Our armchair, affectionately known as the “lollipop” chair, was made circa 1880. George Jakob Hunzinger (born 1835 in Tuttingen, Germany), was a progressive designer out of New York who was often influenced in his designs by machinery; their geometry and patterns of repetition in their elements. This is a Hunzinger original, a family piece, which has weathered more than a century of continuous use.
The chair needed re-upholstery, as the seat was failing. Our clients chose a leather similar to the original leather, which our client remembered.
There were issues with
the mortise and tenons at
all four leg stiles, front and back.
As is often done,
previous upholsterers
and repair people had
“made do” in their repairs,
using screws and and nailing
through tenons instead of
creating a proper repair.
EXCAVATION, a Slow Strip!
As always, we begin with a deliberate slow excavation.
Excavations are often rushed by inexperienced upholsterers,
but there is so much to be gleaned from paying attention
to everything as you take apart each layer.
The chair is turned upside down: webbing is removed,
and the springs are set free to bounce!
“The underside of a chair is often neglected
because it is the least seen and therefore often taken for granted!
It’s importance, of course, is the foundation for support and comfort of the sitter, but it is also, from the standpoint of the engineer, the focal point for the distribution of load.
Springs had not been used in chairs more than a few decades
before Hunzinger built his chair and the introduction of
their buoyant properties revolutionized upholstery! The introduction of springs to upholstery frames created a stress dynamic which most designers and upholsterers now take for granted, and those stresses from lateral motion of the spring under tension impact the way a frame performs, especially if careful thought has not been taken to include bracing: the frame must withstand the stresses.
Notice how modest the depth of the seat apron is
relative to the actual height of the coil spring.
The springs are four times the height of the seat frame!
Their coiled energy was expertly compressed within the modest seat frame,
attached between basket woven jute webbing and linen lashing cord.
This technique added comfort and versatility to objects created
for both an aristocratic class and a now burgeoning “middle” economic class
who desired both grace and comfort. This chair has given 150 years of comfort
and is just now begging for attention. Now it is important to make
thoughtful repairs to this modest frame so it can withstand another century,
or more, of spring compression through daily sittings” ~Mitchell
From the top, the current show cover is removed to expose the previous
show cover, a woven coral cotton-rayon from the early 20th century.
The “second stuffing” (term used in the industry) cotton topper and hair are lifted.
The hair is carefully vacuumed through a filter.
Most of the fiber will be reused when the re-upholstery is performed.
The primary seat stuffing foundation’s flax-straw pod stitching is released,
and the fiber pod is carefully set aside for conservation.
The canvas dust membrane which covers the jute webbing is vacuumed;
we find more information about the chair’s history.
Bits of green fiber beneath tack heads and embedded in the stuffings
show at least one previous show cover.
“Modern upholsterers used inappropriately large tacks,
which are seen next to the smaller tack, and this is a common mistake. The seat frame, as well as the legs, though they look to the casual eye to be crafted from mahogany, or walnut, are actually birch which makes both joinery and tacking surfaces less robust than most hardwoods.
The rabbeted tacking margins, including the underside of the seat frame, are peppered with excessive holes and splits from over-takings.
If not properly repaired, they will in time make upholstering impossible
without causing catastrophic damage to the frame.” ~Mitchell
The last bits, the springs, are dangling from the frame.
We measure and set aside, to begin to repair the frame.
REPARATION
SECOND ASSESSMENT
After excavation, we began a second thorough assessment, focusing on the frame and stuffing which we can now see.
The Lollipop chair was badly distressed from repeated poor repairs.
The barrel shape was originally created to have it’s tenons compress into the seat mortise without the aid of glues, screws or lag bolts, all of which were installed over the years, splitting the original tenons which required reparation.
To restore the historic chair to viability while remaining true to the
original decorative design intent required developing and applying
extraordinary woodworking strategies relative to structural repairs outside the upholstering effort, some of which will be described in text, or shown through step-by-step photographs of structural work performed below. ~Mitchell
DESIGN
The chair was originally designed with a modest seat apron height; in spite of this, it was fitted with tall medium gauge coil seat springs. The interior structural base and outer decorative seat frame and legs were created from Birch, a softer hardwood.
All leg mortise’ and tenon connections were badly bruised from compression wear, preventing the joinery from having tight connections. All this caused stability issues within the seat to leg structure and formed loose connecting points between the decorative rungs and seat.
I surmise an early 20th century upholsterer set lag bolts through the front corner blocks penetrating the front arm mortise and inside arm tenons in an attempt to mitigate the problem. This may have saved the front legs from breaking away from it’s original joinery, however, the repair person who introduced the lag screws failed to make repairs to the multiple abandoned leg stump mortises. Because of these issues lag repair and leg stump connections ultimately failed and damaged the front mortise and leg tenons.
This changed the order in which we upholstered and reassembled the chair. The back assembly (spindles and turnings) had to be properly affixed to the seat prior to upholstering, requiring us to carefully tack upholstery to the side rail tops. This made the process far more difficult, as we had to take care not to mar the back turnings.
Also, Huntzinger designed the barrel shaped chair seat with double continuous mortise’ which bored deeply into the side/rear seat apron, companion rear leg stump and adjoining rear barrel shaped seat apron. The long oak dowel tenons were softened inside a steam chamber then slipped into the continuous mortise’ while still pliable until all connecting points tightly aligned. The seat was then placed into a band style clamp and allowed to remain in stasis until the joints cured and stabilized. Once the moisture evaporated from the mortise and tenons the wood substrate shrank tightly, requiring no glue to maintain a tight connection.
Huntzinger’s modest seat design resulted in a flaw in the seat joinery engineering. The upper dowel to mortise connections are placed at one-eighth of an inch below the tacking surface, causing upholstery tacks to continuously penetrate through the mortise wall and into the dowel tenon, splintering the tacking surface and undermining the tenon to mortise connection. Decades and multiple upholsterings simply eroded the tacking surface, exposing the structural tenon to environmental issues and subsequent damage from poor repairs.
As these connecting point issues evolved, small losses turned into significant damage causing other major issues to develop. Most of the damage to the chair began within the seat and extended into other elements (legs, spindle connections, etc.) as the chair endured multiple damaging upholsterings and poor subsequent repairs. Over time the chair began to list while in use, causing the birch wood legs to move under stress whose forces then extended along unintended grain lines, eventually causing the rear legs to split at stress points. Over time both structure and upholstery became untenable making the chair nearly unusable without risking either the chair or one’s limbs during use. ~Mitchell
OVERVIEW OF DISASSEMBLY
Corner Blocks.
Disassembly of the frame began with the removal of the original corner block hardware. In general, after lag bolts and corner blocks are removed, above, warm water is introduced into mortise and tenons to allow the various glues to soften. When the glue softens, the frame top is gently wiggled until loose; the tenons can be pulled upward, removing them from seat frame.
Note regarding glues used throughout: Many eroded joints were amended with additional same species woods in order to mitigate gaps, but bruising was so severe that not all gaps could be dealt with through amending. I mixed a 50/50 slurry of hide glue from Patrick Edwards’ “Old Brown Glue” combined with Garret Wade’s PVA gap filling glue. This specialized glue amalgam was created and applied at room temperature, rather than warmed, in order to slow the curing process and assure the glue did not contract pulling away from coated joinery elements during cure. This insured good adhesion of the repaired parts, a reliably bulky and compression resistant consolidate filling minute gaps, while also maintaining the possibility of future reversibility.
LEFT-FACING FRONT LEG
An example of the damaged and carelessly placed nails which were carefully extracted from both the decorative left-facing front leg face and the historic blocks. Note the nail, previously carelessly driven through a finished leg in a haphazard repair, which we removed, image two.
Note: Each damaged part was filled or otherwise repaired in all cases prior to reassembly.
LEFT-FACING BACK LEG
The left facing rear leg shattered in the early 20th century and was poorly repaired with screws which penetrated the leg. The screws are still in the leg, countersunk beneath wood plugs. I removed the upper plug in order to remove the screw penetrating the seat joinery and inside back tenon, allowing us to remove the back from the seat frame! ~Mitchell
Note the split leg, repaired below.
The screws and badly placed reparation nails which secured the large rear corner block to the internal seat apron were removed and freed the rear apron joinery for disassembly. Once the corner block was removed and set aside, the extent of the seat tacking margin disintegration became apparent, along with the recognition of the necessity to amend portions of the element.
Finally the corner block is released
Above, the old reparation of the split left-facing leg with the exposed screws.
Two brass wood screws bisected the break on both the exterior and interior leg. These repair screws, though originally unsightly and poor substitutes for a traditional contextual repair, were successful in acting as structural tenons holding the leg together for several decades without concern.
Ultimately it was decided to leave the screws in place after treating the break for losses within the crack. It was determined the only other solution for repair was to re-create the historic leg and in so doing lose the authenticity of the aged finish patina.
RIGHT-FACING BACK LEG
As with the left facing rear leg’s blocking system the right-facing corner block was removed then set aside, thus freeing the entire rear joinery to slide open and separate for disassembly and restoration of joinery surfaces.
RIGHT-FACING FRONT LEG
The right-facing front leg was the last section of the chair seat to require disassembly and in many ways the most revealing relative to the structural dynamics because of the close intersecting points between the seat front, seat side and upper turning mortise proximity.
Note the split front apron requiring reparation.
As the front leg and apron connection was revealed inspection of the joinery made clear the extent of damage from abuse from poor repairs and the extent to which we must go relative to developing strategies for repairs which would not either compromise the elements for future use and repairs, nor impact the original design and aesthetic intent.
BACK REMOVED AND TENDONS REPAIRED
Unfortunately some images were lost in a file transfer of the removal of the wooden flush cut plugs and the countersunk screws holding the back onto the seat frame.
Back removed, and the left-facing rear tenon and right-facing front tenon
show extreme damage; replacement shown below.
The decorative back/arm unit was removed from the seat mortise connections and unit set out for assessment of the tenons/rungs.
The individual rungs with their decorative turnings were strung together like a necklace by the threading of two multi-strand copper cables at the crest and lumbar which terminated with countersunk brass screws which acted as decorative elements on the front of the arm-fronts, shown left.
The left-facing rear tenon and right-facing front tenon showed extreme damage from a century worth of losses in girth from compression within its seat mortise connection. As the tenons wore, the back unit became increasingly loose.
Unwise and damaging repair attempts exacerbated the loss in joint connection. At least two of the tenons had been shattered by the inclusion of the lag bolts which were installed to stabilized the seat, shown below.
Left-facing rear tendon. The lags penetrated the tenons, fragmenting the left facing rear tenon into two sections, thereby reducing the ability of that tenon to function as a reliable anchoring point.
I replaced this entire section of the left facing rear tenon, above, because it was not only fragmented by the lag bolt penetration, but also completely eroded from repeated upholstery tackings from above. As mentioned in the seat reparation section the original joinery design left little margin between the mortise and tenons relative to the decorative surfaces. Repeated expansion and contractions of wood substrate of the connecting joinery elements were prone to damage from not only repeated flex from load, but also from disturbances (tackings and repairs) which were attempted from the decorative surfaces.
The right-facing front leg tendon was also repaired, above.
The back/arms after reparation, above, and ready to be reattached to the seat frame.
LEG MORTISE AND APRON REPARATION
Once the accretions of numerous applications of various glues (original hide glue, injections of carpenters glue, etc.) within the joinery and poorly repaired breaks was removed from surfaces, we could begin repairs. The breaks within the legs were opened, glue was installed and the legs were placed in clamps to cure.
Irregular mortise bores were fitted with fresh doweling then re-bored to precise diameters. Areas of losses between the upper leg mortise and tacking surfaces were excised then fresh wood crafted to fit within the void. Once the glue within the fresh repairs cured we shaped the joinery to conform with the historic dimensions.
The left-facing rear leg which had been broken during the early 20th century then repaired using screws, was deemed supportable using the original screws. A decision was made to retain the historic leg rather than re-crafting the leg with fresh material since the repair was now considered historically contextual to the piece and the aged finish/patina of the leg to the surround was identical. Therefore, we removed the screws, opened the break, cleaned the wood substrate of old glue, set fresh pins into the screw bores, re-drilled the bores, then re-glued the leg break and re-installed the historic screws.
Huntzinger designed this barrel shaped seat with long dowel tenons which slid through the rear leg mortise, bending slightly to conform to the shape of the barrel which anchored well into the rear and side seat aprons. When the joinery components were seated the boring trajectory formed one long tunnel-like tenon to mortise connection. The slightly wider diameter (relative to the mortise diameter) dowels were cut to conform to the mortise length then placed in a steam box in order to make the wood bendable during assembly. Once the steamed dowels were thoroughly pliable the dowels were set into the leg stump and convex shaped seat mortises and the entire seat slipped together tightly as one unit.
Huntzinger chose a soft hardwood, Birch, as the primary lumber, not uncommon in the Depression era when designers were looking to find less costly woods with which to work. As seen in photos above, this became problematic because of the stresses upon the frame and the demand of multiple reupholsterings. This resulted in stress fractures, warping, twisting, and poor repairs by individuals without woodworking training. These repairs caused additional problems.
The eroded leg stumps were excised of debris, then dados were mortised into the leg stumps which bisected the original upper dowel mortises. Fresh wood was crafted to fit into the new crafted faceted voids and glued into place, then pinned. Once the glue cured the leg stumps were re-bored to configure to the original dowel mortise. The leg was once again viable at its connecting points.
“Modern upholsterers used inappropriately large tacks, a common mistake.
The entire frame was peppered with too many holes.
If not properly repaired, they will in time make upholstering impossible
without causing catastrophic damage to the frame.” ~Mitchell
The frame’s tack holes were repaired by inserting hardwood picks into the large tack holes after injecting warmed Old Brown Glue hide glue, shown above. The picks were allowed to set undisturbed for 48 hours in order to cure, then the picks were trimmed level with a chisel, and sanded smooth.
Corner blocks were installed, glued and clamped to cure for 24 hours, above. Reinstallation of screws were installed, and finally the internal seat frame retrofit was created and installed, above and right, then set to cure for 48 hours.
The internal seat frame retrofit created an internal structural dynamic which would allow add surface space where upholstery foundational materials (webbing, spring lashing, burlap, & stuffings) could be tacked.
This saved the historic tacking surfaces for the leather show-cover instalation only, and created a multi-faceted fulcrum which would prevent the frame and legs from future flexing thereby saving the elements from more damage during use.
The internal tacking rail would assure that the historic tacking rail could endure future upholsterings without invasive woodworking treatments. ~Mitchell
The restored joinery was coated with glue, and clamped to cure. The back of the frame was carefully reassembled.
FINISH
The chair before finish treatment, below right. We performed a modified “mechado” treatment to preserve and revive the historic finish.
HISTORIC FINISH OVERVIEW
“Mechado” in this case refers to a blending of compatible materials as well as a technique for preservation and restoration. Our intent is to preserve and revive the historic coating using a layering approach which when thoughtfully applied, in a circumspect order, will not only bind to the historic coating but also become an amalgam (combine as one) with the historic coating. Our layering “Mechado” approach is described below.
VARNISH FORMULA COATING APPLICATION
The original finish, shellac varnish over asphaltum oil glaze, was thoroughly cleaned using VM & P Naptha. Once the cleaning distillate had thoroughly evaporated the surfaces were gently skip sanded using a wet/dry lightly abrasive paper, concentration especially upon areas of odd accretions and losses in varnish elevation. The surfaces were again treated to a wipe-down using naptha then set aside for 24 hours to completely evaporate.
MPFC creates our own shellac which are true to original formulas, occasionally choosing to infuse the shellac with compatible tree resins which will add either stretch, or hardness, or both to the coating. For this chair’s base varnish we chose a blended shellac and tree resin varnish which we prepared, in-house, from a traditional recipe. The varnish was chosen for its ability to balance hardness with flexibility. Lab grade isopropyl alcohol was decanted into a wide-mouth glass jar while mixing into the distillate fresh beige toned shellac flakes, copal resin and larch sap turpentine.. These ingredients dissolved over the day and rendered a brush-able 1lb. cut solution of varnish. The fresh varnish was then brush applied to all surfaces, melding with the historic varnish and sealing the damaged varnish surfaces. The fresh shellac coating produced a reliable foundation for amending the damaged historic surfaces with blended waxes and resins after we performed selective infills into losses.
We allowed this coating to cure for several days.
TOUCH-UP AND COLOR INFILL
In order to blend distressed areas of color losses from careless usage and indiscriminate wear, MPFC created a repair, spirit- varnish (shellac varnish infused with earth based pigments which would infuse into shellac while remaining transparent and non-textural), which maintained a viscosity which could be manipulated with artist brushes in order to blend into historic surrounds. Once the infill varnish was satisfactorily applied and waiting a day for the material to fix to surfaces we re coated those areas of touch-up with the primary varnish in order to seal the pigments prior of the wax applications.
Hard carnauba wax infused with a small percentage earth based pigment, bees wax and tree resin was melted into our specialized applicator, then drizzled into areas of loss. Once the hot wax fills cooled they were selectively contoured then leveled to match surround.
Finally, a warmed wax slurry was applied over the varnished surfaces before the varnish had the opportunity to harden to the point of rejecting the wax infusion. As the wax solution spread over the surfaces they were assisted in flow by using over-large artist brushes which were dipped into odorless mineral spirits.
Once the surfaces would no longer absorb wax the finished elements were both buffed with soft brushes and rubbed with cotton diaper cloth and woolen rag.
The final polish was turned into a semi-gloss patina which allowed for normal wear anomalies to assert themselves visually while giving off a rich, well appointed and historically accurate patina.
UPHOLSTERY
Hunzinger’s original design allowed for the convenience of crafting the upholstery without the burden of working around the fixed points of the inside-back and interior arm frame. Unfortunately, the damage caused by previous unskilled repairs ended all possibility of recrafting the traditional upholstery with the decorative spindled back unit separated from the seat. MPFC had to devise strategies to make it possible to perform all the steps required during a traditional upholstery build-up.
Ultimately, after restoring the fiber filled seat pod, it was still challenging to easily tack the show-cover onto the side rails using traditional means (tack hammer and tacks) and so it was decided to secure the leather show-cover to the restored tacking margins using wide chisel pointed upholstery staples shot from a long nosed stapler.
Relative to this upholstery conservation/restoration project our decision was to make certain that the seat build-up could perform as it was originally intended while at the same time preserving/encasing the levels of the original stuffings within the restored set and in that way future generations, when uncovering the seat internals can see and identify the historic pod. To that end we begin the documentation of the phase of this multi pronged conservation effort. We will begin with:
THE HISTORIC FIBER POD
The historic seat pod was cleaned of dirt and debris using a vacuum with the suction level set low. To insure large particles and artifacts did not slip through during vacuuming cheese cloth was attached to the extraction wand. The cleaned fiber pod was set aside for re-installation, during the upholstery phase, after the frame and finish issues were treated and resolved. (below).
THE SECONDARY STUFFING & FIBER POD TOPPER
The seat hair and cotton batting secondary build-up (above) were inspected after vacuuming. It was decided that the two levels of cotton batting ,each representing different times when the seat was reupholstered, were far too damaged to be used as future pod toppers, but the horse mane pod secondary topper could be hand blocked and teased then amended with fresh horsehair when it was reinstalled.
THE NEW SEAT SPRING WEBBING FOUNDATION
Fresh four inch wide jute webbing was applied in the same configuration and position as the original webbing. We were able to establish the original width and position of the spring webbing during the woodworking restoration phase.
MPFC has created a tack and previous repair plotting system which we use on upholstered historic objects which maps and delineates the succession of upholstery. We plotted the tacking positions using clear sheet vinyl over the tacking margins both seat top and bottom. We identified all tack holes by identifying their positions using various colors of a Sharpie marker set upon the surface of a heavy mill clear vinyl. When the vinyl was removed from the tacking surfaces and set onto a white board the transparency allowed us to not only discern the positions of foundational and show cover tacks, but also to detect patterns in the tacking groupings which then allowed us to interpret the patterns as positions and with of webbing and clear understanding as to how many upholstering had taken place.
Additional data gleaned from the tacking surfaces were the style of tacks and fiber trapped beneath the tacks. This allowed us to determine the time frames when the chair was originally upholstered, reupholstered, and what types of show covers had been installed.
The original positioning and width of the tacking pattern also allowed us to determine what the original designer and upholsterer intended relative to the sit of the sprung-up seat, which also determined the intended center of seating gravity and the intended comfort level of the sit.
THE FRESH SPRING-UP AND WEBBING LASHINGS
Seat springs were temporarily clinched into into place over the webbing prior to lashing the springs to the webbing. Once the springs were lashed with twine to the webbing the metal clinches were removed.
Mitchell decided to use a “Number 4” configuration with his lashings, departing from the Holbein configuration.
TYING OF THE HISTORIC, SECOND GENERATION, COIL SEAT SPRINGS
An eight-way double course spring tie was utilized to achieve a stable seat. The second generation springs were reused in spite of their slight distortion from years of listing within the degrading seat primarily because they were of a light gauge no longer available in the height and diameter required for the modest seat footprint. Therefore, the spring-tie we chose included double knotted and overlapping courses of twines around all edge springs forming fulcrums to insure the old springs could not revert to their previous distortions.
THE FIRST SEAT BUILD-UP “First Stuffing”
The tied coil springs were covered with a doubled layer of hessian which was tacked to the retrofitted interior frame aprons. The springs were lashed to the hessian in a Holbein pattern using a waxed linen cord. Teased polished coir was lashed to the perimeter of the seat edge using bridal stitches in order to amend and rectify lost loft and compression in the original worn fiber pod. The cleaned historic fiber pod made of flax tow was hand blocked then placed over the amended seat deck and gently lashed into position around the interior seat perimeter.
The original horsehair second stuffing was hand blocked and set over the historic pod then attached with a running stitch.
.To encase the historic pod within the restored seat a final layer of fresh hessian was then stretched over the historic pod, edges under turned and tacked to the retrofitted tacking margins. A simple lashing was chosen to duplicate the original pod stitching with a simple perimeter roll and a finger stitched edge-roll which transited the front edge.
THE SECOND SEAT BUILD-UP “Second Stuffing”
A fresh layer of horse-hair was teased out on the restored pod then set under the bridal stitching.
The hair was then covered with a half thick layer of organic 50/50 cotton batting. Finally the second stuffing was encased beneath a 400ct cotton percale sheeting muslin, left, then tacked to the conserved tacking margins.
The seat was then ready for the final show-cover upholstery.
THE SHOWCOVER
A “pull-up” style aniline dyed two part colored hide with a waxed surface was chosen to mimic hides from the chair’s decorative period. A grain structure within the hide was chosen to reflect longevity and aesthetic appeal; the leather was cut, image #1 above.
A light layer of organic cotton batting was placed over the percale muslin to act as additional loft and create a buffer between the cloth and the back of the leather, image #2.
The leather was uniformly tacked to the restored tacking margins, above. A straight vertical pleat was chosen to install at the two front edge corners to reflect the typical single vertical pleating from the chairs decorative period. A thin coating of colored shellac was brush coated to the overlapping muslin to leather tacking edge (no image) in order to assure that any bleed or distortions to the upper leather decorative tape could not reveal materials beneath the tape.
A single line of leather was cut and skived to conform to the tacking edges and aged brass decorative nails were installed through the leather tape and into the tacking margin.
THE DUST COVER
Finally, a medium weight polished cotton twill was tacked to the conserved frame underside and the upholstery phase of the chair’s restoration was complete.
A Slideshow of the Chair Restoration and Upholstering:
Slideshow of the Upholstery Phase, above.
The Lollipop Chair completed!!
The Lollipop Chair is Complete!
MPF Conservation (MPFC) is located in Portland, Oregon, USA.
We treat objects all over the Pacific Northwest, down into California,
and west to Idaho, Montana and Nevada.
dkatiepowell [@} aol.com / mitchellrpowell [@] aol.com
503.970.2509 / 541.531.2383
©MPF Conservation. May be printed for your own use.
Notify if you repost: use our url + copyright is used as reference.
