Conservation of an anatomical model of a horse by dr. Auzoux

In recent months I have been working on a special project. I conserved a paper-mache anatomical model of a horse from the 19th century. The model is manufactured by Dr. Auzoux. He had a factory in France where he made anatomical models of paper-mache from both humans and animals for university education.

I carried out the conservation together with some wonderfull colleagues in my studio in Leiden, what a pleasure to work with them: Claire Phan Tan Luu, Ilse Korthagen and Francoise Richard.

I have worked on many other anatomical models by Auzoux, such as the gorilla that can be seen in Rijksmuseum Boerhaave in Leiden.

Before treatment. The horse is approximately 70% life size.

This model of a horse comes from the collection of the Ghent University Museum (GUM, Belgium) that will open (postponed for now) with a completely new exhibition space where the horse will have a central place.

The construction of the horse

The top half of the body can be opened, both by hinges at the tail and by the possibility to detach it completely from the bottom half. The organs in the abdomen become visible. These organs can be individually removed from the abdomen. The left foreleg and head are also removable. The horse is standing on a wooden platform with wheels. The legs are attached to the platform with large bolts and nuts.

The model is built around an iron armature. The fixture is surrounded by gray paper pulp, which is covered with a number of layers of paper pasted into strips. The papier-mâché was made from torn paper, starch, chalk, hemp, and cork powder. This paste was pressed and dried in moulds. The surface of the parts was treated with layers of pigments and as a binder fish glue. The amount of pigment per layer is low. The many layers of pigment and fish glue ensure the natural reproduction of flesh, muscles and veins. The veins are made of iron wire wound with flax fiber and painted paper strips.

The parts are fitted with hook and eye closures and with pins and holes. All parts are numbered with numbers for the nomenclature which is mentioned in a separate printed booklet, ’the tableau synoptique’ (not available with this model). Most Auzoux models are also numbered by labels with indicator hands. The numbers with the hands indicate in which order and where the model should be taken apart.

The horse consists out of 14 removable parts


The entire model was covered with a thick layer of dirt and dust. The surface layers on the outside of the model were in poor condition. There was very strong flaking (small flakes) of the paint layers and there were spots where the paint layer had completely disappeared. The veins on the outside of the abdomen were affected and some of them had become partially detached from the surface. The head lacks much of the paint layer, including the right eye. The abdominal cavity and organs were also very dusty and dirty. It seemed that the model had either been open for a long time, or that the organs had been shown individually.


A vacuum cleaner with a small attachment and a brush was used to remove dust and other dirt that had accumulated on the surface over time. This removed most of the loose dust and dirt. Further surface cleaning was done with sponges and brushes with ice water. The ice water was used to clean the surface without dissolving the gelatin and varnish layers. Wet cleaning removed the dirt from the top layers and returned the beautiful “egg” shine of the surface. This method only worked well for the outside of the model. Cleaning with only water was not enough to clean the intestines. The surface remained very dirty after cleaning with water. We looked at other ways of cleaning. The best and most effective result was cleaning with ethanol (alcohol) in a gel solution (Klucel G) so that the alcohol could be applied in a targeted way and not evaporated immediately. This method of cleaning yielded a good result.

While cleaning with ice water, the paint flakes softened slightly and became soft. This was a good time to proceed immediately with the consolidation of the flakes. For consolidation, the surface was brushed with an adhesive mix of EvaconR (ethyl vinyl acetate) and MC (methyl cellulose). The moisture from the glue softened the flakes even more and the glue itself caused the flakes to stick directly back to the surface. The glue took about 1-2 minutes to work. After this, the excess glue was removed with damp sponges. The surface was pressed directly with non-stick paper so that the flakes could make good contact with the surface. The paper was pressed by pressing with a (natural)sponge or, in some cases, by pressing the flakes with a bone folder.

The paint layers were missing in quite a few places, the paper layers underneath were completely visible. Some of these spots are filled with paper pulp and finished with a layer of Japanese paper. Some spots are only finished with coloured Japanese paper. The places where paper pulp has been applied have been retouched with watercolour (pencil and paint) so that they are less noticeable.

Before and after treatment

Geef een antwoord

Het e-mailadres wordt niet gepubliceerd.

Deze site gebruikt Akismet om spam te verminderen. Bekijk hoe je reactie-gegevens worden verwerkt.