bending stresses would be more likely, imposed by considerable leverages.
Any warp from a flat plane would promote buckling. Such a buckling ten-
dency would pose a long-term bending stress across the joint axes, a condi-
tion greatly exacerbated by thinning.
Evidence of a previous, perhaps original reinforcement exists as
eight sets of three holes each, spaced at regular intervals across the panel
back (Fig. 9d). It was unknown what form this reinforcement would have
taken. Had documentation of the panel back prior to the recent thinning
been available, it may have provided evidence to help with subsequent
treatment decisions.
By contrast, the structure of the Palmezzano is more logical. The
chosen wood is lighter, with the planks disposed vertically. Thus, the wood
bears weight in a more natural orientation, analogous to its mechanical
role in a living tree. With the grain vertical, buckling would be a negligible
concern, even with the panel half its original thickness. Therefore, weight
does not combine with movement across the wood grain to threaten the
Palmezzano’s structure as much as in the Mengs.
The Palmezzano had no evidence of original auxiliary support.^34
At least four different types of battens had been applied at various times
over the long period since the panel was last thinned. Finally, remedial
action took the form of short planks glued over developing splits, includ-
ing old stretcher members taken from paintings on fabric. Eventually the
panel became choked with stopgap solutions. These additions induced
severe distortions, splits, and compression damage concentrated in the
panel’s center. Centralized damage occurred because overall reinforce-
ments tended to concentrate bending stresses toward the middle. This fac-
tor then combined with tension and compression stress overall caused by
restraint oflateral movement.^35 The pattern ofsplits shows howstresses
were interrupted over the cross-grain battens (Fig. 4c).
Also, putties (or fills) had been applied to splits that had not first
been rejoined. This and subsequent wood movement caused compression
stress and distortions in the adjacent paint. Such disfiguring damage is nei-
ther easily nor totally rev ersible. It is better not to put fillers into surface
cracks ifeffective structural work is not done first to underlying splits.
For both panels, attempts to flatten and reinforce them have
instead tended to weaken them further. Such treatment efforts are examples
of excessive, damaging measures that have been used to meet reinforce-
ment requirements of some large panels, as well as to serve aesthetic pur-
poses. The resulting deterioration of paintings withsupports shows that
those requirements must be better understood, and they must be achieved
with better methods that maintain the integrity of the panel painting.
The consequences of thinning a large panel can be critical, mainly
because a heavy weight must then be supported by a structure made rela-
tively weak while still allowing for adequate wood movement under vari-
able conditions. It is worth examining the motivations for thinning, which
have particularly serious implications for preservation of larger panels. In
general (and leaving the question of transfer procedures aside), panels may
have been thinned for several reasons, including:- the mistaken belief that thinning reduces the tendency to
move and warp in response to changes in MC (the reverse is
true);^36
P A S C L P P 461