Some Churches and Monasteries on the Southern Shore of the Sea of Marmara

Author(s): Cyril Mango and Ihor Ševčenko
Source: Dumbarton Oaks Papers, Vol. 27 (1973), pp. 235-277
Published by: Dumbarton Oaks, Trustees for Harvard University
Stable URL: http://www.jstor.org/stable/1291343
Accessed: 09-06-2017 10:03 UTC

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 support@jstor.org.

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

Dumbarton Oaks, Trustees for Harvard University is collaborating with JSTOR to digitize, preserve
and extend access to Dumbarton Oaks Papers

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCHES AND MONASTERIES ON THE SOUTHERN SHORE
OF THE SEA OF MARMARA

CYRIL MANGO and IHOR ?EVCENKO

N important group of Byzantine churches fully accessible to inspection: Pelekete is
and monasteries survives on the southern partly obstructed by chicken houses, the
shore of the Sea of Marmara, in the area monastery of Kursunlu and ?ingyan Kale
delimited by the cities of Mudanya to are almost completely overgrown,3 while
the east and Bandirma to the west (fig. 1). Kemerli Kilise was used in 1972 as a store of
This group comprises the church of the onions. We have not surveyed either the
Archangels at Sige (or Syke), three miles Fatih Camii or Medikion, but have found it
west of Mudanya, two churches within the advisable to set down here some brief notes
town of Trilye or Tirilye (now renamed concerning both of them.
Zeytinba'i), namely the Fatih Camii and the
Kemerli Kilise, two monasteries in the
I. TIRILYE AND VICINITY
vicinity of Tirilye, namely those of Medikion
and Pelekete, and a number of lesser knownThe earliest mention of Trigleia as a
sites between the Karadere (Rhyndacus) locality is probably of ca. 800 and concerns
River and Bandirma, such as Malkara fstii, a goovi TptyAEiaS.4 Toward the middle of
Kursunlu, Timanyo, Manastir (St. Anne) the eleventh century Michael Psellos speaks
near Yenice, and Yenice itself. Of these
latter sites, the ruined monastery to the west 3 We are most grateful to the present owner
of the village of Kursunlu is the most impor- of the monastery near Kursunlu, Mehmet K6se
(also known as Uzun Mehmet), who received us
tant and the best preserved. Finally, ruins very cordially and who, aided by members of
of a fortress, possibly of late Byzantine times, his large family, was kind enough to cut down
are still standing at ?ingyan Kale. some of the dense growth so as to expose the
Of these monuments, the church of the ruins of the church. Throughout our work, we
were helped by the ingenuity of our chauffeur,
Archangels has recently been the subject of a
Hasan Bi~kin. We also wish to thank Mr.
monograph;' the two churches at Tirilye Faik Manyas, of the office of the kaymakam at
were published with some illustrative Karacabey, who provided us with a letter of
material by F. W. Hasluck as long ago as introduction to the muhtar of the village of
1906-7,2 while the monasteries of Medikion, Kursunlu. We were greatly assisted in our
field work by Miss Marlia Mundell, who made a
Pelekete, and Kursunlu have been briefly number of independent finds. At home, we very
described without plans or illustrations.2a much profited from observations on the Kur-
The materials presented here were gathered sunlu inscriptions (cf. pp. 250-52 infra) which
in the course of short excursions in 1968, Dr. Thomas Drew-Bear kindly provided at our
1971, and 1972, and concern mainly four of request. As we aimed at a summary publication,
we made only partial use of these observations.
the above monuments, namely the Kemerli We also wish to thank Miss Jo Peterson, who
Kilise at Tirilye, the monasteries of Pelekete prepared the map for us, and Mlle Marie-Louise
and Kursunlu, and ;ingyan Kale. The plans, Concasty of the Biblioth6que Nationale in Paris
measured in 1972, cannot claim complete for information concerning the provenance of
two manuscripts.
accuracy since all four monuments are not
4 See infra, p. 238. The fullest, if somewhat
unreliable, account of Tirilye is by Tryphon
1 IH. Buchwald, The Church of the Archan- Euangelides, Bp'i etov-Tpiylosia (Athens, 1934).
gels in Sige near Mudania, Byzantina Vindo- A few historical data have been collected by
bonensia, IV (Vienna-Cologne-Graz, 1969). W. Tomaschek, Zur historischen Topographie von
2"Bithynica," The Annual of the British Kleinasien im Mittelalter. I. Die Kiistengebiete
School at Athens, 13 (1906-7), 285ff. und die Wege der Kreuzfahrer, Sitzungsberichte
2a On Dutlimanl (Sykamia), see Additional d. K. Akad. d. Wiss., Wien, Phil.-hist Kl., 124,8
Note infra. (1891), 13.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
236 CYRIL MANGO and IHOR ?EV?ENKO
of taking a ship at Trigleia in company with
provisions to the beleaguered city.12 It was
a monk whose conversation was anythingstill in Byzantine hands in 1337 when news
but edifying.5 No further references was
appear
conveyed from there to Constantinople
to exist until the latter part of the thirteenth
that a Turkish expeditionary force was about
century. When the Emperor Andronicus to attack
II the settlements on the northern
resided at Lampsacus (1284), the purveyors
shore of the Propontis.'s Prousa had already
of foodstuffs for his table made forays
fallenasin 1326 and Nicaea in 1331; Nicomedia
far as Cius (Gemlik), Trigleia, and Elegmi
was captured in that same year 1337. It is
(Kursunlu between Mudanya and Gemlik)
not likely that Trigleia, which was unfortified,
and confiscated chickens and piglets would
fromhave held out much longer. Yet, in 1379
the peasants." we encounter a certain Myron, iEpog6vaXoS,
At the same time, with the development
'apXos TplyKXhEaS, at whose behest a
of maritime, especially Genoese, trade,manuscript
Trig- of Harmenopoulos (now Parisi-
leia gained some prominence and, we nus Graecus 1387) was copied by a monk
assume, prosperity as an exporting center of called Dionysios.'4 The title of exarch prob-
wine and alum. The wine traffic is first ably designates here a patriarchal repre-
mentioned in a Genoese document of 12841 sentative charged with the supervision of
and is confirmed by Pegolotti (ca. 1340).8
monasteries.15
Vino Trillie continued to be exported to
Caffa as late as 1381-82,9 by which time1. Fatih Camii
Trigleia was surely in Turkish hands. The The building has not greatly changed
alum was mined at Uluabat (Lopadion) atsince the days of Hasluck, whose sketch-plan
the rate of 10,000 quintals per year and takenwe reproduce (fig. 9).16 During the occupation
by land to Trigleia, whence it was shipped.10 of Tirilye by the Greek army (1920-22) it was
As a result, Trigleia often appears in portu-
converted into a church, and, says Tryphon
lans and on early maps." Euangelides (a native scholar), "after the
It is not known when Trigleia was cap-whitewash had been scraped off the walls,
tured by the Turks. About the year 1317, there appeared wonderful mosaics which I
when Prousa (Bursa) was being besieged by thedeeply regret I did not photograph for lack
Sultan Osman, the co-Emperor Andronicusof film."" Today, traces of mosaic may be
III landed at Trigleia and was able to send seen in the soffit of the south tribelon: it
consists of very large cubes (about 0.02 m.
5 Ep. 97, ed. E. Kurtz and F. Drexl, Michae-
lis Pselli scripta minora, II (Milan, 1941), 125. square), alternately black and white, thus
6 Gregory of Cyprus, ed. S. Eustratiades forming a checkerboard pattern.
in 'EKKwA1aO~.TIK6s &pos, 4 (1909), 13. Cf. V. The photographs we reproduce here (figs.
Laurent, Les regestes des actes du Patriarcat de 4ff.) will give the reader some idea of the
Constantinople, I/4 (Paris, 1971), no. 1474.
7 G. I. Bratianu, Actes des notaires gdnois de present appearance of the building. The four
Pdra et de Cafla (Bucharest, 1927), 172: item columns of the atrium (fig. 7), capped by
Guillelmus de Varagine iperperos centum viginti reused sixth-century capitals (fig. 8), appear
novem quos portavit in Trigia pro vino emendo. to be standing in their original position.
8 F. Balducci Pegolotti, La pratica della
mercatura, ed. A. Evans (Cambridge, Mass., 12 Cantacuzenus, Hist., 1.45, Bonn ed.,
1936), 26.
I, 220. Cf. G. G. Arnakis, 01 rrpS-rot 'OScopavof
g N. Iorga, "Notes et extraits pour servir a (Athens, 1947), 156.
l'histoire des Croisades au XVe siecle," Revue Is Cantacuzenus, Hist., 11.34, Bonn ed., I, 505.
de l'Orient Latin, 4 (1896), 39, where the term 14 H. Omont, Fac-similds des mss grecs da-
"trillia" is incorrectly explained as "trellis," tds de la Bibl. Nat. du IXe au XI Ve s. (Paris,
although Tomaschek, Topographie (as in note 4 1891), 22. For the date see A. Jacob, "Quelques
supra) had already guessed its meaning. problemes de comput," Revue de Philologie, de
10 Pegolotti, op. cit., 369. Cf. W. Heyd, His- Littdrature et d'Histoire anciennes, N. S., 13
toire du commerce du Levant au moyen-dge (1889), 119.
(Leipzig, 1885), II, 566f. 15 See J. Darrouzbs, Recherches sur les
11 K. Kretschmer, Die italienischen Porto-
699qiKla de l'Eglise byzantine (Paris, 1970), 127ff.,
lane des Mittelalters (Berlin, 1909), 651; A. 162f., etc.
Delatte, Les portulans grecs (Liege-Paris, 1947), 16 "Bithynica" (as in note 2 supra), 288, fig. 2.
241, 285.
17 Bp*bE1tov-TpfyEla (as in note 4 supra), 119.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 237

Hasluck shows a molded door-frame inserted tions quite differently. After observing that
into the central intercolumniation.'s The
both of them started with KiOptE ooijSt, he
interior has been covered once again with a
deciphered the northern one as . .. TEgavcg
thick coat of plaster and whitewash which
rarptpKicp, and the southern one as Mavou~1A
hampers a thorough examination of the
... iTarptKcx . Being unable to obtain a
structure. As for the carved ornament, itladder,
is he could not read the second and
easy to see that it belongs to at least two
seventh monograms. However that may be,
distinct periods. The large capitals of Hasluck's
the interpretation raises certain diffi-
columns, similar to those of the atrium, culties.
are First, it is quite abnormal for the
name Niketas to have been divided between
reused (figs. 10, 11); on the other hand, the
imposts placed upon these capitals, the two monograms. Second, the natural way of
cornice that runs around the interior of the reading the inscriptions would have been
church (fig. 13),19 and the capitals of the from left to right in the order west-south-
south tribelon (fig. 14) are Middle Byzantine east-north, while Hasluck goes west-
work, in all probability contemporary with the east-south-north, so that the Nl- of
construction of the church. The most impor- Niketas is to the right of the termination
tant clue as to the date of construction is pro- -KT1Ta. Third, the word aOXhl-rilKS, though
vided by the monograms that were placed on it certainly existed in the Greek language,
the imposts of the two eastern columns, onemakes little sense in this context; nor is it
on each side. These, unfortunately, have been known to us to have been used as a family
destroyed so that we have to depend on name. Being unable to offer an entirely
Hasluck's copy (fig. 12). convincing alternative explanation, we must
Hasluck himself resolved the monograms content ourselves with noting that the
in the following manner: church was built or restored by, presumably,
Northeast column: KiOpIE 3poil?SEt I Trp 8o6h two patricians, one of whom was called
Michael. It is also quite likely that the dedi-
NTIIflTra I rTaTrptKxiC cation was to Christ.
Southeast column: XploTri poilSE1 I a*XA- As to the Byzantine name of the monu-
-TlKx I MiXath I Ta-rpcxKka.
ment, which appears to have been converted
He went on to suggest that Niketas the
Patrician was the saint of that name who into a mosque in the sixteenth or seven-
teenth century,22 there probably did not
was active at the end of the eighth century
exist in the nineteenth century (when our
and the beginning of the ninth. Today, we
know considerably more than Hasluck did documentation begins) any genuine local
tradition. It has been called the church of
about the career of Niketas the Patrician
St. Stephen23 or the pov T -ro XTivoAdxKKou.24
(d. 836),20 and there is no indication that
would connect him with a church or mona- The second is almost certainly incorrect. We
stery at Trigleia. Furthermore, we are notdo not know exactly where the govil T-ro Xr1-
at all convinced that the name Niketas 22 The date remains uncertain. Euangelides,
appears in the monograms. BpiAQeiov-TpiyXEsia, 32, gives it as 1661
(H. 1039), but on pp. 118-19 he reproduces in
It is somewhat disturbing to note Greek
that translation the Ottoman inscription
B. A. Pan'enko21 (in 1910) read the inscrip-
placed over the western door of the mosque
18 "Bithynica," 286, fig. 1 (1). which is dated by a chronogram to H. 969
19 The cornice is of two types. In the nave (= 1562), assuming the latter has been correctly
it has a palmette design, while in the bema read and computed. The inscription is still
and apse it is decorated with little arches extant.
alternating with darts. The latter type was 23 M. Kleonymos and Ch. Papadopoulos, Bi-
extremely common in the fifth-sixth centuries. SVVIK&X (Constantinople, 1867), 43, 149; T.
It is not clear to us whether these are reused
Euangelides, ITEpi -rivcov &PXaclo--rrov pvLavrlvC'v
pieces or a middle Byzantine imitation. iovCAv Iv BtSvviq, in co-r'lp, 12 (1889), 154;
20 See D. Papachryssanthou, "La vie du idem, Bpi5t7iov-TpiyAsia, 32. Hasluck, "Bithy-
patrice Nic6tas," Centre de Recherche d'His- nica," 287; idem, Cyzicus... (Cambridge, 1910),
toire et Civilisation byzantines, Travaux et 59, was told locally that Fatih Camii had been
Mdmoires, 3 (1968), 309ff. dedicated to the Annunciation.
21 Izvestija Russkogo A rheologiceskogo Instituta 24 Euangelides (as in notes 4 and 23 supra).
v Konstantinopole, 15 (1911), 276. Hasluck rightly questioned this identification.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
238 CYRIL MANGO and IHOR SEVCENKO

ascribe the church of Panagia Pantobasilissa
vo7dKK<ou (made famous by the residence
therein of the Patriarch Methodius) was
situated on an eminence in the western part
of the town. Euangelides calls it the mona-
situated, but in the early nineteenth century
it was thought to have been at Mudanya.25 stery of Trigleia,29 but this is merely an
The most natural identification would be uninformed guess. There is no indication
with the povil TpItyXEia whose abbot, St. that this was a monastic church. The earliest
Stephen the Confessor, suffered persecution
reference to it is made by Dr. John Cove]
(1676), who writes as follows:
in the reign of Leo V.21 This monastery must
have been founded, therefore, before the
There are many Turkes there [at
beginning of the ninth century. It does not
Tirilye] and a pretty large Mosche [the
appear among the signatures of one hundred
Fatih Camii]; though ye Greekes exceed
thirty-three abbots and monks who attended
them in number, and have 5 or 6 churches,
the fourth session of the Council of 787, but
miserable places all, except one which
that is not in itself significant since many of
the monasteries in question are designated they call T-awvrapao-iacrRac [sic], it being
dedicated to ye V. M. It is but very
simply by their dedication, not to mention
the obvious fact that not all monasteries small yet it is very pretty; it is antient
yet very intire. The model is something
sent representatives to the Council."'
of Sta Sophia, that is if you leave out
Several early Byzantine pieces may be
seen today in the immediate vicinity of the ye Isle L-X [see fig. 2]. It hath a cupola
built upon ye 4 pillars apy8, the rest is all
Fatih Camii. These include an impost capital,
similar to those of the atrium and nave, that
arch't. It is made of stone and (most)
brick. ye floor hath been finely tessel-
has been hollowed to serve as a trough (fig.
lated: in ye nave remaine many peices,
19), a couple of other capitals, one of them
one pane is very intire where I have here
Corinthian (fig. 18), a sixth-century abacus
placed it, of black and white marble ....
block decorated with leaves (fig. 17), and
There were windores on ye S. side onely,
three identical closure slabs, also of sixth-
century date, one of which has been reused and those onely at E & L. The rp6o9Ects
in the fountain to the west of the mosque -r, &yiov PIiCa s, ( KEUOq6KlOV, XapT'o-
(figs. 15, 16). quixaK'ov X, were all semicupolas and
had lights in the skirts of ye cupola's.
2. Kemerli Kilise (Panagia Pantobasilissa) It had six pillars placed as in ye draught,
gothick work, and round the side were
It is to the period of the commercial
pilasters. On ye outside to the west, are
prosperity of Trigleia, namely the late
on either side ye door a shallow neech
thirteenth century," that we should like to
and over these are arches in ye wall from
25 See C. Mango, "The Monastery of St. pilaster to pilaster, as the same is quite
Abercius at Kursunlu," Dumbarton Oaks Papers, round. Wall at ye east end is built upon
22 (1968), 174 note 31. broken pillars layd flatt-wise, their ends
26 Synaxarium CP, ed. H. Delehaye, 561,12
(March 26); cf. 12,5 (Sept. 3). This may be the jutting out into street; take it in ye
abbot Stephen to whom Theodore of Studios whole and it seem's to me a very pretty
wrote one letter in the reign of Michael II, Ep. model.30
11.139, Patrologia Graeca, 99, col. 1441. The
synaxaria also record an 6crtos M&pKos 6 Tpry7i- Certain details concerning the history
v6S (Nov. 24): Synaxarium CP, 253,52. of the church in the nineteenth century are
27 For what it is worth, we may note the provided by Euangelides. He states that the
presence at the Council of 787 of the abbot of
dome and belfry (no longer extant) were
the monastery 0ro0 AiAl-roG (J. D. Mansi, Sacro-
damaged by the earthquake of November 10,
rum Conciliorum nova et amplissima collectio,
XIII, col. 152C), which recalls Hasluck's 1855. The dome was restored and slightly
a*ATlcrtK6s. altered in shape. The belfry was rebuilt in
28 The same date is advanced by S. Eyice,
who devotes a few words to this church in his 1883, at which time a women's gallery was
"Monuments byzantins anatoliens in6dits ou - lcorip, 12 (1889), 94f.; BpisEtov-TpfyXEtca,
peu connus," Corsi di Cultura sull' Arte Raven- 15ff.
nate e Bizantina, 18 (1971), 317. 30 British Museum, Add. MS 22912, fol. 266r.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 239

added at the west end of the church, in- reused. The two easternmost columns and
creasing the length of the building by 5.20 m. the south column of the second pair have
to a total of 18.60. Euangelides also mentions similar Ionic impost capitals decorated with
old wall paintings that were nearly effaced, upright leaves, and their volutes with a
a small panel of pavement mosaic consisting diaper pattern (fig. 30). These may have come
of black, white, and green cubes, and a from the same sixth-century building as the
wooden templon containing an icon of Sts. abacus now lying outside the Fatih Camii
Michael and Gabriel signed XElpi Fepaai1ou (fig. 17). The northern column of the second
Kprlyr6s ,acpq' (1569), and another of St. pair, its shaft carved with a cross, has a
John the Baptist dated 1767.31 fifth-century acanthus capital (fig. 28). Fi-
Hasluck saw the same "panel of black, nally, the northern column of the third pair
white and grey opus sectile pavement." has a simple Ionic impost capital decorated
He adds: "One of the columns supporting only with crosses (fig. 29).
the dome-the north-eastern-is reputed to The masonry of the church is charac-
be of an artificial and semi-translucent stone: teristic of the last phase of Byzantine archi-
it is really of very ordinary granite. Many tecture. It consists of single courses of brick
'lords' have attempted (in vain) to steal it, (mostly irregular fragments) alternating with
as it contains gold. For this reason, presum- two courses of rubble, the individual stones
ably, I was forbidden to complete my being often separated by upright pieces of
measurements of the interior of the church."'32 brick, somewhat in the manner of the pare-
Hasluck's ground plan is reproduced in ment cloisonn6 that is characteristic of Byzan-
figure 21. tine buildings in Greece.m The decorative
Paneenko describes certain features of treatment of the lateral blind arches of the
the columns that are not visible today
north and south facades is also worthy of
because of the accumulation of earth inside attention: the voussoirs, both stones and
the church. The first pair of columns from bricks, have been bevelled on the inner
the east, he says, is placed not on bases, but circumference. On the north side (but not
on two capitals with chrisms, volutes, and on the south) these arches are outlined by a
acanthus; the second pair is set upon plainer string of sawtooth (fig. 23), a feature very
capitals without volutes, decorated with common in Greece, whereas it is unknown in
acanthus and a rinceau. The third pair had Constantinople.35 The apses have little pro-
plain octagonal plinths, as also the fourth jection on the outside (fig. 26). The central
pair (set up in 1883), which were capped by one has, above the window, three little niches
similar elements instead of capitals.33 decorated with a fan pattern (fig. 25). In the
The present condition of the building is spandrels, between the niches, are brick
illustrated in figure 22ff. While the nine- crosses. The foundation of the main apse
teenth-century addition to the west is now a contains the column shafts that were noted
roofless shell, the original part has suffered by Covel.
comparatively little damage, except for the The deterioration of the plaster on the
southwest column that has fallen down. The walls of the interior has revealed two layers
ground plan (fig. 20), which is fairly ir-
of painting. The upper layer, which is of very
regular, is essentially a cross-in-square, butpoor quality, is dated 1723 by an inscription
has been extended by an extra pair of that runs around an oval medallion con-
columns to the west. Originally, the churchtaining a representation of Christ in a chalice
appears to have had no narthex. Part of itsin the central apse. It reads (the eucharistic
west facade, namely the northern of the two formula is based on Matt. 26:27):
shallow niches noticed by Covel, is still
standing and bears traces of painting, whose
subject it is now impossible to determine repa1'xS
acrou lep(o)tC(ovaxo)s
Tnav-rET Touro eao-rn To ETcO,'xKy
I.OU To ElSTE
Trrj ..
(fig. 24). The columns and capitals are all
KCIIV1S I81CSlQK! V'iTEp flIlOV .EXuVO~eVOV].
31 As in note 29 supra. 34 See G. Millet, L'dcole grecque dans l'archi-
32 "Bithynica," 292. tecture byzantine (Paris, 1916), 225ff.
33 IRAIK, 15 (1911), 276f. 35 Ibid., 268ff. ("cordon de dents").

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
240 CYRIL MANGO and IHOR ?EVCENKO
Inside the medallion is a further dedication: Finally, in the westernmost bay of the
Ao-rrplts rris [ouArls] church are a number of extremely damaged
funerary portraits which indicate that the
-rou E(Eo)u, OEEoK[....]
K(al) -rcov yCOVsoOV. bay in question was used for burials like a
Much more interesting is the lower layer narthex. On the south wall is a group or
which is in the Palaeologan style and should family portrait consisting of four shorter
probably be dated before the occupation of figures, the second from the east having his
Trigleia by the Turks, i.e., not later than the or her hands crossed over the breast, and,
first quarter of the fourteenth century. The seemingly, two taller figures. The nineteenth-
location of the surviving fragments is indi- century window has amputated the heads of
cated on fig. 20. In the bema are small the "children" (fig. 38). On what remains of
fragments of bishops dressed in their usual the west wall is the figure of a bearded monk
polystauria, who appear to have been moving wearing a cowl, his hands extended to the
towards the center of the apse. It should be right, in the direction of another (holy?)
noted that these bishops' figures are not personage (fig. 37). On the north wall is a
contemporary with the construction of the beardless, possibly female, figure in ornate
church because they overlay the masonry robes, hands extended to the right (fig. 39).
fill blocking the narrow passages between This composition, too, was damaged by the
the bema and the pastophoria (fig. 33). insertion of a nineteenth-century window.
In the north crossing, facing east, above All the early wall paintings are pitted so as
the Corinthian capital, is a representation of to facilitate the adhesion of the subsequent
Joachim's Offerings Rejected (figs. 31, 32). coat of plaster. It is particularly regrettable
Joachim strides forward carrying a lamb in
that so little should have remained of the
his veiled hands. Anna, who also appears to funerary portraits, which probably repre-
be carrying a lamb, turns back toward her sented the local gentry of the fourteenth
husband. The surface of the painting is century. A campaign of restoration would
largely destroyed (except for Joachim's left certainly result in the recovery of many more
leg) revealing the underdrawing. The figure paintings in the church.
of the high priest must have been placed
3. The Monastery of Medikion
around the corner, facing south. This is
normally the initial scene in the cycle of the This monastery, situated about a quarter
Infancy of the Virgin:36 it is strange that it of a mile southwest of Tirilye, was originally
should here read backward, i.e., in an anti- dedicated to St. Sergius, but was later re-
clockwise direction with regard to the named -rcov dyfwov 'a-rIpcov.37 The change
interior. The presence of the cycle confirms,
in any case, the original dedication of the 37 See Euangelides in Icora'p, 12 (1889),
155f., who describes the church as a big rec-
church to the Virgin Mary. tangle, "without columns or roof, like a han
On the south wall of the southwest bay (X&viov) that has been abandoned by the owner."
of the church is a gigantic figure of the He notes in the bema the date Mrivl Map-ricg Ky'
archangel Michael, clad in an elaborate [Trouv Z->MA [sic], which must surely be ,S'>MA',
i.e., 1433. See also the same author's
imperial costume with jeweled loros, holding
BpWIatov-TpiyXtca, 50ff.; A. Hergbs, "Les mo-
a staff in his right hand and a transparent nastbres de Bithynie. M6dicius," Bessarione,
globe in his left (figs. 34, 35). In the lunette 5 (1898-99), 9-21; Hasluck and Pan'enko
above St. Michael is an eight-pointed star: quoted infra. The history of the monastery,
whatever it contained is no longer visible. which attained its peak in the early ninth
century, and later belonged, by way of chari-
On the north face of the pilaster to the right stike, to Michael Psellos, has not been told in
of Michael is a full-length figure of a beardless full. The references to Psellos have been col-
military saint, his right hand holding a lected by H. Ahrweiler, "Charisticariat et autres
raised, bared sword, his left leaning on a formes d'attribution de fondations pieuses aux
shield (fig. 36). Xe-XIe sidcles," Zbornik Radova Vizantologkog
Instituta, 10 (1967), 25. Psellos himself describes
36 See J. Lafontaine-Dosogne, Iconographie the situation of the monastery, Ep. 125, ed.
de l'enfance de la Vierge dans l'Empire byzantin Kurtz-Drexl (as in note 5 supra), 149. An
et en Occident, I (Brussels, 1964), 62ff.
eleventh-century Gospel lectionary cumr syna-

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 241

Hasluck who saw the monastery in the
of name had already occurred by the eleventh
early
century."8 We paid a brief visit to the mona- years of this century:
"The
stery in 1968, and that after dark, so wechurch itself stands in a court: it is
cannot give an eyewitness accounta of its barn-like building with a long nave
rough
remains-an omission we hope to repair divided
on afrom the single (north) aisle by piers
future occasion.38a For the present, we of masonry; the apse retains its semicircular
should
like to reproduce the description and seats.
sketchOn the south side is a chapel with
plan made of it by Covel,39 who writesremains
as of a pavement in opus sectile. The
follows: interior of the church is not without impres-
siveness, due mainly to the gloom and the
On our way just at ye foot of the last
array of old gilded pictures on the screen."40
hill to our left hand (about 1/4 of a mile
A few years later (1910), Panienko noted:
out of ye Town) stands a little mona-
"The buildings of the present monastery are
stery; they call it &ytot TTa-rpEs, in modern. The inscription above the entrance
honour of ye Eo9p6pot -rra-rpE in ye gate proclaims that in 1801 avEKIa1Vi'$Tl EK
councel of Nice. There are belonging to it
about 25 Ka26yEpot; and there are round P3&Spcov }l -rapo0aa a ovil To0 Mrl8tKiou.
However, the church is in part (namely the
about the monastery vineyards, and
chapel of Arabissa)41 ancient and is remarkable
plow'd grounds, and olive and fig trees
for the mass of old icons [it contains]. This
all which these monkes cultivate with
is a veritable storehouse or museum of Greek
their own hands; none are exempt from
church painting.... .42
labour but ye ifyoiEvos [sic] or prior, and
In view of the archaic, basilical shape of
some few preists.There were several young
the church, it would be important to know
lads that came thither to learn to read.
the date of its construction. Hergis suggests
There is an old church there built long
ca. 780,4 which is nearly right for the foun-
wayes, thus [follows the sketch plan:
dation of the monastery, since it was already
see fig. 2]. The top is ruinated, but again in existence at the time of the Council of
(in a poor manner) repaired and tiled, but
787.44 It is stated, however, in the Vita of
I beleive heretofore it may have been
the founder, St. Nikephoros, that when he
arched. At ye bottome it hath been
purchased from a commune of peasants the
checker'd wth black and white marble,
land on which the monastery was to be built,
tesselato op(er)e; as yet some part of
he found there a big dilapidated church of
ye floore remaines.
St. Michael, which he proceeded to restore.45
Of particular interest here is the fact 40 "Bithynica," 293.
that the monastic church was of basilical 41 Pan'enko goes on to describe some of the
shape. The south aisle had disappeared prioricons and manuscripts preserved in the monas-
to Covel's visit; the north one was separatedtery. Among the icons was one of the cTavcayic
'Ap'pa3crcra dated 1688, which accounts for the
from the nave by a row of square piers. Covelname of the chapel.
may have been quite right in thinking that 42 IRAIK, 15 (1911), 274f.
originally the church was vaulted. These 43 As in note 37 supra, p. 11.
architectural features are corroborated by "' Subscription of Nikephoros, hyopEvoS -roo
&yiovu pyiov -ro0 Mrl8KtGvoS: Mansi, XIII, col.
xarion with added Renaissance miniatures, 153A.
now at Horsens (Denmark), once belonged to 46 F. Halkin, "La vie de S. Nic6phore fonda-
the monastery and contains an inventory of its teur de M6dikion en Bithynie," Analecta
books (very few in number) and liturgical Bollandiana, 78 (1960), 413-14. This Vita
furniture: J. L. Heiberg, "Ein griechisches presents a topographical difficulty which ought
Evangeliar," Byzantinische Zeitschrift, 20 (1911), to be stated even if it does not admit of an
507 f.; O. Volk, Die byzantinischen Kloster- obvious solution. It informs us (pp. 408-9)
bibliotheken von Konstantinopel, Thessalonike that Nikephoros received the tonsure at the
und Kleinasien, diss. Munich, 1954 (typewritten), monastery of Herakleios, whose abbot Joseph
180f. then dismissed him to one of his family estates
38 See infra, p. 261. situated west of Katabolos (KTc-r& 8avo&p T-rota Kca-
38a See Additional Note at the end of the
racp6Aov) that he might purify his soul. Nike-
present article. phoros founded a first monastery (on his
39 British Museum, Add. MS 22912, fol.estate
266r.?). When the number of his monks had

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
242 CYRIL MANGO and IHOR ?EVCENKO
This church must have become the katholikon of Tirilye on a slight eminence close to the
of the monastery, since it was in it, on the seashore. Though briefly noticed in several
left-hand side of the narthex, that Nike- fairly old publications," these remains have
phoros was buried a few days after his death never been adequately described and illus-
on May 4, 813.46 His successor, St. Niketas, trated.
was later buried in the same tomb.47 If these The identity of the monastery is not
indications pertain to the church that was subject to serious doubt since it retained its
still partly preserved in the early years of name and continued functioning until the
this century, it should be regarded as being early part of this century. Today it is still
considerably earlier than the 780's. known as Ay Yani. Furthermore, it was
visited in 1676 by Covel who drew a rea-
4. St. John the Divine of Pelekete sonably accurate sketch-plan of its church
The remains of this monastery, which (fig. 3), and gives the following account of it :49
played a prominent role in the iconoclastic
period, are situated about three miles west About 2 1/2 mile beyond the town
[of Tirilye] stands another monastery
grown to twelve, he felt the need for larger just upon ye sea over against KaA6Atkvoso
quarters and moved to Medikion, which is upon ye bottom of ye cliff. It is dedi-
described (p. 413) as being inland, but not far cated to CCytoS 'ICAoxvvrs SeoA6yoS, and
removed from the sea, from which it was 7.5
cal'd govf TTi 7TEXEKaI-rs [sic].51 It is
stadia distant, as it was also from the first
older they say then that of ye &ytot
monastery (&dK1riq ptov iv pEaoyaicp -rtl oi wrap&
TTaTrpEs.62 There are 20 KcA6yEpot there
-rroA 8ts5py6pEvov rij~ SaXad&a Ka Kcr i -rij~ rrpc'-rlS which labour and maintein themselves in
PovijS, o'Ei &TrTr y1E ptOV C tu rTaSfous KcTkppSEV tic-
Kt1aOIVOV, MESiKlv 68' 6vopaLC6pEvov). A gloss con- like manner as the other, there being
tained in the older of the two manuscripts of the excellent fruitfull ground about them.
Vita informs us that 7.5 stadia were equivalent to There church is much in the forme of
one mile, which is, indeed, the actual distance
from Medikion to the sea. On the other hand, that at ye town (of ITavTaPaat1iaooa
the monastery of Herakleios is known to have [sic]) onely there wants the W. isle. The
been at Kios (Gemlik) and Katabolos was the square then is much bigger then the
name given to the coastal area immediately other, and ye cupola stand in a manner in
west of Kios: see K. Amantos in 'EXArVIKaK,
6 (1933), 149-50; Mango, "The Monastery" ye middle. In ye sacristia on ye E. side
(as in note 25 supra), 172; Papachryssanthou, ye Altar are rays'd three semicircular
"La vie" (as in note 20 supra), 319. Halkin steps (I have seen ye like in ye monastery
(op. cit., 408 note 3) tries to resolve the diffi- in ye seraglio in Brussa) aloft ye p(at)riarch
culty by postulating the existence of not only
two, but even three monasteries of Herakleios, sits and on ye rest his m(et)ropolites
one at Kios, the second on the Bithynian and others, and I question not but this
Olympus, and the third 7.5 stadia inland from gave occasion to call ye whole place
Medikion. For the second he refers to B.
(over ye Altar) &ytov P3ifa. This mona-
Menthon, L'Olympe de Bithynie (Paris, 1935),
156, who, however, merely mentions the exis- stery as also the other of ye &ytot TraT&-
tence of Byzantine remains at a place called pEs are under ye p(at)riarch, he makes ye
Erikli Yayla (="Plum Pasturage"). For our new gyofi*voS, or rather sell ye place;
part, we are inclined to believe that there was
only one monastery of Herakleios, namely at 48 Euangelides in w-n'ip, 12 (1889), 275f.;
Kios. The distance between Kios and Trigleia repeated with some additions in the same
is, however, about 30 kilometers as the crow author's Bp 2sxtov-TpiyX~ta, 35ff.; A. Herg6s,
flies, so that our hagiographer must be guilty "Monastbres de Bithynie," Echos d'Orient, 1
of inexactitude or unclarity. On St. Nikephoros, (1897-98), 274ff.; Hasluck, "Bithynica" (as
see also F. Halkin, "S. Nic6phore de Medikion in note 2 supra), 293, and Cyzicus (as in note 23
d'aprbs un synaxaire du Mont Sinai," AnalBoll, supra), 61; Pan6enko, IRAIK, 15 (1911), 271f.
88 (1970), 13ff. 49 British Museum, Add. MS 22912, fol. 266r.
46 AnalBoll, 78 (1960), 424. Note that where- 50 On Kalolimno, also called Kalonymos
as this text gives the date of the demise as (Imrall Adasi), see Hasluck, "Bithynica," 301ff.
May 4, the Vita of St. Niketas has April 4 51 Above the line Covel wrote: "ye rock was
(Artemisios): Acta Sanctorum Bollandiana, cut & hewn plain (from thence [illegible word])
April, I, App., xxiii, chap. 24. ye monastery set upon."
47 Vita of Saint Niketas, xxvii, chap. 49. 52 The monastery of Medikion.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 243

they are therefore cal'd ac'aupoTr'5tov53 It has often been stated that the mona-
as being founded by ye p(at)riarchs stery of Pelekete was founded in 709.56 For
authority of weh elsewhere. this we can find no authority other than that
There were many young lads which of the worthy Euangelides. The earliest
are sent thither likewise to learn to read mention of this monastery occurs, if we are
and there were severall old women not mistaken, in the Vita of St. Stephen the
(KOxhoypaicSEs) though they belong notYounger. During his last incarceration in
to ye foundation, but come for severall the Praetorium (hence, presumably, in 764)
uses and necessityes; I spyed one hand- Stephen found himself in the company of
some young woman wel clad but all three in hundred forty-two imprisoned monks,
black, setting close in a corner talking among whom was Theosteriktos, the aged
wth a lusty young KccA6yEpo; at our abbot57 of Pelekete. This unfortunate cleric,
coming they both vanisht; I askt who whose nose had been cut off and whose beard
she was. They told me, a widdow who had been burned by the iconoclasts, told the
came to confession. following story. On the previous Holy
We had been informed that there Thursday (hence either in 763 or 764), while
were here many antient stones and old he was celebrating divine service in his
M. SS. but the deuce of one we met wth monastery,
all; the notorious Lachanodrakon,
they have many old church bookes, but "governor of the Asiatic land," broke in with
not one of any account in ye world. In this
a band of soldiers; he put thirty-eight monks
monastery of St. John in ye corner of in yefetters, while others were flogged, burned
court next ye sea, we goe down aboutor
20(like the abbot) had their noses amputated
and their beards tarred and set on fire. Not
or 30 steps into a kind of a grotto, where
content with this, Lachanodrakon had the
is an i'aoya,54 ye water was very good,
but in a manner lukewarme, yet they entire monastery burned to the ground,
including the stables and "the churches"
say in ye heat of sommer it is very cold;
it lyes deep and runs wth a full pipe of 4
(-rcTv AKKriaIncv). The thirty-eight arrested
inches. To come and see this placemonks in were conveyed to Ephesus and buried
may or june must needs be exceeding alive in the vaulted chamber of an ancient
pleasant. bath.
From this account it may be deduced
The only modern author who has question-
that in the middle of the eighth century
ed the identity of the monastery is Father
B. Menthon, who confidently places Pelekete
farther inland, southwest of Tirilye, but he retrouv6es exactement dans la direction et A la
does not give any valid reason for his opinion.55
distance marquees par les textes, la vaste
6tendue de ses ruines et la description d6-
taill6e des lieux donn6e par l'auteur de la vie de
53 Read o-raupornilylov. S. Macaire, ne laissent pas subsister l'ombre
54 Read &yiaopia. d'un doute." The text he has in mind appears,
55 L'Olympe (as in note 45 supra), 37f., 44ff.
therefore, to be the Vita of St. Makarios to
The author asserts that "d'apres les anciens which we shall presently turn. In fact, the Vita
textes," Pelekete was situated southwest of contains no statement whatever concerning the
Tirilye. In that direction, a few hundred meters
situation of Pelekete. It is Van den Gheyn, the
from the road leading from Tirilye to Uluabat modern editor of the Vita, who quotes Euange-
(Lopadion), he found some shapeless ruins on lides to the effect that Pelekete is southwest of
a high platform, which he confidently identifies
Tirilye at a distance of three quarters of an
with the monastery. As to the church under hour: AnalBoll, 16 (1897), 141. Yet Menthon
discussion here, he says: "Le petit rectangle
uses this very statement to refute EuangelidesI
sans style, ressemblant plut6t A une pauvre 66 This date has been repeated in all sub-
masure, que l'on montre au bord de la mer, sequent literature down to H.-G. Beck, Kirche
non au sud-ouest, mais exactement d l'ouest de laund theologische Literatur im byzantinischen
bourgade, n'6tait jadis autre chose qu'un Reich (Munich, 1959), 210.
ermitage d6pendant de cet illustre monastere; 57 He is called wpeoparr7pos Kal yipaXoos - ro
et il est mmkne int6ressant A visiter comme
Eayo0S ~ppovToT-rP1ovpfo ri HTTAEK11Tti (PG, 100,
sp6cimen d'ermitage, car c'est le seul qui ait col. 1164 D), but the synaxaria designate him
subsist6 jusqu'd nos jours." On p. 45 he asserts as flyo?ipEvos: Synaxarium CP, 496,46.
once again: "Les traces de ce monastbre 68 PG, 100, cols. 1164D-1165D.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
244 CYRIL MANGO and IHOR SEV(ENKO
Pelekete was a large establishment with and rather vacuous document we can glean
upward of forty monks and more than one the following facts. Makarios (whose secular
church. There does arise, however, the fol- name was Christopher) was born of good
lowing difficulty: Michael Lachanodrakon, family at Constantinople. He lost his parents
whose iconoclast zeal is well known, was at an early age and was brought up by an
governor of the Thrakesian theme,59 while uncle. His education completed, he retired
Pelekete (unless we are dealing with another to a modest room near a church whose priest
monastery of the same name) was clearly in became his only companion. Since he mani-
the theme of Opsikion. Would the energetic fested a desire for the monastic life, the
strategos have made a raid on a neighboring priest directed him to Pelekete, which the
theme, and this by imperial command ?60 We young man found to his liking. At this point
are content to pose the question. As for the biographer inserts a brief descriptive
Theosteriktos, his subsequent fate is not passage which may be rendered, more or
recorded. His memory is celebrated on less, as follows: "A rock rises to a height on
February 17.61 both sides and is reached by a single ascent
It has been maintained that Theosteriktos which is not easy. Standing on higher ground
was succeeded by St. Hilarion, whose memory than all the adjoining valleys, it enjoys a
is celebrated on the 27th or 28th of March. purer air; and it is smooth all around and
This is merely a guess. All we can say con- well-turned as if (?) hewn with an axe."66
cerning Hilarion is that the Canon in his Here, then, we have the etymology of the
honor was composed by Joseph the Hymno- name Pelekete; its appropriateness is today
grapher (816-66); and the only piece of difficult to judge since the bluff on which the
concrete information we can extract from monastery is built is thickly overgrown.
the Canon is that Hilarion's tomb performed Christopher was readily admitted by the
miracles.0 Nor is it at all certain that the community. A monk named John instructed
abbot Mark of the monastery of St. John him in the monastic rule, while the abbot
Theologos who attended the Council of 787" (unnamed) set him to copying books. So
was abbot of Pelekete. good was his progress, that he was unani-
Toward the end of the eighth century mously chosen to receive the tonsure and
Pelekete found a worthy abbot in the person was renamed Makarios. Eventually, he was
of St. Makarios, whose Vita was written by promoted to oikonomos and, soon thereafter,
his disciple Sabbas.65 From this rhetorical to abbot. At the same time he began to
acquire powers of healing. He cured of a
59 Theophanis Chronographia, ed. C. de Boor, lingering disease the patrician Paul, who was
I, (Leipzig, 1883), 440,27, and 446 for the ex- governor of the district-to be identified with
tirpation of monks from the Thrakesian theme.
Paul, patrician and comes of Opsikion, who
60 Synaxarium
PG, 100, col.CP,
1165A:4-r7 ij To aath(Feb.
osEEK 5at. suffered a defeat at the hands of the Arabs
61 470,17; cf. 496,46
29). in 799.67 He also healed at Constantinople
62 Ibid., 564,6; cf. 565,37, 573,33; M. Gedeon, Paul's wife as well as the wife of another
Bulavrtrvbv op-ro,6ytov (Constantinople, 1899),
patrician called Theognostos. His fame reach-
87.
ed the Patriarch Tarasios (784-806), who
63 The Canon, which has the acrostic Xo00 roCs
IAapo*s alviaco -rp6wovu, 1wrrEp. 'Icoai), may be
conferred upon him the diaconate. After
found in the menaia under March 28, and returning to his monastery, Makarios per-
extracts from it in PG, 105, col. 1085. Cf. formed further miracles: he cured a para-
ActaSS, March, III, 728-30.
6 Mansi, XIII, col. 153D. 66 Ed. Van den Gheyn, 145,5. The text, as
65 Ed. Van den Gheyn, AnalBoll, 16 (1897), printed, appears to be in need of emendation:
140ff. This Sabbas should be distinguished
from his homonym who wrote the Vitae of iS IYoS Iv~LV y&p &KcarTpcov T CV PEp'CV ipprat wrrpav
[Trr-rpa ?] pv ria v Ti vo6ov, Kc al-rTrrTlv OIK vapoXSov
St. Ioannikios and St. Peter of Atroa. See
V. Laurent, La Vie merveilleuse de S. Pierre EXovaa, avc)Klo'apv [&vco KESIvV ?] 8EIoG
W1T&VTrV "rrp6
aCoriv KOfAcov, &Kpactqw~o-rpOv -rv &pa XXErac
d'Atroa, Subsidia Hagiographica, 29 (Brussels,
1956), 16-17, and H. Loparev, "Vizantijskija Adia rE -r& wrpt1 TruyYXcvovaac KCx ErTopvoS TrEAEKT -
rilv [read 1TEAEKTYIfI or OCiTEp WrrEEKTyr1~].
zitija svjatyh VIII-IX vekov," Vizantijskij Vre- 67 Theophanis Chronographia, I, ed. de Boor,
mennik, 18 (1911), 137ff. 473,29.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 245

lytic boy and caused rain to fall in aperiod time ca. 815-18.71 At the time Makarios
of drought while performing a procession was under arrest or in exile, and Theodore
to a neighboring shrine of the Prophet encouraged him to persevere in the right
Elijah.68 faith. The extremely respectful tone in which
Upon the outbreak of iconoclasm (in 814 these letters are couched-Theodore calls
or soon thereafter) Makarios was summoned himself the "son" of Makarios and addresses
to Constantinople. John the Grammarian the latter as "father"-may suggest that
(Iannis) tried in vain to convert him to the Makarios was the elder of the two, i.e., that
iconoclastic doctrine. Makarios was impris- he was born before 759. Another letter,
oned and then exiled, somewhere in the addressed to Ignatius, bishop of Miletus,
area of the Bosphorus. After the murder of cites Makarios of Pelekete as one of a few
Leo V (820), he enjoyed a brief respite andabbots who had not succumbed to icono-
spent some time with the exiled Patriarchclasm.72 Finally, a letter addressed to the
Nikephoros.69 He also began building a "brotherhood of Pelekete" (probably dating
monastery at the place of his former banish- from ca. 823-24) informs us that the commu-
ment, i.e., "at the straits of the Propontis."nity, finding itself leaderless after the death
It is implied by the Vita that he did not
of its abbot-hence surely not Makarios,
return to Pelekete. After the accession of who lived on after 829-had elected a certain
the Emperor Theophilus, Makarios was once Sergius, apparently an anchorite, since he is
again arrested and imprisoned; while in described
jail as "coming down from a hole in
he converted a Paulician heretic, one of a the rock." Theodore approves the election
group of Paulicians who had been condemned provided Makarios does not object, and
to death. He was then exiled to the small pending the return of normal conditions when
island of Aphousia close to Prokonnesos,70
the matter can be settled more canonically.
where he caused a church to be built and He then exhorts Sergius to keep to correct
where he eventually died (August 18, yeardoctrine and hints that there had been some
unknown). His tomb, constructed by defection
his in the community.73 The nature of
disciples Dorotheos and Sabbas, remainedthis
on defection is explained a little more
Aphousia and performed many miracles. clearly
In in another letter addressed to the
the concluding paragraph, Sabbas (who
abbot
is Sergius (presumably the same). His
the biographer) intimates that he succeeded
predecessor had "fallen" and dragged down
Makarios as abbot, perhaps of the monastery
with him the whole community. Theodore's
on the Bosphorus, and not necessarilyadviceof is not to disperse the monastery, but
Pelekete, as has usually been assumed. rather to strengthen it by the strict obser-
The meager data furnished by the Vita vance of orthodox doctrine.74
may be to some extent supplemented from
71 See C. Van de Vorst, "Note sur S. Macaire
the correspondence of St. Theodore of Stu-
dios. Theodore wrote five letters to the de P616cete," AnalBoll, 32 (1913), 270-73. The
letters in question are in A. Mai, Nova Patrum
"abbot Makarios," all of which fall in the
Bibliotheca, VIII (1871), nos. 87 (815-16), 151
(summer-autumn 816), 262, 271 (probably 818);
68 As noted by Van den Gheyn, 152 note 1, and Ep. 11.20, PG, 99, col. 1177ff. (late 816-
Euangelides in Xco-rIp, 12 (1889), 279, postu- early 817). On the dates see A. P. Dobroklonskij,
lates the existence of a monastery of St. Elias
Prep. Feodor, ispovednik i igumen Studijskij,
between Trigleia and Sige. Even if this is notII/1 (Odessa, 1914), 272f., 318f., 386f., 391f.
a figment of the Greek author's imagination, In Mai, no. 38, p. 31 (letter to Naukratios
it need have no connection with the church dating from 815-16), Theodore mentions two
mentioned here. former monks of Pelekete (not lignorum olim
69 The Patriarch was at the time dwelling in
caesores as the Latin translation renders it!)
the monastery of St. Theodore on the Asiaticwho appear to have joined the Studite commu-
side of the Bosphorus. See J. Pargoire, "Anity.
propos de Boradion," BZ, 12 (1903), 476; 72 Mai, op. cit., no. 188, p. 160. Cf. Dobro-
P. J. Alexander, The Patriarch Nicephorus of klonskij, op. cit., 334ff.
Constantinople (Oxford, 1958), 148. 73 Ep. 11.146, PG, 99, col. 1457ff. Cf. Dobro-
70 On this island see M. Gedeon, TTpolK6vvarlro klonskij, op. cit., 480f.
(Constantinople, 1895), 58ff., and, on the exile 4 Ep. 11.179, PG, 99, col. 1553ff. Cf. Do-
of Makarios, ibid., 71. broklonskij, op. cit., 490ff.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
246 CYRIL MANGO and IHOR ?EVCENKO
From the above data we may draw summarized
the in those of 1788 and 1794)
following conclusions. First, the monastery
related that a certain Arsenios of Triglia had
was certainly rebuilt after its destruction:
found the monastery in a ruined condition,
this may have happened in the last repaired
two it, endowed it with fields, vineyards,
decades of the eighth century. Second, upon
and olive groves, and constituted a koinobion
the removal of Makarios ca. 815, anotherthere. In order to preserve its independence,
abbot (unnamed) was appointed and he fell
placed it under the direct control of the
into heresy. The latter died ca. 823 and Patriarchate,
was to which the monastery was
succeeded by Sergius. These circumstancesobligated to deliver seven okes of oil annu-
explain why Makarios did not return to This was the situation encountered
ally.77
Pelekete in the reign of Michael II (820-29),
in 1676 by Covel who, as we have seen, found
but instead set up a new monastery to which
twenty monks in residence. The provisions
he may have attracted some of his formerof the sigillion of 1658 were partly upset by
disciples. that of 1788 which placed the monastery
Theodore's exhortations appear to have under the supervision of the Kaminars78
borne fruit for, in the reign of Theophilus, Antonios and Demetrios. So disastrous was
Pelekete (described as a "very big monas- the interference of these two archons, that
tery") became once more a center of oppo- the brotherhood made an appeal to the
sition to iconoclasm, and its monks refused Patriarchate and obtained the sigillion of
communion with the Patriarch-probably 1794 which re-established the status quo ante,
Antony I (821-37). This came to the atten- including the payment of seven okes of oil.79
tion of the Emperor, who directed the Finally, in 1797, the Patriarch Gregory V,
spatharios Kallistos to proceed to Pelekete as part of his general policy, reaffirmed the
and have the "panel-worshipping" monks "stauropegiac" rights of the monastery, but
(arcavitrrtorol) scourged and expelled if commuted its annual dues to sixty piastres.80
they persisted in opposing the imperial As we shall see presently, the church of the
doctrine. Kallistos (who later died for the monastery was severely shaken by earth-
faith together with the Martyrs of Amorium quake in 1855 and rebuilt the following year.
in 845) did not answer the imperial com- It was burned in 1880 and restored to some
mand.74a At this point the Byzantine history extent thereafter.
of Pelekete comes to an end.
Turning next to the architectural form of
We pick up the story again in the seven- the church, we may begin by observing that
teenth century. There exist four patriarchal in Covel's time the structure was still entire
sigillia referring specifically75 to the mon- except for the narthex. The following eye-
astery of St. John the Divine of Pelekete. witness accounts dating from the end of the
These are dated 1658, 1788, 1794, and 1797, nineteenth and the early years of the twen-
respectively, and they confirm the autonomy tieth centuries add a few specific details.
and "stauropegiac" status of that monastery,
Euangelides (whose recollections or notes
i.e., its direct dependence on the Patriarchate
dated back to the 1880's):
and not on the metropolitan of Prousa. Of
these documents the last three have been 77 T. Euangelides, lacrrplapXiK6V liyiAMiov, in
published in full. The sigillion of 165876 (as evoq&vnrS, 1 (1897), 333-36; D. A. Zakythenos,
'AvKiOT-r WCrra-ptaPX1Ka& -yypaga, in "'EMjvlK&, 6
74a Skazanija o 42 Amorijskih Mudenikah, (1933), 141.
ed. V. Vasil'evskij and P. Nikitin, Mdm. de 78 Officials of the Phanariot r6gime in Ro-
I'Acad. Imp. des Sciences de St.-Pitersbourg, mania who were responsible for the collection
VIIIe s6r., Cl. Hist.-phil., VII/2 (1905), 25 and of taxes on alcoholic beverages and tobacco.
comment on p. 156.
79 Ed. Zakythenos, 'AvhKSo-ra, 140ff.
75 K. N. Sathas, MECat(lcvlK BipMoSf)Krl, 3 80 Ed. with some serious mistakes by D. A.
(Venice, 1872), 587, cites a sigillion of 1652, Trrepi Petrakakos, 'AvKiOTrOV alyflhov rpnyopiov trov
-rov Xcopiov TptyAfaS Kai 'EAyp&v6v, 65rl o-ravpo- E', in 'EKK.$)p., 3 (1909), 94ff., from a manu-
wfyi& EsliT, Kal OOX)(1 TOKEImpevcX -r'j lpo"tISc , but it script at Andreaskiti, Mount Athos. Cf. P. G.
is not clear whether this refers to the monastery
of Pelekete or to that of Medikion, or to both. Zerlentes in NoSno loijv, 2 (1920), 243. Euan-
76 Ibid., 594 (text not given); cf. M. Gedeon, gelides, Bp0AAEtov-Tpfy?,sta
133, claims (as in
to have found the same note 4insupra),
sigillion
'Ta-rplapXnKO vrrivamKE (Constantinople, 1890), 588. 1887 in the hay barn of the Medikion monastery.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 247

There remain [of the original struc- yofrOat av Trcp EOCp ~8paicos [sic] oa-r'hos
ture] only the marble pavement, a few
10 xai 18o0" y FyipCOv 1K TCOV pEiTriCOV
capitals and column bases inserted avEyipSrl [sic] KQT-ra pfiva 'Arrp. 1856.83
into the walls during the rebuilding. However, both in the church and
After so many centuries, the church hasnear it are antiquities of the Byzantine
become a small rectangle lacking anyperiod. The apse and the two columns
style and resembling an ordinary poor
next to the sanctuary are ancient. The
building. Around the church is a court-
capital closest to the sanctuary is
yard with cells in which the abbot andidentical with one that stands today
monks reside. next to the entrance of the church. The
He then mentions the fire of 1880 and adds latter measures 0.70 m. between its vo-
that, following this, the monastery was not lutes-the normal size of big capitals....
repaired in any way. The acanthus leaves form one row, and
"Today, there are preserved in it only a each has three points and two longi-
small box with holy relics and a bu- tudinal grooves. The four volutes come
cranium immured in the outer wall on together at an acute angle; in the center
the right-hand side.81 of the arc-like depression between them
there is, on each side, a rounded-off pro-
Hasluck (who travelled in the area in jection. Between the pylon of the trium-
1902-6): phal arch and the closest column is
It [the church] is now a poor modern inserted or lies in situ an ancient marble
building, having been damaged and cornice with rosettes. Outside, in front
repaired after the earthquake of 1855 of the entrance is immured a fragment
and burnt and rebuilt in 1880: a fair of a marble frieze consisting of two
amount of old detail, including fragments bands: a horizontal branch with leaves
of a marble pavement, a Byzantine and, underneath, a row of parallel acan-
cornice, and some old capitals, has been thus leaves, pointing down and placed
used in the reconstruction. It is tenanted diagonally with respect to the branch.
by a single priest. A rock-hewn hermit's The most important object in the
cell near the church explains the name church is a carved wooden iconostasis
of the monastery.s2 in two storeys. The carving is very
complex and meticulous, especially on
Panc'enko (1910):
the half-colonnettes; the ornament is
At one hour's walk from Triglia is everywhere vegetal, consisting of dif-
situated the monastery of St. John the ferent types of acanthus. A carved
Divine T-riS TEAEKTi'S. The church stands wooden throne is decorated with an
in the middle of a rectangular courtyard acanthus-like vine.84
formed by residential structures with
wooden galleries, and is itself new, as In recent years the ruins of the monas-
may be seen from the inscription over tery have been used as a chicken farm, but
the entrance: today (1972) they are deserted except for
two savage watchdogs. Chicken houses in
'O -rrav86rpdrcTp Xp6vos 6 "rp6ycov "rTavTra various stages of dilapidation obscure much
T-rv va6v T-rourov pEfleSr Sauxpacic
83 These feeble verses may be rendered as
6 oaEmlap6s OpCoS 6 0rpopEpc cuTva Vtr follows: "Time, the all-subduer, who consumes
Tr% "O lpoI\1 Trpo Evios:x Tra:voE.9picS, everything, had miraculously spared this church.
5 KaTo'TpEYE E KCGi TOITOV E1K EIkEiCOV, Yet, the earthquake which caused such dreadful
copv acTOv Trotdilocas, <pEG, EpEiTriOv. events at Prousa, destroyed it, too, down to
the very foundations and made it, alas, into a
6 "rrTps Ta .Ei2a TO rlyoupivou I.os heap of ruins. The abbot's zeal for things divine,
KaCi TnpoSupICa Kai ouvpoi ivPO vicov the eagerness and the contributions of certain
men have risen up like a stout pillar-and
81 BpXXstov-TpiysXia, 39. On p. 40, however,
he speaks of three reliquary boxes. behold: raised from ruin, it has been erected in
82 "Bithynica," 293; cf. Cyzicus, 61 (as in the month of April, 1856."
notes 2 and 23 supra). s4 IRAIK, 15 (1911), 271f.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
248 CYRIL MANGO and IHOR SEVCENKO
of the interior and the exterior north wall Among the carved elements, we may
of the church. Otherwise, the architectural notice, first, the southeast column (fig. 47).
features of the building are more clearly Both its shaft, which is decorated with a
revealed than they must have been in the cross in low relief, and its capital are spolia.
days of Hasluck and Panienko. The capital (fig. 49), which is of the fine-
The church forms a rectangle measuring toothed acanthus type, may be dated to the
on the outside roughly I5 X 12 m. (fig. 40). middle or second half of the fifth century.85
The eastern end is original and stands to the Another large capital with four volutes, but
full height of the bema vault; the rest was of somewhat different form, lies buried next
rebuilt in the nineteenth century, but to the north wall of the nave. Imbedded into
follows the lines of the Byzantine walls. The the projecting corner above the southeast
narthex, which disappeared prior to Covel's column is a richly carved cornice of the Early
visit, must have communicated with the Byzantine period. It has a wave pattern on
nave by means of three doors: the northern its cimatium, different leaf forms on the
jamb of the north door together with part of modillions and rosettes between the modil-
the arch it subtended are still plainly visible lions (fig. 48). Fragments of a similar cornice
in the west faqade, the rest of the west wall may be seen in the corresponding position
being modern (fig. 41). The south wall on the north side of the church (fig. 45). The
(fig. 42) is also modern, except for traces of continuation of this cornice into the bema
ancient construction that may be seen at and apse is, however, plain and dates from
its base along a stretch of 4 m. starting at the construction of the church. The door
the southwest corner. The two side apses, leading from the bema into the diaconicon
which must have been semicircular, have has a molded marble frame (fig. 51) next to
been cut off, on the outside, along a straight which there remains in situ a piece of molded
line. skirting, an indication that the church was
The central apse was lit by three windows, once reveted with marble. In the apse of the
but the two lateral ones have been walled up. prothesis a piece of late antique cornice
Each one has a brick arch of two voussoir decorated with a wave motif and modillions
rings (fig. 43). Traces of the synthronon arehas been reused to form a shelf (fig. 52).
still observable. The pastophoria are coveredLying loose in the apse is a chancel slab of
by domical vaults as are also the corner the fifth or sixth century decorated with
bays of the nave. The north and south walls three crosses, the central one in a circular
of the nave appear to have had wide openings medallion, the two lateral ones poised on
in the form of tribela, as in the Fatih Camii
curving shoots of ivy (fig. 50). Outside the
of Tirilye (fig. 9) whose ground plan showschurch we have found two small impost
considerable similarity to that of Pelekete.capitals decorated with crosses (figs. 53, 54).
Note, next to the northeast pier, the spring- Finally, built into the south wall of the
ing of an arch of narrow span which wouldchurch are two fragments of a late Roman
have been appropriate to a tribelon (fig. 46).sarcophagus or frieze with ox and ram's
Of the four columns that once supported the heads and a garland (figs. 55, 56).
dome (whose approximate diameter was
4m.) only the southeast one remains in II. KURSUNLU AND VICINITY (fig. 60)
place.
The Byzantine construction exhibits five 1. (7ingyan Kale
consecutive courses of brick (height of five ?ingyan Kale (the forms of the name vary:
courses and five joints, 0.57 m.) alternating we heard it pronounced as "?ingyen,"
with four courses of roughly squared stones "Tsingyan," or even "?ingene" Kale, i.e.,
(height of four courses and three joints, Gypsy Castle) is situated near the highest
0.75 m.) The bricks measure 0.34 to 0.39 m. point of the Karaday range about fourteen
in length and 0.05m. in thickness. The
mortar joints are 0.065 m. thick and have 85 It is quite similar to those of the nave of
the Acheiropoietos at Thessalonica: R. Kautzsch,
been wiped flush with the vertical surface
Kapitellstudien (Berlin-Leipzig, 1936), 134 and
of the rwalls. pl. 26, no. 432.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 249

couldand
kilometers southwest of Plaj (Yenik6y), have been an observation post or a
place
at 54 degrees south of east, measured of refuge for those who were being
from
attacked
the mosque in the village of Kursunlu. from the sea. We are inclined to
It is
connect
accessible only by forest roads, and the ascent(ingyan Kale's construction or at
least
by Land Rover takes about an hour. its use with the wars which the Latins
Today,
of brush
the castle is hidden by substantial Constantinople and the Byzantines of
growth and large old trees, some of Nicaea
them waged in the area between Lam-
psacus
growing out of its walls (fig. 59), but orig-(Lapseki) and Lopadion (Uluabat)
during
inally it had a view of the sea and the the reigns of Theodore Laskaris and
coast
below. It is trapezoidal in shape, with the
John Vatatzes. In 1233, to give an example,
northwestern corner cut off at anthe Latins raided the seacoast reaching from
angle
(text fig. A). In places, its walls rise to a as far east as the area of Cyzicus,
Lampsacus
while the Emperor occupied the foothills of
the mountain, i.e., the coastal range. The
Latin raid was of little avail, because, as
Akropolites tells us, the Emperor "had
anything that was useful stored in places
that were higher up."8s6 ingyan Kale may
have been one such stronghold.

2. Village of Kursunlu
In the village of Kursunlu, the most
interesting spot in terms of antiquities is the
S ,0 2P0 50 04 O
house of Bay Lokman Kasar on the west edge
A. ?ingyan Kale. Approximate
of the village.8" InPlan
the back yard of the
house, near the back door, there are rem-
height of about 8 m. above nants ofground level
an opus sectile onappar-
pavement,
the outside, and about 3 m.
ently of on thein inside.
a church, situ. It is not clear
The construction is of rubble, with some whether the church to which the pavement
regularly hewn stones visible on the inner belonged is that of St. Demetrios, the foun-
face of the walls, especially in the northwest dations of which were seen and briefly de-
corner. The north wall has a rounded tower, scribed by Panienko in 1910. According to
which points to the existence of a gate, him, the sanctuary stood on a "steep rock,"
probably to the east of it. The gate may while Lokman's back yard blends into a
have been flanked by another tower, of corn field without a noticeable break. On the
which only rubble traces remain at present. other hand, Panienko saw fragments of an
The east side of the enclosure is about 85 m. altar screen having three crosses with flaring
long, and the average thickness of the en- ends on the site of St. Demetrios, and at
closure is 1 meter. Lokman Bey's we recorded a fragment of a
The name 0ingyan or Tsingyan Kale similar screen having one cross with flaring
reminds one of Sigriane, the Byzantine ends (fig. 76).88 To be sure, even if our frag-
name of the mountain range on which it is ment is a part of Panienko's altar screen, it
situated. This name points back to a hypo- could have been transported to Lokman's
thetical form Kastron tes Sigrianes, and to house after 910o.
Byzantine times. However, it would be too
risky to date this fortress by its name alone. 86 Georgii A cropolitae Opera, ed. A. Heisen-
berg, I, Teubner (1903), 47,16-25.
Among the rubble inside, we found some 87 On the capital used at the fountain in the
marble fragments and a brick, about 0.02 m. center of the village, cf. p. 257 and fig. 118 in/ra.
thick. The latter find could be of the Late s8 Pan'enko in IRAIK, 15 (1911), 270: on
Byzantine or Early Turkish period. account of its location on a rock, the church of
It is no less difficult to determine the St. Demetrios had only one lateral aisle; length
of the central nave up to the apse: 24 paces;
purpose of our fortress. As it occupies width
a of nave and aisle: 9 steps. Measurements
commanding spot on top of the range, it of the altar screen: 12.54 (?) x 0.84 m.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
250 CYRIL MANGO and IHOR ?EVCENKO
put on
In his back yard, Lokman has gathered a them by representatives of the kay-
lapidarium, which includes a number makam
of of Karacabey, the closest town.
capitals, namely one Doric (fig. 69),Lokman
one also has a collection of small antique
complete and two fragmentary specimensobjects,
of among them a lead cast, which once
the composite order (figs. 72-74), andconnected
one the foot of a marble statue to its
Byzantine acanthus capital of approxi-base-thus suggesting the existence of statu-
mately fifth-century date (fig. 71).ary
Thein the area of Kursunlu89--and an
octagonal lead weight, ca. 0.06 m. in dia-
composite capitals as well as an anta capital
that now lies on the seashore (fig. 75) appear
meter, with the name XaptSilpou (fig. 70).
As a service to classical epigraphists, we
to have come from the same antique building.
Furthermore, there are funerary reliefs
reproduce the following five inscriptions, with
(figs. 61-62) and inscriptions (figs. 63-68).
one possible exception dating from the im-
Some of the stones bear numbers, probably
perial period, in the custody of Lokman Bey:

I. List of Victories at Games (fig. 63)

Provenance: ca. 50 m. west of house of Lokman Kasar
Marble slab, right side missing; height 0.46 m.; length 0.37 m.; maximum thickness 0.12 m.;
height of letters 0.025-0.03 m.; interlinear spaces 0.01-0.014 m.
Letters: X, Q, A, E
Date: second-third century A.D. ?

'IEp6v K6pprs [
'O7i rrta ~ 'Ev B[
as ~ 'Ev NE1xKO[pr8EIa
'Ev TTEpivS9 ~ [- - - 'iE-
5 p6v K6prs ~ 'E[v
BEtSuvfias 'E[v - - - TraavTai-
ous x(ai) EIplt-rc a[vntaIous VtKl(crca
M

On the cult of Kore in the Cyzicene area and on inscriptions connected with that cult, cf. F. W.
Hasluck, Cyzicus ... (Cambridge, 1910), 210-13, 270.

1 and 4: On the Cyzicene games called iEp6v K6prs iowrristov, cf. J. H. Mordtmann, "Zur
Epigraphik von Thracien," Archaeologisch-epigraphische Mitteilungen aus Osterreich-Ungarn, 8
(1884), 219 = no. 49,4-5, reproduced in A. Dumont and Th. Homolle, Milanges d'Archdologie et
d'Epigraphie (Paris, 1892), 392 = no. 74s.
2: The 'OAuiirrma mentioned here are most probably those of Cyzicus, in which case our
inscription is later than A.D. 135. Cf. J. Marquardt, Cyzicus und sein Gebiet (Berlin, 1836), 144-45.
On these Olympia, cf. Corpus Inscriptionum Graecarum, nos. 3665,4, 3675,4-5, 3676,6; Mordt-
mann, op. cit., no. 49,8; idem, Mitteilungen des Deutschen Arch iologischen Instituts, Athenische
Abteilung, 7 (1882), 255, no. 26,10; L. Moretti, Iscrizioni agonistiche greche (Rome, 1953), 266;
L. Robert, Revue de Philologie, de Litterature et d'Histoire anciennes, 3rd Ser., 3 (= 55) (1929),
153 and note 3 = idem, Opera minora selecta, II (Amsterdam, 1969), 1119 (Philippopolis, with
parallels and bibliography); idem, Comptes-rendus des seances de l'Acadimie des Inscriptions et
Belles-Lettres (1970), 20,20. On the basis of the passages in CIG, Mordtmann, and Robert just
cited, in lines 1/2 one might restore 'A8ptavcd or 'A8pptdvela] I 'OA~rOgta.

tion,'Ev
cf.B[:
CIG,perhaps Ev
no. 3676,14. B[vLavrip,; for a Cyzicene agonistic inscription mentioning games at Byzan-
2/3: One could restore Kotvbv 'Aor]laS. For this game, sponsored by the Kolv6v of Asia and
sometimes held in Cyzicus, cf. a Cyzicene inscription (now in Paris), CIG, no. 3674,3-4: VEtxKlaaS
KOIV6V 'AaiaS iv KuliKp; cf. also Moretti, op. cit., no. 76, p. 223.

89 The lead cast is said to have been found along the shore several kilometers west of Kurgunlu.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 251

3: The
media. restoration of bv NE1KO[1&1Eia, is plausible, since there were games in neighboring Nico-
4: For games in Perinthos (Marmaraere'lisi), cf., e.g., Mordtmann, op. cit., no. 49,1-2: 'AK-
Tla Ev TTEpiv$Nc, 1-TTOia v rrEpiv9cO.
5/6: We know of games sponsored by the Kotv6v of Bithynia and held at Nicomedia: cf.
Inscriptiones Graecae ad Res Romanas Pertinentes, IV, no. 1645,3-5 (Kotvbv BE1iuvias v NEIKo-
riE8ia); Robert, CRAI (1970), 20,15 (Delphi: Kotdv BEtuvviaS Ev NEKOTpilEia); hence, one might
restore E[v NEiKOrlj8Eici KOiV6V or KOiva`] BEIQuviCas; however, one could also think of ?[v NEiKafI
KOIVaX] BE1Suvias, cf. Robert, in RPh, 3rd Ser., 4 (= 56) (1930), 35,21-23 -= idem, Opera minora
selecta, II, 1135 (inscription from Tralleis: NEIKaIV KOtVXa BEt$uviaS, KOiVw 'ACtfas).
6/7: The parallel is provided by CIG, no. 3676 = IGRRP, IV, no. 161, from Cyzicus: Tahcav-

[TiaiouS Kal] ES1ITC4aaVTIf1ous vE.K[4icras. Reference is to minor contests in which the prizes
amounted to one or one-half talent. On these contests, cf. Moretti, op. cit., no. 69,18-19 - p. 191
and pp. 195-96.
8: M: a minimum of forty victories of "one or one-half talent" category? Cf. PKA at the
end of the agonistic inscription from the Cyzicus area, Mordtmann in AM, 7 (1882), 255 = no. 26,
and Moretti, op. cit., no. 86,15 = p. 253 and p. 256 (Anazarbos: yCOvas Tccavataious Kai lptl-
TracavTiaiouS I.').
II. Funerary slab (sarcophagus?) (fig. 64)

Provenance: ca. 50 m. west of house of Lokman Kasar, presumably reused in a church as a
pier capital
Marble slab, cut at left and right; height (apparently original) 0.86 m.; maximum length
0.86 m.; maximum thickness 0.14 m.; height of letters 0.04-0.09 m.; interlinear spaces
0.05 m.

Letters and ligatures: , @, 0, *
Date: second-third century A.D. ?

] A
]XOV yEpouCacraT[as
?a]vuc Ki(al) T- &o'uv p[Tcp I yuvaIK

Trol]s U horl-rois d'rra[yopEr.o" Ei S, "ris "TroP4icE1
5 EirE]pov KaTaQSE'9ai [co'YEi T-r Talai),
]oixla * ,Bct

2: ]Xov is a genitive of the patronymic. For yEpovalacrTs, i.e., "local senator" in our area,
cf. CIG, no. 3687,2 (Erdek, near Cyzicus), no. 3749 (Nicaea).
3: For doayKpi-rToS in a funerary inscription, cf., e.g., IG, XIV, no. 1743, L. Moretti, Incriptiones
Graecae Urbis Romae, II (Rome, 1973), no. 499, and K. Wessel, Inscriptiones Graecae Christianae
Veteres Occidentis (Dissertation Halle, 1936), no. 168 (Rome); for the same as a laudatory epithet,
cf. CIG, no. 3493,12 (Thyateira); Mordtmann in AM, 12 (1887), 175 = no. 7,2 (Prusias ad Hy-
pium, Bithynia). Cf. also J. and L. Robert, BullEpigraphique (1948), no. 102; (1951), no. 219; (1954),
no. 285; (1959), no. 524.
4: For the formula TroiS 8E hXorrois cTrayopEOco, characteristic of funerary inscriptions from
the shores of the Marmara Sea, cf. B. Keil, "Ober kleinasiatische Grabinschriften," Hermes,
43 (1908), 523 note 1, and L. Robert, Etudes anatoliennes (Paris, 1937), 222 note 2.
6: ]9ia has no upsilon within the omicron (as it does in line 2 twice). If it had, one might
have restored "ri yEp]ouvia, "will pay the local council 2500 denarii." Cf. this formula in an
inscription from the Cyzicus peninsula, ed. F. W. Hasluck, "Inscriptions from the Cyzicus
Neighbourhood," Journal of Hellenic Studies, 24 (1904), 33, no. 47,4.

III. Funerarary slab (sarcophagus ?) (fig. 65)

Marble slab, broken on all sides, five bottom lines erased; height 0.73 m.; length 0.675 m.;
thickness 0.08 m.; height of letters 0.065-0.075 m.
Date: second-third century A.D. ?

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
252 CYRIL MANGO and IHOR SEV(ENKO
KaTEO-EO'a]icEV [aUTC-
K(al) -o], fTEKVO1S

-N-

IV. Catalogue of names (fig. 66)

Marble slab, broken on top and left; height 0.38 m.; length 0
height of letters 0.011 m.
Date: late Hellenistic ?
List of names in two columns:

ArlmnTpi9s.T9. ... iou
Aov'los OiOcpios Kpio-rros
'AWnroWA68copos BaKXfiov
'AWnroW'68copos ATrl.rlrpiou
5 'Airoaacvios MrlvoqpGv'ros
]ov TEIPI63Eos I'rrrovEiKouV
o]v M6oxos 'AXailoG
o]v BCi'Xios 'ATro7Xcovi8ou
Arllp7lTpos V-EPIpyEvoV
10 'AXcal6 MEvav8poV

For a similar catalogue from the gymnasium of Cyzicus, cf. Robert, Etudes anatoliennes, 199-
200, where the names 'ArroAh68copos and B&KX1to occur. Robert dates his inscription to the
Hellenistic period; here, however, we have a Roman name in line 2. For parellels to -ov in
f-Eptiyvov as a genitive ending, cf., e.g., T. Drew-Bear in Bulletin de Correspondance Helle'nique,
96 (1972), 468 and note 203. - If our list is connected with a gymnasium, the names may be
those of kolakretai (financial officials), also connected with gymnasia or games. For a similar
list of kolakretai of Cyzicus, cf. CIG, no. 3660. However, our names may have belonged to some
other association, e.g., a religious society, cf. Hasluck, in JHS, 24 (1904), no. 58 (Manastir near
Yenice).

V. Funerary stele (fig. 62)

Marble relief of a funerary banquet, complete; height 0.535 m.; length 0.40 m.

-aFTa -r'pcoro2p[Evous, -i~6ous, or -&yXov
XaipE

For the rare name Tacrr's, cf. a funerary inscription of unknown provenance, recorded in GIG,
no. 2143g (Addenda, p. 1017), as being in the Aegina Museum, and in IG, IV, no. 116, as "Aeginae
in demarchio": Auv0Trropts 'laTr XaTlpE. The name Avh0rrropti is Thracian, cf., e.g., Mihailov,
Inscriptiones Graecae in Bulgaria Repertae, III, 1 (Sofia, 1961), no. 967 (Plovdiv): AchErropts.
VaQS seems Greek.

VI. We reproduce without comment two parts of one slab (figs. 67-68), height 0.41 m., length
(combined) 1.28 m., thickness 0.095 m., height of letters 0.06-0.07 m., as the only inscription
from the Byzantine period in Lokman Bey's possession. Both parts of the slab have been re-
carved; the lettering is of crude quality. The only whole word legible is pvipa.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 253

3. Kurqunlu, Manastsr: Description of Site The local inhabitants call the mon-
The remains of this monastery lie aboutastery..."imperial" and ascribe its
foundation
one mile west of the village of Kurgunlu on a to a daughter of the Em-
peror Constantine
steep bluff overlooking a sandy beach (fig. 78). who, being afflicted
The monastery complex occupies part of withtheleprosy, was exiled to this wild
farm of Mehmet KSse. The ruins indicate a was healed by drinking from a
spot and
spring indicated to her by wild boar
prosperous foundation of the middle Byzan-
(youpo'vra)
tine period in the construction of which a that are still common in the
Karadag.
great deal of antique material was used. Accordingly, when the "Em-
peror
Hasluck, who travelled in these parts in Constantine," in gratitude for the
healing
1902-6, gives; the following account of the of his daughter, had constructed
monastery: a splendid monastery on the site of the
source, he directed that marble heads
I found at Kurshunlu not only a
of these youpoiivia be placed on the
Byzantine church with remains of a once
monastery's enclosure wall. Indeed, one
magnificent marble tessellated pave- acroterion that has survived in the
ment, a massive precinct wall on the
enclosure represents the head of an
seaward side and a ruined gateway of
animal, not a boar, but probably a lion,
some pretensions, but many ancient
and may have been brought from the
remains; these included several large
ruins of Cyzicus.
fragments of marble lions, which suggest
The enclosure of the monastery or,
that the monastery occupied the site of
to be more accurate, the seaward wall of
the temple of Placiane. At the same time
the terrace on which the monastery is
I should hesitate to place Placia at
situated, stretches along a considerable
Kurshunlu, which, lying under the
distance and is roughly 5 m. high. At
highest point of the Kara Dagh range,
intervals of three paces are buttresses
does not possess land enough for its own
two paces wide. On one of these but-
support, but lives by the export of char-
tresses is the afore-mentioned head [of
coal to Constantinople. It is much more
an animal] in rounded relief; on another
probable that Placia was at Yenije....
there was until recently the head of a
If the Panagia is indeed the successor
K6pTI, probably from some temple of
of Cybele, we find a curious repetition
Cyzicus. It has now vanished.
of history in the legend that the great
The monastery is situated on a
picture now preserved at the mon-
level part of a gently sloping spur of
astery of the Phaneromene in Kapu
ground between a stream and the sea.
Dagh90 was stolen from the monastery
It is nearly deserted and has a miserable
of Kurshunlu, to the great prejudice of
the latter, which is now in ruins, while the
little chapel built on part of the ruins
of the monastic church. There live here
picture brings a large revenue to its rival.91
one old monk with one or two workmen
Panienko (1910) gives a much more de- who till the rich land of the monas-
tailed account of the monastery which is
tery's fields-a charming, shady spot,
worth quoting almost in full:
abounding in spring water and suffi-
As we approach the village of ciently high above the sea to be free
Kur'umli, we see from afar, on the most from malaria.
level and convenient area of the slopes As one enters the monastery [Pan-
of the Karadag, the ruins of the mon- &enko enters it from northeast], one sees,
astery of Theophanes92.. .-r To Msy&?0ou first of all, a fountain of clear, but not
mineral water. The fountain, which is
90'Aypov "rfimonastery
On this ItyptavqS. ....
see Hasluck, Cyzicus square, is ancient, but has been restored.
(as in note 23 supra), 24ff.
Behind the square structure there branch
91 Ibid., 52f.
off thick walls whose direction and
92 Pan enko identified the monastery of
Kursunlu with MEyas 'Ayp6S of Theophanessignificance could easily be established
Confessor, cf. p. 263 infra. by means of superficial excavation.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
254 CYRIL MANGO and IHOR SEVCENKO
Above the arch of the fountain is inser- The former are semicircular on the
ted a small (0.32x0.16 m.) and badly outside, the latter is three-sided both
preserved flat relief representing a man inside and out.3 The internal width of
spearing an animal which, judging by the central apse from one pylon of the
the shape of its head, is a lion rather triumphal arch to the other is six
than a boar. There is no inscription. paces.
Next to the fountain several archi- From these brief indications it may
tectural fragments of marble are laid be seen that the plan and sections of
together under the trees: a double fluted the church of the monastery of Theo-
pilaster, the shaft of a column 0.38 m. phanes can be measured and delineated.
in diameter, part of a (chancel?) parapet Furthermore, nearly all the foundations
with a cross (0.31 m. high), and others. of the church are still preserved and
We may now turn to the ruins of the visible even without excavation. The
church. Not only all the foundations, new chapel is a miserable affair with a
but also its walls have been preserved wooden roof. On its west side a section for
to a height of 1 to 2 m. In the north women has been fenced off by means of
aisle the arches of the doors are still a simple partition. The sanctuary has
standing. The south aisle is buried been narrowed down and the walls
under a pile of earth and stones, over- whitewashed. The terminations of the
grown with bushes and trees, and is walls of the former diaconicon and
certainly preserved, perhaps in better prothesis stick out.
condition than the nave, which is occu- Inside the church stands a Roman-
pied by the chapel, and the north aisle, Corinthian capital, of the type called
which stands open. The walls are about Theodosian, exactly like the one in the
0.70 m. thick and are built of squared monastery of St. Anne." Its width from
limestone alternating with four or five the end of one volute to the other is
courses of brick. The bricks are on an 0.70m. Above the entrance, several
average 0.38 m. wide. Their length is not marble fragments are built into the
known since I have not encountered any wall, namely a double meander, a
complete specimens. They have no Roman ornament composed of palmet-
stamps, only parallel finger marks tes in the form of the [Cyrillic] letter (K,
running diagonally in crooked lines. and acanthus leaves in relief under an
The monastic church had three arc-like fillet. There is also a curious
aisles and two annex chapels, as indi- antique relief of good workmanship
cated by the existence of a wall parallel representing two busts: Hermes in a
to the north wall of the left-hand aisle brimmed cap holding a staff with an
which, on the east, is joined to the same entwined snake (the young god is turned
aisle at a distance equivalent to about to the left) and, to his right, the bust
one third of the latter's total length of a beardless man in a Phrygian cap
counting from its apse. Through this seen almost en lace ....
outer wall of the annex and the adjoining Inside the new church, i.e., in the
north aisle (now ruined and standing central nave, are preserved parts of a
open) an entrance has been broken Byzantine mosaic floor consisting exclu-
into the present chapel constructed on sively of geometric motifs executed in
the foundations of the central nave. The porphyry, green marble and a yellow
length of the north aisle is twenty-one and white composition. On the central
paces. There remain its west door and axis, approximately midway the length
in its outer (north) wall two big arches of the nave is a circle 1.80m. in di-
that were filled up in ancient times as ameter filled with parallel zigzags. It
well as one window near the apse. is inscribed in a square frame which
The foundations and, partly, the
93 As far as we were able to observe, the
walls of [the apses of] both lateral central apse is semicircular on the inside.
aisles and the nave have been preserved. 94 See infra, p. 272.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 255

does not touch its circumference. The fragments and other finds which are
resulting corners are filled with a geo- assured. A clandestine excavation in the
metric ornament consisting of the ten- monastic graveyard has recently pro-
drils of a half-palmette. The square was duced a gold pectoral cross of con-
framed by circles which are preserved siderable size, with a chased represen-
only in part. Inscribed in each circle is a tation of St. George (his name written
rhombus with concave sides; this con- vertically), embellished with one or
tains a second rhombus of similar shape, several emeralds. This information is
its corners touching the middle of each derived from the peasant who found
side of the larger rhombus. Another the cross and sold it to the local rich
fragment pertains to the frame of the man; the latter did not show it to me,
pavement near the north wall and being afraid, no doubt, of losing this
consists of little circles, 0.10 m. in sacred object.
diameter, filled with a checkerboard The western entrance gate (rr6p-ra)
ornament of triangles and eight-pointed of the monastery is almost completely
stars or rather rosettes with pointed preserved. This is important considering
leaves. the possibility that the monastery of
The pavement in that part of the Theophanes may have been built after
church where the women's vestibule the model of that of Studios. The
has been set up is different. It is an location of the gates of the latter is
interlace of the same basic type as in the known."9
floor of Imrahor.95 The details are The gateway is five paces broad
minute and complex, but, once again, and has on each side three niches capped
exclusively geometric. There were no by semicircular arches. The niches are
[figural] insets; at any rate, none have 0.70 m. wide, the spaces between them
survived. Next to the interlace is a are 0.45 m. wide, and the width of the
series of four bands of different colors. projecting walls on the western and
The remainnig parts of the pavement eastern sides of the gateway is 0.70 m.,
are well preserved and the colors of the so that the length of the entire passage-
stones are surprisingly bright. way formed by the gate is 4.50 m.97 The
Also worthy of attention is a height is approximately 3 m., but the
marble fragment, 0.24 x 0.22 m., con- ceiling and upper courses of stone have
sisting of four leaves, each having five not survived. The exterior face of the
lobes. These leaves are attached to a gateway, after a stretch of wall 1.35 m.
little circle at right angles to each other. wide, forms an angle 1.20 m. deep, and
Between them are inserted four more then is visible for another four or five
leaves, each having three lobes. This paces. In the wall on either side [of the
ornament is contained in a double gate], beyond the corners, are inserted
frame, which in the middle of one of the
reliefs taken from an ancient pagan
sides curves outward, forming a kind structure. The relief on the left side
of loop, of which only the beginning is
(it has fallen down and is lying in a
preserved. ditch) represents a head en face. The
In addition to the church, varioushair is rendered by means of two con-
foundations of monastic buildings, cov-centric semicircles connected by a series
ered with earth and overgrown, project of radii, which gives the impression of a
from the ground. The foundations of allnimbus with rays inside, as in Egyptian
the monastic buildings lie underground monuments. Under the head is a gar-
probably in their entirety. An exca- land of leaves, lightly delineated on the
vation would enable one to reconstitute
the plan of the monastery of Theo-96 Panienko's repeated references to St. John
of Studios are due to his involvement in the
phanes, not to mention architectural
exploration of that monument undertaken by
the Russian Archaeological Institute in 1908.
5 I.e., St. John of Studios. 9 Read 5.50 m.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
256 CYRIL MANGO and IHOR SEVCENKO
background. On the right is the faint as is often mentioned in Byzantine sources.99
outline of a figure with one wing visible. The two reliefs seen by Panienko on either
The relief on the right side of the gate side of the gate have disappeared.
represents a monster (Erichthonios?
a Giant?) with a human trunk and the
lower part of the body in the form of
a snake without wings ....98

In the more than sixty years that have
elapsed since Pancenko's visit many features
(in particular the mosaic floor and most of the
reliefs he describes) have disappeared, while
others have come to the fore. The outer enclo-
sure wall is still standing on the seaward side
along a stretch of some two hundred meters o 1 2 3 4 5m
(fig. 79). It follows the contours of the bluff
B. Kursunlu Monastery. Gateh
as far as the stream (to the northeast of the
church) and then turns inland (fig. 80). The
Within the original monastic
construction is of rubble, and, therefore,
one encounters the fountain which is still
difficult to date, but would appear to us to
be of the later rather than of the middle
flowing. It is a Byzantine brick construction
and is rectangular in shape, measuring
Byzantine period. As noted by Paneenko,
2.51 X 2.05 m. Inside is a tank, 1.20 x 1.12 m.
the wall presents a series of rounded but-
The arch and relief have vanished. Next to
tresses, about 1.85 m. wide, at intervals of
the fountain lies a marble sarcophagus lid
about 2.25 m. A stretch of similar wall may
also be seen some distance south of the (2.15 x0.74m.) of slightly gabled shape,
decorated with four crosses in low relief.
church. We were informed that a relief
One of the crosses is contained in a circular
representing the Virgin Mary holding the
medallion, another has a semicircular arch
Child lies buried under the sand of the beach,
over it (fig. 106). Before one reaches the
and is occasionally exposed by the waves.
fountain from the northeast, one passes a
We did not have the good fortune of seeing
spot, hidden by brambles, in which there are
it. Nor did we see a big relief representing
some remnants of vaults and a column shaft
"a snake" which, according to Mehmet
(diameter 0.35 m.); according to local infor-
K6se's father-in-law, once was to be seen on
the beach. mants, the place once was full of bones. It may
have been the ossuary of the monastery.
The gateway (figs. 81ff. and text fig. B)
The church (fig. 108), now largely over-
has been incorporated into Mehmet Kbse's
house and converted into a room. In the grown, is practically of the same size as that
of Pelekete, and exhibits the same form of
process, it has lost one of its six niches. The
construction, namely five consecutive courses
ceiling is modern. The original structure is
of brick (height of five bricks and five mortar
built of reused blocks and is, therefore, of
joints 0.60 m.) followed by four courses of
indeterminate date. Nevertheless, it provides
rubble (height 0.80 m.). The bricks measure
a good example of a monastic rrvu'cbv, such
0.32 to 0.38 by 0.04-0.05 m., and the mortar
98 IRAIK, 15 (1911), 267ff. In July 1897, joints 0.065 to 0.08 m. The latter are vertical
W. Judeich spent two days at Kursunlu, which and have been scored with a blunt implement.
he found to be a "purely Greek village of ca. In the zone of rubble there are both horizon-
thirty houses." He cursorily described the ruins
tal and vertical score marks in the mortar.
of the monastery and a number of antique
fragments (including the lions immured in the The most visible part of the church today
retaining wall). His notes, much shorter than is the east half of the north aisle. At the
those by Pandenko, usually overlap, but some-
center of the prothesis apse (fig. 113) is a
times complement, the latter. Cf. W. Judeich,
"Bericht iiber eine Reise im Nordwestlichen
9 Cf. A. K. Orlandos, MovaarptaKi &PXITEK-
Kleinasien....," Sitzungsberichte der Akademie TOVIKfi2 (Athens, 1958), 17ff. and esp. fig. 20 for
der Wiss., Berlin (1898, II), 552-55. a gateway with internal niches.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 257

between the two, and exhibits four courses
little cruciform sinking, such as is also found
in the north church of the monastery ofofLips brick followed by a good deal of rubble.
at Istanbul.100 Both prothesis and diaconiconSouth of the church is part of a high wall
had a little semicircular niche in their outer constructed entirely of rubble and, pre-
walls. In front of the prothesis is an arched sumably, of fairly late date. There is also a
recess, 2.17 m. long and 0.64 m. deep, per- cruciform fountain to the southeast of the
haps intended for a sarcophagus. church.
Originally, the dome of the church was All over the area of the monastery there
supported on four columns. Part of one are a great number of carved fragments,
shaft remains (diameter 0.51 m.), but it is Roman, early Byzantine, and middle Byzan-
not in situ. We have found the footing of tine. Some have been transported to neigh-
the northeast column (fig. 114), which is oc-boring houses, others have found their way
tagonal (0.77 m. across and 0.32 m. to the to the municipal building of Karacabey in
side) and has, next to it, tiny portions of the whose garden a small lapidarium has been
mosaic pavement. One column base remains installed, and reportedly also to the Bursa
(not in situ) in the southwest part of the museum. We shall mention here only a few
nave (fig. 115). It has an octagonal plinthof the more important pieces. First, there are
and was intended for a shaft whose lowerseveral fragments (now divided between the
monastery and Karacabey) of what must
diameter was 0.59 m. The capitals were not
Theodosian, as Panienko claims, but Roman.
once have been a very handsome "Asiatic"
One battered specimen remains in the churchsarcophagus with seated and standing
(fig. 119), another, well preserved, has been
figures and a podium bearing an elaborate
transported to the village of Kurgunlu and decoration of unusual form101 (figs. 90-97).
The dividing colonnettes appear to have
placed at the fountain in front of the mosque
(fig. 118). It is 0.66 m. high, 0.76 m. widehad Ionic capitals, if the fragment illus-
at the top (from volute to volute), and hastrated
a in fig. 96 belonged to the same
lower diameter of 0.50 m. The interior of the sarcophagus. We may also note two pieces
church was originally revetted with marble of the same coping (from a sarcophagus
as is proved by several fragments of serrated lid?) decorated with leaf-and-dart and other
dividers such as were normally used to frame motifs (figs. 98-99), a bench support deco-
marble panels. Of the furnishings of the church rated with the figure of a harpy (figs. 100-101),
we have found the lower part of one colon- two incomplete funerary reliefs (figs. 102-
nette pertaining to the templon (fig. 120). 103), a battered lion's head (fig. 105), part of
We have been able to locate the southwest a column shaft with spiral fluting (fig. 104),
corner of the narthex (fig. 112). Abutting on etc.

its west face is what appears to have been a The number of early Byzantine pieces is
staircase. Two treads are visible, but Mehmet comparatively small. In addition to the
K6se informed us that there were eight. He sarcophagus lid already mentioned, we may
also claims to have found many burials in note a piece of wall skirting (?) decorated
this area, including three tombs one on top
of the other. A funerary stele and the en- 101 The normal podium ornament of colum-
tablature with the feline, now at the house nar sarcophagi consists of a deeply-cut guilloche
(sometimes alternating with sections of meander
of Stileyman K6se, Mehmet's son (cf. p. 258 and garland) and Lesbian cymatium, as on the
infra), are said to have been found at this sarcophagi of Melfi, Palazzo Torlonia (Rome G),
spot, "at a depth of 4 m." etc. See C. R. Morey, The Sarcophagus of
Claudia Antonia Sabina and the Asiatic Sarco-
Built against the north wall of the church
phagi, Sardis, V/1 (Princeton, 1924), figs. 39-41,
is a chapel having a fairly large semicircular 83-84, 92; M. Lawrence, "Additional Asiatic
apse and, within it, traces of a synthronon. Sarcophagi," Memoirs of the American Academy
The chapel is later than the church, as one in Rome, 20 (1951), figs. 31, 42; G. Ferrari,
may ascertain at the point of juncture II commercio dei sarcofagi asiatici (Rome,
1966), pls. 1.1, 3.1-2, 3.6; H. Wiegartz, Klein-
asiatische Sdulensarkophage (Berlin, 1965), pls.
100 See Th. Macridy, "The Monastery of 5b, 27d, 29a, 32b. We do not know of any
Lips ..," DOP, 18 (1964), 260 and figs. 6, 25, 26. columnar sarcophagus with Ionic capitals.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
258 CYRIL MANGO and IHOR SEVCENKO
with a meander (fig. 125), perhaps the same
44:2, ['H wn1 aorT-r KeaEK IvLEVo T Eo'raTCt, OoK
piece that was seen by Pantenko. More
numerous are the remains of the middle
avolxSiloeTat Kai o]v8eis [sic] 090 pil tX9r1i
St' aO"ris (fig. 131), a text traditionally asso-
Byzantine period. The largest of them ciated is a with the Virgin Mary.'04 The lettering
complete, but lidless sarcophagus (length is black on a white ground and each line is
2.07 m., height 0.72 m., depth 0.75 m.), separated from the next by two ruled in-
whose front is decorated with three rather cisions. The fragment probably belongs to a
crudely incised crosses (fig. 107). The centralscroll held by the prophet Ezekiel. Palaeo-
cross, which is the largest, stands on a graphically, the inscription has to be dated
vestigial orb, and its bottom arm is deco- in the thirteenth or fourteenth century, at
rated with a series of diamonds placed endwhich time, therefore, the church must have
to end.102 Today this sarcophagus is used as a been repainted.
fountain and has been placed in front of Finally, a few hundred meters southwest
Mehmet K6se's house. In the neighboring of the monastery, walking uphill, one en-
house of Stileyman K6se is part of an enta-counters the foundations of an apsed chapel
blature (height 0.38 m., width 0.32 m.)or church, completely overgrown with bram-
decorated on its underside with a square bles (fig. 135). Lying inside it is part of the
panel containing a feline attacking a donkey curved parapet of an ambo of approximately
or a hare (?) (fig. 124). There is a series ofsixth-century date (fig. 134). Close by we
drill holes along the feline's back, and thehave found part of a crudely carved mullion
eyes of both animals are likewise marked capital (fig. 133) and, rather surprisingly, a
by drill holes. To the right of the panel is piece
a of a porphyry column.
square sinking which accommodated the top Our epigraphical gleanings at the farm of
of a post or column. A small, slightly curved Mehmet K6se were meager. In addition to
marble element, built into the south wall of the eighth-tenth-century letters A and Q
Mehmet Kbse's house has the letters A and f9 and the thirteenth- or fourteenth-century
carved in relief (fig. 88). Two fragments of inscribed fresco fragment already discussed,
marble cornice decorated with crosses and we found only three inscriptions there in
palmettes (figs. 121, 122) recall similar work1971. The first of them was a loose fragment
at the monastery of Lips at Istanbul.103kept near the entrance to the house; it was
There are also a number of pieces of in-still intact in 1972. The remaining inscriptions
determinate destination decorated with a are in fixed positions, the second forming the
rinceau in low relief (figs. 126, 127, 130). threshhold of the back entrance, the third
Of greater interest is a small piece of being immured in the back wall of the
painted plaster bearing the text of Ezek. house.

I. Fragment of a funerary inscription (fig. 86)

Height 0.185 m.; length 0.23 m.; height of letters 0.025 m.
Date: Late Empire?

rrr6]vqa
ymKEX]apayA[v - -
KAYIHN
NOE TO

102 The crudeness of the carving and the 103 Cf. C. Mango and E. J. W. Hawkins, "The
"degeneration" of the orb suggest to us thatMonastery of Lips.. ..Additional Notes," DOP,
this is a Middle rather than an Early Byzantine 18 (1964), 307 and fig. 15.
work. On this topic, see O. Feld, "Mittel- 104 Cf. G. Babi6, "L'image symbolique de
byzantinische Sarkophage," Rimische Quartal- la 'porte ferm6e' A Saint-Cl6ment d'Ohrid,"
schrift, 65 (1970), 158ff. Synthronon (Paris, 1968), 145ff.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 259

1: The term OTr6pvlrpa, "funerary monument," is characteristic of the Cyzicus
regions, cf., e.g., CIG, nos. 3690,1, 3692,3, 3693,1; H. G. Lolling in AM, 9 (1884
Hasluck, Cyzicus ... (Cambridge, 1910), 242-43; idem in JHS, 24 (1904), 20-32
29, 30, 40, 42; and J. Kubiniska, Les monuments fundraires dans les inscriptions g
Mineure (Warsaw, 1968), 23 (examples from second-third century and furthe
For nTr6i'vr1pa, "tomb," in Christian inscriptions, cf., e.g., CIG, no. 9257.

2: For EyKEX]apayji[v - -, cf. Kubiniska, op. cit., 15: 8ta& rT)s itypayiS -r
I'vrs Ti5 o op~j (Smyrna).

II. Fragment of a sarcophagus slab (fig. 87)

Length 0.76 m.; height 0.65 m.; height of letters 0.045-0.09 m.
Date: Early Byzantine

v[irn6ovilva
11a[OAou

Bibliography: F. W. Hasluck, "Inscriptions from the Cyzicus Neighbourhood,"
36 (fragment in facsimile). Hasluck saw a sarcophagus slab 1.75 x 0.83 m., whic
the door." He read:

YHOMNHMA
ITAYAOY PA..
Thus, in almost seventy years, our inscription has remained in very nearly the same spot, but
has been reduced to three letters.

III. Lower left side of a tabula ansata (fig. 89)

Height 0.30 m.; length 0.57 m.; height of letters 0.025-0.035 m.
Date: sixth century?

.U[
Y[
O[

5 o[
AP[

We found no trace of the inscriptions nos. 54 (a sarcophagus of Aur. Asclepiades) and 56 (a
fragment with the name Aur. Maior), published by Hasluck in JHS, 24 (1904), 35-36, or of the
inscription, probably containing the iTrr6p~pa formula, published by W. Judeich, "Bericht
uiber eine Reise im Nordwestlichen Kleinasien ...," Sitzungsberichte der Akademie der Wiss.,
Berlin (1898, II), 552-55.

4. Kurpunlu, Manaster: Megas Agros or owned by Mehmet Kdse. A local Greek
Polichnion? tradition identified this complex with the
monastery -ri iTrepayicaS SeoT6KOU T -r i-
We shall now discuss the Byzantine name yptavils or TroL MEy&6ou 'Aypovi, of which
of the complex west of Kurgunlu, presentlyTheophanes Confessor (d. 818) was the first,

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
260 CYRIL MANGO and IHOR ?EVCENKO
and the only known,104a abbot. However, them
we is the long Vita written by Patriarch
Methodius (d. 847); several other Vitae are
shall presently see that this tradition cannot
be traced back beyond the seventies ofderived the from it. Their dependence extends
last century. Moreover, another establish- from literal borrowings017 to the excerpting
ment which our sources situate close by of essential features (including factual in-
Megas Agros may have been the predecessor formation)-the latter is true of texts re-
of the present Kurgunlu complex. This is presenting the Vita which was included in the
Polichnion, also known as Polychronia, or so-called "Imperial Menologium."108 In addi-
in Slavic Polihron, a monastery connected tion to the Vitae, we have two independent
with the names of Theophanes Confessor and witnesses: the Panegyric on Theophanes,
of Saint Methodius, apostle of the Slavs.105 which, if it is by Theodore of Studios (d. 826),
We therefore must examine the relevant is the earliest extant source on the Saint,101
texts and first report what they have to say and a letter by Michael Psellos (d. 1078), de-
on the respective locations of Megas Agros scribing his voyage along the coast of the Sea
and Polichnion or Polychronia. of Marmara and his landing at or near Agros.110
Various Vitae and laudations of Theo- Pieced together, the topographical passages
phanes are our main source of topographical in these texts give the following picture:
information.106 The earliest and the best of 1) Megas Agros, or simply Agros,111 was
104a This statement may need modification 107 This is the case of the Vita printed in C.
in light of the recent publication of a Dumbarton de Boor, Theophanis Chronographia, II (Leipzig,
Oaks seal mentioning "Christopher, monk and 1885), 28-30; example: 28,24-25 repeats the
abbot of Agros," cf. V. Laurent, Le corpus des Vita by Methodius (=BHG3, no. 1787z), 11,3.
sceaux de l'Empire byzantin, V,3, Supplement In subsequent notes, this latter Vita will be
(Paris, 1972), no. 1938. The editor assigns the referred to as VM; other Vitae will usually be
seal to the tenth century. He sees in it not quoted by the BHG3 number and the page of
a regular seal, but an expression of veneration the edition described in that number.
which the monks of Agros showed for Chris- 108 BHG3, no. 1791; cf. also no. 1788. For a
topher, the mentor of Theophanes, in whose discussion of various Vitae, cf. also BHG3,
monastery the future Confessor had spent some no. 1792, 585-96.
time prior to the founding of Agros. Chris- 109 BHG3, no. 1792b=Ch. Van de Vorst,
topher's monastery, as we shall presently see, "Un panegyrique de S. Theophane le Chrono-
was contiguous with that of Agros, and, accord- graphe par S. Th6odore Studite," AnalBoll, 31
ing to the testimony of Metaphrastes, was called (1912), 11-23. The identity of the author is
"the little Agros." Laurent's is an enticing, if established mainly on stylistic grounds. It is
complicated, exegesis; however, Metaphrastes' conceivable that the Vita by Methodius bor-
testimony, on which everything hinges, seems rowed from the Panegyric, cf. Panegyric, 23,7-8
unreliable to us. Christopher of the eighth and Vita, 3,12 and 4,4-5, where the question
century could not have been an actual abbot of the Saint's baptismal name (Theophanes
of Agros, because in his time Agros either did rather than Isaac) is treated in the same way.
not exist or was led by Theophanes. On the On the other hand, the Panegyric, 21,31-36
other hand, if the Agros of the seal is our differs from all the other sources in having
Agros (rather than a monastery of Cyprus or Theophanes tonsured on the island of Kalonymos
Asia Minor, cf. note 111 infra), and if the seal rather than at Polichnion. If the Panegyric is
is of the tenth century, then it may refer to an by Theodore, this would be a considerable slip.
actual tenth-century abbot of Agros by the name 110 Eps. 12-15 and 140, ed. Kurtz-Drexl (as
of Christopher. Father Laurent himself appears in note 5 supra), 13-19, 167-68.
to admit the existence of a tenth-century abbot 111 So VM, 17,2-3, 38,25-26, and BHG3, no.
Christopher of Agros, "otherwise unknown." 1791, ed. Krumbacher, 398,19. - There exis-
105 Menthon, L'Olympe (as in note 45 supra), ted at least two other monasteries by the name
second map at the end of the book, locates of Agros or Megas Agros: (1) in Cyprus (pos-
Polichnion just west of Kursunlu-that is, at sibly two: near Nicosia and in the diocese of
our complex-but he puts "Polychronion" Kition), and (2) on Mt. Latros, near Miletus,
(which is surely the same place) some distance where Agros occurs jointly with Stylos. This
further west. Unfortunately, the author does must be kept in mind in any attempts to
not explain these localizations, nor does he reconstruct the monastic library of our Agros
indicate his source for the date of the foundation in the Hellespont. Cf. J. Darrouzbs in Revue
of Polichnion (p. 198: "about 779"). des Etudes Byzantines, 8 (1951), 177, and 15
106 For a list of the Vitae of Theophanes, (1957), 135-36 (concerning Parisinus Graecus
cf. F. Halkin, Bibliotheca Hagiographica Graeca3, 533, Athens, National Library, codex 842, and
nos. 1787z-1792e.
Jerusalem, St. Sabas, codex 259).

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 261

that fish some years ago;"" the name
situated in a picturesque spot in the foothills
of the Sigriane (Karada') mountains,112probably refers to the sturgeon, for, according
a range extending along the southern coast
to the eleventh-century author Symeon Seth,
of the Marmara Sea from the Rhyndacus the hyska's flesh has the same nutritive value
(Karadere) river in the direction asof the
pork, and the taste of pork is attributed
Cyzicus peninsula. to some parts of the sturgeon in a modern
2) It was at or near the seashore. This can
encyclopedia.117 If hyska was a sturgeon, one
be shown in several ways: Agros had awould
harbor
understand why the seashore near Agros
of its own;113 Saint Theophanes was traveled
unlikely to yield this surprising catch.
between Agros and Polichnion (which 3) was
The Agros property was contiguous
surely on the seashore) by sea;114 monks of
with that of the monastery of Father Chri-
stophoros,
Agros did not have to go far to catch fish, where Theophanes dwelt for
as we learn from the following story,awhile
trans-after his return to the mainland from
mitted in the Vita by Methodius: the island of Kalonymos (Imrah Adasi).118
"On one occasion, when Theophanes 4) was
Agros was some distance west of Medi-
visited by some fathers or rather brethren,
kion. Medikion, we may remember, was also
he welcomed them with generous handknownand
as the monastery of the Holy Fathers
soul, opening the soul through the hand (as
(of Nicaea).'19 When Psellos traveled by
for the hand, he always kept it open).boat to "the Holy Fathers"-that is, was
Forth-
on the way "up" to Medikion, his chari-
with he sent for fish, and had a large quantity
stike120-he kept "our" Agros "opposite and
of many sorts of them brought. He magnani-
mously told his guests: 'Lo and behold,
to the right"; originally, he intended to sail
by,forth
blessed fathers, your prayers did bring for he hoped that Agros would receive
a luxurious fare.' One of the guestshim on his
gave a way "down," that is, on his
quick and witty repartee, 'I do notreturn
wishvoyage.121
to Whether we assume that
partake of these. I would rather have a
hyska.' When he who was firm in his11" "Les Kordakia de Nic6e, le combustible
hopes
de Synnada et les poissons-scies. Sur des lettres
and farsighted in his faith and perfect in his
d'un m6tropolite de Phrygie au Xe sikcle.
chastity heard this, he right awayPhilologie ordered et r6alit6s," Journal des Savants
one of his disciples to go and bring a (January-June 1962), esp. pp. 62-67.
hyska, and added, '(Let it be) on account of 117 Symeonis Sethi Syntagma de alimen-
the fathers, since they so desire....' The torum facultatibus, ed. B. Langkavel (Leipzig,
1868), 111,14-19; D. N. Papagiannopoulos, s.v.
disciple who had been dispatched left, and
brought him an enormous hyska, as he had 'AKt-rrrljfov, in MEy&xc1] EN7NV1iKia 'EyKVKXOWiai8E10a,
3 (1927 ?), p. 108, col. 1.
been ordered to do. This (was accomplished) 118 VM, 16,22-17,6 (cf. the word 6popo0v-racs,
in spite of the fact that the place was al- 17,5).
together unsuitable for the catch of fish of 119 Cf. p. 240 and note 37 supra. Psellos, too,
this kind; for they caught fish not from some knew that Medikion had another name: Mr81(-
spot opposite [on the opposite shore of the KtOV... 6v6pa-roS iv Trap& "rroXXoTsioi . t~pvov, iol
84 A&vcbvvuov 1rTv-nr 8OKET, Ei peh rI Mr8GifKOV 'rtco-
Marmara Sea?], but from their monastery v6poa-ra, Ep. 140, ed. Kurtz-Drexl, 167,10-13.
proper, that is, from the seashore."115 The 120 Cf., e.g., Ahrweiler, "Charisticariat" (as
in note 37 supra), esp. p. 25.
identity of hyska has not yet been estab-
121 The crucial phrases in Ep. 13, ed. Kurtz-
lished, in spite of the learned excursus Drexl, 15,21-24 and 16,6-8, are: (a) TropS~Eiov
which Professor Louis Robert devoted to
8 TniPrrlhs -r 'v Els robs 'Ayiovus ITa-rpac [the
editors do not capitalize these words] 686v
112 Cf. the Panegyric ascribed to Theodore crEXXA6~rjv vayoVivcp ydp i.pot irv r' KEIVOVU
of Studios (as in note 109 supra), 22,6-8, where
'Ayp6 cvaixas tfrrl S~Etd& Tp'rEpos, 8 8o K:arax-
the area of Megas Agros is called rrrc6bpaEiv Tiva yopEvovus TrflE-ro 8tia8lEacSa, and (b) c0s 5rfiS
rtoo ~iyplaviKOi 6povS.
E*0IS TC-Ov T-a-ripcov K6<XaSopvovu sTr i'6 ro To Trrpot-
113 'E-rivEov, vecbpiov, Psellos, Ep. 13, ed. prnivov 'Aypoi irrfVivov Kac-rpact... itrrefyESat. -
Kurtz-Drexl, 16,7-8. Psellos traveled toward the end of May, for he
114 VM, 25,18-25; cf. BHG3, no. 1791, ed. sailed on a Sunday on which the Feast of the
Krumbacher, 395,28-396,2. Holy Fathers was celebrated, cf. ibid., p.
115 VM, 24,17-32; cf. BHG3, no. 1791, ed. 16,26-27. The usual date of the Feast of the
Krumbacher, 395,17-23. Fathers of Nicaea is May 29.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
262 CYRIL MANGO and IHOR ?EV?ENKO
In only two respects does the identi-
Psellos was traveling to Trigleia-Medikion
from Constantinople or, as we prefer to of Megas Agros with the ruins at
fication
believe, from the area of Cyzicus, theKurqunlu
ex- present some difficulty. In the
pressions "on the right" and "expectedfirst to
place, the term pyas &yp6s presup-
receive us on the way down" put Agrosposes
westa large stretch of tillable land, while
of Medikion.122 the arable area behind Mehmet K6se's farm
is of modest dimensions. In the second
5) Finally, Agros lay six semeia from
Polichnion-Polychronia,1x and twelve se- the Kurgunlu ruins are right on the
place,
sea-the
meia from its proasteion at Hiereia124 (a church itself can be reached from
semeion being roughly a mile125). the beach along an ascending path no more
Thus, the information which the texts than two hundred meters long-while there
give us about Megas Agros agrees with arethe
various indications to the effect that
location of the present complex of Megas ruins Agros may have been some distance
west of Kurgunlu: that complex is on the from the shore. When Leo V's henchmen
sea
(see point 2 supra); it lies west of Medikion arrived to summon Theophanes to Constanti-
(see point 4 supra); and it is near the remains nople (814), the Saint, gravely afflicted with
of a church, situated a short distance higher kidney stones, had to be taken to the shore in
up on the slope in a southwesterly direction. a covered carriage, and put on a boat there.126
These remains would then be the remnants The Vita by Methodius seems to imply that
of the monastery of St. Christophoros (see the procession which escorted Theophanes'
point 3 supra). body from Hiereia to Agros twelve miles
away made this journey by land.12' Finally,
122 If "The Holy Fathers" and Medikion on his turbulent voyage to Medikion, Psellos
are the same, then the itinerary as described in
Ep. 13, ed. Kurtz-Drexl, 14-17, is compatible first disembarked at the harbor of Agros,
with the assumption that Cyzicus, i.e., most and only then ascended to Agros.12s
probably Artake (Erdek), was its initial and Admittedly, these difficulties are not
terminal point. The party used a small vessel serious. What seems to us a skimpy area of
holding "over" twelve passengers and a crew
tillable land may have appeared large to
of three; the initial route was along the shore
(to round the peninsula of Cyzicus, we submit); Theophanes' contemporaries; and Theo-
afterward the boat moved onto "high seas"; phanes' attendants may have had to use a
Agros was to the right; the plan was to go to cart for a distance of a few hundred meters of
Medikion first and to land at Agros on the way steep ground, in order to alleviate the ailing
back; however, when contrary wind and rain
started, the party "forgot" about Medikion and Saint's discomfort; Methodius' passage is
landed at Agros, apparently without ever vague; and Psellos does not indicate the
getting to Medikion; they started the return distance between Agros and its harbor.
voyage from Agros; they walked on dry land We come now to Polichnion or Poly-
after another landing on the way back from
chronia, the second possible Byzantine
Agros-if the point of destination had been
Constantinople, such a walk would not be antecedent of the Kurgunlu complex. This
easy to explain. The city in which both Psellos monastery was located close by Agros, being,
and his addressee lived at that time (cf. Eps. 13 as we just saw, six semeia distant from it;
and 14, pp. 15,20 and 18,10-16) was a coastal it, too, lay along the Sigriane range;129
city, but not necessarily Constantinople: "on
both sides" it had groves, gardens, and meadows and it was on the sea. The latter may be
(Ep. 14, pp. 17,26-27 and 18,2). If Psellos did inferred, for instance, from the following
travel from Constantinople, his intended itin- miracle wrought by Theophanes and reported
erary was: the capital-Medikion-Agros-the ca-
pital; however, having been blown off course 126 VM, 29,27-29 (Xapmwifv nPASiS...dpdgeS
(but how could a contrary wind do that to pIrcrreSES E5I rr7poi&piv). Cf. BHG3, no. 1791,
someone proceeding from the north ?), he landed
ed. Krumbacher, 396,20-22 (?E' d &rs T-rSe-rat
at Agros and thence returned to the capital. ... KdV TcO alytia KcTOCr&yE-ra).
123 VM, 25,19. 127 VM, 38,19-27: r.... .KP K6iV Kal UStva-
124 VM, 37,11-14.
6rwTCV Kait lJcapccpic KVKkOT-rEpr' TOo wi'ljSouvs
125 After some discussion, E. Schilbach,
ca(pXovros p-rijpav a*rr6v EiS -nTv wrrap' a*iroG
Byzantinische Metrologie... (Munich, 1970), KITmOEaav Ov1'v r6v 'Ayp6v.
32-36, opts for the value of 1574.16 meters for 128 Ep. 13, ed. Kurtz-Drexl, 16,7-10.
the uiXtov or onlaelov. 129 VM, 25,18-19, 15,20-21.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 263

in Methodius' Vita: one day, the Saint identification
visited by Nikodemos of Cyzicus, but
Polichnion, but was not able to return on curiously enough he credited himself with
account of a prolonged storm which pre- being the first to connect the ruins with
vented anyone from landing (Trpoo-rrs'cat)i) Theophanes' monastery.133
at Polichnion. Theophanes sent out one of Neither Nikodemos nor Euangelides
his disciples and had him order the sea to offered evidence in support of his identi-
calm down. The sea obeyed, and the Saint fication.Y34 Hasluck, who visited Kurgunlu
returned by boat to Agros.130 sometime between 1902 and 1906, took the
Thus, on the strength of evidence offered attribution for granted-incidentally, as
by the texts alone, the ruins at Kurgunlu we already saw, he found the monastery in
could be the former site of either Megas ruins.135 The Greek Calendar ('HsEpoh6yIov)
Agros or Polichnion-Polychronia. Nor does of the National Philanthropic Establish-
the size of the present complex provide a ments for 1905 put Megas Agros in the
decisive argument in favor of one or the vicinity of Kurgunlu, which it described
other monastery. To be sure, since Agros as a partially Greek village with a Greek
was the main monastery of Theophanes, school and church.136 Panienko, who sur-
where the Saint had a church built for his
veyed the site in 1910, had no doubt as to
community, it probably was bigger than the
its identity with Megas Agros on account of
monastic enterprises which stood on his
"the very name of the monastery"-but
what was the authority for that name ?-
other proasteia, such as Polichnion and Hie-
reia; and the Kurgunlu complex is the mostand "the texts of the Vitae, describing
Theophanes' journey to Cyzicus"-but the
extensive ruin to be found today between the
mouth of the Rhyndacus and the village of Vitae, which we already discussed, allow
Yenice to the west. However, we shall see only the conclusion that the monastery was
that by 861-63 Polichnion-Polychronia had situated somewhere in the area. Pancenko
seventy monks-and thus was big enough to explicitly stated that "nobody had estab-
have occupied the site of the present ruin.
lished the location of (Theophanes') monas-
To decide, we are left with the Greek tery in print." This observation is impor-
tradition to which we alluded at the be- tant; it shows that he was unaware of Niko-
ginning of the present section, and we demos
must of Cyzicus, of Euangelides, and of the
examine it briefly. To our knowledge, Calendar of 1905. Consequently, when Pan-
identification of the monastery near Kur-
cenko spoke of the name of the monastery,
he relied on a different, oral tradition, which
?unlu with Megas Agros was first proposed
he had learned either from the Christian
in print in 1876, when Nikodemos, metro-
inhabitants of Kurgunlu (who traced the
politan of Cyzicus, published an Akolouthia
in honor of Aimilianos, his predecessor ofmonastery's
the origins back to the time of
iconoclast period, and prefaced his work
Constantine the Great) or from the single
monk
with remarks on the history of the diocese of who resided in the ruined monastery
Cyzicus.131 In 1892 Euangelides visitedatour
the time of his visit.137 Hasluck, too, may
site and left a cursory description of it; he, learned the name Megas Agros on the
have
too, called it Iov'i *rs XItypiavfi MEy(hkou
spot, although he quoted the work of Niko-
'Aypoii.ls Euangelides was aware of the
Theologische Literaturzeitung, 22 (1897), esp.
130 VM, 25,18-25. cols. 309-10.

131 'AKoXovuita TroO &v &yiol TrcTp6s i$lpoOV133 Compare Euangelides, 'H povi Tri~ Itypia-
viis, 12, with 11 and 13.
AiPthavo0 -Tiiaxo6irov KUiKov roV T 6pooyyl-roi...
-rrr6 -ro0 MT-rporrolT-rov KvlfKov NtIKOS OV (Con-184 Nikodemos, 'AKoXovf9ca, says only that the
stantinople, 1876), esp. p. ss': the monasterymonastery's
of ruins bear witness to its past
Megas Agros is situated "two hours <on footglories.
?>
north [sic] of the mouth of the Rhyndacus,
135 Cyzicus (as in note 23 supra), 52-53.
near the Christian village of Kurgunlu." Cf. p. 253 and note 91 supra.
132 T. Euangelides, 'H ov TIs-i Xtypiavis " Troi 136 'E.SVIK& (y1Xov.gpCOWtK KO7CT'O1TO cp VtiTaT Av Kcov-
MEydXovu 'AypoG (Athens, 1895), esp. pp. 11-13. aTVvrTvov-r6X7st, 'HIsepoX6ytov TroO v-rovs 1905
Uncritical summary of Euangelides' pamphlet (Istanbul, 1904), 181-82.
in the review of BHG3, no. 1792, by J. Draseke,137 IRA IK, 15 (1911), 267-68.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
264 CYRIL MANGO and IHOR ?EVCENKO

demos of Cyzicus in the bibliographyhaving previously distributed his earthly
of his
book.138 goods among the needy-and was able to
We have no assurance that the oral borrow two and one-half pounds of gold to
tradition which we postulate as Panienko's cover the price of purchase.141 In order to
source in 1910 was independent and au- recruit monks for the new foundation,
thentic, as it may have been itself derived Theophanes made a trip to Pandemos, the
from the learned guesses which Nikodemos Hellespont, and Bithynia.142 On the site of
and Euangelides had committed to print. Agros, he had a church built for the com-
In spite of these caveats, we provisionally munity.143
identify the complex at Kurgunlu with the Except for occasional absences, either of
Megas Agros of Theophanes Confessor. We an official nature, such as his journey to the
turn now to the history of that monastery.139 Second Council of Nicaea (787), or of a
Megas Agros was founded by Theophanes, private character, such as his trip to some
who acquired the land for it shortly before hot springs (probably those of Bursa or
787;140 by that time, he was thoroughly Yalova),144 Theophanes stayed in the Agros
familiar with the area. He had been entrusted
with supervising (and financing) the building 141 Cf. VM, 11,1-4, 11-21; 15,20-22; 16,3-16,
of fortifications at nearby Cyzicus by Leo IV23-28; 17,7-14. The price of two and one-half
pounds of gold does not seem exorbitant.
(d. 780); during his stay at Cyzicus, he had
About 1050, Constantine IX gave St. Lazarus
traveled to Sigriane and discussed his plans of Galesion 720 nomismata, or ten pounds of
for retiring from the world with the local gold (true, in all probability, in the debased
anchorite Germanos. About 781, he was currency of eigtheen carats of his reign), "from
tonsured at Polichnion, which had been his which nomismata the major part of the con-
struction of Pausolype [a monastic complex]
property. From there, he moved to the was covered"; cf. the Vita of Lazarus the
island of Kalonymos opposite the mouth of Galesiote, ActaSS, November, III, 584E.
the Rhyndacus, where he remained as a 142 VM, 19,4-9. Pandemos is mentioned in
simple calligrapher-monk in the monastery Vita Ioannicii by Sabas, BHG3, no. 935, 341B
which he himself had founded on a piece and 344B; Sabas locates it in the Hellespont.
The Eriste monastery was at Pandemos.
of land inherited from his father. Six years 143 VM, 21,7. Some of the reused elements
later he returned to Sigriane and joined the traceable to the church at Mehmet K6se's
monastery of Father Christophoros. Agros, farm, such as the large capitals (both those at
we recall, was a piece of property adjoining the farm, either whole or in fragments, an
that monastery; when it was put up for sale, that at the fountain of Kursunlu, cf. p. 25
and fig. 118 supra), are likely to be spolia from
Theophanes overcame his financial diffi- Cyzicus. Theophanes, the former supervisor o
culties-he was no longer a wealthy man, public works at Cyzicus, could very well have
ordered antique material from that city to be
138 Cyzicus, 318. brought to his new church. On spolia from
139 The best treatment to date of the history Cyzicus at the St. Anne monastery near
of Agros and of its location is by J. Pargoire, Yenicek6y, cf. Pan'enko, IRAIK, 15 (1911),
"Saint Theophane le Chronographe et ses rap- 265; on the same in Constantinople, cf., e.g.,
ports avec Saint Th6odore Studite," VizVrem, C. Mango and E. J. W. Hawkins, "Additional
9 (1902), 31-102, especially sect. VI: "PositionFinds at Fenari Isa Camii, Istanbul," DOP, 22
du monastbre de Th6ophane," pp. 42-49, and (1968), 182.
sect. XVI: "Tombeau de Th6ophane et son 144Nicaea: e.g., VM, 18,22-19,4; BHG3,
monastbre de Grand Champ," pp. 91-95. Theno. 1792, 612,15-24. Neither Theophanes' name
present sketch is much indebted to Pargoire's nor that of Agros occurs among the signatories
article. - For some good, if succinct, remarks of the Council of 787. - Trip to hot springs
on the location of Megas Agros, cf. also Toma-for reasons of health: VM, 22,15-16. In the
schek, "Topographie" (as in note 4 supra), 14.forties of the tenth century, Theodore, metro-
140 Cf. VM, 17,1-19,9, where the sequence ofpolitan of Cyzicus-a locality close to Megas
events is (a) purchase of ground for Megas Agros-visited the thermal establishments
Agros, (b) participation of Theophanes in theof Pythia-Yalova; cf. J. Darrouzes, Epistoliers
Second Council of Nicaea, (c) Theophanes' byzantins du Xe sikcle (Paris, 1960), 326,17-21.
trip to Pandemos, the Hellespont, and Bithynia Two circumstances qualify the value of this
to recruit monks for his monastery. Little time visit as a parallel to Theophanes' trip: (1) the
must have elapsed between (a) and (c), and context of Theodore's letter indicates that
a fortiori between (a) and (b). he went there on his way back from Con-

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 265

monastery, of which he became abbot, until
816, since the monastery of Sergius and
814. In that year, Leo V commanded abbots
Bacchus, in which he was for a while con-
of important monasteries to come to Con-
fined, could have provided him with the
stantinople, and Theophanes, an ailing man
sources necessary for the compilation of the
of fifty-four, was one of those summoned.145
Chronicle, and later on he could have worked
Theophanes was never to see Agros again, in the prison of the palace of Eleutherios.
though his body was returned there However, in perusal of the last pages of the
triumph in 819, after it had lain in state and
Chronicle leaves no doubt that Theophanes
wrought miracles in nearby Hiereia for wroteone it before his imprisonment by Leo V.
year. He was buried at Agros in a monument In these last pages, Leo-then still a mere
patrician-is called "pious"; this usurper's
"on the right side of the church," in the spot
proclamation as emperor is "most lawful";
where he used to sit while singing psalms.146
His grave became a pilgrims' goal and a Theophanes reports the iconodulic
and
source of miraculous cures from the very
Patriarch Nikephoros' advice that Michael I
beginning. Theodore of Studios seemsshould
to abdicate in favor of Leo.149 Such
have visited Agros-unfortunately, welanguage
do could not have been used by a
not know at what date;147 as early as 825,
martyr of faith detained at Constantinople
Saint Ioannikios came there to worship
after Leo V had shown his true colors.
before Theophanes' tomb and relics "gushing The Chronicle, then, was finished before
with miracles."14
814; it follows that it was written entirely
Theophanes' Chronicle is his main claim at Megas Agros. This throws some light on
for recognition by posterity. As we learn the bibliographical resources of a ninth-
from the preface to that work, he conceived century provincial monastery, for Theo-
the idea of writing it towards the end of phanes himself mentions his having "sought
George Synkellos' life-George left an un- out many books" for the purpose of writing
finished chronicle (which stopped at the his Chronicle. This work draws on a sizable
reign of Diocletian) and urged Theophanes number of sources, some extant and familiar
to continue it up to their own time-but to us, such as Theodoret, Procopius, Malalas,
started work on his own XpovoypacpEov George of Pisidia, and Theophylaktos Simo-
probably only after George's death in 810-11. kattes, and some postulated, such as the
He finished it sometime after 813, which is p?yaS XpovoypdpoS and a Melkite source,
the date of the last events recorded in the written in Greek in Palestine after 780.150
Chronicle; but where did he write it? The early days of Megas Agros seem to
One might be tempted to speculate that have been its most glorious ones; we hear
the writing was done during Theophanes' relatively little about it in subsequent
forced stay in the capital, between 814 and centuries. Two letters of Patriarch Nicholas
Mystikos (probably dating from the years
stantinople to Cyzicus, rather than from 906-11) are addressed to an ex-magister
Cyzicus; (2) the thermal baths in Yalova were Ignatius, who had become-or had been
restored by Leo VI, and it is not known whether made-a monk at Megas Agros. There,
they were in operation at the end of the eighth Ignatius excelled in piety, but did not limit
century.
145 Combine Theosteriktos, Vita Nicetae Medi- himself to contemplation: he wrote Nicholas
ciensis, BHG3, no. 1341, chap. 38, p. 25, with memoranda about the Tetragamy affair and
VM, 29,13-15. - On summons issued to denunciations of some prelate.151 According
Makarios of Pelekete about 814, cf. p. 245
supra. 140 Theophanis Chronographia, ed. de Boor, I
146 VM, 38,19-27. (as in note 59 supra), 502,4, 10-12, 24.
147 Assuming Theodore to have been the 150 Cf. E. W. Brooks, "The Sources of Theo-
author of the Panegyric, BHG3, no. 1792b, ed. phanes and the Syriac Chronicles," BZ, 15
Van de Vorst, 22,7-8: o0i <i.e., the beautiful (1906), 578-87, and G. Moravcsik, Byzan-
monastery of Theophanes) T-rs -rpTrv67rlrot i tinoturcica2, I (Berlin 1958), 531-32.
6 P 5 cpaxov 8t6loKcGoS. 151 Nicholas I, Patriarch of Constantinople,
148 Sabas, Vita Ioannicii, BHG3, no. 935, Letters, ed. and trans. R. J. H. Jenkins and
360BC (aKi<vo5.. . Suc.pT-r63pT-rov). For the date, L. G. Westerink, I (Washington, D.C., 1973),
cf. 359B.
nos. 126 and 143 = pp. 420-22 and 456-57.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
266 CYRIL MANGO and IHOR SEVCENKO
to Constantine Porphyrogenitus, Magisterhave been merely an indication that the
Leo Katakylas was a monk in that mon- place was known both to Psellos and to his
astery under Leo VI or soon afterward. addressee; however, it is more likely that
Katakylas may have been a mediocre writer,
Psellos made the sea voyage on business'"
and his style full of solecisms, but heanddidthat he counted Agros among the several
treat a topic on which Constantine couldcharistikai which he possessed in the area
find no bibliography in the Imperial Palace,
at one time or another. Some of them, such
namely, the procedures to be followed during
as Artigenous and Muntania, were under the
imperial military campaigns. What is more,
jurisdiction of the metropolitan of Cyzicus;
he seems to have written on this subject at such as Medikion, lay nearby.155
others,
Agros; at least it is there that Constantine
After 1204, the region of Agros may have
discovered a copy of Leo's treatise.152been
Leo affected by the intermittent warfare
was, then, the second author residing at the Latins of Constantinople waged
which
Agros, and his work another-and rare--
against the emperors of Nicaea. To be sure,
volume produced or at least kept insources
that do not mention Agros specifically,
monastery's library.1" the Sigrene mountain of George Akropolites
and
In the eleventh century, Psellos spoke ofTheodore Skutariotes seems to be west
"our" Agros in his letter about the intended
of the Cyzicus peninsula,156 and the Latin
journey to Medikion. This possessiverazzias
may were conducted mainly from Lam-
psacus (Lapseki) and Pegai (Karabiga).
s52De Cerim., I, Bonn ed., 456,13-457,13, cf.
However, the Latins did occupy Panormos
(Bandirma)
especially 456,15-17: wEpi TrorMrcov wrr6pvrlpa 1v and Cyzicus in 1204 and 1206-
T ~ iov Ti) KaAOUIpiV13 l typtacvi EOPEiV l8uvvlASl-
20, respectively,157 and therefore could easily
pev. Cf. P. Lemerle, Le premier humanisme
byzantin.... (Paris, 1971), 273-74, and A. have reached Agros. Thus, if the tradition
Toynbee, Constantine Porphyrogenitus and His that the monastery suffered "from the
World (London, 1973), 578-79 (the author crusaders" or "from the Pope" has any
speaks of a 'Sighrianes' monastery). - Pargoire, foundation in fact, the damage may already
"Saint Th6ophane" (as in note 139 supra), 94, have occurred in the first half of the thir-
is inclined to identify Nicholas Mystikos' Igna-
tius with Constantine's Leo Katakylas, presum- teenth century.'"s
ably because they both were magistri and lived
under Leo VI. The occasion which sent Leo- 154 Ep. 13, ed. Kurtz-Drexl, 15,5-19; Psellos
Ignatius to Agros would be the Tetragamy speaks of a "wedding" which came to an end
affair. For our purposes, it should be noted (we do not know whether it was an actual
that in the early ninth century Agros received wedding or a symbolic wedding to a monastery,
a monk, if not monks, who had enjoyed high i.e., receiving a charistike and giving it up, cf.
social status as laymen.
wTrrov... pot Tv vv(ptlKiv WrrTO-r&6(c, meaning,
153 Two more volumes must be assigned to "give me a monastery in charistike," used by
our Agros. One of them is the present Parisinus Psellos in his letter to the Metropolitan of
Graecus 216, a tenth-century luxury manu-
script of the Acts and Epistles (with commen- Cyzicus; Sathas, MEcaitoVlK' BtoS10~Kri, 5
[Venice, 1876], 457). Initially, he intends to
tary). According to a (thirteenth-century ?) return to his studies; however, the devil
remark on fol. 2r, it once belonged to the mon- changes his mind, &wrocAaiaots vcatrrwiSov oyf-
astery; cf. the catalogue of the Bibliothdque
Nationale, Byzance et la France mddidvale (EacxI KcXt iETr06Pods, which we translate "persua-
ding me to calculate advantages and barters."
(Paris, 1958), 9-10 = no. 13 (with bibliography), Psellos decides to sail to Medikion and Agros.
and J. Darrouzes, "Notes d'Asie mineure," in 155 Cf. Ahrweiler, "Charisticariat" (as in
'ApXETov Ti6vrov, 26 (1964), 30-31 and 35 note 37 supra), 24-25, with source references.
(reproduction of owners' notes). Ca. the thir- 15s Akropolites, 46,10-11 and 68,6-7, ed.
teenth century, the library of Megas Agros Heisenberg (as in note 86 supra) = Theodore
possessed thirty-five volumes. Unfortunately,
the list of titles itself has perished. Cf. Darrouz6s, 7Skutariotes, ed. Sathas,
(Venice, 1894), MEocovKiKl
476,29-30, BiPAlOS.fKTi],
and 488,31-489,1.
op. cit., 35. The other manuscript is Parisinus 157 Cf. Hasluck, Cyzicus (as in note 23 supra),
Graecus 1538 (eleventh century; Vitae of 197-99.
saints); it belonged to Agros in the thirteenth 158 The tradition is related by Nikodemos of
or fourteenth century, cf. DarrouzBs, op. cit.,
33, 38. - For manuscripts attributable to other Cyzicus, 'AKohovuSia
and vigorously denied(as
byinPargoire,
note 131"Saint
supra), As',
Th6o-
monasteries by the name of Agros, cf. note 111 phane," 94-95, who points to the Catalans as
supra. possible culprits.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 267

Since the addressee of the letter in which
This damage, if real, did not interrupt
life at Agros for long. We know thatthis thereport occurs was Andronicus II, it is
rhetor Manuel Holobolos was disgraced and that the offending imperial portrait
unlikely
publicly humiliated by Michael VIII inwas his. It must have been that of his pre-
1273;
and we know further that in 1283, after decessor, Michael VIII, and it was put up
Andronicus II inherited the throne and in gratitude for the chrysobull granted to
changed the Empire's religious policy, Athanasios
Holo- of Alexandria. The thirteenth-
bolos returned to the capital from Megascentury (?) fresco fragment of a scroll which
Agros.159 We therefore can assumewe have mentioned (p. 258 and fig. 131) is
that
probably connected with Athanasios of
Holobolos spent his ten years of banishment
at the monastery. Granted, some of Holo-Alexandria's renovation of the monastery.
bolos' fulsome versified praise of Michael VIII
The story of the substitution of images at
dates from 1279-80 and 1281, respectively,
Megas Agros is the last known mention of
and one of his poems was to be delivered
thatat
monastery in Byzantine times. We can
only speculate on the damage which the
the time of the imperial prokypsis ceremony,
hence in the capital.160 However, Holobolos
Catalans may have caused to Agros either in
did not necessarily have to be present in when their company, still in the
1303-4,
Constantinople on those dates: he may Empire's
have service, spent the winter on the
composed his poetry at Agros and sentCyzicus it to peninsula,163 or about 1307, when
their troops, by then foes of the Empire,
the capital in hopes of mollifying his imperial
tormentor. raided Artake (Erdek) on that peninsula.164
When Holobolos resided at Agros, the We found no reference to the monastery of
monastery was not under the jurisdiction ofMegas Agros in patriarchal documents of
the patriarch of Constantinople, but underthe post-Byzantine period.
that of Athanasios, patriarch of Alexandria.
About 1272, that prelate obtained from 5. Malkara Ustii and Timanyo
Michael VIII a chrysobull granting him the If Megas Agros was in fact situated on
revenues of Megas Agros. Athanasios I ofthe present farm of Mehmet Kise, the mon-
Constantinople (first Patriarchate: 1289-astery Polichnion-Polychronia should be
93) would not tolerate this infringementsought nearby. Unfortunately, Methodius'
upon his rights. A quarrel ensued, whichVita does not specify whether Polichnion
ended with the return of Agros to the foldlay west or east of Agros; we recall that he
of the Ecumenical Patriarchate before 1293.161
gives only the distance-six semeia or miles-
It is to this quarrel that we owe an interesting between the two.
glimpse of Agros. We learn from a letter We found Byzantine remains in two areas
which Athanasios of Constantinople wrote situated at a distance of roughly ten kilo-
about 1305 that a representative of Atha- meters east and west of Mehmet Kbse's farm.
nasios of Alexandria had a life-sized image The place to the east of the postulated site
of Christ at the monastery of Megas Agros
of Agros is Malkara Vtstti (Upper Malkara),
replaced by a portrait of the Emperor.162 so called to differentiate it from Malkara

159 Pachymeres, Hist., I, Bonn ed., 392,10-
394,18; II, 25,13-15. On the affair, cf. M. Treu,Alice-Mary Talbot. The relevant passage runs
"Manuel Holobolos," BZ, 5 (1896), 538-59,
esp. 545-47. as follows: -r6v Kiplov ripCev Kai SE6v 'Irao0v
Xpit-r6v v 1EK6V IOT-r&vpEvov EIS 76 T poKUVEiT.(at
160 Cf. Poems 2 and 17, ed. J. Fr. Boissonade, iv -ri povi3 -roi MEy&Aov 'Aypo0, 6 81Kcdic oa-racXS1
Anecdota Graeca, 5 (Paris, 1833), 160-61, 178-79.
Trrap' CaVTroV o0K Epp?Ev 6 SEoarvyhS 8Icacr'rracCX Kic
161 Pachymeres, Hist., II, Bonn ed., 203,6-10.
T-r~lv P~ x aaihlKv KoXacKEI 7 ~Vwta TIOTvo- c1 TvqAoi-
Cf. Laurent, Les regestes (as in note 6 supra),
pEvos. A note of caution: 81Kaic means "as a re-
no. 1614, esp. p. 406; F. D61lger, Regesten der presentative of (Athanasios)," not "a dikaios
Kaiserurkunden des Ostri'mischen Reiches, 4 (of Athanasios").
(Munich-Berlin, 1960), no. 2179 = p. 23. 163 Ramon Muntaner, Cranica, chaps. 203-4,
162 Cf., e.g., Vaticanus Graecus 2219, fol. 49r; Colleccid Popular Barcino, 145 (Barcelona,
Laurent, Les regestes, 404, 406. The letter 1951), 25-32; Pachymeres, Hist., II, Bonn ed.,
bears no. 69 in the forthcoming edition of 399,6-400,2.
Athanasios I's correspondence by Professor 164 Pachymeres, Hist., II, Bonn ed., 529,1-2.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
268 CYRIL MANGO and IHOR SEVCENKO
below, about eight kilometers east of theinhabited in Byzantine times. This is
edly
village of Kurgunlu. attested by several fragments found in or
Malkara seems to be an alternate name near
for the structure built by the mayor of
Magara, "cave," as a large natural cave,Bandirma:
now a column shaft, part of an Ionic
accessible from the shore, was hollowed out
capital, and a chancel slab (figs. 137, 138).
by the sea not far from the starting point Malkara
of tstil and Timanyo are thus two
the path which leads up to Malkara possible
Ustli. sites for Polichnion-Polychronia. Of
One reaches the site after slightly lessthe
than
two, we prefer Timanyo, for, like Polich-
one hour's ascent; the descent takes about
nion, it is close to the sea, while Malkara
twenty minutes. A spring is only about tVstii is farther inland. Even if neither of
fifty meters away from the site, and these
must places should turn out to be Polich-
have determined its choice. The place has the presence of Byzantine remains
nion,
been thoroughly plundered for building
there attests to the density of monastic
material-a deep trench in one spot testifies
establishments along the south shore of the
to this. Consequently, very few remains are
Marmara Sea at the foot of the Sigriane
visible at present: an almost completemountain range.
quadrangular slab (0.57 x 0.65 m.), several
We know next to nothing about the history
broken slabs, a number of pieces of worked
of Polichnion. The very name of the monas-
tery is transmitted in several forms: in
marble, fragments of Ionic impost capitals--
one of them with a cross and an egg-and-dart
Greek, it appears as Polichnion, Polychronia,
motif on the opposite face-and a fragmentand perhaps Polychronion.166 Polichnion is
of a chancel slab. These fragments, several
theoforiginal form, as it is the only one that
which were photographed together (figs. 57, in the Vita by Methodius. However,
occurs
58), may be as early as the sixth century.we
Thedo not consider Polychronia or Poly-
fragments at Malkara Ustfi, sparse aschronion
they errors, but rather early variants of
are, prove that a church-and probably
the a
monastery's name. Otherwise, it would
monastery-existed at roughly the be
same
difficult to explain the appearance of the
distance east from Mehmet Kbse's farm formas Polihron in two early Slavic texts: the
Polichnion was distant from Agros. Vita of Saint Methodius, the apostle of the
The place lying about ten kilometers to written in the late ninth century, and
Slavs,
the west of the postulated site of Agros the is
Laudation of Saints Cyril and Methodius,
called ?iftlik or Timanyo; the latter name written about the year 900, possibly by
seems to go back to the Greek (T-rdptpov =
Clement of Ochrid (d. 916).16s Polichnion was
timar ?). Timanyo is a tillable plain extending
situated on Theophanes' property. Shortly
about five hundred meters along the after shore 780, the Saint was tonsured there in
and five to six hundred meters inland. Today,
the presence of his wife by Father Strategios.
it has an artificial harbor, too dangerous to
166 Polichnion: e.g., VM, 15,21 and 25,20
use in rough weather (fig. 136). A sizable
(the only form occurring in that text); BHG3,
stream runs along the western edge of
no.the
1791, ed. Krumbacher, 392,7; Polychronia:
plain. Within living memory, Timanyo
Vita of Theophanes, ed. de Boor (as in note 107
belonged to an Albanian; about 1920 itsupra),
was 29,5; Polychronion does occur in BHG3,
bought by the mayor of Bandirma, who no.put
1790, as edited by Goar, cf., e.g., PG, 108,
col. 28D, but de Boor's edition of the same
up the stone structure still standing on the
text, 18,38, has TrohiXvicp in the corresponding
shore; at present, it is owned in common
spot. by
the villagers of Yenicek6y.e65 A modern I'l Vita of Methodius, apostle of the Slavs,
tile
with Greek letters found near the shorechap. 4, ed., e.g., P. A. Lavrov, Materialy po
points to Greek occupation of Timanyo istorii vozniknovenija drevnejgej slavjanskoj
before 1922. It is not clear whetherpis'mennosti
the (Leningrad, 1930), 71; Laudation
of Cyril and Methodius, ibid., 81: postavisa i
foundations of houses on a terrace in the
igumena
northeast area of the plain are mediaeval orv, manastyri
Polihrons, v njemble ile naricajetbsf
bfge o(tb)Cb 6.imenbmb
For the
modern. However, Timanyo was undoubt- claim that the Laudation is by Clement of
Ochrid, cf., most recently, B. St. Angelov,
16 Information obtained orally at the coffee-
K. M. Kuev, Hr. Kodov, eds., Kliment Ohridski,
house of Yenicek6y in August 1972. S6brani ssinenija, I (Sofia, 1970), 444-47.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 269

Methodius, however, does not give any
Soon afterwards, Theophanes donated Polich-
nion to Strategios and his community.les details about Polihron's location. The phrase
We hear for the second-and last-time "Olympus, where Holy Fathers live" and
about Polichnion from the two Slavic where Methodius went to become a monk
sources just quoted. These two are occurs reallyin a different passage and, logically
one, since the Laudation depends on enough,
the refers to an earlier period, that
Vita. The latter text has this to say: "... preceding the Khazar mission of 861.173
the Emperor and the Patriarch ... urged Even if Polihron had been assigned to Olym-
him <i.e., Methodius) onto the way of God169 pus by the authors of the Slavic texts, their
so that they might have him consecrated usage would not have been unparalleled: we
archbishop of a worthy spot, where a man find this usage in a tenth-century Greek
like him would be needed. When he refused, source, the Dithyramb on Theophanes.174
they compelled him and ordained him abbot When the author of the Dithyramb came to
in the monastery called Polihron; its worth speak of the monastery where the Saint was
[?] is twenty-four pounds [?] of gold, and tonsured, he called it "a certain place in the
it has more than seventy fathers."170 The area of Olympus"-and we already know
author of the Vita inserts the passage about that the tonsure took place in Polichnion-
Polichnion between Methodius' (and his Polihron.175
brother Cyril's) missions to the Khazars and
173 Vita of Methodius, chap. 3, ed., e.g.,
to Rostislav, prince of Moravia, the former
dating from 861, the latter, from 863. It is Lavrov, 71: i gbd, vw Alimbb, ideie ivutb
s(ve)tii o(tb)ci, postrig se oblRe vs 6brny rizy.
quite likely that Methodius was made abbot - Vita of Constantine, chap. 7, gives the same
of our monastery between these two years; chronology: when Constantine is reported to
join his brother in Olympus (mv Olimbb be ed
if so, the patriarch who was instrumental in
k,6 Me/odiju bratu svoemu), the Vita has him
this nomination was Photius himself. Al-
sojourn there before the Khazar mission.
though Methodius was never to return to174 BHG3, no. 1792. Chronological limits of
Polichnion after the Slavic mission began theinDithyramb are given by the passage in
863, he must have retained the title of abbot which the author refers to the emperor of the
day (an adult man) as a relative of Theophanes
for a while, for he is given it in the speech (617,22-23). Since Theophanes was a great-
which the author of the Vita has Emperor uncle of Zoe Karbunopsina, mother of Con-
Michael III deliver on the eve of the Thes- stantine Porphyrogenitus, the emperor in
salonian brothers' departure for Moravia.171 question could have been Constantine himself
(an adult co-emperor by ca. 925; d. 959), the
There is no need to doubt the identity of
latter's son Romanos II (d. 963), or his grand-
the Slavic Polihron with Polichnion-Poly- sons Basil II (d. 1025) or Constantine VIII
chronia of the Greek Vitae. In secondary (d. 1028). - For slightly narrower limits
literature dealing with the Vita Methodii, (920-59), cf. Krumbacher's edition of the
Polihron is sometimes located on or near Dithyramb, 596. - If we understand the
author's obscure style correctly, the congre-
Bithynian Olympus.72 The Slavic Vita of
gation was celebrating Theophanes' memory
168 VM, 15,20-22, 15,32-16,1. (feast day: March 12) on the Sunday of Ortho-
169 Cf. Matth. 22:16, Mark 12:14, Luke 20:21,
doxy (609,1-5, 12-15, 24-26). Between 920
Acts 18:26. and 1028, the Sunday of Orthodoxy fell on
170 C(6)s(a)rb i patriarhk... na b(o)9ii put'
March 12 only in 976. However, we must
b6dida i, da by a i s(ve)tili arhiep(i)s(ko)paabstain from dating our Dithyramb to that
na cbstbnoe mlsto, idele estb potrlba takogo year; an important saint, such as Theophanes,
mula. Ne racdbju le, unudida i i postavisa wasi commemorated on the Saturday or Sunday
preceding the day of his feast; cf., e.g., A.
igumena vb manastyri, iUe naricajetbse Polihrons,
Dmitrievskij, Opisanie liturgiceskih rukopisej...,
jemu2e jestb s, mnra k i d spudove zlata, a o(tb)cb
I (Kiev, 1895), 428. Thus the Dithyramb could
obile 5 v, njemb jestb.
171 Vita of Methodius, chap. 5, ed., e.g., have been pronounced any time, say, between
Lavrov, 72 (Emperor speaks to St. Cyril): March 7 and 12, and the author's allusions
i poims, brats svoi igumen, Mefedii, idi ?e. cannot be used for dating it.
172 Cf. F. Dvornik, Les Idgendes de Constantin 175 Cf. BHG3, no. 1792, 611,9-21: Ki -r~lv pv
et de Mdthode vues de Byzance (Prague, 1933), [i.e., Theophanes' wife]... ov&'dLv w-rrEXE-ro
210-11. - Hr. Loparev in Viz Vrem, 18 (1911),
69 note 2, considers the evidence of the Slavic P3ios, r6v sU [i.e., Theophanes] Ki aoTi-rov 9p'
6pofI ri r 7rpouvpviq. .. 1rp6 -riva -rO v 'OhuXvpriCv
Vita Methodii spurious. We disagree.
ppcov y~pov [a lacuna follows]. This passage

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
270 CYRIL MANGO and IHOR SEVCENKO
The size of Polihron's community-more
support a community of over seventy monks.
than seventy fathers-puts it in theNorsame
can we render "its yield is twenty four
modioi," for we learn from the Vita of St.
class as two other important monasteries
of the area, Sakkudion and Medikion,Peter
eachof Atroa that Balentia, a rather small
of which had a community of roughly one
monastery in the neighboring region, needed
hundred monks about the year 800.176 One
thirty modioi of wheat per month for the
hundred must have been a considerable upkeep of its community of fifteen.179 How-
number, for the biographer of Niketas of in
ever, our passage sppdt, "modios," is
Medikion praises his hero for having recruited
followed by zlata, "of gold." We suggest the
that many monks for the monastery. By meaning
way "twenty four pounds of gold" and
of comparison, let us recall that the foun-
apply it to the worth of the monastery,
dation charter of the Great Lavra of Athos rather than to its yearly revenue. To be sure,
foresaw a community of eighty monks, that
a monastery at Nicaea had forty-two monks, the meaning ?idrpa, "pound," for sppd, is not
attested, but we know-to give an example-
that the three monasteries of Mount Gale- that the sum total of revenue of the Nicaean
sios for which St. Lazarus (d. 1054) wrote monastery, to which we just alluded, was
his charter had to have sixty-four monks, measured in "pounds of hyperpera."180
and that the size of an average Byzantine Although the passage of the Vita Methodii
monastery has been recently assessed at ten is probably corrupt, its general message is
to twenty monks."' clear: in the sixties of the ninth century,
The words of the Slavic Vita which Polihron was a large and wealthy establish-
precede the information on Polihron'sment. size--At about the same time, it almost
jemuge estb szmera k. i I. sppdove zlata-arebecame the shrine of an important relic:
the body of Saint Constantine-Cyril, who
difficult to understand. S7imbera corresponds
both to pirpov, "measure," and to orra2p651,died in Rome in 869, was to be brought to
Polihron to be buried there. Such was the
"weight"; this is the first ambiguity. Usually,
sppdz translates p'68to5. In the New original
Testa- injunction which Constantine and
ment, this word denotes a dry measure
Methodius had received from their mother:
corresponding to about nine liters. In"And
By-Methodius, his brother, asked the
zantine times the modios as a dry measure
Apostolic Father, saying: 'Our mother ad-
jured us that whichever of the two of us
had varying values; often it was equivalent
to forty litrai, or pounds, which in turnbe called to Judgment first, (the
should
corresponded to about twelve kilograms of brother) should take him to his,
surviving
wheat. However, the modios was also a the brother's, monastery and bury him
surface measure, the extent of which fluctu- there."' However, Roman bishops insisted
ated between about 900 and 1280 squarethat Cyril be buried in Rome, and their
meters;178 this is the second ambiguity. counsel prevailed.s18
Translating "its area is of twenty-four 179 Laurent, La Vie merveilleuse (as in note 65
modioi" will not do, since this would give supra), 167 = chap. 50,7-8.
Polihron about three hectares, too little to 180 Chr. Papadopoulos, 'H NKMltcX..., in 'O
iv KovoTroavrtvovnr6Xst 'EXAAXrjvltYK6; XoXoylK6i XiX-
XoyoS, 33 (1910-11, printed in 1914), 138: OMOY
corresponds to VM, 15,20-26, describing Theo- AYTP(AI) YTTEP(TTYP&YN).
phanes' tonsure at Polichnion and his wife's 181 Vita of Constantine, chap. 18, ed., e.g.,
taking the veil at Prinkipo. Lavrov, 35-36. - For the sake of completeness,
176 Medikion: BHG3, no. 1341, ActaSS3, we report that a note in Barberinianus Graecus
April, I, chap. 10, p. xx. For Sakkudion, cf. 319 (date: 1039), fol. 174v, states that Monk
Menthon, L'Olympe (as in note 45 supra), Barthelemew t-ro TTovXxp6vt bought the manu-
162. - We were not able to find Menthon's script for three nomismata in Jerusalem in 1168.
source for this statement. Barthelemew was a pilgrim; but it is not clear
177 Cf. P. Charanis, "The Monk as an Ele- whether he was the son of a Polychronios, or
ment of Byzantine Society," DOP, 25 (1971), whether he came from our monastery. The
esp. 69-72, and (for Mt. Galesios), ActaSS, former is more likely. Cf. K. Lake and S. Lake,
November, III, 585D. Dated Greek Minuscule Manuscripts to the
178 Cf. Schilbach, Metrologie (as in note 125 Year 1200, fasc. VII (Boston, 1937), 15 = no. 285
supra), 66. and plate 518.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 271

6. Yenice and Manastir Hiereia, mentioned in the Vitae of Theo-
The village of Yenice is situated thirteenphanes, that the Saint's body lay in state
kilometers east of Bandirma, and ca. twentyfor about a year, having been brought there
kilometers west of Kurqunlu. In 1971, itby ship from Samothrace soon after Theo-
could be reached from either place only by phanes' death in March of 818.184 While we
boat. The village has two natural coves; thedo not know whether Polichnion lay west or
present harbor is located in the western one.east of Agros, we may safely assume that
In 1910, its inhabitants were Greeks and Hiereia was west of it, since a ship leaving
Lazes from Samsun; in 1972, they were from Samothrace and presumably bound
Turks from Thessalonica and Kavalla, re- for Theophanes' monastery would deposit
settled there after the war of 1922.18s2 As a her precious cargo in a place lying between
consequence, the antiquarian and topo- that island and Agros.
graphical knowledge of the present-day We recall that Hiereia, lying to the west
residents of Yenicekdy does not reach beyond of Megas Agros, was twelve miles distant
from it. If Mehmet K-se's farm near Kur-
that year. Nevertheless, some memories
have been preserved; thus the site of the ?unlu is in fact the former monastery of
St. Anne monastery, lying about twenty- Megas Agros, then Yenicek*y, situated about
five minutes' walk east of Yenice, and de- twelve miles west of Kurgunlu, may be the
scribed by Panienko in 1910, is called Mana- site of Hiereia, with its martyrium of St.
stir today. Procopius. However, as long as no trace of
In Yenice proper, Pancenko found no that martyrium is found at Yenicekby, we
must consider other candidates for the loca-
archeological remains, except for small ar-
tifacts of the Roman period which peasants tion of Hiereia as well. Dutlimanl (Paneenko's
dug out from graves in nearby wooded hills Sykamia, lying between YenicekSy and
and offered for sale.s83 We did come across Bandirma) is one such candidate, since it
some fragments of Byzantine church archi- has an ancient or mediaeval harbor, and
tecture. In the village fountain a sarcophagus some Byzantine architectural remains.185
serves as the trough, and a mullion and the Nor should Bandirma itself be eliminated:
colonnette of an iconostasis parapet are used its Byzantine name, Panormos, is no ob-
as stepping stones (fig. 139). The mullion stacle to its identification with Hiereia, for
and the colonnette are said to have been the sources do not mention the toponym
Panormos before the thirteenth century.186
found in a field adjacent to the road leading
west from the village, about five minutes'
walk from the fountain. We visited that
The site called Manastir today is at a
twenty-five minutes' walk northeast of
site, and saw there a large threshold block,
Yenice, and a quarter of an hour's walk from
fragments of column shafts, and a column
the seashore. The place itself has an abun-
base. Moreover, it is reportedly from there
dant supply of water, at present channeled
that a parapet slab had been transported tointo a modern fountain, and lies in the midst
one of the houses in the village. The slab,
of a grove of plane trees. This is the former
0.86x 1.16 m., displays the usual cross and
monastery of St. Anne. The fullest account
vine motif on one side, and a cross in a circle
of it is that given by Panienko,'87 who found
on the other, and should be assigned to the
sixth century (figs. 145, 146). We were told 184 E.g., VM, 37,10-15 (time of body's
of inscriptions both east and west of Yenicearrival: Easter), 38,19-20 (body stays in
and at Timanyo, but we cannot vouch for Hiereia for one year); BHG3, no. 1791, ed.
Krumbacher, 398,4-6 (time of body's arrival:
the reliability of this information.
Christmas; body moved to Agros at Easter-
Given the presence of Byzantine remainsthe passage is corrupt).
at Yenice, the question arises as to the 185 Cf. Panienko, IRAIK, 15 (1911), 262. -
mediaeval identity of the site. It is at It appears now that Hiereia should not be
sought at Dutlimani. See Additional Note infra.
182 Cf., on the one hand, Pan'enko, IRAIK, 188s This according to Hasluck, Cyzicus (as
15 (1911), 264, and, on the other, information in note 23 supra), 50.
obtained orally on the spot in 1971. 187 IRAIK, 15 (1911), 264-67. Cf. also the
183 Pan'enko, ibid. 'HvaEpoh6ytov (as in note 136 supra), 181.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
272 CYRIL MANGO and IHOR SEVCENKO

there in 1910 a small church tendedwas
byofone
the Byzantine period, transcribed by
old monk. Round about were considerable both scholars. As Hasluck realized, it was
ruins, all overgrown. Various antique and broken into two fragments. The text, we add,
Byzantine fragments had been either de- is in dodecasyllables.
posited or built into the church (of which we
no longer found the foundations in 1971). Fragment 1:
Among antique fragments Pancienko listed
the following: 1) A funerary stele representing Above, cross on orb, in angles of cross X 9
four recumbent figures, presumably members (= p9Cs Xptloroi paivEl mT Tr )

of a thiasos (=-our fig. 140). The stele appears 'EvraiiSa - I Xo0v KI c Ti 76Sv iO-,KOV
to have been complete at that time and had crapKiOV
no inscription. 2) A piece of ornamental
sculpture with a series of circles superim- Fragment 2:
posed upon diamonds. 3) Several fragments [VE ........ I
of columns, namely an Ionic base of the
a-yovoaa "nn1TTV[1] 1 A3op& So Vri1v I 'wlppav I
Hellenistic period, a Doric capital, and part SK-m) X hus, I KCaTOVTad I i Tffp=Tf.
of a slender shaft of red marble with fluting.
4) Four composite capitals similar to those The date was broken off, since, to judge
in the Kurgunlu monastery. Two of these from the approximate facsimile reproduced
measured 0.55 m. between the ends of their by Hasluck, the inscription was certainly
volutes and comprised three rows of acanthus later than 992 (= 6500). Pan'enko, who did
leaves (=our fig. 143). The other two mea- not connect the two fragments, guessed from
sured 0.72 m. in width and were of a different the character of the letters that the first one
design. They had three (read "two") rows must have been of the twelfth or thirteenth
of complex acanthus leaves and an acanthus century. However, since the two fragments
ornament both between the volutes and on belonged together, their date must fall
the fillet (= our fig. 141). We may add that within the years 6500 and 6600 of the
the latter pair, with their double leaves of Byzantine era, that is, 992-1092. The frag-
fine-toothed acanthus, were certainly of ments should probably be dated to the second
fifth-century date.188 half of the eleventh century.'91
Among Christian remains, Pan'enko enu- In addition to fragments recorded by
merates: 1) Part of a cornice with a cross Panienko, we found remnants of at least
flanked by half palmettes that formed a four mullions, and a number of blackened
pointed arch over it. 2) A capital in the form bricks, the latter suggesting that the whole
of a truncated pyramid having a square complex may have been burned at one
abacus, 0.52 m. wide. On the sides of the time.
abacus was a continuous vine without Twenty to thirty meters further down
leaves. On two sides of the capital was a the scattered ruins, there are traces of
from
a regular retaining wall with rounded but-
cross with flaring ends, flanked in the upper
quarters by two doves, their heads turned tresses similar to those of the retaining walls
inwards, and in the lower quarters by two
on Mehmet Kdse's farm near Kurgunlu.
pomegranates. On either side of each cross
was a branch of acanthus. On the lateral
190 Pandenko reads Trov, Hasluck, rTv.
sides of the capital was a six-leaved rosette 191 For the sake of completeness, we mention
inscribed in a circle. here another funerary inscription found "in the
Panienko noted a number of inscriptions, rustic chapel of a monastery (Ala Anna)," and
published by J.H. Mordtmann, "Zur Epi-
some of which have also been published by
graphik von Kyzikos. III," AM, 10 (1885), 211.
Hasluck.'s9 We reprint the only one that This, like other fragments at Manastir, must
have been a spolium brought from Cyzicus and
188 Cf. Kautzsch, Kapitellstudien (as in note used in the construction of the monastery. The
85 supra), 133ff. inscription is of the usual type, in which for-
189 "Inscriptions from the Cyzicus Neigh- mulae such as Ka-rEKo<EaoEv ?a.rrc4 Kal' TOTS
bourhood," Journal of Hellenic Studies, 24
(1904), 36-38. IRAIK, (1911), 266-67. rT4KVOI, El 8S TilS T"OXAfjO110 E'EpOV K-raSiaSa0t, T-r
TaPaIEqC can be easily reconstructed.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
% C.6=

ERDEK$
Artake 4
.. (-000,
.. rlCY /\CUS
Ar.takejll
BA DIR A
11lot

E D N C

- DIR A 9.. o
, ', 9)

/' B N I A* A.
E EDNC DE K "i- -

MANYAS GOLU

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
iMRA

Me ta rorp

Ol ,

_ _ __ _ _
_---- %- . . -:A

,Sl'/iane'
751
C" i tNGAN

0Sigr
KALE

S%

0 Q
E EME C t

404A

MULUABAT -

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
AP(
IMRALI KOYU

IMRALI ADASI
Kalonymos
Metamorphosis mon.
BOZBURUN ARMUTLU
Cape Triton

14 pi
e I Mediion' PMUDAN Y

C , f $

.Olt

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
C ZCU

,I" ?., t t "-
10.

6AERDEK - ___

T .?

4 i
-for_

BAMA

5 4 3 2This 1 0 5
content downloaded from 10
194.27.33.66 15
on Fri, 09 Jun20km
2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
Me t or p

ko

A,,". "
'4' 4'C_
-,: ..(%\
,.,
r' ian 04
S8 'L i

0 1

Mih a/ic ULUABA -__________
Lopaon -_______________

:? AP(
r 'Y :10 ~La
4 AP

7 AK A R AC A B EY

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
9IMFMALI \UTYU

IMRALI ADASI
Ka/onymos
Metamorphosis mon.
BOZBURUN ARMUTLU
Cape Triton

,,Medi on~MUDANYA-
1 4
Apollonia * \' ,iiii

Lake Apo//onias

**

APOLYONT GOLU ona

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
1.
All use subject to http://about.jstor.org/terms
0 40 ILAl~ ~ ;I~
-tan s t"4 4
- PI~~LLLL~~~ r~' C r I~C4 (a> V~r4 ~ff4(t va
4.,,

o4-t L4W4- yP4

--4 nt &. [. / yi"- s 1, _,Al wta ,

0.4 , uIW
t17~ .f4 d-..ae,.j4_
11"
4g, D
(FI-L ;*Itat SI.C-vck?,v ;oCO L'J Jrt A L A . 40L

Ito af le ^4wJL & wko-
Va -04WY4;o 4VAL k.Ck

&-... , f- . 4 "."
WL./!t4,A- - V-00' .0 '' "-7"010re
4w%.&*4g.& ," . , ,

~~~ ~kr r44 irMri
.- II% , Iz - . - - -. . . .
444-ti bR.%t 1 Q S,.SA 4 # R.

2. John Covel's Diary. British Museum, Add. MS 22912, fol. 266r

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
0
? Jvv M5 -?IA41'k- " AAO 441tf-r n ' m4

(7 4 ?4 M ,44
,r " -"* --t-.a"
6rw CU flj4P'w
%'4A~*' 4t~j fl . 4, w Y 4k
r?? AJ L Qtn% el*,a'
- L-
"< ? "" /aV
"t"g e.Lak titntv , tKL/rrw

'o'
W?~AO

" "p K O' i p, f- t ,,ri , ,

:r re,.mw,. ,,&
4 IL Si dr/Ct? , ._-r ,- OrQ-
4;k tIL,
8,,,,,kb-., -,4 cr .t "< ,,,L
C t- n- %,, cLr, .
*.... t ,.-,a,= ..t . C ,I , - I i

vA1r Pactin ?,~c l>4 CG-3 j~% At Z~)CC64: 4f~&-i 4 Ct -ry 44."i

i. :,t 6" '/ / f "k /"
"I'01, . , 9, "A"
C~~ C~c ~~~I~?
- )i(42,oat..LAC
" -'.,. j.P'u it , ,r '-
Vt~kas7O~a&-cA. #vC A r44L

. ,. kxQ < .~ vS'. 4ta ~... (4Li71L2 . .,w A - ~ . " - / 4,i , , ,
eA A

;1 .-,, Li - 9k4tt&&c
,,,.'.? ?ieiZ .. Z'_..', L _c '0. 3 _' cl ?' CC5

n;?lf~~ ,.. ..G. G u; r?I?/ r'l
lw,,,LM Cr a9Z t~r ivM . ,Ct--, I Iwrr41 ?Ac
4z. .&Z\J "ltlX~ " c&U .'_ o ,:4 cA
I t-" jVt- , j, AI
o b t" ' '#% W *A A to MI t4W"tI 4,r

3. John Covel's Diary. British Museum, Add. MS 22912, fol. 26

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
4. Exterior from Southwest

5. Exterior from Southeast

Tirilye, Fatih Camii

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
6. Exterior from Northwest

7. Atrium Colonnade

Tirilye, Fatih Camii

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
?. /

?I

A LLERY 01 ER
'~~SP

5 T R E I
M(NARET

N ARTH EX

FORE CO U RT

5 T RE ET

8. Capital in Atrium 9. Ground Plan after Hasluck.
Scale 1: 300

10. Interior, looking Northeast

Tirilye, Fatih Camii

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
lbl
North-east Co/urn

DD

, . ... A? i .? ??; ~-
6 r I
~ I)

49. .. . . . . " South-east Column.

rV
r1!I~l ,V0?~1 ,4 'r l
4e ve

D W.

WV. E.S.N

11. Interior, Northeast Capital
12. Mon

- .r -Q.. , ,.?

. .... ?.! :

'i" :i'i lb-

?' * ? ? " . "o .
t7-
i :...-... . ?+_: ,, f;
:"-,,.'.., - .,:,- :::.?Yrqsz?
;,.-, ..,,, ..:= . ...":.':.... ,?
.., ., . i; ? I 7$T teT
M , ...

_ IZ

. .. .. .. ....

13. North Crossing. Detail of Cornice 14. South Tribelon, C

Tirilye, Fatih Camii

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
. ...:..
? .. . ..,
.. *. 5,.
.- . ...

r -, i

.:.- .. . .'

., -~ ??- "

I :.

,. _ ...
a. .4L

16. Par
15. Fountain in Street to West of Mosque

'., ,.i, '"i J
!,.
...

....~=dc . ". r; '+?

'~q - ;
?..i?' " ,

17. Abacus to North of 18.
Mosque 19

Cap

Tirily

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
MODERN HOUSE

17O7 -_.- --'
Funerary portraits

I I:
I I I I

6 - [DOME
IMODERN
I S i i t
SB

Funerary groupportrait--
1,j =:', ORIGINAL CONST

o s 6 7 8 9 0 RGIA AL

OLD AD

19th CEN

20. Tirilye, Ke

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
S T R E a t
! ..
" 11
)III

, l ii

* *
dg

e e

I

A,-

ecslm i ovr,.
GALuMY: OVER
-di

GAL o- R

. . . . . .
BarRYY r Wp
?!?-

21. Ground Plan after Hasluck.
Scale 1:300
4W

-1wr ~
z, ne r~iL'r

At~~

22. Exterior from Southeast

Tirilye, Kemerli Kilise

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
c+ r, :.a
,,4 wr. L-;
f4 1 It . .. . . . . .. . , , . ,
tr Or-

-Iw

i '" ' "' ~ii ~ .
-.,,.Z . ?-- . ..

? Ell
-AM

.: lL - I-. V

-,W ..r.
ZI

,i t lm~~d "I r?
23. Exterior, North Wall

': ,, ; .N "
. .. .." ~ -,
to-. f

~ ~ ~ -
~ ~ ~,~ . ,
. "?I,", ." .,

14, -," ' ,. , . +? ~
. ! .. 4, - !
r +:. I' ." SI- ;., .,:
Ir +. , +
z.: . . W .

." i? -? . ~r;
' ~I i? " "+:;:-P., .' .t.-.
, ? ~ ~",* ':::?' -.- "':',
41
,. ?"~L 4L; .? ;i.

- ' .. .. "
,~ ~ ~ ""; .. .....,, .
": . -. ..~r . ..? . .. :re,:re

. ! .. ? , "!= ,
"! ~v ........ I ' ' ....
iC.?. r. + a
t_ I ?. r ... t. +u..
... . . . . . i,.. . ..'-.
? L '+ " " .... .. < .. -k;
_,,~ ~ : .- ., ,.3 ,

24. West Fagade, North Niche
Tirilve. Kemerli Kilise

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
0
66.

ici

tiL .I
Cl)
0

.O

tit

,v-

Jrit I A
Am
av

I'dH ~ L l i

. ,. : "" .
4k It-l _?
L \j
? ,v-i

'43 j ;d
Ir W,
. '.:'"i ..

",. . ? " i
.. - .. .

.. ?. .. ?..

?'

r,

~1
???1~ p~u?.~ I1. r.
I 1

:+.:-~x I~t;T?: r

`~:,. s? I ,

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
"C'- _...~.

r-

i
: *

C:

'??.?~~~ . .~,.?PL -L. i'
?? `~f:
EJ. ?~~ P ? 'r

?I-~yyy- I ? - I

~~?? L
I~
?L.

d.P? d -rl -r s
r I;i Z
L C

~? ~I?
*"I~ f ;?

i
- ?orrr r ~.?

,i

~~P~ ~EIIP~L~Ylrl -~V~C'6-?

?u ?~

??1-~?
ii; 1?
I~l~y ~sr ~F~.=
r
?- ,
I'

I I I - ? *
?Y.

-? ~c:
~'??

'I

r

,?d~ C"

I
~.??f~4~. i~?
r 1-
L- r; ?.

i
3T r ':I
;I, JC:
i!
*i.. ~?lt~
r,.

J
:P?? F ~'??
I*/
?..

cirii;r 1

1: ?r.r I Y

,t
i

;t. I
"'

.'?

s~s
;Ip?

i:
h~ ~? ~ "~C ?'??

*~?;;C.
r' t ?:??.II~,
%? i''

.?

,

V
-- ?L~
r

CI
i ~e
"'
1Wvr
~' ?
- I
Bih`Z~'~
? ~~ r?? 7 wur:
I-: ;
??

.,..
?i I r.?
:: a~i~e
?? ~I~ .? .'I
1
r
j' ? ? i
~

I; ~; j r
t

r?; g-; r t,
.?

,1~D~~a~?c ??

i.

.?

27. Tirilye, Kemerli Kilise. Interior, looking Northeast

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
29. Northwest Column
28. Capital under Northwest Corner of Dome

30. Capital under Southeast Corner of Dome

Tirilye, Kemerli Kilise

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
.. . " , .. , ..

{ r , * :?I
$?' . :::11~~ '.i .
?." . -1 .. ,.:;,..,-
r.' ,, ! .

!!I, . i.. ,, L Z

! .'?'I*

32
? ..c

I~'~~C ~; ?~

,. . .? 1QF;
i

i
I':~ r
L??.
i

r
r

\??,

?-~?:? ~
:;~ ?r
"*. .I .
' ?*-
1

??:

. 6

i
i:

r 1?? '
''
fs
:~?
*; c

~?,

33. Bema, North S
Figure of

31. Joachim's Offerings Rejected

Tirilye, Kemerli Kilise

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
34. St. Michael and Military Saint 35. St. Micha

Tirilye, Kemerli Kilise, Naos, North Wall, West Bay

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
37. Narthex, West Wall. Funerary Portrait

36. Naos, North Wall, West Bay.
Military Saint

38. Narthex, South Wall. Funerary Group Portrait

39. Narthex, North Wall. Funerary Portrait

Tirilye, Kemerli Kilise

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
This side of wall plastered .....
;%-" ''' ~ '" ~?.??''
' ?.: ,,i.

.??:~ I II I I I
" ,:=
? ?i?I I \
I I I I' ii I"
i \I I I! 'I I \I I

''
..-:.... .,

::(... : ? , r . . - -
-...:..: , !; ";'
?. ?. .

-""' Thi sie
:-:-... ofwallnotexamned I
??. oo..-
'' L :": ::-...-'-'=2 .. - -.:i"::""::.....-':::,:.L'.--.",; -..:..:.:,'.", ;..;':;::''': : ; :;; ? ;';: : :i
0' 1 7 8 9 lr

40. St. John of Pelekete, Ground Plan

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
41. Exterior, West Wall

42. Exterior, South Wall

St. John of Pelekete

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
43. Central Apse from South, Exterior

44. Interior, looking East

St. John of Pelekete
This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
45. Interior, looking Northeast

46. Interior, Naos, Northeast Bay
St. John of Pelekete
This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
47. Interior, Southeast Column, looking South

48. Cornice above Southeast Column

St. John of Pelekete

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
49. Southeast Column, Capital 50. Parapet Slab

51. Bema, South Side 52. Prothesis, Aps

St. John of Pelekete

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
A X

- ' 0,0,

. i" *~ ; h~ r cl

I;
I. 'a 1%

A4.l

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
57. Carving, Fragments 58. Capital, Fragment

Malkara Ustii

59. ?ingyan Kale. Tower seen from North

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
60. Karadag Region. View of Seacoast. Kurqunlu Village in Distance

61. 62.

Kur?unlu Village, Lokman's House. Funerary Reliefs

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
'' itr , 1' -i A
3 '

A 1^
i

, . .,?

:IN

SPo . NKt. ..:,~ ~ ~~ ~. ..;':.: , ;.
. .. r.. . . "

f YiT.

64.
"? :.]v

ZUl
63. ":. ' ),
?r*
. C "., :, t
? : bt .. . .." "
w?.
:?-~
C.

?r?.ur~i r?r
.,,, .v~ r-.,.
~
? t,:,'.'-"`,-- , ' ' 2
.:e

..
.: ..',::"
;::" ",:...
- "" . .,,
- CJ:
. . -.
I ~ ~ ? .-, -; .. '" ,, .? -, *-;.-
,?

- . . . , .- . ? . ;- " . Z "
., - - , . : ; . . ? . , ",

jrC~la~z~%--r3~ ~R~e(?L4--~G~Y~~ 4f~~i:

: ? ~TJP~i~-=e~.r~

r
'~~ ~C.C?I~

C' .~.j: ?t Ii~r?lrr ~uHu~~urau~r;
:._.. ;? -?* r?.Pil ~:?'"~;Y
;:i ?1 q~
ir
?? ?,
L-1 ?~
r~ t
r;
*???
..
~7 , b ,r4

u??F~- ; ? IcLir)r, :,,, .

/1
?,

65. 66.

Kursunlu Village, Lokman's House. Inscriptions

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
68.

67. Inscriptions

70. Lead Weight

69. Doric Capital

Kurqunlu Village, Lokman's House

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
73.

72.

74.
75.

Kur?unlu Village, Lokman's House. Capitals

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
77. Entablature, Fragment

76. Parapet Slab, Fragment

Kur?unlu Village, Lokman's House

78. Kurgunlu Monastery. General View from North

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
79. Circuit Wall along Seashore

80. Circuit Wall turning Inland from Seashore
Kur?unlu Monastery
This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
81. Monastery Gate Transformed into Dwelling House
"ow'l, w' i'oi - -I. WAI U -I

83. Mullion reused in D

84. Carved Fragment reused in Gatehouse

82. Gatehouse, West Side

Kur?unlu Monastery

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
85. Gatehouse, Interior, North Side
.. 7~--?~

- h? *?;I.

W .?.~~~1 ""?'.~9.-
-4 '~C~
?? ?r
".II' 'P 3
rrr~~..c.i' ?L;f~~?.~T . Ci~?C ~?~~
r. ??

'4'. i~i~t~l ''r
:?~i ~? 'rr,*~F~-r.l~~i~? ::-:?'?
'r'' 3?.';'

?Y L I: .r?'ii?? ?*?' '; r
i-r. ?~i???I ?*7 0.

r 1;~ ... ? - r ?!
AaU "' i ?

?-- ...

~rlpX~:
'1
*r
~? c
:"''
Ir
'
C ? ''?

. t k. ?? ?
?~*
?'?I???? 'f
't? ??-?:
?'
~

.Y ? ;5 i i?
i '.
-4ki .?*??
.. ?:~;: .??* *. i
E '?'?~ .? i +, *
Be~: ?~ .* *\F4 ~~ , r ?'`?
;:i I~i ??--? ?-t?a3 "
?~ ?r' d ???
?-: ? .~~Y - ' ?,?? ?? _a?l??
- : 1?
:i''
ut
-fi .....
?-
AsP '?-;1 ?'+?..
?;e i ;???
i ~ ' ~??

.?j?
?71~?~ r
? v? '?

C,"L~I 1!I?,?i ;* ~- c
- ' o Ad.~ :r

?I
L. ? ??
-I?i?* . ?L '~?C
" ' ??C~~~f , ,i-I~
:?''c,?'
..
,.? ?.

r
'.*:.~r,

87.
86.

?'?~'*'i:r -Y~i: -;-1"t?
.7.
.- . .. .-.. ... .-.
AV-
:o ,;o Q: *??;
?.i " Awy? . . . ? .At

Nw. ? s, ,I, " . .. ,.: ?" , :
-?; Zt
:I,?I ?jg

89.
88.

Inscriptions, Fragments

Kur?unlu Monastery

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
90. 91.
Kurqunlu M

92. 93. 94.
Columnar Sarco

96.

97.
95.

Kurqunlu Monastery. Columnar Sarcophagus, Fragments

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
?_?_I_? ? ?1~~Y1 C?UI?I~?UI???II~ C?-?~L?Y~~ III~LIU~~II

).*
Slb~-

L
?? ~5~I?'
?': i
:c~:?
i
~?~i? `?;;:?: ;~ '~r;. .':.:=:??"i . .~._:Y" p -iY~,~Pn~:~P1Lfl.l~

A. L 4 6 L A AL ALAL
C1 R i

:r.b.'T Qr? ???
C~I , ~MAL

??

'I:

98. Antique Coping, Fragments 99.

. ? ,." . .? ,,, . . . . , .
. . r. . . .

.o,
.vRla ,"w"l',
ir~. . . .I ~ .* t
..I ?"? .

I .. .

100. Bench Support, Fragments 101.

;C? rrr
?~F .?~r ?
c~~ ~; .O*?. '~.s ? ?s
~- ;
C~4.
n.~? i:??
r--

I?
I?I:S-????' -\,
.~???
': ?? r

?~ Y" IP ,,~
.J E"a~'' :: u
b'f:-'?- : I ) ?e
r\_

i ? . ..z~~glC~T-. ~~y~'" Y , ?- I:
?!. r.
r~~:t*
'! i
1? i ,,
*3,
i-?
.. ;;

~I ? .,. ?i..I ?ir U? f. _ L-?i~i"~~l"-~e~F

? 'V j(l !r; iP

'C ; *

~*~?1 f l?'r
103.
??~

i?c
JJ~.
ij'
jli~
b
'' d' .i.,?
I? i..

Y.?
i. , Y
I r?

102.
Funerary Reliefs, Fragments

Kurgunlu Monastery

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
105. Lion's Head

104. Fluted Column, Fragment

106. Sarcophagus Lid next to Fountain

107. Sarcophagus outside Dwelling House

Kur?unlu Monastery

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
Li
SPHASE 1 VISIBLE
HYPOTHETICAL

SPHASE 2

PHASE 3

r-j
I

O 1 2 3 4 5 6 7 8 9 10mn

108. Kursunlu Monaster

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
109. Interior, looking North. Column Shaft not in situ

110. Interior, looking West

111. Central and North Apses, Exterior
112. Narthex, Southwest Corner

Kur?unlu Monastery, Church

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
114. Northeast Column Footing

113. Prothesis Apse, Interior

115. Column Base not in situ

Kursunlu Monastery, Church

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
118. Kur?unlu Village. Capital from Monastery
119. Kursunlu Monastery. Church, Capital

121.

120. Post from Chancel
Screen 122.

Cornice Fragments

123. Door Lintel

Kursunlu Monastery

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
124. Kurgunlu, House of Siileyman Kose. Entablature, Fragment 125. Kurgunlu Monastery,
Church. Wall Skirting (?),
Fragment

127.

126.

128. 129.

Kur?unlu Monastery. Byzantine Carving, Fragments

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
5,.,

L.... . " r o""
- ._ " P 7
" ' .. -

131. Kurg
Fragment

132.

Kur?unlu Monastery. Carving, Fragments

133. Mullion Capital
Rk 'n ,K- n- 4ME Is ? m? A 1mom

135. Interior, looking East

134. Ambo Balustrade

Kurgunlu, Chapel Southwest of Monastery

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
136. View of Valley, looking South

137. Ionic Capital, Fragment Timanyo 138. Parapet Slab, Fragment

139. Sarcophagus and Architectural 140. St. Anne. Funerary Relief
Fragments at Village Fountain
Yenice

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
141. 142.

143. Yenice, St. Anne. Capitals 144.

146. Parapet Slab, Rear View

145. Parapet Slab

Yenice Village

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
i~~- iI; iii? i
N A-*

I-IF

? ..; ..

: -'~-" ~ rel~ CCla - Aw

I

(r E.- . .....

A,-*

4u,
.'i.

II

%0 '

147. Iznik

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
IN E
1IIII
- F'
ii iil I ' q.ABu~'m m

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
149. Gate

150. Gate, Lunette. Maximos Inscription

Tirilye, Medikion Monastery

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
151. Gate, Lunette. Inscriptions

152. Courtyard, Ionic Impost Capital

Tirilye, Medikion Monastery

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
153. Side A 154. Side B

Kurqunlu Monastery. Parapet Slab, Fragment

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 273

III. SOME REMARKS ON ARCHITECTURE monuments, since it is generally held that
the first clearly documented example of a
The churches we have been studying in
cross-in-square church was the Nea Ekklesia
this paper belong to three architectural
of Basil I (880) and the earliest standing
types: basilica on piers (Medikion), cross-in-
example the Theotokos church of the mon-
square (Fatih Camii, Pelekete, Kurgunlu),
astery of Lips (907).194 The latter can hardly be
and extended cross-in-square (Kemerli Kilise).
regarded, however, as marking an early stage
Regarding the first, we refrain from comment
in the evolution of the type: it is, with its
pending an examination of whatever remains
roof chapels, cantilevered passageways, and
may still be in existence.191a The cross-in-
scalloped pastophoria, a highly complex
square or four-column churches (to avoid the
building. There are, indeed, several indi-
hideous term 'quincunx' that has gained some cations that four-column churches were
currency in recent years) raise a more inter-
current at an earlier date: it is enough to
esting question. None of these churches is
quote here the church at Side which is prior,
strictly dated, yet there is considerable
perhaps by a considerable period, to the
circumstantial evidence that we are dealing
ninth century.195
with monuments of the late eighth or early Another consideration that has not been
ninth century.192 Pelekete was rebuilt at
generally taken into account is that the
about this time: Megas Agros, if it is the
construction of four-column churches de-
same as the monastery at Kurgunlu, was
pended on the supply of column shafts and
founded ca. 785, and, if Kurqunlu is the same
large capitals. In all three churches of this
as Polichnion, it was in existence by 781;
group that we have been studying these
the monastery of Trigleia, if it is the same
elements are reused. Even in a church as
as Fatih Camii, was in existence by 815. The
luxurious as that of the monastery of Lips
last exhibits a number of unusual, probably
the large capitals have been taken from a
archaic features, such as the single barrel
building of the fifth century and carefully
vault (instead of the normal three vaulted
mended, while the marble blocks used for the
bays) that covers the narthex, the projection
cornice were removed from a pagan cemetery
of the prothesis beyond the line of the north
of Cyzicus.'19 This suggests to us that the
wall,193 and the fact that-if Hasluck's plan
quarries of Prokonnesos may not have been
is to be trusted-the apses of the pastophoria
protruded further to the east than the main exploited at the time or, perhaps, exploited
apse. There is, furthermore, ample evidence on a very small scale.'97 The area of Trigleia
in contemporary saints' Vitae, such as those
and Sigriane, being situated very close to
the ruins of Cyzicus, had a ready supply of
of St. Niketas of Medikion, St. Ioannikios,
and St. Peter of Atroa, of an intensive large marble elements, and this may explain
the early adoption there of the four-column
building of monasteries in Bithynia, be-
type.
ginning about 780 and continuing through
The Kemerli Kilise is interesting in another
the first half of the ninth century.
context since in some respects-the length-
Historians of Byzantine architecture may
be surprised by our early dating of these 194 R. Krautheimer, Early Christian and
Byzantine Architecture (Harmondsworth, 1965),
191a Nothing remains of the foundations of 245ff., who misdates the Nea to 881 and the
the Medikion church on the surface. See Addi- Theotokos church to ca. 930 (p. 263).
tional Note infra. 195 See S. Eyice, "L'6glise cruciforme byzan-
192Hasluck, "Bithynica" (as in note 2 tine de Side en Pamphylie," Anatolia, 3 (1958),
supra), 290, had already suggested that the 35ff.
Fatih Camii was built between 780 and 813, 196 See C. Mango and E. J. W. Hawkins in
but he did so for the wrong reasons. DOP, 18 (1964), 303f.; 22 (1968), 182.
193 One may imagine that originally there 197 A study of these quarries has recently
were side aisles, somewhat after the manner of
been initiated by Dr. Nezih Firatll of the
the Koimesis church at Nicaea and St. Sophia,Istanbul Archaeological Museum. C. M. was
Thessalonica, in both of which, however, the privileged to visit them in Dr. Firatlih's com-
pastophoria open into the aisles. A more pany in the autumn of 1972 and did not observe
thorough examination of the building mayany evidence of mediaeval exploitation later
throw some light on this question. than the sixth century.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
274 CYRIL MANGO and IHOR ?EVCENKO
ening of the nave and the presence of firstsix,
sight to be of later date, demonstrate
instead of the traditional four columns-- this new trend in Bithynia.203 Here the stone
is laid in single courses, separated, by one,
it offers an analogy to the so-called "Mistra
type."'98 One may note also the interpene-two, or three courses of brick. This, essen-
tration of nave and sanctuary areas'99tially,
which is the system employed in the Kemerli
Millet considered a distinguishing featureKilise, and it reappears in certain Palaeologan
of the Greek as opposed to the Constanti- buildings at Istanbul, such as the south
nopolitan school.200 We have alreadychurch com- of the monastery of Lips and the
mented on the use of sawtooth which is outer narthex of Kilise (Molla Giirani) Camii.
particularly common in Greece. TheseThe con-prevalence of this kind of construction
siderations may be taken, within the termsin Bithynia is particularly significant since
of Millet's classification, to indicate a Greek
it was in Bithynia that the Ottoman state
was founded and Ottoman architecture
influence on Kemerli Kilise. For the present,
however, it would be wiser not to make any
formed. It is not surprising, therefore, that
such assertion since our knowledge ofthe Las- earliest major Ottoman mosque that
carid architecture in Asia Minor is almost nil,
has come down to us in good condition, the
not to speak of the questionable validity Muradiye Camii at (ekirge (Bursa) of
of
Millet's regional "schools." ca. 1366-85, should exhibit this charac-
A final point of interest concerns the teristic
type late Byzantine kind of masonry
of masonry used in the Kemerli Kilise.(single
The courses of stone alternating with
traditional Byzantine system, such three
as wecourses of brick) and even strings of
have encountered it at Pelekete and Kur- sawtooth used as a framing motif for arches.2?4
?unlu, was the parement arase, consisting
Onceofadopted by the Turks, this technique
(usually) five successive courses of brick
continued to be widely used as late as the
alternating with several courses of stone.
eighteenth century.
This method began to be modified in the
eleventh century, first, by the introduction ADDITIONAL NOTE
of the "concealed" brick courses,201 and
second, by a more haphazard alternation In
of late May of 1973, Ihor 8evienko made
a
brick and stone. The domed pastophoria brief visit to Medikion near Tirilye and
Kursunlu-previously
of St. Sophia at Nicaea (fig. 147), which inspected by both of
us--as well as to the village of Dutliman
Schneider ascribes to the eleventh century,202
situated between Bandirma and Yenice,
although their masonry would appear on
which we had not visited in 1968-72. The
198 See H. Hallensleben, "Untersuchungen
following remarks are intended to comple-
zur Genesis und Typologie des 'Mistratypus',"
Marburger Jahrbuch fiir Kunstwissenschaft,ment
18 the information contained in our article.
(1969), 105 ff., who challenges some previously
held views. An important monument in this 1. Site of the Monastery of Medikion205
connection is the Siileyman Pasa Camii at At present, the site of the monastery (cf.
Vize dated roughly in the ninth century by
C. Mango, "The Byzantine Church at Vize .. .," 148 for a general view from the north-
fig.
Z VI, 11 (1968), 9ff., and in the thirteenth- west) is used as a chicken farm. Nothing
fourteenth century (because of its similarity to remains from the basilical katholikon described
the "Mistra type") by S. Eyice, "Les monu- by Covel, Hasluck, and Panienko. According
ments byzantins de la Thrace turque," CorsiRav,
18 (1971), 293ff. The large size of this church, 203 Cf. also for the thirteenth century the
its heavy proportions and stark exterior, entirely Lascarid palace at Nymphaion, the upper
devoid of plastic articulation and ornament, portion of which is built of single courses of
speak against a late date. irregular stones, often separated by vertical
199 The eastern arm of the cross was clearly bricks, and an uneven number (one to five) of
included within the sanctuary. The templon brick courses: S. Eyice, "Izmir yakininda
must have followed the line of the two eastern Kemalpasa (Nif) da Laskaris'ler sarayl," Belle-
columns as shown by Covel (fig. 2). ten, Tiirk Tarih Kurumu, 25 (1961), 1ff.
200 L'dcole grecque (as in note 34 supra), 55 ff. 204 See A. Gabriel, Une capitale turque:
201 See p. 122 supra. Brousse-Bursa (Paris, 1958), I, 56, 60; II,
202 Die rdmischen und byzantinischen Denk- pl. xvff.
mdler von Iznik-Nicaea (Berlin, 1943), 14f. 205 See pp. 240-42 supra.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 275

Funerary Inscriptions of Maximos
to one local informant,206 a large building
(presumably the church of Medikion) was
Reused part of marble architrave, cut to
still on the site of the monastery as late as
accommodate the inscription; height
1942; at that time the building was occupied
0.13 m.; length 0.67 m.; height of letters
by the military. In 1951 or 1952, a printing
0.015-0.025 m.; cross centered in upper
shop was constructed at Tirilye from the
part of block (this cross is contemporary
stones of the structure. The present owner
with the inscription, as line 2 runs over
of the site, Orhan Gazi, bought it in 1953.
its lower part).
In 1972, he erected the large chickenhouse
on the emplacement of the razed building Letters: I; ; E1; E; V; W and 0W.
so that that emplacement is now "inside the
Date: fifth century?
chickenhouse and toward the middle of it."
The present compound follows the line of old 1 OI ONION OLKO cross 5 Map0ou Kiai
2 EvurpoyljS co ya<(~P)pp>os pou KUvov
walls, which, as we shall presently see, must [ou8Ev EX i Els To P opIV ToUro- To Ev
date from 1801 at the latest; its approximate
3 T-n yuvOaK1t KI TOItS TEKVOIS
length is 70 meters, and its width 40 meters. 4 -mlSov To E91KEV EV TCO lV0pl1C0
The only preserved Byzantine remains of [Xapiv auro v Errwac
Medikion seen in 1973 were parts of the To make sense of this inscription, which
monastery gate and an Ionic impost capital. we consider to be essentially complete, we
The gate (cf. fig. 149) has two inscriptions propose to read its lines in the sequence 1,
in its lunette (cf. fig. 151). The larger inscrip- 3 (lines containing large letters), followed by
tion, which has been recorded with varying 2, 4 (lines containing small letters), and offer
degrees of precision by Hasluck and Pan- the following tentative translation:
&enko, refers to a restoration of the monastery
"Permanent abode of Maximos and for
"from the very foundations" in 1801.207 How-
ever, the wall on both sides of the gate, up his wife and children. Eustorgios my son-
to the level of the wooden lintel, seems to be in-law [?] has nothing in common with
late Byzantine or early Turkish, as it is made this tomb; the one child which he deposited
of alternating courses of brick (one to three in the tomb, I did it as a favor to it
bricks thick) and of stone, with some bricks [him ?]."
standing upright; the wall above, of irregular We imagine that Maximos' inscription
construction, may be considerably later. Still, first contained lines 1 and 3 only; in the
it is remarkable that the mdchicoulis over usual fashion, he reserved his tomb for him-
the gate (it has its counterpart on the inside self, his wife, and children. At some later
of the gate) should have been constructed as date, one of his grandchildren died and was
late as 1801, when the monks could no longer buried in the tomb. To make sure that this
have entertained thoughts of resisting an ene- precedent would give no right of burial to
my. Such features are more likely to have been his son-in-law Eustorgios, Maximos added
put up in times of mounting Turkish danger, lines 2 and 4 of the inscription. In spite of
in the thirteenth and fourteenth centuries. his name, Eustorgios must have been disliked
The second inscription (fig. 150) was insert- by his father-in-law.
ed in the lunette above the one of 1801, 1: OI~ NION or INION, if the first O is
probably in that same year. It escaped Has- an ornament, corresponding to a similar
luck and Panienko, perhaps because it was feature at the right end of line 1, presents
plastered over in their time. We reproduce difficulties. We conjecture that it stands for
it here:
6 aickvioS.
206 An officer of the Belediye of Tirilye, Mr. On olxos as a technical term for a tomb
Hasim Aks6z. or a part of a tomb, cf. J. Kubiliska, Les
207 Hasluck: Cyzicus..., p. 61; Pan'enko,monuments f/undraires dans les inscriptions
IRAIK, 15 (1911), 274; cf. p. 241 and notes
grecques de l'Asie Mineure (Warsaw, 1968),
40 and 42 supra. The inscription is now partly
whitewashed; we tried to clean it up. It runs113-14, and passim; see index s.v. For the
formula "eternal abode" on tombs, cf. J. Keil
as follows: 1801 KT-ra piva MCafov I dcxKevfQiSa
and A. v. Premerstein, "Bericht ilber eine
9v pcapov 0 rTapoiocoa 'CVo TOiJ M151K1I\IOV/.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
276 CYRIL MANGO and IHOR ?EVCENKO
zweite Reise in Lydien," Denkschriftenmust
der have followed the plan of St. Michael's.
K. Akademie der Wissenschaften, Wien, In 44
such a way, the truncated basilical form
of Medikion's katholikon would find its most
(1911), no. 155, line 2: EIS advEoov olKov, and
natural explanation.
four further examples (one of them Christian),
ibid. Cf. J. and L. Robert, Bull. Epigraphique
(1972), no. 507. 2. Kurfunlu: Village and Mehmet Kbse's
Farm210
2: Evo-rpoyrlj stands for E0o-r6pytoS, and
co yapos for 6 yappp6s. The word is some- In the village a fragment of the base of
an antique altar or a statue (lower dimen-
times misspelled; cf. yppp6s, yppp6o in
Keil-Premerstein, op cit., nos. 153, 154. sions: 0.83 x 0.86 m.) was recorded. At
pip6ptv: for wrvp6ptov "tomb," cf., present,
e.g., it stands near the village fountain,
J. and L. Robert, Bull. Epigraphique (1972),
but it was brought there from Mehmet Kdse's
nos. 453 and 473, and the pep6pIov farm.ofThere it may have served as the base
a deacon, BCH, 17 (1893), 290 (Phrygia),
for the altar in the bema of the church.
and of a layman, MAMA, 5 (1937), no. In
77the courtyard of Lokman Bey's house
and pl. 28 (area of Eskisehir); cf. Keil-Pre-
a roundel of green marble, probably a part
merstein, op. cit., no. 174 (pvrlj6ptov). of an opus sectile pavement, was recorded.
The remnants of the pavement in situ
4: For the spellings rnrm8fov and rrirTbca,
cf. our article, "Some Recently Acquired behind the house (cf. p. 249 supra have been
Byzantine Inscriptions at the Istanbul almost
Ar- entirely destroyed, with one tiny
chaeological Museum," in Arkeoloji Mi zeleri
fragment and some mortar bed remaining.
Yillhgs (1973), commentary to nos. 13, line 7,
The most interesting new find at Mehmet
16, line 3, and 28, lines 3/4. Cf. also n'rrlv'icav
KSse's farm, the putative site of Megas Agros
in an inscription from Phrygia, BCH,
or 17
Polichnion, was a fragment of a parapet
(1893), 289 = no. 96. slab carved on both sides. It was discovered
The Maximos inscription, which we date
southeast of the church, near a big fallen
to the fifth century, must have come from
fragment of the church wall. The ornament
the site, or at least the vicinity, of Medikion,
on both sides seems to be contemporary. On
and attests to the existence of a burial ground
one side (cf. fig. 153) is carved a cross with
there centuries before the foundation of the arms and circular loops attached to
flaring
monastery by Nikephoros. the ends. Two leaves sprout from its base.
The large Ionic impost capital (fig. 152)
On the other side (cf. fig. 154) is a simplified
lying today in the court of the present chrism in a disk.21' Although we are aware
enclosure tells a similar story, as it attests
of the pitfalls in dating Byzantine carved
slabs between the sixth and the late ninth
to the existence of a large sixth-century208
church on or near the site of Medikion.centuries,i22
This we are tempted to connect our
church, we believe, was the very church of
210 Kursunlu was visited in the welcome
St. Michael which Nikephoros, the founder
of Medikion, found in ruins and which company of Dr. Nezih Firatll of the Archaeolo-
heMuseum
gical in Istanbul.
rebuilt as the core of his new foundation211 Dimensions of the fragment: maximum
height 0.77 m.; maximum length 0.66 m.; thick-
ca. 780.209 In his reconstruction, Nikephoros
ness 0.075 m.; diameter of disk 0.50 m.; length
208 For parallels to our capital, cf. Kautzsch, of existing upright arm of the cross 0.63 m.
Kapitellstudien (as in note 85 supra), no. 566e 212 With the exception of two recently dis-
(Ephesus); I. Nikolajevi6-Stojkovi6, Ranovizan- covered (gallery ?) slab fragments from St. Irene
tiska arhitektonska dekorativna plastika u Make- at Istanbul, datable to the years 741-75, there
doniji, Srbiji i Crnoj Gori (Belgrade, 1957), is no securely dated Byzantine carved slab
fig. 65 (Stobi); eadem, "Chapiteaux de Carihin between the late sixth century and the year
Grad," Actes du Xe Congrhs d'Etudes Byzan- 873-74 (church of Skripou, Boeotia). Cf. Thilo
tines (Istanbul, 1957), fig. 1 (Cari6in Grad). Ionic Ulbert, Studien zur dehorativen Reliefplastik des
impost capitals are dated to the first half of the dstlichen Mittelmeerraumes (Schrankenplatten des
sixth century, cf. F. W. Deichmann, Studien 4.-10. Jahrhunderts) (Munich, 1969), 46 and
zur Architektur Konstantinopels (Baden-Baden, 61-62. On the slabs of St. Irene, cf. also idem,
1956), 48-49. "Untersuchungen zu den byzantinischen Relief-
209 Cf. Halkin, "La Vie de St. Nic6phore" (as platten des 6. bis 8. Jahrhunderts," Istanbuler
in note 45 supra) and p. 241 supra. Mitteilungen, 19/20 (1969-70), 349-50 and pl. 72.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms
SOME CHURCH BUILDINGS ON THE SEA OF MARMARA 277

the of
slab with the original church structures jetty and the Byzantine tower which
were
Polichnion or Megas Agros, that is, with thethe main ruins noticed by Panienko.
period between 760 and 800. According to a local informant, the site of
the tower is now covered by the school, a
Among other objects recorded at the farm,
we may mention a fifth-century capital barrack
(in made of sheet metal situated near
the shore. A new breakwater is being con-
a field southeast of the house) and a cross
carved in relief on a square of marble structed
(side: near the school, and fragments of a
0.17 m.) and immured into one of the mullion (0.49 m. high, 0.39 m. wide, and
buttresses of the retaining wall at the sea- 0.20 m. thick) and a doorpost with moldings
shore. lay among stones to be used in the construc-
The ruins situated southwest and uphill tion. Near the school were found a fifth-
from the farm, which may be those of the century capital and two fragments of a Late
monastery of Christopher (see pp. 258 and Antique tomb, one of which bore the inscrip-
261 supra) are now completely overgrown; tion [KTrEOKnE]lraoEv ' racv Tj. Its lettering
the growth obscures the fact that originally points to a third-century date.215 The kastron
the complex looked down on a perfect little on the seashore was the second ruin described
harbor which connected it with the outside by Panienko. Its site seems to lie about
world.
three-quarters of a mile east of Dutlimanim.
Sculptural fragments from the farm which All that was found there was a column shaft
in 1972 were preserved in the garden of theof grey stone and a fragment of worked
municipal building of Karacabey (see p. 257marble (entablature?). Worked stones lying
above) are no longer there; reportedly, theyunder water may be the remnants of a harbor.
are now in the possession of the kaymakamrA plateau, now a wheat field, extending above
of that town. In the garden of the municipal- the shoreline of the "harbor" may have been
ity, only one inscription which had not beenthe site of that kastron, unless it is that of
previously there was found; it mentions thethe monastery of St. Theodore, the ruins of
Emperor Trajan.213 which, according to Panienko, were dis-
mantled in 1909-10.
3. Dutlimant (Sykamia) In the village proper were to be seen a
The small settlement of Dutlimanl (40follis of Justinian struck in Cyzicus in 546;
houses, 250 inhabitants-some of them a fragment of a capital, identical with that
resettled from Drama in Greece after 1922)found at the school, lying in a courtyard;
communicates with the outside world mostly and two inscribed marble slabs of the past
century, one being a tombstone, one prob-
by sea, as the road from Bandirma, not quite
completed, is not passable by car in badably commemorating a construction.216 Both
weather. Dutlimanl is the former Greek stones were imbedded in thresholds of houses;
we were told that they come from "the
village of Sykamia, briefly visited by Pan-
cenko in 1910. In Panienko's time, Sykamia
monastery," surely that of St. Theodore. In
was merely thirty to forty years old, but
conclusion, Dutlimani stands on or near a
it must have been settled on an old site,
Byzantine site. Judging by our brief visit,
for he found some, though few, Byzantine
this site was not significant and there is no
remains there.214 Today nothing remainsreason
of to look for Theophanes' Hiereia there.
213 It is a relief with the lower part of the
figure of the Emperor above the inscription, spoke only of a "new, utterly poor church"
and a row of chlamydati and a dance scene and stated that ruins considered by the local
below it. Dimensions: height 0.895 m.; width peasants to be those of a church were in fact
0.51 m.; thickness, 0.14 m. The first two lines those of a tower. He does admit that some
run: 'Yw-p -ri" afrroKpdropoS [T]patav[oO Kc]il foundations near the ruins of the tower "may"
aapos v(dKil K(al) alcoviou 8tapxovijs. have belonged to a church.
214 IRAIK, 15 (1911), 262-64. The Calendar 215 Dimensions of the inscribed fragment:
of the National Philanthropic Establishments for height 0.027 m., length 0.84 m., height of
1905 (see p. 263 and note 136 supra), 181, letters 0.05 m.
mentions the village "Skamniais" and records 21 The tombstone is of 1884; the other
a "most ancient church of St. George" there-- inscription runs: 0fT6 ijyouvIvov NtKT1p6pOV
apparently functioning. However, Pan'enko 'HAtdS6ov acy(<o0>a. 6 1894.

This content downloaded from 194.27.33.66 on Fri, 09 Jun 2017 10:03:24 UTC
All use subject to http://about.jstor.org/terms