Roads and Forts in Northwestern Attica Author(s): Eugene Vanderpool Source: California Studies in Classical Antiquity, Vol. 11 (1978), pp. 227-245 Published by: University of California Press Stable URL: http://www.jstor.org/stable/25010733 . Accessed: 08/12/2014 16:03
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].
.
University of California Press is collaborating with JSTOR to digitize, preserve and extend access to California Studies in Classical Antiquity.
http://www.jstor.org
This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions EUGENE VANDERPOOL
Hp6cKEiTatTij; 3COpacfipCl O6pj esydXa, KaIKcovTa iffi Tiv Botioiav, 6S' )v Eiei TV Xcpav ooao60t Cs vai E KCai IpodavTEtS
Xenophon, Memorabilia 3. 5.25
Roads and Forts in Northwestern Attica
In recent years I have done a good deal of walking, accompanied by various members of the American School of Classical Studies, in the mountainous country of northwestern Attica between the upland plains of Mazi and Skourta and the coastal plain of Eleusis.' The peaks in this region, which are covered with a forest of pine, rise to heights of over seven hundred meters above sea level, their sides are steep and often precipitous, and they are separated by deep valleys in which flow the two streams, the Kokkini and the Sarandapotamos, which unite to form the Eleusinian Kephissos just before they emerge from the hills into the coastal plain (figs. 1 and 2). This country is almost entirely without roads except for a track that runs along the Sarandapotamous valley and a rough road that has been carved out of the mountainside and runs from the village of Krora (now Stephani) on the Skourta plain down through the hamlet of Kok kini to the plain of Eleusis. These roads are quite unimproved and are passable only in dry weather with a sturdy vehicle.2 Apart from these roads, which are very little used, there are only paths through the region. Two of these paths, however, are rather special ones. Carefuly laid out, with gentle grades, with retaining walls on the downhill side, and with well-engineered zig-zags on the steeper slopes, they had obviously been designed as roads even though no longer used as such. Feeling that they must be ancient, we traced their course with keen interest.
227
This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions 228 Eugene Vanderpool
THEBESS oT TANAGRA
A S P O S
Dervenosalesi o PLATAIA Skourta Tsoukrati Limiko Plain -CT HAER ON PANAKTON Kavasala \ )OINOE \^ oKrora o 0o~~~~~~~~~~~~~~~ "^--j ''^^^ s--1412 Villia 0 S , }/ok Krorc Kokkini PHYLE
Sarandc'potamos R.-S2-.^^ j^.
Ag. Georgios Ag. Vlasis Plakoto
mt. Pateras
ELEUSIS
MEGARA
ATHENS SALAMIS o
A. C.
Figure 1. General Map of Western Attica.
THE OINOE ROAD
A short stretch of this road, about five hundred meters long, is shown in red and marked "Strassenspuren" on Sheet XXV of the Karten von At tika just east of the saddle between hills 532 and 513. Captain Winterberger, one of the officers who did the survey for this sheet of the Karten, mentioned it briefly in a lecture but did not consider the traces sufficient to belong to an important road. They are also mentioned
This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions Roads and Forts inNorthwestern Attica 229
OSiOs5 0 KAKO NISTIRI SKOUR TA PLA /N MELET MEL E T OS ?530 ?O
PACONAKTOUR INTERVASAL <
SPRING2.
680
i fL (o 1th e A W ~Wi C~ 1970s ^^^9~~/\-^--~if r LA/ OFELEU~SI~S~i.
CONTOUR INTERVAL I01 M. OF ELEUSIS WWj.19.70_PLAIN Figure 2. Map of theHill Country of Northwestern Attica.
briefly (on the basis of the map reference and not autopsy) by A. Milchhoefer and L. H. Chandler, and by I. SarrisSarris who adds that there is another section of the road north of Agios Viasis.3Vlasis.3 Repeated visits to the area have shown that this road can in fact be traced for a distance of about six kilometers.
This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions 230 Eugene Vanderpool
A person looking for the road for the first time would do well to follow the modern track up the Sarandapotamos valley as far as Agios Vlasis and then strike up the hill to the north for a few minutes until he comes to the ancient road which will be clearly recognizable by its con spicuous retaining wall of rough stones, here averaging a meter high, running across a fairly open area at the boundary between the olive groves and the pine trees (pl. 1:1). From here, if he wishes, he may easi ly follow it back eastwards for nearly a kilometer to the point where it first appears. At this easternmost point the retaining wall of the road consists only of a single row of rough stones which might easily be mistaken for field terracing did it not fall exactly in line with the road we have been following. Returning to the point above Agios Vlasis, we follow the road westward (pl. 1:2). It rises along the flank of the hill at a gentle gradient and we soon come to the best-preserved stretch where the retaining wall is as much as two meters high (pl. 2:1). The road itself may have been about four or five meters wide, but there is no road surface preserved, the fill behind the retaining wall at this point having mostly been wash ed away. About a kilometer west of Agios Vlasis we come to a zig-zag, and the road, having gained altitude, soon crosses a saddle leaving the low hill 345 to the left. The retaining wall, which has hitherto been on our left now appears on the right. Near the center of the road bed at one point are clear marks of wheel ruts 1.50 meters from center to center. The road now climbs steeply up the flank of hill 513 in a series of short zig-zags. This section has been much damaged recently when an under ground pipeline or cable was laid, but fortunately the actual turns are clearly preserved on either side of the disturbance. The road now runs westward along the flank of hill 513 with the retaining wall again on the left, making for the saddle between hills 513 and 532 (pl. 2:2). This is the stretch noted on the Karten von Attika. It has been rather badly damaged recently, apparently by the people who laid the pipeline or cable. Emerging on the saddle, traces of the road are harder to discover as no retaining wall, or only a small one consisting of a single row of stones was necessary. Enough is preserved, however, so that with a sharp eye we can follow its line, and at one point wheel ruts may be seen in the rock. It goes westward on fairly level ground until the hill begins to drop off. Then it turns sharply to the right and descends in a series of irregular zig-zags. The modern path follows a more direct course. On the lower slopes we gradually lose the traces of the road until on
This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions PLATES
This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions Plate 1 Vanderpool
0~~~~~~~~~~~~~~~~~~~~~~~ * _
INO Air-
,,i 1. Thet_ ... ..Oinoe ..:, road running along flank of hill west of Agios Vilas'is.
~.,
_!1 w l*
r e 2.:" The ...... '~ , ~ .. *.? :*: '~',~.. . Tei On a I Vat A -;i:i? i,...... ' 2__ 2.The a Oinoe road unn_if lR westhl ak of Agios Vlasis.
This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions Vanderpool Plate 2
A., .
7..**, ?f- ,"~. . * . *
1. Retaining wall of Oinoe road west of Agios Vlasis.
2. Looking SE from Oinoe to hills 513 and 532. The Oinoe road crosses saddle between the two hills.
This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions Plate 3 Vanderpool
-AM
? ,* I ..
...?...... ?..?
2'. zzar :.,, ai;~~~~~~&a A.... . ?~~~~~~~~~~~~~~~~ ""~?. l.~.
,,'* ,:..' ..
1.I. The PanakronPanakton road with Panakton in the distance.
:~~~~~~~~~i?:: '.:::. ::., . , f ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.i?L "??~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:.::' :.. 14 f: .a ? _ ? ~~~~~.?' : ?':'?" ,. .' :. . ":....: . ::4:,~[:' 4: ::..~:,
j~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -~ . . ~.~:' '":~-.'',~:i'~::'" .d~'* ' : z.. ." . , ' . : ': .... zig-zag.2.A the~~~~~~~~~~~~~~~~~~~~~on Paanra. ~.. ..~::~~i::'~..:, ,'.l
:-U???~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~.. !1? . -~...~rlE ~ 2.A zig-zagon the Panakton road.~~~~~""
This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions Vanderpool Plate 4
, ...:.:::d :: r.....:
| ?, 2 l i
1.The ..Pn?": t Ro . ?~~~~~~~~". . . ~ ,. . l 111 _ i a j-tf ; ~ : S:,,:,,:,~~~~~~~~~, ,,,% ::.:, _ | l i' l \,, ? * i : 'rlsl~et,gkla~LPsl, i ' J'ih+#';> N%=~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.~ "~ :f ~~~gE;~~~~~e
, ;~~~ ~ ~ ~~~~.. :~~ ? . . ':_
1hv'. " . i , .'"* ? .~acc~~ " ': ....',':::; ? ?.*'.; ,?.' ~' 1. The Panakton ?~~~~~~~~~~,road. ;C~, ..PR .r:'.? ".) ; ?- . .).:. ~ ....~: ~
.. -,.;.'..rp' s .
?s6~`~6~`~1~ 6~I~?l"~~:. ,a..
"~, ...... ? ..... ?h Panakton r..
2.?tirTe Paaln od
This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions Plate 5 Vanderpool
, ~'~? f: . ~ ~ ~.. i' ;
10
aM~~~~~~~~~~~~~~~~~~~~~~~~" _r;' -,, *
Panakto:ase2. for....a. site'e .i ; .. ~?
..,~ad1 y s~ ! ijd~~~~~~~~d ,.~? d,, f =j= ~., ,, .. u w t. ~ . "' +~~~~~~~. ', "" l. Panakton. Section of wall near main gate'
j ...... ~,.... , -.,..,...... ,*..,. ;~. :,' ..
_~~~: . . i .l II ~~~~~~~~~~~~~~~~* -;Ki _,4~~~~~~~~~~~~~~~~~~~~~~' ...,~
t _.W;XI-?C_ lx --r *1:1L41 '81 ' I . .~...i,_'X, ? B U l _ | _ I ? ?
2. Panakton. B ase for a stele.
This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions Vanderpool Plate 6
'PS.'~..:":g'~
?:t
,...
'"i: ~;ans~?,n~r.. .. ~~~~~~~~~~~?,..:."-.':, ?i.:~.,, " ~2 /d.~,:, .:;.~<~"rd :,~?, ...... '~`,,"' on 1. Oinoe.I~ Tower west ~~~side, trapezoidal limestone. ~~~~~~~~~~~~~~iiJ?:i, .c-~~* m.'~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~r 2...ino . Toe'a'W : .orner, as...r:nglomerate .. An, ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~-?
2.1.Oinoe. Tower atNWon corerts,wes sidne, tashzdlatapezoial climesatoe. cnlomeratoe.
This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions Plate 7 Vanderpool
;......
1. Round tower on hill 532.
__.([i~:~ '_ -~i~iii:iiii.t.x*...... r::!:b:i;'~!:;....~,~^~ii' .0...::!{iir:':.~ ...... ".'.....'...?.. %:,:...,;:,.;..:_ . . ' : ;.::::"'. .':.::''::~'::'~:::,:::3 :... 1 '., ...... ,~.? ~.`~`~..~~~i`~?_~~.?`k.`:.~`~~``~.`.~~~.`.U~.' .... .'...... '.....'~~'~: .....':. ~'':,~ I_...... _ _.''?~:~...~'~.~,:~~i::..." ~" - B ~~~~~ ~ ~ ~:':i"'~~::~....'...:....?.:.."::~~~,."~'~~..:::, .:~'.~..:.:.., '
2. Looking NE from hill 532, Panakton on skyline left of cent. Hill 513 in foreground.foreground.
This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions Vanderpool Plate 8
.:: .: 3....?
_2 . .imio tower from *.
Limik*-. t . . '.:. '.
2. Limiko tower from SW.
"~:-~. ,. '_St..@ _ "
3. Limiko tower from West.
This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions Plate 9 Vanderpool
, i
...
~~~~~~~~~~8.'~::
1.Tsoukrati tower from East. t_ f \-, . ..: U_...... ??~~~~. U: ;~:i::. ...:.::: : : * .::: : :: :. .: : ..... :: :.: : : ::: .:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:i::-iii:::?-;:.i:.i::.:?::?:?: : ~i Ia~. : .. :~:: ? .. : . : .: :::: ~ ~ , ~ 1K~ii8 ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i:::; :::..ii.:.: .N *.: _ 2.Tsoukrati tower from ESW. .Tsokrt to e fo l Y. This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions Roads and Forts inNorthwestern Attica 231 reaching cultivated land, it disappears altogether. Here, however, its line is picked up by a modem dirt track which runs across the fields and eventually makes its way to the modern Athens-Thebes highway at the forty-fifth kilometer, near the church of St. George. A branch of this track turns north into the main Mazi plain and passes the ruins at Myoupolis which I believe to belong to the ancient fortified deme of Oinoe as will be explained in the following section. PANAKTON Several years ago I reviewed once more the evidence for the location of Panakton, an Attic fort on the Boeotian frontier, and the related prob lem of the locations of Oinoe and Eleutherai. I came to the conclusion that the old assignment of names to sites was correct, namely that Gyphtokastro was Eleutherai, that Myoupolis was Oinoe and that the fort on the hill at the south edge of the Skourta plain above the village of Kavasala (now called Prasino) was Panakton. This assignment of names occurs for the first time on the French staff map of 1852, the ex istence of the fort above Kavasala having been unknown previously. It is accepted and supported with arguments by Milchhoefer and it is also accepted by Hitzig-Blttmmer, Frazer, Chandler, and many others. It suits nicely all we know about each of these places and seems about as sure as such things can be in the absence of inscriptions.4 Gyphtokastro corresponds exactly with what we know of Eleu therai. It is on the main road from Eleusis to Boeotia, and its territory marches with that of Plataia. It has houses and a fortification wall, both of which are mentioned by Pausanias who saw them in ruins, and the place is located a little above the plain on the slopes of Cithaeron just as he describes it. There are traces of a lower town and also the founda tions of a small temple in the plain below the fort to the east. Just to the west of the fort there is a strong perennial spring, no doubt that noted by Pausanias.5 Myoupolis suits what we know of Oinoe. It is a fortified town or deme site on the border of Attica and Boeotia.6 It shares with Gyphto kastro a small plain which is divided in two by a low ridge just west of the modern village of Mazi (now called Oinoe), and thus corresponds with what we know of the relationship between Oinoe and Eleutherai from a fragment of Euripides's Antiope quoted by Strabo.7 The fortifi cation walls at Myoupolis show two periods of construction, an earlier This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions 232 Eugene Vanderpool trapezoidal style in limestone and a later ashlar style in conglomerate, of which the former was no doubt in existence when the place was un successfully besieged by Archidamos at the beginning of the Pelopon nesian War. Archidamos's invasion route was across the Megarid and over the mountains via the "Road of the Towers,"8 past Eleutherai to Oinoe and thence by our Oinoe Road (or a predecessor or variant) to Eleusis. This is a perfectly good marching route if somewhat circuitous as compared with the direct Megara-Eleusis coastal road; but Archi damos was not in a hurry, in fact he was deliberately stalling in the hope that the Athenians would decide not to fight. Finally, the fort above Kavasala corresponds to what we know of Panakton. It is an important fort on the boundary between Attica and Boeotia, large enough and important enough to have been exchangeable against Pylos in the negotiations for the Peace of Nicias.9 It is not far from Parnes and thus suits the evidence of the scholiast on Plato, Kritias, 110e, which says that Parnes is a mountain between Attica and Boeotia, near Panakton.10 It goes nicely with Phyle as part of the Eleusis command. The fort dominates the Skourta plain which is prob ably the district of Panakton referred to in the old treaty invoked by the Boeotians according to which neither they nor the Athenians would settle the district but would share the use of it or graze it in common.1 The Skourta plain is a bleak upland district about five hundred meters above sea level. It has no outlet to the sea but is drained by sink holes. Large parts of it are marshy inwinter, though the situation is now much improved by drainage works. Other parts were wooded and many oak trees grew there in the seventeenth century.13 The area was uninhabited in antiquity, at least no ancient settlements have been reported to my knowledge, with the possible exception of some ruins at the church of St. George at the eastern edge of the village of Dervenosalesi (now Pyli) where a large squared block is to be seen and also traces of a rough enclosure wall north and east of the church.'4 The plain now supports five small villages known collectively as the Dervenochoria. Another district near Panakton was known as the Drymos, for we read inDemosthenes'5 that the Athenians were going out in arms to de fend the Drymos and the district of Panakton. "Drymos" is defined by Hesychius in general as a forest or wooded area, and specifically as a locality and fort in Attica. Harpokration describes it as a city between Boeotia and Attica. The Drymos appears to have been cultivated to a small extent in the fourth century B.C., perhaps by or on behalf of the Athenian garrison at Panakton, for we read in an inscription from This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions Roads and Forts inNorthwestern Attica 233 Eleusis16 that in 329/8 B.C. the general Philon offered to the goddesses as first fruits from the Drymos a medimnus and two choinikes of barley and two medimnoi, five hemiektea and a choinix of wheat, which is a very modest amount. The Drymos may have been either some part of the Skourta plain, if the word is to be taken as referring specifically to an oak forest, or perhaps the whole great area wooded with pine bet ween the Skourta and Mazi plains and the plain of Eleusis if it is to be taken in a general sense simply as a forest. In the latter area there are occasional cultivated patches as around Kokkini and Agios Vlasis. These identifications, reasonable as they seem, are no longer generally accepted, however. They were challenged by K. J. Beloch who argued that Gyphtokastro could not be Eleutherai but must be Panakton. The location in a lonely mountain valley was unsuitable for a town such as Eleutherai but ideal for a fort controlling a pass, and the apparent lack of houses in the fort or below it (as he thought) seemed to bear this out. Beloch placed Eleutherai at Myoupolis and moved Oinoe down to the north edge of the plain of Eleusis below the fort at Plakoto. As for the fort above Kavasala, it was too far out of the way, he thought, to be the important fort of Panakton. 1 Beloch was followed by U. Kahrstedt who elaborates some of the arguments in his study of the land frontier of Athens.18 W. Wrede, however, soon introduced an im portant variant. According to him the fort at Gyphtokastro should be Panakton but the town below it should be Eleutherai; Oinoe should return to Myoupolis.'1 This view of Wrede's is now widely accepted.20 Yet it is open to serious objections. A double identification such as this is unlikely enough in itself, but to have the fort Panakton right above the town of Eleutherai is in direct contradiction to Pausanias (1.38.9) who does not distinguish between fort and town but calls both Eleutherai: 'EXtEu0EPDv5e 7iv gi/v TItTOi TEsiXoVS,IV 'sKaic oilcov ipeita' 8Xil| Ue To6Totqgczti n6Xt oXiyov OREOpTzo eb6ioUnpbg ZT Ktiatp6ivt oiKctieiaa. But perhaps the most serious objection to this current theory is that it deliberately downgrades the fort at Kavasala and makes it seem small, remote and unimportant. This might pass unchallenged in the old days when the Skourta plain lacked road connections and was inac cessible so that hardly anyone ever visited the Kavasala fort whereas every traveller to Greece passed along the modern highway from Athens to Thebes and could see for himself the impressive walls of Gyphto kastro towering above the road. The worst offender in this respect is Kahrstedt, who writes of the fort at Kavasala:21 This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions 234 Eugene Vanderpool Es handelt sich um eine recht rohe, die Kuppe umziehende Steinsetzung von gegen 300 m. Umfang (vgl. die Skizze bei Chandler, J.H.S. XLVI, 1926, p. 6, fig. 2) mit einem isolierten Mauerzug weiter abwarts den nur ein sehr guter Wille als Befesti gung einer Unterstadt bezeichnen kann. Die antike Siedlung an der Stelle ist durch die herumliegenden Scherben sicher, aber von einer Festung, die auf den namen Panakton Anspruch erheben kann, ist keine Rede. Vor allem kann das kleine Anlage, wenn sie iiberhaupt militarischen Charakter hat, ebenso gut boiotisch gewesen sein. He later22 identifies the place as the Boeotian village of Skarphai. This description and estimate could hardly be wider of the mark. To be sure, the Chandler sketch to which Kahrstedt refers is inade quate, but Kahrstedt had obviously visited the site himself and should have seen that the ancient circuit was complete in spite of some gaps and in spite of being badly damaged in places, and that "the isolated run of wall which only the greatest good will could call a fortification of the lower town" was an integral part of the whole and contained a massive gate, the main entrance to the whole fort. The fort is in fact a large and important one. Though smaller than Gyphtokastro, it is com parable in size to Myoupolis, and much larger than Phyle.23 I offer a sketch plan of it and also one of Myoupolis (fig. 3) accompanied by a few photographs (pls. 5:1, 6:1 and 2). Both sites deserve to be further studied and excavated. A stele base (pl. 5:2) in the ruins of a chapel near the center of the Kavasala fort holds out the hope that inscriptions may one day be found; they are so far almost entirely lacking in the whole northwest frontier area. Having reestablished the fort at Kavasala as a large and important one, we must also show that it is Attic, not Boeotian, before we can restore to it its proper name, Panakton. The identification with the in significant Boeotian village of Skarphai, which is usually associated with the Boeotian Asopos and so should be down in the plain, can be dismissed out of hand,24 and there is no evidence (pace Kahrstedt, p. 18) that the Boeotians as a whole, or Thebes or Tanagra or any other Boeotian town ever maintained a fort such as this on the Attic frontier. On the other hand, if it is to be Attic it should have a good connection with Eleusis and not be isolated. The tangled mountain country, inter sected by deep valleys, that one sees as one looks from the Kavasala fort in the direction of Eleusis is forbidding and seems to make communi cation difficult if not impossible. We hesitated a long time before This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions MEDIEVAL " TOWER 7181 / \ I \ t Y II \ \ ' PANAKTON 50 ' loM. OINOE ...\ I 1I I i 3 I II: \ WWC 1970 Figure 3. Plans of the Panakton and Oinoe Forts. This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions 236 Eugene Vanderpool deciding to plunge down into it. But our fears were groundless. The path we picked as the most likely course proved to be a well-engineered road with retaining walls and zig-zags just like the Oinoe road but even better preserved and more dramatic. This led us easily and surely through the hills, across a valley, over a saddle and down into the Kok kini valley and so out to the plain of Eleusis. THE PANAKTONROAD As one looks from the Kavasala fort in the direction of Eleusis one sees two good paths, one leading southwest, the other southeast. The former, after crossing several broad-backed ridges, drops down steeply into the Kokkini valley and turns upstream towards the Mazi plain and Oinoe. The latter goes more in the direction of Eleusis, and as soon as we set out along it we realized we were on a road like the Oinoe road (pls. 3:1 and 2, 4:1 and 2). Starting at a saddle above the village of Kavasala and a little east of the fort, it runs along the south flank of a hill supported by a retaining wall of rough stones. After about five hundred meters, however, it is joined by a dirt road coming in from the Skourta plain via another saddle a little further east. The dirt road, suitable only for tractors and carts, has recently been "improved" and widened by a bulldozer. It follows almost exactly the course of the old road for about two kilometers and has obliterated it except where the old retaining walls are still visible. At several points along the way, however, the courses of the two roads differ slightly and the old road can then be seen clearly. The best example of this is at a point where the footpath cuts steeply down through a gully, the modern road swings out and back in a wide loop, and the old road follows an intermediate course, making a smaller loop. About half a kilometer farther on we cross a low divide and the modern road swings off to the right towards some upland fields and pastures and the modern path bears left and drops steeply down along the flank of the hill. The old road continues straight, running along the flank of the hill and dropping gently. It is supported on the left by a re taining wall built of small unworked stones and iswide enough for three people to walk abreast comfortably. Its average width may be two and a half to three meters, but occasional outcroppings of native rock narrow it somewhat and would seem to eliminate the possibility that the road might have been used for wagons; I have not seen any wheel ruts on this road. It continues in this fashion for a distance of about seven hundred This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions Roads and Forts inNorthwestern Attica 237 meters, dropping about sixty meters the while. At this point the flank of the hill which we have been following ends abruptly in a steep cliff, and the road doubles sharply back and descends in a series of six zig-zags, always supported by retaining walls, to the bottom of the valley forty five meters lower down. Crossing the stream and a small tributary that comes in from the left, the road makes two zig-zags then swings out in a wide curve around the west flank of a hill, climbing gently but steadily until it reaches a saddle. This saddle is part of a ridge that runs westward for about a kilometer and a half. About midway along the ridge is a rocky outcrop (520 m.) on which is located the small rubble fort called Korynos, first described by Sarris.25 On the way out along this ridge to the fort there is a rough road, passable for vehicles, with a bit of terracing along its downhill side and perhaps, therefore, a branch of the road we have been following which leads out to the little wayside fort. Just before reaching the fort, the ridge dips sharply and forms another saddle over which passes another north-south path noted by Sarris (Kiapha Delera). This other path is unimproved and lacks retaining walls. Returning to the saddle where we left our main road, we start the long descent into the Kokkini valley. The situation is obscured at the start by two modern roads, an older one and a more recent one. Both come from Krora on the Skourta plain and continue on down through Kokkini, but here at the saddle they criss-cross and then take quite dif ferent courses. The more recent road descends the slope in a wide sweep to the west. The older modern road seems to follow the ancient road; at least we can see several stretches of the rubble retaining wall which we have taken to be characteristic of the ancient road. This can be traced at intervals for a kilometer or so until we come to a flowing spring, called Asprovrysi on Sarris' map, and some farm houses. From here on down the slope is cultivated, chiefly with olives, and there is much ar tificial terracing for the fields, making it impossible to sort out ancient from modern. We may suppose, however, that the modern road con tinues to follow the ancient way down through the hamlet of Kokkini where there is another flowing spring, a chapel, and some scattered farm houses,26 and so to the valley and on out to the plain of Eleusis. DATE AND PURPOSE OF THE ROADS Thus far we have simply assumed that the two roads we have been describing are ancient. It must be admitted, however, that there is no This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions 238 Eugene Vanderpool direct proof of this. The retaining walls that support the roads are very roughly built of unworked stones and have no stylistic character by which one might be able to date them, even approximately. The chance of finding any pottery or coins in the filling of the roads, if one were to dig a section of them, is rather remote. At several places on the Panakton road where the retaining wall had collapsed recently or the road had been cut through inmodern times we were able to see that the filling consisted of shale-like rock from the neighboring hillside and was completely without artifacts. I have the impression that one could dig away meters and meters of road without finding anything that would help to date it. On the Oinoe road most of the filling seems to have been washed away, but where it exists it seems to consist of broken stone and gravel. One stylistic feature may, however, help us in a negative way. There is no stone cobbling anywhere along our roads. Cobbling is characteristic of roads of the Turkish period, and lack of it here gives an indication that our roads are not of Turkish times. There is, of course, in any case no real liklihood that the roads are Turkish. Captain Winterberger, who surveyed the area for the Karten von Attika in 1891, noted only a short section of the Oinoe road which shows that itwas not in use as a road in his day and had not been so used for a long time previously.27 The few early travellers who pass this way fail to notice it at all because the path they are using takes a slightly different course, following the valley of the Sarandapotamos and passing south and then west of hill 532. This is the route shown in the French map of 1852 cited above (n. 4) and also the route followed (in the reverse direction) by Leake in 1806 when going from Hosios Meletios to Athens.28 He notes on reaching the Sarandapotamos that "on the summit of the mountain on the left, one mile distant, may be perceived the lower part of a Hellenic round tower." This is clearly the tower on hill 532 which is oc casionally mentioned or shown on maps but has never been described. I offer a photograph here (pl. 7:1). The tower commands an extensive view of the whole northwest Attic territory (pl. 7:2). It seems clear then that what we have called the Oinoe road is not modern or Turkish. The same arguments hold true for what we have called the Panak ton road. It is unknown to modern investigators, to the German map makers, and even to Sarris29who discovered the Korynos fort but who took the slightly more westerly path called Kiapha Delera, closer to the fort, as "the most direct route from the Thriasian plain and the lower valley of the Eleusinian Kephissos to the ancient fort of Panakton and This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions Roads and Forts inNorthwestern Attica 239 the Skourta plain." Travellers who pass through the Skourta plain on their way from Attica to Boeotia do not come this way at all but cross the Parnes-Cithaeron range either via Phyle or via Hosios Meletios. A modem path, to be sure, follows the line of the Panakton road for much of its length, but itmakes no use of the road's special feature, the zig zags, which lessen the grade. The path disregards these and cuts more steeply up or down the slope, as does the path that follows the Oinoe road. We conclude then that the Panakton road was not made in modern or Turkish times. Other periods are less easy to eliminate, but the fact is that the on ly period that comes seriously into consideration as a date for our roads isClassical antiquity. These roads are major construction projects. They were not built by the local residents, the shepherds, the resin gatherers, the charcoal burners, the few farmers of the district. They must have been built by the state and built to fill a special need. They lead from the plain of Eleusis (and hence from Athens) through the hills to two major ancient forts on or near the frontier with Boeotia, Oinoe and Panakton. They are undoubtedly military roads, built to assure quick and easy movement of troops and supplies between the center and the border forts. Furthermore, there is no evidence that our roads con tinued on to Boeotia; at least I know of no such well-built roads, but on ly paths on the north side of the Parnes-Cithaeron range. Even the im portant "Road of the Towers" (above, n. 8), the main road from the Peloponnesus to Central Greece, is a second class road, a mere path, by comparison.3s Our roads are Attic roads, not "international" highways. They lead to areas where there were important Attic forts and do not go beyond. They must therefore have been built when the defense of these forts was important, namely in Classical and Hellenistic times. Both Oinoe and Panakton figure prominently in the Peloponnesian War, and the forts no doubt existed for some time before that. They also figure in the history of the fourth and third centuries B.C., after which we hear no more of them. The history of the roads probably parallels that of the forts, and we may assume that they existed and were used through the Classical and Hellenistic periods. The Oinoe road with its gentle grades, wide turns, and generous width was suitable for wagons, and indeed we have noted wheel ruts at several points along its course. It was probably used by civilian as well as military traffic, being a section of one of the easiest and most direct routes to Boeotia. The Panakton road, on the other hand, was probably more strictly military, its course being less advantageous as a through This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions 240 Eugene Vanderpool route and serving poorer country. Its grades are steeper and longer, its turns sharper and its width less than the Oinoe road, and wheeled traf fic seems to be excluded. Two SIGNALTOWERS I add a few words about two signal towers on the northern frontier of Attica, at Limiko and Tsoukrati. Both towers are known; they have been mentioned in most discussions of Attica border forts and their position has been marked on maps but they have never been described. They appear to be twins, built to the same plan and specifications, and they have some unusual features. The Limiko tower is almost perfectly preserved, the Tsoukrati tower is half ruined. The Limiko tower stands on a low, broad-topped eminence three and a half kilometers as the crow flies NNW of the highest peak of Parnes and at a very much lower level (altitude 814 m.), being about at the bottom of the fir-tree zone (pl. 8:1, 2 and 3). It commands a wide view of the hill country north of Parnes with glimpses of the Boeotian plain below. Panakton is visible W by S and Tsoukrati W by N. The tower is eight meters square and rises to its full height of 3.50 meters in six courses of masonry. In addition, there is a distinct euthynteria course at the bottom which projects out a few centimeters beyond the face of the wall above it. The masonry is isodomic, mostly ashlar but with a few trapezoidal joints. There is drafting at the corners. The blocks have quarry faces except in a few cases where there is a broached face. The interior arrangement is peculiar. At the NE corner there is a small room about 2.45 m. square at the level of the top of the second course. It is entered through a door near the north end of the east side. This room communicates with another room at the SE corner at the level of the top of the fourth course. These two rooms together occupy the eastern half of the tower. The western half is solidly packed with rubble to the level of the top of the sixth course and forms a sort of plat form (fig. 4). The tower never rose any higher. There are only three blocks missing from the top course and they lie on the ground below. The Tsoukrati tower stands on a peak (altitude 736 m.) at the eastern end of the line of hills that bounds the Skourta plain at the north (pl. 9:1 and 2). It commands a wide view in all directions. It is eight meters square and rises to a maximum height of 2.60 meters in five courses at the north and east sides. Pry holes show that there was This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions Roads and Forts inNorthwestern Attica 241 Limiko Tower Solid packing Level of top of sixth course Level of top Level of top of fourth course of second course East face, elevation N ; 1--i 0 2m. Figure 4. Plan and Elevation of the Limiko Tower. once a sixth course. The west side is preserved to a height of four courses at the south, but the north west corner and the whole north side are gone and only the dressed bedding for the blocks remains. The interior arrangement is not very clear because of the ruinous condition of the tower, but it seems to have been like the Limiko tower. The west half was certainly filled with rubble packing to form a plat form. The eastern half had a room at the south entered by a door in the south wall but the exact dimensions are not clear. This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions 242 Eugene Vanderpool These towers are to be thought of as Attic outposts, manned by a few soldiers and designed to signal the approach of danger by a fire lighted on top of the platform or by a torch or series of torches displayed there. POSTSCRIPT: THE GYPHTOKASTRO FORT When the manuscript of this article was virtually complete, I received from Luigi Beschi a copy of his article "La Fortezza Ellenica di Gyphto kastro," reprinted from the Acts of the VIII Scientific Congress of the International Castles Institute, "Les fortifications depuis l'antiquite jus qu'au Moyen-Age dans lemonde mediterraneen" held at Athens 25-29 April 1968. Beschi, without excavating, has made a careful study of the visible remains of the fort and reaches the following conclusions. The earliest phase, represented by the big rectangular building, 16.40 m. x 11.10 m., on the top of the hill, which had been dated in the late sixth to the mid-fifth century B.C., is to be eliminated. The style of the masonry and the drafted corners seen in this building cannot be earlier than the fourth century B.C. It therefore dates from the same period as the main fort and, judging by the plan of the interior rooms, is perhaps to be thought of as a dining room or as a barrack. The main fort shows two construction periods, the north wall with the imposing line of towers be ing of the second half of the fourth century B.C., the southern wall with its use of conglomerate being a repair perhaps of the third century B.C. Thus, according to this analysis, we are left without any structural remains of the fifth century B.C. This becomes embarrassing when the question of identification arises, for Beschi, following the current modern view, wants the Gyphtokastro fort to be Panakton. Yet how can this be if we find no fort there of the fifth century? "Our sources are silent on this matter" he says (p. 17) "but if Thebes, on orders from Sparta to give up the fort, hands over an area of ruins levelled to the ground, we cannot exclude an Athenian occupation of the site precisely during the first phase of the Peloponnesian War." Surely, however, the unprejudiced observer would conclude that if there are no remains of a fifth century fort, Gyphtokastro simply could not be Panakton. If on the other hand we allow the fort at Gyphtokastro to be Eleutherai, then Beschi's observation that there are no remains there earlier than the late fourth century B.C. makes it easier to understand why we have no men tion in our sources of a fort at Eleutherai. This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions Roads and Forts inNorthwestern Attica 243 Beschi's treatment of the Kavasala fort (p. 18) takes its cue from Kahrstedt. 'The sparse remains of walls which encircle an area of reduc ed size on top of a mountain near Kavasala do not seem to correspond to the importance indicated by our sources for Panakton. The castle of Kavasala is excellently situated as an observation point . . . but as a road control point it is not so well located: it duplicates at a short distance the function of Phyle. .. ." I have already commented on the erroneous estimate of size. Here I would simply add that Kavasala in no sense duplicates Phyle, nor is the distance between them short, be ing about three hours on foot. American School of Classical Studies Athens NOTES 'The reconnaisance was done in the late 1960s and the text was written in 1970. In preparing the article for publication I have been actively assisted byW. Willson Cummer III, Frederick A. Cooper and Merle K. Langdon who took part more than once in the reconnaisance and helped with measurements and checking details. Mr. Cummer drew two of the maps, and Mrs. Abby Camp the other two. 2Another road has recently been built from the Mazi plain following the Kokkini river downstream a few kilometers to the point where drilling has commenced at the outlet of the Cithaeron tunnel of theMornos aqueduct. This roadwill no doubt eventually be con tinued down the Kokkini valley to the plain of Eleusis when the corresponding section of the aqueduct is built. 3E. Curtius and J. A. Kaupert, Karten von Attika, Erliuternder Text von Arthur Milchhoefer, Heft VII-VIII, 18. Captain Winterberger, AA, 1892, 123. L. H. Chandler, JHS, 46, (1926) 15. I. Sarris, Arch. Eph. 1927-1928, 118. 'Carte de la Grece, 1:200,000, Public par le Depot de la Guerre (Paris 1852) sheet 8. A. Milchhoefer, text to the Karten von Attika (supra n. 3), VII-VIII, 15-17 and IX, 35-37. H. Hitzig and H. Bliumner, Des Pausanias Beschreibung von Griechenland I, 280 and 358. J. G. Frazer, Pausanias's Description of Greece II, 325-326 and 515-518, and V, 513 and 537-538. Chandler (supra n. 3) 6-12. SPausanias, 1.38.8-9; 9.1.1 and 6. Strabo, 9.2.31. Xenophon, Hellenica, 5.4.14. Lucian, Dialogues of the Dead, 27.2. bThucydides, 2.18 and 8.98.2. Cf. Herodotos, 5.74. 'Strabo, 8.6.16. Euripides, fr. 179 (Nauck). Cf. also Harpokration s.v. Oinoe. The Strabo passage, and particularly the Euripides quotation it contains, is given in slightly fuller form in the Vatican palimpsest of Strabo: Wolfgangus Aly, Franciscus Sbordone, de Strabonis codice rescripto (Vatican City 1956) 16, 212 and 281. I quote it here as it has not yet appeared in the topographical literature. 'eovord6?TO(Aegina insula) 6' Oivcbvn rd~.ai, 6$ov6uo; Puai flaotq Tfi 'ATTIIKS, TO Ie npOb 'EXeuOepaiqS' EXEIV onr6 uot 6toftqS eor76trI 6' o6 Oiv6r; o67yXopTavaiet neSia Taioa' 'EXeu09paig This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions 244 Eugene Vanderpool See also Bruno Snell, Scenes from Greek Drama, 71-72, and Jean Kambitsis, L'Antiope d'Euripide, edition commentee des fragments (Ath6nes 1972) p. 1 Fragment I, and commentary on pp. 21-34. 8Described by N. G. L. Hammond, BSA, 49 (1954) 103-122, and in Ham mond's book Studies in Greek History, 417-446. The towers which give the road its name are not military; they belong to farms. J. H. Young, Hesperia 25 (1956) 122-146, esp. 145. 9Thucydides, 5.3.5, 5.18.7, 5.35.5-6, 5.36.2, 5.39.2 and 5.44.3. ?1I. L. Merker has called attention to this scholion: cf. AJA 69 (1965) 171. 1'IG II2, 1299, 1303, 1304, 1305, 2971. Oinoe, which is not mentioned in this connection, was no doubt a separate command. 'Thucydides, 5.42.1. '3George Wheler, A Journey into Greece, London 1682, p. 334 reports that on January 26, 1676, "After Dinner we passed over a Plain, for two hours and a half; and by the way, I observed many Ponds up and down in it, with plenty of Wild Ducks, and Teal in them; one or two of which we shot. Here are also many of those Oaks I before described at Troy, with great Acorns they gather Vellania from." For a description of the plain see Alfred Philippson, Die griechischen Landschaften I, 2, pp. 531-533. "4H. N. Ulrichs, Reisen und Forschungen in Griechenland II (Berlin 1863) 72 noted here several ancient blocks and also a small relief of white marble with a representation of Trophonios as Chthonic Hermes conducting a soul to Hades. He also reported seeing a long stretch of the ancient military road from Athens to Thebes via Phyle running straight across the Skourta plain. 'De falsa legatione, 326. '1G, II2, 1672, lines 271-272. '7Klio 11, (1911) 436-439. Beloch attributes the basic idea to Otfried Mfiller who, on his map of Boeotia in his book Orchomenos (1820), marked Gyphtokastro as Panakton without argument.Whether the idea really originated with Miller I do not know, but it did have some currency in the first half of the nineteenth century for K. G. Fiedler unequivocally identifies Gyphtokastro as Panakton (Reise durch ... Griechenland [Leipzig 1840] 1, 87), and Ludwig Ross is at pains to deny the identification (Wanderungen in Griechenland [Halle 18511 14-15). But these early opinions carry little weight as their proponents did not know of the existence of the fort above Kavasala, an essential factor in the argument. It is perhaps worth recalling that it was in this same Klio article that Beloch launched another topographical heresy, that which would make St. George rather than Lipsokoutali the island of Psyttaleia off Salamis. '8AM57 (1932) 8-28. '"Walther Wrede, Attische Mauern (1933) 32-33. 20E.Stikas, Praktika (1939) 44-52; Alfred Philippson, Die griechischen Land schaften (Nebst einem Anhang von Ernst Kirsten) I, 2, (1951) 525; J. Wiesner in R.E., s.vv. "Panakton" (1949) and "Oinoe" (Suppl. 8, 1956); E. Kirsten inR.E., s.v. "Plataia," col. 2260 (1950); Ernst Meyer in Kleine Pauly, s.vv. "Eleutherai" (1967) and "Panakton" (1970); W. K. Pritchett, AJA 61 (1957) 17, n. 45; L. Beschi (see the "Postscript" to this article, p. 242); and no doubt others. It is accepted, but hesitatingly, by A. W. Gomme, Historical Commentary on Thucydides III, 632-633. N. G. L. Hammond has a combination all his own. He leaves Panakton at Kavasala and Eleutherai at Gyphtokastro (History of Greece. fig. 13), but placed Oinoe at the modern village of Villia where no ancient remains have ever been reported, let alone a large fifth century fort (BSA, 49 119541120-122 and Studies in Greek History. 444-446). I have not discovered what he would call the important fortified town site at Myoupolis; the map in his History, fig. 19, simply writes the modern name Mazi in the general area. 2AM 57 (1932) 17-18. 22AM 57 (1932) 27. 23Size, of course, is not an absolute criterion of importance, as Polybios observes (IX.26a). Recall too Thucydides' remarks in his classical comparison of Athens and Spar ta (1.10.2-3). The terrain likewise often dictates the course and extent of a circuit. This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions Roads and Forts inNorthwestern Attica 245 24Thiswas Kahrstedt's suggestion,AM 57 (1932) 27. On Skarphai see further, S. N. Koumanoudis, RPh 35 (1961) 99-105. 25ArchEph (1927-1928) 112-119. J. R. McCredie, Fortified Military Camps in Attica (Hesperia Suppl. 11) 83-85. The fort above Kavasala cannot be seen from Korynos, pace McCredie 84. In the sherd collection of the American School there are a few sherds picked up at the site. One is a black-glazed fragment from the wall of a closed pot. The others are nondescript but clearly ancient Greek. This fort is best thought of as an Athenian outpost of the years 422-420 B.C. when the Boeotians held Panakton. The scarcity of sherds and the rough, makeshift construction indicate a short period of occupation. '6Outside the chapel at Kokkini I noticed a stone base of the kind used for wooden columns 0.50 m. in diameter and a block with a picked face. Milchhoefer (text toKarten vonAttika VII, p. 18) reported several ancient blocks and some fragments of marble. There was no doubt a small ancient settlement here, whether a way station on the road to Panakton, some farms, or one of the many small demes of the coastal trittys of Hippothontis it is not possible to say. 27CaptainWinterberger considered the road ancient, but this observation is of course not decisive for, excellent map makers though theywere, theGerman officers who surveyed for the Karten von Attika were no archaeologists, and things marked in red (i.e., as ancient) on their maps are often ruins of much later date. This caveat is sometimes overlooked. It has been emphasized by W. Wrede, Attika, 30, 8"W. M. Leake, Travels in Northern Greece, II, pp. 379-380. 29ArchEph (1927-1928) 114. 30BSA 49 (1954) 103-120. It lacks retaining walls for most of its length though it runs along a steepmountain flank. The only place I have observedwell-engineered zig-zags with retaining walls is on the stretch where the road leaves the Small Vathichori in the direction of the Large Vathichori. This content downloaded from 137.22.1.233 on Mon, 8 Dec 2014 16:03:34 PM All use subject to JSTOR Terms and Conditions