removing the dark
glue and other foul matters from the irregular surface.

This requires the application of warm water, by means of a stiffer
implement, used in fact as a small scrubbing brush; the moisture
between whiles is pressed in and out by the fingers with repeated
wipings and re-wettings.

After a while, being satisfied that the surfaces of the fracture are
clean to the desired degree, James allows the wood to partially dry.
In the meanwhile, he cuts a couple of pieces of wood to fit the back
and front of the table, so that with folded paper as a pad the parts
may be pressed together. All being in readiness, fresh strong glue is
inserted all along the opening, the repeated pressing being kept up
until he is sure that the glue has penetrated every part. The
superfluous glue oozing from the inside, after a slight squeeze along
the course, is lightly wiped off, the moulds applied back and front,
and the cramps fixed.

Other little odds and ends of repairing have occupied the attention
of the assistant during the time of waiting for the drying of the
before-mentioned upper table of the Stainer. This latter being of full
model, although not resembling the swollen or bolster-like form of many
imitations of the master, requires special attention with regard to
fitting in of the fresh wood or core.

The fresh wood must not be bent, or the matching of the grain with the
old material will be impossible, and the repair when completed will
be strikingly conspicuous. It must therefore be inserted in such a
manner that when pared down, the direction or flow of the grain will
exactly coincide in all respects with the rest of the table. The fact
must be recognised that although the threads may be perfectly straight
from end to end, yet they may rise higher at one end than the other
or not run level with the plane of the table. (Diagram 37.)

[Illustration: DIAGRAM 37, showing half length of core inserted before
being pared down, and with grain.]

This being duly calculated by James when cutting the piece that is to
be inserted as a core, the table with its opening extending nearly from
end to end is placed on an even surface facing upwards, as when on the
instrument. The core is then tried in the aperture. Perhaps a thin
shaving or two is found necessary, when finally it is sufficiently
exact. The next stage is that of getting ready the means of holding
or pressing all the parts together till dry after glueing. Taking a