SARAH

SYMONS

ANCIENT

EGYPTIAN

ASTRONOMY:

TIMEKEEPING
IN THE

COSMOGRAPHY AND NEW KINGDOM

Submitted for the Degree of Doctor of Philosophy

Leicester Mathematics University Computer Science, of Departmentof and

1999

Abstract
The first part of this study analyses and discusses astronomical timekeeping methods used in the New Kingdom. Diagonal star clocks are examined first, looking at classification of sources, decan lists, and the updating of the tables over time. The date list in the Osireion at Abydos is discussed, and issues concerning its place in the history of astronomical timekeeping are raised. The final stellar timekeeping method, the Ramesside star clock, is then examined. The conventional interpretation of the observational method behind the tables is challenged by a new theory, and a system of analysing the tables is introduced. The conclusions of the previous sections are then gathered together in a discussion of the development of stellar timekeeping methods. The small instruments known as shadow clocks, and their later relatives the sloping sundials, by the The hypothesis the that are also examined. established shadow clock was completed addition of a crossbar is challenged and refuted.

The second part of this study is based on New Kingdom representationsof the sky. Two Book beginning the diagrams in detail, texts discussed of with major and several celestial are Nut, which describesthe motions of the sun and stars. New translations of the vignette and dramatic text are presentedand discussed. Portions of the Book of the Day describing the behaviour of the sun and circumpolar group of starsare analysed.
Finally, celestial diagrams dating from the New Kingdom are described. Their composition is diagrams discussed is behind framework the the recreated. and and significance conceptual By introducing new theories and analysis methods, and using a modem but sympathetic knowledge of to the to this our original sources, approach update and extend study attempts these areas of ancient astronomy.

ACKNOWLEDGEMENTS
I wish to expressmy deepestgratitude for the help, guidance, and support I have received throughout my researchfrom my supervisor,Dr Michael Dampier. I also wish to thank Dr Mark Collier, Fellow of All Souls College, Oxford, for teaching me texts the his Egyptian for concerning so much about advice grammar and nuance, and translated within this study; and Dr Allan Mills of the Astronomy Group, University of Leicester,for his interest and support for my work. My study was supportedby a University of LeicesterStudentship.

CONTENTS
INTRODUCTION PART I: TIMEKEEPING SECTION A: RISING STAR CLOCKS SECTION B: NEW KINGDOM STAR CLOCKS SECTION C: RAMESSIDE STAR CLOCKS SECTION D: DEVELOPMENT OF STAR CLOCKS SECTION E: SHADOW CLOCKS AND SLOPING SUNDIALS PART II: COSMOGRAPHY SECTION F: THE BOOK OF NUT SECTION G: THE BOOK OF THE DAY SECTION H: NEW KINGDOM CELESTIAL DIAGRAMS 4 16 16 52 69 107 127 152 155 176 190

CONCLUSION APPENDIX
BIBLIOGRAPHY TABLES FIGURES

216 221
222 228 229

INTRODUCTION
The study of astronomy in Ancient Egypt has traditionally fallen between the fields of Egyptology, astronomy, archaeoastronomy, and history of science. Researchersfrom all four disciplines have looked at the area from different perspectives. In addition however, have not tackled the field: Egyptologists have felt there have been reasonswhy researchers it required too much technical knowledge while other researchers have believed that those capableof reading the texts are most suited to studying the material. Also, many feel that the Egyptians did not reach a high enough level of scientific knowledge, nor develop rigorous enough methods of observation and recording, to make the study interesting or
Egyptian have felt linked is to the that the too religion others valuable, while subject strongly

for a scientist to work on comfortably. These reasonsare now being discarded: Egyptology has become more tolerant to multidisciplinary approaches, and the traditional labelling of anything non-classical as `primitive'

has to a large extent disappeared. This study tackles the subject with, it is hoped, a mixture of the scientific and the Egyptological, the analytical and the sympathetic, and aims to examine some existing theories,presentnew hypothesesand introduce new ideas into this interesting and somewhat neglectedarea. The term `astronomy' means `the scientific study of celestial objects, of space,and of the Egyptian (Concise Oxford Dictionary). Throughout religious universe as a whole' physical literature we find a plethora of celestial references. We find depictions of the sky on the ceilings of temples and tombs. We find written predictions and records of certain

for find architectural that events and we used evidence astronomical celestial references were

and timekeeping purposes.

The Egyptian use of celestial objects differs from our own. We have devised a series of specialiseddisciplines to study various aspectsof the sky. In our every-day lives, celestial
bodies play only a very small role. The presence or absence of daylight is the most

noticeable intrusion of the celestial realm. Those living or working with the sea are also affected by the tides, which we now know are causedby the gravitational effects of the moon. Apart from admiring a clear night sky, this is as close as the majority of present-day westernerscome to the cosmos. The serious study of such things is left to scientists, of
different have we many which kinds: astronomers, astrophysicists, space scientists,

lines dividing drawn have In and cosmologists, meteorologists etc. strict other words, we

have developeda taxonomy in which to fit the universe and the study of the universe. The ancient Egyptian view of the sky appearsto be somewhatdifferent. We cannot know has information Egyptian but he looked the thought the worker sky, what when up at heavens. have little how the the of perceived survived which shows us a educatedpriest may
We find that the line between the physical and the spiritual is difficult to draw: the priest

describe their light, He look but at visible points would of would would seespirits and gods.
in journeys in terms their of spiritual and explain motions existence terms of myth.

One of the exciting aspectsof reading religious texts which describecelestial objects is that, unlike most spiritual literature from ancient Egypt, we can still seethe objects to which the texts refer. Although the effects of precessionhave altered the orientation of the sky, we can calculate what the heavenswould have looked like thousandsof years ago. The celestial bodies themselvesare still there and still behave exactly as they did then. Unlike ancient lost the location, to us, or characters or events, whose are cities, significance, and appearance What has changed,however, is our human perception of it. remains. sky

It is unfortunatethat a study such as this should have to begin by drawing limits around what is and what is not `scientific study of celestial objects'. In itself, the act of dividing somepotential for further understanding. However, the choice has beenmade to suppresses concentrateon two linked areas: timekeeping and cosmography. `Timekeeping' is easily defined: using celestial bodies to mark periods of time. Previous researchinto the area has defined a small number of distinct methods which the Egyptians developedfor this purpose. Texts dealing with thesemethods are readily identified by their contents,but are rare. We shall discuss five timekeeping methods, three based on the stars and two on the sun.
Other timekeeping devices which do not have an astronomical basis (such as burning wicks

be be but this the not and will study and water-clocks) will mentioned remit of are not within dealt with extensively.
`Cosmography', `a description or mapping of general features of the universe' (Concise Oxford Dictionary), however, is a far more difficult area to define. Clearly, the celestial

diagrams found on ceilings in temples and tombs fall into this subject area. Further texts indicated they that by location, by were style, content, were chosen which and primarily To diagrams. limited been this This has to small selection. related to the celestial study literature Egyptian to religious open the study celestial referenceswithin the wider corpus of two between the been have subjects, the a massive undertaking, and relationship would timekeepingand cosmography,would have beenobscured.
Cosmography is an area distinct from cosmology, both in the modem scientific sense and in

the Egyptological senseof a creation myth. Cosmography in essencedeals with a state belief This distinction that is Egyptian by than even the a process. slightly rather clouded
We the have the motion of sun, must repetitive events, such as cause at each occurrence.

thereforefind similarities betweena `cosmology' describing the creation of the universeby a god, and a `cosmography' which describesthe annual journey of a star (for example, the similar imagery betweena creation myth as related in the cosmological text The Book of the Divine Cow and the description of the motions of the sun and the stars in the cosmographical text TheBook of Nut). Egyptian cosmologiesfall within the scopeof theological studiesand have no conceptual link with modem scientific cosmology, whereas cosmographicaltexts contain conceptswhich we recogniseas physical, observableprocesses. The languageused to describe these processesis alien to us (both linguistically and conceptually) but the themselvesare familiar, offering us the opportunity to gain a clearerunderstanding processes of thesetexts than of other, more spiritual works.
The aim of studying Egyptian cosmography is to further our knowledge of the human perception and classification of the sky. The relationship between ourselves and celestial bodies has produced both philosophical and scientific thought, and led the Egyptians to

This both study their the to purposes. understandingof apply sky spiritual and practical lead to timekeeping the methods on practical aspectsof cosmography,and uses concentrates
into the study of the depiction of the heavens. The link between timekeeping and cosmography is evinced by the sources to be discussed in

this study. The group of timekeeping stars known as decansevolved from a presumably Grecothe diagrams into in until a symbolic group present astronomical practical origin Romanperiod, illustrating their prominenceas a recognisedpart of the sky. Thesediagrams
timekeeping for the other sources contain major study of cosmography, are and also Ramesside for dynasty XXth diagrams include the star clock, the tables example of methods,

dynasty the diagram, is decorated XVIIIth most and water-clock a with an astronomical
interesting of all astronomical texts, the Book of Nut, contains within it both stellar and solar

timekeepingmethods. The time period chosen,the New Kingdom, is selectedas the areaof Egyptian history during which the majority of timekeeping instrumentswere known and in use, whether practically or decoratively. Diagonal star clocks were already ancient, but the earliest surviving examples of other types of stellar timekeeping methods, shadow clocks, and water-clocks
date from the New Kingdom. Foreign influences had not begun to enter the area of

astronomy, but it was a time of increased interest in categorising and describing the world, been has It the time the not of great astronomical ceilings which we shall also study. and

definitely desirable to texts to originated possible nor restrict research methods and which within the New Kingdom, as the exact dates of origin of most of the timekeeping methods and texts are not certain. This study doesnot include the related fields of chronology and stellar alignment, nor does it aim to cover the wider issues of the basis of the Egyptian religion. These fields are

devoted to have distinct from in the research recent areascovered this study, and somewhat has does identify It Egyptian which them. also starsor constellations, not attempt to specific
been the ultimate goal of many researchers entering this field. In fact, this study shows

be fail knowledge, or either clearly why such attempts must, with the present state of

first certain is identifications, make The that to must one reason make unprovable.
that even Egyptian this shows study timekeeping and stellar methods, about assumptions

some of the more widely accepted assumptions are not as probable as was previously
how the be It that that understand we thought. must accepted until we are reasonably sure timekeeping methods were constructed and used, we cannot begin to explore the area of stellar identification with any hope of success. Also absent from this work is research into recent theories concerning the relationship

betweenconfigurations of starsand configurations of constructions,and other `astronomical' hypotheseswhich have been put forward in the present wave of pseudo-scientific `nonfiction'. These publications have highlighted the vulnerability of scientific Egyptology to

attack from outrageoustheories, and have identified the need for Egyptian astronomy to be re-established as a serious,well-understood,and well integrateddiscipline. This is not the only reasonwhy the study of Egyptian astronomy is important. The subject history intersection disciplines including history the of stands at of many of astronomy, science and scientific thought, history of scientific instruments, philosophy, anthropology, theology, Egyptology, and the study of the Egyptian language. Each of theseareasneedsto
integrate what knowledge we have of Egyptian astronomy into its own field, and each one is

enrichedby the incorporation of this early, written knowledge. We must also acknowledge that although the amount of material which has survived from in Egyptology is Every it is incomplete. waits times to the area present enormous, ancient with anticipation for further sourcesto be discovered,and in the meantime theorisesabout the missing parts of our knowledge. Each study should therefore representthe best that we
the level knowledge most using sources, of of can achieve with our present and availability

but of analysis, should acknowledge that we are part of a continuing methods up-to-date searchfor increasedunderstandingand that further discoveriescould overthrow our carefully definitive. be No be to work. study can ever said complete and constructed Previous work in this field has been undertaken by some of the greatest names in Egyptology: Champollion, Petrie, Brugsch, and Borchardt have all made major

Neugebauer Otto in 1969, Egyptian to the More study of astronomy. contributions recently, Parker Richard completed their great study Egyptian Astronomical and Texts in three

has become and clocks for the star the of which study standard volumes reference work

celestial diagrams. Their work summarisesall previous studies and publishes and analyses 1 In Volume history. Egyptian to throughout the most sourcespertinent astronomy courseof they present analyses of diagonal star clocks and the Book of Nut. Volume 2 is entirely concerned with the Ramesside star clock, while Volume 3 consists of an exhaustive catalogueand analysisof astronomicaltexts from monumentsdating from the New Kingdom in knowledge is this best Their the the of onwards. work still representation existing stateof field and will be referredto frequently in the presentstudy. More recently still, Marshall Clagett published the second volume of his series Ancient Egyptian Science which is entitled Calendars, Clocks, and Astronomy. He summarises little but Parker, Neugebauer presents and previous work, covering a wider subject area than new information and analysis.
Apart from these major works, there have also been publications dealing with narrower However, the be in quantity the to study. present subject areas, many of which will referred in for lower is far that, than example, of publications on the subject of Egyptian astronomy Egyptian mathematics, and to a large extent the publications have been tentative and have to made for have, been isolated. There example, papers published recently which claim ' These the identification between certain stars and certain Egyptian names of constellations. been has difficult to is see what identifications seem to be basically conjectural and it often

by knowledge the publication of such theories. to the sum of our added In contrast, Christian Leitz's recent work on Ramessidestar clocks2is a methodical attempt
at analysing a timekeeping device, and ultimately produces some possibilities for

identifications of Egyptian stars. Such a study is useful because Leitz's clearly presented
Egyptian Ancient the identifications 'Probable Identification of Locher 'New of Egyptian constellations', Sheep'. the Constellation Constellations', 'A Conjecture Concerning of Circumpolar the Early Egyptian and

2 Leitz AltaegyptischeSternuhren.

10

he his his the the reasoning, problems worries concerning validity of results, and encountered with his method, all add to our understandingof the device. Also, we know that one line of enquiry has beenthoroughly investigated. The presentstudy aims to approachthe subject methodically, dealing with each timekeeping method in turn before moving on to the areaof cosmography. We start with analysisof preNew Kingdom diagonal star clocks. After briefly reviewing the sourcesavailable to us, we discusswhether a classification systemis a useful aid to the analysis of the sources. We look at decanlists in somedetail, and will try to find out if, how, and when the diagonal star clock was updated. We shall also collect information about the possible motivation and construction of the diagonal star clock in preparation for our later attempt at tracing the developmentof such timekeepingmethods.
In the next section, we move to the New Kingdom and look at instances of decan lists within

timekeeping contexts. We shall see how the diagonal star clock survived into the New Kingdom, and introduce an important text, the Book of Nut. Whilst looking at one section of
the Book we will begin to analyse the unique date list it contains.

Section C deals with a type of star clock which occurs only in the tombs of pharaohsfrom
the XXth dynasty: the `Ramesside star clock'. The section will question some of the

detail have the problems of analysing which and will gained wide acceptance, assumptions these observationally-basedtimekeeping tables. A new theory will be presentedtogether
it be supporting evidence which, will explained, solves some of the problems associated with with the currently accepted method of usage of the tables. The next section will take our findings from the three preceding star clock sections and

between the links for, development stellar to the outline motivation attempt of, and
timekeeping methods. The previous chapters will have highlighted some weaknesses in

11

current theoriesconcerning`transit decans' and thesewill be extensively discussed. The final timekeeping section will look at sundials, in particular the earliest type of formal been it has 1910, Since has from Egypt, the sundial which survived shadow clock. ancient A have the that the crossbar. shadowclock should an additional member, generally accepted small minority have expresseddoubts about the validity of the addition. This section will be be hypothesis hopefully to the that considered crossbar present evidence against will conclusive. The secondpart of this study will be based on New Kingdom representationsof the sky. Two major texts and severalcelestial diagramswill be discussedin some detail, beginning in section F with the Book of Nut. We will use the text to investigate how the Egyptians how discuss the described We the vignette the will perceived and motions of stars. sun and texts be the heads Book interpreted, translations vignette the of can which read and and new and the dramatic text will be presented. The other major text which will be discussedis the Book of the Day. This New Kingdom
into insight day, describes the an the motion of the sun through the course of text providing

its the daylight of hours, composition the behaviour of the sun, and the classification of the
entourage. A particular portion of the text of the Book of the Day dealing with the

be discussed. translated group of stars will and circumpolar Finally, celestial diagrams dating from the New Kingdom will be described and analysed leading to the introduction of a model of the sky based on our findings in this part of the
study. We will try to recreate the conceptual framework which the Egyptians used to

the of bodies it the glimpse the to small a motions of celestial gain and use understand

Egyptian sky.
Throughout the study, frequent reference will be made to the original sources of

12

the validity astronomicalinformation. It is to thesesourcesthat we will turn to try to assess of assumptions which have been used in previous studies and which have become entrenchedin the popular consciousness; such assumptionsas a strict definition of the time period or `hour' that each timekeeping method was meant to measureand the existenceof conceptualastronomical objects such as the meridian. These assumptionshave even led to imposing a priori conditions on the timekeeping methods, to the `accuracy' of researchers the methods being questioned, and to the whole field being termed primitive and uninteresting. Otto Neugebauer, even after his work on the three volumes of Egyptian
Astronomical Texts, felt that `Egypthasno placein a work on the historyof mathematical astronomy ...
its insignificance be too strongly emphasised Egypt provides us with the exceptional case of a cannot ... ...

highly sophisticatedcivilisation which flourished for many centuries without making a single contribution to i3 the developmentof the exact sciences.

All these difficulties stem from the problem of enabling the modern mind to approachthe instruments completely unbiased by our scientific concepts,methodology, and our ancient The kinds is `myth'. This in timekeeping. modem and true the of own particularly area of only recently established obsession with accurate timekeeping, both at a personal and
national level, makes it difficult for us to tolerate ill-defined and variable time periods, and from find hoping leap `hour' our analysis that out at us ourselves constantly we an even will

information fact, In about the this any of ancient methods. as study shows, we cannot gain time periods without finding out more about the Egyptian timekeeping devicesand methods.
Any consideration of `accuracy' is at the end of a very long list of questions to which we by find first We by diligent survives, this answers. will only achieve must analysis of what sympathetic reconstruction, and by constantly testing our assumptions.
3 Neugebauer History of Ancient Mathematical Astronomy pp. 562-568.

13

This study attempts to learn as much as possible from contemporary documentation, practices, and motivations, before moving on to considerationsof measuredtime periods. This process is aided by recent advances in our ability to translate texts. The new

translations presentedwithin this study owe much to the personal guidance of Dr Mark Collier, and follow his findings on the verb and auxiliary particles4 Since Neugebauerand Parker's work, study of ancient astronomy has been greatly aided by increasedaccessto computers,and by the developmentof software which can plot the stars for date Spreadsheets completing useful and accurately. at any quickly are also particularly repetitive calculation tasks and to provide instantly updated graphical output. These

functions were particularly useful when analysing Ramesside star clocks and shadowclocks. However, the drawback of using computers for such tasks is their demand for accuracy, In this study, such results the which colours our own perception of results of such studies. have been used to indicate trends, to test the likelihood of certain hypotheses, and to illustrate different scenarios. It is hoped that this study will show that some progress can be made towards a greater
by diagrams Egyptian from timekeeping methods and celestial astronomy understanding of

introducing new analysismethodsand new theories,and by re-openingseriousand scientific discussionabout formerly acceptedbeliefs.

Manley " Certain parts of Collier's work have recently been published in Collier in format and an accessible How to read Egyptian hieroglyphs.

14

Conventions
Datesusedin this study are thosegiven by Baines and Malek in their Atlas of Ancient Egypt. A chronology of Egyptian dynasties with special emphasis on the New Kingdom is presented in the appendix. Where individual hieroglyphic signs are described, they are identified using Gardiner's systemof classification from his Egyptian Grammar. Where citations are given in footnotes, the author's surname and the title of the work (sometimes abbreviated for clarity) are given. The full citation may be found in the bibliography. Neugebauer and Parker's Egyptian Astronomical Texts,as the major reference for this study, is cited by title and volume alone.

15

PART I: TIMEKEEPING
Section A:
Introduction
The earliest form of timekeeping device of which examples have survived to the present is

Rising Star Clocks

the rising star clock. A star clock consistedof a table containing the namesof asterismsand individual stars. Thesetimekeepingstars are known asdecans. Each clock consists, in essence,of a table with twelve rows representingthe hours of the
night, and thirty-six columns representing the thirty-six decades (10-day periods) which

hypothetical layout 1 Table the twelve the the of a up make months of shows civil year. ideal rising star clock. The numbers 1 to 36 and the letters A to L (omitting I) which appear in the main body of the table each represent a decan name. The main body of the table caters

for 360 days, and is headedby a date row listing thirty-six decades. The diagonal pattern `diagonal by decan in led has the to the this star clock'. the table name created names part of The decan names representedby the letters A to L are usually called the triangle decans,
be 36 1 decans to The because in will the star clock table. of the shape they make again called ordinary decans.

in decans the decans, leftmost list the four The ordinary columns of the table contain a of all first three columns, the list columns (which do not appearto have any timekeeping purpose, but may have served as a referencelist of decansused in the clock), and the eleven triangle decansA to L plus one extra triangle decanM in the final column. This epagomenalcolumn days, (the for during five days or timekeeping the the epagomenal extra year served whole of `daysupon the year'). The table is divided into quarters by a horizontal strip, containing offering texts, and a

16

vertical strip containing figures of deities associated with the sky.

CC_
2I I TIca na nananaE. ' iar 1 -1 ' ! 2? a a ' co ' Ii aEnaEoa
a co (b

a naE

a c

R.

' 31 as

1 -1 co aE ' nana '

7 E.
0

' -

Q. aaaaaaaaaaaaaaaaaaaaaaaaaaaR. co co co aaaaaaaaaaaaD.

aaaaaaQ. Cl. a Q. naR.

.ne

aaaQ.

Q. Q. aaan

06

naa

A B C D E F G H J K L M

25 26 27 28 29 30

13 14 15 16 17 18

1 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 2 A 36 35 34 33 32 31 30 29 28 27 26 25 24 23 3 BA 36353433323130292827262524232221 4 CBA 36 35 34 33 32 31 30 29 28 27 26 25 5 DCBA 36 35 34 33 32 31 30 29 28 27 26 36 35 34 33 32 31 30 29 28 27 61 EDCBA 7 8 9 10 11 12 FEDCBA GFEDCBA HGFEDCBA JH0FEDCBA KJHGFEDCBA LKJH0FEDCBA 36353433 36 35 34 36 35 36 32 33 34 35 36 31 32 33 34 35 36 30 29 3130 32 31 33 32 34 33 35 34 28 29 30 31 32 33

21 20 19 22 21 20 24 23 22 25 24 23 26 25 24, 27 28 29 30 31 32 26 27 28 29 30 31 25 26 27 28 29 30

18 17 16 15 14 13 12 11 19181716151413121110 20 19 18 17 16 15 14 13 21 20 19 18 17 16 15 14 22 21 20 19 18 17 16 15 23 22 21 20 19 18 17 16 24 25 26 27 28 29 23 24 25 26 27 28 22 23 24 25 26 27

10 987654321 98765432 1211 10 9876543 13 12 11 10 987654 14 13 12 11 10 98765 15 14 13 12 11 10 9876 14 15 16 17 18 19 13 14 15 16 17 18 12 13 14 15 16 17 11 12 13 14 15 16 10 11 12 13 14 15 987 10 98 11 10 9 12 11 10 13 12 11 14 13 12

3119 32 20 33 21 34 22 35 23 36 24

21 2019 18 17 16 15 2221 2019 18 17 16 23 22 21 20 19 18 17 24 23 22 21 20 19 18 25 24 23 22 21 20 19 26 25 24 23 22 21 20

Table 1: Idealised layout of a diagonal star clock

The table would have beenusedin the following manner: An observer looks at the easternhorizon at any moment during the hours of darkness. He looks for a decanwhich is just visible above the horizon. Knowing the date, he looks at the by is headed finds decan either table the twelve rising star clock and nameswhich column of the exact date, or by a date which occurred within the past nine days. He finds in that in decan the decan that he The has the the that of column name of position observed.
first indicating the him first hour The twelve tells the stars position column of of the night. hour of the night and so on until the twelfth position indicating the twelfth hour of the night.

For example, if decan 20 is rising above the easternhorizon and the day is in the middle decadeof the first month of Peret, decan 20 is seventh in that column, therefore it is the seventhhour of the night.

Sources
Existing or recordedstar clock tables in the form of Table 1 are painted on the inside surface
dynasty. XIIth lids date from the First Period Intermediate the and which of wooden coffin The sources are listed below and numbered using the same order as Neugebauer and
S Egyptian Astronomical Texts 1 pp. 2-3. Full details of sources Al Astronomical Texts 1, together with references for each.

5 Parker,

in Egyptian found be A12 to will

17

who used date order and similarity with dated clocks to establisha numbering systemfor the first twelve sources. Four additional sources have since been recognised which are
designated here as A13 to A176. Each source is shown schematically in the manner of

Neugebauer and Parker in the accompanying figures. (Grey dates indicate that a date line is

in not present the original source, decan names refer to the decansin the first column and
bottom row, while grey areas indicate damage' on the original source.) Al Msht (S1C7) IX-Xth dynasty, from Asyut (Table 2)

A2 7t-lb (S3C) IX-Xth dynasty,from Asyut (Table 3)

A3 Hw-n-Skr usurped by Hty (S6C) IX-Xth dynasty, from Asyut (Table 4) A4 Idy (S1T) date unknown, from Asyut (Table 5)

AS M31't (S2Chass)IXth-Xth dynasty, from Asyut (Table 6) A6 r3gyt(T3C) XIth dynasty, from Thebes(Table 7)
A7 7kr (G2T) First Intermediate Period or Xlth dynasty, from Gebelein (Table 8) A8 Hk3t (A1C) date unknown, from Aswan (Table 9)

A9 Hw-n-Skr usurpedby Nht usurpedby Hnn (S3P) IX-Xth dynasty,from Asyut (Table 10) A10 T3w3w(S9Q XIIth dynasty,from Asyut (Table 11) All T3w3w(S5C) date unknown, from Asyut (Table 12)

A12 . ms (S11C) XIIth dynasty, from Asyut (Table 13) A13 Name and date unknown (S#T) from Asyut8 (Table 14)
A14 Name and date unknown (X2Bas) probably from Asyut9 (Table 15) A15 Name and date unknown (S16C) probably from Asyut1 (Table 16) A16 Nbt (S1Hil) XI-XIIth dynasty, from Asyut" (Table 17, presented here for the first time in the format used by Neugebauer and Parker, and Figure 1) A17 Name and date unknown (S2Hil) probably from Asyut12 (Table 18)
6 Certain other coffin lids contain traces of diagonal star clocks: 13L (BM29570 belonging to Sbk-htp, belonging to (BM6655 T2L by Locher in 'Two further lids' has decan coffin matrix); an empty mentioned Mntw-htp); and S1X, which Willems (Chests of life list of coffins) identifies as the coffin of Nny, which is in Section dealt be but in Egyptian Astronomical 1 1 (which Texts is with will probably not a star clock, source H), but which Lesko (Index of the spells) states is an entirely different coffin belonging to D. i-hvpy whose texts have no references to decans. 7 For coffin designations see Lesko Index of the spells. 8 Although excavated in 1908, this source was unknown to Neugebauer and Parker. Details are given in: Locher'A Further Coffin Lid'.

9 First published in Lapp 'Srge des Mittelnreiches' then by Locher `Two further coffin lids'. 'o SeeLocher'Middle Kingdom astronomicalcoffin lids'. " Inv. Nr. 5999 in the Pelizaeus-Museum in Hildesheim. See EggebrechtSuchenach Unsterblichkeit pp. 5861 (including plates) andAntike Welt pl. 33 pp. 41-43.
12See Locher'Middle Kingdom astronomical coffin lids'.

18

t".

M
M "'

e At r. . MMd .o

". -c,
II:, -9,

2
-i

3 '..
Of O rrr rrr ' r t N f0 r, f0 P Co

A. -2 JZ .9 <O in rNM<
m
1 NMR M s V' 1n 1Ci 0h 1f oho f0 fN ti 0OO Or N rr NM N .V' .-r Ln . U) co GD W

c0

CD WOrNM Of O rrr ^NM rrrrrr N M rr Q IA rrrrrN rrer Co t.-r. Co 0O c> mh M rrr OrNM rrrrr Nr

:crew] k(n)m sind srt srt s3wysrt hry bpd srt py-r 3h-Y fmy-ht 31ixry ;may
N -3

' 2

-NM NM M at aT b

of IA (0

10 {p h

CD h OD OD

co
OO-NO

C'l

[crew]
k(n)m

OOrN r_

I-

co rr

"t

sind

srt

m
1 Nm f0 r CO Of

_.1
2 ''4 =3

r mhU
A l0 mh h CO OD O0 O1

U)
c2

.-VVv

Ur, sri

S2 't w0

s3wysrt

IA fp

N cq V 'n to 1, by bpd srt

OOO; CO Ol C a, orr OZNM rrrr rNrr Cl, rrrr

CO O O (V NrN O/

to v to to

co tpy-` 3hwy N imy-ht 3hx'Y

C) 3hx y
rn kd
haw

. m
j
19 O

:5
=3 OCNM Or

2'
of c4 NM Irr el' U) M "' lA (O to IL) IfU) h f0 eh CD

14,W) (U) co e mo r .-I .-Nr

co O
1-rrN CD O

Ed haw rt ry rt
Q

1
r L

N "- 4
C) .-N N r-A N

NN

'" r NM M

Nrrr
Co rrr"r O U, b lD rrrr f0 1O^ h

O^
Co CO O N Of -

CO
NC

rN
N N

NN
NN U) qp

N
OO N

-N
NNN N

NN

art

1
-

hry art

O) 1 f. ) -NNNNN NNNNN

inn fry bwt art wrrt py-e spd a 0

07 aNi

OO O c4

0) N NNNN
co h ao o U) 3 v) rFFFN
FFFNeN N NN NN 'Ir NNNN

rmnbry
cbwt

W) co N NN

O--

NCo 0O
Co Nr N. NNN

m a

N N N

NNNN N -0 N Co -0 NNNM

Co N

hrt wert
V)

C-

NNNNMC. )

. pd
nmt r3wy knmt

Cl F 2 0. M NF "3
N N-

NN

N (4

NNNNN

tPY'esPd
SPd
1. O

NNNNMM p

C C
NNNN NNNN NNNNNN NNNNNN NNNNNM U) CO NNNN co OD NNN NN COO 0) 0 NM 0MNMM MMM CEO P) co R 0) V) l7 61 f'7 r CO 0) O 0) .Cl) Lf) N 0O a0 O) NNNMMM Co 0) NNMM N 0) MMmm 0 0-N

N-

NNNN NNNM NNNNN OO NN CO l0 OD O N CO') O_ MMMM M OR) 0) N M

N)

m n
(N NNNN

knmt
Co M M f7 0) 0)

a A

NM C) M

.-NM

s3wyknmt knmt bpd n Cry hat b3w pkwy Pw

CNO C. ) O) xM C) U) (p MM C7 r, a c9 iA c")

Me to NNNNNN ((N NNNN CI)

iry bpd n knmt F10 -3 tat pw

CN9

l0') Of O C9 t'9 l') ei

CNO C.) f")

>hwyhaw m3t hrt m3t hrt andrsy

fMMMi MMMM N 0) M N2 f'') CC'') (N') Cl) f0 Cl) M CC9 c t'O t+1 tai C7 '' Co

U) Co NNNN

NM C9 M NNM NMNQ 0 Co MM 2 `. '

-3

fm3t hrt
N

1
N

NN 0)

!M MMMM

r' NQ

Im3t hrt sind rsy

OOrNMQ NMMMMM 0 MMMM M NMy MMMM

-NQm NQmU

andmhty iir 43pt

-3

eN)

Q N<

fa

sind mhty ntr43pt rmn (pry lJw tpy-'' spd imy-ht spd

E
1

Co 0

u
M M .-NQ chi

.NQ
NQmUUW QmUOW moOW

to UO

a)

-NQ
N<

CO00
) NW 00 C.

i c N NN

c iOj bj

13w
LL LL 0

N Cl)

py-espd ny-ht spd hwy ,3w

tr d3 pt hwy s3bw

MM

Cl) MrN
8NQ Cl)

co U U

2j W U. WU 0

CIO

L2

=3 2 Co 3
r Q r N

ri chi rNQ
NQmU NQmU

fa

U0W
OW IL LL

LL ( S
O2',

12

NQ
NM M rrr O NN

01 UUW
in P CO CO 1'N

U. ( IYj
CO Oi ONpNN OrN rrrr

PO') M

CN)

CM+fr)

C')

((O') `. '

QmU12

5114 LL

(9Y->M

Z3

19

t2
M " n

12
NMQ NM -3M er U) v :+O mP u) co f0 h CO nOOON ham, CO 0, U) m 1 aD O) to CO fCO O) O r "M <O h t0 0) a-

Z, _ E.

F4 3

c)
1
Q2

.w Q . ..
N
NM

3 `e. ' " -c" c:

Z:

tt

;03
r

t% a f.

N.
w
ON
rrrrr rrr

(3)

N [crew]
-. k(n)m

Cy; '
--

'Ilk
co /. Co 0) C CV M

[crew]
k(n)m

Cornr-N

sind srt
co N srt

-3

MR

CO

rn 0o. - NM
NMVNm

C*4 2
vLo

sind srt srt

s3wy srt

Cor- OD (
e2 w) 0 co 0)
rr

-..

'-N

"'Ir

"''rs3wysrt
tbry bpd srt

Nmv
1 'Z O aD 01 O/ OrN rrrr rNM rrrr NMV 'rrr M rrrr
of rrrr

in (D r` hry bpd srt

t

rrr -C,

co v vo

v> Co r_ Co tpy-t T u'y Co n Co O

V: E !2 m1"CO tPY-`3b"
co V' 1n to
f-

---

Mn y t my-_ ,,,
FA w

r v (o to rrr-N

Co 0

iWrrrr jrrrrrr Co +7 nN ,? rN rrrrr Nf eQ eN Cl) er U) 1R LO co " v) to F-

.3

OO
Or rr rN rF

o0 rr-N

i" co O rn 3hwy
Nr kd N

Co

r-

Co

C:)

1
4 ..

'.

r- Co c"ir' Fj N haw art Co (N N (N -' (N

ry he

V'
Il)

P
<O

ma oNo, g
Co o

NM rF

1'I

pA 4 3w y r; rrt

() e'

r'

UU) r'

o0NcovU) co rN-NNN_

-I
J

1 m 02 3
Tt `SJ) M

F C-4F-

F
0

iNv ' 4 F( '.4

0) -N fV NNN NN9N CO

-ry art

vim
m-

- goo
U) N U1

vNrmnhry CQC',
U) R CO fbwt

N cr)
M rrN It e-

r, (D co
Co rCo . -

rmnhry Os F4C14C(14 ) C. c-4

Cbwt N 0) C14

N l) R NNNNNN

N-

NNN

U)

co

brtWr tPy SPd

rNNNNNNNNN0 NNNNNNNNNMM
0NNNNNNNNNN 0) ZU, G N9NNNNNNMMMM M

knmt 0) s3wy kramt

d w 0 C 0 0

m _ 2

V r 0

r(o co
N

0 co N-

Go a N N
N

hrt wert
tp y .r spd SPd
kn mt

NN

fV C*l
co

NNM N wM <gl.

0 0

3 m1

fV

N
8

laueea7 ,w,

NMMM

-NN

d F

ce F

20

Iii

r - .- ,2-Z.
r S,

Z ,

? " 3 3+

-Z:.-c, -be

CZ -Z. -e-

1 -NMllr
'Z NM', MO0N0 rNmh
m

, m
't In mP

orn-Vj
Co c O Cr rrr rrr r NrMr (4 Vl

[crew]
k(n)m .. Smd 't ") srt
:J

NMVL,
NMe0m m tfi mh m I Co et A

m r- oocmC
r" a0 0 Co p O OrNM rrrrr ;rr -N r3 f In

[crew]
k(n)m Smd srt Srt

3
2

SCt

Co

mm

Co rn

CO le,

mm

s3wy srt

n m it

Co rn

rM

`*

n CO s3wy Srt

3 Co r. in (
1 Co r" wN
2 OD POrNM CO Orrr r

C, 4 mle
1'' -'m1
Nmh 1A m

U)(0N
CO O

hrybpdsrt
120 tpY-f 3hwy

3 mri. 2

aow;
"' -l'7 NM et < 0

mle
v

Ul) 0 f- bry hpdsrt

`. tpY-f 3hwy -N
OM Of N
haw

co c oN

cornr
to OrN rrrrr rrrrr
Nm

-N f0 rrrrNr mP rrrNrN
( CO

m -- f1 aD
oNN

C>imy-ht3bwy
3hwy

Cr

O (OV r

3t1wY , i P:
C

00 O

orNM

st

to

P.

Co

oN

rrrrr 2 NM rrrrrr NM if MQA r0 Kn 0

rrrNr 1-rrN mO WO0 .( Nm (V

(V L of f rt bry fH ..V

2 . 7

rr2M N M" ms

'1ff 0(0

2r f-

rrN 00 N O00N Nr

rfrt

rrrrrr
rrrO

r
N0

(V

(V
V crf NNNN

N
if

N
v)

Z " tL r"
3

rr O. .-

r0 m ~.

rr N

CV fV fV NN

(y

2
3

rrrrrN Nmh rrrr mh rrr

vv>m

-0R
CO O NCO O r4 Ol O! N N

rnNMAU, rNNNN
FV ( MR NNNNN to (p Co

R rmnhry
CO O f

00r, -N
--P r en e-

rr

m0
NNNN

0M; -NNN
NNNNN
V) Om NNNN

#m Y-ht 3hwy bry , rmn c"', t000 N N fbwt

p
Co O N-

NN

bwt
rt w f rt In d : ;

} ?

g fV
Co

hrt wfrt
Mf ! 'Y3..

M Nm of NNNNNN

ONM NN

hwy
O

U7

p
$' OON Nr NNNNM y J NM et NNNNNN uD m0 W ONMMM ON rM NN CN (0 N NNM cO NMa M Co NNm M f MM s3" hrY pt b3w tt3w ICnmt bpd kamt kamt

6P

NNNNNN
C,4 LV NN NNNNm Co MMN M Q) f'7 fMM c0 MMxM NM

d'

NMmm

MMMMM
M2 ci MMM r A Mf'ONQm p U f'9 V) (7 N< N pltwy

lm3t hrt
Emit (art
sind Smd rsy mhty

---

(V

C 2

O 0) NNMMMM NM

f0

N<

in

nir d3pt
rmn bry

ciM
.L M

M `M rMi can
M COO

N<

mUm

NQmO,
<mUNW

W haw
LL. tpy-r spd

MM

N c

.- N<
<m
mU

m0Aw
c owu.

u- 0
Z

fmy-bt spd

0 imy-ht spd

N<m0 2 NQm
NMe

43 x <O 0x nir pt w --i L <,, (D W w (9 z - x pkwy s3bw

&n rti a0 OrN

NNNN2

m Lo M c9 r) chi '-

V)

UDW

LLY

21

xw

.-NM

e_

t0

Co

Ol

ON

[crew]

NMQ NMM M in co P h
OD

t0 CO rCO 01
2

t0 1, Co O-N

rCo O>

Co Cf O-

0ON -NM M-" l; r Nm

fL

2
3

N
M 1n

In
mn

10 tOD

Co OI
Ol O r-

eN

NM
(*I V'

k(n)m
sind srt

2 3 1 2 3 m
jpdsrt

rew] (n)m andsrt rt

le 4,0 co
1-

"ef M. " j

1 eU)0r.

aocm
9--

o--NMVln -N<9

srt s3wysrt
1_ h. ry

'et 1[)
(0

2 In m ti COwF
3 -0W

F11
le bm

Co
C)

O_NMN0
rr -lo (4 rrrrr -e -e00h !rfrr -Co Co -

3wysrt ry bpd srt py-c3hwy my-ht 3hwy d

m Q2
3

Co rn 2-r' NM

m le tn 0r_ vm
N rrr f0 f1 Co O

Co tpy-r 3" 3b" MY-ht i0.

fV M h rrt 'rt

h"

c0

Or r--

co ao
O OrrrFrr rNMQN F rrrrr M FFFFFF U) O .-NM rrrrr NM

'
OD N O ( ( M

Co

0) CD

(4 Cy Me V' rin

c* Co M) FrrrN CO 1 CO

Ln Co 1 CO

(D rCO

ms CD O

CO NJ NN V fV MJ Nc

-3 1 2 -3 1 m

O O .-

NM .-'

1
2 3

Icr
Frr O^ (O h

N l0

tD h rrrNNN 1 CO FrNNNN Co

r` .-

3w K

e-

'V'

rNMR rrrrrrrrN rrrrrNN rrrrNNNN r-

ONNN NNN N

NNNNN C+) `f 1l) O) e-

bly
3. wy

fV

ry ert
1

1
Z

OD

ON NNNNN

hwy mithry

vm rrrrrN
-n (D 99 r-

ti

ao
NN

NN NNNNNrN
NNNNN

_N

rmn h ry
bwt

r0 %0 -t0

Co

NNNNNF "7 to W f0 fV a0 N f

3
SPd
"" O

1_ CO ON'? .--NNNNN

5wt rt wert

Co N NNNN

fV

NNFNN NiNNNM

co brt wrrt
py-'
Ql OF Spd

Q .

2
3 CO oNN NNN

m 2
O O

1
P "

rNNNNNFNNN CO rONNN N-NNNM in cl) Of OV ( f0 O) Ol Ot

'y-e spd pd
nmt

at
0

MN a

V)

mm U) VNNNMM

NNNNF

NNMMS

1
2 3

NNNM NNNNN O N NNNN NNMMM NMMM NNNNNN N NNN N N NN f00 NM O N) Q) M l^7 MM l') M CN9 MM MMNMM N 07 O 1tl NQm M M c4 0 )) Ml+)

N)

M M (O

kamt

NNNNN

co

NMMM M

s3w y kra mt
lt3t b3w
Pkwy 3w

00 d

m1 CL 3 l1 C -3

NNNNNNN Co NNNNN C') er 1A NNNNNNM (0 OJ OOFN ONN NMmM

C')

CM9 9L tai CM

3wyknmt ry lpdnkramt

CM) io (n

bpd n kram 9 ham'

co
0))

_ a

1
2 3 1 2 10

C') N) MMM r1 cr N

it)

3t ly3w hwy f Jw nit hrt

c44 c4 NNNN N NNC N f0) MMM CO

01

V' (a OD Of u, NNNNNMMMMMMF N M C, 4 COV N N COV M

lM7 f7 M

C7 NQ

hrt lm3t -' N lm3t hrt

Smd rsy

Cam') PN')

NMNMM NM 07 ! lM -N N< NQmS S

c0+) 00

nit ljrt

(n l+j

MN

l'9

C) fN') CMS) NM CMS)

C9 (M') ('7

andrsy

3
71 2

t{ CN7 CMS)0) 9 M U) C07 U)

sind mhty
Co U

-3

MMMM

and mhty
trd3pt
mit Iny 3w
LL t I.

NNQ

ntr d3pt
rmn try
MNNQ MNM f'9 CD U07

NMX MMMM

-NQmUD

=3 Z1 O2 =3
1

m mx ; 2N

_"

NQm0Ow Qm0Oww in uoww0 Uowu.

OW LL

b3w tpy-espd Imy-ht spd ox 3hwy

&w

3 Ln,M, xtto ,-N

MMNQ 'n NMFN NQ in NQNNWW GO F

QQCo
CO mU0W UUOW g2 OW LL LL tL
ca ca

Uow1

'y-rspd
ny-ht spd hwy 3w

X en)28 -NQ 2, NQm ,,

Oj . rNQmU

L2 -3 2 3

LL (9 LL 0Z3

Co

2 3

NQmCO QmUow
M OOON9N rrr Kf N CO 1-

W U. =w C _ - Y
Co NN O 2Co N0M V

Y ntrd3pt phwys3bw

ONQ
r) QmUNW

Ga U

U_
S O ' "

nY
YF N r'

tr3pt hwy s3bw

MR EO m NNNNM NN Co

<D O0NV-0gN

NO

p NNrrra

UOYx

ci

LL

QJ-

22

yM +v

MMMM.
1
2 N

ii .. '. v`

W
n
(O

...

. C

1 V

x
ht y -ib wH
N
N MM

N M -0 v
c9 M sY ll)

r-- Co rn a
N

[cre\ ]
k(n)nr

N
C)

m
m

O
MM

( LO M M
M N

tn3t lint
wVtl hk3t1

c ri

n co i-

Co rn 0) a
Z 2

"'

1 (Y) n co r-- Co > Q 2 v n o N- m rn

3 "n (0 N- CD 'j'

T v snrd u' ,srt

(n

LO

MM 10 M -0 -ll M
M M L

hntt

Jirt

12 12 s3wy srt

M c
) C.

mM
(0

M M , co
) 2

Q
-

1 (0 N Co0
2
3

r
co

ao rn
0)

M
Vom' 10 r

l:ry hpd.s,t
D tpy-"
0 inly-lit

) C.

hott )zrt Inn1y

N'. 3t1

31ni'y
3ltwy

CEO

OC)

gym,, ,o co l, (0 (0 i
co r

,n

a,
m

lity-ib wi3
Uk? ti co hntt hrt

1
2 0

v
n

n co
c0

co o rn Thwy' N
m N rn N 41

M m
n

CO LO M

co

Co N rn

r4

N /lilt'

co m

(0 Co m
m
CD

linxy
hry-!
I! I''-!

tIv
CO
M

co
N-

(nm
Cj

m
0)

b wU
B Wi

aD r
I-

r r
N

N
-

O
N

N
N

N
N

r/l
hry rl't

(0

r L

11 O (O GO

CO

co

N O N N O) <N -0 ( N N M N N 1f) N
(0 N

O N N Cl) N

N C) N N

N C N Lr) N (O N CO 0) N
O Cl)

N U) N N co N N O Cl)
M

N (O N 00 N 0) N

N N N 0) N O Cl) M

mill I)Yt't lira

Ily ( am C) M ) m

N -

N (j

/C(I kd

Co

mN
M CO N

0 w 0 0 0 0 J a F

2 00 O N N 0 N Q) O)
N N

l1'l1 f SP[i ,

M 0)

0)

31114)

m N

N N (O N
CO N

L0 N N OD N
O N

fpy'

m f0 m co

m a> N rn 3iiwy liry hpd srt w rn lintw hrw

Q) (D d -

1
2

N N
C') N

SPd

co N
V N

st N
1f) N

M
N M

N Cl)
M M

kamt
s 4v )' k117171

zr D
co

rn rn N c`s3wy td
N

{nth ' 17,14'

23

.MM ~I lyL

Z: r fi CM

y'

r ,

M M m
N

(V M

10Co M M
V) M
M m CO

N N N
N CO

M N (O M U)
M N N N

N N M N ('0
M r> In m 1] N

i N N N N
(O f0 m') C

(O N R N N
CO N f0 M

19 I19 N f0 N N

N N I Jo

kCI IiI'

hurt hrt
' .1

M M

(O (")
M N CO

?-mit

hntt Irrt tms n! ntt silty lmK'y hry'-ib 4'H

M Q

rain hry ,c3

Yt

fj

Q M

kLI

` CO
O C N

'l
N .o
N N

w
CO
M
D N

si`ntvr k(n)m tpy'-" sind sind

' IIYN'

tnl3t lart
t17f I I7rt k11//7I lind,

N N

CO r

CO M

M
0

N
f0 N

w O w 7 O

M (

M chi

co m m

CO N

cN') N

srt
S iN'y Crf . R F

co co CO N cri co r
m
t0 M
lv M

<q r, m r' c
m m 2 M v

N Ifntx hY m tpy-r kamt `O c3xy s7't .

lr tpy-'

lrpd crt iy

0
m M

M Co 10 (13 N ri cm
10 M

f
N M

r `"'

N N
N M

M CO
cU /0 V 1'lif(I

ihwy hraw jim- qr ht

srnc[ Q

(Z M 00

Co m

r, v

m M _ M CO

t()y-r knY)I

tint", ht w qr In kd

rOi

co

Z N

r,

k/imt

s3wykd

f0 Lo M

N (0 CCO

.0 t N in N

0_ . M C N

r7t

m
'R

Co m
M M

m m
N
V

a CO N
M

F"
N r7't

c . r
N M

Cr

)'Hilf

i?
(

1'I7 ,
h (

4-
N N r4 Mm M O Co Ln

`'.
N (0 rc0 rCO

22

M (0

-0
u3

Inltt hrt tnls n hntt spry' Jznwy'

M m

M (

M m

(0 N-

V)

ci

(0

M M
M
N m

r.

N
M

m M
n

a M
W) M fOD m

m j1
r--

r,- Co
Co
r N

Jiry-ib
S, nIl1'

mV
N_

N N

M M

0 2
c0 ( f0 N 2 VM f0

k(n)m
tp)y-' S'nid O O a

92 . M

(p 1Co

M .0 0 m c0 N-

(0 I` 00

Z: an,

( 2

2V Q2 2

S/Ikt sit S3ii'y srt

( N

.. rf

(0

112

cam) V

lunybpd srt
m rn cri n co tpy C 3{fxy

N N N

t/lIH'

lll1'

"v 2 "2u 'm

r.

N N

N N

24

m m
2 '-

co _

`
N

N _a N N

sew'" kd

'Y

v ., _S 0 1 m Lr) m co (n Co (D m

'J
?

Li
v`.

co
A

6 c,

C10

",

bkk

t!

(D

10 C) -C,
O CO

10

/)t A'.

Co ch (D

r' N

tpv-, {iiitf

JINtt

cnrd srf

co

c0

Irr!

04 ()
(V m c,

M
-

V
N

,N
m

o
co

lititt

lirt

tni. s 77 ! hilt

CO

co r-

co m

sptylrrnti

<o

lu -ih ii i

cn

(C) r-

CD -0

enrn cv S. k(r1)irr
cnri/

10 (o N- co
cD N. OD

m72 l

w 0 0 0 R

R F

25

H M -Z" --"

-,
r N C) -T Ul (O

:$ : ' -z .

N Co m
Co 0) rn

[crew]
k(n)m
QL

hry-ib w13

cn

'v

C, v
v

I0 (0 n
co

(0

Ll
V V

'E sind srt 1O srt 'P s3wy srt

Un co

2
3

U, m 1- OD rn
(0 r GD O)

Y2 3

hry hpd srt

V
< 2 OD r r
:I

F N
N

3hwy tpy-r N Amy-ht 3hwy

3h wy

_ 2

s3wy srt

04 (
-N

2
= 3 N

`O
m N-

N ao o N
N N . . N N N

rn NN N co haw
N N N N N NN N Crt

Q2 ti y-I 3hxy

hry rrt
00

N N
rTNN N

N
Cl) N

v) (D rmn hry N N
NN

"bwt

liry rrt
m 2. 3 N. NNNC, NNN NNNN m N

Cc'' N
N N N N N N 0) N

N N

U, N CONN co 0 hrt wrrt

N N 0 C (+) C) (+) N ( ) t py Sp d -r Sp d q. ar 9

N 2 0

CO "1 N N 0) N

N co N

0 c)
M

knmt
knmt

2NNNN
3NNNNN mNNNNN m 2NNNM to ON (N !OmOO CD OO

s3wy C'N) C'Cl) )

h N m ) C
;, 04 04

m 'm C m

airy hp dn knmt hat haw haw pkwy (10) 'tm3t hrt

Cc

m '`, M Cl) mm Cl)

CN') co 04 m

3NNNNMM 1NNNMM
O1 (0 2NNMMMM co p l9 ON

C",,cam, Cam,,

c" tm3t hrt sind rsy sind mhty

cv cj < If, Cl) -Qm

3N

co

MM (C+)

26

as O
O u L

U -O

s e x
L

U N

pQ

bQ
C W

27

Grouping the Tables

The star clock tables Al to A12 have been extensively analysedby Neugebauerand Parker
13 in Copyists' mistakes are rife in the that tables who remark most are some way corrupted.

main part of the table and many of the sources are incomplete. Neugebauerand Parker in by decans tables the the the considering grouped used and arrangementof elements the vertical strip. They identified five groups, which they labelled Group I (consisting of sourcesAl, A2, A3, A4, and A5), Group II (A6, A7, and A8), Group III (A9), Group IV
(A10), and Group V (All and A12). Leaving the consideration of decan lists until the next

briefly let us examine the three layout elements of date row, vertical strip content, section,

and horizontal strip content which contributed to this classification system. A3, A6, A10, All, A12, A13, A14, and A15 do not have a date row, leaving Groups I and II split if this factor is consideredimportant in grouping the tables. The vertical strip14contains four figures, often captioned. In Al to AS, A9, A13, A14, and
A16 the order of figures from top to bottom is: Nut, Foreleg, Orion, Sirius. In A6, A7, A8, A10, and A17 the order is reversed. In All Nut. Orion, Foreleg, is Sirius, A12 the order and

This factor creates three groups of sources, one containing Neugebauer and Parker's Groups I and III and the additional sources A13, A14, and A16, the second consisting of Groups II

A17, IV and the third of the single Group V. plus and The horizontal strips15contain offering texts to various deities and decansfor the deceased,
be A13, in in this (although Re safely can to all cases with an offering starting unreadable Sirius, to then The Orion, Nut, Foreleg, and to strip continues with offerings assumed).

13 Egyptian Astronomical Texts 1 p. 23. 14 Egyptian Astronomical Texts 1 pp. 28-29.

is Egyptian Astronomical Texts 1 pp. 26-28, Locher 'A further coffin lid' (A13) and `Two further coffin lids' (A14), EggebrechtSuchenach Unsterblichkeit p. 59 and Antike Welt pl. 33 (A16). Contentsof the horizontal has been All A15 and not published. strips of

28

various decans. Five sources,Al, A3, A4, A6, and A13, have offerings to Foreleg, Nut, and Orion before A16, decans. A9, Sirius is A5, A2, Four to and omitted entirely. other offerings sources,
have Sirius following Foreleg, Nut, and Orion and preceding the decans. Of the remaining definite has A7 Nut, Orion, is Foreleg Sirius, then the order sources, and probably and which in A8. A10 has only Orion and Foreleg and no decans. A12 has Orion, Foreleg, and Sirius. All has only Foreleg readable as the first offering after Re. A14 has Orion, Foreleg and

then Nut after one decan. Clearly the space available has a great impact on the length of the decanal part of the offering formula painted on the coffin lids. The order and inclusion of the four deities

Foreleg, Nut, Orion, and Sirius is not usually subject to the same constraint (except in the case of A10). Is the order an important factor in classifying the tables? Neugebauer and

Parker's Groups I and II are again split when this factor is considered. Furthermore, so is Group V.
In summary, from the three factors considered above only the order of figures in the vertical This Neugebauer the Parker's by groups. and strip agrees with system not splitting any of

factor, however, producesonly three distinct groups instead of five. These three factors are devices, they importance. In timekeeping the the tables as analysis of primarily of artistic have no relevance. We still needto turn our attention to the major attribute of the tables: the
by is the but five decans; we shall soon see that once again, the concept of groups not upheld

decanlists, and we conclude that such a system of groups is not useful in the analysis of the star clock tables.
This conclusion is clearly illustrated by Locher's diagrammatic attempt at showing the diagram fifteen here Locher's 2. Figure uses seven of sources, reproduced as characteristics

29

factors in addition to Neugebauer and Parker's seriesof groups to position the sourceswithin the diagram. Theseinclude the order of deities in the vertical strip, the inclusion or omission of certain decans,the decanusedin the upper right cell (first decanin the table), provenance, and two epigraphic points (pertinent to sourcesAll, A12, and A13). Locher omits two of

the layout factors which we have already mentioned, inclusion of a date row and the
sequence of offerings in the horizontal strip, which split Neugebauer and Parker's groups. The system was originally designed16 to incorporate thirteen sources (Al to A13), but after the publication of two further sources (A14 and an empty decan matrix), Locher found that his system could not incorporate the inclusion of these extra sources without distortion, '7 and it in subsequent publications. abandoned Kahl18 has recently attempted to classify certain of the sources into a stemma showing

development from an original basis he denotesas a. Locher has combined this approach A17 Al to Neugebauer Parker's five of to and with groups produce a systemof classification plus the empty matrix sourcementioned above (Figure 3). Even before considering the area of prime interest in these sources,the decan lists, we are
identified incorporating has been that to newly see no classification system able capable of indication is This has been an sources without modification, and none universally accepted.

for is a have the that number of examplesof star clock tables which survived not sufficient layout, be devised incorporates of to all points meaningful classification system which
little to The developed have far our very origin, and age. added systems epigraphy, so from have the potentially more diverted the tables to of and seem understanding attention

Egypt. in development ancient of astronomy and timekeeping rewarding study of

16Figure 1 in Locher'A Further Coffin Lid'. 17Locher'Two further coffin lids'.

18Kahl 'Textkritische Bemerkungen'.

30

011ergsfrxtN fl

irrurlrn,! A . u a
try
bA

t1

Z.

'0

0o

a
Vo .nE

V v

sae
f 1' f !`Y

E' a.

Eu

Cl

a an w

't J
Lo 41.4

ca
"IP

Cd

e"+ so
VN Ob 4

W hQ U dU 0

lw. -f d,.

U0

Ca

3 E"
OLs b
w
N-1 N 6/

CT)

s.
on iw
In

V7 T
31

Decans in the Star Clock Tables

The aim of analysing decan lists is to attempt to find some information about the antecedents of the tables in order to assessthe timekeeping abilities of the star clock, and to understand how this timekeeping method developed. It has been accepted that these coffin lids are not

authoritative sourcesfor decan lists becausethe tables were clearly not understoodby the
copyists. We must attempt to strip away copying errors, epigraphic confusion, and artistic

licence to retrieve the conceptsand methodsbehind thesetables.

Neugebauer and Parker list seventy rising star clock decans19two of which do not appear in any of the clocks (ts rrk2o and s3h). The seventy decans and the labels used by Neugebauer be distinguish Parker labels listed in 19. These Table to them confused not should are and with the numbers 1-36 and letters A-M used in Table 1 to introduce the star clock.
1 2 3 tm3thrt tm3thrt 12 "crew" 12a shnw 13 k(n)m 13a tpy-r sind 14 sind srt 14a sind 15 srt
16 s3wy srt

24 rrt 25 (a, y rrt 26 rrnn h,y 26a rmn Ivy s3/1 26b is 'rk 27 rmn Ivry 27a rmn hry sah
27b , mit sah

35 hat haw 35a hat dit 36 liwy haw 36a pkwy d3t A sind rsy B
D

t bk3t w9 3a w. 3t1 3b bk3ti 4

5

lpds

4a sgpt
sb3`sn

sind mhty C ntr d3pt

rmn hry

5a tpy-r Iintt 6 hntt hrt 7 8 9 hart till tms n hnrt kdty

17 hry bpd srt 18 tpy-S314 wy 19 3hwy 20 Imy-dit3hwy 21 bJwy 21a hntw hrw 21b hntw hrw 22 kd 22a s3wytd 23 haw

27c sah 28 "bwt 29 In-t wr,-t 30 tpy-l' spd 31 spd 31a tpy-'*knmt 31b twy . 32 knmt 33 s? wy knmt 34 &, -y bpd n knmt

E h3w F tpy-J spd G Imy-ht spd H J K L 3hwy haw ntr d3pt s3bw

9a spry 9b spry hnwy 10 "hnwy" 10a hnwy

1 11 fry-lb w13

M pkwy s3bw

Table 19: List of decan names with Neugebauer and Parker's numbering system

19 Egyptian Astronomical Texts 1 pls. 26-29. 2D As Neugebauer and Parker noted in their Additions and Corrections to Volume 1 which appeared in Egyptian Astronomical Texts 3 p. 272.

32

c E c C u L O
C E C bA Cd d C
L

u Ou It u O
C O

C Ov bA v cC b

bv
uu jj E _

C C

OL uv

OC "OX

Uu

c N .E E-" C
O bD

E b c
L d L h

C0 N L C

Ir

3'

41 L bD

33

The decans used in each of the seventeen star clock tables are shown in Table 20.

`X'

indicates that a decan is present somewhere in the star clock table. `XA' indicates that the decan only occurs in the epagomenal or list columns of the clock. Grey shading represents parts of the decan list that are missing due to loss or damage, rather than due to omission. Given that the sources are numbered very approximately according to age, Table 20 gives an indication of developments in the decan list over time. A15, A16, and A17 have been placed due in A8 the table to the similarities which they share with the first eight sources. after Some points of interest are detailed below:

Wit bk3t splits into two decansin A9, A10, All, A12, A13, and A14.
Kdty and "hnwy" become or are replaced by spty and hnwy respectively in A10, then merge

to a single decanspty hnwy in All, A12, A13, and A14, while `crew' undergoesa changeof
writing from A10 onwards. The decans labelled `13a' to `17' all refer to parts of a constellation called `Sheep'. Tpy-r in four the sources All appears sind to A14, while sind srt contracts to sind from A9 to A14

(both writings appearin A9 itself). The remaining Sheepdecansare presentthroughout. The four decans `18' to `21' have been placed in the following order by Neugebauerand in Al However, ('18'), ('21'). (119'), b3wy ('20'), 3hwy 3hwy imy-ht 3hwy Parker: tpy-11 and to A9, and A16, tmy-lit 3hwyprecedes3hwy throughout and b3wy is omitted. The surviving that decanssuggestthat this is the casein A15 and A17 also. Neugebauerand Parker state21
`imy-ht' means `following' b3wy be to decan ('19') therefore the 3hwy emended and must

b3wy 3hwy between body difference in the tables, the the and of main as throughout writing is only a matter of the type of bird drawn (S- for b3 and `s. for 3h). This leaves the order
in ('20'), (`18'), the tmy-ht `b3wy' 3hwy (written the by 3hwy with 3hwy clocks, tpy-r mistake)
21 Egyptian Astronomical Texts 1 p. 23.

34

unused decan 3hwy positioned between tpy-r 3hwy and imy-ht 3hwy. Three points argue against this emendation. Firstly, the alleged writing error occurs

uniformly across nine sources (including the clearest and least garbled) which span three of Neugebauer and Parker's `Groups'. This could, however, point to a writing error in a single source used to produce all the tables Al to A9, A15, A16, and A17.2 Secondly, in A10 all

four decans `18' to `21' appear, with b3wy written clearly, although the order cannot be

restored. Thirdly, the translation offered as evidence for the suggestionis not as certain as 23and also very Neugebauerand Parker state. 7rny-htis used in other astronomical contexts
commonly in royal contexts, where it has the meaning `in the entourage of', or

`accompanying'.

The word suggests a close association with, but subservience to, the

does It not preclude the possibility of precedingthe object. object.

Source
Al A6 A7 AS 123456789 123457789 134S6789 25 34567

Contentsof list and epagomenalcolumns

10 11 12 13 14 15 16 17 18 19 20 19 22 24 24 25 26 27 28 29 30 31 32 33 34 35 36 14ABCDEFOK1M 10 11 12 13 14 15 16 17 18 20 22 24 25 26 28 29 30 32 33 34 35 36 1 31A 10 11 12 15 16 18 20 19 22 25 26 28 29 30 32 33 34 35 36 1ACDKM 25 26 10 11 12 13 14 28 16 19 20 19 22 CD 28 29 24032 29 34 35 36 1AFAKLMM BCDEFLKM

Table 21: Contents of the list and epagomenal columns of diagonal star clocks Al, A6, A7, and A8

In the list and epagomenalcolumns of Al, summarisedin Table 21 using Neugebauerand Parker's labelling system,3hwyappearstwice, once before and once after lmy-ht 3hwy,but in the other three sourceswith preservedlist columns, 3hwyfollows tmy-ht 3hwy. 3hwyand Imy-ht 3hwy also appearin the offering formulae present in the horizontal strips of Al, A2, A7, A8, A19, and A13 with 3hwy preceding tmy-ht 3hwy. However, the decanal 24 does formulae in the not comply with the order in the accompanying table, and so order
22 If this is the case, it is another argument against the system of classification Neugebauer and Parker. by groups proposed by

24Removing formulae and other extraneous labels the order of decans in the horizontal strip is usually sind rsy, hrt. unit These hry, hrt, 43 bntt Imy-lit haw, 3hwy, lmy-brt b3w, hat 3hwy, spd, pt, rmn spd, nfr pkwy sind mhty, decans are organised into pairs by the interspersion of the hip dl formula, most pairs having two consecutive decans in list order. However, the pair spd, lmy-ht spd are sometimes replaced by tpy-r spd, [my-pt spd which decans the does not follow list order. overall and order of consecutive, are not

21For example in the dramatic text from the Book of Nut which will be extensively discussedin Section F.

35

cannotbe used as conclusive evidenceof decanalorder.

It must surely be preferable to follow the tables themselves, to read decan `19' as ?hwy as it is written, and to state that the order of decans propounded by Neugebauer and Parker, tpy-c 3hwy, lmy-ht 3hwy, b3wy should be restored to the order tpy-r 3hwy, Imy-bt 3hwy, 3hwy, which appears consistently in Al to A9, and A16. The Orion decans, labelled `26' to `29' in Table 20, are areas of confusion in decan lists up to Greco-Roman times. They make their first appearance in a timekeeping context on these

coffin lids. Al to A9, and A16 have the following order of decansin the main body of the
table: min hry, cbwt, hrt wert (upper arm, cbwt sceptre, under or lower leg). The only

occurrence of the decan rmn hry (lower arm) in any of the sources is at the top of the decan Al, The it hry between Ibwt. reappears column of which places penultimate and rmn in later decanlists from the New Kingdom onwards.

Decan Lists in Al to A9, A15, A16, and A17

Clearly, there is cohesionbetweenthe decan lists employed in Al to A8, A15, A16, and A17 into is two bk3t A9 details: differing 3t in the split portion of w. existing only certain with
decans, w. 3tf and bk3tf, and sind srt becoming simply smd. We shall use the label 0 for the decans in list Al to A9. A15. A16. and A17. ordinary of used general

body in A8 in A7, the of A6, Al, main and are complete that they contain thirty-six columns for decans the four listing the decans three the the table plus or columns and giving
decans days. In individual these appear tables, thirty-four each of ordinary only epagomenal in the main body of the table. The first triangle decan appears where we would expect it from the ideal model (see Table 1). The gap between the final ordinary pkwy b3w and the first triangle decan sind rsy is filled by repetition of the first two decans tm3t hrt and tm3t hrt.

36

A2 and A16 also show this feature in their curtailed tables. It would appear that in the in decans decan for list these tables, the thirty-four and eleven only ordinary source original
triangle decans appeared, out of a presumed thirty-six ordinary decans and twelve triangle

decans. Of the two ordinary decansand one triangle decan that e lacks which would be
like 1, Table the unique appearance of that to of a complete, working clock produce required rmn hry in the list columns of Al hints that this decan could be one of the missing ordinary decans. S3bw makes two similar appearances in the list columns of A6 (where it appears between tpy-r spd and ntr d3 pt) and A8 (between ntr d3 pt and pkwy s3bw which would seem

to be the correct order) and is a candidatefor the missing triangle decan.

Although Neugebauer and Parker postulated that their decan `21' b3wy should appear in

total the decan to A9, decan `19' Al to this add not so would would replace and sources believed decans Neugebauer Parker that they a correctly written used. state and number of for this do but be decans, the two evidence 3hwyto not provide any one of missing ordinary the of Al lists A9. however, is the It that occurrence to to the second of possible, addition is If the b3wy. this for in list Al be 3hwy the columns of could a single writing error name
fore. decans list list of ordinary columns of this source give a complete case, the writing If this

left has decan is no other the not accepted, remaining missing ordinary error

its identity. to suggest evidence

Decan Lists in A10 to A14

decade, for third the both have hry bpd knmt A13 SourcesA12 and n at the top of the column
has A14 A9. in only indicating a closeness age, and also a major revision of the table since it difficult makes which eight rows, decans, the determine intended to the positions of

A13 is decans ordered. A13 table well however, the surviving agree entirely with and the lists decan the in A12, 3hwy, but the existing parts of includes sfpt and which are omitted

37

otherwise agree. The table in A10 is jumbled25 and like A14 only contains four rows, but the decans used agree with those of A12 with three exceptions. Firstly, spry zinwy is split into spty and linwy (the only occurrence of these decans on coffin lids), secondly, tpy-r sind is omitted, and

thirdly 3hwyis included. Like A12 and A13, All has hry bpd n knmt at the top of the third column, but has diagonals from top right to bottom left, the reversedirection to that normally employed. The list used draws elementsfrom both O and the lists of the four sourcesA10, A12, A13, and A14, and
is the only list to retain the decans 'Irt, rmn hry sah, rmn hry sah, rmn sah, and spd. The

26 be third these hry but the fourth is unique hry, the of may secondand sameas rmn and rmn sources. amongtheseseventeen We shall attempt to construct a general list of ordinary decansfrom the lists of A10, All, A12, A13, and A14 which we hope will resemble the original source material used in the five these of sources. We shall label this list 4). construction It will not be as

incompleteness b it due 0 to to the and to construct straightforward as was compile five A10 the to A14. sources within confusion Twenty-four ordinary decans can be placed with certainty into (D. Twenty-one of these decansappearin the four sourcesA10 to A13 and in the surviving portion of A14 (seeTable 20), one (tpy-r 3hwy) appearsclearly in A10, A12, and A13 with tracesreadablein All, and two (tpy-r knmt and knmt) are missing from A13 due to being in a portion of the table which has not survived but can be restored with some degree of certainty. These twenty-four

u Egyptian Astronomical Texts 1 p. 18 fig. 12. 1 See errata to Egyptian Astronomical Texts 1 in Egyptian Astronomical Texts 3 where Neugebauer and Parker in had hry/l: their hry/hry they this were ry rmn that revised opinion to of that and problem suggest rmn state fact four distinct decans.

38

decans form the core of 4). We must now attempt to reconstruct as much of the remainder of 4) as we can. We shall rely mostly upon A10, A12, A13, and A14, but where these decan lists break off we must use information from the disordered All. After considering which decans might belong in (D,

we shall discuss the value of the list which we have created. Discussion of the remaining decans in P can be broken into six problems: 1) Decans included in three of the four sources: tpy-r sind (missing in A10), hntw hrw (traces in A10 and probably present in A13), and s3wy kd (missing in All). We include

hntw hrw in 4) with certainty becauseof its strong associationwith its pair j ntw hrw and
traces of the decan being visible in A10. The other two decans are present in the most

coherentsourcesA12, A13, and A14 so we also include them in (D.

2) S.pt is included in three lists, All, A13, and A14. Neugebauer and Parker (working

decan knowledge A13 A14) it be triangle along to of or an out of place considered without with tpcis and sb9sn, due to the disordered nature of All. However, its inclusion in the

the A13 into A14 bring interpretation raise and their question, and must orderly sources its inclusion in c decan. of as an ordinary possibility 3) The split of spry hnwy which occurs in A10 appearsagain in later decanlists and will also 27 hnwy do so we in lists four The those spty be noted connectionwith sources not split other in it decan in follow them (D. placing as one shall in is (it in A12 the decans: imy-ht is 3hwy 3fiwy. 3hwy 4) The 3hwy only and omitted but tray-ht belonging to likely A14) (D, be decan to and seems an ordinary missing portion of is decan in is A10. It triangle decan or is that this possible 3hwyoccurs only an out of place but there is these other are no neither of as explanations written, satisfactory entirely poorly
27SeeSection H.

39

obvious intrusive triangle decans in A10 and the inclusion of 'lmy-ht' is a strange mistake. We therefore include 3hwy in 1 but reject imy-ht 3hwy, both with reservations. We included b3wy in 4) due to its occurrence in all four sources, but it must be noted that its appearance in A13 only consists of traces in one cell and possibly is a badly written 3hwy. 5) The Orion decans: we have no alternative but to consider All from haw (which also

in appears A13) onwards as our model. This adds to tP up to six decans: b3w and rrt (both
of which also appear in O), rmn fry sah and rmn hry sah (which could be equivalent to rmn fry and rmn hry in O), rmn sah, and spd. We are reasonably comfortable with haw, rrt, rmn hry, and spd and include them in P, but we have not met rmn sah before and rmn hry was

inclusion in be both list decan decans in O. A the possible an a only may of ordinary in both 3hwy ordinary decanand a triangle decan in the samelist: haw, tpy-11 appear spd, and roles in O and rmn hry was also a triangle decanbelonging to O. It is possible that either, both, or neither rmn sah and rmn hry sah are ordinary decans. It is impossible to decide which of thesepossibilities to accept. 6) The inclusion of misplaced triangle decansin the list of All: tpds, sbfsn, rmn hry sah,

The have We discussed 3twy. Ihry other above. s3h already rmn sah, and rmn sah and rmn three seemreasonablycertain to be triangle decansbecausethey occur in parts of the list that in the other three sources. preserved well are Having discussedeach area of difficulty individually, we have now chosena framework for consisting of: twenty-four `certainties' + tpy-r sind + hntw hrw + s3wytd (seepoint 1 above)
+ spt (possibly, point 2) + spty j nwy (point 3)

40

+ 3hwy (possibly, point 4) + haw + rrt + rmn hry sah + rmn hry sah (possibly a triangle decan) + rmn hry (also possibly a triangle decan) + spd (all from point 5) Rather conveniently, this makes a possible total of thirty-six. We now must face three alternatives: either a) our list (D is complete and correct, or b) some All, A10, incorrect `possibles' be five the sources are or all of our and must replaced, or c) A12, A13, and A14 do not form a coherent group and a single list cannot be assembled from

the decanlists preservedin each individual source.

The first alternative is unlikely to be true. Given that we have only four sources, all of them being incomplete and two of them having serious problems in their layout, it would be

be definitive; 1 to to that is, that the contrived list 1 as presentedabove suppose unrealistic is an accuratereconstruction of an Egyptian decan list on which coffin lid star clocks were based. The most likely situation would be that All is missing some `Orion' and/or other

28 decans,and some `possibles' are misplaced triangle decansor mistakes This conforms to the secondalternative, which can also be satisfied by including some decanswhich appear `possibles' decans (spty, in A10 hnwy, lmy-dit again 3hwy) some ordinary with as and/or only being triangle decans. The third alternative must be realistically faced. While the decan lists from Al to A8 are
dissimilarities A9, list occur the to the same nature, and several of allied closely of clearly

between the lists of the five sourcesA10 to A14. The disorder in All

in difficult it makes

A12, A10, but base from theories soundly on evidence gathered this source, the extreme to A13, and A14 do agree reasonably closely with each other. We have already noted the
decan A14, lists that the A10 between A13 A12, the lists and of the and and of and similarity
28 SeeSection H, where the inclusion of s9ptwill be reviewed with referenceto later decanlists.

41

two omissions in the decan list of A12 that occur in comparison with that of A13 are the only differences between those lists. Generally, the five sources display the same major features in their decan lists and seem to have been designed from a similar source. A10 retains certain distinct features, and the

Orion areaholds the greatestuncertainty.

The decan list of A9 displays two features from 1, the splitting of wit bk3t the writing of the decan sind with a single occurrence of sind srt. In summary, we have identified two basic decan lists. We theorise that Al to A9, A15, A16, and A17 were created using a decan list resembling our list O, and that a later list to which

our list is an approximation was used as a basis for A10 to A14. The two lists 0 and are presentedwith brief annotations in Table 22, with disputed or difficult decansshaded. Spd is placed last in the two lists to aid comparison. Comparing 0 and P side by side, we seethat the two lists have nineteendecansin common, but differently four decans with names matching exactly, and whose names are written is decans, to the there Of a possible thirteen the same stars. which could refer remaining between in The li? dJt 4). b3w in hat O omission of t and pkwy correspondence andpkwy and `srt' from the decansind srt in the later sourceshas been discussedpreviously. One of O's decans,w.3t bk3t,is split into two decansin 4). We shall use Table 22 later to discuss possible revisions of the star clock tables, and our in in H decan lists in Section be the our useful seventeensources will also analysis of discussion of decan lists within celestial diagrams. We have divided our sourcesinto two discriminating degree flexibility, than traditions, eachof which retains a stringently of rather between sourceswhich display slight dissimilarities. This is a rather looser approach than but by Kahl, seems more in keeping with the nature of the sources (being that proposed

42

copies for symbolic purposes rather than observational records or primary timekeeping devices). Our findings reflect trends rather than a strict classification system, and the ability of the two lists we have created to include the more recently published sources is an indication that this approach is more likely to continue to be useful if more sources are discovered.
kamt

E)
t)

4) knmt
-r

Restored in A13 Restored in A13

saw kamt hry hpd n knntt hat h3H'

knnit hry hpd n knnit hat d3t

plivvy iw
tmlt hut
tm3t hrt

hw d3t
tm3t hrt
tnt3t hvt Split in A9 HIM

wst bkMt
ipds
sb. sn

bk?ti
s,' pt Omitted in Al0 and A12

hint hit hurt hit tnis n hntt kdty

"linwy

tpy-c hntt hurt hrt hurt hrt tins n hurt

spry lmwy Split into spry and hinny in A10

hr -ib vii? "crew" k(n)m

snid srt Becomessnid in A9

Inry-ib wi3 S917114, k(n)m

t) y-r srml Omitted in A10

sit s3wy sit hry Ppd srt t -C 3hwy

int -ht 3hwy 3hH'y

; y Very tentatively one of the two missing ordinary decans

sntd srt saw sit h )d srt hiy 31wy

t py-e 3hwy
; t )

Traces in Al I Not in A12

kd
haw

hntw hm
hntw ltrw Traces in A10 and possibly A13 Omitted in All
only survives in All and A12

"i t
h! y rrt
hry ! -11111 rum 3H't Itrt x'"rt lny Possibly one of the two missing ordinary decans

kd
s3wy kd
h3w `! t run turn rmn spd try ! 6311

only known from Al 1

Only known from Al l Only known from A11, possibly a triangle decan Only known from All , possibly a triangle decan Only known from Al l

hr)' s3{t sah

spd

ipy-r spd

Table 22: Decan lists a and (D

A Model Star Clock

In order to illustrate the problems which arise when attempting to analyse rising star clock illustration, imagine let the For construction this we of us a new star clock. tables, simplistic 43

assume that the necessary concepts of observation and timekeeping are already within our

grasp, and that we understandcompletely the instrument which we are trying to construct.
On the first day of the first month of the season Akhet, at the first hour of the night, we look at the eastern horizon, and write down the name of a bright star just rising. We wait one hour, then look again and find another bright star rising. We continue in this manner until

the twelfth hour. At the end of the night, we have a list of twelve stars. These form the first diagonal of star clock when headedwith the date. a column Ten days later, on the eleventh day of the month, we follow the samemethod of observation. This time, we notice that the star which had marked the secondhour will now be the first to appear. Similarly, the star which marked the third hour will now appear in the second,and so on until a new star will need to be chosen for the twelfth hour. We use this list as the in the star clock. column second
If the year contained exactly 360 days. continuing observations at ten day intervals would

The list thirty-six decan thirty-six clock stars. clock of columns using a produce a of exactly inaccurate be be hour to the after the sufficient obtain made would of night, and would only had time passedfor the effects of precessionto affect the clock. enough The Egyptian civil year was of 365 days, comprising twelve months of three decadeseach, is that for days `days five the What clock star the this epagomenal or upon means year'. plus (I Shemu decade in does 26th decan in first hour the table the twelfth the not mark the Middle decade) as would occur with a 360-day year. A new star will have to be chosen. This star will be a triangle decan and will lie approximately half way between the last and first ordinary decans29

29For a graphical representationand explanation seeEgyptian Astronomical Texts 1 pp. 108-109 figs. 30,31, 33.

44

decade,a new triangle decanwill be added to the list, until, during the For each subsequent
five epagomenal days, the twelve triangle decans will mark the hours of the night. On the first day of the new year, the first twelve ordinary decans will then come back into use, because there is only a five-day gap between the first epagomenal day and the first day of the first month of Akhet. This model produces an ideal rising star clock table in the form of Table 1.

Using the numbering schemeof Table 1 (which is not equivalent to the numbering of decans by Neugebauer Parker here in is 19, but Table and and repeated used as a convenient used be 1, A, describing for decans the the a general clock) system order of rising of used would B, 2, C, 3, D, 4, E, 5, F, 6, G, 7, H, 8, J, 9, K, 10, L, 11, M, 12,13,14, through to 34,35,36 followed by A, 1, etc. again. The solar year is not 365 days long exactly. The quarter of a day per year by which the Egyptian civil year lagged behind the solar year means that the star clock table remains `reasonably' accuratefor only twenty civil years3 After that time, Neugebauerand Parker 31many new decanswould have to be chosenif the clock was to be kept up to date. contend, Twenty years is sufficient to causedecan 1 insteadof decanB to mark the secondhour of the be first day. table the This final would the that the epagomenal column of means night on C, decan 2 Decan inaccurate twenty years after the table was constructed. would replace by be decan hour D 3 marked decan would replace would and so on, until the twelfth decan 11. A hour. first decan A be decan 36 for the new Neither nor would exactly correct marking decanwould have to be found between decan36 and decanA, which Neugebauerand Parker

30Egyptian Astronomical Texts 1 p. 108. 31Egyptian Astronomical Texts 1 p. 109.

45

labelled `decan36a'.
Neugebauer and Parker suggest that reorganisation of the entire table, not just the

be column, would epagomenal neededtwenty years after the construction of the star clock.
This would mean replacing each decan in the main body of the table. Decan 1 would be

replacedby decan A throughout, similarly, the other decans2 to 12 would be replacedby

decans B to M. Decan 13 would be replaced by a new decan, which rose halfway between

decan 12 and decan 13. Neugebauerand Parker label this new decan `decan 12a'. The
subsequent decans 14 to 36 would all be replaced, using the new decans Neugebauer and Parker call 13a to 35a. Decan A would be replaced by decan 36a and finally, decans B to M decans be by 1 to 11, just like the decans in the epagomenal column. replaced would

The new table would therefore require the choice of twenty-five new decans. The order of 11, decans L, be 10, K, A, 8, J, 9, 1, G, H, B, 2, 4, 5, F, 6, 7, the C, 3, D, E, of would rising M, 12,12a, 13,13a, 14,14a through to 34a, 35,35a, 36,36a, A, 1, etc. After another twenty civil years had passed,the epagomenalcolumn would once again be date. Decan 36 hour beginning first the the would and the epagomenal of rise at out of have be to rewritten as 36, A, B, C, D, E, F, G, H, J, K, L. would column The main body of the table, if corrected at the same time, would then be subject to another `half-decan shift' with decan B being replaced with decan 1 and so on. This new table The forty the only table closely. original star clock of civil years earlier very resembles difference is that a new decan,36a, now marks the first hour of the first decadeof the year, in decans diagonally. the does Decan M anywhere shifted are appear along other not and all table. The processof updating the clock at twenty civil year intervals would continue to keep this

46

timekeeping method viable over the centuries, using a total of seventy-three decans32 It is questionable whether the total revision of the table after twenty years would be necessary. The epagomenal column would be incorrect, but the main body of the clock would still be correct within a time scale of ten days without revision. While Neugebauer and Parker's revision cycle described above is technically necessary to

maintain a workable star clock, we need to compare the theory of a twenty year revision cycle, and of a forty year revision cycle, with the existing sourcesto find out whether such methodical refinement was indeed utilised, and if not, what sort of revisions did take place
basis they were made. on what and

Comparison of Revision Scheme with Existing Star Clocks

Of the seventeensourcesthat we are considering, we can see that Al to A9, and A16 (and hrt A15 have A17 had been tm3t this the tables and as well probably preserved)all portion of as the first hour of the first decade of the year. Al to A8 and A16 provide no evidence in for label We the themselves the time. year which adjustment of shall among clock over the original sourcefor Al to A8 was produced `yearn'. A15 is only partially preservedand from included derive in the A17, but it likely same ten tables that these rows are seems only source. A9 contains a small modification when compared to the year n tables. The third decanw3t bk3t has been split into two decansw33tl and bk3tt. This meansthat although decanstm3t hrt decans following in hrt in the lpcis the remain samepositions as and all year n table, and tm3t has The in is A9 which the writing of sind srt as sind only other change are shifted along. discussed. been already

32That is, 1 to 36 (36 decans) plus A to M (12 decans, as I is decans) 73 (25 36a 12a to making omitted) plus decans.

47

There are two possible explanations for the split of the decan w3't bk3t: copying error or deliberate reorganisation of the decan list. The first would be caused by either the copyists of the other sources (or the copyist of the prototype for the sources) mistaking the name of two decans for one (this would mean that one of the two `missing' ordinary decans of e

discussedpreviously would be bk3tf) or the copyist of A9 splitting a single decan into two
and propagating the shift throughout the table which the error caused. The second

possibility implies that the split occurred in an effort to update the table. A shift of one extra

decanmeansthat this correction would have occurred around year n+ 40. Tm3thrt and tm3t
hrt may have been kept as the first two decans because of tradition. The decan of the first hour of the first decade of the year is difficult to ascertain in A10, All

but has A12 A14 hrt A13, information. both In fm3t shifted and provide and clearer cases, along to mark the first hour in the sixth decadewith a new decan tpy-c knmt occupying the first cell in A12. The position of tm3t hrt indicates a date of n+ 200 years. The four decans between tpy-r knmt and tm3t hrt are knmt, hry bpd n knmt, hat d3t, and pkwy cl3t. It is A6, A2, in Al, is haw for that that the constellation appears gi3t either a new name possible A7, A8, and A16 in the sameposition in the decanlist, or that two new decanshave replaced h3t'zJw and pkwy haw, which are coincidentally another pair of it and pkwy (forepart and hindpart) stars. This implies an order h3t impossible is not which it pkwy pkwy ... ..., ..., ...,

if the decansdo not lie on a belt, but are chosenfor their rising times only. After tm3t girt and tm3t hrt A12, A13, and A14 have the decanswB3tland bk3tt. A13 and A14 then include sfpt, which does not appearin A12 but does appearin the disorderedAll. The

inclusion of sfpt leaves the remainder of the table of A12 out of step with A13 and A14. A date of n+ 240 is therefore also possible for A13 and A14. The dates discussed are

23. in Table summarised

48

Source Al to A8, A15, A16, and A17

A9

Year n
n or n+ 40

A10 All A12 A13andA14

? ? n+ 160 or n+ 200 n+200orn+240

Table 23: Date scheme for diagonal star clocks Al to A17

None of the sourcesshows the major changein decansthat would have occurred in year n+ 20, n+ 60 etc. Although it is possible that additional sourcesmay be found that conform to the pattern of revision twenty years after n or multiples of forty years after that, the current information leadsto the conclusion that this intermediaterevision did not take place.
The revision process as seen in the seventeen sources was more conservative and

observationally-basedthan Neugebauer and Parker suggest. We see that some ordinary decanswere also used as triangle decanswithin the samedecanlist, 33indicating that suitable intermediate decans were sometimes unavailable. We also notice that decans were

34 inserted disturbing but this seemedto be acceptableto the list, the or removed sometimes Comparing decans lists O the P two that which are common certain compilers. and reveals 5 decan different intervals in indicate both list. This that or all some to appearat each could
time just that to would than to so refer a group of star, may stars rather one single names leading to final between first the the the the stars rising of stars of asterism and elapse

differing intervals from list to list.

These factors indicate that the revisions were not achieved methodically by substitution of

had described but for in decan demonstration an the probably another as above, each
factors, to basis human such to and which subject was observational and observational

33For example tpy-l' spd in O.

'A Such as the splitting of wit bkJt into two decans. 35For example, there is a decan(s3wyknmt) between knmt and bry hpd knmt in but not in (D.

49

decans being as certain aesthetically or symbolically favoured, and the weight of pressures tradition. That the star clocks evidently were updated means that the tables were a sufficiently important timekeeping method with usefulness or symbolism somewhat current over several generations 36 that merited being kept

For this reason, they remain of prime Furthermore, the survival of the

importance in the history of stellar timekeeping in Egypt.

decans as a recognised group to the end of Egyptian religion signifies that their importance as a group of objects or as a symbolic concept was never forgotten.

Dating the Star Clocks

The positions of Sirius (spd) in the tables lead Neugebauer and Parker37 to date the origin of been have lists decan lists to lids in The Table 24. the the seem used on coffin as shown compiled only shortly before they were used on the painted coffin
Group I (A1-5) II (A6-8) III (A9) IV (A10) V (A11,12) Date of Sirius as the 12th hour star II Peret 21 or III Peret 11 II Peret 21 or III Peret 11 III Peret 1 ? Peret 21 assume1111 Date BC 2150-2100 2070-2020 2150-2100 2070-2020 2110-2060 1910-1860

38 lists
Dynasty IX-X XI

XII

Table 24: Dates of the star clock tables according to Neugebauer and Parker

It is obvious from the table above that the occurrencesof Sirius in the tables do not provide a basis on which to date the tables on the time scale of forty years. We cannot, therefore, give but it BC. for be between 2020 2150 definite to estimate n, value and a inclusion by from dating difficulties of 23 that, the or Table apart suggests caused omission

36 Borchardt Die Zeitmessung p. 8. A Ptolemaic water-clock decans were not visible. when 37Egyptian Astronomical Texts 1 p. 31.

inscription states that the instrument was used

38In Section H, we will seethat the preservedtriangle decansrelating to list are probably considerablyolder decans list with which they appearon the early coffin lids. the of ordinary than

50

dates: have decan lists far fall into decans, the two considered of we so clusters certain single
n to n+ 40, and n+ 160 to n+ 240. This spread of dates agrees with Neugebauer and

Parker's estimates for the decan lists as shown in Table 24.

We must now apply what we have learned about the composition of the early star clocks to the problem of understanding the decan lists which were used in timekeeping contexts during the New Kingdom, their purpose, and relationship to the earlier lists.

51

Section

B:

New Kingdom

Star Clocks

Introduction
Decan lists in a setting indicating a timekeeping use are rare after the First Intermediate

Period. Only three such occurrencesare known dating from the New Kingdom, two of them derivation. the same clearly of

Sources
B1 Star clock on the ceiling of the Sloping Passage in the Osireion at Abydos39 B2 Vignette from the Book of Nut, ceiling of the Sarcophagus Chamber in the Osireion at Abydos4 B3 Vignette from the Book of Nut, ceiling of the Sarcophagus Chamber in the Tomb of Ramesses IV (KV2)4'

The Osireion Star Clock

A fragment of a star clock table B1 survives on the soffit of a lintel in the Osireion complex

(Figure 4). The contents and disposition of the table have been described by Neugebauer 25. in Table diagram A is Parker. the table the schematic shown of surviving part of and The third column of Table 25 is headed by a month name `I Peret' and the label `Hour Name'. In the six cells below, the first six hours of the night are listed:
Her first period42 of h3wy43 Her second period in bk3t Her third period in bk3t Her fourth period in bk3t 44 ? She/her w. 3w

Her beautiful period in w. 3w

39 Egyptian Astronomical Texts 1 pp. 32-35. 'Egyptian Astronomical Texts 1, ChapterIT, Frankfort The Cenotaphof Set! I, Murray The Osireion at Abydos.
41Egyptian Astronomical Texts 1, Chapter II. 42 The word sp and not wnwt in this and the cells below. The use of wnwt to denote hours of the night is, Parker from from Pyramid 515a) Texts (Utterance known 320 the 251 269a, Utterance which however, (Ancient Egyptian Astronomy) infers that stellar timekeeping was established by the 24th century BC. He also into twelve hours is not confirmed until 2150 BC, refering to the date of the division the the night that of notes it the tables, diagonal be word sp nor the clock although wnwt star must that word remembered neither earliest from that date. in table clock a star appears 4 The words f 3wy, bklt and wJ3w are words for various parts of the night. 44Neugebauer and Parker suggest km3'creates'.

52

c' 0

'9

w =1 a)
rT4

"

co

3 zd

4c

Table 25: Layout of the star clock in the Sloping Passage of the Osireion

Figure 4: Star clock in the Sloping Passage of the Osireion (from Egyptian Astronomical Texts 1)

Neugebauerand Parker show that this table is of the samenature as the diagonal star clocks discussedpreviously, but arrangeddifferently. Eight decansare used and can be placed in
following by order Neugebauer and Parker's rearrangement of the Osireion clock into

diagonal star clock table format. In Table 26 we compare the decans used in B1 with those

in lists 0 and (D.

B1 and 9 have obvious dissimilarities, due to writings (knmw and k(n)m) or different names "crew", the (hnwy "jinwy", Ssmw represent s3pty and same presume we object and which and

53

kdty, sind and sind srt) or the more fundamentaldifference of the omission of tpy-r smd. 1 also displays some writing differences comparedwith B1, smw as opposedto Ssmwand k(n)m instead of knmw. The major disparity between the Osireion list and T is the by list hnwy into distinct decans. A10 differs from the general separationof s3ptyand also
decan, but this also omits tpy-' smd. splitting
B1 9 4)

S? Pty

kdty

bnwy
-1b w13 smw . knmw te sind sind srt h

"bnwy
h -lb w13 "crew" k(n)m sind srt sit

silty lanwy
-lb wt3 sSmw k(n)m tr sind sind srt h

Table 26: Comparison of decans from the Sloping Passagestar clock with those from early sources

The small portion of the list which survives is enough to suggest that the Osireion clock A9, Al to in O-based features A10 A14 the than with to clocks common with sharesmore A15, A16, and A17.

Date of the Decan List of BI

from B1 list in the in hry-lb A15 lid Since tables except as well as wf3 appears all the coffin

dealt have lists decan decan in dates Table 27 we to summarise the of the we shall use this
A12. find We B1 far. be that to appears contemporary with with so However, Neugebauer

for B1, date they dates base behaviour Sirius. To Parker their the obtain a on of and date Decade, Middle so hour IIII Peret and of estimatedthat spd would appearas the twelfth is from the A12, that list being than Osireion the sloping passage as slightly earlier beginning to the middle of the XIIth dynasty. during the Seti I, but is reign of Osireion to B1 the The attributed star clock was painted
list decan the that BC). Our BG) (1224-1214 (2150 to 2020 Merneptah means estimate of n date. 600 indicates the This discrepancy present than archaism years out of more was used

54

in the construction and decoration of the Osireion complex.

What we do not know is or

whether the diagonal star clock was still used for timekeeping (either symbolically practically) and was still being updated in the New Kingdom.

The difference in layout

between the Sloping Passage star clock and the coffin lid star clocks could indicate that the

tableswere still well enough understoodin the New Kingdom to allow such a transformation to be carried out, but of coursethe layout may also date back to the Middle Kingdom.

Source Al to A8, A16

A17

Position of h -ib w13 III Akhet, Last Decade,Third Hour

Only has ten rows

Year n
?

A9 A10 All
A12

IIII Akhet, First Decade,Third Hour Disordered Disordered

I Peret, First Decade, Third Hour

n +40 ? ?
n+160

A13 A14
B1

I Peret, Middle Decade,Third Hour Only has eight rows

I Peret, First Decade, Third Hour

n+200 7
n+160

Table 27: Summary of dating schemefor decan lists

It is more likely that the choice of a six-hundred year old decanlist indicates that this was the oldest available source, and that the timekeeping method was considered sufficiently diagonal be included in that to star the complex, which may mean venerable and arcane in for least time. form have had been the that tables, some at used we studied, not clock

Decan Lists in the Book of Nut

New in is the Nut Book text The a religious of which the earliest surviving example occurs of in 26 25 (Figures Osireion Kingdom on the ceiling of the SarcophagusChamber45 and of the Section F, which discussesthe text in detail). Part of the text also occurs in the tomb of la dating in I Valley IV Kings (Figure Carlsberg the the 27), Papyri and Ramesses of and from the secondcentury AD.

45The entire room representsa sarcophagus(hence the two large Nut goddesses on the ceiling, a characteristic design for the internal surfaceof sarcophagi).See Frankfort The Cenotaphof Seti I p. 27.

55

The main graphical feature of the text is a picture of the sky-goddess Nut being raised by her father, the god Shu, who is portrayed centrally. A decan list is written on the body of Nut in short vertical columns reading from her shoulders to her hips. In each column, one or more

decannamesare written followed by a star (* N14). Under each column, below the limit of the dividing lines, a number of circles appear denoting the number of stars associatedwith the decan(s). We shall call this decanlist a. In the space between the body of the goddess and the ground, a list occurs in vertical
columns. Each entry in the list consists of three dates separated by three labels. Some

line list belonging the to the the wavy also appear on entries arm of goddess and above a denoting sand. Some names of decans are associated with these extra entries, and other

decanalnamesappearas labels in other parts of the vignette. We shall call the list to which all thesedecansbelong P.
a knmt tr kamt liry hdn kamt h3t d3 t P!, dl t tm3thrt + tm3thrt
w. t1

Star circles 3 6 1 3 3 3 B3 __7_(B2) 2 1 3 2 3 3 3

4

a s3wysrt hh pd srt t l' 3hwy 3hw b3wy hnt hr

h hnt -lb

Star circles 4 1 2 4 B2 3 B3 5 3
3 46

bk3t1 sb3wmhw t -c hntt hntt hrt hntt hrt tms n hntt

st hnw

hnt hr kd saw td haw rrt Iwn sah

rmn by rmn h c sah spd s13t B2 wert hrt B3 sTh

1 3 2 15 B2 20 B3 2 6 B2 7 B3
1

hry-lb w13
s. mw knmw t rsmd sind

4
4 2 B2 3 B3 2 0

[nisdr) sah

3 B2 1 B3
1 3 3 3

srt
Break

wert hrt sah + wrrt hit sah

Table 28: Star circles associated with decans in the Book of Nut

Table 28 presentsthe decan list a, which can be fully restoredby looking at the versions of
46Neugebauerand Parkergive 4. B2 has only three, B3 could possibly have four. 47Neugebauerand Parkergive 20 for both sources.

56

the vignette from the Book of Nut in the Osireion (B2) and in the tomb of Ramesses IV (B3). The number of star circles appearing beneath each decan is also shown (with some list by between Neugebauer A break Parker). to the srt and given amendments and occurs distinguish Shu's head. The list forty decans does three consists of not and s3wy srt above

betweenordinary decansand triangle decans. Sevendecansappearin this list that we have not encounteredbefore: sb3wmhw, hry-lb lint,
Iwn sah, msdr sah, slJt, wrrt hrt sah, and wert hrt sah (which could possibly be identified with hrt wrrt which appeared in Al to A9 and A16). This list does not, therefore, belong to 0 nor 4), but is a new type of list. Neugebauer and Parker label the two occurrences of a in B2 and B3 as the only members of a family of decan lists called the `Seti IA Main Group'. Two subgroups of `Seti I A' exist,

differing from the `Main Group' in presentation (having deities, planets, and a formal triangle list, all of which are lacking in B2 and B3, and also differing in the number of stars decan). `Seti IA with each associated Subgroup A' consists of four closely knit XXth

by otherlists'49 dynasty sources plus one later source (Petosiris48 c. 150 BC which is `influenced

has B' `Subgroup decans in list triangle are preserved). and a and which only sevenordinary Ptolemaic). dynasty lid (XXXth to Nekhtnebef5 the early sarcophagus of only one member, None of the decan lists in either of the subgroups is displayed in the same contextual A' `Subgroup lists in `Main Group'. which The four the two the of as members arrangement Valley in the in the of XXth dynasty the tombs pharaohs of exhibit great cohesion all occur layout four decan All the list of the same Kings. occurrences of are associated with

48Number 48 in Egyptian Astronomical Texts 3 pp. 64-67 including a diagram of the ceiling, which is reported to be lost. 49Egyptian Astronomical Texts 3 p. 119. 50Number 39 in Egyptian Astronomical Texts 3 pp. 53-54 and pl. 25.

57

includes Ramesside star clock tables, two types of circumpolar that ceiling, one astronomical s1 decan list, discuss layout later all of which we shall group and another The decan list a differs from the list from the four main ceilings of `Seti IA Subgroup A' and the list of Nekhtnebef ('Seti IA Subgroup B') in three main points: a separates w. 3tt

bk3t1into two decans w. 3ti and bk3ti, includes two extra decans wrrt hrt sah and wrrt hrt sah, list. decans follow the triangle the planets and main section which usually and omits In

includes Nekhtnebef one extra decan tpy-r b3wy which occurs in no other member addition,
of the `Seti I A' family. Neugebauer and Parker identify the appearance of the decan sb3w

mhw as the unique feature of the `Seti I A' family. DecanList Seti IA w3tfbk3tf Triangle tpy-r b3wy

w. 3tf bk3tf

Book of Nut I Source

Ramesside Source

Petosiris

OsirenII 62

Rams s IV 1111 RamessesVI B3 2 occurences

RamessesVII

RamessesIX

Nekhtnebef

Main Group
Figure 5: Relationship

SubgroupA
between sources belonging to the `Seti I A' family

SubgroupB

Figure 5 illustrates the relationships between the members of Neugebauer and Parker's `Seti I A' family. Considering the probable nature of the sources from which the layouts of the

the tying drawn, `Seti I A' be eight with there to of problem are seems a members eight The two family Parker together Neugebauer the suggest. as strongly as and sources of four Nut Book from The ceilings the astronomical of are obviously of one source. examples

51See Section H.

58

from the tombs of three Ramesside kings are based on one other source, which we shall call `the Ramesside source'. The Book of Nut contains, in summary, the following elements: written information about the motions of the stars and the sun; tabulated data about the motions of the stars with a fragmentary decan list; a description of how to make and use a type of sundial and a diagram list labels; instrument; the text of the a and a graphical representation of cosmos with of forty-three decans. The Ramesside source contains the following elements: two different circumpolar groups; two different decan lists with deities, planets, triangle decans, and pictures of constellations;

star clock tables. and twenty-four Ramesside Apart from some similarities between the decan lists, the Ramessidesource therefore bears decan between the differences Book Adding Nut to the the source. no resemblanceat all of lists to this observation leads us to believe that the `Seti I A' family, as described by Neugebauerand Parker, cannot realistically be imagined all to have derived directly from a decan list. is It far four likely that separateoriginal sourcesexisted ancestral more single (one for the Book of Nut, the `Ramessidesource', and two for the lists of Petosiris and Nekhtnebef) which each happen to contain sb3w mhw and share a few other common features.

Decan Names Occurring Outside List a

List a is one of two decan lists which appear in sources B2 and B3. The other list is

fragmentary and, as we shall see, its composition and function are controversial.

decan labels labels Nut contain Short text are scattered over the vignette. Some of these In B2, her dates list Nut to the beneath arms. of and on which occur nameswhich pertain hrt decans body Nut two decan the These are wrrt of namesappearon two under the list a. rrk (under hry hry) lipd knmt B3, is In the appears hat d3t). (under rmn name and sb9sn n and

59

between joint in but B2, Parker rrk Neugebauer rrt. that the is state occurs and also under

discern. difficult to the name make ceiling slabs In the break between the two halves of the decan list a, there is a label
j o 2' spdr5. o .

A*

`The act in the first month of Akhet at the time of the going out of \,j `Knurr lives, with cb and

Near Nut's knees we have

stw' or `Life (of) knnit,with 'b and Stw'. rb could be an abbreviated form of rb9s, an alternative

labels Two `life': other start with of sb9sn. writing

I i

`life of stw' or `stwlives' and

`life of knmt' or `knmtlives'. One further label standing alone in the vignette reads .
ooo,

The first word is unknown, the second is knmt.

We have now identified six decannamesoccurring in labels: wrrt hrt, I'b (sb.sn), is rrk, spdt, knmt, and 3tw. A large proportion of the rest of the Nut vignette is taken up with a list of dates. Each entry in this list comprisesa set of dates in the form: `ec by decan labels The names. are accompanied R R date Mn date'. A few date sets and Mn difficulties, present no reading

Oo, is however, label in The more `Enclosed Duat53' `Birth'. the respectively and meaning `hour' the Hour' Neugebauer Parker `First word difficult. although and use the translation
in text the in body `hour' inclusion The does not appear the occurs only of the table. of wnwt left the to horizontally above Parker label V Neugebauer right text and which runs which body line the main of the table. This text states: under sand Of thatwhich is between thestarwhich makes thefirst hour
and the star which is enclosed in the Duat: there are 9 stars.

Between the star which is born and the star which makesthe first hour: 20 stars
giving 29 of those living and working in heaven.

in the dramatic text54 Nut `head' has in Book the the Tp also which occurs meaning of of

SZ Prt spdt - the heliacal rise of Sirius (see Faulkner Dictionary 53The Egyptian `Netherworld'. 54See Part II.

p. 91).

60

context of meaning the part of the star that is visible to man. For the moment, we shall use the three words `First', `Enclosure', and `Birth' to denote the actions that occur on the three dates in every set. The list is cyclical and reads in an anti-clockwise manner. The labels and date sets (here in 1 for dates furthest line) left the the shown set of are numbered starting at set sand along

Table 29 in the sameorder in which they appearin the Nut vignette.

1 3t1yy lives First IIII Akhet 16 Enclosure II Peret26 Birth I Shemu6 knurl lives 2 First 1111 Akhet 6 Enclosure III Pent 6 Birth 1 Shemu 16 /try 3 4 It d 3t First 1111 Akhet 26 Enclosure 111 Peret 16 Birth II Shemu 16 Pkwy J3t 4a First Enclosure Birth 5 First Peret6 Enclosure Peret6 Birth Shemu26 tm3t hrt ha First IM Akhet 16 Enclosure till Peret 16 Birth 1 Shemu26

hpd
kramt

Table 29: Date sets and decan names from the Book of Nut

Three more date sets (two of them identical) and two decannamesoccur on the arm of Nut. One date set appearsunder Nut's breast, directly under the label sb,sn. These are shown in Table 30. The label sd is thought to be a garbled attempt at ipcis.
Under breast 8 tsn sb. First II Perel 6 Enclosure I Shemu6 Birth III Shemu 16 =7 First ? Peret26 Enclosure III Peret26 Birth III Shemu5 Near shoulder 7 bk3tf First 1 Peret26 Enclosure III Peret26 Birth III Shemu 5 SJ_ Near wrist 6 First 7 Peret26 Enclosure III Peret 16 Birth 11Shemu26

Table 30: Further date sets and decan names from the Book of Nut

The rest of the list occurs in columns which fill the areasto the right and the left of the god
Shu, under the body of Nut. No decan names occur near these entries. The first entry

dates breast (see Table 31). the the set of under repeats
There are thirty-nine entries in the complete list including the entries on the arms and above

leaves This dates. duplicates (=7 Two line. (4a) are the sand and =8) and one contains no
distinct entries. thirty-six

We can begin to construct a decan list made out of the six decanswe collected from labels

61

and the six new decanallabels associated with date sets(fry bpd knmt, hat d3t,pkwy d3t, tm3t
hrt hrt, bk3tf, and sd). We can use other decan lists to reconstruct the order in which the decans should appear, and we note that bk3ti and sd (lpds) are in reverse order on Nut's arm. The order of the fragmentary decan list, which we shall label , is given in Table 32.
=8 9 10 11 12 13 14 15 16 17 18 19 First II Peret6 EnclosureI Shemu6 Birth IIII Shemu 16 First II Peret 16 EnclosureI Shemu 16 Birth IIII Shemu 26 First II Peret26 EnclosureI Shemu26 Birth IIII Shemu 6 First III Peret6 EnclosureII Shemu6 Birth IIII Shemu26 First III Peret 16 EnclosureII Shemu 16 Birth IIII Peret26 First III Peret26 EnclosureII Shemu26 Birth I Akhet 26 Peret6 EnclosureIII Shemu6 Birth II Akhet 6 First 1111 First IIII Peret 16 EnclosureIII Shemu 16 Birth II Akhet 16 First IIII Peret26 EnclosureIII Shemu 26 Birth II Akhet 26 First I Shemu6 EnclosureIIII Shemu 6 Birth II Akhet 6 First I Shemu 16 Enclosure1111 Shemu 16 Birth II Akhet 16 First I Shemu26 Enclosure1111 Shemu 26 Birth II Akhet 26 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 First II Shemu 6 Enclosure ? Akhet 6 Birth III Akhet 16 First II Shemu 16 EnclosureI Akhet 15 Birth III Akhet 26 First II Shemu26 Enclosure 7 Akhet 6 Birth III Akhet 6 First III Shemu 6 EnclosureII Akhet 6 Birth III Akhet 16 First III Shemu 16 EnclosureII Akhet 16 Birth III Akhet 26 First III Shemu 26 EnclosureII Akhet 26 Birth I Peret6 First 1111 Shemu6 EnclosureIII Akhet 6 Birth I Peret 16 First 1111 Shemu 16 EnclosureIII Akhet 16 Birth I Peret26 First III Shemu 26 EnclosureIII Akhet 26 Birth II Peret6 First I Akhet 6 EnclosureIIII Akhet 6 Birth 11Peret 16 First I Akhet 15 Enclosure IIII Akhet 16 Birth II Peret26 First I Akhet 26 Enclosure 1III Akhet 26 Birth III Peret6 First II Akhet 6 EnclosureI Peret6 Birth III Peret 16 First II Akhet 16 EnclosureI Peret 16 Birth III Peret26 Peret6 First II Akhet 26 EnclosureI Peret26 Birth 1111 Peret 16 First III Akhet 6 EnclosureII Peret6 Birth 1111 First III Akhet 16 Enclosure11Peret 16 Birth II1I Peret26

Table 31: Date sets without decan names from the Book of Nut

We must now turn to the question of which decan applies to which set of dates. Although the sequence of date sets is from left to right along the sand line, the direction of writing of line, date decan Along is the left to sand throughout the set and name each right vignette. decan names could either be written at the head of their date sets (that is, to the right of the dates) or be written at the end of their associated date sets (that is, to the left of the dates). For example, set 3 could belong to either hat d3t or bry bpd knmt. The positioning of hry bpd knmt near to set 2 with Shu's legs between the decan and set 3 makes it clear that the first

is likely: the the name of the decanheadsits date set. more alternative

wort hr t spdt 3tw knmt hr bpd knmt hat d3t hw dUt tm3t hrt hr t bk3r1 sd sbSsn Table 32: Fragmentary decan list from the Book of Nut

62

Associating set 2 with hry bpd knmt also fixes the decans Stw (set 36), knmt (set 1), hat d3t

(set 3), andpkwy cl3t(set 4). We next have the label tm3t hrt hrt, which we have previously
55 in e decans, b hrt hrt. We also have the problem tm3t and as two encountered and tm3t that bk3tf appears alone as a label. In every other decan list we have seen, either the single

decanbk3t1is precededby the single decanw 3tl, or the two appearas one decanw. 3t1bk3ti.
The two problems together give four possible combinations: Case 1: tm3t hrt hrt, ww3tl bk3tl (two decans)

Case2: tm3t hrt hrt, w. 3tf, bk3tf (three decans) Case3: tm3t hrt, tm3thrt, w 3tt bk3ti (three decans) Case4: tm3t hrt, tm3thrt, w. 3ti, bk3tf (four decans) Neugebauerand Parker favour the first combination and tie this decanlist to a group of later lists which they call `Seti I B'. They ignore the empty date set 4a, which appearsat the for hrt tm3t position reinforcing the probability that the name label refers to two correct decans. The next occurrence of tm3t hrt hrt written56 as a single decan happensno earlier 57 300 BC However, the same is true for is rrk which makes its first appearance than about in the two New Kingdom sourcesB2 and B3. If we look at the possible decan configurations in combination with the position of labels 4 date find 1 following to based the to above: sets, we cases possible cases, on next Case 1: (tm3t hrt brt, w. 3tf bk3tt) Sd (tpds) is correctly placed and 03tt is both out of place is that is to Sb3sn 8 badly set. first the related and written. placed near set occurrenceof and Case2: (tm3t hrt hrt, w. 3ti, bkRti) The unlabelled set 6 would apply to wg3tt. BkMtlwould be
55Note the changein writing from tm? t to tm3t.
56Figure remains in the tomb ceiling of Osorkon II in Tanis (c. 832 BC). 57 According to Neugebauer and Parker's classification in 'families' this decan should appear in source 35 `Tanis' of Egyptian Astronomical Texts 3 (pp. 44-48 and p1.23) which dates from around 600 BC, but is not preserved.

63

correctly

written

and placed in front of the second occurrence

of its date set 7. Sd (ipcls) is

then out of place as well as being badly written having set 9.

and must be associated with set 8, with sb.sn

Case 3: (tm?t hrt, tm3t hr-t, wflti bk3ti) Tm3t /n-t would be associated with set 5 and tni t !n-t front in leaves label (for bk3ti) This `bk? 6. ? the of the ti' ti correctly placed w. with set 9, 8 7. ds) be S(l (i1 as and sets of set occurrence associated with and sb..sn must second happened with case 2. Case 4: (tnr)t hrt, tnift hit, xw. ?ti, b0i) Both the writing of bkfti and the empty date set are

both accounted for, but the labels bk3ti, sd (ipdr) and sb, s are all out of place. Table 33 illustrates the four cases. Shaded decans indicate that the name appears correctly positioned in association with its date set in the Nut vignette.
appear at all in B2 or B3.
Date set Case 1 Case 2 Case 3 Case 4

Note that `w.o,ti' does not

I 2 3
4

kamt In y hpd kamt hat d?t

plna. v d? t

knnrt lay hpd kamt hat d?t

phui v d3t

knurl hry lrpd kramt h?t c13t

phh'v (131

krnnt lny lrpd kramt hat d?t

/Ir"'y d3!

5
6

lnr3! hri Ort

w.i`11/hkIri

tm31/hrt lnrt
uu'I?!i

tnrlt hit
tnmit l117

halt 111-t
milt hit

7 8 9
10

sd (ip(ls) sb.Nn

bk3tl sd (ipds) SMS11

wVtf bk3tl scl (ilucls) slstsn

Mi _+'. bk3 scl (ipcls)

sb, Sn

Is 'rk 35
36

tv"l-t /n-t spill

is "rk H rrt 1117 spilt

tH''

is irk wrr! 111-t s/rrlt

1N')'

is rrk 1+'`rtII /I Spilt

1 ttvvy 1,

Table 33: Possible combinations

Nut Book in the decan date of sets of names and

As we have already noted, the list produced by accepting case I conforms dating lists mainly of from the Greco-Roman

to that of a family The lists

Period called the `Seti 1 13' family.

64

formed by the other three cases match no other list exactly, although case 4 differs from All only by the omission of a single decan (s9pt).

Given the unique nature of the Book of Nut source, it is not impossible that the list should be unique among the known decan lists, in the same way that we found a to be. The
decan in rrk the is of much later lists only indicates that this name was still used recurrence

for a group of stars and does not prove conclusively that the later decan lists descended
directly from the Book of Nut source. The same argument applies to the possible fusion of

tm3thrt and tm3t hrt into one decan.

It is definitely not safe, given the arguments above, to state with certainty that the decans

be family', from IB this from `Seti later the the as can supplied of members missing 2,3, 4 been have that cases and as outlined conclusively eliminated. presupposes above
It seems that Neugebauer and Parker's argument in favour of case 1 is that a family exists

into which the list (in case 1 form) can be inserted,and their dismissal without discussionof into is 4 based `family' 2,3, which the that exists and again on observation no obvious cases 58 fit thesepossibilities will in has list is This a unique role clearly not an acceptable argument, especially as this lists discussion Indeed, timekeeping of Egyptian astronomical our as we shall presently see. lists decan between within that the overemphasison relationships a and suggeststrongly different layouts is not constructive, and conclusions drawn from associatingsuch decanlists
be family be to ties are questioned. once such weak shown can
58Egyptian Astronomical Texts 1 p. 56 which again mentions tm3t hrt Tartas a single decan concerns portions of Papyri Carlsberg 1 and la. A decan name written dm is used in the phrase `Opposite knmt to dm are they, these five stars'. Neugebauer and Parker take 'these five' to be knmt, luy tipd knmt, h3t 43t, pkwy 43t, and tm3t hrt hrt, /irt hrt, having already decided that this is a single decan. There is no given reason that dm is dm tm3t that and hrt to tm3t only. cannot refer

65

The Book of Nut as a Star Clock

We shall now look more closely at the list of dates. One noticeable feature is the regularity of the dates which occur in the table. With the exception of writing errors, only the 6th, 16th, and 26th days of the month are mentioned. There is a period of exactly three months betweenthe datesof `First' and `Enclosure', two months and ten days between `Enclosure' and `Birth', and six months and twenty days between `Birth' and `First'. No account is days. the epagomenal madeof
Date set First Enclosure Birth

1
2

III Akhet 26
IIII Akhet 6

II Peret26
III Peret 6

I Shemu 6
I Shemu 16

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
35 36

IIII Akhet 16 IIII Akhet 26 Peret6 I Peret 16 Peret26 II Peret6 II Peret 16 II Peret26 III Peret 6 III Peret 16 III Peret 26 IIII Peret 6 IIII Peret 16 IIII Peret 26 I Shemu 6 I Shemu 16 I Shemu 26 II Shemu6 II Shemu 16 II Shemu26 III Shemu 6 III Shemu 16 III Shemu 26 IIII Shemu 6 1111 Shemu 16 IIII Shemu26 I Akhet 6 Akhet 16 Akhet 26 II Akhet 6 II Akhet 16 11Akhet 26
III Akhet 6 III Akhet 16

III Peret 16 III Peret26 IIII Peret6 IIII Peret 16 IIII Peret26 I Shemu6 1Shemu 16 I Shemu 26 II Shemu6 II Shemu 16 II Shemu26 III Shemu6 III Shemu 16 III Shemu 26 IIII Shemu6 1111 Shemu 16 IIII Shemu26 I Akhet 6 I Akhet 16 I Akhet 26 II Akhet 6 II Akhet 16 II Akhet 26 III Akhet 6 III Akhet 16 III Akhet 26 IIII Akhet 6 1111 Akhet 16 1111 Akhet 26 I Peret 6 1 Peret 16 I Peret 26
II Peret 6 II Peret 16

I Shemu 26 II Shemu26 II Shemu 16 II Shemu 26 III Shemu6 III Shemu 16 III Shemu26 IIII Shemu6 IIII Shemu 16 IIII Shemu26 I Akhet 6 Akhet 16 Akhet 26 II Akhet 6 II Akhet 16 II Akhet 26 III Akhet 6 III Akhet 16 III Akhet 26 Akhet 6 1111 IIII Akhet 16 IIII Akhet 26 1 Peret 6 Peret 16 Peret26 11Peret6 11Peret 16 II Peret26 III Peret6 III Peret 16 III Peret26 IIII Peret6
IIII Peret 16 IIII Peret 26

Table 34: Idealised version of Text U from the Book of Nut

66

The regularity of the list makes it easy to construct an ideal list of date sets by restoring lost signs, correcting writing errors and eliminating duplicate entries. This ideal list is shown in Table 34. The period of two months and ten days between `Enclosure' and `Birth' is clearly the is invisibility is decanal which of star and seventy-dayperiod which a characteristic of a 59 The in in `purification the text the accompanying earth' as mentioned a period of The is `Birth' heliacal the the rise. of star astronomical event called reappearanceor `Enclosure' of the star represents the disappearance of the star at the beginning of the period of invisibility. This leavesthe event called `First' to be determined. After `Birth', the star rises each night slightly earlier than the night before, and reachesa higher altitude in the sky before the light of the sun causesthe star to become invisible. Eventually, the star will have risen before night fall and will appear after sunset at some date Parker that the At denotes Neugebauer `First', the the text state which and altitude. as 60 This hour first the transits the action, the the night of meridian at associatedstar end of for First is V in `First' text that the table usesthe word used since which and the assumption
Hour, the table must be intended to mark the hours of the night, has led to this type of list 62 Furthermore, the being called a transit star clock61 and the list 0a list of transit decans.

decans' `transit form family labelled lists `Seti well. I B' the as which subsequently are other The collection of date sets 1 to 36 together with a complete decanlist would createa type have in different from diagonal previously the we nature star clocks which of star clock The is different, table. instead being The list a as of arrangement presentedas a studied.
59Seediscussionof Book of Nut in Section F.

60Parker 'Ancient Egyptian Astronomy' p. 56 states that `Since a star spends 80 in the east before working, it is is D, is it in Section the it [the hour'. transiting In fact, event not star] when marks an as we shall see clear that implies. Parker `clear' as as

61Clagett Calendars, Clocks,and Astronomy pp. 56-59. 62 Egyptian Astronomical Texts 1 p. 115.

67

lack of emphasis on hours in the two date lists (those contained within B2 and B3) is balanced by a greater interest in the yearly cycle of the stars, although the epagomenal days

decan ignored The be by the to entirely. clock would used observing are meridian seewhich
is transiting and, by knowing the date, finding the hour using the date list.

The theory of the existenceof such a method of timekeeping will be examined in detail in the context of the developmentof stellar timekeeping in Section D.

68

Section C: Ramesside Star Clocks

Introduction
A type of timekeeping device consisting of a set of twenty-four tables occurs in three tombs

of Ramessidekings. The tables are duplicated in one tomb providing us with four sources, Cl to C4. All four examplesoccur in the samecontext: as part of astronomicalceilings. Each table contains afigure, a grid, and some lines of text. The figure is that of a kneeling, bearded man portrayed full face, and is placed beneath or in front of the grid. The grid N14 lines fourteen horizontal lines. Usually thirteen stars of seven63 vertical and consists horizontal located between the the the thirteen on grid, each occupying one of are spaces lines and positioned on one of the sevenvertical lines. A set of twenty-four tables is known as a Ramesside star clock.

Figure 6: Example of the layout or a Ramesside star clock table

63Additional vertical lines may be present,but are not significant to the timekeeping method and as such will be ignored in the following analysis.

69

The schematic arrangement of one of the tables is given in Figure 6. The date consists of a 4 half by 16, followed The first table, which `[day] the month'. month name either alone or has II 16, T3 is dated has date label I Akhet, I Akhet T1, T2, the table, the second we shall Akhet, T4 has II Akhet 16 and so on until IIII Shemu 16 in T24. epagomenal days. We 12-nwt. The hours are written as tp grh, wnwt tp, wnwt 2-nwt, wnwt 3-nwt shall wnwt ... beginning in (the hours denoting the by the tables to these of tp conjunction with grh refer (the first hour) as "1, wnwt 2-nwt (the second hour) as -2, and so on to tp night) as -0, wnwt The by So denoted T14.3. hour in 14 be 12-nwt the third table position as would wnwt -12. of each star is indicated by the use of one of seven phrases (discussed below). No table exists for the

No other text, either occurring with the tables or from other sources,deals with this type of star clock. Sources65 VI, KV9 Cl Ceiling of Tomb of Ramesses VI, KV9 C2 Ceiling of Tomb of Ramesses VII, KV7 C3 Ceiling of Tomb of Ramesses IX, KV6 C4 Ceiling of Tomb of Ramesses Previous Interpretations of the Tables

Neugebauerand Parker sum up the study of Ramessidestar clock tables as follows:

1872). in Gensler (so hypothesis [was] hours by that their Champollion's the the '... risings stars represent
Only in 1874 did Le Page Renouf take the coordinate net seriously, give the proper explanation of the he and clock) (though a star took the tables as not and as a calendar of astronomical observations procedure brilliant Renouf's Page Le BC. for it date 1450 designed which was originally as about correctly establish the for long a overlooked while. paper was apparently Brugsch in his Thesaurus (1883) still followed Gensler. In

64Egyptian Astronomical Texts 2 p. 7. 65All are presented in Egyptian Astronomical Texts 2.

70

1891 Bilfinger tried to disprove Renouf's approach and return to the hypothesis of risings in order to obtain a

more even distribution in the length of the resulting hours - at the price of declaring both star charts and target figures as meaningless additions to the texts. Finally, Schack-Schackenburgin 1894 rediscovered the

lines. interpretation Five later, Borchardt, transit the the a years as transits, suggested accepting of significance have been line instrument by known the the used to means of which plumb as could mrit specific procedure ' them. observe

The acceptedinterpretation is that the Ramessidestar clock is a method for finding the hour is by knowing half the the the current year and observing which star either night month of of is line), (represented by or the star appearing on the central vertical exactly on the meridian
just approaching or leaving the meridian (represented by the star occurring on one of the six

other vertical lines, three each side of the central line).

It can easily be seen from

examination of the extant star clock tables that the seven phrases mentioned above each The be lines to the placed. could one of seven vertical on which a star correspond is illustrated below in Figure 7. relationship
II t14
wnmyllmnty

msIdr
wnmyllmnty

trIt
wnmy/tmnty

r rk3 tb

trt
13by

msIclr
t3by

kh
MY

Figure 7: Position labels for the seven central vertical lines from a Ramesside star clock table

The three words k"ii, mscir, and Irt are parts of the human body. Krh is the shoulder, elbow (side) is is left-hand and in or eastern 73by the the ear and eye. msdr can mean either arm, or
The both (side). Imnty 7mnty western. also means the two terms wnmy and mean right-hand

is ck3 tb between imnty. indicating r The the middle position wnmy and phrase writing varies 66 forward which meansstraight The method employed by the clock is therefore that the starspursue their east-to-westcourse it, A line the continues from and the star approaches the crosses sky. central east, across
66FaulknerDictionary p. 50.

71

lines The to watch tables the observer a configuration of star give and vertical westwards. for, and at the moment in time when the star coincides with its predicted position, the hour is fixed. Due to the disparity between the length of the solar day and the sidereal day, the star into line division The twentythe the slightly earlier cross central each night. of year will

four fifteen-day periods, each with a corresponding table, accounts for this retrograde decanal in discussed diagonal Unlike the the stars star clocks previously, the stars motion.
from hour in do Ramesside table the star clock not automatically shift an and uniformly used

to table. This is the acceptedinterpretation of the Ramessidestar clock tables. In order to discussthis interpretation, we must note again that we have no further contemporary information about this timekeeping device at all, so all discussion must be based solely on the four sources. The discussionwill be centred around three elementsof the Ramessidestar clock: the seven lines, the significance of the figure, and the list of starsemployed. vertical

The Vertical Lines

tb has been identified by Le PageRenouf and Schack-Schackenburg as The middle line r k3 literature Egyptian but further the north-south meridian, since, as previously mentioned, no is available on this subject, the identification must also be consideredan assumptionrather than a fact. The nature of the six other vertical lines is open to various interpretationsdue to the problem The figure are suggestions the usual onto a spherical a plane sky. model of night mapping of that the vertical lines representone of the three following possibilities: 1) they are lines of equal right ascension 2) they are some kind of physical grid which was held above the seatedman 3) they are verticals dropped using a plumb line producing lines of equal azimuth

72

The first suggestion implies some means of angular measurement, either mechanical or calculated, for which we have no evidence throughout Egyptian science. The second

suggestion springs presumably from a literal interpretation of the diagram, and the third is 67 instrument, by instrument the existence of a suitable the surveying suggested called a mrjit. The relevance of a seated target figure is not clear for the second suggestion. Such a system

of verticals would be a matter of rigging a frame and strings or thin laths, and would not need a human beneathit. Such a system, if built once, could be left in position for the stars to be consultedat any time. Another possible construction would be to use an opening, such doorway, together with threads. The observer would need to stand or sit at one as a particular point in front of the frame to obtain consistent readings, and the range of stars be could observedusing this system would be increasedas the distancebetween the which observerand the verticals decreased. Considering the evidence available, this solution is not completely impossible, but is also not entirely credible. The third suggestionis perhaps the most likely. The observerwould hold the instrument at (shown body length intersected his that the as the so string partner's arm's required part of the seatedfigure), the right eye, the left ear, etc. Again, the distancebetween the two men have be ideally to place the time they taken, same would each a reading was and would themselvesalong the sameline. (If the assumptionof the central vertical being the meridian is upheld, the observerwould always need to be due north of his partner.) The rangeof stars line, the depends distance between length the the or the the on plumb of men and available line. indicated to line by the capacity the extend mentally plumb observer's For neither of thesetwo suggestions(frame or plumb line) can the starsusedbe further north
than the zenith, and practically, the stars would be between 10 and 70 in altitude.

67This instrument, consisting of a wooden L-shaped handle supporting a plumbline, will be mentioned again in in Section E. clocks shadow with connection

73

These suggestions arise solely from the grid part of the star clock tables and make varying use of the target figure which is included in each table. However, on first glance at the

tables, there is the temptation to think of the grid as positional in both axes, that is, the

is having This bearing the the the of star symbols some star. vertical position on altitude of
be definitely incorrect: however It the the time. vertical also clearly and axis represents may

case that the vertical lines may just be causedby neat presentation (as the horizontal ones contained within the grid certainly are). Both the frame and the plumb line ideas, and to idea the of equal right ascension,spring from a desire to recreatethesevertical extent some lines physically for use during observations. This physical interpretation need not be the case. The text supplied with each table provides the information which distinguishes one hour from the next (the vertical separationbetween stars in the diagram) and also gives the position of each star using one of the sevenphrases. The picture shows thesetwo pieces of information using a grid. This grid simply presentsthis information graphically. The two date information is the that to the and the names of which are needed use clock, other pieces in lines in is It the that the text. therefore entirely possible vertical of the stars, only occur lines, intended that be horizontal for of the the the grid should exactly same purpose as line. indication than clarification, and rather as an of a physical separation This possibility leaves us with two graphical elementsthat have meaning, the stars and the figure. The meaning of the star symbol is self evident and unquestioned,but the meaning of the figure has beenneglectedin the previous literature.

The Figure
A figure appearsunder each table but does not vary significantly from table to table. The figure is dressed as a priest, with kilt and beard, and is shown kneeling in the typical Egyptian artistic pose showing his legs in profile, his torso in three-quarterprofile, and his

74

shoulders frontally.

His head is shown full face. This is a departure from the conventions of

Egyptian art. The use of full face figures is very unusual. If the figure was meant only for decoration or space filling, he would certainly be shown with his face in profile. That the

figure is not included to fill space is clear from his inclusion within the grid itself in source

C4 where starsare omitted or displacedaround him indicating that the figure was considered to be more important than the grid, and by the way that the other three sourcesplace the
figures in an otherwise empty strip below the text and grids. The figure is clearly an

important elementin the star clock, and its significance must be evaluated.
The kneeling posture is unusual, and must be significant as standing figures are more is There default itself important is the element. pose where stance not an common and are

figure kneeling figures for in C1, C3, in C2, the standing so a strip sources and also room for selected spatial convenience. was not The priestly dress is indicative that this timekeeping method was intended for use by the Such an temple to the the precincts. clock use of priesthood and so presumably confines to the materials and space environment was capableof offering undisturbed observation and for It construct and maintain whatever equipment was necessary. also offered a repository information and a source of men who were literate and disciplined enough to carry out is the This to time. probably written proceduresand record observations over a period of in this in developed which this the environment star clock was and also environment which been have used. would star clock dress His dress. his figure face, his his the the To sum up, and main attributes of are posture, is explained by the temple environment in which the clock would have been used. His
face his full find we significant and posture working of the clock. important to the be to considered enough

75

The Hour Stars

The set of stars whose names occur in the Ramesside star clock tables is usually given the name `Hour Stars' to distinguish them from the decanal stars. There are forty-seven distinct hour stars. Neugebauer and Parker have divided the stars into thirteen groups labelled E to R (omitting I). They have achieved this by looking at the order of the stars as they travel from hour to hour throughout the tables, and also by considering their names. The complete list is presented in Table 35. Five groups have just the one member:

F rryt, H sb3n h3w,J sb3n srr, N sb3nw mw, and P sb3 r. 3w.
Group E consists of sixteen stars E1-E16 whose names indicate that they belong to the figure He `giant' strong man nht. or wears two feathers, carries a hd sceptre, and stands on of a The majority of the star names are anatomical. pedestal. of sort some

Group G has four memberswhich are parts of a bird, the two stars of group K are from or L has has Sirius. Group 0 two Orion, to a pair of stars similarly group stars related also near from a lion. (mooring list. has Q to the mnit Two groups complete six members with names relating female hippopotamus. has R named members after parts of eight a post) and Sah Spdt, known from Haw, and Of the thirteen constellations, six are other sources. the M31, decan lists. Mnlt, Rrt in with connected and are names which are regularly appear hour These stars. together twenty-one six constellations contain constellations. circumpolar be Spdt Sah for than constellation the and one that would is used names more It unlikely Orion Kl, in K2, have Ll, that L2 around accepted or occur writers and previous and each, in in list lzJw hour in the Since relation Major. occurs place of n similar sb3 stars a Canis and hour has the been to being lists, identified decan this star also related in as to Sah as the decanb3w.

76

El E2

tpy-r! `wty nt nht Swty nt nht

KI

K2
Li L2 mi
M2

E3
E4 E5

tpy-c hd=f
tp nht tp (n) hd nt nht

tpy-l' (sb3n) sah sb3n sah

sb3 n spdt ly hr-s3 (sb3 n) spdt tpy-r' sb3wy
sb3w

E6 E7 Es
E9 EIO Ell

hd nt nht nhbt nht h36f

mndt=f bgs nht sdh f

N 01
02

sb3w nw mw tpnm31
sd n m31

E12
E13 E14 E16 F

pd nht

P Ql Q2

sb3w "i3w (tpy-r) mntt t3 nfr

sbkf
Pt =f sb3 n s3 pt rt

Q3
Q4

Smsw(n) hat (n) mnit

mnlt

E15 ty s3Pt =f brut nt 3pd

tp n 3pd
k t=

Q5 Q6
RI R2

. msw n mnit 3`mswly hr-s3 mutt

rdwy (n) rrt pd n rrt

Gl
G2
G3 G4

R3
R4

hry-lb n mnty=s

htyt nt 3pd

R5
R6 R7

sb3 n haw

b3h n rrt bpd n rrt mndt n rrt

ns=s

sb3n srr

R8

3`w n rrt

Table 35: The Hour Stars from the Ramesside star clock tables, grouped into constellations

The constellations M3f, MnIt, and Rrt present a problem. The methods of using the star clock tables describedabove, that is the frame or the plumb line, both rely on the hour stars being situated in the southernhalf of the visible sky, that is having declination 8<, where 4 is the latitude of the observer. The region of Canis Major and Orion meetsthis criterion, but the circumpolar region clearly doesnot. To remove the problem, it has been necessary to assume that three other constellations fact the The that M31, Mnlt, in Rrt to the three and ones. addition circumpolar existed named but hippo is labelled helps in Rrt this assumption, not as astronomical ceilings circumpolar That does the hippo Day in identification Rrt the Book not. the the the of circumpolar as of 35) the in (Table hour list the stretches consecutively appear ordered star constellations bounds of coincidence further. The constellation name MnIt means mooring post, a name implications be in thought has the of when can used clear circumpolar which region, which has this North think Celestial the No to Pole. `moored' we of region what as southern as in known text, in have other words we any no evidence for a special southernpolar context

77

region which would have a southern mooring post.

The stars are divided into only two

groups: the Imperishable Stars (northern circumpolar region) and the Unwearying Stars (the Stars is between There Unwearying the and obvious set). no connection stars which rise and have This that the three no we means of circumpolar-type names, another mooring post. evidence for similarly named southern counterparts and some evidence against such a

theory.

A Possible Explanation for the Tables

We have now identified several assumptions associatedwith the accepted usage of the
Ramesside star clock tables: 1) The central vertical line represents the meridian

2) The six other lines representeither a frame with vertical strings or the plumb line of a
mrht

3) Two people aligned North-South were neededto use the clock Some problems and unexplainedareashave also beennoted or implied:
1) The figure is kneeling 2) The figure is full face 3) Only one figure is shown68

in the decanal both starsused 4) There are constellationsand circumpolar constellations 5) The physical interpretation of the vertical lines is open to question the does implies evidence that the with agree This summary acceptedusage not completely
68Bruins `Egyptian Astronomy' outlines a theory for a single-observer using his own fist at arm's length to drawn the that from grid distance interpretation His requires away an established meridian. of units measure him. front in himself) figure be of (who figure is thought of as what the sees behind the seated now the observer left' `observer's from and his for My theory the orientation of He gathers support and wnmy which now suit in disliking Clocks, Bruins' theory (Calendars, Astronomy Clagett 146) and rejects `observer's right'. p. Egyptian highly 'a knowledge astronomers the more sophisticated of the of part assumption on ... particular if be this lack for to body likely', the (ear, used the the of reasons parts of eye, and shoulder) than seems between figure the the relationship and employed, and the grid. were method

78

presentedin the four sources. If we supposethat the constellationsused comprisedboth circumpolar and decanalstars, and that the portrayal of the figure is vital to the understanding of the Ramessidestar clock, a theory presents itself which allows the majority of the problems to be solved, whilst inherent in the the acceptedusage. certain of assumptions eliminating

The Reflection Conjecture

The observationsof the stars for the construction and use of the Ramessidestar clock were into by looking still water. made This conjecture can be comparedwith the list of assumptionsand problems associatedwith the previous interpretations of the clock. The first assumption,that the central vertical line representsthe meridian, may or may not still stand. Nothing in the reflection conjecture identification but the conjecture also the the of central vertical with meridian, precludes The kneel to the at the edge of a pool. possibility. another conjecture observer offers needs is the is distance his edge If the edge straight, the great position along of the stars meansthat immaterial, the important factor is the direction in which he is pointing. If he kneels so that his legs are perpendicular to the edge and looks straight into the water, the reflection of his define him let (which be in head and shoulders the starlit Egyptian night) will would visible his from little himself leaning his By forward backward, a nose. or or raising r rk3 ib using k3 lb R background. his he down haunches may move reflection up and relative to the starry be lb the ck3 in If defined East-West, be this the would the r way. edge of pool were can be the It to in was that the meridian the not accuracy may relation of alignment meridian. if the the meridian was not all, specific goal of this method. at good very The second assumption, that a frame or plumb line was used to measure the six other discarded. by be In defined the is the observer's lb ck3 can that same now way r positions,

79

reflection, it is clear that the parts of the body named in the six other positional phrases: shoulder, ear, and eye; now form their own referencesreflected in the water. No strings, laths, grids, frameworks, nor instrumentsare needed. The observerhimself provides the data points. It can be easily understood that using this method maps the sky, which we describe using spherical co-ordinates, into the positional system described by the star clock tables. The is needed a pool of water with either a straight edge or one particular only equipment position in which to kneel. The third assumption was that two people were needed to use the tables. This has been
by the reflection conjecture. removed completely The lack of a depiction of a second

C2, C1, in in below the text and perhaps using a mrht, person, space sources eachportion of C3, cannot be regardedas proof that a secondperson did not exist, but equally well, the lack his for that in his inclusion theory means a secondperson except of usage of any evidence a existenceis certainly questionable. Moving on to the problems presentedby the previously acceptedtheories, the posture of the
figure is explained very comfortably by the reflection conjecture. It has never been totally

does it frame Indeed, kneel. the not matter the target to or mr/jt usageneeded man clear why for either theory what stance the two observers assume, as long as they are consistent throughout their observations. Kneeling at the edge of water, however, is the most natural way to examine reflections closely. The reason why the figure is depicted with his features face on is also now very credibly face. is his This the the looking into view own observerseeswhen water: explained.
for decanal both is The region and circumpolar stars now possible. The use of circumpolar Egypt extends from the horizon to about 60 of altitude. Sirius reached a maximum altitude

80

of around 44, while the main figure of Orion extended from 43 to 63 at culmination. If we wish to assumethat the constellations indicated by the namesSpdt, Sah,M31,MnIt, and Rrt retain their usual identifications, the observerwould need to be able to view the portion horizon. from 30 horizon 40 the the to the sky around above northern southern of above This correspondsto declinations (epoch 1500 BC) 8> -18. Does the reflection conjecture be to the region entire observed? allow
Finally, the reflection conjecture allows the vertical lines, including the central one, to be form head The than the observer physical entities. abstract concepts rather and shoulders of do but for lines, be the thought not require positional phrases which can of as references

information be heavens. The the to the mapped onto complex procedures grid represents but is text, the within not a drawing of the situation. This was previously accepted contained for the horizontal lines in the grid, and can now be extended to the vertical lines. The
has the to the symbols of star relative observer still meaning. position

All the assumptionsand problems of the older theories have now been dealt with. The new itself in requires, addition to the pool, a re-evaluation of the phraseswnmy/lmnty conjecture and f 3by. for is This 13by left-hand (side) As noted previously, phrase used means either or eastern. looking line any lines at the three the to the central on of occur vertical right which stars the is line In interpretation, the 8). and (Figure the meridian, the conventional table centre Major). Canis in Orion be (including the the area of and southern sky viewed are stars to This means that 13bycannot refer to the observer's left, or to `east'. The remaining is 13by or left. the wnmy to the target's This leaves that readings of refers either alternative
Imnty to mean `to the target's right' (Figure 9).

81

"My il3' or

1 ""

Aby

Figure 8: Configuration of Ramesside star clock tables

ZENIT}{

EAST

wnmy Im1y or = target figure's right

9 Eby = target figure's left WEST

SOUTHHORIZON Figure 9: Configuration for conventional interpretation

NORTH HORIZON ZENITH

WEST

qty or = reflection's right or west

""

13by = reflection's left or east

EAST

SOUTHHORIZON

Figure 10: Configuration for reflection conjecture

The reflection conjecture suggestsanother configuration. If we imagine the observerto sit to the south of the water, either roughly or exactly, we see that the stars near the southern horizon would appearin the water towards the edge, the zenith further out, and the northern 13by is horizon further away still. East to the observer's right. We seethat could either refer Nay indicate wnniy to left leaving `to the east' or to the reflection's or, more probably, left `to (Figure 10). the the west' or reflection's either Neugebauerand Parker do not discuss their consistent use of the term wnmy rather than fmnty. In many cases,the word is written using the single sign P which can indicate both tmnty, the more common usage being tmnty. The hand determinative -- D41 wnmy and

82

which is often used after the fmnty or UUby sign in the star clock tables does not preclude the : 69 Two readings exist for the words containing the sign east/westmeaning of the word western or right-hand. Neugebauerand Parker chose the combination of writing wnmy and because their interpretation of the tables could not admit the meaning right-hand, meaning `western' which is heavily associated with the term lmnty. The role of the target man in the conventional interpretation of the Ramessidestar clock is wholly passive, his right and left ears, eyes, and shoulders, were used to describe the in him. The of stars appearing some used positions way above question of why a man was due to the lack of further supporting evidence. this way has never beenproperly addressed, From external sources,one other piece of evidence can be offered to support the reflection conjecture. The association between stars and water is very strong in Egyptian literature. 7 This The stars were referred to as fish, as birds flying up from the water, and as tears implies a link between the two71which had been noted as early as the time of the Pyramid Texts and makes the idea that stars would be observed in water not as unlikely as would
initially seem.

It has already been noted that the Ramessidestar clock was probably developed and used
only within temple precincts. This means that the sacred lake is an obvious and suitable

location, providing a referenceposition for the observerand a still surface72. observing The reflection conjecture,as a new theory for the construction and use of the Ramessidestar fit interpretation to is The the appears conjecture of star clock tables. clock, an alternative
69Faulkner Dictionary p. 8. Also p. 21 and p. 62 give some variants of imnty and wnmy writings. 7 For example in the dramatic dext which will be dealt with in Section F. 71For example Coffin Texts Spell 62: `you shall navigate on the Winding Waterway and sail in the eight-boat. These two crews of the Imperishable Stars and the Unwearying Stars shall navigate you, they shall pilot and tow you over the District of the Waters with ropes of iron. ' Spell 68: 'the Great Mooring-post speaks to you, a for you from the sea' (Faulkner Coffin Texts). is up set stairway 72From the use of the tables in the royal tombs at Thebes, Karnak itself may have been the place of origin of the tables. The sacred lake preserved there would provide the ideal observing location.

83

information the well with provided by the four sources. Unless further evidence is found, the reflection conjecture cannot be proved or disproved, but is presentedas an alternative
and more sympathetic theory.

Timekeeping Properties of the Ramesside Star Clock

Previous studies have devoted some effort to trying to evaluate the timekeeping properties of

the clock. The assumptionsmade have been that the centre line was the meridian and that a
method using two people was employed, constraining the hour stars to those which

culminated to the south of the zenith. The most recent effort has been made by Leitz. He first conjecturesthat the hours told by 3 the clock were seasonaland usessunrise and sunsettimes to evaluate them. He then finds
pairs of occurrences of a star at either right ear and left ear, right eye and left eye, or right

4 The left time and There in tables. the shoulder. and shoulder are sixty-three such pairs date of an exact culmination can then be found by finding the time and date midway between the two occurrences. If this culmination occurs on the first or fifteenth day of the month (which happenstwenty-four times), this event is sought in the appropriate Ramessidestar to In is found the close table. twenty-four time reasonably the each of a at cases star clock 75 labelled 13by. that predicted and correctly either wnmy or 76 1b. k3 is in A similar process then applied to all 231 occurrencesof stars the position r
This produces seventy-nine culminations on the first or fifteenth day of a month. Again,

theseresults are comparedwith the star clock tables.

If his calculatedtime of transit occurs

be 13by, the by is to transit the time or the table said the of given star clock star and after

73LeitzAltaegyptische 74LeitzAltaegyptische 75LeitzAltaegyptische 76Leitz Altaegyptische

Sternuhren Sternuhren Sternuhren Sternuhren

tables pp. 137 and 140. table pp. 143-145. table pp. 145-146. table pp. 147-153.

77Leitz AltaegyptischeSternuhren table pp. 153-156. Occasionally Leitz's data is faulty.

84

calculated time of transit occurs before the time of transit given by the star clock table and
the star is said to be wnmy, Leitz marks the occurrence `yes'. Also, the occurrence is marked `yes' if the star clock table shows the star as being r rk3 lb. If none of these three conditions

is met, the occurrenceis marked `no'.

Nineteen are found definitely not agreeing with prediction (one falls on IIII Shemu 1, which

is the muddled T23) and two more are marked as questionable.

Certain pairs which have a midpoint on a 1st or 15th day are missing from Leitz's table on

it in One (pair 153-156. 140) the tables, seemsthat and occurrence star clock appears pp. Leitz omitted this by mistake. Pair 183 which deals with star M1 produces a midpoint very by M2. is held hour IIII Akhet 6th 16. In to the tables, this the on position close star clock
Pairs 75,76,129,167,189, and 230 occur when the star in question does not appear on the

from differ for date times table the transit the widely appropriate and estimated star clock hour. calculated any Leitz then uses his data to evaluate the position of each hour star in terms of its right He then likely from the candidates. star a shortlist of possible ascension,and chooses most
his findings. tables the to to clock star assess returns

Assessment of Leitz's results

be to for in The twenty-four results stars which occur right/left pairs are shown 78 Every star appearsin the for 13by and wnmy table the conditions predicted and satisfactory. These final The the table three marks. question column of results shows only are met. k3 lb be the (pairs 25,26, 41) indicate to is at that r the and star supposed question marks later but later (25,35, 31 time, transit respectively). actually much minutes and appropriate to mark, for example, pair 18 as satisfactory when the star However, it seemsunreasonable
78Leitz Altaegyptische Sternuhren table p. 145-146.

85

that is just past the meridian, when the discrepancybetween actual transit should be Irt E3by, and the required hour is over thirty-seven minutes. In fact, if we allow a generouscondition
of t10 minutes for r k3lb, 20 minutes for In, and 30 minutes for msdr and Vz, we find

that only thirteen of the twenty-four results are acceptable. Of course,theselimits have been chosenarbitrarily, but are reasonableconsidering the time scale of on averagesixty minutes betweenone configuration and the next.
Time of transit Seasonal Time (hours) Hour (hours) Difference between predicted time hour transit seasonal of and (minutes)

Pair Star

Table

Position

4 6 7 8 12 13 16 17 18 25 26 28 30 33 37 38 41 42 43 45 47 50 53 55

E10 I Shemu 16 I Akhet 16 G2 G2 I Akhet 1 G4 Akhet 16 J III Akhet 1 J II Akhet 16 IIII Akhet 16 K2 K2 IIII Akhet 1 K2 I Akhet 16 122 II Akhet 1 Peret 16 M1 II Akhet 16 N Peret 1 01 I Peret 16 02 Peret 1 02 Peret 1 02 II Peret 16 P II Peret 1 P I Shemu 1 Q3 III Peret 1 Q3 IIII Peret 16 Q3 III Peret 16 Q3 III Peret 16 Q4 1 1111 Peret 1 Rl

01.26 2028 2123 22.82 23.83 00.59 22.02 2333 00.10 00.25 21.80 0130 22.83 20.77 23.76 01.93 23.86 01.07 20.58 0056 21.32 23.40 00.13 00.20

8 2 3 5 6 7 4 5 7 6 3 7 4 2 5 8 5 6 1 6 3 5 6 6

01.19 20.24 21.29 23.08 23.85 00.72 22.00 22.93 00.72 23.84 21.21 00.85 2230 20.73 2327 10.95 2335 00.23 20.14 0030 21.66 2337 00.25 00.18

tb r k3 lb r k3 trt tJby r 1k3tb r "k3 lb lrt tBby r rk3 tb Irt wnmy In Rby r rk3 lb r "k3ib In wnmy msdr wnmy tb r k3 nzsdrwnmy r 1k3lb lb r k3 msdr wnmy Irt wnmy trt wnmy Irt 13by lb r k3 k"li 13by Irt wnmy

-4.2 -2.4 3.6 15.6 1.2 7.8 -1.2 -24 37.2 -24.6 -35.4 -27 -31.8 -2.4 -29.4 1.2 -30.6 -50.4 -26.4 -15.6 20.4 -1.8 72 -1.2

Table 36: Leitz's seasonalhour predictions for the Ramesside star clock tables

Table 36 presentsa complete list of differences in minutes between the predicted time of transit (using the midpoint method) and the seasonalhours calculatedby Leitz. lb, of the seventy-ninequoted, only 42% are satisfactory Looking at the results for starsr k3 for k3 is hour time r minus actual transit time 10 minutes under the condition of seasonal fb, t20 minutes for Irt, and 30 minutes for msdr and krh. Of the results which must be

86

checked for being before or after the meridian, that is those stars who are given as 13by or

in wnmy the star clock tables, 38% occur on the wrong side of the meridian. Leitz's method relies on three assumptions: seasonalhours, r k3 tb = the meridian, and the observationsbeing made by two people using a frame or mrht. He presentsan interesting and valuable analysis of the tables, but his results are not conclusive enough to lend weight to any of the three assumptionshe uses.

Analysis Issues
The Ramesside star clock tables provide the largest corpus of observational records in the field of Egyptian astronomy. In summary, the clock contains 24 tables x 13 times = 312 data sets concerning the movements of 47 stars. However, the data is somewhat corrupted.

79 largest The area of corruption is T23 which, as Neugebauerand Parker noted, is badly muddled. Each data set consistsof date a a) hour) b) a time (beginning of the night, first hour, secondhour twelfth ...
c) a star name

d) a position (kilt wnmy, mscirwnmy, Irt wnmy, r k3 lb, Irt i3by, mscirf3by,or kch f3by). Parts a) and b) are sequential and present no problems for reconstructing lost data nor but four Part is in the where one or sources. c) often preserved only one source, reconciling C4 in Ti, have For the example, name preserved,conflicts sometimesoccur. more sources has the only complete text. The star of Ti "5 is pt =f and that of Ti "6 is rryt, whereasTi of C3, which has only one star name preserved,has Iryt in the fifth hour. Neither of the other two sourcesCl and C2 contains a readablestar name.

79 Egyptian Astronomical Texts2 p. 65.

87

I Akhet
-3 0 1 2 3 4 5 6 7 8 9 10 11 12 -2

16
-1 0123

T2

II Akhet 1
-3 0 1 2 3 4 5 6 7 8 9 10 11 12 -2 -1 0123

T3

II Akhet 16
-3 0 1 2 3 4 5 6 7 8 9 10 11 12 -2 -1 0123

T41

IIIAkhet 1
-3 -2 -1 0123

T5

III Akhet 16
-3 0 1 2 3 4 5 6 7 8 9 10 11 12 -2 -1 0123

T6

IIII Akhet 1
-3 0 1 2 3 4 5 6 7 8 9 10 11 12 -2 -1 0123

T7I

IIII Akhet 16
-3 -2 -1 0123

T8

0 2
3 4 5 8 7 8 9 10 11 12

I Peret 16
-3 0 1 2 3 4 5 6 7 8 9 10 11 12 -2 -1 0123

T10

II Peret 16
-3 0 1 2 3 4 5 6 7 8 9 10 11 12 .2 -1 0123

T12I

Figure 11: Ramesside star clock tables T1 to T12 Grey text indicates that the star appears in that position in some sources, red text indicates that the position given here differs from that given by Neugbauer and Parker

88

III Peret 1
-3 -2 -1 0123

T13

III Peret 16
-3 0 1 2 3 4 5 6 7 8 9 10 11 12 -2 -1 0123

T141

IIII Peret 1
-3 0 1 2 3 4 5 6 7 8 9 10 11 12 -2 -1 0123

T1 5I

IIII Peret 16
-3 0 1 2 3 4 5 6 7 8 9 10 11 12 -2 .10123

T16

0
2 3 4 5 8 7 8 9 10 11 12

Shemu 1
-3 -2 -1 0123

T17

Shemu 16
-3 -2 -1 012

T18
3I

II Shemu 1
-3 0 1 2 3 4 5 6 7 8 9 10 11 12 -2 -1 0123

T19

II Shemu 16
-3 0 1 2 3 4 5 6 7 8 9 10 11 12 -2 -1 0123

T20I

0
1 2 3 4 5 6 7 8 9 10 11 12

0
1

2
3 4 5 e 7 8 9 10 11 12

III Shemu 16
-3 -2 -1 0123

T22

IIII Shemu 16
-3 -2 -1 0123

T24I

0
1 2 3 4 5 6 7 8 9 10 11 12

0
1 2 3 4 5 6 7 8 9 10 11 12

Figure 12: Ramesside star clock tables T13 to T24 here in in that the the indicates that given appears position position that star some text Grey text indicates sources, red differs from that given by Neugbauer and Parker

89

In T2, preservedin C1, C3, and C4, all three sourceswrite sb3 n rryt instead of sb3 n srr in the ninth hour. Sb3n srr agreeswith the general order of stars, while sb3 n rryt has already beenused as 1'2.5. In the sametable in the tenth hour, in C1 and C3, the star name is given as tpy-r sah,rather than tpy-r sb3n sah,while C4 hasjust sah. From the examplesabove and by considering all four sourcesindividually, we can identify six distinct problems with the star name data.
1) The four sources are damaged, sometimes leaving only one source for a star name. 2) A star name could be written incorrectly in one source. In the case cited above, the star

rryt appears elsewhere in the table, making Neugebauer and Parker state that its in the fifth hour in C3 Tl is a mistake. appearance 3) A star name could be written incorrectly in all the preservedsources. Where the mistake is evident, such as in the case above where sb3 n rryt is clearly misplaced and a suitable, similarly spelled candidateis available, such mistakescan be rectified. 4) The similarity of certain star names, coupled with the omission of certain parts of the
lead dispute. to names can

two data sets into one line of text. 5) C4 sometimescompresses

in C3, T6 two is for in C3. In An T5 table the 6) of sometimes case omitted, entire example copies of the text occurred. These problems arise from three types of scribal error. Firstly, an error in the original source

from the in four is Secondly, copying surviving sources. reproduced all an error which Thirdly, by layout confusion caused source. and graphical considerations. original
Before assessing the impact of these problems on the data sets, the fourth piece of

information must be discussed: d) the position of the star.

Ideally, lines text, figure, consisted table of stars. Each originally of a and a grid containing

90

the position of the star in the text would agree with the position of the star on the grid. This is not always the case. C3 often places the star symbols between the vertical lines. Some tables have two stars shown for one hour. Comparing sources, it is rare that all four grids agree on the position of a star. In fact, of the 312 data sets, only six have complete and clear agreement between the grids of (readable four T4O0 traces in C2), T4.1, T4.2, T21.1 (readable traces in C4), sources: all T21.2, and T21.3. T4.4 also agrees if the placing of the star in C3 is taken to be Irt wnmy

rather than being betweenIrt wnmy and msdr wnmy. Also, T21.5 is consistentapart from the inclusion of an extra star symbol in C1. Looking at the texts for theseeight data sets,only one text is not preserved(C2 T4.4, texts in
Cl, C3, and C4 agree with their grids), six sets of texts agree with their grids (T410, T4.1,

T4.2, T21.1, T21.3, and T21.5) all placing the starsr rk? ib, and one set (T21.2) C1, C2, and
C3 agree with in wnmy but C4 has r rk3 lb. Of the twenty-four tables, C4 omits T7 through to T18 and T24, leaving only eleven tables. The texts of T6 and T15 each occur twice in C3, and the text of T12 is repeated in Cl.

Counting each hour of text as one piece of data, and each hour of grid as one piece of data, is total into detailed there the of a tables taking omission and addition of account above, is figure This data have been included in tables. 2262 pieces of the star clock which should losing the itself, in C4 thus by figure the the the practice of placing grid upon reduced These for information later hours, the by the stars of and various other omissions. positional in Damage to the data. 219 for 10% loss number, account of around a of pieces omissions, lines damage 25% data to has be lost of to causedanother of sources entirely, while partial text and to starson grids has occurred to 14% of the data, although the stars can be restored,
be derived data has data from 51% damaged the can The of texts. remaining and some

91

survived intact.
Agreement between text and grids is not complete. Comparing each pair of one line of grid and one line of text for each of the four sources (C4 omitting T7 to T18 and T24) gives 1079

pairs of which we havejust estimated65% are preservedwell enough to read the position of
the star. Of these pairs 72% agree, that is, the position stated in the text and the position of the star on the grid are exactly the same. A further 13% have either more than one star on the grid, with one in agreement with the text, or have the star in a space between two vertical lines but adjacent to the position agreeing with the text. 15% disagree completely.

Looking at the grids for each hour in each table in each source (that is, comparing the grid for T10 from each of the four sources,then TI -1 from all four sources,etc.) 112 out of 312 data (36%) have 112 in When the the these all sets of grid stars setsof grids sameposition.
trace from 83% the texts texts no contain with relevant of compared are all available sources,

have but damaged, disagreement includes This figure the texts with grids. which are of have from no the traces of words which agree with sourceswhich grid position, and texts 3% further the but A information from of the preserved, agree with grids other sources. grid
112 sets had no accompanying text traces to show agreement or disagreement. The

did data (14% that data) not textual cases sixteen of sets of with grid paired were remaining detailed below. These cases sixteen are agree. 1) T12.1. Cl and C3 have text and grid. Both grids and Cl text have kch wnmy. C3 text has knhi3by. 2) T15.2. Cl and C3 have text and grid. The C3 text is duplicated. Both grids and C3 text
have Irt wnmy. Cl text has Irt t3by. 3) T15.4. C1 and C3 have text and grid. The C3 text is duplicated. Both grids and C3 text have Irt t3by. Cl text has in wnmy.

92

4) T15.6. Cl and C3 have text and grid. Both grids and Cl text have Irt wnmy. C3 text has kc zwnmy. 5) T16.12. Cl and C3 have text and grid. Both grids and C3 text have r ek3lb. Cl text has i3by. 6) T17.1. Cl and C3 have text and grid. Both grids have Irt l3by. Both texts have Irt
wnmy. 7) T18.2. Cl and C3 have text and grid. text have in wnmy. C3 text has Irt f 3by. C2 has a readable grid. All three grids and C1

8) T192. Cl and C4 have text and grid. C3 has readabletext. Both grids and texts from Cl and C3 have in wnmy. C4 text has in i3by. 9) T19.5. Cl and C4 have text and grid. C3 has readabletext. Both grids and texts from Cl and C3 have Irt wnmy. C4 text has Irt f3by. 10)T19-7. Cl has text and grid. C2 has a grid and some readabletext. C3 and C4 have text. Both grids have krh i3by. All texts have msdr IJby, but there are tracesof a possible correction in C2.
11) T19.8. Cl has text and grid. C2 has a grid. C3 and C4 have text. Both grids and text

C4 have msdr wnmy. Texts Cl and C3 have msdr i3by. 12) T21 0. C1 and C3 have text and grid. C2 has readablegrid, C4 has some readabletext. All three grids have Irt 13by.Text C4 has tracesof 13by.Texts Cl and C3 have Irt wnmy. 13) T21.2. All four grids and texts are preserved. All grids, and texts from Cl, C2, and C3,
have Irt wnmy. Text from C4 has r k3tb.

14)1'21.6. All four grids are readable,as well as texts from C1, C2, and C4. All grids, and texts from C2 and C4 have in i3by. Text from Cl has Irt wnmy. 15)121.10. C1 and C2 have text and grid readable. C4 has a readablegrid. All three grids

93

and text from C2 have frt t3by. Cl text has Irt wnmy. 16) T23 0. C3 has text and grid, Cl has a grid. Both grids have msdr 13by. C3 text has
wnmy. In constructing their complete star clock tables, Neugebauer and Parker usually take the decision as being the one to follow. majority 3,4,7,8,9,11,12,14,15, Of the sixteen discrepancies listed above, 1,2,

and 16 all have the majority of texts and grids pointing at wnmy,

disagreeing, two or stating the star is 13by,or vice versa. with one In case 5 the majority of texts and grids have r k3lb, with only one text stating the star is

13by.Similarly, in case13, the majority of texts and grids have Irt wnmy, with only one text stating the star is r ck3lb.
In these fourteen cases where a clear majority exists, Neugebauer and Parker reject that

is for 1,11, doing for 15. three times: Each time, stated. the cases so majority and reason
For case 1 they say `we accept this [kch i3by] in preference to the other extreme position [kch wnmy]

becausethis is the only table which has both Ll and 1.2 in successivehours and not as the virtual alternates they have appearedto be in Tables 2 to 11. They must be stars with transits rather close together and with Ll
have it it is have for to in fb] hour [r ck3 to 0 any L2 delayed [kch 13by] the the to essential order end of +3 of at

We later. length. ' to this shall return point appreciable In case 11, Neugebauer and Parker choose to accept the texts of C1 and C3. They also say

that the text of C2 bearsout this choice, but the transcript they provide for this text shows the information The is [msdr iThy] for decision that This this as unreadable. wasthe reason position
location of El in Table 18 and a shift to the left of four positions for no apparentreasonwould be difficult to accept.'

Finally, in case 15, the reason for ignoring the majority of texts and grids is given as `it is
location the Ib' k3 to as hour accept wnmy correct the at since was and preferable star skips an ...

94

All these three reasons stem ultimately from the model that was chosen by Neugebauer and Parker to explain the construction and use of the Ramesside star clock.

The other two of the sixteen discrepancies,6 and 10, are not so easy to resolve. For case6 for in follow Parker fit Neugebauer text, the to their case again with model, and similarly and 10 where all the texts have msdr while both the surviving grids have kph, Neugebauerand
Parker again follow the text rather than the grids, although no reasoning is given. These sixteen cases are the major points of difficulty in reconstructing a complete set of

tables, from the point of view of reconciling four sources and the information of texts and

is information less has is but Where to there accept what survived, no choice grids. being just lost the or omitted portions of once as correct of and representative preserved
other sources. Neugebauer and Parker's Notes for each entry in each table show that

difficulties data from is the sources a processnot easily completed, with constant extracting
in reading and reconciling the information. Neugebauer and Parker have solved some

in developing by mind. this then the tables model a model, and reading with problems When approaching the tables with a new model, as Leitz did, it is necessaryto re-read the
tables in order to make sure that Neugebauer and Parker's assumptions, based on their own

influence do the not new work. model,

It is also necessary to admit that although

highlighted discrepancies have Parker the tables, Neugebauerand produced a complete set of data thus degree the gathered indicate that the that of certainty say with which one can above
is high. the source original not represents It is difficult if but the problem, to quantify

(estimating from the agreement of information

been have the sources where several of

the data is dubious, in 14% 44 of data 312 reconstruction any of about out of sets preserved) inherent been Add have be to this the erroneous. all tables could problems that may already

in that original source, and the prospects for mathematical analysis of the data contained

95

within the tables are not good.

To compound these problems with the source material, there is also the problem that analysing these tables by constructing a model and then verifying experimental results

against the reconstructed star clock tables involves making several key assumptions. Ideally, a study should solve for one unknown quantity. In the case of the Ramesside star clock,

three unknowns are involved: the intended method of using the tables, the length of time identify For Leitz the to stars. stars and measured, used. example, periods wished ultimately To do this he needed to know method of usage and the time periods marked. He had to both, for so as he acknowledges,his final results cannot be acceptedwith make assumptions
importantly, his conjectures about hour length and method of usage are and more certainty,

lent additional support. not

Is there any way in which just one of the three unknowns can be isolated and studied?

Impact on the Reflection Conjecture

Having stated that there are three key unknowns to deal with in the analysis of the
Ramesside star clock tables, the reflection conjecture aims to satisfy one of them, the

is One undisputed. the the clocks of star certainty about use of problem of method usage. behave the When is for the That the motion of the effects of precession, stars stars. adjusted in the sameway today as they did whenever the Ramessidestar clock was devised. One important difference between the observational scenarioof the reflection conjecture and
The is linked two stars move the to the stars. person method, the conventional, motion of (Figure interpretation during the from east to west course of the night. For the conventional

left) to from the (the that the figure's observer's 9) this means starsmove wnmy right, seated
but from to For the the west, east 13by. reflection conjecture stars, of course, still move because of the different orientation of the observer (Figure 10), this translates into a 13byto

96

wnmy movement,exactly oppositeto that which occurs in the conventional interpretation. Throughout their work in reconstructing the Ramessidestar clock tables, Neugebauerand Parker assumethat the stars are moving wnmy to i3by, which fits in with their model of the fits drawn follow information Frequently, to they the with choose or which written situation. this motion if there is a conflict between sources. For example, we saw earlier how in interpretation, information the towards majority of pointed one certain cases where
Neugebauer and Parker instead chose to follow a single occurrence of the opposite location

becauseotherwise the motion of the star in question would not fit with their model. In particular, they noted (as quoted in case 1, above) that star L2, sb3n spdt, should be at an its from distance usual substitute M. extreme Due to their interpretation of the tables, this

information imply 13by hence that three of position was required, a consistentpieces would in favour discarded, kl'h that the of one piece stating of which stated all star was wnmy, were be kph to kch 13by. In this wnmy that the star was case,the reflection conjecture would allow because the is interest The is of this reason one occurrence particularly of accepted.
between Li L2. the two stars and relationship in One or other of these two stars appears

72 Ti!, both in T12. to tables the and appear of each T2 K2 Ll T3 K2 Ll

Ml

T4
T5

K2
K2

L2
L2

Ml
Ml

T6 17 T8 T9
T10

K2 K2 K2 K2
K2

L2 L2 Li L2
Ll

Ml MI M2 M2
Ml

Tll
T12

K2
Ll

Ll
L2

Ml
M2

Table 37: Occurrences of hour stars Ll and L2

in K2, by In their is 37 except Table summarises appearances. each table the L star preceded Apart be in Tu it K2 had beginning the used not T12 where would as night marked the of . 97

from T2, where Ll marks the twelfth hour, the L star is followed by either M1 or M2. Neugebauerand Parker's reasoningthat Ll and L2 transit close together is sound. No other
in is the tables pair of stars as strongly linked as Ll and L2, although there are other

candidatesfor alternates,such as E3, E4, E5, and R6 (all occurring not more than twice), Q1 (occurs only once) and Q2, R3 and R4, and Q5 and Q6. Ll is r k3 fb, although only one grid survives to indicate this. In T12.1 L2 is k11h In T12O0, in four three of out surviving grids and texts. If the two stars transit close together, wnmy thisindicates. t were movingby _thauhe-stars nmy.

wnmy or Lnnly
EAST

""

Aby
WEST

wnmy or imnly
WEST

""

!iby
EAST

to a passive Starmotionrelative targetfigure

to observer's Starmotionrelative reflection

Figure 13: Motion of the stars in relation to the observer for the two theories for the Ramessidestar clock

T19.8 and T21.10 have already been mentioned as cases where Neugebauer and Parker ignored the majority of available information in order to fit the motion of the star in question into their model. Other instancesof stars not behaving in accordancewith Neugebauerand Parker's model occur: T9.9, T9.10 possibly, T11.2, T11.6, T11.10, T14.5, T14.7, T18.7, T18.9, T18.12, and T210 possibly.

Creation of a System of Equations for Hour Lengths

The sidereal day is 3 minutes 55.91 secondsshorter than the solar day. This means that a days fifteen in day 58.977 in the to the the sameposition sky after minutes earlier star returns have elapsed. We shall denote this time period of 58.977 modem minutes by the symbol i. This time period is the samefor any chosen star, at any time of year, for any position in the

98

sky. In other words, no assumptionsare being made concerning use of the meridian or choice of star.
If we look at a star that occurs at a single position (for instance Irt wnmy) on a number of deduce the time that passes between these events, using our own time can occasions, we system as datum. For example, star F is at Irt 13by in five consecutive tables: T2.5, T3.4, T4.3, T5.2, T6.1. This implies a fifty-nine minute difference between the `fifth hour' on I

Akhet 16 and the `fourth hour' on II Akhet 1 and so on. E14 is r k3 lb six times: n-4,13-3, T4.2, T5.1, T60, and T21.12. Fifteen 15-day periods occur between T6 and T21. This meansthat there is a difference of 15z or 14 hours and 45 minutes betweenthe `beginning of the night' in III Akhet 16 and the `twelfth hour' in III Shemu 1. Many relationshipsbetween hours in different tables can be made this way, building into an incomplete systemof equationsin i, with one group of equationsfor eachstar. Let us look at It 01 tables. 01 in Ramesside times the as an example. star clock appears eleven star Irt T9.7 T5.12 k3 lb, T6.11 Irt trh rk3 T8.8 lb, T7.10 f3by, at wnmy, at r at at r at occurs at T14.1 lb, ck3 rk? T11.4 lb, T10.6 T13.2 kl'h T12.3 at r at r at at msdr wnmy, at wnmy, wnmy, in l3by, and T15-0 at r rk3 lb. From T6.11 to T5.12 is tt. From T11.4 to T6-11,5 x 15 days
have passed, so 5i minutes separate the two occurrences. 01 is at r k3lb twice more: at

T13.2 and T15 O.

We have the following relationships in modem minutes:

T6.11 = T5.12 -i T11.4 = T6.11- 5i T13.2=T11.4-2i T15-0=T13-2-2i four T5.12 fix (for the knew other If we when exactly example) occurred, we could then times. We can also look at the relationship between hours when the star occurs at different

99

positions.

At T7.10,01

is at kIb 13by. The reflection conjecture implies that stars are

moving 13byto wnmy. Therefore the star has not yet reachedr k3lb, so the time between
T7.10 and T6-11 (when it is r k3lb) is somewhat more than t. We cannot quantify the time

do because we not have any certain understandingof what kch13by means. more accurately
At T8.8 01 appears at Irt wnmy. This in turn implies that the time between T8.8 and T7.10 is quite a lot less than i, and that the time between T8.8 and T6.11 is `a bit' less than tc. If

indicate indefinite from 8 by to time to taken one an and variable we use period a star go fb or from msdr wnmy to k"h wnmy, we to r k3 position to the next, for example from Irt 13by
can write down relationships in modern minutes between all the appearances of 01. The

two we havejust discussedare: T7.10=T6.11-i-3680 T8.8=T7.10-tc +46 The sign of 8 is determined by the motion of the stars which dependson the observational
in 8 (13by the for to the positive chosen wnmy make reflection conjecture model would model

interpretation in 13by the would to the equations, while wnmy standard motion preceding
8 that was negative). mean We can repeat this process for each star which appears more than once in the star clock tables, forming a system of equations. Neugebauer and Parker have noted that T23 is

be A correction must also made muddled, so we shall eliminate any referencesto this table. 1. by denoted is for 51/a days in the The the year. to compensate extra relevant time quantity Of the 47 hour stars, 7 (E3, E6, E15, E16, Gi, G3, and Q1) occur only once in the tables, (E6 therefore G3) in the Our T23. two seven and of equations occurring of only system with

is a move of three positions in the 13bydirection, hence 36. Similarly from knh13by 80From r k3 tb to krh 13by four in is the wnmy direction, hencethe term + 45 in the following equation. of positions trt a move to wnmy

100

' 40 254 consistsof equationsand unknowns! The information that can be gained from these equations is more limited than would at first is Although fixed, we can only speculateabout S. 6 is either positive the quantity ti appear. or negative dependingon which observation model (conventional or reflection) is used and is a variable quantity because,depending on the declination of the star, a move from one position to the next may entail a greater time difference for one star than a motion through several positions for a different star. However, 8 can never be very small becausethe difference in position of any star in question must be appreciable to the naked eye. Furthermore,the choice of values for the unknowns is critical. Two approacheswere attempted in order to find some information from the equations. In the first approach,the assumptionwas made that the observationsstarted at some time after sunset,such that the time of the `beginning of the night' was directly related to the time of in the The be hour to the appeared unknowns were star chosen sunset. earliest at which each tables. The stars of T7.0 and T17.0 were two of the seven stars which occurred only once, but the remaining twenty-one `beginning of the night' hours accountedfor just over half the (labelled `a' `u'). to unknowns Resultswere analysedby setting the unknowns `a' to `u' to a set of values which represented
Parker's first fixed Neugebauer Following time and observation at a after sunset. a time of initially analysis, T9 marked the longest night, T21 the shortest, and T3 and T15 the

longest but date the the night. the equations also of allowed alteration of civil equinoxes, The system of equations is shown in Table 38. Times not calculable directly from the
in are shown unknowns grey. twenty-one
40 81There are (13 hours x 23 tables) -5 stars which occur only 294 which of quantities, possible once, giving leaving 254 equations dependent `unknowns' b is by filled 40 our an additional unknown the are on unknowns. and variable time quantity.

101

1O M '

A0 O

10 fV O

10 N C

_ r'1

O 'l

{ O

C-

d>

IAO O p

'r
rG Y

"
C'

rl
dJ

I t

r rQ O 00 + 00 } In 00 + } E m G Q` d 'J f' + f q l" OC r 1, 0' " L

rq ' r1 -. Y rI r t r

N Ic

rq Y

_q f, Y

t 1, r

VI C

a,

ti

4-

{ -o

oll

O
67
i

r0 ,q Y pp 0' p F [r ra Y 1I [10 rY "0 .p ' D " 1'

or

6J

-. -

'0 .0 N rq
r +

67

rq
r

rq

"0

on

e
r4 f

r'
.O a

D L

AQ
ti-

r+ r

110 m t+

0 r. { r 7
a

V 7 -f P
c}a

v,

, rn

Hfl OU

1 I }

Y 1

i E

1 0

rt d

7 i.

`i v

O
, ^ d

J C

rn

r i
`.

a+

V1

`O
+
Y + Y + +

:7

M
/ m m +

o
110
N f. N N N N

00 M

m
N +

N
f

Y F' {I

`
rl I

rn {
i

r')

10

-+

on

a,

"'

,G

11

211

r N l ' 1 4 t , F s i c

1.

f9

t,

,p

oU

cr

F.

Fes-

Ems-

20 F FF FFFI-H IH H H FH H H

102

The only constraint on the results that can be assumedwithout doubt is that the hours of any
one night must follow consecutively. fulfilling to critical was unknowns It was found that the choice of the remaining 19 These unknowns had to be supplied by

this criterion.

few from hour lengths times, throughout the tables a ranged and comparison with adjacent hours. two to modem minutes up nearly No further information about the nature of 6 could be gathered from this attempt, but it
depending that time starting observations at a seemed unlikely on sunset would produce the Ramesside star clocks tables, as the times produced consistently failed to meet the

hours criterion. consecutive The secondattempt assumedthat the hours of a night were roughly equal. Two nights, T3 due the time T15, the to their of chosen and were supposed equinoctial positions, and beginning of the night in 73 was set to zero, providing a datum. The hour length for 73 and T15 was denoted by ?. and respectively. Since it cannot be assumedthat observation beginning the in T15 it did in night T3, time the the of the time same of as altering startedat
in T15 was also tried, the offset between the beginning of the night of T3 and the beginning

if be being labelled T15 Although T15 T3 equinoctial would w. and of the night of
Neugebauer and Parker's date for the construction of the table is correct, this was not

assumed. All hour points in T3 and T15 could therefore be filled. Only time points directly related to Grey 39. in Table T15 T3 cells The and were used. the hours of equations are shown
directly T15. that T3 from time are not points calculable and represent

different limits the to table When these aspectsof generate are randomised within realistic % for be the and A can results analysed of values graphically. range combinations, between 30 and 60 minutes, and w between -120 and 120 minutes fits with known night
103

lengths. From this method of analysis,two results becameapparent:

1) There is an inherent discontinuity in time of `beginning of the night' between T9 and T10. T10-0 always occurs at least 80 minutes and up to 130 minutes later than T9.0. This

trend explains why the earlier attempt, basedon observationsbeginning at some time related to sunset,fails to produce consecutivehours under most conditions. (SeeFigure 14) 2) We can still learn nothing about the nature of S. With other factors such as constraints does being hour length 8 be This the unresolved, sign of cannot yet analysis on ascertained. not, therefore, show that either direction of motion of the stars (that is, for the standard interpretation or for the reflection conjecture) is more likely than the other. Further analysis of the Ramessidestar clock tables might be capable of revealing more information concerning the nature of the night hours, and may resolve the question of how
in their The for place their construction and the tables were originally constructed. motive

Egyptian timekeeping history will be discussedin the next section.

104

A h1 r N + N

N r + rA

7 r

+ r 7 1 i O + 7 f + i 00 +

{ 1 f G 1 a I . I' 1 + vJ ` . " + 1" }

. "O 1 r I 1 1

+
r l + 7 +

O -

i N +

00
i V +

3
to

s
'O

i.

in

-r

m
'q

q 110 + {

q }
O

oo
Y x f-

a
00 p,

r... r

rl <'

+ a

+ s

4-

1-

8
10 r^1

3
10

'Q N ' rI o

1-

'-"

'O Y

i C-

+ +

'7

M r

r }

`o r P v

rI

G H

"q

4 1

+
'.0

00

c .,

-I

+ +

+ +

Y + + !

q + F, N

Y K + 7 + } C_ N ,p 'o rj

N r"

N '

C' i

+ N

ZL f 'q

=I00

a+

'O

'n

f!

'

co

:a

it

oc

f-

.0 4
-

3
eq

C7, -+
3

tq l F

1 +

,q ^

'0

'l

ri

r rl
10

.n
i O M -..

! . 1

<

+ 4

.n

o.

.1

f-

"

r ,

In

w
N

I{
r' + a C + a V r 7 .p 1 _7

q a. Q"

7 T

Y N

.-

.p

oc

3 w
'O

r + F

N '

r -T

7 +

+ +

i +

-1 + }

r +

N +

a }

i +

a I

rC I

3 w

3 'C i
I, I

'j ~

w -

.0 r

N + i

4 " n r1 . C

4 . _

,p Y i

N i

N a

Ic

r~l

r- i

M .1

" c '

.1

7 +

Y +

N }

+-

0 M

N i

q
N

M
r

110
r

In
P I"

"0

rA
N

1
'

i ',

Ic

.1

` r i

%-

-t

oo

' I"
t

vl C: M
ea F'

,c

Ic
q N . N Y rl ` r M N n .

"0
N .

10
n . " r

rq f1 r N r 4 l -14 i rf, r

4 C f . f

10
N + rrN r M

10
r Vl r

,p N

rGi Iq r ^l 'O 'n r r

,C r F' a

r N

N -t

"-

r 0
.

.1

IC

'

.1

.1

10

00

-{-

3
110 N

h
_ F N N M

r^ F

'O

r 1

ri N

.1 M

r!
J

T i+
J

1-

F-"

n F-

Ic"

O 1F F F

N f-

r H

^t .,

.n

oo

ON

c N

N rl

M N

Y N

F-

f^

105

1 700

! 00

600
i

5DO
400

++ +

3DO 200 1pp

137 100

11

13

15

17

19

21

23

X =50=50w=0S=-10
$00
700

X =50 t=50o

=05=

10

600
5w

+I

300 200

too 0
13579 "100 -200 11 13

_
15

--

17 + 19 + 21 + 23

?,=60=40 co=606=-10
800
700

X=60=40co=606=

10

700 800 5

-----

-+`

600

400 300 200

100 ++ +I 300

zoo

100
19 21 23 -100 -200 1379 13 15 17 19 21 23

379

11

13

15

17

100

=40=60W=-60S=-10

a, =40=60co=-606=10

Figure 14: Hours of the night generated by the equations Table 39 of

106

Section D: Development
Diagonal Star Clocks
Although

Star Clocks of

the manufacture of a star clock is not difficult,

the development of such a

procedure is not trivial.

The motivation for the star clock table has always been stated to be

to find the hour of the night. This statementimplies that first, a need to find the hour of the night had been identified, and second that the particular method of a star clock table was
chosen and developed to fulfil this need. No other method of finding the hour of the night which dates from or before the time of these star clock tables has survived 82 In particular, we can be fairly certain that no

seasonally adjusted water-clock existed, becausethe record of that instrument's discovery has survived, and dates from the New Kingdom. 3 The only instruments that could be used to calibrate a new timekeeping method would be those which measured either a certain 84 be length, period of time which could reproducedat will, such as burning a wick of certain or some sort of water-clock which measuredhours which were the samethroughout the year, but were not necessarilyof equal length throughout the night. No recognisabledeviceshave documentation and no survived, about the invention or use of a star clock or any other type of timekeeping instrument is known from the First Intermediate Period or before. It seemsthat the star clock was the first formal timekeeping instrument which could be used

82Wells ('Origin of the hour') proposes that Vth dynasty solar temples at Abusir display alignments which suggeststhat their causewayswere used as a referencefor a `star-clock'. He suggeststhat the requirement to dawn it by for the made imminent at to watching necessary sacrifices sun priests predict the arrival of perform the risings of bright stars. He proposesa system of twenty-four bright stars for each temple, which he lists, in hour's in dawn days for fifteen fifteen days to one notice the of As give of use order per star. we shall see, did from timekeeping to occur nightly at least 1850 BC, but no supporting evidencefor Wells' theory relation has been found, nor for his assertionthat the earliest hours measuredusing the stars were intended to be sixty minutes long. 83From the tomb of Amenemhet in Sheikh abd el-Qurna (time Amenhotep I). Borchardt Die Altaegyptische of ZeitmessungTable 18. 94Bierbrier Tomb-buildersof the Pharaohs p. 50.

107

for finding the hours of the night. It first appears, fully developed to the extent of displaying mistakes and omissions characteristic of an established object, on the coffin lids we have been examining. No contemporary explanations nor earlier prototypes have been found. It has been postulated that the motive for the tables was something other than timekeeping.
De Puydt argues85 that the tables `may have just been an iconographic and textual rendering of the yearly motion of the star sky'. He supports this suggestion by noting that firstly, the tables only

intended in to in he if find than temples tombs they to them rather were where expects occur
be a practical timekeeping method, secondly, he has found `no explicit evidence for nightly ritual
or activity in general for which the precise measurementof the hours was needed', and thirdly, that there

are `too many variables in the behaviour of the star sky that the known structure of the star clocks cannot

by be he As his first an for. explained acknowledges, can must and surely point account in kept in that the method was probably recorded on papyrus and accident of survival, working libraries, whereas the corrupted copies which remain today were painted on wood for the in His lack indicates tombs. motive that of obvious second comment a and placed the into be excludes taken them, which tables should account when analysing an approach Egyptian has discovered motive, the that an modem researcher or understood not possibility both field introduce into seems the which an additional assumption and therefore would and obstructive. unnecessary devices initio intended Puydt's as ab However, De reluctanceto assumethat the clocks were tables the has as have hour Previous the find the examined of night merit. to studies86 87 Discussion hardly development to isolated objects with any reference their and motivation

85De Puydt `Ancient Egyptian star clocks and their theory'. 86Neugebauer and Parker, Clagett, Locher, etc.

87Parker (Ancient Egyptian Astronomy pp. 53-54) attributes them to `somegenius in the early third millennium BC.

108

of thesepoints cannot, with the evidence now available, lead to definite conclusions about
such matters as positive identification of decans but can shed some light on the purpose of the tables and the criteria that governed the choice of decans. The appearance of Sirius as the decan of the twelfth hour occurs at, or during the period just heliacal the after, rise of Sirius. This event was certainly the most important stellar event in

the Egyptian year due to its association with the annual inundation88 Theoretically, the
heliacal rise of a certain star occurs on the same day of each 365V4-day year, until enough years have gone by for precession to have enough effect to change the date of the event. There are several possible reasons why the Egyptians did not alter the length of the year so

that Sirius would rise heliacally on the same day each year.

These reasons include

administrative, political, and religious arguments. There is also the important and often overlooked reason that the heliacal rise of even a bright star like Sirius is a near-horizon phenomenon. Poor observing conditions, location, and human factors could combine to date back fixed from the the the the of event time obscure push predicted of occurrenceand ' heliacal between rises! period Although Sirius gave warning of the flood's imminent arrival, the length of time betweenthe it leaving the the the star and rise of appearanceof waters varied considerably perhaps unclear which phenomenonof the two should take precedenceas a time marker. It is fairly certain that the behaviour of Sirius gives a model for all other decans. This is 90 In Abydos Egyptian from dramatic texts Osireion the from text such the at as clear decanal lasting invisibility that have that text states seventy stars particular, a period of

88Parker Calendars.

89Schaefer`Heliacal Rise Phenomena'. 90Frankfort The Cenotaphof Seti I.

109

days91. Sirius would theoretically have had this period of invisibility

at the latitude of Egypt

92 if 3500 BC, viewing was perfect and observations were carried out and averaged around over several years (see Table 4093). The observational factors previously mentioned would tend to lengthen the period of invisibility, altering the hypothetical date of first calculation

94 invisibility Sirius later, dependingupon the quality of to up to one thousandyears of of the data usedand the number of observationsover which the period was averaged. observational
Right Ascension Declination Period of Invisibility

Date AD 2000 AD 1500 AD 1000 AD 500 Year 0 500 BC 1000 BC 1500 BC 2000 BC 2500 BC 3000 BC 3500 BC 4000 BC 4500 BC

(hours) 6.7528 63853 6.0178 5.6503 5.2836 4.9181 45533 4.19 3.8283 3.4681 3.1089 2.4014 2.3894 2.0267

(degrees) -16.718 -16.134 -15.813 -15.76 -15.975 -16.454 -17.191 -18.177 -19.404 -20.859 -22529 -24.401 -26.457 -28.679

(days) 63 63 62 62 62 63 63 64 65 66 68 70 72 74

Table 40: Position and period of invisibility of Sirius from 4500 BC to the present

There remainsthe question of how the star clock developedfrom the prototype behaviour of

94This period (3500 BC to 2500 BC) ties in with the most likely of three possible dates for the establishment of Clocks, (Calendars, Clagett Sothic by based 2781-2778 BC, given on an analysis of periods: the civil calendar 31). Neugebauer's rejection of an astronomical basis for the foundation of the civil year Astronomy p. and introduction date the Egyptian Calendar' Clagett Origin to the 'The 396-397) leads of (Neugebauer pp. which disappearance lies 3000 BC the of during to still which c. within our the estimate time the of period calendar of The BC. does 2937-2821 Parker's days', '70 (based lasted lunar as estimate Sirius year) of c. on an average be it is because introduction the only would civil calendar of date of the significant to the present argument become the that introduction days in civil calendar natural. of ten counting the would of packets after

9' The earliestoccurrenceof the text is in the New Kingdom, far later (by at least one thousandyears) than the be days in interest may Egyptian the motions of stars, therefore the use of the exact period of seventy origin of funerary and difficulty dating period The the the later the text that seventy-day of origin of means concept. a the that it is however ideal invisibility decans be certain the still of the seventy-day cannot consideredabsolute, invisibility of Sirius was significant. 92Using the heliacal rise and set prediction algorithm presentedby Schaefer `Heliacal Rise Phenomena'and Settings'. Risings into in Heliacal listed Schaefer and 'Predicting hence converted a computer program Angles `Extinction and ('Refraction by Atmosphere' and Schaefer's work on extinction and refraction earth's to the is has implications for horizon pertinent though more Megaliths') also near naked eye observations, study of architectural alignments. 93Similar calculationswere carried out by Ingham and presentedin `The length of the Sothic cycle'.

110

Sirius.
Seventy days is not only the length of the invisibility of Sirius, but is also the time set aside

for funerary preparations after the death of an important person. By the time of the Coffin Texts and the rising star clock tables, the nobility could aspire to an existence after death which was very similar to that reserved for the pharaoh alone during the time of the Pyramid 95 Texts. Sections of the Pyramid Texts illustrate the close links between death and the stars, 96 for example in Utterance 302 which begins `Thesky is clear,Sothislives, I ama living one, because
97 Sothis, the son of and the Two Enneadshave cleansedthemselvesfor me in Ursa Major, the imperishable.My house in the sky will not perish, my throne on earth will not be destroyed, for men hide, the gods fly away. Sothis has caused me to fly up to the sky into the company of my brethren the gods, Nut the great has

from ', Utterance 442 Orion for her `You while states with will regularlyascend uncovered arms me...
the easternregion of the sky, you will regularly descendwith Orion into the western region of the sky' 98

99 links New Kingdom, through These to the continued as evinced by the astronomical the frequently translated New Kingdom beyond. The Duat, as tombs, royal and ceilings of 1'101 `Netherworld', was essentially a celestial domain which surroundedthe material world The stars were visible to living man, but unreachableand intangible. The horizon was a liminal region, a gateway to the Duat, as was the tomb.
95For a collection of stellar referencessee Faulkner 'The King and the Star-religion in the Pyramid Texts'. 96 i. e. is visible (Faulkner Pyramid TextsUtterance302, Note 1). 97 i. e. also a star (Faulkner Pyramid TextsUtterance302, Note 3). 98Both exerptsare Faulkner's translations from his Pyramid Texts.

99 Portions of the Pyramid Texts themselves, as well as texts clearly based on parts of the Pyramid Texts, were (Allen Kingdom New during the still being used as tomb, coffin, and shroud texts throughout Egypt Occurences of Pyramid Texts). from the between distinction 100 passage the is For There visible sky and the Duat. a clear example, consider Ancient Lichtheim in (Davies Aten The Rock Tombs Hymn 6, El Armana to Part translation of the Great (rain, in 2) Aten the associated Vol. Niles: in Duat, Literature three sky where the Egyptian made one one the into lands), three foreign This regions: divides text and one earth. on creation effectively particularly with foreign lands, and the physical Egypt, but also highlights the fact that the tangible sky pt the realm, spiritual Duat. the the as same was not

lot The Osireion itself was thought to have a special connection with the Duat, as evinced by a graffito there LXXXVIII). 3 dit'secret land Osireion Graffiti 1198n to Duat' Cenotaph (Frankfort pl. as the of which refers

111

It is possible that the use of the decans to mark the hours was a by-product of another, more immediately practical, activity: marking a period of days. If the death and burial of a person can be likened to the disappearance and heliacal rise of a star, then the period between death

and reappearance and burial can similarly be linked to the period between the disappearance idea leads to a hypothetical course of development of the rising star clock This of a star. which is outlined below.
The link between the number days of funerary preparations and the disappearance of a star

disappearance dates list to the their the and create motivation of stars of a and provides
reappearance, with the goal of finding stars with a period of invisibility of around seventy

days. If observationswere begun on that basis, it would becomeapparentvery soon that the the large. be Some to task work the simplify was of very short cuts would needed magnitude to Firstly, be brightest the obvious the the most end product. easiest and stars would and observe. Another fairly obvious simplification would be to assume that a star that disappeareda days the days Sirius after that of after same number would reappear certain number of heliacal rise of Sirius. This assumptionis not true: a combination of declination, magnitude, 02 If this were invisibility the assumption the of sun govern star. period of of a and position form in a be located declinations not the and sky the could chosen stars at various made, the stars chosen checked, `belt'. Conversely, if the periods of disappearance were carefully have Parker belt Neugebauer lie and or region of the sky, which on a certain would calculated 103 and plotted

day be have to the Using would need limit set on A number of stars used. one star per would

102 Schaefer`Heliacal Rise Phenomena'. 103 Egyptian Astronomical Texts 1 p. 100 fig. 27.

112

a list of 365 stars. Bright stars are not frequent so a list this long would contain very faint stars and also gaps where no star was suitable. One star per month is too long in comparison day be between Some frequency the to seventy-day period which was one with measured.

and thirty days would be chosen. Exactly how each star was selected is not known, but a combination of magnitude, familiarity, period of disappearance, and a certain amount of human error and leeway would
be list to made with a reasonable amount of ease. a allow

In order to use the list of stars for the original purposeit was intended, one would find on the
list the star which disappeared on or around the date of death. The reappearance of that star

burial, Duat, deceased's time the the to the or some other of or entrance mark of would
important moment in the existence of the dead.

into dates list) decan list a (the The transformation of this and of star names prototype of a the in hours of the observation some addition of a new axis marked and star clock requires be few listed Very would observations stars after their reappearance. movements of the diagonal is the to produce pattern which characteristicof the clocks. needed is during the night The fact that in most of the existing clocks'04twelve hours are marked off decan days and heliacal interval between one ten of the the the to rising of choice as related the major is hour twelve) (ten days It was of certain which not period or one the next. is hours considered the generally the clock, making of when although number consideration followed the if course However, development the the clock star to take precedence. of the may `hour by-product `day opposite that clock', so an clock' was above a of a outlined
104 A10 and A14 mark only eight hours, while A16 shows only ten. A10 is too garbled to ascertainwhether this (like the being likely) hours used to (most due considerations were space the is or whether only middle eight hours, is A14 to when ninth well ordered and appears to represent the second Osireion shadow clock). is A16 does A13. table somewhat to and not appear represent another with of a subsection compared disordered.

113

be true. An important feature of this hypothetical procedure is that the events that happened to the dead person were regulated by the stars and not by the sun nor by any other means of by living in location This time ties the people. used with of the surviving star clock marking tables which were painted on the inside of the lids of coffins or later, as will be discussed, on the ceilings of funerary monuments.

We have now described a hypothetical scenario which provides a motive and a method for the creation of a list of star names that measuredtime primarily in days but could also be '05 has hypothesis hours development The the the to the of measure of night. made highlighted someimportant points about the star clock tables. Firstly, it is unlikely that the decansfirst appearedin the format of a star clock table. This long the decans fact important after that the to group continued appearas a cohesive, mirrors entire genre of star clocks was extinct. inaccuracy to inherently and Secondly, since near-horizon observations106 prone are the the distributed from sky, to across one year another and stars are not uniformly variation in times developed in are the to modem we was not which atmosphere of precision clock star human is to Also, the choice of stars that fit certain criteria subject accustomed. interpretation. Consequently, star clocks do not yield to stringent analysis as much

information as we would expect from a timekeeping instrument. We saw that this was also the casewith the Ramessidestar clock which is capableof producing a massof meaningfullooking equationswithout supplying very much of an insight into what criteria were used to constructthe clock.
105Both Krauss ('Astronomische Konzepte') and Locher ('Middle Kingdom astronomical coffin lids') have decans evolved from `seasonal' markers to `hour' markers. that the theories expressed previously

16 Schaefer`Refraction near the horizon', `Extinction Angles', and `Heliacal rise phenomena'.

114

Thirdly, the Egyptian concept of `hour' does not correlate exactly to any modem concept of `hour'. We have seen how the surviving diagonal star clocks from coffin lids were not

labelled with hour names and how the Osireion passage star clock used the word sp instead

of wnwt. In the analysis of the timekeeping properties of each device we have not found a
uniform hour among them. These three observations indicate that neither the identification of the tables wholly as an (intended) accurate timekeeping method, nor the view that the tables had no function at all

The is than the symbolism, representativeof other available cultural and physical evidence. truth, if we ever have evidence enough to discover it, will probably be somewhere in between. It is this lack of definition that leads to difficulties when identification of decansis information without more about the developmentand use of the tables. undertaken From the earliest postulated use of the decans,through the New Kingdom and beyond, the impression remains that each timekeeping method, and probably each individual timekeeping device, defined time. This is a vitally important point because it alters

those In the these especially studies, way we must approach many objects. completely `How Karnak has been accurately the to the the water-clock, question: on emphasis related did this object tell the time?' If the device was obviously not tailored for equal hours

the immediately Greek hours as the the adopted throughout year, alternative of seasonal was devices. for judging, Egyptian these examining, and ultimately systemof choice For from leads the instant distancing objects. to an This type of assumption of the researcher the begin of problem timekeeping with method such an approach must always each Amenemhet Eber's Texts the show such as the of papyrus and autobiography calibration. texts knowledge these offers length, but was collected of about that some night neither hour that to tables lengths state of throughout the year existed. evidence enough

115

The water-clock is always the favourite candidate for calibrating other timekeeping devices. It is an attractive idea perhaps because water-clocks are the most immediately

comprehensible timekeepers to the modem mind.

To this end, researchers have proposed

that water-clocks provided calibration for diagonal star clocks and were also used in the daylight hours to calibrate sundials. Without further insight into the nature of hours
measured by both water-clocks and astronomical timekeeping methods, such theories cannot be proved. '07 These points, raised by discussion of the development of the first type of star clock, will

be to pertinent throughout this discussion of the development of timekeeping continue methods.

The `Transit Star Clock'

The accepted chronology of timekeeping in Egypt places the `transit star clock' as a

developmentand improvement of the diagonal `rising' star clock. The reasonfor this order `transit is diagonal star two the surviving obvious: surviving star clocks pre-date of events clocks'.

107 Only one water-clock dating from the New Kingdom has survived. The decoration of the Karnak waterlength discussed in Section H, discussed but its be have been at timekeeping abilities analysedand clock will by many researchers. (See articles by Balmer; Cotterell, Dickson, and Kamminga; Fermor, Burgess, and Przybylinski; and Sloley). Approaching the object from the point of view of accuracy, hydrodynamicists are hamperedby the lack of a spout in the preservedportions of the vessel. The nature of the spout is critical to hence flow in and rate of accuracy of measuring seasonalhours. calculations During the course of this study, a replica of the Karnak water-clock was constructedby Dr Allan Mills of the University of Leicester, and decoratedby the present author who also had the opportunity of experimenting diameters it Using before display of Leicester's Museum. New Walk various was placed on the at vessel with holes in copper foil (therefore differing from researchers who have used tubes such as hollow needles)the rate by found be the other to directly This fall the water predictions of was surface contradicts constant. of of the have found that the truncated curve Karnak the approximates conical only who surface clock of researchers for a constantrate. required However, despitethe observation that the Karnak vessel could, under certain conditions with a specific type of jet, measureequal time periods, the study of this particular example's timekeeping is defeatedby the fact that inner inscribed the hour on surface are not equally placed. marks the Without further New Kingdom examples, it is impossible to say whether the marks were placed with care, or fortuitous the and was the vessel particularly both are of chosen, shape these or considerations whether whether this to one example. unique

116

The `transit star clock' offers two potential areas for increasing the accuracy of the star clock. Firstly, the observations would be removed from the horizon area, eliminating many observational difficulties, and secondly, using the meridian (a great circle) as datum would allow equal hours to be measured with ease. However, both of these reasons stem from a desire for accuracy, and almost certainly do not reflect Egyptian timekeeping modem

priorities.
From the argument above for the evolution of the diagonal star clock from a list of stars

days, the to passage measure of created we can see that the date list contained within the
Book of Nut resembles more closely an embellished list of this type (with the event `First'

being addedto the necessary`Enclosure' and `Birth') than an improved diagonal star clock. is Kingdom New date list form The survival of a descendant to the the of original of and star impossible not an concept, especially when it occurs in the Osireion complex where certainly 108 The diagonal for decoration unique type star clock ceiling. also of was used as another a building for location Osireion decoration it the likely was the that of makes nature and have knowledge. Texts the seemedarchaic to which might chosen represent oldest available decorating in the for Osireion be the the time use the construction of of copied out would at for inspiration the been imagine, have the these copies, one can may walls and ceilings, and in Nut Tomb Book Ramesses the IV. the of of use of A date for the star calendarcan be obtained from the heliacal rising of Sirius. By presuming IIII `Birth' 35 Neugebauer on date Parker109 to occurred relates spdt, set that and saw that implies BC, 1850 date list in Middle 16, the Kingdom, and the Peret which places around been have hence list, date more timekeeping it and any that the method represents,could

lollThe Sloping Passage clock, sourceBi. 109 Egyptian Astronomical Texts2 p. 54.

117

than five hundred years old at its incorporation into the decoration of the Osireion. Leitzllo
argues for a much earlier date. He combines an earlier year for prt spdt on IIII Peret 16 = 16th July, a 70-day period of invisibility of Sirius, and culmination of Sirius at midnight on I

Akhet 1 to arrive at a date of around 3323 BC for the compilation of the date list. This date

cannot be confirmed: the use of `culmination' and `midnight' is problematic, and there are difficulties associated with determining an era for the basis of the 70-day period of invisibility of Sirius which were noted at the beginning of this section.
What is indisputable is the remarkable survival of the Book of Nut into the present era in the form of Papyri Carlsberg 1 and la (which date from the second century AD) although the New from is the in documents. If text text these not preserved can survive a star calendar Kingdom to Roman times, it is certainly possible that the basis for, or even the original of,

describes its text the that the that text pre-dates earliest surviving occurrence, and method be may even older. It is therefore conjectured that the star calendar in the Book of Nut is an embellishment of a list of starswhich displayed a common period of invisibility, and as such is not an improved diagonal star clock. has be For this conjecture to given credence,it is necessaryto review the evidence which Neugebauer forward in and been `transit the which put support of clock' star previously Parkerdescribe. has list date the fragmentary instance, list decan first the In the of names associatedwith Also, lists. have decan of later been links none to only shown circumstantial with already lent is to in is found lists later support timekeeping so no conjunction with a these method, from that quarter. theory transit clock the
10 Leitz Studien zur aegyptischenAstronomie 1 p. 52.

118

Far more important is the identification of the event `First' or tpt with concrete, a observable astronomical event: the transit of a star across the meridian at the end of the first hour of the

night.
This behaviour, meaning the transit marking the first hour of the night 200 days after `Birth', is true of Sirius, which has a visual magnitude of does but fainter hold for not stars. -1.60, Although the time of transit in relation to the time of sunset (the condition for `First Hour') is not dependent on magnitude, the date of `Birth' certainly is. Therefore, if Sirius were in exactly the same position, but was a star of magnitude 1.0, the date of `Birth' would be five days later. Similarly, if Sirius had a magnitude of only 3.0, the date of `Birth' would be 17

days later. It is also possible that some decans are fainter than the third magnitude. A difference of 10 days at `Birth' means that transit at the date given by `First' occurs one Egyptian `hour' later than required, and a difference of 20 days at `Birth' meansthat transit at the date given by `First' occurs two Egyptian `hours' later than required. There is, therefore, an inherent problem with the date list if it is to be used as a star clock in the manner which Neugebauerand Parker suggest. The observer would have his meridian "' but for line, the starswould for set up using an external reference south and a mrht a sight be behave `Enclosure' `Birth' to the would as expected with regards not and position tpt. does information but identification list the date not the on which as clock rests correct, of the

Ill See Larson The Tutankhamun Astronomical Instrument'. The instrument was described by Breasted as `A At little 1/z". just 1011Y' long of 1 one end 1/16" ebony wood strip of a thickness over rectangular wide, ... ... the ebony strip is a rectangular mortise hole a little over half an inch long, about 3/16" wide, and a scant '/d' deep. ' Berlin Museum Inv. Nr. 14084 is also mentioned, which is inscribed `A watching stick for determining for in hour(s) their '. The article states that `Using simple tools such as these, festival men all placing and a ... and by keeping careful records of their star observations, Egyptian astronomers were able to predict when a ' However, cross a meridian. the mrlt may well have been a daytime shadow timekeeping would star particular instrument rather than a night-time meridian instrument. Section E will deal with shadow clocks which in but have be way, their they every that used as can marks mrht on resemble upper surfaces which mean Surviving shadow clocks are of stone with engraved marks, but it is possible that wooden mrbt may sundials. have had painted marks and may have been used as sundials.

119

hold. The existence of this problem is a direct result of the assumption that the event `First' occurred on the meridian. No clear evidence exists which proves that the Egyptians had Its place in Egyptian astronomy

defined this imaginary line on the sky in the New Kingdom. derives from two occurrences only:

the `transit star clock' and the Ramesside star clock.

We have already seen that the meridian was simply a convenient modem object to tie to the `r k3tb' position of the Ramesside star clock. The phrase r k3fb, `straight forward' is the only term that has been postulated to mean `on the meridian'. Both the Ramesside star clock

in the conventional usage and the `transit star clock' need, in order to work, a precise

definition of the meridian, yet we have identified no term for it and no external references have beenproved to relate to the meridian.112 The meridian appears to have been introduced into Egyptian astronomy by modern by them intent down devices within the timekeeping placing on pinning ancient researchers, the framework of modem astronomicalterminology.
Finally, some major evidence which has been used to support the theory of the transit star from dating the la from Carlsberg Carlsberg second 1 the two papyri and clock comes century AD. These are copies of a commentary on the Book of Nut, and are apparently

113 hand in the same written

have Nut held have the the vignette It is a great pity that portions of the papyri which would decans to is list It have full accompany that they of possible may contained a survived. not the date table which would be of great interest to the current discussion. `Ua' labelled include interpretation which text the the The surviving portions of papyri of an
112See also ba L'Orientation Astronomique. Parker ('Ancient Egyptian astronomy' p. 51) gives only one inscription Harkhebi in for (that is the term culmination a star crossing the meridian): occurrence of a displays flavoured' 3 Texts 214-216). However Astronomical and is `astrologically pp. the text (Egyptian Babylonian influences. 113In the discussion which follows, we shall use the term `scribe' for the author of the original commentary. discussion. be to the is the out but same person as this copyist of the papyri, not pertinent This may or may not

120

days for 120 before `working' `First' for each star. The nature of the the period uses word this `working' is not fully explained within the text, but it is implied that the star marks the first hour of the night for ten days before the date called `First', the second hour of the night for the ten days before that, and so on up to the twelfth hour for the first ten days of the 120day `working period'. This implies a timekeeping structure similar to the Ramesside star

clock, omitting only the `beginning of the night'.

The scribe breaksthe year of a typical star'14into the following sequence of events: 110 days work
IIII Akhet 26 begins to work

10 days work
`First' I Peret 6 stops work 90 days in the west `Enclosure' IIII Peret 6 70 days in the Duat `Birth' II Shemu 16 80 days in the east In contrast, text V simply states that there are `29 [decans] living andworkingin heaven'and gives information `working'. further about no

is 115 in the infer papyri Neugebauerand Parker that the 120-day working period mentioned hence that V labelled [b3k] in Osireion text and `working' to the the the referred version of for device list from a New Kingdom onwards, as a star clock, the the date was regarded, finding the hour of the night.
is be Nut to Carlsberg 1 and la understood the Book of a star clock That the scribe of Papyri for the be scribe's There to two addressed. explanations needs are possible a point which timekeeping is forming his he from his other of either conclusions experiences comments: difficulty had had he That to the scribe access a more complete source. methods, or with

diagram. from the is Book Nut from texts he the the the of clear way attacks understanding in diagram in he the dealing an Book, the reads For example, as we shall see the section with dates those either 114 of the list the star asph" d3t,but reconstruction decan The scribenames makes the of Parker Neugebauer (as fart hrt and noted)or tmJthit, notpkwy d3t. tm3t
115 Egyptian Astronomical Texts 1 p. 56 commentary 44.

121

illogical

order.

His explanations are not clearly stated and are sometimes obviously

116 The scribe, it must be remembered, from fifteen was around centuries removed wrong. the Osireion text (and up to two thousand years removed from the original text) and so be be to an expert on the original meaning of the text. assumed cannot This makes it

indeed he forming his own understanding of the text using outside that was possible information. Conversely, we have already noted that the Book of Nut as preserved in the

Osireion and the tomb of Ramesses IV is an incomplete document and it is therefore possible that the scribe had a fuller version of the text. However, it is notable that, if this is the case, Kingdom decans do in New is the see of any which we not preserved made no mention does in hours the the to the context of scribe also not other and allude any of night version, timekeeping, which may have helped him clarify his description. This may indicate that the

former case is more likely, but it cannot be determined which of the two possibilities, if

is either, true. is itself in list date from New Kingdom find not the the In conclusion, that sourcesonly, we date list during included the find decan list Firstly, the time the night. the with sufficient to in the Book of Nut is too fragmentary to permit the use of the date list in such a way. lend does information its date list not the the Secondly, contained within arrangement and itself to this use.
in the first determine list date trying to the the presented It seems prudent when nature of New Kingdom to allow New Kingdom sources the primary focus. The label tpt can surely be taken to mean wnwt tpt `First Hour' with the meaning `beginning of the night', without been has It it be table interpreted the also implying that as stands should entire as a clock. 17that the 120 days related to tpt simply during time which the of period represent suggested' hours darkness. in during the is If `culmination' of sense taken of general a culminates a star 116 of theBookof Nut. SeeSectionF for a full discussion
Egyptian `Ancient 117 star clocks' p. 9. Puydt De

122

the highest altitude a star attains, rather than in the precise definition

of transiting the

meridian, this provides a meaning for the third event in the star's year: disappearance, reappearance, and attaining greatest height at the beginning of the night. Such a rough

definition of tpt, and of 120 days working is more in keeping with the nature of the list as a calculated, rather than observed, description of annual events.

It may still be that a true and complete star clock based on the movements of the decans away from the horizon area existed during the New Kingdom and was the basis or a parallel
for this text, but the Book of Nut does not contain the entirety of such a device.

The existenceof a method for determining the hours of the night using the transits of decans is has been disputed Nut Book the that the meridian star calendar and we now see across of descendent to direct diagonal is but the related closely a more of star clock not necessarily for seventydays. the precursorof the diagonal star clock, a list of starswhich disappeared Ramesside Star Clock The final type of star clock that we have examined is the Ramessidestar clock. We have discussedat length the problems of interpreting this type of star clock. We now have one further piece of information: that the existence of transits as a timekeeping method prior to the Ramesside star clock has not been established. In particular, this meansthat the meridian Egyptian be been to has itself proven not a recognisable and well-defined object within in is lb rk3 lends idea This used to the that the which credence r central position astronomy. it. be the Ramesside clock may not star meridian the as we understand
be (assumed date to `First' for in contained the the transit star clock By comparing of spdt in T12 35) of the the beginning date star the night occurrence of of and sb3 set n spd as with Ramesside Neugebauer date Parker the star the Ramesside of clock, star and estimate the 118 is it 1470 BC. This date be not be since to treated caution119 around must with clock
1I8Egyptian Astronomical Texts 2 p. 9.

123

120 is hour decan the star sb3n spd exactly the same as the certain that spdt, and also it is not certain that the definition of `beginning of the night' can be equated between the two date Leitz121 1463 BC. suggests an exact of methods. Concerning this last point, it is clear that the nature of hours marked by the two types of star is different if is be to sb3n spd equivalent spdt, since sb3 n spd r k3ib at the end clock must days its from hour the `beginning thirty the the whereas appearance as of night' second of `transit star clock' places the star at the end of the second hour only twenty days before
`First'. The date found by Neugebauer and Parker therefore remains questionable. It is easy to see the similarities between the `transit star clock' method and the Ramesside in both both the the the meridian. use conventional of used way: make use mrjit, clock star It is difficult to see, however, what improvement the Ramesside star clock offered over the transit star clock that caused the later method to be developed. The Ramesside star clock Ramesside larger The star clock tables, set of stars. extra observer, and a new an needed `transit from in diagonal be clock' star the the or a clock star used memory way not could

had the potential to be used. decans the last that in used We postulated the section that some timekeeping method existed the in high but based meridian. the of the concept they sky, modem was not on were while device this be of Ramesside by a We seethat the star clock used the reflection method would
different set of stars we call `hour stars'. type, using a somewhat

If the notion of `meridian' is removed from the two timekeeping methods, their relationship

119Fermor ('Perceived night length ratios in ancient Egypt') disagrees with Neugebauer and Parker's date of De be while Texts 2 9) believes Astronomical date out, ruled that an earlier cannot 1470 BC (Egyptian p. and dates believes transits so Egyptian than and that the tables mark risings rather star clocks') Puydt ('Ancient the bringing on BC, tables the time painted 1200 date the were to the of creation closer on which them to c. for this His he (note that the tables tombs). arguments the mistakenly places the on tomb ceilings walls of 23 TI 0 T24 to between equals include (such the -12 number of stars counting as interpretation certain events is hour but length is distribution stars hampered this of fact the by the that approach stars) to estimate night (as is the probable case with decans). be to even unlikely 120 See Leitz Altaegyptische Sternuhren. 121 Leitz Altaegyptische Sternuhren pp. 132-137.

124

also falls apart. The discussion of the `transit star clock' left us with only a hypothetical instrument to examine, and its replacement, also hypothetical, with a timekeeping instrument

that existed at the time of writing of the Book of Nut which used decanal stars at some altitude above the horizon, gives us no link with the Ramesside star clock which has previously beenheld to have followed it. As a timekeepingmethod which probably did not survive in any form, except in the tombs of the Ramessidekings, after the end of the XXth dynasty, the Ramessidestar clock seemsto be a unique device, not an improvement or embellishment of an existing method, but an instrument. new entirely
The reasons for constructing the new tables are not certain, but the observational method is capable of reproducing hours which varied throughout the year, and indeed throughout the night. It is tempting to suggest that the new stellar method was created to reflect a

development in the manufacture of water-clocks (Amenemhet's invention of the seasonal just Ramesside before date the star clock the occurred of suggested of origin water-clock in hours the due interest in describing New Kingdom the night to the of tables) or was

funerary literature.

A Chronology of Stellar Timekeeping

We have so far constructed a sequenceof events and developmentsthat fits well with the
timekeeping to stellar and offers place us a structure within which available evidence,

discussed, dates 41). Many in (Table the this schematic chronology, as already of methods interpretations. The illustrates sources, our of the to nature chronology other patchy are open be to too likelihood is that small our sample of astronomical texts probably and the the of rangeof material which originally existed. representative completely decanal into breaks the four development the of find that We chronology sections: the early
little know decanal coffins, on very about; the system we which a period of using system,

125

during which revisions were made and towards the end of which the date list preserved in the

Book of Nut was compiled; then the New Kingdom, where old texts and methods were reused without revision, the Ramessidestar clock was developed, and celestial diagrams introduced, first in in have they or were either used places where survived numbers; were and finally, through the remainder of Egyptian civilisation, the decans become symbolic remnants. The survival of the Book of Nut into the secondcentury of the present era has already been being is it but documents instance in history the only most spectacular of remarkedupon, a kept for centuries before they were used to decorate a monument or object which has survived.
Sirius disappeared for seventy days BC 3500 after

Possible datefor basis of Book of Nut date list Stars were chosenwith invisibility of 70 days (the customaryfunereal period) Stars were usedto find hour of the night at rising
(causing the introduction of triangle)

c. 3323 BC ? ? by 2640 BC 2150 - 2000 BC

2000 - 1800 BC

A triangle had beencompiled Revision of star clock resulting in Al to A9, A15 to All
Triangle for celestial diagrams formed

Revision of star clock resulting in A10 to A14, B1 Decans were usedto find hour of the night while high in the sky Possible datefor construction of Book of Nut date list, /1 Ramesside star clock constructed Osireion built and decorated with Book of Nut Sloping Passagestar clock using BI IV Tombof Ramesses Ramesside star clock in Valley of the Kings Decans usedas group in celestial diagrams Carlsberg 1 and la

1990 - 1760 BC ? if ever 1850 BC 1450 BC 1300 BC 1220 BC 1160 BC 1150 - 1100 BC 1460 BC -AD 100 AD 150

Table 41: A chronology of stellar timekeeping in Ancient Egypt

126

Section E: Shadow Clocks and Sloping Sundials

Introduction
Egyptian shadow clocks are small instruments consisting of a horizontal base rod of rectangular section with a vertical block at one end. The base rod is inscribed with small

for hour block The has hole line scale. making up an suitable circles vertical a and reference the attachment of a plumb bob. indentations. In 1910 Borchardt122,123 proposed that the surviving shadow clocks were incomplete, and that an additional element in the form of a crossbar(a rod of similar dimensions to the base
its mid-point to the top face of the vertical block and extending horizontally rod attached at at right angles to the scale) was originally part of each shadow clock.

The top surface of the block can contain holes or

In this section, the shadow clock will be examined using extant examples and texts which
describe or depict this type of timekeeping method. Borchardt's theory will be extensively

discussed. A second type of sundial, the `sloping' or `Egyptian' sundial, dating from the Late Period the in discussed, be development this and to the method of particular with reference will also between later the types of sundials. earlier and connection

Sources
El Berlin Aegyptisches Museum Inv. Nr. 19744124

125 E2 Osireion Text, part dealing with the shadow clock headedby a diagram E3 Berlin AegyptischesMuseum Inv. Nr. 19743121 E4 Tanis papyrus126
ES Qantara sloping sundial127

"Borchardt 'Altgyptische Sonnenuhren'. 123 Borchardt Zeitmessung.

124 Details of the two Berlin instruments are provided in Borchardt 'Altgyptische Sonnenuhren'. 125 Frankfort The Cenotaph of Seti I. 126 Griffith and Petrie Two Hieroglyphic Papyri. 127 C16dat'Notes sur 1'isthme de Suez' and Kuentz 'Notes sur un gnomon portatif Greco-Egyptien'.

127

The list of objects above is not exhaustive, other sundials and artefacts presumed to be part in have sundials survived and are collections around the world. of The shadow clock first appears at the beginning of the New Kingdom. It is probable that the

128 been had but these long before this time, used to mark time periods shadow of the sun devices would have been informal, non-standardised objects, perhaps resembling the basic

29 in The shadow clock is the earliest sundials seen pastoral societies around the world. sundial used in Egyptian society that has some surviving documentation and aspects of 130 standardisation. El is inscribed with the name of Tuthmosis III. The base rod has five circles making up an

128The Middle Kingdom Prophecy of Neferti alludes to the connection between shadows and timekeeping. Parkinson (Voices from Ancient Egypt) translates the relevant passage as:

`Re withdraws himself from mankind Though he shall rise when it is time, it shall not be known when midday occurs; there is no one who can distinguish his shadow, no one's face is bright when (he) is glimpsed' '29SeeFloyer 'Primitive Sundials in Upper Egypt'. 130 There are a few other candidates for solar timekeeping in the New Kingdom and later. In particular, inscriptions is let down by This which state theory as giant sundials. often cited obelisk are attractive obelisks her for. Hatshepsut future the that obelisks may wonder were of obelisks admirers even predicts clearly what saying: what their purposewas, and preemptstheir guesses `Now my heart turns to and fro, In thinking what will the people say, They who shall seemy monument in after years, And shall speakof what I have done.
Beware of saying, "I know not, I know not: Why has this been done?

To fashion a mountain of gold throughout, " Like somethingthat just happened. I swear,as I am loved of Re, thesetwo great obelisks, regards as ...... Wrought with electrum by my majesty for my father Amun, In order that my name may endurein this temple, For eternity and everlastingness,
They are each of one block of hard granite, Without seam, without joining together! ' (Lichtheim Ancient Egyptian Literature vol. 2) 33401) (No. in Cairo limestone it was a that in thought Museum Egyptian the a white Until recently was model Borchardt's follows Clagett combined with a sloping sundial and a shadow clock. type of stepped sundial `sloping the device, timekeeping theory the the a requires as that object identification of even after commenting is in It be highly widely now to he be knows an east-west used orientation, to unlikely. which portion sundial' is a base for an architectural model. Also recently discredited is the assertion that a this that object accepted from Gezer in Palestine (published in Macalister The Excavation of Gezer and described as a disk ivory small Zeitmesseung and Sloley Primitive Methods of Measuring Time) which bears a cartouche Borchardt in sundial is some radial the New Kingdom and side marks Merneptah one on on vertical sundial. other, a of

128

hour scale. If the length between the vertical block and the first circle is called 1 unit, then the length between the first circle and the second is 2 units, between the second and the third
is 3 units, between the third and the fourth is four units and between the fourth and the fifth is 5 units. This ratio 1:2: 3: 4: 5 is also displayed by the other example in Berlin, which is

around 500 years later in date.

I1T
__

j i k;

UJ ?
Figure 15: Shadow clock from the time of'Futhmosis (from an Aegyplischcs Museum pluilograph) III (LI)

Figure 16: Shadow clock from Fayum (E3) (from an Aegyptischcs Museum photograph)

The ratio is also present in a New Kingdom text, the Book of Nut, in the version which
in Sarcophagus Chamber the of the Osireion appears dealt in he later Nut with a will section. of is E2 here to source as reproduced refer at Abydos. The other parts of the Book

The part pertinent to the shadow clock which we

in Figure 17. In the diagram which heads the text,

from issue L-shaped body lines is depicted. the Five the characteristic of clock vertical short the base rod and above the spaces between these lines are the numbers 3 (closest to the 12. 6,9, block), and vertical Although
instruments

the vertical marks are roughly evenly spaced, comparison with the existing
indicates that the numerals relate to the distance between the marks. Of course III

be 12 9: 1: 2: 3: 6: 3: 4, the ratio which we noted for the Tuthmosis to reduced can the ratio

129

instrument. Also, although the diagram shares with the Berlin shadow clocks the five marks, the surviving shadow clocks do not have the unlabelled gap between the vertical block and the first mark, the distance from the block to the first mark is 1 unit, whereas in the Osireion text it is `1 unit' from the first mark to the second.

The Crossbar Hypothesis

Borchardt first postulatedthe addition of a crossbarin order to make senseof the hour marks
on the two instruments in Berlin. 131 height, With the addition of a crossbar of a certain The theory

Borchardt claimed that the sundials could be made to mark seasonal hours.

instrument be the that used in a certain manner, being aligned exactly east-west. stipulated
In the morning, the crossbar end would be placed to the east. At noon the shadow clock

180, in be be that to the west. turned the the so afternoon, crossbar end would would In 1965 Bruins132 explored Borchardt's theory mathematically and concluded that a set of three crossbarsof differing heights would mean that the time periods measuredby the clock hours. to good approximation very seasonal a were In this section, the arguments for and against the crossbar hypothesis will be examined in the instrument Firstly, context the the then shall we examine written evidence, carefully. hours in Egypt, and for the the seasonal of ancient argument measurement of sundials finally, the issuesraisedby the conjecturedmanner of usageof the crossbarshadowclock.

Textual Sources
12 3,6,9, figures (Figure is 17) headed E2 by diagram has and the The Osireion text a which labelling the gaps between the hour marks. The diagram is simple, but we gain one very V. letter information from it: like device that the our important piece of was shapeof the
indicate (in Die Altgyptische to diagrams Zeitmessung 131 Sonnenuhren') seem `Altgyptische Borchardt's and be block. to the believed the height he casting edge shadow at the vertical same that as surviving 132 Bruins `The Egyptian Shadow Clock'.

130

III

iii iii

it

7Q

ON It

,V 4

OQ 1\

01

I as , y
os 11{I{II 47f! t7tHNun i4) 0 1'1H I 11 p

Figure 17: Shadow clock text from the Book of Nut in the Osireion (from Frankfort The Cenotaphof Seti I)

The Egyptian artistic style portrays three-dimensionalobjects by displaying the characteristic 133 The details the the outline. object with other main shown above or around shape of but the L-shaped, be scale felt the to of shape shadow clock was obviously characteristic The important depicted addition of the and so were above main component. markings were diagram A `T'. of the the a clearly make would characteristic object shape of a crossbar hour drawn be the if looking down marks the with probably would as clock, clock a on such is known diagram base No distort the to their position. rod with no need on such represented
to exist.

133 This is clearly illustrated to be the preferred method even for technical diagrams, such as architectural plan IV' and Davies `An Architect's Plan from Thebes' for example). (Carter and Gardiner `Tomb of Ramesses

131

The text below the diagram is in 13 columns. The manufacture of a shadow clock is first described, followed by instructions for using the instrument. There are gaps of omission in the text. The extract presented below concerns the usage of the clock. The problem of

interpreting the text comes from the unusual vocabulary employed.

If you correctly adjust this st3t rightly in the sun alignment, its head end in the east which is on this mrliyr, then the shadow of the sun will be exactly on this Wt. Then, with the end of the fourth hour you should turn this st3t,

its mrtwt towards the eastaccompanyingthe sun standing at the wpt of this mrbyt.

The entire instrument is called a st3t. The word mrht or mrliyt has been used elsewhere for

but the usual form of instrument to which the word `surveying instrument' in general134 hour but lack in the the marks on of unadorned shadow resembles clock all ways one: refers the upper face of the base rod. It seemsmost likely, therefore, that the word should apply to the L-shaped part rather than to the crossbar. However, in the reworkings of de Buck's 138 136 137 by Parker, Clagett, Bruins, and the translation of mrhyt as translation13s of this text `crossbar' has become established. The word wpt used anatomically means `top of the head', but also has the meaning of `zenith'. Mrtwt is not understoodat all, although Parker indicate instrument. it the to the scale rod of uses The extract indicates that the instrument neededto be aligned correctly in order to produce a 180 The be interpreted in motion turning the two ways: a single act of clock can reading. 139, at noon or a continuous turning towards the sun. A much later text'4 E4 dating from the Roman period is a fragmentary papyrus showing a diagram of a shadow clock (Figure 18).
134 7ba L'Orientation Astronomique p. 55,56 (where he believes that mrht applies to the vertical block). 135 Frankfort The Cenotaphof Seti I. 136 Egyptian Astronomical Texts 1 pp. 116-117. 137 Bruins 'The Egyptian Shadow Clock'.
1- Clagett Calendars, Clocks, and Astronomy pp. 467-470. hour four just fourth hour in 139 the instrument the this text, which describes an Which occurs at end of with five marks we find on existing shadow clocks. instead the of marks

140 Griffith and Petrie Two Hieroglyphic Papyri from Tanis.

132

The diagram appears to show the base rod of a shadow clock, with numerals 6 and 2 above

it. These numbers probably represent the hour indicated when the shadow fell on that
portion of the scale. The diagonal lines converge to a point now lost. The altitude of this

point above the base line has been thought'41 to indicate a gnomon height greater than the block which survives on E3 thus suggesting the addition of a crossbar. However, the
due is flawed to the Egyptian style of graphically representing relationships and argument (as diagram hour in Osireion the the the we noted was not proportions case with marks which were shown almost evenly spaced, retaining their relative position but not their exact position). Thus the nature and dimensions of the gnomon which is now lost from this

papyrus can only remain conjectural.

I" II

r iv j . (f JL-

11n. 9 Z of

I i'

Figure 18: Tanis papyrus (from Griffith and Petrie Two Hieroglyphic Papyri)

The placement of the hour numbers and the convergence of the diagonal lines indicate that Although the in Osireion, like from diagram, the instrument the the the side. one shows this

Lit is lost, likely an diagram, that showed seems this most when complete, gnomon part
T-shaped from a object viewed not object, above. shaped
141 Borchardt `Altgyptische Sonnenuhren'.

133

These two texts are widely separated in date. The Osireion text shows and mentions only four hour marks whereas the surviving shadow clock El which predates the carving of the text has five. It has been argued142that the Osireion text dates back to a time when shadow has included been the had four the text of part only as marks, old originally and clocks Osireion. the style of primaeval

The Shadow Clock in the Context of Egyptian Timekeeping

The shadow clock is the earliest formal timekeeping device for finding the hour of the day
from Egypt which survives to the present. It is probable that simple gnomons were used before the New Kingdom to mark periods of time during the day for such purposes as instrument. but has identified been currently shifts, no object as such an workers'

7(bmmA (a) (c)

L
'uff

(e)

(1)

Figure 19: Ptolemaic hieroglyphs

depicting shadow clocks and sloping sundials

Ptolemaic from the depicting but Hieroglyphs shadow clocks are not common, several exist 43 Two of the Ptolemaic hieroglyphs depict sundial types of which examples have era. survived: diagram (a) of Figure 19 represents a shadow clock similar to the Berlin

142 Egyptian Astronomical Texts1 p. 118. 143 Valeurs Phonetiquesdes SignesHieroglyphiques d'Epoque Greco-RomaineVol. 4 p. 731.

134

instruments, and diagram (e) represents the later type of sloping or Egyptian seasonal sundial, of which several examples survive. Figure 19 (b) shows a shadow clock with a curved scale. This development may have been

made to facilitate the measuring of hours near dawn and sunset and seems to lead to the
design of the sundial (d). Figure 19 (c) has two structures attached to the base rod, offering a possible explanation for the pair of holes on the top face of the gnomon block E3. The in known hieroglyph no appears of a shadow clock. crossbar

Without surviving examples, it is difficult to reconstruct precisely the development of the hieroglyphs The clock. suggestan evolving design leading from the shadow clock shadow to the sloping sundial, whereasthe two distinct groups of surviving sundials, shadow clocks and sloping sundials, have led researchers to treat the sloping sundial as a separate instrument. The observationthat there is a developmentalpath from the New Kingdom shadow clock to the later sloping sundial has important repercussions in the discussion of the crossbar hypothesis. As already noted, the crossbar appearsin no known hieroglyph representinga that has instrument. Neither Borchardt suggested timekeeping nor any subsequentresearcher three bore In for hypothesis pieces to the stand, crossbar. a order sundials crossbar sloping in fit logical together a must way: of evidence 1) New Kingdom crossbarshadow clocks 2) Hieroglyphic representations of sundials 3) Late Period sloping sundials hypothesis have been 3) 1) distinct treated the In past studies, and crossbar as groups, with however, 2), development 1), The for 3). a separate of and addition postulated affecting only into disorder. The Ptolemaic hieroglyphs show that the shadow clock this scenario throws

135

was subject to certain additions and developments which have not survived in concrete form to the present, but which did not change the fundamental nature of the instrument nor, probably, the manner in which it was used. The hieroglyphs also make plain the close

relationship betweenthe shadowclock and the sloping sundial suggestingthat the later was a developmentof the former. For the crossbar hypothesis to hold with the addition of 2), it must be suggestedthat a crossbarclock which is usedby rigorous east-westalignment and may, Borchardt and Bruins
suggest, be seasonally adjusted by raising or lowering the shadow-casting edge of the

develops into a non-crossbar,sun-pointing instrument which is seasonallyadjusted crossbar, by means of additional hour scales. I suggest that the difference between these two instruments is too great for a logical developmentpath fitting the evidence of the Ptolemaic
hieroglyphs to be credible.

Furthermore,it is difficult to seewhy, if the crossbarclock had from its origin measuredthe desired time periods, the conservative Egyptians would have continued to adjust the design to suit the timekeeping requirements of the day. It seemsfar more likely that the original Egyptian device, an L-shaped, sun-pointing, non-seasonal instrument should have been refined over the centuries to cope with the increasing demand for some uniformity of time that would have resultedfrom foreign input and local progress.

Measurement of Seasonal Hours

It is the habit of the modem mind to ask, when approaching an ancient timekeeping device, `How accurateis it? '. This is the question underlying many previous studies of all types of Egyptian timekeeping methods. The question presupposesthat the researcherknows the be by instrument, to time the the measured of nature and will then gain some understanding level how the competence of of by the manufacturers well the object of calculating

136

its task. performed

In such an environment, Borchardt supposed that the shadow clock

should measure seasonalhours, and that to make the instrument perform adequately, a
be added. His arrival at this conclusion is based on the lack of any other crossbar must measure of hours that would fit with the idea of a formalised timekeeping instrument. In

other words, Borchardt started from an idea of `time' and made the shadow clock measure it to his own satisfaction. We must examine whether his assumptions are acceptable. Firstly, was the concept of

We hours the time the the clock? understood and accepted seasonal shadow at of origin of
know that by the reign of Amenhotep I, the difference in night length in winter and summer

in the form of the seasonal water-clock. was known to exist and had been implemented144 day hours 16 I Akhet night hours 8 II Akhet day hours 14 Phaophi night hours 10 day hours 12 III Akhet Athyr night hours 12 Choiak IIII Akhet day hours 10 hours 14 night day hours 8 I Peret Tybi night [hours] 16 day hours 6 II Peret Mekhir night hours 1[8]
[III] Peret day hours 8 night hours 1[6] Phamenoth

[IIII P]eret
[I Shem]u II Shemu

day hours [10

day [hours] 12 day hours 12

night hours 14]

[night hours 12] night hours []

Pharmuthi
Pakhons Payni

III Shemu IIII Shemu

day hours 16 night hours 8 day [hours] 18 night hours 6

Epiphi Wep-Renpet

Table 42: Comparison of day and night lengths through the year from the Cairo calendar

146 42) (Table During the XVIIIth dynasty, a list of day and night lengths was composed145 12 length: day In III Akhet, based on some unspecified unit. the night and were of equal the Shemu, I the in This that Akhet summer III with suggests equinoxes occurred and units. discussion this Shemu 1111 in in the the of II Peret. Clagett and winter solstice places solstice list in a sectionentitled `Tracesof a 24-hour day with Equal Hours' but statesthat `any modem
144 Water-clock text from the tomb of Amenemhet (see Borchardt Die Altgyptische Zeitmessung and Clagett Calendars, Clocks, and Astronomy pp. 457-462).

145 Cairo papyrus 86637, Bakir The Cairo Calendar. 146 After Clagett Calendars, Clocks, and Astronomy Fig III. 58a,b.

137

assumption that they used accurate equinoctial hours from month to month over the whole year ... [is]
exceedingly unlikely'.

Neugebauerand Parker, having discussedthe probable nature of decanal time and produced
formulae for night length based on decanal hours show that from their model, the twelve

decanalhours of the shortestnight span six modem hours. 47 This observation implies that the units used in this table are modern 60-minute hours, and also that night length was based
decans, the while day length was simply `24 - length of visibility visibility of on of decans'.

This definition of `day' is very unusual indeed and is hard to accept against the background 148 Kingdom In the New Kingdom, we find that religious texts treat the New thought. of hours arounddawn and sunsetvery differently from the hours clearly part of the day and part of the night. It is possible that although the day had twelve hours, and the night had twelve hours, the transition between day and night and vice versa was not instantaneous. That is, there were two periods of time, dawn and sunset, when time was not quantifiable. As Neugebauerand Parker noted, decanal hours would have been of varying length, but were found had list 60 Suppose than this that the shorter minutes. author of certainly consistently decanal hour (for example, a vessel resembling a small waterto one reproduce some means his in (perhaps he had time), definition own that that simply a given and emptied some clock find day began We that, began of when and ended, ended. and and when night perception) for example,if the unit had been45 minutes, a period of 3 modem hours around dawn would `the day' `the beginning between the period the similar of a end night' and with of occur is To dawn which a certain extent, a shorter unit and time around and sunset around sunset.

from I 119 Texts Shortest 0 10, time the 147 Astronomical in is p. Egyptian where night occurs when the sun Pio, 029, 9, find length s, hours. To is 10 the the hours, decanal twelve of subtract to we must sunrise sunset hours 10 leaving from 56") 6h 1n' 043. 31(238m 3o,and 14' See Part II.

138

not incorporated into the measuring scheme could explain the large ratio between the length of night and day on the shortest and longest days. However, the most likely explanation of the data within the table is that the pattern which occurred near the equinoxes, that is the

time of the greatest rate of change of day length, was simply extended to its conclusion
leading to the length ratio of 3: 1 at the solstices.

Returning to the connection between the shadow clock and the seasonalhour, we still have hour firm that the seasonal evidence was the timekeeping goal of the shadow clock. no Instead,it would be much more relevant to ask whether the shadow clock supportsor rejects
the theory that some sort of system of equal hours existed and was used in the early New Kingdom. To answer this question, we must examine the shadow clock in its unadorned, L-shaped in to the order the alignment, and also analyse effect of crossbar and east-west configuration,

New in hours day learn the Egyptian anything about the nature of assesswhether we can Kingdom, and whether the crossbar is indeed a necessary,logical and probable addition to the shadowclock. Bruins evaluatedshadow lengths at the solstices and equinoxes at a latitude of 30N for a the his implies the In he that on marks concluding remarks, crossbarclock placed east-west. heights in produced clocks when varying used conjunction with crossbarsof existing shadow is he In fact, hours relating not the to refers great accuracy. with which accuracy seasonal fact the lengths to hour is but to the shadow extant markings, actually referring the seasonal
the lengths between throughout hours shadow similar the at seasonal that ratios remain very

has bearing This `accuracy' merely the and the no clock, on observation shadow of year. hour sundial made in the fashion of the hypothetical crossbarclock that implies a seasonal interchangeable hour marks. only crossbars, of need and no adjustment would The

139

relationship between the 1:2: 3: 4: 5 marking ratio and seasonal hours is not as strong as Bruins and Borchardt initially suggest. At a latitude of 30N, to follow Bruins' own work, we need to find the position of the sun at each seasonal hour through the year to gauge Borchardt's assertion that the crossbar shadow clock measured seasonal hours. To simplify matters, Bruins considered the sun at the

equinoxesand solsticesonly. We need to find the declination of the sun at the solstices, for which we need to know the
obliquity of the ecliptic, E, for the epoch in question. The formula149 for E is e= 232621". 448 - 46". 8150 T- 0". 0059 T2 + 0". 001813 T3 is Julian T the of number centuries after J2000.0. We can use T= -35 to produce an where 150 in for 1500 BC: 23.9. e approximation To find the hour angle of the sun at each seasonal hour between sunrise and sunset, we first

find the hour angle of the sun at sunrise(HAr) cos HA, = - tan4tan6 is latitude 6 the 4 declination (8 the the the E at the and equinoxes, of sun where =0 at the by to HA, divide We the obtain then six solstice and at winter solstice). summer -F, increase hour find To for declination the and the altitude seasonal per any sun. of angular for hour, the sun each seasonal of we have two equations: azimuth 6 +Cos4 cos 5 cos HA; sin sin sin alti =
sin S- sin sin alt; cosazi " -go" cos 4 cos alt i i=0 where to 12, and HA; indicates the hour angle of the sun at the beginning of the first

149 Taff Computational Spherical Astronomy p. 52. ITI inaccurate is increasingly 150 formula becomes large, but is good enough for the present context. The as Bruins apparently did not correct for the obliquity of the ecliptic, but the difference in results is in fact has no impact on the conclusions drawn in this and study. negligible,

140

hour, the end of the first hour, the end of the second hour to the end of the twelfth up ....

hour of the day.

If we now consider the model of a crossbar clock aligned east-west, we can now calculate the shadow lengths for any height of crossbar. Bruins wanted to compare his results with the Osireion ratio. The marks were spaced at 30,18,9, and 3 units from the gnomon. Bruins

therefore scaledhis results so that the shadow length at the beginning of the secondseasonal
hour was thirty units. The results of our calculations, which agree closely with Bruins' own figures, are shown in bold type in Table 43. B ruins discounted the fifth mark present on the left This has fourth the that the scaled results so mark was correct. extant shadow clocks and (18 be From it is third to the three the table, that mark compared. marks clear although only

first in from Osireion is in diagram) the the the and position, gnomon an acceptable units The do indicate degree hours at all. accuracy of not seasonal with any secondmarks clearly in first high for the table the mark especially, and shown errors are unacceptably percentage for the fifth mark which we also include here. It must be rememberedthat the crossbar was added to produced an `accurate' clock. No is be found here. `accuracy' to such
Markings on shadow clock (length from gnomon to mark) Osireion Diagram 3.00 9.00 18.00

30.00

[45.001

Seasonal hour shadow lengths Winter Solstice 4.54 9.82 17.13

30.00 0.00% 30.00 0.00%

30.00

65.71 31.52% 64.64 30.38%

63.38

% error againstOsireion Equinoxes % error againstOsireion

Summer Solstice

33.92% 4.64 35.34%

4.76

8.35% 10.00 10.00%

10.20

-5.08% 17.32 -3.93%

17.54

% error againstOsireion
Table 43: Comparison

36.97%

11.76%

-2.62%

0.00%

29.00%

of existing hour marks with those required to produce seasonal hours

141

Usage of the Shadow Clock

The manner of usage of a clock with a crossbar would require an accurate east-west alignment. A cardinal alignment is not a simple undertaking and certainly could not be produced by the clock instantaneously. This implies that an external reference must be available for the instrument to be used. Moreover, the crossbarclock would also have to be aligned correctly in the other two axes of rotation in order to achieve any consistency and in is Provision for in to the the use. crossbar accuracy adjustment rotation about axis parallel made by the accepted addition of a plumb bob and a reference line, but no method of for the third axis of rotation (along the length of the base rod) is either evident or reference by Borchardt by important the Bruins. This is third as or as proposed axis of rotation equally for the accuracy of the crossbar clock, for if the shadow-casting edge of the two other base is inclined, directly length this the rod, the thrown affects onto crossbar of shadow hours for the around midday. especially We seethat the crossbarclock is by nature a stationary instrument, and would have to be set line datum A datum line be time. turned in to and at noon, consulted situ at any up on a in by for in be datum (such a slot a accuracy all three axes provided could account a might
If be the but in it for is bob this necessary. example) case would unclear why a plumb stone,

datum only provided an east-west reference (using, for example, a reference point on the horizon) the instrument would need not only the plumb line but also stability along the axis is the base by top long existing the of the on which not rod, provided placed crossbar of block. gnomon bob is inconsistent the The pattern of usageof crossbar clock with the well-attested plumb instruments. the the size of surviving and attachment
instruments idea instrument, the to of the In contrast a stationary the small size of surviving

142

and the dimensions which appear in the Osireion text describing the shadow clock suggest 151 by instrument intended be is borne This the the to that out was portable. also strongly by bob in the hole for block E3, the also and attaching a plumb cord gnomon provision of a hieroglyphs in Figure 19 which show plumb bobs suspended from the all the clocks and

by bob instrument be levelled The the observer that to the suggests plumb needed sundials.
quickly, easily, and often.

The evidence presented above, from consideration of written evidence, the place of the
device in Egyptian timekeeping, accuracy and usage, seems to weigh against the crossbar

hypothesis. This is in addition to the fact that no physical evidence for the crossbarsurvives have in holes for blocks, for the the top seen, other the we as of which, vertical except explanationsexist. Until further information about this instrument is discovered,it seemsthat the simplest, most L-shaped, is is portable, the theory that an shadow clock obvious and most easily supported location. instrument time which marks certain periods at any personal Alignment is

device towards the instrument by bob by levelling the the and pointing using plumb provided block 152 the indicated by The be the shadow of gnomon correct alignment will the sun does base not affect the the rod the falling exactly on scale, and slight rotation around axis of 17 in Figure instrument. This as described in the Osireion text passage the function of the
"then the shadowof the sun will be aligned on this scat".

for 'rule' making a between the The simple ratios markings on the scale suggesta convenient '53 for hours. than an accuratemeasure seasonal clock, rather
lsl Macnaughton 'The use of the shadow clock of Seti I', Isler `The gnomon in Egyptian antiquity' L'Orientation Astronomique discount the crossbar but present only short arguments. ist A description of the method is given in Isler `The gnomon in Egyptian antiquity'. is in in Egypt Mesopotamia') `accuracy' clocks 153 shadow and Fermor (`Timing the sun agrees that looking for how the hour to in the is ask it but the 1: 2: 3: 4: 5 marks why are necessary futile not concerning ratio says v Zaba and

143

With no attachments but the plumb bob, the markings on the scale do not measure seasonal hours. Although seasonally-adjusted hours during the night were being measured by waterhours day hours during New Kingdom, the the that the night we cannot confirm and clocks being dawn The the time of same and sunset were as nature. periods around were perceived

in formally be divided have been to transitional to periods and seem not clearly considered New Kingdom times. The shadow clock was possibly the only timekeeping instrument
it is during New Kingdom have the to so secular as well as sacred applications, extant

defined by that for than time the the that purposes other shadow clock was used possible defined by star clocks and water-clocks. Finally, the concept of standardisationof time was
Kingdom. devices in for New timekeeping the a recognised goal not probably

The hypothesis of the crossbar is therefore inconsistent with what is now known about Egyptian timekeeping.

Timekeeping Properties of Sun-aligned Instruments

The L-shapedshadow clock, in contrast to the crossbarclock, doesnot measuretime periods find is It to `hours' easy that are related to the of the present time or of classical antiquity. the lengths of these time periods. We need to know the latitude at which the observations to the hour height the relative the the the marks of took place, gnomon and position of gnomon.
is 154 height, from instrument Berlin g, later find the Museum, the that gnomon For the we

is 191 furthest from mm, distance Li block the vertical mark to the centre of the 28 mm, the is L5 39 La block L2 and 127 76 to the L3 mm, from the vertical next mark mm, = mm, = =

their by since timings to the they found expression, then correct which Egyptians came a close mathematical false Perhaps formula is from the conception of but based some correct. not not empirically timings are stems ' the the path of sun. and geometry celestial

154 Dimensionsare taken from the replica in the Science Museum, London.

144

13 mm. The altitude of the sun for eachhour mark is given by

tan alt; =,g L. where i=1... 5. From this, with the latitude of the observer, , and the declination of the

hour between 8, find hence interval the H;, the the the time can of we angle sun, and sun,
shadow crossing successive marks. S H; 4s sin alt; sin sin cos = cos cos 4i

This analysis, using E3, raises three issues about the sun-pointing shadow clock theory. Firstly, we find that for latitudes within New Kingdom Egypt, the mark nearestthe vertical block will not be reachedby the shadow of the sun for more than half the year becausethe implications this high does While the of not attain a enough altitude. considering sun finding, it must be noted that this behaviour may not have been exhibited by all shadow for the between, The difference few, but example, the time surviving examples are clocks. two Berlin instruments indicates that the surviving shadow clocks representonly a small
sample of the occurrence of this type of timekeeping method. It is likely that practical,

been have preserved. working exampleswere unadorned,wooden objects which would not We therefore cannot statewith certainty that the Berlin instrumentswere madewith the same have textual We bear in that we although must also mind precision as working examples. have E2 3: 4: 5 between 1: 2: for hour might the the which ratio marks, the portion of evidence discussion for height `rule' is in of the the current gnomon missing, us resulting given gnomon shapeand usage. It is, however, easy to find the results of the equations above for a variety of gnomon
heights, and gain some insight into the nature of the time periods that the L-shaped shadow

145

clock would produce. The results for E3 are shown in Figure 20.
250

+ 200
U, d c c Id V 0 E c
t rn c d J

150

+
100

+ 50 + +
0 Ist 2nd 3rd 4th Noon Hour 6th

+ +
7th 8th 9th

Figure 20: Time periods measured by a sun-aligned shadow clock on the summer solstice at Luxor

The secondissue raised in this analysis is the question of what happensaround noon. The (for 10 threw time a shadow clock no shadow clock at noon, and marked periods crossbar by furthest hour first five from the the time mark the crossing of marks) period ranged with fifth the in to the the on the shadow morning until shadow crossed the next mark, and so it block from the the until the closest mark towards shadow moved vertical period when disappearedat noon. The five hours of the afternoon followed the same pattern after for instrument. This does hold for the clock, the sun-pointing shadow system not reversing days the (presumably) two the the nearest the mark of year will sun's shadow reach only on define On days, those five nine points noon. precisely would at a clock with marks gnomon in time: the end of the first period (dawn up to the time when the shadow crossedthe mark (noon), from fifth the the furthest gnomon), end of the second period, period of end ... ...

146

end of the ninth period / beginning of the tenth period. On days closer to the summer solstice than these special days, there would be an additional by time the passage of the shadow from the mark nearest the gnomon marked of period towards the gnomon, turning at noon and returning to this closest mark. On these days,

be defined. time of would regions eleven On days closer to the winter solstice, as we have already remarked, the noon shadow would not reach the mark closest to the gnomon. The remaining marks would define nine regions
of time.

The third issue concernsthe lack of uniformity in the length of the time periods defined by the instrument. It has already been noted that the time periods around midday presenteda long be the Moving hours through the while the year, around midday would problem. long. itself, if it hour from to existed, very would short range negligibly midday From a modem standpoint, these three issues combine to make the sun-pointing shadow clock a very crude timekeeping instrument. Although simple to make and use, the

instrument producesno regular `hours' and seemsto be limited to use in late spring to early least hours, then In a if the at contrast, crossbar only. seasonal clock not produces, autumn hours the entirely. time, passing of and eliminates the problem of noon more regular count of It is easy to see why the crossbar clock has become accepted. It representsthe Egyptian it have been the Even presented though evidence should as clock made and used. shadow before this analysis of hour lengths indicated strongly that there was no crossbar,nor an eastL-shaped the the the sun-pointing points raised above make alternative, west orientation, shadowclock, seemvery unattractive. One final argumentmust be added that may suffice to explain why irregular time periods are the indication than that the is sun-pointing acceptable theory an more shadow clock actually

147

crossbar hypothesis. The sloping sundial is known to exist from the Late Period onwards.

We have already mentioned this timekeeping instrument in connection with hieroglyphs of

shadow clocks, and have noted that the hieroglyphs provide circumstantial evidence that the it by from direct descendant L-shaped the the of shadow clock, evolving sloping sundial was involving falls. the the the reshaping of surface on which shadow a series of steps The sloping sundial has survived in greater numbers due to the use of the device as an device have, We for be this therefore, that there to sufficient examples offering. no question had no crossbar and was used by alignment with the sun with the help of a plumb bob, exactly the same manner of usage as the sun-pointing shadow clock. It is also uniformly

hour different by the that scales sloping sundial was seasonally adjusted using seven agreed

labelled with month names. An example of this type of instrument was found at Qantara155
in Figure 21. illustrated is and Other examples of sloping sundials share the same overall composition, but differ in details A height hour sloping the marks. angle of slope, of gnomon, and presentation of such as 22). dots (Figure Petrie has lines156 from the than museum slanting rather sundial by is the L-shaped into the attested the The survival of sundials shadow clock era of sloping fragmentary Tanis papyrus, and the Ptolemaic hieroglyphs indicate that more variants of both types of sundial existed than those which have survived to the present. We can analyse have in did we the shadow clocks, the same manner as we the surviving sloping sundials fact that for the for hour Again, the the slope of scales. we must allow only to correct

be may only representations of working clocks. surviving examples

Sundial'). Egyptian `An in (Scott it Metropolitan 156 Art the Museum feature A shares with a sloping sundial of Qantara five lists Petrie, examples as MMA, sloping sundials parts and or the of such sundials: Scott's article in Paris Collection lacking in Turin Hoffman from the and a sundial a gnomon originally well as a gnomon Qantara sundial except that it has hour dots between lines instead of on them. (See also Petrie the resembling ) 45-46, Zeitmessung. XXV-XXVI Borchardt Die pp. measures pls. Ancient weights and altgyptische and

issSee Cl6dat `Notes sur l'isthme de Suez' and Kuentz `Note sur un gnomon portatif greco-6gyptien' .

G F_ E

148

11 f1 mm

E
c

E)- 'TY 91 xA
A; Tr

TI-A-f r A '-E

rAx

r
E ' o nT

E E s w 3 3

Figure 21: Sloping sundial from Qantara

Figure 22: Two views of the sloping sundial in the Petrie Museum

El shadow clockL TextL E2Osireion E3 shadow clockL

L to S Variants
PetrieS I E5 Sloping L E4 Papyrus I ParisS

1550 BC NewKingdom

1050 3rd IP

700 LatePeriod

300 Ptolemaic

0 Roman AD

Figure 23: Schematic chronology of Egyptian sundials. `L' indicates a source relating to L-shaped shadow clocks, `S' indicates a source relating to sloping sundials

149

If the angle of slope of the hour scale is 0, and the lengths from the base of the gnomon to the hour marks along the hour scale are l; then the corresponding shadow lengths L; projected onto a horizontal hour scale are given by L; =gl; cos0 g- l; sin0

We can then usethe sameequationsas before to find the altitude and hour angle of the sun at each mark and hencethe time periods marked by the sloping sundial.

100

90

80
c E

70

60
CC

50

40 -*-Phaophi 3rd 4th 5th 6th Hour TT 7th 8th 9th Athyr/ Thoth T---r10th 11th

30 +--1st

--T--T--2nd

Figure 24: Hour Lengths produced by the sloping sundial E5

The results of this analysis for the sloping sundial found at Qantara (Figure 21) are presented

in Figure 24. The latitude usedis that of Qantara,30 52'. Let us comparethese results with our analysis of L-shaped shadow clocks. We find that the Qantaraclock also hastrouble with the time around midday for the majority of the year. The length the occurs with problem of the noon hour, and overall, we see no great same
length, although for the early in hour uniformity and late hours in the day, the variation in length is not too great. 150

We find that the characteristics of the sun-pointing L-shaped shadow clock are imitated in

the sloping shadow clock.

Some refinements have been added, such as the seasonal

adjustments and the reducing of the size of instrument needed: the surviving sloping
(around 120 in instrument length), `problems' the the small all mm sundials are yet of earlier

that we identified previously are all still in evidence.

To conclude, the timekeeping behaviour of the sloping sundial, with its uncontested form instrument later The the sun-pointing resembles shadow clock closely. of usage, manner and

its development from former. be The the to this to comes of a evidence support seems
timekeeping properties and from hieroglyphic signs which imply the development of the flat In bears into the or sloping one. contrast, curved crossbar no relation to the a clock scale

be It Egyptians, the a crossbar surely must questionable whether possessing sundial. sloping device which produced moderately even time periods and could be seasonally adjusted,
develop hours. the then sloping sundial which produced uneven would

The evidence weighs heavily against the crossbar hypothesis, and heavily towards a sunThe device, L-shaped sprang which one shadow clock. and crossbar clever was a pointing
be to impose is but a timekeeping, to seen from a need now clearly modem values on ancient

fiction.

151

PART II: COSMOGRAPHY

Egyptian religion contained a strong celestial component at each stage of its development. Religious literature shows us what aspectsof the physical and spiritual world particularly interestedthe priesthood at different times during history. The earliest preservedtexts, the Pyramid Texts, describe a stellar religion during the Old Kingdom and before, while solar during later Egyptian The the course of civilisation. existenceof the predominated aspects lunar calendar157 shows that interest in the moon occurred at a very early date. Any study of Egyptian religion must touch on this celestial aspect,and many have done so at is look It the to length. this scope not within of study at each occurrenceof a celestial great the New Kingdom in text the to as of matters religious and comment on such each reference include to the the that topics celestial sun, creation pharaoh myths of and other relationship Instead, the this the the subject study concentrates sky, on practical of perception references. that we label `nakedeye astronomy'. This how for have timekeeping We naked eye astronomy was used purposes. already seen is for the to scientific as observation of easy accept modem result mind quantitative Egyptians in it the that the the which star context clocks clear appear makes astronomy, yet For distinction this between `scientific religion. astronomy' and themselves perceived no between astronomy and religion has been drawn, in modem times, boundary the reason, between the method of timekeeping (the star clocks) and the purpose of timekeeping, and
for (their their movements, appearance and between the physical aspects of celestial objects Egyptians. to the their meaning spiritual example) and

dead has distinction The been has and caused a problem. a as studied religion However, this
157 Parker Calendars.

152

to some extent lost perception of a spiritual world. Understanding of this world is easiest
is the subject considered to be abstract. This is the approach that most studies of the when ls$ The have religion chosen. problem with this approach occurs when an area is found that has extant references, as is the case with celestial objects. For example, the sun still looks

it did in behaves the New Kingdom, the stars still appear as points of light, the year and as heliacal the such as rising of stars. These things have not changed, events still contains
by has humans. these the of perception events physical neither What has changed is the

spiritual perception, the understanding,the significance of such events.

It is the intention in this section to discuss how naked eye astronomy was incorporated into the Egyptian religion. To do this, the modern boundary between the scientific and the The task of distinguishing which material

blurred be or somewhat. moved religious will

in is linguistic by included be this the section aided and existence clear stylistic of should

differences between texts which deal with celestial objects as physical entities with spiritual componentsand texts which mention spiritual objects with celestial origins. habit between distinction Egyptian is function the two the The existenceof a categories a of interchangeably deity then the with and using name of concepts an assigned of spiritualising different itself. deities, The between their the chain of connections object the name of be functions them often can and objects, concepts and events associatedwith aspects,their benefit for the the researcher, without modem of an Egyptian set of contexts too complex follow, fill in to to to understand, or thought missing parts. processes, and

rss Only recently has work in the field of Egyptian religion begun to acknowledge that the religion need not be be abstract but can be accepted as a natural human vision of the world. In Daily Life of the to considered Egyptian Favard-Meeks Meeks Gods, present and religion in a refreshing and sympathetic manner. Egyptian As the is that respect and treated such, same acceptance with merit. a current major religion would The religion logic, irrelevance. is truth, to The are relegated consistency a and as presented as world picture such notions The book presents not only a study of the religion, but also an approach to Egyptian reality. valid spiritual for Egyptology. is all aspects of consideration of worthy that thought

153

Where a celestial object is concerned, the researchermust be prepared to combine the physical properties of that object with the spiritual function described by the texts. This has difficult, in general,astronomical aspectshave been neglected. and often proved
In this section, a study of the impact of naked eye astronomy on religion in the New Kingdom will be attempted. A new translation of a major astronomical religious text will be discussed, as will astronomical references in other New Kingdom texts. presented and Celestial diagrams which graphically portray the Egyptian view of the heavens will also be

discussed,together with their relationship to timekeeping and the information they contain
concerning the perception of the sky.

154

Section F: The Book of Nut

Introduction
The most important single source for astronomical mythology is contained in a series of

texts known as the Book of Nut. The texts have survived in four sources,all incomplete, two from the New Kingdom and two from around the secondcentury AD.
The Book contains three sections: a diagram with texts which we shall refer to as the Nut vignette, some text and a diagram dealing with hours and the shadow clock, and a piece of continuous prose called the dramatic text. The portion of the Book dealing with the shadow clock has already been discussed at some

length in an earlier section (Part I Section E), leaving the Nut vignette and the dramatic text to be discussedhere.

Sources F1 Osireion159
In relief on the pitched ceiling of the Osireion, or Cenotaph of Seti I at Abydos. The ceiling (Figure Nut 25), shadow clock texts (Figure 17) and the dramatic text the vignette contains (Figure 28) which appears already corrupted in its latter part.

IV'6 F2 Ramesses
Only the Nut vignette (Figure 27) is used, with some texts missing. F3 Carlsberg 1161

The Nut vignette and shadow clock texts are missing. The papyrus is written in the form of line by to the first texts Nut line the commentary the within part of the vignette and a dramatic text. F4 Carlsberg la162
A fragmentary papyrus containing material similar to F3, and presumed to have been written 163 by the same scribe.

159 Egyptian Astronomical Texts 1, Chapter II, Frankfort The Cenotaph of Seti I, and Murray The Osireion at Abydos. 160 Egyptian Astronomical Texts 1, ChapterII.
161 Lange and Neugebauer 'Papyrus Carlsberg no. 1' and Egyptian Astronomical Texts 1, Chapter II. 162 Lange and Neugebauer 'Papyrus Carlsberg no. 1' and Egyptian Astronomical Texts 1, Chapter II. 163 Egyptian Astronomical Texts 1 p. 37.

155

144

0 !fK. -"

--e

'A
V

a o

v;

r w .E

-0

d a
d pD

ZZ

156

*Z
.r b4

Pl

"

e -2 W
q w~.

t.,

157

., h
67 h

h v

in'r w O .C E" H .i .y V "w

E
l J

\\
_____

---_--

j . WC

.w iE w O d b

0 Q

t-

E V4

158

The Nut Vignette Translation

The Book of Nut has already been translated, and extensively annotated by Parker in Egyptian Astronomical Texts 1. The following new translations follow Parker's translation notes except where stated16a

The vignette consistsof graphical componentsand text labels which vary in length from one figures lines. The to major are the goddessNut, whose elongatedbody arches word several
from her hands and head to the right (north) along the top of the vignette near the apex of the left (south), her feet father is her Shu down to the to the and centrally placed god who ceiling his head. in Nut This its Heliopolitan depicted the above raising action, and context and is in Book is described Divine Cow, the the the the sky of creating where act of cosmology, back (sitting is in Nut the by Re the to the on separate of shape cow) who of a performed 165 from mankind . ageing god

The texts in the Nut vignette divide conveniently into four categories: 1) star calendar,decanlist, and decanlabels 2) short labels
3) daily journey of the sun 4) region beyond the sky have to been first in star the reference category which already mentioned with The texts `knmt D Parker's labelling, S `sbSsn, Neugebauer rrk, and system of is are wrrt', clocks, using W `Stw V, list, date ', Tl decan U T3 list, Horus lives. the the is It that and lives,'M with lb and. tw.
16'The texts have also been translatedby Allen (Genesisin Egypt) who presentsa commentary. Allen's study his texts the therefore the which and commentary of myths, aspects creation concentrates on is concernedwith he (p. 5) his In discussion the the the the of solstice universe. summer of path the of sun at to shape relate in fact farthest day. The farthest sets the that that rises north and sun sun sets on south states mistakenly his discussion invalidates the This sky the of subsequent areas error north-west of and south-east farthest north. in Fl) `south-eastern' it F3 by he follows `north-eastern' in than Dd, reading appears as which rather (from text between for the basis texts, the the the use correlation of vignette about his summer as a solstice comments and between the beginning the Osireion, summer the coincidence and of the civil year. solstice and the date of du The Piankoff la Mythos, livre de `Le Tutankhamun, Mayestre gyptische Der shrines 165 vache of Hornung ciel'. knmt'. life 166 `the of Or

159

lives, knmt lives' or `the life of. tw, the life of kamt' and X `... knmt'.

Three additional texts also belong to the first category. Kk and Ll seem to be labels which
date U6.167 to the set apply
Kk Ll Setting in life in the Duat Coming forth pure of evils

TZ is a label on Nut's body above the head of Shu and refers to a time when Sirius rose heliacally during the first month of Akhet, that is the first month of the civil year. This is ideal or original state. representativeof an
T2 The act in the first month of Akhet at the time of the going out of Sothis

The complete text of V, which has already been discussedin Section B is

V
It is between the star which is bom and the star which makes the first hour 20 decans, giving 29 in those living and working in heaven. One dies, another (lives) for the first decade. It is therefore between the star which is born and the star which is enclosed, 29 decades. Issued under the length, issued to heaven as stars.

Of that which is betweenthe star which makes the first hour and the star which is enclosedin the Duat: there are 9 stars

Horizon', Y `Sand', The second category comprises the five short labels B `Nekhbet', P `Eastern but damaged is (which Gg Jj Q `Western two the texts Horizon' and R `Evening', and and 1) in Carlsberg Osireion few in the version while only a words are preserved readable mostly the describe the of orientation goddess. which
Jj Gg Her western arm in the north-western side, [] south-eastern. Her headin the western horizon, her mouth in the west.

0 N, M, K, J, G, H, hand A, E, F, on Z Cc, the and Bb, right side and and on The texts Aa, belong the to Hh C, Ff, L, Dd, Ee, form third the category, while and left-hand side the below. detail be discussed in last These two categorieswill greater fourth category.

Sun the Concerning Texts

from familiar journey the the texts these presentaspectsof sun which are of The majority of 168 intended, from left, Reading we the to texts. vignette right probably as was other religious
1 87. Texts Astronomical p. 167 Egyptian The led Neugebauer la 1 in Carlsberg Parker texts the order. to treat and and a certain 168 The scribe of head direction the text, the the the Nut of the order of the and at writing vignette of the majority of position of

160

seethat the sun entersthe mouth of the goddessin the `first hour' and enters the Duat in the `secondhour of her pregnancy'. The sun continues to exist inside the body of the goddess
during the course of the night and stars accompany him. This is of interest because the motion of the stars is linked strongly to sun both here and in

the dramatic texts which follows. The stars are seen to move continuously, not just when they are visible to man, and follow the samegeneral motion as the sun.
The notion of purification occurring in the Duat is taken up again in the dramatic text where

the starsare purified in the Duat during their period of invisibility.

Aa The person of this god enters in her first hour of the ?evening. He is enspirited and beautified in the arms of his father Osiris. He is purified therein.

The personof this god sets in the Duat in her secondhour of pregnancy. Then the personof this god commandsthe Westernersand he makesplans in the Duat. Then the personof this god goes upon the earth again developing in the southernland. His powers becomegreat like in the first acts in his primaeval time.
Then the great god develops as Behedet.

When the personof this god sails to the limits, the wonder of the sky is her arms. He entersher in the night in the ?middle hour ... .... When he sails inside her thesestars are in his company.

Texts Bb and Cc contain the samethemes,the vocabulary, style and content of which mirror that of the dramatic text.
Bb The personof this god entersher mouth in the Duat ?open After, he sails inside her.
Just as these stars enter in his company, so they go in his company, they ?run to their quays. These stars are in his company.

Cc

Text Z describesan undergroundaspectof the Duat

Z The housesof pillars is the place wherein is Re.

The remaining texts in this category are located on or near the legs and feet of the goddess is The in form left twice the times: circle three a sun as shown vignette. pictorial the of on

leads me to believe that the texts should not be consideredto follow a strict order (being part described events be absorbedin the same way as an image) but have a general right-to-left sense. This image they should of an is reinforced by the notion that the text is primarily concerned with the stars and events during the night, so describing cyclical is the left. Also, right and end with the the text begin on sunrise sunset with on would be left from Carlsberg to to That would to right read the this against read of scribe cyclical motion. events, and in follow the opposite direction is possibly an indication of the quality of his texts the la 1 and should the text. of understanding

161

and once as a winged scarabflying up the leg of the goddess. The order of the texts is clearly from lower to upper. The texts fall into three groups: those
form Nut leg the one group while the other two are separated by the curved line of on

stretching from below the `sand' line up to Nut's knee. J, H, and K are to the right of this
line, E, F, and G are to the left and M, N, and 0 are on the leg.

The text J talks about the sun preparing to rise. The preparation starts in the ninth hour of the night169 as the sun grows in power. He reachesthe region msktt which is the gateway 170 between the Duat and the physical sky.
Whenever the person of this god goes out from the Duat, these stars go with him to Mesqetet Then he grows in front of Mesqet.

Then he is glorious and is embracedin the presenceof his father in a shapeon the east,the first time in the beginning. Then developsand ascendsto heavenin the (9th) hour of her contentment. Then he is strong and comes into being in the two lands

into flies (text in form H) the the After the preparation, up sun rises of a winged scarab and

the sky in strength (text K).

H K Then (he) has enteredas/into this scarab. Then he developsjust like he came into being the first time in the first land. Then his heart comesinto being, his strength comes into being. Then Geb seesthe offspring Re when he goes forth

Parker E G difficult Neugebauer line, texts to and the and are Below understand. curved Vwhich been has include involved `red'. In G both to to this references have taken reading Nut for Book the substitute of as a throughout used itself Crown Red for the as standing

is `red'. the This by Carlsberg 1 word having the writes suggested which meaning thus and in hieroglyph the original full that in the to the `red' red crown, referring dirt and explains being 1 Carlsberg `red'. In is E there be to taken mean with more confusion171 text, should

170Msktt has usually been translated as 'Milky Way'. Careful reading of funerary literature however, reveals Spell Texts Coffin (for frequently is with portals, opening, associated example, admittance etc. that the region 622 72) 259,336, Spells Chapter (Coffin Texts Dead and as as well with notions of crossing 789, Book of the The do Stars term Milky 58). is the Chapter Way to Dead cannot cross the so. seen never and sun Book of the horizon/gateway/region Way. Milky to likely a than the represent concept rather is far more 1 48-49. Texts Astronomical pp. 171 Egyptian

in the previous footnote that the text is concernedwith the night and 169 the made observation This reinforces describeseventsin order from sunsetto sunrise.

162

little help.
E G he swims in his reddening. He opens .... The red is after birth.

Text F is straightforward in comparison, with the sun finishing his purification in the Duat before returning to the sky. These three texts, by their position and by the content of F are
in final before Duat the the the of sun rising. acts concerned with
F He purifies inside the arms of his father Osiris. Then his father lives and he (Horus) is illuminated. He (Ilorus) is placed before him (Osiris).

Texts M and 0 on the leg of Nut are concerned with actions performed by the sun in the
course of rising
M 0 The personof this god goes from her hindquarters/extremities he opensthe s and eyesof his mother Nut. He himself, ... ... Then he flies to heaven.

is happens The deals to the N these style starswhen actions are performed. with what while in dramatic found the text that to equivalent
N Then entersthe earth thosewho have risen and are bom.

A final text dealing with the sun forms the left-hand border of the Nut vignette:
A This god is in her south-easternside behind Punt.

He is about to travel in front of the dawn.

Texts Concerning the Regions Beyond the Sky

the beyond body Nut of the the the descriptions are representative areas of The of goddess
Egyptian notions of `that which does not exist'. inhabitants of void. and of emptiness embodiments Even where there is nothing, there are Here, primarily in the triangular region

in L forms border see the text the top we the Nut's vignette, of which also and arms above darkness that is the associate we not only with characterised coldness and this region that the by dank, dark sun the but where places Space, moisture and marshiness of also with never shines.
L
its south, north, east, west limits are not known. Those are fixed in Nun as the inert things. There is no light of the soul there. Nothing is known about the land south, north, east, west by the gods and spirits. There is no light there

The upper part of this sky is pitch black,

therefore every place shadow/emptyheaven and earth is the ultimate Duat.

163

Dd

Total darkness, cold water of the sky of the gods, the place from which the birds come. These are from her north-western side, equal in her north-eastern side.

The Duat which opensis on her north side, Her extremities in the east,her headin the west.
Ee These birds' faces are as peoples', their form is as birds. One of them speaks to his other in the words of people. They come to forage and provision themselves in Egypt, They alight under the brightness of the sky. That is the reason they come in the form of birds. The nests which are in the cold water of heaven.

Ff

Text Hh consistsof two cartouches,one with chicks in (probably with the dual significance 172 birds), 'marsh' label the the and also a picture of nest of one empty. meaning of a
Text C is a difficult label appearing on the left-hand side near the picture of Nekhbet
C She appears ...

The Nut Vignette Texts in the Tomb of RamessesIV and in Papyri Carlsberg 1 and la
The layout of the two sourcesis exactly the same in that no text is moved in the Ramesses IV version, although many are omitted. The main differences between the two designsstem from the different proportions of the spacescovered, the Osireion ceiling being long and N1 IV Consequentially, the Ramesses tends to the the version squarer. narrow while the herself defines in is lost in F1 F2 together Nut the with above version, shape which beyond it to the the texts which referred sky. region contained poetic
All the graphic elements of 171are reproduced in F2 with the possible exception of the moon Shu Nut, differ, the wrist, the The arm, P. wears sun's of wings, and and styles vulture near in R4. Also in Jj F2. Hh the two bands cartouches near and are omitted and ankle display legs is in in horizon F2, of to the below spacious and very cramped contrast The area in is due F2. `horizon' being This to the Osireion. in curved the not so texts

is M the fact In Cc. in F2, `labels' together text the are omitted with repetitive All the simple be in for F2 cited also could no apparent reason, although space maybe text omitted only here.
Epigraphically, F1 uses 14 whereas F2 usually employs =. F2 also exhibits a general

246 Dictionary s9. p. 172 Faulkner

164

into cramming of signs groups, a symptom of the lack of space. Comparison with the text included in F3 must take into account the order in which the scribe

(which is in texts the the order of labels assignedto the texts by roughly mirrored presents Neugebauerand Parker) and the fact that some of the texts are said to come from a source 173 The first time the scribe mentions sf is in his commentary to the texts in the Nut called Sf be V, Texts Bb, Z, Aa, Cc, Gg, Dd, Ff, Jj, Ee, to seem part of sf are which and the vignette. dramatic text (first fourteen columns). We see that all these texts occur on the right-hand
(Text into V left-hand the the continues vignette side). of side Prior to beginning sf, the

A, C, dealt H, D, E, F, G, J, K, L (from the book sky174),and T2, all of had texts with scribe

left-hand the side of the vignette, except L which forms the top border of the on which occur in Shu's head T2 body Nut. the the occurs above which middle of and of vignette The texts omitted in F3 consist of the majority of the texts concerning the two decanlists and the star calendartext. An example date set from the calendar is however included F4 seems to include more date information which seems to relate to text U. As for the remaining labels it is in in F3 the the text, texts with all general shorter which are omitted categoriesof 0 N, M, information texts the two three and the orientation containing and short except be is to Hh, Text two the cartouches, expected a not graphical element so would omitted. F3. in text the of occur Table 44 summarisesthe texts contained in the sourcesF1, F2, and F3 in the four categories introduced at the start of the previous section: texts to do with decansand the star calendar, labels, texts concerning the sun and texts about the region beyond the sky. Blue entries indicate that the text is consideredto be part of the book sf. inclusion due how is Nut to spatial in the Book texts the or omission have of seen of We now

224. Dictionary Faulkner p. 173 Yesterday? 174Uplifting, exalting? The names of these two books seem to have to do with setting and rising. Could they the two vignette? for of the sections be names

165

in F3. in F2 to and prominence considerations Decansetc. F2 Fl D D S1 S2 S2 S3 S3 Ti Ti T2 T2 T3 T3 U U V V WI WI W2 x Kk Kk Ll Ll F3 D T2 V Labels F1 F2 B B P R Y G J' F3 G J' Sun F1 A E F G H J K M N 0 z Aa Bb Cc F2 E F G H J K N 0 Aa Bb F3 A E F G H J K z Aa Bb Cc Beyond F1 F2 C L L Dd Ee Ff Hh F3 C L Dd Ee Ff -

Table 44: Concordance between texts from the Book of Nut in the Osireion (Fl), the Tomb of Ramesses IV (F2) and in Papyrus Carlsberg 1 (F3)

The Dramatic Text

The text, a continuous piece of prose in a style usually termed a `dramatic text', appearsin forty-six columns, of which the first fourteen are in good enough condition to be read with here 28). A first is fourteen (Figure translation the new of columns presented certainty falls discussion The the text translation the of an analysis and a together with of content. The do in layout into thirteen to the text the verses, which not relate columns. of naturally in fourteenth in text The the the column at of a suitable portions point. translation stops from F3. brackets translations supplied are consideration of square It is interesting to note that the commentaryto the text supplied in F3 stops in the fourteenth he find book, that the the the scribe could calls sf. comment which no more of column with This is another indication that the version of the text the scribe was working from was no in Osireion. the the than version more complete

166

o J IIe iA v 11I

+ DD

t + n

1 11

d -y

10

oQ it R C05 O
''

Zll-lN
-A

% ,// /x i

F ffl
EL
L-t
I+ 1II
l11

E-3

1II

'1I

p a lls leo
T

nnnn nnn

NO '

/ / Q 111

o 111 7

/ ,/

IL
-PURIP
t1I

ar-

Lie
i

II

ry

I'( 1 III j4 ?I p
1. I4 of Ou

4-.

vq "eR `9n 'i I

VI I llel I

II

III

nnn

*4

1'1

, p t, C

/4/

7/a-

Figure 28: Dramatic Text from the Osireion after A. de Buck

(from Frankfort The Cenotaph of Seti I)

167

Translation
i When thesestarssail to the limits of the sky from her surface in the night they rise and can be seen. ii When they sail inside her in the day
they do not rise and cannot be seen.

iii Just as they enter in the company of this god so they go with him. iv When they sail with him upon the supportsof Shu, they rest in their places after His Personsetsin the western horizon. v They enter her mouth in the position of her head in the west.
She swallows them. vi Then Geb quarrels with Nut because he is angry because she is eating. Her name is called 'sow eating her piglets' because she eats them. vii Then her father Shu, he lifts her and raises her up to his head saying: viii `Be watchful, Geb. Don't quarrel with her because she eats your offspring. They will live,

They will go out from the place of her hind-quarters in the east each day just as she bears[Re]. Her nameis not called `mother of gods' until she gives birth to them. Not one falls from her. ix The one who wandersto earth, he dies then.
[He is purified and gives up his misfortunes to the] earth in 70 days. Her name is not called `releaser' until 7 days.

It goes from the water's surface and screeches to heaven from the sea in its form. The living stars fly, they go from the [Duat] and they fly up to the sky. ' [prince] became Geb Then of the gods. xi

The nameis not called 'living' until he is released. him pure Geb. They are pure and living. makes .... ... They will place their headsin the east. As one becomesdead, anotherwill live for a decade. [They `always' celebratethe first festival in the east. The headof one is placed,] one of them. Then their evils fall to the ground, so that the souls who have fallen embrace[the earth.] in lake begins fish......... life the The star as of a a x

Then Geb he commanded....placed their headsin the east again. ... Then Geb, he said to the gods: `Pick apart your heads!' Then Thoth commandedthat they meet their heads. They lived and their headswent out.
Burial happened for a period of time in the Duat because this is what the dead do. inside the The of the sky at night sailing spirits go xii

it happensthat they go to the limits of the sky in the day: they are thosewhich do not rise in sight there. living by the Those that seen as stars, are xiii
they go and rise in the sky in the hours of the night in beauty in the sky seeing its life. and sail

Discussion
by Earth-bound the the is observer motion of stars with an as perceived The text concerned to diurnal both the the referred annual and stars, deals motion of certain group of a with and `these itself stars'. sb3(w) n(y) as nn in the text

168

The first part of the text (i-v) records the observedgeneral behaviour of the stars during the drama (vi-xi) describing follows the the annual life cycle of a of night, a mythological course (xii-xiii) final both the the annual and diurnal cycles pair of verses summarise star, and
succinctly. The verb skdd is used for the general motion of a star over the sky. The verb is causative,

implying that the stars are not travelling under their own motive power, but are forced to
doubling is The by typical of a second tense verb of motion cause. external some move indicating an habitual action. The choice of the water imagery of the verb `to sail' with a boat determinative is very strongly connected with stars. As we shall see in a later section figures depicting the tableaux, spdt and sah and the superior planets are on astronomical

in in in barques, hour Book Day is the the the a and of sun shown at each shown standing barque with his mode of propulsion changing over the course of the day. The areatravelled by the starsis the liminal region of the sky `drw pt', formed by the skin or
Nut. `haw' the sky goddess of surface The goddess is frequently depicted wearing a blue

dresscoveredwith yellow stars,or as naked with the stars on her skin, as is the casehere. The contrasting use of the active and the passive for the verbs 12r`to rise' and m33`to see'
behaviour being ii links the to in the the seen or action and stars: shining effect of used verse is This be used again at the end of the section in verses xii-xiii. method to shining.

Verse ii mirrors i completing four pairings of words, skin/inside; day/night; rise/not rise; in by introduced the The this a same verb verb of skdd. continued use seen; seen/not believed to the that the not observable shows stars are was star of a motion situation where be continuous whether or not the star was visible. is `this introduces behaviour The the god', an appellation the stars of The third verse of sun. by the that sun correlation of two particular movements: the stars 5k linked of with being

169

him. The usual meanings of these verbs, `to enter' and `to go' the sun and with pr with
respectively, are replaced in astronomical contexts with `to set' and `to rise'. indicates the recurrent nature of the motion. verb of motion prr number of ways, `in the following' The doubled

M-bt can be translated in a

or `in the company' or `in the entourage'. In the present

indicating it is best behaviour loose thought of as an affiliation of positional with a context,

connection. Continuing the parallel with the sun, verse iv shows how, when the stars are high in the sky

175 Shu' `upon the supportsof at the end of the day, when the sun sets, the stars appearin the
is do `in has but that This the the out stars pointing not all rise sun set, are already after sky. their places' at sunset. During the night, the stars pursue their westerly motion, set in the western horizon and are (verse her Having dealt by Nut her to turns the text mouth. v). with skin, now swallowed

Geb, Nut's consort the earth-god, is seen as angry at Nut's apparent consumption of the be Their her, hence Geb's, to regarded as appear and offspring. presumably stars, which father Shu performs his separatingmotion at this point in the text, but it must be remembered is his day throughout the not the continuously action and night, so that stars are setting
is but also continuous. momentary

Shu explains in a long speech(versesviii-x) what happensto the stars who are swallowed. These starsfall into two categories. The majority of the stars, Geb is told, will rise again the `sow Nut diurnal the than be (the day of stars). cycle will rather once more a mother, next happens. follows However, this her the a when star of second category piglets', eating into Nut. ix) body (verse the the through than ground of going rather different path passing
l's See Book of the Dead Chapter 17 `As for that holy gate, it is the gate of the Supports of Shu. Otherwise Otherwise It is Duat. the the door through which my father Atum passedwhen he said., is of It the gate said. horizon of the sky.' From Faulkner The Egyptian Book of the Dead pl. 8. the to eastern proceeded

170

This process is called death, and the purification which follows takes 70 days. After this time, the star is `pure and living' and will rise again. The event `place their heads in the

describing heliacal is the rise of the star which marks the end of the period of phrase a east'
invisibility. Shu also notes the cyclical nature of these events, in that they occur to one star

after anothercontinuously. The speechcontinues in verse x by likening the star during parts of its life cycle to a fish.
This idea of fish usually occurs alongside tears and people and is due to the similar nature of

the words in the Egyptian language. Here, the star flies up from the water into heaven. Verse xi mentions headsagain in an obscure section which relates an interaction between Geb, who commands the stars to rise heliacally, and Thoth, hitherto unmentioned, who be just heads but This their to the the to can appear, star. rest of seems encouragenot likened to what happensafter the heliacal rise of the star. Over the subsequentdawns, the length of time that the star is visible increases. This verse is also vitally important becauseof the link it establishesbetween the period of invisibility of certain stars (70 days) and the customs surrounding death and burial. This ties
decans the that in with the conjecture were originally between the to period chosen mark

from hours development during their to burial, that the use mark death and and night was a function. this earlier Verses xii and xiii are a pair, summing up the diurnal and annual cycles. The starswhich are day inside during the goddess during the night are completely the and are in the sky high be the that those outside goddess are at night can seenas stars. invisible, while

Nut Book the of Cosmography of

Nut Book together describe the the to of of entire celestial attempt sections three The
the rising and setting stars. the and sun of mechanism

171

The text is unique in Egyptian literature becauseits combines the mythological creation
detailed description of the motion of celestial bodies. The text illustrates the with a scenario in direct human described the outside range of were phenomena perception physical way language is different from We Egypt. the that that which we see and style very used ancient

document. in scientific a would expect Discussion of the Book has led to some important information concerning the Egyptian
perception of the sky. Firstly, the cyclical nature of celestial events has been strongly emphasised. The text bases its discussion of the movement of the stars on the behaviour of the sun, which is cyclical on

is The between basis. text that the the daily the not shows sun similarity stars and a describes difference day between the the the the sidereal effect of solar and and complete, day, which is that the starsfollow a secondcycle on a yearly scale.
Secondly, we see that the portion of the sky travelled by the sun and the rising and setting

(the Geb image by Nut. being The Nut the is goddess raised above of stars represented Earth) by Shu (air) is a recurrent and common motif drawn from the creation tradition that
describes Atum as the father of Shu and the grandfather of Nut and Geb. the describe different in daily body the to Nut's and the sun of cycle used stages are Parts of

born, her her they the are In where mouth, sun where and stars set, vagina particular stars. her in Also daily travel they the the within her sail. sun stars cycle, and which over skin and becomes In have however, the the the annual cycle of star which set. stars body after they inside being the body of Nut, but instead goes to the earth and rises is invisible not seenas from the water. information in to the find terms about the references of specific motion stars of Thirdly, we invisibility the itself. that This and of star suggests period calendar strongly the seventy-day 172

`these stars' are the group of timekeeping stars known as decans. It is not clear whether all

stars which rose and set (that is all visible stars except the circumpolar group) were thought to exhibit this samebehaviour. Fourthly, we note that the Egyptian world picture did not only describe things which were visible to man, or state that things which are not visible are wholly spiritual. Instead, the stars which are not visible are still consideredand their motions described. The areaswhich
by beings behaviour to still exist and are populated visible not which similar are exhibit beings. visible existing Finally, the way the Nut vignette is arranged indicates how cyclical events in time are

dimensional two static media. Things that occur in a temporal sequenceare on portrayed depicted as lists, while time is also shown as flowing from the right to the left for the daily
body left flow from in line Nut from left the to the under of a right sand and under cycle and

to right for the annual cycle. The same object is shown in different positions in different (the disk, through time to space motion show and sun as a a scarab,and a winged aspects disk) and graphical elementsare used to separateareas concerning different aspectsof the
The (in for the text to the the serve as parts of example). case of scarab, situation and layout is fusion text, and comprehensive, complex and with a of graphics, arrangement

information far linear is than text more together show a capableof conveying. which

Astronomical Timekeeping Nut Book The and of

indication Nut of the extent of the astronomical knowledge possessed gives an The Book of is New Kingdom. during Contained have the the texts, seen, as we by the priesthood within the stars which is interesting in its selectivity. The stars the of movements knowledge of a
does the This link between the night-time stars and as sun. motion east-to-west an pursue by is the `in the the `in use the emphasised of daytime sun phrase company', m-lit, the

173

entourage'. The difference between the regularity of the sun's reappearance each dawn and

from the sky for a period during the year is noted and explained in a a star's disappearance
death concerning and purification. myth From the vocabulary contained within the dramatic

text that describes this portion of the annual cycle, we derive support for the theory

disappearance decanal linked that the the to the period earlier of stars was postulated between death and burial, and that the diagonal star clock was developed from a list of stars intended to mark the passage of a longer period of time, around seventy days.
The dramatic text states the time period of 70 days explicitly. this period of decanal invisibility was inflexible? Is it reasonable to assume that

The answer is clearly `no'. The reasons

for this stem from the practical and logical barrier to a fixed period being set between two has heliacal Practically, the observations, eye as naked rise and set of a star. near-horizon
dates been heliacal to timekeeping, the reference with stellar noted rise and set are of already

dependent on the nature of the object, the location of the observer, and the observing basic On level, be the more an even observer must physically capableof seeing conditions. averageeyesight) and must be able to identify the star correctly. the star (that is, he possesses
The logical reasons against 70 days being a fixed, immutable figure for the disappearance of

distribution is days' `70 include the the that a round of stars random and argument a star
figure that is likely to be representative rather than definitive.

in Egyptian find the the that religious culture, we Throughout notion a rough representation into Just be in transformed the a world. precise, working object spiritual the real world will bright, how idea illustrating for the a of evenly placed stars use as model a can we as diagonal star clock functions, with the understandingthat a `model' in our modern, scientific ideal the Egyptians simplified version or the of an reality, same apply would represents sense they contained the essential components, properties, and to models: physical reasoning 174

in the of object, so real configuration a sensethey were identical with the real object. It is with reluctance that a modem, scientific mind admits that, in the case of the 70 day
period of invisibility of the decanal stars, as in the case of many promising pieces of `data' in

the field of Egyptian astronomy, no equations can immediately be brought to bear on this piece of information which will for once and all unravel the mystery of identifying Egyptian constellations. Furthermore, this text demonstrates another feature of Egyptian thought concerning the happened Nothing without cause. The separation of Earth and sky was an act of a cosmos. the

The justified day The the the of sun and stars setting was caused night. and and each god. by humans These then ordered star was of a a god. events, which presumably reappearance for in in take the were significant granted, acts occurring a strange spiritual world as now
but being forever. happening once repeated of process The text of the Book of Nut gives a feeling of cyclical continuity which is slightly different from our notion of the continuity of time. There is a slight hesitation, a need for conflict

between the actors, a senseof renewal of the action at each repetition, and the danger of the
slightest imbalance occurring causing the whole process to be altered slightly or

catastrophically. These themesare continued in our next text, the Book of the Day.

175

Section G: The Book of the Day

Introduction
The Book of the Day is one of a pair of religious texts that first appear in the New Kingdom. Its companion text, the Book of the Night, is similar in overall plan and content to works such as the Book of Caverns and the Book of Gates, and contains nothing of an overtly is however, Day, The Book to the the which relevant present study. nature of observational
is of far more interest. As its name suggests, the Book describes the events surrounding the

hours. daylight in the sun Aspects of the text are of pertinence to the study of the daylight timekeeping methods, the important the the text sloping sundial, and and one and portion of clock provides shadow

information the circumpolar about region. unique 176 Sources

VI (KV9), ceiling of the SarcophagusHall G1 Tomb of Ramesses VI (KV9), ceilings of Corridors C and D and of Hall E G2 Tomb of Ramesses in in Taharqa Ramose177 (of text the the tomb time the Later versions of appear of of 178 Harwa (TT37, dynasty), dynasty), XXVth the tomb XXVIth of and in the Ptolemaic 179 Edfu, the graphical elements. without temple of

Layout of the Book of the Day

in is G1 Day Ramesses VIth. twice tomb the Book the contained within appears of The of body by Nut. daytime is back-to-back the This Nut defined the with a of goddess the space is Day Book Book the Night. The text the Nut the the over of arches of time which of night
176 Versions G1 and G2 appearin Piankoff Le Livre du Jour et de la Nuit. la de Nuit fn. 2. du Jour Livre 177 Le p. xi et Piankoff

for CXIV Deir Naville pl. ritual each hour of the day and the night (fragmentary). Hours 179 el-Bahrt See also in Gates divisions Amduat, but little have to and as these used nakedalso relationship works were the night of in dealt this study. with not are so and eye astronomy

178Tiradritti `Three years of research in the tomb of Harwa', Tiradritti). with (private communication

the text is too fragmentary for any analysis

176

arranged on five registers. The uppermost register and the two lowest registers contain

deities. The of secondregister contains barques representing eight hours of the processions
day. The middle register contains the bulk of the text. In G2, the pair of texts Book of the Day and Book of the Night both occur under a single Nut.

The Book of the Day is near the body of the goddessand contains four registers, the upper
lower by band figures from As in G1, the the the pair a of water. of majority pair separated

face towards Nut's head. The barquessail in the secondregister, with their texts in the third.
The first and fourth registers contain processions of deities. Towards the `dawn' end of the text, figures adoring the dawn face in the opposite direction, as they do in G1.

The entire text of the Book of the Day has previously beenpublished by Piankoff, who labels denoting Gi) (1 based 5 layout the to the and text number register on of version with a each from left in barques from birth its the The to the letter. the to west. sail right setting of sun a The majority of figures conform to this left to right orientation, as does the order of the Much hieroglyphic figures face the the to of the text text, right. although within columns of for labels figures, longer the text of while the portions of text relate to effects which consists day. hour texts In be the this during with of study, concerned shall we mainly each occur layout in G1. 29 the Figure the third secondand shows registers of and figures which appear
of these registers.

Hour Barques of the Book of the Day in the Sarcophagus Hall

in Book the text the hours of named of the Day are: The specifically

Text 5c: 2nd hour `bsrt kkw' Text A: `shtp. n=s'

Text 3c: 6th hour called `rrwt-ndrt' Text 3e: 7th hour `3wt-tb'

Text 3f: Text 3g: Text 3j: Text 3k:

8th hour `dw3-lzrt' 9th hour `nbt-mnh' 10th hour `wp .t bi3 sib kbh hpwt' 11th hour `nfrt dgl'

177

a 0

O CD

CC

a 0
L CC Q

. L7

ti

T i. e w om

fnn

b0

k,

178

Text 5c is on the lowest register, text A spans the three highest registers, while the remaining

texts are on the third register.

A total of fourteen barques are depicted in the Book of the Day in the Sarcophagus chamber.

The four smallest unmannedcraft probably are only included as an indication of `water' and occur in two pairs: one pair near Nut's mouth, the other in the water flanking the two barques under the winged scarab at the opposite end of the Book. These manned smallest
two small manned barques are different in character from the remaining eight barques. They face each other with the sun disk containing the image of a small child between them. On

board we see that they have the same configuration: two steering oars, a low shrine with
?leaves growing from it, a goddess (labelled Isis in the lower and Nephthys in the upper

barque) and a shroudedprow. The two uraei flanking the infant sun disk are both labelled
Mehen.

The larger barque (which we shall call the third barque) to the left of this pair is similar to the seven barques of the second register. It is a transformation of one of the two small barques. A pair of gods labelled hw and st3 have appearedin front of the steering oars, the leaveshave developedinto a larger plant, two mats and a bundle of hunting implementshave
been added, and there are now three figures in the prow, which is still covered. The first figure is labelled Shu and is performing the act of raising up Nut, who is above him giving

birth to the sun disk. The secondfigure appearsto be double. Four labels occur between Shu and this figure: gb, 3st,hr and hk3. Since ist seemsto be a misplaced label for one of the Nut, that to the the three we assume attending one of applies names remaining two goddesses in figure the prow and the other two to the doubled figure of the man. hare-headed The fourth barque is on the secondregister. The steering oars themselvesare labelled as hw for figures have these The by gods. no and a naos mats replaced and plant are and s13, 179

have in More hare-headed the the the prow, and people arrived sun-god. god, as containing labelled is clearly gb. we would expect, This process of transformation continues from

barque to barque along the second register. The prow of the barque is covered in the two first barques, large last barques the two the two large barques. This and as as well small lower in dimmer, in the the evening. the morning sun and represents In front of the fifth barque, a confrontation occurs. `The enemy', the serpent Apep, is him (text 2b) he is by Serpents knives (text 2c). The struggle at and spit attacked sighted. lasts three hours, from the sixth hour (text 3c) to the eighth (text 3f). After the eighth barque, in the ninth hour (text 3g) the deities enter the Field of Rushes180 (labels 2g and 20, which is described in texts 3h and 3i in the register below the Field. the of representation The tenth hour sees the barque manned once again, but it is now referred to as the `night barque' sktt (text 3j), and its prow is shown covered. Finally in the eleventh hour, the

described is `as barques 3k). The (text barque descends mooring the of of process uponthe west' identical. hours 11th 10th are the and

Battle with Apep Fieldof Rushes VIII Prowcovered X No barque Al West Sunset East Sunrise IX VII VI V IV III II No barque Prowcovered

Figure 30: Progress of the sun during the day, according to the Book of the Day

ISOA celestial region referenced frequently in funerary literature. Dead. Book the 72 the of of Chapter

For example, in the Coffin Texts Spell 30,

180

Counting back through the hours, the eleventh is representedby the seventh barque on the secondregister, the tenth hour by sixth barque and so on. The leftmost barque on the second
register should represent the fifth hour. In the next section, we shall see why this is probably not the case. The changing crew and cargo of the barque is not random. The bundle of hunting

implements, for example, disappearsafter the serpent is first attacked, becausethese are the in deities in the barque are renamed and redrawn as The the that used attack. are weapons their aspectsor functions change. 11w and stJ, the steering oars, are always present in their barques form figures labels. Throughout the of oars and of a pair sometimes as and pictorial hk3, hr-hknw, the register nbt-wf3, and gb are included in the crew. second of
The small figure of a woman is present in all the second register barques with uncovered

in her label barques She from hour the snryt occurs must all sixth onwards. prows, while therefore be a lookout, who is blinded by the covering or the dimnessit represents. A complete list of the occupantsof the barque from prow to stem is given below: Description/function Name (figure lookout by in final hours) the covered reed mats scryt long (for hare-headed ) with pole sounding or punting? gb
ist Isis, female figure

hr / hr-k3 nbt-wta mart wp-w3Wt hr-hknw hk3 the sun

3ms

a man lady of the boat, with horns and sun disk Maat personified by a feather a man hawk-headed(figure missing in final two barques) man, ?captain or pilot, with a arm raisedbehind him in naos
bundle of knives

s13
hw

hawk-headedpersonification of a steering oar

hawk-headed personification of a steering oar

in in Book Day Corridors C and D and Hall E the the Barques of Hour
The three barquessidewayson to the registers are very similar in disposition and contentsto largest directly the Shu G1, in to rearrangement of slight a with under allowing remain those

181

Nut, and the inclusion of ?mats and the bundle of knives in the two small barques. There are just in barques in G1. Of barques these, the the compared of second register, seven with nine the ninth, tenth, and eleventh hour (which all occur in one section of corridor) correlate hours) do barques (sixth, G1, the three those seventh, and eighth of as preceding exactly with in the next section of corridor which include the fight with Apep. The first barque on the

in G1, is like first barque the the towed, second and correlates register of second register barques therefore it. The two are at variance with the other version. next with

In general, the barques are more populous in G2. Hw and st3 are always personified and
is is the prow covered or not. shown whether srryt

The total crew list is as follows: (figure lookout by in hours) final the covered reed mats s'rryt hare-headed long (for ) with pole sounding or punting? gb
3st Isis, female figure

hr-k3 ? hr-k3 ? wp-w3wt

wp-w3wt?

a man a hawk-headedfigure a man

a man

nbt-wf3 hr-hknw
hk3

lady of the boat, with horns and sun disk hawk-headed(figure missing in final two barques)
man, ?captain or pilot, with a arm raised behind him

the sun s13 hyy

in naos two mats hawk-headedpersonification of a steering oar hawk-headedpersonification of a steering oar

in in knives large barque bundle the the to and others appears of at right-angles In addition, a barques the in the two and barque the register first and small second plants appear of the barque the to the barque others, and penultimate on the secondregister. right-angles large at It damage labels has G1, text than the suffered greater are often missing. and This version of hawk(one the difficult to the two who are ascertain extra crew members of is therefore `men'). three the figure of one and headed between discrepancy in barques two the the the versions. address numbers of We must now

182

There are two possibilities: either G2 inserts two extra barques (the second and third on the
second register) or G1 omits two. The latter is much more likely. The two small sideways

barques are now not part of the run of hours of the day, and the larger sideways barque becomes the representation of the second hour. This means that G1 omits the fourth and

fifth hour barques.

The Course of the Sun

We are now in a position to assessthe ideas behind the Book of the Day in relation to what

daylight know hours from our study of shadow clocks, given that the the about we already
Book the of occurrence of the Day dates from the New Kingdom, as does the earliest earliest

extant shadowclock. The Book of the Day implies the division of the day into twelve hours. The earliest hour of the morning and the last hour in the evening are treated differently in the both versionsof the Book, with no barque being depicted. The barques of the second and third hours, and the barquesof the tenth and eleventh hours have covered prows.
These features indicate that the journey of the sun through the sky was divided into distinct

hours last first day day The the between transitional of and were and night. periods phases. For the secondand third hours, the sun had not reachedfull strength. In the fourth and fifth hours the sun sails unobstructed,but during the sixth, seventh,and eighth hours the struggle in Field triumph land the The Apep the then resulting of over enemy. occurred, at party with Rushes,which marks the beginning of the evening, when the sun decreasesin strength, the the is the barque being is and embarkation covered of night again prepared. prow Apep takesplace around midday. The sun, having risen higher in the sky all battle The with down by the is its The in slowed struggle, to apparently and seems pause ascent. morning,
between the evening and the Field of Rushes, a land connected with the most relationship

183

pleasantand desirablepart of the Otherworld, is similar to our ideals of an `evening of life' of comfortable retirement.
This mythical interpretation fits with our physical perception of the daily journey of the sun:

the ambiguity betweenday and night at dawn which leads to three modem formal definitions
of `twilight', the veiling of the sun through the thickness of the atmosphere and clouds on the

horizon, the ascent of the sun, its similar altitude over the midday period, its descent and in dawn. to the symmetrical a manner compared sunset with events after approach

The difference between the Book of the Day and other texts dealing with the journey of the is The Book is Day describing during the tue observation. night of a process which sun
in daily life. is The deal That texts of aware other with regions. unseen mythical everybody

the Book of the Day uses similar ideas of barques,conflicts, and processionsof gods cannot between Book the the and the reality of the sun's apparentmotion across relationship obscure the sky.

The Circumpolar Group in the Book of the Day

One portion of the text of the Book of the Day, labelled 31 by Piankoff, relates a myth about the figures which comprise the circumpolar group. The text mixes familiar names with

from like Book for texts the words other some and, uses enigmatic writing concepts obscure 181 The new translation presentedbelow allows the text to fall into verseswhich and phrases by thematic headed sentences. are

Translation
The 4 northern spirits are these 4 gods of the entourage

in in the the the time of the great war, storm sky off ward who who seizethe prow-rope, and control the stern-ropein the barque of Re Imperishable Ones, the the of crew with

la du Jour de Nuit. Le Livre Piankoff in et 181 Drioton

184

the 4 northern starsof Meskhetiu. They shine in the midst of the sky, in the southern region of Orion,
they return to the western horizon. As for Meskhetiu, this foreleg of Seth,

it is in the northern sky tied to the two mooring posts of flint by a cord of gold, it belongsto Isis as a female hippo, its guardian, its Water-of-the-Godssurround as stars,the horizon dwellers. Re placed them in its (Meskhetiu's) company with Isis, saying: `Beware lest it goes to the southern sky, to the Water-of-the-Gods. Become Osiris who is in the company of Orion. '
The lords of the [Busirite] nome (Spirits-of-the-Flame are their names), their town is Crete, in a country (Peb is its name), in the God's Northern Land their forms are in the regions of the sky in the northern sea.

It is theirs. This country of theirs is the northern horizon.

Discussion
182 four ,The may refer to the four jackals in the morning side in both northern spirits' Book, in jackals G2. Four the the additionally and on evening side also appearat versions of Night. Book The `entourage' is that of the sun. the of of the sunriseend The `storm' and `great war' contain symbolism of Seth in the ngn-storm, and evoke the Horns-Seth struggle. However, in the connection of the bark of Re, the diurnal battle with Apep is also recalled. This struggle, containing elements of a battle, a sandbank, and
has Osiris in in battle Nedyt the the elements common with of overthrowing of enemies

mysteries. lines few to including the the the nautical nature next As well as references of sun's vehicle, has intrinsic The himself in Egyptian the that no sun motive power mythology. remind us it for his the to he be towed, sailed or make move, rowed, relies on entourage and boat must journey. Here, his for is said to include the circumpolar stars, the he crew needs resources belonging to Meskhetiu. These are probably the four bright stars that four stars specifically

is b3w, for written enigmatically. 182 spirits, The word

185

make up the shareof our Plough, or the shoulder of the Foreleg.

The next line begins with `they'. Since the circumpolar stars cannot possibly be in the

region of Orion, it is assumed that this `they' refers to the four gods of the entourage, which

in be in the therefore the `southern region'. the sun can same area as and move The phrase lzry-tp, here translated as `midst', does not seem to have any concrete
It has being the nuance of meaning. amongst, but no rigorous link to meaning astronomical `in the exact centre' nor any other such exact positioning. The second piece of positional

information mr rsy sah in or from the southern region of Orion (in G2: m 11 sah in or from the

is Orion), much easierto understand. of region RamessesVI reigned from 1151 to 1143 BC. At that time, Orion rose so that the

belt ideal horizon it 260. 1000 the the that of at azimuth was was set so and approximate
Since the angular extent of the figure of Orion is around 20, this means that the portion of

the constellation we think of as representingOrion's head rose and set nearly due east and west.
183

The path of Orion acrossthe sky and the southern horizon bound a southern segmentof the
despite being `northern' in first be to the line the The to called allied seem more spirits, sky.

sun than to the circumpolar group. The next section of the text is thematically labelled (using the particle Ir) as dealing with it is immediately From the used, Meskhetiu. vocabulary obvious that the myth of the
is the the the subject group of section. circumpolar purpose of

hippo female the serves as a guard to the Foreleg. The symbolism of the find that We

decreased, the 183 and that of the rising place increased,for times before 1143 BC setting place The azimuth of If 5 the work millennium. was time, the per written to of rising or rate was referring a an earlier at roughly Since be discussing literary behaviour in southerly. we this more are would general a work, places and setting does not affect our conclusions.

186

tethered Foreleg is very clearly linked to the behaviour of the asterism in the northern sky. Its revolution about a fixed point is very reminiscent of the behaviour of a tethered creature.
Again in this passage we see the cause-and-effect basis for the myth. The behaviour of the Foreleg is accounted for by explaining that the stars are physically prevented by order of the into from other parts of the sky. wandering sun-god The Foreleg itself is the object of

attention, the other membersof the circumpolar group are merely there to enforce Re's edict. The consequences of the Foreleg escapingfrom its region are presumably dire, resulting in a Seth between Osiris in the their celestial forms of the Foreleg old enemies and confrontation
and Orion. The other celestial figures therefore have the vital task of maintaining the

balanceby restraining the Foreleg. The mw-ntr(w), the Water of the Gods, appearsto be the liminal horizon area. The word 184 To the in 31 twice, whom once with male singular possessive pronoun attached. appears is Previously f, clear. s3(w)t not this pronoun refers
Meskhetiu itself. For mw-ntr(w) f

its guardian, is clearly the guardian of

it is possible that the suffix is part of the name that is

be different in Meskhetiu the variants of included or omitted could again writing, or possibly

dominant his Re `he' the texts appearance the or of many could make possessor, cited as here instead of in the next phrase. have is We information is text this the interesting of way positional reported. aspect Another to (the to others some relating physical region) areas circumpolar various regions mentioned, ). Even have (the God's land, Peb mythic etc. aspects or northern areaswhich are mythical it to descriptive the supposedly relate problematic a sky, phrases part of remains these when designated is described. the There how the sky were of no problem and parts to theorise
it is horizons, that these clear are setting areas and associated rising west with and east with
in 4a, in text also connection with the jackals. 184 Mw-n[r{w) also appears

187

is More precision not required in the present context. More problems occur respectively. `north' `south'. Although the terms theseare clear when applied to the flat terrain and with
in horizon, their to the use a celestial setting is not so clear. extending The modem interpretation of a celestial object being `northern' is that the object is closer to

the north celestial pole relative to another object, or that it is in the northern celestial
hemisphere. We are happy with the idea of two star maps labelled `northern sky' and

`southern sky' and would expect to see them overlap at the celestial equator. We are equally

happy with describing Orion as `north of' Sirius even though Orion currently lies on the horizon the the to of attitude objects and changes relative our as they cross equator celestial the sky. Being as literal as possible, and accepting the horizon as the most likely datum for a people `northern' developed `closer the to the skills, on geometrical means point north without horizon'. Dividing the apparentcelestial vault into two halves using a great circle from east
to west on the horizon passing through the zenith gives both north and south halves, and also `north ' `south '. terms In the the to of.... motion of relative and of..... reality meaning gives

(and the celestial pole not the north point on the horizon, since the north the stars around take that the the is the the means stars on region equator) edge of circumpolar at not observer turns at being most northern. have `southern `northern stars' stars' and usually been consideredto be alternative The terms designations for `Imperishable Stars' and `Unwearying Stars'. The circumpolar

in both the the horizon in the Stars the region of of northern are region and Imperishable default, be By the Stars, Unwearying the other all visible must stars, north celestial pole. called southern.
Additionally, decanal how have certain seen stars or asterisms were named hry and hry we

188

`upper' and `lower'. This labelling is again open to several interpretations: that `upper' first, or rose or was consideredto be in the upper part of a celestial meant more northerly, figure imagined to be describedby the pattern of certain stars. The adjective `southern' was (smd in decan names rsy). also used We have already noted the lack of an identification of a phrase meaning `north-south have /iry-tp `in The translated the midst' could conceivably we which as phrase meridian'.
l& but with our current understanding of vocabulary indicate `zenith', in astronomical

designate in to that the terms this say celestial positional only passage seem can contexts, we There is indication `east', `north', `south', than that specific points. no rather general areas

have in know the to the that `west' standard of precision celestial we meant context were and in Old Kingdom for the the terrestrially cardinal alignment of pyramids. was achievable However, the small amount of literature containing information of the kind we see in text 31 does not preclude the possibility that a vocabulary of common words was used with special limited describing for the the of purpose sky. meaning astronomical

bersetzung Iskn `zenith' following 328 Faulkner translates Sethe Kommentar Spell as Texts 185 und Coffin In 11338.

189

Section H: New Kingdom Celestial Diagrams

Introduction
During the New Kingdom, the decanal stars appear in a new context, the celestial diagram or

astronomical tableau,usually occurring on the ceilings of tombs and mortuary temples. The
diagrams themselves have no practical timekeeping purpose, presenting the decans simply as

but be larger decorative scheme involving the sky, of may part of part a a constituent

in forming the externaldecoration for a water-clock. Rarnesside tables or onecase starclock Sources'"
H1 Tomb of Senmut Ceiling of Tomb 353 at Deir el-Bahri. Time of Tuthmosis III. H2 Karnak water-clock From the reign of Amenhotep III. H3 Tomb of Seti I Ceiling of Hall K in KV17. H4 Temple of Seti I at Abydos Ceiling of SecondOsiris Hall. H5 Ramesseum Ceiling of the central aisle of the SecondHypostyle Hall. II at Abydos H6 Temple of Ramesses Ceiling of SecondOcto-style Hall. H7 Tomb of Merneptah Ceiling of Hall J of KV8. H8 Tomb of Tausret
damaged (Hall L) but apparently the same in layout and KV14, from much one Two ceilings (Hall J). better ceiling preserved content as the

I Ceiling IIabu Medinet H9

diagrams, the as well as plates or 186 sources, descriptions all of can be found in Egyptian Astronomical Brief Parker's Neugebauer is follows: source and numbering 10-38, 3 where system as Texts pp. H8=14and15,119=17,1110=18, H11=21, H1=2,112=3,113=6,114=7,115=8,116=11,117=13, 27,1116 26 28 24,1115 29. and and = = = 1112=23,1113=22,1114

190

Ceiling of the Sanctuaryof Osiris. H10 Medinet Habu Ceiling II

Probably from the Second Hypostyle Hall. H11 Tomb of RamessesVI Corridor A Ceiling from KV9.

VI Corridor B H12 Tomb of Ramesses Ceiling from KV9. VI Hall E South Aisle H13 Tomb of Ramesses Ceiling from KV9. VI Hall E North Aisle H14 Tomb of Ramesses Ceiling from KV9. VII H15 Tomb of Ramesses Contains a pair of celestial diagrams, one in the southern half, one in the northern half of Hall B in KV1. IX H16 Tomb of Ramesses Contains a pair of celestial diagrams, one in the southern half, one in the northern half of Corridor B in KV6.

Description of the Sources

Chronologically, the sourcesspan the period from 1480 to 1115 BC. H1 and H2 are from from H8 dynasty, H3 XIXth XXth dynasty, date from to the H9 H16 the to XVIIIth and the dynasty. Each celestial diagram contains a number of elements,as shown in Table 45.
We have already mentioned that certain celestial diagrams have layouts which are obviously between 5. H11, H12, in H13, Figure H14, H15, H16 the traced connection and related, and features: they each form or contain pairs of celestial diagrams, These sources share common having H14) H13 H12, the same two distinct decan lists, the each pair with (H11 with and the Ramesside circumpolar group, of and star an associated two set of arrangements same clock tables.

191

(T

fl, II!
uu I -IF- v

'! r"1.,

;/lam "1
s r

.04

Figure 31: Ceiling from the tomb of Senmut (from Dorman The tombs of Senenmut)

192

Similarly, it has been proved187 that H9 was copied directly from H5. It is also likely that the other celestial diagram from Medinet Habu (H10) which has survived only on a loose
block was similar in layout to the surviving pair. It is possible that there was another similar diagram on the ceiling of the Third Hypostyle hall in the Ramesseum that would also have

belonged to this group.188 The earliest example of this type of celestial diagram layout
occurs as the decoration of the exterior of the Karnak water-clock (H2). 189
W
M
1

cd
vo) 4)

CIS

Source

93

,.,

H1 Tomb of Senmut H2 Kamak Water-clock H3 Tomb of Seti I H4 Temple of Seti I at Ab dos H5 Ramesseum
H6 Temple of Ramesses II at Abydos

H7 Tomb of Memeptah
H8 Tomb of Tausret Two Similar Ceilings

H9 Medinet Habu Ceiling I H10 Medinet Habu Ceiling II

H11 Tomb of Ramesses VI Corridor A

VI Corridor B H12 Tomb of Ramesses VI Hall E South Aisle H13 Tomb of Ramesses VI Hall E North Aisle H14 Tomb of Ramesses VII H15 Tomb of Ramesses IX H16 Tomb of Ramesses
Table 45: Summary of elements contained within celestial diagrams

On the grounds of layout alone, H3, H7, and H8 form a group, as do H4 and H6. We

187Nims Ramesseum sources of Atedinet Nabu reliefs pp. 169-175. 188 Egyptian Astronomical Texts 3 p. 20.

189 The Ramesscumceiling (116) shows a baboon sitting on a djed-pillar in the centre of the lower register. layout (which 43) be this that (Calendars states can p. restored in the diagrams from Medinet Habu (H9 Parker (112) is Karnak the water-clock lunar a schematic 1110) the representation on and of calendar with an and by baboon figure. the The ceiling of Senmut (H1) does not include the `Thoth' represented intercalary month 'wheels' that twelve but figure, are usually accepted to representtwelve lunar months. Clagett shows baboon 22) Astronomy Parker's that Clocks, p. states and assertions concerning the intercalary lunar (Calendars, (including that Sirius was used as rive in inclusion these its sources for intercalation) and a marker are calendar unproved and untenable.

193

therefore have, after very cursory examination of the sources, five groups190 of celestial diagrams (including two groups for the ceilings 1111to H16 which contain two distinct celestial diagrams). The ailing of Senmut (H1) alone seemsto have no immediate parallel. The purpose of placing the diagrams into groups is to facilitate analysis and comparison of the sources. Table 46 summarises and labels the groups. Group A 112 115 119 1110
2nd Ramcsscum ccilin ?
Table 46: Celestial diagrams grouped by layout

Group B 113 117 118

Group C 1-14 116

Group D H11 H13 H15 (South) H16 (South)

Group E H12 H14 H15 (North) H16 (North)

The Circumpolar Group

191 is in It consists of a collection of The circumpolar group a recurring element all groups 192 horizon. In addition to the those figures representing starswhich always remain above the Hippo, Mooring Posts, figure (Ales), Foreleg Lion, hawk-headed figures the of and main labelled An which are always present,Neugebauerand Parker identify the human or animal figures Serkel, Croc, and Sak (who are given legends in some sources) and Man, Bird, and

19The ceiling of the tomb of Tharwas (No 232 at Dra' Abu el-Naga) dating from the XIXth or the XXth dynasty, contains a fragment of a celestial diagram of a layout different from any other sited in the list of least in (Egyptian feet the three figures' towards the the one of registers at ceiling. the middle of sources, with fig. 4. ) 24-26, 3 Texts pp. Astronomical 191 Zinner's early survey of circumpolar groups ('Die Sternbilder der alten Aegypter') includes referenceto two due his inability A3) lids different (Al identify lids to to dynasty and a lXth-Xth coffin such as coffin of the in in is due Foreleg these to to the the liis group' sources a'eitcumpolar appearance of reference type of text. the vertical strip. 192Some attempts at identifying the stars used in the Egyptian circumpolar group have been published: ) identifies An Cygnus; Bull ('An Ancient Egyptian astronomical with ('A constellations, of Wainwright pair bull Hippo between Ursa Major. Biegel the (ZurAstrognosie) and and makes the resemblance ceiling') notes does Locher ('Probable Identification of the Ancient Egyptian Circumpolar identifications as extensive has degree None theories these gained any of general acceptance,with all being generally of Constellations'). by Pogo has been ('Zum Problem der Identifikation') and Neugebauer Biegcl's criticised which ignored except 3 183 In. Tests 2). Davis ('Identifying Egyptian Constellations') also Astronomical p. (Eg)yuian Parker and identifications the decanal but of not only circumpolar of stars also certain other outlined a system her conclusionsare equally unsubstantiated. but constellations,

194

(who are never given names).194 SecondMan193

In the ceiling of Senmut, Bird and Second Man are omitted. In Group A wherever the

find that Man, Bird, and Second Man are omitted. Group B remains, we circumpolar group seems to omit only Sak. Very little of a circumpolar group is preserved in H6 of Group C

but Serket,Lion, and a teardrop-shaped Foreleg are visible. The sourcesof Group D contain Man, Group includes Second E but figures Sak, Serket, Bird, the while except omits a all and
Second Man.

A distinction can also be made between representationsof Foreleg. Groups A, C, E, and

Senmut all show Foreleg as a `teardrop' shape, while Group B and D sources have a

complete

195 ox

Table 47 summarisestheseattributes of the circumpolar groups.

Serket
Hl Senmut Group A

Sak

Croc

Man

Bird

Man 2

Fordeg Shape
Teardrop Teardrop

Group B Group C Group D Group E

Ox Teardrop Ox Teardrop

Table 47: Composition of the circumpolar group in celestial diagrams

Of the constituent figures of the circumpolar group, the Book of the Day passageexamined in Section G explains the mythology behind certain figures. The Hippo, Mooring Posts,and

193 In their list on pp. 183-184 of Egyptian Astronomical Texts 3, Neugebauer and Parker list one type of Man has but In four two Group D, the type two men, group positions. sources one possible of of circumpolar with Neugebauer and Parker say that the second is there mistakenly instead of Serket. H3 has Serket in addition to both types of Man, proving that the two men'are distinct.

194 Isler 'An Ancient Method of Finding and Extending Direction' postulatesthat an additional feature of the from his belief holds the This significance: vertical stems group pole, assertion spike, or spear. circumpolar that both gnomons and stars were instrumental in finding north. He particularly cites the vertical line in the half form from the spearused Ramesses the the VII to southern of ceiling group appears of which circumpolar by the god An (whom he incorrectly labels as Horus) and the vertical spike in the circumpolar group in the (which Pogo 'Astronomical first Ceiling-decoration Senmut Senmut' the to in Tomb was the of of ceiling believe held directional significance). Wilkinson 'New Kingdom astronomical paintings' disputes Isler's findings concerning the solar significance of the circumpolar group, but agreesthat a pole could have some use for stellar observation. 1" Tharwas haspreservedpart of Hippo, An, and Foreleg.

195

Foreleg, together with the chain that is sometimes depicted as linking two or more figures,
are explicitly mentioned. The surrounding `Water-of-the-Gods' either relates to the area

immediately adjacent to the circumpolar region or to the horizon area. In either case, we could possibly imagine a wet region designed to contain the Foreleg and its guardian. The

natural choice of inhabitants for such a region with such a purpose would be crocodiles. This is a possible explanation for the assortmentof this type of creaturewhich appearsin the
depictions of the circumpolar group. The nature or purpose of the other figures is various

196 unknown

Decan Lists
Complete lists or fragments of lists are present in all the celestial diagrams. Unlike the decans in the context of star clocks, the decansin a celestial diagram are the of occurrences deities by and numbersof starsor circles. accompanied Neugebauerand Parker compared and analysed extensively the decan lists contained within the New Kingdom sources labelled 2 to 29 in Egyptian Astronomical Texts 3. Table 48

shows which decansappearin each of the celestial diagrams dealt with in this chapter. The numbers in bracketsrefer to Neugebauerand Parker's classification of the sources,and at the base of each column, a label shows to which family of decans Neugebauer and Parker belong. to source each considered
"6 The use of figures, birds, animals, and objects to depict constellations is a familiar concept. We know that form few our own constellations of shapes that even vaguely represent the object after which they are very it is However, tempting to search for similarities between the figures of the Egyptian constellations and named. arrangements of stars. In particular, the obvious correlation between the Foreleg and the shape of The Plough, between imagineable the similarity our Orion pattern and the Egyptian Sahu figure (in particular the and depiction of the three large stars resembling Orion's belt in the ceiling of Senmut) leads us to suppose that the least behind some of the constellation names could be traceable. It has already been noted that at motivation have (see footnotes this 1 and 192). Three examples are Wells ('Sothis and pursued approach researchers other the Satet temple') who links our constellation Scorpius to a representation of a scorpion and an encircled star in Etoile 'Sirius Beaux et Jeune Horus' who suggests that the name and some attributes of spdt could the temple; formed by Sirius, Rigel, from triangle the shape and Betelgeuse; and Bradshaw The Imperishable Stars arise (as highly is theory part of a which suppositional) the interesting idea that constellations could who suggests have different attributes and hence different figures depending on their orientation in the sky as they rotate around the north celestial pole.

196

E ou b E
L

ca 0

0
4.. a.+

a2

00 "f

CC F-

197

The body of the table is divided horizontally into four parts. The first part contains the main decanlist, the secondthe three superior planets, the third triangle decans,and the fourth the
two inferior planets. Hatching indicates portions which never existed in the source, while shading represents information that has been lost. Concerning the planets, the ceiling of Senmut and the Karnak water-clock omit Mars, as do the other members of Neugebauer and Parker's `Senmut Main Group'. In all other cases, it

is fair to assumethat all five planetswere intended to be present. We know from our discussionof diagonal star clocks in Part I that we would expect triangle decansto appear in a complete list. The most complete triangle list in a celestial diagram
VI, in half in H12, Hall E Ramesses the the the tomb where southern of of ceiling of occurs in decans decans We the triangle that the triangle occur which all are shown. also see seven tpds, this the name, sources one additional are members of with exception of other group

in lists diagrams. from the triangle celestial occurs other which

We therefore have the following fragmentary triangle list: -twy st3tw nsrw s3pt nhs Ipds sbfsn ntr w3.

In the discussion of the diagonal star clocks, we saw that the composition of the triangle decanswas difficult to reconstruct due to the confusion that was seen in this area of the star A M labelled decans to Parker Neugebauer list triangle tables. and provided a of clock (omitting I) which are shown in Table 19.197 We can see immediately that none of the triangle decansfrom the celestial diagrams appearsin this list.
197 The similarly named pairs were distinguished by their writings, and the list is based primarily on that of decan being Al, triangle the with s3bw addedby the epagomenallists of A6 and A8. source

198

Three decansfrom the lists O and 1 appear in the triangle decan lists from the celestial
diagrams: fpds is the ninth decan 9 of O, sbsn is the tenth decan of O, and sfpt is the tenth

decanin 4). Neugebauerand Parker noted that the decan list in H1, the ceiling of Senmut, corresponded to the order of a diagonal star clock table starting with the decan tpy-r knmt as decan of the
first hour of I Akhet 1, continuing with the decans in the order they appeared on that date, decans hour the twelfth the then of of the night for each of the subsequent ten-day with and periods. 198 This provides a list of thirty-six ordinary decans which should then be followed

by eleven triangle decans. If the principle was extended to include the twelfth hour in the list. decan be twelfth the the triangle to the column, of would end epagomenal added 199 includes decans. H1 However, only six triangle Tpy-r knmt was also the decanfor the first hour of I Akhet 1 in the group of star clock tables 22) (Table list We how 4) the therefore (D. relates to the must closely assess containing decanlist of 111(first column of Table 48). The correlation is very clear. The inclusion in
lists between difference decan two list L is the the the up the synthesised of only major s3ppt to the area of confusion around the Orion decans. If Neugebauer and Parker's theory of the origin of the decan order in H1 is accepted, the

date in decan implies between 4) list H1 the and place of proximity and a of resemblance imply lists. for The in H1 the two either that we were omission may of s9pt construction decan in it include in decan 4), be triangle that the to order meaning of must an out wrong All, sources A13, and A14, or that s,pt was included or omitted in a minor adjustment of

198 Egyptian Astronomical Texts 1 p. 22 but also note Egyptian Astronomical Texts 3 p. 272 `Additions and Corrections to Volume 1' where due to their revised idea of the composition and order of the Orion decans, decan Astronomical first in Egyptian becomes triangle the last than (y) the decan rather Stn , as stated ordinary Texts I.
199 Egyptian Astronomical Texts 3 p. 118.

199

the decanlist at sometime. As was noted earlier, no triangle decans belonging to have survived in their correct The

identification decans is impossible. triangle of any so certainty with positions,

decans between the ordinary of H1 and 1 would lead us to believe that the relationship triangle list of 1 could be restoredfrom the triangle list of Hi. Initially, this approachseems to strengthenNeugebauerand Parker's assertionthat s9pt is an out of place triangle decanin All, despitethe additional occurrenceof the decan among the ordinary decansin the orderly A13 and A14, due to the inclusion of spt in the triangle list of Hi.
However, we can check the assumption that Hi can supply possible triangle decans for 1 by

decans first in twelve the the triangle the the that sky as should occur same of area noting in decans in list. the decans The decans 1 triangle therefore contain a of should ordinary first hrt has knmt hntt hrt. O the Similarly, to triangle tm3t the as of which region around hour of the night in I Akhet 1 would have triangle decansin the same region as the decans tm3thrt to `crew'. We seeimmediately that there are problems (Table 49).
0 assumedregion of triangle Triangle from Al etc. which should relate to A (D assumedregion of triangle Celestial diagrams triangle

Wt hrt
tm3t hrt

sind rsy
surd mhty

t -r knmt
knmt

Stw
s13tw

wst bk3t
f ds sb3sn hntt hrt hntt hrt tms n hntt

ntr d3 t
rmn h haw t -r sd fm -ht S Pd 3hwy

hr bpd n knmt
hat d3t hw d3t tmJt hrt tm3t hrt w. 3tf

nsrw
s9 pt nhs t ds sb.'sn ntr w39

kdt h -lb wf 3 "crew"

"ltnwy" V

haw
n_trd3 pt

bk3ti
s3`pt

s3bw phwy s3bw

hntt -r bntt hrt

Table 49: Triangle lists from star clocks and celestial diagrams

The decans of the triangle associatedwith O seem to be gathered from the region of the
implies This 1). (14-31 decan li? if is hrt the w, sah, and spdt tm3t spirits, counted as sheep,

200

that a later 'sheep' decan(/rry jrpd srt is the last 'sheep' member of O) marked the first hour decade in first the the of the year - an occurrencewhich would happen to hry bpd of night is between 2150 2020 BC date in \Ve 640. that the saw earlier n and so of the triangle srt n200'201 2660 BC The triangle at that date would be in the 2790 BC is between 6 and of region: lrry (:pd srt

31my tpy-r
3hty
b3wy kd lr3w rrt hry crt rnrn Irry min Iffy 3Gwt hrt iscrt tpy-r spd

Imy-lit 31r+y

A6, in Al, the that is composition compared with sources survives which acceptablewhen A7, and A8. (which however, diagrams, from Ipcis decans the are celestial sn contain The triangle and sb. from the same region of the sky from which we expected the O triangle decansto come) as In bntt. decan between bk3tf 4 tpy-r was an ordinary of and well as sSpt,which we proposed between bk3t bntt hrt. These three 6, somewhere sJpt wit and this place would terms of in D. O decans to and decans the we can relate with certainty ones ordinary only are triangle

20Neugebauerand Parker reasoned along similar lines using trmy-lt spd to find a date earlier than 2780 BC. 201McKim Malville et al in 'Megaliths and Neolithic astronomy in southern Egypt' publish evidence of Nabta. The in in the the near circle group present a stone who constructed circle were astronomical alignment 10,000 BP (before 'an from the years depression around caused present) until climatic Playa change Nabta [which] 4,800 BP desert have stimulated social differentiation and Nubian years may from the exodus Upper Egypt'. We therefore have an influx of astronomically in pre-dynastic cultural complexity kind in lower tradition Nile the of observational the some arriving with at around valley people knowledgeable development decan lists. have timekeeping the of evidence of earliest that we same time

201

Their closenessin position and the defective nature of the celestial diagrams' triangle list do determine to which, if either, of the two lists O or 4) the triangle lists to not allow us correspond,although the date of the diagrams' triangle list must be close to E) and (D. Summarising, we have three surviving types of decanlist: early diagonal star clock tables: year "" n main list, n- 640 or earlier triangle later diagonal star clock tables:
celestial diagrams:

year -n+
year -n+

160 or 200 main list, no preservedtriangle

160 or 200 main list, circa is to n+ 200 triangle

From this summary, we can see that although the decanswere used in celestial diagrams in the New Kingdom, there is no surviving evidence to suggest that the lists were still being updated at that pcriod ?02

The ordinary dccans contained within the celestial diagrams and shown in Table 48 in the discussed. have be 117 H10 and no decan names preserved, but from top section will now figures, and including the deities which accompany the decans in the celestial diagrams, Neugebauer and Parker placed the decan lists in these fourteen sources into five groups (indicated in the lowest row of Table 48): the Senmut Main Group and SubgroupsA and B, A Seti IA Subgroup the IC Seti of group. the group, and Considering for the moment the names of the decansonly, Table 48 shows us that sources H1,112,115, and 119(Senmut and Group A) have essentially the same decan list as the lists of Jill, 1112,1113,and 1114which are accompanied by a circumpolar group containing a

302 We therefore have evidencefor adjustmentor creation of decan lists from around 2790 - 2220 BC. Looking from New Kingdom, identified Neugebauer the Parker after lists monuments within decan and contained at (i. included in Senmut 'Senmut' their Main Group), type the e. to are two resembled which conform which nine Nekhtnebef), ten lists forming a group into which Neugebauerand (Petosiris lists D and Group the ct and leaving have thus (we this t list association, rejected a group composed entirely of post-New Parker placed from lists Greco-Roman the Period ten lists). decan of group a and four fragmentary 'Miscellaneous' Kingdom decans last be triangle identified in three can any groups the these In with certainty, variation and lists. none of lists allows no analysis of the date of construction of any of the lists, nor the decan first among choice of for list timekeeping adjusted any was any purpose or was even createdwith that goal. whether of determination

202

teardrop-shapedForeleg (Group D Part 2).

These are the lists from Neugebauer and

Parker's `Senmut family' of decan lists. Also, H3 and H8 (Group B) have essentially the
Group decan lists D Part 1, as which are from the `Seti I C' and `Seti IA Subgroup A' same

families. Across all the lists where decans are preserved, we see that each begins with tpy-r knmt. Barring individual omissions, there seemsto be a consensusthat lintw hrw is the twentyin is lists (H13, H15 (South), and H16 (South) of Group decan. This the unity achieved sixth bk3t1, by the insertion of a new decansb3w D) which join wU3tland bk3tl into one decanwB3t1
is in ! (ruts' After there the area of sah resulting in lists shorter than confusion in+', mlrw.

thirty-six decans in the Senmut families and longer than thirty-six decans in the other families. Neugebauerand Parker place 114and 116,the two lists from the Abydos temples, as part of the `Seti I C' family with 113and 118due to the writing of certain decan names. However, the circumpolar group which is thought to belong to H6 has a teardrop-shapedForeleg family `Senmut' lists. the to those accompanying similar Neugcbaucrand Parkeralso consideredthe number of stars or star circles appearingwith the 203 decanal names, the writing of the names, and the deities associatedwith each decan as factors when grouping the dccan lists. As they note, the number of stars assigned to a between be for information decan to this groups, allowing no opportunity varies particular 204 involved identifying The introduction the stars of of such additional used as a mcthod deities is indicative and stars information as numbers of of some underlying processesof In imagine the the case of unaware. number of stars, we can easily which we arc currently

201Full details of the writing of decanal names can be found in Egyptian Astronomical Texts 1 for decans used Astronomical Texts Egyptian 3 for those employed in celestial diagrams. The and diagonal clocks in star 3 which describes each decan list `group' Texts Astronomical details full Eg, also ptian of contains section of dccan, decanal deities. each the with as well associated as symbols star of the number

M4Ea ptian Astronomical Teas 3 p. 2.

203

this to be based on a practical method of improving decan recognition by indicating the in decanal the of stars group. This implies a development of observational practice number for which we have no other evidence, but which we accept is a logical aide to naked-eye in difference The the number of star symbols between decan list traditions is an astronomy. indication that the processof deciding how many starsbelonged to each decan was probably subject to human factors such as varying eyesight, personal preference, and tradition. The by is further confused the practise of sharing stars between decansand other graphical area considerations. Decanal deities presenta more difficult concept. It is an Egyptian habit to have not only lists deity but list. For example the to the a assigned also each member of of objects, of names Coptic names for Egyptian months derive from both the names of months and the names of deities. The one-to-one assignmentof deities to list members is usual and the practice was is Egyptian It highly thought. therefore throughout religious unusual, and widespread deities do decans list to that the assigned not within a single seemingly very significant, decans Deities for diagrams in Groups A E to system. of celestial to this one-to-one conform decanal deities The Horus, four Children 50. Table the in common most of are are shown Imscti, Ilapy, Kebchscnuf, assigned either individually, in pairs, or as a and Duamutcf, in Imseti Children Duamutef Hapy. Where the four. occur pairs, only occurs with group of is the most common Child. Apart from the Children, other decanal deities include Geb, Isis, ), Horus from the Ennead, Eye or Eyes of (sometimes Seth written n and Nephthys, Osiris, Geb, Of Sekhmet, limited 'Ba'. Hathor, Ba to these, Sckhmet, l lathor, and and are h ores, belonging Groups in lists B, C, dccans to E. is Osiris five first and exclusively four of the Isis, is dccans Orion although associated while with spdt as we would expect, linked with decans. for deity other certain the also

204

Group I) Decans q, )'-r knurl knm! hi-' J li ,J kmn/ /7lt dir Pllwy d, t fill, 1/lit lln, r lrrt lr. r, Jy bk It l, .chin, ml, n' rpy-e Iml1 luvt l, r! hlltt In'/
((17.5 17 17,711

Group A, HI Decans t/,)'-r knlnt kmr Illy h! 1d kllml Ir lt dlr p(lu'y di! (1n(t litt tin lt /7,1 u/ Ili hAlii (py-' 1,nrl lit/ I 1777111/! 7
r11IS 71 lliltr

Group 1: Deities Geb Ba Sekhmet Isis Hathor Children of Hours Children of [torus Duamutef, Kebehsenuf Duanlulef, Kebehsenuf fiapy, Duamutcf
[torus

L Groups B and C
l)ccans ipv-` knnnt kllnlt Illy l,prl knnu It it cl. l pln+'y d(t timt 111t timt! 1111 u'"i; /y Ilklty I/I)'-` lll, lr
h711111

Deities Hapy, Imseti Isis Children of Horns Children of Horus Children of I torus Children of Horus Duaniulef Duamulef Duamutef Duamutef, Hapy Horus (u'.C)
N'. (

Deities Hapy, Imseti Isis Duarnutcf Children of Horus Duamutcf Duamutef Hapy Horus Set
Horlls

Deities Gcb Ba Sekhmet Isis Ilathor Irnseti, Hapy Kcbchsenuf Duamutef Duamutef, Kehehsenuf, Duamulcf, Hapy
HORIS

Set
Horus

hntt 111'1
! Ill., ' it 117dt

Set
Ho rus

cl/Ili Jnnvy lily-ih u'i, i'smw . knrnn' Illy-1' sill(/

Isis, Nephthys r n". t l+". Children of Horus Hours

ctpli lnn, y 1,7y-ib wit 01111" s. k/uu' 117y-r sn,d

Isis, Nephlhys Set Set Children of Horus [torus

Isis, Nephthys Ir.t u.t Children of Hours I torus

li lnm'y .cy, Irr)'-il ,r/i tsnnr . knn, ' q,y-r. cmd

Isis, Nephlhys Set Itorus Imsc tl, I1apy' Duamutcf, Kcbchsenuf Itorus

.c/)l
571 .

Hapy
Isis

Sill(]
lit

Hapy
Isis

Hapy
Isis

surd
sit

I lafly
Isis

.c/n'y sit ln'y h/ul. c,! l1/y-1, >Jn+y /nl"y h/ny lrn! (w)Ill n' 111 y-ih hilt(w) )n/(w) ! kc/ !u Ir

Duamutef Kebehsenuf Duarnutef Duamutcf, Kebchsenuf Imscti, Ilapy Hours Horus Horus Horus

sll+y sit !n _vlip(l srt ll, y-r >h(u_y') I/, n_l' hn, l, I(1,

Duamutef Kebchscnuf Duamutef Duamutef, Kebehsenuf Hapy, Imseti Children of Hours

Duamutef, Kebehsenuf Kebehsenuf Duamutcf, Kebehsenuf Duamutef, Kebehsenuf llapy, Inlseti Ito rus Horus

sr! c "1+'y l, i-y 111'al. cr'l t/?v-r (bury

Duamulef, Kehchsenuf Kebchscnuf Duanlutef, Kehehscnuf Duamutef, Kebehsenuf mussti, I lapy

hnr(n')lu'u' /uy-ih lnrt(l+') Jnrt(w) lu lr W

Horus I-torus Hours Imseti, Hapy, Kebehsenuf, Duarnutef

l, nnr) Irnr W Hapy

Children of Floras

s/n,)' k,l
h'", irr
IN7, ih .C

Hapy, Kebehsenuf
Horus Osiris
Hors

kll clnY .
hlu' 'II h?N.rm lu l' rinn nnn stl, 'h sl /I Ir, -rr -rn, v 01 a,

Kebehsenuf
Children of Itonls Gyc of I lotus Children of { I, lms Children of Itorus Eye of Hones Osiris Isis --

Ilapy, Kebehsenuf
Children of Hours Frye of Ihlrus 'hildren of I tones ;1 ('Ili](Ircn of I torus O. iris Isis

Ad ci++'v
hil+' cr1
111'/1 Sill

Kct)ellsenuf, Flatly
In, seii, Hapy, Duamulcf, Kcbchsenuf Eye of Horus
Itorus

nnn in. ) .c; lr /nsdr slh ??? III illy c /II . cill c; h spill
/ .c/;

Eye of Horns Eye of 1lonls Eye of I torus Eye of Horns Osiris Isis

rnu, Illy m.cd/'. cl!, Imn Ivy s; l, rs? h st(, ps/i lu=r ;,r-Itl, '_r

Eye of Horus Children of Horus Eye of Horns Eye of [torus Osiris Isis

l-uy1. -N, u In 431-pt

PN'y

"tut

Hapy,

DIIa111lI1C'I

l; lp\',

I)uallllilef

D118111ulct,

Ilapy

ml, w
/r, c/'N'

Imseti
Eves

7/V/ll

Il)lsl'll

J I llltil' tl

-----

Il, '1'11'

I111sd'll

l. cpl

'l'r

F.yc t

Ikuus

f?1'c of Ihxus

2sl>l

I. ycso1

limas

:.
[torus

il ".r.c
l,., sn Hurus

Holus
Hollis

Itorus
Ito]us

I)uan)nlci ch"K

? III n'ls _
dr

Du, nlulef

Dammniet

rr

n ;.c

Dualnute!

41r .5L,

Table 50: Deities of the decans

205

These deities are clearly not randomly assigned, but the thinking behind their distribution is Neugebauer Parker treat them as objects of curiosity, without remarking that and clear. not

kind logical of purposeor processmust have causedthe associations. We have seenin some the Book of the Day how an initially random-looking crew of figures in a procession of barques, on closer inspection proved to be a carefully planned illustration of the transformationof attributes of the retinue of the sun, some having clear meaning, such as the
personifications of the steering oars, while others remain obscure to us but are nonetheless

all part of an ordered scheme. In view of the observational origins of the decanlists, and the inclusion in the celestial diagrams of new observational data in the form of the number of
stars (which we accept as an attempt at recording attributes of the decans), it appears likely

that the decanaldeities are not simply spiritual metaphorsor mythical personifications of the decans,but present some kind of information for which we at present unable to supply a
meaning 205

The Origin of Celestial Diagrams

We have already discussedthe layout of the Nut vignette containedwithin the Book of Nut in the previous section. We saw how the diagram displayed temporal and positional

information within a two-dimensional space, and how the cyclical elements of the motions of

celestial bodies were treated. The archedfigure of Nut occurs in far more texts than those we have examined in this study. She is the goddessof the sky, and is often present on ceilings and the inner surface of coffin 206'207 lids A motif present on temple and tomb ceilings from the Old and sarcophagus
205It is interesting to note that the gods of the Ennead plus Horus and his four sons are associated with the fourteen fingers of a half royal cubit (for an illustrated example in Mathematics Gillings the fig. 22.1 of see time of the Pharaohs). The other half of the cubit contains fingers associated with some of the circumpolar deities. 206Kozloff `Star gazing in ancient Egypt' likens the stretched out form of the goddess to the `swimming girl' type of spoon. She also however naively attacks Neugebauer (i. e. for 'looking the Parker sky' wrong at and

206

Kingdom through to the latest Ptolemaic temples was the repeating design of five-pointed
(often stars with a central red circle) on a dark blue background. yellow These designs,

while not individually forming part of this study, indicate the ancient basis and significance of the association of astronomical images and texts with the sky. In each case, the

decorationnot only depicts but also representsand takes on the attributes of the sky itself, in 208 far a relationship more spiritual and significant than any modem star chart. Egyptian formal architecture is full of symbolism, mimicry of nature, and journeys through time and space. Decoration forms an integral element of the overall design, with surfaces in keep have to occurring occurred, will occur, and must used show objects and eventswhich
a cyclical eternity. The repeating star pattern indicated that the ceiling was the sky of the

temple.
By the beginning of the New Kingdom, the repeating star pattern as an indication of the sky

improved, or specialised, into the celestial diagram in certain mortuary monuments. was This transformation seemslogically to be a development of the practice of using a diagonal 500 inner lids, but the table years of around a gap on surface of wooden coffin star clock between the last surviving diagonal star clock on a coffin and the first surviving celestial diagram in a tomb leavesa gap in our understandingof what sort of developmentoccurred. One sourcewhich indicates that the link between diagonal star clocks and celestial diagrams Asyut from Heny209 is this the which the missing period earlier end of coffin of occurred at

had in fact have they that they since no the they nowhere explicitly state considered effects of precession) intention of identifying Egyptian celestial objects, their analysis did not rely on precession. Her own theory in Milky Way for Milky Way be the Nut based the to her find desire that to represents seems a metaphor on Egyptian texts. She seems unaware that there are many other candidates than the sky goddess herself, and clearly misunderstands the relationship between Nut and Geb. 208 Describing the decoration of coffins, Willems (Chests of life p. 242) states that `the coffin is turned into a microcosm of the universe'. A similar process is of course at work in the decoration of temples.

201 For a discussionof the relationship of Nut to the coffin, seeWillems Chestsof life pp. 133-136.

20 Egyptian Astronomical Texts3 pp. 8-10.

207

probably datesfrom the Xlth dynasty. This coffin (now perished) is therefore contemporary
with the later diagonal star clocks. One fragment210 contained readable labels for the figures

of the circumpolar group and the namesof some circumpolar deities. It is probable, from the
inclusion of the names of decanal deities on another fragment, that the decans were also present. Some traces of the name of Jupiter were also visible. The four elements decans, planets, circumpolar figures, and circumpolar deities, are the usual constituents of a celestial diagram, and were therefore associated at least as far back as

2000 B C. Our next surviving source,H1 (the ceiling of the tomb of Senmut), setsa pattern for celestial diagrams to the end of the New Kingdom and beyond. The layout is broadly the same throughout the New Kingdom, and there continues to be a strong link with timekeeping. In the case of H1, H2, H5, H9, and H10, the lunar months or lunar feasts were included. H2
occurs on a timekeeping instrument and Hll to H16 occur in association with the Ramesside

star clock tables. This link shows that the sky was, during this period in Egyptian history, of time. consideredto be both the source and the indication of the passage
The figure of Nut appears, as we have seen, in the Book of Nut and in the Book of the Day Night. Book the of and The figure also occurs as a frame for parts of the celestial diagrams Elements of the Book of the Night occur in the

in H15, the tomb of Ramesses VII211

212 in diagram Ramesses IX These links tie together the sources we the tomb of celestial

have discussedinto a group of closely related texts. Each text describesor illustrates some part of the Egyptian sky. Thesetexts lead us to formulate a model of the Egyptian perception of the movementsof the bodies that we seetoday. celestial same
2roBattiscombe Gunn "The Coffins of Heny". 211 Egyptian Astronomical Texts 3 p. 34. 212 Egyptian Astronomical Texts 3 p. 37.

208

Division of the Sky

The celestial diagrams and astronomical texts that we have considered so far show that certain parts of the sky were given greater prominence than others. The diagrams contain little in the way of referencefor the observer,such as an indication of the horizon, and in this sense resemble modern star charts rather than instruments such as planispheres and planetariumswhich show star positions relative to the observer's horizon at a particular time and location. The elementswhich are included are divided strictly into groups. Differences in layout do not alter the individual nature of the groups. We have the circumpolar figures and the accompanying circumpolar deities as one unit, the main decans,five planets, and triangle decansas anotherunit, and the lunar months and Ramessidestar clock tables as two optional units. These component parts never intermingle, and within each unit orders are strictly preserved: the circumpolar deities (barring omissions) always appearin the same order213 The only exception is the location of the label or figure of the goddessIsis which is always located behind the Hippo.214 There is some confusion whether the label was originally intended to belong to the Hippo215 implying that Isis as a member of the circumpolar deities is an intrusion. Equally rigorous in their arrangementare the two setsof decansand the two decans, decans followed triangle by then and the superior planets, are setsof planets: main finally the inferior planets. The internal order of the decan lists is disturbed by the practice is There in decans decan later the two the earlier. one column with placed above of putting is in in Orion decan lists216 the triangle the never complete. area of some and also confusion

213 These deities are listed in Egyptian Astronomical Texts 3 pp. 194-199. The depiction of circumpolar deities in H1 is given in detail in Dorman The Tombs of Senenmut.

21 Egyptian Astronomical Texts3 p. 194. 215 Egyptian Astronomical Texts3 p. 189. 216 This is dealt with extensively in Egyptian Astronomical Texts3 pp. 112-114.

209

This implies that by the time the decanswere being used as part of celestial diagrams, the
triangle was obsolete as a functional group of timekeeping stars, but was retained because, whether or not the names held any significance of their own, their existence in the overall picture was important. The lunar month figures and Ramesside star clock tables had an order that was defined by date. It is interesting that the lunar months are the only representative of the moon that we

seeincorporated into the celestial diagrams. The baboon which appearsin the centre of the in the Ramesseumand Medinet Habu ceilings has been said third register (lunar months)217 to representthe intercalary lunar month Thoth,218 and also was present above the jet of the
Karnak water-clock. The water-clock also contains a vignette depicting the sun and the

moon in the form of the figures of Re-Horakhty and Iah, the only direct allusion to these two bodies associated with a New Kingdom celestial diagram. We noted earlier the cyclical nature of the Nut vignette from the Book of Nut in relation to

both spatial and temporal considerations. The water-clock displays essentially the same layout as the Ramesseumand Medinet Habu ceilings mapped on to a truncated conical layout, inclusion this The top the the of of civil calendar months running along surface. together with the months or festivals of the lunar year on the third register reinforce the idea that this diagram is cyclical, or even cylindrical in essence. This idea fits with the notion of define defined belt' loosely `decanal the that as the ecliptic. the now and we path of planets a The rigid distinction between the circumpolar stars and the decans leads us to take this by further imagine to the a circular capped and cylindrical notion one stage sky as a cylinder circumpolar region.
217 Or representations of the festivals of the lunar month if Clagett (Calendars, Clocks, and Astronomy p. 22) is right in discounting Parker's assertion(Calendars p. 43) that the lunar months are portrayed here. 218 ParkerCalendars p. 43.

210

This cylinder is a conceptual model derived from what we have learned from the celestial diagrams that we can use to visualise the Egyptian sky. It must be stressedthat the model is intended be basis for in to as a used a co-ordinate system or for geometrical any way not is it kind, supposedthat the Egyptians ever used a three-dimensional nor any analysis of figure like a cylinder for this purpose. Such a model is useful becauseit allows us to discard some modern preconceptions such as the celestial sphere' which, despite their wide acceptancenow, are not universal physical truths. It allows us to label different areasof the
in distinguish them to a general sense without needing to quantify our theories in sky and terms of relationships with our own astronomical tools: the meridian, the ecliptic, the altazimuth grid etc.
U^weary^^g Stars
-' pFape($ and

Imperishable Stars

decant OutsideNut

SnN
Horizon - r,

InsideNut The Duat

. . .;; "....

. . ..

'

. '. . . . . . . . . . . .

...

Figure 32: Conceptual model of the Egyptian sky

32. in Figure The illustrated is cylinder rotates around its axis representing the The model the sky' about the circumpolar of rotation apparent region. No particular point is

North Celestial 'the the of' rotation Pole', but the circumpolar area point as distinguished is defined. clearly group the circumpolar containing

211

The continuation of the cylinder's surface below the horizon reflects the dramatic text's indication that the stars were still moving after they had set, and indeed were still moving through the sky (on Nut's body) even when they were invisible due to the brightness of the
sun. We know that the horizon was very important in religious literature. We also know

that it was in the region of the horizon that the most important astronomical events took place: sunrise, sunset, heliacal rise, and heliacal set. We must also note that the texts and the events listed above lead us to the conclusion that the horizon must be visualised as a region rather than a line (the `ideal horizon' of modem positional astronomy). We know that the

horizon was referred to using the cardinal points in order to distinguish probably at most

eight different general regions relating to the four cardinal points and the four intermediate points north-east,north-west, south-west,and south-east. The depiction of a zone containing decanal stars does not imply that the decans lay on a line (either inclined) declination or region or any other great circle, a small circle or of equal that could be mapped on a sphere. Here, it is simply indicated that the decanal stars were
The Stars Unwearying the set. the and rose subset non-circumpolar which some of stars, zone containing the planets and the sun is shown for clarity as being slightly closer to the

but decanal the than the The two or coincide, may overlap circumpolar region zone.
being decanal for decanal `ideal' to the the the zone stars precludes seventy-day entirety of

north of the sun and planetspath. No evidence has been found for the existence of `the ecliptic' or `the obliquity of the in decan in formal New Kingdom inclusion the Egypt. The planets concepts as of ecliptic'
lists in celestial diagrams and the description of the decans being `in the entourage' of the impression the that the tightest definition of the path of the sun and the planets was sun gives that they and the decans were in a similar region of the sky.

212

Identification of the Egyptian Stars

It has become customary to close any work on stellar timekeeping with a proposal of a

system identifying certain Egyptian stars with our modern constellations or individual stars. Neugebauerand Parker throughout Egyptian Astronomical Texts stated that Egyptian texts did not hold enough information of a positional nature to make such identifications possible, have extractedenough to propose theories. but many researchers It has already been noted that various identifications for the Egyptian circumpolar group have beenproposed,yet there is even less information about the Egyptian circumpolar group is definition decans Even hour the the the area than about the circumpolar of stars. and limits beyond the figures from Wainwright219 of the group well placing controversial, with the astronomical circumpolar region viewed from Egypt in dynastic times. identifications have beenwidely accepted. There has been some controversy concerning the nature of the `decanalbelt' between two band the of a (and and the ecliptic, to the tracing of) shape a region south of possibilities: is this large To declination controversy extent, a south of the celestial equator. equal There identifications. by manufactured those who wish to promote various schemesof star is nothing in the diagonal star clocks which suggeststhat decansform a `belt' or were chosen (as we saw in Section D) with any positional factors other than ease of observation and identification in mind. The motive suggestedin Section D was of a temporal nature, as have beenother theoriesfor the motive for the construction of star clocks. is a Locher disputed Neugebauerand Parker's method of locating the `decanalbelt': `There
best known interrelated later diagonal whose texts thousand clocks, than the star roughly one years sampleof 70-day is Carlsberg [sic], Papyrus annual a ascribes and whose crucial astronomical statement example

No

219Wainwright `A pair of constellations'.

213

invisibility

to all the 36 decans. Neugebauer and Parker did not doubt this statement despite the 1000-year gap

between the star clocks and the papyrus, and so they concluded that the belt runs roughly parallel to the

his is difficult interpret. '220 However, have We to argument seen that the surviving ecliptic.
diagonal star clocks were compiled and painted around 2000 BC. The seventy-day period mentioned in Papyri Carlsberg 1 and la is also explicitly Kingdom in New the present

in the in date list (which is Nut form Book the the of the either not present of of of versions 221 date 1850 from 13C. In the Book of Nut, the date list is Papyri) and would appear to associated with decans which appear in the diagonal star clocks. This shows that

Neugebauer and Parker's `assumption' was more firmly based than Locher suggests. Locher had previously presented222 a collection of identifications declination. lie belt on a of equal constellations which for Egyptian decanal

Locher believes that the sequence of

labels presentin several astronomical ceilings 3-nwt ht, 4-nwt ht, 5-nwt ht, 6-nwt ht doesnot
indicate (as is generally thought) `third cluster, fourth cluster, fifth cluster, sixth cluster', but

insteadrefers to the number of stars within a cluster (i. e. `a cluster of three, a cluster of four, is basis Yet the indeed. this five, of a cluster of six') which seemsvery unlikely a cluster of his identification of a group of five stars in a `lemonpip' shape as being representedby a
figure present in the ceiling from the tomb of Senmut (H1, see Figure 31), and hence that the decanal belt is parallel to the equator.

He has not so far revealedhis methods of identifying other starsbeyond remarking that they
223 The former is `based reason figurative in are considerationsand agreement right ascension' mainly on

image that by his illustrated form and that the sheep, of a conviction certain stars clearly Sahu figures Sothis by joined be lines inverted the to and resemble an version of others can
I Locher 'Two further coffin lids'. 221 See Section D. 222 Locher 'New arguments for the celestial location of the decanal belt'. n3 Locher'Two further coffin lids'.

214

in the vertical strips of diagonal star clocks.

The latter reason is weak due to the decans

originally, and perhaps always, being observed near the horizon which does not trace a line of equal right ascension on the celestial sphere. Neugebauer and Parker224modelled the risings of decans by a series of great circles or `horizons', yet the physical realities of near225 horizon observation mean that their model can only be used (as they surely intended) as a conceptual scheme and not as a definitive tool to produce positive identifications. Bker, in his attempt to identify decans, relinquished the idea of a `belt' in favour of a system based on the right ascension of the culminating moon. His conclusions, collated and his death in `ber Namen und Identifizierung after published der gyptischen Dekane', are

draws heavily identifications full discussion his list His of presented without a of methods. on later information, his Again, by stars. and parallels with other cultures' naming of

in have been been by their own have researchers other criticised not accepted and proposals attempts towards identifications. Leitz's work on the hour stars has already been extensively discussed in Section C. arrives at possible identifications He

his analysis methods, and as of as a natural conclusion Yet he himself admits

identification. for the strongly most argued case such, presents stellar that his identifications cannot be accepted as certain, or even probable.

Until enough evidence is gathered to create some consensus of opinion on any identifications Parker's that Sirius, the Orion, The Plough Neugebauer than assertion and and other Egyptian stars will remain unknown will stand. identification to the solution no offers The conceptual model presented above

problem, indeed the removal of the spherical co-

but is frame from this further our grasp, of reference such places ordinate a solution even balanced by the increased freedom from assumptions that the model encourages.

224 Egyptian Astronomical Texts 1 pp. 97-107.

271 Schaefer'Refraction near the horizon', `Heliacal rise phenomena', `Extinction angles' etc.

215

CONCLUSION
We have examined minutely timekeeping methods, some based on familiar concepts and others alien to us, and have also briefly examined a wider perspective of the composition of the sky. Both areas form part of our understanding of the extent to which the ancient Egyptians were interested in the sky. In the course of this study, it has been demonstrated that preconceptionsof methodology, motives, aims, and priorities form a barrier betweenthe modem mind and the ancient sources. In the case of the `transit star clock', these preconceptions have led an assumption to become a generally accepted, previously unchallengedtheory. In the caseof the Ramessidestar clock, one hundred years of sporadic discussionhas produced a theory of usagealmost by default, which certainly deservescloser inspection. A new theory has been offered which demonstrates that the subject cannot yet be by is the ease Further Ramesside tables aided the now consideredclosed. study of star clock The to use of such of analysisusing computerswhich were not available earlier researchers. has in data hypotheses the find to test study the to present trends original methods and within indicated that further information may be gained in this way in the future. In the section concerning the development of stellar timekeeping methods, this new incorporated A into types possible series of events approachwas a survey of of star clock. decans the as a persistentgroup in Egyptian cosmography,and which may which established have produced many more variants of stellar timekeeping methods than have survived, has been outlined and discussed. We saw that this process may well have been a simpler but
longer course of development than the widely accepted but rarely stated `flash of genius'

its in diagonal from first the the fully-formed mind of which scenario star clock sprang
inventor. single The gradual development of such a complex method is intrinsically more

216

likely, but has appealedmore to those meeting the tables for the first time than to those who
study them professionally. Perhaps this is because the complexity of the tables which is

is lost to researchers who habitually deal with the formulae to the obvious unprepared mind

and order of modem mathematicalastronomy.

During the discussion of each timekeeping method, including the familiar concept of shadow length, we have also faced the difficulty itself. That different time of accepting a notion of

is, that the makers and users of instruments had a set of priorities which differ considerably from our own. Nowadays, `time' is a measuring stick for recording the order and relative length of events. It has, as a concept, an existenceand integrity of its own. Egyptian `time' is different. `Time' as a steady,measurable,and abstractflow did not exist it did in the Egyptian world picture. The Nile rose when it did, Sirius appeared and the when sun rose when its battles during the night had been completed. This self-referential nature of time is incredibly difficult for us to grasp fully, as is the recurrent Egyptian notion that there
king to Commonly, in the `time' said was the two types types two of eternity. of sense of are be given life like the god Re for ever and ever. `Ever and ever' is a common phrase in the English language (used in a similar context in the Lord's Prayer, for example) repeating the

Egyptian for However, the `ever' is phrasewhich translation this of emphasis. word a weak in fact uses two different words dt and nhh, that is, linear eternity and cyclical eternity devices is later timekeeping It the to the time, concept, cyclical which relates respectively. discussed. have we As an aside,we can perhapsdraw a parallel between the cyclical time, which the star clocks measuredand which the Egyptians surely would originally have thought of as eternal and due linear to the year the time, liken the and changing, to slow creep never which we can length or the almost imperceptible change wrought by precession that together conspired

217

both to frustrate the Egyptians' attempts at temporal stability and to provide the modern researcher with tools for analysis and dating. Returning to the Egyptian viewpoint, the passing of `time' would only be seen in the

occurrence of events, and if these events occurred at intervals which seem ill-defined or

irregular to us, then that is a problem with our perception, not a question of `inaccuracy' in
ancient methods. The implication of this approach to our analysis of timekeeping

instrumentsis that each instrument must be allowed to define time, not just measureit.
It seems that `hours' were initially divisions of the night in the same sense that `gates' and

`caverns' were divisions of the Duat. They presenteda time framework for the eventswhich
occurred to the sun in his nightly voyage through the Duat, and by extension, the journey of

the deceasedafter death. This connection between night-time events and the dead is apparentthroughout the Pyramid Texts, and continuesto hold throughout the New Kingdom.
All our sources for stellar timekeeping and astronomical diagrams come from objects or

locations associatedwith the dead (with the one exception of the Karnak water-clock). Although the survival rate for funerary objects and texts is higher than that associatedwith
items everyday life, the connection can be shown to be valid by the instances of astronomical

ceilings in mortuary temples where their east-bank`living' temples show only repeating star patterns. It is only in the New Kingdom that we have clear evidence that the formal night `hour' daylight daylight is hours. it that to Of the no applied was suggested not concept course, time divisions existed until the New Kingdom for use in daily life. Ad hoc sundials, waterclocks, and wicks would all serve to measuresome repeatabletime period sufficient to time working shifts. With the collection of sourceswe now have, we thereforejudge that timekeeping during the

218

night using stars was clearly a ritual procedure.

We also must agree that logically,

timekeeping in the dark has little conceivably practical purpose. This completely divides the Egyptian perception of the night hours from our modem idea that all `hours' are of the same, abstract nature. The Egyptian train of thought would not naturally progress from the

motivation of ritual to thoughts of standardisation and of accuracy which we automatically associate with the word `clock'. Releasing the instruments from the duty of measuring

accuratelyan external unit has massive repercussionson the study of Egyptian timekeeping. In the past, each study of a newly discoveredmethod has begun by asking: `How does this instrument measurehours and how accurateis it? ' If modem hours are completely removed
from the initial investigation, by first be timekeeping approached method can each

determining the relationship between the instrument and the reference it used (for example the stars, the sun) and learning as much as possible about that relationship. Only after all

primary information has been gathered, including context and related texts, should the device? ' by this be `What time the the marked periods asked: question was nature of Questions of accuracy are irrelevant until the very last stages of investigation because `accuracy' requires the intrusion of modem parametersthat can colour the perception of the
instrument radically and lead to conclusions that are inappropriate.

in this field has a duty to look first at the Egyptian sources,and to respectat Each researcher information despite damage, the that, times contemporary copying errors, and confusion, all intrusive supersede modern operational parameters. must always This study has attempted to follow these precepts as closely as possible. Using the basic facts thus gathered,new theories have been presentedwhich offer an alternative or improved understandingof the ancient methods and concepts. It is hoped that such work will provide for future discussion basis of the subject areascovered. Two areaswhich may yield further a

219

results are clear: the Ramesside star clock and the field of astronomical vocabulary in
religious literature. The former has already been mentioned, while the latter area was

touched on during research towards this study, and proved to be a huge but potentially rewarding undertaking, linking the subjects of Egyptian astronomy, linguistics, and religion.

It is hoped that with continued research,improved methodology, and new sources,the two areas of timekeeping and cosmography will continue to add to our understanding of the developmentof human scientific thought, and that this study has contributed to that process.

220

APPENDIX
Chronology (after Baines and MalekAtlas ofAncient Egypt)

Late Predynastic
Early Dynastic

c. 3000 BC
2920-2575 BC Ist to IIIrd Dynasty

Old Kingdom Ist IntermediatePeriod Middle Kingdom 2nd IntermediatePeriod New Kingdom 3rd IntermediatePeriod Late Period Greco-Roman

2575-2134 BC 2134-2040 BC 2040-1640 BC 1640-1532BC 1550-1070BC 1070-712BC 712-332 BC 332 BC - AD 395

IVth to VIIIth Dynasty IXth to XIth Dynasty XIth to XIVth Dynasty XVth to XVIIth Dynasty XVIIIth to XXth Dynasty XXIst to XXVth Dynasty XXVth to XXXIst Dynasty Macedonian,Ptolemaic, Roman

The New Kingdom

Aamose Amenhotep I Djehutymes I Djehutymes II Djehutymes III Hatshepsut Amenhotep II Djehutymes IV Amenhotep III Amenhotep IV (Akhenaton) Smenkhkare Tutankhamun Ay Horemheb Ramesses I Seti I Ramesses II Merneptah Amenmes Seti II Siptah Twosret Setnakht Ramesses III Ramesses IV Ramesses V Ramesses VI Ramesses VII Ramesses VIII Ramesses IX Ramesses X Ramesses XI 221 1550-1525 1525-1504 1504-1492 1492-1479 1479-1425 1473-1458 1427-1401 1401-1391 1391-1353 1353-1335 1335-1333 1333-1323 1323-1319 1319-1307 1307-1306 1306-1290 1290-1224 1224-1214 1214-1210 1210-1204 1204-1198 1198-1196 1196-1194 1194-1163 1163-1156 1156-1151 1151-1143 1143-1136 1136-1131 1131-1112 1112-1100 1100-1070

XVIIIth Dynasty 1550-1307

XIXth Dynasty 1307-1196

XXth Dynasty 1196-1070

BIBLIOGRAPHY
Allen, JP Genesisin Egypt: ThePhilosophy of Ancient Egyptian Creation Accounts Yale Egyptological Studies2, New Haven, Connecticut 1988
Allen, TG 'Occurrences of pyramid texts with cross indices of these and other Egyptian mortuary texts' SAOC 27 Chicago University Press 1950

Assman,J Egyptian solar religion in the New Kingdom. Re, Amun and the crisis of polytheism Kegan Paul 1995 Baines,J and Malek, J Atlas of Ancient Egypt Phaidon 1989 Bakir, AM The Cairo Calendar No. 86637 Cairo 1966 Balmer, R T'The formula of the clepsydra' Timecraft 1982 0 pp. 12-17,21-25 BattiscombeGunn 'The Coffins of Heny' ASAE 1926 26 pp. 166-171 Beaux, N 'Sirius Etoile et JeuneHorus' Hommagesa Jean Leclant 1993 pp. 61-72 Biegel, RA Zur Astrognosie der alten gypter Zurich 1921 Bierbrier, M The Tomb-Buildersof the Pharaohs BMP 1982 Bilfinger Die Sterntafeln in der aegyptischenKoenigsgraebenvon Biban el-Moluk Stuttgart 1891 Bker, R'ber Namen und Identifizierung der gyptischenDekane' Centaurus 1984 27 pp. 189-217 Borchardt, L Die Altgyptische ZeitmessungBerlin 1920 Borchardt, L'Altgyptische Sonnenuhren'ZAS 1910 48 pp. 9-17 Bradshaw,J TheImperishable Stars of the Northern Sky in the Pyramid TextsBradshaw 1990 Brugsch, 117hesaurusInscriptionum Aegyptiacarum Leipzig 1883 Bruins, EM 'Egyptian Astronomy' Janus 1965 52 pp. 161-180 Bruins, EM `The Egyptian shadow clock' Janus 1965 52 pp. 127-137 Bull, LS 'An ancient Egyptian astronomical ceiling-decoration' BMMA 1923 18 pp. 283-286 Capart,J 'Astronomie Egyptienne' Ciel Terre 1942 58 pp. 106-107 ' CdE 1941 16 pp. Capart,J 'Est-il possible de determiner les etoiles des tableaux astronomiquesegyptiens? 251-252 IV and the Turin Plan of a Royal Tomb' JEA 1917 4 pp. 130-58 Carter and Gardiner 'The Tomb of Ramesses chatley, H'Ancient Egyptian astronomy' Nature 1939 143 p. 336 Chatley, H 'Egyptian Astronomy' JEA 1940 26 pp. 120-126 Chatley, II 'Notes on Ancient Egyptian astronomy' The Observatory 1939 62 pp. 100-104 Chatley, H 'The Egyptian Celestial Diagram' The Observatory 1940 63 pp. 68-72 Clagett, M Ancient Egyptian Science Vol. 1: Knowledge and Order American Philosophical Society 1989

222

Clagett, M Ancient Egyptian Science Vol. 2: Calendars, Clocks, and Astronomy American Philosophical Society 1995
QEdat, J 'Notes sur 1'isthme de Suez (monuments divers)' RecTrav 1915 37 pp. 33-40 Collier, M and Manley, B How to read Egyptian hieroglyphs British Museum Press 1998

Cotterell B, Dickson F P, and Kamminga J 'Ancient Egyptian Water-clocks: A Reappraisal'JAS 1986 13 pp. 31-50 Daressy,G'Deux clepsydresantiques' BIE 1915 9 pp. 5-16 Daressy,G'L'Egypte celeste' BIFAO 1916 12 pp. 1-34 David, RA guide to religious ritual at Abydos Warminster 1981
Davis, V L'Identifying Ancient Egyptian Constellations' JHA 1985 9 pp. S102-104

Davies, N de Garis 'An architect's plan from Thebes' JEA 1917 4 pp. 194-199 Davies, N de Garis The Rock Tombs of El Armana Part 6 London 1908

de Puydt, L'Ancient Egyptian star clocks and their theory' BiOr 1998 55 pp. 5-44 Dmichen, J Der Grabpalast des Patuamenapin der Thebanischen Nekropolis 1884 Dorman, PF The tombs of SenenmutMetropolitan Museum of Art, New York 1991
Eggebrecht, A Antike Welt Pelizaeus-Museum Hildesheim: die gyptische Sammlung Mainz 1993 Eggebrecht, ASuche nach Unsterblichkeit Mainz 1990

Eisler, Rand Chatley, 11'Egyptian Astronomy - Letters from Dr Eisler and Dr Chatley' JEA 194127 pp. 149152
Faulkner, R O'The King and the Star Religion in the Pyramid Texts' JNES 1966 25 pp. 153-161 Faulkner, R0A Concise Dictionary of Middle Egyptian Griffith Institute 1962

Faulkner, R0 TheAncient Egyptian Coffin Texts Vols. 1,2, and 3 Aris and Phillips 1973 Faulkner, R0 TheAncient Egyptian Pyramid Textswith SupplementClarendonPress1969 Faulkner, R0 TheEgyptian Book of the Dead Chronicle Books 1994 Fermor, JH 'Perceived night length ratios in ancient Egypt' Vistasin astronomy 1993 36 pp. 363-373 Fermor, J H'Timing the sun in Egypt and Mesopotamia' Vistasin astronomy 1997 41 pp. 157-167 Fermor,J 11,BurgessA E, and Przybylinski, V 'The time keeping of Egyptian outflow clocks' Endeavour 1983 7 pp. 133-136 Floyer, E A'Primitive Sundials in Upper Egypt' TheAthenaeum 18953545 Oct 5 Frankfort, H The Cenotaphof Seti I at Abydos EES 1933 Gardiner, Sir Alan Egyptian Grammar 3rd Edition Griffith Institute 1988 Gillings, RJ Mathematics in the Time of the Pharaohs Dover 1982
Griffith, F LI and Petrie, WMF Fund London 1899 Two Hieroglyphic Papyri from Tanis 9th Memoir of the Egypt Exploration

223

Iiabachi, L The Obelisksof Egypt American University in Cairo 1988 I lomung, E Conceptionsof god in ancient Egypt London 1983 Ingham, M F'The length of the Sothic cycle' JEA 1969 55 pp. 36-40 Isler, M 'An ancient method of finding and extending direction' JARCE 1989 26 pp. 191-206 Islet, M 'The gnomon in Egyptian antiquity' JARCE 199128 pp. 155-185
Kahl, J 'Textkritische Bemerkungen zu den Diagonalsternuhren des Mittleren Reiches' SAK20 1993 pp. 95107

Kozloff, Arielle P'Star-Gazing in Ancient Egypt' Hommagesa Jean Leclant 1993 pp. 169-176 in den Pyramiden texten' A In press Krauss, R'Astronomische Konzepte und Jenseitsvorstellungen Krauss, R Sothis und Monddaten Hildesheim 1985
Krupp, E'Egyptian Astronomy: the roots of modem timekeeping' New Scientist 1980 85 pp. 24-27 RecTrav 1916 38 pp. 70-84

Kuentz, C'Note sur un gnomon portatif gr8co-igyptien'

O'Papyrus Carlsbergno. 1, ein hieratisch-demotischerkosmologischerText' Lange, 11O and Neugebauer, SelskabHist. filol. Skrifter 1, No21940 Danske Vuienskabernes
Lapp, G 'Srge des Mitteln Reiches aus der ehemaligen Sammlung Khashaba' A 1985 43 pl. 39 Larson, JA 'The Tutankhamun Astronomical Instrument' Amarna Letters 1992 2 pp. 77-86 Lefebure, G'Le tombeau de Seti I' MIFAO 2 1886

Leitz, CAltaegyptische Sternuhren Leuven 1995 Leitz, C Studien zur aegyptischenAstronomie Harrasowitz, Wiesbaden 1989 Leitz, C Studien zur aegyptischenAstronomie 2 Harrasowitz, Wiesbaden1991 Lesko, LH Theancient Egyptian book of two ways California 1972 Lesko, L 11Index of the Spells on Egyptian Middle Kingdom coffins and related documentsBerkeley 1979 Lichtheim, M Ancient Egyptian Literature Vols. 1,2, and 3 California 1973 Locher, K'A ConjectureConcerning the Early Egyptian Constellation of the Sheep'JHA 19819 pp. S73-75 Locher, K'A Further Coffin Lid with a Diagonal Star Clock from the Egyptian Middle Kingdom' JHA 1983 14 pp. 141-144
Locher, K'Middle Kingdom astronomical coffin lids: extension of the corpus from 12 to 17 specimens since Neugebauer and Parker' Proc 7th ICE 1998 pp. 697-702

Locher, K 'New argumentsfor the celestial location of the decanalbelt and for the origins of the S3hhieroglyph' Acts 6th ICE II Turin 1993pp. 279-284 Locher, K 'New identifications of Egyptian constellations discussedat the 5th international congressof Egyptology, Cairo 1988' Archeologia e astronomia, Rivista di Archeologia 1989 9 pp. 216-217 and pls. 61-63 Locher, K'Probable Identification of the Ancient Egyptian Circumpolar Constellations'JHA 1985 9 pp. S152153

224

Locher, K 'Two further coffin lids with diagonal star clocks from the Egyptian Middle Kingdom' JHA 23 1992 pp. 201-207 Macalister, RAS TheExcavation of Gezer1902-1905 and 1907-1909 London 1912 Macnaughton,D `The use of the shadowclock of Seti I' JBAA 1944 54 pp. 135-137 Malville, J M, Wendorf, F, Mazar, A A, and Schild, R `Megaliths and Neolithic astronomy in southern Egypt' Nature 1998 392 pp. 488-491 Mayestre,C 'Le livre de la vache du ciel' BIFAO 40 1941
Meeks, D and Favard-Meeks, C Daily Life of the Egyptian Gods John Murray 1997 Murray, MA The Osireion at Abydos B. Quaritch 1904

Naville, EH The Templeof Deir el-Bahari Egypt Exploration Fund 1907,1910,1913 Nelson, 1111and Hoelscher,U Medinet Habu Chicago 1929 Neugebauer,0 `Ancient Egyptian Astronomy' Nature 1939 143 pp. 115,765 Neugebauer,0 'Egyptian planetary texts' Transactionsof the American Philosophical Society 1942 32 pp. 209-263 Neugebauer,0 History of Ancient Mathematical Astronomy pp. 562-568 Neugebauer,0 'The Egyptian Decans" Vistasin astronomy 1955 1 pp. 47-51 Neugebauer,0 'The Egyptian Picture of the Sky' Science1939 90 pp. 410 Neugebauer,0 'The Origin of the Egyptian Calendar'JNES 1942 1 pp. 397-403
Neugebauer, 0 and Parker, REgyptian Astronomical Texts Vol. 1 Brown University Press 1960 Neugebauer, 0 and Parker, REgyptian Astronomical Texts Vol. 2 Brown University Press 1966 Neugebauer, 0 and Parker, REgyptian Astronomical Texts Vol. 3 Brown University Press 1969

Nims, C Ramesseum sources of Medinet Habu Reliefs SAOC 39 1977 Parker, RA OUP 1974 'Ancient Egyptian astronomy' in The place of astronomy in the ancient world FR Hodson ed.

Parker, RA The calendars of ancient Egypt SAOC 26 1950 Parkinson, RB Voices From Ancient Egypt BMP 1991

Petrie,WMF Ancient weights and measuresUniversity College 1926 pp. 45-46 Piankoff, ALe Livre du Jour et de la Nuit IFAO 1942
Piankoff, A'The sky-goddess Nut and the night journey of the sun' JEA 1934 20 pp. 57-61

Piankoff, A The Shrines of Tut-Ankh-Amon (=Bollingen Series, 40,2: Egyptian Religious Texts and Representations) Pantheon 1955

Pogo,A 'Egyptian Water Clocks' Isis 1936 25 pp. 403-425 Pogo,A'The astronomicalceiling-decoration in the tomb of Senmut' Isis 1930 14 pp. 301-325 Pogo,A'Three Unpublished Calendarsfrom Asyut' Osiris 1936 1 pp. 500-509

225

Pogo,A'Zum Problem der Identifikation der nordlichen Sternbilder der alten Aegypter' Isis 1931 16 pp. 102114 Ridpath, I Norton's 2000.0 (18th Ed) Longman Scientific and Technical 1989 Schack-Schackenburg, II AegyptolischeStudien Part I Leipzig 1902 Schaefer,BE and Liller, W 'Refraction Near the Horizon' Publications of the Astronomical Society of the Pacific 1990 102 pp. 796-805 Schaefer,B B'Extinction Angles and Megaliths' Sky and Telescope1987 73 pp. 426 Schaefer,B E'Iieliacal Rise Phenomena'JIIA 1987 11 pp. S19-33 Schaefer,BE 'Predicting Iieliacal Risings and Settings' Sky and Telescope1985 70 pp. 261-263 Schaefer,B E'Refraction by Earth's Atmosphere' Sky and Telescope 1989 77 pp. 311-313 Scott, N E'An Egyptian sundial' BMMA 1935 30 pp. 88-89
Sethe, K Die altaegyptisd: en Pyramidentexte Leipzig 1908

Sethe,K bersetzungund Kommentar zu den altdgyptischen Pyramidentexten Hamburg 1936-62 Sloley, RW 'Ancient Clepsydrae' AE 1924 pp. 43-50 Sloley, RW 'Primitive methodsof measuring time with special referenceto Egypt' JEA 193127 pp. 166-178 Sottas,11'Une petite horloge astronomiquegr6co-6gyptienne'RecTrav 1916 38 pp. 1-7 Taff, LG ComputationalSpherical Astronomy John Wiley and Sons 1981 Tiradritti, F'Three yearsof researchin the tomb of Harwa' BEES 1998 13 pp. 3-6 van der Waerden,B L'Aegyptische Planetrechnung'Centaurus 1972 16 pp. 66-91 Wainwright, G A'Orion and the great star' JEA 1936 22 pp. 45-46
Wainwright, G A'A pair of constellations' Studies Presented to F. Ll. Griffith London 1932 pp. 373-383

Wainwright, GA The Sky-Religion in Egypt OUP 1938

Wells, RA 'Origin of the hour and the Gatesof the Duat' SAK 1993 20 pp. 305-326 Wells, RA 'Sothis and the SatetTemple on Elephantine: A Direct Connection' SAK 1985 12 pp. 255-302 Willems, II Chestsof Life: A Study of the Typology and ConceptualDevelopmentof Middle Kingdom Standard Class Coffins (= Mededlingen en Verhandelingenvan het Vooraziatisch-EgyptischGenootschap"Ex Oriente Lux", 25) Ex Oriente Lux 1988 Wilkinson, R 11'New Kingdom Astronomical Paintings and Methods of Finding and Extending Direction' JARCE 199128 pp. 149-154 IS ba, Z L'Orientation Astronomique dans l'Ancienne Egypte, et la Precessionde l'Axe du Monde Prague1953

Zinner, E'Die Sternbilder der alten Aegypter' Isis 193116 pp. 92-101

226

Index of Abbreviations in the Bibliography

AA AgyptologischeAbhandlungen, Wiesbaden

Acts 6th ICE SestoCongressoInternazionale di Egittologia, Atti, Turin AE ASAE BEES BiOr BIE BIFAO BMMA
CdE

Ancient Egypt and the East, London Annales du service des antiquites de 1'Egypte,Cairo Egyptian Archaeology: TheBulletin of the Egypt Exploration Society, London Bibliotheca Orientalis, Leiden Bulletin de l'Institut d'Egypte, Cairo Bulletin de l'Institut Francais d'Archeologie Orientale, Cairo Bulletin of the Metropolitan Museum of Art, New York
Chronique d'Egypte: Bulletin Perio clique de la Fondation Egyptologique Reine Elisabeth, Brussels

JARCE JAS JBAA JEA

JHA JNES MIFAO

Journal of the American ResearchCenter in Egypt, Boston Journal of Archaeological Science,London Journal of the British Astronomical Association, London Journal of Egyptian Archaeology, London
Journal for the History of Astronomy, Chalfont St Giles Journal of Near Eastern Studies, Chicago Memoires Publics par les Membres de l'Institut Francais d'Archeologie Orientale du Caire, Cairo

ICE Proceedingsof the SeventhInternational Congressof Egyptologists 1995, Proc 711; Cambridge
RecTrav Recueil de travaux r6latifs a la philologie et a 1'archeologie egyptiennes et assyriennes, Paris

SAK SAOC ZAS

Studien zur Altgyptischen Kultur, Hamburg Studiesin Ancient Oriental Civilisation, Chicago Zeitschrift fr gyptische Spracheund Altertumskunde,Leipzig

227

TABLES
Table 1: Idealisedlayout of a diagonal star clock 17 .............................................................................................. Table 2: Layout of Al. 19 ........................................................................................................................................ Table 3: Layout of A2 19 ........................................................................................................................................ Table 4: Layout of A3 20 .................................................................................................................................... .... Table 5: Layout of A4 20 .................................................................................................................................... .... Table 6: layout of AS 21 .................................................................................................................................... .... Table 7: Layout of A6 21 .................................................................................................................................... .... Table 8: Layout of A7 22 .................................................................................................................................... .... Table 9: Layout of A8 22 .................................................................................................................................... .... Table 10: Layout of A9 23 .................................................................................................................................. .... Table 11: Layout of A10 23 ................................................................................................................................ .... Table 12: Layout of All 24 ................................................................................................................................ .... Table 13: Layout of A12 24 .................................................................................................................................... Table 14: Layout of A13 24 .................................................................................................................................... Table 15: Layout of A14 25 .................................................................................................................................... Table 16: Layout of A15 25 .................................................................................................................................... Table 17: Layout of A16 26 .................................................................................................................................... Table 18: Layout of A17 26 .................................................................................................................................... Table 19: List of decannameswith Neugebauerand Parker's numbering system 32 ............................................ Table 20: Concordanceof decansused in seventeen diagonal star clocks 33 ...................................................... ... Table 21: Contentsof the list and epagomenalcolumns of diagonal star clocks Al, A6, A7, and A8 35 .............. Table 22: Decan lists e and 4) 43 ........................................................................................................................ ... Table 23: Date schemefor diagonal star clocks Al to A17 49 ............................................................................ ... Table 24: Datesof the star clock tables according to Neugebauerand Parker 50 ... ............................................... Table 25: Layout of the star clock in the Sloping Passage 53 of the Osireion ..................................................... ... Table 26: Comparisonof decansfrom the Sloping Passage star clock with those from early sources.............. 54 Table 27: Summary of dating schemefor decanlists 55 ........................................................................................ 56 Table 28: Star circles associated with decansin the Book of Nut ...................................................................... Table 29: Date setsand decannamesfrom the Book of Nut 61 .............................................................................. Table 30: Further date setsand decannamesfrom the Book of Nut 61 .................................................................. Table 31: Date setswithout decannamesfrom the Book of Nut 62 ........................................................................ Table 32: Fragmentarydecanlist from the Book of Nut 62 .................................................................................... Table 33: Possiblecombinations of decannamesand date setsin the Book of Nut 64 ........................................... Table 34: Idealisedversion of Text U from the Book of Nut 66 ............................................................................. Table 35: The Hour Stars from the Ramessidestar clock tables,grouped into constellations 77 ........................... Table 36: Leitz's seasonalhour predictions for the Ramesside 86 star clock tables............................................... Table 37: Occurrencesof hour stars Ll and L2 97 ................................................................................................. 102 Table 38: Systemof equationsfrom Ramessidestar clock tablesbasedon time of sunset ............................. Table 39: Systemof equationsfrom Ramessidestar clock tables basedon equal hours during one night...... 105
Table 40: Position and period of invisibility

star clocks and celestial diagrams ..................................................................... Table 50: Deities of the decans ........................................................................................................................

Table 46: Celestial diagrams grouped by layout 194 ......................................................... ..................................... Table 47: Composition of the circumpolar group in celestial diagrams 195 .......................................................... Table 48: Dccan lists from New Kingdom astronomical diagrams 197 ................................................................. Table 49: Triangle lists from
200 205

Table 45: Summary of elements contained within celestial diagrams .............................................................

Table 41: A chronology of stellar timekeeping in Ancient Egypt 126 ................................................................... 137 Table 42: Comparisonof day and night lengths through the year from the Cairo calendar ............................. hours Table 43: Comparisonof existing hour marks with those required to produce seasonal 141 ....................... Table 44: Concordancebetweentexts from the Book of Nut in the Osireion (Fl), the Tomb of Ramesses IV (F2) and in PapyrusCarlsberg 1(F3) 166 ......................................................................................................
193

of Sirius from 4500 BC to the present ....................................... 110

228

FIGURES
Figure 1: Hildesheim Inv. Nr. 5999..........
Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure 2: Locher's diagram of diagonal star clock classification ...................................................................... 3: Kahl's stemma superimposed on Neugebauer and Parker's grouping system by Locher .................. 4: Star clock in the Sloping Passage of the Osireion .............................................................................. 5: Relationship between sources belonging to the `Seti I A' family ...................................................... 6: Example of the layout of a Ramesside star clock table ...................................................................... 7: Position labels for the seven central vertical lines from a Ramesside star clock table ....................... 8: Configuration of Ramesside star clock tables .................................................................................... 9: Configuration for conventional interpretation .................................................................................... 10: Configuration for reflection conjecture ............................................................................................ 11: Ramesside star clock tables Ti to T12 .............................................................................................

27 ..........................................................................
31 31 53 58 69 71 82 82 82 88

Figure 15: Shadow clock from the time of Tuthmosis III (El) ........................................................................

89 Figure 12: Ramesside star clock tables T13 to 124 ........................................................................................... Figure 13: Motion of the stars in relation to the observer for the two theories for the Ramessidestar clock 98 .... 106 Figure 14: Hours of the night generatedby the equationsof Table 39 ............................................................
129 131

129 Figure 16: Shadowclock from Fayum (E3) .....................................................................................................

Figure 17: Shadow clock text from the Book of Nut in the Osireion ...............................................................

Figure 23: Figure 24: Figure 25: Figure 26: Figure 27: Figure 29: Figure 30: Figure 31: Figure 32:

Figure 21: Sloping sundial from Qantara ......................................................................................................... Figure 22: Two views of the sloping sundial in the Petrie Museum ................................................................

133 Figure 18: Tanis papyrus ................................................................................................................................. 134 Figure 19: Ptolemaic hieroglyphs depicting shadow clocks and sloping sundials ........................................... 146 Figure 20: Time periods measuredby a sun-aligned shadow clock on the summer solstice at Luxor .............
149 149

Figure 28: Dramatic text from the Osireion

149 Schematicchronology of Egyptian sundials ................................................................................... 150 Hour Lengths produced by the sloping sundial ES ......................................................................... 156 Vignette from the Book of Nut in the Osireion ............................................................................... 157 Index of texts in the Nut vignette in the Osireion ........................................................................... 158 IV Index of texts in the Nut vignette in the Tomb of Ramesses .....................................................
..................................................................................................... 167

Hall of KV9 178 Layout of the secondand third registersof the Book of the Day in the Sarcophagus 180 Progressof the sun during the day, according to the Book of the Day ........................................... 192 Ceiling from the tomb of Senmut ................................................................................................... 211 Conceptualmodel of the Egyptian sky ...........................................................................................

229