1

Ancient Eclipses and the Fall of Babylon,

Boris Banjevic

Institute of Informatics and Statistics of the City of Belgrade

INTRODUCTION

The creation of an absolute chronology for the ancient Near East

depends upon identifying the recorded observations of ancient astronomers.
Astronomical data can form the basis of an absolute Near Eastern
chronology, but there are other solutions at least as acceptable as those
previously proposed. These new solutions result from a reinvestigation of
the observations of ancient solar eclipses using a computer program
Emapwin1. In our opinion, the Middle Chronology is too long; the Ultra-
Low chronology is too short. The chronology proposed here means that
Babylon fell in 1547 BC.

Our investigation depends essentially on Assyrian and Babylonian king-

lists. However, in the search of absolute date assignments we depend on
ancient astronomical texts absolutely. These texts consist mainly of
observations of lunar and solar eclipses and of planetary phenomena.

The earliest records of astronomical omens date from the Dynasty of

Akkad, where the preference to omen astrology may indeed have arisen.
Nevertheless, the most extensive collection of celestial observations was
undertaken during the Kassite period and evolved into a series known as
Enūma Anu Enlil. While many of the eclipse observations found in EAE are
considered as essential for the absolute dating of the early dynasties of

1
http://www2c.biglobe.ne.jp/~takesako/cal/emapwin_eng.HTM. The program is based on investigations of
Stephenson (1997). It was used by Leo Dubal on the Solar Eclipse Coference (SEC) in August 2004.
 2

Mesopotamia, a caution appears necessary with regard to all the recorded

details (Huber et. al. 1982, 40).

1. THE DECELERATION OF THE EARTH'S RATE OF ROTATION

The most significant factor for the calculation of astronomical events is

the known deceleration of the Earth's rate of rotation (ΔT). While 'tidal
friction' has long been identified as a major cause of the Earth's general, but
variable, deceleration, numerous 'non-tidal' mechanisms have also been
identified. The result of the various tidal and non-tidal processes is a gradual
slowing down of the Earth's rotation, but at an irregular rate. To determine
the past rotation rate of the Earth, highly detailed solar eclipse records must
be used. These are July 17, 709 BC (the total eclipse at Chu-Fu); March 4,
181 BC (the total eclipse at Ch'ang-An); April 15, 136 BC (the total eclipse
at Babylon. The temporal discrepancy for the first eclipse demands between
5.62 and 5.88 hours; the third eclipse 3.11 and 3.37 hours (Stephenson &
Morrison 2000, 62f).2

The value of the acceleration parameter (c) is expressed by ΔT = ct 2=

ET-UT where: ΔT is in seconds3, t is measured in centuries since AD 1830,
ET is Ephemeris Time (calculated time) 4, UT is Universal Time (observed
time). In Emapwin calculation lunar orbital acceleration is -26 arcsec/cy 2.
According to lunar laser ranging, the correct value is in the range of -25.8 to
-26 arcsec/cy2.

In this study, using the recently designed eclipse program capable of

receiving the input of alternative values for each eclipse calculation, the
range of c necessary to register totality for several solar eclipse records was
investigated, and the results are given in the Table 1.

Site Date UT/LMT M c

2
This is a good test for Emapwin. Analysis of the various observations over the past 2500 years shows that
the average rate of lengthening of the day has been 1.7 ms/cy (milliseconds per century).

3
The discrepancy in time ΔT is the measure of the cumulative drift in time of clock varying in the rate in
accordance with changes in the length of the day (Stephenson & Morrison 2000, 58).

4
ET has recently been replaced with Terrestrial Time (TT).
 3

Chufu (Qufu) July 17, -708 7:41/15:31 1.03 32.1

E 117, N 35.65

Chang’an (Xi’an) March 04, -180 7:41/14:57 1.02 30.3

E 108.9, N 34.3

Babylon April 15, -135 4:27/7:25 1.02 30.2

E 44.5 N 32.5

Table 1

LMT - Local Mean Time; M - Magnitude of the eclipse; c – acceleration parameter.

From this table, it can be seen that in all instances the average value c =
31 s/cy2 appears to be satisfactory. The rate of change of length of day
(d[l.o.d.]/dt) is 1.7+/-0.1 ms/cy, which is equivalent to an acceleration
parameter between 29.2 and 32.8 s/cy2 (Stephenson & Morrison 2000, 74).5

There are still more valuable eclipse records. These include records of
observations of eclipses either observed from or mentioned in documents
found at various sites including a solar eclipse from Ur, lunar eclipses of
Gutium and Ur, a solar eclipse from Assur, lunar and solar eclipses from
Babylon, a lunar eclipse from Tell Muh.ammad, and a lunar eclipse from
Mari.

2. ASTRONOMICAL EVIDENCE

The Venus Observations of Ammis.aduqa (EAE tablet 63 et al.)

Among the more reliable EAE astronomical texts, we can include the
observations of the appearance and disappearance of the planet Venus
covering 21 years of the First Dynasty of Babylon. They are believed to
belong to the reign of king Ammis.aduqa. Since Kugler identified the “Year
of the Golden Throne” in 1912 AD, with his 8 th regnal year several possible
matches to the Venus data have been suggested. The currently favoured dates
are 1702 BC, 1646 BC, 1582 BC and 1550 BC for the first year of
Ammis.aduqa. The corresponding chronologies are called the ‘High’,
‘Middle’, ‘Low’, and ‘Ultra-Low’ respectively.6 We give Huber’s translation

5
The value for c is calculated from d (l.o.d.)/dt=0.0548*c (Stephenson & Morrison 1984, 56).
 4

of the main portion that was used to assign the series to king Ammis.aduqa
of the First Dynasty of Babylon (Huber 1977, 123):
“(Year 8) If on (month) IV (day) 25 Venus disappears in the west, remaining
absent in the sky 7 days, and on (month) V (day) 2 Venus appears in the east,
there will be rains in the land; desolation will be wrought. If on (month) XII
(day) 25 Venus disappears in the east. Year of the Golden Throne.
(Year 9) If on (month) III (day) 11 Venus disappears in the east, king shall
send greetings {var: declaration of war} to king.”
Huber (2000, 173) has concluded that the High Chronology best fits the
data. The periodicity of Venus is 8 sidereal years. Nevertheless, this is only
approximate because there is a -4.10 day shift in the 8-sidereal period or 99
synodic months. Babylonian astronomers also used a second calculation
which assumes a -4 day shift for the period (Gasche et. al. 1998, 73).
Besides, any solution using the data in omens 11 to 21 and 34 to 37 must
drastically emend the text (Hunger&Pingree 1999, 37). The most interesting
for the history of mathematical astronomy is section II, designated source δ.
It is probable that it was already included in another source α compiled
within a century or two of 1000 BC. This was simply a list of omens in
which the synodic period of Venus was 587 days.7 As Gurzadyan (2000, 181)
assumed only the relative sequence of the inferior and superior conjunctions
is reflected in the tablet and not the absolute lunar calendar. They lie at the
base of his approach to the 8-year cycles. This approach opens several
possibilities for the fall of Babylon.

The Solar Eclipse of Ur

240-251

“The king … On that day, in the foreign land … His roar … the hills …
The city which Enlil has… which An has ..., which Nintud has …, which
Enki has … good wisdom. Nanna has … the heights of heaven, Utu has …
on the horizon, Inana the lady of the battle has frowned (?) on it. The people
of the rebel lands … like a dead reed … The great and terrible battle of
Šulgi…”8

6
The term “Ultra-Low” is widely accepted for the Fall of Babylon in 1499 BC instead of the previous one
(in 1507 BC).

7
The Babylonian synodic period is comparable to that of Mayas for whom it was period of 584 days.
 5

A passage from Šulgi’s poem (Šulgi D) describes one of his wars

against Gutium. Klein (1981, 59; 81) has noticed that the passage expresses
two eclipses, of the moon and the sun respectively. The former eclipse is
connected with Nanna, the god of the moon, but the information available is
insufficient to draw any conclusions. The latter eclipse is connected with sun
god Utu. During the battle, there was an eclipse of the sun. It was possibly
preceded by the lunar eclipse. We found that only five eclipses of 2053 BC,
2044 BC, 2021 BC, 2006 BC and 1973 BC satisfy these conditions. They
match with 1579 BC, 1563 BC, 1547 BC, 1532 BC and 1499 BC
respectively (Table 6).

The Eclipse of Gutium (EAE 21)

In the famous battle of Utu-hegal (who according to king-lists is the

sole member of the Fifth Dynasty of Uruk) the king defeated Tirigan, a king
of Gutium, and so ended the long rule of the dynasty which had begun
immediately before the fall of Akkad, in the reign of Šarkališarri. According
to EAE, the battle and subsequent downfall of the Guti was attended by a
lunar eclipse (Huber 2000, 171; Rochberg-Halton 1988, 238):
“If an eclipse occurs on the 14th day of Du’ūzu (month IV), and it
begins in the west and clears in the south, variant: north; it begins in the
evening watch and clears in the middle watch. You observe his eclipse and
bear in mind the west. The prediction is given for the king of the Guti: The
downfall of Guti in battle. The land will be totally laid waste.”

The eclipse occurred on the 14th day of the 4th month, began in the
evening watch, and cleared in the middle watch.9

I do not pay much attention to this eclipse because the lunar positions
appear untrustworthy for further analyses.

The First Ur Eclipse (EAE 20)

8
The ETCSL project, Oriental Institute, University of Oxford, trans. t.2.4.2.04, (http://www-
etcsl.orient.ox.ac.uk/section2/tr24204.HTM; Klein (1981, 50-123).

9
The duration of a 'watch' is thought to be the difference in time between sunset and sunrise divided into
three parts (i.e. 4 hours).
 6

The next EAE account deals with the early part of the Dynasty of Ur
III; it is generally thought to mark the end of the reign of Šulgi (Huber 2000,
168).
“If an eclipse occurs on the 14th day of Simānu (month III), and the
god, in his eclipse, becomes dark on the east above, and clears on the side
west below,…(The eclipse) pulls out (issuh) the first watch, and 'touches' the
middle watch (so recension A; B has: ‘equalizes’ (imšul) the first watch). …
The king of Ur, his son will wrong him, and the son who wronged his father,
Šamaš will catch him. He will die in the mourning place of his father. The
son of the king who was not named for the kingship will seize the throne.”10
There are two versions of the text. In recension A, all eclipses begin
“above” and end “below”. Huber (2000, 168) gives a list of eclipses which
satisfies criteria for the recension A. Gurzadyan (2000, 178ff) confirms that
the eclipse of June 27, 1954 BC satisfies all criteria for the recension B.
The eclipse takes place on the 14th day of the 3rd month, beginning in
the first watch in the east, and ending in the west at the beginning of the
second watch. Three candidates are possible: in 2019 BC, 2002 BC and 1954
BC (Table 6).

The Second Ur Eclipse (EAE tablet 21 and 20)

The following lunar omen marks the end of the Dynasty of Ur III. The
eclipse is therefore attributed to Ibbi-sîn, the last king of the dynasty (Huber
2000, 168; Rochberg-Halton 1988, 248):
"If an eclipse occurs on the 14th day of Addaru (month XII), and it
begins in the south and clears in the north, it begins in the evening watch and
clears in the morning watch. You observe his eclipse and bear in mind the
south. The prediction is given for the king of the world: the destruction of
Ur... [variant: or Ur will be] destroyed, variant: an order to destroy its city
walls will be given. While the barley is being heaped up, the devastation of
the city and its environs (will occur).”

The eclipse takes place on the 14th day of the 12th month, beginning in
the south during the evening watch, eventually clearing during the morning
watch in the north. There are three eclipses of 1976 BC, 1962 BC and 1912
BC. The eclipse of 1976 BC begins in the south, but does not match the First
Ur III eclipse. The beginning of the third eclipse is in the north-east. It is not
10
Rochberg-Halton, (1988, 189; 191) translated issuhma with “passes” instead of “pulls out”, imšul with
“half” instead of “equalizes”.
 7

decisive argument for rejecting, so it will be discussed. Koch (1998, no.132)

has concluded that the latter eclipse did not satisfy all conditions. An
objection is the magnitude 0.57, quite low to be recorded as a major
historical event. However, it is likely that it was the partial eclipse because
of lacking word “dark”. We propose the lunar eclipse of 1962 BC (see §3).

There is another lunar omen which reports about the end of Ur (EAE 20
§ II) (Rochberg-Halton 1988, 187f):

“If an eclipse occurs, on [the 14th day] of Ajaru and the god, in his
eclipse, becomes dark on the side north above, [and] clears on [the side
so]uth below; The east wind (blows, and) the first watch [until (its) ris]ing(?)
…is visible with the sun. [Observe his ec]lipse, (that of) the god who, [in]
his eclipse became dark on the side north above, and cleared on the side
south below, and bear in mind east wind. […] there will be famine; the land
[will have] riches. The prediction is given for Ur: In Ur, Mars will [rise (?)
…]. […] Mars; there will be abundance; [The king (?)] of Ur’s reign will
end.

The end of line 7 is Uri palûšu iqatti which rather corresponded to the
end of dynasty of Ur III than to the end of a king’s reign (Hunger 2000,
155). The best fit is the lunar eclipse of May 16, 1961 BC (see §3).
 8

The Solar Eclipse of Assur

The solar eclipse over Assur during the reign of Narām-Sin was
mentioned in the year of the eponym Puzur-Ištar from the Mari eponym
chronicle (MEC). According to Veenhof (2000, 149) it comes one year after
the birth of Šamši-Adad. All solar eclipses magnitude higher than 0.8, from
1850 to 1750 BC were computed (Table 2). According to Huber (2000, 174f)
σ-error in computing an eclipse is half an hour in 1500 BC and one hour in
2000 BC. In this case, σ-error should be about 40 minutes (Tables 2 and 3).

Date Assur* Mari

43.26 E 35.46 N 40.9 E 34.55 N

-1846.04.02 0.486 0.509

-1832.06.24 0.995 1.028

-1817.09.06 0.819 0.777

-1807.08.16 0.888 0.909

Table 2
-1803.11.28 0.528 0.695
* M =
-1794.11.19 0.690 0.714

-1790.09.07 0.913 0.933

Date Assur* Assur
-1763.10.08 0.836 0.870
ΔT1=ΔT-2400 s ΔT1=ΔT+2400 s
Magnitude of
eclipse. M -1846.04.02 0.326 0.665 for a total
eclipse has the value
1. A partial or a near-
total eclipse -1832.06.24 0.958 0.988 has value
<1. The year -1846 is
1847 BC. All -1817.09.06 0.905 0.718 dates are
calculated by
Emapwin. -1807.08.16 0.886 0.855

-1803.11.28 0.771

-1794.11.19 0.670 0.755

-1790.09.07 0.881 0.952

-1763.10.08 0.770 0.926

 9

Table 3

* This value is more possible according to analysis of the eclipses in §3. For ΔT1= ΔT
+1800 s the solar eclipse of -1803.11.28 has M=0.911.

From the Table 3 we can see the magnitude of the solar eclipse in
Assur and Mari. From the Table 3 we find four eclipses to be much affected
with ΔT: 1847 BC, 1818 BC, 1804 BC, 1795 BC and 1764 BC. Only the
first eclipse is of relatively small magnitude.

The Eclipse of Mari (Tablet HC-A.25-115)

The next eclipse record (in chronological order) was discovered by

Parrot in the city of Mari, and translated by Dossin (1951). The lunar eclipse
from Mari occurred during the eponym Asqûdum which would be Year 11-
12 of Hammurabi (Warburton 2000, 59). Asqûdum was also the diviner of
the king11. Since the EAE archive only contains later copies of the originals,
this tablet is actually the oldest known lunar eclipse record (R. Chadwick,
pers. comm.):

“To my lord speak thus: Thus (speaks) your servant Asqûdum. On the
14th day (of the month), an eclipse of the moon took place. The taking place
of this eclipse was evil, (but) the omen for my lord was good and the omens
were good for the upper district (meaning the upper part of the Euphrates
river valley). Now (from) where my lord is, he should take these omens for
11
Asqudum was a diviner of the king Zimrilim, but the letter could have been sent to Yasmah-Adad at the
time of Šamši-Adad’s death or after (Warburton 2000, 62). This is the English translation, from the original
translation of Dossin, by professor Robert Chadwick with some minor corrections.
 10

his own good and for the good of the city of Mari, and let not the heart of my
lord be troubled. Also, would that my lord sends me a response to my tablet
so that he will put my heart at ease.”
The word for eclipse is antalû (from the Sumerian word AN.MI)
(written an-ta-lu-ú-um and an-ta-le-e-em, lines 5 and 7) (Dossin 1951, 46-
48; cf. Rochberg-Halton 1988, 32).

There are several total eclipses of the moon: September 01, 1781 BC,
May 31, 1748 BC, August 23, 1734 BC, March 10, 1716 BC, July 03, 1694
BC (before sunset) and December 6, 1684 BC (see possible matches in Table
6).

The Eclipse of Babylon I (EAE tablet 20)

The next omen presages the end of the First Dynasty of Babylon, and is
therefore attributed to its last king, Samsuditana (Rochberg-Halton 1988,
210f):
“If an eclipse occurs on the 14th day of Šabatu (month XI), and the god,
in his eclipse, becomes dark on the side south above, and clears on the side
east (var: west) below; the north wind (blows, and) (the moon) begins the
last watch, and he (the moon) is seen with the sun. His horns bend (toward)
the sky. He (the moon) did not obscure (cover up) his entire šurinnu, and
disappeared. You observe on the 28th, and the eclipse is near; begins and
becomes red; it (the šurinnu) will show you the eclipse. Observe his eclipse,
(that of) the god who in his eclipse was visible and disappeared, and bear in
mind the north wind. The prediction is given for Babylon: the destruction of
Babylon is near.”12

In the above, the lunar eclipse on month XI, day 14, begins in the last
watch in the south and is partially eclipsed in the west. Two weeks later on
month XI, day 28, a solar eclipse occurs and is visible throughout its course.

Huber et al. (1982, 40) noted the extreme rarity of eclipse pairs
comparable to the above description.13 We found only three eclipse pairs
between 1600 and 1490 BC: in 1547 BC, 1532 BC and 1508 BC. The
12
Line, concerning the eclipse on XI 28, would be replaced by the more accurate translation of Huber et al.,
40: “On the 28th you observe and an eclipse is close by; it begins and makes full (its time)”; Line § XI.2
regarding the lunar eclipse has impossible direction for that time of year. Huber replaced it by “west”. The
term šurinnu refers to the appearance of the moon while it is eclipsed in that point at which the eclipsed
moon looks like one of the crescent phases (Rochberg-Halton 1988, 62 n. 147).
 11

eclipse pair of the year 1532 BC gives very late beginning for the year; the
lunar eclipse in 1532 BC begins very late at 5:50 and ends after sunrise. The
solar eclipse in 1508 BC has a relatively small magnitude, M=0.41 and gives
early beginning for the year. According to 8-year cycle, the former eclipse
pair best fits the evidence.

Fall of Babylon M1* Date M2

-1546.02.15 0.81 -1546.01.31 0.25

-1531.04.29 0.44 -1531.04.14 0.81

-1507.12.26 0.41 -1507.12.10 -0.11**

Table 4
*
M1 - Magnitude of the solar eclipse, M2 - Magnitude of the lunar eclipse

** The lunar eclipse was penumbral one, which differs from the partial eclipse as cited
above.

The Eclipse of Tell Muh.ammad

The lunar eclipse is found in the economic texts from Tell

Muh.ammad, a site near Tell H.armal in Baghdad. The first tablet mentions
the lunar eclipse, the other having a year-name 38th year after Babylon was
resettled. These texts are from months Nisan and Abu (Gasche et al. 1998,
86; Gurzadyan 2000b, 45). The lunar eclipse of the month Abu has an
impossible day, 10, where 14-15th day of the month is expected. An interval
of years is assumed after it was occupied by the Hittites. We found the
13
I do not wish to agree with the view of Rochberg-Halton (2004, 271f) that these Babylonian omens
consists of signs and their divine decisions whose connections are variable and not dependent upon the
regular or even one-time observation of their co-occurrence. She also states that formulation “if X then Y”
in this context expresses only the belief the gods determine Y whenever X, unless the performance of an
namburbi ritual persuaded them to change their determination. In the earliest astrological texts usually
referred to as ‘astral omens’, to distinguish them from horoscopic astrology, the astronomical phenomena
were not yet predictable. We suppose some trained observer looking at the sky and watching the signs of
this kind. He knows an omen text from EAE tablets to be connected with the sign and he gives a prediction.
However, the caution is needed (see n. 24). In this case, the sun eclipse could have been predicted because
it was preceded by the lunar eclipse in the previous syzygy. There is also the claim in Tablet 23(24) that
certain features of the Sun’s appearance on the 1st day of the month allow one to predict a solar eclipse at
the end of the month (Hunger&Pingree 1999, 44).
 12

eclipse of the moon in 1459 BC to have begun too early. If we consider

duration of eclipse 111 m, it ends at 19h 40 m only one our after sunset 14. As
we do not have the time of the eclipse, it is not conclusive for the
chronology.

3. ANALYSIS OF ASTRONOMICAL OBSERVATIONS

In the previous section, we reviewed the astronomical evidence from

ancient Babylonian sources. In the following, we shall analyze the ancient
eclipses and give our interpretation and results.

There are a number of astral omens from EAE tablets which might have
been used for the chronology. Huber argues that the lunar eclipse omens of
Ur III were “so detailed and unsystematic that they appear to contain actual
records of observations” (cit. in Warburton 2000, 66). Koch (1998)
confirmed that one of the eclipses matched with 1954 BC and that the other
match was only partial, but the date was correct. These two omens were
based on observations. There is disagreement between Huber (2000) and
Gurzadyan (2000a) about the time of the beginning of the first eclipse and
the entrance angle of the second eclipse. Gurzadyan (2000a) takes into the
consideration the exit angle and watch time which is more objective
argument than the beginning of the eclipse. There are two problems with
these conditions. One must know the exact time of the beginning of the first
or the last contact (lunar eclipse has its penumbral phase, which is also
important). If it were wrong, we should not know which information was
valuable: the time of the eclipse, the entrance angle or the exit angle. If an
eclipse begins in the last watch, the entrance angle is important. If an eclipse
begins in the evening watch, the exit angle is more valuable. The basic
problem of selecting of an eclipse is which factor is decisive: the time, the
angle or even the magnitude. We should extend entrance and exit conditions.
The terminology of the omen texts is open to dispute. The terms “above” or
“below” do not indicate a geographical direction. Therefore, this details are
of no help for selecting of the eclipse (if it began on the “lower” east (or
“below”) the eclipse could not have been observed”) (Gasche et al. 1998, 74
note 296). In EAE 20 almost every eclipse begin “above” and end “below”
(Huber 2000, 169). Therefore, we must turn to observed eclipses and then
14
Lunar eclipses are placed on Espenak’s Eclipse Home Page:

http://sunearth.gsfc.nasa.gov/eclipse.HTM.
 13

choose eclipses from EAE tablets. If we find a sequence of eclipses, which

correspond to historical events for a long period from Dynasty of Ur III to
the fall of Babylon, it would confirm our chronology. The best way is
finding solar eclipses on the abovementioned interval and tries to fit lunar
eclipses into it.15 Why is it better way? Solar eclipses above one place are
more infrequent event than lunar eclipses. We presume solar eclipses, which
were reported in the most ancient sources, to be total or near-total.16 The
second problem is ΔT. For the uncertainties in the extrapolation of ΔT one
takes σ-error of ca. 30 minutes by 1500 BC and ca. 60 minutes by 2000 BC
(Huber 2000, 174) which is equivalent to an acceleration parameter c
between 28.5 and 33.5 s/cy2. The average value of c is 31 s/cy2. This is
slightly extended interval according to changes in the length of the day (see
§1).

Solar eclipses

If we analyze the Table 6 the best score gives year 1547 BC for the end
of Babylon. The first solar eclipse of August 1, 2021 BC is from the poem of
Šulgi. It was central-annular eclipse with M=0.79 at Ur (46.2 E, 31 N). The
solar eclipse was possibly preceded by the lunar eclipse. Šulgi began his
great wars with the destruction of Karahar in his 23 th year. He destroyed
Simurrum and Harši up to his 26 th year. He also waged war at Simurrum and
Karahar between Year 30 and 33, at Simurrum, Lullubum, Urbillum and
Karahar between Year 43 and 44 (Sigrist&Damerow 2001). If we take
northern location, near Kirkuk (44.39 E, 35.47 N) and Lower Zab River the
magnitude of the solar eclipse would be 0.7. There were also pairs of
eclipses in 2013 BC and 2006 BC. According to Tables 5 and 7 Šulgi began
his reign in 2048 BC. The former eclipse might have happened in his 35 th
year, which is not compatible with our source. The latter eclipse could have
happened in his 43rd year, which would be compatible with our source. The
latter solar eclipse was annular-total with M=0.9. The solar eclipse of 2021
BC could have occurred in his 26 th or 27th year depending on his first year. It

15
An Akkadica reviewer has kindly pointed that the crucial logic of the final solar eclipses for the end of
the First Dynasty of Babylon need to be elucidated. If one can actually find historical records with a
specified number of years between various solar eclipses this would clinch any chronological debate.

16
The following description of the area of obscuration is given in the omen protases of the eclipse series:
“If the eclipse, in its middle, becomes dark all over and clears all over” (Rochberg-Halton 1988, 49). This
omen obviously refers to the total eclipse. “If an eclipse begins in the south (quadrant) and clears in the
north (quadrant)”. In this case, eclipse became dark only over the left portion of the lunar surface. The
second Ur eclipse (EAE 21) was probably partial one.
 14

is important that all eclipses could have happened in the interval between
Year 23 and Year 48 in the time of his battles.17

The solar eclipse from Assur in 1808 BC begins at 7:05 and ends at
10:01. It was central-annular eclipse with magnitude 0.89. There are more
eclipses in the list: -1837.03.24, -1831.06.14,-1767.12.20 and -1764.04.24,
but they are much more affected with ΔT than eclipses in the Table 3. The
solar eclipses in the Table 6 have relatively high magnitude in most cases.
Other eclipses are inserted because they are described in the literature (see in
Gurzadyan 2003, 16). According to the Table 3, the solar eclipse of 1833 BC
was presumably the total one, yet only the partial one in accordance with
changes of ΔT.

The Babylonian solar eclipse of 15 February 1547 BC has a relatively

high magnitude, M=0.81.18 It begins at 12:10 and ends at 15:24. More
important factor is that Babylonian eclipse coincided with the 8-year cycle
of Venus. Šamši-Adad appears to have lived between 1809 and 1728 BC. He
died between Year 12 and Year 18 of Hammurabi, possibly in Year 17
(Charpin&Durand 1985, 306-308). The latter reign began in 1744 BC. The
year matches with Venus cycle for year 1547 BC. The interval from the
beginning of his reign and the end of Babylon is 197 years.

Lunar eclipses

One should differentiate eclipse reports about partial and total lunar
eclipses. The eclipse report EAE 20, Month XI has description of the partial
eclipse where the moon was seen with the sun. There is also report in EAE
20, Month I where eclipse began in the last watch, and set while eclipsed. It
is not the case with the Babylonian lunar eclipse.

17
It is common for Year-name or Year formula to refer to the previous year (Year-name N= Year N-1).
Simurrum is placed at Lower Zab near Altun-Köpru (44 E, 36.2 N). Karahar (Karhar) and Harši were
located in southern Kurdistan (Gadd 1971, 601).

18
The solar eclipse of Assur is little affected with ΔT. The solar eclipse of Ur has magnitude 0.79 for ΔT
=46370 s. For ΔT=43970, which is 2400 s less than calculated value with Emapwin, M=0.9 and c=29.7.
The solar eclipse of Babylon is central-annular eclipse for ΔT =35978 s and M=0.81. For ΔT=33798, which
is 2180 s less than calculated value, estimated magnitude would be 0.96 and c=29.7. We presume all solar
eclipses to be total or near-total (M >=0.96). It is obvious that ΔT has less value for that period than
calculated by Emapwin where c has average value 31 s/cy 2. According to the current theory of Stephenson
and Morrison (2000, 74) parameter c is between 29.2 and 32.8 s/cy2.
 15

The first lunar eclipse of Ur III of July 18, 2002 BC started on the east
(117°) at 20:24 and ended on the west (242°) at 23:46.19 Reporting of the
second lunar eclipse of Ur III has problem with duration: from the evening
watch to the morning watch. This might be the scribal error or only a
schematic time. The maximum duration of a total eclipse is six hours (from
the first contact until the moon leaves the penumbra). It can be, however, be
understood as an extention of the schema for eclipse duration where eclipses
spanning the first to second watch of the second to third watch are collected.

The interval between the end of Ur III and the end of Babylon is 409-
412 years. We should search the second lunar eclipse of Ur III in that
interval according to alternative chronologies. There is also problem with the
month of the eclipse. There is prediction in EAE 20 for the second month,
given for Ur, which describes its destruction. This was the total lunar eclipse
of May 16, 1961 BC, started on the south-east (107°) before sun set, ending
on the south-west (257°) at 19:02 about half an hour after sun set 20. The exit
angle agrees with the eclipse report. More detail that is important is
conjunction of Moon and Mars on the same day (2.2°). This eclipse perfectly
fits the report. There is also the total lunar eclipse of May 27, 1962 BC
which might have matched our sequence. The latter eclipse started on the
south-east (115°) at 22:16, ending on the south-west (249°) at 2:00. The exit
angle of the eclipse does not fit very much (it should be > 260°). There
might have been two lunar eclipses which presaged the end of Ur. The end
of Ur III could have happened in the next three years. Ibbi-Sîn made for
Inanna a harp in his 21st year in order to induce mercy in the goddess and
moon god Nanna (Sigrist&Damerow 2001).21 It might have been connected
with the lunar eclipse. The Mesopotamian year officially started with the
first appearance of the moon in month 1. In certain texts, particularly from
the Old Babylonian period, the 15th of month 12 (or full moon) was ideally
vernal equinox (Brown 2000, 107).22 In the particular case, the vernal

19
North = 0°, East=90°, South=180° and West=270°. East is on the right. According to recension A the
eclipse ‘touches’ the middle watch (the first watch (19:00-23:00), the second watch (23:00-03:00), the third
watch (3:00-7:00)). The beginning of the first watch is after sun set.

20
If we include standard error σ of ca. 60 min for ΔT the result is better. Moon was seen with the sun in its
rising. The lunar eclipse had to start later to have been seen with the sun (at twilight). It is significant for the
history of astronomy because ΔT demonstrates smaller growth at the beginning of the second millennium
than we would expect according to its formula. The same case could have happened later. See Solar
eclipses above.

21
Sollberger (1976, 4) noted a lunar eclipse before Year 22 of Ibbî-Sin.
 16

equinox was around April 8. If two months were intercalated (April-May),

the eclipse would fit.23

Were one to identify a lunar eclipse which would correspond to the date
of the end of Guti, according to our chronology this eclipse would be more
probable. According to our chronology, the beginning of UR III was ca.
2066 BC. The king-list gives a reign of seven and a half years for Utu-hegal,
and the eclipse is believed to have occurred during his accession year. An
alternative scenario has been proposed whereby Ur-Nammu was actually a
close relative of Utu-hegal, governing Ur on his behalf (Hallo 1966).
However, there is a strong possibility that the Guti were defeated about the
10th year of Ur-Nammu (year-name k) of Ur III Dynasty (Sigrist&Damerow
2001). According to our interpretation, the only eclipse that might have
corresponded to the eclipse report was that of June 06, 2056. It started at
22:31 and ended at 2:06. It does not fit very well.24

According to our chronology, the lunar eclipse from Mari perhaps

recorded during the reign of Yasmah-Adad and therefore dated to Year 11 of
Hammurabi occurred on August 23, 1734 BC. It was a total lunar eclipse
which started before midnight, lasting for two hours.

The lunar eclipse of February 31, 1547 BC begins in the last watch in
the south and is partially eclipsed in the west. It started on the north-east
(65°) at 1:18 and ended on the north-west (294°) at 3:06. Magnitude of the
eclipse is 0.26. The moon is seen with the sun in the last watch (3:00-7:00).
For the first contact, the typical accuracy of the prediction is 1.12 hours,
improving to 0.95 hours after 550 BC. The mean error of the lunar eclipse
times predicted by the Babylonians in last contact is 2.63 hours (between
731 and 77 BC) (Steele&Stephenson 1997, 125-128). The error for earlier
22
In EAE 14, the equinox, indicated by a night-time 3.0 UŠ = 6 bēru, on day 15, is assumed to fall on the
fifteenth of the equinoctial month (1 UŠ= 4 min, 180*UŠ=12 h is measured time from sunset to moonset)
(Hunger&Pingree 1999, 45ff).

23
According to the Drehem calendar, the beginning of the year falls in the range 328° to 41° (the vernal
equinox is 0°). There were intercalations in the first two years of Ibbi-Sîn (Huber et. al. 1982, 39 and 82). It
is possible that in the last chaotic years of his reign there were intercalations in three consecutive years.

24
The question of empirical veracity has been raised by Leichty in his treatment of the series Šumma izbu.
His opinion was that most of the birth anomalies described in the protases (if-clause) could be identified
with attested birth abnormalities. Consequently, he said that we do not wish to agree that all the omens in
the series were actually observed. He also noticed that in the expression and redaction of omen collections,
additional omens were introduced, not on an empirical basis but based on the requirements of formal
schemata into which phenomena were arranged. It is difficult to judge what may be the implications of the
creation of the schemata in the terms of omen texts (Rochberg-Halton 2004, 251).
 17

predictions was greater. If we include standard error of ca. 30 min for ΔT it

is obvious that the lunar eclipse started very close to the last watch and
ended at dawn.

Tell Muh.ammad eclipse matches perfectly 1547 BC. It was the total
lunar eclipse of May 26, 1506 BC which started at 21:40 and ended at 1:36.
It occurred 41 years after the end of Babylon and 38 years after its
resettlement.

4. HISTORICAL EVIDENCE

Long Chronology

The Long Chronology is not supported by any evidence except the

astronomical data and statistics. The Mari correspondence mentioning Hazor
and Cypriot copper must postdate the Egyptian Twelfth Dynasty. According
to Beckerath (1997), the beginning of the dynasty is placed at ca. 1991 BC
(without coregencies) or ca. 1976 (with coregencies) and the end at ca.
1795/3 BC. The Lower chronology for the Twelfth Dynasty would set the
beginning of the dynasty, according to Krauss, at ca. 1951 or 1938 BC (with
coregencies) and the end at ca. 1759 BC (Warburton 2000, 41). Kitchen
(2002, 11) has suggested that the length of the Second Intermediate Period
would be 240/250 years. The beginning of the Eighteenth Dynasty is placed
at the interval between ca. 1550 and 1530 BC, independently of lunar dates.
The dates are based on the Sothic sightings and are generally accepted
(Kitchen 2002, 11f).

Assyrian Calendar

The Assyrians were using two calendars before the time of Tiglath-
Pileser (1114-1076 BC). They were using a solar calendar for appointment
eponyms, and most probably kings, in the second millennium. Koch assumes
that, in the absence of ‘intercalation rules’, in that period the vernal equinox
was to correlate the calendar year with the solar year. Larsen has established
the fact that the Old Assyrian calendar started with Bēlet-ekallim as first
month, which was synchronous with the sixth month of the Mari spring
calendar, IGI.KUR. Hence, the Assyrian calendar and eponymy year started
with the first new moon after the autumnal equinox (Veenhof 2000, 142).
The Assyrians also used a purely lunar calendar. The date when a solar
 18

calendar was adopted is not known. Since the lunar year is 11 days shorter
than the solar year, this would mean that 100 years in the king-lists, before
ca. 1100 BC, correspond to a real total of ca. 97 years. The problem is still
open.
From the letters from the Middle Assyrian period it is evident that
during the period, ca. 1270-1200 BC, spring harvests and the repayments of
corn loans occur during the months of Sin (4) and Kuzallu (5) and that
autumn corn loans are negotiated in the months of Muhur-ilani (10) and
Abu-šarrani (11).25. If, over these c. 70 years, the calendar consisted of only
12 lunar months then, these activities should have cycled twice through the
lunar year. However, they do not. Harvest and planting occur at the same
time in the Assyrian calendar. Therefore, during this period, the Assyrian
calendar was solar and the intercalation must have occurred.
We assume the use of a solar calendar for regnal years. If this were
wrong, this would extend reigns of Aššur-nādin-ahhē and Aššur-rabî I ca. 15
years.26 We could also lower the reign of Šamši-Adad as much as 5 years
(from Year 12 to Year 17 of Hammurabi, 1733 to 1728 BC according to our
chronology).

Old Assyrian Chronology

The discovery of Kanish Eponym List (KEL) has serious consequences

for the reconstruction of the Old Assyrian chronology. This list overlaps with
the so-called ‘Mari Eponym Chronicle’ (MEC). The solar eclipse was in the
126th year of KEL. According to Veenhof (2000, 139) the temporal distance
between the accession of Irišum I and the death of Šamši-Adad I was 199
years. He also used ‘Distanzangaben’ of Esarhaddon and Shalmaneser I: 126
or 159 years (the difference is the rule of Šamši-Adad). Veenhof accepted
these years because only four years should be added to the Kanish Eponym
List (KEL) for the lacking eponyms. According to the reconstruction of
KEL, this eclipse comes 126 years after accession of Irišum I and one year
after the birth of Šamši-Adad. Šamši-Adad died in the 199 th year of KEL at
the age of ca. 75. Veenhof tried to fill few gaps in the last part of KEL.
According to him, the first manuscript KEL A has 129 years. It overlaps with
25
http://www.geocities.com/farfarer2001/assyrian_letters/ urad-serua/urad_serua_corn_loans.HTM.

26
There are several possibilities: For Šamši-Adad’s reign, 1765/60-1733/28 BC, two missing reigns would
be placed at 1467/62-1423 BC (according to the lunar calendar) or 1459/54-1430 BC (according to the
solar calendar). Two hundred and seventy four years separated Išme-Dagan from Āššur-nādin-ahhē
(according to the SDAS version of the Assyrian King List) or 266 years according to the lunar calendar. If
we raise our chronology 5 or 10 years according to lunar calendar, this will extend missing reigns too
much.
 19

MEC A *1-*23 (=KEL 103-125). After the 129th year, there is a gap of 4
years. Then begins MEC B *1-*31 overlapping with the Assyrian Kinglist
(AKL). If we examine the list closely, we notice that in the reign of Ikūnum
we must add two years. There are two names on the position 40 in KEL B
(40A and 40B) Šuli and Irišum. Consequently, Edzard (2004) noticed that
the sum of years in KEL A amounted to 131 years. Veenhof (2003, 71f)
places Iddin-Suen, brother of Šuli, at the position 41B after Šuli.
Consequently, we could extend his chronology with additional years, or we
could reduce it, as there is a gap between KEL and MEC. Nevertheless,
Veenhof (2001, 144; 2003, 13f) chooses a solution which has Šuli being the
eponym of that year (as in KEL A 41A) but during that year he was replaced
by Iddin-Suen, his brother (KEL A 41B). This would be the only time in the
list that an eponym is called brother of X, rather than son of X. He
eliminates Irišum as an error in the list. If the list were accurate, there would
be only one eponym for a year. Veenhof (2000, 139) yields the following
scheme according to the Middle Chronology:27

Irišum KEL 1-40 40 years (ca. 1982-1943)

Ikūnum KEL 41-54 14 years (ca. 1942-1929)

Šarrukīn KEL 55-94 40 years (ca. 1928-1889)

Puzur-Aššur II KEL 95-102 8 years (ca. 1888-1881)

Narām-Suen KEL 103-129 27 years (ca. 1880-1854)

Narām-Suen gap 1 4 years (ca. 1853-1850)

Narām-Suen/Irišum II MEC B *1-*26 26 years (ca. 1849-1824)

Irišum II MEC B *27 (ca. 1823)28

Missing MEC *28-43 (ca. 1822-1807)29

27
The reader is certainly confused by Veenhof’s use of the Middle chronology dates in a chronological
discussion, and such must be avoided (pointed by Akkadica reviewer). Hence, we shall use the relative
chronology of KEL

28
If Šamši-Adad took Ekallātum in MEC B *31 his first year would have been in the missing line
MEC B *32. In that case, his first year in Assur would correspond to position MEC B *35. There
would be one year more in the list.
29
According to kne KEL G list of eponyms. That list starts at Kel A 111 in the eponym Samaya
1834 BC after Middle chronology and ends in the eponym Ibba-Ištar 1748 BC. List was
 20

Šamši-Adad AKL (Ekallātum) 3 years (ca. 1811-1809)

Šamši-Adad AKL (Assur) 33years(ca. 1808-1776/5)

According to Veenhof, Šamši-Adad is born in Year 125 of KEL. A solar

eclipse took place in Year 126 of KEL, one year after the birth of Šamši-
Adad. It is possible to include new eponyms between KEL 129 and MEC B
or between MEC B and AKL. Šamši-Adad died between Year 12 and Year
18 of Hammurabi possibly in Year 17 (Charpin & Durand 1985, 306-308). It
depends on the gaps in KEL. He was born in the Year 125 and died in the
Year 199 of KEL. He died at a minimum age of 75 years and one can only
force this number upwards. This number should not have been much higher
than 80. We are not certain about the missing years in the reign of Narām-
Suen, but they should not exceed 10 and the age of Šamši-Adad ca. 81 (cf.
Warburton 2000, 61). According to AKL, he went to Karduniaš during the
time of Narām-Suen. Then he captured Ekallātum. He resided for three years
in Ekallātum. He removed Irišum II, son of Narām-Suen, from the throne
and ruled for thirty-three years (Grayson 1980-83, 106). The number of
years he spent in Babylonia is not known. It is generally believed Narām-
Suen from Ešnunna to have conquered Assur and Ekallātum. Šamši-Adad
had to flee to Babylon. He returned to Ekallātum after 7 years (Mieroop
2003, 101). It is likely that Ešnunna occupied Ekallātum, but not for a long
time. The last mention of Ipiq-Adad of Ešnunna is in MEC *25. The
departure to Babylon should have been in the next year when Narām-Suen
of Ešnunna seized Ašnakkum (Sigrist&Damerow 2001). It is possible to
insert additional years between MEC B *31 and AKL. If Šamši-Adad
returned to Ekallātum after the end of MEC B *31, we should add two years
at least. However, we may not prolong the life of Šamši-Adad very much. If
we believe in the statements of Shalmaneser and Esarrhadon, error in years
might have been insignificant. It is not likely that he would die at the age of
85 according to the Ultra-Low Chronology (Table 5). We have two
alternatives for the sun eclipses in 1808 BC and in 1804 BC. As we can see
from the first table, 166 years separated the death of Irišum from the death of
Šamši-Adad. The distance in years differs from the ‘Distanzangabe’ by 7
years. In the second case, the difference is only 3 years.30

published by Veenhof. I haven’t copy of his article, but I have found it on internet site Ancient
Near Eastern Chronology Forum under ‘MEC and KEL’, Anonymous answer to Joe (July 16,
2012).
30
Including the solar eclipse of November 28, 1804 BC in our calculation would reduce the age of Šamši-
Adad only 4 years. This will not affect our chronology very much. ‘Distanzangabe’ of Esarhaddon is
doubted. This would be 5*25 +1 for 5 generations between the death of Irišum and Šamši-Adad (cf. Gasche
 21

CHRONOLOGICAL SCHEME ACCORDING TO SOLAR ECLIPSE IN 1808 BC

Death of Irišum KEL A 40 ca. 1895 BC

Accession of Ikunum KEL A 41 ca. 1894

Accession of Naram-Suen KEL A 103 MEC A 0 ca. 1832

Birth of Šamši-Adad KEL A 126 MEC A 23 ca. 1809

Solar eclipse KEL A 127 MEC A 24 in 1808

End of KEL KEL A 129 MEC A 26 ca. 1806

Naram-Suen 130-133 (gap) MEC A 27-30 ca. 1805-1802

Naram-Suen/Irišum 134-160 MEC B*1-*27 ca. 1801-1775

Accession of Šamši-Adad MEC B*8 ca. 1794

Šamši-Adad in Ekallātum AKL ca. 1763

Šamši-Adad in Assur AKL ca. 1760

Death of Šamši-Adad AKL ca. 1728

# MEC – ‘The Mari Eponym Chronicle’, Birot 1985; Durand&Guichard 1997, 42f,
mentions the solar eclipse from Assur. We have used some information from the
translation of Glassner (1993, 157-160). The variants A and B are used as in Veenhof
2003, 47-50.

et. all., 157-161).

 22

Assyrian King-List

Concerning the reconstruction of ‘Distanzangaben’ Reade (2001, 4f)

has recently supposed that Aššur-rabî I and Aššur-nādin-ahhē I reigned 30
years altogether. This figure is accepted in the table bellow. He assumed that
96 years had been assigned to 6 tuppišu rulers.31 From the statement of
Shalmaneser 580 years elapsed between him and Šamši-Adad I. According
to the statement of Esarhaddon 580 or 586 years elapsed between the
restoration of Ashur Temple and Shalmaneser, 679 + 586=1265 BC. If we
add 484 years to Shalmaneser we obtain ca. 1749 and 1717 BC for the
beginning and the end of the reign of Šamši-Adad. If we emend the reign to
ca. 1747 and 1715 BC there is a difference of 13 years between our
chronology and the latter date32. If we assign 36 years to Aššur-dān, 13 to
Ninurta-apil-ekur and 24 (instead of 14) years to Puzur-Aššur, the difference
will be only 3 years.33 However, there are some problems with the reigns
after Šamši-Adad I. His son Išme-Dagan I was on the throne at least for 11
years (Reade 2001, 57). He is assigned 40 years in the Assyrian King List
(AKL). According to the Assyrian King List Fragment (VAT 9812) his
descendants were Mut-Aškur and Remu-x. Puzur-Sin defeated a
descendant of Šamši-Adad named Asinum. Asinum was most likely a
grandson of Šamši-Adad or Remu-x himself. He might have reigned at
Ekallātum near Assur (Reade 2001, 6). It appears that Išme-Dagan I reigned
from Ekallātum until his 11th year. Reade identified IB.TAR-Sin with Puzur-
Sin and reduced AKL ca. 46 years. There is no need for reducing AKL. It
would appear that all descendants of Šamši-Adad have reigned for 40 years

31
The same principle might apply for the total number of 576 years (and 9 months) assigned to 36 Kassite
kings in the Babylonian king-list.

32
Shalmaneser I reigned 1273-1244 BC, Aššur-dān 1178-1133 BC (Brinkman 1977). If we assumed for
Ninurta-apil-ekur to have reigned for 3 years instead of 13, Shalmaneser I would begin 1263 BC.

33
The possibility of reconstructing 46 in Nassouhi King List is rather difficult since “40” would have been
written by four wedges in two rows. Ninurta-apil-ekur is assigned 13 years in the same list. It would
probably be preferable to accept a total of 49 years for these two reigns rather to pick the higher figure in
each case and arrive at a sum of 59 years. Brinkman (1977) follows higher figures for both rulers.
 23

until Aššur-dugul.34 AKL recorded 6 kings (eponyms ?) who reigned bab

tuppišu, presumably in the time of Aššur-dugul. We should assign them 0
years. Nevertheless, we don’t have any information about Puzur-Sin in the
main version of AKL. He might have ruled only briefly because he was not
included in the later lists.

Babylonian King-List

According to our chronology, Agum II returned the statue of Marduk

after 24 years in 1523 BC.35 His successor Burna-Buriaš I might have
reigned between 1500 BC and 1485 BC or little later. According to the
Synchronistic History (I 5’-7’) Puzur-Aššur III was a contemporary of
Burna-Buriaš I from the Kassite Dynasty (Grayson 1975, 157f). According
to our chronology, Puzur-Aššur III reigned from 1499 to 1476 BC. If 1499
BC were chosen then his reign would be ca. 1475 BC. It would not be
compatible with 14 years of Puzur-Aššur (1489-1476 BC) and 36 years of
Aššur-dān (1168-1133 BC). If the total of regnal years of Aššur-rabî I and
Aššur-nādin-ahhē I were reduced to 20 years, it would be comfortable for
Burna-Buriaš. The only complication is the identification of Agum III with
Agum II (Kakrime) which is not tenable. According to the Chronicle of
Early Kings, Agum III was a son of Kaštiliaš, who campaigned against
Sealand (B r.15-18), which had been conquered by Ulam-Buriaš, 36 his
brother (B r.13f) (Grayson 1975, 155). An inscription of Agum II (Kakrime)
did not mention Sealand (which is supposed to be near the Persian Gulf): “(I
am) a king of the Kassites and Akkadians, the king of the broad land of
Babylon, the one who (are) settled the land of Eshnunna with an extensive
population, king of Padan and Alman, king of the Gutians” (Gasche et. al.
1998, 89). In his inscription, Agum-Kakrime stated that he was a son of
Urzigurumaš (Tazzigurumaš) (Grayson 1980-83, 134). Therefore, he should
not have been identified with Agum III.37 Two added rulers, Hurduzum and
Šiptaulzi were separated ca. 50 years (III and II stratums at Tell

34
According to AKL Irišum I ruled for 40 years. Veenhof’s (2000, 139) reconstruction of KEL yields 40
years for Šarrukīn (Sargon I). The figure for Išme-Dagan I may have been artificial because there was a gap
in the list of eponyms.

35
Agum II (Kakrime) may have restored statues of the gods Marduk and Zarpanitum (Sarpanitum), which
had been removed from Babylon by previous Kassite ruler (Leick 2002, 250).

36
There is a possibility that Ulam-Buriaš was only a ruler of Sealand, because his inscription called him
“King of the Sealand” (Rowton 1970, .233). According to Grayson (1980-83, 121) reading of traces in K.
L. 12 I 22 is uncertain. He quotes Weidner who remarked that only –l[am] was “einigermassen
wahrscheinlich”.
 24

Muh.ammad) (Gasche 2003, 216, Table 1).38 Their presence in the list of
kings fills the gaps in chronology at the end of the Old and the beginning of
the Middle Babylonian period. According to Gasche et. al. (cit. in Gasche
2003, 216, Table 1) maximum six generations of the Kassite Dynasty lasted
for ca. 120 years (1489-1369 BC). It is compatible with ca. 20 years for one
generation. According to our chronology, Agum II (Kakrime) returned the
statue of Marduk and Sarpanitum to Babylon in 1523 BC, which makes 154
years for seven generations or 22 years per generation. The Babylonian list
of rulers is not conclusive for the chronologies in question.

Hittite chronology

In view of the references to Hittite rulers and their campaigns against

Syria we have to consider a number of generations within the period
between Muršili I and Muršili II. The eclipse of the sun took place in 1312
BC while Muršili II managed military operations against Azzi in both his
ninth and tenth years (Beckman 2000, 22). According to our calculation,
only one eclipse of the magnitude greater than 0.9 is possible. The eclipse of
June 24, 1312 BC was visible in the northern part of Anatolia. Beckman has
calculated an average generational tenure of the Hittite monarchs. He
divided period between the death of Hantili I (9 years after the fall of
Babylon) and accession to the throne of Muršili II (1321 BC) among eleven
generations. He concluded that the Middle Chronology best fits the evidence
(Beckman 2000, 25). The generation should be counted as between twenty
and thirty years. The Ultra-Low Chronology yields an average tenure of
15.36 years per generation; our chronology 19.72 years per generation; the
Low Chronology 18.27 years; the Middle Chronology 24.01 years per
generation; the High Chronology 29.18 years per generation. It is also
possible to change the number of generations. The filiations of Zidanta II
and Huzziya II are unknown. Tudhaliya I was probably the son of Huzziya
II39, but Bryce (1999, 132) suggested him to be the grandson of Huzziya II.
The number of the generations could have been from 10 to 12 depending
37
His inscription is doubted. It appears to have been a copy of the Old Babylonian inscription from the
seventh century BC (Astour, 1986). The title of Agum Kakrime “king who causes the four world quarters to
exists permanently” is of disputed authority and do not fit into the early Kassite period (Sassmannshausen
2004, 64). Grayson (1980-83, 121) has identified only Agum I (K. L. 12 I 11) in the Synchronistic King List
(SKL). Agum II should have been placed in K. L. 12 I 18 of SKL as Weinder had proposed. The missing
reign in K. L. 12 I 20 could have been the reign of Šiptaulzi.

38
Sassmannshausen (2004, 64) states that Cole’s placement of Tell Muh.ammad texts after Agum (III)
seems forced. Šiptaulzi is dated 30-50 years after the fall of Babylon, ca. 1470-1450 BC (Gasche 2003,
216). The former writer dated Tell Muh.ammad texts ca. 1540-1490 BC. Therefore, Šiptaulzi may have
reigned before Burna-Buriaš I.
 25

upon the generations of Kizzuwatna kings. It appears that the Ultra-Low

chronology is not possible, but we have also several possibilities with other
chronologies.40

Gates (1981, 37) has achieved the similar conclusion with Alalah. She
has yielded ca. 1525 for the fall of Babylon without help of astronomical
data. Alalah VII ended in the first quarter or the first half of the 16 th century
BC, shortly before the appearance of Cypriote Bichrome Ware pottery
(Gates 1981, 17-22). At Tell el-Dab‛a, Bichrome pottery occurs in the Late
Hyksos stratum D/2 and eighteen-dynasty strata D/1 and C/3 (ca. 1560-
1500) (Gates 2000, 88; Bietak&Kopetzky, 2000, 22f). There are different
opinions about the duration of Alalah VI (from Hattušili’s attack on Aleppo
to Muršili’s attack on Alalah), but ca. 35 years would be appropriate (van
Soldt 2000, 103, 108, 116). Therefore, we could take ca. 1575/1550 to
1540/1515 BC for Alalah VI. Muršili attacked on Babylon after he had
captured Alalah.

Albright (1965, 42) originally identified the name written Yantin-

Hammu in the Mari texts (dating to the reign of Zimrilim) with Yantin of
Byblos whose name is associated with Neferhotep. Ryholt (1977, 408)
places Neferhotep at 1742-1731 BC, but his reign might have been lowered
for 10 years (Warburton 2000, 63 no. 14). Zimrilim ruled Mari from
Hammurabi Year 18 to 32, 1727-1713 BC (Table 5). One generation before
him, s.htp-ibr‘ with Horus name s.wsh-t³wy was contemporary of yu-ki-in-ili
or Yakin-el (Ryholt 1990, 110 no. 35), ca. 1783-1781 BC (Ryholt 1997, 408)
or ca. 1773-1771 BC. Although this synchronism with Egypt is tentative, it
should not be rejected.41

Now we shall present alternative chronologies concerning problem of

fixing the Babylonian chronology:

39
Beckman 2000, 27.

40
I am not convinced that there were Tudhaliya I and Tudhaliya II between Huzziya II and Arnuwanda I as
proposed by Freu (1994, 39). Hattušili II might have been a coregent of Arnuwanda I, not a father of
Tudhaliya II (Bryce 1999, 154).

41
There is another synchronism with Egypt. Seals which postdated Ur III can be linked to Amenemhat II
(1914-1879/6) via Tod treasure (Warburton 2000, 51 and n. 7).
 26

Beginning of Ur Death of Šulgi End of Ur Šamši-Adad42 Hammurabi Fall of Chronology

(max. age) Babylon

2113(2112) 2048(204743) 2006(2004) 1848-1776 1792-1750 1595 Middle

2097(2096) 2032(2031) 1990(1988) 1834-1760 1776-1734 1579

2084 2019 1976 1819-1744 1760-1728 1563

2066 2001 1959 1809-1728 1744-1702 1547

2049(2048) 1984(1983) 1942(1940) 1792-1712 1728-1686 1531 Low

2025(2024) 1960(1959) 1918(1916) 1765-1688 1704-1662 1507

2018 1953 1911 1765-1680 1696-1654 1499 Ultra-Low

Table 5

5. DENDROCHRONOLOGY

Dendrochronology date offers 1774 +4/-7 BC for the construction of

the Sarikaya Palace at Acemhöyük. Sealings of Šamši-Adad I and his
officials were found in an archive collection and must postdate its
construction (Manning et. al. 2001). The destruction of the Sarikaya Palace
was after 1766 +4/-7 BC. According to the Middle Chronology, commercial
contacts with Acemhöyük lasted some sixty to seventy years, between ca.
1800 and 1730 BC ten years after the beginning of Šamši-Adad’s reign. The
earliest date for Šamši-Adad should have been 10 years before 1774 +4/-7
BC, 1784+4/-7 BC and the latest date should have been after 1776+4/-7 BC.
Therefore, we could have assigned ca. 1770 BC as the latest year for the
beginning of his reign. It would be compatible with our date for greater σ-
error.44 However, there are some problems with dendrochronological dates.
42
It is maximum age according to the solar eclipse.

43
Dates for Ur and Babylon follows CAH, Chronological Tables and Brinkman, 1977 for the Middle
Chronology. The Low Chronology is 64 years lowered. For the Ultra-Low Chronology see Gasche et.al.,
1998.

44
. There is not a problem only with dates. As a statistician, I disagree with a statistical method that is used
for σ-error. The most commonly used method for statistically matching tree-rings relies on what are called
“t-scores” and “g-scores”. The statistical method used in Anatolian tree-ring studies was “D-score”. D-score
combines the t-score and the g-score in the following formula g*t-t/2. The formula is an arbitrary one and it
 27

According to the dendrochronology the level II of kārum Kanish (Kültepe) is

represented by a 521-year tree-ring chronology which lasted from 2544 to
2024 BC, built from the juniper door-threshold timbers of rooms in the old
palace of Eski Saray. It is contemporary with the level II of kārum Kanish
(Newton & Kuniholm 2004, 166f). The level II of kārum Kanish came to an
end ca. 1836 BC having lasted for 138 years, from ca. 1974 BC (Veenhof
2000, 139f; 2003). The difference between historical and
dendrochronological dates is at least ca. 50 years. If a half-century of rings
missing from 2024 BC (as authors presume), it would be compatible with
Middle Chronology. Disregarding it, the Anatolian tree-ring date appears to
be too early regarding any here proposed chronology. Gasche (2003, 207f)
has recently pointed at the similar problem with the 14C chronology and a
thermoluminescence.

SOLAR AND LUNAR ECLIPSES

Mari (lunar)
Ur (lunar, I Ur (lunar) II Ur (lunar) Tell Assur (solar) Fall of
solar) Muh.ammad Babylon
(lunar)

1579
-2052.04.13 -1832.06.24
-2052.04.29
-1747.05.31
1563
-2043.04.20* -2018.06.26 -1975.03.04 -1520.03.13 -1817.09.06

-1733.08.23
-2020.07.17 -2001.07.18 -1961.05.27 -1505.05.26 -1807.08.16 1547
-2020.08.01 -1960.05.16 -1803.11.28

-1715.03.10
-2005.09.29 -1790.09.07 1531
-2005.10.14

-1693.07.03
-1466.04.15 -1763.10.08 1507

-1683.12.06
-1972.01.17 -1953.06.27 -1911.03.16 -1458.05.16 1499

Table 6
has no mathematical derivation (cf. Keenan 2005).
 28

* There were only solar eclipses in 2044 and 1973 BC.

Conclusion

If we analyze the Tables 4 and 6, we find the Middle chronology to be

excluded. Only one solar eclipse in 1847 BC with small magnitude is not
sufficient for accepting it. If we consider the solar eclipse in 1833 BC we
have better result for the reign of Šamši-Adad. This result is closer to the
estimated age according to Veenhof’s chronology (ca. 75 years). The lack of
other eclipses makes this chronology less probable. The next eclipse in 1818
BC offers better chronology. If one excludes two Babylonian eclipses, that
chronology has almost the same probability as the Ultra-Low Chronology.
The latter chronology has a good result, with the problematic solar eclipse in
1764 BC, which extends the age of Šamši-Adad to 85. Our chronology has
the best score.45 There are two advantages: the first is compatibility with
AKL, the second is the great solar eclipse in 1547 BC. Hunger (2000, 158)
has made an objection with EAE eclipses especially with the eclipse of
Babylon. An omen for the fall of Babylon said that Ur had conquered
Babylon, which is impossible statement for the period in question. This
statement could have been added later.

The following chronological scheme could be reconstructed within the

framework of our chronology:

45
Sassmannshausen (2004, 64f) has recently reached the similar conclusion. He has accepted AKL as a
reliable source and yields 1725 or 1715 BC for Šamši-Adad’s death. According to the same author, the end
of Babylon would be 1544 or 1534 BC, independently on the Venus observations.
 29

Kings of Assyria Kings of Babylon

 30

Aššur-dān 1178-1133 Meli-šihu 1181-1167

Ninurta-apil-ekur 1191/81-1179 Adad-šuma-us.ur 1211-1182

Enlil-kudurriī-us.ur 1196/86-1192/82

Aššur-uballit I 1363/53-1328/18 Kurigalzu II 1327-1303

Burna-Buriaš II 1354-1328

Aššur-bēl-nišēšu 1417/07-1409/1399 Kadašman-Enlil I 1369-1355

Aššur-nādin-ahhē -1431/21 Kara-indaš ca. 1420

Aššur-rabî I 1450- Agum III ca.1450

Ulam-Buriaš (Sealand) ca.1470-1450

Puzur-Aššur III 1499-1476 Kaštiliašu III ca.1480

Burna-Buriaš I ca.1500

Aššur-dugul 1684-1679 (Šiptaulzi) ca.1510

Puzur-Sin 1687-1685 Agum II Kakrime ca.1523?

Asinum (=Remu… ?) -1688 (Hurduzum)

 31

Samsuditana 1577-1547
Mut-aškur, Remu 1716-
Ammis.aduqa 1598-1578
Išme-Dagan I 1727-1717

Šamši-Adad I 1760-1728

Old Assyrian
Hammurabi 1744-1702
Ur III (2066-1959)
Old Babylonian
Ibbi-Sîn 1982-1959

Šulgi 2048-2001

Guti 2147-2056?

Table 7
* For all reigns of Assyrian kings is given an alternative of 10 years according to the reign of Ninurta-apil-
ekur (1181-1179 BC). I give the minimum years regarding the synchronism of Enlil-kudurrī-usur (1186-
1182 BC) and Ninurta-apil-ekur (1181-1179 BC) with Adad-šuma-us.ur (1211-1182 BC) Babylonian king
(Brinkman 1972, 272, table 1). One year is not comfortable for the synchronism. Therefore, we could lower
the reigns of the Babylonian rulers for a year or so. The first year of Meli-šihu is dated between ca. 1183
and 1178 BC according to the tablet from Emar (Boese 1982). Aššur-nādin-ahhē with Aššur-rabî I were
assigned 30 years instead of 40 as in Brinkman (1977, 344). Names are taken from Grayson (1980-83).
Years of the reigns of Babylonian kings follow our chronology until the reign of Kadašman-Enlil I. After
that, we take those from Gasche et. al.
 32

BIBLIOGRAPHY

ALBRIGHT, W.F., 1965: «Further Light on the History of Middle-Bronze Byblos», BASOR 179,
38-43.

ASTOUR, M., 1986: «The Name of the Ninth Kassite Ruler», JAOS 106, 327-331.

BIETAK, M., KOPETZKY, K., 2000: «Quantitative Seriation Model Study for the Assessment of
Stratigraphic Shred Collections (Relative Chronometry II) » in BIETAK, M. (ed.), The
Synchronization of Civilization in the Eastern Mediterranean in the Second Millennium
B.C. Proceedings of an International Symposium at Schoss Haindorf , 15 th-17th of
November 1996 and at the Austrian Academy 11th-12th of May 1998, Wien, 22-26.

BIROT, M., 1985: «Les choniques ‘assyriennes’ de Mari» in Durand, J.-M., Margueron, J.-Cl.
(Eds.), A propos d’un cinquantenaire: Mari, Bilan et Perspective (= MARI 4), 219-242.

BOESE, J., 1982: «Burnaburiaš, Melišipak und die mittelbabylonische Chronologie», Ugarit
Forschungen 14, 15-26.

BRINKMAN, J.A., 1972: «Chronologies in Old World Archaeology. Archaeological Seminar at

Columbia University 1970-1971, AJA 76, 271-281.

BRINKMAN, J.A., 1973: «The Nassouhi and Assyrian Kinglist Tradition», Or 42, 306-319.

BRINKMAN, J.A., 1977: «Mesopotamian Chronology of the Historical Period» Appendix in

OPPENHEIM, A.L. (rev. ed. E. Reiner), Ancient Mesopotamia. Portrait of a Dead
Civilization, Chicago, 335-348.

BROWN, D., 2000: «The Cuneiform Conception of Celestial Space and Time», Cambridge
Archaeological Journal 10, 103-122.

BRYCE, T., 1999: The Kingdom of the Hittites, Oxford.

CAH = Edwards, I.E.S., Gadd, C. J., Hammond, N.G.L., 1971: Cambridge Ancient History3 I/2.
Early History of the Middle East, Cambridge.

CHARPIN, D., DURAND, J.-M., 1985: «La prise du puvoir par Zimri-Lim», MARI 4, Paris, 293-
342.

DOSSIN, G. 1951: «Lettre du divin Asqudum au roi Zimrilim au sujet d'une éclipse de lune» in
Compte-rendu de la seconde rencontre assyriologique internationale par le Group
Françoise Thureau- Dangin, Paris, 46-48.

DURAND, J.-M., GUICHARD, M., 1997: «Les rituals de Mari (textes n° 2 à n°5)», Florilegium
Marianum 3 (=Mémoires de NABU 4), Paris, 19-78.
 33

EDZARD, D.O., 2004: Review of Veenhof 2003, ZA 94, 304-306.

FREU, J., 1994: «Histoire d’un people et d’un empire» in MASSON, E. (ed.), Les Hitittes:
Civilisation indo-européenne la fleur de roche, Dijon, 26-39.

GADD, C.J., 1971: «Babylonia c. 2120-1800 BC», in CAH I/2, 595-638.

GASCHE, H., 2003: «La fin de la première dynastie de Babylone: une chute difficile», Akkadica
124, 205-220.

Gasche et. al. 1998 = GASCHE, H., ARMSTRONG, J.A., COLE, S.W., GURZADYAN, V.G.,
1998: Dating the Fall of Babylon. A Reappraisal of Second-Millennium Chronology (=
MHEM 4), Ghent, Chicago.

GATES, M.-H., 1981: Alalakh Levels VI and V: A Chronological Reassessment (= SMS 4/2),
Malibu.

GATES, M.-H., 2000: «Kinet Höyük (Hatay, Turkey) and MB Levantine Chronology, Just in
Time. Proceedings of the International Colloquium on Ancient Near Eastern Chronology
(2nd Millennium BC). Ghent 7-9 July 2000 (= Akkadica 119-120), Bruxelles, 77-101.

GLASSNER, J.-J., 1993: Croniques Mésopotamiennes, Paris.

GRAYSON, A.K., 1975: «Assyrian and Babylonian Chronicles» (= TCS 5), Locust Valley, NY,
Glükstadt.

GRAYSON, A.K., 1980-83: «Königslisten und Chroniken. B. Akkadisch», RlA VI, 86-135.

GURZADYAN, V.G., 2000a: «On the Astronomical Records and Babylonian Chronology», Just
in Time. Proceedings of the International Colloquium on Ancient Near Eastern Chronology
(2nd Millennium BC). Ghent 7-9 July 2000 (= Akkadica 119-120), Bruxelles, 177-186.

GURZADYAN, V.G., 2000b: «Astronomy and the Fall of Babylon», Sky & Telescope 100, 40-45.

GURZADYAN, V.G., 2003: «The Venus Tablet and Refraction», Akkadica 124, 13-17.

HALLO, W.W., 1966: «The Coronation of Ur-Nammu», JCS 20, 133-141.

HUBER, P.J., 1977: «Early Cuneiform Evidence for the Existence of the Planet Venus» in
GOLDSMITH, D. (ed.), Scientists Confront Velikovsky, Ithaca, New York, 117-144.

HUBER, P.J., 2000: «Astronomy and Ancient Chronology», Just in Time. Proceedings of the
International Colloquium on Ancient Near Eastern Chronology (2 nd Millennium BC). Ghent
7-9 July 2000 (= Akkadica 119-120), Bruxelles, 159-176.
 34

Huber et. al. 1982 = HUBER, P.J., SACHS, A., STOL, M., WHITING, R.M., LEICHTY, E.,
WALKER, C.B.F., VAN DRIEL, G., 1982: Astronomical Dating of Babylon I and Ur III
(=Monographic Journals of the Near East, Occ. Papers 1/4), Malibu.

HUNGER, H., 2000: «Uses of Enuma Anu Enlil for Chronology», Just in Time. Proceedings of
the International Colloquium on Ancient Near Eastern Chronology (2 nd Millennium BC).
Ghent 7-9 July 2000 (= Akkadica 119-120), Bruxelles, 155-158.

HUNGER, H., PINGREE, D., 1999: Astral Sciences in Mesopotamia, Brill, Leiden, Boston,
Köln.

KEENAN, D.J., 2005: «Anatolian Tree-rings Studies are Untrustworthy» (web site:
http://www.informath.org/ATSU04a.pdf).

KITCHEN, K.A., 2002: «Ancient Egyptian Chronology for Aegeanists», Mediterranean

Archaeology and Archaeometry 2(2), 5-12.

KLEIN, J., 1981: Three Shulgi Hymns: Sumerian Royal Hymns Glorifying King Shulgi of Ur,
Ramat Gan.

KOCH, J., 1998: «Neues von den UR III-Mondeklipsen», NABU 1998/4, 126-129.

Kuniholm et. al. 1996 = KUNIHOLM, P.I., KROMER, B., MANNING, S.W., NEWTON, M.,
LATINI, C.E., BRUCE, M.J., 1996: «Anatolian Tree Rings and the Absolute Chronology of
the Eastern Mediterranean 2220-718 BC», Nature 381, 780-783.

LEICK, G., 2002: Mesopotamia: The Invention of the City, London, New York.

Manning et. al 2001 = MANNING, S.W, KROMER, B., KUNIHOLM, P.I., NEWTON, M.W.,
2001: «Anatolian Tree Rings and a New Chronology for the East Mediterranean Bronze-
Iron Ages», Science 294, 2532-2535.

NEWTON, M.W., KUNIHOLM, P.I., 2004: «A Dendrochronological Framework for the Assyrian
Colony Period in Asia Minor», Türkiye Bilimler Akademisi Arkeoloji Dergisi 7, 165-176.

READE, J.E., 2001: «Assyrian King-Lists, the Royal Tombs of Ur, and Indus Origins», JNES 60,
1-29.

ROCHBERG-HALTON, F.R., 1988: «Aspects of Babylonian Celestial Divination: The Lunar

Eclipse Tablets of Enūma Anu Enlil», AfO Beiheft 22, Horn.

ROCHBERG-HALTON, F.R. 2004: The Heavenly Writing: Divination, Horoscopes and

Astronomy in Mesopotamian Culture, Cambridge.

ROWTON, M.B., 1970: «Chronology. Ancient Western Asia», in CAH I/1, 193-238.

RYHOLT, K.S.B., 1990: «A Reconsideration of some Royal Nomen of the Thirteenth Dynasty»,
GM 119, 101-113.
 35

RYHOLT, K.S.B., 1997: The Political Situation in Egypt During the second Intermediate Period
c.1800-1550 B. C. (= Carsten Niebuhr Institute Publications 20), Copenhagen.

SASSMANNSAHUSEN, L., 2004: «Babylonian Chronology of the 2nd half of the 2nd Millennium
B. C.”, », in HUNGER, H., PRUZSINSZKY, R. (eds.), Mesopotamian Dark Age Revisited,
Proceedings of an International Conference of SCIEM 2000 (Vienna 8th-9th November
2002), Wien, 61-70.

SIGRIST, M., DAMEROW, P., 2001: «Mesopotamian Year Names» (web site:
http://www.cdli.ucla.edu/dl/yearnames/HTML/T6K1.htm).

SOLLBERGER, E., 1976: «Ibbī-Suen, RlA V, 1-8.

STEELE, J.M., STEPHENSON, F.R., 1997: «Lunar Eclipse Times Predicted by the
Babylonians», Journal for the History of Astronomy, 28 119-139.

STEPHENSON, F.R., 1997: Historical Eclipses and Earth’s Rotation, Cambridge.

STEPHENSON, F.R., MORISSON, L.V., 1984: «Long-term Changes in the Rotation of Earth:
700 B.C. to A.D. 1980», Philosophical Transactions Royal Society London A 313, 47-70.

STEPHENSON, F.R., MORISSON, L.V., 2000: «Historical Eclipses and the Earth’s Rotation»,
Science Progress 83, 55-76.

van de MIEROOP, M., 2003: History of the Ancient Near East: Ca. 3000-323 BC, Oxford.

van SOLDT, W.H., 2000: «Syrian Chronology in the Old and Early Middle Babylonian Periods»,
Just in Time. Proceedings of the International Colloquium on Ancient Near Eastern
Chronology (2nd Millennium BC). Ghent 7-9 July 2000 (= Akkadica 119-120), Bruxelles,
103-116.

VEENHOF, K.R., 2000: «Old Assyrian Chronology», Just in Time. Proceedings of the
International Colloquium on Ancient Near Eastern Chronology (2 nd Millennium BC). Ghent
7-9 July 2000 (= Akkadica 119-120), Bruxelles, 137-150.

VEENHOF, K.R., 2003: The Old Assyrian List of Year Eponyms from Karum Kaniš and its
Chronological Implications. (=Türk Tarih Kurumu, Ser. 6, No. 64), Ankara.

von BECKERATH, J., 1997: Chronologie des pharaonischen Ägypten (=Münchener

Ägyptologische Studien 46), Mainz.

WARBURTON, D.A., 2000: «Synchronizing the Chronology of Bronze Age Western Asia with
Egypt», Just in Time. Proceedings of the International Colloquium on Ancient Near
Eastern Chronology (2nd Millennium BC). Ghent 7-9 July 2000 (= Akkadica 119-120),
Bruxelles, 33-76.