Christopher Walter

IC XC NI KA. The apotropaic Function of the victorious Cross

In: Revue des études byzantines, tome 55, 1997. pp. 193-220.

Résumé
REB 55 1997 France p. 193-220
C. Walter, IC XC NI ΚΑ. The apotropaïc Function of the victorious Cross. — Dans cet article, l'auteur tente de réactualiser la
recherche que les regrettés Anatole Frolow et Gordana Babic avaient entreprise sur la place de la formule IC XC NI KA dans l'art
byzantin. Il fournit un important répertoire d'exemplaires de cette formule dans divers contextes, ainsi que des légendes ou des
acronymes qui peuvent la remplacer ou la compléter. Il rejette l'hypothèse qu'elle ait été originellement associée à la pratique
eucharistique et à la propagande iconoclaste. En suivant l'évolution de son usage, il observe que son emploi premier, pour
affirmer une victoire, généralement mais non nécessairement impériale, a conduit à des pratiques apotropaïques. Bien que
certains des acronymes, qui de plus en plus fréquemment accompagnent la formule dans l'art byzantin tardif, soient difficiles à
déchiffrer, l'auteur ne considère pas qu'ils aient été conçus en règle générale comme des énigmes.

Abstract
In this article the author attempts to bring up to date the research undertaken by the late Anatole Frolow and the late Gordana
Babic on the place of the formula IC XC NI ΚΑ in Byzantine art. He provides an extended repertory of representations of this
formula in different media, together with other legends or acronyms which may replace or supplement it. He rejects its early
association with Eucharistie practice and Iconoclast propaganda. In tracing the development of its use, he observes that its early
exploitation as an affirmation of victory, usually but not necessarily imperial, gave way to apotropaic practices. Although some of
the acronyms which increasingly accompanied the formula in later Byzantine art are hard to decode, the author does not consider
that they were intended regularly to be arcane.

Citer ce document / Cite this document :

Walter Christopher. IC XC NI KA. The apotropaic Function of the victorious Cross. In: Revue des études byzantines, tome 55,
1997. pp. 193-220.

http://www.persee.fr/web/revues/home/prescript/article/rebyz_0766-5598_1997_num_55_1_1940
 IC XC NI ΚΑ
THE APOTROPAIC FUNCTION
OF THE VICTORIOUS CROSS

Christopher WALTER*

In Memory of Anatole Frolow and Gordana Babic

Résumé. In this article the author attempts to bring up to date the research undertaken
by the late Anatole Frolow and the late Gordana Babic on the place of the formula IC XC
NI ΚΑ in Byzantine art. He provides an extended repertory of representations of this
formula in different media, together with other legends or acronyms which may replace or
supplement it. He rejects its early association with Eucharistie practice and Iconoclast
propaganda. In tracing the development of its use, he observes that its early exploitation
as an affirmation of victory, usually but not necessarily imperial, gave way to apotropaic
practices. Although some of the acronyms which increasingly accompanied the formula in
later Byzantine art are hard to decode, the author does not consider that they were
intended regularly to be arcane.

It may come as a surprise to the passerby in an Athenian subway when

he observes, among the panoply of graffiti , the letters IC XC NI KA
placed between the arms of a cross. It is less surprising to find the same
device imprinted on the prosphora used in the Greek liturgy. The
problem here is rather to know when the practice began. The device is
certainly ancient, and its basic meaning - Jesus Christ conquers - has not
changed. Nevertheless, it has been used with varying connotations and in
different contexts, to which may be related the medium in which it
appears. It is now forty years since A. Frolow published what is still the
best general study of the device '. Curiously a number of scholars who

* I gratefully acknowledge help received from Suzy Dufrenne (Paris), Nicole Thierry
(Étampes), Gojko Subotic and Radivoj Radie (Belgrade) and George Gerov (Sofia). For
Greek accentuation I follow the conventional rules for Byzantine texts.
1. A. Frolow, IC XC NI ΚΑ, Byzantiiwslavica 17, 1956, p. 98-1 13.

Revue des Études Byzantines 5.\ 1997, p. 193-220.

194 CHRISTOPHERWALTER

have written on one or other aspect of the device since Frolow's time
were apparently unaware of the existence of his study. Their ignorance
of it may be in part responsible for the implausible nature of some of
their proposals. Furthermore, much new material is now available that
was not so in Frolow's time. It seems, therefore, that an attempt at a new
synthesis is justified. The uses to which the device was put will be
grouped according to the medium in which it appears. Although a
developed repertory is provided, the author does not, of course, claim
that it is exhaustive.

The origins of the device

The Cross, whether represented figuratively or merely as a sign, was
perhaps the most pervasive subject in Byzantine art 2. Like its relics and
like the gesture of the sign of the cross, it was deemed to be powerful
apotropaically. The earliest legend with which the Cross was combined
was probably Constantine's celebrated formula : τούτω νίκα 3. This
legend, the first to be stamped in Greek characters on Byzantine coins,
was used on ihefollis of Constans II from 641 to 658 4 and recurs even
as late as the reign of Basil II on the miliaresion 5. It may be observed on
a cross at the entrance to a ruined church at Akören, fifty-two kilometres
north of Adana. The church may be dated by an inscription to 525. 6
(Figure 1). It may also be observed, with a cross, on the reverse of a
Solomonic amulet 7. No certain date can be given to this object, which is,
however, unlikely to be later than the period of Iconoclasm. The earliest
surviving representation of Constantine's vision, with the legend, is in
Paris, graec. 510, f. 440 8. The scene here is a narrative one, while the

2. For a detailed general account, see Erich and Erika Dinkler, Kreuz I, Lexikon zur
byzantinischen Kunst 5 (1991), p. 1- 219 (vorikonoklastisch) ; G. Galavaris, Kreuz II,
ibidem, p. 219-284 (nachikonoklastisch).
3. Eusebius, De vita Constantini (Clavis 3496), I 28, PG 20, col. 944. The earliest
reference to the prodigy is made by Lactantius, La mort des persécuteurs, I, xliv, 13, edi
ted J. Moreau, Paris 1954, p. 127. In his account of the battle of the Milvian Bridge,
Eusebius, Historia ecclesiastica IX, ix, edited J. Bardy, Paris 1958, III, p. 60-65, espe
cially p. 61 note 6, says nothing.
4. C. Morrisson, Catalogue des monnaies byzantines de la Bibliothèque Nationale,
Paris 1970, p. 331.
5. ibidem, p. 609.
6. Tabula imperii biz.antini, V, Kilikien und Isaurien, edited F. Hild and II.
Hei.enkemper, Vienna 1990, 1, p. 168-169, 2, fig. 40.
7. Ν. Thierry, Mentalité et formulation iconoclastes en Anatolie, Journal des savants,
avril-juin 1976, p. 101-104 ; Eadem, Le culte de la Croix dans l'empire byzantin du Vile
siècle au Xe siècle dans ses rapports avec la guerre contre l'infidèle. Nouveaux témoi
gnages archéologiques, Rivista di studi bizantini i slavi, Miscellanea Agostino Pertusi, 1 ,
1980, p. 215, fig. 7.
8. H. Omont, Fac-similés des miniatures des plus anciens manuscrits grecs de la
Bibliothèque Nationale, Paris 1902, plate LIX, p. 31 (about 880). Constantine's vision
was not often represented, but see the late example (dated 1466) in the church of the Holy
Cross, Platanistasa, Cyprus, A. et J. Stylianou, By This Conquer, Nicosia 1971, p. 69
 THE APOTROPAIC FUNCTION OF THE VICTORIOUS CROSS 1 95

two crosses with the IC XC NI Κ A legend in the same manuscript, f. Bv

and f. C, are rather to be interpreted as apotropaic 9. This matter will be
discussed later. The manuscript dates from about 880, but the device was
certainly already being used before that date and was beginning to
replace the Constantinian formula. Frolow wrote of a «simple
changement de graphie» l0. However, even if this explanation is
plausible for its origin, the practice of abbreviating Jesus Christ and of
placing the letters between the arms of the cross, while not universal,
was too common not to have been deliberately standardized. When the
temple of Philae in Egypt was converted into a Christian church under
Justinian, a cross was carved by the door of the nave on the south side
and beneath it the legend : ό σταυρός ένίκησεν άεί νίκα. Clearly this is
close to the device". Another Justinianic monument is the cistern at
Madaba, with the device represented at each angle, and a dedicatory
inscription : άνεκαινίσθη ύπο Ιουστινιανού" Αύτοκράτορος των
'Ρωμαίων 12. Frolow was sceptical, perhaps unduly so, as to the
authenticity of the «Justinianic» crosses 13. For him the earliest authentic
example of the device occurs in an inscription commemorating the
restoration of the ramparts of Constantinople under Leo III and
Constantine IV in 741/2 14. Frolow did not know that they also
introduced the device on their joint silver miliaresion, minted after the
coronation of Constantine IV as co-emperor on March 31st, 720 15.

The device on coinage

At the beginning of the eighth century Byzantine emperors had the
legend victoria Augusti stamped on their coins 16. This was the case for
Philippicus Bardanus (711-713), Anastasius II (713-715), Theodosius III
(715-717), and, on his solidi, for Leo III (717-741) 17. Leo's introduction

fig. 2. In a segment of the sky the legend EN ΤΟΥΤΩ ΝΙΚΑ accompanies a cross made
up of stars.
9. Ibidem (Omont) plates XVII, XVIII, p. 12-13.
10. Frolow, art. cit. (note 1), p. 102.
I 1 . See below, note 104.
12. H. Leclercq, Madaba, DACL 10 (1931), 860; P. Gatier, Inscriptions de la
Jordanie. Il, Région centrale, Paris 1986, p. 126-127 ; M. Piccirillo, Chiese e mosaici di
Madaba, Jerusalem 1989, p. 118. neither Gatier nor Piccirillo takes up the question whet
her the device is later in date than the inscription.
13. Frolow, art. cit. (note 1), p. 109 note 66.
14. H. Lietzmann, Die Landmauer von Konstantinopel, Berlin 1929, p. 31, n° 13,
referring to A. Van Millingen, Byzantine Constantinople, the Walls of the City and
Adjoining Historical Sites, London 1899, p. 98. In Van Millingen's time the tower had
already been destroyed, so that the inscription was only known from a photograph.
Van Millingen notes another example of the device, p. 101, accompanying another resto
ration of the walls after the earthquake of 995.
15. Morrisson. op. cit. (note 4), p. 450-45 1
16. Ibidem, p. 437-439.
.

17. Ibidem, p. 445-447.

196 CHRISTOPHERWALTER

of the cross and IC XC NI ΚΑ exclusively on the silver miliaresion, no

doubt minted after his son's coronation for a largesse, started a precedent
which was followed regularly by his successors on the same occasion :
Constantine V (741-755) with Leo18, Artabasdus (742-743) with
Nicephorus l9, Leo IV (775-778) with Constantine20, and Constantine VI
with Irene21. For Nicephorus I the miliaresion is not known. However,
he had our device stamped on the solidus and tremissis 22. The practice of
stamping the device on the silver miliaresion was taken up again by
Michael I (811-813) with Theophylact 23, by Leo V (813-820) with
Constantine24, by Michael II (820-829) with Theophilus25, by
Theophilus (829-842) with Constantine 26, his elder son who died young.
After Constantine's death, Theophilus continued to mint the silver
miliaresion but with his name alone, a departure from earlier practice 21 .
Thus a century after its inception either the original practice of
minting the silver miliaresion when a co-emperor was crowned was no
longer considered to be relevant or its original purpose had been
forgotten. Nevertheless our device continued to be stamped on coins for
some time. Michael III (842-867) used it on his miliaresion, bearing his
name alone or also those of his mother Theodora and his aunt Thecla28.
Basil I (867-886) associated his elder son Constantine (died 879) with
him but not his younger son Leo29. Leo VI (886-912) figured alone or
with his son Constantine30. Alexander I (912-913) figured alone31. For
Constantine VII (913-959) five types of miliaresion with the device are
known : Constantine figures alone or with a co-emperor32. There is no
miliaresion for Romanus II (959-963) 33. There are several types for
Basil II (963-1025), which vary according to the co-emperor who was
reigning at the time 34. There also exists for Basil II, as already noted, a
variant with the Constantinian formula.
The miliaresion disappeared after 1080, along with the gold nomisma.
Moreover, it was not replaced subsequently to Alexius I's coinage

1 8. Ibidem, p. 469.
19. Ibidem, p. 480.
20. Ibidem, p. 486.
21. Ibidem, p. 492.
22. Ibidem, p. 499-500.
23. Ibidem, p. 503.
24. Ibidem, p. 507.
25. Ibidem, p. 520.
26. Ibidem, p. 525.
27. Ibidem, p. 526. According to Ph. Grierson, Byzantine Coinage, Washington DC
1983, p. 10, the miliaresion now became a regular part of the currency.
28. Ibidem, p. 534-535.
29. Ibidem, p. 543.
30. Ibidem, p. 553.
3 1 . Ibidem, p. 559.
32. Ibidem, p. 571-573.
33. Ibidem, p. 580.
34. Ibidem p. 591,595, 609.
 THE APOTROPAIC FUNCTION OF THE VICTORIOUS CROSS 1 97

reform of 109235. For the intermediate decades between Basil IPs death
and 1 080 the miliaresion was minted but our device was replaced by an
invocation of the Virgin. Finally attention should be drawn to a puzzling
anonymous coin, which Hendy attributed to the early years of the reign
of Alexius I (1081-1092) 36. It has on its obverse a representation of
Christ's bust, and on the reverse our device.
A survey, however tedious, of all the imperial mintings which used
our device has the advantage of enabling us to make confidently some
positive as well as negative affirmations. First of all, while not excluding
necessarily the use of our device in more restrictively religious contexts,
it was certainly exploited imperially, for it is notorious that coins were
one of the principal instruments of imperial propaganda. However, to
take the step, as some scholars have done, of attributing the introduction
of our device on the silver miliaresion to Iconoclast emperors is
temerarious. Constantine IV and his father Leo III admittedly espoused
the Iconoclast cause. However, as Cécile Morrisson has pointed out, the
silver miliaresion stamped with our device was minted six years before
Leo Ill's first decree against images37. Subsequently it appeared on the
coins of Iconophile rulers (Artabasdus, Constantine VI and Irene) as well
as Iconoclast rulers. Convenient as it may have been to have available an
image of Christ which was at once orthodox and inoffensive to
Iconoclasts, this would not explain adequately why our device was
introduced on Leo Ill's miliaresion. It is much more plausible,
particularly if the use of the device on Leo Ill's coins is taken in
conjunction with its use in the inscription on the walls of Constantinople,
to suppose that he - and his successors - were affirming their confidence
in victory over the Arabs 38. As André Grabar observed long ago, neither
Iconophiles or Iconoclasts used coins for propaganda purposes against
their antagonists 39. A specific connection between our device and
Iconoclast teaching on images can be safely ruled out.
At first there must have been some reason for associating this gesture
of defiance to the Arabs with the coronation of a co-emperor. Possibly
the explanation was that the coronation of a co-emperor was in itself an
affirmation that the dynasty was secure of its future, another act of
defiance to the Arabs. By the early ninth century the original reasons for
associating our device with the coronation of a co-emperor may well
have been forgotten. Moreover the device may have been losing its
imperial connotations of triumph and victory. The religious mentality of

35. Griekson, op. cit. (note 26), p. 1 1 ; J.-C. Cheynet, Quelques remarques sur le culte
de la Croix en Asie Mineure au xe siècle, Histoire et culture chrétienne, Mélanges M.
Marchasson, Paris 1992, p. 72.
36. Morrisson, II, p. 605 ; M. F. Hendy, Coinage and Money in the Byzantine Empire
(1081-1261), Washington DC 1969, p. 75, plate 2 n° 22.
37. Morrisson, p. 450.
38. The silver miliaresion may have been originally an imitation of the dirhem of Abd-
al-Mali (695-698), ibidem.
39. A. Grabar, L'iconoclasme bvzantin, dossier archéologique, Paris 1957, p. 1 19-129.
198 CHRISTOPHERWALTER

Byzantium was undergoing a change. The confident affirmation of

certain victory, thanks to the protection of Christ and his Cross, gave
way to a more modest prayer for help. This change coincided with the
growth of the cult of the Mother of God. In order to seek her
intercession, her image replaced our device on the last issues of the
silver miliaresion.

The device in the Byzantine liturgy

In the contemporary Greek Byzantine liturgy, according to Brightman,
the pieces of the prosphora destined to be consumed at communion are
placed crosswise on the paten according to the formula IC XC NI Κ A 40.
The earliest reference which Frolow could find to the use of the device
in the liturgy was in the celebrated edition of Goar dating from the
seventeenth century41. Goar must have been following an earlier Greek
text, but this has not been identified. There is no reference to the device
in medieval texts of the rite of the prothesis. References are indeed
known to the celebrant making the sign of the cross with his lance over
the prosphora, for example in the Historia ecclesiastical. Theodore
Studite also refers to the sign of the cross over the prosphora, a practice
the origin of which he attributes to Basil (not to John Chrysostom) 43.
Pseudo-Sophronius wrote of the crucifixion of Christ on Golgotha «as in
the holy prothesis»44. It requires some stretch of the imagination to see
here a reference to our device being stamped on the prosphora. There
remains only the difficult phrase of Symeon of Thessaloniki : ό σταυρός
η αυτός ό Σωτήρ έξεικονιζόμενος 45. Should one translate : «the Cross
or the Saviour himself (is) typified» rather than «represented» ? Either
way there is no clear allusion to the letters IC XC NI Κ A . Frolow noted
only one reference - and that a Western one - to an image of Christ on

40. F.E. Brightman, Liturgies Eastern and Western, Oxford 1894, p. 393, και τίθησιν
αύτας έν τω άγίω δίσκω σταυροειδώς IC XC NI ΚΑ.
41. Frolow, art. cit. (note 1), p. 109 note 66; J. Goar, Εύχολόγιον seu Rituale
Graecorum, Paris 1647, p. 117.
42. F.E. Brightman, The Historia Mystagogica and Other Greek Commentaries on the
Byzantine Liturgy, Journal of Theological Studies 9, 1908, p. 263-264, especially § 28,
§30, § 31b.
43. Theodore Studite, Adversus Iconomachos, I, PG 99, col. 489, To δε
σφραγίζεσθαι την προσφοράν.
44. Pseudo-Sophronius, Commentarius liturgicus, 8, PG 87, col. 3988-3989.
45. Symeon of Thessaloniki, De sacra liturgica, 88, PG 155, col. 265-268. See in
general, P. De Meester, Les origines et le développement du texte grec de la liturgie de
S. Jean Chrysostome, Χρυσοστομικά, Rome 1908, p. 305; Idem, Genèse, sources et
développement du texte grec de la liturgie de S. Jean Chrysostome, Rome 1908, p. 63-64 ;
R. Engdahl, Die Proskomodie der Liturgien des Chrysostomos und Basilius während des
Mittelalters, Beiträge zur Kenntnis der byzantinischen Liturgie, Berlin 1908, p. 87-149 ; II
commentario liturgico di S. Germano patriarca Constantinopolitana e la versione latina di
Anastasio bibliotecario, edited Ν. Borgia, Studi liturgici, I (Roma e oriente, I),
Grottaferrata 1912, p. 19-20.
 THE APOTROPAIC FUNCTION OF THE VICTORIOUS CROSS 1 99

the host46. Honorius of Autun drew an analogy between the imperial

denarius, on which the emperor's name and image were inscribed, and
the host : «Ita imago Domini cum litteris in hoc pane exprimitur, quia in
denario imago et nomen scribitur, et per hunc panem imago Dei in nobis
i-eparatur...»47 It is to be noted that in each of the three cases the verb
changes, from scribitur for the emperor on the coin to exprimitur for
Christ on the host, to reparatur for the image of God in us. It is not clear
whether or not exprimitur should be translated, like scribitur, as
inscribed.
It is against this background of extremely tenuous evidence for the use
of our device in the liturgy that the subject of bread stamps should be
examined. Since Frolow (to whose study they make no allusion) two
scholars have written on this subject. The first was George Galavaris,
who prudently observed with regard to the hypothesis that the prosphora
were stamped with our device from an early date, that «of the number of
bread particles and whether the bread was stamped with special symbols,
we know nothing that can be proved» 48. The second was James
Breekenridge who readily committed himself to some rather imprudent
hypotheses 49.
The practice of distributing stamped bread, known as hygieia, at
pagan shrines was identified by Perdrizet, and subsequently by Dölger,
as the precursor of the later Christian practice of distributing eulogia 50.
Here we are on fairly certain ground. A stamp in the British Museum,
which was first published by O. M. Dalton, was found in Cyprus 5I. It is
marked with our device. (Figures 2 and 3). Galavaris considered that it
was not later in date than the seventh century and that it was intended for
stamping the liturgical prosphora 52. Regrettably, it is not possible to date
the object with any certitude. Moreover the use to which it was intended
to be put is not any more certain. Was it intended to be used for stamping
bread ? If so, was this bread for eulogia or the liturgical prosphora ?
These questions can only be left open, because the earliest solid evidence
for the practice of stamping the liturgical prosphora is provided by a

46. Frolow, art, cit. (note 1), p. 109 note 65.

47. Honorius Augustodunensis, Open, III, Liturgien, Gemma animae I, 35, PL 172,
col. 555.
48. G. Galavaris, Bread and the Liturgy. The Symbolism of Early Christan and
Byzantine Bread Stamps, Madison (Wisconsin) 1970, p. 70.
49. J. D. Breckenridge, The Iconoclasts' Image of Christ. Gesta, 1 1 2, 1972, p. 3-8.
50. P. Perdrizet, vriA ZCJH ΧΑΡΑ, REG 27, 1914, p. 266-267, citing a pre-Christian
text about pilgrims obtaining hygieia at a pagan shrine. J.-D. Dölger, Unser tägliches
Brot, Antike und Christentum: Kultur und religionsgeschichtliche Studien 5, 1936,
p. 201-210, plates 13-16. In his example of a refrigerium (?), the breads are marked with a
cross. See also G. Vikan, Art, Medicine and Magic in Early Byzantium, DOP 38, 1984,
p. 69, fig. 4
51. O. M. Dalton Catalogue of the Early Christian Antiquities and the Objects from
.

the Christan East in the Department of British and Mediaeval Antiquities and
Ethnography of the British Museum, London 1901, p. 172. n° 973.
52. Gai.avaris. art. cit. (note 48\ p. 74-75. fis. 38,
200 CHRISTOPHER WALTER

dated stamp (1265-1266) at Mount Sinaï53. That it was intended for

liturgical use is made probable by the fact that the words of consecration
are inscribed around the edge. However, the letters which were actually
stamped are not I C XC NI KAbutIC XC MH 0V.
Breckenridge's approach to the subject was anything but hesitant. He
readily accepted the British Museum stamp as being intended for the
liturgical prosphora and pre-Iconoclast in date 54. For him «there can be
little question that, by the time that the iconoclastic controversy began,
the employment of this design (= our device) as the most correct one for
marking the host was universal in the Eastern Church» 55. Furthermore
«the special form of the Eucharistie bread stamp, with its formula IC XC
NI ΚΑ...» was being «established by the liturgy of St. John Chrysostom,
whose fully elaborated program was introduced throughout the Eastern
Church during the sixth century» 56. If Breckenridge can support these
two statements, he must have access to source material which is
unavailable to other scholars. However his ingenious conjectures do not
end here. He goes on to enlist the silver miliaresion of Leo III and
Constantine IV in his argumentation. He sees in it a «sharply stated
assertion of Iconoclast doctrine» 57. In consequence, it must have been
minted «considerably later than Constantine V's elevation by his father
in 720, a date when there is no evidence of manifest iconoclast activity
within the imperial government»58. Thus Breckenridge arbitrarily
interprets the miliaresion of Leo III and Constantine IV as an Iconoclast
manifesto and arbitrarily redates it from 720 to the period after Leo Ill's
first promulgation of a decree against the cult of images. Now that
M. F. Auzépy has shown that there are serious grounds for calling in
doubt the authenticity of the account of Leo Ill's replacement over the
Halki Gate of Christ's icon by a cross 59, Breckenridge gains less by
citing this account in favour of the cult of the Cross as central to
Iconoclast piety. Further, it would seem that Byzantine art historians tend
now to interpret differently the Iconoclast fashion in decorating
churches: rather than substitute crosses for figurative subjects,, they
simply eliminated the latter60.
Sadly, therefore, one seems obliged to dismantle and reject
Breckenridge's elaborate assembly of hypotheses. On the one hand, the
use of our device on coins, not being restricted to Iconoclast emperors,
cannot be interpreted as an Iconoclast manifesto. On the other hand,

53. Ibidem, p. 87-89, fig. 42.

54. Breckenridge, art. cit. (note 48), p. 4.
55. Ibidem.
56. Ibidem.
57. Ibidem, p. 6.
58. Ibidem.
59. Ibidem, p. 8. M. F. Auzépy, La destruction de l'icône de Christ de la Chalcè de
Léon III : Propagande ou réalité ?, Byz. 60, 1990, p. 445-492.
60. Thierry, Le culte de la Croix, art. cit. (note 7).
 THE APOTROPAIC FUNCTION OF THE VICTORIOUS CROSS 201

there is no convincing evidence that our device had acquired liturgical

connotations at this early date.

The device in manuscripts

1. The two earliest examples known, Paris, graec. 510, f. Bv and f. C,
dating from about 880, have already been noted61. Placed at the
beginning of the manuscript, full-page, they were executed with
considerable sophistication. Their refined and confident execution might
suggest that they follow an established tradition, were it not that all the
illustrations in this manuscript attain a high level of sophistication.
2. Closely resembling the preceding two miniatures are those in the Leo
Bible, Vatican, reg. graec. 1, f. 2 and f. 3v, which are generally
considered to date from the tenth century 62. They too are placed at the
beginning of the manuscript, full-page, and executed with considerable
sophistication.
3. Frolow noted five other dated examples of manuscripts containing our
device published by Kirsopp Lake, who, regrettably, did not specify the
number of the folio on which the miniature was executed 6\ The earliest,
Lavra cod. 19, is dated 984 64. It is placed at the end of the manuscript.
The cross is not drawn but constituted by the arrangement of the text in a
cruciform pattern, although, as usual, the letters of our device are placed
in the angles. Thus Lavra cod. 19 is the earliest example of this practice.
Four other Greek letters are added : 13 Χ Ψ Π. We shall return to the
practice of adding other letters later. The manuscript is described by
Lake as a Tetraevangelion.
4. Paris, graec. 375, a collection of Gospels and Epistles, is
dated 1021 65. The cross is small in size and placed in the top right hand
corner of a folio next to the title of the text. It is ornamental, resembling
an illuminated initial letter, made up of twisted strands. Two dragons
hold its foot in their jaws.
5. Florence Laur. plut. 11. 9, also dated 1021, is a Chrysostom 66.
Immediately after the title and a note by the copyist including the date,
there is a simple one-barred cross. In the corners above the bar are IC
XC ÎUC XV, and below it NI ΚΑ ΐ ΑΙΝ6Ι nACIN. Other legends, on
the cross itself and below, abound. In the margin, two small crosses are
made up οϊ letters. There is also another ornamental cross and a bird.

6 i See above, note 9.

62. C. Stornajolo, Miniature delta Bibbia cod. Vat. Regin. Gr. I, Milan 1 905,
.

plates 3, 6. A new edition has been prepared by P. Canakt, S. Dufrenne and C. Mango.
Frolow, art. cit. (note I), p. 107.
63. Frolow, ibidem.
64. Kirsopp and Silva Lake. Dated Greek Minuscule Manuscripts to the Year 1200,
Boston, III, 1934, p. 107, pi. 163.
65. Lake, IV, 1935, p. 12, pi. 256.
66. Lake, X. 1939, p. 10. pi. 693.
202 CHRISTOPHER WALTER

6. Jerusalem Panagia cod. 1 , f. 2, dated 1061, is decorated with a full-

page cross as frontispiece to a Lectionary 67. The one-barred cross is
highly ornamented. Placed around it are nine circles (one is lost) all
containing legends. The four largest, placed in the angles, contain the
letters IC XC NI KA.
7. Sinait. 341, f. 2v, the liturgical homilies of Gregory of Nazianzus, is
dated to the end of the 11th century68. Our device serves here as a
frontispiece.
8. London Additional 36 654, a Menologium, dated 1103 (?) ends with
the text set out in a cruciform pattern with the letters of our device set in
the angles 69.
9. Athen. 167, f. 220 and 220v, 12th century, Lectionary, contains two
richly ornamented crosses with our device. They are placed neither at the
beginning nor the end of the text 70.
10. Sinait. 221, f. 40, dated 1175, Lectionary, with, above the headpiece
for the first Monday of Pentecost, our device 7I.
11. Athen. 152, f. 144, Gospels, thirteenth century, has a band-shaped
headpiece surmounted by our device72. This folio is neither at the
beginning nor the end of the text.
12. Athen. 2509, f. 10, Gospels, fourteenth century73; the device is
placed at the beginning of the text.
13. Athen. 175, p. 3 and p. 152, Lectionary, fourteenth century74. The
device is placed at the beginning of the Gospels of John and Mark, but in
each case the and A are missing.
I

14. Athen. 108, f. lv, Gospels, fourteenth century, with a full-page cross
which serves as a frontispiece75. The letter A is missing from the device.
(Figure 4).
15. Sinait. 339, the liturgical homilies of Gregory of Nazianzus, dating
from about 1 136-1 155, differs from the preceding manuscripts in that on
our folios the text is set out in a cruciform pattern with letters placed in
the angles 76. On f. 90 they are \C XC NI ΚΑ, but on f. 73 and f. 197
they aret Χ ί Π while on f. 396v they are 6 e e 6.

67. Lake, V, 1936, pi. 21, p. 371.

68. K. Weitzmann and G. Galavaris, The Monastery of Saint Catherine at Mount
Sinai, The Illuminated Creek Manuscripts, I, From the Ninth to the Twelfth Century,
Princeton NJ 1990, p. 108-109, fig. 336.
69. Lake, II, 1934, p. 16, plate 133.
70. A. Marava-Chatzinikolaou and Ch. Toueexi-Paschou, Catalogue of the
Illuminated Greek Manuscripts of the National Library of Greece, I, Athens 1978, p. 221,
fig. 616.
71. Weitzmann and Galavaris, p. 167, fig. 662.
72. Ciiatzinikolaou and Pasciiou, II, Athens 1985, p. 40-41, fig. 55.
73. Ibidem, p. 148, fig. 321.
74. Ibidem, p. 226, fig. 434.
75. Ibidem, p. 131, fig. 295.
76. Weitzmann and Galavaris, p. 140-153, fig. 583-586.
 THE APOTROPAIC FUNCTION OF THE VICTORIOUS CROSS 203

Although this list is certainly not complete, there are sufficient

examples to make it clear that, even if there was no rigorous obligation,
the device was normally placed at the beginning or end of the principal
text in the manuscript. However, equally, there are far more manuscripts
without the device than those with it. Of these fifteen manuscripts, only
four, (4) Paris, graec. 375, (9) Athen. 167, (10) Sinaït. 221, and
(11) Athen. 152, have the device placed elsewhere than at the beginning
or the end of the principal text.
The formula IC XC without the NI ΚΑ, is also fairly common. Here
are three examples of manuscripts in which the cross is accompanied by
IC XC and placed at the beginning of the text: Sinaït. 500, f. 4v, a
Metaphrastic volume dating from about 1063 77; Sinaït. 172, f. 1, Four
Gospels, 106778; Athen. 70, f. 3, a Lectionary, dating from the end of
the thirteenth century 79.
Quite different formulae are sometimes used or added, normally with
just the initial letters of the words, although they may be written out in
full: Vatican, graec. 463, f. 21v, the liturgical homilies of Gregory of
Nazianzus, Α Π Π C 80 ; Messina Salvatore 73, f. 1, a Lectionary dated
1172, IC XC VC Θ V 8I ;Athen. 133, f. 1, Lectionary and Epistles, Ψ [χ]
ί Π 82 ; Athen. 152, f. 11, Gospels, Ό Χρ(ιστο)ς αύτό(ς) έστιν άρ(χή)
και τέλος 83.

Minor objects
The number of minor objects on which our device is inscribed is
limited. As far as I am aware it figures on only one icon, on the reverse
of that of Saints Zosimus and Nicholas at Mount Sinai, which
Weitzmann dated to the first half of the tenth century 84. Two earlier
icons at Mount Sinaï which Weitzmann dated to the sixth or seventh
century have the letters IC XC VC 0V placed around the cross 8\ The

77. Ibidem, p. 74, fig. 199. Note that the miniature with our device published by
G. Galavaris, 'Ελληνική Τέχνη. Χεφογραφική, Athens 1995, p. 231, fig. 87, is incor
rectly named Sinaït. 500. Unfortunately I have not succeeded in establishing its proper
identity.
IS.'lbidem, p. 80-81, fig. 220.
79. Ciiatzinikolaou and Paschou, II, p. 33, fig. 28.
80. G. Galavaris, The Illustrations of the Liturgical Homilies of Gregory Naz.ian~.enus,
Princeton NJ 1969, p. 251, fig. 69.
8 1 J. Spatiiarakis. Corpus of Dated Illuminated Greek Manuscripts to the Year 1453,
Leiden 1981. n° 159. fig. 304.
.

82. Ciiatzinikolaou and Paschou, II, p. 33, fig. 28.

83. Ibidem, p. 40-41, fig. 55.
84. K. Weitzmann, The Monastery of Saint Catherine at Mount Sinaï, The Icons, I.
From the Sixth to the Tenth Century, Princeton NJ 1976, Β 52, p. 83-85, pi. 108b. To this
could be added the metal plaque on the reverse of an icon of the Virgin and Child in the
Hermitage, on which our device was executed apparently in the thirteenth century, A.
Banck. Byzantine Art in the Collections of the Soviet Museums, Leningrad / New York
1978, n° 248.
85. Ibidem, Β 22. 23, p. 47-48. pi. 70.
204 CHRISTOPHER WALTER

Sotirious recorded a calendar icon, which they dated to the second half
of the eleventh century, upon which the following letters are inscribed : 2
ZCK,AnriC,eeee,XXXX86. There remain the two
icons, possibly dating from the time when the chapel of Saint Neophytos
was decorated (1 183). Various letters - not our device - are inscribed on
them, but they are best considered in the context of the chapel itself,
which is particularly rich in such inscriptions.
A single lead seal, which has been attributed to the seventh or eighth
century, has on one side an inscription invoking the Theotokos in favour
of John the Apohypatos ; on the other side our device is inscribed 87.
Of ivories carrying our device, the Harbaville triptych (10th century ?)
is outstanding.88. The letters IC XC NI ΚΑ are all placed, unusually,
above the arms of the cross. The sophistication of this superb carving
recalls the frontispieces to the two earliest manuscripts carrying the
device. Three other less impressive ivories should be mentioned : the
Halberstadt triptych, partially destroyed, the Borradaile triptych in the
British Museum and a plaque in Turin (Gualino collection).
The steatite of Saint John the Baptist in the collection of the Armoury
of the Kremlin, Moscow, dated to the late fourteenth or fifteenth century,
is unique89. Our device, which is not carved in the same way as Saint
John on the other side, is placed above a coat of arms and may have been
added later.
Considering the great number of reliquaries of the True Cross which
are known, remarkably few carry our device. Frolow noted only three.
One (n° 574), dating from the late thirteenth or early fourteenth century,
is now lost 90. A second (n° 663), in the Treasury of San Marco, Venice,
possibly dating from the thirteenth century, has the Crucifixion
represented
(n° 872) in the
on Accademia,
the face and
Venice,
our dating
device probably
on the reverse91.
from 1446-1459,
The third
has
on the reverse, besides our device, the letters 13 Χ ΐ Π 92. Frolow noted
two carrying inscriptions other than our device : n° 340, the twelfth
century reliquary at Esztergom, with the letters X X X X 93 ; a reliquary

86. G. and M. Sotiriou, Icônes du Mont Sinai, Athens 1 956-1958, I, pl. 136-143 ; II, p.
123.
87. Iskusstvo viz.antii ν sobranijah SSSR, Moscow 1977, I, n° 258.
88. A. GoLixscHMiirr and K. Weitzmann, Die byzantinischen Elfenbeinskulpturen
des
n° 53,
XLp.- 41.
XIII.
A. Jahrhunderts,
Cutlhr, The HandBerlin
of the
1934,
Master,
n° 33,Princeton
p. 34-35,NJn° 1994,
60, p.p. 43
235,
; n°28138,note
p. 36;
32,
fig. 152,239,242.
89. Iskusstvo viz.antii (note 86), III, n° 1015; I. Kalavrezou-Maxeiner, Byzantine
Icons in Steatite, Vienna 1985, p. 236-237, n° 174 (no reproduction of the reverse).
90. A. Frolow, La relique de la Vraie Croix, Paris 1961, p. 443.
91. Ibidem, p. 485-486; Ioem, Les reliquaires de la Vraie Croix, Paris 1965, fig. 48 ;
A. Grabar, etc., // tesoro di San Marco, II tesoro e il museo, Florence 1971, p. 26, n° 24.
Since the piece is no doubt Venetian loot from the Fourth Crusade, Frolow' s suggested
1 3th-century date for this piece is hardly tenable.
92. Frolow, La relique, p. 526-527.
93. Ibidem, p. 33 1 -332 ; Les reliquaires, fig. 4 1 .
 THE APOTROPAIC FUNCTION OF THE VICTORIOUS CROSS 205

in the cathedral of the Annunciation, Moscow, dated 1382, on which the

Greek letters have not been deciphered 94.
The small amount of evidence which we have suggests that the
presence of the letters IC XC NI ΚΑ on minor objects had no special
connotation. They were merely an adjunct to the cross.

Sepulchral monuments
The cross figured regularly in funerary art from the earliest times.
However, the use of our device was restricted. It does not, apparently,
figure on sepulchral monuments in Syria, a territory so rich in Christian
inscriptions. The nearest equivalents are on a tomb, dated 410, at Deir
Sanbil, two chrisma with the letters AU under the bar and above the
words TOVTO ΝΙΚΑ 93 and another, undated, with the words το σημίων
τοΰτο νίκα %.
Another tomb, undated, probably from Herment, Egypt, carries the
name IHCOVC XPICTOC written in full above the bar and NI KA below
it. An accompanying text calls on the Lord to redeem the soul of
Theodoros 97.
It is only in the eleventh century that examples of the device become
more numerous on tombs, and these are only recorded for the modern
territory of Greece 98. A pseudo-sarcophagus from the eleventh century
at the Great Lavra is decorated with three crosses, all-double-barred ".
The cross to the right is inscribed above and below the upper bar with
the letters I C XC NI ΚΑ, that in the middle with X X X X, while the
right hand cross has no inscription. Other examples are at the Panagia,
Episkopi, Ano Volos (1274/6) l0°, the Moni Petras, Portana Piliou of the
same date l01, again at Ano Volos (after 1276), with the lower part of the
cross destroyed l02, and, at the same place, with the cross complete (also
after 1 276) 103.

94. La relique, p. 526-527.

95. L. Jalabert and R. Mouterde, Inscriptions grecques et latines de Syrie, IV,
Laodicée, Apamée, Paris 1955, p. 130, n° 1437 ; M. De Vogué, Syrie centrale, Architecture
civile et religieuse du 1er au Vile siècle, Paris 1865, I, p. 7, II, p. 108-109, 154.
96. Jalabert et Mouterde, IV, p. 1 17, n° 1404. See also p. 350, n° 1984, a lintel, pos
sibly from a nearby basilica (undated) at Tell 'Abd el-'Aziz: σ(ταυ)+ρ(έ) ες άε(ί) ν(ίκα).
The authors read this as an imperative. Bernand, op. cit. (below, note 104), «corrects» to
the indicative (νικά). But if σταυρέ is in the vocative case, surely the authors, not
Bhrnand, are more likely to be right. See below p. 206.
97. G . Lefebvre, Recueil des inscriptions grecques-chrétiennes d'Egypte, Cairo 1907,
p. 94, n° 5 13.
98. Th. Pazaras, Ανάγλυφες σαρκοφάγοι και επιτάφιες πλάκες της μέσης και
ύστερης Βυζαντινής περιόδου στην Ελλάδα, Athens 1988, especially p. 114 et seq.:
σταυρός.
99. Ibidem, p. 28. n° 17, pi. 13b.
100. Ibidem, p. 38-39, nu 45a, pi. 30b, 32a.
101. Ibidem, p. 39, n° 45b, pi. 31, 32b.
102. Ibidem, p. 40, n° 46a, pi. 35a.
103. Ibidem, p. 65. n° 88. pi. 35c.
206 CHRISTOPHER WALTER

Again, as on minor objects, our device cannot be shown to have any

special connotation, the letters being merely an adjunct to the cross.

Churches

1 . Early churches.
The earliest datable example of a cross, accompanied by a legend
close to that of our device, is in the converted temple of Philae 104.
Theodosius II promulgated an edict ordering that pagan temples should
be destroyed and marked with a cross l05. It is well known that this edict
was not universally observed. Deichmann noted eighty-nine examples of
pagan temples converted into churches 106, while some probably, like
Philae, continued to be used for pagan cult. The temple of Isis at Philae
was only converted into a church, dedicated to Saint Stephen, under
Justinian between 535 and 537 l07.
An undated, but no doubt primitive, legend in a church at Telanissos,
Syria, which is not accompanied by a cross, runs XC NI I08. Finally an
undated example of our device was discovered under a ninth-century
painting in the catacomb of Saint Januarius, Naples 109.

2. Cappadocia
In spite of the importance of crosses in decoration of churches in
Cappadocia "°, only four examples of our device have been recorded. It
is, however, significant that, in each case, the device is associated with
an apse. They all date from the ninth or tenth century, and Cheynet has
suggested that they may be associated with the Phocas family '". In
Açikal Aga kilisesi, the cross is placed on the wall above the altar. Only
the word νίκα has survived from the legend "2. In Ke§i§lik kilisesi,
Pantanassa (Akköy), the device is placed in a niche in the centre of the
apse, with a cross carrying the same legend to left and right 113. In the
south chapel of Zelve 4 the device figures on the north pier of the arch

104. Lefebvre, op. cit. (note 96), p. 110, n° 590; H. Leclercq, Philae, DACL 14 ,
1939, 700; P. Nautin, La conversion du temple de Philae en église chrétienne, CA 17,
1967, p. 14, fig. 8 ; E. Bernand, Les inscriptions grecques et latines de Philae, II, Haut et
Bas empire, Paris 1969, p. 256-259, n° 201, pi. 49.
105. A. Frantz, From Paganism to Christianity in Athens, DOP 19, 1965, p. 187;
Codex Theodosianus, XVI 10 26(435).
106. J. Deichmann, Christianierung II, Reallexikon für Antike und Christentum 2, 1954,
1 230- 1 234 ; Dinkler, art. cit. (note 2), p. 1 35- 1 38.
107. Nautin, art. cit. (note 103), p. 6, 8.
108. Jalabert and Mouterde, op. cit. (note 95), II, p. 232, n° 414.
109. H. Leclercq, Cujus nomen Deus seit, DACL 3, 1914, col. 3185.
1 10. Thierry, Le culte de la Croix, art. cit. (note 6).
111. Cheynet, art. cit. (note 34), p. 74.
1 12. C. Jolivet-Lévy, Les églises byzantines de Cappadoce. Le programme iconogra
phiquede l 'abside et de ses abords, Paris 1 99 1 , p. 327-329.
113. Ibidem, p. 148.
 THE APOTROPAIC FUNCTION OF THE VICTORIOUS CROSS 207

before the apse il4. In Hacli kilise the cross is placed in a niche in the
centre of the apse. Of the accompanying legend only the letters IC XC ..
.A have survived "3.
It should also be mentioned in passing that the legend, along with
other inscriptions, was written on the wall of the hermitage of the monk
Symeon at Zelve "6. However there was no accompanying cross.
It is evident that it was by no means de rigueur that a cross should be
accompanied by our legend, or, indeed, by any inscription. Moreover
these examples in Cappadocia do not take on the form, already
established in manuscript illumination in particular, whereby the letters
are placed above and below the arms of the cross. That the device should
be placed in or near the apse has its parallel in later churches, where,
however, il was not exclusively placed there.

3. Later churches not Serbian.

The church of the monastery of Saint Laurence, Pilion, has two
crosses on the façade "7. One, placed over the door, has the letters IC XC
carved on the bars. The other, a double-barred cross, has our device
inscribed above and below the lower bar. Since this church has been
restored more than once, difficulties arise as to the dating of the crosses,
one of which could, however, be as early as the end of the
twelfth century.
The decoration of the apse of the hermitage of Saint Neophytos can be
dated to around 1200 "8. Even if the two icons of the templon, of Christ
Philanthropos and the Virgin Eleousa, are slightly later, the legends on
their reverse are evidently in harmony with the decorative programme of
the apse. Certainly legends other than IC XC NI ΚΑ exist on the reverse
of a Sinai icon perhaps slightly earlier in date than Saint Neophytos "9;
the latter is the first monument in which our device is only part, although

1 14. Ibidem, p. 7, pi. 18, fig. 1 ; N. Thierry, Haut Moyen Age en Cappadoce, Les
églises de la région de Cavuçin II, Paris 1994, p. 356. In the main church on the triumphal
arch the cross figures between two fishes with, to the left, the letters HS XC, while, to the
right, only X is legible, Jolivet-Lévy, p. 6, pi. 17, fig. 1 and Thierry, p. 352, fig. 108,
pi. 1 85a. Jolivet-Lévy' s description of the legend of our device as «la célèbre acclamation
de victoire introduite au Moyen Âge dans la liturgie» is hardly accurate !
I 15. Jolivet-Lévy, p. 5 1, 53 (date) ; Thierry, p. 248.
116. G. De Jerphanion, Les églises rupestres de Cappadoce I 2, Paris 1932, p. 594,
inscription n° 108, possibly contemporary with the church of Saint Symeon, 10th century,
Jolivet-Lévy, p. 7-12: Thierry, p. 323-324. No information on the hermitage subsequent
to that of De Jeri'hanion is available.
117. R. Leonidopoulou-Stylianou, Παρατηρήσεις στο μοναστήρι του 'Αγίου
Ααυρεντίου Πηλίου, Αελτίον της Χριστιανικής 'Αρχαιολογικής 'Εταιρείας, per
iod 4, 9, 1977-1979, pi. 97a. The cross above the door with a legend referring to the foun
der(Ignatios) is accompanied by a date 1551, p. 234, fig. 96a; the cross to the left of the
door possibly dates iron; the twelfth century, p. 243, fig. 97b.
1 18. C. Mango a;,d E.J.W. Hawkins. The hermitasc of St. Neophytos, DOP 20. 1966.
p. 162-163.
i 19. Sec :ih(ve. note 86,
208 CHRISTOPHER WALTER

usually the central part, of a more developed progamme of apotropaic

signs. Surely this programme was the personal choice of the eccentric
patron, the hermit Neophytos, who expressed his originality in his choice
of other decorative themes in the chapel. The legend of our device
appears six times in the sanctuary : four times on the wooden cross
above the altar, on the front and rear of each branch (abbreviated to IC
XC NHK), once in the niche to the right of the central window and once
in the alcove in the south-west corner of the bema, but not on the reverse
of the two icons. In three instances our legend is accompanied by others :
in the south-west alcove, left Τ Τ Ν Τ, centre? Χ ? Π, right IC XC
[?Χ] Ψ Π ; on the reverse of the icon of Christ Philanthropos IC XC
0C 0V and X X X X ; on the reverse of the icon of the Virgin Eleousa
again IC XC and Μ Γ Χ Β.
It is worthy of notice that on the two icons the older form of IC XC
VC 9V replaces our legend. Also, with one exception, Τ Τ Ν Τ in the
south-west alcove, all the inscriptions are readily deciphered, at least by
the literate and the initiated. Other legends will be noted in due course
which are not easily deciphered, as well as a few in which the artist
seems to have made a mistake. Yet others occur to which several
interpretations (but usually with regard to the same theme) may be given.
Consequently it is unlikely that these legends were intended to be
deliberately arcane. More probably they belong to a fashion or a genre,
with which may be compared the contemporary craze for acronyms
(s ig les).
Three fourteenth-century churches, all on the territory of modern
Greece, at Thessaloniki, Kastoria and the Meteora, should now be
mentioned. That of Saint Nicholas Orphanos, Thessaloniki, dating
from 1310-1320, contains four examples 12°. Around the upper bar of the
cross in the apse is inscribed the legend of our device. Below the central
bar are the letters 6 ? Θ T, and to the left and right of it CT(au)POC
Ι(ησο)Υ Χ(ριστο)Υ ΥΙΟΥ TOY Θ60Υ 121. Three other examples are to
be found in the narthex : that on the east side of the window in the south
wall is virtually identical with the preceding example 122 ; that on the
west side of the same window has the legend of our device above the
upper arm of the cross and, below the central arm Π(ας) XPICTIANO(ç)
ΠΡΟΟΚΥΝ(εΪ) CT(om)PON Χ(ριστ)ΟΥ Τ(όν) ΤΗΝΙΟΝ m. The cross
at the south window of the west wall is accompanied by the letters Ψ Χ
f Π l24. These legends are easily read. Moreover the fact that two are

120. A. Xyngopoulos, Οι τοιχογραφίες τον αγίου Νικολάου 'Ορφανού

Θεσσαλονίκης, Athens 1964, especially p. 24 ; Idem, Νεώτεραι ερευναι εις τον "Αγι,ον
Νικόλαον Όρφανδν Θεσσαλονίκης, Μακεδόνικα 6, 1964-1965, ρ. 93. Largely super
seded by A. TsiTOURiDou, Ό ζωγραφικός διάκοσμος του αγίου Νικολάου 'Ορφανού
Θεσσαλονίκης, Thessaloniki 1986.
121. TsiTOURiDou, p. 217, pi. 108.
122. Ibidem, p. 217-218, pi. 117.
123. Ibidem, p. 217-218, pi. 118.
124. Ibidem, p. 218, pi. 119.
 THE APOTROPAIC FUNCTION OF THE VICTORIOUS CROSS 209

written out fully, instead of being reduced to the initial letters of the
words used, makes it clear that they were not intended to be arcane.
The two examples in the church of the Taxiarchis, Kastoria, probably
belong to the series of painting executed in 1359/60 l25. They are placed
on the walls either side of the entrance from the narthex to the nave.
(Figure 5). The crosses are three-barred. The legends which accompany
the north cross are easily read. Around the uppermost bar is our legend,
abbreviated : IC XC Ν Κ . To the left and right of this bar are the letters
e e 6 e . Below the central bar are three series of initial letters : Ψ Τ
Τ Χ,'ΐ5 Δ «f Π, and Ρ Ρ Ρ P. The third is not problematical. The first
and second Orlandos amended toi Χ ΐ Π and Τ Τ Δ ΐ. The legend
of the south cross are less well preserved (and were not published by
Orlandos). Apparently they are not greatly different from those on the
north side : IC XC Ν Κ above, then ί Κ, followed by Β ί( ?) Κ, and Ρ
Ρ Ρ Ρ . An entrance-way, whether to the body of the church or to the
sanctuary, was a strategic position for placing an apotropaic sign.
The church of the Presentation at the Meteora, built according to an
inscription over the door in 1366/7, with the hieromonk Nilus as
principal donor, contains several of these devices l26. In a niche on the
south wall around the upper arm of the cross may be read the letters XC
NK. The cross on the east wall of the south entrance has the following
legends : IC XC Ν Κ, 9 Χ <t> Π, Χ Χ Χ Χ, and 6 6 6 6. On the
west wall the cross is accompanied by the legend Τ Τ Δ Ψ, and another
legend which has not been deciphered: CT CT Δ Π(?). On the east
side of the north wall are four legends : 15 ΧίΠ,ΑΒΔε,ΤΧΛ
Ρ, Ρ Ρ Ρ P. The second and third of these remains obscure. Gojko
Subotic considered these crosses to be associated with Christ's sacrifice,
but in the fourteenth century there is no explicit evidence for this.
Another church in Cyprus, dated 1 494, and consequently much later
than the hermitage of Neophytos with which it has no evident
association, is that of the Holy Cross near Platanistasa with, apparently,
several crosses on the west and north walls l27. A three-barred cross, the
only one published, has its legend disposed in an unusual way. Above
the cross are the letters IC XC, followed by ΥΘ ; then C Κ C Τ Τ,
followed by Ν Κ . Between the upper and central arms is the legend A K
Χ Τ (amend to A Χ Χ Π ?), and below the central arm Τ Τ Δ Ψ.
Since the church is dedicated to the Cross, it was perhaps not considered
necessary to place the devices in strategic places.

1 25. A. K. Orlandos, Τα Βυζαντινά μνημεία της Καστοριάς, Άρχείον των

Βυζαντινών μνημείων της 'Ελλάδος 4, 1939, ρ. 69-70; S. Pelekanides and
M. Chatzidakis. Kastoria, Athens 1 985, p. 1 02 (date).
1 26. G. Subotic, Poceci monaskog zivota i crkva manastira Sretenja u Metcorima,
Zboniik zxi likovne mnetuosti, 2, 1 966, p. I 72.
1 27. A. and J. A. Styi.ianou, The Painted Churches of Cyprus, London 1 985, p. 2I0,
pi. 1 20.
21 0 CHRISTOPHER WALTER

The parecclesion of Saint John the Almoner at the church of the

Phaneromeni, Trikkala, undated, but not, apparently, earlier than the
sixteenth century has a cross on the left door of the templon leading to
the prothesis 128. Around the central arm is inscribed our device. Below
this, descending to the footrest, are six legends ιΙ'ΚίΠ,ΤΚΠΓ
(alluding to Adam's skull represented beneath the cross), Θ Θ Θ Θ, 6
666,ΑΠΠ0Τ,ΤΤΔ Ψ The disposition and number of legends
recall the device which would be represented on the monastic schema l29.
A final example is a plaque from Arkadia, Crete, also undated.
Inscribed on it are six legends : Ψ ΧίΠ,εεεε,ΑΠΜΟ,ΧΧ
Χ X, IC XC NI ΚΑ ι3°.
This paltry list of monuments provides, nevertheless, the basis for
some generalisations. Firstly the apotropaic value of these device
becomes explicit in the late twelfth century. They were not intended as
imperial invocations for protection against and victory over imperial
enemies. Rather they invoke the protection of the Cross against the
power of evil. Such an understanding of the Cross was not, of course,
new. However, from this period, our device, certainly the most
widespread and possibly to be understood as the focal point of a network
of apotropaic devices, gains extra force. Sometimes the older legend IC
XC YV ΘΥ replaces IC XC NI K.A m. Also one or two earlier
examples of the multiplication of devices in other media are known, for
example on the reverse of the late eleventh-century icon at Mount
Sinai 132 and in the manuscript Sinai't. 339 133. However it is clear that,
from the earliest dated example at Saint Neophytos in the late twelfth
century onward, the placing of a network of prophylactic devices in a
church was common at least in some regions, notably northern Greece
and on territory governed by the Nemanjic family.

4. Serbian churches.
It would be tempting to provide here a list of the Serbian churches in
which our device is represented. However this would be supererogatory,

128. N. Giannopoulos, Αί παλαιά! έκκλησίαι Τρικκάλων (Θεσσαλίας), BZ 27,

1927, p. 360.
129. Th. Provatakis, Ό διάβολος- εις την Βυζαντινήν τέχνην, Thessaloniki 1980,
fig. 270.
130. Κ. Kalokykis, 'Ανέκδοτοι έτπγραφαί και χαράγματα έκ μεσαιωνικών
μνημείων Κρήτης, Κρητικά Χρονικά 5, 1951, ρ. 338-339. One could also add a curious
slab, built into the wall of the church at Kardzali (Bulgaria), on which, together with other
apotropaic signs, our device is represented three times, N. OvCarov, Sur l'iconographie de
St. Georges aux XF-XIF siècles, Byzanünoslavica 52, 1991, p. 121-129 ; Ch. Walter, An
apotropaic sequence at Kardzali, to appear in a number of Zograf dedicated to the memory
of Gordana Babic.
131. For example on the icon of Saint Neophytos, in the apse at Saint Nicolas
Orphanos, and, doubled with Ν Κ , at Platanistasa.
132. See above, note 86.
133. See above, note 76.
 THE AP0TR0PA1C FUNCTION OF THE VICTORIOUS CROSS 21 1

for the late Gordana Babic has not only published a comprehensive list
of these churches but also a study in depth of the devices used in
them l34. The reader is referred to this study. Here we shall only
summarize Babic's conclusions. (Figures 6, 7, 8).
Babic has listed eleven Serbian churches, all dating from the end of
the thirteenth century (Arilje 1296) to the end of fourteenth century
(Monastery of Marko, Susica, 1380-1382). She has catalogued twenty-
six of these devices, in which our legend IC XC NI KA figures
invariably, usually along with other legends. The devices are normally
placed in a «strategic» position : near a door or window or in an entrance
passage, places where evil forces might pass. Of these devices, fourteen
are either inside or at the approach to the sanctuary of the church. Thus it
is for Serbian churches that we have the most fully documented
repertory and possibly the most abundant. These Serbian churches
follow in the tradition of the earlier, more scattered examples that we
have noted, particularly those at Saint Neophytos, Cyprus, the
Taxiarchis, Kastoria, and Saint Nicholas Orphanos, Thessaloniki.

Inscriptions
As mentioned earlier, some legends have not been deciphered, while
others may have more than one meaning. However, generally speaking,
their meaning is clear, although their frequency varies considerably. A
list now follows of fifteen of these acronyms, with the object or church
on which they figure l35.
1 . Α Π Π C - 'Αρχή Πίστεως Μυστηρίου Σταυρός
Calendar icon, Mount Sinai, Saint John the Almoner, Trikkala, Pec
(twice).
2. A X Χ Π - 'Αρχή Χριστοί) Χριστιανικής Πίστεως
Platanistasa ( ?), Pec.
3. 6 6 6 6 - Ελένη Εύρε 'Ελέους 'Έρεισμα
Calendar icon, Mount Sinai, Saint John the Almoner, Trikkala,
Taxiarchis, Kastoria, Presentation, Meteora, Gracanica, Saints
Constantine and Helena, Ohrid l36.

134. Gordana Babic, Les croix à cryptogrammes peintes dans les églises serbes des
XHIe et XlVe siècles, Mélanges Ivan Dujcev, Études de civilisation, edited S. Dufrenne,
Paris 1979, p. 1-13.
135. Among the numerous lists of acronyms, none complete, see particularly, G.
Lampsakis, Μελέται, έργασίαι και περιηγήσεις, 1 893, Αελτίον της Χριστιανικής
'Αρχαιολογικής 'Εταιρείας, period 1, 2, 1894, ρ. 49. In the present article the acronyms
are set out in Greek alphabetical order.
136. Babic notes, p. 6, six alternative readings for this acronym, which also appears on
some representations of the Cross upheld jointly by Constantine and Helena, at Donja
Kamenica (Serbia), for example, in the churches of Theoskepastos and Saint Sabbas at
Trebizond, G. Millet and D. Talbot Rice, Byzantiine Painting at Trebizond, London
1936, p. 46, 70 and at Berende, E. Bakalova, Stenopisite na c'rkvata pri selo Berende,
Sofia 1976, p. 53-54, illustrated fig. 46, p. 71. This last example has our legend on the
upper bar of the cross and € 6 £ G on the central bar.
212 CHRISTOPHERWALTER

4. 6 Υ Θ 6 - 'Ελένης Υιός Θαϋμα Εΐδεν

Staro Nagoricino.
5. 6 Ω Θ Τ - Ελένη "Ωφθη Θεοΐ5 Τάφος
Saint Nicholas Orphanos, Thessaloniki.
6. Θ Θ Θ Θ - Θέα Θεοΰ Θείον Θαΰμα
Saint John the Almoner, Trikkala.
7. 1 Χ Υ Θ Υ - Ίησους Χριστός Υίός Θεοΰ
This acronym was probably in vogue before IC XC NI KA. It is found
on the reverse of two early icons at Mount Sinaï. Babic suggests that it
derives from John 4, 15. It figures also on the icons at Saint Neophytos,
at Platanistasa, six times at Staro Nagoricino, and there is a slight variant
in Saint Nicholas Orphanos 137.
8.H Γ Χ Β - Μαριάμ Γεννήσασα Χριστόν Βασιλέα
Saint Neophytos on the icon of the Virgin Eleousa.
9. Ζ Ζ C Κ - Ξύλον Ζώης Σωτηρία Κόσμου
The calendar icon at Mount Sinaï.
10. Ρ Ρ Ρ Ρ - 'Ρητώς 'Ρήτορες 'Ρητορεύουσι 'Ρήμα
Taxiarchis, Kastoria, Gracanica.
11. Τ Κ Π Γ - Τόπος Κρανίου Παράδεισος Γέγονεν
Saint John the Almoner, Trikkala, Gracanica.
12.T Τ Δ t - Τοϋτο τον Τύπον Δαίμονες Φρίττουσι
Presentation, Meteora, Saint John the Almoner, Trikkala, Platanistasa.
13. Τ Τ Τ Δ - Τούτο τον Τύπον Τρέμουσι Δαίμονες
Arkadia, Crete.
Η.ί Χ t Π - Φως Χρίστου Φαίνει Πασιν
This was the most popular legend. Saint Neophytos, Saint Nicholas
Orphanos, Thessaloniki, Saint John the Almoner, Trikkala, Sopocani,
Gracanica, Arilje (twice), Saints Constantine and Helena, Ohrid 138.
15. X X X X - Χριστός Χάριν Χριστιανοϊς Χαρίζει
Calendar icon, Mount Sinaï, Saint Neophytos, icon of Christ
Philanthropos, Presentation, Meteora, Pec (exonarthex), Saints
Constantine and Helena, Ohrid.
These legends may be distributed into three groups. There are those
which are concerned with Christ, firstly as God or the Son of God (6, 7),
secondly as the primary object of faith (2) and thirdly as the source of
light and joy (14, 15). The next group contains those concerned with the
Cross. Either they are dogmatic, concerned with the Cross as the
beginning of faith (1) and its redemptive function (9), or they are topical,
concerned with Helena and the invention of the Cross (3, 4, 5) and
Calvary as the place where Adam's skull was buried (11). The third
group is concerned with the apotropaic power of the Cross, terror of
demons (12, 13). The relevance of 10 is not clear.

137. Cxaupoç Ίησοΰς Χρίστου Υίοΰ τοο Θεού.

138. G. SuBOTiC, Sveîi Konstantin i Jelena u Ohridu, Belgrade 1871, p. 111. Babiî
derives the phrase from the liturgy of the presanctified.
 THE APOTROPAIC FUNCTION OF THE VICTORIOUS CROSS 213

Only two of these legends appear to derive from official sacred texts :
7 from the New Testament and 14 from the liturgy. However, 12 and 13
recall an epigram of Theodore Studite addressed to a cross at the
entrance to a church. It is described as a guardian and the fearful enemy
of demons 139. Their composition would seem to be, for the most part,
the fruit of a monastic jeu d'esprit.
And the game continued, particularly at Mount Athos, where,
unfortunately, the legends of the devices have not been as yet
systematically catalogued and studied 140. One late example in the
trapeza of Iviron (1848) may be indicative of the spirit in which these
acronyms were composed : 9 9 t3 Ψ - Φίλε, Φέρε, Φάγε, Φεΰγε 141.

Conclusion
The focal point of this study is the device made up of a cross and the
legend IC XC NI KA. Its primary element is the cross, whose
significance, according to its context, is rendered more precise by the
presence of this, and possibly other, legends. These secondary legends
add a new connotation by associating the Cross with Saint Helena or
Calvary, or by emphasizing its redemptive and apotropaic powers.
In studying the theology of the Cross cognizance must be taken of two
themes : its value as a sign or symbol of Christ (being his principal relic),
and its own specific qualities, albeit derived from Christ.
As a sign of Christ it was ubiquitous in the late fourth century. John
Chrysostom wrote of its presence «everywhere most frequently. It shines
on the walls of houses, in books, in cities, villages, in deserted and
inhabited places» 142. There was, indeed, room for discussion as to the
relationship of the sign to the person. Theodore Studite insisted that a
cross signified Christ but did not represent him l43. Frolow considered
that the distinction was too subtle to be made by the non-professional
theologian l44. His opinion is supported by the fact that legends 1 and 2
attribute the beginning of faith to the Cross and Christ respectively.
Another legend (9) attributes to the Cross the salvation of the world.
A second quality attributed to the Cross, the power to repel demons,
dates back to early Christianity, in which it took over the function of the
pagan amulet. Cyril of Jerusalem, when instructing his catechumens,
stressed the prophylactic power of the sign of the cross : ό σταυρός

139. Theodore Studite, Epigram n°47, PG 99, col. 1796b; Theodoras Studites
Jamben auf verschiedene Gegenstände, edited P. Speck, Berlin 1968, p. 199.
140. See particularly P. Uspenskij, Pervoe putesostvie ν Afonskie monastyre i skity, II
2, Moscow 1880, p. 22-26, 180-181 ; G. Millet, etc., Recueil des inscriptions chrétiennes
de l'Athos, Paris 1904, n° 212, n° 275, n° 393, n° 543.
141. Millet, n° 275. Sec also above note 129.
142. John Chrysostom, Contra Judeos et Gentiles, quod Christus sit Deus
(Clavis 4326), PG 48, col. 826.
143. Theodore Studite, Refutatio et suhversio impiorum poematum, PG 99, col. 457.
144. Frolow, art. cit. (note I), p. 103.
214 CHRISTOPHERWALTER

είναι μέγα φυλακτήριον... σημεΐον πιστών και φόβος δαιμόνων ι45.

It is noteworthy that a similar expression was used by the painter of the
parecclesion of Saint John the Baptist at Saint Sophia, Ohrid (1347-
1350) in the inscription which accompanies his representation of the
Cross. He invokes Christ crucified as his invulnerable protector
(φυλακτήρ μου άρρηκτε) 146 (Figure 9).
The secondary legends do not, apparently, reiterate the victorious cha
racter of the Cross. This was left to the principal legend, which, incident
ally, attributed victory not directly to the Cross but to Christ. In this res
pect, our device is intriguingly discreet. The Cross was indelibly connected
with victory from the moment of Constantine's vision, although some obs
curity remains as to the precise form that the legend took : τούτω or εν
τούτω νίκα ? The imperative form of the verb is not in doubt. Thus in
Eusebius's Life of Constantine the formula τούτω is used 147. In Paris,
graec. 510, f. 440, εν is added l48. This, together with the imperative form
of the verb, was to become standard. However, when Christ's name is
introduced, as in our device, the verb becomes indicative, (νίκα), with
Christ's name in the nominative, not the vocative, case. The contrast is
clear in the De cerimoniis, with the formula Χριστός νικά (indicative),
while, in the acclamation of the emperor, the verb becomes imperative (εν
τούτω νικάτε) ι49. It seems that, where Christ and the Cross are concer
ned, the verb is invariably indicative.
Although the victorious formula may well have been imperial in
origin, it was early used in an ecclesiastical context, as, for example, at
the council of Ephesus (431) : «Christ, our master, it is you who have
conquered ! Oh Cross, it is you who have conquered !» Here Christ and
the Cross receive the same status I5°. What can be argued more
convincingly, in the absence of contrary evidence, is that our device
became an iconographical formula in imperial art. On the miliaresion
and, perhaps, on restored fortifications, the device acquired what was to
become its habitual form in the early eighth century 151. It served as a
challenge to the Moslem Arabs. The invocation of the Cross (along with
that of the warrior saints) would have continued in battle, long after our
device had disappeared from the imperial iconographical repertory 152.

145. Cyril of Jerusalem (4th century), Catéchèses ad illitminamlos 13 36

(Clavis 3585), PG 33, col. 816.
146. G. SuBOTii, Ohridski slikar Konstantin i njegov sin Jovan, Zogra/5, 1974, p. 44-
47.
147. See above, note 3.
148. See above, note 7.
149. De Cerimoniis II, 43, edited J. J. Reiske, Bonn 1829, p. 650; ibidem, I, 69,
p. 324; O. Treitinger, Die oströmische Kaiser- und Reichsidee, Darmstadt 1956, p. 178
and note 70.
150. P. Battifol, Un épisode du concile d'Éphèse, d'après les actes coptes de
Boudant, Mélanges Schlumberger, Paris 1925, 1, p. 32.
151. See above, p. 195.
152. For example, Joseph Genesios, Historia de rebus constantinopolitanis IV,
PG 109, col. 1109 (mid-10th century): the victorious Saviour's Cross is acclaimed in
 THE APOTROPAIC FUNCTION OF THE VICTORIOUS CROSS 21 5

A clearly dated legend close to ours, accompanying a cross, appeared

at Philae in Justinian's reign 153. Here its triumphal connotation -
Christian cult superseding pagan cult - is evident. However, when the
whole device is attested in Paris, graec. 510 towards the end of the ninth
century, it has acquired an apotropaical function. There is no adequate
evidence that it had long been used on Eucharistie bread stamps, any
more than in iconoclast polemics l54.
The apotropaical function would continue to be characteristic in all
religious media, on tombs, reliquaries, icons (extremely rarely), and
above all in manuscripts and churches. Other legends might modify or
supplement the message of the device, but its basic significance did not
change : the victorious cross on which Christ was crucified, the all-
powerful protector against evil.

Christopher Walter
10 avenue de la République
94 300 Vincennes

battle against the Saracens, το Σωτήρος νενίκηκεν. See Thierry, Le culte de la Croix,
art. cit. (noie 6), p. 208.
153. Sec above, note 104.
154. Sec above, p. 200.
 216 CHRISTOPHER WALTER

Illustration non autorisée à la diffusion

Figure 1 - Akören

Illustration non autorisée à la diffusion Illustration non autorisée à la diffusion

Figures 2 and 3 - Stamp with IC XC NI ΚΑ

THE APOTROPAIC FUNCTION OF THE VICTORIOUS CROSS 21 7

Illustration non autorisée à la diffusion

Figure 4 - Athens cod. 108,1". Γ.

218 CHRISTOPHER WALTER

Illustration non autorisée à la diffusion

Figure 5 - Taxiarchis, Kastoria

 THE APOTROPAIC FUNCTION OF THE VICTORIOUS CROSS 219

J>-,

Illustration non autorisée à la diffusion

Figure 6 - Sopocani

'4'

Illustration non autorisée à la diffusion

Figure 7- Gracanica
 220 CHRISTOPHER WALTER

Illustration non autorisée à la diffusion

Illustration non autorisée à la diffusion

Figure 8 - Saints Constantine and Figure 9 - Parecclesion of Saint John

Helena, Ohrid the Baptist, Saint Sophia, Ochrid

*** Credit lines : Figure 1, Nicole Thierry. Figure 2 and 3, British Museum. Figure 4,
Greek Academy, Athens. Figure 5, George Gerov (Sofia). Figures 6 and 7, B. Zivkovic
(Belgrade). Figures 8 and 9, G. Subotic (Belgrade).