Most if not all of the Delft-tiled fireplaces at Winterthur are probably not original to the house. The Museum's Registration Department holds copies of Henry Francis du Pont's daybooks, journals where he meticulously recorded each and every antique and piece of art that he purchased. By comparing entries in his daybook to receipts from antique dealers, I can hypothesize that du Pont bought all of the tiles in his collection sometime between 1920 and 1940. These dates also correspond to when many of the rooms in the house with tiled fireplaces were installed.
Tiles are traditionally attached to fireplaces using a combination of a mortar (a thin sort of cement) and lime plaster. However, when the Winterthur tiled fireplaces were installed from the late 1920s to the early 1940s, builders bypassed the mortar, instead using gypsum plaster and a creative variety of "modern" materials.
The rather culturally insensitive Chinoiserie tiles from the 1750s in the Bertrand Room were installed using Portland cement in the 1930s. According to tile expert, Lesley Durbin of The Jackfield Conservation Studio, Portland cement is an inappropriate material to install Delft tiles. By their composition, tin-glazed earthenwares like Delft tiles are susceptible to the efflorescence of soluble salts. Soluble salts are normally held in the ceramic body, but when exposed to high humidity, they crystalize on the surface, damaging tiles and causing glaze to spall off. Portland cement also is much harder than tin-glazed ceramics, making the prospect of ever removing tiles from this fireplace daunting.
The tiles in the Patuxent fireplace surround were probably attached with DUCO® cement is a fast-drying, cellulose nitrate-based commercial cement. When viewed under long wave ultraviolet light, the grout and mortar fluoresce a light greenish yellow, a classic indicator of cellulose nitrate. The adhesive tends to discolor overtime, giving the grout between these tiles a dark brown stained color.
The front of this innocuous-looking tile masks a mysterious secret. When flipped over....
...whatever this is is revealed. This tile is in storage at Winterthur-- not mounted in a fireplace--but this tile was too crazy to not include in this post. It was coated in a cellulose nitrate adhesive, then a black resin, and finally 16 adhesive tabs were attached over top--all of which retain their plastic barrier layers. Why and when this was applied remains a mystery...
Exploratory Excavation of Vauxhall Fireplace
Since building records are eluding me at present, the only way to definitively know how the fireplaces were constructed is to perform exploratory excavations. Vauxhall fireplace on the fourth floor of the museum was damaged in a flood in the 1980s. While it appears structurally stable, the plaster surrounds are powdery and delaminating (below, left). The tiles show evidence of soluble salt damage such as cracking and spalling (below, right). Chemical spot tests reveal that sulphates, common in gypsum plaster (hydrated calcium sulphate, CaSO4•H2O), are the probable culprit.
Part of my fellowship project may include dismantling the tiles in Vauxhall fireplace. This would allow me to treat the tiles in the conservation lab (much more ergonomic working conditions as you shall see). In order to conclusively determine how the tiles were mounted in the fireplace, my supervisor, Associate Objects Conservator and Assistant Affiliated Professor, Lauren Fair and I decided to remove one tile from the lower proper right corner of the fireplace.
During my internship year, I participated in a project at the British Museum removing Medieval floor tiles from a panel (above). Based on my experience, Lauren and I decided that using a hammer and chisel was the best course of action to remove tiles from the Vauxhall fireplace.
After a two-hour delay due to snow, we began the removal process! To protect ourselves from the plaster dust and chips, we wore dust masks, protective goggles, and nitrile gloves and vacuumed dust and debris as we went to minimize disruption to the rest of the objects in the room. To remove the tile, we decided to cut a channel behind the lower tile using a stone-carving hammer and chisel. This was accomplished by sitting or lying on the floor and crouching to strike the plaster at the proper height. Because the tiles are so tightly stacked, the only way to access the tile was from the proper left side. When we finally chipped most of the plaster away, we realized that the tile above had also become detached.
To Deinstall or not to Deinstall...
When the tiles were removed, we could clearly see that the fireplace had been constructed of brick covered in about an inch of plaster (above, center). The tiles were directly stuck into the plaster without a barrier to protect them from the movement of soluble salts. We still haven't decided whether or not we're going to remove all of the tiles to conserve them. A further consideration of the risks and rewards of deinstall needs to be undertaken. Stay tuned!
Thanks for checking back! Be sure to check Twitter for updates under #WeirdTileoftheDay and #WeirdTileWednesday. Tune in on March 15th as I delve into my struggle to find where the tiles came from!
For more information see:
Durbin, Lesley. 18th Century Delft Tiles: English and Dutch Tin Glazed Tiles Circa 1650-1790. http://www.jackfieldconservation.co.uk/18th-century-delft-tiles/
Photos by author unless otherwise stated.
On my first day at Winterthur, I met with Senior Curator of Ceramics and Glass, Leslie Grigby to see the unmounted tiles in the museum's storage areas. 160 of the 515 tiles in the collection are kept in storage and 22 are displayed in the Ceramics and Glass Study Area (below). A major component of my project at Winterthur involves conducting a condition survey of all of these tiles.
Where to Begin?
Before I went into the collection to start my survey, I had to come up with a way to record my data. I decided to create a survey spreadsheet on Microsoft Excel rather than using a paper form. This way I could take my laptop into the collections and enter data right away rather than having to type up my observations later. The information I collected would be used to prioritize my conservation treatments.
One of the major challenges in creating a database or survey form is standardizing the nomenclature, or wording that you use. Consistency makes information much easier to find and sort during data review. To this end, I made specific categories and wrote in set terms to organize the information I wanted to collect.
I placed an image of each tile in my spreadsheet in order to make sure I was assessing the correct tile. The next field is "Object ID," where I placed Winterthur's accession number, for example 1969.4732.005. An accession number identifies each individual object in the collection and relates it to similar objects based on the year they were collection and their “accession group number.” For example, a group of objects donated by the same person at the same time would be part of the same accession group.
"Other Number" records any number on the back of tiles from other cataloging systems. "Condition of Tile" is a free text field in which I recorded any condition issues with the tile, such as structural instabilities and historic repairs. "Condition of Mortar/Mounts" and "Condition of Fireplaces" applies more to the mounted tiles in the house. "Priority" sorts the tiles by minor to major condition problems and will help me prioritize which tiles to treat first. "Maximum Dimensions" gives the measurements of each tile. "Current Location" shows where the tile is located in the collection.
A “Priorities” field in a condition survey allows conservators to easily rank objects by conservation needs. For example in my survey, I sorted the tiles into four groups:
Low Priority: treatment not necessary, minor aesthetic repairs, dirty surface, structurally stable
Medium-Low Priority: treatment suggested, more significant aesthetic repairs, chips, small cracks
Medium-High Priority: treatment recommended, significant aesthetic repairs, structural cracks, spalling, chips, large areas of loss
High Priority: treatment necessary, ceramics is actively deteriorating, extremely disfigured by historic restorations, or unstable
A majority of tiles will probably fall into Medium-Low or Medium-High priority.
Organizing the Tiles
The first thing I noticed when I got into the collection was that the tiles were stored and stacked seemingly randomly. Because my spreadsheet was sorted by object number, step one involved sorting the tiles by number.
Once I finished sorting the tiles, I was ready to start my survey! Tune in next week as I talk about some of the major condition issues I noticed as I began to look at the individual tiles more closely.
Watch this blog for weekly or biweekly updates and follow me on Twitter to learn more about some of the individual tiles I’ll be looking at on #WeirdTileWednesday and #WeirdTileoftheDay.
Winterthur Postgraduate Fellow in Objects Conservation