The choice of positions for the retainers was based partly on a regular
distribution across the panel and partly on the location of relatively flat
gluing surfaces.
Normally, the removal of sliding battens from a panel requires a
space of twice the batten length. The required space was twice the panel’s
height in this case and made a modified means of removing the retaining
strips desirable. It was possible to narrow the strips at intervals equal to
the vertical distance between the retainers. This allowed the strips to be
placed directly against the lattice battens (Fig. 17b, c) and between the
retainers; they were then slid down and engaged in a functional position.^53
Built thus, the structure provided adequate reinforcement while it
was still flexible enough to bend with warp movement. This reduced the
risk of restraint of lateral movement by friction and locking. It protected the
edges against mishandling and accidents. While attached, the structure still
permitted examination of much of the panel back. Most of the structure
could be quickly removed to access all of the back surface except beneath
the retainers. The retainers could be removed mechanically with relative
ease because low-density wood was used. The glue used to attach the retain-
ers could be easily swelled with water and removed with spatula and swab.^54
Since the structural work was completed, the panel has been
monitored for at least two years to determine the effectiveness of the rein-
forcement and other aspects ofthe treatment. Due to RH variations,
changes in deflection at the middle (in relation to the side edges) have
been measured at up to 30 mm—about halfthe deflection that was
observed under a similar RH range when the panel was structurally con-
solidated but not reinforced. The two central retaining strips have shown
increased friction as the panel has become more convex (viewed from the
front), but lateral movement has not been excessively constrained, as
occurs frequently with more rigidly battened or cradled panels of this
nature.^55 The panel appears to be adequately reinforced and moves with-
out any obvious detrimental effect.Framing, hanging, and transit
Old wooden panels are continually subject to movement—probably nearly
as much as when they were first painted (Buck 1952; Laurie 1967:55; Klein
and Bröker 1990; Mecklenburg and Tumosa 1991). Therefore, allowance
should be made in the frame for potentialpanel movement. Of course,
ex cessive frame restraint would negate any capacity of the panel’s rein-
forcement structure to allow for movement. Considerations related to the
frame retention of panels are similar to those relating to auxiliary support.
Many paintings do not remain in a relatively constant, well-controlled
environment. Passive controls are not always sufficient, and active controls
can malfunction in even the best-maintained buildings.
Therefore, an allowance must be made by sizing the frame rabbet
for cross-grain expansion of the panel wood. Otherwise, a “bound-in-
frame” condition occurs as the panel expands to press on the rabbet’s
outer walls. Also, it is important that framing not restrict warping move-
ment with overly rigid retention. These stresses can easily break the panel
or the frame (Museum1955:159–60). Of course, competent framers allow
space in the rabbet to avoid this possibility, but the degree of panel move-
ment can be underestimated, especially in larger panels.470 Brewer