PMG Preservation of Glass in Photographic Materials
In progress: Seeking additional comments and images to develop this section
Photographic Materials Conservation Catalog
Preservation of Glass in Photographic Materials
The chapter addresses the philosophy and methods for the preservation and conservation of glass used as part of and housing for photographic materials.
Date initiated: 2007
Compiler: Katharine Whitman, Amanda Maloney
Contributors (Alphabetical): Contributors names in alphabetical order Your name could be here! Please contribute.
First edition copyright: ?. The Photographic Materials Conservation Catalog is a publication of the Photographic Materials Group of the American Institute for Conservation of Historic and Artistic Works. The Photographic Materials Conservation Catalog is published as a convenience for the members of the Photographic Materials Group. Publication does not endorse nor recommend any treatments, methods, or techniques described within the chapter.
TABLE OF CONTENTS:
7.2 Factors to consider before treatment
7.3 Glass as a photograph support
7.4 Equipment and materials
7.5 Techniques and Methods
7.6 Preventive Conservation
7.1.1 To physically stabilize the photographic material.
7.1.2 To restore the image closer to original appearance and function before damage occurred.
7.1.3 To provide preventative care during storage.
7.1.4 To consider traditional and current methods of glass stabilization and other cosmetic treatments.
7.2 Factors to Consider Before Treatment
7.2.1 Purpose and focus of the object
Archives, “fine art”, artist’s intention…
7.2.2 Type of object – Binder considerations
Repair strategies will depend on binder: albumen, collodion or gelatine...
7.2.3 Format of the object
Repair strategies will depend on size…
7.2.4 Condition of object
Repair strategy will depend on the extent of the damage.
7.2.5 Past treatments
Dealing with the removal of past treatments…
7.3 Glass as a Support for Photographs
7.3.1 Basic chemical make-up
In its pure form, glass is a transparent, hard-wearing, strong but brittle, and biologically inactive material that can be formed to have very smooth and impervious surfaces. The properties of glass can be modified by varying the heat treatment or by adding other components. The basic ingredients are: amorphous silicon dioxide (SiO2), soda (sodium carbonate Na2CO3) or potash, (a compound, or flux, to lower the melting point), and a stabilizer such as lime (calcium oxide, CaO, to restore insolubility). Glass produced with these ingredients contains about 70% silica and is called a soda-lime glass. Soda-lime glass accounts for about 90% of manufactured glass.
7.3.2 Forms of deterioration
Glass is characterized as a brittle material and therefore, subject to brittle fractures with rapid crack propagation without significant plastic deformation. An impact break is one that is caused by a crushing blow to the material. This variety of fracture may have a bright, granular appearance and is characterized by an impact cone, where the most damage has been done, surrounded by radiating arcs.
Cracks will run perpendicular to an applied stress: that is, the shard of a broken sheet of glass will have smooth fracture surfaces. Blind cracks are characterized by breaks that do not carry through the whole shard. These have to be stabilized primarily to avoid further damage to the support.
Knowledge of stress states and how they affect the binder, or image layer, is needed. Sheer stress that is parallel to a face of the material, normal stress that is perpendicular to the face of the material, tensile stress induced by pulling forces, or even compressive stress that can rupture the material are all potentially very harmful to glass with blind cracks and breaks that have not broken the image binder. If the break has not torn or damaged the image binder, these shards will need special attention and stabilization to prevent further damage.
“Sick glass” has always been a problem for conservators. There are several contributing factors affecting decay, the major ones being those of glass composition, poor founding practice and environment. Others are exposure to aggressive leaching solutions, time and temperature. The usual symptoms are weeping and the formation of crystals. Weeping is often caused by the excess of alkali and the lack of sufficient lime and/or other stabilizing material in the batch. In general, glass with a high proportion of silica to modifiers will tend to be stable whereas if the proportion of modifiers is high then the glass will be much more readily attacked. UV light is known to increase surface corrosion and the presence of sulfur in the atmosphere is known to increase weathering problems. The chemical nature of decay is varied and very complex. Some of it, such as the weeping and crizzling caused by an excess of alkali and a lack of stabilizers in the glass, is fairly simple. The results of attack by water can be much more complicated. Water is a primary agent in the deterioration of glass. The surface of glass tends to react with water or even with a humid atmosphere and starts a continuing process in which the effect progresses further into the glass. Poor storage in the very damp conditions that often prevail in cellars or attics can be a considerable contributory factor to this form of decay.
7.4 Equipment and Materials
7.4.1 Temporary treatment housing
When undertaking the treatment of a broken plate, the shards should be housed in a shallow box, binder side up, on a cushion of Ethafoam. Textiles such as woven polyester should not be used because the fibers can catch on loose binder or pieces of glass. If there are many small pieces and blind cracks, it may be advisable to assemble the pieces in PhotoShop first (see section 7.6.2), to minimize handling and avoid chipping fragile glass edges.
The difficulty in assembling a photograph on glass by this method is in choosing an adhesive. There is not a perfect adhesive available at this time. Ideally, the adhesive would need to have a refractive index that matches the artifact’s, would have a working time and be available with a low enough viscosity enough to wick into a shard interface while providing strength in the final cure, and most importantly, be completely reversible.
Paraloid B-72 is an acrylic copolymer, made of ethyl methacrylate and methyl acrylate, with a refractive index of 1.48. It is soluble in a variety of organic solvents and has a glass transition point of 40°C (approximately 104°F). A solution of 50-70% B-72 in a solvent is recommended for the assembly of glass fragments. A small amount of fumed silica (approximately .01 by weight) should be added to the solution to improve application and drying properties of the adhesive. The B-72 will set hard in 1-2 hours (possibly longer for very thick glass). A problem with B-72 is that when it is used with a solvent (an aromatic), as the repair cures and the solvent evaporates, “snowflakes” appear in the repair areas, therefore making the adhesive inappropriate for glass plate photograph repair where an invisible, clear repair is desired.
Epoxy resins used in conservation are typically composed of two parts, a di-epoxy component and a polyamine cross-linking agent, both of which are combined with diluents and catalysts and have a refractive index of approximately 1.55. The adhesives are very strong and has low shrinkage upon curing. The disadvantage is that this adhesive tends to yellow with time and exposure to light. It is not advisable to use Epoxy on a photograph with a collodion binder because of reversibility issues. To reverse an Epoxy repair the plate must have a methylene chloride (CH2Cl2) poultice applied to the fracture for 2-4 hours to soften the adhesive, which can then be removed mechanically. Collodion binders, the varnish layer and the japanning tend to be very sensitive, and they may be severely damaged by this treatment.
Cyanoacrylates “snap bond” when placed in contact with alkaline surfaces, such as glass. The resulting bond is extremely brittle; therefore cyanoacrylate adhesives are generally unsuitable for glass restoration, except for effecting temporary repairs. Acetone will reverse any applications.
For the application of adhesive to the interface of the shards, a wooden or glass applicator should be used. Metals can interact with the adhesives, making them less affective. Capillary action pulls the required amount of adhesive into the interface with minimal excess. An alternative to this tool could be a dosing bottle and tip. These bottles and tips cost about $17 USD for ten bottles and $12 USD for ten of the smallest size dosing tips, and are available from Cyberbond Incorporated. Or brush
7.4.4 Backing material during treatment
184.108.40.206 Polyester film (Mylar Type D)
When the adhesive is applied to the fractures sites and the shards are placed on the polyester film, the adhesive is pulled out of the repair site by capillary action, onto the glass side of the plate where it is in contact with the polyester film. Once the adhesive has cured for 24 hours, this can be easily removed with a razor blade and appropriate solvent. However, there also tends to be some overflow onto the emulsion side of the negative, where it was also in contact with Mylar during drying.
220.127.116.11 Silpat™ sheet
A Silpat™ mat is made of silicone and fibreglass and has an evenly textured surface. This textured surface provides enough air pockets to prevent the capillary action that can occur on the bottom side of the plate during treatment, removing the adhesive from the shard interface. This method of repair produces very satisfactory results, with no trauma to the emulsion side of the plate, or excessive adhesive use.
A “light-line” creates a straight line of light, via a fibre-optic array. When the light is directed onto the shard interface any misalignment will be marked by a crooked line. As the pieces are brought into alignment, the light-line will become a straight line. A less expensive alternative to the light-line could be a flashlight fitted with a snoot that creates a straight line of light or any straight-edged beam of light will suffice.
7.4.6 Final secondary supports
In some circumstances a repaired glass plate will require additional support. As a rule, a glass plate that has been broken into many pieces, or is larger than 5 x 7 inches, will require a secondary support as part of its housing.
Of concern with this housing is the possibility of glass deterioration occurring between the support glass and the image glass. Deterioration crystals can form over the surface. The reason for this phenomenon is a combination of interacting factors: the use of inferior sandwiching glass and trapped moisture migrating within the sandwich. When an inferior glass, that is more prone to deterioration, is used, the likelihood of the alkali leaching out of the glass structure in humid conditions is much higher. Placing two pieces of this material into intimate contact compounds the problem.
P-4 clear silicone, sourced from Silicones Incorporated, has passed the Photographic Activity Test and is appropriate for use with photographic materials. This silicone can be used as a barrier layer between the glass side of a repaired plate and a piece of backing glass. The theory being that it will prevent glass deterioration from occurring between the plate and the backing glass.
7.5 Techniques and Methods for Repair
The most common cause of breakage in the lab is the object slipping out of the conservator’s hands when wearing cotton gloves. Neoprene gloves should be used instead to protect the emulsion from fingerprints that will cause deterioration over time. Glass fragments should be handled as little as possible. To that end, virtual assembly in PhotoShop software may be necessary to determine the arrangement of the shards (see section 7.5.2).
A padded (foamed polyethylene) and tight weave tissue or Sintered Teflon lined box should be used for fragments. Do not let the fragments come into contact with the foam; any lifting binder will be very susceptible to snagging. Handle glass plates by two edges and carry flat. Do not handle by the corners. Work on a clean, dry, padded surface: such as a sheet of thin Ethafoam covered with blotter. Place glass plates on a flat surface, binder side up. Never place any pressure on the plate: label sleeves before placing the plates in them. Never stack bare plates. High use plates should be duplicated.
7.5.2 Virtual assembly with PhotoShop Software
In the case of an object that has been broken into many shards, initial assembly should be performed by scanning the shards into files that can be manipulated in PhotoShop to determine their positions. This minimizes handling, protects the binder and any mirroring and reduces the chance of further damage to the glass. Blind cracks may be very susceptible to separation. To assemble the shards in PhotoShop, the pieces must first be photographed or scanned on a transmissive illumination flatbed scanner. It may be more convenient to photograph or scan the shards all at once so that they will have a precise size relationship to each other. Each shard should then be selected and made a separate layer. Select the first layer to be manipulated, and select the rotate function under edit > transform > rotate. Rotate the piece until it reaches an approximation of the area in which to be moved and drag it into place. Select and manipulate the rest of the layers until the object is assembled. This method of assembly will not only protect the delicate emulsion and glass from the further damage that can occur by physically manipulating the shards, but will give the conservator an idea of any other problems with the piece, such as missing shards.
7.5.3 Inclined assembly
In this assembly method, adhesive is applied to the shard interfaces, which are then assembled on an inclined plane on a sheet of Mylar or a SilpatTM. The edges of the assembled plate are shored up with matboard, and an additional light weight is placed on the top side of the plate during curing. This method requires cleaning of the bottom side of the plate, which was against the inclined plane because capillary action tends to suck the adhesive out of the fracture interfaces.
When a Silpat™ mat is utilized on an inclined stage over lights to highlight the fracture interfaces, the broken plate can be assembled on the stage emulsion side down. The adhesive can then fed into the fracture with a brush while monitoring the progress of the adhesive. When enough adhesive has been applied, a weight was placed on the top surface and it is allowed to dry.
18.104.22.168 Application of adhesive
22.214.171.124.1 Direct application
When repairing a broken plate on an inclined plane, the adhesive is applied directly to the fracture interface and the shard is placed directly next to its corresponding shard on the inclined plane.
126.96.36.199.2 Wicking application
When a drop of adhesive is applied to the shard interface, capillary action pulls the required amount of adhesive into the interface with minimal excess. Describe how to apply the adhesive. Brush, applicator bottle
7.5.4 Vertical assembly
Cracks will run perpendicular to an applied stress: that is, the shard of a broken sheet of glass will have smooth fracture surfaces, and therefore, two adjacent shards of glass will fit back together perfectly and in only one proper way. With experience, a conservator can learn when this position is attained. This method of assembly will also insure that the binder side of the plate is disturbed as little as possible.
The basic, traditional procedure for assembling glass is as follows: A stable shard of glass is supported perpendicularly to the working surface so that there will be no movement during assembly. When a bubble level is used to insure the shard is perfectly vertical, this position will insure that gravity, the constant force pulling down on the shard. (compare this to the not flat surfaces of most negatives) The object is then assembled, using pressure sensitive plastic tape, or sticky wax, to hold the shards together on the glass side. The shards are aligned by visual inspection and feel. It is important that the object is assembled completely before any adhesive is applied to insure precise alignment of all shards. In the case of an object that is broken into many pieces, the smallest misalignment will be magnified in the assembly of the subsequent pieces that will lead to a major misalignment by the end of the process. When adhesive is not applied until the last step, these misalignments can be recognized and corrected. Once the object has been assembled completely, adhesive is wicked into the shard interfaces from the glass side and allowed to cure. The object is then removed from the vertical support, the tape is removed and excess adhesive is cleaned up.
188.8.131.52 Use of the “Light-line”
A “light-line” can be used to aid in the alignment of the shards: if they are misaligned in any direction, it is instantly known because the light line will not be straight. Additionally, the use of fiber-optic lighting on the edge of the glass illuminates the shard interfaces during assembly and allows the conservator to know when the proper amount of adhesive has been applied.
184.108.40.206 Pressure sensitive tape
Plastic tape, while easy to use and completely removable, has a flexible plastic carrier that gives minimal support.
220.127.116.11 Sticky wax
As the pieces are assembled, sticky wax, such as that used for lost-wax casting in jewelry making, is very useful for holding the shards in place. Sticky wax becomes soft with low heat and hardens at room temperature, providing a stiff support for the assembled pieces. It is also very easily removable once the adhesive has been applied. The wax comes in sticks that can be cut into lengths as needed. Using a pin to hold the wax, warm the wax slightly over an alcohol lamp and place it on the glass side of the assembled shards. Enough pieces of wax should be used at sufficient intervals to support the glass as much as possible without impeding the wicking-in of the adhesive later in the treatment: one piece of wax for every four centimeters of shard interface is suggested. When the plate is completely assembled, the light-line should cast a straight line across all of the shard interfaces.
7.5.5 Loss compensation
Epoxy may result in a repair with the same refractive index as glass, so the reapir is clear, but has debatable reversibility issues. Methylene chloride (CH2Cl2) will soften Epoxy to a gel that can be scraped off, but is highly toxic and must be used in a fume hood. The advantage of using Epoxy is that it will fully cure in a week with a very strong bond. Describe filling method It is highly recommended that a form-fitting secondary support of matboard be created for objects with losses. This will ensure that the loss edges will not cause tearing of the enclosure. Further research needs to be done regarding the creation of fills for gelatin glass plate images on glass.
7.6 Preventative Conservation
7.6.1 Housing considerations
18.104.22.168 Intact plates with no binder damage
House glass plates in four-flap enclosures and envelopes. All materials should pass the Photographic Activity Test (P.A.T.) The four-flap enclosure will insure that any incipient flaking will not be exacerbated by inserting and removing the plate while sliding it out of the envelope. If four-flap enclosures are not feasible, the plates should be placed in the envelope binder side away from the seam to ensure there is no abrasion of the binder and that the adhesive cannot cause deterioration. This enables the object to be put into and removed from the enclosure without the risk of scratching that can result from sliding the object into and out of an envelope. Plates should be stored vertically in document boxes, on the long edge. Interleave every inch with acid-free cardboard to support the plates. The plates should be housed vertically, on their long edges, in a partly filled box with a spacer to minimize jostling during handling. For plates over 10 x 12”, use over sized legal boxes. Only partially fill the box to prevent the box from becoming too heavy. Acid-free corrugated board should be used to fill out the box to avoid shifting of the contents. Boxes of glass plates should be stored on lower shelving and never above about four feet to prevent someone from having the lift the plate down from above their head. The boxes should further be labeled clearly: "FRAGILE," "HEAVY", and "GLASS".
22.214.171.124 Intact plates with binder damage
If the flaking is minor, i.e. a few small edge losses with no flaps or flaking towards the center of the plate, the plates should be housed in four-flap enclosures within their envelope. The envelope should be labeled: FLAKING EDGES remove with care. More severely flaking plates, i.e. plates with hanging flaps and/or cracking overall with incipient flaking in the center, should be duplicated and housed in four-flap enclosures within their envelope. Plates with more extensive flaking should be stored in sink-mats and stored horizontally. Housings should be labeled accordingly.
126.96.36.199 Broken plates
Broken plates deserve special attention. It is very important that broken shards do not come into contact with each other. This can cause damage to the binder and the glass edges such as chipping and additional breakage. Create a form-fit support by cutting 3 pieces of 4-ply mat board and place one on the shards on one of the pieces of mat board, emulsion side up, and trace the edges of the shard. Remove the shard and cut out the form, leaving two 2 pieces that will fit to the shards. Attach one of these pieces to a full size board and the second form to the second piece of board with wheat starch paste or 3M #415 double-stick tape. The shards should sit level to below the top surface of the mat. The objective is to support each piece so that additional damage will not occur by placing the shards in contact. These are then placed in separate 4-flap enclosures and housed flat, and marked “broken plates / carry flat”. An alternative housing for plates that have been broken into many shards is suggested in a 1991 paper by Constance McCabe of the National Archives in Washington D.C.: “Broken negatives are assembled in proper orientation for duplication but are housed in sink mats with the components separated with paperboard spacers attached with adhesives to avoid mechanical damage to the glass pieces… [Paper tabs are inserted] to assist in lifting out large pieces. These enclosures muse be carried horizontally or the glass will slip away from its support… Negatives housed in sink mats are stored horizontally in stacks of three to six (depending on the size and weight) within storage boxes. Boxes are of the drop-front style with metal stays… Each box is marked with the cautionary label: “Caution: Broken glass. Carry Horizontally”.” It is extremely important that broken negatives housed in a sink-mat include pieces of paperboard to separate the shards. If this is not done, the shards will rub together, causing flaking of the binder and grinding of the glass. The sink mat should be constructed of mat-board and care should be taken that if there is any flaking binder that it is not exacerbated by rubbing on the edge of the sink-mat. A sintered Teflon lining will minimize this possibility. Cased images should be hosed in four-flap enclosures with a thumbnail photograph of the object on the outside of the box so that it can be determined what is in the box without opening the enclosure. Objects should be stored flat in larger boxes or padded drawers, or vertically in padded slots in boxes or drawers. Do not house glass plates in plastic sleeves.
7.6.2 Environmental Considerations
Control of Relative humidity offers a significant improvement in the long-term preservation of photographic materials. A slight increase in RH will lead to the deterioration of the silver, binder, varnish and glass support, and a slight decrease on RH can lead to flaking of binder, and dehydration of the glass. Connie McCabe recommends 30-40% RH. Below 30 and the binder will desiccate and above 40 the glass will start hydrating. Stephen Koob of the Corning Museum of Glass recommends that the following guidelines be followed: No fluctuations in conditions Avoid temperature extremes Avoid spot-lighting (leads to uneven heating) Avoid cycling. Cased images have their own requirements when it comes to their preservation. The base recommendations are 18 – 20°C and 40-50% RH. Do not store cased images <40% RH to prevent embrittlement of the case, and do not store them >50% RH to prevent brass mat and cover glass deterioration. For all photographs on glass, the light levels should be kept below 50 lux (5 foot-candles) when they are on display.
Barger, M. Susan. 1985. Characterization of Deterioration in Glassplate Negatives. Second International Symposium: the Stability and Preservation of Photographic Images, Ottawa, Canada: 134-137
Barger, M. Susan, Deane K. Smith and William B. White. 1989. Characterization of Corrosion Products on Old Protective Glass, Especially Daguerreotype Cover Glasses. Journal of Materials Science 24: 1343-1356.
Brill, Robert H. 1975. Crizzling - A Problem in Glass Conservation. Conservation in archaeology and the applied arts. Preprints of the contributions to the Stockholm Congress, 2-6 June 1975: 121-134.
Brill, Robert H. 1988. General Notes on Glass.
Brill, Robert H. 2000. No, It Doesn't Flow.
Bunker, B.C. 1994. Molecular Mechanisms for Corrosion of Silica and Silicate Glasses. Journal of Non-Crystalline Solids 179: 300-308.
Clark, Susie. 1998. The Conservation of Wet Collodion Positives. Studies in Conservation 43: 231-241.
Culham, Charlotte, Janet H. Notman and Norman H. Tennent. 1999. Tension Released; An 'Exploding' Chinese Vase. The Conservation of Glass and Ceramics: Research, Practice and Training: 167-170.
Davison, Sandra. 2003. Conservation and Restoration of Glass. New York: Butterman Heinemann.
Davison, Sandra. 1999. A Review of Adhesives and Consolidants Used on Glass Antiquities. Adhesives and Consolidants: Preprints of the Contributions to the Paris Congress, 2-8 September 1984: 191-194.
Errett, Raymond F., Merrill Lynn and Robert H. Brill. 1999. The Use of Silanes in Glass Conservation. Adhesives and Consolidants: Preprints of the Contributions to the Paris Congress, 2-8 September 1984: 185-190.
Fonatine, Chantal. 1999. Conservation of Glass at the Institut Royal du Patrimoine Artistique (Brussels): From the Earthquake in Liege to the Stained Glass of Loppem. The Conservation of Glass and Ceramics: Research, Practice and Training: 199-207.
Gillet, Martine, Chantal Gardnier and Francoise Flieder. 1986. Glass Plate Negatives: Preservation and Restoration. Restaurator 7:2: 49-80.
Hearn, Charles W. 1878. A Few More Remarks about Printing - Treatment of Broken Negatives. The Practical Printer: A Complete Manual of Photographic Printing: 92-95.
Koob,Stephen P. 2006. Conservation and Care of Glass Objects. London,England: Archetype Publications Ltd.
Koob, Stephen. 1995. The Conservation of Archaeological Glass, with Special Reference to Chinese Glass. Proceedings of the XVII International Congress in Glass: Volume 6: 475-480.
Koob, Stephen. 2004. The Corning Museum of Glass: Object Handling.
Koob, Stephen. 1986. The Use of Paraloid B-72 as an Adhesive: Its Application for Archaeological Ceramics and Other Materials. Studies in Conservation 31: 7-14.
McCabe, Constance. 1991. Preservation of 19th Century Negatives at the National Archives. Journal of the American Institute for Conservation 30, no. 1: 41-73
McCormick-Goodhart, Mark H. 1992. Glass Corrosion and its Relation to Image Deterioration in Collodion Wet-Plate Negatives. The Imperfect Image: Photographs their Past, Present and Future: 256-265.
Munier, I., R.Lefevre and R.Losno. 2001. Atmospheric Factors Influencing the Formation of Neo-crystallizations on Low Durability Glass Exposed to Urban Atmosphere. Glass Technology 43: 114-124.
Newton, Roy and Sandra Davison. 1996. Conservation of Glass. Boston: Butterworth-Heinemann: 135-283.
Oakley, Victoria. 1999. Five Years on: A reassessment of aspects involved int he conservation of glass objects for a new gallery at the Victoria and Albert Museum. The Conservation of Glass and Ceramics Research, Practice and Training: 217-228.
Pasquariello, Giovanna, Fabio Talarico, Maurizio Coladonato and Simona Favetti. 1996. Problems in Conservation of Historic Glass Plate Negatives: State of Research and Possible Developments. International Conference on Conservation and Restoration of Archival and Library Materials: 241-260.
Romich, Hannelore. 1999. Historic Glass and its Interaction with the Environment. The Conservation of Glass and Ceramics: Research, Practice and Training: 5-14.
Salem, A.A. and M.Grasserbauer. 1994. Study of Corrosion Processes in Glass by a Multitechnique Approach. Part 1. Glass Technology 35:2: 89-96.
Salem, A.A., R. Kellner and M. Grasserbauer. 1994. Study of Corrosion Processes in Glass by a Multitechnique Approach. Part 2. Glass Technology 35:3: 135-140.
Taylor, Thomas H. 1999. In situ Repair of Archaeological Glass. Adhesives and Consolidants: Preprints of the Contributions to the Paris Congress, 2-8 September 1984: 202-204.
Tennent, Norman H. and Joyce H. Townsend. 1984. The Significance of the Refractive Index of Adhesives for Glass Repair. Adhesives and Consolidants: Preprints of the Contributions to the Paris Congress, 2-8 September 1984: 205-212.
Whitman, Katharine and Ralph Wiegandt. 2007. Case Study: Repair of a Broken Glass Plate Negative. Topics in Photographic Preservation 12: 175-181.
7.8 Online Resources
Mazzini, Mauro-Julian: http://glasnegspath.wordpress.com/