The layers of paper were first of all dusted off using a soft brush, then gently wiped with a latex sponge (which picks up dirt without water). Next the surface was cleaned using moist Japanese vellum compresses and/or cotton swabs.

Various aqueous gels were used for cleaning :

• methylcellulose, Benecel A₄C and Methocel A₄M (Kremer) 0.5-2% in demineralised water
• Funori (Kremer) 0.5-1% in demineralised water
• Funori and methylcellulose mixtures

The bits of paper that had come unstuck were re-glued with methylcellulose, Methocel A4M or Benecel A4C 0.5-3%, depending on the required viscosity.

The unvarnished areas of paper were singled out using methylcelluose and then protected with a thin coat of dammar varnish dissolved in sangayol (Kremer).

Next, the missing spots were touched up with water colours, gouache (Schmincke) or dammar and dry pigments (Kremer).
In the places where there was no more paper on the spheres, the gaps were first of all filled with edible gelatine sealant (8% in demineralised water) laden with Champagne chalk (6 parts), Bologna chalk (4 parts) and kaolin (1 part). These fillings were then touched up the same way, first with water colours and/or gouache then with pigments and dammar resin. Sometimes Mussini colours (Deffner and Johann) were also used, made of finely ground resin-oil-based pigments, but only on the sealed spots.

The stands were first of all dusted off and then cleaned with a Wishab sponge and lukewarm demineralised water. The badly soiled areas (feet and compasses) were cleaned using cotton swabs.

The large nicks and the particularly visible smaller ones were touched up with gouache, dry pigments and dammar pigments dissolved in sangayol. The worn black edges were also lightly touched up with the same colours.

To reposition the spheres as best as possible in relation to the horizon rings, small pieces of wood were inserted in the notches for the meridians, thereby forcing the spheres towards the side opposite the rubbing. Unfortunately they are still off-centre in the horizon rings, so considerable care is needed when they are rotated. The meridians can moreover be pushed manually away from the area of rubbing during rotation.

The terrestrial globe was raised on its metal shaft using a small aluminium ring about 3 mm high. This ring, like the one on the celestial globe, was coloured with dammar and pigments.

To position the terrestrial globe better, the small cylindrical support holding up the meridian was also corrected by sawing 5 mm into one side of the notch and compensating on the other side with a small piece of balsa wood coloured with brown gouache.

The holes left by the samples taken for the C14 tests were sealed and touched up, but a small cavity was deliberately left on the terrestrial globe.

Similarly, the hole in which the printing press character was discovered (just below the equator in present-day Gabon at longitude 40°) was left unsealed. It was simply covered over with coloured and varnished Japanese vellum.

Overall construction

The sphere of the terrestrial globe has slipped downwards slightly on its metal shaft; leaving it off-centre in relation to the wooden horizon ring. The celestial globe suffers from the same problem but its position was corrected in an earlier restoration by adding a small metal ring to hold it higher. Even so the spheres are not centred, and because of this they rest on the wooden ring and rub against it.

The tip of the terrestrial globe’s metal shaft is broken.

The meridians are oxidised and stained. Moreover, in an earlier restoration they were varnished (mainly the edge) very unevenly, causing runs.

Gores

The paper strips and calottes that cover the surface of the spheres are peeling slightly at the joints, probably owing to strong tension within the layers of varnish. The detached paper has been weakened and bits are missing in these areas, exposing the preparatory chalk layer underneath. The surface also shows a number of scratches, most of which have been caused by sliding the runner along the latitudes.

The colours have deteriorated; they have faded and the greens have often turned yellow-brownish. This wear could be due to the globes’ use over time and to their natural degeneration, but it may also have been exacerbated by cleaning the varnish in an earlier restoration.

Surfaces

The outer varnish (not original) is in rather good condition but oxidised, and has a marked yellow appearance. Moreover, in some areas the surface is worn and the varnish has all but disappeared. In these spots the surface of the paper, no longer protected, is badly soiled. The varnish also has also lost its gloss in places.

The remains of older varnish are very heavily oxidised, forming dark stains where the paper is pitted.

Paper layer

The paper strips glued to the horizon rings are badly damaged, far more than the gores covering the spheres. This is due to their horizontal disposition, which causes an accumulation of grime. Moreover, these areas have probably been touched considerably during the globes’ use.

In particular, the paper has peeled and crinkled along the edges. There are tears and numerous pieces missing, as large as 6 cm2. The printed surfaces are moreover so worn and chafed in places that the ink has disappeared and only a very thin layer of paper remains.

Water damage has displaced the dirt in some places and left moisture stains.

Surfaces

The varnish on the rings is heavily oxidised and has turned brown (much more than on the spheres). The surface of the rings is moreover very dirty, particularly on the unvarnished band where the accumulation of grime in the paper fibres has often made the printing illegible. The layer is so thick that it has come off in places, taking the surface of the paper with it.

There are drops of white paint on the surfaces, especially that of the celestial globe.

Wood

The edges of the wooden rings are worn. This wear runs deeper in some places, and there are also nicks and cracks. Some of the cracks running cross-wise have been closed using small nails (not original). For the moment they do not pose any risk to the rings’ stability.
The dove-tailing is no longer even and can often be seen through the paper layer.

The rings, like the stands, have been taken off and put back on in the past. A dove-tail on the celestial globe was replaced by a piece of metal (not original). The moulded rim was painted over with opaque black paint that sometimes covers over earlier wear of the wood.

The black moulded rim is spotted with traces of white and turquoise paint.

In an earlier restoration, the stands were taken apart and then reassembled using glue, dowels, nails, screws and pieces of metal. While the stands were disassembled, the positions of the balusters were marked and numbered from 1 to 4 with pink chalk.

The dowels holding the balusters to the horizon ring were replaced and the joints re-glued. This restoration showed little regard for the original assemblies and the pieces were glued clumsily. Beads of glue spill over from the joints and the pieces are not always flush (e.g. the planks of the circular base are slightly misaligned). The black parts of the stand were likewise coarsely painted over. It is difficult to know what colour these parts were originally, but that coat of paint was probably much thinner and applied more thoroughly than the later one.

The stands also show natural wear, particularly along the edges and in the corners. Some of this wear was painted over in black.
The paper layer with the red-green markings has worn very thin and in some places right through to the wood.

The needles of the compasses have disappeared along with the glass that probably covered them.

On the circular base of the celestial globe, there are remains of glue from an earlier label.