Introduction
An exceptional French historic monument, the Château de Versailles is without doubt one of the most famous buildings in the world. Its construction during the seventeenth and eighteenth centuries greatly influenced the course of European architecture and art history. Thousands of books and articles have been written about it. Given that the site is so well known, what could a modest archaeological field evaluation, lasting only a few days, possibly add to the sum of knowledge?
Quite a lot, as it turns out, because the trenches were opened in the outer courtyard of Louis XIII's first lodgings at Versailles. The lodgings themselves were built in 1623 and completely demolished in the period 1631-1634. The same fate befell the courtyard, laid out in 1624 and destroyed in 1662, without a single drawing having been made. The discovery of structures located on or around the courtyard would therefore add to our knowledge of the origins of the castle and of the principal entrance to Louis XIII's residence.
The test trench was opened in spring 2006 in the royal courtyard in front of the principal castle façade (Figures 1 and 2). One hundred metres long, aligned east-west, the trench was in the royal courtyard sandwiched between the forecourt (the Ministers' yard) and the small yard giving onto the king's apartments (the Marble Courtyard).
View of the evaluation trench (photograph: Gaël Pollin).
Partial view of the evaluation and of the shored services trench (photograph: Jean-Yves Dufour).
A projected underground services gallery in the royal courtyard provided an opportunity — just before and during the first days of construction work — to examine a length of 57m and a width of 6m of the courtyard, that is 342m² or 19 per cent of a courtyard which covered an estimated 1800m² (Figure 2). There is a clear conflict between the prestigious nature of the site and the limits within which the archaeology could be tested and this obviously restricted what could be discovered about this major site. Nevertheless much has been learned about the enclosure wall of Louis XIII's residence.
A late medieval path
The substratum, encountered at 1.6m below the royal courtyard, consists of Fontainebleau sands: it is light grey in colour and contains siliceous pebbles. This level was cut through by a path aligned NNE-SSW. Pottery sherds dating from the fourteenth and fifteenth centuries were found in its associated occupation layer. Four wheel ruts spaced at 0.4-0.6m intervals cut through the track. The left side of the track may have reached 1m in height and shows that the path was lower than the surrounding fields. In the sixteenth century it was called the 'hollow way' or the 'path of the hollow way'.
By good luck therefore, the excavation trench revealed a former medieval path whose existence, but not its location, was known to historians. This path led from the village of Versailles to Marly and then on to Saint-Germain-en-Laye. The location of this path is important since it fixed the eastern limit of the land on which Louis XIII built his first hunting lodge in 1623. In fact the royal works crept onto the path, but this was of no great consequence for those who used it, as it was simply moved a little to the east in front of the residence.
The enclosure wall of Louis XIII's residence
The medieval path described above was lost to view in 1624 when the ground was made up in preparation for new royal constructions. These include the foundations of two parallel walls, aligned north-south, seen in the excavation to cross the middle of the royal courtyard (Figure 3). Wall 17 was visible for a length of 8.5m. It was founded in a make up of fine yellow sand, more than 1.3m deep, which covered the old hollow way. The wall foundations, 0.7m wide, comprise two courses of meulière blocks, 20-30cm² in size, and a core in a yellow-brown sand matrix. Disturbed in the middle of the courtyard by a later construction, wall 17 continued into the northern half of the yard where it was called wall 96: 0.75m wide, it consisted of irregular courses of meulière blocks in a silty sand matrix. Where it covered the hollow way, this wall also cut into more than 1m of sand, but its alignment diverged from that of the path. Parallel to wall 96, wall 90 was 3.2m to the east: 1m wide, it was built with large meulière blocks laid flat, and a core of small stones, all bound by a beige sandy mortar.
Plan of the structures found within the limits of the future services trench in the royal courtyard.
These walls correspond to a detailed description of 28 July 1631, which states that, for the wall enclosing the garden and the park, 'the foundations shall be dug to a depth sufficient for a breadth of two feet until ground level and this will be filled with good stonework built using lime mortar, and on this construction two inches narrower on each side...' (Arch. Nat., Minutier Central; étude XIX 403). The two parallel walls are connected to other, perpendicular, walls and to the foundations of a circular tower.
A tower
The only image known of the courtyard of Louis XIII's residence is on a small engraving by Gomboust (see Figure 5 below). Two round towers with pepper pot roofs are connected by the enclosure wall, and seem to be bonded to the exterior angles of the side buildings of the courtyard. Before the test-trenching the existence of these towers was in doubt. The only written contemporary confirmation mentioned a 'tower' which was near the north-east angle of the courtyard or 'near the gate'.
Wall 90 was perpendicular to the tower (Figure 4) and both foundations were contiguous. Only half of the plan of the tower was seen in the evaluation trench. The circular tower foundations (SU 104) were 7.4m in diameter, with internal measurements of 5.6m in diameter (17 feet), probably a more accurate reflection of the builders' intentions. The area within the towers measured 230 square feet or 24.6m². The masonry foundations were 0.8m wide and comprise courses and irregular layers of meulière blocks, with a beige sandy mortar.
Plan and photograph of the tower on the castle enclosure wall in context.
In the southern part of the courtyard, symmetrical to the north tower, a vast pit full of demolition debris was found (pit 13, Figure 3). A segment of this circular pit was visible over a length of 5.5m and a depth of c. 2.5m. The fill comprised layers of six-sided floor tiles, roof tiles, bricks and plaster separated by layers of brown sand. This rather commonplace material is in keeping with Louis XIII's residence, and in contrast with its sumptuous successor. The location and contents of this structure strongly suggest that these are the remains of a second tower, built directly over the hollow way, which explains the depth of the pit.
Development
One of the few representations of Louis XIII's residence in Versailles is to be found on Gomboust's Une Vignette extraite des Maisons royalles et remarquables aux environs de Paris (Figure 5) which is to be found together with his plan of Paris in the Bibliothèque Nationale (Cabinet des Estampes). On this engraving, the walls and the towers recently found are seen as part of a projection which separates the outer courtyard and the park from the public way. The excavation has shown that these towers protrude from the wall-line.
The evaluation results seen in plan (bottom) and superimposed (top) on an extract from the engraving 'Maisons royalles et remarquables aux environs de Paris', 1652 (Bibliothèque Nationale, Cabinet des Estampes).
The two enclosure walls belong to different construction phases. The courtyard wall was built in 1625 whereas work on the wall on the park only began in 1631. They looked different. According to Gomboust, the park wall was lower than the courtyard wall on which are represented four elements which may well be pavilion roofs with finials. In the middle, the monumental entrance was topped by a pediment also decorated with finials. The construction of a basin at a later date seems to have destroyed all archaeological evidence for the entrance.
A possible interpretation for the four roofs can be proposed. The similarity between the walls 90 and 17/96 suggests a continuous pentice roof. No wall perpendicular to the enclosure was found during the evaluation. Nothing confirms the hypothesis of four small buildings next to one another. Rather there was one building, 10 French feet wide, probably used for storage.
The results of the evaluation add detail to Gomboust's drawing: the courtyard wall is accompanied by a second wall, which tends to suggest that there was a building against the inside of the courtyard wall. This is likely to have been a long narrow wooden building, eleven bays long and on average 9 French feet wide. There would have been a central bay for a pediment, four bays for the pavilions with hipped roofs and six bays with a pentice roof. Alternate pentice and projecting roofs, flanked by the pepper pot roofs of the two towers would have produced an interesting decorative effect. Was this building a gallery providing shelter between the stables and the lodgings or was there a number of small rooms separated by wooden partitions? The second possibility would solve the problem of the location of certain items that the Inventory of the king's and janitor's property (Arch. Dép. Yvelines, 1F32; B3586) places in the courtyard but which can be neither in the stables nor the lodgings. These include the musketeers' quarters, the building where an oven is to be found, eighty fowl, the small store under the tower and the small pigeon loft next to the enclosure wall. In these locations the janitor kept materials such as wood for building and heating, kindling, bundles of hay, pigeon excrement, baskets, bags, small kegs, trestles, planks, boards, doors, tools and other material.
Louis XIII's royal place of residence in Versailles was a modest affair, built cheaply in a few months (Le Guillou 2000: 91), and the enclosure and tower(s) are in keeping. The stonework and foundations are of squared blocks in a light mortar and of no great quality. In fact the lack of care taken in the preparation of the foundations is astonishing when compared to other, less prestigious, buildings. After the digging of the ditches in 1624, the upcast, a considerable quantity of sand, was used to bury the path and make up the ground to a level surface, to serve as a courtyard in front of the residence. The fact that the new buildings were founded on fine sand led inevitably to problems.
The extremities of the lodgings
Fragments of three walls, aligned east-west, and perpendicular to the courtyard walls were also found in the royal courtyard (see Figure 3). They are clearly the limits of the lodgings.
The south building
The eastern part of wall 5 was strongly built, existed to a depth of 1.5m and was 1-1.1m wide. The fact that it was built over the former hollow way explains the depth of the made up ground, and the deep, reinforced, nature of the foundations. Twenty feet (6.67m) north of this wall, a robber trench for a second east-west wall (wall 6) represents one of the buildings drawn by Gomboust. The complete absence of mortar in the wall section visible in the evaluation trench suggests a 2 feet wide dry built wall.
The techniques used to build these two walls conform to what is known about local post-medieval practice. Stable ground was sought so that foundations could be laid. In the absence of stable ground, relieving arches were used or the foundations widened. As little mortar as possible was used. No surface or occupation layer was found. The occupation levels of both the tower and the courtyard were higher than the excavation level and were removed during the major works undertaken for Louis XIV.
The north building
A distance of 100 feet (33m) separates the south building from the fragmentary remains of the north wing of the lodgings. The width of the courtyard (100 feet) is an important discovery. The residence built in 1623-1624 was square (18 toises or 108 feet) externally and 104 feet internally. The courtyard wings were well placed, setting off the residence without masking it. The courtyard too was square. A documents of 1630 refers to the paving of 'the courtyard path towards the residence, being about 16 toises [96 feet]long and about 3 toises wide' ( Arch. Nat, Minutier central étude CV pièce 444, en date du 6 août 1630).
Wall 103 represents the southern façade of the north wing of the lodgings. Aligned east-west, wall 103 touched the exterior wall of the enclosure (wall 90) and also supported the internal wall of the same enclosure (wall 96). It was therefore an integral part of the protection of the residence. Its foundations were made of 10-40cm meulière blocks. Two courses of large rough hewn stones separated a core. The entire structure is bound together by a beige sandy mortar. The building technique is the same as that used for the courtyard wall. The eastern part of the wall dropped deeply into the ground and in so doing followed the contours of the hollow way below.
The two wings of the lodgings must have been identical. Had one been larger than the other, their roofs would have reflected this: the southern roof would have been very high and the northern one very low. Or the southern roof would have been stunted and the northern one would have soared, a highly unlikely proposition. If, as is most likely, the roofs had the same profile, the wings would have had the same width. Between 1623 and 1632, the wings of the residence (where the future Marble Courtyard would be) were 15 French feet wide within and 19 feet without. It may well be that the wings of the courtyard had the same dimensions.
Discussion
At the beginning of the seventeenth century, the site of the future residence was a valley, with a mixture of cereal fields and meadows in which there were four ponds. Until 1622, Louis XIII visited Versailles occasionally to hunt. He nonetheless maintained the rabbit warren belonging to the local lord, Jean-François de Gondi.
In 1623, the king came into possession of about 7 acres of land. Of this, he used around 2 acres to prepare the ground in front of the entrance gate as well as a by-path. On the rest of the ground he had his house built; he surrounded it by ditches, created a courtyard for the stables and the lodgings and also established a flowerbed. Taken together these elements formed a rectangle 100 toises long, from the courtyard entrance to the end of the flower bed, and 60 toises wide. Building work on Louis XIII's residence was completed by 1624. Its internal courtyard corresponded to the Marble Courtyard visible today.
The sandy upcast from the ditch around the house was used to fill in the hollow way over which the enclosure wall was partially founded. The remains of the tower are part of an enclosure wall intended to hide the residence from the gaze of outsiders. Versailles was the king's private property, paid for through the entertainment (Menus Plaisirs) budget. At this time, the king spent a lot of his time in the castle of Saint-Germain-en-Laye, hunting in the nearby woods. Occasionally, he and his companions went deep into the forest, and surprised by the falling light, were obliged to take refuge in a forester's hut or a wayside inn. The idea of a house in Versailles grew from the need to have proper accommodation near to where he hunted. He alone chose the house and which of his hunting companions could stay there. There was no room for his wife, or his mother or brother. In 1626, when he considered the works finally over, he invited the queens to dine and to hunt. But when evening fell he accompanied them back to nearby Saint-Germain and went back to sleep alone in Versailles.
Soon it became his refuge from the Court and its incessant intrigue. Over a period of 20 years, few people found the king's favour and few were invited here, perhaps less than a dozen. When the king was not in residence it was possible to visit the building, probably with the agreement of the janitor, François Mongey, or the head gardener, Hilaire Masson. However given the absence of works of art, it is unlikely that the place held much interest, although the architecture and the gardens were of interest to professionals. Indeed in 1635 Jacques Boyceau's book appeared, mentioning the flowerbeds and groves of Versailles, which were presented as examples for gardeners and amateurs to follow.
The dimensions of the wall tower are unspectacular. It was neither decorative as are the small angle towers of fortified farms, or massive as in medieval castles. The tower does not simply mark the antiquity of the landholding as royal land. Walls and towers were a symbol of power. They spoke of feudalism and were sufficiently symbolic in the first half of the seventeenth century to be used in a royal project. The medieval image of the castle still existed in the post-medieval period. Not all circular towers date from the Middle Ages. The architectural transformation of the great aristocratic buildings in Ile-de-France did not begin until the end of the reign of Louis XIII. Until then, classical genius was expressed by the interiors rather than by the conception of the buildings.
Behind the wall, stables and lodgings give onto a courtyard which is in place by July 1624. Gomboust's engraving shows long, narrow buildings aligned east-west. When the residence became important, during the seventeenth century, the refuse from the farm buildings (stables, barns and stores) was thrown out, onto the courtyard. Separating the master from his service quarters had become a sign of luxury.
The archives show that the lodgings wing is on the north side of the courtyard, and thus the stables were inevitably on the south side. The Inventory of the king's property dating from 1630 lists an apothecary's chamber, a small room near the door, Monsieur de Montalan's (captain of the carabineers) chamber, the chamber of the maître d'hôtel and of the valets — an office, the Chamberlain and the Chamber of the king's maîtres d'hôtel. This is clearly what is to be expected in a royal household. A note concerning the repair of the tiled roof of the lodgings dating from 1639 mentions a cistern in or next to this building (Arch. Nat., O1 2127 f°254). The 1630 inventory of the king's property does not mention the stables. However we should remember that 'to be in best condition, they should be in single file' ( Reference SavotSavot 1625: 109). So that their paintwork was in good condition, coaches were usually stored where there was a northern opening. The stables were probably paved.
Conclusion
Test trenching was restricted to the area within the perimeter of the projected underground services gallery in the Royal Courtyard. Despite this, the results are instructive about the layout and construction of the wall around the forecourt of Louis XIII's residence. The remains of the enclosure wall and of a tower have been clearly identified. It was not possible to gather sufficient information about the lodgings to treat the subject in depth. Nonetheless the discoveries confirm the accuracy of Gomboust's engraving, the only known picture of an outer courtyard which, according to Jean Héroard, Louis XIII's doctor, had been the idea of the king himself. On Tuesday 2 July 1624 'he went to Mass, fed deer meat to his dogs, came back to the residence, had his musketeers exercise, then decided on the lay out of the courtyard for his house at Versailles'. Post-medieval archaeology can add new information even about a site which has been as closely studied as the Château de Versailles. The poor state of knowledge about post-medieval residences other than Versailles is an indication of what remains to be done in the domain of post-medieval archaeology.
