wood is weakened by splits or degraded by woodworm damage, and when
glued joints are embrittled by age. It is doubtful that any calculation could
bedevised to give a value for the elasticity of the wooden panel at right
angles to the grain—there being so many variables and features peculiar to
each panel. It seems that framing systems that exert minimal restraining
force at the edges and parallel to the wood grain are least likely to cause
damage. A framing system to achieve such a goal would have to be
designed on an individual basis, with the construction and inherent stresses
of each panel, as well as its display and travel requirements, taken into
account. It is an important consideration that systems that hold the panel
while allowing maximum movement often provide insufficient support
when the panel is moved.
Three case studies illustrate techniques employed at the Hamilton
Kerr Institute. The first, the framing of a small panel of three vertical
planks, the Portrait of Elizabeth Altham,has been mentioned above. The
second is the framing ofthe large altarpiece Noli Me Tangereby Anton
Raphael Mengs, with horizontally aligned planks; the frame of that work
required considerable strengthening. The third is an altarpiece, attributed
to Pietro Gerini, which had been reframed in the nineteenth century.
Even though the frame was causing splitting in the three panels, the client
wished the altarpiece to remain unchanged in appearance after treatment.The 1617 panel, one of a pair of seventeenth-century portraits, was in its
original frame of oak painted black and partly gilded. The joints of the
panel had opened. After gluing, the panel was observed to assess the maxi-
mum curvature it would develop. It became slightly convex at an RH of
50%. It was decided to increase the depth of the frame rabbet from 8 mm
to 18 mm. Strips of dimensionally stable spruce were cut and angled to the
outside edges of the frame to make the addition inconspicuous. Strips 25
mm in width were mitered, stained dark, and attached with screws to the
back of the frame. (The use of glue was rejected as less reversible.) The
addition is not visible when the frame is hanging, and only aslightly larger
gap between frame and wall is evident. The panel fitted quite closely to
the rabbet, which did not require any addition of shaped sections to follow
the panel’s curvature. Rabbets can be adjusted to remove the gap between
the sides of the panel and the frame edge, which can be visually distract-
ing, especially where light can be seen between the picture frame and the
wall. However, a curved rabbet can itself restrict movement if the panel is
subjected to higher RH levels; in such a case the sides of the panel across
the grain will press against the outer edge of the rabbet and the nearest
top and bottom fixing points, exerting pressure on weak areas and joins.
The panel must be able to assume a less convex profile; this is facilitated
either when space is left for movement at the edges or when the central
fixings are designed to compress, allowing the panel to move away from
the frame rabbet in the center. Compressible curved additions to the rab-
bet could be made. However, any material that can be accurately shaped
and that presents a visually acceptable surface, such as Plastazote or
Evazote, is likely to be too rigid to conform to changes to the configura-
tion of the panel.
The panel was then fitted in the frame. The central vertical plank
was set on a thin, 15 mm strip of hardwood. This raised the lower edges of
the panel away from the bottom edges of the frame, allowing free move-Portrait of Elizabeth Altham
440 McClure