add that the broad edges, which
you join afterwards, making the wood for the upper or lower table
look like the roof of a house, are at the outer part of the tree,
springing from the centre, where are the broadest rims, as is
natural, seeing that youth is there, vigorous and full of sap;
whilst the rims decrease to the outer, or bark part, in some cases
very decidedly in width, in others more slowly. So you may gather
from this why we have the narrow bait, or reed, where the bridge
comes, the open reed at the edge. At least, I hope you can see the
reason, which is, as generally admitted, and is certainly _my_ view
of the matter, the strength is most wanted at the centre of the
violin, as at the bridge, which the closeness of the reeds and
narrowness of cell passages would supply. The broad reed is more
volatile, and we put it to the edges, where it throws off the
gathered activity of resonance, recurring so rapidly, which we
increase by reducing the thickness of the plate there, bringing
about that timbre so rich to the ear of the listener. These remarks
apply to the belly, and are offered "on the quarter" only. Wood "on
the slab" is never used for the front table in any case, as, cut
that way, it would be far too weak.

For "on the slab" means that our orange and our tree are cut
_through_ the rings or reeds in flat layers of equal thickness (as
required); and it is at once obvious that, in the upper plate, would
be not only ugliness of broad, irregular figure of wood, with now
and again snatches of the bait as it should be, and as I have
endeavoured to show it is; but apart from its general weakness, it
would be most irregular as the main vibrator or soundboard, so is
entirely discarded.

But both ways of cutting are employed for the back; I have heard
tone as fine from one as from the other, yet I think, as a rule, I
prefer the quarter to the slab, as being somewhat more resonant and
of finer timbre.

At the last moment I have selected the "brother" of my "Elephanta"
violin--I mean its back--in one piece, but on the "quarter" in
preference to the one on the "slab," so we will now proceed to
active work upon it.

I clamp the selected wood to this bench, having the flat side
uppermost, and so that I can plane it to a perfect level surface,
first at the narrow end, then at the broad, but _across_ the wood,
as, being sycamore, it is very liable to cut very roughly if done
lengthwise. This I do with a twenty-two in