Conservation of pastels

HOME

ABBREVIATIONS

ARTISTS

AUTHOR

BIBLIOGRAPHY

COLLECTORS

EXHIBITIONS

GENEALOGIES

MUSEUMS

SITTERS

SUBJECTS

Dictionary of pastellists before 1800

CONSERVATION OF PASTELS

This note provides informal practical guidance for owners of pre-1800 pastels and discusses many of the questions which commonly arise. The traditional aversion to holding pastels among most museums is well known, although many of the issues are manageable when properly understood. The evidence of the extent of these hazards appears contradictory. Many in the art trade will report that pastels are routinely handled and shipped without any visible loss, and some doubt that museums' concerns are well founded. On the other hand, as any visit to the salerooms will confirm, a great many pastels today show extensive, irreparable losses, and only a minority preserve the full delicacy of the original surfaces. Those however have the huge attraction that (unlike oil paintings where varnish must be renewed every generation or so) they will have been untouched since they left the artist's studio.

Construction Ideal display/storage conditions Shock and vibration The spacer Glass Frames Transport and loans Photography Further reading

Construction

Pastels are almost invariably framed, often in their original frames, frequently with the original glass and occasionally with backing boards that have never been unsealed. These are all part of the work and should be left intact unless there are overriding considerations.

The basic construction of most of these works is similar (see simplified diagram): canvas is fixed to a strainer (a rectangular wooden frame that today is called a stretcher, as they are now fitted with wooden wedges at the corners which can be driven in to restore tension in the canvas: but this is very rare in eighteenth-century pastels). Most strainers are made of pine or spruce: they can be very roughly cut or quite highly finished; some have cross-bars at the corners, while larger ones may have central bars. Ovals, except for the smallest, usually have cross bars. Joints may be mitred or butt.

The paper onto which the pastel is applied is first pasted ("marouflé" is the French term) onto the canvas, before the pastel is commenced. The whole surface is typically covered with pigment, applied dry. When finished the work is placed in a frame: ideally the rebate should be large enough for the whole strainer to fit inside (it will often need to be extended or commonly the backing is held in place by angled brackets), but first a sheet of glass is placed next to the sight edge, then a wooden spacer or fillet is cut to lie within the rebate. The pastel sits behind the spacer, isolated by acid-free lining paper or thin card; and finally a backing board is fitted and closed up. Lining tape (not shown in the figure) or other arrangements are made to ensure the pastel is protected from dust and insect infestation (the adhesion of lining tape can fail over long periods, and should be relied upon neither for mechanical support nor dust exclusion).

There are numerous variations of this configuration, including microclimate assemblies; simpler structures in which thin battens are fixed to the sides of the strainer, projecting beyond its front surface but onto which the glass is taped; a variant of this known as the Lepeltier box (an L-shaped wooden moulding incorporates the spacer with outer wrap for the strainer: once favoured by the Louvre, but now rejected as they involve enlarging the rebates in original frames).

Some pastels - especially pre-1700 or those of smaller dimensions - are not mounted on canvas and strainers, but may be pasted on board or even made on loose sheets of paper, like conventional drawings. Some pastels are made on parchment sheets fixed onto strainers without an additional canvas support. Parchment's weight and uneven thickness can cause particular problems. Pastels have also been made on supports as varied as silk, wooden panels and even copper plates. Even mid-sized pastel are frequently made on several sheets of paper joined carefully (normally this was simply because large sheets were not available, but some artists, La Tour among them, made a practice of drawing heads on separate sheets which were later pasted in). Such sheets are more liable to come adrift than single sheets wrapped around the edges of the canvas/strainer, but can sometimes be carefully refixed to the support. (Where separate sheets become completely detached and pass over the surface of other sheets of the pastel, irreparable damage can occur.) Joined sheets of parchment, without a support on which to rest, are far trickier: they are held together only by paste along the overlap, and can very easily separate with movement from transport or change in atmospheric conditions.

In what follows we discuss the conventional pastel made on paper pasted onto canvas fitted over a wooden strainer.

Ideal display/storage conditions

Pastels are intended to be hung on walls like other paintings. If stored, they should still be kept vertical, on walls or fixed shelving (not on sliding racks which are pulled out). Light levels for any works on paper should be controlled, although they are normally far less of a problem with pastels than with drawings and watercolours where the paper is exposed and can be bleached and become brittle with exposure to too much light. No work of art should ever be in direct sunlight. Most pigments in eighteenth century pastels are light fast: the exceptions are the lakes (particularly the red favoured by artists from Liotard to Cotes, and the yellow often mixed to produce green, thereby leaving so much foliage a naked blue) which have often faded from crimson to a pale pink or grey (but the same effect is found in Reynolds paintings).

Humidity levels are important: an RH level in the range 40-50% is ideal, but rapid variations and extremes are the main enemy. In an English climate with central heating, you will need humidifiers in winter (whenever the external temperature is close below say 5° this becomes really important) and dehumidifiers in summer (particularly with wet weather). Basements should generally be avoided. Mould readily grows on pastels (it often shows a preference for black areas), but it is one of the few interventions that can satisfactorily be made by competent paper restorers. Flood damage, in contrast, is almost impossible to repair satisfactorily.

Shock and vibration

The contemporary literature demonstrates a constant hunt for methods of fixing pastel, an impossible quest since successful fixatives alter the optical properties of the medium. It is safest to assume that pre-1800 pastels have not been fixed and remain vulnerable to damage from shock and vibration. Shock is a sudden force in a single direction, while vibration involves the transfer of energy from a cyclical source over a longer period. There is no certifiably safe level of either: shock levels below 0.5g are thought not to present any immediate hazard, but this is a level likely to be exceeded when they are moved with all but the most careful handling.

Shock and vibration can occur most obviously during transportation. During conservation screw fixings should be used; pins may not be hammered, nor may staples or nail-guns be applied. Shock and vibration can also occur in situ, from nearby building work or even traffic. When in 1912 a new omnibus service was introduced in the faubourg Saint-Honoré, Henri de Rothschild's famous La Tour pastel of Duval de l'Épinoy was sufficiently jeopardised that he built a new house in a quieter neighbourhood.

It is natural to assume that the effects of shock and vibration will be immediately apparent, as debonded pigment will fall from the surface and, even if blank areas of the paper are not immediately visible, the pigment dust should be seen lying on the lower fillet. Unfortunately this isn't a reliable indicator: the outcome may not be a neat binary alternative of damage/no damage. Pigment can be loosened without falling immediately, and the nature of the adhesion (at micro- and macro-scopic levels) is highly complicated. Although this is in area in which research is progressing (notably by a group at the Rijksmuseum), one theory likens the adhesion of pastel subjected to vibration to the behaviour of metal in aeroplanes, where nothing is seen until failure occurs after a certain number of cycles. Metal fatigue produces catastrophic results: whether the same phenomenon occurs with pastel has yet to be proved, but few will want their pastels to be tested. The theory could explain the apparent contradiction between evidence of the effect of single journeys and the observed condition of many pastels today. Furthermore the variety of techniques employed by different artists makes inference from samples unreliable (although it is clear that some artists' work was more fragile than others). It is inevitably easier to demonstrate that there is a risk than to prove that there isn't.

Minor in situ shock and vibration can be mitigated to some extent (my experiments suggest an attenuation of about 60%) by the simple expedient of placing a small (2.5 cm) cube of soft plastic sponge (ideally Plastazote LD33) between the wall and the bottom of the frame. Vibration can still travel down the picture chain, but two degrees of freedom are damped; shock from neighbours' building work etc. is most likely to be in a horizontal direction. The sponge also helps keep air circulating round the work and avoid humidity being trapped; this is particularly relevant if pastels are hung on external walls (if so, monitor temperature differentials to ensure there is no risk of condensation forming on the glass).

The spacer

One of the most important conservation considerations (and the one most often ignored) is the spacing between the surface of the pastel and the inner face of the glass: there needs to be a clear gap over the full area of the work, normally achieved by having a spacer or fillet (ideally of at least 6 mm) all the way round the border. This will normally be hidden behind the rebate, and the side edge blackened so that it can be hard to see. A simple way to measure the gap without disassembling the work is to use a torch and observe the shadow of a particle of dust, an imperfection in the glass or a piece of paper placed on the front of the glass: moving the torch beam from 90° to 45° incidence will trace a line on the surface equal to the distance from the surface to the spot. Allow for the thickness of the glass (typically 2 mm). Inadequate spacing is a common problem (not only in historical assemblies: most framers today need special instruction) and can lead to serious problems when pastel surfaces meet glass: the danger is during transport, not only where careless handling pushes through the backing but also, particularly with larger works, where the support has lost tension, and movement occurs. It can also occur with increased humidity when paper expands. If the humidity is so high that moisture condenses on the inside of the glass, the droplets may also bridge the physical gap.

The spacer should be made from a low-density wood such as obeche with little risk of resin seepage (museum mounting board can be used, but will need to be laminated to achieve the necessary depth). It should be cut and mitred to fit the rebate so that it is not in danger of sliding across the pastel if lining tape adhesive dries out. It was common for spacers to be pinned in place, but pins often rust and can cause problems. Ovals present special problems. In older assemblies the spacing devices may be blocks of cork placed at intervals or even pieces of thick cord.

Glass

My personal view is that original glass should always be preserved wherever possible; minor blemishes should be regarded as part of the work's charm. Several difficulties can arise, some of which (such as breakage) require the glass to be replaced. Old glass can sometimes grow mould which causes a fogging effect that cannot be reversed (this is a fungal infestation from the Aspergillus genus or similar air-borne microorganisms). Hand-made sheets are rarely perfectly flat, but neither are frame rebates, and fillets may need to be sculpted to ensure a closer fit and proper, even support for the glass to avoid additional breakage hazard on refixing and transport. Hand-cut oval sheets of glass are especially at risk as the jagged edges can result in a single point taking all the pressure during transport if lining tape has lost its adhesion; this risk is compounded as the mechanical strength of oval frames is compromised by their construction (see below).

When replacement glass is required, I favour normal 2 mm picture glass. There is a wide range of newer glazing materials on the market, said to have benefits ranging from ultraviolet light reduction to being shatter-proof. Some of these claims need to be examined carefully. Apart from price, the deficiencies of such products can be serious. Many are not colour-neutral. Some are very much heavier than normal picture glass, with the consequence that frames can be compromised and the weight of the refitted picture can result in more severe shocks during transportation.

By far the biggest concern lies in the electrostatic properties of some of these new materials. Pastel is vulnerable to lifting by low levels of electrostatic forces. This can be seen when tape is used to protect glass from breakage during transport: when the tape is peeled off, particles of pigment can often been seen jumping from the surface of the picture onto the inner surface of the glass (this is another reason why deeper fillets should be used than framers often think: a 6 mm gap is an order of magnitude better than 2 mm since the force obeys an inverse square law). Newer plastic glass can do this without the tape. While some very expensive products are marketed as "low static" the phenomenon is not eliminated. Problems can arise even when cleaning the glass with a duster. It has been suggested that this risk can be managed by using an ionizing air blower (a machine developed for use in the manufacture of electronics components): this may well be the best solution in a restoration studio, but is unlikely to be practicable in most collections.

Frames

Original frames should be preserved wherever found. French rococo frames achieved an extraordinary level of sophistication and are often worth more than the pastels they frame (no doubt many have been stripped out for this purpose: but leaving pastels unprotected out of their frames is probably the biggest cause of their destruction). But such frames can be very fragile, not just because fine details (e.g. decorative corners and especially ribbons projecting above the top rail) can easily be broken in handling, but also because carved giltwood frames have an inherent problem: the oak carcass has a tendency to shrink over long periods, while the hard gesso onto which the gilding is laid does not: this means that frames should only be handled on the flattest part. (Composition frames, particularly favoured in the nineteenth century, are also vulnerable loss of large sections of decoration. They are also far heavier than carved wood frames, exacerbating the problem of shock in transportation.)

As with any piece of period furniture, control of humidity levels is important.

Oval frames have these problems to a far greater degree: while a straight piece of wood can be cut along, or across, the grain for a consistent response, the four pieces of a traditional oval frame unavoidably include pieces cut at different angles to the grain. Joints on original oval frames have almost always been broken and repaired frequently. (In Eastern Europe it was common to use octagonal strainers and frames to avoid these problems, while in Germany oval pastels were often made on rectangular strainers with rectangular frames so that only an oval slip need be cut.)

While these problems are no different from damage to all period picture frames (and while it is relatively straightforward to repair them), the difference between pastels and oil paintings lies in the hazards pastels face when taken out of the frame. This should not been attempted in a gilding workshop, where a fine layer of gesso dust is likely to settle on any exposed surface. Pastels should be unframed by a skilled paper conservator and immediately stored flat in a dust-proof solander box before the frame is taken offsite.

A related problem (again far worse with ovals) is the possibility of failure of the wooden strainer. When this happens with an oil painting, a straightforward repair can be made; with a pastel, however, the consequences can be catastrophic (particular care should be taken not to take a pastel out of its frame if the strainer shows any risk). It should also be noted that most strainers will have lost some tension since they were originally made: this can result in the canvas billowing during any movement, thereby loosening the bonding.

Transport and loans

In 1747 Oudry sent a pastel (a precious landscape which he had exhibited at the Salon) to comte Tessin as a mark of his friendship and esteem. In the accompanying letter (Uppsala Universitetsbibliothek) he wrote:

J'ai fait fermer la caisse et toutes les séparations avec des vis, parce que les coups de marteau detruisent entierement le pastel en le faisant tomber. ... Quand le tableau en sera tiré, il sera à propos de prendre un canif, de couper tout autour le papier qui est colle derriere pour tenir le pastel à la bordure, ôter le tableau, essuyer bien la glace, et remettre le pastel et aussi recoller les bouts de papier, parce que le transport détache toujours quelque partie qui s'attache à la glace et ternit l'ouvrage.

Unfortunately the pastel was last recorded in the Swedish royal collection in 1911, and we cannot assess just how severe the losses were. Perronneau inscribed this more optimistic observation on the back of one of his pastels (Mme Schweighäuser, 1767):

Si il arrivoit que l'on voulu transporté en voiage ce tableau il faudrait faire une caisse ou bouette qui ne ferma qu'avec des crochets et non avec des clouts parcequ'il ne faut pas frappe crainte de gatter le pastelle et lier la caisse avec une corde et bien l'emballé alors on ne risque rien.

Another example is provided in a 1771 letter from the Duke of Argyll to his framer insisting that to avoid damage his Read pastel "be carried on a man's back" from Edinburgh to Hamilton Palace (over 40 miles).

Exactly the same concerns remain today. Packing cases have to be fastened by hand (although staff in logistics companies will instinctively reach for labour-saving power tools unless supervised). Triple-lined museum cases may seem to offer better protection than soft packing, but they become much heavier and harder to move without greater shock levels than careful hand transport. Hydraulic lifts on purpose-built pantechnicons, which boast air-cushioned rides, can only be operated with the ignition running, with engines which create alarming levels of vibration when idling. The vibration levels experienced in air travel are obviously undesirable. Maintaining climate control during international travel is itself a challenge.

The debate about how to move pastels safely has not yet been resolved. It is particularly relevant in the case of loan exhibitions, where many museums operate a strict refusal policy. There is nothing new in this: the impossibility of lending pastels was not questioned in the Chardin (1978) or Boucher (1986) exhibitions, although the scientific consequences were deplored, and - since they are self-reinforcing - serve to maintain the neglect of this field. Considerable experience of safe handling of nineteenth-century pastels has now been built up, but whether earlier works can safely be moved, and if so how, remains in dispute. The General principles on the administration of loans and exchange of works of art between institutions, issued 1995 originally by the Réunion des musées nationaux (revised 2002, and adopted by the Bizot Group of some 67 major museums worldwide) observes that the potential dangers of damage are "very real" and concludes that "unfixed pastels are usually too fragile to travel". The Bizot Group principles have been revised in part, but the section about pastels appears not to have been amended.

If exceptions are to be made, further decisions must be taken. Should pastels travel vertically (if so, in the direction of travel, or at right angles?) or horizontally? Should glass be taped to guard against the devastating consequences of the small risk of breakage? Is the removal of such tape (if used) safe in an ionizing air flow? Should old glass be replaced with toughened sheets? Should the billowing of strainers that have lost tension be addressed by inserting a quilted lining between the backing and the canvas? All these perfectly sensible questions, and the many intelligent suggested responses, do not currently meet with agreed answers or fully tested solutions. For the moment risk assessment will depend upon the skill and judgement of conservation and curatorial staff. Each case is different, but the risks are greater for ovals, works on multiple sheets or on parchment, and compound greatly with size. The only safe advice is not to allow pastels to undergo unnecessary travel.

Photography

I am frequently sent photographs that are almost useless because of reflections in the glass: yet it is quite simple to take good quality images. It is quite unnecessary to remove the glass (and for the reasons discussed above, I strongly recommend that you do not do so). Ideally photographs can be taken in a studio with multiple light sources. But an adequate image can be obtained far more simply, just by using flash, and this doesn't require the picture to be taken off the wall. You simply need to take the photograph at an angle so that the flash is reflected outside the camera field. To find the best angle to minimise distortion, imagine a rectangular box extending into the room, one end of which is the face of the picture; get the camera as close as possible to the middle of one of the longer sides with the flash source just outside the exclusion zone of the imaginary box.

You should close curtains etc. to remove other sources of light. Make sure that the whole of the frame sight is captured. The resulting image can easily be straightened up using software (but I will need the exact dimensions of the sight to do so). The technique works with ovals, but without an understanding of conic sections the adjustments are less intuitive. One good photograph of the whole pastel (to the sight edge of the frame) is of far more value than dozens of details, but take images also including the frame and of the back, as well as of any inscription or old label.