1. Art and Diseño

ART
THE
292
BULLETIN
It is hoped that some day further researchwill bring ern type of Castelseprioand the Carolingian parallels.
to light facts which will prove conclusivelywhether or I proposed further the view that the frescoes, though
based on Byzantine art, were not of the tenth century,
not this piece was in the Valenti-Gonzaga collection.
as Weitzmann and others maintained, but belonged
rather to the second half of the eighth century at the
UNIONMUSEUM
COOPER
earliestand probablyto the late eighth century, because
of the relations to early Carolingian art and to Italian
works of the end of the eighth and the early ninth
NOTES ON CASTELSEPRIO
century.' This connection with the art of the CaroMEYER SCHAPIRO
lingian period I had already remarked in reviewing
the book of Bognetti and de Capitani d'Arzago in
I. THE THREE-RAYED NIMBUS
1950.4
In an article in Cahiers drcheologiques (vII, 1954,
pp. 157-159), Professor Grabar writes that in citing
his observationof the peculiarsymptomaticcross nimbus
in the frescoesof Castelseprio,with the lines of the cross
extending beyond the circle of the nimbus, "M. Schapiro a pens6 pouvoir en r6duire la port&een joignant
aux exemples carolingiens et ottoniens de ce genre de
nimbe crucifere deux exemples byzantins." He goes on
to argue that these two examplesare not at all relevant
to the problem, for in one case-the Cotton Genesis
(Fig. I)'--the arms of the cross are broad bands and
not thin lines as in the frescoes (Fig. 3), and in the
other case-Athens Ms 211, a work of the tenth century-?there is not even a cross nimbus, but rather a
candlestick in the form of a cross, independent of the
nimbus and illustrating a sermon of St. John Chrysostom (Fig. 2). Hence the detail at Castelseprioremains
a Western peculiarityand an evidence of the late date
of the frescoes, since it is found in this form only in
Carolingianand Ottonian works.
I hope this detail will not seem too small and unimportant to warrant further study and the reader's attention; and I hope I shall be forgiven if I take this
opportunity to restate my views on the subject of
Castelseprio.
Although Professor Grabar links my name with
Morey and Bognetti as proposing dates "sensiblement
plus anciennes que l'epoque carolingienne," the conclusion of my article was preciselythat the fresco cycle
could not be of the seventh century, as Morey and
Bognetti believed, since it contains elements which are
first known to us in the art of the Carolingian period.
I regarded the type of cross nimbus as one of these elements, and in citing the two Greek examples I distinguishedtheir form from the more specificallyWest-
In contrasting the form of the cross nimbus in the
Greek and the Western examples, the first with a
"solid, material" cross, the second with "a thin line
[of the cross], an effect more suggestiveof the luminous
and emanatory than of the instrument of the Crucifixion," I wished also to offer an hypothesisto explain
the distinctive features of the cross nimbus in the
frescoes. I supposed that the three ray-like lines, resembling the raysissuingfrom the nimbusof the phoenix
and the personifiedsun in Roman and early Christian
art, symbolizedthe emanatoryaspectof the divinebeing,
the threefold Godhead as a "superessentialray," "an
originating beam," "an overflowing radiance"-metaphors found in the writings of the pseudo-Dionysus.
But such imagery was not peculiarto this Greek writer;
the underlying idea was common enough and had been
expressedearlier and in a manner more pertinentto the
frescoes by a Latin author, Tertullian, in speaking of
the Logos and the Incarnation: "When a ray is projected from the sun it is a portion of the whole sun;
but the sun will be in the ray because it is a ray of the
sun; the substanceof the sun is not separatedbut extended, as light is kindled from light. . . . So from
spirit comes spirit,and God from God .... This ray of
God, as had always been prophesiedbefore, descended
into a Virgin and having been incarnated in her womb
was born a man-God." (Apologeticum xxi. 12-14)
Philo in the first century had already written of the
Logos as "the radiant light of God"' and in the New
Testament the Son is describedas "the brightness of
God's glory" (Hebrews 1:3). In the controversies
during the fourth century over the consubstantialityof
the Father and the Son, the relation of the ray to its
source was the most cogent example of emanation and
of distinct forms with a common substance.6Writing
i. After H. Omont, Miniatures des plus anciens manuscrits
grecs de la Bibliothkque nationale du VIe au XIVe sikcle, Paris,
1929, pl. opp. p. iv.
2. After A. Grabar, Cahiers arch~ologiques, VII, 1954, pl.
G. P. Wetter, PHOS, Uppsala, 1914; Fr. J. D*lger, Antike
379;
und Christentum, I, 1929 ("Sonne und Sonnenstrahle als Gleichnis in der Logostheologie des christlichen Altertums"), pp. 27129o; idem, VI, 194o ("Das Sonnengleichnis in einer Weihnachtspredigt des Bischofs Zeno von Verona. Christus als wahre
und ewige Sonne"), pp. 1-56; R. Bultmann, "Zur Geschichte
der Lichtsymbolik im Altertum," Philologus, xcvII, 1948,
pp. 1-36.
Since the heretical Arians rejected the metaphor of light in
explaining the Trinity (A. Harnack, History of Dogma, London, 1905, Iv, pp.I- , 41), the three-rayed nimbus might be
regarded as more characteristically Western in the fifth and
sixth centuries, for Arianism was more important then in
Italy and Spain than in the East 5 and the frequency of the
three-rayed nimbus in the Carolingian period might be con-
LIV.
3. ART BULLETIN,
XXXIV,
1952,
pp. 156,
I6o,
I62,
163.
4. The Magazine of Art, December, 1950, pp. 3i2f.
5. De Somniis i. 13, 72; cited by Harry A. Wolfson, The
Philosophy of the Church Fathers (I, Faith, Trinity, Incarnation) Cambridge (Mass.), pp. 300, 301.
6. On the analogy of light in early Christian theology and
particularly in the doctrine of the Trinity, see besides Wolfson,
op.cit., pp. 3ooff., 359ff., Cl. Bauemker, Witelo (Beitrige zur
Geschichte der Philosophie und der Theologie des Mittelalters,
III, 2), Miinster, 90o8, pp. 357-433 and especially pp. 371-
NOTES
293
of the Trinityin the symbolicdecorationof the church the light is set abovea cross."The artistof the Athens
of Hagia Sophia,the poet Corippusexclaimed:"Sub- manuscript,or perhapsthe painterof his model, had
sistite,numinafulgent! Natus, non factus,plenumde anothertype of light in mind: one in which a cross
surmountsa lamp." The thoughtunderlyingthisearly
luminelumen."'
In the backgroundof these theologicalmetaphors Christiantype is often expressedin the Greek phrase
was also the conceptionof Christas the true Sun, re- inscribedon the lamp:"The light of Christshinesfor
all."
placingthe pagansolargods.
In the Athensminiaturethe bustof Christas well as
The heavy armsof the crossin the two GreekexamplesI had cited did not seem to me relevantto this the crossissuesfrom the lamp in Adam'shand. The
explanation,although they extend, like the rays at questionis whetherthe crossherealsobelongswith the
Castelseprio,beyond the circle of the nimbus. But nimbusas an attributeof Christor only symbolizesthe
after reading ProfessorGrabar'scommentson my metaphorical
lampstand(or candlestick)of thehomilist.
article,it occurredto me upon furtherstudy that the If the latterwere the case,it would be hardto undermassivecrossesin the Greek examplesalso originated stand why the cross shouldbe set above ratherthan
in a metaphorof God as light and that both typesof below the lamp and why the horizontalarms should
cross nimbus--with broador with thin ray-likearms extend beyond the nimbus; they scarcelylook like
andin any casethe comextendingbeyondthe circle-go backto earlyChristian partsof a standor candlestick,
binationof lampand candlestickseemsstrange.When
models.
Grabarhas observedthat in the Athens National the artistwisheson the samepageto representa candle
LibraryMS211 the miniatureshowinga bustof Christ in the handsof the angels,he drawsit clearlyenough,
with threearmsof a crossreachingbeyondthe nimbus and wherehe wishesto depicta crossdistinctfromthe
(Fig. 2) rendersvery preciselythe wordsof a sermon nimbus,as in his drawingof the Holy Spiriton another
by St. John Chrysostomon the parableof the ten leaf, the cross is set behind rather than within the
drachmas(Luke I5:8-Io). Christ,accordingto Gra- nimbus,in passingbeyondit."2Grabarinterpretsthis
bar, is the tenth drachmaand is held by Adam in a crossas an actualmetalobjectplacedon an altar; but
lamp; the other nine are the accompanyingangels. he ignoresthe materialprototypesof the lamp in the
The bust of Christwith the doubletraversecrossbe- miniatureof the tenth drachma.
hindhim illustratesthe wordsof the homilistfor whom
Is it not moreplausibleto supposethat the artist,in
Christis the Divine Wisdom "whichlights the lamp order to symbolizethe illuminationby Christ, "the
and placingit in the lampstandof the Cross,bearsthe light of the world," the "lucernaardenset lucens"
light and leads the whole world to piety." Christ (Augustine),and to renderthe homilist'simageof the
"comesfrom heaven,takesthe clay lampwhich is the crossas the lampstandor candlestick,
hascondensedthe
body,lightsit with the light of divinity,and sets up the chain of metaphorsin a cross nimbus of decidedly
candlestickof the cross."What I had taken to be a radiantaspect?The color re-enforcesthis effect: the
cross nimbusis therefore,in Grabar'sopinion,quite nimbus,the cross,and Christ'sapparelare all painted
anotherthing; the cross, symbolizingthe candlestick yellow.
and renderingliterallythe text of the accompanying That light could be representedin this more solid
sermon, "has nothingto do with the crossnimbus."8 form and even with a thickeningof the raysor bands
I am not convincedby ProfessorGrabar'sinterpre- at the outerend is confirmedby the imageof Heliosin
tation.Is not the tenth drachmaAdam who holdsthe the recentlydiscoveredceilingmosaicin the pre-Conlamp?9and is not Christthe light by whichAdam,the stantinianmausoleumunderneathSt. Peter'sin Rome
lost drachma,is found? In identifyingthe cross of (Fig. 4)." The purplishrays,formingmassivebundles,
Christ with the lampstand(lychnia) that bears the extend beyondthe halo as in the Cotton Genesisand
light, the author of the sermon describesa familiar the Athensmanuscript.
But even withoutthe testimonyof this mosaic,one
object of his time: a candelabrumor stand of which
nectedwith the controversiesover Adoptianismand the Filioque chritienneet de liturgie,s.v. "Lampe,"cols. Iio6ff., I202,
formula at that time, although these two questionswere in- x209, I2xo; Fr. J. D6ager, Antike und Christentum, v, 1936,
dependentof the old Arian issue.
pls. I, II.
12. On fol. 56, cf. Grabar, loc.cit., 1932, pl. xviii, 2 and
7. For the whole text, see A. Heisenberg'sarticle in Xenia,
Hommage international ai l'Universite nationale de Grice,
pp. 263, 278. Cf. for a similar effect the cross with a medallion
Athens, 1912, p. 152.
8. Op.cit., p. 158.
9. As the sermon clearly states (Migne, Pat. gr., LXI, col.
781) and as Grabar recognized in his publication of the miniature in 1932: "Un manuscrit des homilies de Saint Jean
Chrysostome a la Bibliotheque nationale d'Athenes (Athenensis
211 ) ," Seminarium Kondakovianum, Recueil d'ktudes, Prague,
v, I932, pp. 259-297, especially p. 272. Although attributed
to St. John Chrysostom in the manuscript, this sermon is published in Migne among the spuria.
0o. For an example see H. Leclercq, Manuel d'archdologie
chritienne,
Paris,
1907,
II, p. 570, fig. 379 (Cairo
ix. For examples see Cabrol, Dictionnaire
Museum).
d'archiologie
bust of Christ in front of the cross in Paris, Bibl. nat., Ms gr.
20, fol. 7.
13. See Esplorazioni sotto la confessione di San Pietro in
Vaticano, ed. by B. M. A. Ghetti et al., Vatican City, 195p,
I, p. 41, pls. B, C (in color, ii, pl. xI from which our Fig. 4
is reproduced) 5 Jocelyn Toynbee and J. Ward Perkins, The
Shrine of St. Peter and the Vatican Excavations, London, 2956,
pp. 42, x 7, pl. 32. Toynbee and Perkins date the mosaic in
the mid-third century and call the figure Christus-Helios: "It
may be no accident that the rays to right and left of his nimbus
trace strongly accented horizontal lines, forcibly suggesting
the transverse bars of a cross."
THE
294
ART
BULLETIN
could have surmised the possibilityfrom other works."'
There is one especiallyin which the connection of the
cross with solid bands of emanating light is perfectly
clear; the mosaic of the bema of the church of the
Dormition at Nicaea.15Here eight broad bands of light
issue from a cross set on the throne of the Hetoimasia;
four of these prolong the arms and vertical parts of the
cross, the other four issue from the angles. They all
traverse three great concentric circles which probably
symbolize the Trinity. In the same church the mosaic
of the apse repeatsthe theme of light in a more explicit
Trinitarian sense: three rays in the form of bands issue
from the hand of God, passing through two concentric
circular segments into the semicircular golden space
below."1
Let us note too that these unquestionablerays of
light are of changing color: white, grey-blue, and greenblue in the first mosaic; rose, grey, and green in the
second. Professor Bognetti has objected to the interpretation of the three lines of the cross nimbus in Castelseprio as rays or symbols of light that they are painted
in a darkcolor." This is due, I believe,to the fact that
they are set on the light groundof the nimbus.In the
mosaicbelow St. Peter's, as in the later mosaicsin
Nicaea and in many other works, rays of light are
representedin dark tones (grey, blue, green, and
purple)."sBlackraysare not uncommonin mediaeval
sun."9The colorof raysvaries
imagesof the personified
in mediaevalart just like the colorof the nimbus,anothersymbolof light.
I maypointoutfinallythatin Carolingian
artthesolid
crosswith armsprojectingbeyondthe nimbusis sometimes formedof parallellinear rays closelyaligned.20
Hence from the pointof view of theologicalmeaning and origin-though not of style-the Carolingian
examplesof the nimbuswith the linearray crossneed
not be separated,as I oncesupposed,
fromthe solidtype.
Both dependprobablyon modelsof the fourthor fifth
century--the periodduring which the solar analogy
was common and Helios was representedwith rays
issuingfrom the nimbusin the two formsthat I have
The theologicallysignificantchoiceof the
described.2'
14. For the same type of rayed nimbus as in the mosaic
under St. Peter's, cf. the painting of Helios in the Vatican
Virgil (lat. MS 3225), (Fragmenta et picturae Vergiliana,
Rome, 1930, pl. 6) the pavement mosaic of the phoenix in
Antioch (Doro Levi, Antioch Mosaic Pavements, Princeton,
1947, II, pl. LXXXIII); the mosaic of Christ on the arch of
triumph of S. Paolo f.l.m. in Rome, which may be of the late
eighth century (M. van Berchem and E. Clouzot, Les mosaiques
de l'Institut de philologie et d'histoire orientales et slaves
(Universit6 Libre de Bruxelles) Ix, 1949, pp. 593ff. Visions
of the luminous cross are also mentioned in Gnostic-Christian
writings (Acts of John and Philip) (R. A. Lipsius, Die
apokryphen Apostelgeschichten und Apostellegenden, 1883900oo,I, pp. 452, 5235 III, pp. 9, 16, 37). In the Acts of
John, which were read at the Council of Nicaea, Christ shows
John a luminous cross which he calls Logos, Nous, Spirit and
Life, as well as Christ.
I6. Schmit, op.cit., pl. xx, pp. 29, 30.
17. See Cahiers arch~ologiques, viI, i954, p. 143 n. 2.
18. Cf. also the sun in the scene of Joshua's miracle in the
mosaics of S. Maria Maggiore in Rome; the miniature of
Joseph's Dream in the Vienna Genesis (Fr. Wickhoff, Die
Wiener Genesis, Vienna 1895, pl. 29); the painting of Apollo
in Pompeii cited in note 14 above; the bust of the sun in
British Museum, Harley Ms 647, fol. 13v (Catalogue of
Astrological and Mythological Manuscripts of the Latin Middle Ages, III, Manuscripts in English Libraries, by Fr. Saxl
and Hans Meier, edited by Harry Bober, London, 1953, II,
pl. LVII, fig. 148); the sun in Madrid, Bibl. Nac. Ms 19
(A. 16) (ibid., I, fig. 9).
19. Cf. the Leyden Aratus (Voss. 69) (Thiele, op.cit., p.
121,
fig. 46 [three black and three gold rays])5 British
Museum, Cotton Tiberius C. I (Saxl, Meier, and Bober, op.cit.,
chritiennes,
Geneva,
1924, figs. 100,
101, p. 89);
the Romano-
British relief of Sol Invictus in Corbridge (T. D. Kendrick,
Anglo-Saxon Art to A.D. 900, London, 1938, pl. XIII, I:
here the slender rays crossing the nimbus may be regarded as
lines rather than as solid bands). More common are the solid
pointed, i.e. peaked or convergent, rays passing across the
nimbus: the figure of Apollo in a Pompeian fresco (0. Brendel,
R'mische Mittgeilungen, LI, 1936, p. 57, fig. 8); the relief
of Mithras in Nimrud Dagh (Fr. Sarre, Die Kunst des alten
Persiens, Berlin, 1923, pl. 56); the solar personifications in
Mithraic art (Fr. Cumont, Textes et monuments relatifs aux
mystires de Mithra, Brussels, I894-1900, II, fig. 29, p. 202
[no. 18] and passim); coins of Antoninus Pius (Daremberg
et Saglio, Dictionnaire des antiquitis grecques et romaines, s.v.
"Nimbe," fig. 5321); for mediaeval examples, cf. the sun in
Joshua's miracle of the sun in the Vatican Joshua Roll, the
bust of the Sun in Phillips Ms 1830 (Thiele, Antike Himmelsbilder, Berlin, 1898, fig. 72).
Theodor Schmit, Die Koimesis-Kirche von Nikaia,
i5.
Berlin and Leipzig, 1927, pl. XII, p. 21; O. Wulff, Die
Koimesis Kirche in Niciia und i/re Mosaiken, Strasbourg,
1903, pl. I. For other examples of bands of light extending
from the cross beyond the enclosing circle, cf. the fresco of
the sixth century in the apse of the annex to the cathedral of
Rusafa-Sergiopolis (J. Lassus, Sanctuaires chretiens de Syrie,
Paris, 1947, fig. Io9) ; fresco in tomb chamber in Sofia, sixth
century (E. K. Riedin, The Christian Topography of Cosmas
Indicopleustes [in Russian], Moscow, 1916, fig. 2, p. 6) ;
mosaic of Hagia Sophia in Salonica (van Berchem and Clouzot,
op.cit., p. 182); a newly uncovered mosaic in Hagia Sophia
in Istanbul (Dumbarton Oaks Papers, Ix and x, I956, fig. iio,
after p. 300). For the possible connection of the luminous cross
in painting and mosaic with early Christian legends about the
luminosity of the true cross and the miraculous appearance of
a radiant cross in the sky over Jerusalem in 351 (or 353), see
O. Wulff, op.cit., pp. 241, 243 and A. Grabar, Martyrium,
Paris, 1946, II, p. 276; for the occasion and date of Cyril's
vision of the celestial cross, see J. Vogt, "Berichte iiber
Kreuzerscheinungen aus dem 4. Jahrhundert n.Chr.," Annuaire
I, fig. 147).
20. Cf. the Egino codex, Berlin Phillips Ms
1676, fol. 24
(E. Arslan, "La pittura e la scultura veronese dal secolo VIII
al secolo XIII," Milan, 1943, pl. 42)5 the Lorsch Gospels
(Boinet, La miniature carolingienne, Paris, 1913, pl. xvI, A) ;
the Gospels of Soissons (Bibl. nat. MS lat. 8850) (in ibid., pl.
xx); the Smyrna Physiologus, bands of triple rays from the
bust of the Sun (J. Strzygowski, Der Bilderkreis des griechischen Physiologus, Leipzig, 1899, pl. Iv). Such bands are common in images of the Anastasis, issuing from Christ's body. Cf.
the mosaic in the chapel of S. Zeno in Sta. Prassede, Rome
(J. Wilpert, Die r6mischen Mosaiken und Malereien, Freiburg
im Br., 1917, pl. 114, 4).
21. For the Sun or Helios with single linear rays, cf. the
Joshua scene in the Rabula Gospels, and the relief of the Sun
in Corbridge cited in note 14 above. Both types of rays, the
solid band and the linear, occur also in images of the phoenix:
for the first, see the mosaic in Antioch cited in note I4; for
the linear type, see examples I have cited in ART BULLETIN,
1952, p. 156 n. 56; for a list of examples on Roman coins, see
L. Stephani, "Nimbus und Strahlenkranz," Mimoires de
l'Acadimie Impiriale des Sciences de St. Pitersbourg, 6e sBrie
na t-?
•s
!
?w~~i
00 Cta.40,
o~C"
h~o•••.•
::':•:i:::
•i
uaoEr0
•.IU,
d.•!
•'
.'.,/K
i,)..
'4
. ...
'
.,/
41b
.•.•.•; ..,
IT01 - I
,
' "
.
-," -
4
"t....,,II
(A..
i. Creation, The Third Day
(detail)
Cotton Genesis (Drawing after L::
destroyed1kminiature)
2.
!a
%4
X,,'.
The Tenth Drachma, Homilies of St. John Chrysostom
Athens, National Library, ms z i i, fol. 34.V
1~1*
~~!
.. .
A7
0
'.'.•
I.
-
~,"r~rj
3. Adoration of the MagiCatlero
Sta. Maria (Drawingafefrso
i-t...
.:: r,.
I
,k
a~
~
ebN
t: tA
-rc
miniature)'
destoye
?/
,F
t
..
5.
.1exnde
/
•
;,•
edalio
Gratbroze
th
Rom,
Vticn
Mseu
5. Alexander the Great, bronze medallion
Rome, Vatican Museum
4. Helios, ceiling mosaic of underground mausoleum
Rome, St. Peter's
6. Dusseldorf, Landesbibliothek, Ms
3
N
-
•i~i!
ir_
al
i
j\"i.•
%LL
8. Crucifixion. Engraved Gem. London, British Museum
7. Christ Enthroned. Vercelli, Biblioteca Capitolare,
MS CLXV
YE)
9. Annunciation to Joseph. Castelseprio, Sta. Maria (Courtesy Frick Art Reference Library)
iIi
fA
A
o. Mounted Amazons Silk
Lyons Collection C. Cote
Tapestry.
Giraudon)
(photo"
iA
s s. The EmGperor Theodosius II. Vercelli,
Biblioteca Capitolare, MS CLXv
NOTES
three rays of Christ's halo may be likened to the seven
rays of the sun in Mithraic art, also a theological attribute.22
The recurrence of the solid emanatory cross in the
Creation scenes of the Cotton Genesis agrees with the
theology that speaks of the divine light especially in
connection with the Creation. In the same miniature
(Fig. i) the nimbus is marked with many small rays, a
device that reappearsin Carolingianart.
(I have assumed until now that the three rays are
parts of the cross and merely a variant of the "cross
nimbus." But the facts presented here suggest another
interpretation:that besides the true cross nimbus and
monogram nimbus in which all four bars were more
or less clearly drawn-as in the mosaics of Sta. Maria
Maggiore and the door of Sta. Sabina23-there existed
very early a three-rayednimbus of which the rays were
at first conceivedless as partsof a cross than as a symbol
of the Trinity. In time this emblem was conflated with
the cross in the familiar form in which the fourth part
is covered by Christ's head. It is a remarkablefact that
although Christian art develops from naturalistic to
increasinglysymbolicmodes of representationand tends
to present its symbols as distinct surface forms, the
cross in the nimbus remains a symbolic object in real
space overlappedand obscuredby the head of Christ.)
The tiny stroke at the end of the rays in Castelseprio
(and in several other Western works) seems to Bognetti incompatible with the interpretation of the lines
295
as rays.2"This is hardly a decisive objection since the
rays in the image of Helios beneath St. Peter's in Rome
(Fig. 4) are thickened at the end. In early Christian
mosaics stars are often, if not usually, represented by
radiating lines with knobbed ends;25 this, in fact, is
how the painter of Castelsepriohas drawn the great
star of Bethlehem in the scene of the Nativity. In pagan
and Christian works the rays of the sun and of solar
divinities end in dots or knobs that might symbolize
the stars or planets,26 a device which survives into the
Middle Ages, especiallyin pictures of the Apocalyptic
Woman "clothed with the sun" (Revelations 12).27
On a medallionof the late fourth century in the Vatican
Museum, the emblem of the sun accompanyinga head
of Alexander the Great is a circle from which radiate
wheel-like spokes with short transverse lines at the
ends (Fig. 5).28 This schema for rays of light is not
uncommon in Christian art. Similar nail-headed rays
appear on the stars in the mosaic of Sta. Agnese in
Rome29 and in the frescoes of Civate and SaintSavin;so they may be seen too in the personificationof
Day in the Gumpert Bible in Erlangen."s The nailheaded form in Castelsepriomay therefore be regarded
as a variant of an older convention for the luminous
cross. Even if the end strokes are derived from a type
of cross, the arms extended beyond the nimbusmay still
be interpreted as ray-like elements which owe their
characterto the theological metaphorof Christ as light.
I am not at all sure that the cross with projecting
(Sciencespolitiques,histoire et philologie), Ix, 1859, pp. 444- pp. 84ff.); the nimbedbust of the Sun in VaticanMs gr. 699;
446. For the nimbus with linear rays on solar figures carved Cosmas Indicopleustes (C. Stornaiolo, Le miniature della
on amulets, see Campbell Bonner, Studies in Magical Amulets,
Ann Arbor, 195o, pl. xi, no. 236, pl. Iv, nos. 83 and 86; and
the same writer's article "Amuletschiefly in the British Museum," Hesperia, xx, 1951, pl. 97, no. 32.
topografia cristiana di Cosma Indicopleuste, Milan, g908, pl.
52) ; Sacramentaryof Henry II, Bamberg (A. Goldschmidt,
German Illumination, II, pl. 75); solar emblems on Gaulish
coins (A. Blanchet, Manuel du numismatique franpaise, I, 1912,
fig. 107). F. D5lger, Antike und Christentum, vi, i, 1940,
22. An interestingparallel to the three rays of Christ is the
description by the Carolingian poet, Sedulius Scottus (carm. p. 31, interpretsthe end-pointsor knobs of the rayed nimbus
3'), of an image of the personifiedMedicina painted in a or solar wheel as points of light (apices) rather than as stars.
27. Cf. the Bamberg Apocalypse (H. Walfflin, Die Bamhospital-she is representedwith three rays issuing from her
brow: "Haec regina potens rutilo descendit Olimpo. . . berger Apokalypse, Munich, 192i, pls. 29, 31); Valenciennes,
Fronteque florigera cui lumina terna coruscant." (J. von Bibl. mun. Ms 99, fol. 23 (Bulletin de' la Sociite Franpaise
Schlosser, Schriftquellen zur Geschichte der karolingischen
Kunst, Vienna, 1896, p. 381, no. 1027.) The analogy with
Peintures, 6e annie,
pour la Riproduction des Manuscrits
1922, pl. xxIII); Paris, Bibl. Nat., Nouv. Acq. lat. Ms 1132,
schrift, xxx, 1929-1930,
pp. 587-595, and especially p. 593.
24. See Cahiers archiologiques, viI, pp. 143, 144.
25. Cf. the mosaicsof Sta. Maria Maggiore, Rome, Adoration of the Magi (van Berchem and Clouzot, op.cit., fig. 51);
Naples, Baptisterydome (ibid., fig. i i9); Albenga, Baptistery
(ibid., fig. 128); Ravenna, S. Apollinare in Classe (ibid., fig.
the Cambrai Apocalypse (Boinet, op.cit., pl. io6 B);
202);
28. See Andreas Alfoildi, Die Kontorniaten, Ein verkanntes
Propagandamittel der stadtr6mischen Aristokratie in ihrem
Kampfe gegen das christliche Kaisertum, Budapest, 1943, I,
p. 133, no. 34, 11, pl. xLIII, 7; Jocelyn M. C. Toynbee, Roman
Medallions (Numismatic Studies, No. 5, The American Numismatic Society), New York, 1944, p. 236, n. 34, and pl.
Christ dependedperhapson the metaphorof Christ as medicus fol. 17. Cf. also angels with rayed nimbus and dots or little
or archiater common since Origen. See R. Arbesmann,"The circles at the ends of the rays, in Trier MS 31 (Apocalypse),
Concept of 'ChristusMedicus' in St. Augustine," Traditio, x, fol. 22; Cambrai Ms 386 (Apocalypse), fol. 13; Madrid,
Bibl. Nac. HH58 (Beatus), fol. 96v, 97; Annunciationby
1954, pp. Iff.
23. On the early types, see E. Weigand, "Der Monogramm- Filippo Lippi, National Gallery, Washington, D.C. (rays
nimbusauf der Tiir von S. Sabinain Rom," ByzantinischeZeit- with knobbedends).
the Milan gold altar of Vuolvinus, Nativity and Christ in
Glory; Codex Egberti,
lumination, II, pl. 6).
Magi,
(Goldschmidt,
German Il-
xxxxx, 5 (photograph). The same nail-headedrays appear on
amuletsof Chnoubis,a lion-headedsnakewith solar affinities-see Bonner, in Hesperia, XX, I951, pp. 339, 340, no. 65, and
pl. 99, and copies in nos. 66, 67. Bonner regards all three as
26. Cf. a Chnubisamulet reproducedby E. R. Goodenough, modern works based possibly on old models (see pp. 3o8f.).
29. Van Berchem and Clouzot, op.cit., fig. 247.
Jewish Symbols in the Graeco-Roman Period, New York, 1953,
30. In Civate, in the Apocalyptic Vision of the Dragon; in
III, fig. Io96; a statue of Attis from Ostia in the Lateran
Museum (F. Cumont, Les religions orientales dans le paganSaint-Savin, in the Creation of the Sun and Moon.
isme romain, 4th ed., Paris, I929, pl. IV, opp. p. 66); a
31. Georg Swarzenski, Die salzburger Malerei, Leipzig,
Gallo-Roman bronze figure of Dispater in the Walters Art 1914, 11, pl. xxxiv, fig. I 4.
Gallery (The Journal of the Walters Art Gallery, x, 1947,
296
THE
ART
BULLETIN
arms, whether solid or ray-like, was in every case designed to expressthe metaphoricalluminosityof Christ.
In some examples it might have been repeated as a
fossilized type without thought of its figurative meaning, much like words of which the original sense has
been lost. That the old meaning was still alive in the
region of Castelseprioat the end of the eighth century
seems evident, however, from the radiant aspect of the
cross on the medallion bust of Christ in the Egino
codex, a manuscript of homilies of the Latin church
fathers, which was written in Verona between 790 and
The three arms of the cross passing beyond the
799•82
circular frame are formed by bundles of rays. In another Lombard manuscript of the ninth century, the
Homilies of Gregory in the Chapter library of Vercelli
(MS CXLVIII),an ordinary closed cross nimbus is inscribed LUX, with one letter on each of the three solid
members of the cross.33
For the problem of the frescoes of Castelseprio-their date, their place in mediaeval art-it does not
matter perhapswhether the ray-like form of the cross
was conceived explicitly as a symbol of the divine light
or was only an inherited convention adopted because
of its congenial form. The rendering of the arms of the
cross by a rapid stroke reaching beyond the nimbus
seems to accord better with the impulsive sketchy style
of the artist than would a massive form. In the medallion bust of Christ in the same series, the cross is of the
more common type and is confined by the nimbus; but
this painting of Christ already suggests another attitude
of the artist-he is reproducingan iconic image, frontal
and severe, and acceptsthe conventionsof the canonical
model which are in principle opposed to the forms of
his own art.
Grabar believes, however, that the varieties of the
rayed cross nimbus have some historical significance
and provide a means of dating the frescoes. In another
article in which he deals more specificallywith the different types of cross that project beyond the nimbus,
Grabar distinguishesnot only between the thin and the
solid form but also between the thin cross with the
terminal stroke, like a nail head, as in Castelseprio,and
the thin cross without this element."4Carolingianartists
knew the latter, but Grabar has found examples of the
Castelsepriotype only in Ottonian works. He asserts
that while Carolingian artists invented the projecting
linear type, it was in the Ottonian period that the thin
cross first acquired the terminal strokes. From this he
concludes that although the painter of Castelsepriowas
strongly influenced by the art of Constantinople,these
frescoes are related to the Ottonian Renaissance and
must be more or less contemporarywith the Ottonian
miniatures in which the same type of cross appears."5
This is a surprisinginference, since it would bring
the date of the frescoesinto the second half of the tenth
century or even into the eleventh--the Salzburg manuscript which he has cited as the Ottonian parallel
(Morgan MS 780) was made about 1070. Elsewhere
he has spokenof the frescoes as works of the ninth century, and has admitted the possibilityof a still earlier
date in the Carolingian period."6
But it is unnecessary to go beyond his article in
order to criticize his reasoning or the fluctuationin his
dating of Castelseprio. For precisely this form of the
nimbed cross with the nail-head ending of the rays
occurs in Carolingian manuscripts:the Stuttgart and
Utrecht Psalters, a fragment now in Diisseldorf which
comes from the same center or school as the Utrecht
Psalter (Fig. 6),8 and, in the region of Castelseprio,
the Canons of Councils in the Chapter Library of
Vercelli (Fig. 7)o." All these, except the latter, are
listed by Grabar in his article, but he has overlooked
the terminal strokes of the cross and classifiedthe examples in these manuscriptswith the plain Carolingian
types.
The criterion on which Grabar built his conclusion
should requirehim then to date the frescoesearly in the
Carolingian period. But it would be imprudentto base
the dating on a single detail of which the history is still
so little known. If the frescoes are to be placed in the
later eighth century, as I have supposed,it is for a number of reasons beside the evidence of the rayed cross.
The latter has some value, however, since of the many
surviving works with the cross nimbus, none before the
Carolingian period shows the peculiar form that occurs
in Castelseprio,and the Carolingian works that present
this form share other traits with the frescoes.
Future discoveriesmay discloseexamplesof this cross
older than the Carolingian works. To Grabar's belief
that the simple linear cross extending beyond the
nimbus, but without the terminal strokes, was a
Carolingian invention one can oppose an engraved
gem in the British Museum attributed to the early
Christian period (Fig. 8).89 It is a red jasper found
32. Berlin, Staatsbibliothek, Phillips MS 1676 (E. Arslan,
op.cit., pp. 42, 43)33. See Noemi Gabrielli, "Le miniature delle omelie di San
Gregorio," Arte del primo millennio (Atti del IIO Convegno
per lo studio dell'Arte dell'alto medioevo, Pavia, i950) Turin,
n.d., pl. 149, pp. 3oIff. The same inscription LUX appears
also in Vercelli, Bibl. capitolare MS LXII, fol. zzv, Io3r.
34. "Les fresques de Castelseprio et l'occident," Friihmittelalterliche Kunst in den Alpenlandern (Akten zum III. internationalen Kongress fuir Friihmittelalterforschung, September
9-14, 1951), Olten and Lausanne, 1954, pp. 85-93.
35. Ibid., p. 89.
36. "Les fresques de Castelseprio," Gazette des Beaux-Arts,
July 1950 (published in 195x), pp. o17-xx4, and especially
pp. x 13, 14. See also Grabar's book, Byzantine Painting, New
York, 1953, p. 86.
37. Landesbibliothek, Ms 113, a fragment of Rabanus Maurus, De institutione clericorum; another example in a missal
of the tenth century in the same library (Ms D 3), in a drawing of the Crucifixion. Cf. also Carolingian examples in the
Marmoutiers sacramentary, Autun, Bibl. mun. Ms I9 bis
(Boinet, op.cit., pl. XLI), and Nancy evangiles (ibid., pl.
xxvii). For a clear example of the nimbus with nail-head rays
in the Utrecht psalter, see E. T. DeWald, The illustrations of
the Utrecht Psalter, Princeton, 1932, pl. xcI (fol. 57r).
38. MS CLXV.Cf. N. Gabrielli, op.cit., (note 33 above),
pp. 303ff. from which our Figs. 7 and I I are reproduced.
39. See A. de Longpirier and E. Le Blant in Mimoires de
la SocietWdes Antiquaires de France, xxx, 1866, p. III. The
gem is now in the British Museum (no. 5623x) and has been
NOTES
297
silkswith the same motif of the Amazons."'I cannot
regardas relevantMorey'sobservationthat the davus
"is of the samecharacacrossthe thighin Castelseprio
ter as the waistbandworn by David in the miniature
of the Penitencein the ParisPsalter."
One can hardlybe sure that thispeculiarbandwas
first usedin the eighthcentury;but until now I have
foundit in no worksthat can be securelydatedbefore
thistime.It couldhavebeena fashionin Constantinople
(or some other center) for a shorttime in the eighth
century,from which little of the art producedin the
capitalhas come down; such a form might have per2. THE CLAVUSON THE THIGH
sistedlonger in the art of provincialregions,like Italy
I passto another detail of the frescoeswhich has been and France,where it had been acceptedas a pictorial
a matter of dispute.In my articleon CastelseprioI elementratherthan as part of real costume.The fact
called attention to a peculiarityof costume that might thatamongsurvivingworksthe transverseclavusspanserve as an indicationof the date and perhapsthrow ning the thighfirstappearsin the West after the comlight on the origin of the work. It is the transverse mon use therein the seventhand eighthcenturiesof an
ornamentalbandsewn acrossthe middleof the thigh older Easterntypeof appliqu6on the thigh," seemsto
on the long tunic or the mantle (Fig. 9), an element confirmits laterorigin.
that is frequentin Carolingianand Italianart of the
For the possibility
that the Carolingianartistscopied
late eighth and ninth centuries."It was precededin this detail from a much earliermodel there is some
the seventhand eighthcenturiesby a relatedform-the indicationin a manuscriptthat I have cited:the illusshorterparagaudae-thatsurvivedinto the ninth cen- tratedSeduliusin the Mus6ePlantinin Antwerp (Ms
too from anotherorna- 126). The modelsof the miniaturesin this book,actury; it must be distinguished
the
on
horizontal
hose." Morey objected cordingto ProfessorWilhelm Koehler,our bestjudge
bands
ment,
to usingthisbit of costumeas an evidenceof a date in in thesematters,were Italianpaintingsof about
500.'8
the eighthcenturysincehe believedthat it had already But betweenthe Italianprototypes
and the Carolingian
occurredin a work of the sixth."5But in the figuresof copy there were undoubtedlymediatingversionsof
mountedAmazonson the silk fabricin the Cote Col- which one was preservedin England; an inscription
lectionin Lyons whichhe has cited,I find no traceof in the Antwerpcodexreproducedfroman earliercopy
such a transverseclavus acrossthe thigh (Fig. io). namesan owner Cuduuinus,whom Traube identified
Morey has mistakenfor the element I had described as the Bishopof Dunwichbetween716 and 731 ." The
a simplehem at the lower edge of a shorttunic that transverseclavusin theseminiaturesmay belongto the
reachesto the thigh.The sameformreappears
in other model of 5oo00
or to an Italianor insularcopy--whatin Gaza, with a scene of the Crucifixionand an undecipheredinscriptionthat Le Blant believedwas of
gnosticorigin." I cannotsay more aboutthis object;
its dateis uncertainand nothingdecisiveseemsto have
beenaddedto the commentsof de Longperierand Le
Blant publishedin 1867." On the other hand, the
broad-armedcrossextendingbeyondthe nimbus,as it
appearsin the Cotton Genesis,is less rarein the early
Christianperiodthan I had supposed.It is foundalso
on two engravedgemsin CambridgeandThe Hague.'2
the late eighth century.
44. As on the hose of a Persian Magus in the Adoration
scene in the Menologion of Basil II in Vatican
MS gr. 1613
(Weitzmann, The Fresco Cycle of S. Maria di Castelseprio,
40. A. de Longpirier and E. Le Blant, loc.cit.
Princeton, 1951, fig. 66), and as on the costumeof lay figures
41. For later commentsee, besidesBonner, loc.cit., L. Br&- in the tenth centuryAthens MS2 x (reproducedby Grabarin
hier, L'art chritien, 2nd ed., Paris, 2928, p. 8 i; R. Zahn and Recueil d'itudes, Seminarium Kondakovianum, Prague, v,
J. Reil, "OrpheosBakkikos," Angelos II, 1926, pp. 63, 64; 1932, pl. xvIII, 2).
J. Reil, Christusam Kreuz in der Bildkunstder Karolingerzeit,
45. "Castelseprio and the Byzantine Renaissance," ARTBULLeipzig, 1930, p. 3.
LETIN, XXXIV, 1952, p. 199 n. 7o.
42. See Cabrol, Dictionnaire, s.v. "Gemmes,"col. 846, fig.
46. Cf. the piece in Cologne--Peirce and Tyler, L'Art
reproduced by Bonner, op.cit., Hesperia, xx, 1951, pp. 336,
337, no. 54, pl. 98. For this informationI am grateful to Mr.
John Beckwith and Dr. A. A. Barb. Our Fig. 8 is from the
engraving in Cabrol, Dictionnaire, s. "Gemmes,"fig. 4945.
5088 (The Hague, Museum, Entry into Jerusalem), fig. 5090
Byzantin, nI, pl. I85, and also Ars Orientalis, I, 1954, opp.
(Lewis Collection, CorpusChristi College, Cambridge,Cruci- p. 192 for an example in DumbartonOaks and one formerly
in the Sangiorgi Collection. Dr. Florence Day (ibid., p. 240)
fixion). On the latter see also E. Babelon, Bulletin de la Sociite
des Antiquaires de France, LVII,
pp. 194, 195-he at- cites the shroud of St. Fridolin as the earliest example of the
tributes the gem to Syria in thex896,
seventh or eighth century. Amazon motif in silk, dating it in the sixth century,but notes
Note also a trace of the same type of rayed cross nimbus in that the type continuesinto the eighth centuryand into Islamic
the encolpium of a gold cross in the Dzyalinska collection at art.
Goluchow, an Italian work attributedvariously to the sixth,
47. See ART BULLETIN, XXXIV, 1952, p. i6o, for a brief
seventh and eighth centuries (Cabrol, Dictionnaire, s.v. "As- account of its history.
somption," col. 2993, fig. 1027).
48. "Die Denkmailerder karolingischenKunst in Belgien"
43. See ART BULLETIN,XXXIV,1952, p. I6o. To the exam-
in Belgische
Kunstdenkmdiler, Munich, 1923, pp. 9,
figs.
x9,
ples listed there should be added the ValenciennesApocalypse 8, 9.
(Ms 99, fol. 3) and the recently discovered frescoes of the
49. In the Neues Archiv der Gesellschaftfifr altere deutsche
ninth centuryin St. John at Miistair (Switzerland) (L. Birch- Geschichtskunde,
9xxviI, 09o, pp. 267ff. A. S. Cook has identiler, "Zur karolingischenArchitekturund Malerei in Miinster- fied the name with another person ("Bishop Cuthwini of
Miistair," Frilhmittelalterliche
Kunst in den Alpenliindern,
Leicester [68o-691],"
Speculum, II, 1927, pp. 253-257);
but
Olten and Lausanne,1954, pp. 167-252, fig. 96). An example W. Levison, England and the Continent, 1946, pp. 133, 134,
in the Morgan Library Beatus, Ms 644, fol. 9v (also fol. has rejectedCook's opinion and supportedTraube.
I8Iv), early tenth century, may go back to the model of
298
THE
ART
BULLETIN
ever its date-or it may have been introduced for the
first time in the Carolingian version." The initial P
on folio 8 is clearly of the later eighth or ninth century
and the furniture, too, points to this period. If the
miniaturesdepend ultimately on prototypesof the early
sixth century, this fact does not entail the same date for
the ornament of the costume.
The same questionarisesfor the Trier Apocalypse,a
work attributedby Goldschmidtto the end of the eighth
century."5The transversethigh band occurs here often
in the form which we observe at Castelseprio; and
since the illustrations in this manuscript were copied
from older models, perhaps of the sixth century, one
must consider the possibilitythat this detail was part of
the original work. There are two reasons, however,
for doubting this. First, the thigh band, so frequent in
the Trier Apocalypse, does not appear in the Carolingian sister manuscriptin Cambrai (386) which surely
descended from the same prototype." In the second
place, the thigh band does occur on a flyleaf of the
Trier manuscript in an early Carolingian drawing of
Christ Treading on the Beasts, which is independent
of the illustrations of the Apocalypse and by another
hand.5"It is close in style to the Genoels-Elderen ivory
plaques in Brussels, where the same thigh band is
" these
rendered;
ivory carvings are of the late eighth
century and combine insular and Italian features-they have been commonly placed in or near the region
of the Ada School."5I had noted before their relation
in iconography to the frescoes of Castelseprio (Annunciation, Visitation);" to which may be added-as
of some interest for the North Italian, as distinguished
from the Roman, connections of Carolingian art-the
resemblanceof the draperyforms in these ivory panels,
and particularlyof ChristTreading on the Beasts,to the
stucco figures of the eighth century in Cividale."7
There is a North Italian work which supports the
possibilitythat the motif of the thigh band, as it appears
in the Antwerp manuscriptof Sedulius and the Trier
Apocalypse,alreadyexistedin the first half of the eighth
century. In the sculpturedpanels of the altar in Cividale, dedicated around 737 to the memory of Duke
Pemmo (d. 734) by his son Ratchis, the transverse
stripes on the robes of the angels"5seem to represent
triplebandsof clavilike thoseon severalfiguresin the
Sedulius manuscriptand particularlyon the flying
angel in the sceneof Daniel who in other respectsbetrays a family resemblanceto the Cividaleangels."5
In the primitiverelief carvingof this altar, it is not
easy to say what representscostume ornamentand
what is ornamentalstylizationof folds. But a connection with the claviseemsto me possible.
EarlierI had noted the occurrenceof the clavuson
the thighin imagesof Lombardkingsin a manuscript
of laws in La Cava (Badia Ms 4), apparentlycopied
from a model of the ninth century."0
This suggested
the possibleoriginof the motif in royalcostume.It is
not at all characteristic
of the dressof Byzantineor
Carolingianrulers,though it does appearon the costume of kingsand crownedMagi in Ottonianart. In
Castelseprioand in Carolingianworks the clavus is
found more particularlyon the thighsof angels. But
in drawingsof the ninthcenturyin a nativemanuscript
in Vercelli(BibliotecaCapitolareMsCLXV,Canonsof
Councils)"6the clavus, decoratedwith dots or beads
is appliedjustbelowthe knee on enas in Castelseprio,
throned figures whose foreshortenedthighs are not
visible;theseare the emperorsConstantineand TheodosiusII (Fig. I1).Y2 Here one may assume,I think,
thatthe bandbelowthe kneehasbeentransferredfrom
the thigh. The transverseclavusin this manuscriptis
not limited to the emperors;two seatedclericsat a
What makesthese
councilwear the sameornament."6
for
even
more
interesting the problemof
drawings
that
in one of the miniatures(Fig.
Christ
is
Castelseprio
linearrays crossing
nail-headed
with
a
nimbus
has
7)
the circle."'All the thronesare of the simpleblock
type that appearsin the paintingof the Hetoimasiaat
Castelseprio.The drawingsin this manuscriptshow
that several distinctivemotifs of the frescoesexisted
elsewherein Lombardyin the ninthcentury.
It may be arguedthat the two elementsconsidered
in these notes-the rayedcrossnimbusand the thigh
band-occur in Ottonianas well as Carolingianart,
and that the presentstudy thereforepointsonly to a
probableterminus post quem for Castelseprio:the fres-
coes are not earlierthan the eighth centuryand may
well be of the tenth. If I inclineto the view that the
50o.The formhad alreadyreachedthe Northin the second 57. A. Haseloff, Pre-RomanesqueSculpturein Italy, New
half of the eighth centuryin late Merovingian art; it appears York (1930 ?), pls. 48-50, and articles by Hj. Torp and
clearly in a manuscriptof Corbie,Paris, Bibl. Nat. lat. 11627, L'Orangein the Atti del 2o Congressointernazionaledi studi
fol. iv (E. H. Zimmermann,VorkarolingischeMiniaturen, sull' alto medioevo, Spoleto, 1953.
Berlin, 19x6-x9x8, plates, vol. II, pl. 109).
58. C. Cecchelli, I monumentidel Friuli del secolo IV all'
I, p. 9, fig. 5 (StadtbibliothekXI.
51. Die Elfenbeinskulpturen,
MS 3x, fols. 5, 6, 9, x7, etc.). On this manuscript see also W.
Neuss,Die Apokalypsedes HI. Johannesin der altspanischen
Miinster(Westphalia),
Bibel-Illustration,
und altchristlichen
I, Cividale, Milan and Rome, 1943, A
pl.
.
59. Koehler, op.cit., fig. 8.
60. ART BULLETIN, XXXIV, 1952, p. i6o.
6x. See Gabrielli, op.cit. (note 33 above), pp. 303, 304,
pls. CLVIII,CLIx. The author dates the manuscript in the last
1931 , I, 248ff.
52. See Neuss, op.cit., pp. 248, 249.
quarterof the eighth century,but the script points to the ninth.
62. This detail does not appearon ecclesiasticalor imperial
dress in the painting of a council under Theodosius I (362)
54. Ibid., I, pl. x (the angel of the Annunciation).
55. For the most recent views, see W. F. Volbach, "Ivoires in Paris, Bibl. Nat. MS gr 51o (Omont, Manuscrits grecs
mosans du Haut Moyen Age originaires de la region de la de la Bibliothdquenationale, pl. L).
Meuse," in L'Art mnosan(Journkesd'etudes, Paris, February,
63. Gabrielli,loc.cit., pl. cLxIII (Council of Constantinople,
53. Reproducedby Goldschmidt,loc.cit.
1952),
43-46.
56.
Recueil prepare par P. Francastel, Paris, 1953, PP.
ART BULLETIN,
XXXIV,
1952,
pp. I53,
154.
381).
64. Ibid., pl. cLxII.
NOTES
299
(Figs. 1-3). This is probably the chief reason why
the pane has been overlooked for nearly a hundred
years. The window in questionis the first from the left
of the three windows in the middle choir chapel and
the pane is the first on the.left in the second row from
the bottom. Another reason why it has not been
noticed is the difficulty of gaining access to it from
the outside. If one obtainspermissionto enter the little
garden, however, one can clearly see from the patina
which of the panes are original. But it is also possible
to see from the inside, in the ambulatory,that the pane
and the tendril band to the left of it are old, since they
After this paper was written, I observedwhat appearsto be are darker in color than the rest of the
glass (Fig. 2).
the motif of the transverseclavus on a work of the sixth cen- No one will
the fact that these portions are
question
It
is
the
set
of
of
the
that
originally
fragments
ivory panel
tury.
formed the counterpieceof the leaf of the Murano diptych ancient. The theory that while the glass is old the denow in Ravenna. (See W. F. Volbach, Elfenbeinarbeitender sign is new, and that the whole pattern is rearrange,
is untenable because the faces of the two women and
Spdtantikeand des frfiken Mittelalters, 2nd ed., Mainz, 1952,
p. 65, nos. 127-129, pl. 39, 4o, 45.) The thigh band is most the lines on their garments are different and very
evident and most like the form in Castelseprioon the figure much better in
quality than those elsewhere in the
of Saint Anne (in the fragmentin the Hermitagein Leningradsame
window.
Volbach, no. 129, pl. 40). But this detail belongs to the painted
The design represents two women without halos,
ornamentationof the ivory panel together with the gold stars
on the background.Although the painting is old, there re- both facing right and each with her right hand pointing
mains a doubt concerning its age; we are not sure that it is
(Fig. I). A little above knee height are letters
of the sametime as the carving.The motif of the applique upward
of
which
a capitalS is legible on the left and a capitalC
thighbandis not foundin plasticformin otherworksof the
samegroup or schoolof ivory sculpture.Yet on one of the on the right. Above the letters following the S is a confragmentsof the same diptych(formerlyin the Stroganoff traction sign. The first of the small letters is perhapsan
Collectionandnow in privatehandsin Paris) representing
the i, followed by what may be a b cut off by the new
the Trial by Waterand the Journeyto Bethle- lead. This
Annunciation,
portion of the inscription may be read as
hem, of which the last two are uncommonthemes and
"Sibylla."
Following the C on the right is an illegible
with some remarkable
similarities
occur also in Castelseprio
in the Journey(Volbach,no. 128,pl. 45), there letter, perhaps a v with the remains of a contraction
of conception
as the thigh bandin sign above it. The C may probablybe construed as the
are incisedlinesthat may be interpreted
question,unlesstheyare, as Dr. Volbachsupposes(in a letter beginning of Cumae.'
to the writer), simplyschematicgroovesrepresenting
folds;
The two sibyls are standing against a blue backthe photographsare not clear enough to permita definite
ground in an architectonicframe. A pointed clover-leaf
judgment.
arch rests on slender red shafts and is surmounted
COLUMBIA UNIVERSITY
by a yellow Gothic gable with crockets and (restored)
finial which cuts across a horizontal gallery of six
yellow tracery windows. On either side of this Gothic
UNNOTICED FRAGMENTS OF
frame are green frontal buttresses and two lateral
in profile. Each buttress is made up of four
buttresses
OLD STAINED GLASS IN
segments separated by Gothic cornices. Blind tracery
NOTRE-DAME DE PARIS
appears on the frontal buttresses, and a pattern of
PAUL FRANKL
joints on the lateral ones.
From these two figures the nineteenth century
It may be assumedthat at the time of their construc- painterdeviseda compositionof twelve sibylsand twelve
tion (1296-1330) the east chapels of Notre-Dame in apostles (Fig. 3). The window has four axes of equal
Paris were furnishedthroughoutwith stainedglass win- breadth. In the first and fourth, a and d, he placed six
dows. These were destroyed, perhaps during the En- sibyls and in each of the two middle axes, b and c, six
lightenment or the French Revolution. When Viollet- apostles.The two lower rows contain two figures each,
le-Duc restoredNotre-Dame (i1845-1856) the chapels the two upper rows, one each. This distributionmay
were provided with new stained glass similar in style correspond to some original pattern since there are
to that of the thirteenth century. Glass paintersof this pieces of old glass above and below the section which
period liked to set into their new works bits of old glass is preservedin its entirety.2The old pane is only about
which were left over after a buildinghad been restored. half as wide as the window axis. For the apostles the
Here, however, a whole pane was employed and was painter developed a frame by adding narrow open
used as the basis for the design of the entire window arches with gables. For the sibyls he employed the refrescoes are of the eighth century, it is because their
style fits better between the classical forms of S. Maria
Antiqua and the classicismof the Carolingian schools
than between the Byzantine miniatures of the tenth
century and Ottonian art, where others have placed the
frescoes; and this earlier dating is supportedby a number of details, including the painted uncial inscriptions
which seem so much older than the inscriptionsrecording the ordination of a priest, added in the tenth century, that a palaeographerlike Lowe could date them
in the sixth century.
I. Concerning the sibyl theme, cf. Pauli-Wissowa, RealLexikon der classischen Altertumswissenschaft,2. Serie, II,
Stuttgart, 1923, cols.
2073-2183,
and Wilhelm VSge, Jdrg
Syrlin der litere und seine Bildwerke, Berlin, x95o, pp. I5ff.
2. This observation was made by Louis Grodecki, who
recognized as old three other portions of the window: in the
first and third rows in the secondaxis and a piece of tendril in
the first and second rows in the fourth axis. Elsewhere also
little fragmentsof old glass were used.