Anciet Egyptian SCience Vol2

Memoirs
aflM
ANrlcan Philosophial1 Sodtty
Held 01 Philadelphia
for Promoling UsifuJ Knowledge
Volume 214
Copyrighl c 1995 by !he American Philosophical Society for its

Memoirs series, Volwne 214
""
ConI' WUltntioo: An'anj:emenI of the northern comtellatiOOl
on !he aauooomical ceiling of corridor 8 on the tomb of
Ramcsscs VI. Prom: NCUIebauer and Parter, Etyptkm
Astronomical TtxU, Vol. 3, Fig. 30.
Library of Congress CaIaloging in Publication Dati
Clagett, Marshall
Ancien! Egyptian Science, Volume 2
includes bibliography and iodellOes
iIIuslraled
ISBN 0-87169-214-7
I. Egypt 2. Science, anciem, hiliOf)' of

3. calenders. ancient 4. clocks, ancient
ISSN: 0065-9738 .........

PREFACE
Preface
This volume is the second of the three volumes on
Ancient Egyptian Science which I hope to complete. I
have not included everything which I projected for
Volume Two in the Preface to Volume One. The
chapter and documents regarding mathematics are
missing. It would have greatly increased the length of
the volume to have included them here and accordingly
I decided to shift them 10 Volume Three. That shih
maKes no difference in the progression of subjects
originally planned for the whole worlc. for I believe that
the lack of theoretical discussions of mathematics by
the ancient Egyptians in their rudimentary science made
it imperative first to outline the principal uses of
mathematics by the dwellers on the Nile before
discussing its structure and content. One possible
benefit of including mathematics in Volume Three is
that it can be more closely related to my discussion of
Egyptian techniques of representing nature and within
that topic to appraisals of the ancient Egyptian lad: of
a direct angular or arcal measure to quantify stellar
displacements and thus produce more accurate celestial
diagrams. of the absence of an effective method of
geometric projection. and finally of their lack of any
extended use of perspective (which. however. surely
did not hamper their considerable artistic skilD.
The organization of the current volume is
self-evident. Again I have given a lengthy introductory
chapter which attempts to synthesize the three main
subjects included in the volume: calendars, clocks. and
astronomical monuments. It summarizes the principal
-vi-
ANCIENT EGYPTIAN SCIENCE
conclusions which we can draw from the eighteen

documents and the Postscript that constitute the hulk
of the volume. The order of those documents follo ....s
that of the three subjects mentioned. There is,
however, no hard and fast isolation of the topics one
from another. Because of this there is much skipping
around from date: to date: in the corpus of documents.
bul within each area of treatment there is fair
chronology evident as befits a historical work covering
three millennia of activity.
In the case of every individual document the effort
is made to supply a meaningful date or dates. It is true
that sometimes the carrier of the document, say a
temple ceiling. has a date of construction or execution
that is orten much latcr than the document itself, as,
for example, is the case of Document 1I1.12 where the
decanaJ transit tables (marked in my document by the
letter ·U") found in the ceiling of the Cenotaph of $eti 1
(ca. 1306-L290 B.CJ date from at least as early as the
reign of $esostris III in the 12th dynasty (ca. 19th
century B.CJ. Similarly, the earliest copies of the
Ramcsside Star Clock (Document 111.14) are found in the
Tomb of Ramcsses VI (ca. 1151-43 B.C. in the 20th
dynasty), but the carC£ully reasoned date implied by the
document itself is some time between about 1500 and
1470 n.c. in the 18th dynasty.
I have given more than 150 pages of illustrations.
For the most part they include the hieroglyphic (or
rarely, hieratic or demotic) texts, some from a single
legible copy, others from an edited text based on
several copies. These illustrations will allow the reader
who controls the Egyptian language to have ready
access to the texts that lie behind my translation. But,
as in the first volume, it is my hope that the
-vi-
PREFACE
translations themselves will give readers without

detailed knowledge of the original language. i.e., most
students of the history of science, a good sense of what
the documents intend. In regard to the illustrations. I
should notc that occasionally 8 magnifying glass may be
needed by the reader studying them. But even in the
cases where considerable text appears on a single
iIIustntion, the reproductions are remarkably clear as a
result of the careful photocopies prepared by my
secretllry. Ann Tobias. who often improved the contrast
and clarity of the originals from which the illustrations
were made.
I have given very full notes 10 illustrate the
historical steps taken by earlier scholars to advance our
knowledge of the subjects trealed in this and the
succeeding volume or my work, This was done not
only to give the reader a good sense or the
development or scholarship over the last two centuries,
but also to give honor and credit where they are due.
Since the appearance or Volume One, two towering
Hgures in the study or Ancient Egyptian Astronomy
have died: Otto Neugebauer, whose help and rriendship I
have acknowledged in the PreFace to Volume One, and
Richard Parker, a premier student or the Egyptian
calendars and Neugebauer's coauthor or the penetrating
and inrormative Egyptian Astronomical Texts in three
volumes. Their respective talents complemented each
other exceedingly well: Neugebauer's superb analytical
powers and Parker's philological skill and extensive
knowledge or the texts. The reader will be well aware
or my debt to them. Among earlier authors, Renour.
Lepsius, Brugsch, Meyer, Sethe, and Borchardt stand
out, but the reader will also Hnd mentioned the works
or many other later scholars (e.g.. Hornung and Barta)
and younger ones (e.g.• Krauss) who have darined and

solved many of the puzzling problems concerning the
topks of this volume. Unfortunately. it was only after J
completed this volume that I obtained a copy of
Christian Leitz's Studien zur Igyptischen Astronomie
(Wiesbaden. 1989), and so I was unable in this volume
to give it the careful study which it deserves (but see
Chapler Three. notc 49), I must also notc with
gratitude that James nay·) O. Mills of the Nekhen
Excavations team has allowed me to publish as II
Postscript most of his unpublished paper on a
petroglyph with possible astronomical significance. This
petroglyph was discovered by Mr. Mills and Ahmed
Irawy Radwan during II survey operation in 1986. As in
the case of Volume One. special thanks are tendered to
Dr. Robert Bianchi for his helpful reading of this
volume in its first version.
Closer to home. I must again thank my wife. Sue. to
whom the whole work is dedicated. for her expert
editorial help and. above all. for her constant
encouragement. I have already thanked my secretary.
Ann Tobias. for the magic she has worked with the
illustrations. But she also undertook the formidable
task of reading and rereading the manuscript numerous
times. much to its improvement. I also repeat here my
thanks to my home institution. The Institute for
Advanced Study. for all of its intellectual and material
assistance: and especially am I in debt to the library
starr. who uncovered copies of even the rarest articles
and books. And finally I want to express my
appreciation to the American Philosophical Society for
continuing to publish this large work. In that
organization thanks must especially go to Herman
Goldstine. its Executive Director. to Carole Le Faivre.
-x-
PREFACE
its Associate Editor. and to Susan Babbitt. who
skillfully copy-edited this volume. All have contributed
to its appearance. its accuracy. and its publication.
Once: again I have provided the Society with
camera-ready copy. using Printrix to print the copy and
Fontrix to create special fonts (see the Preface to
Volume One). But since the publication of the Cirst
volume. I have prepared addition!!l. smaller fonts to
represent the consonantal. phonetic transcription of
hieroglyphs and to indicate 8 large collection of
accented letters and letters with various diacritical
marks in order to print accurately the nolcs and
bibliography. My hieroglyphic fonts are those of the
first volume designed by me and greatly improved and
extended by Ann Tobias. though many more glyphs
have been rashioned ror this volume.
Marshall Clagcu
Proressor Emeritus
The Institute ror Advanced Study
Princeton. New Jersey
-xi-
CONTENTS
Table of Contents
Preface yll
Pari I, Chaptu Thr«' Cattod.rs, Clocks, and Astronomy
Ell.ypllan Calendars, Introductlon to EgyptIan C",ltndars 1
Parker', Accollnt or the Old lunar Calendar 7
Th,. I al"r I liner Calendar 23
The Orilio of the CIvil Calendar 28
Solblc Dales and the Eberl Calendar 37
The Nlr.kl HouTi 48
Deanal Clock. 53
Transll Dcnoal aocks 56
The Rlme-ulde SII[ Clock 59
Qul[]a,", Waler Clocks 6S
Inflow Water Clocks 77
Shadow Clock. 83
ElYpllan Sundial. 95
Tracts of .. 24-hour Day .... lIb Equal HOUri 98
Astronomical erllla&1 and Other Monument. 106
The Ceiling of the Secret Tomb of Seam"l III
The Va.,lttd Cellini of Han K In Sell 1'. Tomb 121
Egyptian Zodiac. 126
ConciudoD 127
Notto to Chapler Three 131
part II, Document.

Document IIU. FuS! Lilli In the Old and Middle Klnadoms
Document III 2 The Eberl Calendar 193

Document 111.3. The Astronomical Celllng of the Secret Tomb
of Senmut 217
Document iliA. The V~IIJhed Celllni of Hall" In the Tomb
of Sell I 235'
Document 111.5. Extracil from the Calendar of Ihe Temple at
MedIn .. "Ibn In Western Thebes 253
Document 111.6. The Names of the 30 [FUlt-Oays! of the
II IInld Monthls! 279
-xi-
Document 111.7. A Table of the LeDl.th. of the Daylll.ht and

of the Nll.hlllme at Monthly IDtervab. I.., .. OD the Flnt of each
Month (from Cairo Muuum Papyrus No. 866371 291
Daeument 111.8. A Table of the Lengths of the Daylight and
of the NllI.hlllme at Semimonthly IDterYab, I.e., on the Flrll and the
Flrtrcnth of each Month 293
Document 111.9. An EUplhlO 2S-Year Lunar Cycle 295
Document 111.10. The Hellacal RI.lna of Sirius and 50thlc
Period. In Ancient Egypt J.07
Document 11111 The Dreanal Clock nn the I td of the Cnffln
of M".het 341
Document 111.12. The Book of Nut that AccomPAnies the
FIlI.ure of Nut In her Arched CelCltlal Posillon, TOBether with
EKlncts from a Commentary on It 357
Document 111.13. The Dramallc Text In the CeDotaph of Sell I
3D
Document III 14 The Ramcsslde Star Clock 405
Document 11115 Amenemhet's Water Clock 457
Document llI.16. The Shado., Clock described In the Cenotaph
of SeU I 463
Document 111.17. The Rectangular Zodiac from the Temple of
Khnum IEsna Al and the Round Zodiac from the Temple of
Hatbor (Dender,. RI 471
Document 111.18. InlCrlptlon. on Ihe Statue of Ihe Astronomer
Uarkhebl 489
Postscript. A Petroglyph DlscoYered at Nekhen with Pos.lble
Astronomical Significance 497
Part 1lI, Blblloaraphy and Indelees

AbbreYlalions Used In Tellt and Blblloa.raphy 507
Bibliography 507
Indell of Earptlan Words and Phraus 521
Indu of Proper Names and Subject. 53S
part IV, lIIu.tratlons

A I lsi Dr lIl11str;tllpos 567
Illustrallpns After 515
-xiv-
Part One
Chapter Three'
Calendars, Clocks, and Astronomy

CALENDARS. CLOCKS. ASTRONOMY
Chapter Three
Calendars. Clocks. and Astronomy

Introduction to Egyptian Calendars
As we have seen in Volume One. Chapter Two, the
Egyptians were wont to call the creator god. whoever
he was. Lord of the Years or Lord of Eternity and
Everlastingness. Though these titles primarily
designated the creator god's creation of time as an
aspect of or limit to world order or maat, they may also
have oblique reference to the supposed role played by
the creator god (or at least by the king with whom he
was identified) in the establishment of a systematic
calendar. However it originated. the civil calendar of
365 days was securely in place by the time of the Old
Kingdom. An investigation of the origins of this
calendar and its relationship to one or more lunar
calendars is fraught with uncertainty and difficulty, but
such an investigation will be most useful for anyone
attempting to understand the steps taken in Pharaonic
times to organize society and express its religious
culture in a satisfying, efficient and productive way.
Hence we must pursue this investigation as far as the
scanty evidence permits.
In the last generation or so, the most inrJuential
work on the subject of Egyptian calendars has been
that of Richard Parker.' Though Parkers views have
wide acceptance, they are anything but certain, as we
shall see presently. Parker believes, as did L Borchardt
+
in II tentative way earlier. that the Egyptians used the

civil calendar of 365 days and two diHerent lunar
calendars, He describes the first lunar calendar and the
civil calendar in II succinct manner:2
Like all ancient peoples, the protodynastic
Egyptians used a lunar calendar, but unlike
their neighbors they began their lunaT month.
not with the first appearance of the new
crescent in the wcst at sunset but rather with
the morning when the old crescent of the
waning moon could no longer be seen just
before sunrise in the east. Their lunar yeaT
divided naturally. following their seasons, into
some four months of inundation. when the
Nile overflowed and covered the valley. some
4 months of planting and growth. and some 4
months of harvest lind low water. At 2- or
3-year intervals. beclluse 12 lunar months are
on the average II days short of the natural
year, a 13th or intercalary month was
introduced so as to keep the seasons in place.
Eventually the heliacal rising of the star
Sirius. its first appearance just before sunrise
in the eastern horizon after a period of
invisibility, was used to regulate the
intercalary month. Sirius. to the Egyptians the
goddess Sopdet or Sothis. rose he1iacally just
at the time when the Nile itself normally
began to rise, and the reappearance of the
goddess heralded the inundation for the
Egyptians. The 12th lunar month. that is the
4th month of the 3rd season. was named
[Wp-rnpt] from the rising of Sothis and a
simple rule was adopted to keep this event
CALENDARS, CLOCKS, ASTRONOMY
within its month. Whenever it fell in the last

II days of its month an intercalary month
[Thothl was added to the year. lest in the
following year Sothis rise out of its month....
This IURi'stellar yeaT was used for
centuries in early Egypt and indeed lasted
until the end of Pagan Egypt as 8 liturgical
year determining seasonal festivals. Early in
the third millennium o.c. however. probably
for administrative and fiscal purposes, a new
calendar year was invented. Either by
averaging a succession of lunar years or by
counting the days from onc heliacal rising of
Sothis to the next, it was determined that the
year should have 365 days. and these were
divided into three seasons of four 30-day
months each. with 5 additional 'days upon the
year' or 'epagomenal days', This secular year
which is conventionlllly termed the 'civil' year
remained in use without alteration to the time
of Augustus when a 6th epagomenal day
every 4 years was introduced. That the
natural year was longer than the civil year by
II quarter or a day was of course known to
the Egyptillns rllirly soon after the civil year
was inaugurated but nothing was ever done
about it, Still it is a great achievement of
theirs to have invented a calendar year
divorced from lunar movement and to have
been the [irst to discover the length of the
natural year, which eventually led to the
Julian and Gregorian calendars....
The first year in Egypt, the lunar one, had
divided the month into 4 'wee\cs' based on
-3-
'first quarter', lull moon', and 'third quarter,

The new [civil] month of 30 days was not
divisible into 4 even parts but conveniently
divided into 3 'weeks' of 10 days each. from
later texts called ·first'. 'middle' and 'Ias(,
Thus in the entire year there were 36 weeks
or decades, plus the 5 days upon the year.
Parker here maintains that the first of the Egyptian
lunar calendars came into existence prior to the
invention of the civil calendar. But he also suggests
that a second lunar calendar developed as the result of
the invention of the civil calendar to keep in step with
it, In this second lunar calendar a thirteenth month was
intercalated when the New Year's Day of that lunar
year would have fallen earlier than the New Year's Day
of the civil year without such intercalation.
Before considering Parker's views on the structure
of the two lunar calendars, let us briefly review the
Egyptian civil calendar. The details of this calendar are
firmly known and not in dispute, except for the time
and manner of its invention, which will be discussed
later in the chapter,
As noted in the quotation from Parker above, the
civil year consisted of twelve months of thirty days
each plus 5 epagomenal days, totaling 365 days. The
months were grouped in three seasons of four months
each, perhaps taken over from the seasons of an earlier
lunar year as Parker suggests.3 but more likely drawn
from an early 365~day calendar that was seasonally
oriented, such as one based on the rising of the Nile.
The first season was named Akhet <U11 !O,
transliterated as 'bt. and usually translated "inundation"
since originally, when the year coincided with the
seasons. this season and indeed the year began with the
-4-
sudden rising of the river and for much of that season

the land lay under water), the second was Peret (Si>JQ.
prl, meaning 'emergence: no doubt for the emerging
land and plant life), and the third was Shemu (l::JssO,
$mtV, meaning 'Iow water' and the time of "harvest"l
Arter indicating the regnal year (see Vol. I. Document
Ln. dates usually specified the day. month. lind season
in the form of month-number. season. and day-number.
For example. "IV ~mw 20' signified the fourth month of
the season of Shemu, day 20. Note that each month
was divided into three 'weeks" of tcn days each (j.e.,
decades), both the month of 30 days and the 360 days
of the twelve months of the year being neally divisible
by that number 10. ThaI IO-day week dictated the use
of 36 decans (stars or groups of stars) in the star clocks
which we shall discuss later.
Month names instcad of numbers were somctimcs
used in the New Kingdom and later. 4 The late forms of
these names, based on their Greek spelling, arel First
Seasorr. I. Thoth, 2. Phaophi, 3. Athyr, 4. Choiak; Second
SellSOrr. 5. Tybi, 6. Mechir, 7, Phamenoth, 8. Pharmuthi;
Third $ellSOrr. 9. Pachons, 10. Payni. 11. Epiphi, and 12.
Mesore.
Now the heliacal rising of Sirius became, at an
unlcnown time, the event marking the beginning of the
civil year, and because the ycar between successive
risings of Sirius was about 365 1/4 days in Pharaonic
times,S it is evidcnt that thc civil year of 365 days was
about 1/4 day short of the Sothic year each year, so
that the former's New Year's Day was about one day
earlier after 4 years (hence 1460 Sothic years • 1461
civil years). Or, to put it in another way. the stellar
event of the heliacal rising of Sirius occurred in thc
-5-
civil year about one day later aher being on the same
day for four years and thus in the course of about 1461
civil years it fell successively aher every four years on
each day of the civil year until finally once again the
rising of Sirius took place on New Year's day of the
civil year. This long-term procession is called a Sothic
cycle. Now if a sixth epagomenal day were to be added
to the civil year at the end of every fourth year in
order to halt this march almost entirely, the civil year
so modified would then be known as a Sothic fixed
year, and the disparity of Sothic and civil years would
almost disappear (again see footnote 5). Despite the
belief of some earlier Egyptologists that the so-called
Ebers Calendar (Doc. 11I.2) is evidence of such a fixed
Sothic year, there is no sure evidence that the
Egyptians widely or regularly used a separate calendar
of the fixed year to which a sixth epagomenal day was
formally added every fourth year, lit least not before
the time of Augustus. (We may except from this
statement the abortive fixed-year calendar of Ptolemy
III promulgated by means of the Decree of Canopus in
238 B.C.. see Doc. 11I.10,) Still the Egyptians obviously
knew that the New Year's Day of the civil year,
initially set or better reset as I Akhet I by the rising of
Sothis. was, quadrennium by quadrennium, steadily
receding from the day of the rising of Sothis, appearing
almost always one day earlier than that rising after
each quadrennium, for at times they recorded (most
surely as the result of observation during most of the
Pharaonic period but perhaps occllsionally by
calculation by the time of the Ptolemaic period) the
rising of Sothis on different days of the diHerent civil
years, liS we shall see later when discussing Document
11I.10.6 Furthermore, it could well be that, as was
-6-
probably the case of the Ebers calendar, other specific

calendars or tables correlating a specific Solhic year
with some given civil year were used for finding dales
in the civil year of seasonal prescriptions or for the
construction or use of water clocks where hour scales
varied from season to season. These uses we shall
discuss Ialer.
Parker's Account of the Old Lunar Calendar

Now let us examine the evidence thaI Parker
presents for his view of the old lunar calendar.7
Certainly we may agree that prior to the invention of
the civil calendar a lunar month was in use, since the
term for "month" (llxJ) employed in the civil calendar
was not only written with 8 crescent moon
surmounting a star (X-=-) but appeared in civil dates in
-
abbreviated form as a crescent moon ( ) with the
seasonal number or the month stroked below it (see Fig.
111.2. second register rrom the top where both
hieroglyphic rorms or the months or the seasons lire
given. but without the phonetic complement "dO [0:=0] or
the rirst rorml Hence the lunar rorm or the term was
used even though the 30-day month or the civil year
was not tied to the waxing lind waning or the moon, as
were the lunar months or 29 and 30 days.
The reader should be rirst reminded that ParkerS
(and before him Drugsch, Mahler. and Sethe)9 concluded
that an Egyptian lunar month began with the rirst day
or invisibility or the waning crescent, i.e., in modern
terms the day of mean conjunction or the moon lind sun
with the earth. This conclusion was arrived at by an
anlllysis or the names or the days or the lunar month
rirst collected by Drugsch (and I shall comment on this
-1-
analysis in Document 111.6 below) and confirmed. Parker

believed. by his analysis of a late 2S-year lunar cycle
(see Document 111.9). But granting the origin of the
term "month" in a lunar calendar docs not tell us
anything about the nature of the old lunar year, i.e.,
whether it was a lunar year of twelve lunar months
without intercalation (as for example was the case with
the Moslem calendar later) or whether a system of
intercalation was in use and if so whether it was the
system, suggested first by Borchardt and improved on
by Parker, in which intercalation of a thirteenth month
was prompted by the heliacal rising of Sirius in the last
eleven days of the twelfth month. This system would
have produced a normal year of 12 months and an
expanded year of 13 every three, or occasionally two.
years.
Parker's conclusion that the old lunar year had an
intercalary month was drawn, in the first instance, by
analogy from the lunar year implicit in the 25-year
lunar cycle given in Document 111.9, which was written
down in 144 A.D. or later but which Parker calculates
was composed in about the middle of the fourth
century B.c. In that 25-year table there were to be 16
"small years" of 12 months and 9 "great years" of 13
months. We shall talk about the details of that
document later. Suffice to say now, it surely reveals
the existence of an intercalary scheme at this later
period of Egyptian history, but of course tells us
nothing about the old lunar calendar, since the purpose
of this later lunar cycle was to determine the position
of lunar months within the civil year and hence the
lunar calendar implicit in this table postdated the
invention of the civil year.
Parker hoped to strengthen his analogical case for
-e-
the earlier use of intercalation by noting references to

festivals of the "great year" Bnd the 'small year' in the
calendar of Beni Hasan, II document of the Middle
Kingdom (sec Document lII.n that may confirm that at
Ihlll time there: was a lunar year of 12 months (the
small ycar) and as well 8n intercalary one of 13 months
(the great ycar).10 But even this interpretation is not
entirely certain. for the 'small year' referred to may be
the lunar year of 354 days and the 'great ycar' the civil
year of 365 days. Hence there is no completely sure
way to tell whether the lunar calendar of the Middle
Kingdom including these festivals was intercalated in
the same way as the later lunar calendar, or what its
relationship was to the old lunar calendar.
However, even if we assume that the analogical
case for some kind of intercalation in the old lunar
calendar is sufficiently strong, we still need to look at
the evidence for the actual system of intercalation
accepted by Parker, namely one depending on the
heliacal rising of Sirius. The nrst type of evidence
presented by Parker is the oft-expressed relationship
between Sothis and the year or the opening of the year
" , 11
\liP rnpt ("W). In the Pyrltmid Texts (Volume One.
Doc. ILl, Section 965) we read the following statement
addressed to Osiris: "It is Sothis (Sirius), your beloved
daughter, who prepares the yearly sustenance for you
(or, has made your year-offerings or year-renewals) in
Ihis her name of . Yeltr' (rnpO: Then at the other end
of Egyptian history, in the Temple of Hathor at
Oendera, a reference to Isis-Sothis, says that "years are
reckoned from her shining-forth:
Parker, in his CItJendars, follows Sethe and
Borchardt and points to an earlier possible reference 10
Sothis and wp rnpt on a fjrst-dynasty tablet from the
-9-
reign of Horus Djet (see Fig. 11I38),12 He and his

predecessors believed that on the right of the tablet the
recumbent cow represented Sothis and that she has
between her horns two vertical signs. the rirst one
being an old form of the hieroglyph for 'year' (rnpt>
and the other III simple stroke. If this is correct, then
the horns combined with the rnpt glyph suggest the
meaning 'opening or opener of the year: rendering the
combined signs W. Furthermore. if it is assumed Ihllt
the cow represents Sothis, then the whole figure ought
to be translated as ·Sothis. opener of the Year: Below
the cow are other signs that, they thought, ought to be
read as the glyphs for '/It, the first season. i.e..
Inundation. And still below them these early scholars
thought they saw in the fragmentary copy of the tablet
(see Fig. lll.3b) traces of a crescent with at least two
strokes below it. suggesting the meaning 'month 2:
However this is doubtful and no such traces can be
seen on the main tablet itself. To the left of the cow is
first a vertical spear-like bar that may be a separator or
it may be a glyph representing the word 'first" (tpy) or
the number "I." Farther to the left are glyphs that
relate to King Ojer and a town that was earlier
interpreted by Petrie as Oep (~O) but is more likely
to be read as Ap (--"01 see note 14 below), a town that
is not otherwise known. Following this interpretation
of the tablel. the whole expression. in a revised version,
would then mean, if we discount the so-called month
traces at the bottom right of the fragmentary copy of
the tablet which do not show up on the tablell l:The
Year of the] first [celebration of) Sothis [as] Opener of
the Year [in] Inundation [by] the Horus Djer [til] Ap:13
Dut let me hasten to say that this interpretation can
-10-
probably no longer be accepted (even in the revised

version), and indeed the earlier version was later
abandoned by Parker himself (see notc 12 above),
mainly as the result of the reexamination of the purport
of the tablet by Gerard Godron.l 4
Godron's principal conclusions may be summarized:
the inscription is not a dale al all and has nothing to do
with the opening of the ycar: the recumbent cow is
Sekhet-Hor instead of Sothisl the uprights between the
horns are not an earlier form of the 'year" sign f plus a
simple vertical stroke. but nther. as Petrie llnd Griffith
thought. Il single "feather" sign P since the 1'0110 vertical
signs aTC in fact connected at the lopI the month signs
on the lower right in the fragmentary copy of the
tablet cannot be seen on the main tablet and so simply
do not exist on it. Lastly, the sign interpreted on the
one hand as the glyph for the season of Inundation or
on the other for "marshes" could be either, and thus
cannot be surely identiried as the one or the other.
But such rderences connecting Sothis with the
regulation or the beginning of the year, even including
the doubtful reading on the tablet from Djer's reign, do
not necessarily apply 10 the role of Sot his in
determining the intercalation of II thirteenth lunar
month. since the coincidence of the rising of Sirius and
the inundation that signalled the beginning of the civil
year was probably known from lln early date even if
that rising was not responsible for the determination
and invention of the civil year of 365 days. as used to
be thought. Hence all of these passages relating Solhis
and the year could simply constitute recognition of that
coincidence.
One further line of reasoning followed by Parker
-11-
involves the fact that on occasion the twelfth month

(at least some time from the New Kingdom on) had the
name Wp rnpt instead of Mesore, which former name
Parker believed to be the name of that month in the old
lunar calendar.6 He reasoned that it was so named in
the lunar calendar because this was the month in which
the reappearance of Sirius would herald the succeeding
month as the first month of the later lunar year if it
happened prior to the last cleven days of the month. or
if it occurred in those last eleven days it signalled the
intercalation of a thirteenth month in order that during
the next year the rising of Sirius would remain in the
twelfth month rather than shifting to the first month
and thus after the New Year's Day. Hcnce in this
argument wp rnpl simply mcant the actual appearance
of Sirius and was equivalent to the later expression prt
Spdt, "the going forth of Sothis: 16 So. according to
Parker, it was only after the establishment of the civil
calendar (and in fact perhaps not until the Middle
Kingdom) that wp rnpt was used in its literal meaning
of the New Yeal's Day of the civil calendar. and when
this happened both meanings of wp rnpt were in use,
Hc finds support for this double use of wp rnpt by
slightly misinterpreting the opening reference to I
Akhet I in an incomplctc Ramesside calcndar edited by
Dakir. saying that on thai day the sccond fcstival of wp
rnpt (wp rnpt sn-nw) was celcbrated.17 The inference
that Parker draws from this information is that another
<that is, a "first") celebratien of wp rnpt (wp rnpt tpy)
tool: place to celebrate the actual appearance of Sirius
in the twelfth month of thc preceding lunar year. Out
this evidence of a "second' festival of wp rnpt does not
ensure that what Parker calls the first celebration refers
to the use of the appearance of Sirius as an intercalary
-~-
device to keep the lunar calendar in step with the

seasons or that it refers at all to 8 lunar calendar.
Balcir's comment on the passage in question and on
Parker's interpretation of it is of interest,l8
Wpt rnpt plV ph snnVll In the first place,
it seems that either wpt rnpt or wp rnpt is a
possible reading... Jt is clear from the reading
of the whole passage that I differ from
Parker"s point of view....who regards the
addition of sn-nw 10 wp-rnpt as a designation
for £the 'second" IVp rnpt, that is] the "first"
day of the "civil" year. He reads Wp-rnpt
snnw [omitting pw and ph after wp-rnpt] and
[so] translates 'the second wp-rnpt: Wpt
rnpt coincides, in this BOOK {of the Papyrus],
with the "first" day of the year. Or as my
reading rightly claims. there were two
celebrations [on the same dayJ: one for the
'first' [i.e.,J tpyday. and the other for the
'opening' [j.e.•J lVpt of the yeaL It is also my
contention that pw exists here [in the text
and should not be ignored as Parker
apparently doesJ lind is to be regarded as
copula since wpt rnpt is never written with an
additional p and w [before pb whether pb is
its possessor or its determinativel Thus my
interpretation of this pasSllge runs as
suggested: 'the second feast is the opening of
the year: Furthermore, if we go back to the
inscriptions of the O. Kingdom mllStabas, we
find offerings presented on two separate
festivals on the New Year's Day -the wpt
rnpt and the tpy rnpt. To quote H. Winlock...:
"the first of these festivals, in the Xllth Dyn.
ANcrENT EGYPTIAN SCIENCE
calendars, is also the 'coming forth of Sothis:

the second is, in all likelihood, the new year
invented for the calendar when it became
derinitely and obviously separated from
nature:
A quite different explanation of the signi£icance of
Wp rnpt as a name (or the twelfth month has more
evidence supporting it than that presented by Parker.
Parker had viewed the Ebers Calendar (Doc. 111.2) as
being an effort to correlate a schematized old lunar
calendar with the civil calendar in the ninth year of
Amenhotep I's reign. He believed that the £irst entry
which identified wp rnpt with prt SpOt is evidence that
we have a lunar calendar in which the rising of Sothis
is an indicator of the lunar New Year's day to (ollow in
the next month. On the other hand, Kurt Selhe (see
Document 111.2 below) and, following him closely,
Raymond Weill,l9 presented the view that the month
names must be distinguished from the eponymous (east
days from which they took their names, and that the
eponymous feast days were not in their homonymous
months but were in fact the culminations of those
months and hence took place at the beginnings o( the
succeeding months. In this connection Weill (depending
on Gardiner. Meyer [see especially Fig. 1II.6a], and
Sethe> presents a list. developed hom a number of
calendars and calendaric statements dating from a
period extending from the Middle Kingdom through the
Ptolemaic-Roman epoch, which shows that almost
invariably the (east days were celebrated in the months
following those to which the feast days gave their
names (also see Fig. 1II.6b). This would explain why
the Feast of Wp rnpt begins the Ebers Calendar even
though the month of Wp rnpt is later the name of the
twelfth month. Hence the Ebers Calendar loses its

significance as evidence for Parker's reconstruction of
the old lunar calendar since in this year of the Eberl"
Calendar the Feast of Wp rnpt is correctly the first
entry of the year occupying the first column. It is not,
as Parker believed, an event of the twelfth month
placed at the beginning of the calendar to inform us
that the next entry (the Feast of TekhD is the
beginning of a schematized lunar year. I shall elaborate
my view of the Ebers Calendar laler. Now I simply say
that I find Selhe's presentation convincing and hence, in
discussing Document 111.2. I shall accept the view that
the Ebers Calendar docs not present a correlation of the
old lunar calendar (even in its schematized form) witb
tbe civil calendar. but ratber is an effort to correlate.
ror civil year 9 or the reign or Amenbotep I. an ad boc
Hxed Sotbic year with tbe civil calendar. a correlation
needed when seasonal dates have to be converted to
civil dates (such as is the case with seasonal medical
prescriptions and in the construction and proper use or
water clocks, a view I shall elaborate laterl
Be that as il may, Parker turns to the listing or
reast days in Old-Kingdom lombs 10 support his view
Ihat Ihe Hrst use or wp rnpt as a rderence to Ihe
appearance or Sirius in the twelhh month was as a
determiner or intercalation in the old lunar calendar:20
So rar we have dealt in the main with
generalities. We have established the
reasonableness or a primitive lunar calendar
based on the rising or Sothis and also the ract
that the rising was termed wp rnpt. In the
rollowing pages I shall present the evidence
which has led me to the conclusion that the
year which was opened by Sothis' rising
-5-
cannot have been the civil year or the fixed

Sothic year £i,e" one with a sixth epagomenal
day every fourth year] but must have been
the natural lunar year....
Beginning in the 4th dynasty [and
continuing through the 5th dynasty] the
mastabas [j.e.. tombs] of the Old Kingdom
frequently exhibit a JJtp-dl-nsw [or offering]
formula involving Anubis <at times Osiris) and
requesting..:invocation·offerings· on certain
festivals [listed thereinl If a number of these
lists £see Doc. III.!] are examined. it will be
found that the feasts [in 21 lists] tend to
follow a derinite order [6 lists including all of
the feasts in the same order lind the others
containing anywhere from 5 to 10 feasts
preserving this same orderl ...:
Order of reasts
(l) wp rnpt
(2) [}Qwtyr
(3) tpy rnpt
(4) wi'
(5) Vb 5kr
(6) fJb wr
(7) rk(l
(8) prt Mn
(9) (/bc/J (n) 5t!
(10) (tp) 1M
(II) tp smdt
(12) fJb nb r C nb or variant....
The exceptions to be found are cerlainly
insufficient to weaken the overwhelming
evidence of II strict order to the calendar of
-.,-
feasts in which the dead expected to take

part. This order. it is easily demonstrated. can
be nothing other than chronological.
Whatever may be the exact meaning of IVp
rnpt and tpy mpr. they clearly belong at the
head of 8 list. If we now check feasts 4 to 8
against the later temple calendars (Medinet
Habu and the Greco-Roman temples), we rind
the following dates in the civil year on which
they were celebrated,
(4) 'IV~ I 'br 18
(5) bb SKr 1111 fIJI 26
(6) pb lilT II prt 4
(7) db (wr) JI prt 9 (Edfu) III prt I
(Il1ahun)
(8) prt Mn J 5m IV II (Med. Habu)
Nos. 9-11 were monthly feasts. celebrated
at least twelve times 8 year....
There seems full justification for
considering the Old Kingdom list 10 be
arranged chronologic8l1y. Moreover. there are
reasons ..Jor thinlcing that all the feasts, with,
of course, the exception of wp rnpt. were
lunar at that rime. The wW"feast can be the
movable lunar feast for which there is
evidence from the Middle Kingdom...: db as
the name of a lunar month cannot be other
than lunar...; and the lunar character of prt Mn
is brought out in the Medinet Habu calendar
and later....The monthly feasts of Ibd and ~mdt,
and so probably of sId, were lunar.
Furthermore, there is no other phlUsible
explanation for the sequence of wp rnpt.
Dbwtyt, and tpy mpf than the assumprion
-17-
that the latter two also were lunar....

The proposed original lunar calendar fits
the chronological order perfectly, and I know
of no other explllnlltion. Wp rnpC was the
rising of Sothis, the event which opened the
new year but which. in itself, did not form
part of it. Tpy rnpt was the first day of the
new ycac, the first day of the month (by in
which fell the fellst of w(g1 and the remaining
feasts followed in chronologiCliI order. As for
the feast of Ilbwty. between lVp rnpt and tpy
rnpt, this can be nothing other Ihan the fcast
of the intercalary month which would occur
al three-{at times two-)year intervals. As 1I
special month it was fittingly dedicated to
Thalh, the moon-god.
Now what are we to make of Parker's explanation
of how these popular Old Kingdom feast days fit
chronologically into and so help to support the specific
intercalary lunar calendar described by Parker? It is
certainly true that if the first three feasts were
celebrating separate. distinct events on three distinct
days which were items in a chronological order. then
Parker has indeed a good case in view of the fact that
the remaining annual feasts (Nos, 4-8) are quite
probably in chronological order, For if the festival of
the rising of Sothis had to fall before the festival of
Thoth and that of the first of the year. then one could
not simply answer that it was the rising of Sirius that
had strayed from its position at the first of the civil
year that was being celebrated before the first day of
the year because if it were that civil-related rising of
Sirius that was being celebrated even after moving
away from the civil New Year's Day one would expect
-18-
it to follow after the civil New Year's Day, as it inched

through the civil year. But what about the festival of
Thoth that appears regularly between the rising of
Sothis and the New Year's Day? Was not Thoth the
name of the First Month of the Year and should it not
accordingly have followed the festival of the New
Year's Day? No. Parker says, at the time of the old
lunar calendar Thoth was the name of the intercalary
month of the lunar year rather than that of the first
month. Unfortunately Parker has had to assume what
he should have been proving. That is to say, he first
assumes that the first three feasts are chronological and
relate to the old lunar year and then in order to support
this he has to assume that Thoth was the name of the
intercalary month rather than of the first month as it
was later and for which there is a great deal of
evidence. Thus in order to justify the chronological
assumption he has to make a further assumption for
which there is no other hard evidence, namely thai the
festival of Thoth must indeed precede rather than
follow the First Day of the year. The kind of
reasoning that Parker applied to the first three old
feasts may leave the reader uncomfortable. But. in
defense of Parker, it must be said that the evidence
presented by Parker for the rest of the list's being
chronological is substantial (though based on
considering the order of these feasts celebrated in the
civil year). and thus he believes that the first three
feasts must also be chronologically ordered.
Furthermore. since many of the feasts on the list appear
to be lunar in origin, the first three festivals must refer
to feasts in the old lunar calendar. If that is so, then
the position of the festival of Tholh between the
festival of the rising of Sothis and New Year's Day
-19-
necessitates its reinterpretation in a lunar calendar. and

the only sound explanation for the order of the three
feasts in 8 lunar calendar would result if it assumed
that tbe rising of Sirius signalled the coming New
Year's Day at the beginning of the next lunar month
and that the festival of Tholh was a festival of the
intercalary month. This festival of Thoth (for which
there is no other compelling evidence)21 would be 8
quite different fcast from the ordinary feast of Thoth
celebrated in the civil calendar on I Akhet 19 (see
Document 11I.5 below),22 The objection implied by
Gardiner that unlike the other festivals that of Thoth in
the lunar calendar would not be an annual festival,23 is
simply set aside by Parker. Convinced that aU of these
twelve feasts were chronologically present in the old
lunar calendar. he of course rejects Gardiner's quite
different suggestion that the names of festivals (I}-(3)
might reflect "three separate aspects of the beginning of
the (civill year. Wpt rnpt Viewing it as the birthday of
the sun-god Re c (Mesorel. [}pwtlt as under the sway of
Yhoth, the initiator of the year of 360 • 5 days. and
Tpl rnpt as the most appropriate dale for the accession
of the earthly king:24 Even if the aspects suggested
by Gardiner are themselves wrong. the idea that more
than one festival representing the beginning of the year,
each with a different name, was celebrated on the same
day (especially at different places) is certainly
reasonable considering the calendars edited by Bakir and
the evidence assembled by Schott.25
On considering further Parker's treatment of the
first three festivals mentioned in the old mastabas, I
remind the reader of an even more telling refutation of
his explanation of the character and order of these
festivals. Our examination of the Ebers Calendar shows
[below and in Document 111.2] that just the reverse of

Parker's assumption appears to be true. namely, (I) the
Feast of Wp-rnpt was not merely an indication of the
rising of Sirius to hail the next month's beginning of
the year or the necessity of intercalating a thirteenth
month (Parker's view), but rather itself marked the
beginning of the year, and (2) the Feast of Thoth
mentioned as the second festival on the mastabas was
not at all a celebration of an intercalary thirteenth
month dedicated to Thoth which followed the rising of
Sothis (within II days of the end of the twelfth month)
but rather was the culmination of the annual
celebrations of Thoth beginning before the end of the
rirst month and culminating on the first of the second
month of the year. The Feast of Thoth thus did indeed
follow the Feast of the New Year but not as a festival
of intercalation, and certainly not before the beginning
of the year,
Without detailing the other lines of reasoning that
convinced Parker of the correctness of his description
of the old lunar calendar but which do not produce
certain conviction. I believe we can reasonably conclude
that he has given us an account that is only barely
possible and is quite speculative in detail and not
convincing in its over-all argument, His often used
rhetorical expressions like such-and~such ·cannot be
other than: ·can be nothing other than: 'there is no
other plausible explanation: ·What can be more natural
than: "It is a natural assumption: etc, when in most
cases there could be alternative explanations make this
reader uneasy,
In brief, it appears to me that Parker's opinion that
the old lunar calendar was intercalary may be correct
<though not certainly so) but that (I) the use of the
-21-
Sothic heliacal rising as the mechanism of intercalation.

that (2) the intercalary month (if it existed) was named
"Thoth: and that (3) the lunar calendar in scbematized
form is that given in the Ebers Calendar and in the
astronomical ceilings of Senmut's tomb and the
Ramesseum - 8rc all unproved and indeed untenable,
Also. strictly speaking, the conclusion that the lunar
month in the old lunar calendar commences on the first
day of crescent invisibility is not certainly proved.
although that seems to be the case for the later lunar
calendar and appears to be in accord with a religious
statement about the conception and birth of the moon
god Khans (see Document 11I.61 I say "not certainly
proved" because the three main pieces of evidence on
which the conclusion is based. namely the names of the
days of the lunar month. the 2S-year lunar cycle of
Papyrus Carlsberg 9. and a series of double dates (the
lunar month days in the civil calendar) all date from a
period more than a millennium or two later than the
date when the old lunar calendar flourished. Indeed
though some ten of the protective gods of the feasts of
lunar days are found in astronomical ceilings of the 18th
and 19th dynasties and the names of eight of the feasts
of lunar days appear in the calendar of Medina Habu of
Dynasty 20 (and perhaps earlier in the Ramesseum in
Dynasty 19)-see Documents 111.3. iliA and 11I.5. in fact
it is only in the Greco-Roman period that the full list of
30 lunar day-feasts appears, as is also true of the
2S-year lunar cycle and the double lunar-civil dates.
On the other hand, it does seem almost certain that the
Egyptian day began at dawn the hour before sunrise
rather than at sunset during ancient Egypt's recorded
history,26 and this fits beller with a month whose first
day is the first day of invisibility of the waning
-22-
CALENDARS, CLOCKS, ASfRONOMY
crescent before sunrise, We shall return to this

question presently when we discuss the later lunar
calendar,
One final word concerning Parker's reconstruction
of the old lunar calendar. By means of it he has
attempted to substitute a luni-stellar seasonal calendar
for a fixed Sothic seasonal calendar. That is to say,
Parker has replaced with his description of a
schematized 'old lunar calendar the view of early
Egyptologists that in order to keep track of their
seasonal and other feasts the priests in the temples used
some kind of fixed Sothic calendar whose New Year's
Day was the heliacal rising of Sirius and to which a
sixth epagomenal day was added every four years.
Both reconstructions found the ·heliacal rising of Sirius·
as crucial and both employed the evidence of temple
calendars (see Document 11I.5), astronomical ceilings,
and the Ebers Calendar as useful. Needless to say, to
question Parker's construction is not to affirm the
Sothic fixed year, at least not in the complete and
exaggerated form which held that a sixth epagomenal
day was added formally to the year every four years.
Now two further subjects remain to complete the
discussion of ancient Egyptian calendars. the later lunar
calendar and the origin of the civil calendar and its
occasional correlations with the rising of Sothis.
The Later Lunar Calendar

The essential features of the later lunar calendar are
much more firmly established than those of the old,
Key to the understanding of this calendar was the
discovery and editing of Papyrus Carlsberg 9 by
Neugebauer and Volten in 1938 (see Document 11I.9).
As Parker has noted, this is "the only truly
mathematical astronomical Egyptian text yet

published: 27 He continues with the follo....ing
characterization of the document,
This papyrus was written in or after A.D.
144. and it furnishes a simple scheme. based
upon the civil calendar, for determining the
beginning of certain lunar months over 8
25-ycar cycle. Underlying this cycle arc the
[acts that 25 Egyptian civil years have 9.125
days and that 309 lunar months (divided into
16 years of 12 months and 9 years of 13
months) have 9,124.95231 days. The carliest
cycle mentioned in the text began in the sixth
year of Tiberius. A.D. 19. and the latest in
. year 7 of Antoninus. A.D. 144. Actually. the
calendar does not indicate when every month
began but lists only six dates for every year.
those falling in the second and fourth months
of each season. It also indicates the years of
13 months ("great" years) according to the
following scheme, 1st, 3d, 6th. 9th, 12th, 14th.
17th, 20th, and 23d year of each cycle.
Table 3 [equivalent to my Fig. 111.& taken
from Neugebauer and Volten] gives the
complete cycle as stated in the papyrus.
Anyone using this calendar would begin a
lunar month. without regard for actual
observation, on II Ibt I (which for A,D. 19
would be September 18; for A.D. 44, September
II; for A.D. 69, September 5: and so on up to
A.D. 144, August 17), on JIll I!lt 30 (for A.D.
19. December 16, for A.D. 144. November 14),
on II prt 29 (for A.D. 20. February 13: for
A.D. 145, January 12), and so forth. What is
-24-
significant about the cycle for our purpose is

that it gives a continually recurring series of
lunar dates, easily and accurately translatable
into the JulillR calendar for this period.
The first use that Parker malees of this lunar cycle
is to tcst his view (propounded first by Brugsch and
followed by Mahler and Selbe) that the lunar month
began with the nrst inVisibility of the waning crescent
moon in the morning. The original editors of the text
of the cycle. aher analyzing the values for the year 144
A.D.. arrived at the conclusion that for this table the
lunar month began with new crescent visibility in the
evening.28 Though this seems to go against Parker's
view, derived originally from an analysis of the names
of the days of the lunar month, he agrees that the cycle
when applied to the year 144 A.D. does indeed seem to
show that the month is starting with the £irst visibility
of the new crescent.29 But he goes on to observe that
the cycle produces an error of one day in about 500
years. Hence if it had been in use for about 500 years.
the accumulated error would produce values that
coincided with the day of £irst crescent visibility rather
than with £irst invisibility of the waning crescent.
Thus it becomes crucial to kno..... when the cycle was
invented and thereaher continued in use. Since the
document shows no evidence of contemporary
Hellenistic astronomy, it seems a safe bet that its
fundamental structure and calculating techniques arc of
more ancient Egyptian origin. Analysis of a series of
double dates <i.e., lunar dates speci£ied in the civil
calendar) pushes bael.:: the use of the cycle into
Ptolemaic times. 30 Furthermore, if we apply the table
to the lirst two years of the cycle beginning in 856
B.C., we £ind (Parker, ClJlendlJrs, p. 17) that 'the month
.25-
starts in every case exactly one day berate the morning

of invisibility and two or three days before the evening
of crescent visibility: Hence it is clear thai the table
would be of no use that early and thus had not yet
been invented. But if we apply the table to the first
two years of the cycle beginning 500 ycars later in 357
B.C.. (ibid) "ten out of twelve months start on the
morning of invisibility. and only one on the evening of
visibility: If one applies the table to the first two
years of the seventh cycle in the period between 357
B.c. and A.D. 144. i.e., of the cycle beginning in 207
B.c.. <ibid.) 'seven months out of twelve begin with the
morning of invisibility, one begins with crescent
visibility and the other four are in between. It is clear
that, the closer in time the cycle approaches A.D. 144.
the less frequently will the months start with
invisibility and the more frequently with visibility:
Thereupon, considering also his examination of the
double dates already mentioned. which push the use of
the cyclc bad: into Ptolemaic timcs, Parker reasonably
concludes <ibid, p. 23) that the time most appropriate
for thc cyclc's inauguration was around 357 8.C.. 'at
least in the form in which we have it. with invisibility
of the old crescent as its foundation: 31
Satisficd with the yield of Document 111.9 to the
problem of when the lunar month began. Parker makes
other contributions to a description of the lunar
calcndar that underlay this table. In the first place. he
uscs the techniques apparent in the cycle's
determination of the beginnings of the six even lunar
months of each year of thc 25-year cycle and specific
lunar dates to produce values for the odd months, thus
giving us the completed 25-yel'lr cycle (see Fig. 111.9),32
Having already statcd that the document with the
-26-
25-yc:ar cycle tells us which years have twelve months

and which contain thirteen. Parker then deduces with
little trouble the principle of intercalation followed in
the lunar calendar underlying the 2S-year cycle:
'Whenever the first day of lunar month Thoth would
fall before the first day of civil Thoth. the month is
inlc:rcalary:33 A few final observations by Parker may
be quoted,34
The outstanding fact of the cyclic
calendar is that it begins on I '!Jt L We have
come to the decision that in Ihis form it
originated in the fourth century B.C....Despite
the fact that in A.D. 144 the cyclic calendar
was clearly no longer in agreement with lunar
phenomena. it W6S not corrected. This is
certain, since lunar date I falls in A.D. 190 and
still fits exactly into the scheme. Obviously.
then. there was no provision in the calendar
itself which required that periodically it be
adjusted. Furthermore. if the present cycle
were the result of a correction in the fourth
century B.c. of lin IIlready existing cycle,
instituted. let us say 500 years earlier. that
earlier cycle would have had to begin with I
'!!t 2. so that its correction would result in
Jlbt I. This is exceedingly unlikely....
The question may still be legitimately
asked. however. whether any sort of
schematic lunar calendar was in earlier use....35
We have decided that the first schematic
calendar was introduced in the fourth century
B.c. and that prior to that time there is no
evidence that IIny other method than
observation was used to begin the month. We
-27-
have discovered the rule that regulated

intercalation in the 25-year cycle and have
seen that it is one which could easily have
been operative before the cycle was installed.
The essential point to observe is that the
lunar calendar was governed by the civil
calendar, since whenever the first day of lunar
Thoth would fall before the first day of civil
Thoth, the month was intercalary. Since the
civil calendar moved forward through the
natural year. the lunar calendar attached to it
must likewise have moved with it. Moreover.
Ihis IUnltr ca/endllr annat have existed before
the civil clt/endar WIlS introduced.
The Origin of the Civil Calendar

I have alrelldy outlined the main chancteristics of
the remarkable civil calendllr of 365 dllyS (12
schematized months of 30 days each, llrranged
according to four months in each of three sellsons. plus
5 epagomenlll days) invented by the ancient Egyptillns,
The most difficult problem regarding this calendllr is
the time and circumstance of its inauguration, We
should first examine the ellrliest historical evidence of
its existence. Setting aside the Djer tablet of the first
dynasty, which we earlier characterized as probably not
containing a reference to Sothis and the opening of the
year, we can point to two indications in the early
Annills that seem to be plausible evidence that the civil
calendar of 365 days was in use at least by the time of
the beginning of Shepsekaf's reign in the fourth dynasty
and irrefutable evidence for that of Neferirtare's in the
fifth (see Volume One. Document 1.1. note 13), since the
year shared between each of these pharaohs and his
-28-
predecessor (probably in the first case and certainly in

the second) adds up to twelve months and five days.
Another probable reference to the civil calendar occurs
in the Ann/lIs for the reign of Neferirkare where it is
noted that in the first year of his reign ceremonies took
place for the "Birth of the Gods: which in all likelihood
were ceremonies devoted to the gods of the epagomenal
days (ibid, n. 112). The full civil year, perhaps headed
by the epagomenal days (6 ~f 0) and including the
three seasons with their months, may be found in an
annual schedule of priests performing services for
Hathor in the fifth-dynasty tomb of Nekankh at Tehne
(see Document 111.1. line 11 of the extract from that
tomb; and see the contrary opinion of E. Winter in note
2 to the Introduction to Document 11I.1). There is as
well a reference to the epagomenal days in the Pyalmid
Texts (Sect. 1961b-c, Pepi Irs pyramid at the end of the
sixth dynasty), "He sees preparation for the festival of
the burning Wt., making) of the braziers, the birth of
the gods before you in the five epagomenal days
<f6}O:J:36
Accepting these references as firm evidence that
the civil calendar of 365 days was in use in the Old
Kingdom, we are brought to the question of when it
was first used. The exposition of the details of the
calendar by Eduard Meyer in 1904 37 had wide
acceptance for a generation or longer. His fundamental
conclusion was that the civil year with its structure of
twelve months of 30 days each, ordered into three
seasons, plus the five epagomenal days remained
essentially unchanged from the time that it was adopted
and recognized to have a New Year's Day at the
hcliacal rising of Sirius throughout the whole Pharaonic
-29-
ANCiENT EGYPTIAN SCIENCE
period and indeed until 139 A.D. when, we arc: told by

Censorious 100 years latcr (sec Document 111.10, Section
6. caps. 18 and 2D. the rising took place on the first day
of the civil year, Le., an Apoht8St8Sis ("re-establishment
or return [of SiriusT> took place on the first day of the
month Thoth. and hence this was the first ycar of the
Great Yellr (·period) called (cap. 2)) "Solar Year, Year
of the Dog-slar [i.e., Canicular or Sothic Yearl or Year
of gOO: 38 To this was added the assumption that the
rising of Sirius was determined at one place (most likely
Memphis or Heliopolis) and then promulgated
throughout Egypt.
A second assumption was that the Egyptians
concerned with the calendar recognized after some time
that the rising of Sirius was delayed one day after
every four years in the civil ca\endllr (which is very
close to the actual delay during the Pharaonic period)
and consequently thcy lcarned how to calculate simply
the day whcn the heliacal rising of Sirius would occur
in any given civil year, and hence observation of the
rising was no longer the basis for determining and
disseminating the day of rising in the civil calendar,
thllt determination being calculated on the basis of an
assumed delay of one day in the rising after every four
years.
If these assumptions were correct, and given the
fact that a Sothic year calculated on the basis of a
one-day delay after every four years of successive
risings of Sirius is nearly identical in length with the
Julian year of 365 1/4 days. we could detcrmine. from
the Egyptian historical records of the risings of Sirius in
the civil year. the Julian dates of these recordings so
long as we knew which Sothic period the record fell in.
If we work backward from 139 A.D. we would find the
-30-
possible quadrennial dates of the beginnings of the

three Sothic periods pr~cedjng the one beginning in that
'year: B.C. 1321-1318. 2781-2778. and 4241-4238.39
Meyer believed that the period beginning rrom
2781-2778 8.C. was too recent for the establishment of
the calendar and so he accepted the earlier period. He
thought that he had confirmed this when he deduced
from rather inconclusive considerations of the Gregorian
dates for recent risings of the Nile that it was in the
Sothic period beginning from 4241-4238 that the date of
Sirius' appearance best conformed with the date of the
Nile's rising. 40 However. with the development and
acceptance of the 'short chronology' of Egyptian
history which can be coordinated with the short
chronology elsewhere in the Near East. the general
opinion has switched to the third millennium for the
establishment of the civil calendar since, at the
beginning of the earlier period. Egyptian society was at
an undeveloped level of sophistication and was not a
unified state. This was given a significant boost by
two influential articles of Otto Neugebauer. 41 In the
second of these articles he summarized the main
conclusion of the first,42
...the simplicity of the Egyptian calendar is a
sign of its primitivity; it is the remainder of
prehistoric crudeness. preserved without
change by the Egyptians. who are considered
to be the most conservative race known in
human history.
....there is no astronomical phenomenon
which possibly could impress on the mind of a
primitive observer that a lunar month lasts 30
days and a solar year contains 365 days.
Observation during one year is sufficient to
-31-
convince anybody that in about six cases out

of t .....elve the moon repeats all of its phases in
only 29 days and never in more than 30; Ind
rorty years' observation of the sun (e.g.. of
the dates of the equinoxes> must make it
obvious that the years fell short by 10 days!
The inevitable consequence of these facts is, it
seems to me. that every theory of the origin
of the Egyptillfl a/cndlJT which IIssumes lin
astronomic,,1 foundlltion is doomed to (lli/ure.
four years ago I tried to develop the
consequences of this conviction as far as the
Egyptian years arc concerned. I showed that
a simple recording of the extremely variable
dates of the inundations leads necessarily to
an average interval of 365 days. Only artcr
two or three centuries could this 'Nile
calendar' no longer be considered as correct,
and consequently one was rorced to adopt a
new criterion ror the £load, which happened to
be the reappearance or the star Sothis.
He went on to point out that the seasons or the
Egyptian year are agricultural and not astronomical.
The adoption or a unirorm month or 30 days is not to
be laid to astronomical observation but to a deliberate
schematization or the real lunar months ror
administrative purposes in a highly centralized society.
He rurther noted that a schematic calendar or twelve
months or 30 days is also round in Babylon, though the
practical needs in Babylon, organized in city-states, did
not lead to a single centralized civil calendar.
Following Neugebauer's rejection of the
astronomical origin of the 365-day year in favor of its
origin in the seasonal rising of the Nile and adopting the
-32-
increasingly short chronology for the establishment of

the unified state, we could arrive at the condusion that
sometime during the first dynasty a 365-day calendar
was adopted. say about 3000 B.C. (sec the chronology
in the Appendix to Volume One), and that it was
maintained for about 220 years until it was decided to
use the heliacal rising of Sothis al the beginning of
Inundation as the New Year's Day. A strong point in
support of giving a role to the Nile's rising is the fact
that from at least the reign of Djcr in the first dynasty.
a yearly record of the Nile height was kept (and latcr
recorded in the Annals which we have given as
Document I.I in Volume One; see that document, notes
13 and 14>' The first of the year and the rising of the
Nile are identified in a later rderence (see note 25,
quotation from a Cairo Calendar of Lucky and Unlucky
Days). To be certain about these conclusions would be
foolhardy, since we have piled conjecture upon
conjecture, starting with Meyer's assumptions about the
Sothic cycles. the initial observations of the heliacal
rising of Sirius at a given. unchanged location, and the
adoption by Egyptians of the scheme of calculating the
risings of Sirius upon the rule that it was delayed one
day after every four ycars. 43
Before leaving the problem of the origin of the civil
year, I must recount in more detail the views of Parker
on its origin. 44 views which were alluded to in our
long quotation from Parker at the beginning of the
chapter. He remarks first that prior to Neugebauer's
articles. the students of the Egyptian civil calendar had
helieved it to have been astronomically determined in
that it was inaugurated with the heliacal rising of Sirius
as its New Years Day, and that despite the fact that
New Years Day fell progressively in advance of Sirius'
-33-
rising, no attempt was made to adjust or correct the

calendar. He also noted that Sethe believed that the
lunar year and the civil year were parallel in their
development. As for Neugebauer's views. Parker rinds
them unsatisfactory. primarily because Neugebauer had
not known the structure of the old lunar calendar based
on the rising of Sirius, that is, the old lunar calendar
discerned by Parker (and which we have characterized
above as wanting in surety). Here then is the
explanation of the origin of the civil calendar in
Parker's own wordsl 45
During the protodynastic period the only
calendar in use was the lunar calendar already
described. By this time Egypt had become a
well organized kingdom, and the economic
disadvantages of a lunar year of now twelve
months, now thirteen, all of which began by
observation. must have pressed themselves
upon the government, In an effort to alleviate
the situation and to proVide a simple and
easily workable instrument for the
measurement of time. they hit upon the idea
of a schematic lunar yeM, or as it might be
termed an averaged lunar year, There are two
ways by which the length of this schematic
lunar year might have been determined as 365
days. On the one hand, since the current
calendar was based on Sothis, one might
simply have counted, for one or two years,
the number of days between successive
heliacal risings of the star. On the other hand,
the lunar year might have been averaged. It
would have been little trouble to refer to the
various records of one kind or another from
-34-
which data on the number of days in recent

calendar years might be derived.
To show how this might have happened. Parker
gives II hypothetical table of the lengths of 2S
successive lunllr years by which he suggests that by
accumulative lIveraging from ycar to ycar it will soon
be clear that 365 days is the proper length or that by
averaging after each year in which the intercalary
month was added the SlIme rigure will be evident even
aher only 11 years. Parker continues,
Thus easily, by either method. could the
Egyptians of the protodynllstic period have
arrived at 365 days as the proper length for II
schematic year which could be adopted for
administrative lind economic purposes. After
the analogy of the ordinary lunar year the
schematic year would be divided into twelve
months and three seasons. each month having
30 days for simplicity and regularity. The
extra five days of the schematic year were
regarded as an abbreviated intercalary month
and were placed before the year. just as the
intercalary month headed the lunar year
whenever it occurred.
Now as the new year was after all only a
schematic lunar year. it must have been
planned to have it run concurrently with and
as far as possible in concord with the old
lunar calendar. which was by no means to be
abandoned. Just what were the actual
circumstances of its introduction we shall
probably never know. It is possible. at least.
that it was introduced in a year with an
intercalary month and that the five
-35-
epagomenal days were concurrent with the

last five days of that month. so that the first
day of both the old lunar year and the new
schematic year fell on tpy rnpt, the first of
(by; or it may have been handled without
regard for the epagomenal days, with the ne....
year. 8t its installation. simply having its first
day coincide with tpy rnpt.
If we cannot be certain of the
circumst8nces of the introduction. we can al
least set a range in time within which it took
place. According to the rule for intercalation
in the lunaT calendar, when the first month
after wp rnpt began .....ithin eleven days of
that event it was intercalary. Tpy rnpt then
might be 8S close to wp rnpt as the twelfth
day following or. with 8 full intercalary
month beginning on the eleventh day
following. as rar away as rorty-one days. It is
almost a certainty that the rirst day or the
nrst month or the civil year. as we shall
hencdorth term the schematic lunar year. also
rell within these limits. As this Hrst day had
come to be the date or the rising or Sothis in
ca. 2773 B.C.. the civil calendar must have
been introduced between ca. 2937 and ca. 2821
B.C.• with the probability that it was in the
direction or the rormer rather than the latter
date. 46
This is a remarkably consistent account. It has the
advantage or dovetailing nicely with what we know
about the historical circumstances and chronological
limits or early Pharaonic Egypt hom other sources.
The only major diHicuhy with it is that it is entirely
-:J8-
dependent on Parker's usc of the beliaeal rising of Sirius

to govern the old lunar calendar. and. as I have already
remarked. this is anything but certain.
One last piece of evidence: may bear on the origin
of Egyptian knowledge of the length of the solar year
as 365 days. This is 8 petroglyph. most probably of
Predynastic date. which was discovered at Nekhen by
Ahmed Irawy Radwan and James O. Mills. It may
indicate 8 record of annual solar risings from solstice to
solstice (see the Postscript at the cnd of this volume),
The length of the solar year would have been much
more quickly established in this way than by averaging
the wild swings of Nile-risings. Needless to say. even if
the solar year was determined in this manner. we still
cannot be sure when the onicial civil calendar with its
beginning assigned to the heliacal rising of Sirius was
established.
Sothic Dates and the Ebers Calendar

We have gone as far as the evidence allows in
considering the origins of the civil calendar. It now
behooves us to note briefly the three extant double
dates that refer to the heliacal rising of Sirius in the
civil calendar (see Document 111.10), In our discussion
we no longer stick exclusively to the original
assumptions noted above in describing Meyer's
chronological procedures. but apply to the formulations
and tables of P.V. Neugebauer the various possible
values for the latitude of the observer C.> and the
so-called arc of vision Cll>. i,e., the minimum angular
distil nee between the sun and Sirius after which Sirius
is first visible on the horizon in the dawn,47 The latter
is somewhat misnamed. for at the minimum angle the
star is not visible and the subsequent visibility in ract
-37-
depends on variable conditions of the observation. as

we shall see,
The first of these Sothic dates (proceeding from
earliest to latest), is based on two entries from a temple
register from llIahun (see Doc. 111.10 and Figs. 111.928
and III.92b), 'Year 7 [of the reign of Sesostris liB,
[Month] III [of Season] Peret, Day 25....You should
I::now that the Going Corth <i,e" heliacal rising) of Sothis
takes place on [Month] IV [of Season] PeTet. Day 16"
and ·Year 7 [of the same reign], [Month] IV [of Season]
PeTet. Day 17.... Receipts [from the] Festival orferings
of the Going Forth of Sothis 'M' Loaves. assorted. 200;
beer. jars. 60 ...: It has been widely accepted that the
monarch in question in both references is Sesostris III.
Thus we have two entries for the seventh year of his
reign, the first of which predicts the rising on the
sixteenth day of the fourth month of Peret and the
second of which notes on the seventeenth day of that
month the orrering receipts for the festival of that
rising [which presumably did take place the day before,
i.e" on the sixteenth day]. We do not know with
certainty that the prediction was made as the result of
the technique of calculating by means of assuming a
one~day delay of rising after every four civil years. or
whether it was made as the result of a continuously
kept record of observed annual risings on which
accurale guesses were made beforehand. In the last
case the remarks of Edgerton are pertinent for the
determination of the Julian date to which it
corresponds,48
The view that a heliacal rising of Sothis
was predicted for the sixteenth day of the
eighth month in the seventh year of Sesostris
III seems, therefore, to be historically well
-)8-
CALENDARS. CLOCKS, ASTRONOMY
attested. Judging by what is known of the

history of Egypt during and before the
Twelfth Dynasty, it is almost certain that the
underlying observations must have been made
somewhere between laritude (4\) 29.2 0
(eH.ahun) and 30.10 (·Heliopolis). I pointed
out five years ago that the modern
experimental determination of the IIrCUS
visionis (ll) for Sothis in Egypt rests on
inadequate data: we must continue to use the
resulting figures (p may range at least from
8.60 to 9.40 ) but we must bear in mind that a
wider range is not excluded....
The earliest possible equation for the
Sothic date, if 4\-30.1 0 and p"9.5°. is 1876 D.C..
July 18. Julian: while if 4\"30.1 0 and p"9.4°, the
earliest possible equation is 1875 n.C.. July 18.
Julilln. The latest possible equation, if 4\"29.2 0
and p"8.6° is 1864 B.C.. July 15 Julian. The
sixteenth day of the eighth calendar month in
the sevenlh year of Sesostris JJI may have
fallen in any July from 1875 (less probably
1876) 10 1864 n.c. inclusive. The result may
conveniently be expressed as 1870 D.C. + ca, 6
years.
As a matter of fact, Parker, by considering lunar
dalll, arrived at what he thought was a more certain
date. namely 1872. with the first year of Sesostris III
thus being 1878, a date rather generally accepted for the
next generalion. 49
The second Sothic date is that of the ninth year of
the reign of Amenhotep I, IIJ Shemu 9 (see Document
11I.2, the Ebers Calendar below). Again Edgerton,
-39-
following the considerations of P.V. Neugebauer's

Astronomische Chronologie and the article the latter
co-authored with Borchardt <see notc 47>, has given us
the highlights of determining the limits of that dale in
the Julian calendanSO
The most important unkno....n quantities
Hor determining that Julian date] appear to be
the /lTCUS visionis...and the terrestrial latitude
of the ancient obscrvations....Tbc experimental
determination of the IIrcus v;sion;s, made in
1926 on Borchardt's initiative. marked I!l great
advance on previous knowledge but still falls
faT short of certainty. We arc given only five
individual observations. made by five different
persons at five different places. Since the
Ilreus visionis varies with shifting atmospheric
conditions and also with the eyesight or the
observer, it is evident that a very much larger
number or observations must be made berore
any historical conclusions resting on this basis
can really deserve the high respect which we
habitulllly record them. But ror lack or
anything better, we shall proceed to use what
we have.
Borchardt and [P.V.l Neugebauer
calculated that the actual 4rcus visionis ror
their rive observers in rive diHerent parts or
Egypt in July and August, 1926, ranged
between 8.60 and 9.40 , with an average or
9.00 , while they believed it possible that
under more ravorable weather conditions an
IIrcus visionis as low as 8.5 0 or even 8.3°
might have been possible....
I shall assume wilh Borchardt that the
-40-
ancient observations were made at Heliopolis.

latitude 30.1 0 [and use his range of values for
Pl... The conclusion, for historical purposes, is
that 9. II. W. n.e., III Shcmu 9] of year 9 of
Amcnhotcp I may have fallen in any July
from -1543 to -1536 [Julian] inclusive. if the
andent observations were made at Heliopolis.
It is possible that the actual 6rcus visionis of
the ancient observations may have been either
greater than 9.4 0 or Jess than 8.6°, in which
case the actual year mllY have been carlier
than -1543 or later than -1536.
If the ancient observations were made at
Thebes. the year might lie anywhere between
-1525 (earliest year if 1J"9.40 ) and -1518 (latest
year if p-8.6°). Other localities in Egypt
would yield yet other results. Sais liS an
observation point would yield earlier years
than Heliopolis, and Assuan would yield later
years than Thebes. No one believes that the
observations were made at Sais or at Assuan
[however, now see Document 11I.10, n. 4, for
Krauss's viewsl, but this disbelief rests on
purely historical considerations. which have
nothing to do with astronomy.
For this same Sothic date, Hornung gives the same
limits 1544-1537 a.c. <i.e.. expressed in historical rather
Ihan astronomical forml if the observations were made
at Memphis or Heliopolis, and the limits 1525-1517 B.c.
if they were made at Thebes.Sl The chronology I have
given in the Appendix of Volume One assumes Thebes
as the cenler of observalions and the lower limit of
1517 B.c., thus suggesting 1525 B.c. as the first year of
the reign of Amcnhotcp I. Needless to say, if the
-41-
observations were made at Aswan, as Krauss proposed.

then the rising in question could be as late 85 1506
(Document 111.10, n. 4), which date rests on the doubtful
assumption that the Ebers Calendar correlated a lunaT
and 8 civil year.
Whichever exact Julian date is appropriate for this
second Sothic date. the document on which it appeared.
namely the Ebers Calendar (Document 111.2), was the
object of much speculation and interpretation. Before
the time of Borchardt most Egyptologists believed that
the Ebers Calendar had the purpose of specifying in
civil year 9 of Amenhotep I first days of the twelve
schematic months (each having 30 days) of 8 fixed
Sothic year (see the Introduction to Document m.2
'below, where the views of the earlier Egyptologists are
described), In fact. even before the discovery of the
Ebers Calendar. Lepsius and Brugsch seem convinced on
the basis of very scanty evidence that the early
Egyptians kept a Sothic calendar of 365 1/4 days along
side of their civil calendar of 365 days.52
The Ebers Calendar on the verso of the rirst column
of the Ebers medical papyrus was discovered (or at
least its signiricance was discovered) in 1870 (see Figs.
111.10-12 and the discussion in the Introduction to
Document 111.2 below). Arter the rubric specifying the
regnal year of Amenhotep I, twelve lines follow. The
first column has a series of twelve monthly festivals
beginning with that of wp rnpt, the second had the
corresponding civil dates (each falling on the "ninth"
day of successive months), and the third is occupied on
the nrst line by the expression the -Going Forth of
Sothis" (-the hetiacsl rising of Sirius) and on each
succeeding line by a dot. The usual interpretation of
this document before Borchardt was that it gave
-42-
evidence of the Sothic fixed year regularly coexisting

with the civil year, since it began with the festival of
the rising of Sothis and proceeded by monthly festivals
successively thirty days apart. One obvious difficulty
was that the day of each successive festival fell on the
SlIme day of each month of the civil year (the ·9th")
without taking ·account of the epagomenal days inserted
in the civil year. One would hllve thought that after
falling on the 9th day of the fourth month of Shemu, it
ought next to have fallen on the 4th day of each of the
remaining civil months because of the insertion of the
five epagomenal days at the end of IV Shemu. There is
no completely satisfactory explanation of this
inaccuracy, but Edgerton. following the early views of
Lepsius, gives us a possible one,53
But the inaccuracy is one of the given
facts in our document, and instead of trying
to interpret it out of existence by
philologically improbable translations, it is our
du'ty to face the fact and to try to account
for it. The explanation may well start from
the purpose which led the ancient scribe to
write out this calendar in a prominent and
readily accessible place (the first page) on the
back of a medical treatise. The purpose, as
Lepsius pointed out, was to give the physician
an easy means of knowing at what seasons in
the year certain prescriptions were to be used,
No prescription in the entire document is
restricted to a shorter period than two
months. In calculating the calendrie
equivalent of a season of the year which was
two months long, an inaccuracy of five days
would probably not seem very serious to the
-43-
Egyptian medical practitioner. I still do not

understand why the scribe was so thoughtless
8S to write "9. 1. WI-Wandering. i.e., Civil
Year]: etc., where "5. I. W.: [! "4. I. W: etc.l
would have been just as easy. and obviously
more precise-but I think I have seen more
deplorable examples of thoughtlessness in
modern works whose reputation for accuracy
stands deservedly higher than that of the
Papyrus Ebers.
Up to the time of Borchardt's Mittel the date for
each civil month was read 8S "9" {psI!'. But Borchardt
read the hieratic sign used in these dates liS standing for
"New Moon Day' (pSr}nlyw) or sometimes perhaps
pS!i-nbb (where the sign for -nbb. i.e.. _. could be a
determinative) rather than for "9" lind accordingly he
believed the festivals of the first column to be those of
twelve successive lunar months beginning with the
festival of IVp rnpt (see Fig. 111.121 But since the
hieratic sign. without an accompanying sign 'for _.
which was interpreted by Borchardt as 'New Moon
Day: is precisely the same as the unadorned sign used
for "9" in the regnal date given in the rubric. and its use
for "New Moon Day" (when unaccompanied by the sign
for-n(lb) cannot be attested in any hieratic documents.
Borchardt's reading must certainly be rejected. 54 While
recognizing the cogency of this rejection. Parker
retained Borchardt"s belief that the first column gave
festivals in the lunar year, as we have already notcd. 55
In explicating his view, he properly felt the need to
explain why the festival of the month IVp rnpt,
identical later with that of Mesore, the last month, was
in the lead position. while t!Jy. the usual first month,
-44-
was shifted to the second position, to be followed by a

consequential shift of one position of all of the months.
His solution to this problem also, he thought, disposed
of the problem of the missing epagomenal days. Here is
what he says (ClJJendlJrs, p. 42):
The event which regulated the original
lunar calendar was the rising of Sothis, called
wp rnpt. The dale of this event would, then.
correctly go at the head of a calendar
governed by il. BUI this event also gave its
name to the last month of the year. In the
first column of the Ebers calendar. thererore,
the last month of the year appears at the head
of the months merely because its eponymous
feast determined the following year. The
correct interpretation of the second line of the
calendar seems to me to be that the date III
§mw 9 is common both to the going-forth of
Sothis and to the beginning of the lunar
month wp rnpt. From this date as a
starting-point was projected a schematic lunar
calendar of full months of 30 days. The
failure of the scribe to reCKon the epagomenal
days is accordingly not deserving of blame. as
it would be were we concerned with a fixed
ycar. The schematic lunar calendar was
regarded liS exactly that; it was to be merely a
guide to the proper identification of the lunar
month which included the 9th day of any civil
month in its first few days.
First consider Parker's last argument against the
omission of the epagomenal days. Thc problcm is not
that of the omission of the epagomenal days from the
first calendar. the schematized. lunar calendar according
-45-
to Parker, but rather is why the epagomenal days were

omitted from the civil calendar witb which it was being
correlated! The main part of his explanation. which
serves to explain the appearance of the last month first.
is also lacking in conviction, for if we: look at the first
'month" listed. namely wp rnpl, we see that it is not the
month (IV) that is given but rather the 'Teponymous]
festival of wp rnpt ClQI"_): as Parker well Ime.... and as
all the early discussants of the Ebers Calendar realized.
Indeed. on the basis of the evidence analyzed by Selhe
and Weill (see Document 111.2), it seems to me certain
(as it did to them) that not just the first but all of the
t .... elve cntries in the first column of the Ebers Calendllf
ought be considered as the eponymous feast days rather
than the months themselves, If this is so. then the
problem of why the months are out of order docs not
exist, For as Sethe has already shown. the eponymous
feast day ordinarily appears in the month following as
the culmination of the preceding month (see Fig,
1lI.6bl.56 Furthermore. the £net m that the Ebers
Calendar (in the third column) identifies the heliacal
rising of Sothis with the eponymous feast day of wp
rnpt, which indeed was the first day of the civil year
when the latter was organized or perhaps reorganized in
the third millennium B.c., and the fact (2) that that
feast day could not have been the first day of the old
lunar calendar outlined by Parker (which. unlike
Parker's suggested schematized lunar calendar presented
here, was itself not schematic but had months of
variable length). seem to dispose of any view that the
calendar which was being correlated with the civil
calendar was the old lunar calendar with its
intercalation scheme, In £net, all that remains of
-46-
Parker's old lunar year in the Ebers Calendar are the

names of the cleven monthly fcast days succeeding the
Feast of Wp-rnpt and indeed there is nothing to
distinguish the Ebers Calendar understood by Parker 8S
a schematized lunar calendar from an ideal. ad hoc
Solhic calendar. And hence what appears to be the elise
is that we have in the nrst column 8 truncated Solhic
year beginning with the rising of Sothis and containing
II more monthly feast days 30 days apart but Jacking
any epagomenal days.
As the result of this discussion we have so far
concluded that the year being correlated with civil year
9 of the rcign of Amenholcp I was an ad hoc fixed
year embracing twelve monthly feast days 30 days
apart, which year began with the day of the feast of
the Opening of the year. namely the day on which
Sirius rose heJiacally. I sayan ad hoc fixed year rather
than the Sothie fixed yelJr based on regularly adding 8
sixth epagomenal day e.'very fourth year because there
is no evidence whatsoever of the use of a sixth
epagomenal day in Egypl until the abortjve calendar
stipulated by the Decree of Canopus of 238 n.c. This
calendar was apparently ignored but perhaps served as
a model for the establishment of the Alexandrine year
(see Document 111.10, Decree of Canopus).
The effort by Weill,57 who firmly believed that the
Egyptians maintained a fixed Sothic Year of religious
feasts along side the civil year (see the Introduction 10
Document 111.5, note 5) is nol convincing. Still, as the
Sothic dates being here analyzed reveal, the date in the
civil year of the rising of Sothis was recorded on a
number of occasions, and thus was probably kept track
of on a regular basis. This was most probably by
observation but perhaps also occasionally by
-47-
calculation. even though no trace of a formal calendar

with a sixth epagomenal day added every fourth year
or with any more complicated scheme of correction
based on yearly observation clIn be found.
In concluding my account of the Ebers calendar. I
wish to mention what I think. may be a correlative
correct reason why the Ebers calendar was composed.
namely to assist in the construction or use of a water
clock.. an idea first suggested to me on reading SiegFried
Schou's Altigyptischc Festd/lten. but which I have
developed in a somewhat different direction. I shall
describe bow the Ebers calendar would have been used
for this purpose later in the chapter under the rubric
'Outflow Water Clocks: and at that time it should be
clellt that the major arguments against accepting the
Ebers calendar as a correlation between an ad hoc fixed
Sothic year and a given civil year can be answered.
Leaving the Ebers calendar. we may complete this
treatment of Sothic dates by merely referring to the
last such date from Pharaonic Egypt. namely that found
in a calendar in Aswan from an unknown year of the
reign of Tuthmosis III, III 5mw 28 (ca. 1464 8.C.).56
Further details concerning heliacal risings of Sirius in
ancient Egypt and Sothic periods arc presented below in
Document 111.10.
The Night Hours

So far we have concerned ourselves with [he grand
units of time represented in calendars with their years.
seasons. and months determined in some fashion by the
motions of the moon, the sun. and Sirius. Now it
behooves us to consider the day and the night and their
division into hours. For the early Egyptians nighttime
and daylight were each divided into twelve hours. thus
-48-
producing hours that varied in length according to the

seasons (I shall have more to say about this when I
discuss water and shadow clocks and the lengths of
days and nights for which such might be useful). But
we do not lenow the precise time when the number of
divisions became 12. We saw in Volume One, Chapter
Two and its documents, that by the time of the New
Kingdom a series of works had been composed that
described the passage of the sun through twelve
divisions of the Duat or Netherworld, each division
occupying a single "hour: But long before these worles
the Egyptians had divided the night into hours. Two
passages in the Pyramid Texts taken from the wall of
the pyramid of Wenis, the last king of the fifth
dynasty, confirm this.59 The first (Utterance 251. Sect.
269a) reads: 'Oh you who lire over the hours, who lire
before Re. prepare a way for Wenis: The second
rellltes (Utterance 320, sect. 515a) the following: 'Wenis
has cleared the night; Wen is has dispatched the hours:
The hours mentioned in these passages are no doubt
nighuime hours, since the contexts seem to confirm
this. In both passages the word for hours (wn'IV() is
determined by three stars, suggesting to us that the
most primitive meaning of 'hours' was "nighttime
hours: 60 Later, daytime hours were sometimes
determi,ned by the sun (0), as in the hour names listed
in Fig. 111.42, or determined by a star and complemented
by the sun when 'hour of the day" is meant (e.g., see
Fig. 111.38, line n.
It seems certain that the division of night into
twelve hours arose from the 12-month year and the
10-day week and thus was a consequence of the
installation and acceptance of the civil calendar. There
were of course 36 decades (i.e., lO-day ·weeks") in the
-49-
course of the 12 months embracing 360 days. At some

time during the third millennium o.c. star watchers
divided the night into twelve hours, determined by the
heliacal risings of stars or groups of stars called
decans. 61 Parker describes the invention and
development of the decanaI star-clock system
succinctly,62
We do not know exactly when the stars
were first used to tell time at night but it was
certainly by the 24th century D.C. (evinced by
the quotations from the PyrlJmid Texts above]
and quite possibly soon after the introduction
of the civil calendar....
It was not until roughly 2150 D.C. that we
know for certain that these night hours
totalled 12, This we learn from diagrams of
stars on the inside of coffin lids, which were
earlier termed 'diagonal calendars' but which
are better called 'star clocks'. Figure I (omy
Fig. 11I.13] is a schematic version of such a
clock. followed in some degree by all our
examples. It is to be read from right to left,
A horizontal upper line T is the dale line,
running from the 1st column, (j.e.,l the 1st
decade of the 1st month of the lsi season, to
the thirty-sixth column, (i.e.,l Ihe last decade
of the 4th month of the 3rd season, With
decade 26 there begins in the 12th hour a
triangle of alternate decans to lell Ihe hours
through the epagomenal days....
The mechanism of such a clock was the
risings of certain selected stars or groups of
stars that conventionally afe termed 'decans',
at 12 intervals during the night, and at IO-day
-so-
intervals through the year, If we now

examine the stllr dock of Idy...[my Fig. 111.14]
even though it lists only 18 decades we can
see very graphically how the name of a
decanal star in Bny hour is always in the next
higher space in a succeeding column [i.e.. in
the next decade), so that a star in the 12th
hour rises over 120 days to the first hour and
then drops oul of the clock....
We have already seen in connexion CD
with the star Sirius-Solhis thai it eventually
disappears because it gets too close 10 the sun,
and then afler some days it reappears on the
eastern horizon just before sunrise, its he\iacal
rising....
Of course, 8 slar which has just risen
hcliacally does not remain on the horizon bUI
every day, because of the Enrth's travel about
the Sun. rises a little earlier and is thus a little
higher in the sky by sunrise. EventulIlly
another star, rising helillcally is likely to be
called the Morning Star. In the early third
millennium B.c.. we may conjecture. the
combination of IO-day weeks in the civil
calendar lind the pattern of successive morning
stars led some genius in Egypt to devise II
method of breaking up the night into parts, or
'hours', He observed a sequence of stars each
rising heliacally on the first day of a decade
or week. from a text in the cenotaph of Seli
I at Abydos [see Documenl 111.121.. we learn
that stars were chosen which approximated
the behavior of Sothis in being invisible for
70 days. The star which had risen heliacally
-51-
on the 1st day of the first decade. thus

marking the end of the night. would 10 days
latcr be rising well before the end of night.
The interval between its rising and the new
star's beliacal rising would be an 'hour',
Inevitably this pattern of heliacal risings at
IO-day intervals would lead to a total of 12
hours for the night.
Parker goes on to observe that all of the decans
marking the hours would fall in a band south of and
parallel 10 the ecliptic. This is illustrated by Fig. 111.15.
where Orion and Sothis are marked on the band. The
hours so marked off by the successive risings of these
decans would obviously not be of equal length. for as
the length of tbe night increased the length of the
hours would increase and as it decreased the hour
lengths would decrease. As Parker notes. it is the
combination of the lengthening and shortening of the
night with the periods of morning and evening twilight
and the oscillation of the star clock "that explains why
the night was divided into 12 hours though there were
always 18 decanal stars from horizon to horizon in the
night sky:63 The time from sunset to the appearance
of very bright stars (the so-called civil twilight)
averaged approximately 1/2 hour, while the time from
sunset to that when all the stars were visible (the
astronomical twilight) may be over an hour and the
decans determined the time of total darkness
exclusively. The first hour of the night was of
indeterminate length. beginning with total darkness and
ending with the appearance of a decanal star in the
eastern horizon. That hour became progressively
shorter each night or the decade, until it was replaced
by the rising of another decan as the determiner or the
-52-
first hour during the next decade. The end of the

t .... elfth hour. at the beginning of a decade, was
determined by the heliacal rising of its deean and
followed soon therclIftcr by morning twilight.
Obviously. as the decade advanced the twelfth hour
was followed by a progressively longer period of
darkness, which was apparently incorporated into the
twelfth hour, its length being considered as lasting until
light.
Deeana! Clocks
Now we are prepared to examine in somewhat more
detail the early diagonal clocks based on the risings of
deCORS. Our generalized decaRal clock given in Figure
1JI.l3 is based on the detailed analysis of twelve father
incomplete and defective clocks on cornn lids that date
from about the time of Dynasty 9 or 10 to that of
Dynasty 12. The twelve clocks can be placed into five
groups, each of which has been cleverly dated by
Neugebauer and Parker.54 The least corrupt of the
texts on the various lids are those of Coffins I and 2
(sec particularly the coffin of Meshet given in
Document III.II and also Fig. 11I.16 for Coffin 2), which
have 34 decans instead of the full 36 decans. Indeed in
Document IIUI we are told in column 40. lines 1H2.
that 36 decans ought to be there (Iotal of those who
are "'in their places', the gods of the sky, 36").55 The
first line of the diagonal clocks, labelled T on the model
clock of Fig. 111.13 and reading from right to left,
contains the headings for the thirty-six decades or
IO-day periods we conventionally call "weeks" of each
month (labelled "first: "middle: or "last">. The months
are grouped into the three seasons of four months
which we have described above when discussing the
-53-
civil calendar. Seven of the twelve corrins contain the

dale line (e.g.• see Document 111.11 and also CoHin 2 in
Fig. 111.16).
Then under each decade-column ....e rind t .... elve
decans. I£ .... e look carefully at the clock of Mcshet as
depicted in Fig. 111.86. which is described in detail in
Document 111.11, and also at another sample decanal
clock, that of Idy. shown in Fig. 111.14. we notc that the
deean in the twelfth row of the nrst column. whose
heliacal rising marks the end of the twelfth hour of the
nights of the first decade and whose hieroglyph is
composed of two pairs of facing men Ccrew?'), we see
that the decan shifts up to the eleventh row in the
next decade. while a new decan. namely knm. occupies
the twelfth row of that decade. In the twelfth decade
or column the decan of the two pairs of men has
reached the first row of the twelfth decade and thus
marl:.s the end of the first hour during that decade. The
declln then passes from the eastern sky. Skipping over
to the bottom of the 25th column in our idelll clod
schemlltized in Fig. 111.13, the 36th declln (p/Hvy !Jlat,
shown in the bottom of the 23rd decllde in Fig. 111.16)
would hllve lIrisen heliacally to marl:. the end of the
twelfth hour during the 25th declIde. That decan will
ultimllte1y marl:. the end of the first hour during the last
or 36th decade of the normal yellr of 360 days. The
twelve decans appearing in the twelfth row
successively from the 26th decade onward (j.e.. the 11
decans from the 26th through the 36th decades and the
12th decan originally no doubt in the 37th column but
later on in the 40th column) serve to determine the
hours of the five epagomenal days. Hence that last
column (whether the 37th or 40th) strictly speal:.ing is
a pentad (j.e., a period of 5 days) rather than a decade.
-54-
These twelve additional decans form a kind of triangle

on the diagonal chart as they progressively rise through
the hours from decade 10 decade, and the hypotenuse of
that triangle is marked in Fig. 1lI.!3 by II bold step-like
broken line. The result is thai the twelve hours of the
night during the last II decades of the regular year of
360 days and the pentad of the epagomenal days are
marked by some combination of the regular 36 decans
and 12 supplementary decans (slarting with 8 sum of II
of the former and I of the latter in the 26th decade lind
progressing to the sum of I of the former and 11 of the
latler in the 36th decade) and rinally in the last column
<the pentad of the cpagomenal days) by the twelve
supplementary decans alone. Inddenllllly. as J have
said, it seems likely that the last column was initially
merely the 37th column. But apparently this was
expanded into 4 columns, the first three occupied by a
list of the regular 36 decans and the 40th by the 12
supplementary decans.
The- identification of the various decans is
extremely difficult (see Fig. 111.18 for a table of the
decans based on the star clods>. Not only are all of
the extant star clocks corrupt textually to some degree,
but even when we are certain of the decanal names and
their translations, their proper astronomical identity is
for the most part unknown. 66 Decans numbered 26-29
are related in some fashion to Orion (as are the
epagomenal decans C and m and 30-31 to Sothis (as are
decans F and G).67 One interesting detail of the star
clocks is the strip (V) of four sky-related figures
depicted between columns 18 and 19. As shown on
Coffins 2 and 3 (see Figs. 111.14 and 111.16-111.17), they
are the goddess Nut. holding up the sky; the Foreleg of
an Ox (-the Big Dipper); Orion. looking back toward the
Foreleg and holding an antlr(Cn[J-Jsign in his right hand

and a 1V"s-(IV~Jscepter in his leh l and Sothis. the
goddess of the star Sirius, hcing Orion and holding a
was-scepter in her right hand and an 8ntlrsign in her
left. Finally we should note that between ro....s 6 and
7, each coHin lid contains II strip of non-astronomical
offering prayers for the deceased.
The fact that the diagonal clocks were organized to
tell night hours during the civil year of 36S days
instead of the Sotbic year of approximately 365 1/4
days guaranteed that they ....ould quickly lose even the
moderate usefulness they might have had ....hen •
particular clock was composed. Decans would have to
be shifted in place and some new ones used. Evidence
for a revision toward the end of the twelfth dynasty
exists, but by that time a new system based on the
transiting of the meridian by decans was developed, as
we shall see below, Be that as it may, it must be
remembered that the extant diagonal clocks on the
coffin lids were funerary in purpose and hence largely
symbolic, such a clock being merely one more necessary
convenience for the deceased to have at hand for
eternity just as he also had at hand the offering prayers
to sustain his supposed physical wants during that same
eternity, Hence the accuracy of the corrin clocks was
probably of little moment.
Transit Occanal Clocks

We learn of the new system of using the transiting
of the meridian by decans. i.e., their culminations, to
mark the nighttime hours from later texts which may
be collectively called the Book of Nut or the Cosmology
of Scti I and Ramesses IV (see Document 111.121 These
texts accompanied a large vignette of the sky~goddess
-56-
Nut in a vaulting position..... ith her feet in the east and

her extended arms in the west, both ends touching the
earth. Bnd her body supported in the center by the
air-god Shu. Examining Document 11I.12 will give us a
good picture of the transit clock. They tell us that the
decen stars, on the model of Sirius, disappear in the
Duat. i.e., the Otherworld. for 70 days. after which time
they rise again. The account is of a simplified scheme
of 360 days. an old star dying and 8 new star being
born every 10 days. After spending 70 days invisible
in the Duat. 8 decan rises and spends 80 days in the
eastern beaven before it culminates or tURsilS the
meridian, after which it 'works' for 120 days, its worlc
being [0 mark a given hour by transiting for tcn days
starting with the 12th hour during the first decade and
ending with the first hour during the 12th decade.
Then. its ·worl:" being over, it spends the remaining 90
days in the western sky before dying once more. The
change from rising to transit resulted in changes in the
decanal list. as Parker observes when describing and
evaluating the transit clods.58
Since 8 star spends 80 days in the east
before working. it is clear that it is transiting
when it marks an hour. This change to
transiting required a wholesale readjustment
of decanal stars. When rising, two decanal
stars could mark an hour between them but if
they were at opposite sides [i.e.• boundaries]
of the decanal belt, thus some distance apart
on the horizon. they could, when transiting,
either pass the meridian together or so far
after one another as to result in an extremely
long hour. or the 36 decans of the star clocks
nearest in time to the transit scheme only 23
-57-
remain exactly in place in the transit list. The

other 13 have each a different place or drop
out and are replaced by new decans.
The simplification of the clock on the basis of 360 days
reveals to Parker69
the strongly schematic character of such a star
dock and poses the question of its real utility.
In actual practice one may speculate that all
an observer would need to do would be to
memorize a list of 36 decllnal stars, preferably
rising ones as being the easier to observe,
watch to see which one was in or near the
horizon when darkness fell and then use the
risings of the next 12 to mark off the night
hours. Whether this was done in fact cannot
be said with certainty. bUI it is true .that
while we have lists of decaRs on various
astronomical ceilings or other monuments to
the end or Egyptian history, we have nothing
at all approaching a star clock in rorm aher
the time or Merneptah (1223-1211 O.C.), and
that that was a purely runerary relic is
indicated by the ract that its arrangement or
stars dates it 600 years earlier into the
TweUth Dynasty.
However long the practice or telling time by the
observation or deC8ns lasted, the title or "hour-watcher"
(Imy wnwt or wnwty) persisted until the Ptolemaic
period, but perhaps with the general meaning or
astronomer once water clocks, shadow clocks. and sun-
dials came into common use. 70 This is borne out by an
inscription hom a sighting instrument made ror an
observer named Hor, or the sixth century B.c.. which
has on its bottom the rollowing statement, "I knew the
-58-
movements of the two disks (j,e., the sun and the moon)
and of every star to its abode; for the ka of the
hour-....alcher amy wnw/} Hor. son of Hor-wedja: 71
Still, even if the title had a morc general meaning. one
would suppose that the astronomer might have assisted
in overseeing the remains of the old star clocks that
were added 10 temple monuments.
The Ramesside Star Clock

Before examining waler clocks (in most of the
ancient period used exclusively for telling time at
night) and shadow clocks and sundials which allowed
for different systems of the measurement of hours in
the day, we can nOle the passing by the middle of the
second millennium D.C. of the decanal clocks based on
risings or transits for anything but decorative purposes.
Still we should briefly examine a new form of star
clock usually designated as the Ramesside Star clock
because it appears on the ceilings of a number of royal
tombs of the Ramesside period. Two scts of the tables
appear in the tomb of Ramesses VI, and one each in
those of Ramesses VII and Rilmesses IX. 72 Though all
of the sets arc corrupt in some fashion, a prototype
clock can be constructed from them (consult Document
111.141 Recall that the ancient decanal clock in its ideal
form had 36 columns of 12 hours whose ends were
marked by the rising of decans and that in its modified
form the hours were marked by successive transiting of
the meridian by decans. But the new Rllmesside clocks
had instead 24 tables (two per month) of 13 stars, with
the first star marking the beginning of the night.
Unlike the earlier transit docks it did not confine itself
to the transiting of the meridian but also included the
transiting of lines before and after (but presumably
-59-
near) the meridian.

Each of the 24 tables was accompanied by the
rigure of a seated man Cthe target figure) with a grid or
chart consisting of nine vertical lines and thirteen
(sometimes erroneously only twelve) inner horizontal
lines (see Figs. 111.19110 and 1II.19b). The inner seven
vertical lines are lines of transit. the fifth (of the nine
lines) representing the meridian circle. and the other six
inner lines representing lines of transit before lind lifter
the meridilln; these lines Ire appllrently It equal
increments of distance from the meridilln. The inner
horizontal lines represent first the beginning of the
night. for which the first star is given. lind then.
successively. the ends of the twelve hours of the night.
The twelve hours are IIccomplnied by the names of
stars. each one mllrking the end of an hour when it
crosses one of the vertical lines that Ire related to the
target figure positioned within or below the grid.
Apparently a wllter clock was used to set the
original hour ends. the north-south meridilln having
been established. At the end of In hour. as determined
(without great accuracy) by the water clock. a marking
star was looked for, preferably one at the meridian but
if there was no significant star there on the meridian
then one transiting one of the other six inner vertical
lines that were mentioned above as being before or
after the meridian. Preceding the list of stars. each
table bore a date in the civil calendar. the first table
having the date "I Akhet [1]: which signified that it was
valid for the period from the first through the 15th of
that month. the second " Akhet 16, [2nd] half-month:
signifying its validity from the 16th day of the month
through the 15th day of that second hair of the month
(j.e., the 30th day of the month: see Document 111.14,
-00-
notc 7), Though they are not separately mentioned. it

seems probable that the epagomenal days (along with
the lasl fifteen days of the fourth month of Shemu)
....ere to be covered by Table 24.
From (I) these dates as given in the four extant
copies and (2) the notation that Sothis culminated at
the beginning of the night in Table 12. dated II Pere! 16,
and (3) hom our knowledge of the -relevant Sothic
cycle. we: can deduce a dale of about 1470 B.C. for the
stu clock. i.e., more than three centuries earlier than
the Ramesside monuments on which the clocl::
appcars. 7J
After the date in the tables a phrase 'Beginning of
the night" is given. followed by the name of a star
transiting at that time:, and this refers to the first
interior horizontal line. Then the names of the twelve
hour stars are listed over the succeeding horizontal
lines. The name of each star is followed by a brief
phrase descriptive of its position relative to the seated
figure, i.e., Ihe target figure: ·opposite the heart" (r et '
Ib) if on the centcr line or mcridian, 'on the right eye·
(/Jr Irt wnmy) if on Ihe first line left of Ihe center line,
·on the lert eye' ((lr Irl flby) if on the first line right of
thc center line, ·on Ihe right ear· (pr msgr wnmy) if on
the second line left of the cenler line, ·on the left ear
(pr msgr !'by) if on the second line right of the center
line, ·on the righl shoulder· (pr tC(I wnmy) if on the
third line left of the center line, and ·on the lert
shoulder (pr tC(I "by) if on the third line right of the
center line. These verbally indicated positions are in
addition marked on the accompanying grid by star
marks (see Figs. 1II.l9a, 1IJ.19b, 1II.99b, and 1IJ.99c).
Parker presents his view of how these transit
clocks were used: 74
-61-
.... On a suitable viewing platform. probably a

temple roof, two men would sit facing one
another on a north~south line. The
northernmost would hold a sighting instrument
like a plumb bob (called by Egyptians a mrbt
[see Fig. 1II.20a and Fig. III.20b]) before him
and would call out the hour when a star had
reached either the meridian or one of the lines
before or after 8S sighted against the target
figure. The effort for such precision points to
the use of the ....aler clock as an independent
means of marking when an hour had ended,
and emphasizes 8S well the reluctance of the
Egyptians to abandon telling time at night by
the stars. Indeed. a water clock of the
Ptolemaic period [Borchardt's Auslaufuhr 3]
has an inscription on its rim that its purpose is
to tell the hours of Ihe night only when the
decanal stars cannot be seen (Borchardt
[Zeitmessungl. 1920. p. 8 [d. also p. 9 for
Auslaufuhr lOll
If the lists of stars in the 24 tables are examined
(see Document 111.14). it will be evident that only three
of the stars used in the Ramesside cloclr: are the same as
or near to those used in the decanal clocks, ·Star of
Sothis" (sb ' n spdtJ. i.e.• Sirius. "Star of Thousands" (sb ' n
!J'W), and "'Star of Orion" (sb n Sip), the last two not
'
identifiable with any specific star. Incidentally, it will
be seen from the list of the stars found in the 24 tables
that many of them are stars from or related to other
well-known Egyptian constellations or stars, such as Ihe
"Giant" (16 stars), the ·Bird· (4 stars). "Orion" (2 stars.
one known as the "Predecessor of £the Slar of] Orion·).
and ·Sothis" (2 slars, one of which is called "The One
"62-
Coming After £the Star of] Sothis·). The conclusion of

an inquiry into the star list seems to be that for the
most part the stars of the Ramesside tables lay roughly
parallel to but (except for slight overlapping) outside
and south of the decanal belt (see Petrie's effort. here
reproduced as Figs. 1II.84a and 1II.84b, to reconstruct
the constellations and stars of the Ramesside Clock
along with the northern constellations and some of the
decanal stars).75
So far we have seen that the decanal and Ramesside
clocks differ (I) in all but three of the stars they used,
(2) in the periods during which each list of twelve hour
stars was effective (36 ten-day periods in the decanal
clock and 24 half-month periods in the Ramesside
clock), and (3) in the fact that the former (in its transit
version) used only transits of the meridian to mark the
hours, while the latter used transits of the meridian and
of lines before and after the meridian. Notice further
that the decanal star marking successively the twelfth
hour, eleventh, tenth, etc. through the first hour in
successive ten-day periods in the decllnal clock docs not
change, but that in the Ramesside clock the hour stars
do not always move regularly up the hour-scale from
one half-month period to the next. (4) In addition, the
hour lengths themselves (however much they vary from
each other), once established, remain essentially fixed in
the decanaJ clock but often varied in length in its
successor as the result of star substitutions "and shifting
hour boundaries on. before or arter the meridian: 76
Incidentally, as in the case of the earlier star clocks, the
twelve hours of the night marked in the tables or the
Ramesside clock, like those marked in the decanal
clocks, were not equinoctial hours but were variable.
seasonal bours that divided into twelve the period of
-63'
total darkness (without twilight> as that period varied

from half-month to half-month. 77 Neugebauer and
Parleer are careful to point out the inaccuracies inherent
in the procedures of transiting in the Ramesside
c1ock,78
When we describe the Ramesside star
tables as lists of transits the reader should not
be misled into associating with this term any
astronomical accuracy we are accustomed to
assume for transit observations. Actually the
sources of inaccuracy in the Egyptian
procedure are only too apparent. That two
persons, sitting opposite each other, cannot
resume exactly the same position night after
night is clear. To fix accurately the moment
of transit. when even very small motions of
the eye of the observer will displace the
apparent position of a star, is impossible.
Even if we assume optimistically that the
target figure was replaced by a fixed statue
the least shift of the observer's location would
influence badly the accuracy or timing.
Schack*Schaclcenberg assumed that a frame
with vertical strings was used to give
precision to such vague terms as 'on the
shoulder: 'on the ear: etc. Though we have
no evidence to support such an assumption the
results would still be crude, if only rrom the
lad or definition of the position of the
observer's eye with respect to distance,
altitude above the horizontal level and
deviation from the exact meridian line. And
whatever resuhs might have been obtained in
one place could not be repealed wilh any
-54-
•
CALENDARS, CLOCKS, ASfRONOMY
accuracy by another observer and target

somewhere else. To this has to be added the
well-known inaccuracy of time measurement
by water docks. which must have been tbe
instrument on which the star tables were built.
In shortl Just as in the case of the decans
the crudeness of the underlying procedures is
so great that only under severely restrictive
assumptions could numerical conclusions be
abstracted from the given lists. If we add the
fact of obvious errors and carelessness in
details in the execution of the texts as we
have them one would do best to avoid all
hypothetical structures designed to identify
Egyptian constellations from the analysis of
the Ramesside star clocks.
A more recent, quite diHerent <and I believe less
satisfactory) analysis and evaluation of both the data of
the Ramesside star clock and the procedures used in
establishing and using it has been given by E.M.
Bruins.79 As will have been evident to the reader. the
account I have given here and more fully in Document
111.14 rests primarily on the pioneering treatment of
Peter Le Page Renouf in 1874 and the thoroughgoing
editorial and analytical eHorts of Neugebauer and
Parker.eo
Outflow Water Clocks

In the description above of the Ramesside slar
clocks. it was suggested thaI waler clocks may have
been the instruments with which Ihe hours of star
tables were established or at least they may have been
used as substitutions for star clocks. Two general types
of water clocks were known and used in Ancient Egypt,
-65-
Outflow Clocks and Inflow Clocks. with remains of the

former outnumbering those of the latter.SI The earliest
and most complete specimen of an outflow clock was
one found in broken pieces at Karnak in 1904 by G.
Legrain and restored for the Cairo Museum (see Fig.
1l1.21a).82 About 14 inches high. it dates from the reign
of Amenhotep III (ca. 1391-1353 B.C.>, is made of
alabaster. and has the form of an inverted truncated
cone. Thus it is shaped rather Iilce 8 flo....er pot. The
vessel was filled with water which Clowed out
gradually from a small aperture or oriHce near tbe
bottom of the vessel. The Egyptians used the truncated
cone for the form of this and other outflow c1epsydrac
to compensate for de<:reasing water head as the water
level descended and so to obtain a more even flow. By
noticing the dropping water level against a scale of 12
hours (in fact 12 scales, one for each month) on the
inside of the vessel the hour was obtained. However,
the conical form only approximated the paraboloidal
interior surface that would produce steady outflow and
thus a better estimate of the time. In fact a vessel with
a height of 14" and a conical shape like that of Karnak
would have produced a better approximation to an even
flow (j.e., an equal descent of the water level in equal
times) if the angle of slope of the vessel had been
slightly less (103 0 instead of about 11(0).83 We shall
discuss the form of the clock in more detail later.
let us examine the outside of the vessel in the four
views comprising Fig. 1I1.21a. Astronomical and other
decorations of colored stone and fayence inlays are set
out in three rows or registers (see Fig. 11I22) that
comprise the so·called celestial diagram, which we shall
examine in more detail when we examine Egyptian
astronomy later. The first two rows are interrupted by
-66-
a vignette of King Amenhotcp III making an offering to

Reharakhti with Thoth behind the Icing. The uppermost
row (starting on the right) represents the decans. Just
past the midpoint we see Isis as Sothis (Sirius) in a
baric. Then continuing to the left we see the planets
and the so·called triangle decans. The middle row has
the more prominent northern constellations in the
center plus some deities on both sides. The bottom row
includes six frames or fields that display the Icing with
the twelve gods of the months, two gods per field.
Between the first (just left of center) and the sixth (to
the right) was located the outflow aperture.
Examination of later specimens of waler clocks and of
the astronomical ceiling of the Ramesseum of Ramesses
II (Fig. 111.2), which like other such ceilings mimics to a
considerable extent the exterior decorations of water
clocks, makes it probable that a small statue of Thoth as
the cynocephalic baboon and thus 8S the patron of time
was once mounted over the aperture (see also Fig.
111.23>.84
Turning to the inside of the Karnak clock (Figs.
1II.24a-b), we see the aforementioned twelve scales of
varying total length, each divided by depressions more
or less equally spaced that divide the night in the
specified month into 12 hours. The month for which
each scale was originally intended 8t the time of its
construction (or better, at the time of the construction
of the water clock from which it was copied) is
indicated on the top rim of the clock (and thus with the
hieroglyphs inverted) in the conventional form of
months of the civil year: Month I, Season Akhet; Month
2, Season Akhet;...Month 4, Season Shemu. No effort
was made to give separate scales for the epagomenal
days. Presumably either the scale for the twelfth
-67-
month or that for the rirst month was used to tell the
hour during the epagomenal days. Under each month
the hour scale consists, as I have S1Iid. of a linearly
distributed series of small circular depressions each
about 1/5" in diameter.BS When the water level is at
one of the depressions it indicates the hour at that time
of the year to which the scale belongs. While the
depressions of each scale are by no means precisely
placed, the total lengths of the scales were quite
accurately measured, as we shall see later. The longest
scale is about 14 ringerbreadths. which is applicable to
the· month containing the winter solstice and thus the
longest night of the year. The shortest scale is about
12 ringerbreadths and is used for the month containing
the summer solstice and thus the shortest night. Each
fingerbreadth is equal to 18.75mm (j.e., about 3/4"). The
14,12 ratio is not a very good approximation of the
length of longest night to shortest night for Egypt.
The ratio 14,10 would have been a betler one. I shall
have more to say about the 14,12 ratio later. But we
should observe now that. since the civil year was about
1/4 day shorter than the rixed seasonal year, such a
clock progressively becomes out of synchronism with
the solar seasons (j,e., a shih of one month in about 120
years). Hence as time went on the scales were
necessarily being applied to months other than those to
which they were assigned in the clock. In fact, since
the shortest scale of 12 ringerbreadths was applied in
the Karnak clock to the civil month II Shemu, i.e.. the
tenth month of the civil year. it is evident that when
the clock was built in Amenhotep Ill's time it was
already out of date. since the assignment of its shortest
scale to the tenth month was not valid for the period
of Amenhotep III but rather for that of Amenhotep I
'68-
morc than one hundred years earlier (see Fig. 1II.37)

when a certain official named Amenemhet describes
having constructed such III clock and dedicated it to
Amenhotcp I. So it is evident that the maker of the
Karnak clock simply copied the monthly assignments of
the twelve scales hom an carlier clock, and thus the
scales were about one month off. We shall speak more
of Amenemhefs "invention" shortly. Incidentally. it
would have been easy to keep a clock like that of
Karnak approximately up to date merely by shifting the
month names on the rim of the clock after about 120
years. There is. however, no evidence that any effort
was made to cover over the month names on the rim of
the t(arnak clock itself and replace them with new
month names.
Sioley sums up the use of the Karnak clock as
follows,86
In use, the vessel was filled to the full-line,
not visible in the Karnak example. where it
had probably been painted on. Traces of the
line are to be seen in the other fragments. At
the end of the first "hour" as measured by the
clock. the water level would have sunk to the
first mark. and so on successively through the
various "hours" to the last mark. Except in
the case of one month. the last mark is not
shown as its position is covered by the row of
alternate signs [j.e.• glyphs signifying] LHe and
Stability which decorate the interior [as a
bottom register! The rlow of water from the
orifice averaged 10 drops a second,
In fact when the vessel was filled and then started
to empty when the aperture was opened, the water at
first rlowed in a fairly constant stream; but, towards
-69-
the bottom. it slowed and dribbled out. Hence true

equal hours were not produced. as we shall discuss
further on. As Sioley further notes:87
Although the hours were nearly correct at the
middle of the scale. the earlier were too long
and the later too short (Fig. 16 [d. my Fig.
1I1.3iJ). The Egyptians did not know this.
They had no [sure] means of determining
whether their hours were equal or not. To
divide a space of time equally is a very
difficult problem without [0] regularly moving
mechanism or some precise means of observing
the movements of the stars.
There are two pieces of evidence of written
material that deserve examination in our effort to
understand the development of water clocks. The first
document bearing on water clocks is an inscription from
a ruined tomb near the top of the hill of Sheikh cAbd
el-Gurna in Western Thebes. This was the tomb of the
official mentioned above. Amenemhet. who lived under
the reigns of the first three kings of the 18th dynasty:
Ahmose. Amenhotep I, and Tuthmosis I (see Document
11I.15 below and fig. 111.25). In the much worn
inscription the deceased tells us (Jines 7-9) that "while
reading in all of the books of the divine word" (j.e.. the
whole of Egyptian literature) he found lthe (longest)
night of wintertime to be] 14 [hours long] when the
[shortest] night of the summertime is 12 hours UongT
and that the hour lengths from month to month
increase and then decrease. This [its in precisely with
the Karnak water clock as we have described it: the
monthly scales of the lengths of the nighttime hours to
be used in a water clock vary from 12 fingerbreadths on
the summer solstice to 14 fingerbreadths on the winter
-70-
solstice with the intervening monthly lengths first

increasing uniformly and then decreasing uniformly
(with the fingerbreadth • 114 palm· 1/28 cubit· ca.
3/4">' Compare this description with the measurements
of the monthly hour scales in the Karnak clock which
have been detailed by Borchardt.BS and reported
succinctly by SloJey:89
The fundamental scale (12fl is the shortest
of the Karnak clock.... The aperture of the
Karnak clock is a distance of 4f below the
lower end of the lZf scale.
The variations in the scales from month to
month must now be considered.
In the Karnak example. the scalc lengths
in terms of fingerbreadths "r: are
approximately as followsl-
r
10th month....11
11th and 9th monlhs .12 1/3 f
12th and 8th months .12 2/3r
1st and 7th months .13 r
2nd and 6th months .13 1/3r
3rd and 5th months .13 2/3 r
4th month....14 r
The errors are almost negligible except in
r.
the case of the rigure 13 r
which is .25 too
great.
These rigures show that a uniform change
r
of 1/3 per month was intended.
The second piece of literary evidence that might bear
on the development of the water clock dates from
about the same period as the rirst, i.e., the reign of
Amenhotep I. is less certainly connected with a water
clock of the Karnak type or in fact any water clock.
However, a good case for its pertinence to water clocks
ANClENT EGYPTIAN SClENCE
can be made. This is the Ebers calendar. which has

been mentioned earlier and translated below as
Document 111.2. This document has a first column
consisting of the eponymous feast days of 12 months of
a truncated Sothic year beginning with the rising of
Sirius (and, like water clocks, without epagomenal days)
and a second column listing the ninth day of each of
the corresponding months of the civil year, this year
being the ninth year of the reign of Amenhotep I. I
believe that this calendar or equivalency table was
composed to aid in constructing or using a water clock
like the Karnak clock. As in the case of the Karnak
clock the prospective clock maker could have placed the
12 months of the civil year about the rim of the vessel.
Then he could have found out from the Ebers or similar
Sothic-civil double calendar which civil month in the
ninth year of Amenhotep's reign is equivalent to the
Sothic month that includes the shortest night. i.e.. the
month that includes the summer solstice. No doubt he
would know that this would be in the 11th Sothic
month. the first month of that year being the time of
the appearance of Sirius. Then he would simply put the
shortest monthly scale (say the 12f scale) under the
proper month name on the water clock he is
constructing. that is, under the current civil month II
Shemu, as the Ebers table would show him. He could
then do the same thing for the longest hour scale (say.
the l4f scale), and for all the intervening scales,
assuming. as the maker of the Karnak clock did. that the
hour scales uniformly varied in length by 1/3 f each
month. This in fact may have been what Amenemhet
did, i.e.• construct such a clock of the Karnak variety
after studying the literature and devising the monthly
scales. Perhaps its originality lay in his placing all
-72-
twelve scales in one clock with a single aperture.

Or it is possible that the hour watchers at this time
had at their disposal a kind of fixed water clock where
the scales had been inserted under the rubrics of the
months of the fixed Sothic year. Then the watcher
could simply look in the second column of the Ebers or
similar calendar for the civil day and month he was
using the clock. Opposite it he could then find the
equivalent Sothic month and thus was able to know
which hour scale on the water clock to use. The only
difficulty in this last reconstruction is that so far we
have not found any early water clock of the fixed
season variety. Furthermore. if such a clock with
monthly scales of hours already eXisted. we must
wonder why Amenemhet would have bothered to single
out his construction of a water clock as an achievement
to record in his tomb.
A few remarks regarding other outflow clocks are
in order. Borchardt has shown that the fundamental
scale of 12 fingerbreadths that was used in the Karnak
clock for the shortest night appears in other clods to
have been used as the scale for the months embracing
the equinoxes. so that the scale for the shortest night
was 11 fingerbreadths. for the equinoxes 12. and for the
longest night 13.90 Although in the Karnak clock the
monthly scales vary in length by the constant amount
of 1/3 of a fingerbreadth per month, these variations do
not represent the actual changes of the lengths of the
night from month to month, regardless of whether one
defines the night as lasting from first darkness to last
darkness or sunset to sunrise. The clocks that have the
12 fingcrbreadths scale as the length of the hourly scale
at the time of the equinoxes-all from the Alexandrine
period <Borchardt's clocks nos. 2-4}-attempt to correct
·73-
the constant changes of the scale lengths to take into

account the fact that the length of night chllnges most
slowly during the months immediately preceding and
following the shortest or the longest night. i.e., the
nights of the summer and winter solstices. and most
rapidly in the months immediately preceding and
following the months containing the equinoxes. Instead
of the constant relationship 1:1:1 they have what appears
to be an incremental relationship of m,2:3.91 i.e., where
the increments or decrements of the scale lengths of the
months preceding or following the lengths of the scales
of the solstitial months are 1/3 of II fingerbrclldth.
followed by 213 of a ringerbreadth. then by I
fingerbreadth. or of course the reverse relationships
where the increments or decrements are related as
3,2:l11 preceding or following the equinoctial months.
Another fragment of an outflow clock (Borchardt's
no. 9; see Fig. 111.27) yields a curious set of scale
lengths, ? II. II 112, 12. 13. 13 2/3. 14. where 12
fingerbreadths give the length of the equinoctial scales.
The second half of the set of scales shows the 3,2,1
relalionship noted in clocks 2-4. DUI the top part of
the scale deviates from it. and I see no ready
explanation for this deviation, for one would expect it
to show the same 3,2:1 relationship as the bottom half.
There is always the possibility that the clock was
carelessly constructed or that it was simply made to
serve as part of the tomb furniture for some worthy
deceased and hence was not actually meant to be used
in this world.
In discussing the monthly hour scales I have paid
little attention to the distance between the hourly
marks. except to note that certainly in the early
outflow clocks and probably in all Egyptian water
-74-
clocks the assumption was that they were to be equally

spaced. that is, that the water level sank equal distances
in equal times. However this was not a fact. the
segment of the vessel bearing the scales being a
truncated cone that could only produce an
approximated even rIow instead of being the segment
of a paraboloid which would yield steady flow.
Presumably the users of such a clock must have
realized Ihllt the flow was not uniform. Indeed there is
evidence that points in that dircC[ion from a papyrus of
the third century A.D. which sought to determine the
volumetric discharge of water between the various
hourly marks. 9Z The text is corrupt and includes some
wrong numbersl it lakes 1C as 3 and the basic formula it
uses for the volume of a frustum of a cone is incorrect.
The pertinent part of this determination begins in
Grenfell and Hunt's text in line 31t
They give the number(sl for the
construction of horologes as follows, making
the (diameter of thel upper part of the
frustum 24 fingerbreadths. hhat of the basel
12 fingerbreadths, and the height 18
fingerbreadths (see Fig. 111.28).
The volume of each segment of the frustum having a
height of one fingerbreadth is determined, according to
the papyrus, by finding the mean sectional area (j.e., 2
x T x 1/4 x 21fT· 1f?, with the value of ,.. taken as 3)
and multiplying this by the height (j.e., I fingerbreadth).
This involves the incorrect assumption that that volume
of the frustum is equal to a cylindrical segment whose
radius is equal to that of [he mean area of the segment
of the frustum. Since the over-all height of the
frustum is 18 fingerbreadths. the volume would be the
sum of all the segments of I frustum height. It is not at
-75-
all cleaT how this incorrect volumetric analysis was to

be used. but possibly the clock makers thought they
could adjust the hour markers to produce what they
thought were equal quantities being discharged in equal
times, though of course that would not be completely
accurate so long as they were using the wrong Cormula
for the volume of a frustum of a conc.
But whatever the explanation of this curious
fragment, it certainly seems that in Pharaonic Egypt the
hour-watchers thought the hour markings ought to be
equidistant and that the shape of their pot produced
this uniform flow. As we have said earlier. precisely
equal flow (other things being equal) could only have
been produced by 8 vessel with 8 paraboloidal surface
(see the parabolic section of such a paraboloid in Fig.
111.291 However. as further shown in Fig. 111.29. a
vessel with a slope of 2,9 and thus an angle of
inclination of about 770 measured between the pot wall
and the base extended (see the left side of the rigure)
would produce a very good approximation. Less
satisfactory is a pot with a slope of 1,3 <like that given
in the Oxyrhynchus papyrus; see the right side of Fig.
111.29). And even less satishctory is the angle of
inclination of the Karnak clock, namely 700 25'. The
angles determinable for the other clocks whose
fragments we possess are of still lesser inclination to
the base extended than that of the Karnak c1ock.93
Borchardt completes his discussion of hourly scales
by showing quantitatively that neither the Karnak clock
nor the clock discussed in the Oxyrhynchus papyrus
will produce equal hours. This is clear from the table I
have reproduced here as Fig. 111.30 and the linear scales
of Fig. 111.31. The question of comparative hour lengths
and their applicability to the possible day and/or night
-76-
periods leads to his principal conclusion thai the Karnak

clock measures the hours of what he calls the
"astronomical night: i.e., the period from the beginning
of darkness to its end, with the morning and evening
twilights excluded,94
I call the rcader's attention to one last point
concerning the various outnow clocks discussed by
Borchardt. namely their dating. where possible. by the
position of the shortest and longest monthly scales in
the civil calendar Hor 8 useful table showing where the
seasons fall in the civil year at dirrerent Julian dates
see Fig. 111.37).95 There is no particular need to review
that discussion here. since I have given the specific
dates as we have mentioned the various outflow c10clcs
and have morc than once indicated how they were
found. most particularly when J described the probable
use of the Ebers calendar for the construction and use
of water clocks.
Inflow Water Clocks

The only Egyptian water clock of the inflow type
is that found by 'G. Maspero at Edfu in 1901 (see Fig.
111.23), though a number of small models. which were
probably used as votive offerings. are extant (Fig.
111.32).96
The Edfu clock is a cylindrical vessel of limestone
unadorned on the exterior except for a small
representation in relief of Thoth as the cynocephalic
baboon placed immediately above an aperture for
draining the water in the clock.97 On the interior
surface of the vessel is a poorly drawn grid embracing
vertical monthly hour scales intersected by horizontal
sloping lines at the hour points (see Fig. 111.33). As an
inflow vessel. the Edfu clock performs its time-telling
-77-
duty by having waler flow into the vessel from a

reservoir and reading its level against one of tbe
monthly hour scales. The great advantage that such a
clock has over the oulnaw type is tbat if the reservoir
is kept full constantly the head of inflowing water will
remain the same and hence in theory the water flowing
into the clock will rise at 8 constant Tate. Before
slarting the clock at the beginning of the night. tbe
waler is drained from the vessel to the lowest marlc on
the hourly scale appropriate for the month that it is
being used. It would. of course. be very difficult to
read the watcr level against the scales on the interior
of the cylindrical vessel. Hence the bour-watcher
would surely benefit from some kind of movable float
with an extended pointer which could be set at the
proper month. This pointer could then indicate the
hour on the exterior surface of a column that possessed
a duplicate hourly grid and was attached to the top of
the vesseJ.98 However no such attachments have been
found.
Turning to the grid with its vertical monthly scales
and intersecting sloping lines. we must first note that it
preserves still the Old Egyptian system developed by
Amenemhet and appearing in the Karnak clock of
having 12-fingerbreadth and 14-fingerbreadth scale
lengths for the months containing the summer and
winter solstices. and 13-fingerbreadth scales for the
months with the equinoxes. But a significant change in
the distribution of the vertical scales, that is, in the
horizontal distances between the scales, is signalled by
the interior diagram on the &lfu clock, where the
distances vary erroneously as 3:2:1 1 1,213; 3:2,1; 1:2,3,
beginning from a 14-fingerbreadth scale, though
according to the phenomenon of slower increments and
-78-
decrements before and after the solstitial months the

distances ought to hllve been 1,2:3; 3,2,1; 1:2:3: 3,2,1. It
was Alexander Pogo who first explained the
significance of this feature of distance displacement.
suggested its possible origin in some earlier prismatic'
inflow clock. and revclIled the error made by the &Uu
clock maker in applying the 3,2,1 relationship to the
distances from the solstitial scales,99
[Following the discovery of the 12- and
14·fingerbreadth scales and the concurrent
assumption that the monthly scales varied
uniformly in length from month to month by
1/3 of a CingerbrclIdth,] The next step in the
development of waler clocks -made.
apparently. by a designer of an inFlow c1ock-
was the correct observation that the length of
the nights changes slowly during the month
immediately preceding and during the month
immediately following the shortest or the
longest night; the advanced design of sun dials
made such an observation possible at a
relatively early date.
Here is a simple method for incorporating
this observation into the scales of a prismatic
inflow clock with a square basel "see figure I
[my fig. 111.34], The corner scales of the
prism ADGJ are 14. 13. 12. and 13
fjngerbreadths, respectively; each corner scale
is divided into 12 equal parts, and the
corresponding points [of division] are joined
by inclined straight lines which form 'stripes'
on all four wallsl additional vertical lines are
then drawn through the centers, C and K. and
E and J. of the sides of the prism. and form
-79'
I
I
the 13 1/2 and the 12 112 scales. respectivelYI

finally. the distances AC and AI(. and GE and
01. are divided into thirds. and the vertical
lines Band L. and F and H, form the 13 5/6
and 12 1/6 scales respectively.
Here is another method for drawing the
diagram of Figure I [my Fig. 111.34] on the
inside of 8 square inflow clock. The corner
scales of the prism CEIK are 13 1/2. 12 112. 12
1/2 and 13 112 fingerbreadths. respectivelYl the
equinoxial m scales. 0 and J. occupy the
middle of the side walls of the prism, while
the solstitial scales, A and G. with their
satellites Band L. and F and H. are drawn on
the inside of the front and bad: walls.
respectivelYI the inclined lines form 'chevrons'
on the front and back walls, and ordinary
'stripes" on the side walls....
When prismatic inflow clocks were
replaced by cylindrical ones, the diagram
remained unchanged. and the distances
between the scales, 1:2/3 1 3121h 1:213; 3;2;1,
preserved on the cylindrical diagram or the
Edru clock. point. it seems to me. clearly to
the prismatic prototypes or the cylindrical
inrJow clocks. The Edru diagram, Figure 2
bee my Figs. 11I.33 and 111.35], is a somewhat
careless and not too intelligent copy or a
diagram or the type reproduced in Figure I
[my Fig. 11I.34]; it was copied either directly
rrom an obsolete prismatic inrJow clock, or.
more likely. hom II cylindrical inrJow clock
which was 'out-or-date: The ract that the
Edru diagram begins the 1:2:3 count hom the
-8D-
autumnal-equinox scale placed over the
cynocephalos orifice seems to indicate that the
maker of the clock was trying to copy II
diagram which originated on the inside of a
prismatic clock of the 'chevron' type. CEIK.
where the cynocephalos orifice was placed. for
reasons of symmetry. below a solstitial scale:
the engraver of the Edfu diagram did not
realize that the h2:3 count of distances
between the scales must begin at a solstitial
scale. and that the position of the
cynocephalos orifice is irrelevant: he saw that
the diagram he was trying to copy started the
1=2:3 count from the orifice. and he faithfully
reproduced this irrelevant detail. The lengths
of the Edfu scales increase and decrease by
the traditional amounts, 1112, 1/6. and 1/4 of
the difference between the longest and the
shortest night. but the wrong starting point of
the 1,2,3 count leads to slow changes of the
lengths of the scales ncar the cquinoxes. and
to rapid changcs ncar thc solsticcs-an
absurdity which failed to shock the maker of
the Edru clock..Iwhol did not understand the
meaning or the 1,2,3 diagram.
While there is no archaeological evidence to
support Pogo's ingenious reconstruction other than the
existence or models or prismatic inrJow clocks (which
or course had no interior grids), it seems to me
inherently probable. However. we might well temper
somewhat his judgment that the inflow clock with its
interior grid was a truly original discovery and thus an
outstanding event in the history or science (sce note
103 below). for he does not mention the fact that the
makers of some extant outflo.... clocks (earlier than the

Edfu clod) had already altered the lengths of their
monthly scales so that the increments follo....ed a h213
relationship. thus making a start to....ard recognizing the
observation that the lengths of the nights changed most
slowly before and after the solstices and more rapidly
before and after the equinoxes.
Furthermore. Pogo is less than successful in trying
to push the invention of the prismatic inflow clock and
interior diagram back to Dynasty 18. While recognizing
that the models of the inflo.... clocks are more than a
millennium later than the Karnak outflow clock. he
·argues the possible earlier existence of a prismatic
inflow clock of the ·chevrons· type from a relief in the
Luxor temple dated to the time of Amenhotep III <See
Fig. 111.36).100 In this relief. Amenhotep III is shown
presenting a votive oHering (lC) to the goddess Mut
which Pogo says "might be described as a prismatic lin
receptacle. with a squatting cynocephalos attached to
its front wall. presented on a nb [sign. """"1" Pogo finds
it ·difficult to escape the impression that the object on
the nb sign represents a prismalic outflow clock:
Looking closely at the lin prism (or so he calls it). t1'. he
sees the V-shaped notch on top of it as a possible
"allusion to Ihe 'chevrons' characterizing a CEIK
[prismatic] diagrami" and if Ihis is so. then we have
further confirmation of his belief that the oHering is a
prismatic clock of the ·chevrons" Iypc. the chevrons
being on the side walls rather than on the front and
bad walls. His Inlerpretation of the nolch at the top
of the lin sign in this manner seems to me entirely
fanciful since the V-nolch is strictly a part of the sign
from its earliest use in the Old Kingdom. lOl which
-82-
certainly predates any known use or record of inflow

clocks. The ....hole votive offering in the Luxor relief is
rderrcd to as a !bt in the column of glyphs under the
offering. This column is partly worn away but what
remains may be rendered, •...the giving of the ~bt upon
the land so that he CThoth?> may cause the giving of
life forever: Hence we cannot with any assurance
describe the clock (if that is what it be)102 as II
prismatic influx clock. at Icast not of the 'chevrons'
variety. All we really know from this offering scene is
that it involves "the giving of life forever: Since the
(In sign was used often from Dynasty 18 onward to
mean an extended or infinite time period,103 we could
perhaps say that the three signs comprising the votive
offering, m. meant "Thoth. Lord of extended time:
This rendering would underline its significance in the
presentation scene as concerned with "the giving of life
forever:
Because of the clear link between the monthly
scales of the out now and inflow clocks. it seems
probable that in Egypt, at least during the Pharaonic
period we are concerned with, the innow clocks were
primarily used to tell the time of Borchardt's
"astronomical night: i.e., the period of total darkness
without the morning and evening twilights. However
when the innow clocks spread from Egypt [and no
doubt especially from Alexandria] through the classical
world,104 changes and improvements were made so that
this was no longer true. Sioley has described briefly
these later water clocks and mentions their use in
public places.1OS
Shadow Clocks
The determination of hours in the daytime during
-83-
the Pharaonic period ....as based on the sun's motion and

its projection of shadows. Hence the basis of marking
daytime bours is quite distinct hom that of marking the
nighttime days. which. we have seen. was rooted in the
risings or meridian transits of stars and in the outflow
or inflow of watcr from water clocks. The only
assumed similarity is that the length of daylight
(initially including morning and evening twilights) was
assumed to be divided into 12 hours just as was the
length of nighttime.
There are two main types of sun-clocks. The first
is an instrument to measure the length of shadows
projected upon it and is usually simply called a shadow
clock: the second measures the hours by the changing
direction of a shadow projected during the daily course
of the sun's motion; they are of course known as
sundials. Let us first examine shadow clocks.
Telling time by shadows appears to be alluded to in
a well-known passage from The Prophecies of Neferti
quoted above in Volume One (page 402) to the effect
that during the time of unhappy conditions preceding
the Middle Kingdom "Re separates himself from men; he
shines that the hour [of dawn] may be told, but no one
knows when noon occurs, for no one can discern [or
measure?] his shadow:106 The use of "measure· or
"number" instead of "discern" by some authors may
result from their desire to push the history of the
shadow clock at least to the Middle Kingdom.
The first (and indeed the only) Egyptian technical
description of an ancient Egyptian shadow clock is
found in an inscription in the cenotaph of $eti I (ca.
1306-1290). This is given in translation below
(Document 111.16), with the text transcribed in Fig.
111.38. The shadow clock (SI'r) so described (see Fig.
"84-
111.39) consists of a horizontal board (mrtwt) 5 palms (-

20 fingcrbreadths • ca. IS") in length with a vertical
head (tp) aHixed at one end. Attached to the vertical
head is a crossbar (mr!Jt)07 2 fingerhreadths in height
whose vertical plane is perpendicular to the horizontal
board. Marked ("branded") on the board are signs that
measure the hours. JOB The first is 30 units from the
vertical head, the second 12 units neater, the third 9
ncarer, the fourth 6 nearer, and the distance from the
fourth sign to the vertical head is 3 units. the signs
being placed to measure four "hours: The horizontal
board was first aligned to the east with the head
pointing in that direction. Accordingly. the crossbar
then pointed toward the north and the horizontal board
was probably leveled by a plumb line, as is evident
from the shadow clocks that have come down to us,
though no such procedure is mentioned in the text.
When the shadow falls on the first mark it begins to
measure the first of four hours before noon. When it
reaches the second mark the first hour is completed, and
so on until the shadow length on the horizontal board
comes to an end and noon has been reached. Then, as
the text tells us, the shadow clock is reversed so that
the head points to the west, and the lengthening
shadow falling on the marks indicates the passage of
the next four hours. Now it should be noticed that
only eight hours have been marked off for measurement
by the clock. But the text goes on to tell us that °2
hours have passed in the morning before the sun shines
[on the shadow clock] and IWO hours [will] pass before
the sun enters [the Duat or realm of darkness] to fix
the place of the hours of night [j.e., before darkness
arrives and the hours of the night begin]." Hence it is
clear that the author of Ihis piece assumed that the
-65-
daylight (with the twilights) was divided into( 12 hours.

The selection of t ....elve hours seems to have been made
in imitation of the division of the night into 12 hours
necessitated by the civil calendar with its division inlo
36 decades. According to Neugebauer and Parker.109
the "most plausible assumption [concerning the four
extra hours not measured by the clock] is that sunrise
and sunset, .... hich are observable phenomena. mark the
dividing points between the two extra hours before and
after the central 8: This leads to their belief that this
account describes a clock that is much earlier than the
time of the cenotaph of Seli I, for the first extant
shadow clock. preserved from the time of Tuthmosis III
(ca. 1479-1425 B.C.l, though very much like the shadow
clock here described. has 5 instead of 4 hour marks,
thus supposing a 10 hour clock with 2 extra hours and
thus a day of 1.10.1 hours. the 12-hour day perhaps
beginning with sunrise and ending with sunset. Or it
could be that the 5 marks simply indicated a decimal
division of the day with two additional intervals to
connect it to the 12-hour nighttime.
Now it is evident that the maries on the Seti clock
(those accompanied by Roman numerals on straight line
E in Fig. 11I.40) indicate the central hours at the time of
the equinoxes, while the marks with Arabic numerals
marlc the central hours in the 5-mark clock from the
time of Tuthmosis III. But even if these marks
correctly measured equal hours at the equinoxes (which
they did not), they would not have accurately marked
the lengths of those hours at other times of the year in
view of the changing declination of the sun throughout
the year. This is evident again from Fig. 1IJ.40, where
the marks of the hours fall on points on the hyperbolic
curves Wand S representing in each case the shadow
-86-
end's movement at the winter solstice (W) and the

summer solstice (5), and the shadow end passing
through a point on the curve marking a given hour at
either of those times of the year would not fall on the
horizontal board. Le., on line E. in such a way that it
would pass through the corresponding hour mark on E.
Hence the original hour markings would not serve to
mark the hours correctly at the summer and winter
solstices. However, since. the Egyptians did not appear
10 be exacting about the equality of the hour lengths.
as we have already shown to be the case for the hours
measured by water clocks. they probably used the early
shadow clocks during the whole year without concern.
But see Document 111.16, notc 2, for Bruins' clever
suggestion that in Seti's c10cle a properly thin strip was
added to or removed from the crossbar to change its
height and thus to accommodate the c10cle to the
changing declination of the sun at the times of the
winter solstice, the equinoxes, and the summer solstice.
Still, it should be remarked that none of the extant
shadow c10cles included a crossbar or thin strips to be
added to that bar but only an end block with a groove
for hanging a plummet line on it. In fact it is not clear
from the tcxt in the cenotaph of Seti I that the shadow
clock included a cross bar mounted on the vcrtical end
piece as is usually depicted (see Fig. 1I1.39). For the
name mrf1t translated in Document 111.16. line 10. as
"crossbar" could just as well refer to the vertical block
alone. especially if the block contained a groove for the
plummet line. Furthermore, the illustration of the
shadow clock which accompanies the text (see Fig.
111.38) does not reveal any crossbar distinct from the
vertical head; it merely shows the whole instrument in
profile as an L with a short vertical upright. Of course.
-87-
if the vertical upright and a8 attached crossbar 'lVere

the 'same thickness. the crossbar might well not show in
a profile view. But the principal rcason for assuming
that II crossbar (much longer tban tbe vertical block)
was attached to the vertical block was that the shadow
clock was apparently used throughout the year. And so
in all probability 8 longer crossbar was needed to
ensure that part of the line terminating the shadow
always fell on the ratber narrow horizontal board
during the four measured hours at all times of the ycar.
Now 1 am prepared to examine actual shadow
clocks that have been preserved from ancient Egypt. IIO
I have already brierly mentioned the earliest one dating
from tbe reign of Tuthmosis III (see Fig. 11I.41. West
Berlin Mus. Nr, 19744. with separate eodpiece above it
from the reign of Nebmaatre, i.e., Amenhotep Ill, No.
14573). A clock from Sais (Fig. 11I.41, West Berlin Mus.
Nr. 19743) dates from 500 or more years later: it is
quite similar to the Tuthmosis clock but larger and more
intact. Following Borchardt's treatment,lll I shall
concentrate on the latter.
The base is a ruler of about 30 em. in length. At
one end is a small vertical, rectangular block mounted
perpendicularly on the base ruler. The vertical block
has been pierced through its length by a narrow boring
and a groove has been scratched out on one of the
pierced surfaces, the boring and the groove being used
to mount and guide a plummet line for the purpose of
ensuring that the shadow clock lies on a level surface.
Borchardt surmises that the plummet would not be
peeded very orten since a known level surface such as
the roof of a temple would have been readily available.
On the upper surface of the base ruler we find five
circular markings that indicate the passage of the hours
-88-
(no circle is found at the end of the sixth hour since

the vertical upright serves instead), the intervlIls
between the hours decreasing in lincar fashion from the
mark of the nrst hour toward the end of the rule up to
the upright which marks the end of the sixth hour (j.e.•
the intervals are diminished by decreasing multiples of
the distance from the vertical upright to' the marlc
ending the fifth hour). Accompanying these markings
are the names of the first six hours (that is, the names
of their protective goddesses), the first hour lasting
until before the sun projects a shadow on the rule
which reaches the first marking. The names follow (see
Fig. 111.42 for their hieroglyphic signs), 1st Hour-The
Rising One. 2nd Hour-The Introductory (or Guiding)
One. 3rd Hour-The Protector of Her Lord. 4th
Hour-The Secret One. 5th Hour-The Flaming One, 6th
Hour-Thc Standing One,lll
It is obvious by the description I have just given
that this shadow clock is used in a manner closely
similar to that of the Seti clock we have already
discussed. The base ruler lies in the east-west direction
and the upright block with an assumed linear crossbar
mounted on i! accordingly lies in the north-south
direction. Borchardt assumes such a crossbar mounted
on the vertical blod. for. as I have said before. the
vertical block by itself is so narrow that the shadow
would at all hours fall on the hour maries only at the
periods of the equinoxes. At the periods of the
solstices it would fall widely north and south of the
narrow baseboard and therefore not on the hour marks.
Therefore. Borchardt suggests that perhaps a cross bar
ought to have been mounted on the vcrtical board so
that in all seasons some piece of the shadow of the
crossbar will fall on the graduatcd baseboard (see
-89-
Borchardt's drawing given below as Fig, 111.43),

Borchardt further points out thai the graduation of
hours on the baseboard does not produce equal hours
even at the times of the equinoxes. This is immediately
evident from the fact that the increments of the
intervals between the hour marks decrease by linearly
decreasing multiples of 8 constant distance. This is an
approximation much too simple to represent the
complicated changes in shadow length (and their
intercepts on the baseboard) produced by the angular
changes in the sun's height in the course of the sun's
apparent daily motion. This is not surprising in view of
the complete: absence of simple trigonometric
calculations among the ancient Egyptians. for in order
to determine shadow lengths and above all their linear
intercepts on the ruler with the hourly graduations at
the times of the equinoxes we would need to know the
trigonometric functions of 150, 300, 450, etc. Their
values could not be exactly expressed using a simple
arithmetical series of numbers. We might, however.
Ienow more about the ancient theory lying behind the
graduation of the later shadow clods if we had an
intact version rather than charred remains of a papyrus
found in Tanis and dating from about the first century
A.D. (see Fig. 111.44). They depict a shado..... clock, ....ith
the numbers of the hours marked above the rule and
under them are names of the protective gods of the
hours which are inscribed on the rule. The vertical end
piece which casts the shado.... is not sho.... n but the rays
proceeding from it to the hour marks are shown. the
heavy lines picture those portions of the rays present
on the fragments.
Returning to the Berlin shado..... clock analyzed by
Borchardt. we sho..... In Fig. 111.45 the results on his
-90-
investigation of the approximate deviation of its hour

graduations from equal hours. Moreover. in Fig. 111.46
is shown 8 system for using II 5-edged beveled crossbar
(Borchardt's 'sacred cubit') in three dirrerent positions
to produce shadows which mark equal hours, At
sunrise the crossbar would be in position I. The first
hour would be completed when the shadow's end fell on
the first hour mark. The bar would then be changed 10
position 2. and the second hour would be completed
when the shadow fell on the second hour mark. At this
time the crossbar would be shifted to position 3. For
the fourth hour the bar would be shifted back to
position 2. for the fifth back 10 position I, and fot the
sixth hour it is left in position I or shifted to position
3. The diagram of Fig, 111.46 indicates how these
positional changes produce the appropriate shadows,
One cannot emphasize too strongly, however, that there
is no actual evidence that such a beveled crossbar was
used, and I suspect that the apparent indirrerence of the
ancient Egyptians to the exact divisions into equal
hours of any of their docks makes their use of this
device unlikely. Indeed. Borchardt himself does not
insist that such a crossbar was used, though he believes
that its use was a possibility.1l3
Before leaving our treatment of shadow docks with
their base rules oriented in an east-west direction.
mention must be made of a model of a sun clock
prepared in sort white limestone and now in the Cairo
Museum (No. 33401).114 This is later than the Berlin
shadow clocks we have already described. It represents
three dirrerent kinds of shadow clocks (see Figs. 111.47
and 111.48>. the first with the scale on a flat, level
surface like the Berlin shadow clocks. the second in
which the shadow falls on stepped surfaces. lind the
-91-
third in which the shadow falls on an inclined plane.

The clocks with 8 stepped or inclined surface fot the
reception of the shadows makes possible the reduction
in the length of the shadow-receiving surface that was
necessary when using a fat. level baseboard to measure
the early and late hours. Notice that in representing
the clock with the level surface. the vertical shadow
producing block bas no crossbar attached. But
presumably no such baT was needed (at least from the
end of the second through the sixth hours) because the
graduated shadow-receiving surface is not narrow as in
the case of the Berlin shadow clocks but is instead
wide. so that some portion of the shadow's end~line
would probably fall on the surface regardless of the
season of the year if we had a real clock rather than
merely an illustrative model. In view of the fact that
we have only this model and we do not have any idea
as to the latitude for which each of its three clocks was
designed for and that we are not sure how faith£ully
the model re£lects the proportions of the real clocks it
represents. any extended analysis of real or potential
hourly graduations like that given by Borchardt can
hardly be more than educated guesswork iJlustrating the
way in which the clocks functioned. However, after
noting that the hourly graduations on the level-surface
and stepped-surface clocks. even if they were correct
for the equinoxes, would be incorrect for any other
days of the year. Borchardt suggests that the
inclined-plane clock (without graduations of any kind
on the model) could h8ve had scales that were to be
mounted and usable for other days of the year as well
(see Fig, 111.49).
It should be realized that the particular
inclined-plane clock of the Cairo model like the other
-92-
two clocks represented on it was also to be fixed in an

cast-west orientation so that like them the hourly
graduations represented the east-west intercepts of the
shadow produced by the sun falling on the vertical
edges. However there are evidences of another type of
inclined-plane clock that was meant to be movable and
was not maintained in an cast-west orientation (see Fig.
111.50 for several examplcsl 115 At whatever time the
clock was to be read. its vertical block was pointed
directly at the sun so that the sun's rays were always
perpendicular to the shadow producing edge of the
vertical block with the consequence that the shadow of
the block fell completely and exclusively on the
inclined plane (see Fig. 111.51>, and indeed the lack of
any side shadow would be an indication that the
sun-pointing clock was properly oriented, In view of
the orientation of the clock directly toward the sun, the
hourly scales on the inclined surface measured the
direct length of the shadows rather than the intercepts
of those shadows on an east-west oriented rule as was
the case in the earlier shadow clocks we have been
describing.
The first of these sun-pointing shadow clocks <the
one from Qantara, Fig, 111.50, J) is datable to about 320
B.c, (plus or minus 60 years). It is shaped like the
halves of the inclined-plane clock represented on the
Cairo model. On the Icft vertical block from which the
inclined plane begins to rise there is a groove for a
plumb line used to level the clock, On the inclined
surface are seven scales of varying length, each of
which has six hour points, Beyond the incline is 8
right hand block on the lOp surface of which are the
names of the Egyptian months given in Greek so that
the longest scale (on which the midday shadow would
-93-
be the shortest> is beneath the month f'(!Phlarmuhhil,

the second scale beneath the months P(!Ph)amc(nothl
and Pach[onl. the third beneath the months Mechi[r]
and Pay[nil. the fourth beneath the months Tybi and
Epe£iphl, the fifth beneath the months ChoiIlU:) and
MelsoreJ, the sixth beneath the months Hathyr and
Thoyt. and the lasl scale. i.e.. the shortest one (which
has the longest midday shadow), beneath the month
Phaophi. The fact that the longest scale (which would
be the scale to be used at the summer solstice) is
assigned to the month Pharmuthi of the 365-day civil
year indicates that the dock scales were designed for
the year of 320 B.C. (with a leeway of 60 years before
and after that date), as I have said.
The only other of these sun-pointing clocks
complete enough to date even roughly is that of Paris
clock shown in Fig. 111.50. no. 2. The Egyptian month
names on this clock are also written in Greek. The
longest scale there is for Payni and the shortest for
Choiak. Since the summer solstice fell at the end of the
10th month (Payni> in the fixed Alexandrine year. it is
evident that the Alexandrine calendar underlies the
scales of the Paris clock. Hence all we can say about
the date of that clock is that its scales must date to the
Roman period sometime after Augustus' introduction of
the Alexandrine calendar into Egypt in 25 llC.
One last point can be made concerning these
sun~pointing shadow clocks. It evident that the
Egyptians constructed the various scales so that
straight lines could connect the corresponding hour
points. e.g.• a straight line would connect the points
designating the first hour in all seven scales. as seems
to be confirmed in a fragment in the Petrie Museum at
University College. London (see Fig. 111.50. no. 3 and
'94-
Fig. 111.521 In actuality this ought not be the case, 8S

is indicated in the dotted lines drawn in Figs. 1II.53a
and 1II.53b through the theoretically correct points on
the scales of the Qantara and Paris sun-pointing clocks
(the first taking scales for the Qanlan clock assumed 10
be constructed for latitude 31° north and the second
scales for those of the Paris clock with an assumed
latitude of 2(0),116
We note finally that hieroglyphics representing
some of the shadow clocks we have discussed were
used in the Ptolemaic period either as delcrminatives
for the word "hour' or by themselves as ideograms for
that word (see fig, 111.54),
Egyptian Sundials
The sun clocks described in the preceding section
measured shadow lengths projected by the sun orr the
edge of a vertical upright and they depended on the
altitude of the sun. The other major kind of sun clock
found in ancient Egypt was the vertical-plane sundial.
which. of course. measured the hours by the changing
angular direction of the shadow of a gnomon produced
by the sun during its daily motion. I shall consider
here two examples which are clearly of Egyptian origin.
The fragment of a third dial found at Dendera. which
includes only the radial lines marking the six hours
arter noon (Fig. 111.558). seems to be completely similar
to the rirst two examples and so I shall not examine it.
The rirst of the Egyptian sundials is 8 small ivory
disk found at Gezer in Palestine l11 (see Fig. 1II.55bl It
dates from the reign of Merenptah (ca. 1224-1214 o.c.).
whose cartouche it bears on the back of the disk and
who is also pictured there making an offering to Tholh.
Above the scene are the cartouches of the king. On the
-95-
rrant of the disk we see the remains of 10 radiating

lines ((rom a set that almost certainly originally
contained 13 such lines. with the angles between t .... o
successive lines each being about 15°). The.diamcter of
the dial is about 2 1/4 inches. with its centcr being
evident from II small circle on the back. One can sec
from the edgewise view in Fig. m.5Sb that the disk ".s
bored through its entire length so that the dial could be
hung up by a chord passing through the bored channel.
The inscriptions of figures and radial lines ....ere
originally filled with green enamel. some of which
remains. Presumably. when intact, a style (j.e.. a
gnomon) must have been set into a hole in the centcr
of the diameter of the semicircular dial face for the
purpose of casting the angularly changing shadow.
The second dial (i.e., its face) is just like the first
one but more complete (Fig. 11I.56>. It is made of a
green fayence that suggested to Borchardt that it ....as
probably fashioned in the Greco-Roman period. na The
dial was acquired at Luxor and is a part of the West
Berlin Museum collection (No. 20322). Since the dial is
intact (except for the gnomon) it contains 13 lines
radiating from the center hole. Tbe first and last line
together make a straight line. to which the 7th line
(marking midday) is almost perpendicular. The dial is
without any decoration. As one can see in the drawing
of Fig. 11I57. left. the dial was hung vertically by a
cord supported from two eyes in the rear rather tban
from a cord passing through a channel bored lengthwise
through the top of the clod. as in the case of
Merenptah's sundial. A now missing gnomon for
projc<:ting the movable shadow was inserted in the
center hole and so stood perpendicular to the dial face.
Tbere remains above the center hole a rectangular
-96-
recess. That recess apparently allowed for hanging a

plumb line necessary for positioning the dial correctly.
It is clear from the way the sundials were made that
for their use the straight line consisting of the first and
last radial lines had to be level and that the dial surface
had to bang in an east-west plane. The plumb line
accomplished the first of these conditions when that
line was parallel to the 7th radial line. The second
condition was met by using it in a place where the
cast-west line had been previously established.
This clock like all the other clocks we have
examined did not mcasure equal hours at lIny time of
the year. At the time of the equinoxes, the 1st and 12th
hours were each too long. as were the 2nd and the 11th;
the 3rd and the 10th hours, as well as the 4th and 9th,
were each approximately one daylight hour long; and
the 5th and the 8th hours. as well as the 6th and the
7th, were each too short. This is clear from fig, 11I.57.
right, which indicates the results of Borchardfs
investigation of the Berlin clock for an assumed latitude
of 25.50. In that figure the bold radial hour lines are
those inscribed on the clock and thus end at the
circumference of the disk. Near the end line of each
hour (except the 6th and the 12th) is a set of three
extended lines: two broken lines and one that is
continuous. The broken line at the greatest angle
indicates the proper marker for the hour at the winter
solstice, the middle continuous line represents the
desired marker for that hour at the time of the
equinoxes, and the broken "line at the least: angle is the
suggested marker for that hour at the time of the
summer solstice. This figure then not only shows us
that the radial lines did not mark equally long hours at
the time of the equinoxes. but. as expected. they also
-91-
I
I
failed to do so at the solstitial times.

It is evident from my treatment of the various
kinds of Egyptian clocks (star clocks. water clocks.
shadow clocks. and sundials) that there is great
diversity in their accuracy and their methods of telling
time. Indeed these various kinds of clocks share
virtually nothing in their theoretical underpinnings,
except of course the division into day and night hours
and the seasonal variability of hour lengths (which is
ignored in all but the water clocks aod the
inclined-plane. sun-pointing clock), It seems apparent
that hom at least the New Kingdom onward all of the
types of clocks we have described were in use despite
their diversity. This is one more instance of the fact
that emerged in the first volume in my treatment or
cosmogony and cosmologYl the Egyptians ohen
accepted essentially contradictory and mutually
inconsistent practices. doctrines and modes or thought
without great concern.
Traces or a 24-hour day with Equal Hours

As I have more than once stressed. the division or
hours generally accepted by the ancient Egyptians and
rerlected in the development or their clocks consisted or
seasonally variable hours. with separate systems or 12
night hours and 12 day hours. the division or daylight
into 12 hours perhaps arising by analogy rrom the 12
hour division or the nighttime required by the structure
or the civil calendar. But I have also mentioned. in
describing Amenemhet's invention or his variable scale
water clock in the reign or Amenhotep I, Document
111.15, note 4. that the ract that the length or the
longest hour of a winter night was compared to that of
the shortest summer night as 14:12 perhaps hinted at a
-98-
later theoretical discussion of the possible division,

month by month. of the whole period of day and night
into 24 equal hours. It is this latcr division that J now
mention. Two documents (and the fragment of a third)
refer to such divisions. The first (my Document 111.7
but only given here in the body of Chapter Three and
not in the section of documents) is II table added on
folio XIV verso of Cairo Papyrus No. 86637. a
Ramesside papyrus of about the twelfth century 8.C,I19
The date of the hour table given in the text is earlier,
falling between about 1400 B.C. and 1280 8.C.12O Its
hieratic text is shown in Fig. 1II.58a and ils
hieroglyphic transcription is given in Fig. 1Il.5gb. The
table may be rendered as [01l0W5,121
I. I I/Jll hours (wnw/) of daylight (hrw). 16;
hours of nighttime (grf,), 8.
2. II I/)/l hours of daylight. 14; hours of
nighttime. 10. Phaophi.
3. III Ih/l hours of daylight. 12; hours of
nighttime. 12. Athyr.
4. IV I/)/l hours of daylight. 10; hours of
nighttime. 14. Khoiak.
5. I pr/l hours of daylight. 8; [hours of]
nighttime, 16. Tybi.
6. II pr/l hours of daylight. 61 hours of
nighttime, 1(8l Mekhir.
7. (Ill) pr/l hours of daylight. 8; hours of
nightlime. 1(6). Phamenolh.
8. (IV) (p)r/l hours of daylight. (l0l; (hours
of nighttime. 14). Pharmuth.
9. (I 5m)Wl (hours) of daylight. 12; (hours of
nighttime, 12). Pakhons.
10. II 5mIV: hours of daylight. 12 (sio, but
should be 14); hours of nighttime. (Mmk; but
-99-
should be 10>. Payni.

11. III S'm\Vl hours of daylight. 16; hours of
nighttime. 8. Epeiph.
12. IV S'mw: [hours on daylight. 18; hours of
nighttime. 6. Wep·rcnpet.
Aside from implying the division of the whole day
into 24 hours. this calendar also appears to assume that
equal hours are being used. We have seen that the
Egyptians were unable in their clock to produce truly
equal hours. But they perhaps produced the ina~curate
approximations (indeed their preposterous data) 'by the
consistent use of II water clock. during night lind
daytime aher the division into 12 of both parts had
been already established: 122 Most puzzling is the fact
that the table reports values that seem to yield 311 8S
the ratio of the longest daylight to the shortest. 8 ratio
that Clerc declared to be a fantasy. As Neugebauer and
Parker note,123 this ratio corresponds to no locality in
Egypt "if 'day' would mean the interval from sunrise to
sunset." However they suggest that if the longest
daylight was merely the complement of the shortest
"night" of 12 decanal hours and the latter is worked out
in accordance with their analysis of the elapsed limes
of decanal hours, this would produce a shortest night of
about 6 hours in agreement with the table we have
been examining. However it still does not explain the
extreme ratio of 3~1. "particularly in view of the scales
in the ratio of 7 to 6 of the Karnale water clock which
malees the difference between winter and summer night
smaller than it should be: Further note should be taleen
of the table's technically incorrect assumption that the
length of daylight from month to month increases and
decreases in a linear fashion, which, as we have seen,
was 81so assumed in the scales of water clocks. The
-100-
constant monthly increment or decrement in our table is

2 hours. At any rate the fact that in none of their
clocks did the ancient Egyptians produce series of truly
equal hours. makes any modern assumption that they
used accurate equinoctial hours from month to month
over the whole year in this and the succeeding table
exceedingly unlikely.
A second table of the lengths of daylight and
nighttime at IS-day intervals (my Document 111.8,
presented only here in this chapter and not in the
section comprising documents) appears on a stone
plaque discovered in Tanis in 1947.124 Though the date
of the plaque is not certain. it probably dales from the
time of Necho II in Dynasty 26. 125 The incomplete
plaque and another fragment discovered in 1948 contain
four Texts. Text No. II is the table under
consideration. A drawing of the table and its
restoration by Clere is given in Fig. 111.59. Two
additional columns. numbers II and 12 (not shown in Fig.
111.59). contain a small part of a similar 24-hour table.
of which not enough remains for any useful analysis. 126
In translation. the table of Text No. II runs as
followsl 127
Knowledge {r/.J} of the quantity {C~w}} [j.e.,
the length] of the Daylight (mtrftJ) in
relationship to the Night {gr/J}.
I [j.e.. 1st Month of] Akhet, Day b
daylight, 10 1/4 hours (wnw/); nighttime, (13
3/4 hours)
(I Akhet, Day 15: daylight), 11; nighttime,
13
II Akhet. Day); daylight. 11 1/2: nighttime.
(12 1/2)
(II Akhet. [Day] 15: daylight. 12);
-101-
nighttime. 12
III Alchet. Day I, daylight .• 12 1/4;
nighttime. II 3/4
Oil Akhet. [Oayl IS, daylight....; nighttime.
... )
(IV) Alchet. Day II daylight. 13 3/4;
nighttime, 10 1/4
IV Akhet, <Day IS, daylight, ... 1 nighttime,
_. )
(1 Per)el, Day I: daylight. 14; nighttime. 10
I Perct. Day 1(5), (daylight•...; nighttime.
... )
(II Peret,> Day 11 daylight. 14; nighttime. 10
II Peret. Day 15, daCylight .... 1 nighttime•
... )
(III Peret. Day., daylight. 13;) nighttime. 11
III Perel, Day 15, daylight. 12 1/4 1/6:
nighttime. (II 113 114)
(IV Peret, Day I, daylight, ...; nighttime.
... )
(IV Peret. Day) 15, daylight. 12; nighttime.
12
I Shemu. Day I: (daylight, ...1 nighttime.... )
(J Shemu, Day 15: daylight. II 1/3 1/4;
nig)httime 12 1/4 116
II Shemu, Day I: daylight. 22 (sic);
nighttime. C. )
(J) Shcmu. Day 15: daylight•...; nighttime.
... )
(III Shemu. Day I, daylight•...; nighttime.
... )
(III Shemu. Day 15: daylight. 9 1/3;
nighttime 1)4 112 1/6
IV Shemu. Day I: daylight. L; nighttime.
-102-
... )
(IV Shemu. Day J5l daylight•...; nighttime.
...
)
Clere suggests that unlike the Ramesside table,
which is a schematized table where the values seem to
have been arbitrarily chosen, with their succeeding
increments and decrements being 8 fixed two hours
each month. the Tanis table despite its gross
incompleteness and inaccuracies has no trace of
schematization of the values. In elaboration he
continues: 128
One does not rind either regular progression or
parallelism in the sequences. The few
observable concordances (for example, the
same duration of 12 h. 25 m. for the daylight
one month before. and for the night one
month afler the autumnal equinox. 15.IV.prd
are certainly fortuitous-except. doubtless. for
the duration of 12 hours. which. placed at a
6-month interval and a mid-distance between
the solstitial days. ought to have been
intentionally regularized to bear witness of
the understanding that the Egyptians had of
the phenomenon of the equinoxes"..The
duration-values of the text of Tanis certainly
rest upon observations and. if they differ from
the actual values. it could only be that some
of the errors were owing to the imperfection
of the measuring instruments employed to
determine them. There could, of course, also
have been copyist crrors.
It was also observed by Clere that the use of fractional
hour measurements in this table demonstrated that the
Egyptians did not simply use successive divisions by
-Kl3-
two or three in their tables as some people have

thought. but that their hour divisions were
duodecimally based. doubtless in imitation of their
division of the year into twelve months and the day
and night into twelve hours.129 The obvious difficulty
of measuring accurately time units as small as twelfths
of an hour ....hen the Egyptians could not even
accurately measure the length of an hour itself strongly
contradicts the suggested observational base of the
table and points rather to some more arbitrary
mathematical schematization of the table that Clerc
initially rejected. Finally we can notc that Clerc
produced a table of his own comparing the values for 13
days that survive in the Tanis table with actual values
of the duration of daylight on those days estimated for
several cities in Egypt (see Fig. 111.60.
We can complete our consideration of the Tanis
table by examining the analysis of it made by
Neugebauer and Parker. It stands on its own and is
worth quoting in fuJI. especially since in the course of
their analysis they reject Clere's suggestion that the
table was established by means of observation. Here
then are the remarks of Neugebauer and Parker:130
This list concerns the length of daylight
and night. expressed in equinoctial hours.
Unfortunately only 13 of the original 24 data
for the length of daylight are preserved.
They are indicated in our Figure 10 [see my
Fig. m.60], by blaclc dots when both daylight
and night are given. and their total is 24
(hours]. In all other cases only one of the
two values is preserved.
Four values of the thirteen conform to 8
pattern one would expect for lower Egypt.
-104-
The longest daylight. at the beginning of I and

II prt. is 14h, corresponding to the standard
Hellenistic norm which assumes a ratio 14h:l()h
for the extrema. The equinoxes are located in
II '!Jt 15 and IV prt 15 in conformity to the
solstice data.
It would be natural to expect from this
arrangement a linear scheme for the remaining
values. following a well-known paltern of
Hellenistic astronomy. Such a scheme is
indicated in our figure by the dotted line. It
would require a constant difference of ± 24
minutes for each IS-day interval. None of the
preserved data, however, agrees with such a
scheme. A constant increase occurs only once
(J 'hr
- 15 to II 'ht
- 15) but with 30 minutes
instead of 24; a constant decrease of 2S
minutes per month Onstead of 48) is found
between III prt 15 and I 5mw 15. The
remaining values follow no recognizable
pattern, not even as far as the necessary
symmetries are concerned. The minimum is 9
1/3h instead of the expected l()h; finally 22 h in
II 5mw I is an obvious scribal error (for II?),
The absence of any strict regularity
caused Clere to assume that the data of the
text were based on observations. It is, of
course, impossible to disprove such an
explanation. On the other hand 8 month by
month observation of lengths of daylight and
night with relatively high accuracy (down to
twelfths of hours. Le., 5 minutes) falls
completely outside the experience we have
elsewhere with the handling of the problem of
-105-
variable daylight. The norm is al....ays a

simple ratio of longest to shortest day and for
the intervening dala either linear variation or
linear variation of the rising times. The
occurrence of small fractions also speaks more
in favor of some arithmetical process. perhaps
incorrectly applied. than of direct
measurements with very finely calibrated
water clocks.
That the lengths of daylight and night
were not observed independently is obvious
since such observations could not result in a
constant total of exactly 24 bours. Hence
either daylight or night would need to be
observed and the other part found by
subtraction from 24. Since the water clock
seems to be the only instrument available for
this purpose one could perhaps consider the
length of night as the basic interval. But
whatever the measured interval may be it
seems extremely unlikely that observations
would result in equinoxes spaced exactly six
months apart or in a neat longest daylight of
14h from I prt 1 to II prt I. An original poorly
computed and badly copied text seems to be a
historically more likely hypothesis.
Astronomical Ceilings and Other Monuments

The last subject of this chapter centers on the
development and existence of a series of astronomical
registers and related insertions. These appear primarily
as reliefs or paintings on various surfaces. including
coffin boards. tomb and temple ceilings. water clods.
and other objects. Neugebauer and Parker catalogue 81
-106-
relevant astronomical monuments and objects.l31 The

subjects represented on the various registers include the
declIns in a belt south of the ecliptic (i,e., some varying
number of the standard decimal stars and constellations
used on the diagonal clocks we discussed earlier),
planets. certain northern constellations with lIttending
lind flanking deities. Various of these objects are also
found on zodiacs which are combinations of Hellenistic
concepts with traditional Egyptian elements; the zodiacs
all date from the Greco-Roffian period. Needless to say
I shall only describe and analyze a few representative
monuments, emphasizing what the monuments tell us
about the detail and extent of ancient Egyptian
knowledge of the heavens.
Before expanding my discussion of the standard
arrangement of astronomical elements in celestial
diagrams. a few pertinent words on the orientation of
monuments from Neugebauer and Parker are in order: 132
The term ·orientation" should not be
understood in an exact astronomical sense.
Temples were usually directed with their main
axis perpendicular to the Nile. thus generally
east-west or west-east. But the course of the
Nile can deviate considerably from its
northerly direction and thus may bring the
"northern" side of a ceiling decoration actually
to the west. as for example in the temple of
Hathor at Dendera which lies south of the
Nile on its great bend from Qena to Nag C
Hammadi and so looks almost due north.
.."rather strict rules governed the position
of the various elements of an astronomical
composition with respect to the cardinal
points. The principal areas are north and
.-107-
south of an east-west axis. When the areas

of a ceiling or other monument are east and
west of a north-south axis, then the eastern
area has the same decorative elements as
would the southern area, and similarly for
west and north. To state it in another
fashion. north and west oppose soutb and
east....
On astronomical monuments which are
fixed in place, then. we consistently rind the
northern constellations on the north or west
or northwest. while the decans. since they lie
in a belt south of the ecliptic... , are
consistently to the south or east or south-east.
The arrangement of the various astronomical
elements developed into an almost standard form that
we can with some looseness call tbe Ancient Egyptian
Celestial Diagram. In fact there are about six £llmilies
of the standard form. but not much wiJI be gained by
describing all of the divergencies. Hence ....e shall
confine our preliminary discussion to the standard
arrangement. with only general and occasional remarks
to family dirrerences.
Initially we should recall. from our treatment earlier
in the chapter. that the principal astronomical diagrams
before the development of the standard celestial
diagram were those of the diagonal docks of the
coffins. We remember that, in the case of the docks
dependent on risings (that go bad to the ninth or tenth
dynasty), the ideal diagonal clock (never realized but
inferred from those coffins that remain) in its columnar
structure included the names of 36 decans applicable to
the 36 decades of the year plus 12 decans for the
epagomenal days. Between the 18th and 19th columns a
-108-
strip (designated as the ·V strip") included depictions of

Nut. the Foreleg of an Ox (tbe Big Dipper), Orion, and
Sothis (the goddess of the star Sirius). Some dirrerent
dccans were selected for use in the clocks dependent on
their culminations (meridian transits).
In the ordinary celestial diagram which we are now
considering all 36 decans of the diagonal clocks (or
more or less than the 36) were listed but not in the
dillgonal form. used on the coHins (see Figs, 111.2. iliA,
111.21, 111.22, III.65b and the detailed des<:ciption of the
ceilings of the tombs of Scnmut lind Sct; below).
Usually onc (or occasionally more than one) star
accompanies the decanal name as II determinative. I(
more than one, the stars indicated an attempt to
represent a deea" consisting of more than one star.
Furthermore liS our detlliled presentation in Documents
111.3 and iliA show. groups of stars appear in the
columns below the decans. However. such star
groupings or clusters are of little assistance in
identifying the constellations they represent because the
number and arrangements of stars for each group varies
widely from monument to monument. Exceptions are
those below the decans related to Orion and Sirius.
which can be identified. at least roughly. The names of
the deities associated with the decans are usually
included. but again not with great uniformity.
Sometimes depictions (and not just the stars from which
depictions were conceived) of the gods or possibly
larger constellations associated with a number of the
decans were also included. e.g.. a bark, a sheep. an egg.
Orion in a bark. and Isis-Sothis in a bark.
In addition to the main 36 decans. the artists
included the names of some (and perhaps in the original
celestial diagram. if it were extant. aID of the 12
·109·
triangle decans used in the old star clocks for the hours
during the epagomenaJ days: these triangle decans were
also accompanied by stars. their associated gods. and
less frequently other material (such as a depiction of
the constellation of the Two Tortoises under the decan
§/lVy). After the last regular decan, that of Isis-Sothis
(j.e.. Sirius). and before the triangle deities the names of
the three outer planets <Jupiter, Saturn, and Mars, or
sometimes only Jupiter and Saturn, Mars being omitted)
were given and following the triangle decans the two
inferior planets Mercury and Venus were added. As
depictions of the exterior planets we find two or three
figures of Horus in a bark. For Venus the deity names
b l/) and lVslr (Osiris) are given and a heron is depicted.
So much then for the celestial bodies south of the plane
of the ecliptic (those of the decanal belt) and the bodies
which in their motions intersect the plane of the
ecliptic Cthe planets). All of the decanal and planetary
material ordinarily occupies one grand southern panel.
In addition, many of the monuments include in a
northern panel a group of constellations oriented about
the Big Dipper. These are usually called the "Northern
Constellations: though the changing arrangements of
these constellations and their accompanying deities from
monument to monument makes their identification
(except for the Big Dipper) extremely difficult if not
impossible. Hence without II sure identification of these
constellations, we cannot assert that all of them are
necessarily circumpolar. as is usually said. though I
suspect that they are since the Egyptians
conventionally distinguish two groups of stars, namely
the Imperishable stars, which are the circumpolar stars
that do not set, and the Unwearying Stars, which are
the decans and planets (see Volume One, Index of
-110-
Proper Names. under 'stars') and this is just the

distinction that seems to be made in the northern and
southern panels of the celestial diagram.
On each side of tbe northern constellations is a
register of panels which contains protective deities of
the days of the month with disks on their heads.
beginning on the right side with Isis (sometimes), the
four sons of Horus. and four more protective
day-deities. and on the left side as many as ten or
eleven other protective day-deities. We owe to
Brugsch the discovery of the origins of the assignment
of these day-gods. OJ
Below the northern constellations we occasionally
rind a depiction of Thoth as the cynocephalic baboon,
and we have already discussed its significance on the
Ramesseum astronomical ceiling and on water clocks
(see notes 21 and 84). Flanking Thoth the standard
diagram (beginning on the [ert side of Thoth) includes
panels for the twelve months showing the deity of each
month fadng the king.
So much then for a generalized account of the
so-called celestial diagram. Now we should look briefly
at two specific monuments, which are described in
greater detail in Documents 11I.3. and lIlA.
The Ceiling of the Secret Tomb of Senmut

The oldest monument 134 that includes enough of
the celestial diagram to allow us to get a good sense of
the whole is contained on the ceiling of the tomb of
Senmut CTheban tomb no. 353).135 This sO'called
·secret: unfinished and empty tomb of Senmut was
discovered near the entrance of the temple of
Hatshepsut during the excavations of the Metropolitan
Museum in western Thebes in 1925-1927.136 It is, of
-m-
course, not to be confused with his completed tomb

(Theben tomb no. 71) on the hill of Sheikh cAbd
el-Qurna. Presumably it remained unCinisbed bttause
Senmut. who supervised the building of Hatsbepsut's
temple and apparently had long and close service ·on
the queen's right hand: fell out of favor with his queen.
One of the most interesting features of the tomb is
a sketch of Senmut on the .....all of the stairway (clIding
to the first chamber (see Fig. 111.63), Next to the
sketch we read ·Overseer of the Estate of Amun.
Senmul: But it is the astronomical ceiling of the
chamber al the bottom of the third flight of stairs that
principally concerns us (see Fig. IlIA and Document
111.3). It consists of a southern panel (Fig. llJA. top)
lind a northern panel (Fig, lilA. bottom) separated by
nvc lines or prayers ror Senmut. the middle or ....hich
includes the titulary or Queen Hatshepsut. The
southern ptlnel is devoted to the decans (listed rrom
right to left in Fig. iliA), the northern ptlnel to the
so-called Northern Constellations ....hich occupy the
center or the panel and ....hich are £Ianked at the top by
....ell dra....n circles representing the 12 months or the
year (with their names superscribed) and at the bottom
by the day deities which I mentioned above.
So rar as the de<:ans or the southern panel are
concerned, the relatively early date or the Senmut
ceiling decorations <Ca. 1473 D.C'> in comptlrison to the
other monuments containing the celestial diagram seems
to suggest tbat the Senmut decorations represent the
prototype ror a decanal ramily.....hich comprises 18 lists
that can be grouped into one main group and t....o
subgroups.137 The earliest extant list is that on the
Senmut ceiling and the latesl list is that or Harendotes
dating rrom about 246-221 B.C. A carerul analysis or
-112·
the decans of the Scnmut ceiling by Neugebauer and

Parker resulted in their reconstruction of the original
thirty-six decaos (one of which is missing in the
Scnmut ceiling), and most importantly in the
straightening out of the confusion in the decans
associated with Orion (mentioned in more detail in
Document JII.3).I38 The reader will notice in Fig. 111.4.
top. that the decan names have stars as tbeir
determinlltives. For example. in the nrst column
(starting from the right) the name of the first dccan
(tpy-C knmtJ has onc star as a determinative. This is
followed by the names of the gods associated with the
decam Hapy and fmseti which are directly followed by
two stars. Then in the second column at top appears
the third deean (lJry tJpcI knmtJ with the second decan
(knmt) below it. each decan followed by a star. Below
them the name Isis appears and below it. covering both
columns in a descending line. appear five stars plus a
sixth one below in column 2, and considerably below
them near the bottom of the two columns is a
descending line of five more stars. These star groups
apparently indicate some general constellation
embracing all three of the decans connected with Immt.
I leave the details of the remaining columns of regular
decans to the considerations found in Document 111.3.
except to note that at the bottom of several columns
we see what are apparently grand constellations with
pictorial representations and accompanying lines of
stars. Thus under columns 7-12 is depicted the Ship.
14-16 the Sheep. 21-23 the Egg (bul considerably above
Ihe line of the other grand depictions), 24-28 Orion in a
boal, and 29 Isis in a boal. with a feathered crown
surmounted by a disk. She represents Sothis. whose
name here. as often in the decan lists. is written at the
-113-
top of the column as Isis-Sothis (Ist-spdtJ. The ultimate

origin of the decan list of this ceiling certainly lies in
the decans arranged in some diagonal star clock. This is
shown by the fact that. beginning with column 7. a
horizontal line is drawn all the way up to the 29th
column and each of the columns so underlined contains
a successive 12th hour decan from a diagonal star clock.
This line then represents the bottom line of a star clock.
The next column (col. 30l in the southern panel
following those containing the main decans is devoted
to the planet Jupiter. and we read in the hieroglyphics
"Horus who bounds the two hmds'is his name" followed
by a disordered phrase which probably should be
rendered "southern star of the sky: Below is the
falcon-headed Horus in a boat and bearing a star on his
head. The column next to Jupiter is that of Saturn.
where we read "Horus bull of the sky is his name:
followed by a disordered epithet. which should probably
read 'the eastern star which crosses the sky: Once
more the falcon-headed Horus with a star on his head is
depicted in a boat at the bottom of the column. A
column devoted to Mars often follows in other
monuments (see Fig. 1ll.65b and my discussion of the
ceiling of the tomb of Seti I below). In the Scnmut
tomb it is omitted.
Following the exterior planets are six of the 12
triangle decans in five columns. I mention only the
first decan (f/wyJ. which has at the bottom of the
column two tOrloises. and the second decan (nsrwJ.
which has below its name the commentl "It is a cluster
(bl pwJ." This is no doubt an indication that this decan
is a group of stars. Finally the southern panel
terminates with columns devoted to the interior planets
Mercury (sbgJ and Venus (/JI). At the boUom of the
-114-
columns of Venus a heron with a star on its head is

pictured.
Turning to the northern panel of the Senmut ceiling
(Fig. 111.4 bottom) we should first concentrate on the
northern constellations in the middle of the panel (cr.
Fig. 111.66>, those usulIlly designated as the circumpolar
constellations. At the top center we find the Big
Dipper or Great Bear represented here as II bull with a
recognizable bull's head joined to a curious ovoidal
body, with two short protruding legs llnd II tail of three
stars. the last being a red encircled dot hom which two
diverging vertical lines extend to the bottom of the
panel (see Fig. 111.66 [0]). Presumably the use of this
form of the bull liS well as a fully formed striding bull
Gn some copies of the celestial diagram) and Il hybrid
where the bull's head is joined to a buWs foreleg Gn
other copies of the diagram> merely reneet the earlier
Egyptian representation of the Great Bear or Big Dipper
as a buWs foreleg (ms!!tyw), as it is always depicted on
the Middle Kingdom coHins (see Figs. 111.14, 1ll.16 and
111.17. where the four stars of the dipper and three of
its handle represent the basic points of the foreleg).
Pogo, by referring to a slcetch of the northern
constellations with the horizon at Thebes. dated about
2000 B.C. (see Fig. 1II.64a; and d. Fig. 1lI.64b for the
northern constellations at about 3500 n.c. not long
before the beginning of the dynastic period when the
"hour watchers· began to delineate the main features of
the celestial diagram>, commcnts on the representation
of the Grcat Bear as follows,I39
The three stars attached to the bull of the
Scnmut ceiling correspond to thc position[sl of
Delta, Epsilon, and Zcta Ursae Majoris.
Around 3000 B.c., Zeta was the only 2nd
-IIS-
magnitude star within ten degrees of the pole;

the upper culmination of Eta Ursac Majoris
coincided. for the latitude of Thebes. with the
setting of Sirius.
Observations of ethel culminations of
circumpolar stars for the determination of the
meridian were certainly made by Egyptians
when Eta and then Zeta Ursae Majoris were
the nearest of the bright circumpolar stars.
The transit staff in Berlin [see my Figs. 1II.20a
and 1II.20b] is of relatively recent date, but
the plumb-line holder which apparently
belongs to it looks like a distant descendant of
the cross~shadow ruler [i.e.. shadow clock]
which beaTs the name of Thutmose
[Tuthmosisl JJl [and is described above in my
account of shadow clocks];...
The cord+stretching procedures accompanying the
foundation of temples no doubt used such
determinations of the meridian. I4O Cord-stretching goes
bad to the rirst dynasty and is mentioned in the
Palermo Stone (see Volume One, p. 50 and the Index of
Proper Names: "Stretching the Cord").
Continuing our perusal of the constellations that
accompany Meskhetyu at the center of the northern
panel of the Senmut ceiling (Figs. 111.4, bouom, and
111.66, [0]). note that there are seven rigures that surely
represent constellations of stars: [l1 Serket (the scorpion
goddess above Meskhetyu and looking down on
Meskbetyu and the other figures), [21 Anu (a
falcon-headed god below and facing Meskhetyu. with
arms extended and holding a cord or spear directed at
and touching Meskhetyu), [3] lsis-Djamet (the
hippopotamus goddess-with a crocodile on her
-116-
bad-standing at the bol1om of the group and to the

tight. with Cront paws holding on to 8 mooring post and
a small. vertical c,rocodile). [4] Hotep-redwy (whose
name seems to be an epithet of the Crocodile god
Sobek-"restful of feet"> is the crocodile which is toward
the left and opposite the bottom of the hippopotamus
and which is lunging al [5] Man (with upraised arms as
if to spear the crocodile; to the left behind the crocodile
is II small figure of a man thal seems not to be 8 part of
the whole group>. [6] Divine Lion (with a crocodile-like
tail. resting on his haunches above the figures of the
crocodile and Man). (7a] Haqu-the Plunderer (8
crocodile with II. curved tail. located above the Divine
Lion); the figure follows the name perhaps as a
determinative; since [78] appears only on the Senmut
ceiling, it perhaps rdlects or has some relationship with
the next Figure. The last constellation, [7] Saq (another
crocodile with a curved tail, located at the top of the
panel, to the left of Serqet). As the result of the size
of the two flanking rows of the monthly circles, there
is considerable space between the group at the top
comprising Meskhetyu and Figures m. (2), and (7) and
the group at the bottom including (3), (4), (5), (6), and
(7a). One supposes that the totality of Figures
representing the northern constellations in its original
form was a much more compact group, and indeed
Neugebauer and Parker have attempted to reconstruct
....hat they believed to have been the original, more
tightly knit grouping (see Fig. 111.66>'
As one examines the various extant versions of the
celestial diagram given by Neugebauer and Parker in the
third volume of their Egyptian Astronomical Texts,
Chapter IV, and my remarks in Documents 111.3 and 111.4
below, it is evident that there was considerable
-117-
variation in the arrangement of the northern

constellations in their depictions on the sundry
astronomical monuments (c.g.. compare Fig. 111.66 with
Figs. 111.67-69»)41 Hence. as in the case of decans given
on the southern panel. it is exceedingly difficult (if not
impossible) to identify the Egyptian constellations
exactly with modern depictions. except, of course, for
the certain identi£ication of Meskhetyu with the Big
Dipper. Still. the identifications proposed by Pogo are
reasonable. if not certain.142
The slar surrounded by a circle on the
Senmut ceiling corresponds, obviously. to the
early recognizable bright staT. Zeta Ursae
Majoris-the one with the conspicuous
companion. The scorpion goddess Selqet
(Scrqed stands behind the bull Meskheti in
such a way that it seems a~ if she were trying
to grasp the two cords stretched from the
culminating star Zeta-over the invisible
pole-down to the northern horizon....
By 1500 B,C•• the celestial pole was closer
to Ursa Minor than to Ursa Majorl
culminations of Beta Ursae Minoris could be
-and possibly were- used for the
determination of the meridian, It is therefore
not surprising that craftsmen of the XIXth
dynasty began to consider the meridian cords
as mere reins attached to the tail of the bulh...
The hippopotamus with the crocodile on
her back probably corresponds to the stars in
Draco indicated in our sketch (Fig, 4 [omy Fig.
1II,64a; cr, Figs, 1II,64b and 111.708 for the
locations of Draco in 3000 B.c' and the
present day]), One of the hands of the
'118-
hippopotamus rests on the "dipper" of Ursa

Minor. which is just east of the meridian with
our polaris near the horizon of Thebes. The
hand of the hippopotamus always rests, in
later representations. on an object which is
wider at the lop than at the boltom [j,e.• a
mooring post]; the reins always lead to that
support of the hippopotamus in more or less
fancy curves-a tradition blindly followed by
craftsmen who could not be expected to
realize that in the dim past a vertical line
joined Zeta or even Eta Ursae Majoris, when
they culminated, with Ursa Minor near the
horizon.
The identification of the circumpolar
constellations...which appear west of the
meridian. near the horizon. is father difficult.
The stars Omicron and 23 Ursae Majoris may
have something to do with the crocodile, and
the stars Lambda-Kappa-Alpha Drllconis, with
the man facing the crocodile (his left hand, his
head, and his right hand respectively),
The identification of the falcon-headed god Anu
with Cygnus argued by Wainwright 143 seems unlikely
because of the considerable distance of Cygnus from the
Big Dipper (cL Fig, 1II,64a and Fig, 1II.70a),144 J
suspect that the same reasoning puts into doubt some of
the identifications of Herbert Chat ley, which he calls
"probable"; the Hippopotamus with Bootes (the
Guardsman), Hercules, Lyra, as well as Dracol the
crocodile and accompanying god or man with Cassiopeia
and Perseus; the Lion with Auriga (to which he appends
a "1"); and the Goddess Serqet with Coma Berenice
(again with a 7).145 The methods used by Rekka
-119-
ANCIENT EGYPTIAN SCiENCE
Aleida Biegel to establish the stars that make up these

and other constellations in the Egyptian celestial
diagram 146 have been shown by Pogo. in a severe
review of her dissertation, to be quite: fallacious and
inadequatc:.147
Not much more need be said in this brief account of
the celestial diagram as represented on the Senmut
ceiling. I have mentioned the twelve circles
surmounted by month names that appear on each side
of the northern constellations. They proceed in the
first row from right to left and then in the second row
from left to right. Hence the four circles at the top
right refer to the four months of the season Inundation.
the two at the top left and the two beloW' them on the
leFt refer to the season of Peter. and the four circles on
the right to the season of Shemu. Each circle is divided
into 24 unidentified sectors. Since the time of the first
descriptions of the tomb the sectors have been thought
to mark in some fashion the 24 hours of the day and
night periods, but there have not been any satisfactory
explanations as to how they marked the hours.148
Perhaps the sectors were meant to carry two sets of
12-hour meridian or other transits of bright stars, as in
the Ramesside transit clocks, which we have seen
probably originated at least as early as the time of
Senmut in the 18th dynasty. If so. then each circle was
meant to contain the star-transits marking the 12
nighttime hours of the rirst day of the month in one
half the circle and the star-transits marking those SlIme
hours of the 16th day of the month in the other half of
the circle.
There is no need to discuss here the nature and
number of deities that £lank the lower part of the
central northern constellations since they seem to have
-120-
no astronomical significance. The reader may compare

the Senmut ceiling with the illustrations of two other
members of the Senmut family of depictions given in
Figures 111.2 and 111.22.
The Vaulted Ceiling of Hall K in Seti I"s Tomb.

We have talcen some pains to describe the Senmut
ceiling because it represents an early and influential
family of the Egyptian celestial diagram. A few words
can be added concerning another inrluential depiction of
that diagram found in the ceiling of hall K in the tomb
of $eli I 0306-1290 B.c.> in the Valley of Kings in
Western Thebes (see Document iliA for 8 more detailed
account), The sepulchral hall with the whole
astronomical ceiling is shown in Fig. 111.658. It is
evident that the northern and southern panels (often
designated as the Eastern and Western Parts; see
Document lilA belo.....) are separated by an empty band
in the center. Let us look first at the decan list of the
southern panel (see Fig. III.65b). Neugebauer and
Parker call it "the prototype for the third family of
rising decans: l49 The list includes 39 decans, but three
of them appear to be triangle decans that have strayed
out of position, Sometimes the columns are ..... idened to
include more than one decan (e.g.• columns 11, 12, 20.
and 22). no doubt because the decans included are part
of a larger constellation <the prime example being the
various decans in column 22 that are a part of the
constellation of Sah. i.e., Orion>' As in the Senmut list
mistakenly omitted decans are occasionally inserted
belo..... the top decan (or decans) in a column (see
columns 11, 14. 20, and 22). In most cases the omitted
decan is placed beneath the decan it should precede.
But not al.....ays. For example, in column II. the added
-121-
decan (fms Inl lmte) was properly squeezed in below

the two decans (!!ntt lIrt and bntt brtJ given at the top
of the column.
The orientation of the southern panel in the $eli
ceiling is more astronomically correct than in the
Senmu! ceiling. as Pogo has remarked:15O
As stated before, the orientation of the
southern panel [on the Senrnut ceiling] is such
that the person in Ihe tomb looking at it has
to lift his hcad and face north. not south. The
list of Ihe dccans preceding the Sah-Sepdct
[Orion-Sirius] group occupies. therefore. the
castern par[ of the southern panel. whereas
the planets and constellations which follow
Sepdct in the traditional arrangement are listed
in the western parl. The southern strip of the
Ramesseum. like the southern panel of Scnmut.
must be read. in the temple [see Fig. 1lJ.2l. by
a person facing north. On the ceiling of Set J.
on the other hand. the orientation of the
southern panel is astronomically correct. so
that Orion precedes Sirius in the westward
motion of the southern sky.
The irrational orientation of the southern
panel has caused some confusion in the
representation of Sah [Orion] on the ceilings
of Scnmut and of the Ramesseum, both of
which obviously follow the same tradition.
On the ceiling of Seli I - which reflects
another tradition - Osiris-Sah. participating in
the nightly westward motion of the sky. is
running away from Isis-Scpdet; he turns his
head eastward to look at the pursuing goddessl
the bow of her boat is almost ramming the
-122-
stern of his. With the reversed orientation of

the southern panel [in the Senmut traditionl,
Orion. the most conspicuous constellation of
the southern sky, appeared to be moving
eastward. i.e., in the wrong direction; to save
the situation, Orion was turned around-within
or with his boat. The Senmut and the
Ramesseum ceilings represent Orion in this
'reversed' position. adapted to their orientation
of the southern panel: the Senmut draftsman
made the bows and the sterns look alike; in
the Ramesseum. the bow of Orion's boat
approaches the bow of the boat of the
Sirius-goddess; the element of pursuit is lost in
both cases; instead of looking back at the
bright star behind him. the god is turning his
head away from the goddess. Mythologically.
both the Senmut-Ramesseum and the $eti
traditions may be equally valuable;
astronomically. the $eti representation is far
more satisfactory.
The first 25 columns (starting from the right) are
divided into three registers and the last 10 into 4
registers. In the decanal columns. the names of the
decans are included in the first register; stars indicating
(but not always accurately) the number of decans in a
column are in the second register; and the names and
depictions of most of the patron deities of the decans
(having anthropomorphic bodies and a mixture of
human and animal heads) are in the third register.
There is considerable crowding of the figures of the
gods and goddesses so that some of the decans are not
represented and some of the figures do not represent
accurately the decan given above it. The large figures
-123-
of constellations mentioned in our description of the

southern panel of the Senmut ceiling are mostly present
on the Seli ceiling, except for the figure of the sheep
(under columns 14 16 in the Senmut list) and the figure
4
of the two tortoises under the triangle decan 5/wy (in

column 32 of the Senmut list). One other diHerence in
the two lists is the inclusion in column 26 of the $eli
list of the planet Mars ("easlern star of the sky, his
name is Horus of the Horizonl he travels backwards
[j.e.• is in retrograde rnotionJ")ISI, with the depiction of
the f8lcon~he8ded god in 8 bark, while this planet is
missing from the Senmut panel. Hence $etrs list
includes all five of the easily visible planets, Jupiter,
Saturn. and Mars before the six triangle decans and
Mercury and Venus after them. J52 The reconstructed
list of triangle decans on the Seti ceiling is the same as
that of Senmut.
Finally, in regard to the southern panel. there is
considerable divergency in the number and arrangement
of stars that accompany the main 36 decans. As in the
case of the Senmut ceiling. it is extremely difficult to
identify with any certainty the decans listed on the Seti
ceiling, other than those connected with Orion and
Sirius at the end of the list, and even those surely
connected with Orion are not identifiable with specific
stars in the Orion constellation. 153
A few words remain to be added concerning the
northern panel on the Seli ceiling (see Fig. 1II.65b.
bottom, and especially Fig. 111.69). Among the
differences in the northern constellations of the Seti
ceiling from those on the Senmut ceiling. are the
folloWing: (I) Meskhetyu, the constellation of the Great
Bear. is now represented by a complete bull striding on
a platform; (2) An (written as An rather than as Anti).
-124-
the falcon-headed god, appears to be supporting the

bulrs platform: (3) behind the bull a Man (or god) with
a disk on his head has been added: he holds reins that
connect the bull with a mooring post resting on the
bottom of the figure; (4) in front of the bull a falcon
has been added; (5) both paws of the Hippopotamus
appear to be on the mooring post; there are seven
circles (slars) running down the 1·lippo·s extended
headdress; (6) the crocodile on the back of the Hippo
here is much larger than it is on the Scnmut ceiling; (7)
the other crocodile on the lert is more horizontal: as
before. the crocodile lunges at Man (78) in whose
outline are 10 circles (stars); the Lion (8) above the
crocodile is almost completely outlined by stars; (9)
there is only one depiction of the crocodile with a
curved tail and it is locatcd in front of the Lion and
acts as a determinative for its namc; (10) the scorpion
goddess Scrqet is at the top left rather than above the
bull; and all of the figures except the crocodile on the
back of the Hippo, the crocodilc on the bottom lert, and
the Falcon, have circles within them representing the
stars on the basis of which the constellations were
imagined: those in the rigures> with anthropomorphic
bodies being more numerous than those in the figures
with animal bodies (see the use of the circles as stellar
connecting points for the construction of the
constellations by Biegel mentioned in notes 146 and
14n
I ignore once more the deities nanking the northern
constellations on the northern panel of the Seti ceiling
since they do not seem to have any astronomical
significance. Finally we can note that there is no
register on the 5cti ceiling that includes the months of
the year and their deities.
-125-
Egyptian Zodiacs
One last kind of astronomical decoration on
monuments (temples. coffins. and tombs) needs to be
treated, Egyptian zodiacs. all of which originated in the
Greco-Roman period. The earliest known Egyptian
zodiac (the so-called una A zodiac) was 8 rectangular
one hom the temple of Khoum 2 1/2 miles northwest of
Esna, that is. from the part of the temple originating in
the reigns of Ptolemy III-V. Hence it dated from about
200 D.C. That temple was destroyed in 1843 to build a
canal and the zodiac disappeared along with it. But the
zodiac is still preserved in a plate from the Description
de I'~gypte (see Document 111.17 and Figs. 111.758 and
1II.75bJ.
The pictorial representations of the zodiacal signs in
the Egyptian zodiacs (though nol their full range of
Hellenistic additions) were surely of ultimate
Babylonian origin. This is true of both the early
rectangular form of the zodiacs and of the round
zodiacs where the signs constitute a circular belt (as in
the Dendera B zodiac presented in Document 111.17 and
depicted in Figs. JII.76a and 1II.76b). Though it is
tempting to think that the Egyptian round zodiacs
represented the divisions of the zodiac in terms of
degrees (as some of the earlier investigators seem to
have assumed). it is evident from the analysis of the
celestial diagrams I discussed above that there was no
use of a system of degrees in ancient Egypt to measure
celestial arcs. The Egyptian zodiacs not only included
the signs of the zodiac but the old hour decans. the
planets. and other Egyptian stars or constellations as
well (like the Big Dipper and the Hippopotamus). By
the time of the preparation of these zodiacs. the risings
-126-
or transits of decans were probably no longer used to

mark the hours of the night. In the earliest Egyptian
zodiac. the rectangular zodiac Esoa A mentioned above
(Fig. 1II.75a) and also in the rectangular zodiac in the
ceiling of the outer hypostyle hall at Dendera
(designated as Dendera E), the decens for the most part
seemed to be associated with or represented divisions of
the areas assigned to the zodiacal figures. each sign
being divided into three decans (except in a few cases
of four decans), which. when absorbed in Greek and
Roman zodiacs. were merely the names of the three
IO-degree divisions of each of the zodiacal signs.
Hence. in this new form the decans continued to playa
part in Ihc later astrology of Greece, Rome. India, Islam.
and medieval and Renaissance Europe.54
A more detailed examination of the Egyptian
elements in these zodiacs is given in Document IlI.l7
below. We can simply note here that these elements.
and particularly the decans. were drawn primarily from
two families of the celestial diagram, those called by
Neugebauer and Parker "the Seti I B· and the "Tanis·
families and a few constellations from those we have
discussed as northern constellations on astronomical
ceilings. And indeed the earliest complete Greek
version of decanal names, that of Hephaestion of
Thebes (fourth century A.DJ. appears to have been
chosen exclusively from the Seli I B and Tanis families
(see Fig. 1I1.103).
Conclusion
So far I have discussed the principal Egyptian texts
and monuments bearing on aSlronomy. The reader may
have been conscious of the fact that almost all of the
documents have been anonymous. A spectacular
-121-
exception was the tract on the waterc10ck composed by

one Amenemhet and dedicated to Amenhotep 1 (see
Document 111.15), Another interesting example is given
below as Document 111.18. where in an inscription on his
statue (ca. 2nd century. B.CJ an astronomer and snake
charmer named Harkhebi enumerated his astronomical,
calendrical. and timc·telling activities, including the
observation of the stars and announcements of their
risings and settings. his purification on the days when
the deean Akh rose helillcally beside Venus. his
observations of other helillclil risings. lind particularly
his foretelling of the heliacal rising of Sirius at the
beginning of the [civil] ycar. and so on. These lire
essentially the activities we have described in this
volume. and the documents that now follow and
constitute· the major part of this volume will throw
further light on them.
It should be clear from my summary account that
the ancient Egyptian documents do not employ any
kinematic models. whether treated geometrically or
arithmetically. However they did use tabulated lists of
star risings and transits (as is revealed clearly in
Documents IlI.II. 11I.12. and IlI.14). all tied to their
efforts to measure time by means of the apparent
motions of celestial bodies,
On more than one occasion in this chapter. I have
remarked on the absence in early Egyptian astronomy
of the use of degrees. minutes. and seconds to quantify
angles or arcs. though slopes were copiously used in the
construction of buildings. water clocks and shadow
clocks; such slopes were measured by linear ratios,1S5
And obviously. in the above mentioned tables for hour
charts. right ascension was being approximated in terms
of hours (though we are not completely sure how
-128-
precisely), So far as I can determine, there were no

standard. quantitative evaluations of declimltions given
in the tables. But of course in determining meridian
transits the observers were keeping track of the
maximum height of the transiting stars each night
throughout the ycar.
Finally. the reader should realize that by the time
of the Ptolemaic and Roman occupations of Egypt there
were numerous other late Hellenistic and Roman
astronomical and astrological texts which. though
including some Egyptian elements, are not discussed in
my volume. This is because they were primarily
motivated by Greele and Roman astronomy or astrology
(e.g.. documents such as planetary tables consisting of
the dates of entry of planets into zodiacal signs). I
consider them outside the limits of my account of
ancient Egyptian astronomy, but the reader can find
expert treatment of such texts, or guidance to them, in
the last chapter of Volume Three of Neugebauer's and
Parker's Egyptian Astronomical Texts. 56
(For notes to Chapter Three. see pIIge 131)
-129-
•
NOTES-CHAPTER THREE
Notes to Chapter Three

l. Sec Th~ C~/rnd.rl of Anl"lrnt Egypt IChlngo, 1950l:
•Anclent Egypl1an Astronomy: Phllorophlt:al Trannctlons of the
Roy.1 Soc~ty of London. A.276 Cl974}, pp. 51-65: :and iI somewhiill
more recenl relter.tlon of his main condu,lonl. ·Egyptlan
Astronomy. ArtrolOf,Y, and Calendrical Redonlna: C.C. Gillispie,
~., DktJonary of SdentlfJe Biography, Vol. 15 (Ne.... York. 1978),
pp. 706-10, full.rtld.., 706-27.
2. "Anelent Egyptlan Aslronomy: pp. 52-53. Parker In hi.
C.lend"rs. pp. 30-]1, a1vu • hlslork.1 pr«ls of the earlier thead".
of the original lunar calendar. Indeed he d"cl"n lhl! "Borchardt
11935 I-Mittel etc., pp. Sfr.. 24, cf. allo OLZ, XXVIIl, 1925, 620.
and ~s. LXX. 1934. 98·99JJ .... :a. the last chronolO&cr 10 discuss
.n Form. of the Eayptlan yur. both civil and lunar. According to
him lDorchardtl. the first yur was lunar. and II began with the
next lunar month after the hellacal riling of 5othll. When,
becaule of Ihe yurly Ihlft forward of Ihe lunar calendar by lIbout
ele..en daYI Iitl yur a ..engln; only JS4 daYI In Jen;thl, the flrlt
month of the ycar would fall before Ihe rilln; of SlrIUI, the month
wal. Inllead, Intercalary. The name of Ihll lnterealuy monlh wal
wp ropt, Proof of the ellistence of luch a calendar WlIS 10 be
found In Ihe Eberl calendllr Isee Doc. 11I.21 and In the namel of
certain of the lunar monthl. After this yellr had been In ule for
lome lime. the cl..11 yur WllS Inaugurated, with III bealnnlna
marked by Ihe hellacal riling of 5othls. SIIlI later. the lunar year
concurrent with the clvll ycar wal developed. By the end of the
New Kln&dom. or ponlbly rater, the original lunar calendar fell
entirely out of Ule and wal luperseded by the later lunar calendar,
which ""al probably present In the Middle Kingdom and eerlalnly
present In the New Kingdom.· AI Parker notel laler In Q/eodars.
p. 74, n. 17, "In the very same contellt Borchardt quallfled hll dear
ltatement that the orl&lnal lunar Yellr be&an with the flrsl month
afler the rllln& of 50lhll by placing III be&lnnln; 'around' the
riling of 5olhll, or 'uound' the lon&elt day of Ihe year:
J, There II a hymn to Amun In the temple of Hlbll, datlnl
from Ihe time of Darlul J, In which Ihe moon as Amun'l left eye al
nlll.ht II laid 10 be ·rulu of Ihe Ilarl, who dlvld ... Ihe 'two tlmel'
-131-
(I.e., day and night). the months, lind the yeln: See H. BrulSCh.
Th~saurus loscrlpt/anum ~gyptillclI,.um. 2. Abthellunl 10r.oz, 1968.
unaltered repro of Leipzig ed. of 1885-911. p. SIt. Parker noler lhlt
pusage as one indlcattnt. the relationship of the moon to the yeu
ICalendar" 32). The only reeron I quote II here Is the posslblllty
that puhaps the dual form of "lime" or "perIod" (drllli or trwlJ
Indicated here rather peculiarly by the addition of 1.... 0 .olar sign,
surmounting the normal plunl form of three Ilroke. (~:f ,e,)
instead of the common dual form of two Slroke. might be an error
for the form wllh one lolar Ilgn lurmounllnll the three .troke•. If
10. then It would bream.... plural form for "seasons" and ....e could
Iransl.le the end of the plluage as • .... ho dlvldu the seasons, the
months, and the yean: and Indeed that Is ho.... Brugseh translated
It. But In any cue. the hymn Is 100 laic 10 be of any slgnlncance
as evIdence for the old lunar calendar.
4. J. Cerny. "The Origin of the Name of the Monlh Tybl:
ASAE. Vol. 43 (19431, pp. 173-81. CHing A. Erman.
"Monatlnamen aus dem neuen Reich: Z)(S. Vol. 39 (l9OU. pp.
128-30, he says (p. 173), "Only more recently has It been possIble 10
trace thele month-namu back to aclual names occurring In bUllneu
teKtl of the Ramenlde period: He also mentions the earlier
monlhly festival names of Ihe Eben calendar and those of the
Calendar of Lucky and Unlucky Days that Abd el-Mashen Baklr
....al later to edit, The Cairo Calendar No. 86637 (Cairo. 19661. p.
54, Veno XIV. The usC' of the month name Instead of lis number
In a rC'gnal date occurs only latcr. An Inscription on a statue of III
(an offlclal of the reign of KlnS Shabah of the 25th dynutyJ In
the British MusC'um (No. 244291 may be the first recorded rC'gnal
date In .... hlch a month (Paynl! Is named. See T.G.H. James and
W.V. Davies. Egyptian Sculpture (London, 19831. p. 53.
5. I remind the reader of the followIng more precise values,
The Julian Year Is 365.25000 dayl. Becaule of the aSlronomlnl
eondltlons during the whole Egypllan period, the 50thle year Is
virtually Identlnl In length to the Julian year !the total delay of
the riling of Sot his bet ....een 4231 B.C. and 231 B.C. Is aboUI one
day, after that time the progrcsslon of delay accelerates 10 that by
1926 A.D. an IIddltlonal delay of about 2.64 days hal occurred, see
also the data mentIoned In note 7 of the Introductlon to Document
111.101. The solar, natural (better, troplcall, year, I.e., the Interval
between two consecutive solstices. Is 365.24220 days. The
difference bet ....een the Julian and tropical yun Is 0.00780 days
-132-
NOTES-CHAPTER THREE
p<!' year. It II this dlrrerence thai the Gregorian reform of 1582

sought to accounl for by dropping len days from Oclobu and by
femoyl", the extra leap year day when .. century year II not
dlylslbl.. by 400 ....lthout fractional remainder. Incidentally the
tropical year I, shorler lhan the .ldcrnJ year of 365.25636 day.,
which hllter year I. the Interval biol........ n consecutive conJunction.
of the earlh and the ,Un .... Ilh .. fixed Ila... The difference
between the tropical and the sidereal yun rnulls from the
procession of the equinoxes, which amounts to .. displacement of
the equinoxes In the sidereal year of about I day In 72.5 yean.
6. There II one In.crlpllon recorded In the relln of Ptolemy IV
at An... n th.,l seemS to 'p".k of the procession of Irl,"Sothl.
throulJ,h • tot.l of 730'360'12·1/2 fl.e., 1102 1/21 yearR, ....hlch
accordR ....1011 ... lth the assumption that Ihe currenl 50thlc cycle of
....hlch the Ptolemale perIod ..... a part belJ,an In the quadrennium
1321-18 B.C. of the Julian calendar. This strengthens the yle .... that
the ancient Egypllans, al least by Ihe Ume of the PtolemaIc perIod,
kepi an accurate record of the annual 50thlc risings oyer long
perIods of time. See belo..., Document 111.10, SectIon 5, and Fig. 1ll.1
as "'ell.
7. Sec Puker, The Calendars of Ancient Egypt, Chaps. I and
Ill. and his reply to Gardiner. 'The Problem of Ihe Month-Names.
A Reply: Revue d·Egypto/ogl... Vol. II 119571. pp. 85'107. See also
Parker's concise accounl In "The Calendars and Chronoloay: In
The L~.cy of Egypt, 2nd I'd" edited by J.R. H.rrls {London. 19711,
pp. 13-26, ...here all thr« calendars are dlKussed, the old lunar
calendar, the clyll calendar. and the 'aler lunu calendar. By lis
very conciseness. this account Is the cleuest of his seyeral efforll
to summarize calendlrlc deyelopmenll In AnCient Egypt.
8. c.l..ndan. Chap. I..... here he gives a hlRtory of earlier
lrealments of the beginning of the month and a detallfll discussion
of the Egypthw names of the days of the lunar month, Cr.
Document m.b belo.....
9. Ibid., pp. 9 and 70 fnotes 7, 8. and 91, clling trealments by
Brugsch and Mahler, and SethI'. Sec also E. Mahler. Eludes .ur Ie
ca/endrler igyptlrn (Anna/rs du Musie GuJmet, BJbJJDth~ue
d"itude•. Vol. 24), Paris. 1907, p. 7.
10. Parker, C./endar•. p. 72. n. 49. and 'The Problem of the
Month-Names: p. 91. For the list of fents In the Benl Hnan
nlend.r, see K. Sethe, HlstorJsch-bJographlschc Urltunden de.
Mlttler..n R ..lchu. I (·Urltunden VlIJ (leipzig. 19351. pp. VII 29-30,
-133-
and Document 111.1. -rhe Middle Kingdom" (Section n.

II. C.kndars. pp. 32-34.
12. ParkC'r, Caknd.n:, p. 34, -rho; right half of • tabId of the
ht dynasty...be..rs the rlgllre of '" r«umben! eow lIsls-Sothld which
h.1 the 11&" for 'ynr' .... \Ih • stroke between III horns....
Underneath Is ••lgo whleh Is apparently Ibll and the whole un be
plausibly rnd 'SothlJ, the opener of the ynr: the Inundallon': But
laler lin O. Neugebauer .nd R.A. Parker, Egyptian A,tronomlnl
TCKts, Vol. 3 (ProvIdence and london, 19691, p. 2011 he .C~ptl the
condullon of Godron that the co... is Sekhet-Hor (sce foolnole 14
below) nther than 5othl•. The casc for Identifying the glyphs on
the tablet with Sothls and the beginning of the ynr ...,U lucclnctly
made earlier by L. Borchardt. D~ Anna/t!'n und die ~ritllche
Fntkgung de~ Alt~n Rekhn der )(gyptf$Ch~n Ge$Chfcht~ U~erlln.
19171, p, 53. n, I, He concludes In this note, "Mlr schelnt es, d.n die
Insc::hrlrt bedeuten soli. dan d.. Hundnternneujllhr zum enten
Male (der Sirich neben dem Jarhreszelchen z ....lsc::hen den H6rnern
der Kuh?1 In den z ....elten MOn.1 de. W.ndeljahre. flel, .!so
d.nelbe, ..........Ir oben au~ dem 'ersten Male de. Fe.leI der
Zettordnuna' abaelesen h.lten:
13. In order to m.ke the best c..e for the nrller Interpret.tlon
of thl' t.blet lthouah as I later sunest there Is scarcely any
support for the !nslc tdea thai the t.blet represent•• reference 10
Sirlu. and the apenlna of the yurl, J have lert out .ny reference to
the .o-called month slanl, and J have deliberately Ult the
translation In • form that SUlaelts thai the t.blet .....s II marker
that dated the particular obJe<:t to .... hlch II "'al allached to the
year of the kin,'. relan .... hlch ...... named from a s".,cl.1 actlvlty of
Ihe kina. n.mely from his first celebratlon at Ap of 5othl. . . the
opener of the year. Presum.bly he ....ould celebr.te the co'"
aodden In othu ynn, but th.t thll "'as the "flnt" year of his
dolna '0. Examples of such t.bletl are nOled In Vol. J, Chap. I.
The name that I have sunested doci not mean th.t Sirius actu.lly
nne In Akhel al the tlme of Ihe celcbratlon but merely that 50thls
as the h.llowed opener of the Year In Akhet ....s celebraled by
Djer In Ap for the nnt time. The farm I. clearly like that
recorded In the Annals translated .. Doc. 1.1.
14. G. Cadron, ""tudes lur l'Epoque Archalque: OIFAD, Vol.
57 (l958J. pp. 143-49. full arUcle. pp. 143-55. His evidence Is
persuallve and thus makes II qulte doubtful that the lablet Is •
reference to the 50thls as the opener of the year. SUII, If the older
-134-
NOTES-CHAPTER THREE
Interpretallon I. correct. the" this t.blet ....ould tell u. eUher that

the civil calendar of 36S days .... lIh rome lie to Ihe rlslna of Sirius
.. lis one-Ume Ne .... Year', Day .....as In existence url)' in the nr.'
droll.l)' or thai the lunt-,teU.r cel"nd.r proposed by Puker ......s
being used at Ihe time of Djer. ~e al.o W.M.F. Pelrle, The Royal
Tom6:1 of the arllnt Dynntles, Pari II (London, 19011. p. 22, .... ho
makes no mention of the possibility that this tablet II concerned
....lth the openl"a of Ihe ynr. Regardlnll it. mnnlng he u)'.'
"They IIhe 'wo rngmenl. of the lablet) leem to name 'Halhor In
the manhe. of Kina Zer. clly of Oep,' or ButD....The flau,,, of
Hathor wllh the fealher bel........n the hornl II alrudy known." So
the upright Itrokes bet.... un the horn. "'ere for Petrie and ",1'0
GrlrfUh .Imply the sian for "fnther" rather than a urlpped p-Im
branch (for "ynr"J with an accompanylna vertical ,troke. Grirrith
in the Petrie volume, p. 49, read, the glyph for the city not as
"Dep' but as "Ap".
IS, Calendars, pp. 41, 45. The earllC'St source In which Wp
rnpt I, dearly the twelfth month Is the astronomical ceilina of
Senmut (see Fla. 1lI.4), which Parker believed repreRnted the old
lunar calendar, The latC'St ,ource I. on the w.lI of the Temple at
Edfu. See H. Bruasch, 'Eln neue. 5othl.-Oatum: Z){S. Vol. 8
(18701. pp, 109-10, In between Is a geographical fraament from
Tanl. In which Ihe lut month of the year (IV rmw) is called Wp
rnpt isH Fig. illS),
16. Parker. Calendars, pp, 33-34, give. a serlC'S of paslage. In
which the meaning of wp rnpt and prt Spdt .re equivalent. But
aranllna that equivalence doe, not mean that the reference Is to
the lun.. r calendar since In every case the reference might be to the
New Year', Day of the elvll year when the rlslni of Sirius .. nd
New Year's O.. y were colncldental at the bealnnlng of a 50thlc
cycle. a coincidence that continued to be celebr.. ted on the New
Year's nay of the civil ynr, long ..fler the Inilial coincidence of
the rising of Sothls and the firu d.. y of the year. Obvlou,ly at any
other Ume the ..clu..1 rising of 50thls would not be the New Ye.. r".
O.. y but mention of II In the phrase "opening of the year" could
refer to the Inlll..1 colncldenn .. t the beginning of a cycle, an event
worth keeping track of for season.l celebrations,
17. Parker, 'he Problem of the Month-Name.: p. 93. Cr.
B.. klr, The Cairo Calendar, Pla,e lA, line 2. The text actually ,ay.
"Wpt rnpt pw lib snnw:
18. Baklr, The Cairo C.fmdllr, p. 61.
-135-
19. &~. m~thod~. ~t ri.ultat. d~ I. chronoJ~Je ,-yptkflM

(Paris, 19261. pp. 112-26.
20. Puker, C.knd.,.s, pp. 34-36.
21. One piece of .o-nlled evidence for the elolblence of Ihe
toluca',,,)' month .:lIed by Parker le.lemurs, p. 43) I. Ihe
appear.nce of the Tholh symbol of • cynocephallc: bilboon re.ted
on a d,Jed column In Ihe ..cotnl dividing penllton of the penultlmate
lower r"ahler of the R.messcum ..Ironomlc.' uLllns (Fla. 111.21.
Now there .re thirteen dlvislonl of Ihe ceilina. twelve of ""hlch are
.nlgned to month. on the ...cond upper rcalster; the thlrl~nth
dlv1l10n In Ihe middle of that rell.ter contains nO mo... th
de.lln.tlon but I. empty. And rlnce that division Icp.,.tec Ihe
twelve month. of • ynr I. ynr which Parker by comp.rlng II with
Ihe yur liven on the Se:nmut celllng .,urt, I, the old lunar yurJ
Puker eono:ludCl that the blank dlvl,lon II th.t of the lnternl.ry
month whose n.me Thoth i' confirmed by the Thoth .ymbol In the
lower regl.tu. But,.1 I Ih.1I remuk later, the month .nd senon
dulgn.tlonl In the second regllter ue limply those of the civil
ulend.r and furthermore there II .nother expl.n.tlon for the
appear. nee of the Thoth Iymbol In ito eentr.1 polllion In the lower
regllter, n In H, .ppe.r.nce over the lpoUt. of water doc", It
merely .ymboilul Thoth·. role •• the Invenlor .nd p.lron of Ihe
dctermln.Uon of time.
22. For the ordlnuy Thoth-Feast on I Akhet 19, al.o ICC S.
Schott, Altlgypthche.Felltd.ten, In Abdemle der W/$IIen«h.ften
und der Llter.fur In M.lnz, AbMndlungen der Geister und
Soz/.lwlllRnsch.ftlichen K/a$lle, J.hrg.ng 1950, Nr. 10, p. 962.
23. C.rdlner, "The: Problem of the: Month-Namu: p. 23.
24. Ibid., pp. 27-28.
25. Scholl, Altlgyptlsche: Fntdaten, pp. 890-91, 959-60. See:
.Iso Baklr, ed.. The Cairo Caknd.r, p. II, where for I Akhet I of
the Incomplete clliend.r we re.d, "O.y I, Second Fent Being the
OpenIng of the Ye.r...Fealt of Ollrll; Fealt of hll: Fent of every
god; FUlt of Sobek..... The .uthor of thil nlendar m.y well be
confusing the full. of the epagomenll dly' tin five of these d.y.
being conlldered ••• kInd of extended New Yearl Eve: ICC Schou,
pp. 886-87) with the ordln.ry rent or fentl of New Ynrl O.y.
Note Ih.t In the lecond e.lendar of the C.lro document, we find on
I Akhet I IIbld.. p. 13), "The Birth of Rf·-I;I.r.khte; ablutlon
throughout the entire land In the water of the beginning of the
High NIle which comes forth al frelh Nun" Thll latter Itatement
-136-
•
NOTES-CHAPTER THREE
may ....ell r"neet the Yery ancient view that Ihe )lnr war to begin
with the .udden rl.lng of the Nile. In Ihe many rcf"ren".... liven
for I Akhet I by Scholl, he 0;11.... Hat.hepsul's remark IUrl:unfhn IV
2611 thai her r.ther -knew that. crownlns on New Yur', Day
....ar good Ii.e., hvonbld: thus lending mrne support to Gardiner',
Idea thai New Year', Day war considered an appropriate day for
Ihe accession Clf the king, and .. sueh produced .. feul
dl.llngulsh"ble from other fealts on that day. For Ihe "oonection
of Thoth with the reckoning of time. IU the eollecUon of hll
epllhets thai I hIve assembled In Volume One. p. 304.
26. Ie Set he, "Ole Z"ltrec:hnung der alten Xgypler 1m
Vuhllinl. zu der der andern Vlilker," N.chrkhlen von der
K6nl611chen Geullsch.rt du Wluenschaften zu G6ttlngen.
Phllologl«h-hlstorl$Ch~J:la~ au, d~m Jah~ 1920 (Berlin. 192m,
pp.I30-38.
27. Ca/~ndal7, p. 13.
28. O. Neui~baue~ and A. Volten. 'Unl<trsuchunp,en zu~
antlken ASlronomle IV: Qu~II~n unci Stud/~n zur (k~hlcht~ tier
Math~matfk, Astronomle, und Physik, Abt. B, Vol. 4 119381. pp.
401-02.
29. Pa~ke~, C./~nd.r~ pp. 15-16.
30. Ibid.. pp. 17-22. He conclude' (p. 221 by saylnp, that 'Of
Ihe enll~e HYenleen calculallon., nIne Ilndlutel month. Ithall .tl~t
on Ihe mornlnp, of old crneent Invl.lbillty, fIve on the day of neW
c~escent YI.lblllty, Iwo on the day In between. Ind one on the day
before the mornlnp, of InYblblllty. Thl. I. p~ecliely Ihe .art of
l~relular1ty that one "'ould expe<:t f~om a cycle scheme, ~lgldly
adhffed 10, In contnst 10 a method of strict obse~vallon for
slartlng the month ....!n my mind thue I. not the .llghtest doubt
thai the cyclIc calendar Wa' In use during Ihe pedod coYe~ed by
the lunar dues:
31. O. Neulebauu In hi. laler ,.,o~k A History of Ancient
Math~matlcal Astronomy. Pari 2 (Berlin, Heldelbera, and New
York. 19751, pp. 563-64, accepl. bolh Ihe Ylew of Ihe redonlna of
lunar monlh. from thO" lime of InYI.lblllty and Parker'. aenenl
t~ellmenl and dating of the 25-yur cycle. He does. however, add
Ihe follo""lna .tltemenl' "As "'e hlye ~emarked before, the dales
furnished by the cycle ag~ee with the bellnnln,. of Ihe Egyptian
luna~ month. In lhe fourth cenlury B.C. Since dales p~oYldcd by
Ihe cycle Ind Iclual lunar date. ,.,111 only .gr~e In the mean and
.Ince the dIfference. belween cycl~ dates and faCti wlll vary only
-137-
v~ry .Io....ly, one cannot exclude. date of orlsln of the cycle. "y,
In the firth nntu...)' B.C. If Ihl. dale ....er.. correct II would
con,tltute • curlou. parall"l to the c:onlemporuy development of
mathematical astronomy In Mnopolaml.. In "either case I. tkere
the .llghte.t lodlc.llon of Greek Or any olher forell" Influence.
Nevertheless U look. aI If the creation of the Persian empire
.Ilmul.ted intellectual life c ...ery.... here In Ihe anclent world of
which Ihe HellenistIc world was to become the heir: Regardlnl the
beslnnlnl of Ihe lunar month ""uh ,"re'unl Invl.lbUUy h" says Ip.
563, n. 31 that it -'I In .n probability C:lIund by the EJyptl.n
reckonlnl of the day from lunrlse.• procedure .... hlch In itself LI
mOil n.tunl and d~1 not re:qulre any •• tronoml".1 motlvaUon:
32. Cillmd.rs. pp. 24-26.
33. IbId.. p. 26.
34. Ibid.• pp. 27. 29.
35. Parker. Ibid.. pp. 27-29. rejeel. the schemn prnenled by
L. Borchudt and G. Whl:l:ll:r. W. Buta. "OIl: Jgyptlschl:n
Mondaten und del' 25-Jahr'Zyklus dl:. Papyrus earlobeI'I 9: ZXS.
Vol. 106 119791. p. 10 (full artlcll:, pp. HOl nSC'nually a,rl:e.....Ith
Parku and doubt. E. Hornun,'s opinion that Ihe 25-yur cycle
mll!.ht havl: alrudy Ml:n kno .... n In Ihl: 18th dynasty.
36. Ed. of K. Scothe, Ole .lllgyptf«hm Pyr.mldmteKle, Vol. 2
lulpzlg. 191OJ, UtteranCe' 669. Sut. 1961b-c, p. 472. Scol: alJO
refl:rencu to thl: I:p.gomen,,1 day. In the Introduction to Doc. lll.l
nn. 2-3, and 10 the document 1t.l:lf. Mid. Kln,d .. Il:ct. I, from Ihl:
tomb of Khnumhopll:p al Benl Ha.an and that of Amenemhet at
TheM•.
37. Xgypllsche ChronologIe IBerlln. 1904l; N.chlrlge zur
Xilyptlschen Chronologie (Berlin. 1908).
38. De die nal./I, ceo 18. 21. See Document IIlJO. Sut. 6, for a
translallon of the Latin teltl of Cen.orlnus· dl.cunlon. of thl:
Sothic year. There has Men a ,rut dul or dl.cuulon of thue
pa.sa'.... and I rcfu the ruder to R. lepslus. Ole Chronologie dcol"
Acogypur. Eln/coltung del" corrlCOI" Theil. Kr/tlt dcor Quco/Icon (BerlIn,
1849l, pp. 167-71: Ml:yu. Xgypllsche Chrono/oglco, pp. 23-29; l.
Borchardt, Quellen und Forschungen zur Zellbntlmmung der
Igyptlscht'n Gl:schlehle, Vol. I. Dlt' Ann./con und die ze/tllehco
Fat/egung des .lten Reichel del" 'gyptlschen Geschlchtco. p. 55; and
10 Wl:lIl. Chronologk l-gyptlenne. pp. 9-15. 59-60. 65-66.
39. 1 say "polilble quadnnnial datn" since thl: filled Sothle
yur Is 365 1/4 days and hence thl: ri.lng of Sirius markin, the
-138-
NOTES-CHAPTER THREE
Mllnnlna of • Sathlc period (and particularly of Ihe flrst Solhlc

period hen the clyll calenda........ made dependent on Ihe rlslnal
could h c taken place on lin)' one of the four rnn when It fell on
Ihe flul day of the ne .... civil )'('8t. The quadrennium Is u.... ally
ulled "Ielractcrls" from III Gr~k dc,lanation.
40. Meyer, J(gyptische ChronoJOSlc. pp. 42-44.
41. "Ole 8f:deutung.loslghlt de,. 'Sothllpcrlodc' rut die Sllcllc
SKYptlschc ChronolOlle: Act. orknl.II., Vol. 17 (938). pp. 169-95,
and "The Origin of the Egyptian Calendar; Journal of Ncar
ustun Studlcs, Vol. I (1942), pp. 396-403. Both artlcln ....ere
reprinted In O. Ncugcbauu. Aslranomy anel JI/#ory. &1«tN
any$, (Ne.... York/lkrlln/Heldelbrrarrokyo. 1983J.
42. "The Orlaln of the Egyptian Calcndlr: pp. 396-97.
43. Neugebauer', o... n vie.... of the Introduetlon of the elvil
calendar ere epltomlud a' follow. (Ibid.. n. 3. pp. 397-98), "The old
.tory about the 'crullon' of the Egyptian calender In 4231 B.C.
un now be eon.ldered ;os definitely liquidated. An objection hal
been raised agaln.t my theory of a 'Nile-year' resulting from
averaging the ,trongly fluetuatlng Interval. between the
Inundation.. Thlt objection II that there I. no proof of the
elodstence of 'Nllometerl' at 10 urly a perlod.... However no precise
Nilometer I. requIred for my theory. The sole requirement I. that
comebody recorded the date when the Nile was cludy riling. As a
matter of fael. ev"ry phenomenon which occurs only once a year
Iuds to the Same average. nO matter ho .... Inaccurately the dale of
the phenomenon might be defined. The averagln& proce.. of a few
year. will automatically eliminate an Individual fluctuations and
lnaccuraele. and re.ull In a year of 365 day •. Fraction., however .
....ould be obtained only by much more e"ten.lve recording and by
accurale calculallon. The aclual averaging must. ho ....ever. be
Imagined .. a very slmpl.. proc.... baled on th .. primitive counting
methods as r .. fI ..cl..d In the Egyptian numb.. r slgnl' the ..lapse of
one. two. or three day. recorded by one. t .... o. or three .trok....
After ten .trokes are aecumulated. they are replaeed by a ten-.lgn.
therufter t.. n t.. n-slgnl by a hundred symbol. etc. ThIs Is th..
....ell-known method of an Egyptian calculation.. Thl. method
finally reducel the proc.... of averagln& to the equal dIstrIbution of
th .. few markl which are beyond. say, thr ..c hundr..d-llgns and five
ten-.18n" In oth.. r word., ther .. I. no 'calculatlon' at all Involved In
d.. t.. rmlnlng th.. av..rag.. l..ngth of th .. NII..-years. Of cours.., ......
ne..d not even a•• ume th.. proc.... of countlnl!. all th.. sIngle day.
-139-
every yeu: the aver_ling of the execsr num~r of days over any
interval of conlt.n! lenllih lny twelve lunar months} ,II/n the
lame rcsult. This equal dlstribullon of counllng-marks finally
mahs It dear that no fraction .....Ill be the relult of the
process....lpp. 401-02, n. 17), When I revle .... ed thlt c:onten! of lhl.
paper al the meetlo. of the American Oriental Society In Botlon.
Profusor H. FranHort ..ked whether the Instltutlon of the
","hematic calendar c:auld be a..umed to belonl to the relao of
Djoser. I think that no "rioul objectlQn can be ral.ed a,alnsl luch
an allumption. because the only eondltlon for the nUlion of the
schematic calendar Is • lufflclently well-organized and dCYeloped
Konamlc life. On the other hand, muns 10 determine luch • date
by a.tronomful cansidcrllUanS do not exlsl. The problem of the
inl/entlon of the .chematlc month. mUlt nat be confulll!d ...lth the
problem of the period In which the 365-day year wa. Introduud.
The t....o in.llIullon. ue IlbsoJutely Indepo:ndcnl - lit leal! In
principle. The 365-day year mu.t have been created al a period
... hen the Inund.Uon ..oln..lded rouahly ....lth the Rasan ...lIed
'InundatIon'. Su..h a ...,lncidcn"l! hl!ld for thl! unlurll!" Ilraund
4200 and aallin for Ihe cl!'nlurle. around 21KXl. The lalter date
Ii.e., the time of DjolerJ h.. been "onlldered by Winloc:k...•• the
dalc of Ihe deflnlle establllhmeni of the ElJ,ypUan year: At this
point Neugeb'auer refer. to the paper of A. St:huff. HI.torl«he
Z"ft«hrJft, Vol. 161 (J9391, pp. 3-32.....hlch araun Ilgalnlt the
earlier date.
44. C,'"nd.,... Chllp. IV.
45. C.Jl!nd~rs, p. 53.
46. The'l! posllble datc. for the Inaugur.tlon of the sc:hl!mlltlc
year are b..ed on the coincidence of the hellacal ri.ina of Sirlul
and the New Year'l Day In 2773 B.C. For it II el/ldenl that if the
Ncw Year'. Day of thc cll/ll year .dl/anced 41 day. (I.e., II • 30
day.) It will hal/e beaun that adl/ancement 164 11.c.. 41 X 4)
calcndar yur. earlIer. thai II In 2937 B.C. But If iI adl/anced 12
dayl. It began Ihat adl/ancement 48 (I.e.• 12 x 4) yun euller. that
II In 2821 B.C. But an of thll resll on hi. ,uaaelted mechanllm of
lnterCllllltlan or II thlrtecn month, I.e.. lu..h II month ...... lidded
.... hcn Sirius rOSI! ....ilhin the laot Ii day. of thc twdfth month.
Parker flnd. lupport for hIlI/Ie..... by IInllly'Zlng the 59 dellles that
occur flnt on the baR of II relalll/cly late .tatue of Mut of the
22nd dynlllty lIbld.. 54-561. Ue .uggclli thai Ihe 1111 rCl/ul1 the
dual character or the concept of 'year: R.I(. Kuuss.....haR
-140-
NOTES-CHAPTER THREE
dllurl.llon 1. mcnllon£d belo..... In notc 49 and dlKuned In

Document 111.10 Unlro. n. I and Se<:t. J of the Documentl. on pp.
24-29 of that dissertation dlscu"" Parku'l theory of the origin of
the 365-day ~.lend.r and, on p. 28. find. It 'contradlclory and not
pracllcable:
47. P.V. Neulebauer. A.trcmam/St:he Chronologie, Vol. 1
IBerlln. 19291, 149-55. 160-62; P.V. Neugebauer and L Borchardt.
"S£'Obachlungen de. FrUhaufga"v dCI SlrluJ In Kgyplen 1m J.hre
1926: Or/Mt.lJst/sche Lltullturultung. Vol. 30 11927), ct. 441-48.
Su .1.0 the brief, but nut lummary In E. Hornuna.
U"tuluchungen zur Chronologie und GeschJchte drs Neuen
Re/t:hes (Wlcsb.den, 19641. pp. 17-21. Conlult allo Weill. &'.1,
mEthode. d rlsu/t.t. de I. chronol06/e igypticnne, pp. 189-204.
48. W.F. Edaerton. ·Chronology of the T ....elfth Dynasty:
JNES. Vol. 1 (942), pp. 308·09 (full artlde. pp. 307-141.
49. W.C. Hayes, ·Chapter VI. Chronology: The C.mbriclge
Ancient History. Vol. l. Put I IC.mbrldge. 19701. p. 182 s.y. -rhe
date In question II.e.• the 50thle date] I. preserved for u. In a
temple papyrus from El-Lihun and Un ....lth great probllbillty be
plnpolnled to 1872 B.C. by referenee to four lunar dates eontalned
In doc:uments from the same archive: As I holYe said, It w",.
Parker. C../encliJIrs. Excunus C, p. 66. who used the lunar dates to
fix the 50thle date as July 17. 1872 B.C. tJullan Yearl. Parker's
eonduslons have been disputed by R.K. Krau.., Probleme cle•
• ltlgyptlSf;hrn K./rnclers und der Cho.noJoglr des mIUJer.:n und
m~uen Reiche. In Xgypten lOin. Berlin. 1981l, ....ho u.ed .lx lunar
references and proposed 1836 i l l the year of the riling. having
a..umed th.t Elephantine ...... the site of the observ.tlon. For a
.ummary of the arguments of both authon. ue Document 11I.10.
note 3. A method of deddlng bet....een Puker .nd Krauss.....hlch
rests on original proposals by L. Borchardt .•nd .... hich (....lthout my
yet bt.lng able to eumlne the evidence) I .:annot properly evaluate.
has been proposed by R.A. Well•. "Some A.tronomical Reflections
on Parker'. Contrlbullons to Elyptlan Chronology: In L.H. Luko,
ed., £gyptolog/c'.1 Studies In Honor of Rlch.rc/ A. P.rl:er lHanover
and London. 1986). p. 170 (full artlde. pp. 16S-7l), "Nevertheless
there Is a potentlally very Important astronomical means available
.... hlch may nol only a..lst scholln Ln choosing bet....een Parker's or
Krau..·• d.tes but also provide a firmer network of dates for a
number of olher dynastlc periods. 1 refer to Ihe possibility of
computing the foundation dates of certaLn tempi........hose principal
-141-
axn may have bee" aUaneod towud specific rl.lnl or KIll"1 Itars,
by using Ihe equations of preCflslon which ilt;COunl for motion of
the earth', axl••bout Ihe North Celestial Pole: Well. note. the
utle of hI. prcllmlnu)' Invnllgallon of the IUuclted method.
Vcry Ilronl support for Ihe usc of s'a.. rlsln'l or tra"slll
with u:aclly m"iJ.ur"d temple .Uaomenl. 10 determine foundation
d",tn II found In the more recent Stud/cn zur l/lypllschcn
AstronomJc (Wlcsbadcn. 1989) of Christian Leitz, who usn this
technique and other ..lranomlc:.1 c:onsldenllon. Independent of the
conYentlonal employment of Ul the hell.cal rlslna of Sirius on liven
days of the civil yur In yulous reigns and (21 known lunar and
clvU-year double dates, lh.l 1_, Ihe main techniques I dncrlbe here
In this volum~. J not~ thaI Leitz bell~"~r thai h~ har rhown that
th~ r~cent effortr of Krau$I and otherr to loat~ th~ place of
obrer"atlon and thul th~ cal~ndar referenc~ point at AI....an .Ire
Incorr~ct (p. VIIl, "DI~ "orl1ea~nd~ Arbelt brlnal 8 "on~lnand~r
unabhlnala~ ntronomlsche Araum~nl~ fUr Unlu'aypl~n, soda" In
d~n Augen der VerfaSRrr dar Probl~m dec "_lend~rbezu&lpunkt~1
all !!.elllst errchdnt -nlcht not ....endl!!.er....elle aber du der absoluten
Chronologie Alt1ayptenr: My flrrt reactIon Ir that I cannot aarM
....Ith all of hlr "ariOUI arrumpllonr and analYln of pertinent lextr,
but I rhould note thai I had completed my "olume when I finally
obtained a copy of Lell",'r book. Hence to be fair. I mUlt Iea"e my
erllmate of the roundnerr of hlr work to a future date. I do nole,
howe"er. that my late colluaue Olto Neup;eb.uer "ehemently
r~..Iected Cfor the morl partl the "alue of templ~ aUgnment data for
there dllcurrlonl,
50. "On the Chronology of Ihe Early Elahteenth Dynasty
CAmenhot~p I 10 Thutmor~ lin: The American Journa' of Semitic
unguagn and Literaturel, Vol. 53 119361, pp. 192-93 Cfull article.
pp. 188·971.
5!. Unt~rruC"hungen zur Chronologie, p. 21.
52, R. Leprlur, DIe Chronologie der Acgypter eBcrlln. 18491, p.
lSI-56, 165-67; H Bruasch, MlI,l-rJ6ux pour serylr. la constrllC"lIon
du ca'endrlu del anc/enl l-gyptlenl ILclpz1a. 19641. pp. 28-29.
L~prlur arau~r from Hor_pollo'r r~ference to Ih~ EaypUanr' "year
of th~ four y~lIrr: .... hlch he CLepslur) takes 10 be Ihc quadrennium
of thr"", yean of 365 daYI and one of 366, and from Ihe referencer
In the EXnl Haran calendar 10 the ferUyal of New Year'l Day. 10
that of the "EX"lnnlna of th~ Sun Year" lhtr tranllallon of "p
rnptl, and to the ferll"all of Ihe areal and Imall yurr lice Doc.
-142-
NOTES-CHAPTER THREE
111.1 belo....l. thai the Earpll.nl kne .... of the quadrennium of

Intercalary day. and thus had stde by side the clvll year of 365
days and the Sothlc: rur of 36S 1/4 days,
.53. Edlcrlon, ·Chronoiozy of the Early Eighteenth Dynuty:
pp. 191-92. For the pertinent arllde by Lepsh.... see the
Introductlon 10 Doc. 11I.2. ODie 4.
54. Edgerton. Ibid., pp. 190-91.
55. Parker, C.fMd.rl, pp. 38-42.
56. If .... e .tudy FIg. IIUO we lee under the a1yph _ In " ' _
enough Ip.ce to IUIKell that ,hal determinative _ ....as to be
undentood In the nelll Une and Indeed In mOlt IInel that luccew lI.
Note that In the lIsl of Fig. 11I.6.. We see mentlon of the Feast of
Re or Re-Ibrakhtl, celebrated on the flr.t day of the first month
rather than II' th.. flnl day of the twelfth month Mesore. This of
course become...qulvillent to hilvlng the FetUval of wp rnpl as the
nut day of the nut month rather than as the first day of the
t....elfth month when that month Will known liS wp rtIpl. It is the
latter fealt day which II present In the Eberl Calendar. See Weill.
Chronolagle t-8yptlenne, pp. 114-15.
57. Chronologie /-gyptlenne, pp. 8, 127-33, 145-58, 175.
58. Bruglch. Thesaurul In«rlptlonum. p. 363a ('my Fig.
111.931, cr. R.A. Parhr. "Sothlc Oat" and Calendar 'AdJustment':
Revue d'Egyplologle, Vol. 911952J. p. 103.
59. As In Volume One I use the text of K. Sethe. DJe
.ltSgyplJsC"hen Pynmlelenlexte, Vol. ) ILelp",lg, 19081, pp. 145. 264.
60. See K. Sethe. Obersetzung unel Kommenlar zu elen
a/t/igyptlsc:hen Pyramlelentexten (Hamburs, 1%2), Vol. l. p. 276:
Vol. 2. pp. 383-84.
61. While the nam" of the d«:ans of these Egyptian IIU
cloch provided the sourct' for the d«:ans found In Hellenistic
zodiaCI, the laltt'r no longt'r refer to the same temporal decans but
simply refer to dlvtslons dividing the ",odlac. I shall dlsculs the
relallon of the Egypllan decans with ",odlacs below when dllcusslng
the later astronomical monuments.
62. Parker. "Ancient EsypUan Auronomy: pp. 53·54.
63. Iblel.. p. 55. For a detailed and lucid dllcusllon of the
utronomy that lies behind the diagonal Itar doch, see Neugebaut'r
and Parhr, Egypl/.IfI Astronomlc,,1 Texts, Vol. 1. Chap. 3. and the
somewhat nrllt'r IIccount by Neugebauer. "The EgyptLan 'Deeanl':
Vistas In Astromomy, Vol. I U9551, pp. 47-51.
-143-
ANCIENT EGYPTIAN SCiENCE
64. Neuacbauer and Parker. Egyptian Astronomical Tcxt6,

Vol. l. pp. 29-32. The authors trnt nch of the 'welve eorriol In
detail, m>tlnl the n;ome and Iltles of the deceased. the dale. the
provenance, prel.mt loc8tlon, pnvlour publlnllon and dlscunlon.
IChemllie dlilaram of the arrangement of the dK8dn_ and crUleal
apparatur. Platn are liven at the end of the volume. In my
account here I have Ignored the decan lisl from the crlllo. of the
tomb of Senmut {which ....1lI be treated rater In the chapted and the
le'llu..1 (raiment from the CQlmology of Sell I and RamnleS IV
(Ibid•• pp. 22-23 and 32-35). For extracb from the Jailer, ICC
Document 111.12 below.
65. Ibid.• p. S.
66. NeUlebaul'r and Pnker, Egyptfan Astronomical Texts.
Vol. I, pp. 2-3 and 23-26: S. Scholl. -Ole .ll.ligyptll~hen lkbne: In
W. Gundel, D~ltan~ unJ lkbnsternbilder, Eln B~Jtral zur
G~s~hi~ht~ dcr Stcrnbllder dcr Kulturv"llt~r IGlUckstadt und
Hamburg. 19)6). pp. 1-21. Neugeb.uer, before hi. de.th, lold me
th;ot he had flO ~oflflden~e In the eh.bor;ote Identlfl~atloo. of Ihe
de<;;Jol proposed by R. BOker In his po.thumoul publication, ·Uber
Namen und Identlflzierung der lIgyptllcheo lkkane: Centaurus.
Vol. 27 (1984), pp. 189-217. BOker attemptl 10 locate the J6 decanl
by ~onsldering the positions of the full mooo at IO-day loteruls
and noUog thaI the Umlts between de<;ans were 10 many uses
exactly preserved 10 Greek altrooomy. But as I haye said at the
end of the ~hapter. the EgypUan decan-names .. used In Greek
utronomy had Utile conformity ....lth the star and constellation
de<;ao. used by th" ao~I"nl Egyptian Itargu'''rl to mark the hour-
of the nlsht. IloCe by the tim" of their use by the Greeks the n.mes
....ere merely u.ed to mark the IO-degree dlyi.lon. of the zodlaul
Ilso., aod the lielof the de~afl namel with the old deufl SUri uled
w"re lurely not determinable with any exactitude. The remark. I
quote from Neugebauer and Parker In the Introdu~llol'1 to
Document 1Il.17, t"xt to whi~h note 4 pertalnl, sUlIgest furth"r
why ex;o~t astronomical loc;oUons cannot be found from the
eyldeoce we pos""II.
67. Neugebauer and Parker, Egyptian Astronomical T~Kt..
Vol. l, pp. 24-26.
68. P;orker, ·And"nl Egyptian Allronomy: p. 56.
69. IbId.
70. Neug"b;ouer and P;orker. EgyptiJm Astronomical Texts.
Vol. 3. p. 213, glyn rehren~es to .tar obseryer- or hour-watch"r-
-144-
NOTES-CHAPTER THREE
In Dynasly 26 and In the Ptolemaic period. See .1,0 Document

111.18 below. and note the earlier references for the Middle Kingdom
In my flnt volume, Chap. J. n. 25. Consult also W. Heick.
Urtund~n des '.yptfst:h..n A/tutumll IV, Abl. Urtum/en drr /8.
Dynast/e, Heh 19 (Berlin. 19571, p. 1603. line 4: Heft 21 (958). p.
1199. linn 17-18. and p. 1800, linn 10-11.
71. L. Borchardt. "Elo alllSyptl,chu ulronomllches
In.trumml: ZJ(s, Vol. 37 (1899), p. 11 h"hole .rllele. pp. 10-171.
72. For • detailed lfnlment of these clocks. consull
Neusebauer and Parker, £gyp"." Altnmomk.' Texts. Vol. 2
(Providence, R. I.. and London, 19641.
73. Neuleblluer and Parker, Egypthtn Altronomfc.J Texts,
Vol. 2, p. 9.
74. Parker. "Ancient Egyptl.n Astronomy: p. 58.
75. Parker, "Eayp... n Astronomy. Adrolosr. and Calendrlul
Rec"'onlng: p. 715.
76. Neulelnuer and Par"'er, EgyptJ.n Astrorwmlc.1 T~~ts.
Vol. 2, p. 4.
77. Su the dlscunlon of the astronomlul considerations that
bear on the lengths of these hours and other aspects of the
Rameulde sur clock Ibid•• pp. 9·18.
78. IMi., p. x.
79. "Eayptlan Astronomy: J.nus, Vol. 52 119651. pp. 161-80
Ibut particularly, pp. 173-801. Bruins anumes a more hlah1y
sophlsUuted kno ledae on the part of Eayptlan astronomers of the
relationship bet een solar and sidereal moUons as they bear on "the
equation of Ume" than seems likely from the nature of the textual
remains of the clocks and of the few survlvlnl rather confused and
slmpllsUc panages dlsplaylna the technical, astronomical "'no""led,e
of the ancIent Eaypllans. In terms of procedures Bruins suaaests
the use of the so·ulled nst·decan. that 11 the use of the knuckles of
the extended nst as a slahUng "Instrument: a usaae never
mentioned. so far as I ... no...., In the ancient Eayptlan literature. His
araument leads to the conduslon that the slahUnas of transits ""ere
not made by t""o astronomers sllilna fKe-to-face, as Is usually
believed, but that only one astronomer was Involved Ipp. 173-741,
"Once a 'merldlan' for the transits has been flxed the nst·decan
aUo""s (one! to measure Ume differences of 4 min. from mldnlaht.
T ....o remarks are then to be made' flrslly that luse of) the
nst·decan Invites Il.e.. encourages the observed 10 (anume) ;I
division Inlo three domains left and right from the mld· ... nuckle lind
-145-
thus provldn u, wllh an InterprdaUon of the orilin of the ~"en

'Internal vertle_1I lines around land Includlnll thC' meridian In •
c::cnlnl line! in the coordinate syltem of Ihe liar clocks: and
sc<:ondly Ihal If such abse-rv.tlaru u" mlde their Interpretallon can
never be ,lven correctly nea1e<:!lna the lime equal10n al Inn
tran.lt. Ue observe<! .... uh • hl&h precision, an errOr of some
mlnulnl Consldenbly more ."..urate oburvatlonl can ,tut Il.e.• be
undertaken when) carried out by one utronomer. Here we should
remark th.t we completely disagree .... uh Neugebauer and Parker
lin their two-penon theoryl....1t seems obvious to us Ihat the
observations ....efe carried oul by One observer, .... ho. havlna at hi.
dlsponl the North-South direction has only to .Irelch hl. arm
forwud In that dlrectlon to be able to measure about 10 alonl the
equ..lorl In our opinion Ihe 't~rlet rIlure' or the stu docks II not
an ass/ltant Dr the observln" astronomer, but the ~.tronomer
h/m~/fI The p,alnter depleted Ihe sealed Ulronomer .nd what he
:lN1 11. Independently, drawn 'behind him' In the chartl,.. A Slronl
lndlcallon or the Independent representation or the sealed
astronomer .nd ·.... h.t he sces' Is contained tn the ract, lhat In the
descrIption Dr the Itar clocks, In the ch.rt the 'Iert' and the 'rlght'
.re JUlt 0ppollte to Ihe 'lert' and 'rlght' Dr the ·t.rget rIgure', The
painter plcturt'd tht' astronomer 'en race', then turned around and
plcturt'd what he sa....: II havt' added a rc.... braclrt'ted phrases 10
Imooth the rio.... or the Engllsh.l
Surt'ly this Is a pervt'rse theory to t'llpiain why tht' puts or
the body were used to mirk tht' tranllt I1nt'l, IlnCt' with u:tt'nded
rIlt the observer h.s no .ccur.te sen.e or .... here the lines crOll hi.
eyt'. or ear, or Ihoulder. And one cannot really Ult' tht'
non-perspectlve aspects or Elyptlan repre",ntatlon to explain the
contrary ",n",1 or right and lert on the grid and the human body.
ror the Elyptlan palnler had no dlHkulty In corrt'ctly orlenlln"
rJluru In the same or parallel planes. Startlnl rrom his
unsupported and Impl.uslble Interpret.tlon Dr the .llhlln&
procedure, Brulnl lOCI on to ellpl.ln how completely he dln&rees
with Neu&t'~uer and Parker, 'Where Iht'se aUlhors assume crude
procedures and til-conceived schemel, which allt'mpted to combine
conlradictory elemenl•• we, on Iht' contrary, conclude hom the
mere exlstenct' or the charll that lime lnlerval, or aboul a twelrth
or a decan-a. Ihe liar. are In the chari. pUI 'upon' and '~t"","n'
the coordinate lines-some sidereal minutes WlI're considered and
qulle a dlHerenl problem Dr observallon rrom comblnln& liar
'146-
NOTES-CHAPTER THREE
IransUs and seasonal hour Un at the base of the stu clocks. That
small partl of " decao were C'on.ldcred slaods 10 rUlon, If the
dlfh'rcnul ..ovcrnl by Ihe chari' werc more than. denn... the
nlronomer should have observed the stu of Ihe nelll or former
dennl The observallon. ue therefore undoubtedly meant to have
an accuracy of fome minutes of sidereal time. As we h,.vc alrudy
lndlcatfl:!. an aJ£umpllOn of • conslanl duration of twlUght Is
Identlcal with removing from observed facti Ihe change In dur.lIon
of the day and night In Ihe automatically con.tant dC'C'anal houn.
Moreover, true star po.ltloru "£ed to Imply Jlu/omat/cally the 11m.:
..quatlon and ;II Neugebauer and Parker neglected this lime
~u;.Uon completely their method of Interpretallon Is completely
ruined by their flrst Issumptlon Ind even If lhls were nOI the case.
1111 Is I"aln ruined by their nealecllnl the time equltlon:
80. See the InVoductlon to Document 1ll.14 Ind Neulebluer
Ind Plrker. Egyptian Astronomical TeJC/~. Vol. 2. pp. Ix'x.
81. For I detilled descrlptlon of both types. see L. Borchardt.
Ole I/tlgyptl~"e Zeltmenung {Berlin and leipzig. 19201. pp. 6-26.
To Borchardt's list of fragments of outflow docks {see next note
belowl mlY be Idded Ihose cited by Didier Devluchelle In W. Heick
Ind W. Weslendorf, eds.. LeNllon der A·gyplo/ogle. Vol. 6,
(Wlesbaden. 19861. ee. 1156-57 l"W..seruhr-J. Cr. Neugebauer and
Plrker, Egyptian A.tronomJr:11 TeNt., Vol. 3, pp. 12, 42, Ind 60.
See Iiso R.W. Sloley, "Ancient Clepsydrle: Ancient Egypt, Vol. III
09241, pp. 43-50. Ind by Ihe same luthor, -Prlmillve Methods of
MeasurIng Time with Speclll Reference to EIYP"- The Journ61 of
qypflan Archlt!OJogy, Vol. 17 (19311, pp. 174-76 (full article, pp.
166·78). Incidentally. in FIll.. 1lI.2Ib. I hive ginn conslde",bly
better photographs of the exlerlor, Interior, Ind edge of wlter
clock No. 938 II the Brltl,h Museum In London {Borchardt',
Ausllufuhr 61 lhln Borchardt'l lingle photograph of the exterior.
Note that Borch.rdt 11.1,," • drlwlnll. of those II.radullions of the
Interior surface of Auslaufuhr 6 lhal rem.ln (-my FIll.. 11I.2Icl. I
h....e 01150 .dded. In FIll.. 1I1.21d, exterior and Interior views of I
frlll.ment or I second wlter clock INo. 933) at the Brltllh Museum.
82. Borchardl, Ole altlgyptlsche Zeltmessung, pp. 6-7. On pp.
7·10 Borch.rdl descrlbes brlerly fnll.menls of 12 olher wlter
docks. See al50 the lilt given by Devluchelle In the article
mentIoned In the preceding note. See A. Wiedemann. -Bronze
Clrde:ll and Purlflcallon V"sels In ElI.yplian Temples: PSBA. Vol.
23 (June, 1901l, pp. 270·74 (full arllcle. pp. 263-741. where
-147·
fragments of Inrlow docb ue Interpreted.., purlfk::.Uon ycswls.

Sf:" the- descrlptlonr .nd depletlons of four of the already known
fragment" A. Roullet, The Egyptilln .nd Egyptlan/zlf16 Monuments
of Im~r/.1 Rome ILelden, 19721. pp. 145-46. 11,,11I nOI. 326-328.
330, lind rI". 334-36, 337-38. 339-42. 344. Incidentally. Dr.
Kad-Heln:fl: Prine, the Dir«tor of the Xlypllsche Munum
(Slaalllche Mu~n zu Bedln) t.,lI, me that he hll' no remalnln,
evidence of the clock wlih the Inventory nO. 19556, mentioned by
Borchardt III -Ausillufcr II (Berlin)" and by Roullet ., her Item
326.
83. Sloley, 'Prlmilive Methods: p. 175.
84. Borchardt, Ole ./l6ByptJsch" altmeuung. p. 6. Su .Iso
note 21. above. The representaUon of Thofh In the center of the
third resister of Fla. 11I.2 show. the Tholh b.boon .urmounllnl •
djed column, which I.tter h.. Ihe munlng or •...blllty· or
·endur.nc.'- .lind If p.rl Dr Ihe decor.tlon or the bottom or the
Interior or the K.. rnak dock where It .ppears belo... every other
monthly .e.le without h.vlng .nythlng to do wllh the dock'.
purpo.e or telllni lime. except In a very general w.. y. Thl. Ihen
expl.ln. It• • ppe.r.nce below Ihe b.Iboon In the R.meneum cellini
.hown In Fig. m.2. In FII. 111.23 deplctlnlJ, the lnrlow clock round
.It Edru by G. Mupero In 1901 we ue Ihe Thoth baboon Itlil
.lI.ehed to the venel. Cf. the Thoth b.Iboon prelenl on •
rr.r;ment or • w.ter dock datlnlJ, rrom Oynally 26 (Neur;ebauer
and Parbr, Egyptian A~/nHJom/cal TeKt~, Vol. 3, p. 43 .lind plate
22BI, where Ihe b.Iboon appears In Ihe .ame posilion In the bollom
reglfter. SlmU.rly the three model. or Inflow dock. presented by
A. POlJ,O, "E&ypllan W.ter Clock.: '~f~. Vol. 2S (19361, p. 416, Fla.
4, .lind pl.te 4 opposl1e p. 416 Hull .rUde pp. 403-42S) fef. my
FIlJ,. 111.321, III hive the Thoth b.Iboon, .. doc. the Cllro model
depleted by SIDley, 'Prlmltlve Method.: plale XXI.1. In the
HluoglyphlclI or Horapollo. cd. or C. Leeman. (Amsterdam. 183SI.
Bk. I. Clip. 16. We .Ire told th.t 'In their wlter docks the Er;ypU.n.
Clrve .I .uted dog-heed b.Iboon .lind they mlb water flow rrom
his member: G. Oaressy, 'Deux clep.ydres anllquu: Bulletin de
L 'In~lItul Egypt/en, Sme SErle, Vol. 9 (I'AnnEe 19\5J, p. 7, n. I Hull
.rllcle. pp. S-161, .tate~ that ".ccordlnlJ, 10 • lJ,cnenl belief In
.nUqully, the cynoceph.llc baboon urln.tel .It relJ,ul.r Intcrvils
twelve Umes • d.y: One would .uppou that the. merely means
th.t In expeilina water rrom the clock benealh the le.Ited b.Iboon,
the bllboon I. eHeclively 'urlnaUnlJ,' the lwelve hours of the day.
"148-
NOTES-cHAPTER THREE
Honpollo, Hi"l"Of/lyphica, 1.16. says that the baboon ".Ione of .11

"nlmall, during the cqulnO>o;C'I, crle. Qut eaeh hour Iw"l ..." tImes ..
day: But the lundry references to the role of the baboon In
announcing the hOUri do not Iholt the b.lboon'. cUltodlal role of the
houn 10 the equinoxes.
8S. SIDley. "Andent Clepsydra,,: p. 45.
86. Ibid. Sioley claim. In II noll' InD. SI to thIs pan"s" that
Honpollo 1I,16} 'slales that .. metal lube of narroW bore was fItted
to the "pcTtur,,: But It I. "vldent that Hor.pollo II describing ..
deYlce like .. spigot 10 control the flow of ""ater Into an Inflo....
clock Tath"r than .. spout fltle-d to the aperlure of an outflow dock.
a. the edlloT of the tnt, C. Leeman., IUlllt'lted In hi. descriptIon of
• model of an Inflow dock In Lelden (ree Pogo, "Egypllan Water
Clocks: p. 4161. H~n<;e I believe that Gcorse Boas' translallon of,
the p:usaSe hal II riSkt (The Hieroglyphics of Horapollo IN~w
York. 195OJ, p. 69), "In order that the water may not be too
copiou....a <;ontrlYanu ~)Clsts (to re!!.ulate Itl. Throulth this the
Wiler Is let Into the clock. not too fine a strum. for Iher~ I. nen!
of both ({Ine and broad!. The broader m.kes the .... ter flow
,wlflly and dar. not mealur~ off Ih~ houn pro~rly, And the
thinner strum lives a flo ... which Is too sm.1I .nd Ilow, (Hence!
they arran!!.e 10 loosen the duct by • halr's breadth and they
prepare lin Iron plug for this use, a<;eording to the thickness of the
slrum (required),"
87. Sioley. "Prlmltiye Method.: pp, 175-76.
88. Bor<;hardt. Ole altSgyptls<;he Zeltmeuung. p. 12.
Borchardt's table of measurements Is alv~n In millimeters and
ftncerbreadlhs. wllk approllimate erron In bolh me'Sures.
89. Sioley, "Anelent Clepsydrae: p. 47.
90. Borchardt, op. dl. In note 88. p. 13. Thl. I. Ih~ Cal~ for
Borch.rdt's oUlflow clock no. 2 and prob.bly for nO. 3 .nd no. 4.
Ineldentally In elock nO. 2 the aperture for the outflow Is 3
flnlerbreadths below the bottom of the 12-finlerbreadth .cal~.
91. Th~ number III I. a proper deduction from Ihe numben
.yallable from the fr'lments of those clocks. See Fla, 11I.26.
92. Thl. fragmenl Will flnl published by D.P. Grenfell .nd
AS. Hun!. The Oxyrhynehus Papyri. ParI III (London, 19(3), No,
470, pp. 141-46. II wu examined In great delail by Borchardt. op.
cit., pp. 10-12, wllh some chana'"' In transcription. Cf. Sioley.
"Anelent Clepsydne: p. 46. Borchardt. p, 14. sUllell. th.t the
.0-c.lled "sacred cubiU (ruler-like boards)" preserved In yarlous
-149-
temples_ which conl.ln a 'able' that may Indicate monthly varlatlon5

of .h.dow lenath' for ,h.dow cloch.•1'0 have' • table' that may
expru. monthly volumetrle c:hanr;el In the ...aUr content of
outflow ...aler clock••• they empty, althouah he hal no explanation
for the Ipeclrtc: munloa of the lum. of unll fnellon. thai
constitute the table.
93. See Borchardt, Oleo ./tlgyptlst:he ult_Uf>6. p. II, where
he lives the anales for the \/arioul doch. SIDley .Imply alvei the
dlnct anale of Inclination of the .....11 with the bue, whkh Is of
course merely 18QO mlnu. the angle liven by Borchardt, so that the
~.t anale of • conical clock would be .boul 1030, while the .clual
enaln of the extant clocks renae between 109" and 112° (See
Stalely, AnC'IMt Oepsyr/rae, pp. 45-471.
94. Borchardt, DJ~ .1t18yptl$Ch~ Z~Jtm~nun8. pp. 16-17.
revle ...s th~ vulou. dertntuon. of day and night hOUri. assuming
th~ follo .... lng poSllbllltl~I' (I) the daytlme bet.... nn lunrl.... and
.un.et. re.ultlng In the "elvll d.y: or (2) the nlghlllme hours
bet....een .unlet and lunrlse. resulting In the "civil night: Or (3) the
duration of d.yll,ht that Includel morning _nd evening t ....lII,ht. In
hi. terminology the "..tronomlul d.y· or fln.lly (4) the Um~ of
darkness after the lubtr.ctlon of morning and evening t....lII,ht. A.
Borchardt •• y•• th~oretlc.lIy ....e could me..ure .ny on~ of the.e
four dur.tlon. by a ..... ter clock. But .ctually ex_mln.tlon of the
clock fragment••nd their 11;.1.....ho.... 1 UI th.t lome Ire clearly for
nighttime In .ome h.hlon. n.m~ly clocks numbered 1. 4. 7. and 8•
.... hlch dllplly the .tu dec.n. In lOme form In the top exterior
register or explicitly .... y that the doc!:. Is to be us~d It nl,ht. In
.ddltlon.....e can further deduce for the K.rn.k dock (no. 11 that
the hlse position of the end of the .Ixth hour In the K.rn.k scale
....ould Immediately be evident If used In the daytime lI.e .• the
midday ob.ervatlon of noon ....ould be orr: see FI,. 1II.31J.
Borchardt .ummarlu. his vie .... coneernln, the Karnak dock as
follo.... 1 lp. 19), "Als .Icheres Ergebnll diCier Unterluchun, I.t .ber
hervor,etreten. d.n die Au.laufuhr I (K.lro) nlcht fUr bUr,crllchr.
IOndern fUr •• tronomllche Nlchte odcr Tage 1!'U gebr.uchen ....u.
O.SI lie nlcht fUr altronomllchc Nlichte und T.le 1!'U ver enden
..... r. erglht die .... ellen Er .... Jgung. da.. lie dann Ie die
ZUlimmen.teliung 1!'clgt. T.g und N.rht 'J;ul.mmrn 'J;ur Zeit der
WlnterlOnnen ende 'J;U kur1!' (bel .12 0 um 0.2Sh • IS'). 'J;ur Zeit der
SOmmerlOnnen ende 'J;U Illng Cbel -)20 um 0.2Sh • IS'} .ngegeben
hllle...... 'J;U elner unh.lIb_ren Venchlebung des T.ges- und
-150-
NOTES-CHAPTER THREE
Nachtanhng•• aim der Zeltpunkte. an denen die W;ossl!ruhr frisch

ZU fUlien war, hllte rUhTe" mUssen. Ole MeSiung und Tellung de•
••Irooombe",," Tages hat nun wenig Wert, wohl abcr 1st dl., de•
••lranomlsch,," Nae"t ""lehllg. WI. kllnne" .Iso yon den belden
Mligllchhllen ruhll die elne raUen laueo und behauplen, dass die
Au.l.ufu". I O::.lro) nur fUr ulronomllche Nuhl be,llmml war:
95. Borchardt. Die .ltlgyptl~hl! aitmnsung. pp. 19-21. The
helpful table 1 have glyen .s Fig. 111.37 was tahn from A. Pogo,
"Egypllan Waler Cloch: lsi., Vol. 2S (1936), p. 410. (Noll! Iha,
the: full .rllde occupIes p-gn 403-425J
96. The mont complete trealment of ElypU.n Inrlow docks II
thai of Pogo mentioned In the preceding note. As We .hall see. lis
most lnlue.llng section Ireat. of the grid diagram fo... nd on th~
Int~rlor .... rfac~ of th~ Edf ... clock, A I~u compl~t~ acco ... nt of
Inrlow c10ch II fo ... nd In Borchardt'. D/~ .Itlgyptln:hl'
allmenung, pp, 22-26, Altho ... gh th~y hav~ ~n .uperseded, th~
original artlcler flnt dercriblng the EcH ... clock by G. Dar~uy .,~
worth m~nllonlng' 'Grand VaI~ en pierr~ IV~C graduation.:
Ann.ln du ServIn du Antf'lullE., Vol. 3 (19021, pp. 236-39, and
·o..... x clep.ydr~. anllq ...e.: Bul/l'fln d~ I1nstltut Egypt/l'n, Sm~
.hl~, Vol. 9 1l91S, Cairo, 19/61, pp, S-16, Th~ ce<:ond of these!
.rtlcle. nJpened~. th~ nnt.
97. S~~ Ih~ dncrJptlon by Dar~uy, 'Deux c1~p.ydr" Inllque.:
pp, 6-7, 'L~ VI'~ ~.t ~nller..,11 re compore d'un fllt en calcalr~ dur
sen.lblement cyllndrlq...e de 0 m. 285 mill. de dlamllre ;l I. ban et
d'une hauleur de 0 m. 30 cent. Au-deslu. une moulure hll Ie lour
de I. partie .... pirieure, el .u IOmmel 10 m, 38 c~nt.lle diamltre e.t
pont. 0 m. 345 mill. Inlhl~... r~ment on a d'.bord ... n hldemenl
clrculalr~ d~ 0 m. 05 cent. d~ hlul~ur ~I 0 m, 225 mill. de
dlamelre, puis II lilrlJ,~ur du creux re rMuil il 0 m. 169 mm.•ur a m.
275 mill. de hauleur. Sur Ie caTfi, au nlvu ... du fond, un trou e.1
perd, lrav~r..nt II plrol el ltgolrement incllnfi ven Ie bas; II 10m.
005 mill. de dlametre il l'Intfirleur mil. r'fil.rgll Ju.qu', Ivolr 0 m.
OJ c~nt, ~xttrl~uremenl. L'accrollSem~nt d~ di.mo!tr~ n'"t pll1
rfgull~" II .ugm~nl~ brurquement prll de la lortie, comme .1 t'on
IVIII dQ nxer lil un t... be au ... n a ...lre app.reil. Cet orifice ~It j 0
m. 043 mill.• u-d~"ul d~ I. b..~, et lmmfidialement a ...-de""u. ell
.culptfi en rtllef un cynocfiphaie u.l. de 0 m. 10 cent, de hauteur:
98. See Borchllrdt, Dlt! .llIgypt/rchl' Zl'/tml'ssung. Tarel 10.
99. Pogo, ·Egypll.n Watl'r Clocks: pp. 407-09, 411-12,
100. Ibid., pp. 420-22, For another example or the
-ISI-
presentallon of this clepsydra-like objco;t, lee H.H. Nelton (and

edited by W.J. Murnane). The Great HYJ'Mty/e Hall a, K.r~t.
Vol. 1. PilIr! " The W." Reiter, (Chicago, 19811, plate 191 (Thr
U,,/ver,fty of Ch~.go Ork"'.' Institute PublJaftlons Volu~ 100.
Here $ell I le•. 1306-1290 8.C,) pre.""!. Ihe objec:t Ca. In Ihe: nllef
of Amenhotep III called • lod to the lodden of ma,lc
Weret-Hekau.
101. WO, Vol. 3. p. 100. referenn 12: K. Sethe. Urkuntkn tin
A/ten Reich. (ulp-.dg. 19331. p. 152. Doc. 41. line 2.
102. Didier [Hvauo;helle, in the artlcle 'Waueruhr: ux/ton
der Xgyptologle. Vol. 6, 0;:, 1156, IUllcst. that the ulual
idenllflcatlon of the ,ht wUh ......ter clock perhaps should be
rejected, .nd he clter llteralure thai tend. to support this view.
See fH'rtlcuhllrll' R.A. Camlnos. The New-Kingdom Tempk. of
Buhen, Vol. 2 (London, 19741, p. 82, n. 4, .nd lhe FH'perr of C.
Sambln-Nlvel given below In lhe blbllogr.phy.
103. Wh. Vol. 3. p. 106 ~ulm. The combined sllnl
representlng the votive offering, Jt}, appnr u • determh'l.tlve for
~hl In lhe New Kingdom .ccordlng to Wb, Vol. 4, 438, Item 8-n.
In lhe Middle Kingdom the determln.tlve IhoWI only Thoth .1 the
cynoceph.llc b.boon In • n.os (IbId., 8-ml.
104. Pozo, "Egyptl.n W.ter Clock.: pp. 412-14, Indlc.tes, as
.n example of the probable Influence of Elypll.n cloch .nd hour
delermln.tions on c1.nlnl .uthors, the form.lIon of the duodeclm.1
rule given by Cleomedel, ....ho lived somellme belwnn the end of
the firlt century B.C.•nd lhe end of the r1rst century A.D.• "Our
prlsm.llc dl.gr.m reproduc«l In Figure 1 (my Fig. 111.341 accounts
nol only for the cyllndric.1 diagram of th" Ecifu clock_It throws
neW light on several p....ge. of c1alSlcal Ilter.ture de.ling with the
sn.on.1 rat" of Incr".,,, and d""rn,,, of th" I"ngth of the d.y .nd
of lh e night. Cleomedes (KUK).Hcn 9Wlp(a. ~uaSpc.JV. I. 61..-
not"s lh.t lhe dlfferenc" bel ween longes' .nd shortes' d.y Is thus
distributed on the sIx month. following winter ,.,Istlce'
during the 1st month. the day Increu" by 1/12 of the
difference,
during the 2nd month, lhe d.y Increase. by 1/6 of th"
difference,
during the 3rd monlh, the d.y Increasn by 1/4 of th"
difference,
during the 4th monlh, the d.y Increun by 114 of lhe
-152"
NOTES-CHAPTER THREE
dlrrerern:c;
during ... SOh month. ... d., Increases b, 1/6 •r ,h•
dlffer!!m:e;
durlnl
difference.
,h. 6.. month. ... d., lncrcues b, 1112 .r ...
The numeratOrl of thue fractlons lall belna unllyl
Immediately sUBaul an Egyptian orlalo of Oeomedes' rule:. A
,I.nu at our dlagrams...sha..... thu these shl f,actlon. ue: the:
direct rClult of the 1,2,3 prD<;UI of dlllributlna the Iwelve: monthly
real" on the four Inside ....all. of • squue: lnflo.... dodt; It ",al the
Intention of the ElypUan Invenlor of the ),2,3 dlagnm to produce
lICale• •howlng .10....er monthly vuleUons near the lolstlces and
hr!"r onel near the equinoxes; the monthly nln, 1/12, 1/6, and
1/4. happen to be an lnevll.ble consequence, for the ardln.ln of
the dla,,.m, of • simple and rymmetdcal geometrical construction
Hred on the f.vorlte Ecyptlan fractions, 1/2 .nd 1/3, .pplled to the
.bsdnae....
...: both the Elyptl.n Inventor of the 1,2,3 dlaKr.m for
prllmatle Inflow docks, .nd the EKypU.n astronomer who rlrst
fOrfl'lu!.ted the duodecimal rule concernlnK the monthly rate. Were
so dOle to the true lolullon of the problem that their .chlevementl
duerve a pl.ce of honor In the history of EKypUan sc:lence:
lOS. "Ancient Clep"ydrae: pp. 49-S0, and "Prlmllive Methodl
of MeasurlnK Time: pp. 176-77.
106. ConcernlnK the word /nl here rendered as "discern lor
mUlure 11: Borchardt, DIe altSgyptische Z ..ltm..rsung. p. 27. and
n. I. uses "zlhh: Thll usuaUy tr.nll.tel /nw ("numberl"', .... here..
the translation "discern" more often renders Jnl. Borchardt was
obvioully looklnl for. meanlnl more precllely connected with the
meuurlnl of Ihadow lenlths, I.e., where the sh.do.... lenlth Irows
conlinually Ihorter up to the time of nOOn. Indeed Borchardt Ices
on to IUllelt th.t the meai'lli'll of the passaiC II th.t becaule of an
edlple of the lun It W.I not ponible to fix the time by means of
shadow lenlth numbers. The p.suKe may however limply mean
that the sun-lad .t the lime or IOCI.I upheaval wllirully projected
too we.k • light to produce. Ih.do..... ay by produclnK overc.st
days. In either case It does .ssume that lime tellln(l, was associated
....lth Ihado.... leolth.
Borch.rdt (Ibid.• pp. 27-321 dlscuslel t ....o leIs or monthly
tablel seemlnKly related to Ih.do.... lengths for the lelllni of tlme.
-153-
The nut II contalnlOd on Ihe- ",Icr.od cubltr or rules I menlloned

above In note 92. the oldest specimen belnl from the lime of one of
Ihe o.orkonl HOth-9th centuries S.C.l. It con,alns (",hen complete)
thru .hada,", lcnlth. for nth month expressed In cubiti and pall'll:
bUI .... hy only the shadow ,,.0lth, for three hOUf. per month .re
liven and how the Ihado..... arc produced and measured are not
clear. The second II • table from Ihe temple a' Talr. In Northern
Nubia that date. from laic Roman ltmn which ha. Ihe EaypUln
month. In Greek lellen and Ihe shadow len!tlhs for the ".(Iou_
hours ue ellpruiJed In "fect: Aglln e do not know how the
shadow leoath, are produced and for h.' location Ihey ~re
orlalnally prepued and Indeed whethu Ihey hiVe any relaUon 10
anclent Egyplliln technlquel of telling Ume by Ih.do... lengthl.
Borchardt m.kes an herolc but admlnedly f.lling eHort to expl.ln
the.e I.blel.
107. AI I h.ve .Iready noted the word mrh"t or III equlv.lent
mrht ...al not only employed In connection ""lth the IlghUng .nd
leveling In.trument u.ed with ••h.dow elock (.1 It II herd but llso
i l l • term for the ....ter clock of Amenemhet. lind thul perhlpl i l l I
genenl Itrm for any clock hee Document llUS, line 14J.
108. See Fnnkfort, Th~ C~rlot.ph of ~tll. Vol. I, pp. 78-79.
109. Neugeb.uer and PlIrhr. qyptMn Astronomlc.1 Text.,
Vol. I, p. 118.
110. cr. Borchilrdl. Ole altlgyptlsche z"ltmt!llung. p. 32, lee
also hll utlclc "Altlgyptllche Sonnenuhren: ZXS. Vol. 48 U910l,
pp.9-17.
III. Ole ~/tSflyptfsche z"ltmenung. pp. 32·36.
112. Sec Ilso H. Bruglch, Thuaurus Inscrlptfonum
ilegyptf.carum. I. Abthellung lreprlnt, Gr.'Z, Au.lrll, I968J, p. 31,
for a compilrlt1 ..e table of hour nlmel and their eponymous
goddul'" from Dender.. cr. J. DUmlchen. "N.men und
Elnlheliung der 51unden bel den .Iten Xgyptern: ZXs. Vol. 3
08651. pp. 1-4.
113. Ole .IISgyptlSCM z"lImNlunB. p. 36.
114. IMI.. pp. 37·43. The clock ..... flnt publl.hed by c.c.
Edgar, Sculplors' Studies and Unffn/slK-d Wort.. Cata/OtJue BEnEral
des anliqultl-s I-gypllennt!'s du Mush du C.'re, Ncn. 33301-33506
(Cairo. 19061, p. 5' .nd pl.le 21. It ....1 correctly described
phYllully by Edgar, but he ...al unable to Idenllfy It a• • clock
model.
liS. Ole .I"gyptlsche Zeltmnsung. pp. 43-tH.
-154-
NOTES-CHAPTER THREE
!l6. DI~ .ltlgyptlKh~ ultm~nung, pp. 45-46. W.M.F. Petrie,

Andent Weight. lind M,,"surel, p. 45, men lions Ihe Paris dock as
well as Ihe London clock. Concerning Ihe leiter he lay" "Below
thl. copy lof Ihe Pari. Icales On Plate ">tvl) Is II dlalll."" " shadow
clock! cul in black Ilullte, Ihe foil inscrlptlons on .... hlch ue copied
In pl. xxv, II ...... made for Sennu, .... ho held mlny prlellhoodl, bUI
Ihe Inscrlptlon don not relate to Ihe dial lIself. Al Ihe lo.... er point
" mall hal been broken orr whIch rose up. doublleu to carry Ihe
edge which Was to casl Ihe shadow on the Ilope....The gradu.llon II
nol el<lIct and the 1..lllude nonal be deduced from Ihe maximum
re-dlng•. When Ihe dial ..... moved about. II ...... provided with"
plumb bob. hanging down Ihe projection .... hlch I. no .... lost, thl.
en.bted the dial to be N't Uprllht:
117. This dial wu rtrlt descrIbed by R.A.5. Macaliiter, Th~
ENcavanon of G~zer 1902·1905 and 1907-1909, Vol. I (London,
19t2), p. 15; Vol. 2 (London, 1912), p. 331. but he merely called It .n
I.. ory pectonl and did not realize Itl purpose as a lundlal. The
belt delcrlptlon of 11 Is by Borchardt. D/~ a'tlgyptlsch~
ultmCRung, p. 48.
118. Ibid. Hli descrIption and .molylls of the &rlln sundial
occupy pp. 48-50. See .110 his urll... delcrlptlon of the clod.
"Elne RellelOnnenuhr aUI X&ypten: ZA's' Vol. 49 {l9lll. pp. 66-68.
where he IUlleued that thll type of sundl.l wu of Greek origin.
an opinion he .bandoned In hI. Z.ltmnrun6 In conlldentlon of the
Cairo example dlKo ..ered In Paleiline. which be.rl the cartouche Dr
Mer.. nptah (N''' n. 117. and the text .boye Itl. It II worth noting
that the conltructlon of Yertlcat-pt.ne lundlall of the Egyptl.n
type (the lo-called "protractor" type) contlnued .mong thc Gr..ek.
and Romani (see S. L. Glbbl. Grrek and Roman Sundla'. INew
H....n .nd london. 19761, pp. 45-46. 348. 354. 356, 360. J661.
There II no trace amonl the early Egypl1anl of the rar more
lophlltlcat..d Gr..ek .nd Roman pl.nar, Ipherlcat. and conlnl dlall
delcrlbed and ca"lolued by Glbbl.
119. Th.. tabl.. WiSS flnt d..lcrlbed by J. Cern9, "The Orilin of
the Name of the Month Tybl: ASAE, Vot. 43 1190131, pp. 179-81;
the rull artlcl.. coyerl pp. 173-81. The whole pllpyrul ....as ..dlted by
A.M. Enklr. The CaIro Catendar No. 86637ICalro. 19661. For the
tabl... see pp. 54 and Plat..1 XLIV and XLlVA.
120. Neugebauer and Parker, E6yptlan Astronomical Texts,
Vol. I. p. 119.
121. I hay .. Indicated the main erron and omlilions Ihown In
-155-
Flil. SSe-b. A. for the month-names at the end of nth line Ibut
the flnt), I have not attempted to rpectfy ""hal elements of the
month- ....ords and their determinative' are mlssina bUI have limply
used the conventional lpeltlngl. In Ihl. translalion I have in lOme
place•• for typolnphlul rUlon., lubrllluled parenlhet<" for
.quare bracketl. But It should be clear to the reader from the
diagram. of the texl ""hal parts are mlnioa and .... h.t part••re not
readable.
122. Neugebauer and Parker. Egyptian A$lronomka' Text$"
Vol. I, p. 120.
123. Ibid. For the earlter rem.r'" of J.J. ellr.:. ue 'Un text.:
ulronomlque de Tanis: Klml. Vol. 10 119491, p. 10 {whole .rllcle
pp.3-271.
124. ellr.:. op. cit. in nole 123. p. 3.
125. Neugebauer and Parker. Egyptian A$lronomleal Texts,
Vol. 3, p. 44.
126. The figure Ibid.. Plate 23. add, column. II and 12 and
marh them a. ITextl 1lI. A brIef treatment of theft column. II
given by Cilre, op. cU. In note 123, pp. 20-21, and by Neugebauer
and Parker. Egyptian A.tronomlca' Text•. Vol. 3, p. 47.
127. Compare the tran.latlons of Cllre, op. cit. In note 123, p.
9. and NeuBebauer and Par hr. Egyptian Artronomlca' Text., Vol.
3. pp. 45-46. Again, 81 In the earlier table. I have used. for
typosr.phle,1 rea.ons, parenthuell Instead of .quare bracket. In my
translations. The reader can easlly see from the figures ...hlch
readlnBs are dlrrlcult to read and ...hlch are .Imply mining hom
the fragment. CIlre', clear discussion {pp. II-18l of the termlnolOlY
u.ed In the lllble I. helpful. He .ho.... the "senllal equivalence of
mtrt, whleh I have rendered a. "daylight" though II commonly
mun. "noon: wilh hrw. the common word for "day used In the
RamelSlde table above. which. Incldentally. I have also translated a.
"dayltght:
128. J.J, Cilre, "Un texle aslronomlque de Tanl.: pp. 10-11.
129. Ibid.. pp. 16·17.
130. EgypUan A.'ronomlcal Tex,.. Vol. 3. pp. 46-47.
131. IbId" Vol. 3, pp. 1-104.
132. Ibid.. pp. 4-5.
133. Brugsch, Thesauru. In~rlptlonum IIlegyptlaCllrum. I. Abt.,
pp. 52-53, relates the name. of the feaus of the protective gods of
the day. of lunar month. and lhe plctorl,l reprelientaUons of these
day'gods at the time of Dynasty 19. In Brugseh·. chari on p,ge 53
-156-
NOTES-CHAPTER THREE
of his volume, th.. Roma" numerals show ""hlch days of the month
Ihe deities belong to. Sn also Document 111.3 (followlna the ~l;lIon
on Ihe Northern ConstellaHons) and nole. to Document JIl.6 below
for the nrller referencC'I to lome of thele' prolc<:tlve god. on the
cellinI'. of Ihe tomb of Sf:nmut and other ulronomlc"l ~llln&l.
134. I have omlued any dllcunlon of Ihe nrller loose boards
from Ihe coffin of Hcny. which probably date from Dyna.ly 11 and
thul IUlp,ul that the I'andud fOl'"m of thl! 8Slronomlclll diagram
Orlllo"led on coffins In Ihe 11th dynasty, or before. For their
dlscoYery In 1922. Ie'," G.A. Wainwright. "A SulnJdlary Burial In
!;Iap-Zerl'. Tomb ,,' Asslut: ASA£ Vol. 26 (1926), pp. 160-66, and
for Ihe IStronomlcal Inscriptions 1« B. Gunn, Ihe Coffins of
lJeny'- Ibid., pp. 166-71. See aim A. Poao. "The Astronomical
Inscriptions on the Coffins of I;teny (Xllh dynnty1): lSI$, Vol. 18
(1932), pp. 7-13, and the brief trealment In Neugebauer and Parker,
Egyptian Astronomical T~xts, Vol. 3, pp. 8-10. Th~ pl!rllnent
ntronomical [nscrlption. are on board nOs. [ (Inside, see Fig. 1ll.62a)
and 4 (Inllde, Fig. 1Il.62b). Board no. I hal the remain. of Ihe
Norlhern Con.lellallons arranged about the Big DlppI!r (with lome
of the flanking day-dellll!lJ and baud No.4. we are laid by Gunn
(p. 169, para. 4, and ue Fig. 1Il.62bl, had "lracu of lhrll!l! regillers
divided up by vertical IlnCI' In the uppl!rmolt, names of Itan or
con.lellatlon•...: In Ihe Rcond reglsler. Illegible Ilgn. In red: In the
lowelt, varying numberl of Ilarl In blue." In the vilible parI of
board no. 4. we can Identify lhe Ilgns 11~ for Sepedet ISpdtJ or
50thll {-Sirius!. The Ilgn. before thll name ousht 10 have lome
relallon.hlp 10 Ihe decans made up from Orlon (bul lhls cannol be
confirmedl, while thOR afler Sirlul appear 10 be partl of the nam...
of the planel Jupiter. Unfortunately the boards appl!u 10 have
be...n reburied afler Ih...lr In.crlptlon. were tran.crlbed by Gunn
and hence are not now available.
135. Aside from the description In Neugebauer and Parker
(Ibid., pp. 10-12, and pa•• lm throughoul the volume). we .hould elle
the mosl Imporlanl of Ihe earlier dllCUlllonl, namely that of A.
Pogo, "The Altronomlcal Celling-decoralion In the Tomb of Senmut
(XVllIlh Dynaslyl: Isis. Vol. [4 11(30). pp. 301-25.
136. H.E. Winlock. "The Egyptian El<pedltion 1925-1927,"
Section {{ of the Bufl..t1n of the Metropolitan Mus...um of Arts
(February, 19281, pp. 32-44, full arllcle pp. 3-58.
137. Neugebauer and Parker. Egyptian Astronomical Texts,
Vol. 3, pp. II. 105'06.
-151-
138. /bld.• pp. 106-111.

139. Pogo, "The A.tronomlcal Ceiling-decor.llon In the To",b
of Sen",ul IXVlIIth DymutyJ: pp. 309-10. Though there h•• long
been no doubt .bout the Identity of the bull. or prl",ltlvely the
bull'. foreleg, ....lth the Grut &ar, the .u",,,,.ry of the evldenu
for the Identlflcatlon Is neatly pre.ented by C.A. W.ln .... r~ht. "A
P.lr of Con.tell.. tlonl: Studies PresentN to F.LL GrIffith (London
/ Ollford, 1932), Pl" 373·75 (fulllrtlck 373-82). A. W ..lnwrtaht
nolel. the .Itern.tlve. e.r1y Egyptl.. n represent.tlon for the
conltell.llon II .1 .n .dze (.... hlch I. n.",ed, like the conltell.tlon.
mslJtywl. the Inurumenl used for the cere",ony of the Opening of
the Mouths of the dud .nd of the .I.tuel of god...nd Indnd the
conltell.llon looks some.... h.t like .uch .n Inltrumenl ( ...... j when It II
high up In the sky .. bove the pole. In fact ..... hen the conltell.llon I.
menlloned In the PyramId Texts. Seci. 4Sac. It I. determined by .n
.dze •• well .... slar. After .. r.ther lonl period In ....hlch the Big
Dipper .....1 deplcled I I lome form of • bull, In lhe Plolem.1c period
11 ..g..ln beg.n to be depicted ..s the Foreleg. lnterelllnily enough,
not lona before the Ptole",.lc period. thlt II In Ihe time of
NeCI.. nebo II 1360-43 B.CJ, .. buil"l larcophasu. from Abu Yliln
hll t ....o plrallel stripi (I.e .. t ....o realilenl th.t purport to show the
polilloni of the dipper (depicted ..... forelegl for the bealnnlng.
mlddlc and end of the night on the flrst night of each month of the
civil year Iset: Fla. 1ll.74l. A. Neugeblluer and P.rker point out
(Egyptian Astronoml<:.1 Texts, Vol. 3. p. Sil. the .trlps Ire
.stronomlc.lly uselell Ilnee they do not .ho.... the proper steady
rot .. Uon of the dipper' "Obvloully the posltlons In e.ch triple should
represcnt • rota lion about the pole...of apprOlllmately 9QO between
the first .nd the second and the lime bet....et:n the second .nd Ihe
third. lnlte..d ....e find Identlnl posltlonl. e.I .• In Ihe flnt ltop!
It rip field. 2 .nd 3; " .. nd 5: roo II and 12; 14 .. nd IS, 16 .nd 17; .nd
In the s«ond Ilo ....er! .Irlp field. I and 2: 3 and 4; 6 .nd 7: 10. II.
.. nd 12: 14 .nd IS. Similarly the evenlnll, posltlon. In monlh. 1 .nd
II of 'ot cannol be .Imost lSOO different. etc. Since e.ch fleld In
the lecond Ilrlp Ihould represent ••llu.llon .111 monthl laler th.n
In the field .bove II. one Ihould nnd poslllon•• bout lSOO different
from one another. In hct thll II only three time. the c.. 1<! (fIeld. 7,
13 .nd 17). lienee the archetype of our tellt c.. nnol be dated
astronomlnlly and II mllhl .ntedate the fourth century B.C. by ..
considerable Inlerv..l:
140. Borch.rdt. "Eln .. Itllaypll.chn UlrOnomllchn
."".
NOTES-CHAPTER THREE
lnllrument: pp. 12-13 mentions lnKrlptlons from Ptolemaic temples

that Indude refefencu 10 observations of the stan bod
parllcularly to the Grn' Bearl during the ceremonlcs of IIrclchlng
the chord, ·Spanne" de. Schnur 1m Tempel zwischen den belden
Fluchlsliben. Zu opfern doe G.n.....Zu Ipr~hl!n' lch hue den
Fluch.tab. paf;ke du ende des Schligels und cq~relrc die Schnur
zusammen mil de. W"I.heU1ll.liltln lI.e.. Sclh.t~ Icc Volume One,
Fig. 1.3lc ;and the present volume, Fig•. 1II.73a-bl. lch wcnde meln
Geslcht nach dem Gans", dCT Sterne. Ich rlchlc meloe Augen nach
dem Gronen Bbcn', Der...llchl ncben 11) Iclnco Zelger
l...lmrl1tl...J. Ich I"tl.e die vier &:k..n delo" Tempel. fesl:
'Borchardt has ·kleln..n Blren" but the IC)(1 hu msbtYI/II. I.....
the Great Bur. By the time or lhe Plolemalc period II could ....ell
be that the Liltle Bnr was used in observations but nol In
quolatlon or the lradilional lexl lcL Pogo, 'The Astronomical
Ceiling·decoratlon: p. 310. n. 251. Hence there can be little doubt
lhal in the early dynasUes the Great Bear ....as used In both
obllervallons and texl. So the traditionaL earlier statement or lhe
king "Zu sprechen...: in the tnnsLallon abovel Was continued on
Inlo lhe Ptolemaic period. This tradilional lext ..... as treated earlier
by BrugKh. Thes"urus Insorlptlonum flqryptl"cflrum. I, Abth .• pp,
84-85,
141. Though lhe treatment or lhe northern con.lel1"Uon. by
Neugebauer and Parker Is by rar the most complete. the reader
should realize thlll Pogo's dllcunlon In 'The Aslronomlcal
Ceiling-decorallon In lhe Tomb or Senmul (XVlllth Dynalty): pp.
308-312, ..... as a rundamenlal step rorward. In addllion. allentlon
ought to be called to the lenglhy, carly Ireatmenl by BrugKh,
Thesaurus Insorlptlonum aegyptl"cllrum, 1. Abth., pp. 121-31, who
published a number or drawinl!,s or the norlhern conslellallons rrom
the dlrhrent monumenls and hlghlll!,hled their dlrrerences.
142, Poao, 'The Astronomical Celling-decoralion In the Tomb
or Senmut IXVlIlth Dynllslyl: p. 311.
143. Walnwril!,ht, 'A Pair or Constellations: pp. 375-79. L.S.
Bull, 'An Ancient El!,yptlan A.tronomlcal Celllnl-dttoratlon:
Bulletin of the Metropolitan Munum of Art. Vol. 18 119231, p. 286
crull arllcle. pp, 283-86), says. concerninK the bull and the god or
man he calls 'the personage grasping lhe 'reins" bet ..... een lhe bull
aod the head or the Hippopotamus In the depiction or the northern
conslellatlons In Sell rs tomb Csce Fig, 111.69), 'A glance al the
'stars' 00 the Hgure or the bull and or the personage grasping the
-159·
'relnc' rho.......tron, resemblance bet....een their rd.tlve posltlon.

and thon of the hmlllu .tafs ...hlch form the Great Bl!ar." He
add. that he I. lndebtC'd to Georae Ellery Hale, Director of the
Mount Wilson Observatory. for this oORrv.Uon. Incldentrlly. the
rein. holder perhap. UOR from the earlier depletlon of Serqet In
tn!! family of dJa.ramr reprl'Sented In Senmut', tomb, thouah in the
Sell I diagram the (llure of Serqd ,Iso .ppcu.....lth her name on
the other side of the bull. while Ihe reIns holder bI .n enUre!)'
distinct fl.un that h.. no name.
144. Neugebauer and Par hr. qypUan Astronomical Texts.
Vol. J. p. 183. n. 2.
145. H. Chatley. "EgypUan Astronomy: The Journal of
Egyptilln ArchaHlogy, Vol. 26 119401, p. 123 {full ullcle. pp.
120-261.
146. R.A. Biegel, Zur Artrogmxk J.,r .1I.,n Xgyp'u (Zurich,
19211. She declarn (pp. 15-16, referring 10 her rlgurn 6a-b and 7
appended aher p. 36; cr. my Figs. JII.71a-b and 111.72), "Slmtllche
helleren Slern" lind berUekolchllgt word"n, d,,~u rail "II"
Sl;hwlch"r"n Sterne 1m B"r"leh d"r Zelehnung. Wlr sehen, dUI
d.n Slebengnllrn del Grossen Blren ~erlegl wlrd In dal Vlereek
aPYlJ und den Tell £Cl1. 01" Flgur 6 I- my Figi. 1II.7Ia-bl komml
auf der Sternhrte m"hr nach reehtl ~u II"g"n. Ole Gestall der
Stl"rhUI"rln lI.e., HlppopotamulI d"ckt lleh umg"flhr mit unserem
Sternblld Bootes, dal Krokodll, Flgur I I-my Fig. JII.691, ....lrd
geblldet "UI "lnlg"n SI"rn"n von Bootn und "lnlg"n von Corona
Borealis und S"rpen.. o"s Her~ dn Uw"n fillt mil elnem Stern
(-27 Lynell, Gr6sse 4.91 uns"rel Slernblldu der lynx ~usammen.
dl" vord"r" Klaue ldenllrtzl"rt" Ieh mil f Un"e majorls IGrlll:le
4.91. Ole SI"rnehen....."khe auf d"r Igyptlsehen Danlellung d"n
KlIrper umgeben, lind wohl aile Sterne der Lynxgruppe. 01" allen
VlIlk"r belrachl"l"n dlese Grupp<! all unwlchllg. lie .... urde nlehl
ben;lnnt.....01" grllsll"n St"rne d"r Canes V"nallel (OC, B und d"n
Slern 141 habe Ieh all T"n" der Kelt" IdentlHzlerl. Das M"lIer (11.
!Iuf w"khC'l dcr Nllpferd slch IIUtZI, II"gl In der Gruppe Coma
BH"nlea,,:
147. A. Pogo, -Zum Problem der Idenllrthllon der nllrdllch"n
Slernbiider der allen ,l(gypter: Isis, Vol. 16 (1931>. pp. 102-14. He Is
highly erilleal of Blegel's belief Ihl! the Egyptians eonlldered Ihe
Iky al a heavenly globe ....lIh the observer at Ih" e"nter and allo of
h"r Ielf-Iervlng use of som" bright Ilan .... hll" Ignoring otherl and
h"r use of v"ry hint It;lrs as sl"ll;lr conn"cUng polntl 10 eonltruct
-160-
NOTES-CHAPTER THREE
lh~ network of liner constituting E&yptlan conttc:llauon•. She ,1'0

use. the mirror Image of • common representation of the northern
conslen.llon. tn the celestial diagram found In the lomb of Sell I.
This Is equivalent to the unngement found In the "'Iuhed cellini
of hilll J In the lomb of Tausert 11198-96 S.C.!. for .... hleh See
Neugebauer and Puker. Egypt/an Astronomical Text$.. Vol. J.
Plate 9. Concerning the paSSlilIt' of B1e&el'.....ork ldentlh'lna the
northern consleU.llons which I have quoted In the previous note, he
cay. Ip. 114), ·Ole Lage der Urn Malor In den roten Llnnlenndzc:n
6h und 7 (-my Figs. 1II.71b and 111.721 entspr..:h! der unteren
Kulmlnatlon del Slcbenletltlrns. Ocr Hlmmel.abKhnltt von 2h bl.
14h und von _25 0 btl -800 IEpoo;he -130m, dCS$cn vnzerrlc
Polarprajektlon In ZIt!chnung 7 (-Biegel Fig. 7 and my Fig. 111.72)
dargesleUt lsi, befand slch lelder unter dem Harlzant van .Kgypten
zur Zitlt der unteren Kulmlnatlon van Ursa Malar. D.. Blcaelsche
Nilpferd 1m B001er lag auf dem RUcken ader auf dltm Biluch-adcr
ItS Itand gilr auf dem Kopf ,-wenn es den .Kgyptern rlchtbar seln
kannte: ein aufrecht Itehendel Nllpferd haben die Agypter In die
Bootes·Sterne Ubcrhaupt n!cht hlneinphantulltren k6nnen. Wenn
eln slernkundlger .Kgypler den Blegeischen Hlmmeisabschnltl
zWllchen Ursa Malar und dem .KquatarlalgUrtel IChen wallle-und
z ....ar am gullrnlen Himmel und nlcht etwa auf elnem
lIImmellilabus der Zukunf1.-muSile er die Zell der aberen
KulmlnaUan \Ian Ursa Malar abwarten und Ilch auf den RUcken
lcaen, den Kapf nach Norden, die FUne nach SUden, HaUe er nun
dIS Splegelblld der !>cbnnten Darstellung der Sternblider In der
Nlhe des Nardpoll 1m Ged~chlnlo und !>crass Itr ebensa\llel
E1nbl1dungskraft wle R. A. Biegel, 10 konnte er z ....lschen Ursa
Malor und dem unit del Nl1pferd 1m Bootel erbl1cken, auf eln mit
Coma-Sternchen geschmUcklel Mener gellUlzt; den Harul An In
Coma und Lito Mlnor-den halbwegl zum .Kquillor \lerschobenen
Wender der Polgegend; und anderer Wunder mehr, auf deren
Schllderung Ich hler \lerzlchte. SollIe nun dleser auf dem RUcken
llcaende Xgypter auf den Gedanken kammen, aufzuslehen und Ilch
die nle unlergehenden Geltlrne des Nardens anzusehen, 10 wUrde
er dill Siebengestlrn (Schenkel. Slier-O\lald, Slied In oberer
Kulmlniltlan nach links dilhlngleltend erbllcken, darunler die Lan:r.e
des Wenden (Anul. die Gegend des Pals durchllcchend. nac:h
....eller nach Unlen. rechts \lam Nordpunkl des Horlzants. elne auf
der Spltze Itlthende Slerngruppe (Pflad. Messer. lIelfn
Krakadllchen, Genn us ... .l und ....eller rechu \lam Meridian du
-16>
aufrcchl clnhuschrdtendc Nllpferd ...: We .1'0 noll' the attempt

(wholly errooeou., I bc:1It'yd of an unnamed reader of the
ehron/que d'E6ypte, No. 32, JulUc! 1941. pp. 251-52. 10 Identify
the northern constellatlon. under discussion.
148. H.E. Winlock, "The Egyptian Expt'dltlon 1925-1927: p.
37, "Acton the sky the t ....e1"e .nclent monthly full",.!. arc drawn.
nch al • circle with II. round of t ....enty-four hours...: cr. G.
Roeder. "Elne ncue OarltcUuna de. Ielllrnlcn Himmel. In Xsyplen
au. dc, Zeit Urn 1500 Y. Chr.: 0.. Wcltllll, 28. Jahraana (19281.
Hert I. pp. 3-4 (.... hole .tlkl.., pp. )-51..... here he .uUesl. lin.
rather confused ruhlo'll that the .utor. of circles were to Include
data ldentlfylni. for the flnl day of each month. the daytlme hours
(In thc upper half circled drawn from sundl~b, ~nd the nlghtUme
hours (In the lower semicirde) drawn from utronomtul tables or
monthly sl.r-ch.ru lpeclfylng the stan that culminate e ...ery hour
between sunset and sunrbe on the first day of each month. But
this 11 ... ery loose and It Is by no mnns de.r how the 24 sectors
would be used to specify both d~y and nlllht hours. It would be
more likely that each clrde cont~lned star transils for the first and
.Ixteenth day of nch month lI.e.. 24 tnnsltd, .s I lunell In the
text. See also POlo. "The Astronomic~1 Celllng-deC"ulion In the
Tomb of Senmut (XVlIlth Dynasty): pp. 312-13, ·Whether the
unfinished monthly circuli., tablel on the Senmut uiling represent
an ~mbltioul atlempt to Introduce a new feature In the decoration
of .epuichr~1 halls or a hesltanng atlempt to continue a dying
tradillon, only future finds could .how....lncldent.lly...fthe clrdes of
Fig. 11I041 show that the drartsman InKrlbed his circles In squares,
the dialon.ls colncldinl-u a rule but not wuhout e>lCeptlons-wlth
the eorrespondlng lubdlvlllon., poort of the blame Ihould 10 to the
SUlht curvature of the cellini:
149. Neugebauer and Parker, £6yptJlJln A.tronom/c,,' Text~
Vol. 3, p. IS.
ISO. "The Astronomical Ceiling-decoration In the Tomb of
Senmut (XVlllth Dynastyl," pp. 315-16.
151. It WaS only from the Ptolemaic period th~t hbrl wal
called by the EgypUans "Horus-the-red" Il,Ir-cJdrP.
152. Pogo, "The Altronomlc~1 CellinI-decoration In the Tomb
of Scnmut IXVUlth Dynasty''" pp. 322,23, 32S, comments on the
appeararn:e of the pl~netl on the three cclUnKs lthose of the tomb
of Senmut, the Ramesseum, and of the tomb of Sell II as follow.,
"Jupiter. the "'anderlng .t.r of the south, Is the nrst pillnet after
-162-
NOTES-CHAITER THREE
Sirlu. on the thr« c.,lIIng•. The uloclated divinity I. Horul, In ..

boat. wllh .. liar on hll head. Saturn, the ,,",xt planet, I... wCltern
..... ndcrlnB .ta, on the Sell ceilIng: Ihl. 'bull of the huYe,,' I,
....anderl"a In the us, on the ..elllng. of Senmut and of the
Ramulcum. The a.soclated dlvlnUy I. h.gal01 Horus, In .. boat,
...,lIh a .Iar on hil head. Marl 11 Ih, 0.,)(1 pillnet on the cellini" of
Sell lind of the Ramesl"um. In both casel, he I. In retrograde
motion: he I. a" casl'rn .Iar on the Sell celling... western Itar In
the Ramcucum. Mar. I. not mcntlolu:d on the Scornul ceiling:
apparently. he was In conjunction with the Sun. whcn the
decorallon ...... golng on. The associated dll/lolly lonce Igalnl 11
1I0rul, In .. boat, ....lth ...tar on hi. h.. lld. [t I. probable thllt Mira
was In conJuncllon with the Sun, and lhat Saturn ....al an easlern
eyenlna-,tar, hen Senmut'l draftlman ....al at ....ork: that Salurn
...... In the elt, and Mnl In the eall, at ,unnt, .... hen the
lnl<;rlpllons ....ere added to the rlgures on the Setl ceiling: and that
Saturn ...... an e..tern and Marl a ....estern evening-star..... hen Ihe
Ramesleum crahsmen Were al ....ork on lhe planetary part of the
celling. The laconllm of the Jupiter rKtangle of Sell I, and the
erasure In the Ramesseum of Ihe reference to the south mlghl be
due to conJunction. w1th the Sun or to the emharrasslna
appearance of Jupiter as an eastern or wlltern eYenlna-ltar, when
the craftsmen of the t....o astrographlc monumentl of the X1Xth
dyna,ty ....ere dallng their m..lerpIKII.... lp. 325) The planet
Mercury comes neltt lafter the triangle dKans, caned by Pogo
·meta·Sothlc conltellatlonl'l on the three ceilinas, The associated
dlvlnHy, on the Senmut monument. II Setlhl. AI could be IItpected,
this dcvlllsh divinity Is replaced In the Ramclseum hy a Horul,
although a reminder of Setlhl Is apparently preserved In the
Icepter-head allached to the name or the planel. On the Setl
ceiling, the name of Set(hl appearl over the ,lulttaposed Itars of the
slanl-..ared divinity associated with the planet. The planet Venus
occupies lis traditional poslt1on at the end or the 1111: the aSloclated
divinity Is always Oslrll. Iymbollzed hy the phoenix Bennu:
153. For errorll to Idenllry the denns, 1« note 66 above. As
....e said. Neugebauer and Parker were skeptical of establishing
certain Identlflcatlons of the decanl hecaule of the great
dlvergenclel apparent In the many monuments. Pogo. In his
accounl of the differences In the selection, order, and grouping of
the decans bet ....een the Senmut-Ramllseum and Sell I C traditions
of the decans ('The Astronomical Celllng-decorallon In Ihe Tomb
-163-
of Senmut IXVlllth Dyn.~tyl: p. 3211. make. lome lnterullnl

remarh, •...our 6'1,,~h celnUal alobe (",uh sevent pain of ,ohl'
holes drilled to corre.pond to lnlerval. of lime sho..... lng _ppredable
proc;esslonal changn on the celnll.1 Ilobe. 10 ,hilt the: support of
the pol., aKIs could be modified to pnmlt the ",.,'11100 and
mnsurement of the ailltude of the polel has permUted UI to
Identlfy, tent, lively.•cver.l d«"an' wllh certain conspiCUOUS stan
or afOUl'. of Ilan, which appnre'd In ten-day Inlerval. on the
horizon of Thebe••bout 3OCK) B.C. and ,boul 1500 B.C. 'e.,.,
,,00Ile1l,lIon of the ShIp (den"s 12-17) on the Senmut "elline seem.
the
10 overlap with our Scorpl01 the K"omu liar. of de<:iI" 17 nem to

belonl to the Milky Way In S'lltlarlul, elc.J. A .ranc:e at Fl._ 3
I-my Fig_ 1II.65b; ,f. Fill. 111.65" ,olumn 201 111 sho.... thlt Ihe
'Ell' Iconitellillon) hi, Ihe .hape of I loop of .Ilrs on the Sell
,elllni. The V ·.h.. ped lroup of .lln, bet.... ~n Orion Ind the
Icolumn belo.... Ihe bealnnlng of the! 'Eag' Ion the Senmul «Illnl.
column. 21-241. rnppeln In the shipe of I V on the Set! celllng
Icolumn 211. Our 6-lnch globe Iho.... 1 Ihll .. boul 3000 B.C...... hen
Ihe vernll eqUinox ...... In T .. uru. "nd the nielUII equltor p"'ned
IhroUlh the Pleladel. the hell I'll rlllni of Ihe Plel..de. Ind of the
HYldes o"urred Ilmoll eXlclly nSI of Thebes Ind ......
follo....ed-I ....o dends Ifler Aldeblnn or four de<;lds Ifler the
Plelldes. and prlctk..lly In the .Ime plrl of the horizon-by the
hellacal rlslng of Betclaeuze. With Ihese exptlnltlonl In mind. II
....111 not be dlfflcull 10 see on the ,elilng. of Senmul .. nd Sell I lind
especl..lly on the laller! I fllrly correct represenlltlon of Ihe
PleI..des Ind Ihe Hyldes, preceding Orion .. nd Sirlul. The generll
confu.lon .... hkh chu"'lerlzes, In III declnologuel, the Orion
rellon-so rkh In stirs Ind .Iu-Iore-frultrlles III Illemptl of a
closer Idenllncltlon of the IndiVidual declnl:
154, See W. Gundel. JRbn~ und Debn't~rnbf'dcr(GlUchlldt
Ind Hamburg, 19361. 2. Abl .• Ind Ihe monognph of his" IOn H,G,
Gundel, W~/lblld und A,'ro/OIlIe In den grlechfschen ZlubrrfMPyrl
(Munich. 1968), MUnchcncr &/trlge zur Papyru,forwhung und
Inttken Re<;htlgcKhfchte, 53. Iteh (19681.
ISS. F. Pelrle. WI,dom of the Egyplflns (London. 19401. p. 21,
"Allgu'.r melsure, Mnluremenls of angles ....ere un.no .... n In
Egypl or else""here lilt perhlps 200 B.C.. Ifler ""hl,h EnlOllhenel
used" divided circle .... It """I problbly lome Blbylonlln ""ho mlde
Ihe Idvlnce of I gndulted Circle, 11"ln& direct mellurement. of
Illllude, ""hkh h.. culmlnlled In the modern Innlll circle .. nd
-164-
NOTES-CHAPTER THREE
~qu.torili. AI Babylonian d~reel lert no trace In Earpt. It II only

by hour divisions thai the right acenllon II delJanllcd. and there II
no rKord of decUm,Uon:
156. For a brief account of EgypUan astralos)'. see Parker.
-Egyptian Astronomy. Astrology. and Calendrical Redonln,: pp.
723-25. ~e .Iso the ""orks clIed In the preceding note.
-165-

Anciet Egyptian SCience Vol2

Uploaded by

Copyright:

Available Formats

Anciet Egyptian SCience Vol2

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Anciet Egyptian SCience Vol2

Uploaded by

Copyright:

Available Formats

What are the main subjects discussed in the volume?

What are the main subjects discussed in the volume?

What is the organization of the current volume?

What is the organization of the current volume?

Memoirs

Copyrighl c 1995 by !he American Philosophical Society for its

Library of Congress CaIaloging in Publication Dati

I. Egypt 2. Science, anciem, hiliOf)' of

ISSN: 0065-9738 .........

conclusions which we can draw from the eighteen

translations themselves will give readers without

and younger ones (e.g.• Krauss) who have darined and

part II, Document.

Document III 2 The Eberl Calendar 193

Document 111.7. A Table of the LeDl.th. of the Daylll.ht and

Part 1lI, Blblloaraphy and Indelees

part IV, lIIu.tratlons

Calendars, Clocks, and Astronomy

Calendars. Clocks. and Astronomy

in II tentative way earlier. that the Egyptians used the

within its month. Whenever it fell in the last

'first quarter', lull moon', and 'third quarter,

sudden rising of the river and for much of that season

probably the case of the Ebers calendar, other specific

Parker's Account of the Old Lunar Calendar

analysis in Document 111.6 below) and confirmed. Parker

the earlier use of intercalation by noting references to

reign of Horus Djet (see Fig. 11I38),12 He and his

probably no longer be accepted (even in the revised

involves the fact that on occasion the twelfth month

device to keep the lunar calendar in step with the

calendars, is also the 'coming forth of Sothis:

twelfth month. Hence the Ebers Calendar loses its

cannot have been the civil year or the fixed

feasts in which the dead expected to take

that the latter two also were lunar....

it to follow after the civil New Year's Day, as it inched

necessitates its reinterpretation in a lunar calendar. and

[below and in Document 111.2] that just the reverse of

Sothic heliacal rising as the mechanism of intercalation.

crescent before sunrise, We shall return to this

The Later Lunar Calendar

mathematical astronomical Egyptian text yet

significant about the cycle for our purpose is

starts in every case exactly one day berate the morning

25-yc:ar cycle tells us which years have twelve months

have discovered the rule that regulated

The Origin of the Civil Calendar

predecessor (probably in the first case and certainly in

period and indeed until 139 A.D. when, we arc: told by

possible quadrennial dates of the beginnings of the

convince anybody that in about six cases out

increasingly short chronology for the establishment of

rising, no attempt was made to adjust or correct the

which data on the number of days in recent

epagomenal days were concurrent with the

dependent on Parker's usc of the beliaeal rising of Sirius

Sothic Dates and the Ebers Calendar

depends on variable conditions of the observation. as

attested. Judging by what is known of the