News Articles Library Event Photos Contact Search


Sunday, October 22, 2017

Weathervane Finish Analysis

Waethervane Finish Analysis by Jennifer L. Mass and Julie Lindberg
by Jennifer L. Mass and Julie Lindberg

Weathervanes have been part of America’s architectural landscape for hundreds of years. Used to indicate wind direction and as architectural ornaments in both secular and sacred contexts, they have their origins in ninth-century Europe.1 The desirability of weathervanes to folk art collectors is driven by form, history, surface, and rarity. As collectors evaluate possible purchases, they consider many factors; the overriding one of which is generally visual appeal.

The aesthetic impact of a weathervane is often related to the weathered surface, which takes on various appearances, whether verdigris, gold leaf, yellow gilder’s size,2 or—most appealing—a combination of all three (see sidebar). An historic surface does not necessarily mean that which was applied at manufacture. It also relates to what occurred during the natural life of the vane. As vanes were often damaged, repaired, resurfaced, and returned atop buildings, the resulting multiple layers of paint and gilding denote a rich history. In addition, the blending of the surface with the elements resulted in a mellowed, worn appearance. It is to be expected, then, that outer layers of material will be chemically different from those applied earlier, and by analyzing these layers we can attach an approximate age to the different surfaces.

Weathervanes have been part of America’s architectural landscape for hundreds of years. Used to indicate wind direction and as architectural ornaments in both secular and sacred contexts, they have their origins in ninth-century Europe.1 The desirability of weathervanes to folk art collectors is driven by form, history, surface, and rarity. As collectors evaluate possible purchases, they consider many factors; the overriding one of which is generally visual appeal.

The aesthetic impact of a weathervane is often related to the weathered surface, which takes on various appearances, whether verdigris, gold leaf, yellow gilder’s size,2 or—most appealing—a combination of all three (see sidebar). An historic surface does not necessarily mean that which was applied at manufacture. It also relates to what occurred during the natural life of the vane. As vanes were often damaged, repaired, resurfaced, and returned atop buildings, the resulting multiple layers of paint and gilding denote a rich history. In addition, the blending of the surface with the elements resulted in a mellowed, worn appearance. It is to be expected, then, that outer layers of material will be chemically different from those applied earlier, and by analyzing these layers we can attach an approximate age to the different surfaces.

Weathervane Surfaces: Metal Preparation Layers
The stratigraphy or layer structure of a gilded weathervane surface begins as follows:

1.) The copper alloy vane substrate is coated with an oil-based primer, often a lead-based paint such as red lead (Pb3O4) or lead white [2PbCO3.Pb(OH)2 ].

2.) The next layer is a bright yellow gilder’s size, used so that any voids or losses in the gilded surface are not readily apparent. This bright yellow layer is commonly comprised of the vivid yellow pigment chrome yellow, also known as lead chromate (PbCrO4), which has been available since the first decade of the nineteenth century.

Gold Leaf<br>Oil or oil resin gilder's size<br>Yellow ochre-based gilders's size<br>Oil-based prmer<br>Copper alloy vane
3.) The second yellow compound found in a traditional yellow gilder’s size is yellow ochre, a mixture of clay minerals and a yellow iron hydroxide mineral goethite [FeO(OH)]. The Gilder’s Manual (1876), describes this preparation as “gold size” made up of “a mixture of boiled linseed oil and ochre, well ground up together.”12

4.) The binder for a gilder’s size is traditionally a drying oil or plant resin, but shellac (an insect resin) has also been observed.

5.) There are many variations to this basic layer structure, but the gold leaf was typically applied directly over this yellow size, or over a second layer of size comprised of a plant resin or drying oil.

The technologies used to finish metal objects are different from those used on woodwork or furniture since surface materials designed to withstand weathering for decades are chemically different from surfaces intended for interiors. These materials have changed over the centuries, so in order to determine when a surface was laid on a vane, an understanding is needed of what was used and when it was available.

To date, unlike furniture, there has been no systematic examination of weathervane finishes, nor has there been an attempt to understand the makeup of these often intricately layered structures. We conducted a study of thirteen weathervane finishes, subjecting them to molecular and microscopic analysis, in order to gain a greater comprehension of their finish materials and stratigraphy (layer structure). One of the driving forces for this research was our desire to address the common misconceptions about weathervane finishes, such as: how long does gilding survive on architectural ornaments exposed to the elements? Were weathervanes typically stripped before refinishing? What are the effects of marine and inland environments on verdigris surfaces?

Regading the first, there is a common belief among collectors, that original gilding on a vane would have completely weathered away within the first five years. In fact, gilded surfaces can survive, albeit weathered, between thirty to fifty years with the right surface preparation of the copper alloy—typically a primer of an oil and lead-based paint.3 The importance of proper surface preparation both before and after the gold leaf is applied is emphasized in The Gilder’s Manual (New York, 1876): “...no gilding exposed to the extremes of summer and winter, wet and dry, cloud and sunshine, should ever be varnished.” The Manual also recommends coating copper with a layer of shellac before leaf gilding, and using an oil gilder’s size for exterior gilded ornament.4

An additional misconception about weathervanes is that they were always stripped before re-gilding. While this was certainly true for some instances, several of the vanes we examined had rich finish histories showing multiple gilding campaigns as well as gilding over polychromy and polychromy over gilding. As to the effects of environmental factors, that will be discussed shortly.

A second motivation for our research was to examine treatments used for artificial aging including acids; painted-on verdigris-like corrosion layers; modern gilding distressed to appear weathered; or painted-on yellow finishes intended to imitate gilder’s size. In some cases, gilded surfaces have been made to appear aged by the use of gold flakes applied to the vane with a paint binder such as a natural or synthetic resin. One does see numerous examples of weathervanes that have been entirely re-surfaced with bronze paint. Bronze paint is actually a paint prepared from brass (a copper and zinc alloy) flakes suspended in a slow-drying medium, typically a natural oil or plant resin. While this material is often associated with later finishes or restorations, and considered undesirable because if uncoated it rapidly dulls/tarnishes, bronze powder paint was in fact mentioned by Stalker and Parker in their 1688 Treatise of Japanning and Varnishing, and so it would not be impossible to encounter this as an early finish or even an original finish on an eighteenth- or nineteenth-century vane.5

All of the vanes in this study were crafted from copper or copper alloy. The beautiful matte verdigris surface observed on many vanes is copper corrosion from the vane’s metal substrate, which forms over years of exposure to atmospheric corrosion agents such as sulfur dioxide, which is converted to dilute sulfuric acid in urban environments. This leads to the formation of copper sulfates such as brochantite and antlerite. Marine environments will contain high concentrations of chlorides from sea spray, leading to the formation of copper chlorides such as atacamite and paratacamite.

In our methodology for gathering samples, we removed an average of six pinhead sized flakes from each weathervane’s finish, taken from unobtrusive locations using a scalpel; this was done to attain a complete finish history. All of the areas selected (such as the inside of a horse’s ear) are those exposed to minimal weathering and so should have the most complete finish histories. It is also important when sampling to avoid areas prone to restoration (such as seams and where a tail is joined to a horse), so that these are not mistaken for later finish campaigns. Each sample was mounted in resin, polished, and viewed in cross-section under high magnification using ultraviolet light, visible light, and, when higher magnifications were required, an electron microscope to “read” the layers of surface applications. Further analysis included subjecting the samples to molecular fingerprinting techniques such as Fourier transform infrared spectroscopy (FTIR).

Cross-section microscopy and molecular analysis have proved to be invaluable tools for identifying and evaluating these surfaces, but they must be combined with research into historic forms and the connoisseur’s eye for a complete understanding of each piece.

Case Study 1: Goat weathervane
Period appropriate finish history and materials
Fig. 1: Goat weathervane, form dating to ca. 1875, unknown maker, possibly Massachusetts, in a gilded finish with verdigris corrosion visible on the upper surface.  INSET: Fig. 2: Goat finish stratigraphy—visible and ultraviolet light images.
Fig. 1: Goat weathervane, form dating to ca. 1875, unknown maker, possibly Massachusetts, in a gilded finish with verdigris corrosion visible on the upper surface.

INSET: Fig. 2: Goat finish stratigraphy—visible and ultraviolet light images.

This weathervane (Fig. 1) was found to have a rich and extensive finish history with four gilding campaigns in evidence; an expected result for an architectural ornament exposed to the elements. All of the yellow sizing layers found in the vane’s finish history were prepared with period-appropriate materials, including lead chromate yellow, and all of the metal leaf layers were gold.

The use of a shellac-containing size in the first gilding campaign (identifiable by its orange fluorescence in ultraviolet light) (Fig. 2) is common for outdoor architectural gilded finishes, and is also mentioned in the period literature.6 All of the non-pigmented gilder’s size layers fluoresce in ultraviolet light, confirming the use of natural plant and insect-based resins. These are period-appropriate technologies that one would hope to observe in a gilded architectural ornament more than 140 years old. Some of the materials identified in the gilding campaigns allowed us to identify the dates after which the campaigns were applied. For example, the first gilding campaign contains a barium-based pigment, barium white, which was introduced after 1820.7 The identification of both barium and zinc in the primer of the second gilding campaign suggests the presence of lithopone in this layer, a pigment first produced in 1874.8

Case Study 2: Cock weathervane
Period appropriate finish history and materials
Fig. 3: Cock weathervane with gilded surface and visible small patches of verdigris corrosion, ca. 1880. A popular form, rooster vanes were manufactured by Cushing and Sons, Waltham, Massachusetts, as well as by several other firms.  INSET: Fig. 4: Cock finish stratigraphy—visible and ultraviolet light images.
Fig. 3: Cock weathervane with gilded surface and visible small patches of verdigris corrosion, ca. 1880. A popular form, rooster vanes were manufactured by Cushing and Sons, Waltham, Massachusetts, as well as by several other firms.

INSET: Fig. 4: Cock finish stratigraphy—visible and ultraviolet light images.

This weathervane (Fig. 3) has as many as five separate finish campaigns, including four layers of gilding and one layer of bronze paint (Fig. 4). The first gilding campaign has a traditional lead chromate and iron ochre yellow gilder’s size in an oil-based binder applied over an oil-based priming layer. The second campaign is a black-pigmented oil-based finish, followed by gold leaf. Black paint was commonly used for weathervane finishes because of the dramatic appearance of the black finish against the bright blue sky, although it does not appear that this black was intended to be a presentation surface. The third finish campaign includes a layer of natural resin priming (see the bright fluorescence of this layer in figure 4) followed by an oil-based yellow gilder’s size, a thick natural resin-based red lead priming layer, and then a third layer of gilding. This use of a red lead primer after a yellow gilder’s size is unusual, and may have been done to provide extra weather-resistance and longevity to the subsequent gilding. The fourth finish is a traditional campaign of a yellow gilder’s size followed by a second natural resin-based size and then a layer of gold leaf. Finally, a thick layer of yellow gilder’s size is followed by a layer of bronze paint, easily identifiable by the flaky nature of the layer in cross-section, as well as its reddish color in comparison to the gold leaf. A drying oil such as tung oil was also identified as the paint binding medium in the first two gilding campaigns. Tung oil was an ideal choice for a drying oil on an exterior architectural ornament because it is strong, tough, waterproof, and also wear-resistant.9 These findings are consistent with vanes prepared in the nineteenth century (chrome yellow was introduced in the first decade) that has been subjected to substantial use/weathering and repeatedly refinished.

Case Study 3: Angel Gabriel
Later surface, no finish history, and a possible artificial patination
Fig. 5: The angel Gabriel in flight and blowing a trumpet, form prior to 1900. Surface gilded with some blue/green, possibly artificial patina.  INSET: Fig. 6: Angel Gabriel finish stratigraphy—visible and ultraviolet light images.
Fig. 5: The angel Gabriel in flight and blowing a trumpet, form prior to 1900. Surface gilded with some blue/green, possibly artificial patina.

INSET: Fig. 6: Angel Gabriel finish stratigraphy—visible and ultraviolet light images.

This angel Gabriel weathervane form should date to prior to 1900 (Fig. 5). However, none of the materials found on the surface of either the horn or the angel are consistent with period practice. The metal leaf was found to be brass (also known as Dutch leaf) rather than gold. The period literature recommends not using this material because of the rapid dulling/tarnishing that would result from outdoor exposure of this surface.10 The gilder’s size was found to be polyvinyl acetate, a synthetic resin, rather than a natural resin or drying oil such as pine resin, tung oil or linseed oil (the period gilder’s size formulations). In addition, there was no evidence for a yellow gilder’s size layer, nor was there evidence for any previous surface finishes.

The sizing layer beneath the gold, polyvinyl acetate is a synthetic resin that was not invented until 1912.11 As such, this resin, a primary component in Elmer’s glue, cannot represent the original, pre-1900 gilding. It is, however, a hard resin with good water resistance and weathering properties, which might explain its use in this context. To prepare the finish, a thick layer of polyvinyl acetate was applied over the surface of the vane, and then brass leaf was applied to this layer while wet. The brass leaf appears to be in the resin layer as well as on top of the resin layer (Fig. 6). The verdigris corrosion layer on the vane was found to be the copper chloride atacamite. Copper sulfate-based patinas are the most commonly observed natural corrosion layers for outdoor architectural ornament, although evidence of copper chloride corrosion can be expected (often in conjunction with copper sulfates) in coastal towns due to saltwater spray in the air. The copper chloride patina finding, in conjunction with an out-of-period gilding technology and the lack of any copper sulfates, however, suggest that this vane may have been patinated with an acid treatment. While the identification of copper chlorides does not conclusively prove artificial patination, we have been able to identify copper nitrate- based patinas on a number of similarly- decorated vanes, the result of an artificial/chemical patination rather than a natural weathering.

Conclusions
Of the thirteen analyzed vanes, our study revealed several features of weathervane finishes that are indicative of an appropriately aged surface. One of these is an extensive finish history. Two or more finish campaigns not only provides a vane with a rich aged appearance, but also acts as a valuable historical document of gilding technology. Other features of period finishes include the use of natural drying oils and plant resin binders for the sizing layers, lead chromate and yellow ochre pigments for the gilder’s size, and shellac or oil-based priming layers applied to the metal substrate.

The majority of the vanes examined were found to have evidence of a traditional, historic yellow gilder’s size (based upon the identification of chrome yellow, yellow ochre, kaolin clay, or a combination of these materials). One vane was found to have a faux yellow gilder’s size (a yellow acrylic paint colored with a synthetic dye). Four vanes had copper sulfate-based corrosion surfaces resulting from environmental exposure, while another four had a copper chloride-based corrosion surface in the form of atacamite. As noted above, while copper chloride-based finishes could be the result of marine weathering, in these instances it was found on vanes that also had modern synthetic materials applied to them, suggesting that the copper chloride might have been a result of an artificial patina.

The vanes with modern finishes show evidence of modern materials such as polyvinyl acetate, which was used as a gilder’s size; gold flakes suspended in acrylic or urethane polymers used to imitate a naturally aged gilding; and a yellow acrylic paint used to imitate a period gilder’s size. In one case an acrylic resin coating had been applied over a copper sulfate-based verdigris finish. This coating may have been a restoration treatment applied to preserve a powdery surface, and so one must be cautious in the interpretation of modern materials when they are identified. In addition, many restorations to correct damage are carried out with period gilding materials to blend the aesthetic appearance, hence they will not always be identifiable with molecular analysis. In this case the restoration layer may be identifiable by its position in the stratigraphy, if it goes over a bullet hole repair, for example. It may also be identifiable because it is only found in one location on the vane’s surface.

There can also be circumstances in which a vane’s surface is correct but the form is modern; achieved by means of using a naturally weathered piece of copper to construct a new vane. Conversely, vanes with period copper substrates and modern finishes also occur. As has been noted, while not all weathervanes were stripped prior to re-gilding, some undoubtedly were, resulting in weathervanes that, while period, have only one finish campaign.

Bullet holes and repairs are another feature often present in weathervanes. In cases where such repairs exist, it is important to determine whether the finish was applied before or after the repair. Also examine the angle of entry and exit. Shot from the ground, the bullet would enter and exit at a steep angle, and not pass straight through as if shot point blank—unless the vane was being used as target practice on the ground. There should also be weathering to the edges of the hole. Check also the soldering on the seams—the absence of at least one split seam on a period vane is possible but highly unusual.

This work represents our initial attempt to understand these beautiful and complex finishes. Our results provide a better idea of the types of surfaces one encounters and reveals some of the most important features of each. Cross-section microscopy and molecular analysis have proved to be invaluable tools for identifying and evaluating these surfaces, but they must be combined with research into historic forms and the connoisseur’s eye for a complete understanding of the piece. There are certainly more questions to answer in the future, and other promising analysis techniques to explore (such as metallography to examine how the corrosion layers interact with the copper alloy substrates). The beauty and excitement of finding that perfectly aged surface, whether of verdigris, gilder’s size, polychrome, or gold, is certain to keep us interested in the functional, sculptural, ever-present weathervane.


Jennifer L. Mass PhD, will will present her findings on vane surface analysis on August 13 at the Rufus Porter Museum in Bridgton, Maine, in collaboration with an exhibit on weathervanes. For information visit www.rufusportermuseum.org.

The complete analysis and study will soon be available at www.winterthur.org.


Jennifer L. Mass PhD, is the Senior Scientist at Winterthur’s Scientific Research and Analysis Laboratory and teaches in the Winterthur/University of Delaware Program in Art Conservation. Julie Lindberg is an antiques dealer specializing in folk art and weathervanes. She is a founder and curator at the Rufus Porter Museum in Bridgton, Maine.


1. Robert Charles Bishop and Patricia Coblentz, A Gallery of American Weathervanes and Whirligigs (New York: Bonanza Books, 1984), 11.

2. Size is a tacky and viscous coating that acts as an adhesive and substrate for gilding.

3. Michael W. Kramer, “Architectural Gilding on Exterior Metal: An Overview of Materials and Methodology,” in Gilded Metals: History, Technology and Conservation (London: Archetype Publications, in association with the American Institute for Conservation of Historic and Artistic Works, 2000), 351–361.

4. The Gilders’ manual Picture Frame Gilding, and Gilding for Interior Decoration. Preparations Used in Gilding (New York: Jesse Haney and Company, 1876).

5. John Stalker and George Parker’s Treatise of Japanning and Varnishing (London, 1688) (Reprint: Tiranti Publishing, London 1971).

6. The Gilder’s Manual, 26.

7. R. Feller, “Barium Sulfate,” Artists Pigments, Vol. 1, R. Feller (ed.) (Washington DC: National Gallery of Art, 1986).

8. Nicholas Eastaugh, Valentine Walsh, Tracey Chaplin, Ruth Siddall, Pigment Compendium, Vol. 1 and II, (Oxford:Elsevier Butterworth-Heinemann, 2004).

9. R. J. Gettens and G.L. Stout, Painting Materials, A Short Encyclopaedia (New York: Dover Publications, 1966).

10. Franklin B. Gardner, The Painters’ Encyclopædia (New York: M.T. Richardson, 1887), 37.

11. C.V. Horie, Materials for Conservation: Organic Consolidants, Adhesives and Coatings, (Oxford: Butterworth-Heinemann, 1987).

12. The Gilder’s Manual.

Events