The

BIBLICAL ARCHAEOLOG

Published by
THE AMERICAN SCHOOLS OF ORIENTAL RESEARCH
126 Inman Street, Cambridge, Mass.

Vol. XXXIII September, 1970 No. 3

Fig. 1. Airview of Megiddo from the southwest. The water shaft is in the foreground, with
gallery 629 coming toward the viewer at the very bottom.

Contents
Megiddo of the Kings of Israel, by Yigael Yadin .......................................................66
66 THE BIBLICAL ARCHAEOLOGIST (Vol. XXXIII,

The Biblical Archaeologist is published quarterly (February, May, September, December)

by the American Schools of Oriental Research. Its purpose is to meet the need for a readable,
non-technical, yet thoroughly reliable account of archaeological discoveries as they relate to the
Bible.
Editor: Edward F. Campbell, Jr., with the assistance of Floyd V. Filson in New Testament
matters. Editorial correspondence should be sent to the editor at 800 West Belden Avenue, Chica-
go, Illinois 60614.
Editorial Board: W. F. Albright, Johns Hopkins University; G. Ernest \Wright, Harvard
University; Frank M. Cross, Jr., Harvard University; William G. Dever, Jerusalem.
Subscriptions: $3.00 per year, payable to the American Schools of Oriental Research,
126 Inman Street, Cambridge, Massachusetts 02139. Associate members of ASOR receive
the journal automatically. Ten or more subscriptions for group use, mailed and billed to the
same address, $2.00 per year apiece. Subscriptions run for the calendar year. In England: twen-
ty-four shillings (24s.) per year, payable to B. H. Blackwell, Ltd., Broad Street, Oxford.
Back numbers: $1.00 per iccue and $3.75 per volume, from the ASOR office. Please make
remittance with order.
The journal is indexed in Art Index, Index to Religious Periodical Literature, and at the
end of every fifth volume of the journal itself.
Second-class postage PAID at Cambridge, Massachusetts and additional offices.
Copyright by American Schools of Oriental Research, 1970
PRINTED IN THE UNITED STATES OF AMERICA, BY TRANSCRIPT PRINTING COMPANY
PETERBOROUGH, N. H.

Megiddo of the Kings of Israel

YIGAEL YADIN
The Hebrew University

(Professor Yadin broke the news of his restudy of the Megiddo stratigraphy
in the BA ten years ago, after the first of three short campaigns he describes
in this article. The report was so imlportantthat we squeezed it in tiny type
with only three photographsat the end of BA, XXIII.2. The article you are
about to read contains much that is new and exciting; for that reason we
have left a certain alolint of repetition of the 1960 article in place so that
the entire picture can be painted. - EFC)
The endeavors of the spade to unearth the building remains of Solo-
mon, greatest builder among Israel's kings, are part of the enthralling web
of the excavationsin the Holy Land during the last seventy years. No doubt
the crowning glory of Solomo1n's enterprises is the Temple he built in Jeru-
salem, to which, understandably,whole chapters in the Bible are dedicated.
David, who spent his life warring even beyond the borders of Israel, had
no time to build fortified cities (which his offensive strategy in fact made
unnecessary), let alone the Temple in Jerusalem."You know that David my
father could not build a house for the name of the Lord his God because of
the warfare with which his enemies surrounded him, until the Lord put
them under the soles of his feet" (I Kings 5:3).
Despite the detailed descriptionsof the Temple's plan, its measurements
and its holy vessels, scholars for the past few centuries have struggled to
reconstructits plan and form, and their opinions differ to this day, mainly
because we are unfamiliar vith some of the basic architecturalterms. In-
1970, 3) THE BIBLICALARCHAEOLOGIST 67

deed, these struggles,as W. F. Albrighthas so aptly characterizedthem,

amountto the "sufferings of Tantalus."
When, at the beginningof the century,the disciplineof archaeology
vigorouslyjoined the other disciplinesdealingwith biblicalcomprehension
and interpretation,scholarsrealizedsoon enough that it wasn'tthe length
of a biblical passagewhich conveyedits importanceto the archaeologist.
Often a brief and concisedescriptionwas an importantclue, if appliedto
the tools of archaeologyin generaland excavationin particular.Indeed,it
seems that there is no examplein the historyof archaeologywhere a pas-
sage helped so much in identifyingand dating structuresin severalof the
most importanttells in the Holy Land as has I Kings 9:15: "And this is
the accountof the forced labor which King Solomonlevied to build the
house of the Lordand his own house and the Millo and the wall of Jeru-
salemand Hazorand Megiddoand Gezer..."
The excavatorsof Megiddohad uncovereda largecity with a complex
of stablesand a magnificentcity gate of six chambersand two towers.That
city was attributedto Solomon;but at Gezer,excavatedat the turn of the
century,no clear Solomonicremnantswere discovered.Even on othersites
which the Bible does not mentionin connectionwith Solomon'sbuilding
activities,scholarstried occasionallyto relate to Solomonsome structures
found during excavations.But their inability to substantiateexact dates
turnedtheseeffortsto a bone of contention.It was thereforenaturalenough,
when we selectedHazoras a large-scalediggingsite in 1955,that we hoped
to succeedin uncoveringSolomon'scity as describedin the Bible, which
would in turn help clarifyhis buildingenterpriseselsewhere.Indeed,then,
one of the moreimportantdiscoveriesat Hazorwas Solomon'sfortifications
in the tenth stratumfrom the top. Our decisionto attributethat layer to
Solomonwas basedprimarilyon the I Kingspassage,the stratigraphy and
the pottery.But when in additionwe foundin that stratuma six-chambered,
two-towered gate connectedto a casematewall identicalin plan and measure-
mentswith the gate at Megiddo,we felt surewe had successfullyidentified
Solomon'scity.
The Mystery of Gezer
As is well-known,Gezer had been excavatedby the Britisharchaeolo-
gist Macalisterat a time when archaeologywas still in its infancy,and be-
fore sophisticated,up-to-datemethodsof diggingwere known. In addition,
Macalisterservedas his own administrator, architectand recorder,which is
why deciphering the resultsof his excavationsis one of the most difficult
problems in Palestinian archaeology.Some compensationmay be found,
perhaps, in Macalister'sdetailedthree-volumereport,which he lavishedwith
drawings. Macalister did not observe in the Gezer fortificationsany case-
68 THE BIBLICALARCHAEOLOGIST (Vol. XXXIII,

mate wall attributable to Solomon, but the discoveries in Megiddo and

Hazor induced me to re-examine his report. As BA readers know, I was
able to identify in the plan on page 104 of his Volume I a casemate wall
and even a structure similar to Solomon's gates at Mlegiddoand Hazor. Mac-
alister had marked these structures "a Maccabean Castle" and this was
accepted by scholars.Of the gate, only the western part had been discovered.
The resemblance to Hazor and Megiddo prompted me to suggest that this
whole complex was no Maccabean castle but rather a part of the Solomonic
fortificationssystem which had been built in a breach in the earlier wall.
Since, as is well known, the Hebrew Union College in Jerusalemhas
dug extensively at Gezer under William Dever's direction (see BA, XXX
[1967], 34-62), and one of the objects of that expedition was to clarify the
problem of the Maccabean castle. I am happy that the recent excavations
have completely vindicated my theory; not only has the eastern part of the
gate been discovered, but also pottery found on its original floor was from
the second half of the 10th century, that is Solomon's time. Thus, with the
aid of the brief biblical passage from Kings, the Solomonic fortifications,
identical in plan in the three cities, were located and dated.
Megiddo: Fly in the Ointment
Despite the absolute similarity between the gates of Hazor, Megiddo
and Gezer, one strange and worrying fact remained: while in Hazor and
Gezer the gate had been connected to a casemate wall, the Megiddo wall,
attributedby its excavatorsto Solomon, was of a quite different type, a solid
wall, with "insets"and "offsets,"i.e. its interior and exterior walls had pro-
trusions, 20 inches deep, with niches between them. The wall had been
so built that opposite every offset in its exterior wall there was an inset in
its interior wall and vice versa. The wall was built of flat, not too large,
stones, which, according to the excavators,had been laid as in brick build-
ing; the top of the wall's foundation was even and one could assume with
certainty that it carrieda brick wall. There was no doubt that this wall was
contemporarywith the famous stables.
Why would Solomon's engineers in Megiddo resort to a fortification
method different from the casemate? It was hard to accept the suggestion,
made following my article on Gezer, that Megiddo had been fortified in a
special way because it was a chariot city, while Hazor and Gezer had been
store cities. As I wrote in BA, XXIII (1960), it is axiomaticthat the strength
and characterof a wall is determined by the tactics, strength and siegecraft
of the enemy against whom it is erected. Gezer in the south and Hazor in
the north, could theoretically have been fortified against two different po-
tential enemies, yet both had the same type of fortifications.There was no
reason why Megiddo in the center of the land should be protected by dif-
1970, 3) THE BIBLICALARCHAEOLOGIST 69

ferent fortifications. The only possibility that seemed conceivable to me

was that perhaps the offsets/insets wall formed a part of a later set of forti-
fications and that during the Solomonic period there had also existed a case-
mate wall, one perhaps hidden beneath the solid wall; when the solid wall
was built, the former had been filled up (as in Hazor). The only way to
verify that assumption was to make a trial trench across the offsets/insets
wall in order to establish if in fact there had been a casement wall beneath
it.
Further Difficulties in the Stratigraphy of Israelite Megiddo
Before I describe the surprising results brought forth by three short
seasons of excavations carried out at Megiddo, I should describe briefly
what Megiddo looked like in the early Israelite period and the additional
difficulties, which have direct bearing on the subject and are well known
to all who have studied the results of the Megiddo excavations. Megiddo
had been excavated over a number of years and under the direction of sev-
eral excavators, starting with the destructive Schumacher dig in 1903-1905
and ending with the systematic excavations of the expeditions of the Chi-
cago Oriental Institute from 1925-1939. Despite the lattcr'smethodical exca-
vations, quite a number of stratigraphicalproblems remained without satis-
factory solutions, mainly because of the changing directorship of the ex-
pedition, which impaired their continuity: the efforts to bridge the dis-
crepancies between the various expeditions were not always successful. As
far as our specific subject is concerned, the foremostproblemcan be defined
as follows (see Fig. 2). The main buildings attributed by the excavatorsto
the Solomonic period were discovered in Stratum IV from top, which they
were compelled (for reasons to be explained below) to call "IVA." This
stratum included the two famous stable complexes; the first (1576) on the
southern side of the western half of the tell ("the southern stables") and
the second (407 - following the number of one of its halls) on the northern
side of the eastern half of the tell (the "northernstables"). There was no
stratigraphicaldoubt that the stables were contemporarywith the offsets/
insets wall, and since this wall was attributed to Solomon, it followed that
so should the stables.
Other structures, too, were ascribed to that stratum, of which the most
outstanding was an elaborate building (338 - on the east side of the tell).
Below this stratum, an Iron age stratum was discovered by the excavators,
who marked it "V." Further excavations made it clear that in fact they had
struck two strata and not just one. Since by that time they had designated
the next stratum "VI," they were compelled to designate the two strata
under the stables "VA" (the upper) and "VB." Beneath these was, as men-
tioned, stratumVI which also turned out later to consist of two strata,mark-
70 THE BIBLICAL ARCHAEOLOGIST (Vol. XXXIII,

ed respectively VIA (to be discussed in detail further on) and VIB. Under
the latter, the last Iron age stratum (VIIA) was discovered,with its famous
ivory treasure and inscriptions of Ramses III and Ramses VI, that stratum
which established the earliest possible date for stratum VIB, in the second
half of the 12th century.
The great complication in attributing stratum IV to Solomon arose still
earlier, as a result of a surprising discovery on the south side of the tell,
east of the southern stables. Here a building (1723) was discovered - a
palace or fort - measuring twenty by twenty-two meters, built of ashlar,
similar in style to the Solomonic gate. To the astonished excavators it be-
came clear that the so-called Solomonic offsets/insets wall (325) had been
erected on the ruins of that palace. Hence, to judge by its pottery, it must

?: .:-':" ')' '

/1325

"Ii~i? iI

ezajl //
o^oo | 48y
1007 157 8

/?????- ? 325

Fig. 2. Plan of Megiddo in the days of Solomon and of the early kings of Israel. Structures
rendered with dots belong to Stratum IVA, including the offsets/insets wall, the two
stable complexes 407 and 1576, and the water system 925 and 1000. Structures of the
Solomonic period, rendered in diagonal hatching, include the palace 6000, buildings 1723
and 1482, and the gallery 629. All will be discussed in what follows.

be later than the elaborate building belonging to the 10th century. More-
over, even west of the palace (1729), but contemporarywith it, the "south-
ern palace," a huge, well-conceived structure (1482), had been discovered.
Its west part was under the foundations of the eastern stables of complex
1576, so it too preceded the stables and adjacent wall!
Once it became clear that one could not ascribe both these buildings
to stratum IV, and since V had already been used to mark other ruins, the
Megiddo excavatorsascribed the palace and structure 1482 to Stratum IVB.
1970, 3) TIHE BIBLICAL ARCHAEOLOGIST 71

But here they encountered one of the most disturbing problems, which they
could not explain: since they were quite certain that the stables and the
offsets/insets wall (325) had been built by Solomon, and that also the
palace and structure 1482 were Israelite and approximatelySolomonic, they
were left with one of twroexplanations, neither of which was logically and
historically very tenable. They could claim 1) that the palace was built at
the beginning of Solomon's rule, before the city had been fortified. At that
stage the palace (or fort) had been a single structure on the tell, serving
perhaps as the governor'sresidence, with an excellent vantage point on the
whole region. Later, when Solomon's engineers were about to fortify the
city, this palace was in their way and so they demolished it (and structure
1482 with it, when they built the stables), and built the offsets/insets wall
on top of its ruins. This explanation was coupled with the assumption that
perhaps the palace had been demolished before its construction was com-
pleted. Or they could claim 2) that the single palace had been built by David
only to be destroyed by Solomon when he rebuilt Megiddo.
Both these explanations assumed that Solomon himself destroyed the
two grandest Israelite structures existing in Miegiddo,in order to build the
offsets /insets wall and the stables. The first assumption is certainly illogical
and the second is impossible also for historical reasons.The Bible says clearly
that Solomon built Megiddo, Hazor and Gezer, whilst David did not en-
gage in building enterprises and couldn't even build the Temple. He cer-
tainly did not build a city in Gezer; and at Hazor, too, there were no forti-
fications prior to Solomon, but only a poor, unwalled hamlet. Had David
built a large city in Megiddo, the Bible would surely stress the fact. Or
would it say that it was Solomon who built Megiddo? Moreover, even were
we to assume that David was the one to build stratum IVB, and Solomon
not the one to destroy it, we would have to conclude that the palace had
been destroyedby some enemy in David's times or at the beginning of Solo-
mon's times - a conclusion that has no support in the written sources as
far as this region is concerned.
Albright and Wright made valiant efforts, some ingenious, to clarify
the stratigraphyemanating from these discoveries.They proved conclusively
that not only should the palace and structure 1482 be ascribed to stratum
IVB, but also a number of other structureswrongly attributed by the exca-
vators to stratum VA. Thus Albright and Wright introduced a new stratum,
which they called IVB-VA, comprising the stables and the offsets/insets
wall; we shall designate it VA to distinguish it from IV. But even they did
not succeed in overcoming the main difficulty, namely the attribution of
stratum VA, because they accepted the assumption of the excavators that
the Solomonic stratum was IV. They were therefore also compelled to as-
72 THE BIBLICAL ARCHAEOLOGIST (Vol. XXXIII,

sume that the palace and the other buildings were built in David's time,
which, as we have seen, was impossible historically. From all the above it
became clear to us that the stumbling block in attributing the offsets/insets
wall to Solomon was not only its difference from the Solomonic walls at
Hazor and Gezer (casemates) but also its stratigraphy.
The New Excavations at Megiddo
In order to clarify the problem of the \lcgiddo walls in Solomon's
times, I excavated there briefly in 1960 (13-15 Jan., and again for a few
days later in the month). I asked the late I. Dunayevsky to join me as ar-
chitect, feeling that I might have preconceived ideas and wanting to ensure
that every conclusion would be independently and objectively checked
from the stratigraphicalview point, and who was more competent for that
than Dunayevsky? I asked him in fact, to be the "devil'sadvocate"and bring
up as many difficulties as possible. His participationin that particular dig
and the two subsequent ones was of decisive importance, and all the con-
clusions that follow were shared by him, indeed, were reached with his
assistance.
The area we selected for our trial dig was located in the northern side
of the eastern half of the tell, cast of the "SchumachcrTrench" and the DD
area of the Megiddo excavators.In that area the offset/insets wall had not
been removed by them and was thus well preserved.Futhermore, an aerial
photograph published by them showed that under and near the offsets/in-
sets wall, some walls cxisted that had not been marked on the plan, nor
were they discussed in their report. We attributed particularimportance to
the fact that here too was located the northern stable complex (407) which
had been only partially excavated, and its remnants were still strewn around
the area.
This brief dig brought interesting surprises,of which the most import-
ant was the discovery of a second palace, also built of ashlar like the Solo-
monic gate, with a casemate wall on either side. Consequently a partial
report of the campaign was published on the problem in general.
My subsequent archaeological activity in 1960-61 in the Bar-Kochba
caves and the enormous excavations at Masada in 1963-65, delayed the de-
tailed report of our short Mlcgiddodig. When we were finally ready for it
in 1965, I realized that it was imperative to re-examine the stratigraphical
problems related to our discoveries and especially the date of the famous
water system, which in accordancewith our finds, seemed to require fresh
examination. At the end of a short dig in 1966 (31st July - 12 Aug.) with
students of the Institute of Archaeology of the Hebrew University, many
of the problems were decisively clarified. Those which required further
investigation were the strata under the casemate wall. Sure enough, dur-
1970, 3) THE BIBLICAL
THE BIBLICALARCHAEOLOGIST
ARCHAEOLOGIST 73

ing the third season, March 26 to April 7, 1967, we were lucky - on

the very last day of the dig - to come up with the decisivestratigraphical
discovery.The descriptionof this and our conclusionswill now be presented
by subject,with occasionalreferenceto what had been found in everysea-
son in relationto the respectivediscovery.
52 'I7

35\ IVA
mi
sY ~ VA-IVBM
0 2 4 VB M
I i I VIA -
the northerntier of rooms. The various trenches of the excavatorare indicated, so that
one senses how the dimensions of the building were determined. Beneath the building,
in the upper left, are traces of earlier strata.
The Additional(6000
Northern Palac
The firstday of the excavations
broughtwith it our firstsurprise.
As
we startedmarkingthe exactspotof ourintendedtrench- standingout-
side the offsets/insets wall, on the slope north of it - we realized that what
seemed to be the lower part of the foundations of the wall was actually
built in a straight line (without offsets and insets) measuring twenty-eight
meters.It also becameapparentthat while the wall itself had been con-
structedof field stones, or small dressedstones, to strengthenthe comers
of the offsetsand insets, this stretchof wall was built of ashlars,some of
which had marginsdressedin the mannerof Solomon'sgate and southern
palace, and previouslyunknownin Megiddo.That stretchof ashlarwall
(52 on Fig. 3) stoppedabruptlyin a straightand perpendicular line at its
westernend. At thatpointthe ashlarswereparticularly largeand well dressed,
and laid in the header-stretcher fashion,similarto the six-chambered gate.
74 THE BIBLICAL ARCHAEOLOGIST (Vol. XXXIII,

While contemplating this fact, it occurredto me that at that particularspot

the offsets/insets wall must have been built on the foundations of some
external wall of a palace (or fort) that had been built on the tell's edge,
like the southern palace (1723). In order to verify that assumption,we de-
cided to carefully peel off the foundations of the offsets/insets wall above
the beautiful ashlar, at what appeared to be the northwest corer of the
building. Should our assumption prove correct, then the western wall of the
building must be found under the foundations of the offsets/insets wall, con-
tinuing south into the city. Sure enough that wall was found as soon as the
foundations had been removed. It was about 1.5 meter thick, built of ashlars

0 5a I
t0a.
20
I Meters
Fig. 4. Plan of the palace of Zinjirli, comparable to palace 6000 at Megiddo. Note the corner
tower (lower right), the long central court, and the smaller rooms surrounding the
court.

laid header and stretcher fashion both inside and outside, with the gap be-
tween them filled up with small stones, similar to the building method of
the southern palace. We followed that wall into the city and discoveredthat
it stretched not only under the offsets/insets wall, but even under the foun-
dations of the northern stables complex (407)! During the 1966-67 cam-
paigns we continued excavating this large structure (trenches only) and
succeeded in ascertaining its plan. We were particularly impressed by its
1970, 3) THE BIBLICALARCHAEOLOGIST 75

southernfront,in which the thicknessof the beautifullybuilt wall reached

over two meters.
The building is rectangularin shape: its length from east to west is
abouttwenty-eightmetersand fromnorthto southabouttwenty-onemeters,
occupyinga built-upareaof 600 squaremeters,slightlymorethan the south-
ern palace (500 squaremeters). The building'sfront was to the south;on
its north(on the edge of the tell) it had five roomswhile the east and west
sides had elongatedhalls. Its southeastcorer must have had a tower,and
the southwestcornera squareroom.The northeastcornerwas built in the
shape of a big offset. Althoughthe excavationof this palace has not yet
been completed,its plan can be clearlyreconstructed and a glimpseat it
proves that it resembles the of
greatly type palaces known fromZinjirliand
other sites on the Phoeniciancoast,all fromthe early centuriesof the first
millennium.This type - known in Assyrianas bit hilani - servedas a
ceremonialpalace;those enteringit from the wide front found themselves
in a centralcourtsurroundedby roomson threeflanks.The bit hilani type
also as a rule had a towerat one corner(see Fig. 4). This factis particularly
interesting,becauseit addsfurtherevidenceto the biblicalassertionof Phoe-
nician influence on Solomon'sbuilding activities.But let's return to the
stratigraphical problem.It becamequite clear that here, as on the southern
edge, monumental buildingsexisted, identical in building style with the
"Solomonicstables."In other words,what was found was not an isolated
structurebut a largecity with magnificentbuildings.In someof the palace's
rooms,we found much potteryin situ (store jars, dishes,juglets,cooking
pots) in a state that indicatedsuddendestruction.All of them were charac-
teristicof the 10thcentury.
Robbers' Trenches
Through careful examinationof the ruins in the stratumunder the
stables,we cameup with an importanttechnicalphenomenonwhich helped
us not only to locate anotherhuge structureof the same period,south of
palace 6000, but also to comprehendwhy the earlierMegiddoexcavators
had failed to locatethe structuresof the magnificentcity.
South of the palace,acrossa ratherwide street(ten metersor so), we
discoveredthe remainsof a largestructureresemblingthe buildingstyle of
otherstructuresin VA-IVB.However,exceptfor a few placesin which the
lowerstoneswerepreserved,we noticedbeautifulbeatenchalkfloors,stretch-
ing practicallyto the assumedline of the walls, then abruptlystopping.It
becameevident that most of the beautifulashlarstones of structuresIVB
had been pulled out and robbed.No doubt this was done mainlyby the
buildersof the stablesand the offsets/insetwall. This also explainsthe fact
that a numberof stonesbearingmason'smarkswere discoveredin IVB and
76 THE BIBLICAL ARCHAEOLOGIST (Vol. XXXIII,

IV. Once we realized this phenomenon, we managed by closely following

the floors and robbers'trenches, to discover the remains of a huge structure,
with a front at least fifty-five meters long, and rooms and courtyardsbranch-
ing southwards.This structure was on the other side of the street, and it is
a pity that most of it, in the south, vanished in the previous excavations
which had reacheda greatdepth at that spot.

Fig. 5. A Stratum VA-IVB foundation of a square pillar (perhaps a staircase). Note especially
the termination of the plaster floor in a straight line, showing the line of a robber
trench by which another Solomonic building at Megiddo was discovered.

The stratigraphicalattribution of the new palace (6000) and the huge

structure to its south, in relation to the offsets/insets wall and the stables, is
clear enough if one counts the strata from top to bottom, as in the case of
the southern palace (1723) and its adjacent structure, 1482. The offsets/in-
sets wall and the stables belong to stratum IV, while the whole complex be-
neath them belongs to IVB, since there are no structureswhatsoeverbetween
the two. But it is our task to prove that such attribution is correct too, if
counted from bottom to top, since theoretically it might be argued that the
discoveredbuildings - even though they lay beneath the stables and wall -
might not belong to stratum IVB but to an earlier period or stratum.
Already in 1960 we discovered two levels of walls (excavated by the
Megiddo excavatorsbut not marked on their plans) outside the palace and
continuing beneath it. By their characteristicsand based on what could be
seen in a trench west of the palace, we ascribedthese walls to strataVB and
1970, 3)
1970, TIHE BIBLICALARCHAIEOLOGIST
THE BI[BLICALARCHAEOLOGIST 77

VIA. But this still neededto be proved,and a goodpartof our 1966efforts

were spent tryingto do so. We decidedto dig underthe floorof the north-
westernroomof the palace,in the hope of hitting the remainsof the huge
conflagrationof stratumVIA which were evidentall over the tell. And this
indeedoccurred.Beneaththe palacefloor,we firstcameacrossrelativelypoor
walls (VB) and immediatelyunderthem we uncovereda structuretypical
of stratumVIA. It was completelycoveredby a thicklayerof ashes,in which
we found a wealth of typicalpottery(many vesselsintact) includingthe
famous"beermugs"of the decadentstyle - as definedby T. Dothan-
ascribedto the third phase of the Philistineculture,that is, to the second
half of the 11th century.Thus we corroborated absolutelythat the palace
did indeed belong to stratumIVB-VA,by counting from either the top
downwardsor the bottomupwards.

~'
:' q

6. Contents
Fig. 6.
Fig. the pocket-book
of the
Contents of under the
found under
pocket-bookfound floor of
the floor StratumVIA
of Stratum VIA ruins at Megiddo.
ruins at Megiddo.

An Unexpected
An Treasure
Unexpected Treasure
Our check dig
Ourcheck was, as
dig was, above, aimed
explained above,
as explained at uncovering
aimed at stratigra-
uncovering stratigra-
facts:
phical facts:walls,
phical levels and in situ
walls, levelsandin situpottery.
pottery. But
But even
even on
on this
this kind
kind of exca-
of exca-
vation, excitement
vation, excitement among
among the
the diggers
diggers soared
soared when
when they
they hit
hit a veritable
veritable trea-
trea-
78 THE BIBLICAL ARCHAEOLOGIST (Vol. XXXIII,
sure which shed light on one of the interesting cultures of Megiddo. When
we removed the level of ashes from stratum VIA with its pottery and all,
on the threshold of the room and below the floor we encountered a woman's
treasure in a cloth bag (of which remnants were preserved) protected by a
few surroundingsherds. Into this small bag (ten by fifteen cms.) the woman

Fig. 7. Two bronze weights from the

managed to squeeze an enormous number of objects that must have been

precious to her, such as nine ivory spindle whorls, two pomegranate-shaped
pendants taken from a tripod typical of the time, iron bracelets(!), a ring,
hundreds of tiny beads, semi-preciousstones and two bronze weights in the
shape of a horned animal and a squatting monkey. Before the haul was
cleaned we thought the monkey was of the prevalent type with one arm on
his mouth and the other on his ear but after cleaning, it appeared that the
monkey was holding an apple-like object in one hand. We joked about what
seemed like an artist's effort to depict Adam holding the apple with one
hand and using the other to block his ears against Eve's temptation.
1970, 3)
1970, THE
THE BIBLICAL ARCHAEOLOGIST
BIBLICAL ARCHAEOLOGIST 79
79

This find, in addition to the enormous quantity of metal vessels dis-

covered by the Megiddo excavators, testifies to the high standard of the
Philistine-Canaanite metal craft in Saul's times (as we read in the Bible).
It is interesting to note that, also in Tell Qasileh, the first metal utensils
appear in stratumX, which is contemporarywith stratum VIA in Megiddo.
,- 6002
6001
6003

..... ~ solo
407Asmentioned, our primary objective in 1960 was to find out
already
of
stratum
covery the IVBin
palacewas a very VA-IVBnk
in the cla
0 5 0
lol1 i-BB -~ ~ VB Mg
VIA
Fig. 88. The plan of palace 6000 and its i" adjacent casemates (6001-6003, 6007-6009) after re-
movat of the offsets/insets wall in the northeast portion of the Meiddo tell. Note por-
tions of earlier strata under the casemates on both sides of the palac.

The Eastern Casemate

As already mentioned, our primary objective in 1960 was to find out
if there was a casematewall hidden beneath the offsets/insets wall. The dis-
covery of the palace in stratum IVB was a very important link in the clari-
fication of the Solomonic stratum and its character,but our main problem
had been the city's fortificationsduring that period. So we selected the sec-
tion of the offsets/insets wall lying east of palace 6000. It became evident
in no time that the portion close to the corner of the palace was completely
demolished: both the offsets/insets wall and the upper courses of the corner
were missing. Clearly this must have been the result of relatively late des-
truction. We started our trial trench, therefore, about ten meters east of the
palace corner. After carefully peeling off the foundations of the offsets/in-
sets wall, we hit a fill of earth and field stones, and immediately beneath it
a casemate wall appeared! During that season we only managed to cut a
few trenches, but they sufficed to give us the estimated plan of three case-
mates. However, in 1967 we removed further portions of the offsets/insets
wall, so now we have (see Fig. 8) not only the plan of the above three
80 THE BIBLICALARCHAEOLOGIST (Vol. XXXIII,

casemates,but alsotwo moresmallercasematesnearthe fort,altogetherabout

thirty-fivemetersin length.
The casematesin that sectionare built of none-too-thick walls (exter-
nal wall aboutone meter) with an averagedistanceof two metersbetween
them. They are made of field stones(as a foundationfor a brickwall) like
the Hazorwall. The length of the ordinarycasematesis aboutsevenmeters,
but where the wall turnsat an obtuseangle towardsthe fort,there are two
smallercasemates.The casematewall here is less strongthan the one at
Hazor,which is easily understandable since the slope here is much higher
and steeper,and in fact it is impossibleto attackthe city fromthat side.The
1967 dig proved- after we removedfurthersectionsof the offsets/insets
wall - that althoughoccasionallydwellingswere built inside the city as a
straightcontinuationof the casemates'crosswalls (similarto the casemate
wall of Tell Beit Mirsimwhich also dates to the 10th century), there was
no connectionbetween them (see in particularcasemates6008-6009-6010).
After the 1960 and 1966 campaignsone problemstill remainedwhich
I deemednecessaryto clarifythoroughlyand decisively.Since we uncovered
the casematewall only in sections,we realizedthat its foundationslay on
the burntlayerof stratumVIA. In otherwords,althoughwe could attribute
it (counting downwards)to stratumIVB, one could argue that the case-
mate wall was in fact part of stratumVB, having been built on VIA. In
that case the wall had been integratedinto and re-usedin the IVB fortifica-
tions. This appearedillogicalbecause,as we shall see later, there was no
evidencethat city VA had been fortifiedat all. At any rate it was impera-
tive to clarifythe point.To that end we formeda specialtaskforcein 1967
headedby the late I. Dunayevskywho was assistedby two of the ablestIn-
stitute of Archaeologygraduates,their missionbeing to come up with an
independentverdict on the stratigraphyof that section. Work here was
deliberatelyslow, and carriedout with the most delicateof implements.
Those who have never participatedin archaeological excavationsmay
think that the greatestjoy of the excavatoris to find a hidden treasureor
objectsof artisticmerit;this is not so! The big discoveryof 1967 at Megiddo
was a few wretchedwalls seeminglyof no significance.In our case their
importance was great.
On April 6 (as usual on the last day but one beforethe end of the
campaign)we discoveredthe walls of stratumV beneaththe walls of the
casematewall and above the burnt layer of stratumVIA. Thus the circle
closed:here too, as in the palace,we had proofthat the casematebelonged
to stratumIVB, countingboth upwardsand downwards.Moreover,a slow
and carefuldig even provedthat a stone-paved areain the joint betweenthe
casematewall and the palace comer (the destroyedsection) touches and
1970, 3) THE BIBLICAL
ARCHAEOLOGIST 81

.; %

, ..
- Il

N . , .

*01"!? --.2IL

.- ai

-t

Fig. 9. The western casemates running away from palace 6000, the western wall of which
appears in the foreground.
82 THE BIBLICAL ARCHAEOLOGIST (Vol. XXXIII,

leans directly on the palace, without any gap, and is built in accordancewith
the expected turning of the casematewall.
The Western Casemate
From the start of our excavationwe discerned a wall-top made of ashlar
stones and constructedmainly of headers, which appearedwest of the palace
in a level below that of the solid wall and stables. That wall had already
been uncovered by the Megiddo excavators, as can be seen in the aerial
photograph published by them, although it was not marked on their maps.
At the beginning of our dig there, it became apparent that the wall (56)
was interruptedby openings at fixed intervals. Additional examinationproved
that these openings lead to casematesof which the external, northern wall is
built of large field stones, partially dressed, and extends in a westerly direc-

Fig. 10. A typical casemate to the west of palace 6000, looking north (from inside the city
outward). Note the entrance at the lower left and the headers and stretchers in
the wall construction. The outer wall is built with field stones.

tion towards the city gate as a direct continuation of the northern wall of the
palace. This wall too, was of course built below the offsets/insets wall. Dur-
ing the 1966 and 1967 campaigns we excavated the whole set of these case-
mates, even below their floors, until we hit upon stratumVB. Altogether we
uncovered two complete casemates (6002-6003) and one half of another
(6001; see Fig. 8). We did not manage to uncover the western part of 6001,
because it was here that area DD of the Megiddo excavatorsbegan, and
that is but an extension of the Schumacher trench. But by examining the
1970, 3) THE BIBLICALARCHAEOLOGIST 83

Fig. 11. Aerial view of the Megiddo mound in its northeast sector. The Schumacher trench is
the dark gash at the left. The offsets/insets wall can be seen all along the northern
edge of the ruins, but a careful look will reveal fragments of the casemate wall and
of palace 6000. Photograph courtesy of the University of Chicago's Oriental Institute.
84 THE BIBLICALARCHAEOLOGIST (Vol. XXXIII,

plans of the American excavators,it is possible to reconstruct the casemate

and even discern the traces of a fourth casemate. We do not know what
Schumacher discovered and nothing can be leart about this from his pub-
lications, but there seems to be no doubt that these casemates continued
westwards to the gate. Theoretically, as we stated in our first report, it could
be argued that the casematesbelonged to the northern part of a court which
might have surrounded the fort. But during the 1966-67 seasons, after the
southern comers of the fort and particularlythe remains of the huge struc-
ture to its south had also been discovered, it became clear that this did not
make sense, since no turn of that imaginary court had been found. Was it
possible that Schumacher and the American excavatorshad come upon the
casemates between our trench and the gate, without discerning them? It is
possible, since as we have already observed some sections were marked in
their plans, while others uncovered by them were not. It is feasible, though,
that the wall in this particularsection was deliberately demolished by the
builders of the offsets/insets wall for its stones. From this particularportion,
dug up by us, a very slanting slope begins as far as the gate, and it is con-
ceivable that when the offsets/insets wall was built, the earlier wall could
not serve to support it and the whole area had to be reorganized.The fact
that here the casemates had been built on the very edge of the tell indi-
cates clearly that they belong to the general line of the IVB fortifications.
The casemates'floorsin this area yielded not only a whole complex of typical
pottery (similar to the pottery in the palace) but brought to light the re-
mains of a large structure from stratum VB, built completely of bricks on
a course of field stones. Its plan, in addition to the other finds in this stratum
which were unearthed by the Megiddo excavators (as well as the walls we
had discovered below the eastern casemates), testify clearly that the poor
city of stratum VB had not been fortified and certainly had no city gate. The
houses were close together even on the edge of the tell and their external
wall served as some kind of defense to this town.
The Megiddo Gates
Once we had established that Solomon's real VA-IVB city had been
properly fortified and contained many splendid public buildings, we could
try, even briefly, to clarify the complex of the Megiddo gates in the Israelite
period. The problem is not easy, not only because most of the gates and
especially the wall adjacent to them were destroyed by the excavators as
they deepened their search, but also because the area had been excavatedby
two separate excavators and some important details do not figure in their
records. In order to grasp the complexity of the problems and our proposals
concerning them, it might be best to follow briefly the developments of the
fortificationsin Megiddo in the period preceding and following Solomon.
1970, 3) THE BIBLICALARCHAEOLOGIST 85

Since the offsets/insetswall was the only one to last to the end of
Megiddoin the Iron age, and since on the other hand (as we shall see
later) post-Solomonicgates were discoveredby the excavatorsin various
strata,it was assumedrightly that the new gates had from time to time
been integratedinto that wall. This is a naturalprocess,since generallythe
surroundingsof the gate are destroyedwhere the enemy makeshis main
break-through. Hence the offsets/insetswall carriedon throughstratumIII.

325

325

Em IVA
m VA-IVB
"VIA
^^ ~ ; meters

Fig. 12. The complex of city gates along the north edge of the Megiddo mound. The sturdy
VIA gate appears under the six-chamber Solomonic gate.

The city in stratumII was unwalled,for a large fort was discoveredsouth-

east of the northernstables,which had been partiallybuilt above the wall
of the last city, namely the offsets/insetswall. Similarly,the last city in
stratumI, of the Babylonianand Persianperiods,had been open and un-
walled.
As for the Ironage periodsbelow the Solomonicgate, in stratumVIIA,
the magnificentsix-pilaster(or four-chambered) Bronzeage gate continued
to exist. In stratumVIA a small but well-constructed gate was discovered
(see Fig. 12). Directlyaboveit Solomonbuilt his gate, which was the first
86 THE BIBLICAL ARCHAEOLOGIST (Vol. XXXIII,

to be constructedof drafted ashlaror diagonally dressedstones, as was typical

at the beginning of the Israelite kingdom. The excavatorsassumed that the
gate of VIA continued through VB, but had no facts to substantiatethe as-
sumption. Let it be said right away that in a check we made in 1967 we as-
certained that no gate existed in the VB city, a fact which fits in well with
its unfortified character.Up to the time of our dig the principal difficulty in
ascribing the Megiddo gates to the city walls was mainly in the discoveries
above the Solomonic gate.

Fig. 13. Eastern half of the Solomonic gate as it looks today. Note the two phases; the original
is founded on the ruins of the VIA gate, while the upper phase was built on the
filling of the gate's chambers. One of the fillings can be seen between two piers; above
it is the threshold of the gate of the latest level.

Above the ruins of Solomon's six-chamberedgate, P. L. O. Guy (the

discoverer of the stables) discovered two additional gates. While unearthing
the remains of the first, Guy thought it should be attributed to Solomon
because it was bound to the offsets/insets wall (which, as mentioned, he
had ascribed to Solomon) and was built of ashlar similar to those of stratum
IV. But since in the 1935-36 campaign under Loud the six-chamberedgate
had been found, the excavatorshad no doubt that the "Guy gate" was post-
Solomonic. However, as they ascribed the late phase of stratum IV to Sol-
omon, they attributed the structureto Level III, the last fortified city. Then
1970, 3) THE BIBLICAL ARCHAEOLOGIST 87

a new difficulty arose when it was established without doubt that that gate
(500) comprised in fact two gates, which from the stratigraphical-technical
viewpoint belonged to separate periods and had different plans. Since both
gates could only be ascribed to stratum III, the older of the two was labeled
gate IIIB and the later gate III. As it was impossible to ascribe them to one
single level from the historical viewpoint, the writers of the season's report
came to the strange conclusion that in fact only one gate existed, gate III,
while gate IIIB was taken to be a misconception, a mistaken beginning
abandoned in due course in favor of the gate III plan. They seemed to think
that all these alterations between gate IIIB and gate III were a result of
sudden decisions arrived at in the course of building. Some scholars who
disagreed with this and attributed gate IIIB to stratum IV found themselves
then without a Solomonic wall.

Fig. 14. Megiddo gates of Strata IVA and III (the former in dashed shading and the latter in
dotted shading), shown as they relate to the Solomonic gateway underneath, shown in
outline.

Now, with the discovery of the casemate wall, it seems that the con-
fusion can be dispelled:
The six-chamberedgate (Solomon) - casemate wall- stratum IVB
The four-chamberedgate (IIIB) - offsets/insets wall - stratum IV,
main phase.
The two-chamberedgate (III) - offsets/insets wall - stratum III.
88 THE BIBLICAL ARCHAEOLOGIST (Vol. XXXIII,

This development in gate structure fits in well with what we know from
other sites.
In our first report following the 1960 digging season, we assumed that
the offsets/insets wall was first built together with the four-chamberedgate
(IIIB). After the 1966 and particularly the 1967 seasons (when we made
small sections in the vicinity of the gates), we arrived at the conclusion
that while the offsets/insets wall was first built by Jeroboam I (?) follow-
ing Shishak's destruction, it was still leaning on Solomon's gate (which had
apparently not been demolished by Shishak, a fact further attested by its
condition when unearthed) except that its level had had to be raised. For
the moment we will designate it IVA1. This conjecture is further substan-
tiated by the fact that the gate does indeed have two separate floor levels.
The Megiddo excavators assumed that the magnificent construction, mostly
of well-dressed ashlars, was only the gate's foundation. But this assumption
(based mainly on the lack of a wall in its main phase) had one great stumbl-
ing block: not only were the chambersdiscoveredto be filled up and bordered
by secondary walls (of field stones) blocking them off towards the gateway,
but even the gate's entrances, external and internal, were blocked by thick
supporting walls made of field stones. Had the builders really planned the
portion of the gate which was discovered as a mere foundation, they would
certainly have built it differently, incorporating the supporting and revet-
ment walls into the structure. Moreover, they would certainly not have built
it of well-dressed ashlars (as is known from examples in Samaria, Ramat
Rahel, etc., where the foundations are built of drafted stones and only the
walls on top of them of well-dressedstones). On the other hand, when the
offsets/insets wall was joined in the higher level, its' builders had no al-
ternative but to block the gate and raise its level. Accordingly the Megiddo
excavatorswere bound to find a lower level in the gate's construction,which
indeed existed. Below the level of the road leading to the later gates, the
excavators discovered a road made of well pressed chalk. That level leads
towards the lower courses of the six-chamberedgate. However, since they
assumed the gate structureto be but a foundation, the excavators attributed
that level of pressed chalk to the stratum VB gate. But we have already as-
certained that that level had neither fortificationsnor gate. In fact, the lower
courses, partly covered by the road, are the very foundation courses of
Solomon's gate, and not of gate V, which as explained, did not exist at all.
Thus Solomon's gate was well founded on the foundations of gate VI, and
its structurewas solid, a fact which had been also recognized by Miss Ken-
yon. The problem is complex indeed and we are able here to show only
some schematic plans preparedby the late I. Dunayevsky. The final report
will have detailed maps and plans in which the remains of the structurein-
1970, 3) THE BIBLICAL ARCHAEOLOGIST 89

Fig. 15. Gallery 629 seen from the air. Note the typical header-and-stretcher method of con-
struction.

side the city (near the gates) will also be incorporatedin accordancewith
our above conclusions.
The Water Systems - Israelite and Not Canaanite
I would like to end the description of our short excavations with the
problem of the dating of the gigantic water system of Megiddo. To illumin-
ate the problem and its reference to our excavations let us briefly describe
the system and the dates attributed to it by the previous excavators.
90 THE BIBLICAL ARCHAEOLOGIST (Vol. XXXIII,

One of the most interesting and spectacularstructures discovered by

the Megiddo excavators in the southeastern edge of the tell, west of but
near the southern stable complex (1579), was labeled by them gallery 629.
This gallery is only a narrow passage, slightly over a meter in width, lead-
ing from the city outwards towards the southwestern slope of the tell, at
the bottom of which the well-known spring-caveexists. The passage walls
(revetment walls of which only the internal face was well-built) were con-
structed of well-dressed ashlars, laid header and stretcher fashion (Fig. 15).
The discovered walls stood up to two meters high. One may assume as the
excavatorsdid, that the passagewas coveredand camouflagedand led towards
the spring. Stratigraphicallyspeaking, the passage is under the offsets/insets
wall (which was considered Solomonic) and for that reason it was assumed
that the "gallery"had been built before Solomon; however, since it was
impossible to ascribe this magnificent structure to the relatively poor strata
of the 11th century, it was attributedto Rameses III, namely stratum VIIA,
in which Megiddo still retained some of its erstwhile splendor. Fixing the
gallery's date was a cardinal point of departure for R. Lamon, the archae-
ologist who published the report on the Megiddo water systems, as well as
for the dating of the famous Megiddo water tunnel. That tunnel consists
of two parts: a vertical shaft (925) and a horizontal tunnel (1000) con-
ducting the water towards the shaft. The upper part of the shaft was cut
through layers of accumulation of the earlier strata, and its walls were lined
with stone. Its lower part, on the other hand, was rock-cut. To achieve its
main purpose, namely water supply in time of siege, it was necessary to
block the cave opening from the outside, so as to prevent the enemy from
poisoning or blocking the spring water. Sure enough, the cave entrance was
discovered by the Megiddo excavators to be blocked by a wall of huge
stones. At first they sought to determine the date of the tunnel's quarrying
through the identification of the latest stratum through whose ruined layers
the shaft was dug. It became apparent,though, that the highest stratumwhich
carried clear evidence of destruction by the shaft contained sherds from the
Late Bronze age, the 14th-13thcenturies. The excavatorscould not check con-
ditions above that stratumsince the strataabove the excavatedshaft collapsed
in antiquity together with part of the stone lining of its walls. They con-
cluded, therefore, that the shaft was quarried later than the 14th-13th cen-
turies or at their end. Another point of departure that was important in
the dating, was the fact that by the cave entrance (inside the blocking wall
1007) a man's burnt skeleton was found. Lamon assumed that the man must
have been the cave's guard who was killed in enemy attack and since there
was no point in guarding the cave after it had been blocked, the guard most
probably died before the shaft and tunnel were cut, while the citizens of
1970,
1970, 3) THE BIBLICALARCHAEOLOGIST
ARCHAEOLOGIST 91
91

Megiddo still had access to the spring by an external way on the slope. The
fact that the latest pottery found belonged to the 12th century brought about
the assumption that the cave's blocking and therefore also the quarrying of
the shaft and tunnel were executed in the middle of, rather than before, the
12th century.
But let us resume the problem of the gallery. Lamon assumed rightly
that the gallery's function was to make the spring accessible before the
quarrying of the tunnel and shaft. Therefore, in his opinion it belonged to
the first half of the 12th century.

629

Fig. 16. Plan of gallery 629, showing how it cut structures of VIA (darker shading) and the
skimpier remains of VB (lighter shading).

A New Attack on the Problem

Having ascertained already in 1960 that the offsets/insets wall was

post-Solomonic, there was no further reason to assume that the gallery pre-
ceded Solomon. On the contrary, the gallery's stratigraphicalcircumstances
were identical with those of the southern palace and all the other structures
in stratum IVB. Moreover, the manner of its construction and stone dress-
ing matched to a surprisingdegree the constructionmethods of that stratum.
In 1960 we scraped the vicinity of the gallery a little, just enough to realize
that its foundation was built into the fallen burt bricks of stratum VIA.
We assumedtherefore again that it was built after the destructionof stratum
VIA but before the offsets/insets wall, and so we said in our preliminary
92 THE BIBLICAL ARCHAEOLOGIST (Vol. XXXIII,

report of that campaign. It follows that the quarried water system was built
at the end of Solomon's or Rehoboam'speriod, or - as seems to us more
likely now - at the beginning of the Divided Monarchy. The dating of
the water system is of paramount importance and for that reason we de-
cided in 1966 to check our assupmtions by excavating. The method was
simple enough: since it was obvious that the gallery had been dug into
previous strata (a fact that was evident also from its walls, which, as prev-
iously mentioned, were actually revetment walls, well finished where they
faced the passage and roughly built towards the edges) we selected a square
adjacent to the gallery on its south, and began to excavate it thoroughly.
Our purpose was to determine the upper strata damaged through the gal-
lery's quarrying. We were fortunate indeed. We succeeded in descending to
stratum VIIA and discovering the whole array of strata familiar from other

j S

Fig. 17. Mason's mark on one of the cut blocks from gallery 629, identical with one of
those used in Solomonic structures elsewhere at Megiddo.

areas, and even in enriching the pottery repertoire of some. It became de-
cidedly clear that the gallery's quarrying damaged both stratum VIA and
VB (see Fig. 16). Thus the gallery was squeezed stratigraphicallyspeaking,
both when counting from top to bottom and when counting from bottom
to top. It was built in stratum IVB just like the other palaces and Solomon's
gate. Had we needed further proof for the attribution of the gallery to
stratum IVB, we now had it by chance. Since the excavation of the gallery
by the first excavatorsof Megiddo, some of its stones had come loose and
fallen to the ground. On overturning one such stone we noticed that it bore
a mason's mark (Fig. 17) identical with marks discovered in the stratum
IVB structuresand in secondary use in stratum IV. It follows that the shaft
and tunnel are later than Solomon and must have been cut at the beginning
of the 9th century, during the reign of Omri's house.
Following the above finds, we succeeded in discovering also in Hazor
a similar water system from the same period. In the 1967 campaign at
1970, 3) THE BIBLICAL ARCHAEOLOGIST 93

Megiddo we made several trial trenches outside the cave entrance in order
to find out the history of its use through earlier periods. The dig was rather
complicated and it is impossible to enumerate all its details here. It would
suffice to state that we have sufficientproof that the cave and spring, together
with the hewn steps leading to it from the outside and supporting and re-
vetment walls, were used largely in the Middle Bronze II period.
An Additional Water System
While were were busy examining the Israelite strata, we decided to
check an additional find of the earlier excavators. Under the city gates
and related to the road leading towards them, a well built staircasewas dis-
covered, which seemed to have been roofed. The Megiddo excavators had
not completed its excavation and in their report they suggest that these
stairs may have been part of a "pedestrian"approach. That interpretation
made no sense and seemed to apply modem problems to ancient times. It
made more sense to assume that the staircaseled to an additional source of
water in the north, which may have been connected with the other spring
of Megiddo discovered near the Jenin-Haifa road. To verify all this we had
to dig at the bottom of the steps. To start with we removed the large earth
dumps with the aid of mechanical equipment. After a few days we discov-
ered the continuation of the stairs turning east at a right angle towards a
well-plastered pool. It was clear, thus, that this was no entrance to the city
but a descent from the city to the water. We did not complete our examin-
ation in this spot and cannot therefore say with certainty whether the water
source was a pool that was filled by canal from the spring or from some
other reservoir. One thing was clear: we had encountered a second water
system, approximatelyfrom Ahab's period, which may have served in times
of peace, since it was located beyond the city's main fortifications.
Conclusions
Let us sum up the dates and characterof the cities from the beginning
of the Iron age. Stratum VIIA, the first Iron age stratum, can be clearly
attributed to the period from Ramses III to Ramses VI on account of the
objects which bear their respective names. Here it was that a considerable
amount of early Philistine pottery was found, and one can assume that this
city had a garrison of Sea people.
The city erected on the ruins of VIIA was VIB, differing considerably
from its predecessor;judging by its buildings and their style, it could not
have been a sizable city. A large amount of middle-phasePhilistine pottery
was found here. This city's earliest possible date is 1150, and probablyabout
1120. It did not last long and it most probably was destroyed in the first
half of the 11th century. The most important town at Megiddo following
the destruction of VIIA and preceding its reconstructionby Solomon was
94 THE BIBLICALARCHAEOLOGIST (Vol. XXXIII,

Fig. 18. Staircase beneath the city gates and leading down the slope near the access roads,
now shown to be part of water system probably from Ahab's time in the 9th century
B.C.

undoubtedly the city of stratum VIA. It was newly planned, and in var-
ious places large and spacious structures were discovered. Noteworthy is
also the density of the settlement at that time. All houses were built of
1970, 3) THE BIBLICAL ARCHAEOLOGIST 95

bricks on stone foundations. Finds here were particularlyrich, among them

lots of metal tools. Outstanding are the Philistine pots of the third, degen-
erate phase. This city was demolished by a fire of which traces are visible
all over. The thick layers of ashes, the burnt bricks and the profusion of
pottery are so typical that it is impossible to mistake its identity in every
area. T. Dothan, in her book on the Philistines, has proved conclusively that
the culture of this stratum is comparable to that of Tell Qasile X in the
second half of the 11th century. It is interesting to note that the earliest
iron tools appear both in Tell Qasile and in Megiddo. There is no doubt
that Megiddo VIA was a developed Canaanite-Philistine city typical of
the Canaanite-Philistineculture of Saul's days, when the Philistines dom-
inated all metal craft (I Sam. 13:19). This city was completely destroyed,
probably by David, as suggested by a number of scholars, notably Mazar.
It was followed by a tiny hamlet in VB, or at least by an open city devoid
of wall and gate.
We come now to city IVB-VA. With the new discoveries there seems
no doubt that this was Solomon's city, not only because the Bible says spe-
cifically that Solomon built Megiddo, but also because all the pottery and
architecturalevidence (the style of stone dressing and the building plans)
point that way. No more are we confronted with a single fort, but with a
metropolis with stately buildings of official and ceremonial character. This
city is surrounded by a casemate wall with a gigantic gate of six chambers
and two towers, and has a secret passage (the gallery) leading to the water
spring. This passage is built in the same style as the other IVB structures
and adds proof to the existence of a wall around the city. Similar to the
Solomonic gate, the structures of this city, too, are the earliest to be built
in the Phoenician style (with its typical margin dressing) widely adopted
in Solomon's days according to the Bible. One might even assume that the
southern palace was the residence of Solomon's governor (Ba'ana son of
Ahilud; I Kings 4:12), while the northern (6000) Bit hilani type served
ceremonial purposes and perhaps even for the king himself whenever he
visited Megiddo. I have already mentioned that we found the palace floors
strewn with many pots, mostly broken, which serves to testify to the enor-
mous destructionof Megiddo in the days of Rehoboam, following Shishak's
campaign. A fragment of a stele of this Pharaoh found on the tell during
the earlier excavations confirms this decisively.
The city that was built on top of the Solomonic ruins, was quite dif-
ferent in character and plan. It was not just an administrativecity but a
chariot city (of up to 450 horses), well fortified, with a water
system for
times of siege, and of course a governor'sresidence (338). In this stratum
(IV) some repairs are visible, and there seems no doubt that it was used
96 THE BIBLICAL ARCHAEOLOGIST (Vol. XXXIII,

over a long period, reaching its height most probably in Ahab's days, as is
hinted in an important inscription of Shalmaneser III. It seems that among
the coalition forces fighting against him in the battle of Qarqar (853 B.C.)
Ahab's chariot force (2000 chariots) was the biggest of the lot!
But even before Ahab's days we know of chariotry in Israel. Zimri for
example (I Kings 16:9) is called "commanderof half his chariots;"it is
possible that that "half" was based in Megiddo. I have already mentioned
when discussing the gates that the beginning of the offsets/insets wall could
be attributed to Jeroboam I or one of the kings that preceded the House
of Omri. It is interesting to note that the Megiddo excavators tended to
ascribe this city to Solomon because of the biblical reference to chariot cities
built by him. On the other hand, it is mentioned nowhere that Megiddo,
Hazor and Gezer were chariot cities. Another interesting fact is that neither
in Hazor nor in Gezer were stables discovered in the Solomonic stratum
(or for that matter in any other strata). Megiddo's greatness as a fortified
city in Ahab's times must have continued - as with Hazor- until its
destruction by Tiglath-Pileser in 733.
During that long period, probably nearer its beginning, the four-cham-
bered gate was built. A new city, quite different in plan, was built on the
ruins of city IV, characterizedmainly by several large buildings with a cen-
tral court in the Assyrian style (as shown by I. Dunayevsky and R. Amiran).
There seems to be no doubt that it must be attributed to the Assyrian con-
quest, when Megiddo became an important military-administrativecenter,
in which the two chambered-gatewas built. City II was an open city, with
a large fort built on top of the wall among its other features. Those who
ascribe this stratum to Josiah may be right.
Thus comes to an end the history of IsraeliteMegiddo and with it, this
short survey. We hope soon to be able to publish the full report of our ex-
cavations with plans and drawings of the pottery and other finds, so that
all scholars can check our conclusions, criticize them, and, we hope,
accept
them. If as a result of our excavations we have deprived Solomon of his
famous stables, we may console ourselves that we have returned to its glory
his real city which was no less magnificent than the cities of the northern
kings of Israel who ruled after him.