4 VARNISH AND GLOSS

One of the initial goals of this study was to try to make sense of the views of American tonalist and impressionist painters on varnishing. We found that varnishing practices at this period appear to vary from one artist to another and from period to period within some artists' careers. Similar variations on the part of French impressionist painters were described in a recent study by Michael Swicklik (1993).

It seems clear that many American tonalist paintings were intended to be varnished, especially the darker, more Barbizon-influenced ones. This is the case with the 33 paintings in the dining room of the Florence Griswold House. According to oral tradition, the artists themselves “went over” the paintings with varnish at intervals, and in fact when we treated the paintings we found that many of them had complicated varnish histories.

Henry Ward Ranger, the leader of the tonalists, was a strong advocate of varnishing paintings. For varnishing he recommended mastic resin, applied in two thin layers rather than one thick one (Bell 1914). In 1914, he derided the fashion of 30 years earlier for matte paintings by painters who imitated Monet. In fact he made efforts to get owners to varnish previously unvarnished paintings, including paintings by Monet, because in Ranger's view the paintings would look better and be better protected from grime and fading. To support his position, he argued that the dealers Durand-Ruel had begun some years previously to varnish their Monets, and he wrote that “these pictures have also gained by the treatment” Bell 1914, 75).

In 1916 Dwight Tryon wrote in a letter that Autumn Night (1916, Freer Gallery of Art) was so dense with oil glazes that he had to wait several weeks before he could apply a final varnish (Merrill 1990). But there are other paintings by Tryon that were not varnished, perhaps by chance or perhaps because the paint was already fairly glossy because the paintings had been oiled out or were painted with a medium-rich paint. Henry C. White (1930) explains how Tryon could have ambivalent opinions about gloss: while Tryon liked the matte effect of paint which was applied directly to wood, he said its brilliance could be restored later by varnishing.

A study of the large collection of pictures by the three generations of the White painters still owned by the family shows that some are varnished, while others have never been varnished. Nelson H. White thinks that his grandfather had no fixed policy on varnishing. He remembers looking at paintings with his father and discussing each one separately, asking whether it would benefit from varnishing or not. Nelson H. White still uses this process to judge whether to varnish his own paintings; he feels that some types of paintings may need varnish more or less than others—a foggy, midtoned scene, for instance, may look better with less gloss.

Such case-by-case decision making about varnishing might be closer to the way that choices were made 100 years ago than many conservators suspect. Those who consult written references only might reach inaccurate conclusions about a painter's varnishing practices.

With the American impressionists the question of varnishing is even more interesting. Did many of them intentionally avoid varnishing in order to achieve “that dead, dull, lackluster, nontransparent look to the surface so much prized by some modern painters, who take special pains to bring it about” (Abendschein 1909, 63), or did they wish to avoid the distortion that a discolored varnish would give to their colors?

Looking at the paintings themselves, a conservator is faced with interesting and sometimes conflicting evidence. In the collections of some museums, usually smaller museums where many of the paintings were donated by the artists, descendants of the artists, or first owners, we have found many unvarnished paintings. On the opposite extreme, in a survey of a corporate collection in which most of the pictures were purchased recently from major dealers, we found that there was a much higher percentage of varnished paintings. We have also noticed that paintings by lesser-known artists are more likely to be unvarnished than works by better-known artists, perhaps because the less well known works have rested quietly in storage for many years.

It is interesting that Roger Dennis feels strongly that the American impressionists he knew did not varnish their pictures and that varnish on an American impressionist painting reflects a decision by a dealer rather than the choice of the artist. But paintings by several artists contradict this observation. For example, we have seen a number of paintings by William Chadwick that have been varnished and then were retouched by the artist over the varnish. In these cases it is absolutely certain, based on the evidence of the paintings themselves, that the artist intended to have these paintings varnished.

A search for written evidence of varnishing on the part of the American impressionists uncovered very few references, except for the case of Childe Hassam, who wrote to his dealer in 1919 about his flag paintings: “I will not let them go anywhere until I have varnished them all this fall or early winter” (Hassam [1919]).

Other artists are silent about the varnish question, and their silence tends to reinforce Roger Dennis's recollection that the artists he knew rarely talked about varnishing. But Dennis believes that artists often added medium to their paint so that they would not have to apply varnish. They wanted to avoid varnishing first of all because they were afraid of the yellowing of the varnish, but they were also conscious of the fact that varnishing, by saturating the colors slightly differently, could change the subtle color and value relationships they saw on their easels.

Roger Dennis believes that, unlike Monet, American painters did not want a dead-matte effect. And in fact when unvarnished paintings have been protected from the effects of light and air by glass or by a wide frame rabbet, the paint is often surprisingly glossy. This glossiness helps to confirm what Dennis described: that the look of the paintings when they were new may have been glossier than is now seen and that a proportion of the matteness of unvarnished paintings may come from an accumulation of grime and from deterioration of their surfaces.

It is well documented that the French neo-impressionists sometimes put their paintings behind glass instead of varnishing them (Herbert 1968), and we have found written evidence that Frank Benson, Bruce Crane (1857–1937), Carleton Wiggins (1848–1932), and Guy Wiggins all asked their dealers to have certain pictures framed behind glass at one time or another (Benson 1926; Crane 1920; C. Wiggins 1922; G. C. Wiggins 1922).

John Twachtman is the only painter we know to have been documented as wanting some of his paintings to look matte. One of Twachtman's students wrote that during the 1890s, when Twachtman was working in Connecticut, “he [Twachtman] deliberately avoided an unctuous, varnishlike effect and would frequently expose his pictures to sun and rain to relieve the pigment of superfluous oil and thus produce a uniform mat or dry surface” (Clark 1924, 58). Twachtman's preference for matteness is especially interesting because he began his studies with Frank Duveneck (1848–1919), who apparently liked very glossy varnishes. In 1879 one of Duveneck's students reported that Duveneck varnished his paintings with “coach varnish which made them shine like mirrors and was the despair of Doll [Duveneck's dealer] who could not get it off” (Duveneck 1970, 79).

John Douglass Hale reported that Twachtman's son confirmed the story about his father leaving paintings out in the sun and rain (Hale 1957). It is difficult for a conservator to believe that many of Twachtman's paintings would have survived if they became rain soaked, although, like Whistler, he may well have put his paintings outside in fair weather to speed the drying of the oil. It is also difficult to explain the contradiction between Twachtman's desire for a dry surface and his addition of mastic resin to his paint, which would have made it more glossy.

A number of Twachtman's Connecticut paintings have remained unvarnished, and they often have a matte appearance, but there are also a few that have been varnished for many years. My House, discussed earlier, had a very old overall varnish, as did The Cascade (ca. 1890s, Newark Museum of Art) (Spanierman 1987).

Horseneck Falls (ca. 1890s, private collection), one in a series of paintings of the waterfall on Twachtman's property in Greenwich, has a moderately matte surface and has never been varnished. It is painted over an ochre-colored imprimatura, which Twachtman often used during this period, that lies over a white ground, which in turn lies over another design. (Twachtman frequently painted over recent compositions [Hiesinger 1991)]. In a number of places the upper painting and second ground have flaked off, revealing portions of the first composition. The paint that was used in the first painting is glossy and has a solubility indicating that resin was mixed into it.

Many of Twachtman's paintings of this period are matte, but the exact degree of matteness intended by the artist is difficult to judge 100 years after they were painted. Apart from any possible intentional aging of the surfaces by Twachtman himself, the intervening years can make an unvarnished surface look more matte. The contradictions between the written evidence that describes Twachtman's preference for a matte final effect and his apparent addition of a resin to his paint also make it difficult to draw more specific conclusions about Twachtman's preferred final surfaces.