@NE\x/s LETTER 17Corpus Vitresrum Technical Committee

Physics, University of York 27 November 1975 Ref. TC/75l8

CONTENTS

1 GENERAL

2 WEATHERING OF GLASS

3 LEADING OF WINDOWS

4 CORROSION OF GLASS AND THE PAINTED LINES

5 NEW ABSTRACTS

1 GENERAL

1.1 CORPUS VITREARUM

Some significant changes were agreed atthe Co:rpus Colloquium held in Paris fnom Bth to12th Septemben.

First, the words trMedii Aevitt have beendropped fi:om the title. Thus the ConpusVitrearum can now concern itself with stainedgl-ass l-ater than AD.1500 and this means thatHolland can panticipate because the imPortantglass at Gouda is of the 16th century; in factoHolland has now been elected as a Member of theCorpus.

Second, more money can be sought (in theUK) for helping in the conservation workbecause some chanitable tr:usts do not givemoney for medieval studies.

Third, new Statutes were adopted. Thesehave been discussed during the yea:r, first byCIHA on 5th April, then by the UAI at itsmeeting on 17th June, again (with modifi-cations) by CIRA on 3rd Septemben and finallyby the Co::pus on 9th September.

The officens of the Corpus wene electedon 12th Seotemben, as foll-ows:-

President: Professor Louis Gr:odecki (Fr:ance)VicePnesidents: Dn RUdiger Becksmann (Federal

Republic of Genmany)Frau Dr: Eva Frodl-Kraft (Austria)Professor Roy Newton (United

Kingdom)Secretary: Madame Frangoise Penrot.

The membership of the Technical- Corilnitteeis being revised under the Chairmanship ofProfessor Roy Newton with M J.M. Bettembourgas Secreta::y. An active programme of work isenvisaged and Czechosl-ovakia has now appointeda representative, Dr Ludvlck Losos who i.s botha chemist and an art-historian.

Other significant changes ane alsoexpected because Professor Roy Newton has madea pensonal appnoach to the European ScienceFornrdation for. support for: r'esearch activitiesin the fieLd of conser:vation of windows. TheESF has responded with a proposal to set up aStudj Group for this purpose and we lookfonwand to a stimulating exer:cise in thisdirection.

I.2 THE PARIS COLLOQUIUM, B-12 SEPTEMBER 1975

I.2.I General

This Colloquium was an unforgettableexpe:rience for all- those who attended it.Underlying the whole week, of couuseo was thebackgr:ound of marvellous Fr:ench hospitalitywhe:reby evenything was arranged for us, thecoaches, the mea1s, the directions, theattractive guides from Champs, the overallplanning, the twin projectons for the lecturersothe opportunity of experiencing a Itbuffetcampagnalerr at the Chateau de Champs, and thesmoothness with which evenything nan.

Each person will bave had his own high-Iights. For me the supnerne noment was that ofseeing the nestoned Jesse window at Chartnesbecause I like to see stained glass windowswhich glow and let light into the bui.lding asthey sunely did in the 13th centur.y. Moneoven,i.t was a great relief to discove:r that thecoating of Viacnyl VC363 on the extenior couldnot be detected; it did not have the shinysunface which might have been expected.Anothe:r faseinating moment was when we sawtPhoenix-li.ke, the new Centne Intennational duVitrai.l arising from the nuins of the }3thcentury building through the effonts of theCity of Chartres, MM. PJ.erne Firmin-Didot',Jearr Tanalon and the chief architect,M.G. Nicot; Chartres will then have a centreof learning and of documentation to add to itsothen i^ronder.fuL attractions o including the6g1ise Saint Pierre de Chartnes which we alsovisited.

I was not able to visit the ateliensopenated by Mme Gaudin and M Gnuber but thosewho did so were fasci.nated by seeing themarvellous panels at close guarters. The nexthigh-spot fon me was the thnee TechnicalSessions; thein contents ane listed inSection 1.2.2 but useful and stimulatingdiscussions we::e evoked by then all.

1.2.2 Th" T""hri..1 P"og

This is merely a list of the contri-butions fo:r the benefit of those who could notbe present; they will all be publishedwithin a few months and wilt be abstracted inthe News Lettens.

Monday 8th Septemben in the aftennoon

t. Dr J.C. Fernazzini trThe influence ofcon::osion on the spread of deterionation ofmedieval glassesrr.

2, Pnofesson R.G. Newton "Using triangulandiagnams to understand the behavioun ofmedieval glassesrr. (See Abstnact No.205.)

3. Ir. J.M. Bettembour:g "Composition andweathering of ancient window glassestr.

4. Professor Collongues, Mme Penez,M. Tilloca and M. Dall-as "Some aspects of thephenomenon of corrosion of ancient glassesfrom Fnench chu:rchesrr. (See item 2.1 below.)

Tuesday 9th Septemben io the monnj.ng

5. Dr. Mar"tha Spitzen Aronson ItA study ofruby glasses fnom medieval windowsrt.

6. Dr W.P. Bauer trThe effect of cleaningtechniques on glass surfacesil.

7. Mn Dennis King ttProblems encountened inthe conservation and restoration of windowstt.

B. Professon G. Mar"chini "The problen ofrestoning paintwonktr .

9. Mn T. Husband rrThe restonation by theCloistens (Museum) of six panels of a stainedglass windowrt.

Twsdav 9th Septemben in the afternoon.

10. Dr Eva Frodl tbaft ilSome nemarks on thescience and craft of restoring stained glassrr.(See Abst:ract No.202. )

11. Dr. Jane Hayward trlnstallation of theSt Leonhard glass at the Cloisters - protectionand conservation". (See Abstract No.2O3. )

12. Pnofessor R.G. Newton "Experimentalstudies of the protection of rnedieval- windowsusing external glazings". ( See AbstractNo.206. )

13. In tI.M. Bettembourg rtThe protection ofancient window glasses and the conservation ofthe paintwonk; the use of synthetic resins".

1.2. 3 Acknowl-edgnents

So we have to .thank our new Pnesident,Professon Louis Gnodecki, and the FnenchInspecton Genenal of Histor,ic Monuments,Jean Tanalon, for having the inspination todevise such a marvellous pnognammeo andICOMOS, ICOM and the Rome Centre for helpingto implement it. Also rI.M. Bettembourg andall the people at Champs for helping to assistit so effectively at every point, and fordemonstr.ating the interesting research wo::k atthe Laboratoine de Rechenche des MonumentsHistoniques, covening not only stained glassbut also stone, paintings, polychromed wood,munals and palaeolithic cave drawings. Specialmention of cor:rse must be made of our charmi"ngpotyglot, Franqoise Perrot, who was anuntining transl-ater, diplonat, pacifier andeven vigilante.

1.3 .KIRKBY WHARFE CHURCH

Page 2, column 1, of News Letter No.llcontained a comnent by Dn Ernst Bacher thatthe lSth century glass at the smal1 par,ishchunch at Kirkby Whanfe does not show thecha:racteristic itEnglishrr type of weatheringbut has the laminar erust which is commonlyfound in Austria. I added a note that a studywould be made of the chemical- composition ofthis glass, and the work has now been done.

Appnoximate analyses were canried out onthe MECA 10-6 equipment (see item 4 of NewsLetter No.15) by courtesy of Mn M. Jocelyn ofLink Systems Ltd., 35 Spring Gandens, Buxton,De::byshine, England. However, the glasscompositions were of a type not usual inEngland and hence no trstandandrr glass wasavailable with a suitable composition but itis neverthefess clear that (a) the glass hadan trAustriant' typ" of conposition and (b) thefive pieces of glass examined fall into twogroups with slightly diffenent compositions.

The genenal composition was unusualbecause only three oxides were pnesent insignificant amounts, silica, potash and limeas shown below, the potash contents of thefipst group being the highest yet neponted.The othen oxides are pnesent in only smallamounts:- soda (Na2o) was low, about 1t;magnesia (Mgo) about 3t; nanganese oxide (Mno)about 0.5% and iron oxide (FeZOS) about lt.

The main oxides wene (weight pencentage):-

Sample Nos.

Sitica (SiO2;

Potash (KzO)

Lime (cao)

r13,r74,r77

50

27

L7

Total 94 93

These compositions are consistent withothen analyses, ::eponted by Dn W.P. Bauenr ofAustrian glasses which form cnusts on weathen-ing (Conpus sanples nos. 9, 47 and 54) andthus it seems that it is the chemicaL nature

'r of the glass (and not something to do with any"Austr.ian weathenr') which is nesponsible forthe special type of detenionation. (See aJ-soFig.6. )

The cost of nernoving the glass fi:om theChunch and then neplacing it was rnet fnom thePilgrim Tnust Special Gnant and we ane gnatefulto the incumbent, the Revenend P. Wood, and theParochial Chur:ch Council for thein help inenabling this intenesting study to be canriedout.

1.4 DANGERS OF USING HYDROFLUORIC ACID

News Lettens Nos. 15 and 16 both containeditems on the precautions to be taken whenusing hydrofluonic acid. I arn now indebted toMr D.B. Honeyborne for dnawing my attention tothe Building Reseanch Advisony ServiceTechnical Infonmation Leaflet TIL 44. This3-page document describes the precautions tobe taken in handling caustic soda, causticpotash, hydrochloric acid and hydnofluonicacid.

The document is particularly usefulbecause page 3 contai.ns instnuctions (inmedical language) to doctons on how to treatinjuries fnom hydrofluonic acid. Copies canbe obtained fnee of charge fncm the BuildingResearch Advisory Senvi.ce, Ganston, Watfond,WD2 7JR, England.

1.5 DIRT IN THE ]NTERSPACE OF THE ISOTHERMALGLAZ]NG AT BERN MINSTER

In item 2.1.5 (p.7) of N.L. No.15 thenewas a conment that pant of the 3o-yean-o1disothennal- glazing at Benn Minster. had beeninspected by Konrad Vetter and Derek White.They had found no dirt on face 3 but there wasa filn of dint on face 2. Derek White wipedsome of the black deposits fnom face 2 with apiece of white nag and then sent the rag to me.

175,176

62

20

11

This bit of old rag then had an eventfuljou:rney, being sent to Zunich wheneDr J.C. Fenrazzini kindly subjected some ofthe deposit to X-ray diffnaction. On heating-the*;deposit to 300oC it glowed as the sootbunnt off and the remainder pnoved to beanhydnite (casO+); the original deposit waspnobably one thind gypsuxn and two thinds soot.Befo:re the nag was sent back to Switzenland,Professon Heavens at York kindly submitted thedeposit to the Isopnobe and detected calcium,silicon, sulphur, iron, zinc, potassium andtitanium. It seems that the bulk of thematerial was air-borne soot and a salt ofcaleium, pnobably limestone dust converted togypsun by the sulphur dioxide in the atmos-phere.

I.6 SPONTANEOUS FRACTURE OF MEDIEVAL GLASS

A still unsolved pnoblen in the conserv-ation of stained glass is the occurnence ofspontaneous fnaeturing of one sunface only ofthe glass (see N.L. No.B, abstnact No.157).All- attempts to expJ-ain it have failed to beconfirmed by experinents made on the glassbut I am happy to say that Dr R.B. Waterhouse,of the Depa:rtment of Metallungy of theUnivensity of Nottingham, has taken an interestin the pr.oblem. He and his colleagues will beusing the ESCA (electnon spectroscopy forchemical analysis) technique to discoverwhethen thene are any detectable differencesin the compositions of the two sunfaces. We

look forwand to learning whether. they discoversomething of intenest and, in this connection,foun nelevant abstracts (Nos. 201, 2O8, 2L2and 213) are included in this issue.

1.7 . ARE PIECES OF MEDIEVAL W]NDOW GLASSTHICKER AT TI{E BOTTOM?

Thene is a widespread bel-ief amongstphysicists and engineens, especially in theUSA, that pieces of medieval stained glass arethicker at the botton than at the top. Thisis assumed to demonstnate that plastic flowof the glass has occunned during the severalcenturies that the pieces have been in thewindow! However, it is also known that plasticflow of glass does not occur at room temper-ature until the stness is one tenth of thebneaking strength of the glass! Thus anytendency for rnedieval glass to be thi-cker atthe bottom will not be due to plastic flow.Does anyone know of any evidence that piecesof medieval glass neally ane thicker at thebottom? If so, what is the reason? (Onepossible reason for glazing the window withthe thicker part of the gJ-ass downwards mightbe to gai.n an additional refractive effect. )0n is it a mistake to believe that the thicker:part is downwar.ds?

Potassium

Sulphur

B Magnesiu m

D Calc iu m

J' **"".i,;,' & ., @

. i:t:' ?:, :l::t, ;:

-44::i;ii

:t !,tit,a.

:.:::

%w' !," &b

,,,

&.4,+

W. " "&<-Cu."y&"! .

a. .:t j?

Aluminium Silicon

Fig. I These electron microprobe 'pictures'of a section through a pit in a piece of glass from Amiens cathedral reveal the sequence in

which corrosion takes place. A and B show that all the potash and magnesia are lost from the pit (because the area of the pit

appears dark whereas the glass at the top is lighter in colour). The pit contains calcium sulphate (gypsum) in region K because

both C and D have a pale area at K, which is dark in all the other pictures. lt is concluded that atmospheric sulphur dioxide

does not attack the glass directly, but it attacks the corrosion products which were formed earlier.

'ta: 4

2 WEATHERING OF GLASS

2.L MECHANISM OF ATMOSPHERIC ATTACK ON

MEDIEVAL GLASS WITii SPECIAL REFERENCE TO

ATTACK BY SULPHUR DIOXIDE

In News Letter No.15, page 12, I made useof the text of a PaPelr' Presented at the PanisColloquium on 8th Septernber by PnofessorR. Cotlongues and l'lrne Perez y Jonba, in onderto discuss their nemarkable demonstnation ofthe effect of sulphun dioxide. I now haveperrnission fi:om Pnofesson Collongues toreprint Fig.5 of that paper and it aPPearshene as Fig.1. I also.have pleasune in statingthat this work was fundird from the Laboratoirede Recherche des Monuments Historiques andfor:ms part of the research Programme at ChamPS-

sur-Marne, unden the general supervision ofM Jean Taralon, Inspecteun Gdndral-e desMonuments Histoniques.

Fig.1 shows six electnon micr:oprobertpictunestt of the same much enla:rged (x260)section through a pit in the glass; each is3OO pn square. The material at the top ofeach picture is the unaltened glass and thesemicircular shape at the bottom and towar:dsthe right is the pit. Fig.lA shows thedistribution of potassiuml it occurs in theglass (the white part) but it is not Presentin the pit because potassium compounds anesoluble in waten and they were washed away bynainwater: (syngenite was not fonrned in thispit ).

Fig.18 shows the distnibution of magnes-ium; again it is pnesent in the glass butnot in the pit because magnesium compounds ar:ealso so.l-uble in water and B looks exactly likeA. The absence of magnesiun comPounds heneneminds us that it is an over-simpli.ficationto assrme that magnesia is exactly equivalentto lime in giving pr:otection to glass againstattack by water.

Fig.1C will be discussed below. Fig.lDshows the distnibution of calcium; it ispnesent in the glass and in the Pit but it isabsent from zone Go close to the glass. Thepre.sent irypothesis is that zone G is free ofcalcium because the attack at the bottom ofthe pit is by wq!Sl_gn.H. The calciumhydroxide thus p:roduced is solubl-e in thewater: and moves towands the surface, so thatzone G tends to be free of calcium. In zone H

the calcium hydroxide reacts with carbondioxide from the atmosphere to produce in-soluble calciurn canbonate (calcite). In zoneK the calcite reacts with sulPhun dioxide topnoduce calcium sulphate (gypsum), see be1ow.

Fig.lE shows the distnibution of alum-inium; it is uniformly present in the glassand it is afso pr:esent in the Pit but aPPanent-1y in a more concentnated fonm which has manyfissur"es in it. Note that it is absent fr:onarea K at the top of the pit. We do not yetknow in what fonn the aluminium exists in thepit.

Fig.lF shows the distnibution of silicon;it is pnesent in the glass and in ttre Pit butNOT in area K. It appears to be somewhatffin"n i.n zone G and thu fi""u"es which can beseen i.n E also appear in F.

We can now retunn to Fig.lC which showsthe distribution of sulphur:. It is notpresent in the glass (at least not in enoughconcentnation to be detected) and it occunschiefly in anea K. Thi.s area must theneforebe gypsun because it occurs in both C and D

but NOT in ArBrE or F. Sulphun also occu:rs inthe fissur.es seen in E and F. The glass was

'rAmiens No.2" (Conpus Vitrearum sample No.27)whi.ch has the weight percentage composition:Si02, 52.6; Na20, 2.29; K20, 13.4; MgO'6.70; CaO,14.4; MnO,0.85; CuO' 0.14;A1203,2.0; Fe203r 0.65; Ti.02' 0.12; P205'5.2; CoO,0.II; NiO' 0.Il-.

Much discussion will take pl-ace overthese impor:tant photographs for many years tocome, but he::e I want to point out that thesulphur is found in anea K and not in contactwith the g1ass. f therefone conclude thatr atleast in this pit, atmosphenic sulphur dioxidedid not attack the gfass directly and the

reaction.AIso, the absence of calcium from zone G

supports the hypothesis that, in the pro-duction of this pit, @was water. Even canbon dioxide seems to have

ffi-Econda::y effectr in pnoducing zone H.

2.2 GLASS WHICH HAS BEEN AFFECTED BY DAMP

SAWDUST

Item No.3 (pages 10-12) of News LetterNo.l6 described an experiment in which asample of No.2 glass, Partl-y fine-polished andpartJ-y ai:rbnaded, had been bur:ied in damp oaksawdust fon six months. The fine-polishedsurface had become pitted but the airbradedsunface had appanentl-y lost about half of itsnoughness.

I have been wornied as to how to explainwhat happened. In Fig.7 (of N.L. No.16) theline ABE su::ely nepresents the original sunfacebut in Fig.8 (of N.L. No.16) the line EGH

probably does NOT repnesent the oniginalsunface. It now seems that the sawdust erodedboth kinds of surface, as weII as pnoduci-ngpits in the fire-polished sunfacer becauseMr Peten Gi.bson had the forethought to framethe sampl-e in lead and the parts of the surfacepr:otected by the lead ane now at a highen levdIthan the remainder. Fig.2 of this neport showsan enlarged (x 12) photogr"aph of the edge ofthe sample; hene A is the Part Pnotected bythe lead, B shows an enoded part of thepolished anea, C shows a pit in the fire-polished anea and D shows the enoded airbradedsurface. It is cLean that the sawdust haspnoduced a gene:ralised attack on the glass inaddition to pnoduei.ng the pits.

Attenpts were made to measune the heightof A conpaned with B usi.ng the Talysurf butthe expeniment was defeated by the generalcurvature of the sunface. The best that canbe said is that the original surface may wellhave been about 10 pm above the pnesentsurface and thus the reduction in noughness

My item in News Letter No.15 ttRewo:rked

old leads or new ones?'r (item 1.7 on page 4)has pnoduced sevenal nepliesr the most eruditeof which was from M Jean Tanalon in Paris who

drew my attention to his important r"emanks onpages 242 to 246 of 'rles Monunents Historiquesde la Francett 1962, No.4. Whereas my quenyhad been along the lines of Aladdints lamp and

the cny of trNew ramps for oldrt ' 'Jean Tar"alonsays ttNo, keep the oJ-d lamPslrr.

He points out that the belief that "1eadsneed renewing every 100 yearsrt is a miscon-

of the airbraded surface by about 9 um couldbe attnibutable to a general attack all overthe surface by the sawdust in the period ofsix months. The pits, a.l-so, would no doubthave been some IO pm deeper in reLation to the

"cirfigTiat- sunface!

$$$$$r$'\ffi

Fig. 2 This is an enlarged view (x 12) of part of the sample of No. 2 glass which had been stored in damp

sawdust for six months. The dark area at A is the original polished surface where it had been protected

by the leading. The surface at B had mistakenly been thought to be the original surface (see N.L. No. 16,

Fig. 8, line G to H) but the sawdust has eroded the surface to a depth of about 10pm and has also

produced isolated pits (as at C) which can sometimes be as much as 60pm deep. The area to the left(at D) is the part which had been airbraded and it also has been eroded by the sawdust, but no pits

were produced in this area, only the whitish spots which can be seen in the photograph. We do rrot

know why pitting occurs only on the part which had been fire-polished and not the part which had

been airbraded.

3 LEADING OF WINDOWS

ception because many cases have come to lightwhe::e.f2th or l-3th century oniginal leading isin a good state of preservationl j-t shoul-d bekept and it may only need to be repaired, notreplaced (16th C and l-ater leads are a differ-ent matter). His articl-e shoul-d be consultedfor: the fu1l discussion but, briefly, hepoints out that these early original leads aredifferent from modern leads, being (a) a

different materi-al which may contain as much

as 3Oeo tin and (b) a differ:ent shape (seebelow) which gives much more rigidity to thewindow.

$r\.\\:\N

Mn Rowan LeCornpte wnote from New Yorkalso to comment on the weakness of the thintwide-flanged, modern lead usually employed inne-Ieading ancient windows compared with adome-flanged lead which again gives nigidityto the window and gneatly neduces buckling.The sections of these leads ane shown, fullsize, below, (A) being the moderrn French feadwith flanges O.3nm thick and (B) being thatf:rom Sens Cathednal of the fi.rst quarten oftire l3th century (both fnom Tanalonrs papen)i(D) is the weak American lead with f]-atflanges and (C) is the domed variety whichgives stiffer" windows and was used by Lawr.enceSaint in Washington Cathednal- in J.934.

Mr Frederick Cole wnote fnom Canterburyto dr:aw attention to a difficulty in saving

Mr Dennis King, of Norwich, Englandr hasshown me an intenesting piece of glass fnomthe chur:ch at Norbury in Derbyshire. Theweathening pr.oduct is a panticula:rly hardcnust which makes the glass quite oPaque. Itoccurs on both sides of the glass and Mr Kingrscleaning treatments diffen.

0n the inside of the glass the ainbnasivei-s nequired to r:emove the hard crust betweenthe painted lines; the nesultant sunface ismatt and a coating of Viacnyl VC363 (handenedwlth Desmodur N75) was applied expenimentallyto about one half of this sunface to impnovethe translucency. There was no paint on theoutside of the glass and it is theneforer:endened transparent by the quicken and mone-effective method of grinding and polishing.

Fig.3 is a photognaph (at a magnificationof x5.5) taken from the outside (face No.l)through the gnound and polished surface and itshows the under"side of two painted lines at A.Note that tiffi are on iace No.2 but seenf::om face No.I. At A the painted lines can

the old oniginal leads in thein entinety. He

says rrWe have in the past tr:ied to pneservethe lead pattern but have abandoned this onaccount of the fnagile glass. The l-ead has to

, bg,c.ut in or"der to reLease the glass and, evenso, the glass is at nisk.tr. If there is aneasy way of nemoving the glass without cuttingthe lead he would Like to know it. Exactlythis point is made by Dn Eva Frodl-Kraft(Abstract No.202) who remar:ks that otd lead-work can only be bnoken by insenting a sha:rpinstrument between it and the g1ass, and it isimpossible to do this l^tithout causing breakagesand cracks. Mr Cole also nemarks that leadsof type B do not, in the long tenm, providesufficient "coveragett. Vibration by the windcauses the glass to ttcreeprr from the leads andhe pnefens the domed type C.

be seen quite clearly but at B and C the linesare r.Iot so clear because the conrosion of theglass has underrnined the paint. The paint hadappeared to be quite firmly fixed when seenfrom face No.2 but Fig.3 shows that it is insome dangen.

This technique of gr.inding and polishingface No.l coufd be used in nesea:rches onmethods ot rrfixing" paint.

Fig.4 shows faee 2 of the glass at aneven higher magnification (x11.3) at a pointwhere the Viacnyl coating has been applied.The dividing li.ne is at D and the area to theleft has been ainbnaded only. To the ::ight ofD the airbnaded surface has been treated withViacr:y1 and it can be seen that the sunfaceis now less neflective. The white lines atErE show whene some of the cnust stifl-remains close to the painted line and it canbe seen that the ainb::asive jet had success-fu11y been guided to within 0.3mm of thepainted tine by the skill of the openator (thescal-e at the top is in mil-limetres).

T

4 CORROSION OF GLASS AND THE PAINTED LINES

Fig. 3 This interesting photograph (x5.5) shows what painted lines can look like when viewed from theirunderside. This glass from Norbury Church has a low durability and both sides are badly crusted. Thecrust on the outside of the glass was removed by grinding and polishing and, as the glass is quitetransparent, it is possible to look through the near outer surface and examine the underside of thepainted lines. At A the paint is clearly visible but at B and C it is not easy to see because the corrosionof the glass has crept underneath the paint.

Fig. 4 This is an enlarged view ( x 1 1.3) of the inner surface of the glass at a point where an experimentalapplication of Viacryl resin had been made on the airbraded surface. The part to the left of D has had

its crust removed by airbrasion and it therefore appears matt. The part to the right of D was alsoairbraded but the surface has been coated with Viacryl; it is now dark and non-reflective (it is in facttranslucent). The white lines at E, E show where the operator did not remove the crust close to theline; he had skilfully airbraded to within 0.3 mm of the painted line, leaving this thin line of whitecrust. The scale at the top is in millimetres.

5 NEW ABSTRACTSr99. BTSHOP, F.L. and MowRxY, F.w. (1952)

t'Deter:ionation of the surface of sheet glasstt,Ceramic Bulletin 1952r 31, 13-15.

The authons give a brief neview of themechanism of the detenionation of the surfaceof sheet glass under atmosphenic conditionsand draw attention to the existence of a thinsurface layer of highen nefractive index thanthat of the inside which is modified' by thewater in the atmosphere, to attain a lowerref:ractive index. This layen is less thanl-0 nm thick.

They devised an accelerated weatheringtest in which moist air at 100"F (38-C) iscirculated over the glass, which is examinedat 60-hour intervals. The pai.r: of samplestouch at the top and are separated by 0.009inch (O.25mm) at the bottom (so that waterwould presumably collect between them - RGN).Other samples were kept in an r:nheated wane-house in order" to test the effects of packagingmateria.l-s. Examination of the weathe:red

":sur:faces under a microscope (xaOO) showed that144 hours in the accelenated test were equiva-lent to 216 days in the wa:rehousel muchsurface deterioration coufd be seen after 2

years in the wanehouse.

Packing of the glass in a highly-absorp-tive, non-alkal-ine papen hal-ved the deter:ior-ation after 2 years. A change of al-kalicontent from L4.27eo to 15.1-19o (alkaline earthsfrom 11.83 to 1i-.15eo) roughly doubled thedeterioration after 2 year:s. Packing theglass in wood fl-our increased the deterionationabout 3 times, compar:ed with packing in paPer.

2OO, ENGLE, Anita (1973a) 'r31000 yeansof glassmaking on the Phoenician Coast",Readings in Glass History, No.1, 1973r Pages1_aa

This article is not concerned withconservation nor with glass windows but it isof inter.est to anyone who wants to learn moreabout the first or:igins of glass, glazes andfaience. Arnong other intenesting points, theauthor makes the suggestion that the B-tonslab of glass found at Beit Shearin may inAD30O have been intended as a source of culLetfo:: other glassmakens.

2OL. FRETMANJa.W. (1974) rrEffect ofalcohiG onGF!:r-gation in glasst',J. Amer:. Ceram. Soc., 1974 !Z' 350-353.

The author was actua.l-ly concerned withdrilling holes in glass and he found that lesscracking occurred in some alcohols than 1nwater; in fact, the water-content of thealcohol was impontant. Water: causes cracks topropagate under certain conditions which mayhave a beaning on the spontaneous crackingthat sometimes occurs in medieval windows.

2o2. FRODL-KMrT, Eva (1975b) rrEinigeBemenffigen-?ffiEchaft und Handwer:k inden GlasgemHlde-Restaunierungtt (Some notes onthe science and practiee of :restoring stainedglass). Paper read at the gth Colloquium of

the Conpus Vitneanum on 9th September 1975;12 pages of typescript.

The authon discussed several questionsneferred to in News Letter No.15 because itcontains comments by Dr Frenzel (item 1.2) andalso abstracts of the papers given at theStockhol-m Conference (Abstracts 186-192), andshe stressed the close inter'-relationship whichis requined between the scientists and thecraftsmen because neither can be effectivewithout the other.

Despi-te the progness made in the last fewyears the scientific unde:rstanding of the prob-l-ems of consenvation is stil1 too :restricted,especially as each new nestonation situationpnesents problems which must be cons.ideredafresh. Anne Moncrieffrs Stockholm paper(Abstnact No.190) is held up as a model of theway in which the chanactenistics of eachrestoration technique are set out so thatanyone can under:stand them. Her approachshould be r:evised from time to time in thelight of conservational experience, inparticular that of avoiding altenations to theoriginal antistry; thus it is not the "mosteffective'r cleaning agent which should be usedbut the one which does the least damage to theglass (especially the painting on the glass),even though it may be harder to use and takelonger to do its cleaning. In fact, there isan urgent need fon a systematic criticalsummary, ItHeute, da sich auch der mit dem FachVertraute kaum mehr einen Ubenblick ilber diel-aufenden Forschungen verschaffen kann, isteine thematisch geondnete, kritische Zusammen-stellung ein dningendes Desidenatum.). (RGN -Penhaps the International Technical Committeecan undertake this as its first task!)

Craftsmanship It must not be forgotten thatthe craftsmants skill is all-impoutant. He

uses coarse bnushes to remove excess cementfr"om the outside of the window and might usethe air-brasive on glass-fibne bnushes in sucha way that the often hardly-discernible haLf-tones are harrned or otherwise. She commentsrathen pithily that few restorers are pr.eparedto have any doubts a.bout thein pensonal skilland sensitivity. (RGN - but no doubt thisapplies to scientists also! )

Cleaning She points out that water must bepresent for the deterionation of glass tooccur and then a::gues that rrdnytt methods ofcleaning are to be prefenred. (RGN - but I am

not convinced that this is necessar:ily sog ifthe cleaned glass can be kept dnyr then thewaten used in ttwetrr cleaning methods witl d::yout; if the glass cannot be kept dny -because it is exposed to the weather - then itwiII get wet an5rway. Neverthel-ess she isprobably right when she says that rrsolutionrlprocesses of cleaning may loosen paint whichis underrnined by co:rnosion products - see Fig.3of this News Letter:. ) She also claims that anymechanical damage encourages weathering.(netl - but I believe that this nequires furthenstudy because some airbraded surfaces seem tobe mone nesistant to weathering, at least whenaccelerated tests are used. )

Mending cr:acks J.M. Bettenbourgts neseancheshave shown (see Abstnact No.181B) that sili-cone adhesives are better than epoxy resins n

and any yellowing is likely to be insignificantprovided the joint is thin and the adhesive isremoved fr"om the face of the glass. (RGN - ifadhesives are accidentally allowed to set har:don the surface of the glass, it seems that theycan be removed with Green-labe1 Nitromors orwith citat L2-r2.)

Plating The authon claims that larninating isunsatisfactory because it cannot be neversedafter several years have elapsedr at least notwithout damaging loose paint. (RGN - but Isuppose that Dr Jacobi will dispute this, aswill the Dommeister at Cologne) (RGN - shequenies what happened at Canterbuny withdefective paintwonk, and the answer is thatMr Cole used to touch-uP defective paintwonkwhen the guide-Iines wene clean but he has nowceased doing this. ) Plating will be an idealmethod of protecting medieval glass providedthe edges can be pnoperly sealedl some hascentainly been well-sealed but there are caseswhere water has entened the cavity.

Leading She asks whether: painted glass isprese::ved better by using new leads or bynepairing the existing damaged leads, andconcLudes that o1d .l-ead-work can only beb:roken by inserting a shanp instr:ument betweenit and the glass;, that it is impossible to do

this without causing br:eakages and cr:acks; andthat it may ther:efor:e be betten to r:epair: theoLd lead (if the panel can again be prope:rlystabilised) than to replace the l-eads by newones. In Austria much oniginal lead-workstill nemains and it is much better to re-usethem. Regarding Mr Col-ets analysis of 12thcentury Canterbury leads, she points out thathigh tin-contents can be obtained if solderedjoints have been melted down;, Dn Bauer:(ozKD 1967 21 2OB) found that most of the ear'Iyleads contained less than 19o of tin but theamount coul-d occasionally be much higher,suggesting contamination by solder.

2O3, HAYWARD,,Jane,(tgZS) "Installationof medieval stained glass at The Cloiste::srt,Papen read at the 9th Colloquium of the ConpusVitreanum on 9th September 1975; 9 pages oftypeseript.

This paper dlscusses the adaptation ofthe medieval glass to the windows of themodern building, the trisothenmalrr protectiveglazing system used, and the means of provid-ing external pnotection in the CloistersMuseum with its extneme climatic conditions(temperatunes nanging fnom 0o to 3soc), highhumidity in the summer, ain pollution, etc.The study concenns the 14th C. St Leonhardglass from Austria, and the problems of adapt-ing the panels to the modern building,including the intnoduction of trace:ry lightsand the incorponation of new stained glassare ful1y discussed (s pages).

The isotherrnal glazing arrangement isnovel, the medieval glass being set in theoriginat glazing gnooves which were made muchwider and deeper to permit ventilation air tofloyr around the edges of the panels. Thegrdbvts were or:iginally 5/8 inch wide (rO mm)

but were widened and deepened to I$ inches(32 mn), and a tee-ba:r just above the sil1level also all-owed air to flow f:reely belowthe window, as was confir:med by smoke tests.Each panel was framed in stainless steel tostiffen it, and this also prevented lightpenetrating anound the widened and deepenedgroove in the stonework. The tracer:y lightswere supported in a differ:ent manner which isexplained fully.

The external protective glazing was itselfa double window, consisting of a 6mm externalsheet of ttPlexiglastt, d 9.5mm ain space and a3mm sheet of ilLexontt which absorbs ultraviolet light; the edges were sealed withsilicone. This assembly is set 3] inches(gOmm) outside the stained g1ass. The author:states that this Plexiglas system ". . . doesnot heat up like glasstt, but there must be amisunderstanding somewhene hene because herfigures show that the stained glass protectedin this way does not heat up as much as doesstained glass protected by an external glasssheet; hence the Plexiglas system mustitself heat up more than an external sheet ofglass wou1d. In direct sunlight on a cl-earday when^the temperatu::e of the outside air

^^ ^v^was 28.6-C, a stained glass window with anoutside protective glass reached 37.4oC, ora rise in temperature of 8.8 deg.C. A sj.milartemperature. was recorded on a window protectedby a single sheet of PlexigJ-as, but onepnotected by the new system showed a rise of1.65 deg.C, ie, reaching a temperature ofabout 30.2"C. This system was only 199o morecostly than ordinary storm glass.

2o4. NEWToN, R.G. (1975b) "The unusualweathering of one of Dn Jane Haywardrs sampledla note pnesented to the gth Colloquium of theCorpus Vitrearum on 8th Septembe:r 1975.

tnis lni"f note, and Fig.5o is intend.edto dnaw attention to an extra-ordinary type ofweathening which the author has seen only once,on a sample of blue glass from The Cloisters.His attention was drawn to this piece byDr J.C. Femazzini and he is indebted toDr Jane Haywand for the opportunity of dis-cussing it. As can be seen from Fig.S, thecomos:-on is characterised by deep nannow pitswhrch are surrounded by a wide shallow circleof corrosion. Superficially it is reminiscentof the cathodic comosion of stainless stee1,which is an electno-chemical phenomenon. Theauthor does not suggest that electro-chemicalphenomena are responsibLe for this type ofattack but he does ask the question whether wepay enough attention to the possibility ofsuch phenomena. Although glass does notconduct electricity readilyo it conductssufficiently for ion-selective electrodes (egpH metens) to function, and hundneds of yeansare available fon any small effects to mani-fest themseLves.

10

Fig. 5 This sample from The Cloisters Museum, New York, shows an exceptional type of weathering whichis discussed in Abstract No. 204. The small deep pits and their surrounding flat areas of corrosion are

reminiscent of the cathodic corrosion of stainless steel.

2O5. NEWION, R.G. (1975c)'rUsing tr:i-angultt diagrams to undenstand the behaviourof medieval glassesrrr a PaPer Presented to the9th Colloguium of the Corpus Vitrearum on BthSeptember 1975.

The authon exPtained how the apparentlycomplicated chemical analysis of medievalg.l-asses can be simplified so that only threeitems need be considered. These three itemsare essential charactenistics of the glass andlangely detennine its durability. As can beseen fron Fig.6, glasses with good dunabilityfal1 nea:r the middle of the ttriangle (A,M,R)and the least durable glasses (H,JrQ) 1ietowards the bottom right-hand corner. Glasseswhich 1ie to the l-eft of the centre also havepoon durability because they "weept' (W) orcnizzle (Z). A similar diagram on a largenscale also helps to clarify the different kindsof weathering encountered among medievalglasses.

206. N.EWTON, R. c. (1975d) I'Experimentalstudies of the protection of medievaJ- windowsusing external glazingrr, a papen presented tothe 9th Colloquium of the Conpus Vitneanum on9th September 1975.

This account of the results obtained fnomthe instnumented window at Youk Minster isexactly the same as pages 4-10 of News LetterNo. 16.

&'

Fig. 6 This triangular representation of glass compositions helpsto characterise them according to their durability. Thosewith good durability (A = modern window glass; M = Saxonglass from Monkswearmouth; R = durable Roman glass)

fall near the centre. The least durable stained glasses(H = Austrian crusted glass; J = crusted glass from Elycathedral; O = poorly durable 1 2th C. green glass from YorkMinster) lie towards the bottom right-hand corner. Glasseswhich lie to the left of centre also have poor durability;those at W 'weep'under museum conditions and those atZlorm crizzles on the surface. See Abstract No. 205.

l1

207, RTTTER, ,lohn E., Jr. (1973) ilStness

connosion susceptibility of polymenic coated,soda-l-ime glassrr, University of MassachusettsReport No.UM-73-5, March 1973.

The author studied tne effectiveness ofresin coatings (acnylico epoxy, and silicone)in pneventing stress-cor.rosion of glass bywater. The coatings limited the availabilityof waten to the sunface of the glass to some

extent but they did not eliminate it entinelybecause stress connosion occunned. Althoughnone of the resins entirely prevented attackby water it may be of interest to know whatthey were. The acrylic coating was preparedfnom Lucite Tnansparent Molding Powder L-149(fr:om Fisher: Scientific Co.r Pittsbungo Pa.);the epoxy coating was made from Conapoxy PA-I22and Conacure EA-011 dilqted with ConaP S-Lthinner and cured at 6O'C for 3 hour:s (a11from Conap Cor:p., Allegany, N.Y.);^ and thesil-icone was PanShield cured at 60-C fon3 houns (from Dow Corning, Midland, Mich. ).

., 208. RoDERICX, Hilli_ (1975) "Pnojectedemiss$i or@ fuel combustionin the 0.E.C.D. area 1972-1985: tne inter-national issues with special neference totransfrontier migration of pollution", a Paperlp::esented to the IPIECA Symposium held inTehr"an in April 1975. 36 pages of typescript.

Dr: Roderick is the Director of Environ-mentaJ- Affairs, OECD, Paris and his paper isnot directly concerned with conservation butit will be of interest to everyone who wantsto know what changes in air pollution aneexpected in the future. After making a numberof assumptions about the changes in tne demandfor" energy, the cost of nemoving sulphur f::omthe fueJ- or from the waste gases, and theprices of oil. and coal-n including the noreexpensive lcw-su1phur" grades he concl-udes thatsulphur emissions are expected to decreaseuntil l-980 after which they may ri.se again ifa greater propo:rtion of solid fuef is used.The major contrj.bution to 1ocal air pollutionis by househol-ds and residential areas' notmajon industny. The harmful effects ofsulpnur oxides (to humans) is believed to belinxed to the plresence of panticul-ate matter(particula::1y sulphates) in the atmosphe::e.

2oe. sPlrzER-1!!!!ff\_$g!$ (1s75a)"EtudE-de v@ b l'aide demicroscope optigue, microscope I balayage avecimage par 6lectrons r6t:rodiffus6s et micro-sonde 6l-ectronique a nayons X" (A study ofmedieval red window glass using optical micr.o-scopy, the scanning electron micnoscope withback-scatter:ed electrons, and electron micro-probe analysis), Ver.res et Refractaires l-97529 l_45-153.

Dn Spitzer-Anonson is well-known for hercareful studies of the nature of medievalcopper-ruby glasses. Her papens are generallynot involved with pnoblems of conservation butthis onen and No.2l1, have a beari-ng on thetechniques used fon making these early glassesand hence perhaps with their conservation.

The present papen shows that the earliestrubies wene mul-tilaye:red through their thick-ness but in the l-4th century these inhomo-geneous nuby glasses were rrflashed" on to a

., eetrourless support glass (and hence deeppitting on the flashed side will lead to thefo:rmation of white spots). Later still themodenn type of fl-ashed ruby came into exist-ence, whene the ned layer is thin and homo-geneous, and is sandwiched between two colour-less glasses, one thick and one veny thin. Inthe 15th to 17th centuries thene we::e rrfalse-flashedrt ned glasses in which there is noclean boundany between the red and whitegLasses.

A fulI desc:ription is given of thesophisticated experimental techniques whichshe used in making these studies, and inexploring the copper contents, mic:ron. bymic:ron, :right acnoss the thickness. Aninteresting point is hen conclusion that somediffusion of copper occurs from a colou:redregion into a col-ourLess one, for about 50 ym

each side of the boundary, despite the abruptcessation of colour at that boundar:y. (RGN -this conclusion and also her finding that thecolourless glass inside "hair:pintt folds of redcol-oun can be richer in copper" than the redg1ass, seems to support my long-held view thatthese glasses wene not made by ::epeated dippingin alternate rrredtt and ttwhitetr melts but abadly-stir:red mixed melt was used, some ofwhich ttstrucktt to give a ::ed colour and somedid not. The glass is so viscous that repeateddipping woul-d not give such thin laye::s, with-out some rtstretchingtt also taking place; Ihope to anrange some practical trials on thispoint. )

2LO. SPITZEWIjSoN, Muothu (197sb)"oiffFion ffie lIarsenicdans 1es vitraux rouges m6di6vaux. f,tudequantitative de la concentration du cuivre pourdes tplaquest de techniques diff6rentesrl(Selective diffusion of coppen and ansenic inmed.ieval- r:ed glasses. A quantitative study ofthe concentnation of copper in flashed glassesmade by different techniques), C.R. Acad.Sc.Panis, 280 (9 June 1975) S6rie C - 1343 to1346.

This paper does not have a bearing onconservation but she found rel-atively strongconcentrations of arsenic in some places. Thenegions where the ratio of arsenic to copperhad a maximum value were black zones wher:emetallic copper had been precipitated. Thepaper will be of interest to those who areconce::ned with the compositions of medievalglasses.

2II. (1975c)"contlFbutiffis vitraux duMoyen-Age. La pr6sence du plomb et du cuiv::eet leur diffusion s6lective dans les vitrauxrouges des cathddrales frangaisesrr (A contri-bution to the undenstanding of medievalglasses. The presence of lead and copper andtheir selective diffusion in French cathednalr"uby glasses)o C.R. Acad.Sc. Panis 280. (August1975) S6nie C - 2O7 to 209.

SPITZER-ARONSON

12

This papen is again pnima:ri1y of intenestfon its chemical content. The distribution ofcopper and'1ead (in l6th C. glasses) wasstudied using sophi.sticated techniques and itwas found that coppen played a protective nolein preventing the neduction of lead oxide byansenicg if the coppel: content was neduced bydiffusion, pnecipitation of metallic leadoccunned with consequent blackening of theglass.

2I2. WIEDERHORN' S.M. (1967) 'rlnfluenceof waTS valifri-ffipropagation in soda-lime glasst', ,J. Amen. Cenam. Soc. 1967 50

407-414.

The author used a novel experimentaltechnique to measure the nate of cr:ack growthin glass, as it was affected by the moistunecontent of the envinonment and the stress inthe gIass. Waten vapoun had a pnofound effecton the nate of crack growth, 10OA humidityproducing about 101000 times the speed ofgnowth which occu:Fed with O.0U9o nelativehumidity. The r:ate also depended on theapplied stness; doubling of stness producedan increase of about 11000 tirnes in the crackgrowth rate. The fastest nates obsenved wereabout 0.Smn/sec and the slowest was aboutO.Ol um/sec (about O.lmm per houn). (RGN -even thi.s speed must be vastly greaten than thenate of spontaneous cnack gnowth in medievalglass and it seens that we have to look forsome additional explanation, such as extnenelylow applied stresses in one face only. )

2L3. WIEDERHORN

57 336-341.

" This is a highly technical paper in which' the nates of cnack gnowth in a vacuum were

studied for^six glasses at temPeratunes from25oc to 775ec. The intenest of the papen fonspecialists lies in the fact that crack growthin a vacuum occunned in sorne glasses but not inothens, and theo::ies of elastic behaviour areused to explain the diffenences. It is hopedto intenest one of these expents in thespontaneous fnacture of some medieval glasses.

.r-J.&lJ..qIr.

NOTE: WilI neaders of these News Lettens!-f"asu draw my attention to any PaPers whichshould be abstnacted hene. It would bepanticulanly helpful if photocopies of thepapens could be supplied. My address is5, Handwick Cnescentn Sheffield, S1l SWBt

England.

NOTES

JOHNSON' H. '

13