The non-conservation of a reverse glass print

Identification, analysis and preservation


In 1774 the Duke of Devonshire lent his collection of 200 drawings by Claude le Lorrain to the publisher John Boydell for the purpose of having them printed. Boydell commissioned the master printmaker Richard Earlom to complete the series. To recreate Lorrain’s works, Earlom started with an etching, then used a mezzotint roulette to create areas of wash tones with varying intensities of darkness and light. The plates were then printed in a sepia ink, similar to the bistre of Lorrain’s original drawings. The completed series was entitled ‘Liber Veritatis’ when published from 1783-1822 [1].

What was thought to be one of these mezzotints was brought in to the Melbourne University Conservation Centre by a private collector of Claude le Lorrain prints, who requested that the print be removed from its glass cover. To the eye, the object did appear to be a highly discoloured mezzotint. However when it was removed from the frame, it was clear it did not simply consist of a flat artwork on paper. 

Fig.1 Reverse Glass Print/ recto / BT. Photographer: Daisy Todd © 17/02/2014

Fig.1 Reverse glass print/ recto / BT. Photographer: Daisy Todd © 17/02/2014

Reverse glass printing is a process whereby a print is transferred to a sheet of glass, varnished and coloured to resemble an oil painting:

  1. The print is first wetted out
  2. A piece of broad sheet glass is coated with varnish
  3. The dampened print is then laid face down onto the glass and allowed to dry
  4. The main bulk of the paper substrate is then removed from the verso of the print by rolling with the fingers, leaving a thin layer of paper and the inked image on the glass
  5. The verso is then varnished using resins of natural origin, such as Venetian turpentine, to give the print a brilliant transparent effect
  6. Finally, the verso of the glass print is coloured with paint media, most commonly oil [2, 3, 4, 5]
Fig.2 The stratified structure of a reverse glass print/ verso / BT. Photographer: Daisy Todd © 17/02/2014

Fig.2 The stratified structure of a reverse glass print/ verso / BT. Photographer: Daisy Todd © 17/02/2014

Reverse glass print manufacture came into fashion in England at the beginning of the 18th century. Mezzotints have a rich tonality, giving a dynamic, ‘painterly’ appearance to the printed artwork. However, during the 17th century, the monochrome mezzotints were becoming monotonous, resulting in an increasing call for colour. This increasing demand for coloured artwork led to the novel production, by professional printmakers and amateurs alike, of the reverse glass print. This ‘progression’ soon went out of fashion when colour printing was introduced in the mid-19th century [4, 5].

Visual Observation and Identification

Reverse glass prints are not widely known to curators and conservators, due to the relatively limited historical record of their manufacture and the fact that relatively few have survived, given their fragile nature. Consequently, there is very little written about their creation and even less about their conservation. Despite there being many articles on related topics such as the conservation of glass, glass-plate negatives, Hinterglasmalerei and varnished papers, there are only a few insightful articles that deal directly with reverse glass prints – the most notable being two by Tremain, published by the Canadian Conservation Institute (CCI) in 1987 and 1988 [4, 5].

Visual observation of the reverse glass print with different light sources showed the object’s stratified structure, the primary support being glass. On top of the glass was a layer of varnish, then a very thin layer of paper, with irregular consistency and scattered lacunae, a second layer of varnish and then a thick application of paint. The visual aesthetic was severely compromised by the distortions, crevices and minor air pockets between the glass and the thin paper substrate caused by the discoloured, embrittled and acidic varnish. The discolouration of the varnish meant that the colour of the paint on its verso was not visible from the recto, affecting the appreciation of the artwork. As such, to a complete newcomer to reverse glass prints, the application of the paint was baffling. Even more baffling were the random brushstrokes, unappealing colour usage and almost careless application of the paint itself.

It must be stated that, in order to avoid hair-raising interventive treatments that may result from these basic first impressions – for example, the separation of the print from the glass, or the removal of the varnish – it is imperative that reverse glass prints are accurately identified prior to their treatment. 

Fig.3 Reverse glass print/ recto / transmitted light/ BT. Photographer: Daisy Todd © 17/02/2014

Fig.4 Reverse glass print/ recto / transmitted light – area of loss/ BT. Photographer: Daisy Todd © 17/02/2014

Materials analysis was undertaken to gain more information on the materials constituting the object. The reverse glass print’s multi-layer structure (fig.2) lent itself to microscopical cross-sectioning. A micro-sample was removed from a discreet location, embedded in resin and examined under magnification (x100) using a Möller-Wedel ® reflected-light stereo-binocular microscope [6]. Photo-microscopy of the cross-section under transmitted light enabled the observation of the sequence, colour and thickness of the compositional layers.

Bildschirmfoto 2015-04-29 um 15.45.20

Fig.5 Reverse glass print/ photomicrograph-cross-section_TL_x100. Photographer: Daisy Todd © 18/03/2014

Further photo-microscopic UV analysis showed the characteristic autofluorescence emission of natural binding materials and oxidised drying oils [7, 8].

Fig.6 Reverse glass print/ photomicrograph-cross-section_UV_x100. Photographer: Daisy Todd © 18/03/2014

 The embedded sample was then stained with Rhodamine B 0.02%, a lipid-soluble dye. The carrier successfully solubilised the dye into the lipid-containing surface, the characteristic red and red-orange colours confirming the presence of drying oils and lipidic fractions within the sample [7].

Fig.7 Reverse Glass Print/ Photomicrograph-cross-section_Stained_x100. Photographer: Daisy Todd © 19/03/2014

Fig.7 Reverse glass print/ photomicrograph-cross-section_Stained_x100. Photographer: Daisy Todd © 19/03/2014

Further non-destructive componential analysis was carried out using Fourier Transform Infrared spectrometry (FTIR) with a diamond attenuated total reflection (ATR) window. Two micro-samples were extracted from discreet areas and mechanically cleaned under stereo-binocular magnification (x40) to remove residual material layers that might complicate the FTIR-ATR process. Each sample was placed in the centre of the diamond window of the Bruker ® Alpha-P FTIR. The IR spectra were recorded by 32 scans at a spectral resolution of 4 cm-1 in the spectral range of 4000–400 cm-1. The data were rapidly collected and processed by OPUS software to enable the visualisation and exploration of the data structure.

The diagnostic bands obtained were comparatively evaluated against those of the range of terpenic resins, including Venice turpentine, Manila copal, dammar and mastic, published by Derrick et al [9]. Spectroscopic analysis of control samples of these selected resins was then carried out. The spectra were characterised by typical absorption bands, according to their molecular functional groups, and visually compared against the unknown varnish spectra to confirm their spectral correlation.

Fig.8 Overlaid FTIR spectra obtained from a control sample of mastic resin and sample (1) of the unknown varnish. Daisy Todd © 20/03/14

Fig.8 Overlaid FTIR spectra obtained from a control sample of mastic resin and sample (1) of the unknown varnish. Daisy Todd © 20/03/14

Table 1 The spectral data set of the organic resins

Bildschirmfoto 2015-04-29 um 13.16.55

The spectral data sets displayed the multivariate scores of the principal components, where each wavenumber is a measured variable. The spectra of the natural resins (table 1) showed a broad band in the 3000 cm-1 region due to the stretching of OH groups, the methylic and methylenic groups showing two sharp, strong absorptions in the ranges 2960–2930cm-1 and 2875–2865cm-1, while the absorption of C=O and C-O groups occurs respectively at 1240cm-1 (weak) and 1715–1695cm-1 (strong).

Samples 1 and 2 and the control mastic resin sample had correlating fingerprint net peaks, suggesting the materials are homogenous [10, 11]. These analogous results, as well as Tremain [4, 5] and Stanley’s [12] evidence of mastic varnishes being commonly used for reverse glass print manufacture, suggest that a mastic varnish was utilised on the glass print.

Materials Characterisation

Mastic is a natural triterpenoid resin long used in varnishes. Its aesthetic function, when applied to the reverse glass print, was to create a ‘brilliant translucency’, allowing the verso colouring to show through – in effect creating the illusion of a coloured print, pre-colour printing [12]. However, its optical functions had dissipated with age, the varnish having undergone an auto-oxidative radical chain reaction, leading to peroxides, which, by thermal and/or photochemical inducement, caused homolytic cleavage, auto-catalysing new radicals.

Photo-oxidation of the mastic film explained the low pH and the low autofluorescence that occurred during visual examination: the oxidative production of acid groups leads to the destruction of mastic’s carbonyl species, which are responsible for its UV absorption and high autofluorescence. These degradation pathways are evidenced by other familiar phenomena including embrittlement, hazing, yellowing, loss of gloss, air pockets and a change in polarity, fluorescence and solubility of the mastic layers. Mastic’s low thermal expansion rates and characteristic retention of solvent (commonly turpentine) were also considered to be the likely cause of the planar distortions visually obstructing the printed image [6, 13, 14].

Treatment Record

The glass was cleaned using deionised water and the object was re-housed in its original frame with the addition of an alkaline buffered backing board. Otherwise, no treatment was carried out. The UK’s Institute of Conservation (Icon) Code of Conduct provided a conceptual framework from which to establish this literally ‘non-interventive’ treatment.

Although the object’s visual aesthetic did not improve, when intrinsic value is considered, the object’s non-treatment was the only viable treatment option [15]. In consideration of the relative rarity of reverse glass prints and good condition that this object is still in, given the susceptibility of glass to physical damage, it was decided that a fully accomplished treatment might only be ensured through the discouragement of varnish maturation processes, through the systematic implementation of direct and indirect preservation techniques [16, 17, 18].

The identification of mastic as the unknown varnish samples enabled further research of the material properties of mastic, deteriorative mechanisms, stability and performance. This aided the development of appropriate environmental recommendations and preventive measures to ensure the long-term stability, and increase the longevity, of the object [19, 20, 21].

The owner of the print, who had requested that the print be removed from the glass, had had no previous knowledge that the print was, in fact, a reverse glass print. However when the true nature of the object was revealed, he was excited and gratified to have added to to the object’s provenance.

Daisy Todd is a graduate of Conservation and Restoration (BA Hons) at Lincoln University (2012), and a graduate of the Conservation of Cultural Materials Masters Course (MA) at Melbourne University (2014). She is currently an intern at the British Library. Previous work and internships include The Smithsonian National Postal Museum, The Victoria State Library and The Acropolis Monuments. Daisy is a conservator of paper-based artefacts, with a specialism in the conservation of philatelic items and artwork on paper, and is currently concentrating on the conservation of parchment, books and photographs. 

Her past critical research projects have involved the identification of organic materials through the use of high-performance analytical techniques, with a primary focus on the characterisation of organic exudates used in adhesives and binders on artworks on paper.


  1. Chisholm, H. 1911. Encyclopaedia Britannica Dictionary. A.S.L.G.I. 11th ed. vol. 33. Cambridge: Cambridge University Press.
  2. Gascoigne, B. 2004. How to identify prints. 2nd ed. London: Thames & Hudson.
  3. Lowenthal, L. 1931. Old English pictures on glass. The Antique Collector. June/July, pp112-113.
  4. Tremain, D. 1987. Two eighteenth century glass prints. CCI Newsletter, Ottawa: CCI.
  5. Tremain, D. 1988. Reverse glass prints: their history, techniques and conservation. In: Burgess, HD. ed. Conservation of Historic and Artistic Works on Paper, Proceedings of a Conference – Symposium 88 Ottawa. Ottawa: Canadian Conservation Institute, 1994, pp 143-152.
  6. Rivers, S. & Umney, N. 2003. Conservation of furniture, Routledge, London, pp390 & 590.
  7. Wolbers, R. 2000. Cleaning painted surfaces: Aqueous methods.  London: Archetype Publications pp169-190.
  8. De la Rie, ER. 1982. Fluorescence of paint and varnish layers. Studies in Conservation, 27, (1). pp1-7.
  9. Derrick, MR, Stulik, D & Landry, JM. 1999. Infrared spectroscopy in conservation science. Los Angeles: The Getty Conservation Institute.
  10. Katon, JE. 1996. Infrared microspectroscopy: A review of fundamentals and applications’ Micron 27 (5), pp303-314.
  11. Nel, P, Lonetti, C, Lau, D, Tam, K, Sagona, A & Sloggett, RS. 2010, ‘Analysis of adhesives used on the Melbourne University Cypriot pottery collection using a portable FTIR-ATR analyzer,’ Vibrational Spectroscopy, vol.53, pp64-70.
  12. Stanley, T. 2006. Short communication mezzotint under glass: A historical review of the glass print. Journal of The American Institute for Conservation of Historic & Artistic Works, 45, (2), pp 147-154.
  13. Colombini, M & Modugno, F. 2009. Organic mass spectrometry in art and archaeology, London: John Wiley & Sons Ltd.
  14. Feller, R & Wilt, M. 1990. Evaluation of cellulose ethers for conservation. The Getty Conservation Institute [Online]. Available from: <> [viewed 30th March 2014].
  15. Munoz Vinas, S. 2005. Contemporary theory of conservation, Oxford: Butterworth-Heineman.
  16. AIC. 2013. Caring for your treasures, [Online]. Available from <> [viewed 2nd April 2014].
  17. Daniels, V. 1999. Imperfect reversibility in paper conservation. In: Oddy A & Carrol S. eds. Reversibility- Does it exist? British Museum Occasional paper No.135. London: British Museum Press.
  18. Oddy, A & Carroll, S. 1999. Reversibility- Does it exist? In: Oddy A & Carrol S. eds. British Museum Occasional paper No.135. London: British Museum Press.
  19. Mecklenburg, MF & Tumosa, CS. 1991. Mechanical behavior of paintings subjected to changes in temperature and relative humidity. In: MF Mecklenburg ed. Art in Transit, Studies in the Transport of Paintings, Washington, D.C.: National Gallery of Art. pp 173-216.
  20. Koob, SP. 1982. The Instability of Cellulose Nitrate Adhesives, The Conservator, 6, pp31-34.
  21. Ogden, S. 2007. The environment. The Northeast Document Conservation Center .[Online] Available from: <> [viewed 2nd April 2014].
  22. Omecinsky, D & Carriveau, GW. 1982. IR investigation of resinous and synthetic varnishes. AIC 10th annual meeting. Preprints. Milwaukee: AIC. pp141-149.

6 thoughts on “The non-conservation of a reverse glass print

  1. Hi I have a john boydell print on glass it has all the signs of reversed glass original it is a view of greenwhich hospital no 16 1751 is it worth anything thankyou


  2. Hi, thanks for the great article.
    I recently worked on a four reverse glass landscape artworks; a pretty unique group of works that I’ve not otherwise seen anything comparable to. The main features of which were:
    – presumed to be Japanese export art featuring Japanese and Roman character text.
    – media applied directly to the glass
    – grit and chaff mixed in with the media
    – fabric weave texture to some areas of media
    – silver albumin print cut outs to depict people
    – crushed mother-of-pearl used as a pigment

    Similarly, treatment was limited to cleaning the front of the glass and some renewal of backing materials to adjust show through.


  3. Thank you for this article, an informative and thankfully even humorous read indeed. Having recently happened on two reverse glass prints, I in fact had the kind of stumped reaction you describe to the pasty paint job on the verso of the prints. Having printed myself, there were no doubts as to the crosshatching I could clearly see through the glass, but I could still see it was not a work on paper (any longer). While I imagined something of the sort, I had never heard of or seen this technique applied, so it was an unexpected surprise to learn of it.
    I’m happy to see that the prints I have are in what appears to be in near optimum condition, neither showing deterioration, disadheasion or lacunae of note. Their only manco us the deep darkening of the varnish that make admiring the image a bit of a strain, but with all lines and colors fully present and visible.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.