9 According to Castelli and coworkers, such a choice may be explained by the fact that wood
placed nearer to the pith (possibly assumed to coincide with the heartwood) was considered
to be of better quality. As a consequence of seasoning, since the cupping convexity becomes
oriented toward the “inner” face, the application of the ground to that face should ensure a
better “grip” (especially during its application, when its high water content makes the wood
swell and then dry again) (Castelli, Parri, and Santacesaria 1992).
10 Defects produce local differences in shrinkage that often result in greater damage to the
painted layer in their vicinity. Because local wood defects were seldom a direct cause of a gen-
eral or widespread deterioration of a painting support, the acceptance of boards with some
localized defects may have been a reasonable choice, considering that negative effects could be
prevented easily by appropriate techniques (e.g., the gluing of a layer of cloth between the
boards and the gesso).
11 Because of obvious geometrical relationships, if the “inner” face of the board is oriented
toward the painted face, wane will appear on the rear face. If, on the contrary, the growth
rings are oriented so that wane is toward the painted face, the wane will need to be “repaired.”
12 The presence of sapwood may not be considered a defect in itself, especially in species in
which heartwood is not distinct, such as poplar, linden, and fir.
13 For instance, Del Serra stresses this possibility for the support of Cimabue’s Maestà (Del Serra
1994). Many Florentine panel restorers use the expression legno frollo(“tender wood”) to
describe partially decayed wood. This expression applies to the early stages of fungal decay,
during which the strength properties of the wood are only moderately affected, while shape is
retained and hygroscopic stability is higher.
14 According to established experience (which is basically the same today as during the
Renaissance), good natural seasoning practice may require years for boards to reach a
satisfactory absence of moisture gradients and settling of internal stresses. The specific num-
ber of years required varies with wood species and their particular permeabilities.
15 This is supported by the fact that given the typical climate in the area, natural seasoning (i.e.,
the traditional drying process, by which boards are stacked and left exposed to natural envi-
ronmental conditions) would have hardly produced a lower EMC. Lower EMC values (around
10%) reached by the paintings that were later kept in heated buildings (after transfer to heated
houses or museums, or after heating plants were installed in their original locations) usually
led to severe shrinkage (Uzielli 1994).
16 When a support is exposed to humidity fluctuations, a greater thickness ofboards contributes
to its dimensional stability, since the consequent MC fluctuations in wood are slowed down
(damped) by the time required for moisture to move into the deeper layers.
17 In many cases, the current thickness of boards differs from the original one. This difference is
due to thickness reduction for cradling, rebacking (usually in the case of severe wood decay, or
to remedy the warp), sawing along the central plane to obtain two separate paintings from a
double-faced panel, or other kinds of intervention, intended either for conservation or for cos-
metic purposes.
18 During recent restoration works performed by Del Serra in 1993–94, the rear surface of
Cimabue’s Maestà (226 3 387 cm) was found to be cambered (cylindrically shaped), measuring
58 mm thick along its central longitudinal axis and growing progressively thinner in a sym-
metrical fashion to the two lateral edges, 40 mm thick. This shape is clearly the result of the
original manufacturing process, since—after removal of the nineteenth-century crossbeams—
there were a number of features to indicate that the present surface is the original one, includ-
ing some remnants of a possibly original red tempera on the surface and red lines made with
the string-snapping technique across the panel width, marking the alignment of the nails that
connected the original crossbeams.
19 Such a rare feature is characterized by boards becoming progressively thinner from the center
to the lateral edges of the panel; it should not be mistaken for the beveled edges that appear
frequentlyin Flemish panels (typically made of radially cut oak boards and thinner than the
Italian panels) and that are intended to allow for an easier fitting of the panel within its frame.
20 In some instances, the edging process might leave some wane, particularly when there was a
need to take advantage of the maximum possible width of the board.H O P-M T C I 131