National Gallery of Canada / Musée des beaux-arts du Canada

Bulletin 17, 1971

Annual Index
Author & Subject

Click figure 5 here for an enlarged image

Click figure 6 here for an enlarged image

Click figure 7 here for an enlarged image

Click figure 8 here for an enlarged image

Click figure 9 here for an enlarged image

A Reynolds Revived

by Mervyn Ruggles

Pages  1  |  2  |  3  

The Portrait of Charles Churchill was cleaned and varnished by F. W. Colley, a restorer under contract with the National Gallery of Canada in 1923. (13) In recent years the natural-resin varnish, dammar, had become extremely discoloured (fig. I), the flesh tone shad taken on a yellowish veil, and the background regions had lost all detail under a dark, dull "craze" state similar to that seen in old chinaware. Because of this and because the canvas, though lined approximately a hundred years ago, was nevertheless weak, worn away at the corners and splitting at the stretcher edges (fig. 7), the National Conservation Research Laboratory initiated a complete conservation treatment after a report in depth was submitted. (14) The report included examination by X-ray (fig. 6), infrared, and ultraviolet (fig. 8). The radiographs revealed that Reynolds had revised the pose of the sitter's left hand, which had originally rested on a parapet, and that foliage had formerly existed in the upper left background and a leafy festoon near the upper centre, above the head. The radiographs also proved that a network of bitumen cracks had been covered over during previous restorations. These were located practically all over the background except in the sky at the upper left. Fortunately, no part of the figure of Churchill had been affected by bitumen. Subsequent tests with solvents indicated that the foliage at the upper left had been painted over by Reynolds himself, but that the branch over the head had been covered over much later by another hand. Only during the removal of the discoloured varnish with a solvent mixture of equal parts of acetone and tetrachlorethylene was the real extent of the overpainting on much of the background revealed; the purpose had been to disguise the extensive vein-like fissures of the traction cracks (fig. 2). Gradually there began to emerge in several sectors a series of white gesso-like islands on top of the original paint film. The gesso, consisting of whiting bonded with animal glue, had been trowelled on to smooth over the voids caused by the receding paint film, and it covered a larger area than was necessary. The photo-micrograph (fig. 3) shows one of these places. Solvent tests in the section of the left hand indicate that the re-vision from a parapet to a tree stump was indeed a change made by Reynolds, but his reason still remains a matter of speculation. A rather different story unfolded when the discoloured varnish came away from the face. Strong white pigment with traces of faded carmine could be detected, and a red colour was observed on the cheeks and lips. It was readily established that these were retouches added during a restoration, because they spanned the normal age cracks in the original paint film (fig. 5).

Under the microscope the primary ground layer was a light grey colour of the type applied commercially, pointing to the fact that Reynolds did not prepare his own canvases but bought them already mounted on a stretcher-frame. Over the whole sky sector he applied an underlay or imprilllatura of blue-black prior to putting in the sky. The tree at the right and the foliage were the next stage. The blue-black underlay seems to influence somewhat the white of the face, producing the ashen-grey colouration that is now visible. Eight miniscule pigment samples (fig. 9), each one half the diameter of a pin-head, were taken for X-ray diffraction and X-ray spectroscopy by Mr. R. Boyer. (15) The results were as follows:

1 Yellow on the tree stump: yellow ochre and white lead

2 Green to the left of the tree stump: yellow ochre, Prussian blue and white lead

3 Green under the left elbow: yellow ochre, Prussian blue and white lead

4 Brown paint near the top of the tree stump: organic brown (bitumen?) and white lead

5 White paint on the frill of the right lace cuff: white lead

6 Yellow paint in the sky near the right elbow: Naples yellow

7 Blue paint at the upper left border: Prussian blue and white lead 

8 Dark blue paint in the sky directly above the head: Prussian blue and white lead

De Wild states that he had found that Naples yellow was in general use only after 1759. (16) However, Kuhn (17) and others have established that it was commonly used much earlier, and the existence of this pigment in the Churchill picture indicates that it must have been available to artists in London in Reynolds' time.

In making further reference to the paleness of the face, Buttery, a London restorer, stated in 1958 that, "Owing to Reynolds' carmine being a fugitive colour, his paintings of the fifties and sixties appear to-day faded and cold. This coldness is frequently increased by injudicious cleaning and lack of understanding of Reynolds' methods at this date". (18) Strangely enough, the flesh tones of the right hand do not exhibit this type of deterioration nor do they have the same consistency as those of the face. The hand still remains pinkish and only slightly faded. Because of this aspect, one could speculate that this hand as well as the left gloved hand, seemingly painted in the more conventional pigment technique, was the work of Giuseppe Marchi, Reynolds' life-long assistant who had accompanied him back from Rome in 1752. Or were they perhaps painted by one of his other assistants - Peter Toms? The hands in the Churchill painting appear to me, at least, to differ from those of several other Reynolds' I have examined. A score of assistants and pupils worked on Reynolds' pictures throughout his career. (19)

After the varnish was removed, a scalpel was used to gently shave away the gesso splatching. Overpainting in the foliated sections at the upper right came away with the varnish coating. The old lining canvas was taken off the back and the glue-paste adhesive was carefully scraped away. The vacuum process (20) was employed to reline the portrait with unbleached linen, using the special wax-resin adhesive formulated at the National Gallery (yellow beeswax, microcrystalline wax W-445, cyclohexanone resin, in proportions of 2 :2: I). A new custom-made stretcher-frame replaced the old one. A minimum of inpainting with pigment in a synthetic resin medium was necessary to match small, old paint losses. Finally, two coats of cyclohex-anone resin in xylene were applied as protection.

The Portrait of Charles Churchill now has a fresh, revived appearance (fig. 4) and, I believe, more closely approaches its appearance when Reynolds painted it 216 years ago.

Next PageNotes

1  |  2  |  3

Top of this page

Home | Français | Introduction | History
Annual Index | Author & Subject | Credits | Contact

This digital collection was produced under contract to Canada's Digital Collections program, Industry Canada.

"Digital Collections Program, Copyright © National Gallery of Canada 2001"