The Corning Flood:MUSEUM UNDER WATER

Aerial view of the Corning Museum of Glass surrounded by flood water, June 23, 1972.

The Corning Flood:MUSEUM UNDER WATER

The Corning Museum of GlassCorning Glass Center

Corning, New York 14830

Editor John H. MartinDeputy Director, Administration

Associate Editor Charleen K. EdwardsPublications Assistant

This project is supported by a grant from the National En-dowment for the Arts in Washington, D. C., a Federal Agency.Part of the research for the project was sponsored by TheNational Museum Act (administered by the SmithsonianInstitution).

A portion of the second chapter first appeared in the Wil-son Library Journal, November 1975, and is reprinted withpermission.

Published by The Corning Museum of GlassThomas S. Buechner, PresidentWilliam C. Ughetta, SecretaryVan C. Campbell, Treasurer

Copyright 1977 The Corning Museum of GlassCorning, New York 14830Printed in U.S.A.Standard Book Number 087290-063-0;ISSN 0022-4250Library of Congress Catalog CardNumber 77-73627

Cover: Islamic Pilgrim flask with dissolved plaster restoration.

ContentsPREFACE

INTRODUCTIONChapter I

Chapter IIChapter III

Chapter IV

Chapter V

Chapter VI

Part 1Part 2Part 3

APPENDIXI

IIIII

IV

VVI

VII

The 1972 FloodThe Restoration of the Glass CollectionThe Restoration of the LibraryThe Restoration of the Photographic andAudio Tape CollectionFumigation and Sterilization: A ReportDavid J. FischerVacuum Freeze Drying: A ReportDavid J. FischerPlanning to Protect an Institution andIts Collections—A ChecklistPrecautions to Be Taken before a DisasterProcedure during the DisasterActions to Be Taken after a Disaster

Museum Staff, June 1972Restoration Staff, 1972-1973-1974Museum Staff, 1976Volunteers Who Helped Reopen the MuseumGlass Center, Steuben, Foundation andCGW Employees Who Helped to Restorethe MuseumExamples of Glass Restoration Problemsand TechniquesMaterials and SuppliersAcknowledgements, References, Chapter IVAcknowledgements, Chapter V

PrefaceOn June 23, 1972, The Coming Museum of Glass wasflooded to a level of five feet four inches above thefloor, possibly the greatest single catastrophe borne byan American museum. The Agnes Flood broke hun-dreds of objects, saturated over half the Library (andall the rare books), ruined equipment, and coveredgalleries, cases, offices, furniture, and files with a thicklayer of slime.

Thirty-nine days later, on August first, the Mu-seum was reopened to the public, and four years later,in June of 1976, the Museum completed the task ofrestoring the glass collection and the Library. Thisbook describes the restoration process and offerssuggestions for disaster planning gleaned from experi-ence. The scientific research which lay behind therestoration of the Library, in particular the researchwhich led to the successful rehabilitation of booksprinted on coated papers through vacuum dryingtechniques, will be available in a separate publication.

The extent of the devastation suffered wasmatched by the help received. Among the first andmost substantial help was that from the insurancebroker, Mr. Huntington Block. His understanding at-titude and faith in our professional abilities permittedus to get the restoration process underway. The claim,painstakingly prepared under Dr. Robert Brill's direc-tion, was paid in full and has subsequently supportedthe Museum through restoration and much replace-ment of objects from the collections. Grants from theNational Endowment of the Arts and The NationalMuseums Act have improved the quality of theresearch/restoration process and enabled us to pub-lish the results—this book is an example of the En-dowment's support.

Literally hundreds of individuals came to help.From local children to eminent specialists from all overthe world, people arrived at the Museum to do what-ever they could—lugging silt-covered debris, hosingdown furniture, sifting mud for fragments, making lists,taking inventory, washing, testing, diagnosing, wrap-ping, readying, repairing—the work went on day afterday. They are too many to list here, but the Appendixlists what was recorded and is recalled.

Needless to say, most of the burden was borne bythe staff and they responded, without exception, mag-nificently. The incumbents as of flood time are listed inthe Appendix. In addition to Robert Brill who becameDirector after the flood and who designed and ad-ministered the restoration program, I would creditthree others for their extraordinary efforts. First, PaulN. Perrot, who was director and who returned from theAmerican Association of Museums annual meeting inMexico City as soon as word reached him. Already onhis way to an Assistant Secretaryship at the Smith-sonian, he faced the apparent ruin of twenty years ofhis own work in building the Museum. He pitched inand saw us through the emergency phase, past theAugust opening and well on the road to restoration.The second is John Fox, Director of the Corning GlassCenter and, therefore, the Museum's landlord. Takenby helicopter from the roof as the flood waters rose, hestarted planning restoration of the building and all itssystems before the flood receded. His continuous at-tention to the Museum's myriad needs made the Au-gust reopening possible. The third to whom specialacknowledgement is due is Raymond Errett, staffphotographer and restorer. He was the first to reachthe Museum after the flood despite the total loss of hisown home. He became and remained a major strengthin all critical matters during the entire emergencyperiod. His judgment, his knowledge, and his mannerprovided the kind of leadership the Agnes aftermathdemanded.

Thomas S. Buechner, DirectorThe Corning Museum of Glass

December 1976

Introduction

Introduction

In June of 1972 The Corning Museum of Glass and itslibrary were suddenly faced with catastrophe; withoutwarning most of the priceless collection of approxi-mately 13,000 objects and 13,000 books was sub-merged by flood waters, the aftereffects of a hurricane.

During the first part of that month, HurricaneAgnes had swept up the eastern seaboard but seemedscarcely a threat to a museum 275 miles from theAtlantic. At the beginning of the third week in June,Agnes (now designated a tropical storm) suddenlyswerved inland and then paused for three days ofcontinuous rainfall along the New York-Pennsylvaniaborder in the western part of these states (Fig 1). Onestatistically-minded observer estimated that theamount of water which fell would have filled a con-tainer twenty-five miles long, one mile high, and onemile wide—more water than had ever fallen before inthis inland area.

In the Coming-Painted Post area of western NewYork, three rivers join to form the Chemung Riverwhich bisects the city of Corning. On June 22nd theriver, normally a placid one- to two-foot deep stream,stood about fifteen feet deep, well contained within itstwenty-three and one-half foot dikes. Worried citizenswho remained awake that night were assured by localofficials on the radio at midnight that there was nodanger.

At 5:00 a.m. on June 23rd, the river suddenlycrested at twenty-seven to twenty-eight feet, toppedthe dikes, and poured into the town. (Fig. 2). Theaction of the current as it coursed over the dikes ateinto the rear of the embankments (Fig. 3). Before longthe dikes were badly breached in five places. Thetorrent which raged through the north side of towntore houses from their foundations, sweeping somedownstream and scattering others in the valley. Waterup to twenty feet deep raced over areas of the town.Without warning the city was split in half (Fig. 4).When flood waters also cut the approaches, Corningwas effectively isolated from the rest of the world(Fig. 5).

The isolation was complete—suddenly the com-munity was without electricity, natural gas, tele-phones, gasoline, drinking water, sewage disposal, ormain highway access. On transistor radios towns-people could hear about the flood in Elmira, eighteenmiles downstream, but it was impossible to learn ofanything occurring in an area only a few blocks away.The next morning, miraculously, the river was back inits banks, but it was almost three weeks before allutilities were restored to most of the city.

Fig. 1. Path of Tropical Storm Agnes, June 1972.

10

Fig. 2. Area covered by floodwaters, June 23, 1972. Note locationof The Corning Museum of Glass.Fig. 3. Floodwaters from Post Creek and the Chemung Rivertopping the dikes.

Unfortunately for the Corning collections, theMuseum was situated in the middle of the disaster area(Fig. 6); water had surged fifteen to twenty feet abovelower level on the west side of the Glass Center. Fivehundred and twenty-e ight of the approximately13,000 objects in the glass collection sustained dam-age. Chapter I covers restoration of the glass collec-tion; Chapter II is devoted to the devastation wreakedupon the library's holdings. Sixty-five hundred of theLibrary's 13,000 volumes and all of the rare bookswere underwater.

On that morning following the flood crest, therewas no question as to how high the flood waters hadrisen: from outside the Museum the mud line on theeast side was distinctly visible ten feet above theground level on the once-white curtains. Within thebuilding, the water line was clearly marked on thewalls and exhibit cases, 5'4" above the floor (Fig. 7). Atwo-inch layer of slippery mud covered the floor andmade walking hazardous; rugs were twisted in muddyheaps at random. So much mud obscured the coverglasses of the exhibit cases that it was impossible to seeinto the lower part of the cases to determine the extentof damage without a close examination of each case.Without electricity, many areas were concealed bygloom; leaden skies offered little hope of more light.Athwart the main gallery lay an eighteen foot longcase, fallen from its upright position, lying face down in

the mud. Within it had been seventeen delicate andextremely rare objects, made between the tenth andtwelfth centuries in Iran, specially exhibited to com-memorate the founding of the Persian Empire 2,500years ago. There was no way to evaluate the damagedone. The cover glass had obviously fallen out as thecase tipped, for shards of the once-clear glass werefound broken.

Groping through the darkened galleries a staffmember found that one case containing aneighteenth-century crystal candelabrum from Ireland(Ace. No. 50. 2.23AB) was missing; it was later locatedsixty feet away in the Glass Center's Hall of Sciencewhere the case had floated, the candelabrum shat-tered from surging back and forth within it. An exhibitcase from the pressed glass collection partially blockedthe entrance to the contemporary glass gallery.

Fig. 4. Coming's Northside and Southside separated from oneanother at flood height.Fig. 5. Eastern approach to the city closed by high water, isolatingCorning and surrounding areas; Big Flats, New York, partiallyunder water in foreground.

Fig. 6. Aerial view of The Corning Museum of Glass and GlassCenter, (a) Normal conditions, (b) Inundated by floodwaters.

11

Introduction

The Museum offices were in chaos. Furniture hadfloated freely and come to rest on top of other furni-ture. Books and papers had been washed about the sixoffices and lay in mud-covered piles (Fig. 8). Theswirling waters had pulled books and periodicals fromtheir shelves, tossed them among glass objects, top-pled a card catalogue, covered the files, and veneeredeverything below the 5'4" level with oleaginous mud.

The offices opened off the Museum's library, andthe library itself was in total disarray. The cardcatalogue had been completely submerged; it eventu-ally became so damaged by mold that it had to bediscarded. The rare book case had tipped over, cas-cading six hundred volumes of ancient vellum andpaper into the slime. Books with coated paper hadexpanded as they absorbed water, and the pressurehad burst their cases, adding those volumes to thefloor. Other books had expanded and locked them-selves into place; their shelves had fallen but the booksremained wedged in place. It would take crowbars toremove them. The Museum photographic collection,some 22,000 prints, and 22,000 negatives had beentotally submerged; sixty percent of the 50,000 colorslide collection had been under water.

Raymond Errett, the Museum restorer, andThomas Buechner, the Museum's President, were thefirst on the scene that next morning; groping his way tothe storeroom to obtain tools needed to open thecases, Mr. Errett encountered a large wooden storagecupboard that had fallen across the room; it had to beclimbed to reach the keys to the cases and suction cupsto remove cover glasses. Glass objects were scatteredacross the mud-laden floors, inviting further damageand making walking a difficult task, but the salvagingof the collection could begin.

12

Fig. 7. Flood level, building interior, (a) Mud line on curtains.(b) Line visible on wall.Fig. 8. Museum offices, contents mud-covered.

Chapter IThe Restorationof the Glass Collection

Chapter I

The large case containing a temporary "Tribute toPersia" exhibition was obviously the first unit to besalvaged. It lay face down across a portion of the maingallery (Fig. 9), its front cover shattered; the pieces ofglass in the mud around it could be part of either thecover glass to the case or of the seventeen ancientIslamic objects which had fallen with the case. Toolsobtained with difficulty from the workroom were usedto open the back of the case. Surprisingly, not all of theobjects had broken; the buoyancy of the air in the casemay have caused it to sink gradually while some of theobjects were floating inside. Other pieces had beensmashed and small fragments were picked up by thehandful. Two objects were crushed between the frontof the case and the floor (Fig. 10). All the mud from thecase and the surrounding floor area was sieved with atea strainer to recover all the shards necessary for theeventual restoration. (Strangely enough, months later,while one of the library books was being thawed be-fore restoration, a missing piece of glass from an Is-lamic beaker (Ace. No. 63.1.10) was discovered withinthe book).

Some of the damaged objects in the Persian casemay have been broken by small wooden pedestals onwhich the objects had stood and which had floatedaround within the case, moved by the currents of theflood water, eventually descending onto the objects asthe water receded. Pieces under the case may havebeen further damaged when it was moved.

With the Persian case out of the way, and itsfragments and objects stored in boxes for later sortingand restoration (pieces were kept with their objects;the inevitable box of "unidentified fragments" cameinto being later), attention was turned to the forty-footlong, six-foot high, and three-foot deep "GermanicTradition" glass case. In contrast to the Persian casewhere pedestals had not been secured, the Germaniccase consisted of a series of back lit shelf units withcurved, translucent, plastic supports on which the ob-jects had once rested and which could be hooked toany part of the back wall (Fig. 11). Since the boxeswere hollow and light, they had risen with the waterand their objects had slid off and floated down to thefloor of the case. As the water receded, the boxes hadeither come to rest on top of the objects or had pinnedthem against the back walls or the front cover glass—acover which had to be removed if the objects were tobe rescued. In one case, the problem arose of trying toremove two pieces of glass pressing in two differentdirections; one piece was successfully removed, but itscompanion piece broke under the stresses upon it.

14

Damage in this case totaled approximately 30%,eleven out of thirty-six objects. The first obvious losswas a reverse painting on glass from which a section ofpaint had come loose. Double-walled and paintedglasses were especially vulnerable to their watery ex-posure because they were held together by water-soluble substances. The lighter pieces had sustainedmuch less damage than the heavier engraved pieces,perhaps because the lighter objects had floated on topof the water or had sunk more slowly.

Once empty, the massive Germanic exhibitioncase was used as a storage area to protect other dam-aged objects as they were found. All undamagedpieces were left in their original cases after they hadbeen righted or emptied of standing water.

Cases of ancient glass in the East Gallery wereopened next. The interiors resembled scrambled glassjigsaw puzzles. Some pieces were piled one on top ofanother; flasks or vases with large bodies and smallnecks were found standing upside down, still full ofwater (Fig. 12). Objects previously repaired beforepurchase with plaster of Paris or water soluble glueslay in heaps on the shelves or the floor of the cases(Fig. 13).

In the South and West Galleries (Fig. 14), theVenetian and American cases contained clear plasticshelves, secured to plastic backs. Although the doorswere tightly fitted, the cases were not watertight (Fig.15). Since the water had risen and receded slowly,damage was not extensive. Some objects were on thebottoms of the cases; others hung by a lip or weresuspended, as if by a casual juggler, between theshelves. Many of the covered objects contained waterbeyond the rim, well up into the lids. It seemed impos-sible to open some of the cases without causing furtherdamage, but careful handling prevailed. Fortunately,not a single object in the Museum is known to havebeen broken during the clean-up process or themonths of restoration which followed.

Fig. 13. Islamic Pilgrim flask with dissolved plaster restoration.Fig. 14. Floor plan, The Corning Museum of Glass, June 1972.

14 Fig. 15. Venetian case with plastic shelves.

15

Chapter I

As noted, there had been two large cases at theentrance of the West Gallery containing a pair of par-ticularly fine late 18th-century Irish crystal candelabra(Acc. No. 50.2.23AB). Because of the darkness, noone had noticed at first that one of the cases wasmissing, having been floated or propelled the entirelength of the gallery (sixty feet), and that the other casestill on its original site appeared to be empty! A closerlook revealed the ironic vagaries of the flood waters.The candelabrum had originally rested on a 3/4"plywood false bottom base. The base had floated,dropped the candelabrum to the bottom, and, as thewater level went down, the base returned to its originalposition—but now on top of the candelabrum (Fig.16). The "disappearance" was complete; the damageto the hidden candelabrum was extensive.

The cases of English and French glass had to beopened from the top to retrieve objects resting onbases slanted toward the cover glass. Objects hadeither hit one another or had struck the cover glass,but damage was amazingly slight. On the other hand,a special exhibition of pressed glass was a victim ofindiscriminate destruction. One of the cases waswatertight, but had not been fastened to the floor. Ithad been pushed, or bobbed like a cork, down thegallery; its cover glass had broken and it had thenfallen over, resulting in damage to every one of itsobjects.

The Study Gallery, an area forty by forty feet withshelves eleven feet high and containing seventy per-cent of the Museum collection, sustained remarkablylittle damage, considering the number of glass objectsit housed, approximately 7,000 objects. The narrowshelves, six inches wide and three feet long, left littlespace for free-floating glass. Two shelves had col-lapsed after the water receded due to the weight of thewater still standing in large objects. Because of thedisplacement of air by water, objects were sometimespiled three high and were leaning against the coverglass, making it difficult to reach the pieces since thecover glass had to be removed to get at the objectsimpinging upon it. Wire hooks were used to separatethe components of such piles as the glass covers werecarefully inched open (Fig. 17).

Two reverse paintings on glass (Acc. No. 53.3.32and 66.6.1) which were damaged were sent to thenoted Conservation of Historic and Artistic WorksCenter, Cooperstown Graduate Center in Coopers-town, New York. At that time the state of the art wasnot such that the paintings could be repaired. Onewhole painted surface of one of the reverse paintings(Acc. No. 53.3.32) had come loose and was in dangerof crumbling as the paint had rolled back and couldnot be made to lie flat. In time, it was flattened andstored with acid-free cardboard holding it in place.Recent technical advances give hope that the paintcan be re-adhered to its glass pane.

Fig. 16. Damaged Irish candelabrum beneath itsfalse bottom.

Fig. 17. Damage in the Study Gallery, (a) Collapsed shelf (b) Glassstill filled with water, balanced on two other glasses. Objectsleaning against the cover glass (c) (d)(e) Filled German drinking vessels balanced against one another.

16

Once the cases in the galleries had been opened,the workroom and glass storage areas were the nextconcern. The open shelving in the storage area hadpermitted glass to move freely and consequently thedamage was heavy. A large collection of Americanflasks had been spread on tables for study; the flaskswere everywhere, and several were broken, unfortu-nately. Careful sifting, sorting, and examination of themud for fragments as well as whole objects were es-sential.

As the cases were opened, an inventory wasmade. All cases were numbered, their contents listed;all were photographed. Within two weeks the inven-tory was complete. The problems of glass restorationaccumulated and they were overwhelming. Lendingurgency to restoration efforts was the decision toreopen The Corning Museum of Glass to the public onAugust 1, thirty-nine days after the disaster. Thosesurrounded by flood damage were not at all sure itcould be done, only that it would be done.Glass Statistics in June 1972Total number of objects in the collection 13,000Total number of damaged objects

in the collection (4%) 528Objects requiring restoration 481Objects not to be restored 38*Objects to be restored eventually if suitable methods

can be found 9

*Either more economical to replace than restore or damage tooextensive to permit restoration.

Distribution of losses according to categories:Ancient 34%English 12%European 36%American 15%Other 3%

Help came from many areas. Volunteers wereeager to do anything asked of them (Fig. 18); a contin-gent arrived from Rochester: Holman J. Swinney, Di-rector, Strong Museum, Rochester, and volunteersDeborah Bacon. Peter Briggs, Jean Farnham, PeterHotra, Lynda McCurdy, Robin and Sandra Stone; W.Stephen Thomas, Director of Development and Pub-lic Affairs, Rochester Museum and Science Center,and volunteers Karen Bennett, Lilita Berg, GeorgeHamell, Robert Martin to wash with care all the objectsstill intact in the Study Gallery. Others came fromToledo: Michael Moss, Assistant Curator, The ToledoMuseum, Toledo, Ohio, and Karen Smith; Drew Oliverfrom the Brooklyn Museum, New York; from Ontario,Barbara Snider of Upper Canada Village; and Newark:Susan and Ron Auth, Curator and Archaeologist at theNewark Museum, to help. A temporary repair shopwas established in one corner of a former supermarketdirectly across the street from the Museum. Dubbed"the Acme Building" in honor of its former tenants,the restoration area became the center of the painstak-ing, deliberate program to reopen the Museum byAugust 1, 1972, with as many of its prized objects onexhibit as possible. Repair of the Venetian dragon-

18C

Fig. 18. Volunteers (a) In the Study Gallery(b) Washing flooded objects(c) Younger volunteers pause before the next task.

17

Chapter I

stem goblet (Acc. No. 51.3.115) was first on the prior-ity list; it had long been a symbol of the Museum'scollection, and was to become the symbol of its recov-ery. The first goal, an August opening, was met, and bythe autumn of 1976, all of the glass restoration wascomplete.

As with all specialists, a glass conservator mustpossess a very special feeling for the material withwhich he or she works. It was, therefore, even moreheartbreaking to survey what had to be done. Fivehundred and twenty-eight objects were damaged, andfour hundred and eighty-one were listed for repair orrestoration. It was estimated that three years would beneeded to complete the task, including time for pho-tography, report writing, and general laboratory sup-port activities. The initial step was for the restorer tohandle each piece and attempt to determine howmuch bench time would be needed to repair or restoreit. All of the fragments of each individual object wereplaced in a box marked with its accession number (Fig.19). The extent of the damage to each piece couldthen be analyzed.

A professional restorer, Rolf Wihr, came from theRheinisches Landesmuseum in Trier, Germany, for amonth to repair specific pieces requiring his particularexpertise. His wife Dorothea accompanied him andhelped in many ways (Fig. 20). The former MasterRestorer at the Metropolitan Museum of Art, W E.Rowe, was called from retirement for a year to utilizehis vast knowledge in restoring sixty-one of Coming'sflood-damaged objects.

Albert Fehrenbacher, a German model maker,refurbished a hand-carved model of a nineteenth-century glass factory, and then began the seeminglyimpossible task of transforming glass bits and pieces totheir former appearance as altarpieces or other intri-cate compositions of the lampworker's art. Severalstudents worked for periods varying from threemonths to six months: Lynda Aussenberg, New YorkUniversity Conservation Center of the Institute of FineArts, New York, New York, Summer 1973; WilliamWarmus, University of Chicago, Summer 1972, 1973,1974; Steven Weintraub, New York University Con-servation Center of the Institute of Fine Arts, NewYork, New York, six months, 1974. Each had somespecial skill to contribute. Two apprentices, JeffreyGaines, December 1973-July 1974, and Kristin Amy-Ion, June 1974-June 1975, also helped reach the goalof complete flood restoration by the end of 1976.

It was the Museum's curators with their intimateknowledge of the collection who decided what couldbe replaced and what had to be restored. A blue Sasa-nian ewer had been smashed into innumerable frag-ments (Acc. No. 65.1.23), but it was irreplaceable andimportant. It was the task of the conservator to restorewhat could not be replaced. The result has been amethodical program of repairing flood damage as wellas improving past restoration practices. Many olderadhesives had simply dissolved during their waterycontacts; pieces repaired with plaster of Paris beforepurchase by the Museum had fallen apart. The first listof single objects to be repaired was quickly revised,

Fig. 19. Fragments classified for damage appraisal.Fig. 20. Dorothea Wihr assisting the glass conservator. Fig. 21. Cleaning broken glass surfaces with ammonia, (a) (b)

18

since it was found to be more useful to restore all thepieces in one exhibit case before those in the next casewere begun. An object was photographed, cleaned,and assembled temporarily with tape; if it lacked afragment, it was set aside until the missing pieces of thepuzzle reappeared.

The enormity of the task was obvious, but thework which began after the flood followed the patternsestablished by glass conservators, disaster or not. Theonly special and unfamiliar problems posed by theflood-damaged objects were those of cleaning. Floodor no flood, glass repair and restoration involve prob-lems of simple breaks, complex breaks, and objectswith missing pieces. A survey of traditional techniquesmay be helpful before a review of specific flood-related problems is made.

Careful evaluation is required before any work isbegun. The nature of the damage, the inherentstrength or fragility of the glass and its decoration, thehistorical value of the object, and the desired finaleffect all must be considered. Repairs, too, range, fromthe simple to the complex.

To repair a break with two or three detachedpieces, the broken surfaces are first swabbed withammonia (Fig. 21), followed by acetone to remove allforeign matter. An even dispersal of ammonia is agood indication that the adhesive will disperse uni-formly. Educated and sensitive fingertips are neces-sary to insure a proper fit. One can "feel" when theglass pieces are correctly seated before the pieces aretaped together. With a fine spatula, an epoxy resin isapplied to both surfaces in extremely thin layers toachieve a close joint. Excess resin is removed before itbegins to harden.

The proper adhesive must be chosen. To addstrength, a permanent epoxy resin is usually used atCorning instead of less permanent adhesives. A lesspermanent adhesive may eventually loosen, and anobject may fall apart on its shelf or in someone'shands, causing further damage to the object or to anynearby object. Araldite AY103 has been used success-fully in Europe for years and at The Corning Museumof Glass since 1965. It has the same refractive index asmost glasses when applied in thin layers.

For objects with complex breaks, it is even morenecessary to assemble the entire piece with tape.Scotch Magic Tape is flexible and adheres well. It isusually applied to the side without any decoration ordesign, always on one side only. Again with a finespatula, a low-viscosity adhesive is then run into thecracks on the untaped side. The fluid resin seeps intothe cracks and fills each void. Excess resin is carefullyremoved, for even a slight shift could introduce airbubbles, weaken the joint, and distort original shape.As the adhesive sets, whether the objects have eithersimple or complex breaks, they are either self-supporting or are supported by the tape, a sand table,or various laboratory stands.

Cone-shaped or narrow-necked objects presentspecial problems. If both sides are not readily accessi-ble, the object can be cut apart at a convenient spot tofacilitate inside access and put back together with tapeto assure proper seating before the adhesive begins toset. After the adhesive has set, the two halves can stillbe separated for cleaning and further restoration.

The Corning Museum of Glass restoration policy isto return an object as much as possible to a semblanceof the original so it can be appreciated for what it was.Integrity of form, color, and decorative effect are im-portant factors. Restoration is based on the un-equivocal evidence which remains and the anticipateddegree of stability.

There is no attempt to duplicate engraving,enameling or any other conjectural stylistic feature.Restoration is never carried to a stage where it wouldbe deceptive or misleading. Restored features shouldnot detract from the overall appearance of the object,but should become visible to the careful observer.Detailed restoration sheets are kept and filed eventu-ally in object folders along with other information con-cerning the object.

If a small area of glass is to be restored, impres-sions from another part of the object are used to makea mold from dental wax coated with polyvinyl alcohol(to prevent resin from sticking to the wax and absorb-ing any of its color) and then transferred to the sectionto be restored. A small opening is left as a pouringspout. Plastogen G is frequently used at Corning; it is acolorless transparent polyester resin which cures atroom temperature in approximately twenty minutes. Ifthe impression is accurate and properly filled with theappropriate resin, the wax can be removed easily andthe replica is complete after removal of the spout.Acetone is never used for surface cleaning of Plasto-gen G because of its tendency to attack the restoredarea. Technovit 4004a is another useful resin becauseof its low shrinkage, apparent mechanical and colorstability, and rapid setting time. Color can be impartedin both of these resins by adding textile dyes beforecasting. Both transparent and opaque colors are pos-sible.

Restoration of an object with many missing piecesmay be much more complicated. All fragments mustbe assembled; the exact shape and size of the originalobject must be established. The existing curves of abowl, for example, may be used to extrapolate thebroken curves. A template can sometimes be made toturn the desired shape on a wheel. With the pieces intheir proper position, restoration can begin.

A clay model of the exterior of the missing part iscovered with a thin coat of silicone rubber thickenedwith Cab-O-Sil to the consistency of cake frosting. Thesilicone is then covered with a layer of plaster to givestrength and to help retain the shape. The clay sili-cone-rubber procedure is repeated when the object isturned over for work on the inside, with the clay the

19

Chapter I

same thickness as the glass. Before the silicone rubberis applied inside the bowl, a clay pouring spout isattached and possible air traps are eliminated by add-ing clay sprues to release the air. Sprues must be verysmall in diameter in case they have to be removed.Plaster is added to the inside, sometimes with a handlefor easy removal; the handle may be made of wire ormay be a large wooden screw embedded in the plas-ter.

After the plaster and silicone in the inside areremoved, the clay is carefully detached and the moldcleaned, leaving the glass in place. Pouring spout andsprues are cleaned before the silicone and plastermold is carefully sealed in place with a small amount ofuncured silicone rubber to prevent leakage. The moldis then filled with resin, preferably in the morning incase of a leak. All resins used must be compatible withthe silicone. Once the resin has cured, the plaster andsilicone may be removed and restoration is complete.

The Corning Museum of Glass flood restorationhas covered what must surely be the full range of glassrestoration and utilization of every conceivabletechnique. Specific examples are described and listedin Appendix IV.

Any discussion of glass restoration techniquesshould include the inevitable list of miscellaneous ad-vice based on years of experience and the practices ofa particular museum. The Corning Museum of Glass isno exception. Its brief list follows;

• Handle every object as though it were broken andwould fall apart in the hand. It may have been crackedor repaired previously, leaving it in an unstable condi-tion.

• Always glue in the morning so that an object may bewatched all day in case it slips, slides, shifts, moves.

• Make certain that the adhesive does not remove anypart of the design. Gilt, for example, must be handledwith special care; it detaches very easily.

• Do not allow resins or adhesives to come in contactwith any objects with crizzled surfaces because thesurfaces will absorb adhesives and resins, removal willbe impossible, and the composition of the surface willbe permanently altered.

• Be as consistent as possible in mixing resins. Bothobject and adhesive should be warmed to approxi-mately 100° F. and temperature maintained untilcured.

• Do not attempt to hide restoration, although it shouldnot be overwhelming. Paint changes color in differentlights; dyes are difficult to eradicate. No attempt ismade to eliminate weathering at Corning Museum ofGlass.

Today's Corning Museum of Glass visitor will findno obvious effects of the flood or its damage. All of theobjects incomparable for their historical importance orfor their beauty have been returned to their assignedplaces in the galleries. Meticulous repair and restora-tion have not only preserved the collection, but inmany instances corrected earlier restoration practices.

Those who ask what would happen if anotherdisaster occurred receive a ready answer. Everyone atthe Museum now has a handbook listing detailedemergency procedures in case of fire, flood, tornado,smoke, power failure, or vandalism. No glass would beremoved from the building. Those pieces most likely tobe damaged by water would be placed in boxes andmoved to an office on the highest level of the building.Test evacuations are held periodically. Greatest flooddamage was caused by floating plywood pedestals andboxes; all are now anchored.

No magic wand will solve problems precipitatedby a disaster. People determined to do what must bedone are the most important resource. There is nosubstitute for the glass conservator who surveys theviolation of years of glass history and of his own effortsto preserve that history, and begins immediately to puteverything back together again.

20

Chapter IIThe Restorationof the Library

Chapter II

Few visitors to The Corning Museum of Glass realizethat for the past quarter of a century the Museum hasbeen developing the library of record on the art, ar-chaelogy, and history of glass. With the exception ofinformation on the current scientific, technical, andmanufacturing aspects, the Museum library acquiresall published books and periodicals issued about glassand allied areas. The collection also includes originalsource material such as trade catalogues, archives,documents, pamphlets, prints, photographs, slides, aswell as supporting material—movie films and printedephemera. In addition, the Museum exhanges printedmatter with countries world-wide and publishes itsown books and catalogues.

Such diversity compounded the range of the res-toration problems confronted after the flood; thesecan best be understood by a listing of the collectionand its post-flood condition:Pre-Flood Collection Number of Flood Damaged Items13,000 catalogued volumes600 rare books3,750 periodicals50,300 slides22,270 photographs2,252 prints20,000 negatives100 films3 drawers documents4 drawers vertical file80 linear feet archivesoffice files

6,500 volumes600 rare books

3,111 periodicals36,400 slides

22,035 photographs2,032 prints

20,000 negatives65 films

3 drawers documents4 drawers vertical file

72 linear feet archivesoffice files

The devastation which the flood waters hadwreaked upon this collection was perhaps even great-er than that suffered by the glass collection. On themorning of June 23, 1972, muddy water had swirledthrough the offices and library to a depth of five feet,four inches. As noted previously, wet books had ex-panded and bent the sides of the stacks so that theybulged into a "V" shape (Fig. 22). These books wereso tightly wedged into the bookcases that crowbarshad to be used to dislodge them. Other shelves, sud-denly sprung free from the sides of the misshapenstacks, had dumped their volumes into the mud andwater (Fig. 23). The rare book and manuscript collec-tion, shelved separately, had fallen into the slime asbookcases overturned.

During the first attempts at salvage, librariansNorma Jenkins and Jinny Wright remembered hearinga lecture by noted conservator Carolyn Horton whorecommended freezing water-damaged books. Shehad written an article for the Wilson Library Journal onthe restoration of flood-damaged volumes after theFlorence flood, and that article, fortunately, was notflooded. It was decided to follow her suggestion tofreeze the books. However, all utilities were knockedout, and without electricity there were no workingfreezers. Transportation and supplies were non-existent, not to mention the lack of running water forcleaning mud from books. Through the assistance ofthe Corning Glass Works, community volunteers, andlocal merchants, supplies were obtained to begin the

24A

Fig. 22. Shelves expanded to V-shape.Fig. 23. Books, periodicals on the floor.

Fig. 24. Flooded Special Collection removed from the mud forstorage in the freezer, (a) (b)Fig. 25. (a) Freezer truck parked at Museum side door.

22

job of saving the Museum library. Bill Warmus, a sum-mer intern, found an empty home freezer in Hornby,outside the flooded area; another was located in BigFlats. Glass Works emergency radio transmitters wereused to find a truck trailer which could freeze materialsas they were loaded into it.

Although it was cool for June, it was still raining,and mold would soon develop on the books. Rescueefforts were thus concentrated at first on the valuableSpecial Collections which consist of medieval manu-scripts, incunabula, and rare materials. These dam-aged volumes were picked out of the mud (Fig. 24),and wrapped in paper towels for ease of handling(because of their slimy covers), to absorb moisture,and to separate volumes from one another when theywere finally frozen. Had the books been wrapped inplastic, which unfortunately was unavailable, theywould have been easier to separate later—and thebooks would have been removed in a state of dampequilibrium. (It was discovered later that books withcoated paper—i.e., paper finished by applying gelatin,starch, casein or some other substance to produce aglossy surface—which had partially dried had sufferedirreparable damage.) Within a matter of days thesebooks were boxed, packed in dry ice, and flown byprivate jet to Carolyn Norton's laboratory for evalua-tion and eventual restoration. The rest of the collectionwas loaded onto the freezer truck parked outside theMuseum doors (Fig. 25). At first, packing was done in

an orderly fashion, but the urgency to freeze materialbefore mold developed eventually led to packing thevolumes without labels or organization. Five dayswere spent boxing the damaged collection of approx-imately 10,000 books, along with pamphlets, researchfiles, catalogues, and Museum office files. By the timethe last volumes were placed in boxes, the first moldhad begun to appear on some of the books.

Items which were beyond repair or those whichwere of current or ephemeral nature were discarded.This included part of the periodical collection—in themisguided expectation that it would be relatively easyto replace. The librarians later realized that everythingshould have been frozen so that decisions to save ordiscard could have been made more rationally at a lessurgent time.

The card catalogue was abandoned, for example.It was muddy and tightly wedged into a warped andbroken wooden case. Its loss was a great deterrent tothe effective use of the rehabilitated library; had itbeen frozen, it could have been dried and cleaned bythe conservation staff. The shelf list, which was equallymuddy, was dried in a staff member's garage (Fig. 26).Photocopies made of the cockled cards became thebasis for restoration efforts and for insurance claimswhich required months to assemble. The accessionnotebooks were also salvaged, which made it possibleto ascertain purchase prices of most of the floodedmaterials.

26B 26C

Fig. 25. (b) Collecting material to be boxed for truck, (c) Sealingboxes of books to be frozen. Fig. 26. Shelf List retrieval, (a) (b) (c)

23

Chapter II

With the books frozen, the library staff turned tonon-book holdings. Corning Community College, lo-cated on a hill above the flood plain, cleared the mainand an auxiliary reading room of its library to provide aplace to work, a place to begin the restoration process,and a space to store the 6,500 undamaged books inthe collection (Fig. 27). A temporary library office wasorganized in one corner of this restoration area (Fig.28).

The saturated art print collection (Fig. 29) and theoriginal document collection (Fig. 30) had not beenfrozen and were taken to the college library. Workingwith meticulous care, volunteers separated the frailpapers, opened picture frames to remove the prints,and blotted the excess water. Carolyn Horton and herassistant, Madeline Braun, traveled to Corning with astation wagon full of blank newsprint and canteens offresh water; both were used to save these unique studyresources.

The slide and photographic collections (over72,000 items) were considered next. The photo-graphic negatives were frozen as they were freed fromtheir flood-swollen files. The mounted prints (approx-imately 11,000 items), which had begun to form asolid mass by the time they were extricated from theMuseum, were immersed in a private swimming poolwhere it was possible to separate them and removemost of the mud (Fig. 31). Unfortunately, the chemi-cals in the pool adversely affected the emulsion andthey had to be destroyed. Following professional ad-vice, volunteers worked on the color slide collection

(36,400 damaged items) at the college library. Seventhousand slides were removed from their aluminiumand glass mounts, individually washed in rain waterwhich had been collected (there was no water servicefor three weeks) (Fig. 32), and hung on lines to dry.Paper clips were used to keep the film, the mount, andits label together for future identification (Fig. 33).Despite this effort, it was decided later that the photo-graphic images had become so degraded by the waterand mud that the flooded slide collection had to bediscarded. Rephotographing to return the collection toits 1972 status may cover a ten-year period.

Of the approximately 11,000 unmounted glossyphotographs which were frozen after the flood, about6,500 prints were cleaned successfully, while 4,000photographs, not frozen until four days after the flood,were too damaged by mold. Some frozen 4" x 5" and8" x 10" color transparencies and color negativeswere also successfully restored. The worst damagesustained by these photographic resources, however,was not due to mud and mold but was caused by thewet glassine envelopes in which they were stored. Thecockling of these envelopes combined with the water-softened surfaces of the film to produce a ripplingtexture on the picture surface. Those photographsfiled in Kraft paper envelopes escaped such damage.The photographic collection is now stored in acid-freepaper envelopes which do not cockle.

Once the holdings within the Museum were inprocess of restoration, attention was turned to materialin storage warehouses. The wisdom of freezing the

31Fig. 29. Print collection spread on floor to dry.Fig. 27. Improvised shelving for relocated Fig. 30. Part of the document collection.volumes in the college Library, Fig.30. Separation of prints from mounts in a swimming pool, the only availableFig. 28. Librarians at work in temporary office, source of water.

24

wetted paper immediately became evident, since thedamaged back issues of periodicals in storage wereruined beyond repair. As the stored back issues ofMuseum publications could not be saved either, thefrozen library copies gained greater importance.

On August 1, 1972, thirty-nine days after the dis-aster, the Museum reopened its doors and welcomedthe public. In late August when the college needed itslibrary for student use, the Museum library moved intoa former supermarket across the street from theMuseum (Fig. 34). The frozen collections werebrought together in a rented freezer-trailer outside thebuilding. That autumn was spent in further reorganiza-tion and in preparing the insurance claim. This lattertask became an eighteen-month project. More timehad been involved in the initial steps to conserve thecollections and to re-organize than had been antici-pated.

It was not until February, seven months after theflood, that work could begin on book restoration. Thesupermarket was divided into work areas for Museumstorage, glass restoration, the undamaged library col-lections, staff work space, and a paper (book) conser-vation area. Sinks, drain boards, tables, a fume hood,and a refrigerator-freezer were acquired. Two specialpieces of equipment were added: a dielectric dryer wasborrowed from the Corning Glass Works ResearchLaboratory, and a fumigation chamber (later modifiedto sterilize as well) was purchased. It was now possibleto set up a restoration assembly line.

Financial problems as well as problems of restora-tion had to be solved concurrently. Until there was aninsurance settlement, it would be necessary to borrowagainst the regular operating budget to support therestoration project. Clearly this could endanger theregular operation of the Museum. But until the restora-tion project began, it would be impossible to estimateeither the time such restoration would take or its even-tual cost. Without such figures there could be no set-tlement with the insurance company to obtain themoney needed to undertake restoration. The decisionwas made to spend beyond the budget, and bookrestoration proceeded concurrently with learning howto estimate the cost of restoration for the insuranceclaim.

Carolyn Horton was asked to restore one hundredof the finest rare books which had already beenshipped to her in a frozen state. While those bookswere receiving out-of-house professional care, it wasdecided to restore the rest of the special collectionin-house. An experienced paper conservator and aphysical scientist were hired, the latter funded by agrant from the National Museum Act. Together theydeveloped techniques for safe restoration of the collec-tion. A staff of local adults was employed for the actualrestoration work using t radi t ional conservationmethods. The scientist was commissioned not only toassist the paper conservator but to develop, apply, andtest new or unorthodox techniques to help reducerestoration time and to bring about as complete arestoration of the collections as possible.

25

Chapter II

The freezer truck-trailer was a continuing prob-lem. It was overcrowded and disorganized, an obstacleto any assembly line (Fig. 35); its refrigeration unit wasunreliable. During one brief failure, the temperaturewithin the truck rose to 30° F. Mold could have de-veloped and damaged the books if the problem hadnot been resolved quickly (Fig. 36). Two large walk-infreezers were therefore purchased and erected withinthe restoration building. These cost no more than twoyears' rent for the truck, and at the end of the project,resale returned half of the original cost. The new freez-ers had a row of shelves against the outside walls and athird double range of shelves in the center, so a logicalorganization and more efficient handling of materialswas now possible.

Individual treatment of every book would havebeen a task of five years or more. There was not thetime, money, or staff for a job of such magnitude, so agoal of Christmas 1975 was set for completion of theproject. Accordingly, a number of decisions werereached.

It was agreed that all volumes from the regularcollection which could be purchased for less than fiftydollars per volume would automatically be discarded,a proposal endorsed by the insurance companies. Theunderwriters concurred that it would cost at least asmuch—and probably more—to restore the individualvolumes. It was also decided to discard the periodicalcollection if replacement copies were readily available.Either these could be replaced through purchase

(second-hand), or a reprint could be purchased inmicroform. (The Kraus Company, a reprint andsecond-hand magazine firm in New York, was mostgenerous in helping in this area, and they deservecredit for their assistance. Swetsand Zeitlinger, Period-ical Agents at Berwyn, Pennsylvania, were equallygenerous in offering to evaluate the periodical collec-tion without fee so that a settlement could be reachedwith the insurance companies.) If journals could not bepurchased, the librarians planned to obtain permissionto microfilm journals in other libraries.

The accessions librarian began soliciting cata-logues from book dealers throughout the world so thatout-of-print volumes could be ordered as they ap-peared. Two staff members went to New York City tobuy in the second-hand market. Foreign languagebook dealers were also commissioned to assist in thesearch, since forty-seven percent of the collection is inlanguages other than English.

26

It soon became obvious that no more than fifteento twenty percent of the collection could be purchasedin the market since the items were so specialized; thusup to eighty-five percent would have to be restored.

In sorting and preparing to move books into thewalk-in freezers, every volume had been examined.The condition and value to the collection were re-corded on a worksheet in multiple copies (Fig. 37),one for the librarians, one for the conservator, and oneto remain with the book. As a book came out of thefreezer, the conservator checked the sheet for the lib-rarian's considered judgement: simply to "clean anddry," or whether to restore and by what method. Hehad a choice of traditional methods and, eventually,new methods. Some books were held for future deci-sion, as it was hoped they could eventually be foundon the open market. If their replacements were notfound before the restoration process neared its end,they would be restored as well.

The use of the two types of restoration—bymethods traditional or innovative—represented amajor policy decision in order to complete the taskwithin the allotted time. As a result, the collection wasdivided into two parts, with the more rare books beingtreated in accordance with established procedure, andother volumes—those of more informational than in-trinsic value—receiving innovative handling. Threeapproaches were used:

1. Traditional Techniques. Volumes of intrinsicvalue in their original condition were thawed (Fig. 38),removed from their bindings (Fig. 39), separated intosignatures and then into pages. These pages weresupported on sheets of non-woven polyester andwashed in water (Fig. 40). Running tap water bathsremoved most of the tenacious mud, water spraysremoved still more, while physical agitation with acamel hair brush removed almost all the rest (Fig. 41)but left a stain where the mud had been. Bleachingcould significantly reduce staining, but it weakenedfibers and thus was not used. The pages were thendeacidified and buffered by a thirty-minute immersionin a magnesium carbonate solution, a simple, inex-pensive, and effective treatment. They were then air-dried on tables or on nylon fishline drying racks (Fig.42).

<«.!

41

Fig. 40. Separated pages on support sheets, (a) (b)

Fig. 41. Use of camel hair brush to remove mud.Fig. 42. Deacidified, buffered pages on support sheets drying onnylon fishline racks.

27

Chapter II

Pages were collated, sterilized in the vacuum-chamber fumigator with a mixture of 12% ethyleneoxide and 88% freon, and wrapped in permalife paperwith acid-free board for storage until they could bemicrofilmed (Fig. 43). If necessary, distorted bookswere pressed between weights to return them tonear-original shape. Each packaged book had itswork-sheet with the complete history of its restorationattached to the paper wrapping.

2. Dielectric Drying. A dielectric dryer was ob-tained from the Corning Glass Works. This device isabout as large as a refrigerator with a chamber approx-imately eight inches by twenty inches (Fig. 44). Insideare two electrode plates between which a frozen vol-ume can be placed. A burst of electrical energy is sentthrough the frozen book to begin thawing it. It wassoon discovered that the covers should be removedfrom all books being treated, for they contained muchwater (Fig. 45); it was also found that unremovedstaples or paper clips caused arcing between the elec-trodes, charring or burning the books.

By trial and error, the duration and the number ofelectrical bursts necessary to dry the varous kinds andsizes of books were determined, and like items weregrouped to simplify the task.

Valuable books were only thawed, not dried, bythe dielectric method and made ready for traditionalhandling. It was not known if permanent damage wasbeing done to the paper with the electrical burst (e.g.,permaturely aging it). In addition, all dried papers and

books were sterilized in a vacuum fumigation chamberin order to prevent further spore growth, a continuingproblem, as all flooded material is likely to harbormold spores.

3. Freeze-thaw Vacuum Drying. Much thoughtwas given to freeze-thaw vacuum drying, but it wasdecided that the technique was not then practical. Tobegin with, there were no vacuum chambers nearby,and most units were too small to handle materials inthe amount on hand. It was feared that books withcoated paper might "block" into a solid mass in thedrying process. Many books with plates were printedon coated paper, paper which tends to become a solidmass after wetting and exposure to air.

Last, and of considerable concern, was the realiza-tion that vacuum drying might break down paperstructure and would not remove the mud or stains onthe papers. It would probably be necessary to rewashall such dried volumes, and it might be impossible toeradicate all the flood stains. Nonetheless, somemedieval parchment manuscripts were sent to the Li-brary of Congress for experimental freeze-drying,since their vellum leaves presented other problemswhich might respond to that treatment.

Microwave drying was also tried, but it was notsatisfactory. Since only domestic microwave ovenswere available, the experiment was dropped as beingnon-conclusive. Experiments in the use of ultrasoniccleaning and drying (a technique successfully used incleaning pieces of glass in the collection) were tried

44B 46A

Fig. 43. Restored books being wrapped for storage.Fig. 44. (a) Dielectric dryer (left), (b) Dielectric chamber.

Fig. 45. Charred book cover after first Dielectric burst.Fig. 46. (a) Truck unloading frozen Corning Museum of Glass li-brary books at Valley Forge.

28

also. Other experiments of a limited nature were alsomade with solvent extraction techniques but were onlya qualified success.

Within the first year, 3,000 books and thirty boxesof files had been dried, cleaned, and readied for mi-crofilming by traditional methods. By this time, thescientist had established the behavior of coated paper(see Chapter V) discovering how and why it turnedinto a solid mass after wetting and then drying. Basedon his experiments, approximately 3,500 frozen vol-umes (primarily of coated paper) were taken by re-frigerator truck to Valley Forge, Pennsylvania, to theGeneral Electric "space chamber" to be thawed andvacuum dried (Fig. 46).

After nine days of treatment, ninety-five percentof the coated stock (2,850 volumes) was saved. Thiswas the first time, to the librarians' knowledge, thatsuch a success rate has been achieved. The five per-cent loss may be attributed to books being partiallydried before they were sent to the vacuum chamber;some may have blocked on the library shelves whilewaiting to be frozen; others may have lost waterthrough sublimation in the freezer. In view of thefreeze-drying problems of the Temple University LawLibrary, following its disastrous fire in July 1972, it isimportant to stress the need to monitor the freeze-drying process. If books are not completely dry whenthey come out of the freeze-dry chamber, they must bereturned for further drying or else they must be air-dried. Some one hundred and twenty-five damp

books were carefully hand-dried on their return to theMuseum Library (Fig. 47). All the dried but still cock-led and dirty books were cleaned during the next eightmonths by erasing the flood dirt with draftsmen's dry-cleaning pads and camel hair brush (Fig. 48). In De-cember, 1974, the book restoration departmentclosed, twelve months ahead of schedule. Almost6,000 books were dry and clean, and thirty cartons offiles were also dry. Most of the art prints, photographs,audio tapes, and sound films were restored to useful-ness as well.

The last phase of restoration then began—themicrofilming of the flooded library collection (Fig. 49).Duplicate copies of the film were turned into micro-fiche format. This was done for a number of reasons:

1) To preserve the information in the restoredbooks in the event that they disintegrate prema-turely.

2) To insure the collection against another catas-trophe; one master copy was stored outside thevalley, and a second copy was kept in mic-rofiche form for use in the library.

3) To provide information to out-of-area research-ers and to decrease physical handling in inter-library loan of rare or one-of-a-kind volumes.

48A

29

Chapter II

4) To make holdings of the library available tospecific libraries in the United States andabroad in exchange for the privilege of mi-crofilming unobtainable items in their collec-tions. Due regard to copyright laws is observedin all cases.

The three-and-a-half year period after the floodwas difficult for the library staff, which continued tosupport all the diverse Museum activities and at thesame time to be involved in a monumental salvage andrestoration undertaking. Museum personnel havetaught conservation seminars so that others mightlearn about their techniques, and the Museum haspublished this book in the paradoxical hope that it willprove useful—but never have to be used.

Fig. 49. Microfilming a restored volume.

30

Chapter IIIThe Restorationof the Photographic andAudio Tape Collection

Chapter III

Perhaps the greatest single loss to the Museum collec-tions occurred in the photographic department (seepage 22 for the listing of this collection). Althoughmassive efforts were made by staff members and vol-unteers in the week after the flood, much of the collec-tion could not be restored to the professionally usu-able state required by the Museum.

As has been indicated previously, those films,negatives, photographs, and slides which were inad-vertently sent to the freezer were in the long run sal-vaged with far less loss. Freezing halted any organicgrowth on the photographic materials and at the sametime provided the needed suspension of time untilthese materials could be worked on and then driedby the vacuum drying techniques to be discussedelsewhere.

If the photographic materials had been wet byclean water and then frozen immediately, further re-storation might not have been necessary after vacuumdrying. When discoloration by dirt or other foreignmatter occurred, these additional restoration steps fol-lowed:

1) Thawing only the number that could be han-dled in one day

2) Washing in clean water from the emulsion sidewith a soft sponge

3) Placing prints in a print flattening solution fol-lowed by a Photo Flo® solution and drying on adrum drier (Fig. 50)

4) Photographs with flaking, loosening emulsionPlacement in Kodak Hardener F-5a for less thanfive minutes soak.

600 cc Water, about 50° C75.0 grams Sodium Sulfite, desiccated235 cc 23% Acetic acid37.5 grams Boric Acid crystals75.0 grams Potassium Alum

Cold water added to make 1.0 literKodak Special Hardener SH-1

500 cc Water10 cc 37% Formaldehyde by weight6 grams Sodium Carbonate, monohydrated

Water added to make 1 liter15 to 30 minutes wash in running waterPlacement in blotter roll or between blotters

If the damage to photographic materials is notextensive or if the damaged materials are few innumber, they can be treated in-house in the followingmanner:

Fig. 50. Restored photographic prints dried on a drum dryer.Fig. 51. Transparencies being immersed in Photo Flo® solution.Fig. 52. Transparency pulled gently through sponge rollers.

32

Preservation of Slides1) Remove slides from original containers, retain-

ing as much identification as possible. Placethem in solution "A"—10 grams magnesiumsulfate, 10 mis of 37% formalin per liter.

2) Following storage in solution "A," disassemblemounts. Gently clean slides with camel hairbrush using solution "B"—10 grams sodiumcarbonate, 10 mis of dequest 2006, and 10 misof 37% formalin per liter.

3) Soak Ektachrome films in a regular Ektachromestabilizer solution.

4) Rinse slides in solution "C"—standard finalrinse in Photo Flo® solution.

5) Air dry slide, emulsion up, on a paper towellined tray. Do not remove film from the mount.

Glass Mounted Slides1) Following storage in solution "A," separate film

from glass if emulsion does not strip. Clean withcamel hair brush in solution "B." Dip in solution"C" and air dry, emulsion up, on cloth linedtrays, keeping mount and film together. If pos-sible, save the glass in a separate container.Ektachrome films require a stabilizer soak.

2) If emulsion adheres to the glass, soak further insolution "B" before final cleaning and drying.

Color transparencies1) Remove transparencies from their sleeves and

place in stainless steel processing holders.Two 8 x 10 processing tanks are needed forwater baths,—one tank for Photo Flo® solution,one tank for hardener solution.

2) Immerse transparencies in holders in the firstwater bath for approximately five minutes toremove all surface dirt.

3) Immerse transparencies in 2% Photo Flo® solu-tion for five minutes (Fig. 51) with agitation toswell the gelatin and release residues. The gela-tin is extremely soft during this step. Careshould be exercised.*

4) Return transparencies to first water bath, thistime with running water and wash for about fiveminutes.

5) Place in hardener solution tank (Kodak SH-ISpecial Hardener for about five minutes. Re-move excess with soft sponge (Fig. 52). Linedry (Fig. 53). Fig. 54 shows two prints madefrom successfully restored transparencies.

"In extreme cases only, a 15% solution of sodium carbonate willswell the gelatin even more and remove all stubborn residues.

Fig. 53. Transparencies hung to dry.Fig. 54. Results of successful restoration—color prints made fromrestored transparencies.

33

Movie Films1) Remove film on reels from cans and store in

Solution "A." Retain identification if possible.2) Reprocess in white light.

• Feed on rollers without emulsion contact.• Start the film in a prehardener solution fol-

lowed by solution "B."• Run through a spray rack if possible.• Depending on trial experience, invert film

and buff emulsion side.• Complete processing; take up dry film on

cores.• Reassemble on reels.

Audio tapes which were damaged were more of aproblem than photographic materials. Mud could beeliminated from the tape with ordinary tape cleaning,but the differential stretching of the tape's acetate basemade playback extremely difficult. Most of the record-ings were interviews of historical value to the Museumlibrary. The material was transferred from reels to cas-settes using a three head, four track reel-to-reel re-corder.

By monitoring, it was found that applying slightmanual pressure to the pads holding the tape againstthe heads would eliminate the periodic dropoutscaused by tape warpage. Although this increased thebackground noise level, it did result in a very usableaudible recording. The use of filters and dolby circuitrycould reduce this problem.

34

Chapter IVFumigation andSterilization of Flood-Contaminated Library,Office, Photographic,and Archival Materials

Chapter IV

As the flood waters receded on June 24, 1972, a layerof mud was deposited on the Museum's library collec-tion, archives, slide and film collections, and office files.Because the mud contained microbial contaminants, aswell as eggs from various insects, all of the flood-damaged materials except the glass objects were frozento preserve them from attack by these biologicalspecies. This freezing action also permitted the Mu-seum's staff to consider restoration plans specificallydesigned for each of the damaged items.

Even though some of the Museum holdings werenot directly in contact with the flood waters, all becamecontaminated during subsequent removal from theMuseum and transfer to temporary storage areas. Sincethe humidity and temperature were not controlled inthese storage areas, there was serious concern aboutthe possibility of biological growth in the materials. Arandom check of the storage area, books in the collec-tion, and correspondence in the office files substan-tiated that this concern was real. Insect, fungal, andbacterial growth were found. Starch, casein, and cel-lulosic materials, constituents of paper formulation, areutilized by some of these species.Approach to Fumigation and Sterilizationof Contaminated Books

As restoration planning evolved, concern for thestored, non-frozen contaminated materials increased.The use of insecticides was considered initially; thesuggestion was eliminated after discussions were heldwith several groups of paper conservators since they feltthat such compounds might leave residual halogens inthe paper. Such halogens ultimately contribute to anacidic environment which in turn accelerates aging ofthe paper.

The need to treat biological contamination necessi-tated the search for and purchase of appropriateequipment. To minimize capital expenditures, the needfor a quick delivery, and the fact that only a smallamount of space was available, a modified VacudyneDocument Fumigator (VDF) was selected. It was man-ufactured with an extra large vacuum pump and hadheating elements installed in its side walls. The unit waspurchased (Vacudyne, Altair, Chicago Heights, Illinois)to establish conditions for the kill of insects and mold,not necessarily as a sterilizer for the kill of bacteria.Since ethylene oxide could be used in its chamber, itwas hoped to modify its normal operating sequence todevelop a sterilizing treatment for the books. Ethyleneoxide, when used under the proper conditions, is notonly a fumigant but also an effective sterilizing agent. Itis used extensively by the pharmaceutical industry tosterilize various plastics and bandages.

Fumigating for destruction of book worms, silver-fish, moth, larvae, mold, and fungal contaminantsproceeds under milder conditions than those requiredfor killing the bacteriological contaminants. It wasnecessary to sterilize all of the Museum's books toeliminate subsequent damage due to biologicalgrowth or transfer of contaminants by users to otherbooks as well as to the users themselves. The followingexperimental approach was developed: In order toobtain the maximum amount of information on thequality of sterilization with the least efforts, attentionwas focused primarily on identifying conditions for killof what was believed to be one of the most resistantbacterial species. After being procured, this specieswas to be placed in the center of closed books, thelocation most difficult for gas penetration. If processconditions could be identified for kill at this location, itcould be assumed that all other, less resistant forms ofbiological life would also be killed. Time and moneycould be saved by not conducting the latter experi-ments. A survey of the Museum's problems confirmedthe fact that mold and bacterial growth were evenmore of a threat than insect growth.Experimental Considerations

FumigationThe initial effectiveness of the VDF was deter-

mined by conducting a series of experiments whichvaried the amount of air being evacuated before theethylene oxide mixture was allowed to enter, the timeof exposure, and the temperature of application to thebooks. The dimensions of the chamber in the fumi-gator were 1.2 by 0.92 by 0.46 meters. A normallibrary cart with three shelves could be moved in andout of the chamber. When these shelves were stackedwith books, approximately two thirds of the volumeinside the chamber was filled by the books and cart.After loading the chamber and closing the door, apre-selected program-evacuation of air, injection ofgas for treatment, evacuation of gas, and back-fillingwith air was begun.

Ethylene oxide at a composition of 12% by weightwith UCON Refrigerant 12 (Oxyfume-12, Union Car-bide Corporation) is non-flammable and can be usedsafely in a closed system. Effectiveness of this com-mercial mixture depends on the amount of ethyleneoxide present, temperature of its application, time ofexposure, and the relative humidity in the chamber.

36

Microbiological Tests and ControlsBecause Bacillus subtilis is a rugged, resistant mi-

crobiological species, spore strips (Castle Company,Rochester, New York) of this bacteria were chosen forstudy. The investigation was supplemented by use ofadditional biological samples involving actual pagesfrom badly damaged books; colonies of mold andbacteria had been allowed to mature intentionally foruse in subsequent experimentation. Later in this re-port, these samples which were placed inside booksare described as muddy-moldy pages. Cultures ofEscheria coli and Aspergillus sp. were also included inthe experiments.Positioning of Biological Test Samples

Initially spore strips and sections of muddy-moldypages were placed inside closed books. All sampleswere held at the center of the page with pressuresensitive tape. When kill was not observed, the sam-ples were attached to a page in the book which was seton its bottom edge and partially opened during treat-ment.

During final experimentation, when better condi-tions for kill were identified, special attention wasgiven to the positioning of biological samples. Theywere attached to pages in the center of books whichwere then placed between two 1.27 cm Plexiglasplates. Four C-clamps, one placed at each corner ofthe plates, were tightened with equivalent torque, 4.0Kg cm, to insure the uniform application of pressure toeach book. This amount of torque was chosen to du-plicate the pressure exerted on books when they aretightly packed on a shelf. Five different locations onthe cart (Fig. 1) were used for the books containingbiological samples.

In petri dishes placed on top of books packed onthe cart, Trypticase soy agar and Sabouraud's agarwere streaked with E. Coli and Aspergillus sp. respec-tively and then exposed to ethylene oxide for fourhours. These samples were used for a special part ofthe second series of experiments.Microbiological Media

Trypticase soy broth (liquid), and trypticase soyagar for bacteria, and Sabouraud's agar for fungi wereused as media. Although the agar media are indenticaland nutritionally comparable to the liquid media, theagar was added as a solidifying agent for cultivation ofbacteria or fungi on a semi-solid surface.

Culturing of Test and Control SamplesBoth control and treated spore strip samples were

cultured at 37° C in trypticase soy broth for forty-eighthours to determine kill. Swab samples, taken fromsections of the surface of the muddy-moldy pagesbefore and after exposure to ethylene oxide, wereplaced in one ml of 0.001 M phosphate buffered salinesolution at pH 7.0, for plating on appropriate media.These samples were taken from a 6.45 square cen-timeter section.

To determine the effect of adding moisture todried contaminated pages, one ml of phosphate buf-fered saline solution at pH 7.0 was added to a sectionof each of several muddy-moldy pages before treat-ment with ethylene oxide. Swabs were taken fromthese selected areas before and after the gaseoustreatment. Duplicate plates were made of all samples.Process Conditions for Ethylene Oxide

Exposure to ethylene oxide was 4, 6.5, and 15hours duration. To keep the relative humidity at thenecessary percent, water vapor was added to thechamber of the VDF.

Back of _^Chamber

nside ofChamber Door

Fig. 1. Location of books containing biological samples on cart.Scale: 1 foot = % inch (9.5 mm).

37

Chapter IV

Exposure of the books to ethylene oxide occurredat either room temperature or at an elevated tempera-ture. The maximum temperature in the chamber in thelatter case was approximately 40° C. Figure 2 docu-ments, during a 6.5 hour treatment, the temperaturechanges of the gas in the chamber after closing thechamber door. Thermocouples were placed at similarlocations both inside and outside of books on the cart.A maximum difference of 11°C was recorded uponentry of the gas into the chamber between a pair ofthese thermocouples positioned on the bottom shelf(see Fig. 1, position 4). The difference was reduced toone degree after eight minutes of exposure.

A comparable discrepancy was observed for theextreme in the chamber's gas temperature after entryof ethylene oxide. The temperature measurementswere made between the gas occupying the volumeadjacent to the gas inlet, located at the bottom and rearof the chamber, and the volume in the center of thechamber. The disparity in temperature was reduced toalmost four degrees after seven minutes of exposure.

The evacuation level and the volume of the booksand cart in the chamber controlled the amount ofethylene oxide present in the chamber. The first seriesof tests was made when the air pressure in the chamberattained a value of 15 inches of Hg. At this pressure aselenoid valve was opened to release the ethyleneoxide mixture. Gas continued to enter until a pressureof one inch of Hg was attained; then the valve wasshut. When the second and third series of experimentswere conducted, evacuation in excess of twenty-eight

Fig. 2. Sterilization gas temperature during exposure; ther-mocouple in center of chamber.

inches of Hg took place before the selenoid valve wasopened for entry of the ethylene oxide to the chamber.After the desired exposure time had elapsed, thechamber was evacuated to this pressure again andthen filled with air before it was opened.Experimental Results and Discussion

Series IThe initial series of experiments was conducted

with placement of the spore strips inside closed booksat five different locations on the cart with exposure forfour and fifteen hours at room temperature. The limitof the evacuation pressure was set at fifteen inches ofHg. Approximately fifty books were positioned on thecart, occupying about two thirds of the volume insidethe chamber, None of the spores was killed by thesetreatments. Furthermore, when the chamber wasopened, Oxyfume-12 could be detected. These tworesults suggested that (1) an increase in the totalamount of ethylene oxide was required to facilitate killand (2) a modification of the VDF program should bemade so that gas left after the treatment of the bookscould be removed before the chamber door wasopened. One change in the VDF program satisfiedboth needs. The level of evacuation was adjusted to apressure of twenty-eight inches of Hg before ethyleneoxide entered the chamber and before the chamberwas back-filled with air.

Series I IIn this series of experiments both spore strips and

muddy-moldy pages were used, some placed in bookswhich were partially opened and set on their bottomedges while they were being treated. Moisture wasadded to some sections of the completely dried con-taminated pages.

The results listed in Table I show that spores ofB. subtilis were not killed in the fifteen hours whenlocated inside closed books. When a book was leftpartially open to permit ready access of ethyleneoxide, kill of all spores was achieved.

The results described in Table II suggest that anecessary minimum of moisture encouraged kill offungi and bacteria at short exposure times; kill wasobserved on the moistened muddy-moldy pages putin the petri dishes on top of the books within fourhours. If the pages were dry, a fifteen hour exposuretime was required for kill of these same species. Thelatter findings, involving exposure to muddy-moldypages taped to pages in partially opened books, sup-ported the data noted in Table I for similar exposureand kill of spores.

38

An additional experiment (see Table III) usedstreaks of moist vegetative cells of bacteria and moistfungal species placed on nutrients in petri dishes.Again only four hours of treatment at room tempera-ture were required to kill these moist species. Similardehydrated forms (see Table I I ) , present on drymuddy-moldy pages, required fifteen hours of expo-sure. The necessary moisture content as well as itslocation were critical. The presence of water at a siteother than biological could produce a reaction withethylene oxide which would eliminate its chance tosterilize.

Series IIIDuring these tests, conditions comparable to the

second series were mainta ined except that thechamber was heated. It is obvious from a review ofTable IV that 6.5 hours of exposure are sufficient tocause kill of spores in the most tightly packed books.There was a high percentage of kill of bacteria andfungi on the muddy-moldy pages (Table V) placed inthe same books and treated at the same time, but thekill was not total. These facts imply that longer expo-sure times are required for complete kill when con-tamination is excessive, mud is present, and the sur-face of the page is dry. Complete kill was obtainedwhen a fifteen hour (Table V) exposure was applied tomuddy-moldy pages with populations of contamina-tion comparable to those treated during the 6.5 hourexposure at the same elevated temperature.

During the Series II experiments, ATI SterilizationIndicator Tapes, #00180 (Aseptic-Thermo IndicatorCompany, North Hollywood, Cal i forn ia) , weremounted beside all spore strips and muddy-moldypages placed in the books used for the 6.5 and 15 hourtreatments. The color change from yellow to darkblue, indicating that sterilization conditions were met,was found in each case. These tapes provided a con-venient approach to determine the general quality ofthe sterilization environment. They could not guaran-tee micro-environmental sterilization when excessivegrowth had left biological debris or when mud hadbeen deposited. Such factors may control reactionswith ethylene oxide or limit its diffusion to the activebiological site.Summary

Sterilization of spore strips and muddy-moldypages in books treated at room temperature was ac-complished in fifteen hours when the books were seton their edges and left partially open.

The importance of sufficient moisture is obviousfrom the results of treating the muddy-moldy page forfour hours after being dampened, as well as the treat-ment of the active cultures of Escherica coli and As-pergillus sp. streaked on moist nutrient in petri dishes.A convenient practical way to insure the supply of thenecessary amount of moisture to each biological siteon each page still needs to be developed.

If library books which had not been exposed tomuddy water or had not experienced excessive bacte-rial and fungal growth are involved, a 6.5 hour treat-ment under the conditions listed in Series III is suffi-cient to sterilize even tightly closed books.

Treatment of the most seriously contaminatedbooks (muddy-moldy pages) required an excess of 6.5and less than 15 hours of processing when the condi-tions outlined in the Series III investigations were met.

The use of the higher vacuum, allowing a greateramount of ethylene oxide to enter the chamber, facili-tated kill of the biological species. After treatment andback-filling with air, traces of gas were not detectedupon opening the chamber door at the end of thesterilization sequence.

The results from the above investigations withbooks have led to the successful application of thisprocess for sterilizing the library holdings, archives,office files, and film and photographic prints of TheCorning Museum of Glass.

39

TABLE I—TREATMENT OF SPORE STRIPS, ROOM TEMPERATURE

TreatmentTime, Hours

4

15

15

Number ofIndividual Samples*

5

5

5

Location ofSpore StripsInside Closed

BooksInside Closed

BooksInserted Inside

Partially OpenedBooks,

Standing on Edge

Results after iunexposed

strip**+

+

+

ncubationexposed

strip+

+

* one at each position noted in Fig. 1** + = Growth

- = No Growth

T

TreatmentTime, Hours

4

4

4

15

'ABLE II-TREATME

Number ofIndividual Samples

5*

5*

2**

5*

N T O F M U D D Y - M

LocationOf Sample

Inside ClosedBook

Inside PartiallyOpened Book

Placed InOpen PetriDish****

Inside PartiallyOpened Book

OLDY PAGES,CountsBacte

BeforeTNTC***

TNTC

60

TNTC

ROOM TEM of Bacteria &ria

AfterTNTC

TNTC

0

0

PERATUREFungi per 6.45

FungiBeforeTNTC

TNTC

TNTC

TNTC

cm2

AfterTNTC

TNTC

0

0

* One at each position, noted in Figure 1** Located at position 1, 2 in Figure 1

*** TNTC = Too numerous to count**** Sample moistened with 1 ml. phosphate butter

TABLE III—TREATMENT OF CULTURES STREAKED ON NUTRIENT,FOUR HOUR EXPOSURE

Species NutrientNumber of Growth after Incubation

Samples Before AfterEscherichia coliAspergillus sp.* + Growth

— No Growth

Trypticase Soy AgarSabouraud's Agar

40

TABLE IV—TR

Type of Book Storage

Partially OpenedPartially OpenedPartially OpenedPartially OpenedPartially OpenedClamped ShutClamped Shut* + = Growth

- = No Growth

TABLE V—EVALUATIONELEVATED TEMPERATURES

Counts of Bacteria and Fungi per 6.45 cm2

BacteriaType of Book StoragePartially opened

Clamped Shut

Closed, Laying on Side

Partially Opened

Clamped Shut

Closed, Laying on Side

Clamped Shut

Closed, Laying on Side

E A T M E N T A N D

Position on Cart

1234514

Position on Cart1234514235123451423514235

EVALUA

+ +

Exposure Time (Hours)

TIO

Controls* Exposed

OF

N O F SPORE STEvaluation of Growth

4

++ -+ -+ -+ -+ -

TREATED MUDDY-MOLDY

44444444446.56.56.56.56.56.56.56.56.56.5

1515151515

RIPS, ELEV After Diff

6.5Controls

+++++++

BeforeTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTC

ATED TEMerent Expo

Exposed—-—-——-

PAGES,

After00000

TNTCTNTC

05000

700

105080

30050

0180

00000

PERATUREsure Times (H

15Controls

+++++++

FungiBeforeTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTCTNTC

rs.)

Exposed—-—----

After0000000000

201010301060

00

1010

00000

41

Chapter VVacuum Freeze Drying:A Report

Chapter V

Among the major restoration problems for the Mu-seum library was the variety of materials which hadbeen frozen. The Museum's library of record for glassincluded literature from countries throughout theworld; the composition of the paper as well as thetypes of printing and binding varied considerably.Both water soluble and insoluble inks were used in therecords and ledgers. There were archival materials: flatworks of art, photographs, newspaper clippings, pat-ents, correspondence, and photostats.

A search of relevant literature and conversationswith scientists and conservators revealed that freezedrying and vacuum drying techniques had great po-tential as initial procedures for salvage and restoration.Scientific laboratory data on drying frozen books bythese methods was too sparse to justify application tothe Museum's valuable collection. The conservatortherefore applied traditional air-drying techniques tothe frozen damaged materials during the early stagesof restoration. A scientific evaluation of freeze andother drying processes was begun with the hope thatthe remainder of the frozen flood-damaged materialscould be dried in a single operation.

Many decisions could not be made until theagglomerates of frozen books were separated, the in-dividual books thawed and opened, and the librarianshad identified the books and established their value.The following categories were established:• Discard—material no longer of value, or replace-

ment had been assured.• Replacement—severe damage, but duplicates

could be obtained immediately.• Recycle—item returned to freezer; replacement

not yet established; restoration postponed.• Microfilm—minimal cleaning with drying re-

quired; discard after filming.• Totally Clean—cleaned, de-acidified, sterilized,

and packaged for subsequent binding.• Completely Restore—same as Totally Clean, plus

salvage of binding for future decision as to itsuse.The conservator could then allocate necessary

money and time for restoration of each item. Ofteneach book presented its special set of problems. At thispoint the conservator and his assistants noted twoextremes during their daily operation. On those dayswhen the librarian identified thawed books to be dis-carded, many books were processed. However, whena thawed book required complete restoration, severaldays and in some cases weeks of effort were requiredbefore restoration was completed. When this hap-pened the number of books processed per day waslow. The periodicals which contained coated paperwere returned to the freezer immediately wheneverthey were discovered, since the conservator had no

practical options other than the time-consuming oper-ation of interleaving the coated paper pages. It wasanticipated that a new, more efficient technique wouldresult from research. The normal restoration operationfor the damaged books was sufficiently complex with-out adding the variety of problems caused by thisassortment of special coated papers.Approach to a Solution

The conservator specifically requested that prior-ity be placed on finding an acceptable mass dryingtechnique. All other conservation research projectswere set aside and full attention was focused on thisproblem. To minimize cost and reduce the time re-quired for obtaining results from a survey of possibledrying techniques, the scientist and conservatoragreed to buy books containing the two extremes inpaper for experimentation. After preliminary tests,separate uncoated and coated paper books wereselected. Experiments showed that normal drying ofthe uncoated paper books caused no problem; stick-ing was a major problem in drying the group of coatedpaper books.

Both sets of books underwent a standard simu-lated flood, drainage, and freezing sequence. Pairs ofthese standard flood-damaged samples were dried bydifferent laboratory processes and evaluated. It wasestablished that Freeze and Solvent Extraction Dryingtechniques could be applied with success to bookscontaining coated paper. Furthermore, the generalquality of the coated paper dried by these two pro-cesses was better than that resulting from the tradi-tional air dry and interleave technique. Since theseinitial laboratory experiments were successful, a pairof flooded, drained, and frozen books was taken to theGeneral Electric Company in Valley Forge, Pennsyl-vania, to be dried simultaneously in their large spacechamber with a mass of wetted, basically uncoatedpaper books by a freeze dry method. After the success-ful drying of this pair of coated and uncoated frozenbooks, the scientist suggested that the conservatorconsider applying this technique to the remainder ofthe museum's frozen flood-damaged collection.Freeze-Thaw Vacuum Drying Process

During initial restoration operations the tempera-ture inside the thawed books was monitored by a flathead thermocouple to help determine the tempera-ture above which sticking would occur unless thepages were interleaved. From measurements on arandom selection of some seventy-five thawed bookscontaining a variety of coated paper, it was observedthat in the majority of the books sticking took placewhen the temperature exceeded 10° C. Sticking wasnot observed, however, when the temperature of thewater in the book was below 4.4° C. This informationreinforced the conclusion of earlier reports. It also

44

indicated that in removing water from the book byevacuation, advantage could be taken of the greatervapor pressure of water in its liquid phase in contrast tothe lower vapor pressure of ice or water at its meltingpoint. Thus the concept of using a mass Freeze-ThawVacuum rather than a standard Freeze-Dry processdeveloped.

Water has a vapor pressure of 6.3 mm at 4.4° Cand 9.2 mm at 10° C. These values are greater than4.6 mm at 0° C for the equilibrium vapor pressure ofan ice-liquid water combination. The potential effi-ciency with which the water is removed at 4.4°Crather than 0° C is increased by at least thirty-sevenpercent. Specifications were developed so that thetemperature of the liquid phase in the books duringdrying would be kept close to 4.4°C and would notexceed 10°C. This was extremely important duringthe early stages of drying when a large volume of waterwas present in each book. It was less important whensmaller amounts of moisture were present.Final Recommendation for Mass Dryingthe Museum's Collection

After considerable discussion and several sup-plementary experiments, both the scientist and theconservator recommended the application of thismass drying technique to the remainder of the frozenitems. The recommendation was accepted by theMuseum's administrative and library staff. There wasstill serious concern about the effect of this process onthe adhesive quality of the mud since there had beenno mud on the research books being dried duringearlier experiments.

At this point, approximately one year of work onthe Museum's collection had taken place. The majorityof the uncoated paper books had been restored; booksand periodicals primarily containing coated paper re-mained in the freezer. No attempt had been made torestore the frozen, flood-damaged periodical collec-tion for several reasons: (1) the high content of coatedpaper made successful restoration improbable be-cause this paper sticks while drying; (2) the applicationof the interleave drying technique was too expensivebecause of the labor-time requirement; (3) it had beenassumed that periodicals would be replaced by acquis-ition. Replacement of both books and periodicals wasnot completely successful since many items were out-of-print and not available on the world market. Archi-val objects and photographs were still in the freezerand were to be included in the mass drying operation.THE PLAN AND ITS IMPLEMENTATIONPackaging and Transportationto the Drying Site

Arrangements were made with a commercial truckagency for the rental of a freezer truck and for a driver

to transport the frozen flood-damaged materials toGeneral Electric's Space Center at Valley Forge,Pennsylvania. The control of the Museum freezer'stemperature was lowered to -28.9° C one week be-fore this trip to insure that the books would be frozenwhen loaded. The General Electric Company wasasked to transfer these frozen materials to their spacechamber for Freeze-Thaw Vacuum Drying im-mediately upon arrival.

The number of books in the truck was limited by itssize and weight. Everything in the freezer which was tobe dried could be transported in the available vehicle,plus a selection of discarded items representing varia-tions in paper type and book bindings. The number ofbooks (3,500) in the truck, it was estimated, would filltwo-thirds of the General Electric Space Chamber.

Several unforseen problems had to be solved.When the freezer truck arrived at the Museum it wasfound that its refrigeration unit was defective. Thecompressor stopped when the truck reached a speedof sixty-four miles per hour; the compressor was foundto work well if the truck was stationary. When thetemperature of the refrigerated section of the sta-tionary truck reached -12.2° C it was decided to loadthe materials described in Table I. They included 240carry boxes (33 x 43 x 28 cm).

A pair of simulated flood-damaged, frozen,coated, and uncoated paper research books wasloaded with these flood-damaged books. They werepacked in dry ice for additional insurance against thaw-ing; their primary mission was to provide quality con-trol for the Freeze-Thaw Vacuum process itself, not tomonitor the transportation storage conditions.

Before leaving for Valley Forge, the owners of thefreezer truck gave assurances that the insulation in thetruck walls would keep the books frozen for ten hourseven if the compressor stopped. This was encourag-ing, but as a precaution it was requested that amechanic be at Binghamton, New York, to repair thecompressor before the longer part of the trip throughthe Pocono Mountains was begun.

The mechanic repaired the compressor, whichcontinued to function during travel on InterstateHighway #81, but as the truck encountered rougherhighway—the Northeast Extension of the Pennsyl-vania Turnpike—the compressor stopped. The truckwas halted and kept stationary until the compressorwas started and the temperature was lowered to— 6.7° C. Only then was travel resumed. Between suchstarts and stops, four times between Scranton andValley Forge, the temperature of the air above thebooks oscillated between -6.7 and +10°C. Thecompressor was started upon arrival at Valley Forgeand functioned well until the frozen books were un-loaded at General Electric's Space Chamber.

45

Chapter V

Unloading the Truck and Loadingthe Drying Racks

Upon arrival at General Electric's Space Centerthe temperature inside the freezer truck was -12.2° C.The truck's temperature measurement was checkedwith a separate thermocouple system. The two werefound to agree within 0.6° C. This confirmed the accu-racy of the temperatures measured while traveling toValley Forge.

The space chamber, a 7.32 meters long cylinderwith a 3.7 meter diameter, had room for three bookcarts. Each cart consisted of a base on which weremounted two rows of 91 x 30 cm shelves, 213cm highand 213 cm deep. This gave each cart fourteen shelvesmeasuring 213 x 91 with 30 cm between shelves, or atotal flat surface area of 27.3 square meters. Each shelfhad a strip heater attached to the underside. Eachheater was connected to a variable transformer so thatthe input power could be varied. Approximately 30thermocouples monitored the temperature. Half of thethermocouples were placed in books at random, andthe other half were to monitor the strip heaters and thechamber functions. The chamber was mounted onconcrete about 6.1 meters above ground level and wasloaded by rolling a cart along a track and onto anelevator which lifted the cart up to the chamber. Thecart was then rolled to the back of the chamber atwhich time the electrical and monitoring connectionswere made.

To diminish labor costs, the Museum's carry boxeswere placed directly on the cart shelves; the bookswere not taken from the boxes, but the lids of theboxes were removed. Two-thirds of the boxes hadbeen packed with the spines of the books on the bot-tom of the box. The books in the remainder of theboxes had been packed both horizontally and verti-cally. Several boxes had frozen agglomerates with thebooks in every position. The oversized volumes wereplaced horizontally on the shelves of the cart, primarilyon the highest shelves. All the frozen material wasloaded onto the racks of the carts and placed in thechamber one and one-half hours after arrival.

The location of eight boxes with solidly frozenbooks was noted especially. Each time the chamberwas opened a check of these boxes was planned todetermine the progress of the overall drying operation.The two frozen research books were also positioned sotheir progress in drying could be checked by quickremoval and weighing. The amount of water absorbedand frozen in each had been determined preciselybefore the trip to Valley Forge.Specifications for Treatment

Since the sticking of some types of coated paper isaccelerated by high temperatures, the first phase of thedrying process was to be run with the shelf tempera-

ture as low as possible. The temperature chosen was21°C. The initial chamber pressure established bymechanical pumps was to be 10 Torr before freoncooled with liquid nitrogen was circulated through thecondenser (13.94 square meters). This coil waspositioned inside, not outside, the space chamber.The condenser was to be kept below -9.4°C anddefrosted when necessary. Drying progress was to bechecked every twenty-four hours unless both partiesagreed to a modification of the schedule.Observations and Action during Drying

The following observations led to successful dry-ing of these materials.Time After Observation and ActionInitial Each Time the Chamber DoorEvacuation Date Was Opened

26 hrs. 4/23/74 The refrigeration coil was de-frosted. Forty gallons of waterwere removed. The uncoatedcontrol research book waschecked and found to have lost44.5% and the coated controlbook 24% of their original ab-sorbed water. Some of theMuseum's books were almostdry; others were still frozen. Thedoor was closed, evacuation in-itiated, and drying started.The condenser's coil held only asmall amount of ice; it was re-moved. The uncoated paper con-trol book had lost 77%; thecoated paper control book hadlost 43.8% of the original ab-sorbed water. More of the bookswere dry, although several re-mained frozen. Adjustments weremade so that the heaters wouldincrease the shelf temperature to32.3° C. The door was closed anddrying was begun.

51.5 hrs. 4/24/74

94.0 hrs. 4/26/74 The small amount of ice on thecondenser's coil was removed.The uncoated paper researchbook was completely dry; thecoated paper book had lost86.2% of its absorbed water. Acomplete evaluation of the booksshowed that 50% were dry. Thesewere removed from the chamber.The remaining frozen items wereremoved from the carry boxesand individual ly placed spinedown on the heated shelves. Itwas noted that the overall re-moval of water was slower thananticipated. Complete drying ofthe coated paper research book

46

126.5 hrs. 4/27/74

156.5 hrs. 4/29/74

180.5 hrs. 4/30/74

could be expected in another 24hours of treatment, but GeneralElectric's commitment to dry atleast 95% of the books would notbe satisfied at that time. Thechamber door was closed and thedrying started again.After the coil was defrosted, thecoated paper research book wasfound to be dry. Many bookswere still frozen. The chamberdoor was closed and the pumpswere set to pull a vacuum of 35mm for the remainder of theweekend. The heaters wereturned off during this period.Many books were still frozen. Theposition of these books on theshelf was changed to facilitateheat transfer for subsequent dry-ing. The power for the heaters inthe shelves was increased to givea maximum temperature of48.9° C. The books dried and re-moved earlier (approximately100 boxes) were placed in a truck(non-refrigerated) for storageuntil their return trip to Corning.Since their environmental temp-erature was about 32.2° C, therewas concern microbiologicalgrowth might take place. A tank ofOxyfume-12 was released insidethe closed truck to decontaminatethe books after a controlled leakto the atmosphere was provided.Several blocks of dry ice wereplaced between the boxes ofdried books as a precaution tomaintain a low environmentaltemperature until a return toCorning was possible.After the chamber door-wasopened, the carts were removed.An individual check of each itemrevealed that about 100 still heldnoticeable amounts of moisture;the remainder were dry. All bookswere packed. The boxes contain-ing the moist books were markedso that they could be given im-mediate attention upon arrival inCorning. The boxes were loadedonto the truck; the trip to Corningbegan.

EVALUATION OF RESULTS AND DISCUSSIONInspection of Each Item

The truck was unloaded in Corning, New York,approximately eighteen hours after it left Valley Forge.Three inspection stations, one each for books, period-icals, and archival items, were established. As eachbox was unloaded from the truck it was opened toidentify its general content and then given to one of thegroups at these stations. Each member of these groupshad been previously trained to categorize each item by(1) the percentage of coated paper pages that werestuck, (2) the amount of wetness present, and (3) thedegree of cockling that was apparent. The percentageof stuck pages was determined by making a minimumof ten specific observations at different locations in thesections of the book containing only coated paperpages. If sticking was observed, an estimate of thepercentage of the total number of stuck coated pageswas made. The inspectors' training also included studyof a standard set of damaged samples chosen to repre-sent three categories of water content (dry, moist, wet)and cockling (zero, some, very).

If a book was identified as being moist or wet, itwas taken to the drying area. It was set on its bottomedge and partially opened to let dry air circulatearound it to remove the remaining traces of water.Dampness was determined by touch. Each inspectorsampled at least ten different sections of each book byrunning his or her fingers over the entire page. If anydampness was detected, the book was rated as moist.If the degree of dampness caused the uncoated paperpages to adhere slightly so that separation was notrapid, the book was rated as wet.

Cockling is defined as irregularity in the flatness ofa paper's surface. In its worst form, a large number oflocalized irregularities are found per unit area. Thedrying of newly-made paper on high speed rolls leavesresidual stresses. These localized stresses are releasedwhen the paper becomes wetted and dried withoutstress. Both the previous history of the paper as well asthe application of the second drying operation play avital role in the amount of cockling which results.

The evaluation of each individual item by thesecrews—six inspectors—took about sixteen hours. Theresults at each inspection station were marked on theoriginal sheet of paper identifying the category of thebook; if these sheets were misplaced, a new sheet wasinserted in the book. These sheets were used by (1) thelibrarians to record the value of the book (2) the con-servator to note the degree of restoration that his assis-tants should apply, and (3) by the scientist to tabulateresults. A total of 1504 sheets (Table II) was collected,covering about 50% of the actual number of drieditems. None of the sheets attached to stuck coatedpaper books was lost.

47

Chapter V

Results (% Wet, % Stuck, % Cockled)Before reviewing the tabulation results, it is impor-

tant to remember the biases which were inherent in thefrozen materials sent to Valley Forge to be dried. Theseare summarized below to aid interpretation of the find-ings:1) Of the frozen books more than 85% contained

coated paper pages.2) Essentially all of the frozen flood-damaged period-

icals contained coated paper.3) A group of oversized books, the largest being 80

cm x 65 cm x 5 cm, had been set aside for dryingby this method.

4) A series of damaged frozen books, ledgers, andalbums containing water soluble ink and dyes wassent for application of this drying technique.

5) All books designated "difficult to replace" duringthe preceding year were included to determinewhether they could be salvaged at a cost lowerthan their most recent replacement price.

6) From the group of books marked "discard," aswide a variety of materials as possible, includingpre-wetted coated paper books, was selected todetermine how versatile this treatment could be.The distribution of percentages identified with the

stuck coated paper pages are listed in Table III. Only7% of this sample of 1504 books had 91% to 100% ofthe coated paper pages stuck. This category of stickingdoes not mean that subsequent restoration was im-possible; it does indicate that some segment of eachpage was stuck sufficiently so that force had to beapplied to bring about possible separation. Fifty-sixpercent of the items did not give any evidence of thecoated paper pages sticking. Ninety-one percent(Table IV) of the books which received the Freeze-Thaw Vacuum treatment were dry. The sheets fromthe books revealed that 119 were identified as moistand only ten as wet. The majority, 56% (Table V), hadpages with obvious cockling. An appreciable fraction,35%, was cockled badly. Only 9% of the books werecompletely free of cockling.

SPECIAL EXPERIMENTSOriginal Drawings

It was evident that some water colors in the freezerhad run before they were frozen. More than 1000drawings were frozen in two large blocks of agglomer-ates in a print drawer; the Museum staff members wereconcerned that by the time the top prints thawedenough to be removed by traditional drying tech-niques the sides of the agglomerates would be de-frosted, increasing the probability of the color washingoff or running. In several books with water coloredillustrations, the binding medium became humidifiedenough to adhere to the facing page as it dried. There

was no additional running of the water colors due tothe Freeze-Thaw Vacuum Drying.Oversized Volumes

Because of their size, large frozen volumes had tobe placed individually on the drying shelf in a horizon-tal position. When dried by this Freeze-Thaw Vacuumdrying technique, they were distorted very little. Thevolumes had returned to almost their original size. Thiswas an unexpected but positive result of applying thisdrying technique.Damaged Coated Paper BooksRewetted and Frozen

Several volumes containing coated paper werefound to be stuck as restoration began. Efforts to sepa-rate the pages resulted in removal of the coating fromthe substrate. These volumes were rewetted and fro-zen until a new drying technique could be applied. Itwas hoped that the slight layer of ice from fresh waterbetween the pages of a frozen book would provideenough pressure to break the bond between the coat-ings. If this ice could then be thawed and the liquidphase evaporated at a low temperature, the pagesmight be separated in a dried state. Similar frozenbooks were to be paired with these books for the massdrying experiment. They, however, were not rewet-ted. Both sets of books were sent to Valley Forge.There was almost no adhesion in the books that hadbeen rewetted; similar books that had not been rewet-ted stuck excessively.Photographs

Two file drawers of photographs, with some cor-respondence interspersed, were sent to Valley Forge.Most of the material in the file drawers was relativelydry because it was the last material to be frozen duringsalvage after the flood. The majority of the photo-graphs, unfortunately, adhered tenaciously to eachother, either emulsion-to-emulsion or emulsion to theback of the adjacent print. These photographs couldbe separated by rewetting, swelling the emulsions, andapplying gentle force. When emulsions had alreadybeen damaged by mold, salvage was laborious, ex-pensive, and at times impossible.

After mass drying, those photographic printswhich were known to have been frozen promptly afterthe flood, i.e., containing much water and no mold,showed no adhesion and only slight cockling. An in-spection of the emulsion showed no sign of damage.No reduction of gloss could be identified.

48

Discussions and ConclusionsMass drying the frozen flood-damaged books

permitted an efficient review of their value and anassessment of the degree of restoration they shouldreceive. The librarians found that this drying methodenabled them to re-evaluate their original decisionsand better estimate replacement versus restorationcosts. Most books were to be restored. They alsofound that books previously designated as discardswere now usable. After fumigation and sterilization,some books were returned to use. Using the librarians'value notation, the conservator could organize allthose books requiring equivalent degrees of restora-tion, instruct his assistants on procedure, and set up amodified production line for their efficient treatment.Estimates of the time necessary for restoration couldbe predicted more accurately.

The period of time for the mass drying was greaterthan the four days anticipated after the previousstudies on drying the pair of wetted and frozen re-search books. Since different types of coated paperwere involved, the temperature control for the cham-ber's shelf heaters was set lower initially. Most of thesecoated paper books were much larger than the re-search books; water diffusion to the edges of the pagestook longer. Many archival items were wrapped inseveral layers of paper, which made water removaldifficult. Another factor which could not be predictedfrom the earlier experiments was the uneven distribu-tion of heat transferred from the shelves to the boxescontaining the books and then later to the individualbooks themselves. The cardboard carry boxes werebelieved to contribute a minimum barrier to heat trans-fer; it is now known that if the boxes are to be used, allbooks should be packed in a vertical position with theback edge of the books on the bottom of the box.Horizontal packing of the books does not permit effec-tive heat transfer to the books in the top of the box.

The surface quality of the dried coated pages wasexcellent. This does not imply that further restorationcould be eliminated, but that restoration could be pur-sued with a high probability of success if justified by thevalue of the book.

The quality of the dried, bound newspapers andjournals was very satisfactory. There were, however,cockled pages and damage to the spines, but theseproblems were independent of the Freeze-Thaw Vac-uum Drying Process.

There was some evidence that the soluble mate-rials in the coating had migrated toward the edges ofthe pages before freezing. This migration did not inter-fere with the drying and separation of the coated paperpages by mass drying technique if the pages had beensufficiently wet before freezing. In several volumes,portions of the coating were completely separatedfrom the paper substrate. This was attributed to the

history of the book before freezing and not to thedrying process.

In general, cockling of the coated paper wasgreater than that of the uncoated paper. The amountof cockling depended upon the extent of wetting bothbook and individual pages.

The inspection of these books was not designed todetermine whether the mud adhered more or lesstenaciously to the paper than after application of otherdrying techniques. However, a large sample of driedbooks containing both uncoated and coated paperpages was checked after the initial inspection wascompleted. Surprisingly, the dried mud could be dis-placed with ease. Instead of this drying process furtherimbedding the mud in the paper, the mud appeared tobe lifted to the surface. Mud removal from the surfaceof the coated paper was less difficult than that from theuncoated paper. Although a systematic controlled ap-proach was not established to measure ease of driedmud removal, both the scientist and the conservator,as well as their assistants, agreed that the massFreeze-Thaw Vacuum process aids rather than hinderssubsequent elimination of dried mud.

It is not known why some of the coated paperswere stuck. It is highly probable that the blocking, i.e.,sticking, occurred before the drying process wasapplied. After the flood the drainage time of the mate-rials was sufficiently long that some of the books hadcompletely or partially dried before freezing; in somecases the degree of drying was sufficient to causeblocking.

Another explanation for sticking of some pagesmay be that the uneven excessive heating of the dryingrack shelves could cause formation of an appreciablevolume of liquid phase. If this happens, significantamounts of starch and casein (adhesives for coatedpaper) are dissolved. The final evaporation of thiswater from the pages results in the previously dis-solved constitutents solidifying between the pages andthereby causing sticking. A daily check during dryingdid not identify this as a primary cause for the sticking.

Even when water soluble inks were used, the de-signer's sketches and the watercolored illustrationsdried with little or no evidence of feathering. Somecolor did transfer to the back of the adjacent sketch,but the quality of detail in the illustration was littleaffected. Several of the archival items were still moistwhen the truck returned to Corning because they hadbeen packed in multi-layers of paper; water removalfrom the center of the package was difficult. After thewrappings were removed, these partially wet itemsdried with minimal evidence of damage.

The results of drying the frozen photographs weremost acceptable if the photographs or prints had beenfrozen quickly after being flooded and if there was noevidence of biological growth.

49

Chapter V

In general, the Freeze-Thaw Vacuum Drying Pro-cess, with the specifications set by the scientist, cansuccessfully mass dry frozen and wetted items. Thisprocess can be applied to uncoated paper materialswith a high probability of success. The quality of thedried product, except for cockling, will be excellent.Uncoated paper books dried by this process are lessaffected by their previous history than coated paperbooks.

Knowledge of the previous history of coated paperbooks is necessary to predict the quality of the driedproducts resulting from this process. If coated paperbooks are wetted and allowed to dry in part, or remainin a drainage position under pressure before they arefrozen, they will probably stick after this mass dryingprocess. When the history of coated paper books isunknown, the books should be dipped or soaked infresh water for a short period of time before beingfrozen if non-stuck pages are expected from this dryingprocess.

Although this mass drying operation primarily in-volved coated paper, two previous Freeze-Thaw Vac-uum Dry treatments on the uncoated paper researchbooks produced pages similar to their original state.The few uncoated paper books that were sent with thecoated paper books for the final mass drying in theGeneral Electric chamber also dried successfully.

Further experimentation is desirable to measurethe true value of this mass drying operation for a widevariety of materials. The results of drying both coatedand uncoated paper materials from The CorningMuseum of Glass Library suggest that those with simi-lar materials and the misfortune of water damageshould consider this mass drying method.

50

TABLE I—FROZEN FLOOD-DAMAGED ITEMS TAKEN TO THEGENERAL ELECTRIC COMPANY VALLEY FORGE SPACE CHAMBER

Approximate Number ofIndividual Items or Boxes Item Type

13001000-1500

10003030

157*

2**Box size, 26 x 30 x 40 cm.

Recycle books stored in boxes*Bound periodicals: Several sets of the complete series of published volumesOriginal drawings of various designers (water colors, pastels, pen and ink, etc.)Extra large volumes containing colored and black-white platesDiscard books stored in boxes: * Damage was severe and all contained coatedpaper; retention was for experimentation.Special damaged coated paper books, rewetted and frozenSeven boxes* of archival items: photographs (1000), patents, ledgers, books,correspondence, notebooksTwo boxes* of approximately 1000 photographic prints

TABLE II—DRIED ITEMS CHECKED FOR STATISTICAL TABULATIONSNumber of Individual Items General Description

554518389996

Periodicals, identification sheets markedPeriodicals, sheets not markedPeriodicals without sheetsBooks, identification sheets markedBooks without sheets

TABL

Type of ItemNon-Periodicals

Actual #Percent

PeriodicalsActual #Percent

TotalsActual #Percent

E I I I—INSPE

Numberof

Items899

554

1453

CTION

0%

43949

37868

81756

F O R

1to

10%

718

5310

1249

STICKING OFDistribution for Various %'s of Coated Pages Stuck11to

20%

506

265

765

21to

30%

425

285

705

T H E

31to

40%

324

173

493

COATED

41to

50%

10011

163

1168

PAPER PAGES

51to

60%

212

41

282

61to

70%

101

61

171

71to

80%

334

41

373

81to

90%

192

20

211

91to

100%

829

204

1027

51

TABLEDistribution f o r Degree of Dampness in Books

Type of ItemNon-Periodicals

Actual #Percent

PeriodicalsActual #Percent

TotalsActual #Percent

IV-

Number of Items

-INSPECTION

863

553

1416

FOR DAMPNESS

Dry

80293

48488

128691

OF DRIED BOOKS

Moist Wet

58 37 0

62 711 1

120 108 1

TABLE V—INSPECTION FOR COCKLING OF PAPER IN DRIED BOOKSDistribution f o r Degree of Cockling

Type of Item Number of Items Zero___ Some VeryNon-Periodicals 904

Actual # 80 479 345Percent 9 53 38

Periodicals 553Actual # 48 341 164Percent 9 62 30

Totals 1457Actual # 128 820 509Percent 9 56 35

Chapter VIPlanning toProtect anInstitution andIts Collections

Chapter VI

Few libraries, museums, or historical societies have theprescience to develop a plan for the handling of adisaster—a disaster which may never come and whichthe odds indicate will happen to someone else. Unfor-tunately, institutions do suffer damage from fire,water, or extremes of nature, and when no plan forhandling such emergencies is available, the loss willbe greater than necessary. The following considera-tions may help in drafting a plan for specific situations:A. Insure

Be certain that insurance coverage is completeand covers the collection, equipment, and building—as well as office furniture, restoration of files, costsof temporary relocation, and the myriad expensesinvolved in returning to the status quo before thedisaster.

The first step in planning is to locate an insuranceagent who can help to provide the proper coverage forall aspects of the institution's activities and holdings.Advice can be sought from local insurance representa-tives or from the appropriate professional organiza-tions such as The American Library Association, TheAmerican Association of Museums, The American As-sociation of State and Local History, or other pertinentorganizations.

Working with an insurance expert is essential, forfew professionals in museum or library work know theintricacies of proper coverage. The kind of recordsneeded to estimate damage or loss to prove a claim aremost important if an adequate settlement is to bereached. The details on which claims might be basedmust be well documented: records of the date of ac-quisition, source, original cost, current replacementvalues, and so forth must be readily available, for therewill be no time to develop such details after a disasterwhen the struggle to maintain services and to restorethe collection must go on concurrently.

Insurance coverage should be considered not onlyin terms of replacement costs but in terms of restora-tion costs as well. A price can be set for replacing astandard dictionary, but what is the restoration cost orloss-of-value factor if a nineteenth-century pamphletor unique Venetian goblet is damaged or destroyed?What is the value of staff time in replacing a damagedor destroyed catalogue of holdings or in reconstructingfiles and records?B. Keep and Duplicate Appropriate Records

The accession records, shelf list, or catalogue ofcollections are among the most important holdings ofan institution, sometimes more important after a disas-ter than the items they represent. No proper inventory,no adequate claim of loss can be carried out withoutadequate and available data when catastrophes occur.

Such data should include a complete descriptionof the object, with size, condition, etc., date of acquisi-tion, source, provenance (where applicable), or origi-

nal cost, current replacement value (this can be ofassistance as well in adjusting insurance coverage asvalues increase or if de-accessioning of duplicates isunder consideration), number of pages (books),number of plates or illustrations (color, black andwhite), and other pertinent information.

Not only must documentation be complete andup-to-date, but it must be available—which impliesduplicate copies. A complete catalogue is of little helpif the only copy is destroyed in the disaster. All keyrecords should be available on microfilm and copiesshould be stored far away from the institution's head-quarters so that they will not be lost also. Storage in abank vault is not good enough—if the bank is in thesame disaster-prone area as the institution. Duplicaterecords may be kept in the general area, but a mastermicrofilm should be stored, preferably in a commercialarchival storage vault where proper humidity control isavailable. Such storage firms can usually produce dup-licate copies of masters if such are needed. Informa-tion on such facilities can be obtained from theNational Microfilm Association.

Once the records of all holdings have been mi-crofilmed, an annual filming of new records should beundertaken to keep the records complete. Often thismeans keeping an additional accession card to be usedfor the filming at the end of the year. Periodically(every five or ten years) it would be well to re-microfilm the master records so as to provide onemaster duplicate file which can replace the originalmaster microfilm and its five or ten supplements.C. Formulate a Plan

A plan for handling a disaster should be developedin consultation with or review by the entire staff of theinstitution so that each person knows his area of re-sponsibility and to whom he reports during theemergency (the emergency chain of command maydiffer from the normal system of reporting). It shouldbe in writing, available to all staff, and should be re-viewed annually for up-dating or change.

The plan's priorities should outline action to betaken during specific disasters. If flood, hurricane, ortornadoes threaten, there may be time to take preven-tive action. Fire or burst pipes necessitate differentresponses, naturally. Thus it is well to divide the planinto three parts:

Action to be taken prior to a disasterActions to be taken during a disasterActions to be taken after the disaster

These three aspects are covered in the followingpages. Some of the types of disaster which should beanticipated include:Bomb Threat Gas leak RiotEarthquake Flood PanicExplosion Water damage Power failurePlane crash High wind Accidents or illnessFire Hurricane on premises

54

Part 1Actions to Be TakenPrior to a Disaster

A plan of action for safeguarding a collection must beorganized well in advance of any threatened disaster.It should provide for:

1) An allocation of manpower and chain of com-mand

2) A listing of individuals and telephone numbers3) An identification of items to be saved by their

importance or value to the institution, and achart of their current location

4) A listing of needed supplies, materials, tools,and their location; information on procurementof vehicles

5) A listing of electrical switches6) A place for re-location if needed7) A set of procedures to be followed in securing

the collections or in moving them to temporarysafety

8) Cameras and film to record damage1. Allocation of Manpower and Chain of Command

As indicated above, an emergency chain of com-mand may differ from the normal roster of responsibil-ity as task priorities shift and new jobs are assignedduring the period of an emergency.

One person should be designated as the coor-dinator for all actions and should be free of all respon-sibility other than the supervision of emergency prob-lems. If ongoing business is to be conducted, it shouldnot involve the emergency coordinator. Specific indi-viduals should be assigned responsibility for movingareas of the collection, maintaining contact with secu-rity, radio and newspaper personnel, supervision ortransportation (if needed), and control of other spaceto which the institution may have to move. Adequate"work gangs" should be assigned to each of thesupervisors according to the needs of each area. Un-derstudies for key personnel should be named. Theirnames should be secured in advance, and they shouldunderstand the kind of work they may be asked to do.Where necessary and possible, additional help may beobtained from friends of the institution or from thecommunity.

A secondary command post should be designatedin advance, with a telephone number, should it benecessary to evacuate the institution. Instructions forthe staff as to when to report to such a post should bespecific.2. Listing of Individuals and Telephone Numbers

An important informational document is an up-to-date list of the names, home addresses, and tele-phone numbers of staff so they can be reached in an

55

Part 1

emergency outside of working hours. The list shouldobviously include police, fire, ambulance, security, orother such agencies. In additon, the names, addresses,and telephone numbers for outside areas should beavailable: transportation (truck to be used in evacua-tion, etc.), plumbers, gas and electric company per-sonnel or technicians, volunteers who will help in anemergency.3. Identification of Items to Be Saved

Each institution must develop its own list of itemsto be saved and the priority of such protection. Thefollowing listing cannot, therefore, be considered to fitall situations.

a. Listing of PrioritiesAfter the priorities have been determined, a listof the order in which items are to be movedshould be compiled and distributed. Wherepossible, names of those responsible for theitems should be appended.

b. Location Chart (Fig. 55)A chart which locates each of the items aboveshould be posted. If certain responsible staffmembers are unable to be on hand during anemergency, it should be possible for other staffto locate holdings in order to carry out theemergency plans.

c. HoldingsThe following list of categories of items to besaved is organized according to the priorities ofThe Corning Museum of Glass.(1) Objects in the collection, listed by their de-

gree of importance and priority of evacua-tion

(2) The catalog, shelf list, and/or accessions list.(3) Library

(a) Rare books, special collections itemswhich are important because of theirage, association value, one-of-a-kindfactor, or limited editions.

(b) Manuscripts, documents, archival ma-terials—which may not fall into cate-gory (a)above.

(c) General book collection. This may besubdivided into the more and the lessvaluable areas.

(d) Periodicals. Same comment as (c)above.

(e) Microfilm rolls and microfilm masters offilmed collection.

( f ) Circulation records.(4) Photographic and audio visual

(a) Films, slides, negatives, prints, trans-parencies

(b) Video tapes, audio tapes, recordings

(5) Art works: Paintings, prints, posters andother two-dimensional items not consideredunder (2) above as objects in the collection.

(6) Office Records(a) Files and general records(b) Correspondence(c) Financial records and tax records

(7) Equipment and furnitureGenerally speaking this is one of the leastimportant since desks, files, etc. can be re-purchased under an insurance claim. How-ever, certain pieces of equipment may begiven priority if time permits because of theirexpense, difficulty of replacement, or im-mediate need after the emergency.This would include(1) Laboratory equipment(2) Photographic and projection equipment(3) Office equipment, i.e., electric typewrit-

ers(8) Supplies—In general, office and library

supplies are the most expendable since theyare easy to replace. Consideration of savingthese items should come only if a maximumof time and effort is available to save them.

4. Emergency Supplies on HandPower failure is one of the greatest handicaps in an

emergency. Buildings should be equipped withenough permanently mounted auxiliary lights to pro-vide illumination in an emergency. An adequate sup-ply of fresh batteries for these lights should be main-tained.

Fig. 55. (a) Location chart prominantly posted in theLibrary, (b) Card catalog with first-priority dots attached.(c) Carry boxes with appropriate category noted.

56

In addition, the following equipment or materialsshould be on hand:

(a) Electrical fuses(b) Flashlights, additional batteries and bulbs(c) Fire extinguishers. Periodic drills should be held

to acquaint all staff with their location andproper use. Staff should have the opportunityto use an extinguisher out-of-doors to put outtest fires.

(d) Crowbars and axes. These may be needed todisengage fallen shelving, panels, etc.

(e) Hammers, wrenches, pliers, screwdrivers,wood and metal hand saws, utility knife andblade, wire cutters.

( f ) Brooms, buckets, mops, wringers, rags. Buck-ets are obviously essential for leaks. All thesetools will be needed to handle water damageassociated with fires.

(g) Plastic sheeting—to protect areas against waterdamage from leaks or runoff of water.

(h) Boxes for packing or carting. Record transferboxes (12" x 16") with inset handles and withlids are the easiest to use, carry, and stack. Theymay be purchased flat and folded into shape asneeded.

(i) Tissue paper, plastic "bubble-pack," and othermaterials to fill partially filled boxes or to protectdelicate items.

(j) Battery operated radio.(k) First Aid kit.

5. SwitchesStaff should know the location of electricity, gas,

steam, water, and sewer cutoffs so that these may beturned off as necessary. Specific individuals should beassigned these tasks and given hands-on training.Drainage outlets should also be designated to permit arapid dispersal of unwanted water.6. Temporary Relocation Space

Should it become necessary to evacuate the build-ing, temporary relocation space should be designatedand, when possible, designated in advance. Obvious-ly, the most favorable locations are in public buildings,churches, schools, gymnasiums and similar structuresaway from the disaster area. Whenever possible, it isbest to keep the institution's holdings together. Scat-tered locations will involve the splitting of staff, a needfor additional security personnel, and will only add tothe problems involved.7. Securing or Moving Holdings

Depending on the nature of the disaster, plans willvary as to the securing or moving of holdings. Thoseobjects on open shelves in the storage areas will besecured by nylon fishnet attached at the top of theshelf framing, ready to be unrolled and stapled as a

safeguard to prevent object movement. It may be pos-sible, in the case of a limited problem (loss of a portionof the roof in a storm) to move the affected collectionto another portion of the building. In the case ofthreatened flood it may be necessary to evacuate partor all of the building. These contingent plans must bedeveloped.

In the case of need for total evacuation, the follow-ing should be considered:

a. A plan for packing should be developed with anunderstanding as to who will be in charge of thepacking, who will pack, who will carry. The useof carry boxes, as suggested in 4(g), permitsloads light enough to be carried by amateurmovers without undue strain and insures hav-ing standard sized boxes so that they may bestacked safely and compactly in a vehicle orbuilding.

b. If shelves, cases, drawers, cabinets, etc., arenumbered in advance, the appropriate numbercan be placed on the box to make retrievalpossible if storage is for a long period or canmake return to the appropriate shelf, etc., easieronce the emergency has passed. Obviously amaster chart of these numbers and their contextshould be made beforehand; this is particularlyimportant if boxes have to be stored for a longtime.

c. If the damage is imminent, i.e., flooding, boxedobjects could be placed in a tractor trailer unit tobe driven out of the danger area until conditionsreturn to normal. The boxes could be left storedin the locked trailer, minus its cab, for as long asis necessary. Obviously, advance informationas to the availabililty of vehicles, supplies, etc.must be ascertained.

d. If materials are to be relocated in another build-ing, a plan for unloading and storage should bedevised with assignments for specific individu-als.

e. A plan for security should be created in advanceso that temporary locations or trucks housingcollections can be safeguarded. This may in-volve the hiring of security personnel or the useof staff on a rotating basis.

f. If water damage is involved, it is well to have alist in advance of freezer plants outside thethreatened area and the conditions underwhich they will accept materials (i.e., will foodlockers accept non-food items?).

57

Part 2Proceduresduring the Disaster

Once an emergency has been declared, staff should beable to follow the procedures listed in the Disaster Planbook issued by the institution.

One important consideration that we kept in mindwas that the safety of people comes first. Staff, super-visors, and management worked from this premise; inmost cases objects can be restored if not obtained oncemore; life and health are not as easily repaired orreplaced.

Outside assistance—police, fire, other public ser-vice units—can be of tremendous assistance. Unfor-tunately they can also do great damage inadvertently.Police can keep outsiders away from the problem areaand so safeguard collections; they can also bar stafffrom the scene at a time when staff knowledge can beof major importance. Likewise, firemen may need in-formation about the building and its collections so thattheir efforts do the least damage and the most good.

It is important, therefore, that firemen and policehave an opportunity to review emergency plans withmanagement and to get to know the affected buildingsso that their efforts during emergencies do not becomecounter-productive. Key staff members should beknown to them, and some system of identification orbadges should be worked out so that staff are notbarred from the scene at the time they can be helpful.

Security during the disaster can be a problem.Volunteer help can be invaluable, but staff or securitypersonnel should plan to keep other unauthorizedpersons from the premises. Once the fire, flood, orother catastrophe is over, the building should be se-cured physically to keep out the curious (there aremany).

Building materials can be obtained to erect bar-ricades; temporary guards can be employed asneeded, often with union help when unions are in-volved.

58

Part 3Action to Be Takenafter a Disaster

The reminders in an institution's "Disaster Plan" cov-ering action prior to the disaster should give a goodindication of the steps to be taken next.

It is essential that planning continue, since no ad-vance plan can cover all emergencies. Objectives,both immediate and long range, need to be set andthen constantly updated as required. Responsibilityshould have been divided among key staff in the origi-nal plan. This now may have to be modified by thesituation or because of the inability of important staffto be present. Policy decisions made at this juncturemay determine the institution's program for the nextfew years; thus it is essential to spend sufficient time inplanning rather than arriving at hasty decisions.

Decide what can be handled by the staff, whatmay have to be turned over to the outside experts.Invite those outside experts to advise you at the verybeginning (insurance will pay for this), and if restora-tion of a technical nature lies ahead, constitute a com-mittee of such experts to review periodically and guidethe restoration progress.

Above all, do not rush into an insurance settle-ment or the institution may settle for less than is war-ranted. It will take time to realize the full extent of theloss, and many unexpected expenses will occur in theperiod of restoration.

Volunteers will no doubt appear, and they shouldbe welcomed—but with caution. Too many volunteerscan create more problems than help; some volunteersare inept and can worsen the damage. Selectivity ofvolunteer help is imperative.

If water damage is involved, freeze all objects in anearby freezer locker or rent a freezer-trailer (with agasoline generator if there is no power available).Whole file drawers can be frozen; sorting can takeplace much later. Frozen collections cannot disinte-grate, mold, or be a hindrance to staff membersotherwise occupied in trying to recreate the institution.They can be worked on at a much later date as time orstaffing permit. If individual objects are to be frozen,wrap them in polyethlene sheets or bags if available sothey can be separated more easily later on.

It would be an ideal situation to know what can berepurchased—repurchasing is often far cheaper thanrestoration, be it furniture, books, or objects in thecollection.

59

Part 3

When outside experts are eventually hired to workwith the institution's staff, be certain to reach agree-ment on seemingly ordinary terms—the word "re-store" means one thing to a curator or a librarian; itcan mean something quite different to a professionalrestorer.

Above all, provide enough time for restoration.Staff will be under strain, and the damage of a fewhours may take years to repair. Patience and planningare essential.

60

Appendix I

MUSEUM STAFF •June 1972Thomas S. Buechner, President,Board of TrusteesPaul N. Perrot, DirectorKenneth M. Wilson, Assistant Director-CuratorJane S. Shadel, Assistant CuratorRobert H. Brill, Administrator, Scientific ResearchMary Catherine Dino, Curator of EducationNorma R H. Jenkins, Associate LibrarianVirginia L. Wright, Library CataloguerJane M. Lanahan, RegistrarAdrian C. Baer, CustodianRaymond F. Errett, Conservator-Photographer

Sherry Glosick, Student Assistant, LibraryDoris Hares, Assistant, Education DepartmentCarol Hull, Assistant for Slides and PhotographsFrances, G. Kragle, Secretary, Education DepartmentClifford Olmstead, Storeroom AssistantMildred Perry, Receptionist and Library AssistantPriscilla B. Price, Secretary to the DirectorLinda J. Randall, Secretary to Administrator of

Scientific ResearchLynette C. Roe, Secretary to Assistant DirectorWilliam W. Warmus, Student Assistant

All staff members worked tirelessly to help reopenthe Museum by August 1, 1972, thirty-nine daysafter the flood.

Restoration PeriodMUSEUM STAFF1972-1973-1974Robert H. Brill, DirectorDwightP. Lanmon, Curator, European GlassSidney M. Goldstein, Associate Curator, Ancient GlassJane Shadel Spillman, Assistant Curator, American GlassJ. Stewart Johnson, Deputy Director

Contemporary Glass (1974)Mary Catherine Dino, Curator of Education (through 1972)John H. Martin, Administrating O f f i c e r ( 1 9 7 3 . . . )Nancy D. O'Bryan, Business Manager (1973 . . . )Norma P. H. Jenkins, Associate LibrarianVirginia L. Wright, Assistant LibrarianPriscilla B. Price, RegistrarAdrian C. Baer, Operations ManagerRaymond E Errett, Conservator-Photographer1972Deborah Barker, Student AssistantSherry Glosick, Student AssistantMimi Cole, Slides AssistantDoris M. Hares, Assistant, Education DepartmentCarol W. Hull, Assistant for Slides and PhotographsDale B. Johnson, Student AssistantFrances G. Kragle, Secretary, Education DepartmentFatma Kassamali, Periodicals AssistantPaul Killigrew, Student AssistantJane M. Lanahan, O f f i c e ManagerDavid I. Leveen, Conservation AssistantJoseph J. Maio, Jr., Storeroom AssistantClifford A. Olmstead, Gallery AssistantMildred H. Perry, Secretary, Curatorial StaffLinda J. Randall, Secretary to the DirectorLynette C. Roe, Library, Serials SpecialistLaverne Schnare, Acquisitions AssistantWilliam W. Warmus, Student Assistant

1973 Staff AdditionsKristin A. Amylon, Conservation Assistant, InternMargaret Belden, TypistPaul A. Donahoe, Student Assistant, ConservationThomas W. Duncan, Paper ConservatorCharleen K. Edwards, Conservation AssistantDouglas R. English, Conservation AssistantAlbert Fehrenbacher, Model MakerDavid J. Fischer, Physical ScientistJeffrey Gaines, Conservation Assistant, InternH. Charles Herring, Student Assistant, ConservationJennifer B. Jones, Student Assistant, ConservationMoira MacAvoy, Student Assistant, ConservationRichard Michaiko, Student Assistant, ConservationMary O'Brian, Conservation AssistantMerle A. Peck, Conservation AssistantKatherine K. Poole, Conservation AssistantEllen P. Roche, Student Assistant, LibraryEdward Rowe, Glass RestorerAnn Southworth, Conservation AssistantJames E. Stein, Conservation AssistantPamela A. Thompson, Student Assistant, RegistrarNicholas L. Williams, Conservation Assistant1974 Staff AdditionsLois D. Hultzman, Museum AssistantBernadette Jacobus, Secretary to the Business ManagerRalph Kragle, Conservation AssistantAnne Maloney, Conservation AssistantBetty B. Sabin, Conservation AssistantJudy Seal, Secretary to the DirectorJudy Stasch, Conservation AssistantTom Steen, Conservation AssistantBonnie W. Wojnowski, Conservation Assistant

MUSEUM STAFF • 1976Thomas S. Buechner, DirectorJohn H. Martin, Deputy Director, AdministrationDwight P. Lanmon, Deputy Director, CollectionsRobert H. Brill, Research ScientistJane Shadel Spillman, Associate Curator, American GlassSidney M. Goldstein, Associate Curator, Ancient GlassDavid R Donaldson, Assistant Curator,

Contemporary GlassPriscilla B. Price, RegistrarCharleen K. Edwards, Assistant, PublicationsNorma R H. Jenkins, Associate LibrarianVirginia L. Wright, Assistant LibrarianRaymond E Errett, Conservator-PhotographerAdrian C. Baer, Operations ManagerLouise K. Bush, BibliographerLois D. Hultzman, Microfilm AssistantErnestine W. Kyles, CatalogerJane M. Lanahan, Secretary, Curatorial StaffJoseph J. Maio, Jr., Microfilm AssistantClifford A. Olmstead, Gallery AssistantGillian H. Paul, Secretary to the

Administratiue OfficerMerle A. Peck, Microfilm, Conservation AssistantBetty B. Sabin, Microfilm AssistantDarlene Scouten, Secretary, Curatorial StaffJudy Y. Seal, Secretary to the Research ScientistNicholas L. Williams, PhotographerBonnie W Wojnowski, Library, Serials Specialist

Appendix II

Volunteers Who Helped Reopen the MuseumJune AlexanderRalph AlexanderDavid AlienOlive AlienDiane AlpertAnne AmesMary Ellen AndrewsJohn AsielloLee AthertonRon AuthSusan AuthDeborah BaconLee BaldwinDave BaldwinDick BaldwinLois BarberRay BarberDebbie BarkerPeggy BeldenKatie BennettLilita BergsMary Ann BertaSue BlackDorothy BlairAnne BlizardHenry BoydMadeline BraunPeter BriggsElizabeth BrillPeggy BrillDonna BrownApril BrownTom BrydgesKaren BublitzBonnie BuechnerMary BuechnerMatthew BuechnerTom Buechner IIINancy BurgettLisa BurmeisterLouise BushShirley CallahanSally CampbellTom ChaseChris ChiodoJennifer ChiodoRex ClevelandMimi ColePatsy ColeRose CrainDwight CrandellEd CummingsSherman CurryHerb DannDick Darcangelo

Lindsay DavidsonDavid DoudBarbara EgathSue EgathJim EriacherPeggy ErrettJean FarnhamBiri FayFrank FehlnerDan FoxSteve FreyEdith GainesDavid GainesJohn GatesHarvey GerryRosalind GoldmanDan GrayWalt GriffinJeremy GuthPolly GuthSabrina GuthDavid HamerGeorge HamelBert HarrisAndrew HerczogMadeline HerczogBill HerthClarey HiltonEddie HoodHal HoodJohn HoodEric HopkinsMark HopkinsPeter HotraCarolyn HortonJames HoughtonLaura HoughtonRobert HoughtonRuth HoughtonAlexander HoughtonTom HullBob IversClint JanesGreg JarraleetChris JenkinsCody JenkinsWes JenkinsEarl JenningsDale JohnsonGreg JohnsonVal JohnsonLeslie KahlIvan KelloggKenny KlaneNancy Kohler

Mary KrampfBob Layton, Jr.Carol LanahanJack LanahanDwight LanmonMark LanningFloyd LattinJoan LeffelMartha LeffelPhil LeffelMarge LewisAnn LingLaurie LiscombGordon LuboldJean LuboldChris MacAvoyEllen MacAvoyPhyllis MacDonnellBill MartinJack MartinPhylllis MartinRobert MartinScott MartinTodd MartinLynda McCurdyPeg McKaleBev McKibbenStan McKibbenEd MishrellGerald MilletMichael MossPatty NealTim NealsonBob NikirkMary O'BrianMary Ellen OakdenAndrew OliverJanet OrrSheila OsterhoudtKay PepperChantal PerrotJoanne PerrotNina PerrotRobert PerrotDick PopeMary Ellen QuigleyBetty RaymondLou RemsAnn ReynoldsEllen RocheMary Kay RussellJef SaundersLiz SaundersJan SchreursClint Shay

Liz ShoemakerDave SmithLyn SmithBarbara SniderAmy SouthworthAnn SouthworthAnnie SouthworthGeorge SouthworthMary Ann SpragueSandra StoneRobin StoneJerry SwinneyHank TaylorBob Taynton, Jr.John TeachmanEd ThomasSteve ThomasLinda ThomsonAnn ThompsonBilly ThompsonKathy ThompsonPam ThompsonSue TongEmily UnderhillFlorence UnderhillAlan UnderhillAnne UnderhillChipper UnderhillEmmy UnderhillAlfred ValerioPeggy VineBernie VigerBarbara WakemanAndrew WatsonMartin WatsonBill WeberBrent WeddingAlan WernerHelen WernerJan WildePhil WildeJean WilsonPat WilsonLori WoodEllen WoodburyJean WosinskiAlison WrightJerry WrightKathryn WrightPeter WrightNancy YenawineDennie Younge

Appendix 111

Glass Center, Steuben, CGW Foundation and CGWEmployees Who Helped Reopen the MuseumDick AndrewsStella BaerDick BesseyFrank BurkeLarry darkAda CopeJohn FoxDick GoltryLarry GreenRaymond J. HallLeona HarrisBertha HolmesShellie HolmesDick JohnsonRed KeckVicky KingMary KnappArt MaioJoe Maio, Jr.Grace MaxwellTilly MirgliaMarilyn MurphyPeter MuthMike NegriDonald NoyesBill PerryKeith PriceTeresa QuailsLouise ResueTed ReydaJoe RossBob ShareFreeman Smith, IIIDave TrippDave WassonLyie WassonPeggy Wilson

ill

Appendix IV

Examples of Glass Restoration Problemsand Techniques1) Dragon-stem goblet with broken finial on cover(No. 51.3.115). (Fig. 1 a , b , c )As the symbol of the Museum, the Venetian dragon-stem goblet was the first piece restored. The gobletand cover were washed with water to remove mud; thesurfaces to be glued were cleaned with ammonia andacetone. Because of the spiral finial, holding the frag-ments in proper alignment was very difficult. A tem-porary cement was used as a clamp around the outsideof the joints. The cement was removed from eachjoint, one per day, and Araldite was run into the joints(Fig. 1b). In this way the object was properly sealed.

2) Enameled German footed beaker with couer (No.51.3.215), body broken into nine pieces and finialbroken off cover. (Fig. 2a, b, c)The pieces were cleaned with water to remove themud. The surfaces to be glued were cleaned withammonia and acetone. The fragments were taped to-gether. Since the outside had a painted design, thetape was applied to the inside, making certain thateach joint was as tight and accurate as possible. Whenthe object was completely assembled, Araldite was runinto the joints from the outside with a pointed spatula.After all the joints were completely filled, the excessadhesive was cleaned off the surface. This was donewith extreme care, for any movement of the fragmentswould weaken the joints by introducing air and possi-bly distorting the shape. The finial was attached withAraldite. (Since this type of repair is not readily visible,those who work with glass objects are reminded thatan object should never be lifted by a finial or a handle.)When the adhesive had cured, the tape was removedfrom the inside and the object was complete (Fig. 2c).

Fig. 1. (a) Venetian goblet before restoration and cleaning.(b) Araldite being run into joints, (c) Dragon-stem goblet afterrepair.

Fig. 2. (a) German beaker before repair, (b) Adhesive applied withpointed spatula along each of the joints, (c) Beaker after restora-tion.

IV

3) Venetian horse goblet (No. 54.3.236) with bowlbroken from the horse's head and one ear broken andmissing. (Fig. 3a, b)The body of the horse was full of mud. A previouslyrepaired section of the mane was removed. The bodywas washed several times with water, mild detergent,and sodium pyrophosphate without much success.There were two very small holes to gain access to theinside, one on the missing ear and another where themane was removed. Sodium pyrophosphate with 300mesh silicone carbide abrasive grit finally removed themud. In each case, the liquid was removed by placing asmall, thin straw in one hole to allow air to enter and toforce the water out through the other opening. Thesurfaces to be glued were cleaned with ammonia. Thehorse's head was protected, not supported, by alaboratory stand (whenever possible, it is best for anobject to be self-supporting). The metal ring, coveredwith felt and closed in front with tape, was not touchingthe object.4) Transparent bottle with applied red enamel decora-tion (No. 69.3.84), made by Maurice Marinot about1914. (Fig. 4a, b)A chip had broken off the lip of the bottle and was notrecovered. The broken portion was clear glass with anopaque fired red enamel. A replica of the missing partwas made in Plastogen G, poured in place with waxbacking to give it shape. After curing, the Plastogenwas finished and polished with small burrs on a flexible

shaft tool. The Plastogen was then removed and gluedin place with Araldite. The finished filled area wascolored with acrylic paint.5) Blue Persian cut ewer (No. 65.1.23), part of thePersian exhibit toppled during the flood. (Fig. 5a, b, c,d)The ewer was crushed beneath the case and smashedinto aproximately one hundred small pieces plustwenty-four main fragments. The main fragmentswere taped together roughly to find where the onehundred pieces belonged. Because of the peculiar wayin which the object was broken (in a zig-zag manner),the fragments locked themselves in place, but they hadto be assembled in order. Sometimes several pieceswere removed before one small piece could be fittedinto place. The ewer was taken apart and put backtogether at least six times before it was ready for anadhesive. Araldite was then run into each joint. Exceptfor some very small slivers, all the fragments wereused. An area 4.0 cm in diameter was missing but wasrestored with Technovit 4004a, colored with dyes tomatch the ewer. Facets were then cut into the restoredarea with burrs on a flexible shaft.

Fig. 3. (a) Horse goblet before cleaning and restoration, (b) Gobletstanding in a box of sand, protected, not supported by a laboratorystand.Fig. 4. (a) Marinot bottle before restoration, (b) Bottle after restora-tion.

50

Fig. 5. (a) Facet-cut ewer before restoration, (b) Detail of fragmentsof ewer. (c) Rolf Wihr examining cut ewer during restoration.(d) Ewer after restoration.

Appendix IV

6) Engraved Roman beaker (No. 66.1.238). (Fig. 6a,b, c, d, e, f , g, h, i)This previously repaired beaker had come apart dur-ing the flood. The object was taken apart and washedin water. Surfaces to be glued were carefully cleanedto be certain that the old adhesive was completelyremoved. Cleaning was checked by holding the frag-ments to the light so that the light reflected off thesurface. The surfaces were then cleaned with am-monia and acetone before being taped together. Be-cause of engraving on the exterior, the tape wasapplied to the smoother inside of the beaker. (Adhe-sive can sometimes seep through the joints and hardenon the tape, making an object very difficult to clean onthe rough engraved surface.) Araldite adhesive wasrun into the joints from the outside. When the joint wascompletely filled, the excess was removed im-mediately. After the adhesive had cured, the restora-tion of the missing area began. Dental wax was usedfor the mold. By warming it over a flame or hot air, animpression was taken from another part of the objecton the inside. The contour was carefully molded to bethe same as the missing area. Wax slightly larger thanthe area to be filled was then placed over the area to berestored and scored with a sharp instrument. This wasdone by holding the object against a light to permitlight transmission through the wax. The shape wasnext cut with a hot knife to prevent the wax frombreaking. The wax mold was coated with a polyvinylalcohol to act as a separator between wax and resinand attached to the glass with a hot knife to make

certain there were no openings for the resin to leakthrough. This procedure was then repeated on theoutside, with one exception. An opening was left to fillthe mold. The placement of the opening was chosenwith care. To make sure that the mold filled properly,all undercuts faced the opening to prevent trapped airpockets. A wax spout was built over the opening toassure a steady flow of resin without forcing air into themold and to act as a reserve tank for shrinkage.

Since the wax is translucent, the condition of theresin in the mold could be observed. If air becametrapped, it could be released by making a small holewith a needle. After the air was released the hole couldbe sealed quickly with a hot knife. The resin used forthis object was Plastogen G. The wax was then re-moved, the spout cut off, and the object was polishedwith a flexible shaft and cleaned. When restoration of amissing area was completed, Araldite was run betweenthe resin and the glass to give a better seal to make theobject more stable.

VI

Fig. 6. (a) Engraved Roman beaker before cleaning and restoration, (b) Objectbeing fitted together, (c) Beaker taped together with other pieces not yet in place.(d) Araldite being run into joints, (e) Inside section of wax mold. ( f ) Outside sectionof wax mold being attached with a hot knife, (g) Completed mold with pouring spout.(h) Mold partly removed showing cured Plastogen restoration, (i) Restored beaker.

7) Sprinkler, Roman Empire, possibly Syria, 2nd-3rd.century A.D. (No. 59.1.148). (Fig. 7a, b, c, d, e, f )The top of the object was detached; the foot wasmissing previously. The object was very carefullycleaned with water. A solution of sodium pyrophos-phate was used to clean the mud from the cracks.Since there was no foot, it was decided to make one.The size and shape were based on a photograph of asimilar object. The foot was turned freehand in clay ona potter's wheel and attached to the object for the finalfinishing (Fig. 7c). After curing, the rubber was re-moved; the clay was cleaned from the mold and theobjects. The silicone mold was replaced on the objectand sealed with a small amount of uncured siliconerubber. The mold could be filled from the top, using astraw as a pipeline to carry the resin into the mold fromtop to base without the resin touching any other areasof the object (Fig. 7d). Technovit 4004a was used; itsnatural color was similar to the object and it would notreact with the silicone rubber. The mold was then cutaway, revealing the molded foot (Fig. 7e). The top wasattached with Araldite.

Fig. 7. (a) Sprinkler before restoration, (b) Foot being shaped inclay. (c) Pouring of rubber mold. (d) Mold in place with clayremoved. Mold will be filled from the top with a straw as pouringspout, (e) Removal of plastic foot from rubber mold. (f) Completedsprinkler.

Appendix IV

8) Islamic beaker (No. 70.1.7). (Fig. 8a, b, c)The beaker was broken into eighteen pieces. It hadbeen restored with plaster, and Hakes of iridescencenot part of the original had been glued to the surface(Fig. 8a). The object was very carefully cleaned toexpose the original glass surface, taped together, andglued with Araldite (Fig. 8b). At this point the objectwas in a very unstable condition, with a number ofsmall pieces missing on one side and about one halfmissing from the other side. All the small sections wererestored first, by taking wax impressions of the outsideportion of the beaker to match the areas to be restored.The wax was placed over the missing portion; Plasto-gen G was carefully laid on the wax from the outside.This was accomplished by waiting until the Plastogenwas nearly ready to jell. After the resin had cured, thewax removed, and Araldite had been run into thejoints between the glass and resin, this section wascomplete.

In restoring the larger missing part on the otherside of the beaker, two problems arose: 1. After a waximpression was made from the already restored areaincluding two of the sharp relief-cut circles, the waxbecame very thin because of its proximity to the sharpedges of the relief-cutting. Another layer of wax wasadded to give more strength to hold the shape. 2.When the shape and design were transferred to themissing part, the cut circles did not line up properly.Instead of making an impression of both circles atonce, the impression of one circle was placed over themissing area, the location of the other circle was

marked, and an impression made at that point. Thefinal impression had to be shaped more by free handthan would normally have been done.

In cutting the wax mold to its final shape so it couldbe filled from the top, which was completely open, thewax was extended well above the top lip. This extraextension allowed overfilling the mold, which col-lected the air bubbles and also acted as a resin reserveto allow for shrinkage. The inside mold, having asmooth surface, could be completed and sealed inplace with only one layer of wax.

The mold was then filled with Plastogen G andallowed to cure. The wax was removed, the extraextension on the lip cut off and polished. The Plasto-gen was tinted with oil paint thinned with xylene.

Fig. 8. (a) Cut beaker with a series of bosses, before cleaning and Fig. 9. (a) Islamic cut bowl before restoration, (b) Shaping theduring restoration, (b) Beaker before restoration, (c) Beaker after exterior of the bowl in clay. (c) Applying silicone rubber to the clayrestoration, model.

vill

9) Islamic cut bowl (No. 55.1.136). (Fig. 9a, b, c, d, e,f , g , h , i , j )An Islamic cut bowl previously restored with plasterhad come apart. There were thirteen fragments whichcomprised approximately fifty per cent of the object.The old restoration and adhesive were cleaned fromthe object. Surfaces to be glued were cleaned withammonia and acetone. The fragments were then fittedand taped together. Araldite was run into the joints.The bowl was placed on a flat surface, upside down,and clay was used to form the outside contour,eliminating the cut design because shape was mostimportant; the design was still visible on the other side(Fig. 9b).

Because the outside shape and size were of chiefconcern, the clay on the inside was left in a very largelump. When the desired shape was produced, the claywas covered with a layer of silicone rubber which hadbeen thickened with Cab-O-Sil so it would not run.This rubber mixture was placed very carefully to pre-vent air between the clay and rubber (Fig. 9c). Notethe excess clay around the edge of the base, later usedas a spout for pouring the resin. After curing, thesilicone was covered with a layer of plaster firmenough to hold the shape of the rubber and to supportthe weight of the object.

The bowl, with clay, rubber, and plaster was thenturned over to shape the inside (Fig. 9d). The thick-ness of the clay was measured by inserting woodenpegs through the clay to the silicone rubber. A layer ofsilicone rubber with Cab-O-Sil added was put on the

inside and covered with plaster to permit easy removalof the plaster in one piece. The plaster and rubberwere removed from the inside (Fig. 9f). All the clay wascarefully removed from the mold (Fig. 9g), withoutbreaking the seal between the glass and rubber whichcovered the outside of the mold. This procedure couldbe reversed if necessary, depending on the shape ofthe object. Experience indicated which side would bethe easier to remove. In this case the inside came offmore easily because the scallops and cutting tended tolock the outside in place.

The rubber mold was then sealed to the objectwith a thin coat of uncured rubber all along the edges.This acted as a gasket to keep the resin from leaking,and it was similar to the wax seal used on Fig. 6 torepair the Roman beaker (Fig. 6f). The plaster waspressed back into the rubber to give it support beforethe seal began to cure. The mold was then turned overand filled with Araldite dyed green (Fig. 9h) throughthe opening left by the excess clay on the base. Thisfilling procedure was done in the morning to preventany leak in the mold. After curing, the mold was re-moved (Fig. 9i) and the object cleaned.

Fig. 9. (d) Shaping the bowl's interior, (e) Completed mold insideand outside, ( f ) Removing inner rubber mold; note plastic insert atleft with metal handle, (g) Rolf Wihr removing clay and cleaningmold. (h) Pouring Araldite into rubber mold. (i) Removal of outermold. (j) Completed restored object.

IX

Appendix V

Materials and SuppliersADHESIVES:

(a) AralditeAY103(b) Hardener HY956(a) RP-103 (Resin)(b) H-956 (Hardener)

17%P.B.WDuco Cement

Ciba Company GmbHWehr, Baden, GermanyRen Plastics5656 South CedarLansing, Mich. 48909Hardware stores

Not available in the United States.

Substitute for Araldite AY103

An easy to obtain, no mixing adhesivefor temporary repairs.

CASTING RESINS:Technovit 4004a

Plastogen G

Polyester-GIEBHARZ GTS

MOLDING MATERIALS:Silastic RTV(A Silicone Rubber)Jeltrate

Plaster of Paris

Baseplate Wax

CLEANING MATERIALS:Lymoff

Kulzer & Co.Frolingstr. 29, POB 261Bad Homburg v.d.Hoohe, GermanyAlfons SchmidtSt. Germanstr. 14Speyer, Bavaria, GermanyVosschemie2082 Uetersen beiHamburgEsinger Steinweg 50Germany

Dow Corning Corp.Midland, Mich.L.D. Caulk Co.Milford, DelawareHardware stores

Dentsply International, Inc.York, Pa.

The Lymoff Co.6728-4th StreetSt. Paul, Minn. 55119

A hard, glassy, room temperature curingpolymethacrylate resin. Cloudy in color.

A room temperature curing polymethacrylate resin.Yellowish in color.

A room temperature curing polyester resin.Clear in color.

A durable, fast-setting silicone rubber.

A fast, pliable, water setting mold material.Must be used immediately.A durable, easy to use, dense mold-making material.Dental wax.

Lymoff should be used mainly on glassware,as it tends to corrode metals.It works most effectively on recentdeposits of lime.

TOOLS:Flexible Shaft Paul H. Gesswein & Co., Inc.

235 Park Ave. SouthNew York, N.Y. 10003

Used for cutting, buffing and polishing.Complete descriptive catalog available.

Potter's Wheel

MISCELLANEOUS:Acryloid B-72

Cab-O-Sil(Fumed silica)Grade # EH-5Spectrum 100

Ultrasonic cleaner

Sodium Pyrophosphatesolution10 grams/liter of HzO

Rohm & HaasPhiladelphia, Pa. 19105Cabat Corp.125 High StreetBoston, Mass. 02110Abbey Materials Corp.116 W 29th St.New York, N.Y. 10001

Protective coating for wherever it is necessary.

Filler or thickening agent. Very goodin silicone rubber to keep itfrom running.A cold setting enamel paint used torestore missing decoration.

Used to clean mud and dirt from unreachable areasof broken objects only, due to its tendency tocause fractures.Cleans mud from cracks.

Appendix VI Appendix VII

Acknowledgements, References, Chapter IVThe author is indebted to Dr. R. H. Brill, Director, andDr. J. H. Martin, Administrative Officer, The CorningMuseum of Glass for their encouragement of theseinvestigations, conducted with the assistance ofRobert G. Kephart. Special thanks are given to Mr.Thomas W. Duncan, paper conservator, and his assis-tant, Mr. Merle A. Peck, for their help in conductingthese experiments. Appreciation is due Mr. Robert M.Organ, Chief of Conservation, Analytical Laboratoryat the Smithsonian Institution, for his information onuses of ethylene oxide. Valuable contributions duringthe initial stages of this study were made by Mr. DaleJohnson, a summer assistant. The gracious coopera-tion of The Corning Community College Administra-tion in allowing use of their laboratory facilities forconducting microbiological tests and analyses was ap-reciated.

ReferencesErnst, R. R. and Shull, J. J., "Ethylene Oxide GaseousSterilization—Concentration and Temperature Ef-fects," Applied Microbiology, 10,1962,pp.337-341.Gilbert, B. L,, Gambill, F. M., et al., "Effect of Moistureon Ethylene Oxide Sterilization," Applied Microbiol-ogy, 12, 1964, pp. 496-503.Opfell, J. B., Wang, Y, et al., "Penetration by Gases toSterilize Interior Surfaces of Confined Spaces,"Applied Microbiology, 12, 1964, pp. 27-31.

Acknowledgements, Chapter VThe author is indebted to Dr. R. H. Brill, Director, andDr. J. H. Martin, Administrative Officer, The ComingMuseum of Glass for their encouragement of theseinvestigations. Special thanks are due to the conser-vator, Thomas W. Duncan; the librarians, Mrs. N. P. H.Jenkins and Mrs. V L. Wright; their aides, and theassistant conservators for their conscientious attentionto evaluating these dried products. Appreciation is dueMr. Richard Shoulberg, Applications Engineer at theGeneral Electric Company, for his prompt attention toour unanticipated problems during the drying opera-tion.

XI