The role of adhesives in many wooden objects is a critical one for both fabrication and restoration. For that reason understanding the characteristics of adhesives, particularly in the areas of application and manipulation, stability and deterioration, is fundamental to the preservation of most wooden objects. This understanding should be universal for all wooden artifacts, from panel paintings to aircraft, although applications of both process and materials are specific to the artifact being conserved. This paper will provide a brief review of adhesives as they pertain to wooden objects, and the conservation treatment of wooden objects whose element(s) have undergone structural damage, either from fracture of the solid wood or failure of the adhesive system, and the selection and use of adhesives during conservation treatments.

When reviewing the selection, properties, and use of adhesives for wood conservation, it is first necessary to answer the following question - What is the adhesive supposed to do? Equally important is the converse question - What should the adhesive not do? Naturally, this inquiry falls within the larger question of the goals and strategies for any particular conservation treatment, which in turn evaluate any ethical constraints on the conservation practitioner, the nature and needs of the artifact, and the users of the artifact.

There are two broad categories of adhesives commonly relative to historic wooden objects, either as part of their original structure or subsequent restoration attempts; natural materials (usually proteinaceous) and synthetic polymer resins. Protein glues include animal hide glues, fish glues, casein and albumin, although the latter two occur with considerably less frequency than the former pair. Other "natural" based adhesives are natural resins, such as shellac, and rubber-based cements. While the number of synthetic resins used for wood glues is large, conservators encounter only a minute portion of those. Prominent among these are the PVA emulsion and reactive glues such as epoxies, phenolics, urea-based and polyester resins, some of which are in the so-called "water-resistant" or "waterproof" glues. Additional adhesives used by the conservator during the treatment of wooden objects include acrylic resins, both in solution and emulsion, and hot melt adhesives of various formulations and forms.

The following survey of adhesives is not a definitive one, but rather serves to familiarize the caretakers of wooden artifacts with the variety of materials encountered in their care. As such it is a brief review of historically important adhesives and traditional and contemporary adhesives used in conserving wooden artifacts. The information presented here is not particularly technical in nature but is a synthesis of observations on the part of conservators who must evaluate adhesives properties and performance. For a more thorough exposition on adhesives and their technology the reader is directed toward the bibliography at the conclusion of this chapter.

Prior to the development of synthetic resin adhesives in the twentieth century the most common adhesive for wood, indeed the glue dominant almost to the exclusion of all others, was protein glue or animal glue. There are a number of glues which fit in this category, that is, proteinaceous animal by-products, such as fish glue, albumin and casein, but in the overwhelming majority of wooden artifacts encountered, the protein glue used is animal hide glue. In the interest of brevity, and as a reflection of the preponderance of use, this section will not deal with either albumin or casein glues beyond a brief description. "Animal glue," "hide glue," and "gelatine glue" will be used essentially interchangeably in the following text.

Casein glue, a powder derived from the curds of acidified skim milk, forms a water resistant, nearly thermoset adhesive when re-mixed with water during preparation and application. Exceedingly strong, casein continues in use for architectural laminae, and was used in the context of panel paintings to join sections of panels during the original fabrication. Albumin glue, derived from blood proteins, is also a proteinaceous water-resistant glue used since anitquity. For the ancients, the coagulating process, which drove the adhesion, required the use of fresh blood for adhesion. However, when the processing of fresh blood to dried blood glue was discoverd early in the 20th century, a more widespread use of this adhesive was possible. It's primary utility was as a water-resistant heat-activated adhesive for the plywood intustry, especially in the fabtrication of early wooden airplanes. However, because of its prominance as an adhesive for the plywood industry, this thermoset glue is very often present as the binder in early plywood panel substrates for paintings.

Animal hide glue is a gelatin adhesive ("animal hide," gelatine, and "rabbit-skin" glues) extracted from the hides, hooves and sinews of mammals, mostly horses and cows. Through a heated aqueous extraction process the protein collagen is removed from the hides and then processed and purified to form gelatin, or glue.[i] Because protein molecules are broken by heat, the temperature at which the collagen is extracted plays an important role in the characteristics of the adhesive. Collagen extracted at lower temperatures has a higher molecular weight and is stronger than collagen obtained from processing at higher temperatures. The designation of this characteristic is referred to as the "gram weight strength" and is assigned by testing the gelatin for deflection under stress. This testing determines the weight necessary to depress the surface of a "glue jelly" a specific amount according to a rigorously controlled protocol.[ii] In general, the gram weight strength of glues normally used for woodworking is in the range of 200-300, although the range available is much broader, i.e. <100 - >400. Specific types of animal hide glues, such as rabbit-skin or pure gelatin, are widely used for particular functions, such as gilding size.

Fish glue, another traditional wood adhesive, is also collagen glue derived from the fish bladders and other by-products of processing fish for consumption. While the collagen derived from fish is very similar to that obtained from horses and other mammals it tends to have a lower molecular weight and is therefore weaker and more easily soluble.[iii]

The procedure for preparing and using gelatin glues is based on the thermal and solubility properties of collagen, which is a thermoplastic, water soluble material (it is relatively unaffected by organic solvents not miscible with water). Hide glues are obtained commercially in either dry or liquid form. Dry glue can be powder, granules or sheets. Regardless, the initial step in preparing hot hide glue is soaking the dry glue in cold water to swell the material and allow the subsequent solution to be made. Longer protein chains (higher "gram strength") absorb far more water per given mass than do shorter (lower "gram strength"). So, for a given viscosity, a low-gram-strength solution will have a higher solids content than will a high-gram-strength mixture, and conversely, a high-gram-strength glue will be more viscous than a low-gram-strength mixture of the identical sold:liquid ratio.

The exact proportions of dry glue to water thus depend on the glue grade and the circumstances for using the glue. Even if the initial recipe proportions are followed, the loss of water by evaporation can alter the mixture, which must then be modified with the addition of more water. In the final analysis the mixture and preparation of glue is usually inexact and depends on the practitioner as much as anything else.

Once the dry glue has soaked an adequate time (all of it has swelled), it can then be heated to a usable liquid form. This is accomplished in a double boiler or a glue pot at a temperature warm enough to liquify the glue but not hot enough to break down the collagen too rapidly. (This latter concern is the reason glue should not be reheated too many times. Only experience will tell the user how many times and at what temperature a pot of glue can be re-used.) For most uses glue is heated to and used at approximately 140*F. Once the glue has reached the proper temperature and viscosity it can be applied to the gluing surfaces. When the gluing procedure is complete the glue pot is allowed to cool and the prepared glue can be reused at a later time (a few hours or days at the most unless refrigerated or modified with preservatives). To prevent mold growth the prepared glue should be stored in a refrigerator if not needed for more than several hours. Traditionally new glue was mixed every day in the woodworking trades.

Hot hide glue dries and hardens in a two-step process. The first step is a transition from liquid to gel, and occurs once the glue has cooled to the gel-point of the mixture. Depending on the glue grade, concentration, and ambient temperature, this can take from a few seconds in a cool room to several minutes if the gluing surfaces are heated. Following the initial gelling the semi-solid mass of glue hardens through loss of water to the surrounding substrate or to the atmosphere. As with all solvent-release processes, the dried glue volume will shrink equal to the solvent loss. With complete hardening the glue is a very hard, durable material which can function as a bonding agent for extremely long periods of time in the proper environment.

The preparation of cold liquid hide glue is similar to that of hot glue with the exception that it is modified with gel suppressants (e.g. salt, urea), which prevent it from solidifying at room temperatures.[iv] Cold hide glue hardens only by the loss of moisture, or, the second step of the sequence for the hot glue, with its incumbent volumetric diminution through water loss. The properties of dried cold glue are similar to those of the hot glue. Commercially prepared cold hide glues are available, but there is no way of knowing the strength of the gel mixture, the solids content, or most important, the shelf life. Gel suppressants reduce greatly the shelf life of the glue mixture. This is probably due to an interaction between the modifiers and the protein molecules which may break the chains reducing their molecular weight, preventing their forming a solid mass. It is common to find commercial cold glues becoming useless in a very short time.

Other modifications of animal glue include the addition of plasticizers, usually glycerine or sorbitol up to 50% by dry weight, for flexibility and increased "tack," and the addition of formaldehyde to yield a water-resistant thermoset adhesive.

There are numerous characteristics about hide glues that should be considered when evaluating either their use or the treatment of objects on which they have failed. Many of these properties can be beneficial or deleterious depending on the exact circumstances involved.

The first set of characteristics, and probably the most important reason that hide glues are so widely used in the conservation of wooden artifacts, is that due to their water soluble thermoplastic nature, they are almost completely reversible. For many fabricators of wooden objects this reversibility is not a factor and the glue is used for other benefits, such as strength, ease of use, and availability. Nevertheless, for the conservator reversibility is a key component. This becomes manifest in two principle areas. Treating damaged or disassembled glue lines, which were originally formed by hide glue, by manipulating and reforming the original material may be possible since it was thermoplastic when applied and may still remain so. If this is not a viable treatment option this characteristic is still important because it makes the removal of old glue and cleaning the gluing surface considerably easier. The second application of this principle is the retreatability, discussed at length elsewhere in this chapter.

The sometimes brief working time due to the gelling properties of hot hide glues prescribe a specific approach to using this adhesive when assembling compound wooden objects. Since the set time can seem rapid, each step in the process must be thought through in advance of initiating the gluing. The brief working time is a drawback for some circumstances and advantageous in others, but in general it is not an undue burden either way, and can be manipulated by heating the gluing surfaces, glue grade selection (higher grades take longer to gel), thinning the glue with water, or adding gel suppressants. Should the glue gel before the work is completed it can be re-liquified by the application of heat from a coil or heat gun.

If the working time of hot glue is prohibitively brief for a particular project the gluing can be done with cold liquid hide glue. The formulation of cold glues provides a working time of up to an hour, which allows very complex procedures to be completed successfully. However, as previously alluded to, if the shelf life of the glue has been exceeded, the adhesive may not harden properly. A further caveat explicitly mentioned on most containers of cold glue is a strict adherence to specific temperature guidelines. If the glue is used in an atmosphere with an ambient temperature below approximately 70* F (depending on brand and formulation) the glue may not harden.

The structure of animal glues suggests a true chemical affinity for wood.[v] Thus, their adhesion is excellent to a wooden substrate. However, an opposing characteristic is that they can be extremely brittle when hard and can easily fracture; they do not function as a particularly cohesive material, particularly in thick applications. Glycerin or other water soluble plasticizers can be added to the glue to increase their flexibility and functional cohesion, but the best solution to this problem is to simply have the thinnest glue line possible.

One final note is that hide glue is hygroscopic and its stability and properties are greatly altered with a change in environmental moisture. If the moisture level drops too low the glue becomes extremely brittle and can be fractured with very little applied stress, thus leading to failure of the bond line. If the humidity rises the glue softens and is susceptible to plastic deformation. In addition, an extreme rise in moisture almost assures an attack of fungi to the surface of the glue. As a protein based material it is an excellent nutrient for much animal and plant life.

The use of natural resins as adhesives is not prevalent but nevertheless encountered in historic objects.[vi] Probably the most extensively used resin was shellac, a thermoplastic exudate of the bug Laccifer lacca, indigenous to India and Indochina. The resinous exudate is refined by a number of processes including heat, solvent, and aqueous extraction, resulting in an amber/orange material of varying purity and composition depending on the specifics of the refining process. (From the experience of the author the primary use of shellac was to glue non-wood veneers to a wooden substrate.)

Shellac is normally used in solution with alcohol and dries by solvent evaporation, although it can be used as a pure material liquified by heat only. As a thermoplastic it is relatively incapable of resisting thermal or chemical attack, but under the proper conditions it can remain stable indefinitely.

A review of this category of adhesives provides a segue into the next section on synthetic adhesives because contact adhesives can (and have) include(d) both natural and synthetic rubbers in solution. Prior to the synthesis of neoprene the use of rubber-based adhesives was limited due to the cost and availability of natural rubbers. The use of these adhesives became more common with the development of "better" neoprene materials following the second World War. Contact cements are usually very sticky and used extensively to adhere laminae. Usually the formulation of this adhesive is neoprene or other rubbers in an organic solvent, although there are now formulations which use water emulsions as the vehicle. In use they are applied to both surfaces to be adhered and allowed to dry. When they then come in contact with each other they are strongly bonded immediately, hence the name "contact cement".[vii]

Due to the primary function as laminae adhesive, the most common application of this material is as an adhesive gluing wood or other veneers to a substrate. Or, they may be encountered in the aftermath of a particularly inept repair to structural elements. These adhesives do not appear to be exceptionally stable over a long period of time.[viii] With deterioration the adhesive the result is delamination of the fabricated structure. Being a rubber-based material contact adhesives are thermoplastic and can be softened with heat and/or organic solvents.

The easiest way to sort synthetic resin adhesives is not through the chemistry of the materials, but rather through the delivery system, or physical formulation for use. Within this description lie four major system types; emulsions, solutions, hot melts, and reactives.

The most widely used general purpose wood glues used in the wood crafts today are those based on aqueous emulsions of (poly) vinyl acetate and are commonly called "white glues. The closely related glues generically called "yellow", "aliphatic", or "carpenter's" glues may (or may not) be based on (poly) vinyl acetate. The attribution of these water based emulsions is because of their appearance as opaque white or yellow liquids, although they become translucent when dry. The drying mechanism of PVA emulsion, like that of hide glue, is a two step process.[ix] The emulsion of PVA consists of spherical oligomers of vinyl acetate suspended in water at a concentration which is low enough to prevent the necessary interaction of the oligomer units required for the formation of longer polymers. As the solvent (water) evaporates the concentration of the oligomer solution increases and the distance between the oligomers decreases, allowing the autopolymerization of the material into molecules of greater weight which can become solid. In yellow glues, sometimes called "aliphatic resin glues", either the oligomer or the polymerizing process have been modified to give polymers of greater size, and thus greater viscosity and quicker set time.[x]

PVA emulsions have several distinct advantages. They are easily obtained and used, providing a moderate work time and easy cleanup with water, and have a good shelf life. They require essentially no preparation; the commercially available product is ready to use straight out of the container. Depending on the formulation and the environmental influences they can be durable adhesives which remain stable for lengthy periods of time. They also remain, to some degree, soluble and reversible. White glues can usually be softened with a water/surfactant solution and are generally removable in a variety of organic solvents. Yellow glues usually require an organic solvent to soften and swell them, but for the most part they are considered to be partially reversible. As with all other glues, the best bond line for PVA emulsion is very thin. Due to their thermoplastic tendencies they can deform if the glue line is too thick and the stresses great enough.

The use of PVA emulsion in the fabrication of composite wooden objects is widespread, and as mentioned previously, is the most common adhesive in contemporary woodworking. It is also used with moderate frequency in conservation. The general tendency is to use PVA emulsion to repair a break or fracture in a wooden element, but should the need arise, PVA and other emulsions bond well to hide glues.

In many respects acrylic emulsion adhesives are much like PVA emulsion in appearance, use, and hardening mechanism. While not widely used in the non-industrial fabrication of objects they are used in conservation in the same applications as PVA emulsion. The advantage of acrylic emulsions is that they can be obtained in a wide variety of formulations with specific properties, including molecular weight ranges and solubility characteristics for a hardened film.

The role of a synthetic resin solution is considerably different than that of an emulsion. It is probably fair to say that synthetic resin solution adhesives are not widely used in woodworking. However, solvent based acrylic resins are widely used in conservation as adhesives for other artifact materials (metal, stone, ceramics and glass, etc.).

A wide range of resins is available, and individual acrylic resins (or blends) can possess very specific characteristics. Of these properties the two most important are solvent specificity and long term stability. Resins can be obtained and solutions designed with solvents which will dissolve the resin but will not effect the surrounding areas of the object. As thermoplastic solutions, they dry through solvent evaporation, and depending on the formulation may remain soluble for a long period of time. The stability of certain acrylics has been well documented in conservation literature.

In addition to those already mentioned, synthetic resin solution adhesives include, among others, cellulose nitrate and cyanoacrylate adhesives. While both of these are solvent borne they which dry (solidify) by vastly differing mechanisms.

Cellulose nitrate adhesive is a solution of nitrocellulose and other film forming materials in a mixture of organic solvents appropriate to the composition of the mixture. This adhesive dries solely through solvent evaporation. Cellulose nitrate is not very effective as a bond forming material with wood and therefore is almost never used as a primary adhesive in wooden objects. It is frequently encountered as a remnant of an inept repair. Nitrocellulose is an instable material and is unsuitable for use in the conservation treatment of any wooden artifact.[xi]

Cyanoacrylate hardens through an anaerobic chemical reaction with nitrogen in the atmosphere.[xii] The lack of working time, poor adhesion to wood, and long term stability render it unusable to the wood conservator. They are, however, used frequently in the wooden musical instrument trade to repair cracks in woodwinds and in the assembly or repair of stringed instruments.

Most thermoplastic materials could be classified as "hot melt adhesives" using a broad definition of the term (e.g. hide glue which begins to harden by cooling and shellac and acrylics which can be used as melted resins) but this section will touch on the group of materials specifically designed to be used in a molten state and harden solely by cooling. Hot melt adhesives as defined in this section are of a wide variety of compositions in several categories of raw materials. The formulation of these adhesives can be very specific regarding the properties of the adhesive, not only when solid but also when liquid.[xiii] The temperature to which these adhesives must be heated to flow is considerably above room temperature in many cases, and since these materials solidify by cooling their use is limited to the penetration possible in a very brief period of time.

Hot melt adhesives come in a wide variety of forms, the two major configurations being pellets which must be used with a heated glue gun, and sheets which can be used any where the gluing surfaces can be heated. Glue guns are normally used when there is a need to inject or apply molten adhesive to a surface and sheet adhesive is used between planar surfaces which will be heated and/or pressed.

Hot melts are increasingly important in the industrial fabrication of wooden objects and are beginning to be used in the conservation of historic wooden artifacts. Of the hot melts, those which remain thermoplastic and soluble are excellent options in treating specific kinds of problems such as delamination, and perhaps even in structural applications. For the most part the knowledge of hot melt adhesives within the conservation field has been limited, and little critical study has been made of their long term stability and other properties.

One group of adhesive materials with only modest impact in the woodworking craft but increasing importance in the industrial fabrication of wooden objects are thermosetting multiple-component adhesives. These materials harden by chemical reaction of the various components with each other. Classes of adhesives in this category include urea-formaldehyde, epoxies, phenolics, polyester, urethane and other formulations. Specific characteristic differences between these materials is of little importance for the purpose of this text.

Multi-component adhesives possess great strength in a wide variety of circumstances. They can be virtually impervious to thermal, physical or chemical attack. Due to their mechanism for hardening there are varying amounts of dimensional change from class to class, e.g. epoxies shrink very little while ureas shrink considerably more. As such, they may be good "gap fillers" in either their raw state or modified with bulking agents.

Despite these qualities, the use of these adhesives in conservation is discouraged. By their very nature as cross-linking polymers they are intractable and therefore not easily reversible. Some of these can be swelled with solvents but usually they must be removed mechanically, potentially causing severe damage to the substrate or adjacent surfaces in many cases. In addition, while they resist chemical and thermal attack when relatively new, they begin to break down in a short period of time (a few years). This degradation takes the form of discoloration and fracturing and is particularly severe in the presence of ultraviolet radiation.

The purpose of adhesives is to bond two or more surfaces together. Regardless of the adhesive material, the processes of adhering wooden substrates are similar enough to be considered as one broad subject. In general, when wood is glued together the surfaces of the wood pieces are (ideally) saturated or impregnated with some liquid material which is between the wood surfaces. When this liquid material becomes solid, since the surfaces of the wood have either been penetrated by this material or exhibit a specific affinity to it, the pieces of wood are bonded.

A successful adhesive system for wood, that is, one leading to bonding, is described as having five zones of function, referred to as "chain links."[xiv]

__________________________________

__________________________________   Substrate (adherend)

__________________________________   Adhesive

__________________________________   Substrate (adherend)

			Figure 1.  The structure of the "chain link" concept.

In order for bonding to occur and remain, all five functions must exist simultaneously and permanently. As soon as any one function fails, the adhesive/bonding system fails.

Although this description fits virtually all circumstances for gluing wood substrates, the exact techniques and procedures vary depending on the adhesive, the purpose of the gluing project, and the condition of the wood surfaces. This differentiation becomes more apparent when describing peculiar requirements of gluing wooden artifacts during conservation treatment. For the fabrication of wooden objects, a wide range of options for gluing members together is available. The same is not true for similar problems faced in the conservation of objects, where the goal of the conservator is to intervene or intrude into the object as little as possible, using materials and techniques to stabilize the object and allow it to fulfill its function. A careless or ill-informed approach to this objective will likely lead to the use of adhesives which are increasingly intractable or unstable, which is directly opposite the conservator's attempt to leave the object in better condition, or at least more stable, and easily retreatable in the future.

Regardless of the purpose of the procedure, there are certain common fundamentals. In addition to the five-zone principle, the pieces must be held in place, preferably in intimate proximity, until the adhesive process is complete. Holding these elements together during bonding can be accomplished as a characteristic of the adhesive, but is more commonly done through the use of clamps or other restraining devices.

The gluing process begins with the design and preparation of the gluing surfaces and terminates with the hardening of the glue and the removal of any devices or tools necessary for the assembling. As stated previously, one of the crucial elements to the success of an adhesive bond is the thinness of the bonding line. In practical woodworking terms this means that the joint (gluing surfaces) should be as planar (smooth) and tight (close together) as possible. In fabricating gluing surfaces this is a relatively simple task. When dealing with gluing surfaces on a historic object, where by definition the treatment involves gluing damaged structural elements or re-gluing surfaces which have undergone adhesive failure, the problem can be more difficult. The pieces may no longer fit together well or the gluing surfaces may be damaged or degraded to the point where a re-gluing effort will be unsuccessful. In these instances the substrates must be altered if the gluing is to be successful.

Before attempting to complete such a process it is important to have determined which adhesive will be used and how it will be introduced to the gluing surfaces, how the object will be assembled (and how fast the object must be assembled), how (and how long) the elements are to be restrained while the glue hardens, and how any excess glue squeezed out during the restraint and compression will be cleaned up. Once the gluing surfaces have been prepared, the object should be assembled without adhesive (or, if there is not "re-assembly" but rather in-place re-joining, as in a partial fracture or crack) to assure that the approach taken will work properly. In addition this "dry run" serves to familiarize the individuals working on the procedure with all the steps involved, and becomes increasingly critical with the complexity of the gluing project.

As mentioned earlier, the restraint placed on wooden elements during gluing is usually provided by clamps. In fabrication the only real limits to the clamping force applied to the object are the crushing of the object or its components, or the squeezing out of too much adhesive, resulting in a "starved joint." In conservation treatments, where the conservator desires the least intrusive interaction with the artifact, i.e. lightest clamping force possible to achieve the treatment goal, a different path must be chosen. This is especially true with decorated surfaces, such as the paint layers of panel paintings, which are inviolate.

In most cases of structural repair, if the clamping force is enough to distort or deform the wooden members, it is too great. This influences the selection of clamps and clamping. One popular clamp form among conservators is the cam-action clamp which has a thin metal spine and flexible wooden jaws.

In almost every situation these clamps provide adequate restraint for gluing. Unfortunately there are practical size limitations for cam clamps, so for larger objects, screw drive clamps must be employed. The options available for inventing or adapting clamping systems is nearly limitless, restricted only by the imagination and abilities of the practitioner and the materials available.

In addition to using proper clamps it is important to consider how the clamping forces are distributed to the object. It is imprudent to have the clamps directly on the surface of the object. In such instances the clamping forces are directly translated from the clamp (whether cam or screw) to the surface of the object through the clamp faces. This invariably leads to crushing the surface directly under the clamp face.

This phenomenon can be avoided by reducing the clamping pressure or more commonly (and more effectively) by employing clamping blocks between the clamp face and the object. Clamping blocks are usually scrap wood or acrylic sheet which are fashioned to fit the space between the clamp face and the contour of the object surface. Clamping blocks are frequently referred to as "cauls", particularly if they are fabricated to fit curvilinear or irregular surfaces.

Since clamping forces radiate hemispherically from the axis perpendicular to the clamp face with an effective angle of 45*, the blocks increase the effective regions of clamping.

Thus clamping blocks serve not only to protect the surface from crushing under the clamp but also to distribute the clamping pressure over the clamped region. In addition the blocks can provide correct orientation of glued parts with one another. Often, thin pads of foam or cloth are used under the block to provide further protection to fragile or variegated surfaces. One other function of blocks is to hold the body of the clamp off the surface of the object and allow more accurate and constant placement of the clamp faces on the gluing blocks.

Release papers and other separating membranes are used under and in conjunction with glue blocks. These include waxed and silicone-release paper, and teflon, polyethylene or mylar sheet. Their function is to assure that the glue-blocks function to provide pressure to the surface without being bonded to that surface. By placing a protective sheet between the block and the surface, the excess glue squeezed out of the glue line is prevented from attaching the block to the object. Silicone papers and expanded teflon have an advantage in that they are vapor permeable, facilitating the hardening of solvent release adhesives. There is some concern, however, that the silicone papers may leave a contaminating residue on the surface of the object.

The ideal adhesive for conservation gluing is one that would be perfectly stable over time, easily applied and manipulated, readily removable if further treatment is required later, and form an adhesive bond strong enough to allow the object to fulfill its function yet be weak enough to be the sacrificial boundary in case of applied stress. Obviously no single material fulfills all of these requirements and thus there is no "wood conservation adhesive." Instead the conservator uses a palette of adhesives with known characteristics. The specific needs of the object being treated are considered, and in many cases placed in a hierarchy, and the adhesive most closely matching those requirements is used. Unless the circumstances are extraordinary, the elements are re-adhered with a thermoplastic material. The most widely used adhesives in this laboratory are hot and cold hide glues, with only minor use of synthetic resin emulsion or solvent-borne adhesives. Crosslinking and multi-part adhesives are almost never used as a replacement adhesive when treating joint failure.

The criterion of object use and the structural stresses placed on the object during that use is of particular importance. Clearly the steps required to stabilize one unsound artifact could be considerably different than those required for a similar one with different utility. For example, a panel painting which is sitting in a display easel may have different stresses than one hanging. Or, the grain direction (hence natural potential for strength or damage) of the panel could affect its exhibition or storage orientation. In addition, the object may serve its function indefinitely in controlled circumstances but only for a brief period of time under adverse environmental conditions. It is against this backdrop that the condition evaluation of the object must take place.

The first step in assessing the condition of the object is to conduct a brief visual and tactile examination. If the object is no longer a whole unit, examine the gluing surfaces to determine their soundness. A great deal of information can be gleaned from this exercise, and frequently the assembly and gluing procedure can be appraised. For this problem the three defining questions to be answered are:

• what came apart, or, what are all these pieces and how do they go together?
• how did it come apart?
• why did it come apart?

If the object is still "intact," or at least all the parts are in some way still connected, the focus of the initial examination questions are slightly different. Are portions of the object in the wrong orientation to the remainder of the object? Do any of the parts feel loose? Does the object move under stress? If so, how and how much?

In addition, it is useful to draw distinctions between the types of damage being repaired. Is it a failure of an aged glue line, a split, in which there is separation but still some degree of attachment, or a break with complete disassociation.

At this point the conservator must begin to formulate a strategy to get the object from point A (the current condition of the object) to point B (the desired condition of the object at treatment conclusion). To do this they must synthesize the answers to the series' of questions already posed regarding use of the object, treatment intent, knowledge of materials, and the conservator's ability. One crucial decision is whether or not an "intact" object needs to be disassembled by the conservator in order to enable the treatment to proceed with the greatest chance of success. Of course, the disassembly itself raises certain risks along with any possible benefits. By carefully weighing the possibilities and limitations inherent in this information, the options for treatment are enunciated.

The reader is admonished to remember that treatment descriptions in this chapter are intended to serve as guidelines rather than formulae. Treatments must be designed individually for each specific problem addressed based on assessments by the conservator of the object's condition, the conservator's knowledge of material technologies, and an honest appraisal of their own abilities.

Probably the first thing that comes to mind in the context of fracture is that "something broke," as opposed to "something came apart." Unfortunately, reality can present the conservator with a wide range of problems, and responses, relative to broken wood. Since this paper is especially pertinent to wood panels, some of the more esoteric circumstances deriving from cross-grain breaks will not be discussed, as the author has never seen this evident in wood panels.

Instead, the focus will remain on breaks which are essentially along the grain, or longitudinal, with respect to the wood orientation of the artifact. Longitudinal fractures fall into two broad categories; those which do not distort either the object or the gluing surfaces and can be repaired relatively simply, and those which distort the object and/or damage the gluing surface and require additional treatment prior to the adhesive assembly.

A simple fracture (where there is minimal distortion to the panel and little damage to the gluing surfaces) whether partial or complete, requires only the introduction of an appropriate adhesive, and alignment and modest compression to complete the re-assembly. The phrase "appropriate adhesive" is intentionally ambiguous, for there are a variety of possibilities available for all conservation gluing -- the selection of which depends on the stresses the object must withstand, the sensitivity of any decorative surfaces of the object (no small consideration when dealing with polychromed panels), the complexity of the damage or the object, skill of the conservator, etc.

A partial fracture, where the object is still in one piece (sometimes tenuously) does not always give easy access to the gluing area, and glue must be applied by either flowing it into the void, or forcing it in under some pressure. The first can be accomplished through gravity or capillary action, the second usually requires some hydraulic device, such as a syringe or a membrane. A syringe is a straightforward tool, acting as a micro-hose to spray liquid into a very specific location.

By the membrane method, hot glue is forced through the cracks at the surface using fingertips as hydraulic plungers manipulating the liquid adhesive. In order to assure that adequate glue is available to be worked into the void the application of hot hide glue is generous. Immediately following the application of the hide glue a transparent membrane is placed over the glue. The membrane, a very thin polyethylene sheet or a similar material has several functions in the procedure, each of which will be enumerated as the description of the technique develops.

Once the transparent sheet has been placed over the glue, the surface is rubbed through it, working the glue through the cracks with a rolling and pushing motion of the fingertips with the sheet acting as a malleable plunger. The membrane also provides the not inconsiderable service of keeping the conservator's hands clean during the treatment. Through the transparent sheet the conservator can observe the progress of the procedure, the end of which is reached when no more glue can be forced into the surface. The transparent membrane also allows the working time to be extended indefinitely through the application of heat from a lamp and the retarding of water loss from the solution. Work sessions of more than an hour are possible for this process.

The complete fracture, that is, one where there are now more pieces than existed originally, presents the advantage of easy and immediate access to the gluing surfaces. Adhesive can be applied directly with a brush, spatula, or whatever tool is best for the task.

With either type of fracture, complete or partial, the conclusion to the process is to align the parts to be unified and apply only that restraint, or clamping pressure, necessary to hold them in place until the glue dries. Frequently the application of clamping pressure squeezes out excess glue, an occurrence which for which the conservator must be prepared to resolve immediately. (This statement applies to all gluing procedures employing liquid adhesives, regardless of context.)

The most vexing circumstance is where the gluing surfaces are no longer adequate for the task of re-assembling the artifact, either through distortion to the panel (planar or otherwise, leaving a void in the alignment), or damage to the gluing surface itself through the displacement ("splintering") of the wood fibers. Alignment and substrate voids are not adhesive problems per se, they are structural. As with a damaged gluing surface, decisions must be reached as to how vigorously the conservator is to intrude into the artifact in order to make it whole. These decisions must be, by nature, ad hoc, but there are some general guidelines for treatment strategies.

The existence of substrate voids, either in the fracture region as a whole or at the glue line in particular, may contribute to the overall structural instability of the object. (Whether or not they contribute to further deterioration depends on many construction, use, and environmental factors outside the scope of this paper.) The degree of that instability, and future circumstances for the artifact, usually determine whether the voids are to be filled or left empty.

If the treatment strategy and objectives require filling the voids, a further decision must be made regarding the degree to which, if any, the artifact is altered to facilitate the fill. In cases of shredded or degraded gluing surface, either consolidation or trimming may be considered. If alteration of the artifact is not possible or desirable, the fill must be made to fit the void exactly.

Methods of accomplishing this include cutting or carving a wood piece to fit the void precisely, casting flexible thermoset material into the void, filling the void with an inflexible thermoset material, or a combination of these. Unless the fill is a tight wood-to-wood system, the gluing surface should be isolated from the fill with an easily reversible barrier film, e.g animal glue, synthetic resin solution, etc. The author finds the use of hide glue to be the most convenient and utilitarian material for this purpose, but other practitioners may well select other materials.

Clean, wood-to-wood repairs generally work well only where there is a relatively "clean" gluing surface, which is not shredded or undulating. In the latter instances wood fills can still be employed, but often with a bulked-out adhesive/fill material used between the gluing surface and the solid fill to eliminate the presence of voids in the repair system.[xv] Whatever procedure and materials employed, it is critical for the fill to not be stronger than the adjoining material of the artifact. If stresses cause damage in the future, it should be to the repair, not to the original artifact.

Treatment options for degraded adhesive systems and associated failures within wooden artifacts are conceptually limited. Regardless of the adhesive used on the object originally, or the specific cause of the failure, the abstract bases for treatment are the same. The decision-making framework for determining treatments presented in this section includes a number of options, generally listed from least intrusive to the object to those which entail the greatest intervention on the part of the conservator. These options include one or more of the following three broad approaches to solving adhesive failure problems.

1. The re-activation of the adhesive by solvents or other means.
2. The introduction of new adhesive which is compatible with the old adhesive.
3. Removing the degraded adhesive and re-adhering the effected substrates with entirely new glue.

In most cases the treatment strategy is limited by the factors:

• What are the properties of the old adhesive?
• What is the probable end use of the object? (designing the treatment to fit the "worst reasonable use" criterion)
• What are the properties of the adhesive to be used in the treatment?

While each adhesive system exhibits highly specific and different characteristics in material and mechanical properties, formulation, use limits and deterioration, the evaluation of adhesive-failure concerns in the treatment of composite wooden objects revolves around the same set of concerns:

• How severe is the failure?
• What is causing the failure?
• How stable is the remainder of the adhesive system?
• What is the adhesive that failed?
• What treatment can compensate for the failure?

The questions just posed are those asked when a conservator examines an object for the purpose of assessing the condition and formulating a solution to the specific problem posed by that object. The reader should be advised that the questions are listed in a rough hierarchy that the author finds useful, but the hierarchy is by no means limited or static. Virtually all observations of objects integrate several of these questions.

Note that the identity of the failed adhesive is only fourth on this list. The specific identification of a failed adhesive is frequently of only modest importance. Certainly this is not always the case, such as instances of the adhesive failing with only minor or no other contributing factors. However, most adhesive failures are a combination of several effects becoming manifest in the specific adhesive system interface. Response to the failure can take one of several paths.

The re-activation of existing adhesive is the most restrained intervention on the part of the conservator. Unfortunately it is usually among the least successful. By definition this technique can be applied only to materials not totally intractable, in other words materials which are thermoplastic. Reactive adhesives, whether catalytic or autopolymerizing, are not candidates for this procedure. In addition, the thermoplastic adhesive materials must not be so degraded as to prevent any useful re-formation of an adhering film. One final factor is the knowledge that even when successful, this approach rarely yields a bond of "structural" strength, and the object may be incapable of fulfilling a "normal use" requirement. Because of these limitations this method is reserved for adhesive materials which remain easily manipulated by solvent or heat, and is applied only to objects that need only support their own weight. As such, it may be perfectly appropriate for panel paintings which are not structural, but unsuited for painted architectural or architectonic decorative panels.

This technique is one where the identification of the existing adhesive, or at least some of its characteristics, is relatively important. Evaluating the possible re-formation of a failed adhesive is possible by trial-and-error solvent testing combined with visual identification and reasonable assumptions, such as presuming brown semi-transparent glassy materials to be dried hide glue. If the re-formation is attempted on thermoplastic resins it is extremely helpful to positively identify the resins before you begin. Analytical methods are preferred in these cases, but without chemical analysis, trial-and-error testing of solvents, whether water or organic solvents, must be undertaken with great care to prevent damage to the object by the solvent.

The most common application of this method is the modification of aged hide glue within the object. The procedure for this treatment is to first clean the surrounding areas of dirt and any other accretions which could contaminate the gluing surface. Warm water is then introduced to the gluing area by syringes or brushes to re-gel the aged glue, and the elements are lightly clamped until the glue re-hardens. For this treatment to be successful the aged glue must still be easily re-soluble and the protein chains must not have degraded to the point that they are no longer sufficient to promote the formation of a viable solid material. At best there will be a re-gelling and re-attachment of the glue surfaces. The fissures of the aged glue may become re-consolidated but there will probably be no thorough re-formation of the bonding material. Re-activation of the entire glue mass usually requires the introduction of considerable quantities of water over a long (several hours) period of time, which is one of the least desirable circumstances for wooden objects. Due to the limitations of the re-formation process, the strength of the resultant bond may be quite low. The re-activation potential of hide glues remains important, however, and will be discussed at greater length later in this section.

Other applications of the re-activation method of treating adhesive failure include the re-dissolving of synthetic solvent-based adhesives. These adhesives, synthetic polymers in solution, are not common in construction as a general rule, but are frequently present in artifacts as part of a previous attempt to rectify damage. It may be possible,though not always desirable, to apply the same technique to these materials by introducing the appropriate solvent to the affected area, re-liquefying the adhesive, and clamping lightly. If the adhesive being re-activated is not original to the object, it is often not considered to be important to that object and its removal may be preferable. In addition, some types of solvent-based synthetic-resin adhesives degrade at a rate which prohibits the consideration of this treatment as a long-term option. As with the treatment of hide glue by this method the ensuing bond may not be strong enough to allow "normal use" of the object. There also exists the very real possibility of damaging the decorative surface of the object by the application of the organic solvents to the damaged adhesive bond. The material technologies of adhesives and coatings are essentially the same in many respects, and solvents which re-dissolve polymeric adhesives will act as paint removers for many coatings.

Certain adhesives can be re-activated not with solvents but with heat. This includes lower weight polymers, especially emulsions, and rubber-type adhesives. Solvents will merely swell these materials and not allow for their re-formation as a whole. Heating these materials may re-introduce increased plasticity on their part and the flow may allow some degree of re-attachment. There are, of course, the associated risks involved whenever a wooden object or a portion of a wooden object is heated. For example, nearby glue lines can fail, the wood can warp, the paint film buckle, etc.

Finally, the technique of re-forming degraded adhesive bonds will not be successful for thermoplastic adhesives which are degraded beyond the point of retaining viability, or with thermoset adhesives. These materials may swell into a gel state but show inadequate cohesion, or may simply crumble apart.

A more intrusive repair method involves adding new adhesive to the glue-line to augment failing adhesives. The primary constraint on this technique is the requirement that the newly introduced adhesive be compatible with and bond to the existing adhesive. Frequently objects are encountered where a repair was attempted by introducing an inappropriate new adhesive into a damaged area, the result being an assembly that is, in the long run, more unstable than before the attempted repair. If the appropriate adhesive is carefully introduced, the chance of success for this approach is greater than that of the previous method. The stability of a joint treated in this manner can approach that of one cleaned and reglued, and will frequently allow moderate levels of use.

The introduction of new adhesive can encompass several procedures, ranging from re-gluing elements which are already apart to those which are restrained or cannot safely come apart even with the most careful effort. The usual methods of introducing new adhesive are injection with a syringe or simply allowing the adhesive to flow from a saturated glue brush down into the crevice, or brushing the adhesive on the gluing surface if the object is apart. In any case the usual objective is to completely fill any voids and provide the necessary degree of strength, and the greatest stability and durability possible.

Once the extent of required stabilizing has been defined, and the compatibility of the adhesive on the object and the adhesive to be used in the treatment determined, a strategy for introducing the new adhesive is implemented, and the elements are lightly clamped until the glue has solidified.

In general, this method of stabilizing the structure is used only for adhesives which are easily soluble in the same solvent and thus can meld together to form a cohesive adhesive bonding unit. As a practical matter, conservation treatments of this type rarely include materials other than hide glues for the same reasons enumerated in the previous section. Water-based emulsion glues can be added to existing hide-glue lines since they will bond reasonably well to existing protein adhesives. However, PVA emulsion glues will likely have different mechanical properties which would make them react to environmental changes differently than would the protein glue, leading to eventual failure of the glue line. In the end, this approach to the problem can create further deterioration. In addition there is the concern of long-term stability of PVA emulsion's themselves. These concerns discourage the use of PVA emulsion adhesives for this, or almost any other, wood conservation purpose. There is growing interest in using synthetic hot-melt adhesives for this type of treatment. Hot melts flow well, adhere to a wide variety of materials, and can be obtained in a number of different formulations, many of which are easily reversible with heat or solvents. While further investigation of this procedure is encouraged, there is no theoretical reason why this treatment option should not be developed.

Depending on the amount of working time required for the gluing and clamping procedure, the glue used will most likely be either hot hide glue, which allows a brief working time (a couple of minutes), or cold hide glue, which allows a lengthy working time (up to an hour). For hot glue, viscosity of the glue mixture is adjusted by heat or water to be the appropriate viscosity to enter the cavity being filled. Hot glue can also be modified by the addition of gel suppressants to allow longer working times.

As with the technique of reforming existing glues, when adding new adhesive it can be advantageous to introduce small amounts of solvent into the area prior to the application of the new adhesive. If the glue is to be injected with a syringe the viscosity must be extremely low and the temperature relatively high so that the syringe will not clog halfway through the procedure. It is useful to have a container of warm water adjacent to the working area in which the syringe can be soaked and the glue in the syringe kept liquid and flowing. For cold hide glue the viscosity can be adjusted through selection of glue grade (the higher the glue grade, the greater viscosity for a given % concentration), or dilution until the consistency is correct. It is then injected in the same manner as the hot glue with the difference that the need for rapidity in the process is diminished.

If the crevice through which the glue is introduced to the space in question is large enough the use of a syringe is not necessary. Simply dip a brush of the proper design into the glue and hold it over the area to be glued, allowing the glue to run off the brush into the void. If possible, the bristles of the brush can be worked into the opening to facilitate the penetration of the glue. Thin spatulas and other similar tools can also be used for the same function. An additional method, described earlier, uses a hydraulic membrane to squeeze adhesive into the voids too small for tools.

If there is an unstable structure which must be treated yet has no opening adequate for the introduction of glue and cannot be disassembled, a channel must be cut to allow the glue to penetrate. This can be accomplished by selecting a drill bit the size of the syringe and drilling a small hole in an inconspicuous area of the intersection between the two structural members being glued. This option is usually undertaken only after extensive deliberation since it involves the removal of existing material from the artifact.

By drilling a hole into the gluing cavity, the new adhesive can then be injected. For assemblies with more than one major gluing surface it may be necessary to drill more than one hole to inject enough adhesive to accomplish the goals of the treatment.

Elements that are already apart or can be considered for disassembly as part of the treatment procedure terminate that portion of the treatment in a straightforward gluing process. The condition of the remaining adhesive is evaluated, and if found to be stable, left in place. The same is true for the substrate to the adhesive; if it's stable it can be left alone. (Adhesive materials which are unstable and need to be removed are covered in the section following this one. The repair of damaged gluing surfaces is a structural problem and is covered in the section on structural treatments.) The pieces of the object are assembled to check orientation and assembly procedure. As with all gluing projects it is critical to have a clear idea of how the procedure will progress once begun. If the pieces all fit together as they should and the clamping methods are worked out, the mock-up can be disassembled and the gluing can commence. Apply the glue with a brush to the gluing surfaces and reassemble the object with the appropriate clamps and glue blocks, etc.

The option of completely removing the aged adhesive materials is obviously available only in cases where the object can be completely disassembled and the conservator has access to all gluing surfaces. The need to remove all the existing glue is encountered frequently with a number of causes ranging from environmental fluctuations to inept attempted repairs with inappropriate adhesives. The result of these factors is a glue whose continued presence on the object contributes not to its preservation but rather to its accelerated deterioration.

The exact identification of the adhesive in these circumstances is of only minor importance. Since the material is going to be removed it is adequate to simply know how that can be accomplished. Two techniques are usually implemented together to remove degraded adhesive; mechanical removal and solvent application. As would be expected, mechanical removal by itself is simply scraping, chipping, grinding, etc., of the adhesive material remaining on the gluing surface. Solvent application attempts to redissolve the adhesive to enable removal by wicking the material into an absorbent pad or swabbing it off with soft wadding or cloths. Removing the adhesive with solvent alone is the optimal circumstance, but usually neither approach is satisfactory by itself. The procedure most commonly employed in the removal of degraded adhesives is similar to the following.

Initial testing with tools such as micro-spatulas and dental picks determine how well adhered the adhesive is to the substrate. Following this examination the adhesive is tested with solvents to determine which (if any) have an effect on the material. This information is then synthesized toward designing the step-by-step procedure for cleaning the gluing surface.

Any adhesive material which can easily be removed without damaging the substrate with the tool would then be removed mechanically. This is done without vigor but with great care to prevent the removal of substrate material with the picks and scrapers. Unless the adhesive can be easily and cleanly removed this way there is a need to introduce solvents to the procedure. Ideally, as stated earlier, they dissolve the adhesive to allow rapid cleaning of the surface. Unfortunately this rarely happens. Generally the adhesive is too degraded or the time limits of keeping the material wetted with solvent preclude dissolution. A more likely result is that the adhesive will swell and/or soften, enabling it to be removed with wooden scrapers or cotton swabs. For most treatments this needs to be repeated to insure a clean surface. Caveats already mentioned regarding the use of solvents during the treatment of degraded adhesive materials by reactivation and reformation should be followed in this case as well.

In cases where the adhesive is so reacted or degraded that it becomes completely intractable, the choices by the conservator become much more difficult. An example would be a crosslinked adhesive which is essentially impervious to adhesive softening. The conservator must then weigh the relative benefits of a clean gluing surface which can be obtained only by aggressive mechanical removal against the possible detrimental effects of such a cleaning. This dilemma must be resolved on an individual basis on no further discussion here will shed light on the solution to the problem.

Following the cleaning of the gluing surface any repairs to the structure necessary for the regluing are then executed and the object is prepared for reassembly.

Adhesive materials on artifacts often reveal vital information about the nature of the artifact, and a historical/material technology base can provide useful clues and direction to the caretakers of them. But, extant or available adhesives are only one component of any gluing procedure, which can range from the mundane to the elegantly complex. Attempting conservation treatments requiring adhesive processes without understanding every component of that process may not be folly, but it is at least a close relative. This cursory chapter is intended to be a useful reference for developing prudent practices for using adhesives during the restoration of wooden artifacts.

Adhesives and Consolidants; Preprints of the Contributions to the Paris Congress, 2-8 September 1984. London; International Institute for Conservation, 1984. (Referred to as "The IIC Paris Conference.)
Allen, K.W. "Adhesion and Adhesives ? Some Fundamentals." Preprints of the IIC Paris Conference, 1984, pp. 5?12.
Anderson, Mark, and Michael SandorPodmaniczky. "Preserving the Artifact: Minimally Intrusive Treatments at Winterthur:, in WAG Preprints 1990. Washington D.C., A. I. C. 1990.
Bradley, Susan. "Strength Testing of Adhesives for Conservation Purposes." Preprints of the IIC Paris Conference, 1984, pp. 22,25.
Called, Bernard and Ostrup, Anna. "Study of Adhesives for Marquetry." Preprints of the IIC Paris Congress, 1984, pp. 129?132.
Cummins, Jim. "Visit to a Glue Factory." Fine Woodworking, Mar/Apr, 1986, pp. 66?69.
DeBeaukelaer, F.L., et al. "Standard Methods (Revised) for Determining Viscosity and Jelly Strength of Glue." Industrial and Engineering Chemistry, vol. 2, no. 3, July 15, 1930, pp. 348 - 351.
Down, Jane, and Raymond LaFontaine. "A Preliminary Report on the Properties and Stability of Wood Adhesives." Proceedings of the CCI Furniture and Wooden Objects Symposium, 1980, pp. 55?64.
Feller, Robert L. and Encke, David B. "Stages in Deterioration: The Examples of Rubber Cement and Transparent Mending Tape." Preprints of the IIC Washington Congress, 1982, pp. 19?23.
Fernbach, R. Livingston. Glues and Gelatine: A Practical Treatise on the Methods of Testing and Use. New York; D. Van Nostrand Company, 1907.
Grozdits, George. Bonding Wood. Washington DC: American Institute for Conservation Wooden Artifacts Group, 1985.
Gutcho, Maria, editor. Adhesives Technology: Developments Since 1979. Park Ridge NJ: Noyes Data Corp., 1983.
Hoadley, Bruce. "Glues and Gluing." Fine Woodworking, Summer, 1977, pp. 28?32.
Kendall, Kevin. "Adhesion: Molecules and Mechanics." Science, Vol. 263, 25 March 1994, pp. 1720 - 1725.
Knight, R.A.G. Adhesives for Wood. Chemical Publishing Company, New York, 1952.
Koob, Stephen P. "The Continued Use of Shellac as an Adhesive ? Why?" Preprints of the IIC Paris Congress, 1984, pp. 103?104.
Koob, Stephen P. "The Instability of Cellulose Nitrate Adhesives." The Conservator, vol. 6, 1982, pp. 31?34.
Kozlov, P.V., and G.I. Burdygina. "The structure and properties of solid gelatin and the principles of their modification." Polymer; vol. 24, June 1983, pp. 651 - 665.
Landrock, Arthur H. Adhesives Technology Handbook. Park Ridge NJ; Noyes Publications, 1985.
Mantell, Charles. Technology of Natural Resins. New York: John Wiley & Sons, 1942.
Marra, A.A. Manual of Gluing. High Point NC: National Furniture Manufacturers Association, 1950.
Martens, Charles. "Emulsion Paints." In Technology of Paints, Varnishes and Lacquers, Charles Martens, editor. Malabar FL: Robert Krieger, 1968.
Mustoe, George. "Glues for Woodworking." Fine Woodworking, Jan/Feb, 1984, pp. 48?50.
Mustoe, George. "Which Glue Do You Use?." Fine Woodworking, Nov/Dec, 1983, pp. 62?65.
National Association of Glue Manufacturers, Inc. Animal Glue in Industry. New York: National Association of Glue Manufacturers, Inc., 1951.
Perry, Thomas D. Modern Wood Adhesives. New York; Pitman Publishing, 1944.
Rice, James T. "Gluing of archaeological Wood." In Archaeological Wood: Properties, chemistry, and Preservation, Rowell and Barbour, eds. Washington DC: American Chemical Society, 1990.
Rose, C.L., and D.W. von Endt. Protein Chemistry for Conservators. Washington DC: American Institute for Conservation, 1984.
Rosser, George L. Animal Glues and Their Use in Woodworking. Ottawa: Canada Department of Mines and Resources, 1939.
Selbo, M.L. "Selecting Adhesives for Wood Products." Adhesives Age, October 1973, pp. 36?41.
Selwitz, Charles. Cellulose Nitrate in Conservation. Santa Monica CA: Getty Conservation Institute, 1988.
Skeist, Irving, Ed. Handbook of Adhesives (2nd Edition). New York: van Nostrand Reinhold, 1977.
U.S. Department of Agriculture. "Animal Glues: Their Manufacture, testing, and Preparation," Report No. 492. Madison WI: 1955.
U.S. Forest Products Laboratory, Wood Handbook: Wood as an Engineering Material (USDA Ag. Handbook No. 72). Washington, DC, 1974.
Wicks, Zeno W., Jr. Film Formation. Philadelphia: Federation of Societies for Coatings Technology, 1986.
Williams, Donald C. and Ann Creager. "Conservation of Paintings on Delaminated Plywood Supports." Saving the Twentieth Century: Conserving Modern Materials. Ottawa: Canadian Conservation Institute, 1993. pp. 231-241.
[i] Numerous sources, e.g. Cummins (1986), Fernback (1907), Perry (1944), Rose and von Endt (1984), and Rosser (1939).
[ii] DeBeaukelaer (1930), Fernbach (1907), Rosser (1939).
[iii] Rose and von Endt (1984).
[iv] Recipes for "liquid" animal glues generally consist of approximately a 5:1 ratio of dry glue to dry modifier, usually urea, calcium chloride or sodium chloride, NAGM (1951). Modifiers in lesser quantities lengthen the gelling time but do not eliminate it altogether.
[v] Conversation with David von Endt, 1986.
[vi] Most sources discussing the technology of natural resins refer to their widespread utility as coatings, sealants and adhesives: Koob (1984), Mantell (1942).
[vii] Steinfink in Skeist (1977).
[viii] Feller and Encke (1982).
[ix] Corey, Draghetti, and Fantl in Skeist (1977), Martens (1969), Wicks (1986).
[x] Hoadley (1977), Mustoe (1984), Selbo (1973).
[xi] Koob (1982), Selwitz (1988).
[xii] Coover and McIntire in Skeist (1977).
[xiii] Gutcho (1983).
[xiv] Marra (1950), revisited frequently, e.g. Rice (1990).
[xv] Anderson and Podmaniczky (1990).

Donald C Williams is Senior Furniture Conservator at the Conservation Analytical Laboratory, Smithsonian Institution, Washington, D.C. The views expressed are solely his and do not necessarily reflect the position of the Smithsonian Institution.

---

---

---