While completing the installation of a new organ in the
Tennessee Valley Unitarian Universalist Church in late 1998, I was drawn into a
conversation between Will Dunklin, the organist, and Marian Moffett, a viol da
gamba player who is a member of a local early music ensemble. Marian indicated
an interest in acquiring a small chamber organ for her home, which would be
appropriate as a continuo instrument for early (particularly English) music.
After briefly discussing prices, both Will and myself commented that an early
American organ (pre-1860) would possess many of the tonal characteristics
required for such a use, as well as providing its own historical interest.
Besides, restoration of such an instrument would likely be quite economical
compared to the price of a new organ.
After checking with the Organ Clearing House, we found
nothing small enough for such a use, and the matter got shelved in the back of
my mind. About a year later, I received a message from Marian that Will had
found a small American chamber organ on eBay, for sale by a doctor in Michigan.
After some negotiation, she purchased the organ and went with Will in a rented
van, returning two days later with said instrument. In such a serendipitous
series of events, then, did this enigmatic and charming little instrument fall
into my hands for the purpose of restoration.
Provenance
Establishing the provenance of the instrument was the first
item of interest; since the organ sat in the shop for a year before work could
commence, it gave me some time to pursue the subject. Alas, despite our efforts,
the little instrument still remains anonymous. The following, however, are some
of the identifying characteristics pertinent to its provenance.
The cabinet holds a number of clues, which help us make some
general conclusions. The cabinet (as well as the chest and internal framework)
is made of eastern white pine, with a smattering of cherry and black walnut.
This clearly identifies it as an American-made instrument. The Empire case,
with its ubiquitous crotch mahogany veneer and late Empire styling, seems to
place it between about 1845-1855. According to Barbara Owen, the cabinet looks
like the work of early Connecticut builders. This dovetails nicely with the
oral history we received from the previous owner, who had been told that the
organ was built for the Lockwood family of Norwalk, Connecticut. Apart from
these general observations, the cabinet holds another clue: the ripple
moldings, which appear in several shapes and sizes. According to an article by
Carlyle Lynch in the magazine Fine Woodworking (May/June 1986, pp. 62-64), such
molding was made by only one company in America, the Jonathan Clark Brown clock
company in Bristol, Connecticut. This company made the gew gaw covered clocks
known as steeple clocks, but after the factory burned in 1853, J. C. Brown
clocks no longer were made with the unique ripple moldings. Such moldings
require an elaborate, slow-moving machine for their manufacture, and the
machine was evidently never rebuilt. If the builder purchased his ripple
moldings from the clock company, then it is clear the instrument was built
before 1853.
The hardware found on and in the instrument provides more
tantalizing hints as to the organ's provenance. The mix of early factory-made
components with other hardware which is clearly hand-made seems to place the
organ on the very cusp of the Industrial Revolution. For instance, the lock for
the keydesk lid bears unmistakable marks of being handmade: all parts were hand
filed out of solid brass, and then fitted together with hand-threaded screws. Yet,
the hinges which occur in various places (e.g., swell pedal, main reservoir)
are all of cast iron and bear the name "Clark's Patent." While a bit
crude (they certainly are not interchangeable), they bear all the signs of
early factory production. An additional item of interest is that one leaf of
each hinge was cast around the pin while the pin was inserted into the other
leaf. This makes it impossible for the pin to ever work its way out; it also
makes it impossible to separate one leaf from the other, short of a sledge
hammer.
The most interesting piece of hardware is the square iron
roller for the swell mechanism. Clearly stamped on the bar is the word CLYDACH.
It turns out that Clydach was a Welsh ironworks established in 1793, continuing
in production until about 1858. I'm not sure what this reveals about early
American sources of iron and steel. Of course, it is possible that the builder
recycled the piece of iron from an older apparatus or structure.
Finally, even the humble wood screws give us some
information. They are a mix of the earlier blunt ended screws and the more
modern pointed screws, and all but one or two were clearly made by a machine.
This also seems to point to about 1850-1855, although I am unsure when the more
modern pointed wood screws became available. The E. & G.G. Hook organ of
1847 in Sandwich, Massachusetts, was put together entirely with blunt ended
machine-made screws, so it seems that modern wood screws came along a few years
later.
One intriguing note is written (sometimes scrawled) on
almost every piece of the instrument. The message "No. 2" can be
found on the bellows, keyboard, backboard, knee panel, etc. The inescapable
conclusion is that there must be (or must once have been) a "No. 1"
lurking out there somewhere, waiting to be discovered.
The reader is left to draw his own conclusions about the
provenance of the instrument. Clearly, the Empire style and the handmade
hardware place the instrument no later than about 1855. The wood screws fit
into the time frame of about 1850. The oral history as well as the general
design of the case place the builder in Connecticut. We were unable to find
information about "Clark's Patent" hinges, and CLYDACH presents more
an enigma than it does an answer. Perhaps a reader will recognize one of these
items and shed a bit more light on the history of this little instrument.
Restoration techniques
The following describes the techniques and materials used
for the restoration. An astute reader will occasionally see the tension which occurs
when the desire to restore the organ to its original state is not always in the
best interest of the customer. Ultimately, we did almost nothing to the
instrument which could not be easily reversed later. Additionally, we took
great care to avoid removing any original material (no pipe tops were trimmed,
and even the finish was not entirely removed).
Cabinet
Failing joints were disassembled when practical and re-glued
with hot hide glue. Other joints were simply injected with hot hide glue and
clamped for 24 hours minimum.
The reservoir and feeder assembly share a common 1"
thick horizontal board which is dadoed into the sides of the carcass. This
board was originally glued into the dados and glued and nailed to the front
rail directly above the two pedals (the self-closing swell pedal on the left,
and the single pumping pedal on the right). Mahogany crotch veneer was then
applied over the nails. Someone had previously done a very nice job of sawing
through the nails and sliding the entire assembly out the back of the
instrument in order to patch the bellows. We decided to leave this alteration,
since it is truly the only way to access the bellows for releathering. Maple
cleats were added so that the 1" board could be screwed securely to the sides
of the carcass.
Stabilizing and repairing the veneer became one of the most
time-consuming jobs. Like many Empire pieces, the crotch burl mahogany seemed
to shed little bits of veneer onto the floor every time one walked past. About
half of the veneer was no longer securely glued to the white pine below, and
the ogee-shaped front board of the folding lid was missing about 70% of its
veneer. The ogee crown molding veneer was almost entirely unglued from its
substrate, although miraculously most of the veneer was still there. The
decision was made to remove the remaining tatters of veneer from the ogee
shaped lid front and use the bits to patch veneer on the rest of the piece. The
lid front was then entirely re-veneered with book-matched mahogany crotch burl.
The crown molding presented another challenge; the veneer
was so brittle that even the slightest attempt to lift it in order to work glue
under it caused it to shatter. Clamping was difficult; since the veneer was
glued over a hand-planed ogee, the shape of the contour changed from one end to
the other, and the molding on the sides of the crown were quite different in
shape from each other and from the front. This precluded any possibility of
making precise blocks to fit the shape of the molding. The solution was finally
to inject fish glue through tiny holes in the veneer and clamp a sand-filled
Ziplock bag firmly over the area. The sand conformed perfectly to the contour
of the molding and distributed the clamping pressure evenly. The fish glue,
being a protein-based glue, was compatible with the old hot glue and adhered
well, though it required long clamping times of about 48 hours. Close
inspection reveals the pinpoint size holes through which the glue was injected,
but it seemed the least destructive way to stabilize and re-glue the very
brittle veneer.
Conservation of the finish required a careful approach.
Rather than subject the piece to the humiliation of being entirely stripped and
refinished, we decided instead to conserve what was left of the old shellac
finish. Parts of the case, such as the underside of the lid, retained the
original finish in excellent condition. Other parts had obviously been covered
with an additional layer of low quality shellac. Besides this, someone had
studiously "patched" every missing veneer chip by the application of
red-primer colored latex paint. Paint ended up on the surrounding intact veneer
as much as it did on the offending gap in the veneer. To address these multiple
problems, the course of action was as follows:
The top layer of accreted dirt and crazed finish was sanded
off using 400-grit sandpaper with paint thinner as a lubricant. This required
removing only a very thin film of finish. Then, a pad of wool and cheesecloth
was filled with shellac and applied over the remaining old shellac. This
smoothed out any remaining "alligatored" shellac. This French Polish
technique was repeated about a dozen times until the surface took on an evenly
covered appearance and began to glow. Then, at the request of the customer, the
shellac was sanded lightly and was covered with two coats of high quality
varnish for durability. On parts of the cabinet where extensive veneer patching
was required (such as the crown molding), the resulting surface was too rough
and the old finish too compromised for conservation; it was necessary to sand
the entire surface down to the bare wood. Then, colored pumice was rubbed into
the grain along with residual sanding dust and garnet shellac, after which the
usual french polish technique was used, followed by the two coats of varnish.
The orange colored garnet-lac returned the "old" color to the newly
sanded wood, making a perfect match. The results were visually stunning; the
mahogany crotch burl fairly leaps off the surface of the piece with three-dimensional
fervor. The keydesk itself is veneered with rosewood, and since the lid
evidently was always closed, the finish on the rosewood required little
attention.
The center panel of cloth was originally a very thin silk,
bright turquoise in color. We found well-preserved pieces of it under the wood
half-dummy façade pipes. Marian decided the original color was
remarkably wrong for her house (I had to agree), and chose a silk of subdued
gold instead. The turquoise silk is still under the dummies for future
reference. Behind the cloth panel is a very small swell front, with shades
which open only about 45 degrees. After listening to the instrument, we decided
that omitting the shades made the organ considerably louder, and virtually
perfect in balance to a small consort of viols. Fortunately, there is a large
well behind the crown molding which provided a perfect storage space for the
shades. Reinstalling them would be the work of a few minutes should a future
owner wish to use the organ in its completely original state.
Wind system
The bellows still had its original leather, but every square
inch of it had been secondarily covered years ago with hot glue and rubber
cloth, probably by the same party mentioned earlier who went to such lengths to
remove the bellows plate from the organ. The rubber cloth and hot glue had
ossified into a stiff, inflexible board-like structure which had caused all
bellows hinging to rip itself apart upon inflation of the reservoir; the single
large feeder suffered the same fate. The bellows and feeder were completely
releathered with hot hide glue and goatskin. The bellows and feeder boards were
rather generously filled with splits, cracks and checks; the worst were
reinforced with butterfly-type patches, and all were entirely covered with
rubber cloth to prevent leakage.
The short wooden wind line which conducts wind from the top
of the bellows plate into the chest was originally simply fitted into place by
friction, but the horizontal members of the cabinet frame did not shrink and
expand in the same direction as the vertical boards of which the wind line was
made; in summer, as the cabinet expanded and lifted the entire upper assembly
away from the bellows, the leakage must have been spectacular. The joints
around the wind line had probably received more attention over the years than
any other part of the organ. Numerous layers of patching (leather, glue, rubber
cloth) attested to the trouble which this particular design flaw had visited
upon those who chose to play the instrument in humid weather. It seemed that a
change was necessary, so four small oak cleats were attached to the narrow ends
of the wind line so that it could be screwed securely to both the bellows top
and the bottom board of the pallet box. The cleats are clearly and
intentionally not a part of the original construction.
Chest
The chest was plagued by innumerable runs, and after some
investigation, they all were found to be caused by a joint in the table. The
front five inches or so of the grid is covered with a thin (1/4") mahogany
table. The rest of the chest is covered by one large pine channel block,
13/4" thick and honeycombed with many channels. The joint between the thin
mahogany and the thick pine channel block is naturally a source of some tension;
even though no crack had opened up between the two, the mahogany had almost
imperceptibly lifted along the joint. The problem was solved by screwing down
the mahogany piece with a screw in every rib, and by gluing a piece of thin
leather in each channel to bridge the joint. Should the joint ever move again,
the flexible leather should absorb the movement and prevent leakage. All key
channels, as well as all offset channels, were poured out with sanding sealer.
Shellac could have been used, but since the work was being performed in the
humid summer weather of East Tennessee, I decided to avoid shellac because of
the tendency of its solvent (alcohol) to absorb water from the air.
The bottom of the grid was originally covered in a thick
cotton covered with much shellac. We chose to replace it with rubber cloth.
Pallets were re-covered with two layers of leather, just as they were
originally, and they were installed in the original fashion, glued with hot
glue at the tail and held down by a small pine slat nailed on by tiny cut
nails. The builder evidently thought it was necessary to provide pallet sizes
commensurate to the wind demand, so the already tiny bass pallets (43/4"
long) were made even shorter at middle C (4" long).
Key and stop action
The keys are mounted on a balance pin rail at a ratio of
roughly 2:5. Thus, the pallets open a small, but nonetheless sufficient,
amount. Under the keyboard is mounted an elegant mahogany backfall (ratio 1:1)
which pushes down on very slender (.047") brass wire stickers. The
stickers pass through the 1/4" mahogany table, which also serves as their
register, and push the pallets open. All the stickers are original and the
action is pleasing to play and surprisingly responsive; in spite of the tiny
pallets, a definite pluck can still be felt in the keys. Key bushings are wood
on round brass pins, and the keys are covered in their original ivory. The
pallet springs are brass, clearly factory-made, and were still all perfectly
regulated when I checked them. No spring varied from all the others more than
1/4 ounce. I left them unchanged. The builder solved one problem with the
keyboard in a rather clever way. Since the keyboard is so short, it is not
possible to place the usual 19th-century style lead-weighted floating thumper
rail behind the nameboard. The builder instead installed the nameboard itself
in loose dados in the stop jambs so that its felted bottom edge simply sits on
the keys, keeping them in tension and making it possible to adjust them
perfectly level. When seasonal changes occur, the nameboard itself simply rides
up and down in the dados. (Of course, since this particular nameboard has no
actual name, it must be a nameboard in name only).
The stop action would seem to need no mention, except for
the stop to the left of the keyboards. The single knob to the right pulls on
the tiny slider for the Principal 4', which leaves the knob on the left with no
job to do at all. However, the builder thoughtfully provided a slotted block so
that the knob, which does absolutely nothing, can be pulled out just like its
brother on the right. The disappointing aspect is that the Principal had its
original engraved ivory disc, but the ivory disc on the left was missing. I
glued in a blank ivory disc for appearance's sake, but I will always wonder
what the label on the dummy knob said. Perhaps it might have even been engraved
with the builder's name.
Pipework
The pipework is unusual from the start in that both ranks
are metal: a Dulciana 8' and Principal 4'. The Dulciana has the usual wooden
bass of the period: large scaled, low cut-up and quinty. No identifying marks
were found on any of the pipes, not even on the seven zinc pipes of the
Dulciana (F18-B24). Early zinc often had an embossed stamp identifying the
(often French) manufacturer. The rest of the pipework is common metal. The
wooden basses were labeled in distinctive block lettering, with pencil, very
unlike the elegant old cursive one usually sees on 19th-century pipes. (I have
seen identical lettering on one other set of New England stopped basses which
the OCH found in an 1890s organ. The pipes were basses to a chimney flute, and
the entire stop had been completely reworked and re-scaled for its second use.
Alas, these pipes were also of unknown provenance).
I can find no rhyme or reason for the varying mouth widths
and variable scales. Surely part of the reason is that the common metal
pipework betrays the hand of a somewhat inexperienced pipemaker. While in
general neatly made, the solder seams are not as smooth and perfect as one
usually sees on 19th-century American pipework. It is particularly
disconcerting to see a pinhole of light shining through from the back of the
pipe when one is looking in through the mouth. These pinholes occur where the
back seam of the body meets the back seam of the foot at the languid, and are
present on several pipes. They did not particularly affect the pipes'
performance, so I left them. It does seem likely that scales were made
deliberately small in the tenor range of both ranks simply so that pipes could
be made to fit in the very cramped quarters. The very fat stopped wood basses
take up a huge amount of space, making it necessary to cram the metal pipes
into a very small area. Both ranks increase several scales in size from tenor
to treble: the Dulciana gets four scales larger, and the Principal increases by
three. (See pipe scale chart.)
From the chart, one can see that the cut-ups are all over
the map. The Principal seems to have a fairly even increase in cut-up toward
the treble, but the Dulciana seems to follow no discernible pattern. Mouth
widths are more predictable, generally hovering between 1/4 and 2/9.
The original pitch was fairly easy to ascertain. The pipes
seemed most comfortable speaking at 21/4"; at that pressure at 70 degrees,
the pitch was about A432. Since the whole point of this project was to make the
organ useful to an early music ensemble, the decision was made to fit tuning
sleeves carefully onto the pipes, and lower the pitch as much as possible. This
is a completely reversible procedure, with the added benefit being that it did
not require tampering with the tops of the pipes at all. The organ pitch is now
A421, not as low as the A415 the early music players had hoped for, but still
low enough that the instruments can tune to it easily.
One remarkable aspect of the tuning is that the Dulciana,
which showed no real signs of having been tampered with, was almost completely
in tune with the pipes at dead length and the few errant pipes brought into
regulation. A few chords quickly revealed that the keys of C, D, F and G were
close to pure, while the remote keys (B, F#, Db) were quite out of tune. This
sparked a lively discussion with Marian about temperament, and after some
research into early music temperaments (research done entirely by Marian) we
decided to tune the organ to Erlangen comma, which yields perfect thirds
between c and e, & d and f#. This temperament dates to the 15th century,
and is particularly suited to use with viols, avoiding the tuning conflicts which
mean-tone introduces between keyboard and viols.
Playing the organ is truly like stepping back in time;
voicing from this era demands less from each pipe than our modern ears
ordinarily expect. The gentle metal trebles in conjunction with the quinty wood
bass is a quintessentially early sound; virtually no one was still building
organs with that inimitable sound by 1860. Adding the small Principal 4' to the
Dulciana is an exercise in judicious restraint more than it is an augmentation
of the sound. All in all, it is an instrument from a different time and place,
built for sensibilities and perceptions unique to its milieu. Other than
changing the pitch, we did nothing to the instrument to make it more relevant
or modern. It so happens that leaving things as they were makes the organ
almost perfect for the customer's use. The subtle tone and slightly unsteady
wind work almost seamlessly with a small consort of viols da gamba. Placing the
instrument in a small room brings the sound into context, and music begins to
make sense on it. It is truly a chamber organ, and is at home in that
environment.
The author wishes to thank Barbara Owen for her gracious and
invaluable assistance in seeking the origins of this instrument; Marian
Moffett, for her research on a multiplicity of subjects; and Will Dunklin, for
his generous help in bringing the organ to Tennessee as well as for insightful
advice during the project.
Pipe scale chart
Principal 4' (labeled "Pr.") TC 42 pipes
Note Diameter
style='mso-tab-count:1'> Mouth
width Ratio
of mouth width Cut-up
style='mso-tab-count:1'> Ratio of cut-up
style="mso-spacerun: yes"> Toe size
C13 41m
style='mso-tab-count:1'> 29m
.225
7.8m
.190
style='mso-tab-count:1'> 3.98m
C25 22.5m style='mso-tab-count:1'> 18m .254 style='mso-tab-count:1'> 4.5m style='mso-tab-count:1'> .200 style='mso-tab-count:1'> 2.99m
C37 15.8m style='mso-tab-count:1'> 12m .241 style='mso-tab-count:1'> 3.0m style='mso-tab-count:1'> .189 style='mso-tab-count:1'> 2.28m
C49 10m
style='mso-tab-count:1'> 7.2m
.229
style='mso-tab-count:1'> 2.1m
style='mso-tab-count:1'> .210
style='mso-tab-count:1'> 2.03m
F54 7.5m
style='mso-tab-count:1'> 6m
style='mso-tab-count:1'> .254
1.9m
.253
style='mso-tab-count:1'> 1.77m
Dulciana (labeled "Dul") 54 pipes
C1 110x90m
90m
21.8m
.242
C13 64x52
52m 11.2m
.215
E17 55x43
43m 10m
.232
F18 58m
45m
style='mso-tab-count:1'> .246
11.8m
.203 6.09m
C25 42.7m
31m .231
style='mso-tab-count:1'> 7.5m
style='mso-tab-count:1'> .175
style='mso-tab-count:1'> 5m
C37 27.5m
21m .243
style='mso-tab-count:1'> 3.9m
style='mso-tab-count:1'> .141
style='mso-tab-count:1'> 3.04m
C49 17m
13.1m
.245 3.4m
style='mso-tab-count:1'> .200
style='mso-tab-count:1'> 2.71m
F54 13.5m
10m .235
style='mso-tab-count:1'> 2.5m
style='mso-tab-count:1'> .185
style='mso-tab-count:1'> 2.38m
The ratio of the mouth width is in relation to the
circumference: .250 would be 1/4 mw and so on. The ratio of the cut-up is a
simple ratio of the diameter.
