Washington Place
Native woods inlaid table attributed to Wilhelm Fischer

320 South Beretania Street
Honolulu, Hawai'i 96813

Wood is koa (leg, column and skirt) with a 496-piece inlaid top made of 12 different native Hawai'ian woods. Secondary wood of top underside is Douglas fir and another unidentified wood.

BASE: One of the three legs was broken apart at leg-to-column joint; two dowels used attaching leg-to-column. Glue residues (excessive) from an apparent earlier repair were on surfaces of finish around joint. Finish on legs and column was stable and composed of orange shellac. 

Joint disassembled.

Above: Separation of leg from column.

TOP: Top finished with what appeared to be a thick layer of blonde shellac or other resin which had delaminated over much of the central area of inlays and along many of the glue lines.  Some areas of loss to the finish had occurred as well as staining into the woods around losses.  Several pieces of the inlaid woods were completely loose and could be lifted out from the table surface.

Presence of staining, delamination of finish as well as the lack of adhesive residue traces on mating surfaces of the inlays would indicate repeated and/or prolonged exposure to water.

A faint pattern of what is commonly referred to as "orange peel" could be seen over the entire top. This would strongly indicate that the existing finish was applied by spraying. The top may have been polished or compound buffed to lessen the orange peel defect. There were brush or rag marks in the finish on the skirt, indicating that finish on this area was applied by hand.

It was apparent that the entire piece had been heavily “polished” with silicone containing waxes which considerably complicated finish work.

Center of top. Delamination of finish along joints and lifted-out/separated piece of inlay.

BASE: Top disassembled from base and leg from separated column.  Joint surfaces cleaned of adhesive residues. Dowel holes were cracked/enlarged and were rebuilt as necessary with a phenolic and silica microballoon filled epoxy mixture.  Dowels were cleaned and reinstalled; all reassembled with hot hide glue.

TOP: In order that as much silicone be removed as practical before other treatments were initiated, a poultice comprised of silica microballoons/amorphous silica/hexane/xylene was applied to the top and apron. Contaminants were absorbed into the poultice layer and were then removed by vacuuming. The process was repeated until no noticeable residues remained in the poultice after it had dried. This procedure enabled partial extraction of the silicone and wax residues, but did not harm the finish.  The finish was first tested to determine type and degree of sensitivity towards various solvents.

After poultice cleaning, degree of cleavage in finish layers more clearly evident.

Once the surface had been cleaned, loose inlays were reset with hot hide glue, consolidation of the loose finish was done with solvents and Acryloid B-72/shellac; areas of loss were progressively built-up/infilled with same.

The finish on the table top was unusually thick. Direct measurement of the finish thickness was not possible without a (destructive) core window being made. Indirect approximation was be made by observing through a 10X loop, the micro fissures that had formed above many of joinery lines of the underlaying wood inlays. The micro fissures emanated from the wood surface along joined sections and generally did not project through to the surface. Where they did, they appeared as distinct cracks in the surface. Film thickness along the fully and partially fractured lines appeared to be approximately 2+ mm thick.

The micro fissures were a result of the finish film failing to respond to movement in the underlaying wood pieces. As the woods are all dissimilar to one another radially, and are concentrically arranged, this type of failure is to be expected in any but the thinnest and most flexible of finishes. In order to lessen this problem, a decision was made by curation to allow the total film thickness to be reduced by means of wet (hexane/mineral spirits) sanding. Sanding dust from finish dissolved at RT mostly but incompletely in ethanol, and not perceptibly in hexane or mineral spirits.

This process decreased the total finish thickness to a more aesthetically acceptable degree but at the same time exposed more of the fissures across the top.

While sanding finish to reduce thickness, no interface or boundary layers were observed at any time. Nor was any such observation made when subsequently applying softening solvents prior to French polishing. This would strongly indicate that the finish was applied in one course, using the same finish material throughout.

Restoration story from Honolulu Advertiser, 2000

Restoration story from Star Bulletin, 2001

Washington Place Web Site


Email Minnick Associates

Conservation of Historic Furniture & Objects
Copyright © 2000 Minnick Associates