘9*2 g56 502 91 £ 806 55i 8* 1*122 858 5oi
£ * Li6 *02 91 2 L06 *5i 8* * «8 *01
t £ 956 £02 91 i ‘££2 906 £5 i 62 £ 958 £01
t z 556 202 91 * 5 o6 251 62 2 558 201
t 1 *5*2 *56 102 Li £*2£2 *06 i 5i 6z 1*022 *58 101
A[nf Apf
φ ο η χ O •dtve αν qioHX O •one αν φ ο η χ S O-CVC αν
I •jjc A I ΙΕΟλ I *-< • ΙΙΕΛ jw A
1 3 3
y y ?
 D- d.
Year Varr. ! . Year Varr.
a.u.c.
1
O
1
Year
Thoth AD
Varr.
a.u.c.
J . Thoth
1
AD a.u.c. O Thoth AD 3

June 267 1020 3 18 284 1037 4 13

251 1004 257,3 22 268 1021 4 17 285 1038 266,1 13
252 1005 4 21 269 1022 262,1 17 286 1039 2 13
253 I006 258.I 21 27Ο 1023 2 17 287 1040 3 13
254 IOO7 2 21 271 1024 3 17 288 1041 4 12
255 I008 3 21 272 1025 4 16 289 1042 267,1 12
256 1009 4 20 273 1026 263,1 16 29Ο 1043 2 12
257 1010 259,1 20 274 1027 2 16 291 1044 3 12
258 IOII 2 20 275 1028 3 16 292 1045 4 II
259 :oi2 3 20 276 1029 4 15 293 1046 268,1 II
26Ο 1013 4 19 277 1030 264.1 15 294 1047 2 II
26l 1014 260,1 19 278 1031 2 15 295 1048 3 II
262 101$ 2 19 279 1032 3 15 296 1049 4 10
263 1016 3 19 280 1033 4 14 297 1050 269,1 10
264 1017 4 l8 28l 1034 265,1 14 298 1051 2 10
265 1018 26l,I l8 282 1035 2 14 299 1052 3 10
266 1019 2 I8 283 1036 3 14 300 1053 4 9

122
NOTES TO TABLE IV

In leap years 23 Feb. = bis VI (bissextilis); 26 Feb.= V; 27 Feb.=IV; 28 Feb. = III

a.d. Kal. Mar.; 29 Feb.=pridie Kal. Mar.
In the pre-Julian Roman calendar Martins, Maiusf Quintilis, and October had
each 31 days, February had 28 days, and the seven other months 29 days each.
For the counting o f days before the Ides see the Julian calendar. For the days
between the Ides and the next Calends, subtract the Roman number o f the
day from the number o f the days in the month and add 2. For instance,
IX a.d. Kal. Nov. will be 31 (die number o f days in October)- 9 = 2 2 + 2= 24
October.
I23
 T A B L E IV

The Roman Julian Calendar

November
September

December
February

October
«
January

April

J >· 1
June
Day

3
*—« <

I Kal. Kal. Kal. Kal. Kal. Kal. Kal. Kal. Kal. Kal. Kal. Kal.
2 IV IV VI IV VI IV VI IV TV VI IV IV

3 m m V m V m V m ni v m m
4 pr. pr. IV pr. IV pr. IV pr. pr. IV pr. pr.
5 Non. Non. m Non. m Non. m Non. Non. m Non. Non.

6 vm vm pr. vm pr- vm pr. vm vm pr. vm vm

7 vn vn Non. vn Non. vn Non. vn vn Non. vn vn
8 VI VI vm VI vm VI vm VI VI vm VI VI

9 v V vn V vn v vn v v vn v v
10 IV TV VI IV VI TV VI IV IV VI IV IV

II ni m V m V m v m m v m m
12 pr. pr. IV pr. IV pr- IV pr. P^·
TV
Pr · pr·
Π Id . Id . m Id . m Id . m Id . Id . m Id . Id .

14 X IX XVI pr. xvm pr. xvm pr. X IX xvm pr. xvm X IX

15 xvra XV Id . xvn Id . xvn Id . xvm xvn Id . xvn xvra

ι ό xvn X IV xvn XVI xvn XVI xvn xvn XVI xvn XVI xvn
XVI xm XVI XV XVI XV XVI XVI XV XVI XV XVI
17
ι8 XV xn XV X IV XV X IV XV XV X IV XV X IV XV

19 X IV XI X IV xm X IV xm X IV X IV xm X IV xm X IV

20 xm X xm xn xm xn xm xm xn xm xn xm

21 xn IX xn XI xn XI xn xn XI xn XI xn
22 XI vm XI X XI X XI XI X XI x XI
vn IX X IX X
23 X X DC X IX X X
24 IX VI IX vm IX vm IX IX vm IX vm IX

25 vm V vm vn vm vn vm vm vn vm vn vm

26 vn IV vn VI vn VI vn vn VI vn VI vn
27 VI m VI v VI v VI VI V VI v VI
28 V pr. V IV v rv v v IV v IV v
29 IV IV m IV m IV TV m IV m rv
30 m m pr. m pr. m m pr m m
Kal. Kal. Kal. Kal.
May July Oct. Dec.
31 pr. pr. pr- pr. pr. Ρ Γ· pr-
Kal. Kal. Kal. Kal. Kal. Kal. Kal.
Fcb. April June Aug. Sept. Nov. Jan.

1 2 5