Luca di Tommè’s Crucifixion in Montepulciano: Its Construction and Function

Fabric Support, Paint Layers, and Inscriptions

Remarkably, the two photographs taken of the Crucifixion in 1929 preserve a considerable amount of information of what the painting looked like when it first appeared: how it was constructed as well as its condition. The Alinari photo (fig. 1) is the only known image which records how the framed painting appeared before the major restoration of 1961, showing the later additions of two oval portraits of St. Francis and Sta. Chiara still attached in the upper corners of the gable. During this restoration, which was carried out at the Pinacoteca Nazionale in Siena, the two corner saints were removed, as well as most of the wood strainer, and the painting was mounted to a solid wood support, thus transforming the painting into one similar to the hundreds of panel paintings from the Trecento that we encounter in museums today.

Unfortunately, the sole Alinari photograph that recorded the Crucifixion as it appeared in 1926 has been heavily retouched, and the printed image has lost detail in the shadows. Because of this, it is less useful as a resource for evaluating the condition and even the structure of the Crucifixion when it first appeared in the early twentieth century. The Sansoni photograph (fig. 2) thus will be our main source for evaluating the structure and condition of the Crucifixion. Although the two corner gable saints were removed photographically when the negative was printed, the resulting image, having less contrast than the Alinari photograph, and being unretouched, preserves more surface detail, revealing the outline of the wood strainer to which the canvas support was attached, damages to the support and paint layer, and even fly droppings on the paint layer. We will refer to this photograph in the following paragraphs.

2

Luca di Tommè, Crucifixion, Montepulciano, Museo Civico (Sansoni photograph taken in 1929 with the two corner saints removed by retouching)

Judging from the Sansoni photo and a detail of the severe damage along the back of Mary Magdalen (fig. 3), the Crucifixion was painted on a single piece of moderately thick, plain-woven linen fabric having a pronounced weave deformation in one direction, which Julia Burke, head of textile conservation at the National Gallery in Washington, has called “warp-faced.”^[7] This single piece of fabric, was glued and not tacked to a six-member wood strainer,^[8] each member being approximately 5 cm wide, whose imprint on the paint layer can be seen in the Sansoni photo. We can clearly see where the edges of the strainer pressed against the canvas support, creating visible linear cracks in the gesso and paint layers. Usually canvas paintings were secured to strainers with small nails or tacks, as described by Cennino Cennini, but for the Luca, the canvas was glued to its strainer, according to notes recorded by Andrea Rothe, the conservator in charge of the 1979 restoration.^[9] Apparently because of the difficulty in removing the strainer from the cloth support during the 1961 restoration, all but the two lateral strainer members were removed (fig. 4). They can be seen as dark vertical strips of wood on both sides of the painting. It is from these two lateral strainer members that we know the strainer corners were of half-lap construction, probably glued together and reinforced with wood dowels. The two, dark, horizontal wood members were evidently added to provide additional support. An estimation of how the original strainer was constructed is shown in fig. 5. As for the strainer for the two gable saints, an imprint of its overall design on the canvas support can be seen in the Alinari photo.

3

Detail of fig. 2, revealing exposed fabric pattern along back of Mary Magdalen

4

Reverse of Luca di Tommè’s Montepulciano Crucifixion before 1979 restoration

5

Drawing of wood strainer pattern on Luca di Tommè’s Montepulciano Crucifixion

Although the ground for the Montepulciano Crucifixion was not specifically tested or identified, it was thick enough to obscure the canvas weave, much like the grounds on panel paintings, and probably consisted of gesso grosso and subsequent layers of gesso sottile, the latter of which Cennino Cennini recommended as the ground for paintings on cloth.^[10] Gesso sottile was a slaked, roasted and powdered gypsum, which was kept suspended in water to prevent the gypsum from congealing. After about a month in water, the slurry was removed, the excess water squeezed out, and the gesso sottile made into small cakes for use. The cakes were then mixed with warm glue size, and thinly applied to the surface so that just the weave interstices were covered. After the gesso had dried, it was carefully scraped to remove all surface irregularities. The composition would have then been drawn on this white surface, and the background completed before the artist started work on the figures, following the same order of completion Cennini described for painting on panel.^[11]

No paint samples were removed from the Crucifixion for identification. From this writer’s familiarity with pigments available to Trecento artists, Luca used a normal range of pigments as described by Cennini,^[12] the only change being a painted rather than a gilded background, and the addition of ultramarine blue for the mantles of some of the saints, particularly those on the right side of the Cross.^[13] The dry pigments would then have been ground in water to a slurry, mixed with thinned egg yoke, and then applied in short paint strokes, typical of the egg tempera technique.

While Cennino mentions gilding the background of banners by having the canvas support firmly placed against a wood backing,^[14] Luca apparently decided that because the canvas support was glued to the strainer of the Crucifixion (though perhaps not as firmly as on a panel), it left those portions not glued too difficult to adequately support underneath. Luca, therefore, dispensed with gilding and painted the background with azurite blue instead, reserving gilding for the haloes of the saints. Other than the Crucifixion being painted on canvas rather than wood, the process of painting seemingly follows the recipes carefully laid out by Cennino Cennini in his Il Libro dell’Arte.

In early 1975, I made a trip to Montepulciano to photograph the Luca di Tommè Crucifixion with color infrared film, which I had been using since the early 1970s specifically to identify ultramarine blue, since ultramarine is highly reflective to infrared rays, appearing nearly white on black and white infrared photographs, and bright red on color infrared transparency film.^[15] Azurite blue, on the other hand, appeared a dark blue on 35 mm color infrared film. The two groups of figures on either side of the Cross on Luca’s Montepulciano Crucifixion (figs. 6, 7) were recorded with color infrared film. Only three small areas of ultramarine blue are found among the female figures to the left of the Cross, whereas ultramarine blue was used more liberally on the right side.

6

Color infrared photograph of left half of Luca di Tommè’s Montepulciano Crucifixion

7

Color infrared photograph of right half of Luca di Tommè’s Montepulciano Crucifixion

The group to the left of the Cross (fig. 6) showed that ultramarine blue had been used for the mantles of two of the background figures, with a third wearing an ultramarine blue cap. And for the two Mary’s consoling the Virgin, the green mantle for the Mary immediately to the right of the Virgin is recorded as greyish pink in color infrared, which I have interpreted as ultramarine ash (a weaker grade of ultramarine that registers more grey than blue) mixed with a yellow, such as lead-tin yellow. Ultramarine ash was extracted at the end of the procurement process, and it was a weaker and less expensive grade of the mineral—more of a grey color than blue. Although drained of blue, the grey color was still used by artists, such as depicting grey beards.

Interestingly, the mantles of the Three Marys on the Crucifixion in Montepulciano show a similar color combination to the three Marys on a Crucifixion predella by Luca di Tommè in San Francisco, dated to approximately 1365 (fig. 8).^[16] In this painting, the Mary to the left wears a yellow mantle that was probably painted with a lead-tin yellow pigment, whereas the figure to the right of the Virgin wears a green mantle with a purple lining. The green in this case is probably a copper-based pigment such as malachite or verdigris.

8

Luca di Tommè, Crucifixion, ca 1365, tempera and gold leaf on wood panel, 41 × 59.7 cm. San Francisco, Legion of Honor, Gift of the Samuel H. Kress Foundation, acc. no. 61.44.6

To the left and right of Christ, extending at an angle to Mary on the left and John on the right, are two fragmentary, raised blackish inscriptions on top of the azurite blue background (fig. 9). The inscriptions could be painted with indigo mixed with a bit of black. Both phrases were supposedly uttered by Christ on the Cross to his mother Mary on the left, and to his closest disciple John, the Evangelist, on the right.^[17] Normally words and sentences are read from left to right. However, the inscription on the left, which extends from Christ to Mary, is a mirror image, meaning the words were painted in reverse, beginning on the right and continuing down to the left. The inscription would appear correct if one were on the other side of the inscription looking up at Christ.

9

Detail of Luca di Tommè’s Montepulciano Crucifixion, showing inscriptions to left and right of Christ

The seven words spoken by Christ on the Cross are rare in Trecento painting, the one on the Luca di Tommè Crucifixion being the only one known from Siena. We do have a similar but earlier example by Bernardo Daddi of a Crucifixion in the Lindenau Museum, Altenburg, Germany, where three utterances are depicted in colored scrolls on either side of Christ.^[18] The one on the right, in a red ribbon, and spoken by a Roman centurion to Christ, reads from left to right: “Truly, this was the Son of God.” And the red and blue ribbons on the left are mirror images citing the same verses as in Luca’s Crucifixion (John 19: 26–27).

Damages and Restoration

No written record of the damages to the Crucifixion was made when the painting first appeared in the late 1920s. However, the Sansoni photograph of 1929 (fig. 2) reveals a number of serious damages both to the paint and the fabric support: Paint and gesso loss on right side of gable, exposing the fabric support; gilding loss in halo of Christ, as well as haloes of angels holding chalices, and the haloes of saints next to Mary; large gesso and paint loss on proper left shoulder of Mary with inverted T-shaped tear in fabric; large “L”-shaped loss of paint and gesso in the Virgin’s mantle, below female saint to left of Mary, with a jagged tear in fabric; a tear with considerable paint and gesso loss, exposing the fabric pattern in the mantle of Mary and along the back of the kneeling Mary Magdalen; a long vertical tear on the right side of the painting, 1.2 cm in from the right side, and 36.8 cm long. Along this tear one can see small holes, with 1.5 cm spacing between them, which probably represent small nails used to secure the painting to the strainer. These holes extend to the top of the transverse strainer member; an “L”-shaped paint loss exposing a gesso layer in the mantle of St. John, the most noticeable among the fewer damages are found in the group to the right of the Cross; and lastly, a deep horizontal fold along the bottom of the painting where, because of light reflection, a lot of detail is lost.

Although the early commenters of the Crucifixion were aware of the condition of the painting, nothing substantial was done to stabilize the paint layer until 1961. In April 1937, Peleo Bacci, the Superintendent of Siena, wrote to the Mayor of Montepulciano describing a visit that Italo Dal Mas, a restorer, had made to the town to evaluate the condition of three paintings in the Museo Civico and give estimates on what it would cost to restore them.^[19] One of the paintings was Luca’s Crucifixion. Dal Mas described the damage to the painting in considerable detail, suggesting at one point that the painting must be removed from its strainer, and new pieces of canvas be attached to the sides and to parts where there were tears. There is no evidence that any restoration was done in response to the letter. Twenty years later, the mayor of Montepulciano indeed wrote to the Superintendent of Siena noting the urgency of treating the Luca Crucifixion, since extensive areas of lifting paint and cracks (“bolle e screpolature”) had been observed.^[20]

Finally, in 1961, the Crucifixion was sent to the Pinacoteca Nazionale di Siena for major repairs, which included removing most of the wood strainer and mounting the painting to a solid wood support, plus filling the gesso and paint losses and retouching them. Unfortunately, no photographs or reports about this restoration have been found, so we are limited in concluding exactly what was done. What we do have are photographs taken of the painting and its back before the last major restoration of 1979, and from these we can determine the major components of the 1961 restoration, which included mounting the canvas painting to a composite panel of four insect-riddled poplar panels, but leaving intact the two lateral original strainer members, which appear dark on the photograph (fig. 4). Evidently, they were too firmly attached to the canvas support to be easily removed.

The restorer in 1961 also added a skull at the bottom of the Cross (fig. 7). There is no evidence that it was part of the original composition, and it was removed when the painting underwent major conservation treatment at the Tintori Laboratory in the Palazzo Pitti in 1979 (see fig. 12).

10

Luca di Tommè’s Montepulciano Crucifixion, condition of painting before 1979 restoration

11

Detail of fig. 2, showing fly dropping spots on light colored blouse of female to right of the Virgin

12

Luca di Tommè, Crucifixion, ca. 1365, tempera and gold leaf on linen, mounted on panel, 116 × 77 cm. Montepulciano, Museo Civico (state after 1979 restoration)

Unfortunately, this restoration failed to solve the issue of flaking of the paint layer, which the Mayor of Montepulciano pointed out in a letter to the Siena Superintendent of 7 September 1961, just four months later, saying that the paint was still peeling off (“si sta staccando tutto il colore”), and he asked that an expert be sent to the Museo Civico to evaluate the problem.^[21] A year later the problem of flaking paint persisted, and a visit by a restorer from the Superintendent’s office in Siena had the painting placed flat. Finally, in March 1967, restorers Primo Senesi and Attilio Galluzzi stabilized the painting by securing lifting paint and filling paint losses.

By 1975, when I saw the painting and photographed it, the areas of cleaving or lifting paint were extensive, but no paint had fallen off. I notified the Siena Assistant Superintendent, Alberto Cornici, of the problem, which may have stimulated action to correct it, for a year later, Piero Torriti, the Siena Superintendent, addressed the problem the Crucifixion faced in a detailed letter to the Dipartimento Istruzione e Culture delle Regione Toscana, focusing on environmental problems within the Museo Civico and problems brought on by the 1961 restoration.^[22]

By 1979, when the Crucifixion was sent to the Palazzo Pitti in Florence for restoration, the paint losses were extensive and disturbing (fig. 10), indicating that the painting had suffered major damage in the four years since I saw and photographed the painting in the Museo Civico. It was during this last major restoration that Andrea Rothe concluded that the canvas support for the Crucifixion had been glued and not tacked to the wood strainer, but how exactly is not known.

When the strainer members and the poplar panels were removed with chisels and a wood plane, traces of wood from the original strainer remained on the support where they had once been attached prior to the 1961 restoration, making it clear that the original strainer had indeed been glued to the fabric. The two lateral strainer members also preserved evidence of how the eighteenth century portraits of Saints Francis and Chiara were attached to the Crucifixion proper. We see this in a photo taken of the back of the painting in 1979 (fig. 4).

Conclusion

From just a handful of sources, we know considerably more about how Luca di Tommè’s Crucifixion was constructed and how it appeared at the time of its discovery in the mid-1920s than we did before. But one issue remains: what was the original purpose or function of the painting? We will address this question now. To do so, we will first examine what Cennino Cennino wrote about banners in his Il Libro dell’Arte, and then compare this with other early Italian paintings thought to be banners, and particularly what we know about the Luca Crucifixion.

Cennino Cennini’s description of painting on cloth is found in chapter CLXII of his Il Li-bro dell’Arte. It describes in brief terms how one should make and paint a banner on a linen or silk support, or what he terms zendado, which was a type of inexpensive silk cloth. He begins by describing how a piece of linen or plain woven silk should be stretched and tacked to a wooden frame, followed by sizing the support with a liquid glue solution. If the banner were to be painted on both sides, then both sides of the support should be covered with a layer of thin glue size. Once this layer had dried, Cennino then describes the application of a thin layer of gesso sottile,^[23] mixed with glue and with small additions sugar or starch for flexibility; if the banner were double-sided, gesso sottile would also be applied to the other side as well. He then specifies that the gesso sottile should just fill the interstices of the weave, since a thicker application of gesso and paint would cause these layers to flake off were the linen to be rolled. Gold leaf could be applied and tooled providing the support were carefully placed on a hard support, such as wood, when burnishing the gold and tooling it. Cennino’s description for painting followed his recipes for wood anconas. Finally, Cennino describes the application of a clear varnish to the paint, since he explains that this would help protect the banner from damage by rain, were it carried in a procession.^[24]

Only one of the paintings described in The Fabric of Images followed most of the procedures described by Cennino for painting banners, and that was the Decollation of St. John the Baptist, a small painting on cloth by an unknown Florentine artist of the late fourteenth century in the Museo dell’Opera del Duomo in Florence.^[25] It followed Cennino’s recommendation of applying a thin gesso ground, barely filling the interstices of the weave. Other aspects of this painting’s construction were not in agreement with what Cennino recommended, but this is not surprising, because individual or workshop preferences in the choice of materials or techniques are found in many Trecento paintings. For example, two paintings by the Master of Figline in the Worcester Art Museum, plus a companion panel in the Fogg Art Museum and another in the Courtauld Institute of Art in London—all parts of the same altarpiece— dispensed with terra verde underpaint entirely for the flesh tones.^[26] And Ambrogio Lorenzetti, the Sienese Trecento artist, used a mixed grey underpaint instead of terra verde for the flesh tones in his St. Nicholas of Bari panels in the Uffizi, in his Madonna del Latte painting, and in his Massa Marittima Maestà, all emphasizing his use of something different, and not subscribing to the recommendations of Cennino. The same color mixture of pigments has also been identified on the other paintings by Ambrogio Lorenzetti mentioned above.

From what we now know about Luca’s Crucifixion, it differs in important and fundamental ways from Cennino’s description of banners as he understood them in the late Trecento. First of all, the Luca painting was executed on a fairly robust piece of plain woven linen, or what is termed “moderate weight,” which was glued to a six-member fixed wood strainer.^[27] The imprint left by the strainer is clearly visible on the Sansoni photograph. With the linen support firmly bonded to the strainer, it was then presumably covered with a thin layer of glue size on the exterior surface before the gesso was applied. Nothing is known about the gesso covering the linen support of the Crucifixion, whether it was a combination of gesso grosso or gesso sottile, but it is obvious from the paint surface that it was thickly applied to hide the plain-woven pattern of the linen weave. At this point, once the gesso had dried, it was carefully scaped to remove any surface irregularities. Since the support had been glued to the strainer, creating a pattern of supported and unsupported areas, it was not possible to place a solid surface underneath the support so that bole and gold leaf could be applied. Instead, Luca painted the background with azurite blue, and locally gilded the haloes of the saints without tooling them.

According to Cennino’s Il Libro dell’Arte, paintings on canvas were constructed much like those on panel, with a gesso ground and an egg tempera paint layer. The major difference between them, of course, lay in the type of support. Wood was solid and inflexible, whereas canvas was not. Thus, a gesso ground on panel could be thick, and it was possible to fully tool haloes on panel in ways that was not possible on a cloth support. The latter support imposed limitations on what could and could not be achieved, and

Cennino clearly laid this out in Chapter CLXII of his Il Libro dell’Arte, where linen or silk supports clearly implied processional banners. As described by Cennino, the canvas, usually linen, was tacked to a wood strainer, and then a thin gesso sottile layer was applied to the surface. Keeping the gesso layer thin was important, because by having gesso just filling the interstices of the weave, the painting could be rolled for storage.

One other aspect of the Luca Crucifixion that has intrigued me is how it might have been carried in a procession, assuming it was made as a banner. This is an area of investigation that has not been thoroughly explored, but then the evidence is meager. Michael Bury, in his article in The Fabric of Images, addresses this issue, but focuses on paintings from Umbria in the fifteenth and early sixteenth centuries. One of those was Niccolò da Foligno’s double-sided framed banner in Deruta, which depicts Sant’Antonio Abate on one side, and SS. Francesco and Bernardino on the other.^[28] The paintings are 221 cm high and 114 cm wide. Also, being framed, they must have been a little unwieldly, as they were carried by a single staff inserted between the two saints on one side.

There are no marks on the two original vertical strainer members still attached to the Crucifixion in 1979 (fig. 4) to indicate how it might have been carried, such as nails or nail holes that might indicate the location of a carrying staff or pole. Besides, the fabric support was glued to the wood strainer, as confirmed by Andrea Rothe in 1979, which would have prevented the attachment of ropes to the strainer members, and then to a pole for carrying.

Most interesting of all is the fact that the Crucifixion was never varnished prior to its discovery in 1926.^[29] Cennino Cennini was quite explicit in writing that banners be varnished, “because sometimes these banners, which are made for churches, get carried outdoors in the rain,” and obviously they needed to be protected.^[30] When the Luca Crucifixion first appeared in 1926, the surface was covered with small grey spots of fly droppings, particularly visible in the light-colored areas as revealed on the Sansoni photograph (fig. 2). These spots never would have occurred had the painting been varnished soon after completion. This evidence, plus the use of expensive ultramarine blue and other materials, and its unusual construction, suggests that the Crucifixion was made for a purpose other than a processional banner, perhaps as a single devotional image. Unfortunately, we do not have incontrovertible evidence what this purpose might have been.

At the end of the last major conservation intervention in 1979, when the painting was safely and securely mounted to a solid wood support, the Luca Crucifixion appeared whole again—but perhaps not quite as the artist originally intended (fig. 12).

A major part of this article was researched and written during the Covid epidemic of 2020–2022. It would not have come to pass without the extraordinary help I received from Lucia Monaci-Moran, widow of my late friend, the art historian Gordon Moran. She made many trips from her home in Florence to Montepulciano, where she was born, and to Siena, to meet with colleagues and to copy documents relevant to this study, which were then provided to me.