prepared tracks and glued in place with a PVA emulsion.^21 The degraded
part of the wood was eliminated by the correction of the edges, and the
level of the painted surface in those areas was realigned.
Inthis operation our intervention was limited to removing only
the parts affected by degradation, so that, as much as possible, the trian-
gular angle ofthe cut was retained. It was considered vitally important
that each wedge be positioned in such a way that the grain be parallel
to that of the panel and that the annual rings be arranged radially with
respect to the plane of the support. Such positioning is more compatible
with the wooden construction and, in case of dimensional changes, guar-
antees less deformation and, thus, a greater bond to the support. The
most difficult part of this operation was the repair of the onion effect seen
in the central plank (Fig. 27). The author and coworkers thus proceeded
with the removal of the wood affected by the phenomenon, following its
irregular disposition. The intervention was carried out in multiple steps,
by alternating opening, leveling, and gluing of blocks in many layers for
the reconstruction of the weak parts. Then wedges were inserted to join
the faces of the opening. This technique made it possible for the surface
to be perfectly adjusted: the multistep integration reduced the forces in
the wooden material, lending greater stability to the intervention. Even
the leveling of this area had to be done in different phases to obtain good
results without the application ofespecially strong force. The solution
adopted for the crossbars followed the concept used in the previously
described intervention—that is, to make modifications only in the bonds
between the various components.
The method ofreducing the rigidity of the crossbars is altogether
valid today when the panel’s planks are subjected to a warping stress that is
thought to be irreversible. Such situations generally are the cause of the for-
mation of cracks and instability of the paint and preparatory layers. These
phenomena can be caused by the aging of the material in relation to the
characteristics of the wood in terms of quality and positioning of the grain,
or by environmental factors that have affected the life of the painting.
Removal of the upper and lower crossbars proceeded with the
cutting ofthe tips of the nails that were hammered over onto the back.
This operation made it possible to observe that two sides of nails had
some space in the horizontal direction as a consequence of the yielding
between the parts (the walls of the wood and the flexible metal) that
occurred over time. The next step consisted of widening the holes left by
the nails in the crossbars (by 3mm) and threading (with a 4 mm pitch) the
uppermost centimeter or so of the nails that protruded from the support
for the entire thickness of the crossbars. The crossbars were then inserted
onto the protruding nails. Because the crossbars had been thinned on
the back to a thickness of a few millimeters, it was possible to reconnect
the crossbar to the support in a stable manner with a nut. For the central
plank, which was missing the original nails, a mechanism was used that
consisted of a brass stop plate with a rectangular slot inside, held onto the
lower face of the crossbar in contact with the panel with two screws and
epoxy resin; a bolt was free to move within the slot but was held in by its
head. A double-threaded brass bushing was inserted into the support—the
external thread to anchor the bushing to the support, the internal thread
to receive the bolt. The bolt passes through a hole made into the crossbar
and attaches itself to the support. Movement is ensured by the head of theT R P P S: S C H 337Figure 27
Lorenzo Monaco, The Annunciation. Ring sep-
aration in the wood on the back ofthe panel
before repair. Also visible is the reconstruc-
tion in the grain direction of the old butterfly
inserts.