climate and the history of egypt -the middle kingdom

48
Climate and the History of Egypt: The Middle Kingdom Author(s): Barbara Bell Source: American Journal of Archaeology, Vol. 79, No. 3 (Jul., 1975), pp. 223-269 Published by: Archaeological Institute of America Stable URL: http://www.jstor.org/stable/503481 . Accessed: 04/09/2013 15:09 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. . Archaeological Institute of America is collaborating with JSTOR to digitize, preserve and extend access to American Journal of Archaeology. http://www.jstor.org This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PM All use subject to JSTOR Terms and Conditions

Upload: alberto-nonimo

Post on 24-Oct-2015

72 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Climate and the History of Egypt -The Middle Kingdom

Climate and the History of Egypt: The Middle KingdomAuthor(s): Barbara BellSource: American Journal of Archaeology, Vol. 79, No. 3 (Jul., 1975), pp. 223-269Published by: Archaeological Institute of AmericaStable URL: http://www.jstor.org/stable/503481 .

Accessed: 04/09/2013 15:09

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

Archaeological Institute of America is collaborating with JSTOR to digitize, preserve and extend access toAmerican Journal of Archaeology.

http://www.jstor.org

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 2: Climate and the History of Egypt -The Middle Kingdom

Climate and the History of Egypt: The Middle Kingdom'

BARBARA BELL

TABLE OF CONTENTS

Introduction and Survey of Dynasty XII ................................... ..... 224

N ile Levels from the M iddle K ingdom .................................................................. .... 226 Reign of Senw osret I

..................................................................................... ......... 226

Evidence from N ubia ........ ............................................................................... ............ 229

H igh w ater levels (H W Ls) ......................................................... 229 Low water levels (LWLs) ............................................. ......236

Volum e of the great Sem na floods ................................................................................ 238

Evidence from Egypt .................. ............................................. ..... 245

Rainfall in the Middle Kingdom ........................................................................... 247 E g y p t ...................... ........ ........................... ............................................ 247 N ubia

....................................... ............... . . 248 Lake M oeris and the Fayum ........................................................... 249

Early Dynastic ........................................ 252 Old Kingdom ....................................... ...........................252 M iddle K ingdom .......................... . ................. 254 Lake Moeris and the levels of the Nile

................................. ......... 255

Some historical implications of the great floods ..................................... 257

Decline of the M iddle Kingdom ........................................ 260

D is c u s s io n....................................................................................................................

2 6 5

References .......................................................... 266

Abstract

The first major section of this paper (p. 226f) surveys evidence bearing on the level of the Nile during the Middle Kingdom and the Second Inter- mediate Period, I991 to ca. 1570 B.C. With the exception of the analysis of the range of a "good flood" in the reign of Senwosret I, most of the evi- dence comes from excavations stimulated by the building of the High Dam at Aswan, from now- flooded sites in Nubia, and from the inscriptions on the cliffs at the Semna region of the Second Cataract, long-known and troublesome because commemorat- ing flood levels 8 to II m. above the modern from some 27 years in the reigns of King Amenemhet III and his immediate successors. Previous hypotheses

are discussed and rejected, and the inscriptions are interpreted literally as indicating actual great floods; the peak volume, in those years for which records exist at Semna, is estimated to have been 3 to 4 times that of the larger floods recorded at Aswan since A.D. 1870. These floods are interpreted as reflecting a climate fluctuation of only a few decades' duration and are not seen as typical of the Middle Kingdom, which appears otherwise to have had floods similar to those of modern times. Review of textual and architectural evidence bearing on rain- fall suggests that the Middle Kingdom had condi- tions similar to those of the A.D. i8oos, with heavy rainfalls somewhat less rare than in the present century.

II take this opportunity to thank Professors William Y. Adams (University of Kentucky), Karl W. Butzer (University of Chicago), and William Kelley Simpson (Museum of Fine Arts, Boston; Yale University) for their generous interest and

encouragement in this work; each of them read the semi-final draft and made valuable suggestions and comments. I received also a number of useful suggestions from Prof. Sterling Dow

(Boston College). The chronology followed in this paper is that of the revised

Cambridge Ancient History, particularly Hayes (i96I), but the chronology of the XIIth Dynasty is known with exceptional exactness, so that there should be no significant difference among various authorities.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 3: Climate and the History of Egypt -The Middle Kingdom

224 BARBARA BELL [AJA 79

The second major section (p. 249f) presents evi- dence bearing on the level of Lake Moeris. It is concluded that the lake was in free connection with the Nile from Neolithic to Ptolemaic times when the level was artificially reduced. It appears that more than seasonal fluctuations in the lake level oc- curred from time to time (especially during the Neolithic), with possible interruptions in the free connection, as during the very low Niles and drought of the First Dark Age. If the connection then was restored with human aid, it appears most likely this was done early in the XIIth Dynasty under Amenemhet I.

The third principal section (p. 25If) discusses possible cultural and historical influences of the great floods recorded at Semna in the last reigns of Dynas- ty XII and early in Dynasty XIII, both while they were occurring and upon their cessation. There is no evidence that the decline in material prosperity and strength of the central government under Dy- nasty XIII was associated with any such severe failure of the floods and famine as brought on the First Dark Age. The few known famine inscrip- tions, from El Kab ca. 1750 B.C., do not suggest the most dire conditions of earlier inscriptions. However it is postulated that the cessation of the great floods, after the Egyptians had become accus- tomed to them, required a readjustment of the irri- gation system in a period of political weakness and uncertainty about the proper order of royal succes- sion, thus creating a sort of vicious circle which made the period "darker" than it need have been from either of these factors occurring alone and, under the Egyptian dogma of divine Kingship with the Pharaoh as "rain maker," or more exactly "flood maker," accounts in some degree for the very numerous and short reigns characterizing Dynasty XIII.

INTRODUCTION AND SURVEY OF DYNASTY XII

In a previous article (Bell 1971), I advanced the hypothesis that the First Dark Age in Egyptian history (generally known as the First Intermediate Period) was brought on by a prolonged and severe deficiency in the annual floods of the Nile. The con- sequent famine, amply attested by surviving in- scriptions, precipitated (I suggested) the collapse ca. 2180 B.C. of the central monarchy of the Old Kingdom, which was already weakened by the operation of social and political forces. I introduced

the further hypothesis that this failure of the floods in Egypt-most severe between ca. 2180 and ca. 2135 B.C., and again for a few years between ca. 2005 and 1992 B.C.-was only part of a widespread climatic fluctuation in the direction of greater aridity. The drought was severe. It played a significant and per- haps decisive role in the collapse of many centers of culture which flourished during the Early Bronze Age, all the way from Greece2 through the Near East to the Indus Valley, and so brought on the First Dark Age of Ancient History as a whole.

In the present paper, instead of proceeding imme- diately to consider a later Dark Age, I propose to continue with another theme implicit in the first paper, viz. the intermediate climatic history of Egypt. I propose, that is, to survey the evidence relating to climate from the founding of the XIIth Dynasty by Amenemhet I in 1991 B.C. to the dis- integration of the Middle Kingdom in the 1700s. In Egypt of course the essential climatic factor is the level of the Nile in the season of its annual flood. If the crops are to grow well, the flood must be sufficient to overflow the fields and prepare them for the sowing of the seeds.

Dynasty XII, ca. 1991 to ca. 1780 B.C., was a pe- riod of strong central government and general pros- perity, one of the high points of ancient Egyptian civilization." "When through the success of the state," and (I would add) through the return of the Nile to generous floods, "the 12th Dynasty Pharaohs demonstrated their capacity to be gods, they be- came once more the arbiters and dispensers of ma'at. To this the Egyptian people were assenting. They were well fed and busy and aware of oppor- tunities for advancement" (Wilson 1956:143)-

There was, to be sure, no ecologically significant revival of the rains over the desert, certainly no re- turn of the Neolithic Wet Phase (see Bell

1971). Although rainfall, occurring only in occasional cloudbursts, was too rare to be useful, the Nile inundations were evidently adequate or more than adequate during Dynasty XII. I shall discuss what can be known of flood levels during Dynasties XII and XIII in a major section of this paper. From

2 In the previous paper (Bell 1971:5-6), I noted that the radiocarbon dates associated with the House of Tiles at Lerna (late EH II) corrected to ca. 2500 B.C., suggesting association of the end of EB II in Greece with the end of the Neolithic Wet Phase rather than with the Egyptian I)ark Age. Recently published radiocarbon dates from the EM II phase at Myrotos, Crete (Q-95o/53, RC 12; and Q-0oo2/4, RC 14) average to

good agreement with the Lerna dates, being 2020 B.C. un- corrected and ca. 2500 B.C. corrected. And Korucu Tepe in eastern Anatolia yields two dates (P-1628, RC 13, and M-2376, RC 14) for the end of EB II, which also average close to 2000 B.C. uncorrected.

a New date, required by Hintze's discovery of an inscription dated to year 13 of Amenemhet IV; see note ii.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 4: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 225

Dynasty XII, Vandier (1936) was able to find only one text referring to famine, viz. an inscription in the tomb of one Ameny, Nomarch of Beni Hasan during the reign of Senwosret' I. The Nomarch Ameny states (Breasted 1906:523):

... When years of famine came, I plowed all the fields of the Oryx Nome, as far as its southern and northern boundaries, preserving its people alive, and furnishing its food so that there was none hungry therein.... Then came great Niles, producers of grain and of all things, (but) I did not collect the arrears of the field (taxes) ...

This inscription no doubt gives a picture of the nor- mal situation in a year of low Nile, which must have occurred from time to time throughout Egyp- tian history, though rarely with such severity as in the Dark Ages at the end of the Old Kingdom and again between Dynasties XI and XII. Vandier notes that Griffith believes that the inscription of Mentuhotep, son of Hepi, dated to year 25 of an unnamed king, refers to the same famine as Ame- ny's inscription; but Vandier himself would place it in the reign of Mentuhotep II, while Goedicke (JEA 1962:25-35) argues for a still earlier dating, under Inyotef II, both kings of Dynasty XI (see Bell 1971 :I6-I9).

But such years of famine were not typical of the reign of Senwosret I, which was in general "a pe- riod of great economic development. The provincial cemeteries throughout the country display the very great wealth of the nomes at this time" (Vercoutter 1967:369). And the king himself was able to build extensively, "from Alexandria to Aswan there is no important site where he has not left his trace," with at least 35 sites revealing architectural ruins from his time; his reign was "one of the most glorious in Egyptian history" (ibid).

So also in the provinces. During all of the first four reigns of Dynasty XII-viz. Amenemhet I (1991-1962 B.C.), Senwosret I (1971-1928), Ame-

nemhet II (1929-1895), and Senwosret II (I897- 1878)5-the provincial nobles continued to build fine large tombs in their several territories. Most of these series of tombs come to an end during the reign of the fifth king, Senwosret III (1878-I843), a fact which is usually taken as evidence that Sen- wosret III was able to break the power of the pro- vincial nobility and replace them by administrators subject to direct royal control.

Outside Egypt, the XIIth Dynasty regained firm dominion over Lower Nubia in the reign of Sen- wosret I, who conducted at least two campaigns there, in his year 9 (year 29 of Amenemhet I, io- year co-regency) and in year 18; and he began the construction of a series of forts in the vicinity of the Second Cataract (Emery 1965:141-52; 1967). Nubia apparently remained peaceful during the reigns of his son and grandson, but Senwosret III, a more military-minded pharaoh than his two pred- ecessors, found reason to conduct in person at least four major military campaigns in Upper Nubia. At the time of these campaigns, Senwosret III ordered the building of additional forts and the enlargement of those already in existence, so that they formed altogether a most formidable array, containing fea- tures of military architecture not previously known before the Middle Ages (Emery 1965:I02). Sen- wosret III established Semna-the most important source of data on the flood-levels of this era-as the official frontier of his kingdom, and pacified Nubia so thoroughly that in later centuries he was worshipped as one of the prime deities of the area. It is now thought that the Nubian forts were built as a defensive measure against Kush, an apparent- ly powerful state in Upper Nubia (Trigger 1965; Emery 1965) about which little is yet known, but which the Egyptians must have considered dan- gerous. In addition the forts served to protect river traffic and trade with the south at points (i.e. the cataracts) where boats were particularly vulnerable to marauders, and where cargoes had to be por- taged or towed through; this indeed may have been the primary function of the forts (Adams 1971).

The reign of the sixth king of Dynasty XII, Amen- emhet III (1842-1797 B.C.), son and successor of Senwosret III, is generally considered to be the most prosperous of the Middle Kingdom. Nubia was under firm control. Egyptian power was ac- knowledged by many of the princes of western Asia (Hayes I96I:48), although the precise relationship between these princes and the pharaoh is still quite uncertain. Mining activity was at a high level, as indicated by many inscriptions from the turquoise mines of Sinai and the diorite quarries of Nubia (Hayes 1961). Amenemhet III was able to build for himself two pyramids, one at Dahshur near that of his father and one near the entrance

4Alternately written Senusret; Greek, Sesostris.

" The overlapping dates are correct, reflecting periods of co- regency between the old king and his heir.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 5: Climate and the History of Egypt -The Middle Kingdom

226 BARBARA BELL [AJA 79

to the Fayum Basin at Hawara, although the rea- son he wanted a second pyramid remains unknown. He also erected two colossal seated statues of him- self at Biahmu, which Herodotus reports as lo- cated in the middle of Lake Moeris-a description which has given rise to much controversy, which we shall discuss below in the section on the Fayum. Amenemnhet III has often been credited also with the erection of the Labyrinth, a structure identified by some as the mortuary temple of his pyramid at Hawara, and described by Herodotus as a "won- der surpassing even the pyramids." Actually, how- ever, Herodotus attributes the Labyrinth not to Moeris (Ny-maat-re, Amenemhet) but to a com- mittee of 12 kings immediately preceding Psam- metichus, the founder of Dynasty XXVI in 664 B.C.6

NILE LEVELS FROM THE MIDDLE KINGDOM

Reign of Sentwosret I (1971t-928 B.C.). In a note elsewhere (Bell 1970), i discussed the most an-

cient known records of the Nile floods, which cover in a fragmentary way the period from ca. 3050 to ca. 2480 B.C. These records indicate that the floods averaged 0.70 m. (or more, depending upon the assumption about the zero of the scale) higher in the vicinity of Old Memphis (near Cairo) during Dynasty I than in subsequent centuries. The next known figures date from the reign of Senwosret I of Dynasty XII, and state that a "good flood" had a level of about 21.5 cubits (11.3 m.) at Elephantine (Aswan), 12.5 cubits (6.6 m.) at the "house of the Inundation" near Old Cairo, and 6.5 cubits (3.4 m.) at Diospolis in the northern Delta (Kees 196i:50; JEA 30:34). The highest surviving figure from the earlier records is about 8 cubits (4.2 m.) and the average from Dynasties II-V is 1.8 m. Because there is no reason to believe the floods were higher in the time of Senwosret I than in the early period, but on the contrary some reason-viz. the location of the valley temples associated with the royal pyra- mids-to believe they were lower, it seems clear

6 The dating of the Labyrinth remains controversial. Argu- ments for the late date have been presented most recently by K. Michalowski (1968 JEA 54:219), while A.B. Lloyd (1970 JEA 56:8I-Ioo) attributes the Labyrinth to Amenemhet III.

The following description of the results of his explorations of the site of the Labyrinth by Lepsius (1853:83, 89-91) may be of interest, particularly to classical scholars interested in assessing the reliability of Herodotus on matters Egyptian (see also note 37):

"Here we have been, on the southern side of the Pyramid of Moeris, since the 23rd May, and are settled among the ruins of the Labyrinth; for I was certain from the first, after we had made but a hasty survey of the whole, that we are perfectly entitled to designate them under this name: I did not, how- ever, imagine that it would have been so easy for us to be- come convinced of this . . I caused some excavators to be levied from the surrounding villages . . . and ordered them to make trenches through the ruins, and to dig at four or five places at once

.. "These lines are written to you from the distinctly recognised Labyrinth of Moeris and the Dodecarchs, ... An immense cluster of chambers still remains, and in the centre lies the great square, where the courts once stood, covered with the remains of large monolithic granite columns, and of others of white hard limestone, shining almost like marble. . . . At the first superficial survey of the ground, a number of complicated spaces, of true labyrinthine forms, immediately presented them- selves, both above and below ground, and the eye could easily detect the principal buildings, more than a stadium (Strabo) in extent. Where the French expedition had vainly sought for chambers, we literally at once find hundreds of them, both next to, and above one another, small, often diminutive ones, beside greater ones, and large ones, supported by small col- umns, with thresholds, and niches in the walls, with remains of columns, and single casing stones, connected by corridors, without any regularity in the entrances and exits, so that the descriptions of Herodotus and Strabo, in this respect, are fully justified. But at the same time also, the opinion, which was never adopted by me, and is irreconcilable with any archi-

tectonic view, that there are serpentine, case-like windings, in place of square rooms, is decidedly refuted.

"The whole is so arranged, that three immense masses of buildings, 300 feet broad, enclose a square place, which is 600 feet long and 5oo feet wide. The fourth side, one of the narrow ones, is bounded by the Pyramid, which lies behind it; it is 300 feet square, and therefore does not quite reach the side wings of the above-mentioned masses of buildings. A canal of rather modern date, passing obliquely through the ruins . . . cuts off exactly the best preserved portion of the labyrinthian chambers, together with part of the great central square, which at one time was divided into courts . . . . the chambers lying on the farther side, especially their southern point, where the walls rise nearly ten feet above the rubbish, and about twenty feet above the base of the ruins, are to be seen very well even from . . . the eastern side; and viewed from the summit of the Pyramid, the regular plan of the whole design lies before one as on a map. Erbkam has been occupied ever since our arrival, in making the special plan, on which every chamber or wall, however small, will be noted down...." (see Lepsius' Denkmdler V.1, Abt. I, BI. 46- 48).

"... The fragments of the mighty columns and architraves

which we have dug up from the great square of the halls, exhibit the name-shields of [Amenemhet III]. . . . We have several times found the name of [Amenemhet III in a chamber which lay in front of the Pyramid beneath a great quantity of rubbish]. . ... The builder and occupier of the Pyramid is therefore determined. But this does not refute the statement of Herodotus, that the Dodecarchs, only 200 years before his time, had undertaken the building of the Labyrinth. We have found no inscriptions in the ruins of the great masses of cham- bers which surround the central space. It may be easily proved by future excavations that this whole building, and probably also the disposition of the twelve courts, belong only, in fact, to the 26th Dynasty of Manetho, so that the original temple of [Amenemhet III] formed merely part of this gigantic archi- tectural enclosure."

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 6: Climate and the History of Egypt -The Middle Kingdom

19751 CLIMATE AND THE HISTORY OF EGYPT 227

.A .ARA S

INAI

..i......ii . :.ii ij•i

.... ......... .......................

ASY T

f. AY.. M ENI .

BDO THEBES

EL KAB 25::iiiii- • ! ii:

ELEPHANTINE 3N

/SLo

/ CATARACT 0/OR/RE

OUARRY KUN K

TOSHKTA AB SIMBEL WoBI HALFA

SCATARAC KERMA -- ARGO K

DONGOLA* JEBAELBE

NAOTA MEROWE 5/h CATARAC

KHARTOUM

MALA LAL 0 10,

A USo MBE .........

MAP I. The Nile Valley and vicinity: open circles, ancient sites; filled circles,

modern towns

that we must infer that the zero point of the scale, if it was ever fixed, was changed at some time be- tween the Vth and XIIth Dynasties. It is most reasonable to suppose that this change, a lowering of the zero-point, occurred during the First Inter- mediate Period (First Dark Age) when, according

to the evidence of surviving inscriptions, some years of Niles so low as to cause severe famine occurred, particularly between ca. 2180 and ca. 2135 B.C. As has been argued (Bell 1971), this drought in all probability precipitated the collapse of the Old Kingdom. The floods ca. 2500 B.C. on the old scale averaged around 1.8 m. (or 1.3 m.) above the un- known zero, so that the water must have failed to reach even the zero level in some of the famine

years. This, together with the collapse of the cen- tral government, may well have led to the abandon- ment of the old Nilometer and the subsequent adoption of a new one with a lower zero point.

We should note the fact that the figures given by Senwosret I describe a "good flood," not any par- ticular actual flood. It is reasonable to believe, however, that in fact the floods were "good," for Egyptologists agree that the reign of this king was a time of high prosperity for Egypt. Moreover his father, Amenemhet I (g1991-962, including io years co-regency) claimed in his Instructions to his son (Wilson 1955:418; see also Bell 1971:18-

20): "The Nile honored me on every broad ex- panse"; that is, the inundations were liberal.

The figures of Senwosret I for the vicinity of Cairo (6.6 m.) would be the same as the average rise of 6.6 m. from minimum to maximum at the Roda Nilometer (also near Cairo) over the period from the seventh century A.D. to 1890, as given by Popper (1951:225); the average rise for the

years A.D. 642-1521 he gives as 6.5 m., while his century-averages range from 6.i to 6.94 m., and the years A.D. 1822-1891 give 6.74 m. Thus it is reasonable to infer that the floods in the vicinity of Cairo in the early part of Dynasty XII were close to those of recent centuries, with the zero-point of the ancient scale being at or near the average LWL (low water level). But see note 22.

By a similar interpretation of Senwosret's figure from Elephantine (Aswan), however, rather higher floods would be indicated. The range for the years A.D.' 1870-1902 is 8.0 m., and the highest Io-day mean (in 1878) is 9.0 m. above the average LWL

(low water level), compared with a value of I1.3 m. given by Senwosret I. Extrapolation8 of the present Aswan gauge-discharge curve, shown in ill. i

7 Numerical data on modern gauge levels and flood volumes, unless otherwise credited, were obtained from The Nile Basin, vols. II, III, IV, and their supplements, produced by H.E. Hurst, assisted in the earlier years by P. Phillips, and later by R.P. Black and Y.M. Simaika, published by the Ministry of

Public Works, Nile Control Department, Cairo, 1933-1963. 8 A word of caution is in order. Throughout the paper, when- ever extrapolations are used, it should be borne in mind that these can be no more than rough approximations, particularly as I have a cross-section of the channel to the highest relevant

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 7: Climate and the History of Egypt -The Middle Kingdom

228 BARBARA BELL [AJA 79

NILE VOLUME, m3 x 108 per day 2 4 6 8 10 20 40 60 80

100//

SE MN A, M K -i - - -

~ 95 E

x F1878

8

.JW ,' --1946- w - AV 1870-99 o

z 90< 1913

SHWL

LWL

85

-

-- AV 1870-99 --

- 1900-I

3 5 I 2 3

m3 x 108 per day

ILL. I. The Nile at Aswan: gauge levels plotted against volume of discharge, from Hurst et al (1946:123) and extrapolation

(dashed lines) above level of modern floods

(from Hurst, Black, and Simaika 1946:I23), sug- gests that a flood of I1.3 m. above LWL at Aswan would have a peak volume of 14x to I6xio"m.3/day,9 about double that of the average for 1870-1940 (Io- day mean - 8.4xio'm.'/day) and some 30 to 40 percent greater than that of the great flood of 1878

(II.4xio"m.'/day) which rose above the average LWL to 9.0 m. at Aswan and 8.36 m. at Roda (Cairo).

This sort of discrepancy between flood levels at Elephantine and at Cairo is found in Greek and Roman times as well. It has been discussed by L. Borchardt in his monograph "Nilmesser and Nil- standmarken" (see also Kees i96i:50). The dis- crepancy could be explained (Lyons i906:315) by assuming that the Egyptians used a zero point be- low mean LWL at Elephantine, which would be a more feasible thing to do among the rocks of the

Aswan cataract than in the alluvial valley of Lower

Egypt, although of course a quayside scale could also perfectly well have a zero below (or above) LWL." It is also possible, even probable, that the zeros of the two Nilometers were fixed at different times. Since Dynasty XII originated from Thebes, which was the capital of the preceding (XIth) Dy- nasty, one could easily conceive that the Elephan- tine Nilometer was established at an earlier time when the LWL may have been lower. If we assume that the "good flood" of Senwosret I was similar to the larger modern floods, which amount to about

i xxo"m."/day, with HWL at about 94.o m., then the zero of the gauge would have been at 82.7 m. Such a zero may have been fixed arbitrarily below the average LWL, but it is tempting to imagine that it was set in a period of exceptionally low LWL such as that just preceding the establishment of Dynasty XII by Amenemhet I in 1991 B.C., when according to the Prophecy of Neferty (Er- man 1927; Bell 1971:71): "The river of Egypt is

empty, men cross over the water on foot." There is however another factor in the problem

of interpreting the levels given by Senwosret I, a factor which in pharaonic times must have served to enlarge the difference between the Elephantine and the Memphis/Cairo readings. This factor is the lake in the Fayum basin. It is generally agreed that since Ptolemaic times this lake has taken off only an insignificant fraction of the Nile flood- waters. But earlier, from pre-Neolithic down to Ptolemaic times, according to some authorities

(e.g. Petrie 1889; Ball 1939), the lake was in more or less free connection with the Nile through the Hawara channel, and served to reduce flood levels in Lower Egypt by draining off a portion of the floodwaters. Water would then return to the river in the low season (Ball i939), and the action of the lake would diminish the range at Cairo, and

everywhere downstream of the Hawara channel near Beni Suef, relative to that at Elephantine.

The level of the Fayum lake in pre-Ptolemaic an- cient times has long been a subject of controversy

levels only for Semna (from Lyons 19o6:26o; and Ball 1903). If the actual channel width suddenly increases at some gauge level within the range of interest, the extrapolations will under- estimate the volume of flow.

9 Explanation of symbols: m' = cubic meters; 1o0 = hun- dred millions = Ioo,ooo,ooo (eight zeros); i6xio8m'/day = sixteen hundred million cubic meters per day. LWL = low water level; HWL = high water level or height of the flood.

10 However Lyons's further idea of a gradient of zero-points differing from that of the river is a rather inherently implausible idea which should not be accepted without a re-examination of the floods throughout the centuries and millennia. Toussoun of the data free from any predisposition to assume a constancy (1925:265) considers it evident that the scales of the various Nilometers found throughout Egypt have no deliberate rela- tion to one another.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 8: Climate and the History of Egypt -The Middle Kingdom

1975 1 CLIMATE AND THE HISTORY OF EGYPT 229

among scholars. I shall discuss it at some length later in this paper because it is an integral part of the climate history of Egypt and can be discussed most appropriately in connection with the Middle Kingdom. We shall there return also to the Nile levels of Senwosret I, and consider to what extent the diversion of water into the Fayum could ac- count for the apparent discrepancy between Ele- phantine and Old Cairo. But first we shall com- plete our survey of the information available concerning the Nile flood levels during the Middle Kingdom.

Evidence from Nubia, HWLs (High Water Levels). Rather more clear and tangible evidence on Nile flood levels is available from the latter part of the Middle Kingdom, between 1840 and ca. 1770 B.C., which yields quantitive measurements of flood levels relative to modern HWL and LWL. The interpretation of this evidence has been sub- ject to much discussion and controversy, leading to divergent views on the level of the Nile floods during the Middle Kingdom. The controversy origi- nated when Lepsius in 1844 discovered a series of inscriptions, on the rocks overlooking the river at the SEMNA region of the Second Cataract, that ap- parently recorded flood levels from the late Middle Kingdom. Vercoutter (1966) re-analyzed the phras- ing of the inscriptions and concluded that they must undoubtedly be considered meaningful rec- ords of high water levels.

Most of the inscriptions, scattered here and there on rocks of both the east and west bank, come from the reign of Amenemhet III (1842-1799 B.C.), but we now have also four from Amenemhet IV and one from Queen Sobekneferu, the last two rulers of Dynasty XII; also four from Sobekhotep I, and two from Sekhemkare, the first two kings of Dy- nasty XIII. Fifteen flood levels from the east bank have been published by Dunham and Janssen (I96o:Plate XXXII), based on the work of Lepsius and of Reisner. In ill. 2 these are plotted against date. Most of the inscriptions from the west bank are on rocks that have fallen from their original

WAD I BUHEN HA

KOR. MEINARTI eDORGINARTI

ROCK OF ABUSIR A

MIRGISSA DABENARTI

GEMAI

MURSHID

ASKUT KAJNARTY

SHELFAK

S URONART

SEM NA • uM

Ao SKUMMA 0 0 20

SEMNA kilometers SOUTH

MAP 2. Sites in the region of the Second Cataract (adapted from Vercoutter 1970: Fig. i)

positions (Reisner 1929a, b) so that they can give no quantitative information; these years (from Dunham and Janssen i96o:129f) are shown in ill. 2 by the broken bars, drawn arbitrarily to the average height of the solid bars. F. Hintze (1972 Jan I6, priv. comm.)"1 has generously made available to me for use in this paper the results of his re-examination of the Semna inscriptions and his additions to the list are shown by the thinner broken lines and included in the total numbers given above. Particularly noteworthy are Nos. 501 and 508, respectively dated to year 13 of Amenem- het IV and year 8 of Sekhemkare, the highest dates known for these kings. The modern floods, at the time of the visit of Lepsius, averaged about 7.3 m. lower than those recorded in the late Middle Kingdom, and from this has arisen the "Semna problem."

Before proceeding with the Semna problem itself, however, we should note that the question of the

11 The list provided me by Professor Hintze (Humboldt- Universitiit zu Berlin) contains his reference number, the refer- ence number assigned by Reisner (I)unham and Janssen 1960) in case of revised translations, and the date of the inscription, as follows:

5oI: (new) Amenemhet IV, year 13 502: ( =RIS 6) Amenemhet III, year 36 503: (z.Zt. nicht greifbar in unserm Archiv) 504: (?--RIS 8) Shm-k'-Re [Sekhemkare], year 4 505: ( =RIS I6) Amenemhet IV, year 5

506: ( =RIS 19) Amenemhet IV, year 6 507: ( =RIS 18) Amenemhet IV, year 7 508: ( =--RIS Io) Sekhemkare, year 8 509: ( =RIS 2) S[m-Re-bw-t;wj [Sekhemre Khutowy]

(Amenemhet-Sobekhotep), year 2 510: ( =RIS 3) Amenemhet-Sobekhotep, year 3 511: ( =RIS 8) Sdf'-k;-Re, year I [probably Sedjefakare,

in the list of Hayes (1962a) the ninth king of Dynasty XIII,

c.1765_5].

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 9: Climate and the History of Egypt -The Middle Kingdom

230 BARBARA BELL [AJA 79

YEARS BC 1840 1830 1820 1810 1800 1790 c 1770 REGNAL YEARS 2 12 22 32 42 2 4 8

24 _

I 5 9 13 I 5 KUMMA TEMPLE AMENEMHET III -IV------ DYNASTY XIII -K U MMA TE MPLE

22 -60

A 20- O MK, MEAN HWL

-""I. 18 -"

16"" " I> 14 I "" I < II I I : I oI

Crr:

12-HWL, LEPSIUS

I=50> "I II 50

o=--HWL 1901, BALL " : : IXII

.. .?? !. ,•

I"II I

::: : r

ii I IIw

2 11" I I

-140

"+ " I

0-LWL, LEPSIUS

-2 LWL, BALL AD-20.t C...

A D 2 0 th C ................ N .-

ILL. 2. The Semna inscriptions: flood levels plotted against year: shaded bars, in situ inscriptions; dashed lines, fallen inscriptions plotted to average

height of in situ inscriptions (from Dunham and Janssen i960); dotted lines, Hintze inscriptions (see n. i i); A, Askut inscription, estimated height at Semna

average modern flood level itself is not without

complexity. John Ball (I903), from his visit to Semna in March 1902, reported that "there can be no doubt of the correctness of the difference of level noted, as the people of Kumma pointed out the precise spot where they go to get water at high Nile. My observation ... confirms that of Lep- sius." "The precise spot" must vary from year to year, however, with the variation in the volume of the modern flood, and indeed the HWLs de- termined by Ball and by Lepsius actually differ by 1.68 m. while the LWLs differ by 1.94 m., the levels given by Ball being lower in each case. This need

cause no disquiet, for the floods of 1843 and of 190o, preceding the respective visits of Lepsius and of Ball to Semna, measured 19.Io and 18.78 m. on the Roda gauge near Cairo, in satisfactory ac- cord with the difference in their levels at Semna. Table i presents a summary of the levels12 mea- sured by Ball and by Lepsius and relates them to absolute levels above mean sea level using the figures given by Vercoutter (1966: I40 n. 43).

In the report of his visit to Semna at the begin- ning of this century, Ball (1903) proposed to at- tribute the apparent decline in flood levels from the Middle Kingdom to modern times to an erosion of

12 Ball (1903) does not give the mean of the Middle Kingdom levels at Semna, but rather a value of I20.8 or 7.9 m. above his HWL of 1901, which represents the "lowest group of in-

scriptions." Fortunately for greater precision, he gives the level of year 23 of Amenemhet III as Io.9 m. above his HWL, which in turn is 12.i m. above his modern LWL; from these data the relation between the scales of Ball and of Lepsius can be worked out. Vercoutter (1966:n.43) reports actual measure- ments only for the two temples, where his figures show a dif- ference between the levels of the temples which is less than half of that published by Lepsius; it is clear that Vercoutter's "modern LWL" was determined from the Kumma elevation using the scale of Lepsius. The elevations of Kumma de- termined by Ball and by Lepsius are in good accord, while Ball gives yet a third elevation for the Semna temple. It is not

surprising that the Kumma elevation was the more accurately measured by Lepsius because all of the in situ inscriptions- the focus of his interest-are on the Kumma or east bank.

Lepsius's error in the elevation of the Semna temple pro- duces an initially puzzling result in Reisner's Plate IX (Dun- ham and Janssen 1960), where the water level of 5 April 1924-which was about 115.25 at Halfa, close to the average modern LWL-appears at 3.90 m. above the zero of Lepsius! From Table i, it should be around -1-.9 m. It would appear that Reisner determined the temple platform to be 34.0 m. above the 5 April 1924 water-level, then used Lepsius's figure of 37-90 m. for the elevation of the temple. If we use Ball's elevation (135-4 m. - 102.9 m. - 32.5 m. on the scale of Lepsius) subtraction of 34.o0 m. gives --I.5 m., in satisfactory agreement with the expected value.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 10: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 231

TABLE 1

Comparison of Nile levels at Semna, in meters, according to the various scales of Ball (1903), Dunham and Janssen (1960, from 1844 observations of Lepsius),

and Vercoutter (1966: I40, n. 43, meters above mean sea level).

Lepsius- meters above Levels Ball Dunham-Janssen sea level

"modern" LWL, Lepsius (1o2.9) 0 138.92 March 1902 LWL* o00.8 (-I.94) 136.98 "modern" HWL, Lepsius (113.8) 11.84 150.76 HWL of 1901 112.9 (0o.i6) 149.08 difference, HWL - LWL 12.1 11.8

Mean Middle Kingdom HWL** (121.9) 19.14 158.06

Year 23 of Amenemhet III 123.8 21.o6 159.98 M.K. mean - modern HWL 9.0 7.30 M.K. mean - modern LWL 21.1 19.14

Base of Temple Kumma 126.0 23.03 161.95 Semna 135-4 37-90 168.90 Semna - Kumma 9-4 14-9 7.0

* In March 19o2 the volume at Aswan was r-, o.52xiosm.3/day (Hurst et al. 1946), intermediate between the average minimum of 1870-1899 (o.58xios) and of 1900-1938 (o.415xIo8); and nearly 2 m. below the LWL determined by Lepsius.

** See n. 12, first two sentences.

the river-bed by some 8 m. in the Semna cataract. With the limited knowledge then available on ero- sion rates, he was able to make a plausible case, and his conclusions were reaffirmed by Sanford and Arkell in 1933-

The erosion hypothesis seemed to be supported by clear evidence that the low water levels in the Middle Kingdom, at least in some years, were close to the levels of today. The various forts built in the vicinity of the Second Cataract by pharaohs of the Middle Kingdom each had a stairway in a pro- tected position leading down to the Nile. Wher- ever these water-stairs are well preserved they ex- tend approximately to the modern LWL, thus pointing to a LWL in the Middle Kingdom close to recent levels. We shall discuss the water-stairs more fully in the section on LWL.

After a re-examination of the geological evidence at Semna, however, Fairbridge (1963) rejected Ball's explanation for the Semna flood records. He concluded that very little erosion had occurred in the Semna cataract since the Middle Kingdom and

that the floods in that period were in fact around 8 meters higher than they are today. He concluded further that the Nile had reached bedrock and ceased any sort of rapid cutting already by ca. 3000 B.C., and that the Semna dike had been cut mostly before the silt lens was deposited in Upper Paleolithic times. Butzer (1971) agrees "that Ball's rates of bedrock incision in diorite are quite incon- sonant with modern observations and that the channel floor must have been cut to substantially its present level long before the Middle Kingdom."

Even before the i960s, however, there emerged various difficulties with Ball's hypothesis quite apart from his over-estimate of the erosion rate. Reisner (1929b) rejected Ball's hypothesis because of a lack of evidence for levels significantly higher than today's in the New Kingdom at Semna, which would imply that all the erosion occurred between 1840 and ca. 1560 B.C. Reisner proposed a different sort of erosion hypothesis involving the sudden collapse of part of the cliff and a consequent widen- ing of the river channel. The amount of widening

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 11: Climate and the History of Egypt -The Middle Kingdom

232 BARBARA BELL [AJA 79

possible, however, appears by no means sufficient to account for so large a change in flood levels.

Moreover the evidence for high floods in the Mid- dle Kingdom is not confined to the Semna inscrip- tions nor to the immediate vicinity of the Semna cataract. Ancient fields associated with the forts at Uronarti and Shelfak are 7 and 6 meters respec- tively above the modern HWL (Wheeler 196i:96). According to Wheeler (i96i:96), who took a lively interest in evidence bearing on ancient flood levels, "It would appear possible that the ancient HWL came over the present river bank [at Mirgissa , and to within a short distance of the foot of the Fort itself. There is mud in the sand here which could not have arrived by other means. . ." In his notes on his 1931-1932 excavations at MIRGISSA, Wheel- er (1961:165) summarizes threefold evidence on the question of the HWL in ancient times:

i. The water-worn surface of the rock is very clear- ly defined at a level of 8.73 m. above 1931 HWL. From this level down to the lowest cleared (4.15 m. above 1931 LWL) the rock is deeply water-worn-all the veins of harder rock standing out from the surface, well polished. The upper limit of this wear, taken at different points, always gives the same level.

2. That part of the rock which was submerged has a surface of salt crystals, which are very thickly distributed from the lowest level up to about 6 m. above 1931 HWL.

3. At a slightly lower point, 6.23 m. above 1931 HWL, there is river mud and sand tightly packed into a pocket in the rock. At the same point was found, wedged in a crack of the rock, a small potsherd, a water-polished pebble, and a small fragment of bone, hardened almost to semi-fossilization by water.

The fluvial planation observed in item (i), how- ever, must have required many centuries, even mil- lennia, and cannot possibly reflect merely the floods recorded at Semna. Most of it probably developed in the fourth and fifth millennia B.C., when there is other evidence for floods some 5 to io m. above the modern HWL (Trigger 1965:29). Regarding item (2), salt can move laterally and vertically through rock by capillarity (Butzer i97i) so this point is inconclusive. Few would dispute Wheeler's conclusion, now supported by evidence from many

parts of the valley, that the Nile flood levels were at some time in the post-glacial past, and for a con- siderable period of time, several meters higher than they are today. Since Wheeler found it "equally certain . . . that it was there during the occupation of the fort," we may suspect that he was strongly influenced by the Semna inscriptions, for a potsherd and bone fragment of unspecified type cannot be considered alone as compelling evidence.

Modern excavations, set in motion in the 1950s by the impending disappearance forever of many sites beneath the waters of Lake Nasser, have brought to light a number of items relevant to the interpretation of the Semna inscriptions and water levels in the Middle Kingdom. Others may be ex- pected as more excavations become fully published.

In the course of their comprehensive excavations at Mirgissa, Vercoutter and his associates of the French Archaeological Mission to the Sudan dis- covered a "Lower Fort" or at least a substantial segment of fort-wall extending to a level some 15 m. below the earlier-known Fort on the top of the hill, and close to their contour"' 155 m. (Vercoutter 1965: fig. I; 1970). On stylistic grounds they con- clude that this fort was built in the earlier part of the XIIth Dynasty, by Senwosret I or his successors, while the well-known Upper Fort was built by Sen- wosret III (Vercoutter 1970:20-22), although the principal occupation remains are from the early part of Dynasty XIII, ca. 1770 B.C. Most important for our purposes, they conclude that "all the earliest structures of the site below the contour level 155 have been washed out by an uncommonly high flood which occurred between the end of the Mid- dle Kingdom and the beginning of the New King- dom" (Vercoutter 1971 personal comm.; see also the chapter by A. Hesse in Vercoutter 1970:51-67). Hesse also reports evidence of structures interpreted as riverside facilities, at a contour level of 149, and this elevation, he concludes, marks the normal HWL of the Middle Kingdom. Hesse (Vercoutter 1970:53) points out that the water damage of struc- tures up to 155 m. is persuasive evidence that such high floods were not typical of the Middle King- dom, contrary to the opinion of O.H. Myers and of Wheeler. The Egyptians would not have built at

13 Vercoutter's (1965: fig. I; 1966:61; 1970) contour levels re- late to a different zero than that used by the gauges of the Nile Control Department because of the pressure of time and lack of a reliable benchmark in the vicinity of Mirgissa to fix the zero; they estimate (Vercoutter 1970:51) that their levels are

too high, relative to a sea-level zero, by 7.8 m. Study of Table 2 and maps of the area indicates to me that the discrepancy with the Kajnarty gauge levels is closer to II to 12 m., but it is possible that some of the discrepancy results from an error in the Nilometer zero.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 12: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 233 levels they could expect to be flooded at HWL, and thus the destruction up to 155 m. must have been caused by an exceptional, a truly extraordinary flood "une crue millinaire," some 6 m. above that typical for HWL in the earlier part of the Middle Kingdom.

It is puzzling that Vercoutter and his colleagues, confronted with this evidence from their site at Mirgissa, nevertheless resist the obvious solution of identifying their "crue millenaire" with the remark- able floods recorded at Semna, apparently because of a lack of independent non-archaeological evi- dence for the climate fluctuation that would be im- plied by such in interpretation. The evidence from Mirgissa, interpreted in conjunction with the Sem- na inscriptions, suggests that during much of the XIIth Dynasty the HWL was around 149 m., sim- ilar to that of the late nineteenth century A.D.14 Although the dating is not absolutely secure, Ver- coutter (1970:20f) concludes that the lower struc- tures at the site were built during the reigns of Senwosret I, Amenemhet II and Senwosret II, while the long-known Upper Fort was built under Senwosret III. All of this occurred before the ear- liest flood inscriptions at Semna, which began only with the reign of Amenemhet III. The channel width at Mirgissa at HWL is around 1700 m. (Ver- coutter 1970:56), and about 400 m. at Semna, so that we should expect the floods to be higher in the latter location; the highest floods, some io m. above the modern at Semna, would be only about 5.7 m. above the modern at Mirgissa." We shall return to this point in a later section. The evidence from Mirgissa of water-destruction up to level 155 m. however appears entirely compatible with interpre- tation of the Semna inscriptions as records of actual

flood levels produced by a short-lived climate fluc- tuation of a few decades which resulted in a num- ber of floods of a magnitude more typical of the Neolithic Wet Phase than of historic times.

Turning now to evidence from other sites, one of the most important items, discovered by A. Ba- dawy (1964 Kush 12:52), is a rock inscription at ASKUT, roughly midway between Mirgissa and Semna, at the north end of the Sarras region where the channel width is comparable to that at Semna at least up to modern flood levels (Lyons 1906:260). Close by is the modern gauge-station of Kajnarty which we shall use presently in an attempt to esti- mate the flood volumes recorded during the Middle Kingdom at Semna. The inscription at Askut re- cords a HWL dated to year 3 of Sekhemkare, the second pharaoh of Dynasty XIII. Badawy points out that it is important for its "very high level" and because it proves that inundations were still re- corded and that Upper Nubia was still under Egyp- tian rule.'" Vercoutter (1966:140, n. 41) reports that the Askut inscription is located at 17.4 m. above the Nile level of 1963 December 16. The level of the Nile at nearby Kajnarty on this date, kindly pro- vided me by the Sudanese Ministry of Irrigation, is 134.54 m., whence the level of the inscription is 151.9 m. From Table 2 it can be seen that this level is about 0.5 m. above the average estimated for the Middle Kingdom inscriptions in situ at Semna (151.4 m. at Kajnarty). The Askut inscription is represented in ill. 2 by the point labelled "A." This inscription is a severe, indeed fatal, embarrass- ment to the erosion hypothesis.

Vercoutter's (1966) excavation of a secondary fort at Semna South has provided clear physical evidence for very high floods at some period after

14 It appears impossible to obtain an exact relation between the 1931 HWL determined by Wheeler and the various fea- tures excavated by Vercoutter and his associates. Vercoutter

(I971 pers.comm.) estimates that Wheeler's level of 7.44 m. above the 1931 HWL is close to his contour 155 m., which would place the modern HWL at 147.5 m. The average HWL of the current century coincides closely with that of 1931; if this falls about 1.5 m. below that judged typical of the Middle Kingdom, the latter may be imagined as similar to the floods of the years A.D. 1870-1898, and almost certainly no larger than the largest floods of the past century.

However Hesse (Vercoutter 1970:54) determined the modern HWL to lie between 145 and 146 m. from two other mea- sures reported by Wheeler. Further, he informs me (Hesse 1972 pers.comm.) that the base of the Upper Fort was so eroded that it was impossible to identify exactly the levels de- scribed by Wheeler and therefore necessary to rely on the other data given by Wheeler. Lacking evidence to the con- trary, Hesse assumes the typical Middle Kingdom flood vol-

ume was equal to the modern average and attributes the dif- ference between 145 and 149 to a change in the riverbed. Unfortunately most of the lower-level structures at Mirgissa were discovered only in the last season before they were flooded by the rising waters of Lake Nasser, so that they could not be re-examined after their importance as evidence on ancient Nile levels was fully appreciated (Hesse 1972 pers.comm.).

15A rise of 9 to io m. at Mirgissa, corresponding to the difference between the "crue millenaire" and a modern HWL at 145 to 146 m. is considerably more than could be expected from a Io m. rise at Semna. This fact tends to support Hesse's postulate of a decline of 2 to 4 m. in the level of the riverbed at Mirgissa since the Middle Kingdom, assuming that his de- termination of the modern HWL is to be preferred over Ver- coutter's estimate (see note 14). In addition, there is evidence that the river flowed closer to the Upper Fort in the Middle Kingdom than it does today.

16 We have now in addition, from Hintze (note II) years 4 and 8 of Sekhemkare at Semna, and year I of Sedjefakare.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 13: Climate and the History of Egypt -The Middle Kingdom

234 BARBARA BELL [AJA 79

the building of the fort. He infers further that it indicates the absence of such ultra-high floods at the time the fort was under construction, but I am less certain than he that the evidence compels this conclusion. At the fort of Semna South, built on "an ancient alluvial terrace of the Nile" about i km. south of the main Semna Fort, Vercoutter found a glacis about io m. wide, sloping toward the river and built of granite slabs. Beneath it was a subter- ranean water-stairway built also of granite slabs and running as a tunnel through the ancient alluvium from the northeast corner of the fort towards the river. Most significantly, the glacis was covered by a thick deposit of Nile silt-a typical water-laid silt (Butzer 1971)-up to a level of about 8 m. above the modern HWL. This silt layer, Vercoutter points out, was deposited at some time after the building of the glacis and the stairway and, there being top- ographically no possibility of local silting from re- peated flooding of a wadi, indicates "a rise of some importance in the Nile level after the building, leading to a heavy silting in places where the sedi- ments could not be washed out by the current dur- ing the subsidence of the river." Only a few arti- facts from the Middle Kingdom were found in the Fort, from which Vercoutter concludes that it was not permanently occupied. He believes it was most probably constructed during the reign of Senwosret III, who is known to have campaigned actively around and above the Second Cataract, to have en- larged and strengthened existing forts and added others to the formidable array. From Vercoutter's (1966) Figure 4, it would appear that the stairway led below the modern flood level, although lack of time and various difficulties prevented him from excavating it completely to the lower end, which by analogy with other forts of the region should be near modern LWL.

The silt-covered construction gives proof, Ver- coutter concludes, that at some time in the Middle Kingdom the flood levels were similar to or slightly less than those of modern times-close to the appar- ent convergence point of glacis and tunnel-roof- and that subsequently there occurred a period of floods some 8 m. higher. The second conclusion, in excellent agreement with the flood-level inscriptions

on the rocks at Semna and Kumma, beginning early in the reign of Amenemhet III, seems wholly compelling. But I am not convinced that Vercout- ter's excavations provide quantitative evidence on the HWL at the time of the construction of the Fort. In order to construct the water-stair, the Egyp- tians would presumably have dug a trench in the ancient alluvial terrace, built the stairway, covered it over and built the protective glacis."7 If they com- pleted this work in one season, I see no basis for any firm conclusion about the HWL at the time of the construction. Since it was not possible to exca- vate the stairs to their end, they cast no light on the LWL, so it is doubtful that Semna South has yielded any firm evidence against the erosion hy- pothesis-which, however, is already convincingly demolished by modern knowledge of erosion rates, and by the archaeological evidence from Mirgissa and Askut. The evidence from Vercoutter's (1970) excavations at Mirgissa also provides more persua- sive evidence that the HWL in Dynasty XII before the reign of Amenemhet III (dated from Semna) was similar to the modern.

Convinced that his excavations disproved the ero- sion hypothesis, Vercoutter (1966) was not inclined to accept the most straightforward (at least to me) alternative of climate fluctuation in absence of inde- pendent evidence for such a climate fluctuation. He postulated that the high levels were caused by a partial dam or rather, because of the Askut inscrip- tion,"8 a series of partial dams, built by the Egyp- tians of the Middle Kingdom with the object of prolonging the period of high water when the cata- racts were relatively navigable. His reasons for proposing this, to me incredible, explanation relate also to the evidence on low-water levels, plus his assumption, which I shall argue is completely un- justified, that a rise of 8 m. in the HWL must be accompanied by a corresponding rise of 8 m. in the LWL.

But first we shall finish with the evidence on HWL. Also relevant here are one (and perhaps two) items found in excavations by the Oriental Institute archaeologists (Knudstad 1966) at SERRA EAST (and Dorginarti), although no quantitative conclusions can be drawn until (unless) details on

17 Some of the 8 m. of silt overlying the glacis may have come from the trench, piled to the sides of the construction and redistributed smoothly by the series of great floods re- corded at Semna.

18 Vandersleyen (1970) proposes to attribute the single in- scription at Askut in year 3 of Sekhemkare to a great flood

which resulted from the rupture of Vercoutter's postulated dam at Semna. This hypothesis, while more attractive than the postulate of a series of dams, is nevertheless invalidated by the inscriptions of Hintze (notes 11, 16), three of which record great floods at Semna at dates later than the Askut inscription.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 14: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 235

elevations become available. At the Serra East fort, apparently built originally in the reign of Senwosret III, is a constructed basin, presumably a harbor. The full depth of the silt-filled harbor could not be excavated, "due to the height of the Nile [in De- cember 1963, 119.6 to i I9.75 m. at Halfa, courtesy of the Sudanese Ministry of Irrigation] and conse- quently of the water table in the silt filling the har- bour" (pp. 173-74). From study of the deposits with- in the accessible layers of this silt the excavators concluded "that the harbour was allowed to silt up possibly from its late Middle Kingdom abandon- ment through transitional periods until the New Kingdom period of return; that it was allowed to remain so during the New Kingdom, and that it was thus not a feature of the later fortress, having been filled to a high and dry level by that time." It is tempting to associate this silting of the harbor with the siltation phase exhibited at Semna South, and to surmise that the siltation got out of control during the reign of Amenemhet III.

Its excavators considered the fort of Dorginarti to be a construction of the New Kingdom because they found there no remains from the Middle Kingdom (Knudstad 1966); the earliest datable material came from Dynasties XIX and XX, or late Empire. Nevertheless the fort has certain fea- tures suggestive of flood damage to the foundations in the early life of the fortress. The damaged brick- work was then apparently given "a stone casing or glacis where bed-rock did not provide such, [the rock] having been piled against the brickwork of buttresses and bays to form an even slope away from them. [The glacis] seems to have been carried well below recent low Nile levels [artificially raised by the Aswan reservoir] and to have been intended as a defense against the Nile as well as the attacker. It also shows that much of the alluvial build-up that gives the island of Dorginarti its recent size came later than the establishment of the fortress. In spite of the lack of evidence for a Middle Kingdom date"1 it is certainly most tempting to associate the water damage with the great floods recorded at Semna. In any case, a contour map of the island should be most interesting as a clue to flood levels.

There is a second and less noticed "Semna prob- lem" that may contribute to the solution of the original one. This problem becomes most striking when the data are plotted, as in ill. 2, against time:

namely, what about the many missing years? If the Egyptians at Semna were recording flood levels, why did they fail to record over half of the years?

Indeed, the motives of the Egyptians for record- ing the flood levels on the cliffs at Semna are ob- scure. Lepsius (1853)-who himself proposed a massive-erosion explanation for the lowering of the flood levels-postulated that the floods were re- corded at Semna as the frontier of the Kingdom, the point where the height of the flood could first be known and transmitted throughout Egypt; he further suggested it had some connection with Amenemhet III's building activity in the Fayum (Denkmdler V:224). Reisner (1929a) pointed out that there is no graduated scale, and that the flood itself would arrive in Egypt almost as soon as any message about it; he postulated that the records were for purely local use, perhaps intended as a navigational aid, to determine the best time for the annual shipping fleet to set sail for Egypt. But this function would seem also to require a graduated scale. When the Semna inscriptions are interpreted as evidence of ultra-high floods, Lepsius's postulate becomes more attractive, and I further suggest that Amenemhet III's famous, but obscure in detail, works in the Fayum were directed mainly toward reducing the destruction by these excessive floods in Lower Egypt, a hypothesis which we shall ex- plore further in a later section. Unfortunately there is no evidence that the Egyptians had any more rapid means of communication than by boat, no evidence that they used smoke signals, mirror sig- nals, drums, etc., which would be necessary to make flood measurements at Semna useful in Egypt proper.

It seems probable, therefore, that the flood levels were recorded only because of their great and un- usual height, as wondrous curiosities. Then, al- though there can be no proof, I suspect the reason for the gaps in ill. 2 is more than human caprice or accident of preservation. I suspect that they re- corded only the remarkably high floods and that most of the missing years had a flood level little or no higher than the modern-and than the earlier years of Dynasty XII-and well below the level of all surviving inscriptions. This is the primary jus- tification for plotting the fallen inscriptions to the average level of those remaning in situ.

In summary, I take the inscriptions as evidence 19 Primarily on stylistic grounds, Adams (1971) indepen-

dently concludes that a Middle Kingdom origin is the most probable for the fort of Dorginarti.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 15: Climate and the History of Egypt -The Middle Kingdom

236 BARBARA BELL [AJA 79

that a number of extremely high floods did occur in the years between 1840 and 1770 B.C., but find no reason to consider these floods as typical of the Middle Kingdom. Such evidence as has come to light, particularly from Mirgissa, suggests that they were not typical. Then there is the evidence for LWLs in the region of the Second Cataract close to modern levels, and this we shall review next.

Evidence from Nubia, LWLs (Low Water Lev- els). Clear evidence has existed for some time (Reis- ner 1929a, b; I931) that the forts built in the region of the Second Cataract by the Pharaohs of the Mid! dle Kingdom each had a stairway in a protected position leading down to the Nile. Wherever these water-stairs are well preserved, they extend close to the modern LWL. At Semna West the steps end just above the level of the Nile on 1924 April 5 (Dunham and Janssen I96o:map IX), very close to the average LWL of the present century A.D. (see Table 2, and note 12). At URONARTI the lower part of the river stairs is a roofed tunnel-in struc- ture the closest surviving analogy to Semna South-- particularly well preserved.20 In 1930 "at the begin- ning of March the water level was about 30 cm. over the tunnel floor at its lowest point" (Dunham I967:20 and map IV); at Halfa the level in early March 1930 was 115-7 m. and at Abu Sir 120.2 m., thus 50 to 70 cm. above the average modern LWL. At SHALFAK the lower part of the river stairway is poorly preserved, but a quay was found at about 7 m. above the water level of 1931 Feb.-Mar,-which was close to the modern LWL at Halfa and at Abu Sir. The Shalfak quay is thus similar in location to the structures identified as riverside facilities at Mir- gissa by Vercoutter (1970) and his colleagues.

At Uronarti was found an inscription recording navigational difficulties encountered by Senwosret in the spring of his 19th year upon his return from "overthrowing Kush" (Wheeler 1931:66; Dunham 1967:34). As translated by Vercoutter (1966:155) it reads in relevant part:

Year 19, fourth month of Akhet season, 2nd day21 ... under the majesty of King ... Kha-kau-Re (Senwosret III) . . . The Lord . .. proceeded northward, having crushed the vile Kush, one had to look for (navigable) water to cross Ishemuk22 (and) to haul (the boats) because of the season,

every shoal (was) likewise. As for (the shoal of) ... (?) it was difficult, its water was (too) light (sic) to get through by hauling upon its . . . (?) because of the time of the year.

The discord between the evidence of the water- stairs for LWLs about equal to the modern and of the inscriptions for floods averaging 8 m. above the modern has long caused difficulty. The difficulty has been compounded by what seems with many scholars an implicit assumption that whatever the conditions implied by the Semna inscriptions, these conditions prevailed throughout the Middle King- dom. But there is nothing in the archaeological data or in the inscriptions themselves to compel such an assumption. Even before the excavations at Semna South and at Mirgissa by Vercoutter (1966, 1970), there was no necessity to conclude that the high floods indicated by the Semna inscriptions and the LWLs indicated by the river-stairs of the forts oc- curred in the same years or even the same decades. We should keep in mind too that the river-stairs of the forts must go to the minimal LWL, not to the average LWL, if the forts are to be assured of water every year. It is of interest that Reisner con- sidered that the lower part of the Semna stair was later than the upper part (Dunham and Janssen 1960:7); perhaps the Egyptians had to extend it in a year of unusually low water. Moreover it is gen- erally agreed that the forts were built in their most developed form under Senwosret III, and several were probably built initially under Senwosret I, in either case well before the years for which we have evidence of ultra-high floods at Semna. From the French excavations at Mirgissa of course we now have good reason to believe that the floods did not significantly exceed the modern during the Middle Kingdom, until late in Dynasty XII or early in Dynasty XIII, Mirgissa itself yielding no exact dating for the "crue millenaire." Thus there is no necessary conflict between the evidence on LWL and on HWL; through most of the XIIth Dynasty each was close to its modern level.

Vercoutter (1966:142) cites in addition, however, an inscription engraved on a rock in the middle of the rapid at Semna, at a height less than 2 m. above the present LWL of the Nile, and which is believed

20 The water-stairs at Semna West were also in a roofed tunnel, part of which survived intact until modern times (Adams 1972).

21 Senwosret III, year 19 IV Akhet 2 =- 186o B.C., 2 March Julian and 14 February Gregorian calendar (courtesy of R.A.

Parker, 1971 pers.comm.). 22 Exact location unknown. However there is a particularly

difficult stretch of rapids just upstream of Uronarti, passable only at or near HWL (Dunham 1967:3), and this may well be the location in question.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 16: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 237

to have been carved in the reign of Amenemhet III, as evidence that the LWLs even in the time of this king were not 8 m. higher than the modern. This argument is open to a threefold doubt. First, the inscription is not dated. Second, even if one ac- cepts it as from the reign of Amenemhet, there is no compelling reason to believe it was inscribed in one of the years for which an ultra-high flood is attested. A notable feature of the flood inscriptions, most striking when they are plotted against date as in ill. 2, is the lack of record for many of the years. If the Egyptians at Semna were recording flood levels, why did they fail to record over half of the years? In the previous section, I suggested that they recorded only remarkably high floods, simply because they were so remarkable as to demand com- memoration. If this is correct, we may suppose that the floods in most of the blank years of ill. 2 were at least lower than the lower ones there shown, or no more than 0-5 m. above the modern HWL. The rock could have been carved in one of these more normal years. And third, I see no justification for Vercoutter's (1966) assumption that a rise of 8 m. in HWL would necessarily imply a rise of 8 m. in LWL.

This last assumption is unsound for two reasons. Flood levels depend essentially upon the summer monsoon rainfall over the Ethiopian highlands while the LWL depends primarily on the rainfall over the East African lake region and the swamps of the southern Sudan, so that there is no necessary a priori correlation to be assumed between the high and low levels of the Nile. It has been asserted

(Brooks 1949) and widely accepted that the levels of high and of low water in the Roda-gauge rec- ords, existing with some gaps from A.D. 622, tend to be positively correlated, although Brooks himself points out that the correlation is far from exact."3

While little is known about LWLs, I consider it probable that at least some rough correlation does exist between the HWL and the LWL of an epoch. But there is no reason to assume that each must rise by an equal amount in meters or gauge level. An equal multiple of increase (or decrease) in vol- ume of water seems more probable, and this would

indicate a rise of 2 to 4 m. in the LWL associated with the great Semna floods, as I shall explain more fully presently.

We noted that, in part because of his erroneous assumption that a rise of 8 m. in HWL required an equal rise of 8 m. in LWL, Vercoutter (1966) was led to postulate that the high levels resulted not from great floods but from a series of partial dams, with open central channel, constructed by the Egyptians of the Middle Kingdom to improve the navigability of the river in the difficult stretches of the Second Cataract. He stated himself that the dams would only extend somewhat the season of higher water, but would have negligible effect at low water when navigation of the cataracts was most difficult. The technical possibility of such a series of structures is for engineers to evaluate de- finitively, but it appears to me altogether unlikely, particularly with partial dams which would be readily torn apart at the unsupported edges of their central openings. To build such a series of dams on a river the size of the Nile, even if it were feasi- ble, seems an incredible labor for a very modest profit. I suggest also we should expect the ancient Egyptians, if they had built a dam at Semna, to have undertaken first some leveling and shaping of the natural rock of the cataract to provide a more stable and level foundation for their building stones.

Vercoutter emphasizes the lack of other evidence for the necessary substantial fluctuation in climate that would be required to produce such great floods. It is however doubtful that any but the most subtle evidence, requiring the most sophisticated tech- niques to discover, could be expected if, as it now appears, the great floods covered less than a century, occurring in only some 20 to 30 nonconsecutive years. I too am aware of no other evidence than the items we have discussed from the general region of the Second Cataract. But there is a growing body of evidence (Butzer et al. 1972) that the levels of the East African lakes have fluctuated substantially in post-glacial times. Although no high (or low) lake level has yet been dated to exactly the epoch under discussion here, one high level at Lake Ru- dolf, some 70 m. above the modern surface, has yielded a radiocarbon date of 3250 BP (Butzer et

23 Moreover, if Popper (1951) is correct in his conclusion, based on extensive study of medieval Arabic and Egyptian texts, that the figures published as "minimum" actually rep- resent the level of the river on 20 June (Julian), this being the date when each year daily readings of the Roda Nilometer were

begun, there is serious doubt that much is known about the long-term behavior of the LWL. (Popper's interpretation of the so-called minima in itself offers interesting possibilities which I hope to explore elsewhere.)

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 17: Climate and the History of Egypt -The Middle Kingdom

238 BARBARA BELL [AJA 79

al. 1972),24 or ca. i6oo B.C. when corrected for sys- tematic error resulting from variation in the C14 content of the atmosphere by the scale of Michael and Ralph (I974). However it would require some decades of rainfall much above the modern aver- age to raise the level of Lake Rudolf by 70 m., so there may well be a closer connection of the rise in lake level with the fluctuation to a wetter climate which produced the great floods recorded at Semna than the dates at first suggest.

A final inscription relevant here, which Vercout- ter (1966:164) cited as evidence for floods of modern size in the reign of Senwosret III, is actually some- thing much more remarkable. It is a record of Nile level, discovered at the Dal Cataract (about 83 km. south of Semna) by the survey team directed by A.J. Mills, dated to year io of Senwosret III and in a position close to the modern HWL. We may wonder why the Egyptians chose to commemorate such an ordinary flood. The answer becomes clear when we have the full date, generously communi- cated to me by Mills as year io, third month of Akhet season, 9th day; which is 1869 B.C. Jan. 24 (Gregorian) (courtesy of R.A. Parker, 1971 pers. comm.). This date is less than a month earlier than that from year 19 IV Akhet 2 at Uronarti which records a troublesomely low water level, and makes it clear that the Dal Cataract inscription does not define a normal flood level but records an exception- ally high level of the winter Nile, about as high as the average modern HWL (to be precise, 3-47 m. above the river level on A.D. 1968 April 16), and about 3.5 m. also above the normal modern level for late January. There was no modern gauge in use in 1968 near Dal, but gauge levels from Argo and Merowe to the south, provided by the Sudanese Ministry of Irrigation, indicate that the river level in April 1968 was close to the modern average (1912-1957) for January 21-31. The Dal inscription thus belongs with the Semna inscriptions in indi-

cating Nile flows substantially in excess of the larg- est values observed in modern times. The Dal in- scription of year io III Akhet 9 (1869 B.C. Jan. 24) together with the inscription from Uronarti, year 19 IV Akhet 2, provides evidence for an extraordinary

variability in the Nile levels of late winter during the reign of Senwosret III.

Volume of the great Semna floods. In this section we shall seek an estimate of the magnitude of the great Semna floods of the Middle Kingdom, and consider the implications for climate history. The accuracy of the estimate will be limited by a num- ber of factors. At once the most important, and now too late to remedy, is the lack of any systematic measurement of "modern levels" at Semna over even a few years. There is thus no basis for deter- mining exactly the relationship between levels at Semna and at modern gauge stations where the channel width and gradient of flow may be differ- ent. We have already noted (see Table i) the dif- ference in modern levels recorded by Ball and by Lepsius. We have a cross-section of the channel at Semna (Lyons 1906), but it shows a difference of about 13 m. between HWL and LWL, and I have not been able to relate it with useful exactness to the levels of Ball and of Lepsius.

The lack of any extended series of measurements at Semna itself can be compensated for to some ex- tent by use of the rather substantial series of mea- surements that do exist from stations in the region of the Second Cataract. Those available are listed in Table 2 with selected data on high and low water levels and on channel-width at high and low water. Particularly useful are the gauge readings from Wadi Halfa and from Kajnarty. The rise and fall of the Nile at Wadi Halfa in modern times averages about 7 meters, and "may be taken as rep- resenting what would take place in Egypt if there were no dams, barrages, or irrigation, to interfere with the natural flow of the river" (Hurst 1952: 239). In Kajnarty we have a station rather similar in situation to Semna, that is with a channel-width not exceeding 420 m., at least to the level of the highest modern flood in 1946. Thus extrapolation of the Kajnarty gauge-discharge curve, shown in ill. 3, to the I8-20 m. level above the modern average LWL (I31.6 m. above modern sea level) should give the best approximation we can hope now to obtain to the peak volume of flow of the great Semna floods of the Middle Kingdom.

As we have noted, any such large extrapolation 24 Indeed some values as low as 3250 BP have been obtained

from samples believed to derive from the Middle Kingdom: A-205, 206, 207, wood from forts at Semna, Radiocarbon 4 & 5, 3290, 3300, 316o BP respectively; from Mirgissa, Gif-295 and 297, 2925 and 3020 BP, Radiocarbon 12; and from Radio-

carbon 7, R-33, 38, 3300 and 3200 BP. Thus it is not impossible that the dated level at Lake Rudolf pertains to the period of the great floods at Semna. However ca. 16oo B.C. must still be considered the most probable date for the sample from the high level of Lake Rudolf.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 18: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 239

TABLE 2

Nile gauge levels, in meters above mean sea level at Alexandria, at stations in the region of the Second Cataract (from The Nile Basin, vol. III & suppls.,

H.E. Hurst et al.)**

Second Cataract 500 m. south of

Halfa Abu Sir Rock Kajnarty Semna

Km south of Halfa 0 12.5 47 Years of data 1890-1957 1920-1957 1931-1957 1901-1912

Average io-day LWL i37

a) 1890-1899 115-5

b) 1905-1930 I15.1 119.4 c) 1931-1955 115-5" 119.6 131.6 Channel width at LWL 520 m. ? 290 m. 40 m.

Average io-day HWL 149 a) 1890-I898 122.5

b) 1905-1930 121.9 125.7

C) 1931-1955 122.35* 126.0 142.4 149 Channel width at HWL 880m. ? 4o0 m. 400m.

Average range, HWL - LWL a) 7.0 m. c) 6.8 m. 6.4 m. 10.8 m. 12.o m.

Selected modern HWLs

90oi, Ball 122.0 142.6 149.1 1931, Wheeler 122.2 125.8 142.3 1946, high 123.8 126.9 144.3 151 1941, low 121.0 124.6 140.1

1946 HWL - 1941 HWL 2.8 2.3 4.2 4

Middle Kingdom Mean of inscriptions 128.0 132 151.4 158.1

in situ year 30 129.5 133 153-3 160.2 year 15 126.5 130.5 149.0 156 year 9 126.0 148.0 155

"Good flood," Senwosret I Aswan LWL 84.5 m. 125.4 129 147.4 154

84.0 m. 124.9 128.5 146.5 153

* Artificially raised by operation of the Aswan reservoir. ** In the years 1949-1957 there was also a station at Gemai, 27 km. south of Halfa, LWL 131.1 m.,

HWL i39-3 m., range 8.2 m.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 19: Climate and the History of Egypt -The Middle Kingdom

240 BARBARA BELL [AJA 79

NILE VOLUME, m3x 108 per day 2 4 6 8 10 20 40 60 80

-' yr 30

/7

SEMNA,MK --K

150/

-yr 15

= //

S45 LI 4

-1946-- u- L1958-

S--1931-

140 •,.'

--1941-

-

- H W L

< /x

135 zX

LWL

/ .--19 4 7

130 - '

-1931

130 - *-, .......... . . ..,_2

___. . . . . . . 3 5 I 2 3

m3x 108 per day

ILL. 3. The Nile at Kajnarty: gauge levels plotted against volume of discharge (see n. 7), and extrapolation (dashed lines)

above level of modern floods

is subject to considerable uncertainty, the more so here where we have little information on the width of the channel at Kajnarty above the HWL of 1946. On the positive side, we do know that the channel- width was increasing only slowly at Kajnarty over the range of modern floods, and that this condition continues at Semna up to the level of the highest record from the Middle Kingdom. Over the range of modern floods too the channels at Semna and at

Kajnarty are closely similar in width. And finally, Kajnarty is just upstream from Askut where that

solitary Nile record from the reign of Sekhemkare was found at about 20 m. above modern LWL. On the other hand, the low-water end of the Kajnarty curve is probably less applicable to Semna, because

the latter has a much narrower channel at low wa- ter. Ball (1903) reports that the low water at Semna flows smoothly through a channel only about 40 m. wide and at least 20 m. deep, so that Semna itself would not appear to have presented navigational difficulties at low water."2 This narrowness of the channel through the Semna dike for a few meters above LWL probably accounts for the fact that the flood of 190o (fortunately a fairly typical twentieth century modern flood) was 12.1 m. above modern LWL, while the modern (1931-1955) floods at Kaj- narty have averaged only io.8 m. It therefore seems probable, if we take the flood of 19o0 pre- ceding Ball's visit as typifying the Semna-modern, that we should disregard LWLs and with ill. 3 use only the difference between modern HWL and the Semna flood levels.

The Middle Kingdom floods recorded at Semna, averaging 9.o m. (see Table i) above the mean HWL (142.4 m.) of the twentieth century, or about

151.4 m. at Kajnarty, would from ill. 3 have a peak volume of around 32xIo8m."/day. This would be about three times the mean peak-volume of the ten greatest floods (-'io.5xIo"m.') of the late nine- teenth century A.D., and about four times the mean peak volume (7.8xio"m.") of the first four decades of the present century. The cross-section of the channel (from Ball 1903 and Lyons 19o6) occupied by the great floods at Semna is about 2.2 to 2.4 times that occupied by the average modern flood. While this value sets an absolute minimum to the volume of the Semna floods, the actual value would be con- siderably larger because the velocity of flow in- creases as the river level rises.

Ill. 4 shows the relation between gauge levels of the river at Halfa and at Kajnarty. Over the 27 years available for comparison, the Halfa values rise progressively relative to those from Kajnarty, presumably due to the operation of the Aswan res- ervoir; the two lines are based on the years 1931- 1935 and 1951-1955, the former being the lower and more relevant to our problem. From the extrap- olated portion of ill. 4, we can estimate the level of floods to be expected at Halfa in the Semna years. (Basic to the accuracy of the extrapolation is the assumption-which may or may not be correct-

25 This appearance of navigability assumes a boat steerable with some precision, and Adams (1972) points out that the Nile nuggar is "one of the world's least maneuverable craft."

Vandersleyen (1970) notes that the principal difficulty at Semna arises from a lack of any points of support for hauling a boat.

This difficulty arises because the deep channel through the dike is only about one tenth the width of the river immedi- ately above and below the dike. In any case, there are other stretches of the river between Semna and Halfa that present more formidable difficulties at LW than does Semna itself.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 20: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 241

yr 30 .1

SEMNAMK--

5/ / I

'

• .•_ -

_

4,

//

S,/-1946 2iz

r--L.1958 d -1931-

D 140

.-1941-

/HWL

so

z

• 35

c\J LWL cj--1947

_ fM -1931

130 0 -3: 0 q

L 1 1 1 1 1 I I_ 2 4 6 8 10 12 14 16

HALFA GAUGE, meters

ILL. 4. Comparative gauge levels, observed and extrapolated, of the Nile at Kajnarty

and at Halfa

that the relative channel widths remain similar at higher levels.) An average Semna flood (151.4 m. at Kajnarty) would reach about 128 m. at Halfa, and the greatest flood (i53-3 m., Kaj.) would come to about 129.5 m. at Halfa. Thus ill. 4 indicates that the great floods of the Middle Kingdom would reach a level 6 to 7 m. above the average twentieth century HWL, and about 4 m. above the highest modern (1946) flood actually measured at Halfa. These expected levels are summarized in the lower section of Table 2; the values for Abu Sir were derived from a similar graph.

Since the channel is wide at Mirgissa, roughly similar to Halfa and Abu Sir, we should expect sim- ilar levels there, or from Table 2 about 5.8-6.2 m. above the 1931 HWL, in good agreement with point (3) of Wheeler (1961) quoted in the previous section. Nearly all the water-wornness of the rock up to 6.73 m. must have developed at an earlier epoch, not only because fluvial planation is a slow process but because Vercoutter's (1970) exca- vations at Mirgissa have yielded evidence that the great floods were not typical of the Middle King- dom. It is particularly noteworthy that the highest

floods, of years 23 and 30, could be expected to have risen about 7 m. above the 1931 HWL, very close to the highest evidence found by the French for water damage at Mirgissa (Vercoutter 1970). This damage, I suggest, can be most plausibly attributed to the great floods recorded at Semna. The climatic history of Ethiopia is surely not known in such detail that we can reject this hypothesis simply be- cause there is no supporting proof for the implied climate fluctuation in these few decades between r84o and ca. 1770 B.C.

Ill. 5 shows the relation between gauge levels at Kajnarty and at Aswan, based on the io-day-max- imum gauge levels at the two stations for the years 1931-1955. With the old Aswan dam, the sluice

gates were fully opened during the flood season, re-

sulting downstream and at the gauge in a flood level quite similar to what would have occurred in the absence of the dam. But LWLs were significantly raised by the regulation of the discharge from Octo- ber onward, so for a minimum point I have plotted the mean LWL at Kajnarty from 1931-1955 against the LWL at Aswan for 1901-1902, which should in fact, from the volume figures, be properly compara-

yr 30-

SEMNA,MK--

150 /1 /

- yr 15

" 145

z / -1946 2 ./ -1931-

8 140 -- 1941 < - HWL

z 0 0

-1947

- ttt O -i N) ?oDCO

130- D0 )O

85 90 95 100 ASWAN GAUGE, meters

ILL. 5. Comparative gauge levels, observed and extrapolated, of the Nile at Kajnarty

and at Aswan

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 21: Climate and the History of Egypt -The Middle Kingdom

242 BARBARA BELL [AJA 79

ble. I have drawn one line through the minimum point and the average of the maxima, and a second to fit best the various maxima without regard to the minimum. From ill. 5 we can see that the great Semna floods would fall within the range from 97 to ioo m. at Aswan. If determined by the line through the maxima alone, 151.4 m. at Kajnarty corresponds to 98.2 m. at Aswan (Elephantine).

Let us now return to ill. I, where we find that a gauge level of 98.2 m. would indicate a vol- ume of about 30x1o"m."/day, in gratifyingly good agreement-considering all the dangerous extrapo- lations involved-with the value obtained directly from Kajnarty and ill. 3. Table 3 lists the peak vol- umes and levels to be expected at Elephantine for selected years including a "good flood" of Senwos- ret I. It would be reasonable to assume an uncer- tainty of at least ten percent in the estimates of ancient volumes. The greatest flood recorded at Semna during the Middle Kingdom had a peak volume about twice that of the lowest flood re- corded, and about four times that of the larger mod- ern floods.

For a "good flood" of Senwosret I, we have of course to assume a LWL. If we start with the mean modern LWL of 84.5 m., we have a HWL of 95.8 m., whence from ill. i a peak volume of

I7x I om. /day, about 50 percent greater than the great

modern flood of A.D. 1878 but just over half that of the average flood recorded at Semna about a cen- tury later. There is of course no reason to believe that the floods were gradually increasing during that century.

It may be objected that I have used too low a minimum to estimate the flood considered "good" by Senwosret I. However Egyptologists generally agree that this king began the building of the sys- tem of Second Cataract forts, and their water-stairs comprise the most solid evidence for LWLs in the Middle Kingdom, and indicate LWLs close to the modern. The modern LWLs were lowest in the pe- riod 1920-1931 when Reisner and his associates were excavating the forts, so one could probably justify a LWL of 84.0 m. for Senwosret I; whence by ill. 1, with HWL at 95-3 m., the peak volume would be around 15xio"m.'/day.26 However we must not overlook the possibility, noted earlier, that the zero of the Elephantine gauge was fixed in the time when "... The river of Egypt was empty, men cross over the water on foot . . ," which could give HWLs similar to that of A.D. 1946. To allow for this possibility, I have included in Table 3 the vol- ume to be expected if the zero was at such a very low level.

As we have noted, Vercoutter (1966) was led into difficulties by his conviction that a rise of 8 m. in

TABLE 3

Estimated io-day peak volume, and gauge levels at Elephantine, for selected great floods in the Middle Kingdom.

gauge levels flood volume Elephantine Kajnarty ma3xio/day

Amenemhet III Average 98.2 m. 151.4 m. 31 '3 Max., year 30 99-4

153"3 42 +3

low, year 15 96.7 149.0 21.5+1.5 year 9 96.1 148.0 19.0- I

Senwosret I Aswan LWL 84.5 m. 95.8 147.4 17 -+2

84.0 95-3 146-5 15-5 I 83-5 94.8 1457 13.5-+1 83.0 94-3 145.1 12.5 82.5 93.8 144.6 o0.8

26Adams (1972) points out that "the amount of subsequent modifications in the Second Cataract Forts [makes it] very un- safe to assume that any of the water stairs as we now know them

date from the original phase of building." Thus we cannot confidently assume that they reflect the LWL in the time of Senwosret I.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 22: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 243

HWL required a corresponding rise of 8 m. in LWL. But by ill. 3 we can see that a rise of 8 m. in the LWL at Kajnarty would occur only with a ten- fold increase (from 0.50 to 5.oxio8m.3) in the LW volume. While I cannot say this is impossible, it certainly does not seem the most likely assump- tion. If we assume a doubling to tripling of the LW volume, we have at Kajnarty a rise of 1.5 m. to 3 m. in the LWL, the latter being close to the excep- tionally high LWL that followed the great flood of A.D. 1878. Although it would be reasonable to expect a rise of 2 to 5 m. in the LWL at Semna after a year of great flood, nothing can in fact be known about the LWLs at the time of the Semna floods from data now available. And of course the LWL could be as variable as the HWL. Particu-

larly if we assume that the years not commemorated by an inscription had HWLs below the lowest of the inscriptions, there would seem to be no lack of opportunity for the inscription emphasized by Vercoutter (1966:142) to have been carved on that rock at about 2 m. above modern LWL during the reign of Amenemhet III.

Although it is premature to attempt, on the basis of data available at present, any extensive discussion of the most probable origin and the climatic im- plications of the great floods recorded at Semna in the latter part of the Middle Kingdom, there are a few points to be noted. In the twentieth century, the flood waters of the Nile originate approximately as follows at maximum (Hurst 1952:242):

White Nile io percent Blue Nile 68 percent Atbara 22 percent

At low water by contrast we have

White Nile 83 percent Blue Nile 17 percent Atbara o percent

Over the year, the volume of the White Nile varies by a factor of 2 to 3, the Blue Nile by a factor of 60 to 70, and the Atbara even more, as the latter two rivers are swollen by the summer monsoon rains over the highlands of Ethiopia.

With our present limited knowledge on climate variation, the great Semna floods could seemingly arise in a variety of ways. The percentage from each main source could remain constant or it could vary widely. The most straightforward hypothesis is to assume a proportional increase in the volume of each of the main tributaries. However study of the relatively high floods of 1946 and 1954 (Hurst, Black, and Simaika 1959:67-71, 92-101, 127) suggests a previously unexpected difficulty. In the flood of 1946, some 15 percent of the peak volume entering the river at Khartoum, mainly from the Blue Nile, was lost by overflowing the channel between Khar- toum and the junction with the Atbara. The extent to which this loss would increase progressively for a much greater flow of the Blue Nile cannot even be estimated without contour maps of the valley, but the possibility should be kept in mind by fu- ture investigators.

If, on the other hand, the great floods result primarily from a further and stronger northward penetration of the summer monsoon rains, it seems likely that the Atbara would increase in volume substantially more than the Blue Nile. The possi- bility of the activation of certain now-dry wadis downstream of the Atbara confluence, entering the Nile from the south and west and implying a sub- stantial increase in rainfall over the western Sudan, should not be altogether overlooked. A detailed analysis by neutron activation of the chemical con- stituents of a core27 of the silts deposited above the glacis at Semna South, and comparison with pos- sible sources, might well cast significant light on the problem and contribute a clearer picture of the climate fluctuation that produced those great floods -and many more great floods in the Neolithic era.28

Although significant penetration of the summer monsoon north of the Atbara confluence, either to the east or west of the Nile, is unknown in modern times, we have evidence that something of the sort did occur in year 6 of King Taharka, 683 B.C. It even rained in Nubia "so that the hills glistened," according to a surviving stele, and the flood was very great (Vandier 1936:I24):

27 Because the site is now under water, special coring equip- ment would be required to obtain samples.

28Excavations by Arkell (I96i:26) produced evidence of floods some 5 m. higher than today in the Neolithic period near Khartoum, and of generally much wetter conditions, attested by

bones of such swamp-loving fauna as the Nile Lechwe, water mongoose, and reed rat. We must therefore imagine that dur- ing the Neolithic Wet Phase the Blue and/or White Niles were substantially higher than today.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 23: Climate and the History of Egypt -The Middle Kingdom

244 BARBARA BELL [AJA 79

Now a wonderous thing occurred in the time of His Majesty, in Hnt-Hn-Nfr,"' never ... seen in the time of the predecessors, and this because his father Amon-Re loves him. His Majesty prayed for an abundant Nile ... his father Amon- Re makes it reality .... When the time came for the flood of the Nile, it began to increase greatly every day; when it had passed many days, increasing each day, it flooded the mountains of the southern lands and the low lands of the north. The land was like an inert primordial Ocean, the banks could not be distinguished from the river.... It happened that a downpour from the sky, in Nubia, made the mountains sparkle as far as their summits [unusual northward pene- tration of the monsoon rains, particularly in so far as any causal connection is implied between the downpour in Nubia and the high flood]. Everyone in Nubia was rich in everything, and Egypt was also plentiful.... Everyone thanked the King. ...

Now this flood, recorded on the quay before the Karnak temple, rose to some 84 cm. above the pres- ent pavement of the Hypostyle Hall, reaching ap- proximately the same height as the crest of the A.D. 1946 inundation when it overflowed the bank between Karnak and Luxor (Nims 1965:76). As-

suming a rise of io cm. per century in the riverbed and floodplain, or about 2.6 m. in the 2629 years between these floods, the flood of year 6 of Taharka was about 2.6 m. higher than the largest of modern floods, and perhaps about the volume of the flood of year 9 of Amenemhet III, the smallest of the great floods of the Middle Kingdom. Such a flood in modern times, before the construction of the High Dam, would be a horrendous catastrophe for Egypt. One may wonder whether it was then really such a glorious event as Taharka's stele asserts, but on the other hand, it is certainly unlike the ancient Egyptians to commemorate an unwelcome event with a special stele.

A great flood of similar magnitude occurred in

year 3 of Osorkon III about a century before and is commemorated-without the suspect enthusiasm

--on a wall of the Luxor temple thus (Vandier 1936:123) :

Year 3, first month of season peret, day 2 under the Majesty of . . . Osorkon III. The water of Ncin rose.. . in this entire land and it reached the two cliffs of the desert as at the origin of the world; the land was in its power, as (in the pow- er) of the sea; there existed no dike made by the

hand of man, that could withstand its force, men were like flies/grains of sand on their town. (The water) was tremendous, it was high in the... like the sky. All the temples of Thebes resem- bled a swamp, the day of the going-out of Amon of Luxor [a festival], when his statue was raised and when he entered the Naos of his sacred bark, in this temple; the people of his city were like swimmers in the water ...

We may imagine a similar condition associated in Egypt with the great floods commemorated at Semna.

Vercoutter (1966:137-38) calls attention to a record of the flood level in year 6 of Senwosret II, dis- covered at the fort of ANIBA during the 1912 ex- cavations by Steindorf. Aniba is roughly midway between Aswan and Halfa. This inscription, in the standard form, was located "on one of the blocks at the base of the main wall" at about 1.40 m. above the "normal level" of the modern flood. Sometime after 1912 the block apparently fell from its origi- nal position, "probably as a result of a very high flood subsequent to the excavations." One would tend to expect the Aniba fort to have been damaged by the great Semna floods, and even by those of Osorkon III and of Taharka cited above, unless the riverbed has risen since the Middle Kingdom in the vicinity of Aniba (as it has apparently fallen in the vicinity of Mirgissa).

The inscription in the Dal Cataract, commemo- rating a water level close to the modern HWL in late January of 1869 B.C. and substantially in ex- cess of anything recorded in modern times in this season, also presents difficulties of explanation. In modern times, the most variable tributary in the winter season is the Sobat, which joins the White Nile just above Malakal, and which in 1918 pro- duced a January flow of about 300 percent above normal and in February over 6oo percent above its twentieth century normal. The Dal inscription implies a volume of flow of about 7.8xio"m."/day, and at its highest (in any season) the Sobat has produced only i.o8xio"m."/day (in February 1918). The highest twentieth century flow of the White Nile at Malakal, downstream of its confluence with the Sobat, is about 2.IxIo"m."/day, recorded in De- cember 1964. It is doubtful that any great increase is possible because of the flatness of the land and the very slight gradient, only 16 m. over a distance well refer to the vicinity of the Nubian capital of Taharka at Merowe near Gebel Barkal.

29 The frontier region (H. Goedicke 1965 Kush 13:ro2f) in Nubia advancing by conquest to Gebel Barkal and the Fourth Cataract by the middle of Dynasty XVIII; thus the name may

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 24: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 245

of Iooo km. south of Khartoum (Hurst & Phillips 1938:17). Indeed "in a year of heavy rain practi- cally the whole country, from the Ethiopian foot- hills to the Bahr-el-Jebel and the slopes of the Lake Plateau ['of East Africa], is under water" (Hurst 1952:245). Such rain occurred in the winter of 1917-1918, producing the highest recorded flow of the Sobat in February 1918, but falling far short of what is implied by the Dal inscription.

This leaves only the Blue Nile. The highest re- corded flow of the Blue Nile in the twentieth cen- tury in December is o.96xio"m.3/day in 1916, but in the flood season it is of course capable of at least 8xiosm.3/day (i946) and probably more. Thus it appears that the Dal inscription should be inter- preted as indicating an extraordinary prolongation of the monsoon rains and/or remarkable strength and early onset of the spring rains over the Blue Nile basin sufficient to produce a January flow of some 5xio"m."/day, plus a flow of at least 2xio"m."/ day out of the White Nile including the Sobat.

Nile Levels, Evidence from Egypt. There is no known geological evidence for very high floods during the Middle Kingdom in Egypt. This lack reinforces the evidence from Nubia, which indi- cates that the great Semna floods were not typical of the period. If they occurred in only some 25 to 40 non-consecutive years, however, geological evi- dence could hardly be expected nor should its ab- sence cause concern. In addition, insofar as the ex- cess water came from the activation of normally dry wadis entering the Nile from the southwest between Merowe and Dongola, that is from one or two heavy rains over the region-as suggested by the stele of Taharka for his great flood--one would expect the abnormally high waters to be of short duration. Insofar as the excess water came primarily from the Atbara and Blue Nile one would expect a longer duration of HWL, with much overflow between Khartoum and the Atbara confluence available to run back and prolong a high level. The shorter the duration of the very high waters, the more difficult of detection would be any geological evidence. We may in any case, how- ever, hope for the eventual discovery of archaeo- logical evidence in the form of inscriptions and/or water-damaged structures in Egypt such as those found at Mirgissa by Vercoutter (1970) and his associates. One such possibility was kindly called to my attention by W.K. Simpson. Among the blocks from earlier temples re-used in the building

of the Third Pylon of the Temple of Amenre at Karnak were found two fragments of a XIIIth Dynasty stele referring to a great flood in year 4 of a King Sobekhotep. The relevant portion reads (Habachi 1974:2IO): ". ... Year 4, IV Shomu, the five epagomenal days under the majesty of this god (Sobekhotep) . . . His majesty proceeded to the hall of this temple, Hapi [the Nile god], the Great, has been seen coming towards his majesty and the hall of this temple full of water. His majes- ty was wading in it together with the work- men... ."

It is immediately tempting to identify this king with the Sekhemre-Khutowy Sobekhotep I, whose year 4, ca. 1775 B.C., is commemorated by a great- flood inscription at Semna. In 1775 B.C., IV Shomu includes 20 Sep. to 19 Oct., with the epagomenal days covering Oct. 20-24, somewhat late in the year for such high waters, but even later dates are known (see n. 45 and nearby quotations from Ball relating to very prolonged high waters in A.D. i371 and A.D. i818), so the date is no obstacle. A very serious obstacle, however, is raised by the fact that the full royal name on the stele is Sekhemre-Seuser- towy Sobekhotep; indeed the second part of the first names is so clearly different that we are obliged to reject the temptation. Habachi follows Beck- erath's (1965) designation of Sobekhotep VIII and dating to the i6oos, which would bring the epago- menal days within the latter half of September. If this be correct, we must regard the stele as com- memorating one of those great floods which occur from time to time, and probably without connec- tion to the series commemorated at Semna. On the other hand we should not forget the paucity of evidence for the actual order of the kings in Dynas- ty XIII (see n. 48) and keep an open mind for an earlier dating of King Sekhemre-Seusertowy Sobek- hotep, making it possible to see the great flood in his year 4 as part of the same climate fluctuation which produced the series of great floods com- memorated at Semna.

Of uncertain significance are two inscriptions from the reign of Senwosret III, cut on the rocks of the Island of Sehel, which record the making (un- dated) and the remaking (year 8) of a canal to facilitate the passage of the cataract at Aswan (Breasted 1906), analogous to the activities of Weni in the reign of Mernere of Dynasty VI before the First Dark Age. There is no indication whether the later canal was made necessary because low floods

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 25: Climate and the History of Egypt -The Middle Kingdom

246 BARBARA BELL [AJA 79

had clogged up the old passages, or because Sen- wosret III wanted to use larger boats than his predecessors for his military campaigns in Nubia. Nor should we overlook the possibility that the channel has become clogged by debris washed down from a local wadi after a heavy cloudburst.

In addition to the evidence from Mirgissa and perhaps Semna South in Nubia, there are in Egypt itself a number of archaeological points which also argue against the ultra-high floods being typical of most of the period, points which concern the height above the river-level of numerous monuments of the Middle Kingdom. In considering these we must take account of the rise in the riverbed and in the surface of the alluvium, which Ball (1939:162-77) estimates as about 3.35 m. in the vicinity of Cairo.

Unfortunately I have been unable to obtain suffi- cient information on the elevation of various rele- vant monuments, so that the following paragraphs are more speculative than I would wish and should be taken primarily as suggestions for future re- search.

One striking fact is the absence of known Valley Buildings associated with the Middle Kingdom pyramids of Dynasty XII; they are either com- pletely destroyed or buried under the modern culti- vation to a greater extent than most of the similar Valley Buildings from the Old Kingdom, although surviving traces of causeways indicate that such Valley Buildings must have existed (Fakhry 1961; Edwards 1961). This suggests-although proof or disproof, if ever possible, must await further ex- cavations or at least magnetometer explorations-- that the Valley Buildings of the Middle Kingdom were constructed at a lower elevation, closer to the modern HWL, than those of the Old Kingdom. This in turn suggests that the floods were lower in the Middle than in the Old Kingdom. If the very high floods had ceased already by Dynasty VI, concurrently with the ending of the Neolithic Wet Phase, their memory would still be fresh and would influence the construction and placing of the Valley Buildings. If the later buildings were erected at a

lower elevation, as I suggest, they could have been more damaged by the ultra-high floods in the reign of Amenemhet III and his immediate successors. Then because of their damaged condition, they could be used as quarries by later generations with less strain on the conscience than one would expect in the case of well-preserved buildings. Moreover it seems unlikely that structures erected in an era of floods substantially higher than today should be so completely buried under the cultivation when the alluvium has risen at most about 3.5 meters.

It may be significant that the burial chambers under the pyramids of Amenemhet I and Senwos- ret I at Lisht have not yet been explored because they are full of ground water that seeps in as fast as it could be pumped out. Use of a strong pump in one of the flooded nearby shaft tombs by the ex- cavators from the Metropolitan Museum of Art resulted in lowering the water in the tomb by 40 cm. and no further. It also lowered the water noticeably in nearby shafts, indicating that the rock is very porous (1914 BMMA 9:207). Unfortunately I have not been able to discover any records of the depth of the water in any of these tombs, nor indeed any indication that soundings were taken in an effort to determine the depth by either the French (Gautier and Jequier 1902)30 or the Metropolitan Museum's excavators at Lisht. At the time it probably did not appear to be of interest. But if the depth of water in any tomb approached 3.5 m., this would pro- vide evidence for floods no higher than the modern when the tomb was built. Whatever the depth of the deepest water, its measurement would con- tribute toward setting an upper limit on the flood levels during the earlier reigns of Dynasty XII. The role of the Fayum lake, to be discussed pres- ently, should be kept in mind by anyone seeking to investigate these various levels.

The excavators of the Metropolitan Museum at Lisht concluded that the Pyramid of Amenem- het I was thoroughly plundered even of its casing stones and reduced to a shapeless mound of mud- brick early in the Second Intermediate Period

30 Gautier and Jequier (19o2) do give figures on the depth of the water below the surface, but their various figures appear incompatible with one another. Within the Pyramid of Sen- wosret I, they excavated a passage ca. 40 m. in length, sloping downward at about 25 degrees, and encountered water at a depth of ca. 22 m. (pp. 5-6). From their Figure 8 (p. 15) the water level appears to be ca. I8 m. below the surface around the pyramid. Beneath the Pyramid of Amenemhet I, a vertical shaft encounters water at ca. 8 m. (p. 94) and the shaft appears

to begin at about ground level (Figure Io6). It is doubtful that these two pyramids differ by anything like Io m. in ground level. In addition, water at the end of a burial shaft of a mastaba nearby the Pyramid of Amenemhet I is reported at 9.50 m. depth (p. o103), but in Figure 123 water appears at a depth of ca. 20 m. It is evident that this problem requires further study by someone with access to both large-scale con- tour maps and to the site itself.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 26: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 247

(Mace 1921 BMMA i6), a conclusion derived from study of the remains of a village found around it and even on its flanks. It may have suf- fered severe damage in the high floods of Amenem- het III's reign, a possibility that could be clarified by fuller information on the elevation of the desert plateau, described as "low" (1926 BMMA 21:33), upon which this pyramid was built. If the level is out of reach of even great floods, then perhaps the pyramid was damaged by a cloudburst, which may have inspired the precautions for drainage taken by the architects of Senwosret II (see below).

RAINFALL IN THE MIDDLE KINGDOM In the previous paper (Bell 1971) I reviewed

briefly the Neolithic Wet Phase-characterized by a rainfall substantially greater than at present over the Sahara, and probably (although not yet ade- quately documented) widely throughout the Near and Middle East-and its gradual decline to ap- proximately the modern level of aridity by the end of the Vth Dynasty, ca. 2350 (Butzer 1958, 1959b, , 1965). There was no ecologically signifi- cant revival of the rains over the desert during Middle Kingdom times, no return of the NWP (Butzer 1959c:58f); representations of flora and fauna in the art of the period indicate rainfall con- ditions indistinguishable from the modern aridity. For completeness and the convenience of future investigators, I shall review what evidence I have found pertaining to rainfall, excluding the floral and faunal material already analyzed by Butzer (1959c), although no firm conclusions can be drawn.

Egypt. The mountainous eastern desert has al- ways received more rainfall than the western (Ball 1939; Trigger 1965). Warm moist air from the southeast, from the Red Sea, is sometimes blown inland, to condense on the peaks near the sea and bring thundershowers, occasionally even violent

cloudbursts which can turn dry wadis into raging torrents for a few hours, as happened in Novem- ber 1923 (Murray 1967:45f; Sutton 1949:74f). On rare occasions, perhaps once every few years, one of these storms may reach the Nile Valley. It would appear that this occurred rather more frequently in the nineteenth than in the twentieth century A.D., for Wilkinson (1835:75) wrote: "Showers fall annually at Thebes; perhaps on an average four or five in the year; and every 8 or io years heavy rains fill the torrent-beds (wadis) of the mountains, which run to the banks of the Nile. A storm of this kind did much damage to Belzoni's tomb some years ago." In December 1843 in Aswan, Lepsius (1853:1i9) experienced a violent and particularly extensive storm: "Heavy rain, and a violent thun- derstorm . . . gathered on the farther side of the Cataracts, crossed with a mighty force the granite girdle, and then, amidst the most violent explosions, rolled down the valley as far as Cairo, and (as we have since heard) covered it with floods of water, such as had been scarcely remembered before.... Rain is, indeed, so rare here, that our guards never remembered to have beheld such a spectacle .. " About a year before, while exploring the pyramid field at Giza, Lepsius (1853:53) experienced another brief but heavy cloudburst, this one presumably from the Mediterranean and connected with the passage of a front. In the present century also, oc- casional heavy rains have been recorded (see Sut- ton 1949 for dates and descriptions).

It is clear that the architects of the Pyramids of the Middle Kingdom Pharaohs considered it ad- visable to take precautions against rainstorms, for the head of a lion waterspout"' which drained the broad flat roofs of the pyramid temple of Senwos- ret I was found among the ruins at Lisht (Hayes 1953). And a shallow trench filled with sand to absorb the rainwater flowing off the pyramid was provided around the base of the Pyramid of Sen-

31 Lion waterspouts in Egypt are better known from the temples of the Hellenistic period. However they existed as early as the Old Kingdom. Fakhry (1969:174) reports of the temple of King Sahure of Dynasty V, ca. 2480 B.C., at Abusir: "Rain falling on the roof was carried off by lion-headed gar- goyles, which projected well beyond the eaves, and fell into open channels cut in the pavement." And from the Empire period, on the Medinet Habu temple of Ramses III at Thebes, Wilkinson (1835:75) noted: "The head and forepart of several lions project, at intervals, from below the cornice of the ex- terior of the building, whose perforated mouths, communicat- ing by a tube with the summit of the roof, served as conduits for the rainwater which occasionally fell at Thebes. Nor were

they neglectful of any precaution that might secure the paint- ings of the interior from the effects of rain, and the joints of the stones which formed the ceiling being protected by a piece of metal or stone, let in immediately along the line of their junction, were rendered impervious to the heaviest storm." The waterspouts are described also by U. H61lscher in The Excavation of Medinet Habu III: "The drainage of rain water from one [roof] terrace to another and finally to the outside was a matter of special importance. Very large waterspouts, shaped like the forepart of a lion, form a conspicuous feature of the exterior walls of the temple" (p. 21; see also pp. 22, 46, and 49; for this reference I am indebted to C.F. Nims).

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 27: Climate and the History of Egypt -The Middle Kingdom

248 BARBARA BELL [AJA 79

wosret II at Lahun (Edwards 1961), lest the bot- tom casing stones slip out of position and the whole casing collapse. Perhaps some such catastrophe be- fell the Pyramid of Amenemhet I, which may have inspired the precautions for drainage taken by the architects of Senwosret II.

These various precautions indicate only occa- sional cloudbursts, such as Lepsius experienced and such as occur in the present century. They indicate nothing about the frequency of rain. How- ever Butzer's (1959c) analysis of the flora and fauna represented in art leaves no doubt that rain was not sufficiently frequent to be of agricultural significance more than it is today.

In year 2 of Nebtowyre Mentuhotep IV, at the end of Dynasty XI, rain in the Wadi Hammamat was deemed worth commemorating by a rock- carved inscription: ". . . Second month of the first season, day 23 [II Akhet 23 1996 B.C. Febru- ary 8, Gregorian calendar, courtesy of R.A. Parker, pers.comm. J . . . The wonder was repeated, rain fell. The highland was made a lake .. ." (Breasted 1906:216-17).

In spite of the rarity of rainfall suggested by this inscription, however, there must have been nomads living in the eastern desert-as indeed there are today and have been throughout historical times

(Murray 1935)-because Henenu took with him an army and sent out scouting parties when he led an expedition to reopen the Wadi Hammamat for voyages to Punt in year 8 of Seankhkare Mentuho- tep III (Hayes 1961; Breasted i906:209).

One cryptic fragment of text might suggest that the eastern desert was, in the later reigns of Dynas- ty XII, typically less dry than it is today. A docu- ment containing copies of several dispatches from one Nubian fort to another was found in Thebes; the dispatches report the comings and goings of Nubians who came to Semna and to other forts to trade their wares, and also mention steps taken to

keep track of the movements in the desert of the

Medjay people (P.C. Smither 1945 JEA 31). In the publication of these dispatches, dated to year 3 of an unnamed king, probably Amenemhet III, Smither remarks: "It is surprising that it should

have been thought necessary to report such trivial activities officially to higher authorities and to other fortresses." Only one of these dispatches, No. 5, from Elephantine, dated to year 3, III Proyet 27 (early June 1839 B.C. if the assumption of Ame- nemhet III is correct), appears relevant to climate conditions. This concerns a small group of Medjay who come from the desert and express a wish to serve the "Great House" (Pharaoh) ". . . A ques- tion was put regarding the condition of the desert. Then they said, 'We have heard nothing at all; (but) the desert is dying of hunger'-so they said." After some argument, they were dismissed to their desert. If this event was deemed worth a report, it would suggest that the eastern desert and Red Sea Hills32 had suffered one or more exceptionally dry winters. It is tempting to see in it a hint too that unusual aridity in the desert may have played a part in the menace against which Senwosret III cam- paigned and strengthened his fortifications in Nu- bia.

Nubia. In a previous section, I advanced the hypothesis that the great floods between 1840 and ca. 1765 B.C. were swollen by a greater northward penetration of the monsoon rains than occurs today. It is therefore of particular interest to consider the fragmentary evidence bearing on rainfall in Nubia during the Middle Kingdom.

In the west Nubian desert evidence has been noted by Murray (i95I) against any significant rainfall from early dynastic times to the present. In particular he calls attention to bodies found, between Aswan and Kuban, in shallow graves from the late predynastic and early dynastic era in so remarkable a state of preservation that details of gross anatomy of the eyes and brain could still be identified. However such remarkable preserva- tion is not usual in Nubian burials. Murray notes further a copper tool dated to the Old Kingdom, found lying on the desert near the Khafre diorite quarry-in the Nubian desert ca. 130 km. NW of Toshka-in a state of excellent preservation when found but now green in Cairo, as evidence of the aridity of the region over the past 4500 years. Mur- ray inferred also a lack of wind and sandstorms

32Adams (i971) points out "The mere fact that people could live in the 'desert' would in itself be sufficient to indi- cate that there was some rainfall; otherwise human habitation would have been impossible . . . the hieroglyphic character which is translated as 'desert' must be presumed to refer to the whole Nile hinterland, including the Red Sea Hills which

have always received some rainfall. These, rather than the genuine desert, have always been regarded as the habitat of the Medjay (modern Beja). Hence the inscription would seem to refer to a failure of the rains in the Red Sea Hills, without reflecting in one way or another on the normal conditions of the true desert."

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 28: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 249

from the fresh and uneroded appearance of in- scriptions on stelae bearing the names of Khufu (Cheops) and Sahure, but the significance of this observation is uncertain, because he did not report the direction the inscriptions were facing.

In these same diorite quarries in the western desert of Nubia have been found inscriptions of the names of Khufu and Dedefre of Dynasty IV, Sahure and Djedkare-Isesi of Dynasty V, but no names from Dynasty VI when alabaster largely re- placed diorite in Egypt (Trigger 1965:80). Murray (1951) attributes the discontinued use of the diorite quarries, apparently ca. 2400 B.C., to the failure of their wells. Perhaps the ground-water around the wells had become depleted by over-use, for it would appear from the bodies mentioned above that the NWP in this general area had ended already by ca. 3000 B.C., but after a few centuries of disuse-- perhaps aided by greater floods and some increase in rainfall early in the Middle Kingdom-had re- covered enough to make reopening of the quarries practical for a time in the Middle Kingdom. The names of Dynasty XII kings Amenemhet I, II, III, and Senwosret I, II have been found at this quarry (A. Rowe 1939 Ann. Service), but signifi- cantly (?) not that of Senwosret III. After Ame- nemhet III the quarries were apparently abandoned for good, and sooner or later were quite forgotten.

I say "significantly (?) not Senwosret III" be- cause the evidence from Semna South may suggest that during his reign the floods were not signifi- cantly greater than modern floods, probably below the "good flood" of Senwosret I and some 8 m. (at Semna) below a number of the floods of Amenem- het III. It is tempting to postulate that the land of Kush, apparently peaceful after its defeat by Senwosret I, was troubled by drought-diminished rain and/or lower Nile floods-in the time of Sen- wosret III, who found it necessary to conduct at least four campaigns against Kush and felt im- pelled to strengthen and add to the system of forts in the region of the Second Cataract. This might be a question of rainfall as well as of lower floods, particularly if we accept the hypothesis that floods substantially greater than the modern level indi- cate a more northerly reach of the monsoon rains than occurs today.

Several investigators have pointed to evidence that the people of the Nubian C-Group culture kept

large herds of cattle as indicating that the climate of Nubia in the Middle and early New Kingdom times must have been rather less arid and thus permitted the growth of more vegetation than at present (Arkell 1961:66; Emery 1965:159). How- ever Adams (1971) points out that "vast herds" may be an "archaeologist's myth." What the ar- chaeological evidence proclaims is not that cattle were abundant but that they had high social and ritual significance-and probably therefore that they were not highly abundant. Whatever else the C- Groupers were, they were a fully sedentary, vil- lage-dwelling people commanding substantial ma- terial wealth-conditions which we do not expect to find among pastoralists."

In describing his excavations of the Middle Kingdom fort at Buhen, begun by Senwosret I, Emery (1965:1o9) further states: "The two arterial roads, which were paved with stone and burnt brick tiles, each had a drain runnel down the cen- tre, which suggests that Nubia must have had a greater rainfall at that time than it has now; for such a feature would be unnecessary at the present day." To the best of my knowledge, this feature has not been found in other forts, although a num- ber of drains running from particular rooms are known at Uronarti (Dunham 1967:24, 30; Plate XVIIA,C; map VI), and at Mirgissa (Dunham 1967:148-52, map XVII; Vercoutter 1966:138, n. 37). We may conclude, provisionally, that cloud- bursts were somewhat more common than they are in the present century, perhaps something more akin to nineteenth century conditions as described by Wilkinson (1835) and by Lepsius (1853) for Upper Egypt.

LAKE MOERIS AND THE FAYUM

It is now appropriate to discuss what is known and what can be most reasonably inferred about the level of Lake Moeris in the Fayum province from the beginning of dynastic Egypt ca. 31oo B.C. to the time of Herodotus ca. 450 B.C. It is ap- propriate that this be done in a paper on the climate of the Middle Kingdom because the pharaohs of this era were particularly active in the Fayum prov- ince and because we need to investigate whether the lake might have functioned as a flood reservoir and thereby account for the large difference be-

33In imagining "vast herds" archaeologists are perhaps in- fluenced by the claim of King Sneferu (ca. 2600 B.C.) to have

captured "200,000 large and small cattle" by war in Nubia (Breasted I906:66).

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 29: Climate and the History of Egypt -The Middle Kingdom

250 CLIMATE AND THE HISTORY OF EGYPT [AJA 79

tween the Elephantine and the Cairo levels given for a "good flood" by Senwosret I.

Egyptologists agree that Amenemhet III built

extensively, including irrigation and/or land-recla- mation works, in the Fayum province. But since

only the most fragmentary remains have been found, there is no agreement about exactly what he built or about the particulars of his irrigation and/or reclamation works, or even about the ap- proximate level of the lake in these years. Herodo- tus (II, ioi and 148-49) credited a King Moeris-- usually identified with Amenemhet III"-with the construction of a vast artificial lake: "Marvellous as the Labyrinth is, the so-called Lake of Moeris be- side which it stands is perhaps even more astonish-

ing." We should note particularly that he says, after visiting the lake in person, that the Labyrinth stands beside the lake, not at some distance from it. He further reports that the lake was 3600 furlongs in circumference and 50 fathoms (about 90o meters) deep in the deepest part, which would make it a vast lake filling the major part of the Fayum de-

pression; and that for six months the water flowed from the Nile into the lake, and for six months flowed in the reverse direction. The Birket Qarun, with its surface at -45 m., below sea level, is the

meagre modern remnant of Lake Moeris. Modern geologists agree that the Fayum depres-

sion itself is a wholly natural formation, not a hu- man construction, which attained its present shape at least some millennia before historic times. But they agree on few other aspects of the history of Lake Moeris. Although the Fayum has been stud- ied by a number of able and distinguished schol- ars, both geologists and archaeologists, the level of the lake at various eras between the Neolithic and the Ptolemaic has long been the subject of contro- versy and incompatible views. Although a defini- tive history of the lake requires further field study with the most modern procedures, it nevertheless appears worthwhile to outline here a history of the lake during the dynastic period that seems most in harmony with presently available archaeological evidence and with our general concepts of climate variation. Ball (1939) provides a convenient his- torical summary of the various conflicting theories of the lake which I shall not detail here.

It is generally agreed that the Fayum lake rose from a very low level in the late Paleolithic to around +18 m. above modern sea level when high floods broke through the Hawara channel, linking the lake basin to the Nile, and that the lake was in free connection with the Nile during the Neo- lithic period. Almost immediately thereafter a divergence arises in scholarly opinion. According to some, Sir Hanbury Brown and Sir Flinders Petrie among them, the lake remained thereafter at a relatively high level, fluctuating annually in free connection with the Nile from Neolithic times down to the age of the Ptolemies. This concept was challenged by Caton-Thompson and Gardner (1929) who found artifacts and habitation sites of what they considered to be an early to middle Neo- lithic (Fayum A) culture just above the +18 m. contour, while they found what they believed to be late Neolithic (Fayum B) artifacts on beaches at +Io and -2 meters. They accordingly argued that the lake, after rising to +18 m. in the time of the Fayum A culture, then declined by stages, with prolonged pauses at the +o0 and -2 meter levels (where they also found some artifacts attributed to Dynasty IV), and never did regain a high level or free connection with the Nile either in the time of Dynasty XII or in the time of Herodotus.

The necessity for postulating a decline in late- Neolithic times has, however, recently been re- moved by radiocarbon dates which indicate that the Fayum-B culture is terminal Paleolithic and about a thousand years older than the Neolithic Fayum-A culture (Wendorf et al. 1970), quite the reverse of what had been generally believed since the work of Caton-Thompson and Gardner. From these new radiocarbon dates, it seems clear that the lake had risen by Neolithic times, in two (or more) stages, stabilizing for a time at the -2 and +-10 m. levels where the Fayum B artifacts were found, then rising later to the +18 m. level, with Fayum-A sites along the shore to the north of this +18 m. lake. And indeed, Said et al.35 (1972), in their re- cent examination of several sites along the north- ern beaches, obtained evidence for four relatively high levels, progressively increasing from ca. 6150 to ca. 3900 B.C. (uncorrected C" dates), and sepa- rated by intervals of lower lake levels. This sug-

34From his throne name, Ny-maat-re. But it has also been suggested that the name Moeris derives not from the name of any king but from Mi-wer, the name of both a town on the lake and of the canal linking the Nile with the lake (Ed-

wards I96x:230). 30 One of the co-authors, Prof. C.C. Albritton (Southern

Methodist University), kindly provided me with a copy of this article in page-proof.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 30: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 251

gests that a number of climate fluctuations affected the volume of the Nile during the Neolithic pe- riod. By the time of the Fayum-A Neolithic cul- ture ca. 3900 B.C. the lake was typically around 16.5 m. and at times as high as 20 m. above sea level.

Ball (1939) had developed a compromise view and suggested that the claim of Herodotus that King Moeris had the lake constructed is best un- derstood as meaning that the king caused the lake to be refilled"6 by clearing out the Hawara channel which had become silted up-in late Neolithic times according to Ball, who to this extent accepted the interpretation of Caton-Thompson and Gard- ner. Such a silting, however, is clearly contrary to the evidence obtained by Said et al. (1972) for a high lake level during the Old Kingdom. It is still possible that the channel was silted up at the be- ginning of Dynasty XII, but if so I suggest it most probably occurred late in the Old Kingdom or during the great drought (Bell 1971) of the First Intermediate Period.

One of the stronger arguments in favor of a high lake level from Middle Kingdom, or even from Fayum-A, to Ptolemaic times is the fact that no pre-Ptolemaic towns or their ruins have been found below +18 m. (Shafei 1960). In the words of Wil- son (1955 INES 14:219): "The ancient sites in the Fayium thus fall into two classes: those running from the Middle Kingdom onward, at or above the 20-meter contour line, and those from the Ptolemaic and Roman period, at or above the o-meter line." This view is supported by the research of Said et al. (1972). In addition we have the eyewitness tes- timony of Herodotus.37 The explanation of Caton- Thompson and Gardner (1929) that Herodotus mistook flooded fields for part of a vast lake is not persuasive when the depth he gave is reasonable and when he had seen many flooded fields in the region of Memphis and Giza.

If the Hawara channel became clogged during

the First Intermediate Period, due to insufficient floods (Bell 1971) and/or drifting sand (Butzer 1959a), the lack of Middle Kingdom sites below +18 m. would indicate that the king who restored the lake to free communication with the Nile by removing accumulated silt and blown sand from the Hawara channel, thus widening and deepening it, must have been Amenemhet I, as Ball (1939) argued, rather than Amenemhet III. Indeed in the time of the ultra-high floods attested by the Semna inscriptions the lake would almost surely have re- covered the free connection by itself without hu- man aid. Once the channel was clear enough to permit free communication with the river, then according to Ball (1939:Ioof):

The lake . . . functioned as a combined Nile flood-escape and reservoir, not only protecting the lands of Lower Egypt from the destructive effects of excessively high floods, but also increas- ing the supplies of water in the river after the flood-season had passed. . ... Amenemhat's [pri- mary] object in improving the channel of com- munication between the Nile and the lake must have been to provide an escape for excess flood- water, rather than to secure increased supplies in the river at the low stage; for the Ancient Egyp- tians, practicing as they did only the basin or flood system of irrigation, are not likely to have been particularly concerned about increasing the low-stage supplies of the river, while they would naturally be anxious for protection against the wide-spread damage, in the shape of breaches of the river-banks and destruction of houses and gardens, that resulted from very high floods....

... One of the great merits of the work carried out by Amenemhat [was] that the higher the Nile flood, the greater would be the proportion of the flood-waters that would automatically es- cape into the lake; for the greater the height to which the river rose, the greater would be not only the cross section of the escaping stream, but also its slope, and therefore its velocity .... Once the canal had been dug, the escape and the return flow would take place automatically, so that there would be no necessity for artificial regulation; and

36 Ball calculated that only about 5 years of average-modern floods would be needed to refill the lake from -2 to +16.5 m. and bring it into equilibrium with the low Nile, if it re- ceived about ten percent of each year's flood waters. Ball fur- ther believed that the lake remained large and in free com- munication with the Nile from early in Dynasty XII until it was reduced to sea level by land reclamation projects of the early Ptolemies in the third century B.C.

7 On behalf of the accuracy of Herodotus and the reality of his visit to the site, it is worth recalling that he mentioned an underground passage into the Hawara Pyramid in the corner of the Labyrinth adjacent to the Pyramid; attention to

this claim would have saved nineteenth century excavators of this Pyramid much labor. For our period he has also a reason- able order of the kings: (i) Moeris, "builder" of the lake (Amenemhet I); (2) Sesostris, the great conqueror (Senwosret I and III); (3) Pheron, afflicted by an excessive flood, see be- low (Amenemhet III). Not surprisingly, he does have them dated wrong, to 900 years before his own time, whereas the reality is around I5oo. Petrie (1927 Antiquity i) pointed out that in general Herodotus's History of Egypt would be vastly improved if the sections 124-36 were placed before 00oo-123, on the assumption that at some time in the past these segments of the manuscript had become erroneously transposed.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 31: Climate and the History of Egypt -The Middle Kingdom

252 BARBARA BELL [AJA 79

indeed it is noteworthy that Herodotus makes no mention of any structures for regulating the flow of water in the channel, as he would almost cer- tainly have done had such existed in his day. Even silting-up of the Channel would be nat- urally prevented; for the silt deposited in it dur- ing the time of slack water in any year would be scoured out again by the rapid rush of the water at the next flood season.

Ball further notes that the first mention of sluices for regulating the flow appears to have been made by Diodorus around 44 B.C., although Shafei (1960) maintains that the flow was regulated throughout Pharaonic times from the reign of Menes. If such a construction as postulated by Shafei actually existed, perhaps traces of it may someday be found by a magnetometer survey of the Hawara channel.

Ball's judgment that the channel, once dug or naturally formed, would keep itself clear and free of accumulated silt makes it doubtful that the chan- nel, once formed in pre-Neolithic times, with a generally accepted high-lake level at about +18 m. in the Fayum-A period ca. 3900 B.C. by the new (uncorrected) radiocarbon measurements (Wen- dorf et al. 1970), could silt up again except in a period of very weak floods. There is an indication -from the early A-group fireplace found by Save- S6derbergh (1964 Kush 12:25) at Askeit buried be- neath some 120 cm. of Nile silt within which were late A-group graves-that the floods were lower for a time in the late predynastic than in the early dynastic period."8 But this evidence does not indi- cate whether they were sufficiently lower to cause siltation of the Hawara channel.

Early Dynastic. Menes, legendary founder of the united kingdom of Upper and Lower Egypt-- ca. 3ioo B.C. by the chronology of the revised Cam- bridge Ancient History-may have done something to restore the connection, or to reduce the inflow, depending upon the then level of the floods and

condition of the channel. Herodotus reports that the priests told him that the Nile used to flow along the base of the sandy hills of the Libyan desert,"'' until King Menes closed this channel at a distance of ioo stades (about 20 km.) upstream from Memphis and diverted the river into a course more or less midway between the two deserts. Sha- fei (1960) believes that this tradition derives from Menes having caused the Egyptians to build an earthen dam or weir at Lahun to control the amount of water flowing into the Fayum basin, either to reduce an excess of flood-water or to pre- vent the Fayum lake from depriving the fields of Lower Egypt. However we have seen that Ball believes that a reasonably open channel would auto- matically act as a regulator of the flood in Lower Egypt, and that a dam would not serve any useful or necessary function. Shafei further considers that the main channel of the river originally ran in the present Bahr Yusef but was deflected to a more easterly course by the works of Menes. Because of the tradition, although evidence is lacking, it seems likely that Menes did do something remark- able about the floods, and whatever he did may have played no small part in establishing the dogma of the divinity of the King of the Two Lands and the popular belief in his control over, and respon- sibility for the Nile flood and the proper order of Nature (see Bell 1971:19).

Old Kingdom. There is good evidence that the lake attained to a level of 20 to 21 m. during Dynas- ty III, IV, and much of Dynasty V, to ca. 2350 B.C. One persuasive item is a man-made road to the north of the Birket Qarun. The road runs about 6.5 km. from a basalt quarry southward to end in a "great elongated dump of colossal, weathered basalt blocks" (Caton-Thompson 1927 Antiquity i) on the north shore of the 21 m. lake, near a tem- ple at Qasr-el-Sagha. Shafei (1960) gives the fol- lowing elevations above sea level, all arguing for

38 Although it is outside the period we are investigating, it is of interest to estimate the magnitude of flood implied by this evidence while we have the gauge levels conveniently at hand. Sive-S6derbergh (1964 Kush 12:24-25) reports that the fireplace was found at about 5.7 m. above the present (1963 March 20) level of the Nile, "on the coarse sand and covered with a layer of 120 cm. of silt. Thus, some time between the Early A- Group and the Late A-Group the level of the Nile must have risen (about 7 m. above the present level) and deposited more than a metre of silt." Now the level of the Nile on 1963 March 20 was I19.55 m. at Halfa (courtesy of the Sudanese Ministry of Irrigation), so that at the time when the layer of silt was deposited the HWL must have been at least 126.5 m. By Table 2 this is seen to be similar to the flood of year 15 of

Amenemhet III, or a little more because the site is a few km. downstream of Halfa, perhaps reading I27.0 m. at Halfa. Thus the deposition of the 120 cm. of silt between early and late A-Group times at Askeit would require floods at least as large as the smaller floods recorded at Semna in the reign of Amenemhet III some 1200 years later. It is to be hoped, that as the method becomes more perfected, thermoluminescent dating of pottery fragments from this site will make possible a more exact dating of this period of very high floods.

" Presumably in the course of the modern Bahr Yusef, now a minor branch of the Nile in Middle Egypt which drains into the Fayum lake; and in the judgment of Butzer (1971) never more than a minor branch.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 32: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 253

a lake level of 20 to 21 meters when they were built or in use, which was most probably during the Old Kingdom:40

paved floor of Qasr-el-Sagha temple +35.00 m. top of eastern Qasr-el-Sagha ramp +20.48 m. top of western ramp +22.04 m. top of the dolerite dump +20.72 m.

Surviving fragmentary records indicate (Bell 1970) that the floods of Dynasty II to early V averaged 70 cm. (or more, depending on the as- sumption about the zero of the scale) lower than those of the First Dynasty. Allowing half a meter or so for the rise in the level of the floodplain and lake bed from Dynasty I to IV, it would appear that the level of the Fayum lake in Early Dynastic times must have been about the same as during Dynasty IV and V, or 20 to 21 meters.

As shown previously (Bell 1971), the collapse of the Old Kingdom was accompanied (whether caused or not) by a number of years of very weak floods and it is possible that the Hawara channel became silted up so that the lake level fell sub- stantially. On this point clear evidence is lacking. Surely relevant, although still lacking precise dat- ing, is the Gisr el-Hadid, an impressive beach ex- tending nearly 50o km. along the southwest of the Fayum depression at RL +22 to 25 m., first recog- nized as a significant feature and investigated by Little (1936:209f). He was subsequently able to trace this feature nearly all the way around the Fayum, with only a few interruptions, and to identify the Idwa Bank as part of it.

Little's more detailed studies were made at a point some 8 km. WSW of Qasr Qarun. There among the sands and gravels of the Gisr el-Hadid, under a few cm. of blown sand, he found frag- ments of pottery (B-series, not to be confused with Fayum-B of Caton-Thompson and Gardner) of post-Neolithic, dynastic but otherwise uncertain date, at RL 22-24 m. And further, Little (1936: 209f) reports:

Underlyifig the Gisr el-Hadid is a stratified sand- rock series containing in places Pottery A, Dynas- tic, RL 17-22 meters. ... We cannot avoid the conclusion that the Pottery A and Pottery B de- posits represent stages of the historic Lake Moe- ris. [Excavations on the reverse slope of the beach ] show that in many localities the Gisr rests on wind-borne material. This is chiefly a rather fine yellow sandrock, structureless on the whole, and probably only affected by the lake to the extent of being slightly consolidated. . . . Evi- dence for flourishing vegetation is seen in very numerous calcified rootlets and parts of the trunks of large trees, which must have grown by a lake side upon deposits of an earlier lake. . . . These underlying deposits are historic [containing frag- ments of Pottery A]. The Gisr, in its transgres- sive character, rests also on deposits probably older than those of historic date. These older deposits are (a) a greenish sandrock, rather bet- ter stratified (but showing a tendency at the top to pass into a yellower type of more wind-borne appearance), with a fauna distinctive from any of the foregoing, and (b) a travertine-cemented conglomerate of beach-origin, the Beach Conglom- erate, composed of limestone pebbles, and richly fossiliferous. The fauna of the Beach Conglom- erate as well as its field relations unite it with the upper, yellowish part of the greenish sand- rock, which latter is stratigraphically the oldest lake-sand deposit preserved [and probably Paleo- lithic].

The features of primary interest to us here are the Gisr beach itself at 23 m., the layer of wind- borne sand beneath it, and the next-down layer of slightly consolidated sandrock with Pottery A de- posits at RL 17-22 m. It is tempting, in absence of more precise dating, to link the Pottery A to the Old Kingdom, the evidence of trees to the Neo- lithic Wet Phase ending ca. 2350 B.C.; the wind- borne sand to the First Intermediate Period and Butzer's (I959a) drifting dunes of sand invading the Nile valley of Middle Egypt from the west at the time of the weak floods (see Bell 1971); and the Gisr itself with Pottery B to the high floods and high lake levels of the Middle Kingdom and per-

40 Because it has no inscriptions, the temple at Qasr-el-Sagha cannot be dated with certainty, but it has been attributed on stylistic grounds to late Dynasty III (P. Gilbert 1944 Chron. d'Egypte).

Also relevant to the dating is the fact that, from his analysis of various Egyptian rocks, Lucas (1930 JEA 6) concluded that the basalt or dolerite used in a number of Old Kingdom monu- ments most probably came from the Fayum quarry at the north end of this man-made road. In particular he noted that this basalt was used in the pavement of the mortuary temple of Khufu (Cheops) and in the sarcophagus of Menkaure of

Dynasty IV, and in the pavement of two Dynasty V mortuary temples at Abusir, and further pointed out that basalt was widely used in Egypt only during the Old Kingdom. It would appear that the basalt quarry was abandoned about the same time as the diorite quarry in the Nubian desert (Murray 1939), at about the end of the Neolithic Wet Phase. It may well be that the abandonment of the basalt quarry was caused by declining Nile floods and a silting up of the Hawara chan- nel, which made transportation of the large basalt blocks too difficult, during the VIth Dynasty. But tangible evidence is lacking.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 33: Climate and the History of Egypt -The Middle Kingdom

254 BARBARA BELL [AJA 79

haps also to the time of Herodotus and his Lake Moeris. Recent findings (Said et al. 1972) of pot- tery believed to date from the IVth Dynasty, and buried within lake sediments at an elevation of 22-24 m. RL at the foot of the Old Kingdom tem- ple of Qasr-el-Sagha, lend support to this inter- pretation. However, radiocarbon dating of shells and fossils, and thermoluminescent dating of pot- tery fragments found in the Gisr and its underly- ing strata will be necessary before the Gisr features can contribute their full potential to the history of Lake Moeris.

Middle Kingdom. If, as I postulate, the lake had fallen to a lower level during the First Dark Age, it had recovered at least to a level of 20 m. by early in the Middle Kingdom. Whether this recovery oc- curred spontaneously with the return of more vigorous floods under Dynasties X and XI, or only with the aid of cleaning operations on the Hawara channel ordered by Amenemhet I, we have no evidence. Petrie (1889) concluded that Amenem- het probably reclaimed from the lake the site of the capital, called Shed "the extracted" or "sepa- rated," on the site of the modern Medinet el Fay- um, and thus established the "land of the Lake." However it now appears that the town of Shedet was already known in the Old Kingdom as a cult center of the god Sobek (Hayes 1961:50). The work, then, of Amenemhet I and subsequent XIIth Dynasty rulers consisted in reclaiming additional land around the town for agricultural use by build- ing dikes and canals for irrigation and drainage. It is considered significant of his special interest in this district that Senwosret II chose to build his pyramid just north of the narrow defile in the hills of the desert which constitutes the Hawara channel, at El-Lahun, a name meaning "the Mouth of the Lake" (Kees i96i:214). However the major development of the Lake District was in all prob- ability the work of Amenemhet III, who was re- nowned there in Classical times, in legend, "for having been the first to make the Fayum accessible to mankind" (Kees i96i:219). Some archaeologists have suggested that the Idwa Bank was part of the dike system built by Amenemhet III, but geologists (Little 1936, Ball 1939, Caton-Thompson and Gard- ner 1929) agree that the Idwa Bank is beyond doubt a natural feature. It remains possible, how- ever, that Amenemhet III could have utilized the Bank as part of his dike system.

At Biahmu, on the Idwa Bank-their location

indeed suggesting he made some use of the Bank- Amenemhet III erected a pair of colossal seated statues of himself, which were seen and reported by Herodotus as seated atop pyramids ioo meters high in the middle of the lake. In fact the colossi were not in deep water, if in any at all, and Shafei (1960:204) offers a plausible explanation for this error by Herodotus. He points out that the lake was divided administratively into a northern and a southern lake, as indicated by a stele found by Daressy (1899 Ann.Serv. 1:44-46) among the ruins of the village of Quta, at the west end of the dynastic lake (above RL +21 by Shafei's map). The left column of this stele reads "Boundary of the southern lake of the god Sobk"; the right column reads "Boundary of the northern lake of the god Sobk"; and the central column reads "This stele has been (erected) on the shore of the lake by the head of the village of Nekht." Thus, Shafei sug- gests, the priests were telling Herodotus that the colossi stood on the boundary between the north and the south lakes, "in the middle" between the northern and southern administrative divisions; whereas Herodotus took "in the middle" to mean in the deepest part-whence he would infer the statues sat atop ioo m. pyramids by adding the 90 m. depth of the lake to visible statue bases of some io m. height.

Another item from Herodotus may be noted here. He wrote that when Sesostris-a king to whom he attributed such incredibly wide-ranging military campaigns that some scholars have seen him as a legendary fusion of Senwosret I and III and Ramses II, and perhaps Thutmose III-"When Sesostris died he was succeeded by his son Pheron [see n. 37], a prince who undertook no military adventures. . ... One year the Nile rose to an ex- cessive height, as much as eighteen cubits (9-4 m.), and when all the fields were under water it began to blow hard, so that the river got very rough. The king so far forgot himself as to seize a spear and hurl it into the swirling waters, and for this act of presumption ... he became blind. . . ." Any attempt at interpretation of this passage must clear- ly refer to the flood below the Hawara channel into the Lake Moeris. If we recall the 6.6 m. given by Senwosret I for a "good flood" in the vicinity of Old Cairo, a figure of 9.4 m. on the same scale for one of the great floods commemorated at Semna seems entirely reasonable.

Shafei (1960:196) estimates that the high and the

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 34: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 255

low Nile, near Beni Suef at the entrance to the Fayum, were about 17.8 and 24.2. m. respectively during the IVth Dynasty, with the lake fluctuating annually between i9.o and 21.0 m., given annual floods of modern volume. These levels would rise to 18.5 and 24.9 m. for the Nile and 18.7 to 21.7 m. for the lake during Dynasty XII, and to 19-7 and 26.1 for the Nile and i9.9 and 22.9 for the lake in the time of Herodotus, due to the rising of the river- bed and floodplain caused by the annual deposition of a thin layer of silt. In the case of the great floods commemorated at Semna between 1840 and ca. 1775 B.C., we would then expect a HWL in Egypt some 4 to 6 m. above normal (from Table 2, Halfa column), and if the high water had any substantial duration-as it must if it came from the Blue Nile -we might expect a lake level 2 to 3 m. above nor- mal, or about 24 m., in good accord with the high- est post-Neolithic beaches found by Little (1936) and by Said et al. (1972).

Let us return for a moment to Amenemhet III's colossi at Biahmu. Only fragments of their bases have survived to modern times. These were ex- amined by Petrie, who noted two items of interest in connection with the ultra-high floods that oc- curred in the reign of Amenemhet III. Around the foundations, Petrie (1889:56) noted that the foun- dation "ground is black mud with strata of coarse sand irregularly through it; the sand is native and is so coarse and clean that it must have been brought in by the bursting of chance dams in the entrance of the Fayum, which let a mass of water in that swept all before it, and brought with it a rush of desert soil." In addition Petrie found a fragment of an in- scription that mentioned restoring works that were injured. Authorities (Ball, Shafei, Caton-Thomp- son and Gardner) agree in giving the elevation of the pavement of the courtyard in which the colossi stood as 18 m. above sea level, and the latter, in their argument for a low lake level, emphasize the lack of Nile silts in the interstices and the lack of water rings on the stones of the bases of the colossi. In view of other evidence for a lake level as high as 23 m., the lack of water rings would seem rather to testify to the formidable dikes-part natural (Id- wa Bank) and part constructed-in the reclamation work of Amenemhet III. The statues, located on

the Idwa Bank, were perhaps intended to guard, with the divine power of the Pharaoh, the re- claimed lands from excessive flooding.41

In view of the extremely high floods recorded in a number of the years of Amenemhet III, it is reasonable to infer that his legendary work of rec- lamation in the Fayum involved both the reclama- tion of swamplands by improved drainage for nor- mal years, and the reclamation of portions of the lowest desert by irrigation works which could be used only in the years of the great inundations.

As to the level of Lake Moeris after Dynasty XII, I have found no information on its condition until the description by Herodotus ca. 450 B.C. However it appears that the Fayum was a popular hunting ground, especially for water birds and marsh birds, during much of the New Kingdom (Kees i961: 218), as indeed it was during most of the pharaonic period (ibid.:226). A "central island" of the Fayum is known from the New Kingdom, referring proba- bly to the reclaimed land around the capital, and suggesting a continued high lake level. While the lake may have had its declines in periods of low Niles, the lack of pharaonic ruins below 18 m. strongly suggests that the lake stood at a high level in free connection with the Nile during all the pe- riods when there was significant building activity in Egypt.

Lake Moeris and the levels of the Nile. In this section we shall estimate the effect of a free con- nection between Lake Moeris and the Nile on the high and low water levels of the Nile downstream of the Hawara channel. In particular, we shall con- sider whether the two figures given by Senwos- ret I for a "good flood" can be reconciled by the ac- tion of the lake, thus whether the "good floods" were probably significantly in excess of the modern or whether it is necessary to assume that the zero of the Elephantine scale was one or two meters below the modern (and ancient?) LWL.

The area of the Fayum basin at the 20 m. con- tour level is 2100 km.2 or 2IxIo8m.2 (Ball 1939:203). If the lake level typically fluctuated over about 5 meters, from 18 to 23 m., it could take in about IooxIo"m." of water from the Nile flood. The loss by evaporation is estimated (Ball 1939) at i80 cm./ year, similar to the modern Birkit Qarun, or about

41 Each colossus stood in a small courtyard surrounded by a wall of 6 courses of stone 12-14 ft. (about 4 m.) high (Petrie 1889: plate XXVI). Thus the tops of the walls would have been at about 22 m., which seems not high enough for

protection from the great floods, nor from the highest post- Neolithic lake levels detected by Little (1936) and by Said et al. (1972), these being at 22 to 24 m.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 35: Climate and the History of Egypt -The Middle Kingdom

256 BARBARA BELL [AJA 79

4oxio"m.3 If we assume that about I5xIo8m." is evapo-

rated during the flood season, then with a range of 5 m.42 the lake could actually take in about Ii5xIo"m." It would then have about 75xio"m3. to return to the river over a period of 9 months,43 or about 8.3xIo"m.3/ month.

Now the capacity of the Aswan reservoir after 1935 and the last raising of the height of the old dam was some 50xio"m.", and this, released gradu- ally over the months February through June acted to raise the LWL of the Nile by about o.8 m. (see Ball 1939:7). It is evident that Lake Moeris could do no more and would probably do somewhat less, because the outflow of the Aswan reservoir is regu- lated to be greater in the months when the natural river is lowest, but the lake would surely drain most briskly when it was relatively full.

Trial calculations44 confirmed that it was not possible, under any plausible pattern of the annual variation of the Nile, to reconcile the Elephantine and the near-Cairo figures given by Senwosret I. The lake is simply not large enough-even at a

range of 6-7 m., and one can hardly conceive of it

having a larger range than the Nile itself down- stream of the Hawara channel-to take off enough water at the flood crest to have enough to return over the remaining 9 months to lower the HWL and raise the LWL below Beni Suef to 6.6 m., if the

range above Beni Suef were io.o m. (11.3 m. at Aswan implies about 1o.o m. at Halfa, typical of

Egypt itself). We are forced therefore to conclude that if the

zero points of Senwosret I's Nilometers at Elephan- tine and near Cairo each represent a LWL, the two Nilometers must have been established at different times. The Aswan Nilometer indeed must have had its scale fixed in a period of exceptionally low LWL, such as that reflected in the Prophecy of Neferty, in the drought immediately preceding the

founding of Dynasty XII, or possibly in the ear- lier drought between ca. 218o and ca. 2130 B.C.

The lowest LWL in modern records occurred in 1922, with a volume of 0.24xIO0 or 24 million m. per day and an estimated gauge level at Aswan of 83.5 m. When "the river of Egypt is empty, men cross over the water on foot," one is justified to ex- pect considerably less than 24 million m.3/day, and one might reasonably postulate a LWL of 83.0 m. or even less. With a LWL at 83.0 m., we have HWL at 94-3 m., and a volume of I2.5xio"m."/day for a "good flood" in the reign of Senwosret I, only slightly larger than the largest floods of the past century.

At some point in his reign Amenemhet I moved his capital from Thebes to Itj-towy in the north (Simpson 1963 JARCE 2:57-59), and it is reason- able to assume that the northern Nilometers were established at that time, after the drought had passed and the LWL returned to near its modern level of 84.0 to 85.0 m. at Aswan, giving an an- nual range of around 9.3 m. there, but I1.3 m. above the zero of the older Nilometer; and with a range of about 8.o m. at Halfa and in Egypt above the Hawara channel, and 6.6 m. below the channel to the lake. Thus we would have in the early part of Dynasty XII above Beni Suef a HWL similar to that of the modern period A.D. 1870-1898 and a LWL similar to that for A.D. 1900-1960; below Beni Suef we would have a HWL similar to that since A.D. 900oo, and a LWL similar to that of A.D. 1870-1899, the somewhat greater range of the floods in the i9oos B.C. being moderated below Beni Suef by the free connection with Lake Moeris.

In the latter part of the Middle Kingdom, as we have seen, there occurred a number of years of ex- tremely high floods, as much as 4 to 6 m. above the modern-normal in Egypt if the excessive waters arose from the usual source of Nile flood waters, the drainage basins of the Blue Nile and the Atbara

42 Actually Ball (1939:139f) estimates that a 7 m. flood would lead to a seasonal fluctuation of only about 2.5 m. in the level of the lake. From its geological section (Little 1936) the Hawara channel, if free of silt, had a width of 500 to 6oo m. from RL 9 to 20 m. and increased suddenly to 8oo m. at RL 20 m. and gradually to ooo000 m. width by RL 25 m. If this section is typical, quite a substantial portion of a high flood could escape into the Fayum lake in the event of a rela- tively prolonged flow close to the HWL.

43 It should be noted that although Herodotus spoke of the lake rising for six months and falling for six months, this is quite impossible with a seasonal distribution of the Nile vol- ume similar to the modern and more than a 2-3 m. fluctua-

tion in the lake level. And we must have a larger fluctuation to allow any hope of reconciling the figures of Senwosret I.

44 The trial calculations were graphical: I plotted the Halfa gauge readings, and the volume, separately for each io-day interval using the mean of I890-i899 and the mean of 1912-

1927. To these graphs I then added conjectural curves for Senwosret's floods, guided by the need to have a range of Io.o m. above and 6.6 m. below the channel to the lake. The difference between the curves would give the inflow to and the outflow from the lake. The needed inflow and outflow substantially exceeded the capacity of the lake over a range of even 7 m.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 36: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 257

in Ethiopia. On the other hand, if any significant part of the excess waters at Semna arose from a few days of heavy rains over the central Sudan south of Dongola, we could expect a much sharper flood crest-with the very high volumes estimated in a previous section lasting only a few days, con- ceivably only a few hours; in this (admittedly somewhat unlikely) case, the volume would be substantially reduced by the time it reached Middle and Lower Egypt as a result of flooding beyond its normal limits all the way down the valley. It is hoped that future research in Egypt will help us choose between these hypotheses.

In closing this section on high floods and the Fay- um, it is useful to consider accounts of the unusu- ally high floods that occurred in A.D. 1817 and i818 in order that we may form some approximate idea of the trouble caused in Egypt proper by the ultra- high floods recorded at Semna in the reign of Amenemhet III and his immediate successors. I

quote from Ball (1939:23I):45

. Ample evidence that the flood of [A.D. 8I81] was a phenomenally high and destructive

one is ... furnished by the contemporary records ... [which] ... inform us that in the year 1233 of the Hegira (A.D. 1817) the Nile rose to such an extraordinary height at Cairo that the island of Roda was completely submerged, so that boats could sail over it, many villages were destroyed, considerable numbers of the inhabitants and their animals were drowned, and there was great lam- entation among the fellahin over the loss of their summer crops ... ; also that in the following year ... (A.D. i818), there was a still more disastrous Nile flood, the inundation not only reaching to even greater heights than in the previous year and sweeping with much violence over the highest banks and destroying all the field-crops, including the cotton, as well as the fruit-trees in the gardens, but also lasting for an abnormally long period, a first slight fall being followed by a renewed rise to still higher levels after Holy Cross Day (Sept. 27), and the waters not subsiding until the Coptic month of Hatur (Nov. io to Dec. 9), when the season for cultivation was past.

[And regarding the Fayum:] Belzoni tells us that when he visited the Faiyum in May 1819 an extraordinarily high flood had caused such a quan- tity of water to pass into the Birket Qarun that it rose some 31/2 meters higher than it had ever previously been known to do by the oldest fisher- men on its shores. [And Linant de Bellefonds re- corded that] in 1819 or 1820 there occurred a breach in the northern bank of the Bahr Yusef near Hauwaret el Maqta, causing a disastrous rush of water down the ravine known as the Bahr-bela- ma, and despite great efforts it had been found impossible to repair the breach till six months later, after the flood had passed (Ball 1939:231- 32).

Particularly noteworthy in these descriptions-and others (Toussoun 1925)-is the fact that some, but not all, destructively high floods are prolonged at a high level, keeping many fields under water be- yond the normal and proper season for sowing. Such abnormal duration of high waters could re- sult in a poor harvest the following summer. Since scholars agree that the reign of Amenemhet III was a time of high prosperity in Egypt we may in- fer that his great floods were not so usually pro- longed as to interfere seriously with the agriculture.

SOME HISTORICAL IMPLICATIONS OF THE GREAT FLOODS

In this section we shall inquire whether the "his- torical truth" of a number of years of ultra-high floods and a period of more than unusually variable flood volumes can illuminate any aspects of the history of the late Middle Kingdom. Four aspects occur to me: (i) the unique royal statuary of Sen- wosret III and Amenemhet III; (2) the "Complaint of Khakheperre-sonbe"; (3) the popularity of the name-element Sobek among the kings of the next dynasty; (4) and the decline in prosperity, without evidence of severe famine, under Dynasty XIII. In a previous section, I pointed out some inferences which might reasonably be made about the legend- ary reclamation work of Amenemhet III in the Fayum. Let us now consider the above four aspects individually.

45 See also Toussoun (1925:510); unfortunately Toussoun was not able to locate any Roda gauge measurements for these years, nor for any of the years between 18oo and 1823 for his table. In medieval times, the highest flood by a substantial margin occurred in A.D. 1360; ". .. The flood that year was of 24 cubits [5-6 cubits or more than 2.5 m. above the normal of its half-century] . . [the ruler] ordered that they cease proclaiming the height of the flood, because they feared a gen- eral inundation. The great waters sustained themselves thus, without diminishing, to 25 Baba (Oct.), which caused extreme grief among the people. The road of the Fayum became im-

passable, the gardens of Elephantine were submerged. ... The waters spread . . . and destroyed . . . dwellings of the Isle of Roda which ended by being entirely submerged. . . . This frightful inundation remained in full force until the end of Baba; never had such a thing been seen in Egypt, neither before nor since Islam. . . . These great waters were followed by the plague, which ravaged all of Egypt. .... In the year A.D. 1371, the flood was excessive and rose to 22 cubits and more; it re- mained at this level up to the end of Hator (Nov.), which caused the Egyptians much disquiet, because the time for sowing was passed. ...

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 37: Climate and the History of Egypt -The Middle Kingdom

258 BARBARA BELL [AJA 79

(I) Many of the finest portrait statues of the last great kings of Dynasty XII, viz. Senwosret III and Amenemhet III, are distinguished by qualities never found elsewhere in royal Egyptian portrai- ture, by an "astonishing realism ... outside the mainstream of pharaonic art" (Aldred 1970:27). In contrast to the norm of eternal youth and serene dignity befitting a god-king, these rulers are por- trayed as men, sometimes with "the brooding latent power ... of an autocrat ... [Senwosret III] who not only so reorganized the Egyptian possessions in Nubia that he was afterward worshipped there as a protector of the region, but also broke com- pletely the power of the landed nobility at home, re- ducing the nomarchs to the status of crown ser- vants"; and sometimes as middle-aged and without illusions, "the careworn shepherd of his people" (Aldred i970:45). In the words of W.S. Smith (1965:o03): "The dominating quality of these heads is that of an intelligent consciousness of a ruler's responsibilities and an awareness of the bitterness which this can bring. . . . A brooding seriousness appears even in the face of the young Amenemhet III ... it is immediately apparent that this man lived in a different time from that which produced the serene confidence of the people of the Old Kingdom." To the present (non-specialist) writer, the young Amenemhet III seems to say that it is a difficult thing to play the role of a god-king, but he will do his best to be worthy of his dis- tinguished ancestors. The limited evidence now available indicates that he succeeded. And Senwos- ret III appears to have had an unusually active and successful, if untranquil, reign. Politically success- ful rulers, as well as the unsuccessful, may of course have personal domestic troubles, even without a harem of wives and concubines, and the Execra- tion Texts suggest intrigues within the royal house- hold in the latter part of Dynasty XII (Wilson 1956:159). But it is unlikely, and unprecedented, that an Egyptian king would allow personal family problems to be reflected in his official portraits. Montet (1968:67) has wondered if they may have had premonitions of the impending end of their dynasty: "The days of the court of It-towy were numbered .... Perhaps Amenemhet III had sensed that his family and all Egypt were to fall on evil days and the sculptors of Karnak caught these forebodings." If they had forebodings, I suggest that an adequate cause is at hand in the erratic behavior of the Nile, evidenced by the Semna in-

scriptions, combined with a haunting memory of the droughts of the First Dark Age (Bell 1971), to stimulate greater-than-usual anxiety about the level of the Nile flood next year, and the next....

An erratic Nile, with wide variations in flood level and a number of years of ultra-high and destructive floods would provide worries enough for any ruler. How much more so, then, for a god- king of Egypt who is responsible not only for rul- ing the social order of the Two Lands but also for the natural order of things, and above all for the beneficent behavior of the Nile. In my previous paper (Bell 1971), I discussed evidence for this dogma of the pharaoh's responsibility for the good order of Nature, and need not repeat it here. Viewed against such a background, the sternly careworn features of the royal statues may seem not only understandable but more appropriate- forebodings apart-than expressions of godlike se- renity, and thus to be encouraged, not merely per- mitted, by these kings. It is not my intention to argue that erratic and excessive Niles be seen as the sole cause for this style of sculpture, but rather to suggest that these natural phenomena merit consideration as a contributing factor which inter- acted with political and cultural factors.

It may be objected that these attributes are more conspicuous in the heads of Senwosret III than in those portraying Amenemhet III, while only the latter-to our knowledge-was troubled by exces- sive floods. In fact there is little evidence on the Nile levels in the reign of Senwosret III. But the recent discovery, at the Dal Cataract, of an inscription commemorating a water level on 24 Jan. 1869, year 10 of Senwosret III, close to the modern HWL, makes it reasonable to think the Nile may have been more erratic in the reign of this king than under his predecessors. When the river remains high, as in that year, long after its normal time, the fields will not be ready for planting in the nor- mal (November) season and the crops may not have time to mature before the next flood.

(2) Turning now to our second point, written evidence for any sort of disturbed condition in the Middle Kingdom is rare. The next item in time after the famine referred to by Ameny of Beni Hasan, and of uncertain significance, is the Com- plaint of Khakheperre-sonbe (Erman 1927:Io9f; Simpson 1972:230f), whose name suggests that he was born in the reign of Senwosret II. He thus probably wrote his Complaint during the reign of

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 38: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 259

Senwosret III or Amenemhet III. Because the avail- able evidence indicates that these reigns were pros- perous, the Complaint is usually taken to be pure literature, an example of the pessimistic writing that became fashionable after the First Dark Age. Even if the Complaint be taken as inspired by a troubled condition, the nature of the trouble la- mented is completely obscure. I suggest it could make sense, however, as a reflection of the disorder, social and natural, consequent on over-long, over-

large floods, or abnormally large fluctuations in flood level from year to year, particularly if we bear in mind the Egyptians' seeming taboo on literature (openly) critically of the Nile. From Erman's translation:

... I am meditating upon what hath happened, on the things that have come to pass throughout the land. Changes take place, it is not like last year, and one year is more burdensome than the other. The land is in confusion, become waste (?) . . . - - Right is cast out, and iniquity (sitteth) in the councilchamber .... The land is in misery, mourning is in every place, towns and villages lament. All people alike are transgressors, the back is turned upon respect. .. [Perhaps looting after a destructive flood?] . . . Come, my heart, ... that thou mayest expound to me the things that are throughout the land, that are bright and lie outstretched [E: in view of all, yet compre- hended of nonef. I meditate upon what hath happened. Affliction is come today - - - and all men are silent concerning it. The whole land is in a great [E: bad] condition. There is none free from transgression, and all men alike are doing it. Hearts are sorrowful - - -... Men rise up early every day to suffering . . .

This quotation is not presented as evidence for anything-it is much too vague-but rather with

the suggestion that it might be better understood against the background of excessive floods, attested themselves by the Semna inscriptions. In particular, it would be of interest to investigate whether the translation could be clarified if it were assumed to refer to disastrous floods.46

(3) Turning now to item three: A unique fea- ture of Dynasty XIII is the frequent appearance of the name-element "Sobek," the crocodile god. At least five kings bore the name Sobekhotep "Sobek is satisfied," as their nomen or personal name, thus presumably from birth. The last known sovereign of Dynasty XII, a Queen Sobeknefru, and two or more royalties of Dynasty XVII a century later, are the only other instances where Sobek was so honored in royal names.

The crocodile-god, Sobek, Sobk, or Suchos, was primarily a divinity of the water and of vegetation, resident in the Fayum, and in other cult places; he became particularly important in Dynasty XII and XIII, being mentioned frequently as a patron of Dynasty XII kings (Hayes 1961). "The crocodile- god Sobek was naturally considered a water-god; he received worship as a patron divinity in towns [and times] the special weal and woe of which were particularly dependent on water, as was the situa- tion on the islands of Gebelein and Kom Ombo, in the oasis of the Fayum" (Steindorff and Seele

1957). We have already noted the Quta stela referring

to the "Lake of Sobk." This god was honored also by at least two surviving monuments of Amenem- het III in the Fayum-the temple to Sobek at Ki- man Faris (RL +24.0 m.) and the temple of Medi- net el Madi (pavement at +25-94 m.: Shafei 1960) in the southwest Fayum to the consort of

46 In a new analysis of this text, Kadish (i973) JEA 59:77-90) gives brief consideration to this possibility and concludes that he does not "see any terms in the text which lead to a clima- tological explanation independent of external information, but the possibility that that was the case is very attractive, especially since it lends some weight to the dating of Kha-kheper-Re- senebu." At the same time Kadish "detected a sense of im- mediacy in Kha-kheper-Re-senebu's distraught state" which argues against taking the work as pure literature. "He is ter- ribly unhappy over the social and political changes that have taken place. They are basic alterations which he regards as morally unacceptable."

Although we are accustomed to focus on the general im- pression of peace and prosperity during the later reigns of Dynasty XII, there is nevertheless, I suggest, in what we be- lieve of the political situation under Senwosret III another neglected possibility for historical significance in this text. As we have noted, Egyptologists generally believe that Senwos- ret III broke the power of the regional nomarchs because most

series of major provincial tombs come to an end in his reign. However he did this, there must have been those among the nomarchs and their families and followers who indeed felt the land had come on evil times. Kha-kheper-Re-senebu may have been one of those. It is certainly unlikely that such a transi- tion in the balance of power between nomarchs and pharaoh could have occurred without severe distress to the losing pow- ers. In this connection it is of interest to note, from the In- struction of Sehetep-ibre (a high official under Senwosret III and Amenemhet III): ". . . Adore the king, Nymaatre [Ame- nemhet III] . . . The one whom the king loves shall be a well-provided spirit; there is no tomb for anyone who rebels against his Majesty, and his corpse shall be cast to the waters" (Simpson 1972:198f). The statement, ". . . there is no tomb for anyone who rebels . . ." in conjunction with the ending of various series of provincial tombs is suggestive; we may well wonder if the nomarchs may have provoked their suppression by a rebellion against Senwosret III.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 39: Climate and the History of Egypt -The Middle Kingdom

260 BARBARA BELL [AJA 79

Sobek (Hayes I96I). The interest of Amenem- het III in the Fayum and the popularity of its god, stimulated by more-than-normal anxiety about the flood levels, may well be reflected in the popularity of the name-element Sobek in the next few genera- tions. The first of the Sobekhoteps was born proba- bly in the reign of Amenemhet III, and the name may well reflect a desire to propitiate this water god. Although the genealogy of Dynasty XIII is still unknown, it seems most likely that its kings were descendants of Amenemhet III and/or his predecessors by minor wives or concubines.

(4) Finally, it seems probable that the cessation of the ultra-high floods played no small part in bringing on the decline known as the Second Inter- mediate Period, since the last recorded high flood comes fairly early in Dynasty XIII. Although the ultra-high floods were probably at first unwelcome and a cause of disaster, once the Egyptians had adjusted to them as a frequent occurrence, their cessation might well bring on a period of poverty until a new adjustment could be made. We shall explore this period of decline more fully, in the next section.

DECLINE OF THE MIDDLE KINGDOM

The conspicuously prosperous period of the Mid- dle Kingdom came to an end about 1797 B.C., with the death of Amenemhet III after a reign of some 45 years. Dynasty XII itself ended about I78047 after the much briefer reigns of his two immediate successors, King Amenemhet IV and Queen Sobekneferu. And a couple of decades later, the records at Semna of great floods come to an end. In recent years Egyptologists have come to realize, however, that the Middle Kingdom did not end in a sudden collapse like the Old Kingdom; it rather slipped by stages from prosperity under a strong government into poverty and disorder, that lasted, in variable degree, to about 1570 B.C.

The next two centuries of Egyptian history, tra- ditionally called the Second Intermediate Period, present a terminological difficulty, since I had re- served the term "Dark Age" for eras of more wide- spread impoverishment, extending throughout the Eastern Mediterranean and Near East. The Second Dark Age of Ancient History, in this full sense, occurred between ca. I200 and ca. 900 B.C. The two centuries here under discussion were clearly in

some respects a Dark Age in Egypt, and perhaps fuller study will reveal at least partially contem- poraneous Dark Ages in other lands. But there is nothing immediately obvious. Hammurabi of Baby- lon forged a substantial empire in (probably) the upper i700s, although his dates are still debated and one does not get an impression of splendid pros- perity from Gadd's chapter in the revised CAH. Further north, this period was a time of prosperity for the Syrian coastal cities, and for Mari on the Upper Euphrates. The period also saw the begin- ning of the Old Kingdom of the Hittites in Ana- tolia. In the I6oos, Crete's palace civilization was flourishing, as were the Hittites, and mainland Greece was belatedly reviving from the First Dark Age; the city states of Palestine were prosperous. But this century is obscure and confused in Meso- potamia. Thus I shall provisionally adopt the term "Little Dark Age" for this era in Egypt, the "Lit- tle" referring to its known geographical extent alone and implying nothing about its severity- which however was almost surely less than that of the First Dark Age.

Although this period of Egyptian history is per- haps the most obscure of any (Smith 1960), a gen- eral impoverishment after the death of Amenem- het III is clearly indicated by the absence of large public buildings and by the small size of the few known royal tombs, as well as by the scarcity of decorated tombs in the provincial centers.

Neither Amenemhet IV nor Sobekneferu left a pyramid that has been identified from written evidence, although the remains of two ruined Pyra- mids at Mazghuna, about 5 km. south of Dahshur, have been attributed to them (Edwards 1961). In spite of, or because of, the some 50-60 kings listed by the Turin Papyrus and other sources for Dynas- ty XIII, the remains of only three additional Pyra- mids of this period are known. There is one at Dahshur belonging to a king with the surprising name of Ameny-the-Asiatic; and two at Sakkara, one of unknown ownership and the other belong- ing to a king with the unEgyptian name of Khend- jer. While these Pyramids are not large, they do possess impressive features, each demonstrating an elaboration of devices first used by Amenemhet III and intended to make them impregnable to thieves, a problem of evident concern to the architects of the Pyramids of the late Middle Kingdom. Ame-

47 Amenemhet IV is now given 13 years, in accord with a year 13 discovered by Hintze at Semna (n. I ). And Sekhem-

kare will be given eight years.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 40: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 261

nemhet III and Khendjer each had a burial chamber carved from a single block of quartzite; the unfin- ished and unidentified Pyramid at Sakkara had such a monolithic chamber estimated to weigh 150 tons (Edwards 1961), no small achievement for a supposed "age of poverty."

Decorated and inscribed provincial tombs, one of the main sources of information about the First Dark Age, are also very rare at this time, suggest- ing that no revival had occurred of the old nom- archic system that had been suppressed by Sen- wosret III (James 1965). Thus the Little Dark Age set in under a distinctly different distribution of political power than existed at the onset of the First Dark Age. The central government around 18oo B.C. was strong, the provincial nobility weak or non-existent.

When we turn to Vandier (1936) for evidence of famine in the years 1800-1570 B.C., we find an- other marked contrast with the First Dark Age. The earlier period gives a much stronger impres- sion of disastrous famines. After describing Dynas- ty XIII as an era full of internal struggles and usurpations, with an artistic as well as a political decline, Vandier (1936:18) concludes: "It is almost certain that the Egyptians of this epoch must have suffered frequently from hunger. Texts are, un- fortunately, rare, and I was able to find only three that allude to a scarcity."

Two of these are tomb inscriptions from El Kab, a town across the Nile from the ancient Nekken or Hierakonpolis in Upper Egypt. In the tomb of one Sobek-nakht of El Kab was found mention of a King Sobekhotep. This king, if correctly iden- tified by Hayes (1962a:9) as the third of this name, belongs at the end of a particularly obscure inter- val between the first two Kings of Dynasty XIII and the relatively well-documented King Nefer- hotep I who a number of scholars agree in dating to ca. 1740-1730 B.C. (e.g. Hayes 1962a, Beckerath 1965). Thus it would appear that the tomb of Sobek-nakht was decorated in the three-year reign of Sobekhotep III, just prior to ca. 1740. The rele- vant inscription, translated from the French of Vandier (1936:114) reads: "... I was a man who protected the afflicted against the powerful . . . who supplied the granaries of the god ... who sum- moned his entire energy every time he saw an insufficient flood (a small water) ...."

From the tomb of Bebi of El Kab, considered by Vandier to date from the same epoch: ". . . I

was a man who amassed grain ... who was vigi- lant during the winter. Whenever a famine came, during various/ numerous years, I gave grain to my city during each famine . . ." (Vandier 1936: 115). And finally, one Horherkhoutef of Edfu, after commemorating his conventional charities to the needy, ". .. I gave bread to him who had hun- ger ..." noted in his tomb (Vandier 1936:115): "I gave grain to the entire country, I saved my town from famine . . . no one has done what I did... ." This tomb is undated, but Vandier con- siders it to be in the style of Dynasty XIII.

Thus we do have evidence for famine and in- sufficient flood during the Little Dark Age, for some uncertain number of years preceding ca. 1740 B.C. But the texts suggest a condition that is much less severe than the tzw-famines of the First Dark Age. It is tempting to associate this famine with the Biblical story of Joseph, particularly as it is the only documented famine in the Second Intermedi- ate Period. Exploration of this possibility would, however, take us far afield.

In Lower and Upper Nubia also conditions were much less severe now than in the First Dark Age. Indeed this later period appears even to have been a time of some prosperity and population growth in Nubia (Trigger 1965:Io04).

Now as to the role of the Nile, we have seen that the Semna inscriptions attest that exceptionally high floods occurred in many years of the reign of Amenemhet III and his first four successors. However Lepsius (1853:525-26) noted numerous remains of temples from the New Kingdom era, more or less at the edge of the water at modern high Nile between Semna and Aswan, and schol- ars generally agree (e.g. Reisner 1929b; Trigger 1965) that the floods in the New Kingdom were rather similar to those of modern times. Thus it seems clear that at some time, either suddenly or gradually, during the Second Intermediate Period, the Nile floods must have declined from the high levels recorded at Semna to levels characteristic of the New Kingdom and modern times.

I postulate that this decline in the Nile floods occurred rather suddenly, about the time the Semna inscriptions cease, and that the decline was a sig- nificant cause for the decline in prosperity, for the onset of the Little Dark Age, traditionally known as the Second Intermediate Period. However I can find no evidence to suggest that very severe famines,

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 41: Climate and the History of Egypt -The Middle Kingdom

262 BARBARA BELL [AJA 79

such as the tzw-famines of the First Dark Age, oc- curred in these years.

The ultra-high floods were probably at first un- welcome and caused much destruction of life and property early in the reign of Amenemhet III. But, as they became a frequent occurrence, the Egyp- tians apparently adjusted to them and even profited by them, as suggested by evidence that the reign of Amenemhet III was the most prosperous period of Dynasty XII. The description of the flood by Herodotus indicates that floods were perhaps simi- larly high at the time of his visit to Egypt about 450 B.C., and no word of his suggests that the flood he saw was regarded as extraordinary by the Egyp- tians of the time: "When the Nile overflows, the whole country is converted into a sea, and the towns, which alone remain above water, look like the islands in the Aegean. At these times water transport is used all over the country, instead of merely along the course of the river, and anyone going from Naucratis to Memphis would pass right by the pyramids instead of following the usual course by Cercasorus and the tip of the Delta."

Thus once the Egyptians, led by the strong and able King Amenemhet III, had adjusted to ultra- high floods as the normal thing, and perhaps in- creased in population as higher land could be culti- vated, it seems not unlikely that a cessation of these ultra-high floods should bring on a period of pov- erty until a new adjustment could be made both in population and in reorganization of the irriga- tion works to best utilize the new levels. I further suggest that this decline in the average flood level combined with political factors to undermine the stability of the government, while at the same time the apparent weakness of the crown slowed adjust- ment to the new conditions, setting up a vicious circle and making the period a darker one than it need have been from climate changes alone. We shall return to this point presently, but first we shall survey the history of the period in an effort to localize as well as possible any periods of drought. For this I follow the chronology of the revised Cam- bridge Ancient History (Hayes 1962a) except for the adjustment required by Hintze's discoveries.

It was at one time believed that Egyptian civili- zation suffered an abrupt collapse at the end of Dynasty XII, but this concept is no longer tenable. It now appears that for over ioo years, in spite of short reigns and frequent changes of kings, "the

power of a single central government continued to be respected throughout most of Egypt itself; royal building activities were carried on in both the south and the north, and, until late in the eigh- teenth century B.C., Egyptian prestige in Nubia and western Asia remained largely unshaken. ... Until about 1674 B.C. . . the seat of the govern- ment remained . . . in the region of . . . Itj-towy ..." (Hayes 1962a). Moreover, "creditable imita- tions of the buildings, statues, and reliefs" of Dynas- ty XII continued to be produced by court artists, although a marked deterioration occurred in the quality of work produced for lesser patrons, and this work "at best was perfunctory and often downright bad" (Hayes i953). However material culture never did fall to the very low level of the First Dark Age (Smith 1965).

King Makherure Amenemhet IV, son and heir of Amenemhet III, appears to have reigned at least 13 years (n. ii). He was probably at least middle- aged when he came to the throne, as his father reigned some 45 years. In spite of a lack of evidence for any brilliant achievements, the reign shows little evidence of any serious decline in Egyptian pros- perity and prestige (Hayes 1962a) except for the lack of an identified royal pyramid. High Niles were recorded in his years 5, 6, 7, and 13 at Semna; the Sinai mines were worked-probably for the last time until the New Kingdom-and Syria ap- parently remained respectful. Amenemhet IV was succeeded by the "Female Horus," Queen Sobek- kare Sobekneferu, probably his sister and daughter of Amenemhet III. Her reign yields one Nile record at Semna, in year 3.

Although the genealogy of Dynasty XIII is still unknown, its kings had mainly Theban names- Amenemhet, Senwosret, Mentuhotep, Neferhotep, Inyotef, and Sobekhotep-with a few outlandish exceptions such as Khendjer and Ameny-the-Asi- atic-and the kings regarded themselves as legiti- mate successors to Dynasty XII, striving to up- hold the same traditions and system of government, and maintaining their capital at Itj-towy, through most of the i7oos. It thus seems probable that at least the first of them, and perhaps all of them, were descendants of the Kings of Dynasty XII, by secondary queens, probably including some for- eign women. It further seems not unlikely that, with the direct line extinct, the priority of claim to the throne among numerous secondary princes would not have been clearly defined, and that this

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 42: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 263

uncertainty played a major role in the great num- ber of Kings apparently reigning within a time-

span too short normally to accommodate them. Dynasty XIII begins48 with King Sekhemre-

Khutowy Sobekhotep I. His ancestry is un- known, but because of the change of Dynasty in the system of Manetho, Egyptologists tend to as- sume that he was not a direct descendant of Ame- nemhet IV or Sobekneferu. But the reasons for a change of Dynasty in the system of Manetho are more often than not obscure to modern Egyptolo- gists; and in the absence of evidence to the con- trary we may assume that Sobekhotep I was the legitimate heir to the throne, probably a son of Amenemhet III by a secondary queen. His reign of 5+ years (Hayes 1962a) yields three inscrip- tions on the flood at Semna, with only that year 4 still in its original position. As we have noted, one high Nile was recorded at Askut in year 3 of his successor, Sekhemkare, and two at Semna in years 4 and 8. No pyramid has yet been identified for either of these kings.

After the first two reigns (about 12 years) of Dynasty XIII, the obscuring mists thicken further. The Turin Papyrus seems to refer to 6 kingless years (Gardiner i96i:i15), then gives some 18 more names, most of them attested at least by scarabs or other minor relics, to be fitted into the period before ca. 1740 B.C. Somewhere in this murky period belongs the King Sedjefakare whose year i provides the last of the high Nile inscrip- tions at Semna; and for this reason one would like to place him soon after Sekhemkare. Also to this period apparently belong the Kings Ameny-the- Asiatic and Khendjer (4+ years). Although proof is lacking, it is both tempting and plausible to assume that the famine mentioned in the El Kab tombs, and floods below the modern, occurred in the "6 kingless years" and that the years were therefore considered "kingless" since no True Horus sat upon the throne, because by definition a True Horus restores Maat and brings good floods (see Bell 1971:19-21).

Shortly before 1740 B.C. a few openings appear in the fog to reveal the three best documented rulers of the Dynasty: Sekhemre Sobekhotep III (3 yrs., 2 mo.), the king mentioned in the tomb of Sobek-nakht at El Kab and containing a refer-

ence to famine and insufficient flood; Khasekhemre Neferhotep I (ca. 1740-1730) and Khaneferre So- bekhotep IV. The latter two were brothers, but not sons of Sobekhotep III nor of any other king. Documents from the reign of Sobekhotep III pro- vide valuable "information on the elaborate ad- ministrative organization of Egypt during the late Middle Kingdom" (Hayes 1962a). An inscription on a rock at the First Cataract appears to attest a visit there by King Neferhotep I, while a steatite plaque with his name was found at Buhen (Emery 1965:167) suggesting that he retained control of Nubia; a relief found at Byblos depicts the local prince doing homage to Neferhotep (Gardiner 1961:154). A giant statue of his brother, Sobek-

hotep IV, was found at Kerma, believed carried there later by the Kushites from one of the Second Cataract Forts.

At or soon after the reign of Sobekhotep IV, it appears that the fortunes of Dynasty XIII declined again, about I720 B.C., when the Hyksos occupied Avaris and much of the eastern Delta (Hayes 1962a). The origin of the Hyksos is still obscure and highly controversial and, although it is not irrelevant, I do not propose to speculate upon it here, because this would require a review of con- ditions in Syria and Palestine that would take us far beyond the bounds of our present investigation of Egypt and its records. No relevant Egyptian records have been found. However, it further ap- pears that about 1674 B.C. (Hayes 1962a) the Hyksos were able to occupy Memphis and Itj-towy, founding the XVth Dynasty and putting a final end to the irregularly declining phase of the Middle Kingdom. Shortly thereafter Dynasty XIII appar- ently expired quietly in the Theban area, where around 1650 B.C. Dynasty XVII "arose to keep alive the embers of Egyptian independence ." (Hayes 1962a:13).

Also about this time, that is between 1720 and 1674 B.C., the Middle Kingdom forts in the region of the Second Cataract were burned and appar- ently abandoned by the Egyptians. Trigger (1965) considers it still uncertain whether the Egyptians left voluntarily or were driven out, and whether the region of Lower Nubia thereafter was independent or controlled by the Kingdom of Kush to the south. That they left, however, is suggested by the lack

48 Because of the paucity of available evidence on this pe- riod, scholars are not in complete agreement on the order, number, and names of the Kings in Dynasties XIII-XVII. Most

relevant here, the three Kings with flood inscriptions at Semna are Nos. 3, 2, and 15 in the Dynasty XIII list of Beckerath (1965:222).

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 43: Climate and the History of Egypt -The Middle Kingdom

264 BARBARA BELL [AJA 79

of evidence for any repair of the fire damage (A.W. Lawrence 1965 JEA 51:72). Emery (1965), Aldred (1963), and others consider that the forts were stormed and destroyed, by an enemy of necessarily considerable military skill and sophistication, al- though Adams (1972) does not believe that the evi- dence compels such an interpretation.

The Kingdom of Kush, centered at Kerma, south of the Third Cataract, appears to have attained its greatest prosperity during this period (Trigger 1965); and the large number of Hyksos seals found in Kerma suggests that it carried on a sub- stantial trade with the Delta. The prosperity in Nubia further indicates that there cannot have been

anything too severely wrong with the floods in the 16oos and that, for whatever reasons, the Kingdom of Kush adjusted more quickly than that of Egypt to the cessation of ultra-high floods.

In spite of the flourishing condition of Nubia, the middle 16oos remained a time of poverty in Upper Egypt, with no evidence of quarrying at Hammamat or Aswan; very few tombs were built around Thebes and even the royal burials were very meagre (Winlock 1947), embellished as they were with only modest mudbrick pyramids (Hayes 1962a). Also around this time a change occurred in burial customs, large rectangular wooden sar-

cophagi, the best of which were made from im- ported Lebanese cedar, giving way to smaller an- thropoid coffins of Egyptian sycamore wood, a change that has been attributed to the poverty of the times and the lack of suitable wood for the large sarcophagi (Winlock 1947).

Recovery proceeded slowly and irregularly. The order and dating of the kings at Thebes in Dynasty XVII is still controversial. Following the chronol- ogy of Hayes (1962a), however, there appears to have been some revival in the 1620os and 16ios when a King Sekhemre Shedtowy Sobekemsaf reigned for 16 years; according to records of the Tomb Robberies that occurred toward the end of Dynasty XX, he and his queen had mummies well bedecked with gold, although Winlock (1947) sug- gests that these records may be a clerical exaggera- tion. Then come several very short reigns, after which we find King Nubkheperre Inyotef VIII, who was an active builder; he embellished his tomb entrance with a pair of small obelisks, and had in- scribed on its wall the famous "Song of the Harp- er," the theme of which is "eat, drink, and be mer- ry, for tomorrow we die" (Hayes 1962a)--one

item of the evidence that learning flourished in Thebes in spite of the general material poverty of the early years of Dynasty XVII.

Conditions in Lower Egypt under the Hyksos are even more obscure, as no records or monuments have survived from the period of their rule. How- ever at least two Hyksos, Seuserenre Khyan, and Auserre Apophis, clearly enjoyed long reigns, 40 years for Apophis, and this in Egypt is generally a sign of prosperity. Scarabs, seals, and other minor relics of these kings have been found widely dis- tributed through the Near East, indicating that they carried on active trade with Crete, Mesopo- tamia, Kerma, and many other places. As foreign princes, moreover, they and their court were proba- bly less influenced, or even completely uninflu- enced by the dogma that the king was responsible for the adequacy of the Nile floods.

Vandier (1936) notes that one might expect many severe famines in the disorder of the Hyksos period; not only could he find no evidence for this, but the stele of Kamose indicates that at least to- ward the end of the reign of Apophis the fields were well cultivated in the north, and cattle were kept in the Delta by the Thebans (Kees 1961). For the Theban courtiers, presenting arguments against war with the Hyksos, remind King Kamose: "The finest of their fields are ploughed for us, our cattle are pastured in the Delta. Emmer is sent for our pigs, our cattle are not taken away . . ." (Vandier 1936; Kees 1961; Erman 1927). This occurred prob- ably around 1580 B.C., and may serve to point up the dependence of Thebes on the rest of the coun- try for any high level of prosperity. The Thebaid is a relatively poor region of Egypt, and the floods need not be particularly low to keep it in a con- dition of semi-poverty when it cannot draw on the resources of a united Egypt.

Thus we need not, and probably should not, in- terpret the fact of Theban poverty in the I6oos as evidence that the Nile floods were significantly be- low those of the New Kingdom. However, intro- duction of the shaduf toward the end of the Hyksos period made watering of selected lands practical outside the flood season (Winlock i947),

and may have contributed to recovery from the Little Dark Age.

Although King Wadjkheperre Kamose disre- garded the advice of his courtiers and began a war to expand the power of his house and reunite Egypt under his rule, he died while still a young man. It

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 44: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 265

remained for his brother and successor Ahmose to capture the Hyksos capital of Avaris and expel the Asiatics from the soil of Egypt, around I574 B.C. (Hayes 1962a). A few years later Nebpehtyre Ahmose, Pharaoh of a reunited Egypt, founder of Dynasty XVIII (proving that a change of Dynasty does not necessarily correspond to a change in the ruling family) and of the New Kingdom, wiped out the vestiges of Hyksos power in southern Pales- tine (Hayes 1962a), whereupon the Hyksos dis- appear from history. Before the end of his reign King Ahmose had also reconquered Nubia as far as Buhen, where he ordered a rebuilding of the fort (Emery 1965).

DISCUSSION

I have advanced the hypothesis that the Little Dark Age in Egypt at the end of the Middle King- dom was brought on-or at least intensified-by a cessation of the ultra-high floods that occurred often (but not every year) in the reign of Amenem- het III and his immediate successors, and a fall to levels around the modern, which various evi- dence indicates prevailed during the New King- dom as well as during the earlier reigns of the Mid- dle Kingdom. I infer that this decline occurred sometime after 1768 B.C., rather abruptly, and that the interval between 1778 and 1745 saw several years in which the floods fell below the New Kingdom levels, as evidenced by references to famine and in- sufficient floods in the tombs of Sobek-nakht and Bebi of El Kab. No evidence was found on climate conditions in Egypt during the Hyksos period, but the prosperous condition of Nubia suggests that the floods were adequate, perhaps close to New King- dom levels. I further suggested that the abrupt de- cline in flood level after ca. 1768 occurred in a time of unusual uncertainty about the proper order of succession to the Kingship and set up a sort of vicious circle, thus increasing the instability of the government and slowing efficient adjustment to the new flood levels. Uncertainty about the order of succession is inferred from the change in Dynasty and from the numerous short reigns in the Little Dark Age.

The reader may recall that a similarly excessive number of short reigns characterized the First Dark Age. In my previous discussion of this phe- nomenon in Egyptian history (Bell I97I:I9-2I), I pointed out, with supporting quotations, that a vital symbolic function of the Pharaoh was as

"rainmaker" or, more precisely, floodmaker. I sug- gested that this dogma, combined with frequent severe failure of the floods, could well explain the abnormally large number of reigns within a few decades of the First Dark Age. As successive kings failed to bring good floods, they were removed or removed themselves from the Kingship. I here postulate a similar explanation for the numerous short reigns of the Little Dark Age or Second In- termediate Period.

The numerous very short reigns to be fitted into the I700oos-some 8 kings in the years between ca. 1768 and 1740 B.C.-and an even more obscure, but certainly vastly excessive, number for the 16oos, led Hayes to suggest that the kingship lost its hereditary character during Dynasty XIII, and that kings were elected for a limited period of time (see Hayes 1962a:5 for refs.). This hypothesis is highly controversial, and to a large extent I agree with those who resist the idea of "elected" kings as entirely contrary to the Egyptian concept of divine Kingship. In possible support of the hypothesis, to be sure, are several bits of information establishing a lack of continuity in the succession, making it clear that certain kings of Dynasty XIII were not sons of any king. But the significance of this fact is obscured by a virtually complete lack of informa- tion on the titles and ranks held by kings' sons under Dynasty XII-XIII. Apparently they possessed no distinctive titles, since it is scarcely conceivable that each king of Dynasty XII had one and only one son (although no more are known), while there is evidence of numerous royal daughters. There is mention, uniquely to my knowledge, of royal princes with the army led by the co-regent Sen- wosret I, in the story of Sinuhe.

But perhaps the seeming contradiction can be resolved if we assume that election to the kingship was not open to just anyone, but only to descen- dants of Kings; and further that election was re- sorted to only when no heir existed with a clearly defined superior claim or talent for leadership. I have suggested previously (Bell i97i) something of this sort for the First Dark Age, where I pointed out that such a situation of uncertainty could easily arise among the descendants of a king who had several secondary wives, in the event of the extinc- tion of the line descended from the Great Royal Wife or primary Queen. There is no evidence that lesser wives were ranked, so it remains entirely obscure how the claims of the sons of such wives

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 45: Climate and the History of Egypt -The Middle Kingdom

266 BARBARA BELL [AJA 79

would be ranked for priority, especially if the King who had fathered them died before such ranking seemed significant. After long reigns, such as the 45 years of Amenemhet III (and the 94 years of Pepi II), there could be grandsons (and great-grandsons) as well as sons to assert their various claims, not to mention claims by marriage to a higher-ranking princess. Thus we can con- ceive how election, with the list of eligible electees, and perhaps also of electors, strictly limited by a hereditary claim, could in certain emergencies be reconciled with the Egyptian dogma of divine King- ship. Under these circumstances also, as I pointed out in discussing the First Dark Age, it could be

relatively easy--especially if the flood should be meagre in the first or second year of the new king -for whomever it actively concerned to decide that a mistake had been made, that the recently selected king was not the True Horus.

Fitting in plausibly with this picture is a further notable feature of Dynasty XIII, that is, a greater stability in viziers than in kings, and probably also a greater hereditary continuity in viziers than in kings. The best documented is one Ankhu, who

apparently served as vizier under several kings, in- cluding Khendjer, Sobekemsaf I, and Sobekho- tep III, and perhaps others yet more ephemeral between these (Gardiner i96i; Hayes 1962a). Be- lieved to be of the same family is the Vizier Iyme- ru, under Sobekhotep IV (Hayes 1962a). Although this greater stability in viziers than in kings has not yet been explained, it would appear that Ankhu was either a man of exceptional political agility (as Talleyrand served both Napoleon and the Bour- bons), or that the rapid turnover in kings was not caused primarily by bitter enmities, by feuds of hostile warring factions, by usurpations and force- able depositions, etc. (Similarly, though on a lower level, in the First Dark Age we saw the case of the priest Merer who "offered for thirteen rulers.") This situation of a single vizier serving several kings is compatible, at least, with the hypothesis that, in a time troubled both by poor floods and unusual uncertainty about the identity of the true Horus King, a king would be secretly murdered or expected to commit suicide when his coronation was not followed in due season with floods ade- quate to avert scarcity. Once well established, with several good floods to his credit, a king might more readily survive a season or two of scarcity.

Since the genealogy of both the kings and the

family of viziers is unknown, it is tempting to speculate that the viziers too were among those eligible, by reason of descent from a King of Dynas- ty XII, for election to the kingship, but that they astutely secured the election of other princes as king. Knowing the fate of kings with controver- sial claims when the Nile was bad, recognizing that it was erratic at this time, they may have preferred to rule as vizier without the responsibility for an erratic Nile, while weaker and less astute princes enjoyed the glory and the hazards of divine King- ship.

It is interesting and suggestive to compare the portraits of certain rulers of Dynasty XIII with those of the last great rulers of Dynasty XII. As noted in a previous section, many of the portraits of Senwosret III and Amenemhet III suggest that these Pharaohs "ruled . . . by virtue of their own personalities.... It was the dominating will of the reigning king that guaranteed the integrity of the state, rather than the institution of monarchy" (Westendorf 1968:98). By contrast, the kings of the succeeding Dynasty XIII "obviously no longer possess the qualities of their predecessors and mod- els" (ibid.). Indeed, to this non-specialist, the statues of Neferhotep I (Aldred 1962:fig. 83) and especially Sobekhotep IV (Wilson 1956:fig. 17a)-among the best surviving works from Dynasty XIII--each ex- press a melancholy resignation that is entirely com- patible with the hypothesis that his reign and life would be terminated if the floods proved too meagre. Neither looks as though he would resist and fight determinedly for the throne. It cannot, of course, be inferred that the brother kings them- selves came to this unhappy end; at most their ex- pressions may reflect only their expectations and their knowledge of the fate of various predecessors.

HARVARD COLLEGE OBSERVATORY

References

Adams, William Y. 1971 Personal communication. 1972 Personal communication.

Aldred, Cyril 1962 The Development of Ancient Egyptian

Art, Tiranti, London. 1970 Some royal portraits of the Middle King-

dom of ancient Egypt, Met. Museum J. 3:27-50.

Arkell, A.J. I96I A History of the Sudan to 1821, London.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 46: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 267

Ball, John 1903 The Semna cataract of the Nile, Quart.J.-

Geol.Soc.London 59:65-79. 1939 Contributions to the Geography of Egypt,

Govt. Press, Cairo. Beckerath, Jilrgen von

1965 Untersuchungen zur politischen Geschichte der zweiten Zwischenzeit in Agypten, Agyptologische Forschungen, vol. 23, J.J. Augustin, Gliickstadt.

Bell, Barbara 1970 The oldest records of the Nile floods,

Geogr.J. 136:569-73. 1971 The Dark Ages in Ancient History: I.

The First Dark Age in Egypt, AJA 75:1- 26.

Breasted, James Henry 19o6 Ancient Records of Egypt, vol. I, Uni-

versity of Chicago Press. Brooks, C.E.P.

1949 Climate through the Ages, McGraw-Hill, New York.

Butzer, Karl W. 1958 Quaternary Stratigraphy and Climate in

the Near East, Bonner Geogr.Abh., Heft 24, Bonn.

I959a Some recent geological deposits of the Egyptian Nile valley, Geogr.J. 125:75-79.

I959b Environment and human ecology in Egypt during predynastic and early dynastic times, Bull.Soc.Geogr.d'Egypte 32:43-87 (a condensed English transl. of 1959c, q.v. for documentation).

1959c Die Naturlandschaft Agyptens wihrend der Vorgeschichte und der Dynastischen Zeit, Abh.Ak.Wiss.Lit. (Mainz) Math.- naturw.Kl. No. I, 80 pp., Wiesbaden.

1965 Physical conditions in Eastern Europe, Western Asia, and Egypt, CAH I, ch. 2 (fasc. 33).

1971 Personal communication. Butzer, K.W., G.L. Isaac, J.L. Richardson, and C.

Washbourn-Kamau

I972 Radiocarbon dating of East African lake levels, Science 175:io69-76.

Caton-Thompson, G. and E.W. Gardner I929 Recent work on the problem of Lake

Moeris, Geogr.J. 93:20-60. Dunham, Dows

1967 Second Cataract Forts: Uronarti, Shalfak,

Mirgissa (excavated by Reisner and N.F. Wheeler), Museum of Fine Arts, Boston.

Dunham, Dows, and J.M.A. Janssen i960 Second Cataract Forts: Semna, Kumma

(excavated by G.A. Reisner), Museum of Fine Arts, Boston.

Edwards, I.E.S.

I96I The Pyramids of Egypt, Penguin. Emery, Walter B.

1965 Egypt in Nubia, Hutchinson, London. 1967 Lost Land Emerging, Scribners, New

York. Erman, Adolf

1927 The Ancient Egyptians: a sourcebook of their writings, Harper Torchbooks, transl. from German by A.M. Blackman.

Fairbridge, Rhodes W. 1963 Nile sedimentation above Wadi Halfa

during the last 20,000 years, Kush i :96- 107.

Fakhry, Ahmed 1969 The Pyramids, University of Chicago

Press. Gardiner, Sir Alan

1961 Egypt of the Pharaohs, Clarendon Press, Oxford.

Gautier, J.E., and G. Jequier 1902 Fouilles de Licht, Cairo.

Habachi, Labib 1974 A high inundation in the temple of

Amenre at Karnak in the thirteenth Dy- nasty, Studien zur altdgyptischen Kultur 1:207-14.

Hayes, William C. 1953 The Scepter of Egypt, vol. I, Harper,

New York.

I96I The Middle Kingdom of Egypt, CAH I, ed. 3, ch. 20 (fasc. 3).

1962a Egypt from the death of Ammenemes III to Seqenenre II, CAH II, ed. 3, ch. 2 (fasc. 6).

Herodotus The Histories, Penguin Classics ed., transl. A. de Selincourt, 1954.

Hurst, H.E. 1952 The Nile, Constable, London.

Hurst, H.E., R.P. Black, and Y.M. Simaika 1946 The Nile Basin, vol. VII: The future con-

servation of the Nile, Physical Dept., Min- istry of Public Works, Cairo.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 47: Climate and the History of Egypt -The Middle Kingdom

268 BARBARA BELL [AJA 79

Hurst, H.E., and P. Phillips 1938 The Nile Basin, vol. V: The hydrology of

the Lake Plateau and Bahr el Jebel, Physi- cal Dept., Ministry of Public Works, Cairo.

James, T.G.H.

1965 Egypt from the expulsion of the Hyksos to Amenophis I, CAH II, ch. 8 (fasc. 34).

Kees, Hermann

1961 Ancient Egypt, University of Chicago Press.

Knudstadt, J. 1966 Serra East and Dorginarti, Kush 14:165-

86.

Lepsius, K.R.

1853 Letters from Egypt, Ethiopia, and the Peninsula of Sinai, Henry G. Bohn, Lon- don.

Little, O.H.

1936 Recent geological work in the Faiyum and in the adjoining portion of the Nile val- ley, Bull.lnst.d'Egypte i8:201-40.

Lyons, H.G. 1905 On the Nile flood and its variations,

Geogr.J. 26:249f, 395f. 1906 The Physiography of the River Nile and

its Basin, National Printing Dept., Cairo. Michael, H.N. and E.K. Ralph.

1974 Univ. of Penn. radiocarbon dates XVI, Radiocarbon I6:198-218.

Montet, Pierre

1968 Lives of the Pharaohs, World Publ., Cleve- land.

Murray, G.W. 1935 Sons of Ishmael, Geo. Routledge & Sons,

London.

1951 The Egyptian climate: an historical out- line, Geogr.J. 117:424-34.

1967 Dare Me to the Desert, Geo. Allen & Un- win Ltd., London.

Nims, Charles F.

1965 Thebes of the Pharaohs, Stein & Day, New York.

Petrie, Sir W.M. Flinders 1889 Hawara, Biahmu and Arsinoe, London.

Popper, William 1951 The Cairo Nilometer, vol. 12 of the Uni-

versity of California Publications in Se- mitic Philology.

Reisner, George A. T929a Excavations at Semna and Uronarti by the

Harvard-Boston Expedition, Sudan Notes and Records 12:143-61.

1929b Ancient Egyptian forts at Semna and Uronarti, Bull. Museum of Fine Arts (Boston) 27:64-75.

Said, Rushdi, C. Albritton, F. Wendorf, R. Schild, and M. Kobusiewicz

1972 A preliminary report on the holocene ge- ology and archaeology of the northern Fayum desert, Playa Lake Symposium, C.C. Reeves, Jr., ed., Intern'l Center for Arid and Semi-Arid Land Studies, Publ. 4, Lubbock, Texas.

Shafei, Ali 1960 Lake Moeris and Lahun: the great Nile

control project executed by the ancient Egyptians, Bull.Soc.Geogr.Egypt 33:i87- 215-

Simpson, William Kelly, ed. 1972 The Literature of Ancient Egypt: An An-

thology of Stories, Instructions, and Poetry, Yale University Press, New Haven.

Smith, William Stevenson 1960 Ancient Egypt, Museum of Fine Arts,

Boston.

1965 The Art and Architecture of Ancient Egypt, Penguin Books, Baltimore.

Steindorff, George, and K.C. Seele 1957 When Egypt Ruled the East, 2nd ed., Uni-

versity of Chicago Press. Sutton, L.J.

1949 Rainfall in Egypt: Statistics, Storms, Run- of, Physical Dept. Paper No. 53, Govern- ment Press, Cairo.

Toussoun, Prince Omar 1925 Memoire sur l'Histoire du Nil, Cairo.

Trigger, Bruce 1965 History and Settlement in Lower Nubia,

Yale Univ. Publ. in Anthropology, No. 69. Vandersleyen, Claude

1970 Des obstacles que constituent les cataracts du Nil, BIFAO 69:253-66.

Vandier, Jacques 1936 La Famine dans l'Egypte Ancienne, Cairo.

Van Seters, John

I966 The Hyksos, Yale University Press, New Haven.

Vercoutter, Jean 1965 Excavations at Mirgissa-II, Kush 13:62-

73.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions

Page 48: Climate and the History of Egypt -The Middle Kingdom

1975] CLIMATE AND THE HISTORY OF EGYPT 269

1966 Semna South Fort and the record of the Nile levels at Kumma, Kush 14:125-64.

1967 Egypt in the Middle Kingdom, ch. io of The Near East: the Early Civilizations, eds: J. Bottero, E. Cassin, and J. Vercout- ter, Delacorte Press, New York.

1970 Mirgissa I, Paris. Wendorf, Fred, R. Said and R. Schild

1970 Egyptian Prehistory: some new concepts, Science 169:1161-71.

Westendorf, Wolfhart 1968 Painting, Sculpture, and Architecture of

Ancient Egypt, Abrams, New York. Wheeler, Noel F.

1931 Harvard-Boston expedition in the Sudan, 1930-1931, Bull. Museum of Fine Arts (Boston) 27:64-75-

1961 Diary of the excavation of Mirgissa fort

(14 Nov. 1931 to 3 Feb. 1932) (ed. by Dows

Dunham), Kush 9:87-179. Wilkinson, Sir John G.

1835 Topography of Thebes, John Murray, London.

Wilson, John A.

1955 Translations from Ancient Near Eastern Texts (ANET), ed. J.B. Pritchard, Prince- ton University Press.

1956 The Culture of Ancient Egypt, University of Chicago Press.

Winlock, H.

1947 The Rise and Fall of the Middle Kingdom in Thebes, Macmillan, New York.

This content downloaded from 190.162.192.202 on Wed, 4 Sep 2013 15:09:32 PMAll use subject to JSTOR Terms and Conditions