Document Sample
  The position of Saint-Etienne as a transitional monument, not only chronologically
but also geographically, is reaffirmed by its sculpture. As with the architecture, the
sculpture exhibits relationships both with the regions to the south and west of Dreux
and with the Ile-de-France. But while the former area seems to have provided the
sources for most of the architectural features, it was mainly Paris and its immediate
area that stood behind the sculptural styles at Saint-Etienne.

The Decorative Motifs
  An early nineteenth-century engraving said to be of the portal in the south wall of
the nave, but which may actually be of the main western portal, demonstrates these
dual affiliations (PI. IVa).77 Apparently without a tympanum or other figural
decoration, the portal was composed of a single round-headed archivolt decorated
with a repeated leaf pattern and supported on two colonnettes with foliate capitals.78
While these elements could easily have come to Dreux from the Paris area, the
continuous chevron label moulding shown decorating the inner edge of the archivolt
and the door jambs certainly did not.79 It was derived instead from the regions to the
south and west, where such single strands of small-scale chevron, sometimes called
rubans angevins,80 were extremely popular in the second quarter of the twelfth
century. To mention only those churches otherwise related to Saint-Etienne, the motif
can be found inside both towers of the west facade of Chartres Cathedral (post-1134),
in the ambulatory and hemicycle of Saint-Lomer at Blois (from 1138), in the aisles
and nave arcade of La Madeleine at Chateaudun (begun in the 1130s), in the choir of
Avenieres (c. 1140), in the nave of Le Mans Cathedral (1137-58) and in the nave of
Notre-Dame-du-Pre in Le Mans (1140s). But this is virtually the only sculptural
motif at Saint-Etienne that was inspired by those regions. Otherwise it was the
decorative trends of the Ile-de-France that predominated.
  The capitals of this portal, for example, are clearly Parisian. As seen in the
engraving (PI. IVa) as well as in the capitals themselves which still survive (Pl. V),
they are composed of long and flaring channelled fronds that enclose smaller rigid
palmettes and terminate in volutes or in leafy sprays. This rather flat and stylised type
of capital seems to have originated in the work of the second master's shop at Saint-
Denis sometime in the second half of the 1130s. It appears in the eastern bay of the
ground-storey narthex (PI. VIA) and in the western bay of the southern narthex aisle
(PI. Vc), the latter example particularly resembling the Dreux capitals.81 The type
recurs throughout the crypt and the choir at the east end of Saint-Denis (completed by
1144), sometimes in more elaborate compositions but always composed of the same
elements.82 It was then adopted in various churches whose construction began in the
1140s, including the ambulatory of Sens,83 the
                                          SAINT-ETIENNE IN DREUX                   103

nave of Notre-Dame in Poissy (PI. VIb), the bays preceding the southern choir
chapels of Notre-Dame in Etampes (PI. VIc), and the nave of Saint-Nicolas in
Meulan (PI. VIIa). What seem to be the latest examples of the type occur in the
ground level of the choir at Senlis (early 1150s).84
  The engraving (PI. IVa) also illustrates another, now-lost capital that was not part
of the portal but, as the inscription states, was Tun des chapiteaux du choeur
(colonne engagee).'85 It too would seem to derive from the milieu of Paris and to be
related especially to capitals at Saint-Denis. The fact that it has a double set of
volutes on its main face ties it to the capitals of Suger's church, where this feature
occurs several times in the work of the second shop, although never in association
with the foliage patterns seen on the Dreux example.86 Double sets of volutes also
appear in the hemicycles of Poissy (where the two central ones intersect) and Saint-
Germain-des-Pres in Paris (begun c.1145).87
  This capital, like those of the south portal at Dreux (PIs IVa, and V), uses flat and
rather stiff palmettes. But in this case they are surrounded not by stylised flaring
fronds but by softly drooping acanthus leaves. The softness of these leaf tips and the
attenuation of their stems is also a feature found at Saint-Denis, especially in the
capitals of the central and northern portals of the facade (PI. VIla), in the crypt and
on the northern side of the choir.
  In sum, the foliate motifs at Saint-Etienne reveal a close association between the
carvers working in Dreux and those active in Paris and its immediate vicinity during
the 1130s and 1140s. The parallels with Saint-Denis, where the work is firmly
datable, are especially important in this regard. Based on the chronological scheme
established above for the simple capital type seen in the Dreux portal (Pls IVa and
V), the execution of the Saint-Etienne examples can be supposed not to antedate the
mid-1130s, when the type seems to have appeared first at Saint-Denis, nor to
postdate the mid-1150s, when it seems to have appeared last at Senlis. It is of further
interest that almost all the churches showing the closest relationships with the motifs
at Dreux were, like Saint-Etienne, centres that enjoyed particular royal favour
(Saint-Denis, Poissy, Etampes and Senlis), a point that will be considered further in
the final section of this study.

The Figural Sculpture
  Four rather sizeable historiated capitals from Saint-Etienne survive today in
Dreux, three in the Musee d'art et d'histoire and another in the church of Saint-
Pierre.88 They are of roughly the same dimensions and all were carved to sit atop
compound piers. This, along with their subject matter, the Nativity, the Adoration of
the Magi, the Entombment and the Three Maries at the Tomb, leads one to imagine
that they formed a Christological ensemble which could most appropriately have
been arranged on the inner faces of the strong piers delimiting the easternmost
double bay of the central vessel.89
  Three hands can be discerned at work on these capitals, although their individual
stylistic tendencies should ultimately be seen simply as variant expressions of the
same overall workshop style. The least inventive of the three carved the capital with
the Three Maries,90 while a second, more dynamic master was responsible for those
depicting the Nativity and the Entombment.91 The work of the third sculptor,
represented by the

104               SAINT-ETIENNE IN DREUX

Adoration capital (Pl.VIIla), stands apart from that of the others by virtue of its high
quality and sheer finesse, although it bears resemblances to the style of the second
master. This artist was clearly the most 'advanced' of the three and it is his work that
provides the basis for comparison with sculpture at other sites.
   The third master's figure style is seen best in the Virgin of the Adoration (Pls
VIIIa and IXc) and the seated Herod from the same capital, now detached and
preserved in the Yale University Art Gallery (PI. VIIIb).92 The bodies are stout and
rounded and project vigorously from the back plane of the capital. The heads are
proportionately large in relation to the bodies and are detailed with bold, assertive
features. The eyes are especially arresting because of the protruding ocular orbs and
deeply-drilled pupils. The hairline is quite low on the forehead and individual
strands of hair are delineated by parallel striations. Strong geometric forms
predominate in the faces: the sharp ellipses of the lids and brows, and the faceted
incisions that define the mouths, accentuated in the face of Herod by a second,
vertical incision in the upper lip. Such forms, along with the overall planar treatment
of the surface, are responsible for the bold, expressive impression of these faces.
   The bodies, on the other hand, are less clearly defined, owing to their hunched
poses and the heavy, enveloping nature of the drapery. The anatomy is suggested
only generally and the drapery folds are formed in traditional ways, using either
simple parallel incisions, as in the flat concentric circles on the breasts, or shingle-
like plates, as on the legs. The hems are carved with a broad central band
(ornamented in the drapery of the Virgin), flanked by narrow, raised borders.
   These stylistic characteristics quite correctly suggested to Cahn a relationship with
the sculpture of the south nave portal at La Madeleine at Chateaudun, a building
whose construction was probably begun at the western end shortly after the
community there was reformed in 1131.93 The western portion of the south nave
aisle appears to be the earliest part of the structure, to judge from the style of the
capitals in the dado arcade;
thus, a dating of c. 1140 would seem to be in order for the sculptural ensemble set in
the outer south aisle wall.94 The figure of an abbot at the bottom of the inner left
archivolt (PI. VIIIc) merits comparison with the Herod from Dreux (PI. VIlla). The
proportional relation of head and body is the same in both, as is the hunched, self-
enclosed pose. Similar, too, are the concentric incisions for the drapery folds, the
treatment of the hems and the attenuated, bony fingers. But the closest relationship
here is in the facial style, which in both figures is characterised by the squarish,
planar treatment of the overall mass and the accentuated, geometric framing of the
brow. The eyes of the Chateaudun abbot display the same pierced pupils and sharply
incised lids seen in the Herod, and the simple gash-like mouth has the same
vertically indented upper lip. These similarities point to what Cahn has called an
'unmistakable bond of kinship' between the sculptures from Dreux and those from
Chateaudun.95 I do not propose that this is the result of the same artists having
worked at both sites (the abbot at La Madeleine is less refined than the Herod from
Saint-Etienne) but rather that these are the products of two workshops whose
training and sculptural attitudes were closely aligned.
   What is the origin of this style? Baratte-Bevillard has found similarities between
the decorative and iconographic conventions of Chateaudun and churches further
west.96 But the distinctive style of the figures there and at Dreux has no real parallels
either in the

immediate region of Dreux and Chateaudun or in those areas to the south and west
from which so many of the architectural elements of Saint-Etienne seem to have
been derived.97 There are, on the other hand, convincing parallels to be drawn
between the Dreux figures and those produced for two of the most important
Parisian monasteries of the period.
  The first of these is the destroyed church of Sainte-Genevieve, where the evidence
suggests that the choir was remodelled during the course of the 1130s.98 One capital
from the abbey, a recarved piece of marble first used in the Merovingian basilica on
the site and representing Daniel in the Lions' Den (Pls IXa and b), is now preserved
in the Louvre.99 Although stylistically distinct from the other capitals made for the
choir, it must date from the 1130s and not the late eleventh century, as has been
imagined, because of its total lack of resemblance to capitals known to have come
from the nave of c. 1100, and also because of its general similarities with other
figural sculpture of the fourth decade of the twelfth century.100 Although it is likely
to have decorated the choir or the crypt, its exact place in the destroyed church
remains unknown.
  The figure of Daniel is related in several ways to the Herod from Dreux: the
hunched pose, the long, expressive fingers, the broad, doubly-incised hems, and the
concentric striations on the drapery of the breast. But the beardless, youthful face is
more appropriately to be compared with that of the Virgin from Saint-Etienne (PI.
IXc). They are close to one another in the overall roundness of the heads and the
pervasive softness of the forms, and more specific similarities may be found in the
drilled eyes, flaring nostrils and gentle, slightly smiling mouths accented by fleshy,
downturned wrinkles at the corners.
  Even closer to the Dreux Virgin is the surviving head of a queen from the west
facade of Saint-Denis (Pl.VIIId), identified by Pressouyre as belonging originally to
the female column statue in the right embrasure of the central portal.101
Significantly, this head shares features with both the Virgin from Saint-Etienne and
the Daniel from Sainte-Genevieve.102 The drilled pupils, the sharply geometric lid
formations, and the pronounced and rigid brows all link this head with that of the
Dreux Virgin, while the subtle expression of the mouth, the fleshy corner wrinkles,
the indented upper lip and the puffy creases extending obliquely from the nose tie it
with the Daniel.103 The head from Saint-Denis and the capital from Sainte-
Genevieve are of extraordinary quality, finer than even the best of the work at
Dreux, but, despite their similarities, the two Parisian pieces still cannot be ascribed
to the same artist. Thus we are dealing here with at least four workshops (at Saint-
Denis, Sainte-Genevieve, Saint-Etienne and La Madeleine) that adhered to the same
stylistic precepts, although each varied the basic formulas in its own way.
  The exact relationships among these workshops, however, and the evolution of the
style they shared, is still unclear. Did the style originate in the area of Chateaudun
and Dreux sometime in the early 1130s and then reach the capital some years later?
Or was it developed first in Paris and then disseminated outward? The lack of
specific dates for Saint-Etienne, Sainte-Genevieve and La Madeleine, taken along
with our ignorance of where in the first two monuments the relevant pieces were
located, makes firm conclusions impossible. Only the head from Saint-Denis, which
can be firmly dated to the years immediately preceding 1140, provides a
chronological point de repere for the group. But even this can do no more than
reinforce the supposition that all four churches were in the course of construction
during the 1130s and 1140s.
106               SAINT-ETIENNE IN DREUX

  If forced to speculate, I would prefer a Parisian origin for the figural style at
Dreux. For one thing, the fact that Paris was so obviously the source of the foliate
sculpture makes it tempting to imagine that the figural style originated there as well.
For another, Paris was the focus of the greater part of the king's attention in matters
of architectural patronage. As at Dreux, he was the direct patron of Saint-Victor and
Saint-Pierre-de-Montmartre in Paris, both of which were being built in the 1130s,104
and he was probably involved in one way or another in the contemporary
reconstructions of several other royal houses in the city (e.g., Saint-Denis, Sainte-
Genevieve and Saint-Martin-des-Champs). The evidence of style suggests that the
workshops active at these Parisian churches were in close contact with each other,
and it seems likely that some capital-carvers worked on more than one project.105
This raises the possibility of a royal workshop of designers and craftsmen in Paris
during the period. And it may well be that it was from this milieu that a team of
sculptors was dispatched to work on the capitals of the king's collegiate church in

   The place of Saint-Etienne in the architectural and sculptural movements of the
period is thus somewhat complex. The decade of the 113os was a time of enormous
experiment in which ideas (and probably also individuals) were brought to Paris and
its region from many other areas, the Oise Valley and the Beauvaisis, Normandy and
the west-central Loire Valley. Paris was the crucible in which these ideas were
transformed, producing in the early 1140s at Saint-Denis the Gothic revolution. But
it was by no means just to Saint-Denis that these regional ideas flowed, nor was it
just from Saint-Denis that the new formulations were disseminated.
   There was in fact a great deal of diversity and experiment in the architecture and
sculpture of the Paris area in the 1130s,107 and this did not end abruptly after the
dramatic appearance of Suger's choir. Instead, the differing stylistic strains of the
1130s continued to have meaning for builders, sculptors and patrons through the
1140s, although sometimes in altered forms. In order to understand the complexity
that surrounds the birth of Gothic, one must seek to separate the various component
strains and their sources, and then to trace the interactions between them.
   Saint-Etienne in Dreux is particularly important in this respect since it seems to
have been one of the principal transmitters of the architectural traditions of the
regions south-west of Paris, acting as a filter between the area of west-central France
and the Ile-de-France. What is striking, however, is the almost total lack of influence
of Saint-Etienne and its source-buildings on Saint-Denis. Its impact is found instead
at other monuments, ones which may themselves reflect the influence of Suger's
choir, but which in many ways stand apart from it and express an alternative idiom
of Early Gothic architecture.
   The cathedral of Sens (Fig. 10) is a good example. It was of course indebted to
Saint-Denis as seen, for instance, in some of its capital and moulding types, in its
use of shafts en delit and broken ribs, and perhaps also in its three-storey elevation.
But in other ways it seems to have been more influenced by Saint-Etienne in Dreux
and the west-central French tradition from which that building was descended, as in
the alternating system, the double bays, and the replacement of the transepts with
lateral chapels, features that were virtually unknown in the Ile-de-France before
Saint-Etienne.108 Yet the destruction

                                              SAINT-ETIENNE IN DREUX                107

of Saint-Etienne makes it impossible to determine its precise relationship to Sens,
that is, whether it exerted a direct influence or was merely a slightly earlier example
of an Ile-de-France church whose sources were partially drawn from the regions to
the west of the royal domain. Nor can it be known whether Saint-Etienne included
other features seen at Sens that seem also to have been derived from that area,
elements such as the dado arcade109 and the chamfered dosseret flanked by two
small engaged shafts.110
  Notre-Dame in Poissy (Fig. 9), a building that has frequently been seen as related
to Sens,111 must also be considered in the present context. It, too, borrowed features
from Saint-Denis, but deviated from that model in many of the same ways seen at
Sens. And in some of these deviations it seems particularly close to Saint-Etienne at
Dreux, as in the very similar arrangement of the lateral chapels.112 The date of
Poissy is not secure, but the parallels with Saint-Denis and Sens would suggest that
construction was begun in the mid-1140s and, therefore, sometime after the
inception of work at Saint-Etienne. Poissy was also a royal abbey and its abbot was
Louis VII's brother.113 It thus may be that its relationship with Saint-Etienne
stemmed from their common source of patronage and that members of the supposed
workshop of royal designers were responsible for the construction of both.
  It can also be argued that elements from the Saint-Etienne disposition reached
Paris before construction began in the choir of Saint-Denis, as suggested by the
appearance of the alternating system at Saint-Magloire before 1138 (Fig. 5). But
since Saint-Magloire cannot be shown to have adopted any other ideas from Saint-
Etienne, it may be that Dreux itself did not introduce the idea of the alternating
system there, but rather that it was in a more general way the west-central French
traditions from which the Dreux design was drawn that were acting on Parisian
  The larger issue raised here is the nature and importance of what should be seen as
the west-central French component in the formulation of Early Gothic architecture,
that is, the impact of the more conservative buildings of the Maine, the Blesois, the
Beauce and the west-central Loire Valley that have been discussed throughout this
study. They had virtually no impact on Saint-Denis itself, as has been seen, but their
influence was strong on a string of monuments located in the southern lie-de-France
and extending from Etampes in the west, through La Ferte-Alais and Chateau-
Landon, and ending with Sens in the east. The west-central French tradition was
clearly a guiding influence at the last of these and then, probably following directly
from Sens itself, influenced Senlis and other buildings, where the older formulations
were recombined with the more Dionysian interpretation of Early Gothic
  In the final analysis, Saint-Etienne in Dreux emerges as a truly 'transitional'
building, bridging stylistic epochs as well as regional traditions. As such, it should
be seen as one of the seminal monuments in the transition from Romanesque to
Gothic architecture in Northern France.
I am deeply grateful for the suggestions offered during the preparation of this study
by Jean Bony, Walter Cahn and Lindy Grant. I am also indebted to Jean Lelievre,
director of the Musee d'Art et d'Histoire de la ville de Dreux, for his assistance and
advice on a number of points.
108                 SAINT-ETIENNE IN DREUX

    This study forms part of a wider programme of research on the architecture of the Ile-de-France
during the 1l30s and 1140s, and several of the points raised here will be discussed in more depth in
my forthcoming book on the subject. The main lines of the present argument were set forth in
lectures sponsored in the autumn of 1983 by the British Archaeological Association, the Courtauld
Institute of Art and the University of Pittsburgh, and in the spring of 1984 by the Delaware Valley
Medieval Association.
    W. Cahn, 'A King from Dreux', Yale University Art Gallery Bulletin, XXXIV (1974), 14—29.
    See further E. de Rotrou, Dreux. Ses antiquites. Chapelle Saint-Louis (Dreux 1864), 15-21.
    On Robert de France, the fourth son of Louis VI, see further A. Du Chesne, Histoire
geneologique de la maison royale de Dreux . . . (Paris 1631), 13—23.
    Three other views of the site are less useful for the purposes of reconstruction. One is the partial
view of the chevet made in 1779 by L. Parizeau (Paris, Bibliotheque Nationale, Cabinet des
Estampes, Destailleur Collection, Va 26k). Two others are to be found in the Musee d'Art et
d'Histoire de la ville de Dreux: a drawing by Constant Bourgeois of the castle under demolition (c.
1819); and a partial view of the chevet and castle made by du Clos and Auvry before 1783.1 am
indebted to J. Lelievre, director of the Dreux museum, for bringing these drawings to my attention.
On the castle at Dreux, see further D. Philippe-Lemaitre, Histoire de la ville et du chateau de Dreux
(Dreux 1850); J. Lelievre, Dreux. La chapelle Royale (Paris 1969);
A. Chatelain, Chateau forts et feodalite en lle-de-France du Xleme au XIIIeme siecle (Nonette 1983),
    The full authenticity of a charter by King Lothar II (954—68), which provides the first reference
to the church, has been questioned, although its substance may be essentially correct. See further
Cahn, 'A King from Dreux', above, n.2, 20; and L. Halphen and F. Lot, Recueil des actes de Lothair
et de Louis V (Paris 1908), l6l-2. The manuscript history by Dorat de Chameulles, Les Antiquitez ou
Histoire de la Ville de Dreux (Paris, Bibliotheque de l'Arsenale, MS 4086, f. 187), ascribes the
foundation to Louis V (968—87). Nothing is known of the form of the earlier church(es) on the site
apart from the testimony of A.-L. Marquis, 'Notice sur quelques antiquites observees a Dreux',
Memoires de la Societe des Antiquaires de Normandie, i (1824), 58-68 (esp. 6o), who reported that
excavations in 1820 'ont revele l'existence, sur le meme emplacement, d'une eglise bien anterieure a
celle que la revolution a vue tomber. A 22 pieds au-dessus du niveau de l'edifice actuel ont ete
reconnus les fondements d'une autre eglise beaucoup plus petite, et qui n'offrait que la forme d'un
quarre long'. Marquis provided no further details about this earlier structure, nor did he speculate as
to its date. No excavations have been carried out on the site since those of the early nineteenth
      Examples are Saint-Mellon in Pontoise, Saint-Guenaud in Corbeil and Notre-Dame in Mantes.
These are to be distinguished from mere castle-chapels, such as those within the royal castles at
Etampes, Senlis, La Ferte-Alais and Paris, since the former were fully-functioning abbeys or
collegiate churches.
      On the enlargement of 1108, see Rotrou, Dreux, 19. On the donation of 1120-9, A. Luchaire,
Louis VI le Gros. Annales de sa vie et de son regne 1081-1137 (Paris 1890), 210-11, no. 451. This
document is datable on the basis of Prince Philippe's co-signature, something that could not have
occurred before his association to the throne in 1119/20 and which would have appeared in a
different form after his coronation in 1129. Philippe died in 1131 without having succeeded his
   - See above n.3. This book is an expanded version of the author's manuscript history, entitled
Abrege historique de la ville et du comte de Dreux, written in 1707. The original, which was formerly
conserved in the municipal library of Chartres as MS 1541, was apparently destroyed in the fire of
1944 (Cahn, 'A King', 21).
      Rotrou, Dreux, 19-20. Philippe-Lemaitre, Histoire, 186, erroneously mentions Saint-
 Victor in Marseille. u Luchaire, Annales, 218-9, no. 471.
      Luchaire, Annales, 218-9, no. 471.
      Dorat de Chameulles, Les Antiquitez, f. 189.
      This document, which was overlooked by A. Luchaire, Etude sur les actes de Louis VII (Paris
1885), was discovered by Cahn ('A King', 20—21) in a fifteenth-century copy of the cartulary of
Saint-Etienne now preserved in Paris, Bibliotheque Nationale, MS lat. 10106, f. i. The king styles
himself dux aquitanorum in the text, a title he gained through his marriage to Eleanor in 1137 but lost
in their divorce of 1152.
      Translation from Cahn, 'A King', 21. The original plaque and the engraved copy reproduced here
(PI. IVa) are both preserved in the Musee d'Art et d'Histoire in Dreux.
      Dorat de Chameulles, Les antiquitez, f. 189.
      The plan reproduced here (fig. 2) is the engraved copy published in 1850 by Philippe-Lemaitre,
Histoire, facing 205. The eighteenth-century original from which it was derived is now preserved in
the Musee d'Art et d'Histoire in Dreux. A second engraved plan, also in the Dreux museum, seems to
be based on the same original and differs from the Philippe-Lemaitre version only in giving the
chapel dedication and in showing the cruciform chapel of Saint-Louis which was built over part of
the site of Saint-Etienne beginning in 1820.
      It is not clear whether the chamber shown to the west of the northern lateral chapel was an
original feature or a later addition. It is identified as the Chapitre on the engraved plan in the Dreux
      On Saint-Denis, see further S. McK. Crosby, L'abbaye royale de Saint-Denis (Paris 1953); and
the same author's forthcoming volume on the church of Abbot Suger.
      This is related in appearance, if not in structure, to the ground-level thick-wall passages in the
main apses of La Trinite in Caen and Notre-Dame in Soissons (the latter now destroyed), although
the passage at Morienval is not taken in the thickness of the wall. All three buildings date from the
1120s and 1130s. See also J. Bony, French Gothic Architecture of the 12th and 13th Centuries
(Berkeley 1983), 26, where the link between the three apses is also acknowledged. On La Trinite, see
further M. Bayle, La Trinite de Caen. Sa place dans I'histoire de {'architecture et du decor romans
(Geneva 1979); on Notre-Dame, C. Barnes, 'The Documentation for Notre-Dame de Soissons', Gesta,
XV (1976), 61-70.
       On Saint-Martin-des-Champs, see further E. Lefevre-Pontalis, 'Eglise de Saint-Martin-des-
Champs a Paris', Congres archeologique, LXXXIII (1919), 106-26.
                                                             SAINT-ETIENNE IN DREUX                   109

      Gallia Christiana, VII, 306-9. The schematic plan published here (fig. 5) is based on various
ground plans made before the building was demolished in the late eighteenth and early nineteenth
centuries. These, along with several pre-destruction views and descriptions will be discussed in detail
in my article, 'The Abbey Church of Saint-Magloire in Paris' (forthcoming).
      The Temple church in Paris, which was begun in the early-to-mid-1130s and was almost certainly
finished by 1147, stands somewhat apart from this group since what might be taken as an ambulatory
there is really only an annular, vaulted aisle. See further H. de Curzon, La maison du Temple de
Paris. Histoire et description (Paris 1888), 71—83.
      Examples are Saint-Maclou in Pontoise, Saint-Germain-des-Pres, Senlis and Noyon.
      The depth of the chapels is paralleled during the period in Paris only by those of Sainte-
Genevieve (now destroyed) (fig. 8), which belong to a phase of construction datable c. 1130-40; see
below, n.98. But there was clearly no ambulatory associated with these chapels on the main level of
the Sainte-Genevieve choir; and if there was one in the crypt, it was completely rebuilt with new
vaults and supports in the seventeenth century, as I will discuss in a forthcoming study.
       The examples mentioned above in n. 23, along with Saint-Martin-des-Champs and Saint-
Magloire, were preceded in date by the tangential chapels of La Trinite in Fecamp and Avranches
Cathedral. And a contemporary example is found in the church at Avenieres (Laval) in the Maine
(fig. 16). The relationship of these to the lle-de-France group has been discussed by Bony, French
Gothic, 52; and the same author will explore the question further in his article, 'What Sources for the
Choir of Saint-Denis?', soon to be published in (ed. P. Gerson), Abbot Suger and Saint-Denis.
      See further H. Hilberry, 'The Cathedral at Chartres in 1030', Speculum, XXXIV (1959), 561-72.
      Regional examples are even more numerous of three radiating chapels that are not as deep as
those at Chartres and Saint-Etienne.
      See further F. Lesueur, 'L'eglise abbatiale Saint-Lomer de BIois', Bulletin monumental, LXXXII
(1923), 37-65; and idem, Les eglises de Loir-et-Cher (Paris 1969), 55—61 (with further
       See most recently L. Grant, 'Aspects of the Twelfth-Century Design of La Madeleine at
Chateaudun', Journal of the British Archaeological Association, CXXXV (1982), 23—34, where the
chapels, however, are not discussed. An opening into the south-eastern of what must have been three
radiating chapels is still visible in the remaining south aisle of the crypt.
      P. Heliot and G. Jouven, 'L'eglise Saint-Pierre de Chartres et l'architecture du moyen-age',
Bulletin archeologique, n.s. VI (1970), 117-77 (esp. 172-3).
      On these see further Heliot and Jouven, 'Saint-Pierre de Chartres', 174; and A. Prache, L'lle de
France romane (La Pierre-qui-vire 1983).
      Hilberry, 'Chartres in 1030', 566 and fig. 6b, Bony, French Gothic, 65—6, has recently come to a
similar conclusion regarding the importance of Fulbert's Chartres in this transmission. Auxerre
Cathedral and Vignory, both rebuilt in the second quarter of the eleventh century, also had lateral
towers flanking the choir, as did other buildings later; see further Heliot and Jouven, 'Saint-Pierre de
Chartres', 172-3. Only the southern chapel at Dreux supported a tower, a disposition not unlike that
found commonly (either on the north side or the south) in Paris during the 1130s and 1140s, as at
Saint-Pierre-de-Montmartre, Saint-Martin-des-Champs, Sainte-Genevieve, Saint-Magloire, Marolles-
en-Brie and La Ferte-Alais.
      Heliot and Jouven, 'Saint-Pierre de Chartres', 174. See further C. Metais 'Saint-Denis de Nogent-
 le-Rotrou (1030-1789)', Archives du diocese de Chartres, I (1895), 1—345, where the construction is
 inexplicably dated to c. 1040—80.
  "Heliot and Jouven, 'Saint-Pierre de Chartres', 174. See further A. de Dion, 'L'eglise du prieure
 Saint-Thomas d'Epernon', Memoires de la Societe archeologique de Rambouillet, xi (1899), 551-68.
 The lateral chapel on the north side seems originally to have had an apsidal termination which was
 squared off only in the fourteenth or fifteenth century (ibid., 567). I would prefer to date the choir c.
 1100—10, as opposed to either shortly after 1050, as proposed by de Dion, or c. 1125, as proposed
 by Bony, French Gothic, 65-6.
      Heliot and Jouven, 'Saint-Pierre de Chartres', 174. The disposition at Avranches is known from a
 ground plan drawn in the 1780s by Lefebvre and published by E.-A. Pigeon, Le diocese d'Avranches
 . . . (Coutances 1888), II, 679. But the two-bay chapels flanking the nave may well have been later
 additions, to judge from a drawing of the south flank made in 1649. For the drawing and the recent
excavations, see D. Levalet, 'La cathedrale Saint-Andre et les origines chretiennes d'Avranches',
Archeologie medievale, XII (1982), 107—53 and pl. IV, See also R. Liess, Der fruh romanische
Kirchenbaudes 11. Jahrhunderts in der Normandie (Munich 1967), 139 and 296-7.
      See further F. Salet, 'Notre-Dame de Cunault, les campagnes de construction', Congres
archeologique, CXXVII (1964), 636-76.
      Heliot and Jouven, 'Saint-Pierre de Chartres', esp. 134-48 and 170-5. Parts of the choir, if not all
of it, were probably complete by 1165, when a translation of relics occurred. Heliot (ibid., 120)
preferred to see construction beginning under Abbot Foucher, elected in 1150/1, but the details of the
structure and decoration of the choir could just as easily be products of the 1140s.
      This chapel, which may have been constructed in the eleventh century, is represented as still
standing in the mid-seventeenth-century plan reproduced by Heliot and Jouven 'Saint-Pierre de
Chartres', 118, Fig. i.
      On this plan type and the probable anteriority of Marmoutier in its evolution, see further C.
Leiong, 'Apercus complementaires sur le plan de l'eglise abbatiale de Marmoutier au XIe siecle',
Bulletin mommental, CXXXVII (1979), 241—7. In England the plan appeared at Saint Anselm's
Canterbury Cathedral c. 1100 and at the Cluniac priory of Lewes.
      On the Romanesque relatives of the plan of Saint-Lomer, see Lesueur, 'L'eglise abbatiale', 38. On
Fontgombault, which was dedicated in 1141, see L. Demenais, 'L'eglise abbatiale de Fontgombault
(Indre)', Bulletin monumental, LXXX (1921), 91-117. A slightly different version of this chevet plan
occurs at La Charite-sur-Loire.
      Bony has also discussed the amplified spaciousness created by this plan type (French Gothic, 56-
 7) and has suggested that the source for Fontgombault may have been Saint-Benoit-sur-Loire.
      See further J. Henriet, 'La cathedrale Saint-Etienne de Sens: Le parti du premier maitre et les
 campagnes du XIIe siecle', Bulletin monumental, CXL (1982), 81-168 (esp. 83—8 for a discussion of
 the various documents of the seventeenth and


eighteenth centuries that assign the beginning of construction to the year 1140). There is nothing in
the fabric of the cathedral that supports the theory put forth by F. Salet ('La cathedrale de Sens et sa
place dans 1'histoire de l'architecture medievale', Comptes-rendus de I'Academie des inscriptions et
belles-lettres (1955), 182-7) that work began between 1124 and 1128. The date of c. 1140 was also
accepted by K. Severens, 'The Early Campaign at Sens, 1140-1145', Journal of the Society of
Architectural Historians, XXIX (1970), 97—107.
    On Poissy, see further F. Salet, Notre-Dame de Poissy (Paris, 1951), although his dating of the
inception of construction to c. 1140 (based on his chronology for Sens, discussed in the previous
note) is probably a little too early. Debate raged during the late nineteenth and early twentieth century
over the dating of Poissy and its role in the transition from Romanesque to Gothic architecture. One
camp, led by A. Saint-Paul, went so far as to assign the beginning of work to c. 1125, thus
positioning it before the choir at Saint-Denis (see further A. Saint-Paul, 'Poissy et Morienval',
Memoires de la Societe historique et archeologique de I'arrondissement de Pontoise et du Vexin, XVI
(1894), 1—21; esp. 13). But several features of the design, especially the form of some of the foliate
capitals, indicate a direct dependence on Saint-Denis. Most of the chevet at Poissy is now the product
of two nineteenth-century restorations, although the general disposition (with the questionable
exception of the axial chapel) reflects the original arrangement.
    The work at Senlis is documented as having been begun during the short episcopacy of Bishop
Thibaut, variously dated as 1150-5 and 1151-6. For a survey of these documents, see D. Brouillette,
The Early Gothic Sculpture of Senlis Cathedral, Ph.D. dissertation, University of California at
Berkeley, 1981, 15 and 35, n. 54.
    Double forebays also precede the lateral chapels at Cunault, Epernon (Fig. 14), Avranches and the
western set of chapels at Saint-Pere. The forebays at Sens, Senlis and Poissy have all been either
modified or rebuilt since the twelfth century, although their original arrangements are still retrievable.
    The monolithic columns of the choir of Saint-Denis were without immediate precedent and did
not make an impact on the Paris region, except for the hemicycle of Senlis and, somewhat farther
afield, the hemicycle of Noyon. Most of Saint-Denis's followers, such as Saint-Germain-des-Pres,
Poissy, Pontoise and Domont, preferred the more standard coursed columnar pier, which had already
been used in Paris c. 1100 in the nave of Sainte-Genevieve. Coursed columnar piers were also the
norm in the Maine, the Anjou, the Blesois and the territory of Chartres, as at Saint-Lomer at Blois,
the nave of Le Mans Cathedral (pl. X), Notre-Dame at Avenieres, and Saint-Pere in Chartres, to
mention only those under construction between 1130 and 1150.
     The epitaph painted on the pier commemorates Chariot Daunoy, a royal painter who died in 1411.
The legend at the bottom of the drawing erroneously locates the epitaph in the church of 'St. Laurent
de Dreux', a dedication otherwise unknown in the town. On the contrary, the Daunoy epitaph is
independently cited as existing in Saint-Etienne by E. Lefevre, Documents historiques sur la ville et
le comte de Dreux (Chartres 1859), 182.1 am grateful to M.J. Lelievre for his kind assistance on this
     The base profile is perfectly acceptable for the 1130s or 1140s. The only curiosity is the absence
of projecting spurs at the corners of the socle.
     The plan is not provided with a scale but, if the relative proportions of the elements depicted are
near to being correct, one can at least estimate the dimensions of the piers on the basis of the
surviving historiated capitals, which will be discussed below. Because they were meant to sit atop the
shafts of compound piers, their width (0.66m at the top) can be used to approximate the width of the
abaci and socles as 0.70m each. This establishes a rough scaling factor of 1135 for the plan as a
whole and produces weak piers 0.66m (2ft) in circumference and strong piers 1.40m (4ft 7ins) wide.
Using the same scaling factor, one arrives at an east-west length of 31.5m (104ft) from the mouth of
the axial chapel to the western nave wall, a width of 3.15m (11ft) for the ambulatory and aisles, a
north-south width of 6m (20 ft) for the central vessel and an east—west width of 3m (10 ft) for the
main arcade intercolumniation.
     P. Combot, Histoire de la ville de Dreux, limitrophe du pays chartrain, Musee d'Art et d'Histoire
de la ville de Dreux, f. 306v. I am indebted to J. Lelievre for his transcription of the portions of
Combot's text relative to Saint-Etienne. This account is also valuable for information regarding the
revenues of the house in the sixteenth and seventeenth centuries and especially for a list of the relics
then housed there (ff. 308r—v).
     On the Rhenish examples and their followers in Belgium and the Low Countries, see further L.
Grodecki, Au seuil de I'art roman. L 'architecture ottonienne (Paris 1958); and L-F. Genicot, Les
eglises mosanes du XIe siecle (Louvain 1972). English examples of the late eleventh and early
twelfth centuries include Romsey and Ely, as well as Durham and its followers (e.g. Selby, Waltham
and Lindisfarne).
     One might also mention the alternation in the four eastern bays of the nave at Laon, in the nave
aisles of Notre-Dame in Paris, in the western nave bays at Saint-Loup-de-Naud and in the double bay
just west of the chevet towers at Saint-Leu-d'Esserent. At Chartres the alternation consists of
columnar cores flanked by polygonal responds and then polygonal cores flanked by rounded
responds. At Bourges the piers all have the same distribution and shape of shafts, but there is an
alternation in the sizes of the piers. Examples of alternation are found outside the royal domain in
Northern France at Ouistreham in Normandy and at Berteaucourt-les-Dames in the Somme, both
from the second quarter of the twelfth century and both still fully Romanesque in style.
      The alternation in the western bays of the nave of Sainte-Genevieve in Paris (Fig. 11),
convincingly dated to c. 1100 by Vieillard-Troiekouroff, involved only the size of the columnar piers,
but not their form, much like the system seen later at Bourges. See further M. Vieillard-Troiekouroff,
'L'eglise Sainte-Genevieve de Paris du temps d'Abelard', Pierre Abelard, Pierre le Venerable. Les
courants philosophiques, litteraires et artistiques en accident au milieu du XIIe siecle (Paris 1975),
745-61 (esp. pp. 758—9 for the attribution of the nave capitals to the atelier that worked also at Saint-
Thibault-des-Vignes between 1097 and 1106). A longitudinal cross-section showing the alternating
piers of the nave is provided by A. Lenoir, Statistique monumentale de Paris (Paris 1867), Atlas.
     See further Prache, L'lle-de-France romane, 399-401.
     These were prepared by Nicolas Davanne, abbot of the house from 1620. All are reproduced in E.
Bories, Histoire du canton de Meulan (Paris 1906), 103-13, as is an exterior view of the south flank,
also published in the Monasticon Gallicanum. A


second exterior view, illustrating the east end of the church, is provided by A. Millin, Antiquites
nationales (Paris 1792), IV, chap. XLIX, pl. I; I am grateful to Walter Cahn for drawing my attention
to the latter.
    For the documentation, see Bories, Histoire, 105—6. I am indebted to Jean Bony for first calling
Saint-Nicaise to my attention.
     See above n.22. The evidence for alternation at Saint-Magloire is provided by a detailed
architect's plan of the crossing and the eastern bays of the nave made probably in the seventeenth
century (Paris, Archives nationales, S4742, Dossier 4). This plan and others will be discussed in my
article, 'The Abbey Church of Saint-Magloire in Paris' (forthcoming), where the problems concerning
the probable campaigns of construction will also be investigated.
     See above n. 35.
     On Notre-Dame-du-Pre, see further A. Mussat, 'L'eglise Notre-Dame-du-Pre au Mans', Congres
archeologique, cxix (1961), 100-18; and idem, Le style gothique de I'ouest de la France aux XIle et
XIIIe siecles (Paris 1963), 87-8 and 92-3. On La Couture see F. Lesueur, 'L'eglise de la Couture au
Mans', Congres archeologique, CXIX (1961), 119-37; and Mussat, Style gothique, 109-14., On Saint-
Vincent, see R. Triger, 'L'abbaye de Saint-Vincent du Mans', Revue historique et archeologique du
Maine, LXXX (1926), 173-220; and Mussat, Style gothique, 88-9, where it is noted that William of
Saint-Calais, the bishop who began the construction at Durham, where alternation is a major feature,
had been a monk at Saint-Vincent.
 "See further F. Salet, 'La cathedrale du Mans', Congres archeologique, CXIX (1961), 17-58; and
Mussat Style gothique,
     Grant, 'Aspects', 28-31, dated the design slightly later and saw it as a derivative of Le Mans.
     On the alternation at Saint-Pere, see Heliot and Jouven, 'Saint-Pierre de Chartres', 134-5; and
Grant, 'Aspects', 29. (i4 See further L. Lecureux, 'L'eglise d'Avenieres', Bulletin monumental, LXXV
(1911), 102-19; and A. Mussat, 'L'eglise d'Avenieres a Laval', Congres archeologique, CXIC (1961),
376-95. Since the design was altered soon after the outer walls of the ambulatory were erected, the
alternating system cannot with certainty be ascribed to the designer in charge at the outset of
     See above n.41; and Lesueur, 'L'eglise abbatiale', 38—9. At Saint-Lomer a change in design above
the level of the main arcade of the choir, occurring sometime in the mid-to-late 1140s, resulted in the
separation and individual vaulting of the two straight bays. The original scheme most likely called for
their being joined together in a double-bay arrangement such as can be seen in the choir of
Fontgombault (compare the longitudinal sections in Lesueur, 'L'eglise abbatiale', 41, and Demenais,
'Fontgombault', 98).
     The probable relationship between Saint-Etienne and Saint-Magloire in Paris will be discussed in
the final section of this article.
     The only evidence suggesting a usable second level in any part of the building is Combot's
description of 1602 (see above n. 51) which mentions ' 'six chapelles' around the ambulatory and not
three, as shown on the surviving ground plans. This could be interpreted to mean that each of the
three radiating chapels was two storeys in height, but it is equally possible that the three additional
chapelles mentioned by Combot were only altars distributed variously on the ground level of the
chevet without regard to the existence of the architectural chapel units, a preferable alternative since
the exterior drawings show the windows and roofs of the chapels rising no higher than those of the
nave aisles (pls II and IIla).
     On the development and dissemination of the so-called false gallery, see further my article, 'The
Nave Galleries of Durham Cathedral', Art Bulletin, LXIV (1982), 564—79; esp. 570—1 and n.27 (with
further bibliography).
     De Dion, 'Saint-Thomas d'Epernon', 566, described the ambulatory vaulting as he had seen it
before its destruction in the following terms: 'Le rond-point avait cinq arcades, tandis que le mur de
precinction concentrique offrait sept divisions, trois chapelles en abside et quatre trumeaux ajoures
d'une large fenetre. II traca la voute d'arete du milieu, en face de 1'abside centrale, en trapeze regulier;
les deux voutes adjacentes furent posees en biais; en face de la chapelle nord la voute d'arete biaise
encore plus et une portion de voute en berceau est menagee entre la fenetre et le gros pilier. En face
de la chapelle sud il s'y etait pris autrement: la voute etait en berceau suivant la courbe et penetree
d'un cote par le prolongement du cul-du-four de la chapelle en abside, de 1'autre par 1'arcade du
choeur, les deux penetrations ne se recontrant pas'.
     The choir-level hemicycle of Fulbert's church is just as likely to have had four columnar piers as it
is to have had the six proposed by Hilberry, 'Chartres in 1030', fig. 6b.
     On the Temple, see above n. 23. Triangular cells of groin-vaulting can be found even earlier in
England, as in the ambulatories of Gloucester (1090s) and Saint Bartholomew's, Smithfield, in
London (founded 1123), and in the chapter house at Worcester (c. 1110—20). I am indebted to Lindy
Grant for calling the first two examples to my attention.
     On Saint-Spire, see further M. Aubert, V. de Courcel and A. Rhein, 'Corbeil et ses environs. Notes
archeologiques', Bulletin monumental, LXXXV (1926), 361—71 (esp. 361—5).
     See further E. Lefevre-Pontalis, 'Eglise de Saint-Martin', Congres archeologique, LXXXV (1919),
32—40; and J. Henriet, 'Recherches sur les premiers arcs-boutants. Un jalon: Saint-Martin
d'Etampes', Bulletin monumental, CXXXVI (1978), 309—29.
     For the possible dimensions, see above n. 50.
     Grant, 'Aspects', 30.
     See above n. 65.
     The engraving is preserved in the Musee d'Art et d'Histoire at Dreux. On the problem of which
door is represented here, see further Cahn, 'A King', 22 and 29. The main portal of the west facade is
partially visible in Parizeau's drawing of
  1779 (pl. IIIb), and a more complete view of the facade, but with much less specific detail, appears
  in Rotrou's drawing of
  1780 (pl. IIIa). The western portal as it appeared in 1814 is described by M.l'abbe de Freminville,
  'Memoire sur les monumens du moyen age du pays chartrain', Memoires de la Societe royale des
  antiquaries de France, IV (1823), 184—5:
'Cette arcade est soutenue par deux pilastres et deux colonnes engagees; elle est en effet decoree avec
assez de delicatesse; on y voit, avec plusieurs rangs de moulures en zigzag, qui caracterisent
l'architecture du temps, et qui regnent dans tout le pourtour de 1'archivolte, des ornements en
fueillage d'assez bon gout; les chapiteaux des colonnes en sont pareillement enrichis'.
[p. 112]
     The colonnettes appear in the engraving as shafts en delit, a feature more widespread in the lle-de-
France in the 1130s and 1140s than elsewhere. The colonnettes supporting the archivolts of the
western portal, however, were described (see previous note) as being 'colonnes engagees'.
      Chevron moulding occurs contemporaneously in Paris only in the western aisle bays of the
narthex at Saint-Denis and the axial arch of the hemicycle at Saint-Martin-des-Champs. In both cases,
it is the thick, heavy zig-zag common in the Oise Valley (as at Bury, Foulangues and the narthex of
Saint-Leu-d'Esserent), Normandy (e.g. Thaon) and England (e.g. Durham), and not the thinner single
strand seen here. The latter type is extremely rare in the lle-de-France, occurring in only a few
examples south of Paris (e.g. the southern choir extension at Notre-Dame in Etampes and the third
level of the royal palace in Etampes).
     Lesueur, 'La Couture', 131.
      On the use of this type of capital at Saint-Denis and on the distinctions between the two
campaigns, see my article 'Two Campaigns in Suger's Western Block at Saint-Denis', Art Bulletin (in
press), esp. n.30 and Fig. 20b. The general form was anticipated in the work of the first shop there in
the western bay of the narthex, although the capitals in that bay are always more florid, more
complex and more deeply carved. See further W. Wulf, Die Kapitellplastik des Sugerbaus van Saint-
Denis (Europaische Hochschulschriften, XXVIII: Kunstgeschichte, 10) (Frankfurt 1979), figs 14 and
35 (right).
     See further Wulf, Kapitellplastik, figs 13, 30 (right), and 54; and Crosby, L'abbaye royale, pl. 83
     For Sens, see Henriet, 'Cathedrale Saint-Etienne', 127, Fig. 46 (first capital left of the central
respond), and 114, Fig. 30a, for an elaborated version of the type, related to the east end of Saint-
     For Senlis, see Wulf, Kapitellplastik, Fig. 118. W. dark informs me that the same type occurs in
the choir of Bruyeres-les-Laon, another building of the 1150s. And related to its appearance there are
the examples of the later 1140s in the choir of the church at Chelles in the Aisne.
     Its dimensions are given as 54cm high and 66cm wide, the same width as the figural capitals now
preserved in the Dreux museum, a relationship first recognised by Cahn, 'A King', 19, nn. 8 and 28.
     For photographs of some of the Saint-Denis examples, see Wulf, Kapitellplastik, figs 52 (centre),
56, 70 (where the face to the right has four sets of volutes) and 73. The double-volute format was
anticipated in one capital carved for the first master on the north pier between the central narthex
bays at Saint-Denis (Wulf, Kapitellplastik, figs 63 and 64).
     For the examples at Saint-Germain-des-Pres, see Wulf, Kapitellplastik, figs 125, 127 and 129.
      On these, see further Cahn, 'A King', passim. There also survives a smaller corner capital
representing the Annunciation, which is now set into the masonry of the castle ramparts at Dreux, and
two other corner capitals, similarly reset, showing: i, confronted harpies; and 2, an angle-figure
grasping foliage. Others seen by de Guilhermy in the nineteenth century have now vanished (Cahn,
'A King', 27-9).
      This would place them in the most prominent position within the liturgical choir, perhaps
surrounding the high altar. It is of course possible, however, that they were set in the aisles and not
the central vessel.
     Cahn, 'A King', 17, fig. 8. This artist was probably also responsible for the smaller Annunication
capital (Cahn, 24, Fig, 14) and the one with an angle-figure grasping foliage (Cahn, 27, fig. 20).
     Cahn, 'A King', 16, figs 5 and 6.
     The arguments for the Herod's having originally been part of this capital are made convincingly
by Cahn, 'A King', 14—19. Compare also the head of this figure with that of the king standing just to
the right of the Virgin in pl. VIIla.
     See L. Merlet and L. Jarry, Cartulaire de I'abbaye de la Madeleine de Chateaudun (Chateaudun
1896), xxiii; and Gallia Christiana, VIII, col. 1318, for Innocent II's confirmation of the reform. On
the construction, see further E. Lefevre-Pontalis, 'Etude archeologique sur l'eglise de la Madeleine de
Chateaudun', Bulletin de la Societe dunoise, V (1887), 299-310;
G. Outardel, 'Chateaudun', Congres archeologique, XCM (1930), 442—60; and Grant, 'Aspects'. The
last of these is by far the most convincing, although I would argue with the contention (25 and 27)
that work did not begin until c. 1150.
     On the sculpture of this portal, see further A. Lapeyre, Des facades de Saint-Denis et de Chartres
aux portails de Laon. Etude sur la sculpture monumentale dans I'lle-de-France et les regions voisines
au XIIe Siecle (Paris 1960), 79-81; and S. Baratte-Bevillard, 'La sculpture monumentale de la
Madeleine de Chateaudun', Bulletin archeologique, n.s. VIII (1972), 105-25, esp. 118-25. The latter
author proposed a date of 1145—50 for this portal.
     Cahn, 'A King', 24.
     Baratte-Bevillard, 'La sculpture monumentale', 124—5. In regard to Dreux, Cahn ('A King', 23)
has noted the use of the fantastical ville sur arcatures on the Nativity capital and has called attention
to the contemporaneous appearance of the same motif in the western portal capitals at Chartres and
elsewhere in the region south and west of Dreux. The occurrence of this convention is rare in the
Paris area, although a good example can be found in the northern ambulatory of Saint-Martin-des-
Champs. On the motif, see further J. Baltrusaitis, 'Villes sur arcatures', Urbanisme et architecture.
Etudes rentes et publiees en I'honneur de P. Lavedan (Paris 1953), 31—40.
     Cahn 'A King', 22-4, emphasises the dissimilarity between the figure styles at Chartres and Dreux,
but wonders to what extent the Chartres style was itself an interloper in the region, and raises the
possibility that the Dreux figures might afford 'at least a glimpse into some constants of local style
before the masterful synthesis achieved by the chief master of the west portal [sc. of Chartres]
changed the course of events'.
     This evidence, which will be discussed in detail in a future study, is partly circumstantial and
partly stylistic. Work on the choir is likely to have been started between 1130 and 1141 for two
reasons. First, the cult of the saint, the patroness of Paris, grew enormously in importance after 1130,
when her relics effected the cure of hundreds of plague victims (Gallia Christiana, VII, cols 708—9;
and R. Giard, 'Etude sur l'histoire de I'abbaye de Sainte-Genevieve de Paris jusqu'a la fin du XIIIe
siecle', Memoires de la Societe de 1'histoire de Paris et de I'lle-de-France, XXX (1903), 41—126).
This, along with vast riches and an exemption from episcopal authority, soon raised the canons of
Sainte-Genevieve to the 'premier rang dans le clerge parisien' (E. Raunie, Epitaphier du vieux Paris.
Recueil general (Paris 1914), IV, 353—416; esp. 365), and it may also have provided an incentive for
the enlargement of the choir, where the body of the saint was kept. Construction was certainly
underway in the crypt during this period since the obituary of Sainte-Genevieve notes that Canon
Geoffroy 'construxit. . .

[p. 113]
ecclesiam Sancti Marie in Cripta' (A. Molinier, Obituaires de la province de Sens (Paris 1902),
I/1,495), and in 1141 established a priest there to celebrate parochial masses (R. DeLasteyrie,
Cartulaire general de Paris (Paris 1887), 273, no. 282: 'ita ut idem sacerdos in altari Beate Marie
quod est in cripta secundum consuetudinem parrochiarum missas celebret'.). A dating in the 1130s for
the choir, whose upper parts were again remodelled at the end of the century, is substantiated by the
style of the architectural details visible in the engraving of Lenoir, Statistique monumentale, as well
as in unpublished drawings of the choir. The same date is acceptable for the surviving capitals of the
choir (C. Giteau, 'Les sculptures de l'abbaye de Sainte-Genevieve de Paris. Moyen age', Memoires de
la federation des societes historiques de Paris et I'lle-de-France, XII (1961), 7—55.
    M. Aubert and M. Beaulieu, Musee national du Louvre. Description des sculptures du Moyen Age
(Paris 1950), I, 22, no. 4. The re-use of Merovingian or Gallo-Roman spolia is paralleled during the
period at Saint-Pierre-de-Montmartre (choir and nave), Saint-Denis (crypt) and Saint-Germain-des-
Pres (choir triforium).
      The late eleventh-century dating is maintained by Aubert and Beaulieu, Description, I, 22, no. 4;
Giteau, 'Les sculptures', 12—13; and Vieillard-Troiekouroff, 'L'eglise . . . du temps d'Abelard', 749.
On the capitals of the nave, see further Vieillard-Troiekouroff, L'eglise . . . du temps d'Abelard',
passim; and Giteau, 'Les sculptures', 13—27.
      L. Pressouyre, 'Une tete de reine du portail central de Saint-Denis', Gesta, XV (1976), 151-60. See
also S. McK.Crosby, et al., The Royal Abbey of Saint-Denis in the Time of Abbot Suger (1122-1152)
(New York 1981), 39-40, no. 3a, fig. 13.
      The comparison with Dreux was first drawn by Pressouyre, 'Une tete', 155-6.
      The noticeable concentric striations of the abdomen of the Daniel seem also to have existed on the
now-lost body of the Saint-Denis queen, to judge from the drawing of Benoit published by
Montfaucon (see Crosby et al., Saint-Denis in the Time of Suger, 40, fig. 13).
      Louis VI founded Saint-Victor in 1113 (Luchaire, Annales, 82—3, no. 160) and provided it with a
rich endowment (Luchaire, Annales 166—7, no. 363). Although construction is likely to have been
underway during at least the 1120s and 1130s, the greater part of the church was reconstructed in the
late Middle Ages. Only the plan of the twelfth-century building is known; see further A. Grimault,
'Rapport. . . sur les fouilles effectuees . . . place Jussieu (decouverte de vestiges de 1'abbaye de Saint-
Victor)', Proces-verbaux de la Commission municipals du vieux Paris (meeting of 28 February
1931), 64—8 (plates and plan published in the minutes of the meeting of 9 May 1931); and M.
Vieillard-Troiekouroff, 'Les anciennes eglises suburbaines de Paris (IVe-X siecles)', Memoires de la
federation des societes historiques de Paris et I'lle-de-France, XI (1960), 17-229, esp. 160—4. Saint-
Pierre-de-Montmartre was founded as a nunnery by Louis VI and Queen Adelaide in 1134 (Luchaire,
Annales, 244—5, no. 536) and the choir was consecrated by the pope in 1147 (de Lasteyrie,
Cartulaire generale, 309— 10, no. 341). The entire church may have been finished when Adelaide
was buried there in 1154. On the architecture, see further F. Deshoulieres, 'L'eglise Saint-Pierre-de-
Montmartre', Bulletin monumental, LXXVII (1913), 5-30.
      These relationships will be detailed in my forthcoming book.
      I have already discussed this possibility in my article, 'The Influence of Castle-Building on
Ecclesiastical Architecture in the Paris Region 1130—1150', in K. Reyerson ed., The Medieval Castle
(Minneapolis, in press). It will be explored further in my forthcoming book.
      On this see further the penetrating comments of J. Bony, 'The Genesis of Gothic: Accident or
Necessity?', Australian Journal of Art, II (1980), 17-31.
      Bony, French Gothic, 64-6, was unaware of Saint-Etienne, but drew attention to its antecedents in
his discussion of the sources of Sens.
      Dado wall arcades are not a feature of either Burgundian or lle-de-France architecture before Sens
(with the exception of the axial trefoil chapel of Saint-Martin-des-Champs), but do occur regularly
from c. 1075 in west-central France (e.g., Le Mans Cathedral, La Couture, Notre-Dame-du-Pre,
Chateaudun etc.).
      At Sens the chamfered dosseret with flanking shafts occurs in the piers situated on the chord of
the hemicycle (choir side). The feature is found contemporaneously in the choir of Saint-Lomer at
Blois and in the eastern bay of the nave of Le Mans Cathedral. It reached the Ile-de-France in the
1140s not only at Sens, but also in the choir of Notre-Dame in Etampes, another building associated
with the west-central Loire Valley. A related disposition occurs in the axial piers of the ambulatory of
Saint-Martin-des-Champs (1130s), and the profile was used for the ribs of the crossing bay at La
Ferte-Alais in the 1140s.
      See for example Salet, Poissy, passim.
      Poissy also has a dado wall arcade in its choir, similar to the one at Sens and Saint-Martin-des-
 Champs (see above, n. 109).
      Henri de France, son of Louis VI and brother of Robert de France (count of Dreux from 1137),
 was born perhaps in 1121 (A. Lewis, Royal Succession in Capetian France: Studies on Familial
 Order and the State (Harvard Historical Studies, 100) (Cambridge, MA 1981), 59-60 and 250, n.
 71). He is cited as abbot of Notre-Dame in Poissy, along with the royal castle-church of Saint-
 Mellon on Pontoise, as early as 1125/6, when he could have been no more than five years old
 (Luchaire, Annales, 166-7, no. 363). He was subsequently, and simultaneously, abbot of the
 following royal monasteries: Notre-Dame in Etampes, Saint-Denis-de-la-Chatre in Paris, Notre-
 Dame in Mantes, and both Notre-Dame and Saint-Spire in Corbeil. He also held the treasury of
 Saint-Martin in Tours, of which the king was the titular abbot. He renounced all of these positions,
 which were transferred to his brother, Philippe de France, when he was elected bishop of Beauvais in
 1147/9. On this and on the king's temporal rights at these foundations, see Luchaire, Annales, cliv-
clv; and idem, Histoire des institutions monarchiques de la France sous les premiers Capetians
(987—1180) (Paris 1891), i, 156 and II, 72.

Shared By: