By Eileen Sullivan
Kress Fellow in Paintings Conservation
The museum was given a grant last year by the Samuel Kress Foundation to fund the conservation of Paolo Veronese’s 1575 The Annunciation .
This year-long project recently came to an end, and the painting has been reinstalled in the 1916 building. Below is a follow-up to the May blog , which described the project and detailed the cleaning process as well as research which would inform the treatment.
Over the last six months, the painting underwent a transformation as the inpainting resolved distracting losses and abrasions. This work varied from integrating straightforward losses, such as the vertical seam that runs through Gabriel’s sleeve and the bed,  to more complex challenges.
Mary’s proper right hand, a focal point of the painting, had suffered a large loss in the center. Initially, inpainting in this area was used to produce an even tonality with the rest of the hand by deconstructing remnants and stains left behind from previous retouchings that could not be removed during cleaning. However, the location and size of the loss, combined with peripheral damage, made missing refinements by the artist difficult to reconstruct. A method I used to aid in conceptualizing the missing information was to reproduce the position of the hand under the same lighting conditions as used in the painting, with a live model. In this way I was able to better understand the anatomy of the hand and subtle effects of light and shadow. These observations, coupled with the knowledge of Veronese’s painting technique, provided enough of the missing information in this critical area to finalize the inpainting. In this way a limited amount of informed inpain!
ting melded the area into the original brushwork of the wrist and fingers.
Ochre  overpaint had truncated the dove’s back, pressing it into a slightly concave curve. Referencing other doves painted by Veronese suggested that the dove might have originally had a fuller, more voluminous body. Under magnification, tantalizing glimpses of white paint was visible beneath the surface, prompting further investigation. An x-ray of the dove clearly showed an original domed back. Using the references and x-ray, inpainting now suggests the dove as the artist intended. 
The most challenging aspect of the treatment was the inpainting of Mary’s skirt. Painted with smalt, a pigment vulnerable to color change, the skirt had undergone a visual shift that caused the robes to lose color, form, and volume, resulting in their unreadability.
An image of The Annunciation taken with infrared photography provided a valuable reference for the inpainting. Infrared imaging allows the viewer to see beneath the paint layer, to any underdrawings or imprimatura on the canvas. In this case, the infrared photograph clearly showed the heavy, deeply contoured drapery of Mary’s skirt. In the initial inpainting, I used this image as a guide to return some of the tonal values to the skirt. This was done using largely darker shades of the existing paint layer. A second step of inpainting was to slightly adjust the hue of the skirt. Under magnification a deep magenta-purple tone was apparent in the low areas of the canvas. This old, original glaze was replicated with a transparent violet glaze and modified the green, further darkening the contours as well as suggesting the artist’s original intent of a blue robe.
After a final varnish application the treatment was documented and the painting was reframed. It was a pleasure to invest so much time into the wonderful painting, and exciting to see it reinstalled in the museum.