So my first challenge was to separate the shards, the process of which was quite like dealing with a puzzle. Once all seventeen pieces were found, the object began to fully reveal what remained of its original shape and appearance (See Fig. 2).

It is a Famillé Rose dish with butterfly motifs on yellow ground with a Guangxu Reign mark, suggesting the production date between 1875-1908 in the Qing Dynasty (1644-1911).

On the interior wall of the dish lie 40 butterflies in a circular and symmetrical arrangement as overglaze enamel decoration. The rim is gilded. There is also gilding on the butterflies. The exterior wall is decorated with three groups of famillé rose floral sprays (peony, lotus, passion flower) on a white surface. The base is inscribed in red with the six-character reign mark “大清光绪年制 (Made in the Guangxu reign of the Great Qing Dynasty)” in regular script in two columns.

The term Famillé Rose was first introduced by French author and art historian Albert Jacquemart (1808-1875). As the name suggests, Famillé Rose decoration refers to the distinctive use of pink or rose as part of a new colour palette in porcelain production. These colours appear as polychrome over-glaze enamel decoration on porcelain, the first of which was produced during the late Kangxi reign (1661-1722). It is characterized by a colour-wash effect, which allows graduations and intensities of colours to happen which makes the subject more three-dimensional.

Luckily, I also found some similar objects in the collection of the Beijing Palace Museum. These objects were all produced during the Tongzhi Reign (1861-1875) in Jingdezhen. Based on their similarities, it is reasonable to believe that my object also shares the same manufacturing origins with the examples in the previous reign. This made me feel even more excited to have this as my first object!

Cleaning 

This project provided a really good exercise for cleaning. The object has been repaired before, on some leading edges there was brownish material, which was possibly the adhesive residue from the previous repair. Besides general surface dirt, there were multiple areas of surface staining, which might also be caused during the previous repair, as they shared the same colour with the material on the leading edges.

Firstly, I cleaned the surface with a soft brush to remove the loose dirt on the surface. Then I carried out series of solvent tests with cotton swabs to find the most appropriate solvent to remove the residue along the breakedges as well as the surface dirt and staining. It turned out that the brownish material could be removed by room temperature deionized water, while warm water could work more efficiently. Thus after I removed most of the previous material by swab I then used a steam-cleaner, which could provide water vapour with high temperature and pressure at the same time, to further clean the leading edges. The technique should be carefully performed with appropriate PPE (personal protective equipment). This technique was only applied on leading edges because it might be too strong for the enamel decoration and the gilding, which are usually fired at a lower temperature than the porcelain body and thus are more susceptible to physical damage. The angle at which the pressure is applied is also important to protect fragile areas such as delicate over-glaze enamels and potential running cracks.

During the cleaning process, I also found that several bonds from the previous repair were still holding the shards together. As the adhesive material might further deteriorate, and the colour was unsettling to the aesthetic appearance of the object, I decided to dismantle the bond. I made a paper poultice with warm water, and then applied this around the shard, initially covered with cling film. After an hour and a half, when I uncovered the poultice, the bond had already been dismantled, and the pieces were separating easily.

After all shards were steam-cleaned, I mapped out a bonding plan and did a dry run formation of the shards supported with tape. This process could also help to check the cleanliness of the leading edges, as any dirt or stain on the edge would show a yellowish or greyish colour along the break line, especially on the white background such as the exterior surface of this object. Thus it is important to ensure that the leading edges were thoroughly cleaned before bonding, as any adhesive applied could only strengthen the colour caused by dirt or staining that remained.

I found that there were still several leading edges that were not clean enough, so I used a commercial detergent powder that works based on enzymes. The powder was dissolved in warm water and applied on those leading edges through poulticing with cotton swabs for twenty minutes. I avoided contact between the poultice and any gilding. The leading edges were then steam-cleaned and, once dry, the shards were ready for bonding! (See Fig. 4)

Bonding

As this object is hardpaste porcelain with a highly-vitrified body, I chose to use an Epoxy resin to bond the pieces through capillary action, as it is strong enough to hold the pieces together and chemically stable to support a long-term treatment outcome. The pieces were first taped together tightly. It was challenging to guarantee perfect alignment when there were so many shards and small areas of contact as the body was very thin. Even tiny misalignment would affect the overall outcome, which would be especially obvious in the rim area. After all pieces were taped together, and the alignment was checked, the epoxy resin was dotted along the bond lines with a cocktail stick and travelled into the bond lines through capillary action. Areas with enamel and gilded decoration were avoided, as the removal of any cured excess resin may cause physical damage to these fragile areas. The whole object was then left undisturbed for two days to let the epoxy resin cure before the tapes were removed.

Here is the current outcome (See Fig. 5). 

As mentioned at the beginning, the project was left at this stage, and remains on-going. Still, I’ve learned a lot through the experience so far, including various cleaning techniques and bonding methods. The next stage will be to prevent future access of dirt into the bond lines as well as to better conceal the bond lines on the white background using the colour-filling technique. I can’t wait to work with the object again and help it to fully resemble its original appearance. Once the project is completed, there will be another post to follow the rest of the treatment, and I hope that you won’t need to wait too long for that!