Restoring a natural skeleton of a premature infant from the early twentieth century

Abstract

Traditional anatomical preparations as they are often found in university collections are both historical treasure and moral responsibility of the hosting institutions. These remains not only represent the tradition and history of anatomical science and education, but above all are the remains of human beings and thus should be treated with dignity. While the discussion on public display of historical human remains is ongoing, the responsibility to keep them in an optimal state of preservation is not in question. However, most time-honored anatomical collections do or did not have the financial and/or technical capacity to keep these high standards, underlining the need for exchange of technical know-how between anatomical institutes and the support of restoration/conservation professionals. Here, we present a state-of-the-art conservation and restoration of a fetal natural skeleton from the early twentieth century along with the professional documentation of restoration-measures. Therefore, we used modern photo- and video documentation, including photogrammetry, as well as UV- and x-ray-examinations to record historical and new restoration procedures. We carried out measures to reconstitute the structural stability of the skeleton. Moreover, we replaced lost parts of the left foot skeleton, and critically discuss the legitimacy of body part replacement in historical specimens. We are convinced, that similar cases of insufficiently preserved specimens can be found in most anatomical collections. Therefore, this article is not only intended to document the state-of-the-art conservation and restoration of a natural skeleton, but also serve as an inspiration for similar campaigns in other institutions in the future.

Introduction

Anatomical and pathological collections, especially in Europe and the USA, keep large numbers of human remains from various historical and medical contexts, often including dry preparations of human skeletal material. Most commonly, skeletal specimens are mounted skeletons (Sceleton artificiale), in which the original bones are fixed by metal screws and wires. Natural skeletons (Sceleton naturale), in the contrary, are skeletons, in which the organic components of the passive locomotor apparatus (i.e., cartilage, ligaments, and joint capsules) are preserved, in an attempt to keep bones and joints in a lifelike topography—a technique particularly often employed for the preparation of fetal and children skeletons.

The preparation of natural skeletons can be traced back to the beginning of the eighteenth century. Thus, the Dutch anatomist Frederik Ruysch (1638–1731) prepared fetal natural skeletons for his vanitas dioramas [1]. A small number of these fetal natural skeletons attributed to Ruysch are still preserved today and are exhibited in the Kunstkamera in Saint Petersburg, Russia. Other early skeletons of this kind are preserved in the Beaux-Arts de Paris in France—orchestrated in the so-called ’’Autel Macabre", which was created in the second half of the eighteenth century [2]. Both, Ruysch's dioramas and the ‘‘Autel Macabre’’, have in common that they are more likely to have served a theatrical staging than strictly scientific purposes. In contrast, numerous anatomical and teratological fetal, and infant natural skeletons can be seen for example in the Meckel collections of the University of Halle-Wittenberg, Germany, some of which can be dated to the late 18th and early nineteenth centuries [3]. These anatomical skeletons were used to demonstrate osteogeny and ossification, i.e. they were used for teaching purposes [3, 4].

An important reference to the scientific and didactic use of anatomical fetal natural skeletons can also be found in historical literature. August Carl Bock (1782–1833) wrote in 1829: "Therefore, neither male nor female skeletons, […], nor children's skeletons, from the tenderest age to the later stages of infancy, may be lacking in the anatomical teaching institutions, because of the gradual bone formation, to complete the illustrative presentation." (“Daher auch in den anatomischen Unterrichtsanstalten weder männliche noch weibliche Gerippe, […], noch Kindergerippe, von dem zartesten Alter bis zu den späteren Stadien des Kindesalters, wegen der allmälichen Knochenausbildung, zur Vervollständigung der anschaulichen Darstellung nicht mangeln dürfen.“, [5]; translation by the authors). Natural skeletons were therefore used in series to illustrate developmental and ontogenetic dynamics [4, 6]. Some larger series, with up to 20 individuals, are still preserved in anatomical collections, e.g. in Pisa, Florence, Montpellier, Halle-Wittenberg, and Greifswald.

Despite their unbroken value for teaching purposes, natural skeletons became increasingly irrelevant for medical research over the course of the second half of the nineteenth century and eventually were considered of little value for more specific anatomical analyses. This is nicely illustrated by Hugo Burtscher in his dissertation in 1877 [7]. Burtscher investigated the exact measurement of the bones and ossification centers of the extremities at different stages of fetal development. In the last paragraph of his preamble, he emphasizes that accurate measurements are only possible in fresh preparations or in preparations carefully preserved in alcohol. In contrast, drying, as it is necessary in natural skeleton preparations, renders the specimens “useless caricatures” (“Durch Austrocknung schrumpfen sie zu unbrauchbaren Karikaturen ein.“, [7]; translation by the authors).

Eventually, the advent of pioneering soft tissue clearing methods, in particular “Aufhellungsverfahren” established and published by Werner Spalteholz in 1911, made the laborious preparation of natural skeletons obsolete [8]. Spalteholz’ method used a combination of chemical bleaching and refractive index matching approaches to render human and animal tissues translucent [9]. The intra-uterine and post-natal development bones could thus be demonstrated and measured much more precisely in transparent specimens than using natural skeletons [8].

With the waning relevance of natural skeletons for didactic and scientific purposes, the handling of them also deteriorated. Improper repairs are one testimony to the loss of importance of these sophisticated preparations. Moreover, the preparation of natural skeletons demands a lot of experience and practice, which conceivably was not passed on the next generation of anatomists and preparators, once less laborious alternatives were available.

This preparation technique of fetal natural skeletons, however, was described in detail at the end of the eighteenth century by the German anatomist Johann Leonhard Fischer [10], which is highly valuable for proper restoration approaches today: First, the cadaver was roughly fleshed down to the joint capsules and ligaments, and decapitated to remove the brain through the Foramen magnum. For older children, Fischer recommends cutting the upper and lower extremities. The spinal cord was removed with a wire. Afterwards, the skeleton was placed in fresh water and cleaned of blood residues. Fischer describes two methods of eliminating residuary organic tissue: [1] the skeleton could be placed for one or 2 days in lime water; [2] with much experience, the dissector could allow to macerate it slightly. To prevent the disintegration of the joint capsules and ligaments, both processes had to be monitored carefully and interrupted in time. Subsequently, the skeletons were cleaned and dried in a wooden frame. For stabilization, a brass wire was passed through the vertebral canal and the skull was connected to it with a piece of cork or wood in the Foramen magnum. The preserved capsules and ligaments commonly were treated e.g., with turpentine or rosemary oil to increase their transparency. Finally, the skeleton was varnished. A translation of two historical German sources on the preparation of natural skeletons (Fischer (1791) [10] and Bock (1829) [5]) are provided in the supplementary material.

Here, we present a state-of-the-art documentation, conservation, and restoration of a fetal natural skeleton from the early twentieth century, which is part of the neonatological collection of the University of Tübingen. The historical background of the skeleton is scarcely documented. It was arguably prepared in the early twentieth century at the “Kaiserin-Auguste-Viktoria-Haus zur Bekämpfung der Säuglingssterblichkeit im Deutschen Reich” (“Empress-Auguste-Viktoria-House for combating infant mortality in the German Reich”) in Berlin, Germany. Interestingly, the renown Finnish pediatrician and pioneer neonatologist Arvo Ylppö (1887–1992) was working at this very same hospital from 1912 to 1921, particularly conducting research targeting the anatomy and patho-anatomy [11], as well as physiology in pre-term children [12]. Moreover, he conducted studies on the growth of pre-term children from birth until school-age [13] making a link between Ylppö and our natural skeleton conceivable. Anatomically, the natural skeleton is a fetal skeleton, arguably before the 10th fetal month as indicated by the lack of ossification centers of the Ossa cuboidea, which normally from at the end of the fetal period or early postnatal stages [14]. Also, there are no signs of pathological changes to the skeleton. Currently, the skeleton does not serve any educational or scientific purposes and is not on public display. However, it might be part of exhibitions of the neonatological collection in the future.

Although the skeleton no longer represents a significant contribution to the education of physicians today and despite the fact that its exact time of origin, but above all its provenance, remains unclear, the University of Tübingen bears responsibility for this specimen. Ethically, we consider it our responsibility to ensure above all a respectful handling of the skeleton, which includes keeping it in a stable condition that prevents damage and preserving the original appearance [see also the recent recommendations by the American Association for Anatomy [15]]. However, despite an undocumented restoration approach in the 1980s, little measures have been undertaken to preserve the condition of the skeleton, resulting in a cumulative, previously undocumented damage pattern over the decades as listed below. Since we are convinced, that similar cases of insufficiently preserved specimens can be found in most medical collections, this article is not only intended to document the state-of-the-art conservation and restoration of the natural skeleton in December 2022 at the University of Tübingen, but also serve as an inspiration for similar campaigns in other institutions in the future.

Research aim

In the present work, we aimed to document the state-of-the-art conservation and restoration of a fetal natural skeleton as it can be found in anatomical collections worldwide. We use modern video- and photodocumentation during and after the conservation/restoration campaign. We critically discuss the legitimacy of restoration measures in the replacement of lost body parts and how modern techniques such as photogrammetric 3D reconstructions can improve the transparency of restoration measures.

Material and methods

Conservation measures

Initial examination and cleaning procedure

Prior to restoration and reconstructive measures, we examined the condition of the natural skeleton visually and with the use of UV-radiation (365 nm). The latter was done to detect previous adaptations to the skeleton that often exhibit different fluorescent spectra compared to the original preparation depending on the materials used. Afterwards dry-cleaning procedures were carried out by using a goat hair brush (Deffner and Johann GmbH, Röthlein, Germany). We used a gauze bandage (Paul Hartmann AG, Heidenheim, Germany) wettened with distilled water to clean the surfaces of the bones and organic components (i.e., ligaments and cartilage).

Structural stabilization

For structural stabilization of bone and ligament fractures, we used glutinous glue, with a viscosity of 4000 mPas/s at 24 °C and a pH 6 (Deffner and Johann GmbH, Röthlein, Germany, product nr. 2,416,000) as an adhesive. In order to restore the connections between individual skeletal parts, the contact points of the organic tissue were first partially flexibilized using synthetic fleece and cotton wool compresses (both by Deffner & Johann GmbH, Röthlein, Germany) wettened with distilled water. Drops of the adhesive were then applied using a 10 ml syringe (B. Braun Melsungen AG, Melsungen, Germany) with a 20 G needle (0.90 × 70 mm, B. Braun). The contact surfaces were temporarily fixed with bulldog clamps (Aeskulap AG, Tuttlingen) and cotton threads until the adhesive effect started. The 10th rib on the right was glued dry and without pre-flexibilization. Excess adhesive was removed with a scalpel and fiberglass eraser before it was completely dry.

Restoration meassures

Reconstruction of missing bones

To reconstruct missing bone structures, we used a mixture of 80% (wt/v) bleached beeswax, 12% (wt/v) dammar resin, and 8% (wt/v) Rügen chalk (all components by Kremer Pigmente GmbH and Co. KG, Aichstätten, Germany), which was modeled on a corrosion-resistant stainless-steel wire (diameter 0.8 mm, distribute by Reidl GmbH & Co. KG, Hutthurm, Germany). The joint capsules and ligaments were modeled with tissue paper (11 g/m², distributed by Deffner and Johann GmbH, Röthlein, Germany). Bonding was carried out using Plexigum^® PQ 611 in white spirit 100/140 °C, viscosity at 20 °C: approx. 100 mPas/s (both by Kremer Pigmente GmbH and Co. KG, Aichstätten, Germany).

The reconstructed digits and 5th metatarsal bone were attached to the metatarsal or soft tissue of the developing ankle joint, respectively, by introducing the excess stainless-steel wire and fixing it with glutinous glue and thin strips of tissue paper. Moreover, all reconstructed skeletal parts were tinted with oil paints diluted with oil of turpentine (both by Boesner GmbH, Leipzig, Germany), and subsequently covered with 2% (wt/v) shellac dissolved in 99.8% (v/v) ethanol (both by Kremer Pigmente GmbH and Co. KG, Aichstätten, Germany) (supplementary Video 6).

Afterwards, we reattached the left foot by flexibilization of the ligamentous organic tissue of the ankle and subsequent application of glutinous glue. Flexibilization was carried out in analogy to the costal cartilages (see above). For exact positioning, we used a customized supportive structure carved from a polyvinyl alcohol sponge (Deffner & Johann GmbH, Röthlein, Germany) (supplementary Videos 7 and 8).

Technical documentation

Photo and video documentation

We carried out detailed photographic documentations before and after the conservation-restoration, as well as throughout the entire process [Panasonic Lumix GH5 (Panasonic) with Olympus 12–40 mm 1:2,8 PRO lens (Olympus)]. Moreover, we filmed selected restoration and reconstructive measures using a tripod mounted Panasonic AG-CX350E (Panasonic) or a KLSMartin MarLED-X Lamp with integrated camera (KLSMartin).

X-ray examination

X-ray examination and documentation was carried out using a Philips BV Pulsera (Koninklijke Philips N.V., Amsterdam, Netherlands) with 40 kV and 0.167 mA. All detectable metal parts were imaged from various angles.

Photogrammetry

In addition to the photographic documentation, we carried out photogrammetry after the conservation-restoration. Therefore, the specimen was placed on a rotary table in front of a black screen. Three LED-spotlights Boltzen Lights (CameTV) were placed around the setup to ensure sufficient and homogeneous lighting. We used a Canon EOS R5 camera (Canon Inc., Tokyo, Japan) with a Canon EF 24–70 mm 1:2,8 L II USM lens (Canon) mounted on a tripod. Camera settings were set to standardized values (focal length: 70 mm; aperture: F/11; exposure time: 1/90 s; ISO 800).

Eight systematic series of photos were taken to collect a maximum of optical angles for later 3D reconstruction with an automatic trigger set to 5 s intervals (supplementary Fig. 1): (1) 71 images with a tripod-mounted camera while 360° rotation of the specimen on the rotary table; (2) five images with a hand-held camera while following an arcuate trajectory in the medio-sagittal plane starting with the thorax from ventral and ending directly above the within the longitudinal axis of the skeleton; (3) seven images with a hand-held camera while following an arcuate trajectory in the medio-sagittal plane starting with the thorax from dorsal and ending directly above the within the longitudinal axis of the skeleton; (4) three images with a hand-held camera while following an arcuate trajectory in the frontal plane starting with the thorax from the left side and ending directly above the within the longitudinal axis of the skeleton; (5) five images with a hand-held camera while following an arcuate trajectory in the frontal plane starting with the thorax from the right side and ending directly above the within the longitudinal axis of the skeleton; (6) 23 images with a tripod-mounted camera with a steeper angle than in the first series while 360° rotation of the specimen on the rotary table; (7) 10 close-up images with a hand-held camera of filigrane structures of the hands, pelvis, and feet while 360° rotation of the specimen; (8) 3 images with a hand-held camera positioned at the base plate in the medio-saggital plane focusing carnially to visualize the interior of the ribcage.

All images were adjusted for contrast and brightness using Adobe Lightroom Classic software (version12.2.1; Adobe Inc., San José, CA, USA) to enable optimal recontruction of the geometry of the specimen. Background was cropped using Adobe Photoshop (version 24.3.0; Adobe Inc., San José, CA, USA) to facilitate the geometry reconstruction. Additionally, shadows were reduced and overall brightness was increased in images using Adobe Lightroom Classic for better reconstruction of the texture of the specimen.

Photogrammetric reconstruction was performed using RealityCapture software (version 1.2; Epic Games, Inc., Cary, NC, USA) with the following settings: Alignement Settings Max Features per mpx: 1 mio, Max Features per images: 4 mio, Image Overlap: High, Downscale factor: 1, Max feature reprojection error: 2, Force component rematch: Yes, Background feature detection: No, Background thread priority: Normal, Preselector features: 2 mio, Detector sensitivity: Medium, Distortion Model: Brown3;

Reconstruction Settings Image downscale: 1, Maximal depht-map pixels count: 0, Minimal distance between to vertices: 0, Maximal vertex count per part: 5 mio, Smoothing: 10;

Unwrap Tool Gutter 2, Maximal texture resolution: 16,384 × 16,384, Large triangle remove threshold: 10, Maximal required textel size: 100 × optimal (1% texture quality), Minimal required textle size: Optimal;

Color and Texture Settings Gutter 2, Maximal texture resolution 16,384 × 16,384.

All other settings were set to default.

Data availability

The photogrammetric reconstructions, both with and without annotations are available on the online repository of the Museum of the University of Tübingen. Access via: https://www.unimuseum.uni-tuebingen.de/de/sammlungen/3d-museum. No datasets were generated or analysed during the current study.

Results

Mounting and general assumptions

The natural skeleton is mounted on a non-ferrous metal rod attached to a historic, black, cast-iron base (Fig. 1A). The rod was introduced into the spinal canal through the 5th lumbar vertebra, with its tip ending just beneath the skullcap as revealed by x-ray examination (supplementary Video 1). It is fixed with a piece of cork in the Foramen magnum. The feet of the skeleton are not connected to the base or rod.

Fig. 1

Condition of the fetal natural skeleton prior to restoration. A’–A’’’’ shows the entire specimen from different angles. The modern glass case was removed to avoid reflections. The skeleton is mounted on a non-ferrous metal rod, which in turn is attached to a historic, black, cast-iron base. The specimen was placed on a modern base plate, to which it was taped by a layperson for transportation before our restoration started. In B and C the thorax is shown from the right in normal light conditions and with UV radiation. Light-coloured encrustations were visible on 7th rib, which exhibited a blueish fluorescence under UV illumination (white rectangles indicate the higher magnification photographs in B’ and C’). The carving marks in the skull shown in D (black arrowheads) were caused during the initial prepration process and are common to natural skeletons. In E, F, and G broken ribs and cartilages of the ribcage are shown (white arrowheads indicate the breaking points). H shows the damage on the left ankle and foot. The entire foot skeleton broke off just distal of the ankle joint and all phalanges II-V along with the 5th metatarsal bone are missing

As a protective measure, the skeleton with its historic base was placed on a modern base plate (black medium density fiberboard) with a glass dome. For transport, the historic cast-iron base was fixed to the wooden base plate with transparent adhesive tape by a layperson before our campaign was started.

Condition of the specimen

The natural skeleton was in a stable overall condition with only minor soiling (Fig. 1A). Smaller, light-coloured encrustations were visible on several bones (e.g., iliac crest and 7th rib of the right side) and exhibited a blueish fluorescence under UV illumination (Fig. 1B, C). The characteristic kerf marks on the skull (Fig. 1D) are caused by the initial preparation and can be found regularly on similar specimens.

Moreover, the costal cartilages and the sternum exhibit obvious deformations (Fig. 1B, E). However, it is unclear whether these changes occurred during the initial preparation and mounting process. On the left side, the costal cartilages of the 5th rib and onwards were either detached from the sternum or only connected with it by minimal residuals (Fig. 1E). Moreover, the sternocostal joint of the 1st rib on the right is disconnected, as well as the vertebrocostal joint of the 10th rib on the left (Fig. 1F, G). The 10th rib on the right is fractured incompletely (Fig. 1G).

The most striking damage is on the left foot and ankle, which broke off just distal of the joint. In addition, all phalanges II-IV as well as the 5th metatarsal bone and all phalanges are missing (Fig. 1H).

The documented history of the natural skeleton is fragmentary, yet, restoration measures, however, without documentation, were carried out in 1985. Moreover, an attempt to reattach the left foot was undertaken by a layperson. Since additional restoration measures are not documented, it is speculative when the following adaptations were done to the specimen: The 1st rib on the left and the 12th rib on the right, as well as the right clavicle, were partially restored with metal wire and a light-coloured putty (Fig. 1F). It is conceivable that the encrustation on the 7th rib on the right and other bones (see above) goes back to the same intervention as it exhibits similar appearance and fluorescence in normal light and using UV radiation, respectively. Additionally, the right clavicle was fractured and readapted using glue. Likewise, a high level of gloss on the right cubital joint indicates another revision (i.e., gluing).

The historical, cast-iron base was lacquered at least once in monochrome black. It is intact, stable, and with only minor defects. Numerous small areas are missing in the black lacquer and slight corrosion can be observed in these areas. The non-ferrous metal rod, the twisted wire fixation of the ribs, as well as the metal mounts of the scapulae also show minimal traces of corrosion.

The modern wooden base plate and the glass dome are in good overall condition and only slightly degraded.

Conservation measures

Cleaning and structural stabilization

After initial visual evaluation and assessment under UV radiation, we employed initial dry cleaning and subsequent wet cleaning with distilled water (supplementary Video 2). We avoided to use organic-solvent-based cleaning procedures to not remove the very thin original varnish that exhibited a slightly orange fluorescence under UV illumination (data not shown).

The aim of the subsequent structural stabilization was to restore all detached connections and fractures between the individual skeletal parts (costal cartilage, ribs, vertebral bodies). However, we did not attempt to reshape the deformed sternum and costal cartilages beyond a minimum that was required to adapt the glueing areas, since it is unclear whether these deformations were caused by the original preparation process already.

We chose to use glutinous glue as a suitable adhesive, considering various requirements: Primarily, the adhesive must not lead to discoloration of the organic tissue. Moreover, since the costal cartilage has to be made flexible and partly reshaped using wet compresses (Fig. 2A, B; supplementary Video 3), the adhesive should also be soluble in water to allow for an optimal adhesive effect after flexibilization. Additionally, glutinous glue has the advantage of a strong adhesive effect that quickly sets in, making it particularly suitable for our application (i.e., gluing fracture points of only a few square millimeters).

Fig. 2

Restoration of the thorax. Shown are representative steps of the restoration on the broken ribs and cartilages of the thorax. As depicted in A, the cartilage had to be made flexible and partly reshaped using wet compresses. Afterwards, glutinous glue was applied using a syringe (B). The breaking point were adapted using bulldog clamps and yarn (C). In D, the removal of excess adhesive with a fiberglass eraser is shown

As shown in Figs. 2 and 4 we successfully reconnected all documented fractures, except for the connection between the 10th costal cartilage and the sternum (see also supplementary Videos 4 and 5). Here, the contact surfaces proved to be too small for effective adhesion.

Restoration measures

Reconstruction of missing skeletal parts and reattachment of the left foot

The reconstruction of the missing digits and one metatarsal bone, as well as the reattachment of the left foot was the most extensive measure of this campaign. The concept of presenting the skeleton with all body parts present at the time of death corresponds to our own entitlement and the general requirement for a dignified and ethically respectful handling of this preparation [[16, 17], see Discussion].

The exact reconstruction of the shape, size, structure and color of the lost bones was considerably facilitated by the fact that all the lost phalanges and metatarsal bones were still present on the contralateral side. Thus, we succeeded in a realistic reconstruction of the lost skeletal parts (Fig. 3). Likewise, we used the contralateral ankle as well as the configuration of the breaking edge as an indicator for the positioning of the joint when reattaching the left foot. Moreover, we took care that the feet are not in contact with the historic metal base to avoid mechanical stress when moving the specimen in the future (Figs. 3, 4, x-ray documentation in supplementary Fig. 2). Additionally, we uploaded two virtual 3D-models of the skeleton after our restoration to the online repository of the Museum of the University of Tübingen MUT, both with and without annotations to document historical changes and the results of our resortative measures (see https://skfb.ly/oRqZq and https://skfb.ly/oRrnU).

Fig. 3

Reconstruction and reattachment of the left foot skeleton. Shown are representative steps in the modeling and painting of the missing foot skeleton and its reattachment to the ankle joint. A shows how missing bones and joint capsules/ligaments were modeled using a mixture of wax, resin, and chalk or tissue paper, respectively mounted on a stainless-steel wire (see Material and Methods). Subsequently, the excess wire along with the reconstructed digits were inserted and glued to the proximal soft tissue or metatarsal bones (B and C). The reattached skeletal parts were pained (D) and fine modeled (E). In F the different stages of the reconstruction of the foot skeleton are depicted (top: reconstructed digits before attachment; middle: reconstructed foot skeleton before paining; bottom: reconstructed foot skeleton after painting). G–I shows the reattachment of the reconstructed foot skeleton to the ankle joint. After the application of glutinous glue with a syringe (G), the exact positioning of the foot skeleton was achieved using a custom-carved polyvinyl alcohol sponge as a supportive structure (H). Excess adhesive was removed using a fiberglass eraser (D)

Fig. 4

Condition of the fetal natural skeleton after restoration. A’–A’’’’ shows the entire specimen from different angles without the modern glass case to avoid reflections. B–E shows details of the specimen from an angle similar to Fig. 1E–H for better comparison. Thus, B, C, and D depict the ribcage with broken ribs and cartilages after re-adaptation indicated by white arrowheads. The white arrow in B indicates the broken costal cartilage for which repair was not possible. E shows the distal lower extremities with the reconstructed left foot and reattached ankle joint

Rework of the pedestal

In order to avoid further damage to the historical iron-cast base, we carefully removed the transparent adhesive tapes. Unfortunately, smaller black lacquer fragments of the historic metal base were already stuck to the adhesive surfaces and could not be repositioned (supplementary Fig. 3).

All metal parts and the surface of the wooden base were then cleaned with white spirit 100/140 °C and a cotton gauze bandage. To fix the metal base on the medium density fiberboard, we drilled a small hole into the modern wooden base and inserted and glued a bamboo stick into it. The tip of the bamboo stick was inserted into the metal base using glutinous glue. The bamboo stick was tinted with black acrylic paint. The glass dome was cleaned with 99.8% (v/v) ethanol.

Discussion

For the development of a suitable conservation and restoration concept, knowledge of the preparation and mounting technique has the highest priority. Thus, a comprehensive understanding of the modalities and materials with which a work/preparation was initially produced is indispensable for a correct interpretation of the current state of preservation. Likewise, later restoration measures must be located and the techniques and materials used must be reconstructed as they can inflict or contribute to secondary damages. Based on these preliminary examinations, a cataloque of restoration and conservation measures is planned individually for each object/preparation.

Methodologically, restoration and conservation research makes use of a large spectrum of material analysis methods such as Fourier transform infrared spectroscopy (FTIR), gas chromatography/mass spectrometry (GC/MS), Raman spectroscopy, X-ray fluorescence (XRF), 3D micro X-ray fluorescence (3D-μ-XRF), portable X-ray fluorescence (p-XRF), Scanning electron microscopy with energy dispersive X-ray spectroscopy (SEM/EDX), and polarised-light microscopy (PLM), as well as imaging procedures using UV-, IR-radiation and X-rays. While some of these methods are non-destructiv (e.g., p-RFA, 3D-μ-RFA, UV-, IR-radiation, X-ray), the use of FTIR, GC/MS, Raman spectroscopy, XRF, SEM/EDX and PLM usually requires the collection of microsamples. However, taking a microsample, even if not easily visually perceptible, changes the object/specimen irreversibly and can therefore only be justified if non-invasive methods do not provide sufficient results (see also [18]). Additionally, taking samples of human preparations or other human remains must be assessed extremely sensitively from an ethical perspective and must be well justified (see e.g., [19, 20], or the Guidelines by the German Museums Association [21]).

In the current case, our preliminary material analysis of the natural skeleton required only minimal, non-invasive procedures (inspection using normal light and UV-radiation, X-ray) and no destructive methods. Moreover, the essential information on the preparation procedure were reconstructed using historical literature and observations on the object itself (e.g., mounting of the skeleton, holes for hollowing the long tubular bones). Examination under UV radiation and X-ray additionally revealed the original varnish and previous reconstructive measures using metal wires. Furthermore, we were able to link the deposits of light-coloured putty (e.g., on the iliac crest) to previous repair works on the left clavicle, the 1st rib on the left, and the 12th rib on the right, as obviously the same replacement material was used. In addition, our investigations strongly indicated that these light-colored deposits were inert and did not react or damage the adjacent organic and inorganic tissue.

When is a more extensive analysis necessary?

If, however, the replacement material or the metal wires used in previous reconstructive approaches had caused any chemical recations or optical impacts to the original, a more extensive material analysis would have been strongly indicated [17, 22]. In this case, even the removal of the old repair measures would have been justifiable from a restoration ethics perspective [e.g., [17]]. In principle, however, such far-reaching measures must be very well justified, because revisions and historical repair interventions on an object/specimen must always be considered as part of its history and should be respected as such—i.e. left as it is, if possible [see e.g., Venice Charter 1964, Art. 11, [23]; E.C.C.O. Professional Guidelines (II), 2003, Art. 8, 15, [24]].

Moreover, additional analysis would be neccessary if the soiling of the surface was so severe that a sufficient cleaning result cannot be achieved by dry and water-based cleaning. If the use of organic solvents is unavoidable (e.g., if fatty acids leak from the bones), sampling and analysis (FTIR and, if necessary, GC–MS) of the thin protective varnish must be carried out to find a suitable solvent that would dissolve the contamination, but not the coating.

Preserving traces of previous repair interventions and revisions on a specimen: Why was the light-coloured putty not removed?

The authors decided to leave the white putty (e.g., on the iliac crest) untouched following a basic principle of conservation and restoration: All traces on an object/preparation are part of its history and as such bear additional information that contextualize the specimens (such as previous revisions, historical repairs, or traces of usage). Thus, if, these traces do not damage the original, they must be preserved. This is particularly important if the origin or the purpose of these traces cannot be determined by the conservator-restorer.

The damage caused by disregarding this principle—or by misinterpretation of traces is impressively illustrated by an example from the Anatomy Museum at the University of Basel [25]. In 1971, the taxidermist had completely removed the historical colouring of the incarial bones (Os interparietale) and several historical inscriptions (partly provenance) from a then almost 300 year-old human skull (1–1-2/27) before transferring it from the teaching collection to the museum's collection. Additionally the skull was bleached. These far reaching measures were justified by the taxidermist as an attempt to guarantee the visitors an unadulterated view of the anatomy of the skull rather than its history/provenance [25].

Is the reconstruction of the left foot skeleton a justifiable procedure from a restoration ethics perspective?

While the removal of historical repairs and revisions would be a legitimate measure under the conditions mentioned above (causing/promoting damage), the reconstruction of the left foot skeleton can certainly be discussed controversially, particularly as it is a constructive measure that does not serve the strucutal integrity of the preparation but rather is an aesthetic addition. A lack of knowledge of the original condition, i.e., concerning the exact appearance of the bones, is particularly problematic as its authenticity may be disputed. However, for proper restoration, verifiablitiy is a basic requirement for comprehensive reconstructions; hypothetical additions to the original should be excluded [see e.g., Venice Charter 1964, Art. 9, [23]].

Therefore, it is noteworthy to explain two main reasons why we decided to replace the foot skeleton:

A) Missing body parts, mostly digits and teeth, along with defilements and improper repairs are among the most common damage patterns found in mounted and natural skeleton preparations in legacy anatomical collections. Yet, these damage patterns are rarely questioned as the skeletons originally mainly served teaching purposes and thus were often regarded and treated as objects of daily use. However, if we consider that anatomical and pathological collections do not contain everyday objects, but the remains of human beings, it becomes obvious that the absence of body parts contradicts an ethically correct treatment of the deceased [16]. The Stuttgarter Empfehlungen (Stuttgart Recommendations), as an important national guideline for the handling of human specimens in museums and collections, state that “the specimens must be displayed in an optimal state of preservation” („die Präparate [sind] in einem optimalen Konservierungszustand zu zeigen “, [26]; translation by the authors). Additionally, a 2021 recommendation by the Coordination Centre for Scientific University Collections in Germany, proposes an optimal state of preservation even outside public presentations [27]. We believe that the completeness of a human preparation is also part of an optimal state of preservation. If this completeness cannot be achieved because body parts have been lost, we therefore suggest that a state that conveys completeness should be achieved using a life-like reconstruction.

B) The natural skeleton is an anatomical preparation, not a pathological case study. However, a missing body part might lead the viewer to falsely conclude that the damage is the result of an injury inflicted during the lifetime of the individual, thus misinterpreting the damaged anatomy as an pathological change [16]. Moreover, since the present natural skeleton is an anatomical preparation morphology and size of the bones can be considered to be axially symmetrical. As the missing phalangeal and the 5th metatarsal bone are still present on the contralateral side, the shape and color of the missing bones of the left foot skeleton could be reconstructed. This reconstruction was therefore not solely based on the subjective interpretation of the authors.

It is notworthy, however, that this approach bears the risk of incorrect conclusions drawn by viewers caused by the life-like, yet never perfect/original reconstruction of the missing body parts. We are well aware of this problem and therefore, in order to ensure transparency, we have documented the additions in a comprehensible manner in accordance with the applicable restoration standards. Moreover, the reconstructed skeletal parts, while having a life-like overall appearance, are still distinguishable from the original parts upon close inspection, making confusions in future investiagions involving this preparation unlikely.

As a modern supplement to the traditional photo- and video documentation, we utilized modern photogrammetry to visualize all changes (historical and new) as a virtual 3D model. Photogrammetrie is a relatively novel technique and is currently hardly used in the documentation of human remains in anatomical and pathological collections. However, it offers easy accessibility of preparations not only for professional conservators and restorators, but especially for researchers from other disciplines (e.g., medicine, medical history, history of art), who might not have the same level of expertise required to interpret the historical and new changes to a preparation. Particularly the implementation of interactive elements such as coloring or annotations shows additional advantages of photogrammetry in terms of transparent documentation. Overall, we are convinced that the technical efford and time invested into photogrammetrie is an asset to the classical photo- and video documentation and might define a new state-of-the-art in transparently documenting conservation and restoration measures on human anatomical and pathological preparations.

Additional considerations

The authors would like to point out that, in particular, the handling, scientific research, and treatment of human specimens that originate from a context of injustice (e.g. victims of Nazi terror, colonialism, slavery) must always be coordinated with the relevant communities [see e.g., [21, 28,29,30]. Establishing contact and respectful dialogue with the communities of origin, descendants, stakeholders etc. should always be the first action before (invasive) measures are taken on these specimens. It is noteworthy, however, that establishing a dialog and/or reaching a consensus involving the relevant communities is highly time-consuming and hardly plannable. Thus, if e.g., a preparation urgently requires structural stabilization or exhibits signs of progressive deterioration (e.g., active chemical reactions, pests or mold), appropriate, reversible measures must not be delayed to prevent further damage [27]. It is self-evident, that these measures must be documented completely and in detail.

Conclusion

Here, we present a state-of-the-art conservation and restoration of a natural skeleton common in medical and university collections, along with a transparent documentation using modern technologies. As the preparation of natural skeletons is a forgotten art, the review of historical sources is indispensable for the planning of conservation-, restoration-, and repair approaches and proper handling. Ethical implications must be considered in each individual case taking the particular damage pattern and provenience into account. Modern technologies, such as photogrammetry can improve the transparency of resortation and conservation campaigns in the future.