Navigation – Plan du site

The newly applied mortars in mosaic restoration

The Dionysos mosaic in Cologne, a case study and an example to face the problems connected to the choice of mortars and how they can change the visual perception of mosaics.
Fabiana Moro et Angelica Pujia


L’intervention de restauration sur la mosaïque de Dionysos à Cologne a permis, dans le cadre du travail de fin d’étude, une recherche sur les problématiques liées au choix du lit de pose des mosaïques detachées et replacées sur de nouveaux supports. Elle a contribué à l’étude des facteurs qui influencent la conservation des mosaïques qui ont précédemment fait l’objet d’interventions de détachement du site originel.

Haut de page

Notes de la rédaction

Istituto Superiore per la Conservazione e il Restauro in Roma – Antonio Iaccarino Idelson

Texte intégral


  • 1  This article is a spinn-off of our degree research in Mosaic restoration, discussed on December 20 (...)

1This study on the Cologne Dionysos mosaic has been prompted by its recent restoration completed in July 2008, thanks to a cooperation between the University of Erfurt unit lead by Dr. Christoph Merzenich and the Laboratory for Restoration of Mosaic and Stucco of the ISCR in Rome1.

  • 2  About the story of Dyonisos Mosaic see Fremersdorf 1956, p. 29; Dunbabin 1999, p. 82- 85; Parlasca (...)

2This German-Italian cooperation is rooted in a previous one: by the end of the ‘fifties, the extremely poor  conservation status of the mosaic prompted the German authorities to request a new expertise and then to commit a new intervention from the Istituto Centrale del Restauro. Cesare Brandi personally supervised the operation, which took place in the years 1959-61; he realized that the only feasible solution for the conservation of the floor mosaic appeared to be the detachment of the mosaic, which was by gently pealing it from the substrate, putting it on a roll and subsequently repositioning it in its original location, which had been duly prepared with an inside ventilation system running below the mosaic setting bed in its final exposition basin (ill. 1)2.

Fig.1 The Dionysos mosaïc

Fig.1 The Dionysos mosaïc

Ill. 1 The Dionysos Mosaic, Römisch-Germanisches Museum, Cologne.

Crédit: F. Moro   

A conservative history of the Dionysos mosaic

3 In May 1941, in the middle of the second world war, the construction was begun of an anti-air raid  bunker along the South side of the Cologne cathedral; in the month of July of the same year, in the course of the digging, a mosaic floor was discovered: it was located under the street level, beneath a layer of 6,50 metres of burnt bricks and potsherd fragments; its size was 7x10 meters.

4       Construction works were immediately interrupted, in order to clear the floor from the debris: the discovery was an unprecedented event for the city of Cologne: indeed, having been inhabited uninterrupted for centuries, the most important Roman city in Germany had surviving ancient buildings, but so far only discovered only in out-of-the-way suburban areas.

5        Until the construction of the bunker was being completed, that is towards the end of 1941, the mosaic was protected by a temporary shed; the floor was covered by a layer of dirt and a concrete sheet, which was meant to save it from further damages, at least until the end of the war.

6In 1946, at the end of  the conflict, the mosaic was unearthed and in the month of October of that year the discovery was made public. A corridor, dug all around the mosaic, allowed visitors to  observe the artifact closely, while the discovery was seen as a proper circumstance for laying the foundations of the Römisch-Germanisches Musem in Köln.

  • 3  See Dunbabin 1999, p. 79-91.

7In the earlier days of the Roman Empire the use of mosaic in the Germanic provinces of the roman empire was generally similar to that in the south. In this region black and white mosaics and other pavement of Italian derivation had been introduced. However relatively, little amount of examples survived. Here, the most productive and most characteristic period began at the end of the second century and lasted until the German invasions in the area, in the 260s and 270s. Crowded ornamental compositions containing figured motifs and groups were the most popular decoration of this kind of pavements. The range of subject-matter found on mosaics from this area is also characteristic: mythological scenes, scenes from literary life, poets, philosophers and muses.3  

  • 4  About the controverso on the datation see Dunbabin 1999, p. 81.

8The Dionysos mosaic decorated the triclinium of a rich roman domus and it is one of the finest products of the Rhineland roman workshops: the field is covered with squares of guilloches interlaced to form octagons; smaller squares are set on the points, while the spaces are filled in with pairs of lozenges. Further, borders of meanders and dentils, outline these compartments. Figures are placed in the octagons and squares, and the segments of these forms along the edge. In total there are thirty-one panels in all. At the centre, a larger compartment is given special emphasis by an additional guilloche border, and shows Dionysos leaning on a satyr; all the rest contain Dionysiac figures and emblems of his cult. There is a contrast between the richly ornamented framework and the lightness of the figured panels, were the figures stand out against a plain ground whose white tesserae are set in an imbrication pattern. The contrast extends to the colouring: bright, variegated tones, with considerable use of glass in the figures and a more muted range in the frames. The figures themselves are classical in form, well proportioned and carefully modelled. Despite controversy over the date, it is probably to be placed in the later Severan period, around 220 AD.4

Technique and mosaic production

  • 5  See Vitruvius, I , VII, ed. Corso 1997; Moore 1968, pp. 60-65. This composition would be diversifi (...)
  • 6  About the observation and study of this layer see Moore 1968, p. 60-65.
  • 7  Fremersdorf 1956, p. 27; Doppelfeld 1962.

9In order to investigate how to choose a correct setting bed for the detached mosaic, it became relevant, to undestand antique technique and methods used by roman craftsmen and workshops in general and to investigate, more precisely, wich materials were used in our case of study, the Dionysos mosaic. Unfortunately, our knowledge is limited because ancient writers showed little  interests in the activities of craftsmen and of manual workers in general; Vitruvius, who wrote a book on architecture during the reign of Augustus is the only one who gives detailled instructions for the construction of foundations of pavements. He starts with the socalled statumen, a bedding of fist-sized stones, followed by the rudus, a mix of rubble and lime and then the nucleus, or upper layer, a fine mortar mixed with three parts of crushed tiles or potsherds5 These general stages he distinguishes are found in most mosaics throughout the roman area; another very fine layer of mortar, the setting bed (object of our study) was required for the actual laying of the tesserae, often composed by simple lime or lime mixed with powdered potsherds which gave a rose colour to the mortar. Even if Vituvius never mentions this final  layer, it was always found in all mosaics of roman and greek origin.6In the case of the Dionysos mosaic, the only information we have, comes from the reports of Fritz Fremersdorf, former director of the museum, who had the chance to observe the setting bed layer of the mosaic before its stripping: he relates that the tesserae were laying on a pink setting bed 2 to 4 cm thick.7     

Our research

10   In 2007 the director of the Römisch-Germanisches Museum asked the direction of the ISCR for new advice; the result was a new cooperation agreement and the establishment of a restoration campus operating throughout the month of  July 2008.

  • 8  Ardovino 1994, p. 33-48; Ardovino 2002.

11During this intervention, a method was developed for the removal of the previous degraded synthetic protection layer which had covered the mosaic for decades and had turned yellowish. The yellow layer had modified the original colour of the mosaic subsequently altering the right chromatic relationship between the tesserae and the interstices. In the light of new and improved knowledge of the composition of the mosaic’s stratigraphy and the chromatic relation between it and the tesserae, we judged it correct to replace the interstitial mortar with a new and more appropriate one.8

12In a mosaic of this size the percentage of surface occupied by the interstitial mortar is extremely  meaningful in the overall perception of the whole composition. At the end of the intervention in Cologne, the replacement of the interstitial spaces with a material  of a different chromatic value has resulted in a significant variation in the chromatic perception ot the whole surface (see ill. n.1).

13This restoration gave the opportunity to put under scrutiny a subject so far not sufficiently considered, i.e. the individuation of criteria underlying the choice of mortar in mosaics stripped and relocated on new surfaces, and therefore deprived of their preparatory and layering strata.

  • 9  Ardovino 2002.

14The study began with an enquiry into the ancient executing techniques, relating to floor mosaics. The question was the following: should the choice of interstitial mortar be guided by our contemporary ideas of chromatic compatibility, or should we rather try to reproduce the ancient execution techniques?9

  • 10  Frova 2004; Gregory 1998.

15Needless to say, part of the knowledge needed for a complete vision of all the factors to be examined is far beyond a restorer’s or an historian’s reach, for it belongs to the general field of physics and optics. The differences in the perception of colors depend on the physiology of the eye and on the brain’s interpretation of visual stimuli and, more important, still they depend on the mutual relationship between different colours, which is properly the focus of our enquiry. Therefore, the research of those interactions plays the fundamental role of enlightening the emerging chromatic combinations, which are responsible for unpredictable results.10

  • 11  See Goethe 1981.

16The study of the theory of colors has the practical and didactic goal to supply guidelines for future restorations by showing the effects of different color juxtapositions11.

17We organized an empirical check of the acquired results along two lines: we examined, in the first place, the changes in the perception of the tesselated surfaces according to their features, and then in connection with the influence of a chosen mortar.

18Then, we measured the variations which interstitial spaces undergo in consequence of a stripping. It should be clear, from this short introduction, that the choices to operate in this practice refer to work in fields that are often far from one another and that do not always admit one solution. The possibility to reproduce ancient techniques, should it be feasible, would free us from choices that might be considered arbitrary; lacking this possibility, it becomes extremely important that one should have the availability of instruments capable to foster decisions based on an adequate knowledge, and respectful of ancient materials.  

19In this perspective it is beyond our aim to suggest detailed univocal directions on the correct methodologies, and even less to distinguish correct from incorrect solutions; our analysis tries to investigate, verify and tie together considerations about the choice of an interstitial mortar and to examine the facts which affect their interaction.

Color perception in a museum environment

20 The analysis of colours and their interaction has been a research field for centuries. It has been a subject matter for physics, optics, and, more recently, neurology. However, as it was ascertained, even a solid scientific ground is an insufficient basis for an explanation of their interaction.

  • 12  Albers 2005, p. 21; Casagrande 2008, Frova 2004, Gregory 1998.

21The notion of harmony and contrast of colors is inherent in the very nature of mosaics, because the final image is the outcome and effect of the juxtaposition of two basic chromatic units12. These are, the tesserae and the interstitial mortar. Chromatic values depend from both the constituent materials, and  the manufacturing techniques.

22The color in a mosaic is not, therefore,  an “intrinsic property” of its materials; it depends largely  from the processing of those materials and the optical effects caused by the geometry of the mosaic.

23The study of chromatic relations among different tesserae is a virtually endless subject and particularly interesting too, the study of execution techniques. In this paper we will only focus on the chromatic connection between the tesserae and the interstitial mortar mainly from an aestethic and visual point of view.

24As far as the study of museal restorations go, dimensions, constitutive materials, the color of tesserae, with their texture, along with their setting bed, are a set of factors in mutual and varying relation. In each restoration one has to cope with new circumstances, which call for different choices.

The Experimentation

25Our experimentation has been prompted by the wish to investigate into the aesthetic fruition of a mosaic artefact, as it is conditioned by different restoration practices. In our study we have taken into account two aspects of restoration: we have valued, in the first place, how the color of interstitial mortar affects the reading of the entire mosaic; subsequently we have investigated the way in which the stripping of a mosaic may produce a variation in the ratio between tesserae and interstice.

26The stripping operations may cause variations in the spatial ratio concerning the interplay of  tesserae and interstices. We need to remember that our discussion is referred to mosaics that were displaced from their original site and placed on a new setting bed and thus by this operation being modified in their proportions and chromatic character.

27In the evaluation of the chromatic element in a mosaic, lacking previous data and research, along with every bit of interstitial mortar, the choice of materials must be well reasoned; it must be also suggested by the mosaic’s uniqueness; which, in turn, is connected with the quality of the original materials, the choice of techniques and the state of preservation of the manufact. These elements must be considered, in different ways, as the starting point of any new intervention.

28The particular section of the Dionysos mosaic that we selected for our experiment is representative of the colours appearing in the larger artifact, and particularly of the geometrical frame enclosing the images, which has a capital role in the entire artifact.

29Therefore, we made three identical copies of a section of the mosaic, which differed from one another only in the colour of their interstitial mortar: our aim was to verify the extent to which this affected the final perusal of the entire image.   

The tests

30The choice of the materials for the tesserae was made in consideration of the need to use colours similar to those of the original tesserae in the Dionysos mosaic. With this in our minds, we selected four types of tesserae: the white ones are in “botticino” marble, the black ones in “nero assoluto” granite, the yellow ones in yellow travertine and the red ones in “rupas” red marble (ill. 2). The three sections, each measurung 30 x 40 cm were placed on a Aerolam basis through a “tecnica diretta”that consists in pushing the tesserae directly in the mortar.

Fig. 2 Samples of the modern tesserae used in our tests

Fig. 2 Samples of the modern tesserae used in our tests

The tesserae chosen for the test were made by stones different from the original ones for origin, but similar in colours.

Crédit: F. Moro

Test n.1 pink mortar

  • 13  Fremersdorf, describing the mosaic in the digging report writes: “ Seine Steinchen sind in einen r (...)

31For the choice of the laying bed mortar, in test n. 1, we followed the information extracted from accounts of the Dionysos mosaic in which the original laying bed was described as made of mortar and potsherds powder. Therefore we used a mortar made of equal amounts of hydrated lime and potsherds powder with the aim of reproducing, as closely as possibile, the visual impact of the Dionysos mosaic before 1959, that is before its detachement, when the mortar was removed along with the mosaic itself (ill. 3).13

Fig. 3 Test n.1 potsherds powder setting bed

Fig. 3 Test n.1 potsherds powder setting bed

The tesserae are layed on a pink mortar, composed of lime and potsherds powder.

Crédit: F. Moro

Test n. 2 concrete

32In our second test we intended to recreate the look of a mosaic laying on a concrete bedding because this is the most frequent condition encountered when dealing with a detached manufact. Almost all detached mosaics, as a matter of fact, are found today laying on a concrete laying bed: this, of course, affects negatively the preservation of the tesserae, besides provoking perceptive alterations connected with the chromatic nature of the mosaic itself. The process of reconstruction is the same as in test n. 1: the laying bed is common grey concrete only (ill. 4)

Fig.4 Test n. 2 concrete setting bed

Fig.4 Test n. 2 concrete setting bed

The tesserae are layed on a cement setting bed.

Crédit: F. Moro

Test n. 3 light white mortar

33In our third test we decided to reproduce the effect of a mosaic laying on a traditional mortar, made of a part of hydrated lime and a part of sand (very fine yellow sand and grey sand in equal quantities; ill. 5).

Fig. 5 Test n.3 white mortar

Fig. 5 Test n.3 white mortar

The tesserae are layed on a white mortar setting bed composed by lime and yellow-gray sand.

Crédit: F. Moro

Some considerations on the aesthetic carachteristics of the tests

34In observing the three specimens we noticed that even in their small proportions the chromatic variations were quite evident. The most relevant differences emerged from a comparison of specimen n. 2 (layed on a concrete mortar) and mosaic n. 3 (layed on a white mortar). The spectrum of reflectance of the three mortars, shows quite a clear difference between mortar n. 2 and mortar n. 3;  such a difference and can be also observed in the other tests. In the test using a concrete laying bed the color of the tesserae appears saturated, while certain other contrasts are visible, namely one between the grey-coloured concrete and the white tesserae. In test n. 1, on the contrary, such a contrast seems softer and the transition from color to color appears rather mild while every color looks unsaturated, despite the fact that identical tesserae were used in the three different tests.

35With regard to the disposition of the tesserae we came to the conclusion that, as was expected, the nature of their arrangement appeared more evident in areas of higher contrast between their color  and that of the mortar: this was particularly evident in the case of test n. 2, but also in the contrast of the black tesserae against the background of the laying mortar in test n. 1 and n. 3.

Colorimetric measuring

  • 14  See Palazzi 1995.

36The variations observed at the end of our work regarding the perception of the mosaic, have been measured with a test named “image colorimetry”  with the purpose of quantifying the color changes of the single tessera against the background of the laying mortar14. We selected a set of tesserae for each of the four colours used in the tests; for each group the l.a.b. coordinates were identified by calculating an average out of a rather large specimen of tesserae. Subsequently we singled out in each test an area of tesserae representing every color used: since each test was carried out using different laying mortars, each area was characterised by three laying beds. In such areas an image colorimetry was carried out again in order to find the l.a.b. coordinates of each area. At the end we obtained two different sets of results: the first one concerning the absolute colour of each set of tesserae, the second concerning the total effect of each set of tesserae in connection with each different interstice. From the comparison of the two measurments we obtained the ∆E with the aim of obtaining the effect of the chromatic variation of the tesserae in connection with the different laying beds.

Fig. 6 Graphic

Fig. 6 Graphic

The graphic shows the variation in perception of the colour of the tesserae laying on different setting beds.

37From the results as shown in graphic in ill.6 we observe that the concrete laying bed and the white laying bed produce fewer variations within the perception of the single tesserae. This result was somehow expected since both mortars have an achromatic carachter, this property in the end results in a very slight variation in the color perception of the tesserae themselves. This in turn is owed essentially to a stronger impact of the laying bed on the brightness rather then on the color tone of the mosaic.

38We observed that the stronger variations are usually connected with the use of the pink mortar, the only one which has a chromatic component.

39This property of the mortar determines certain colour variations in all the sets of tesserae we took as specimens, excluding the black ones. Indeed, as we know, achromatic colours modify the saturation data of a given colour; which concretely is perceived as a higher or lower brightness but never in a variation of color. On the contrary, the justaposition of chromatic colours (as in the case of the red tesserae on a pink mortar) determines a variation in colour shade. We observed a similar effect in all our sets of tesserae which appear to be affected in their colour in the first place by the pink mortar, then by the white mortar and finally, but in a lesser way, by concrete. This effect is changed only in the case of black tesserae on a white mortar, in which the main variation does not pertain in the chromatic sphere but in the sphere of brightness.

Virtual simulation

40Our test proceeded with the use of virtual specimens similar to the preceeding lab tests with the aim of enlarging the range of possibilities in the evaluations of the chromatic characters in a reintegration. With this purpose in mind the virtual simulation was performed through two distinct methods: a photographic and a graphic one. In the virtual simulation of ill. 7, the choice of colours – unusual ones in the sphere of reintegration- is ment to underscore if in a extreme way the importance that such a fact has within the general perusal of such a manufact. While being aware of the extreme nature of such a choice, never than less, we meant this as a mere demonstration intending to show more clearly the question here discussed. In ill. 7 are shown some of the meaningful cases.

Fig. 7  Virtual simulation

Fig. 7  Virtual simulation

This simulation of a mosaic laying on unusual colours shows relevant differences in perception of the colour of the mosaic, depending on the laying bed colour.

Simulation  F. Moro

41We used the same technique to propose possibilities of reintegration on a photographic detail of the Dionysos mosaic where the colour of the interstice (before restauration) was substituted each time with different colours, thus offering a possibility for examining chromatic reintegrations on the original manufact (ill. 8).

Fig. 8 Photographic simulation

Fig. 8 Photographic simulation

This simulation of reintegration is meant to examine different possibilities of chromatic re-integration of the original manufact.

Simulation  F. Moro

42The graphic system of simulation, in turn, allowed us to analyse the way in which the level of the interstitial mortar, in connection to the height of the tesserae, can modify the perception of a mosaic if observed from different viewpoints.

43In the construction of virtual models we started, in this case too, from photographs of laboratory tests: from these we obtained a 3D graphic reproduction in which we might vary the height of the tesserae and that of the interstice (ill. 9).

Fig.  9 Virtual model

Fig.  9 Virtual model

This model is constructed to show the variation in perception of a mosaic depending on mortar thickness.

Simulation  F. Moro

44The reintegration of interstitial mortar in the restauration of detached mosaics is usually aimed at their preservation. We wish to emphasize the concept that, the first datum to keep in mind in this  operation depends on its state of the conservation. The level of interstitial mortar must grant stability and permanence to the tesserae.

45From our experiment emerged that in the creation of the tests, the choice concerning the level of mortar application: underneath or at the same level of the tesserae surface, generates a different perception, when the eye is not perfectly perpendicular to the manufact.

46In ill. n.10 we simulated a point of view of 45° for the mosaic in which the interstitial laying mortar was set under its level on the right side and at the same level on the left side: one can notice immediately that in this latter area the mortar and hence its colour will have a higher relevance in the total composition.

Fig. 10 Perception of tesserae

Fig. 10 Perception of tesserae

The perception of the tesserae  depends largely on the colour chosen for the setting bed mortar.

Simulation  F. Moro

47Another fact which may cause different effects even when using the same mortar, is the state of conservation of the tesserae: to the use of new ones with perfectly sharp borders follows a regular texture of the manufact, in which the perception of full and empty spaces is balanced. However in the area of mosaics from ancient floors the tesserae have usually worn corners and edges, sometimes a consequence of the detachement procedures.

48In this circumstances the laying out of an interstitial mortar on the same level as the tesserae surface might unbalance the proportions.

The detachement of a mosaic

49In order to verify the alterations undergone by a mosaic following to its detachment, we decided to strip and successively replace the texture of one of the mosaics in a lab experiment (ill. 11).

Fig. 11: Lab test of mosaic detachment

Fig. 11: Lab test of mosaic detachment

The detachment test was made using traditional detachment techniques applied on a small dimension specimen, to reproduce the stress suffered by a mosaic during a stripping procedure.

Credit: Fabiana Moro

  • 15  See Fondelli 1992.

50We focused our attention on the analysis of the size variations. These usually appear as an enlargement of the interstitial space, thus provoking an alteration in the texture of the mosaic, a fact which changes the perception of colours of the whole artifact. We measured the consequent alterations through a hybrid monoscopic photogrammetry15.

51We then recorded the images of our test before and after the detachment and subsequently, by means of specific software, we measured the alteration which had occured in the test (ill. 12).

Fig. 12 Variations of measures registered after the detachment of the mosaic.

Fig. 12 Variations of measures registered after the detachment of the mosaic.

The test showed in a small scale the damages that a mosaic undergoes during a detachment procedure.

Credit: Fabiana Moro

52As a result, and as was expected, the whole surface of the mosaic was enlarged, the size of such a variation being proportional to the size of the manufact, in consideration to the dynamics that such an experiment implies. For example, one element worth considering is the strenght of the canvas used for the detachment: the weight of the tesserae and the strength used in the detachment cannot be uniform on the whole surface; besides the pulling force to which the canvas is subject is not easy to control in a large mosaic. In our case, a minor amount of deformation, was connected with the small size of the manufact and therefore a higher control on the whole procedure.

53The surface of our manufact did not undergo any significant changes in size; however, we noticed certain significant deformations. The inner decoration showed the highest degree of deformation, caused by the rolling of the mosaic on the tubular structure used for the stripping. The same sort of deformation occurred in the entire manufact. We can therefore conclude that the detachment had as a consequence a dishomogeneous variation in the size of the manufact, and a change in its symmetry.  


54Our work aimed at registering  the elements involved in the choice of interstitial mortars in mosaics that were detached and re-layed on new support. In being removed from their original support  they suffered from double damage: in the first place their removal involved a loss of information of their original support; in the second place, they suffered considerable damages, connected with their stripping, the treatment they underwent and their replacement on a new support.

55We also investigated the elements to be considered in selecting the right mortar for the reintegration and this because we believe that the primary needs in the treatment of a manufact are its conservation, a correct philological investigation of its history and a right perusal of  a manufact itself. Beginning from these primary needs, a number of elements must be taken into consideration; these, in turn belong in different areas which by their very nature cannot be mastered by a single professional capability. One wonders then if such elements should be considered one at a time or wether they are to be considered simultaneously. The nature of restoration, in consideration of the materials involved, is a matter in which the working out of a general protocol is not possible and which indeed implies each time a new elaboration and the application of different methods.

  • 16  Vitruvius, ed Corso1997; Balmelle-Darmon 1983, Donderer 1989, Dunbabin 1978, Dunbabin 1999, EAA 19 (...)

56Our research began with a philological hypothesis –an approach might appear rather unusual- however we believe on the strenght of our experience, that the interpretation of a manufact and the study of its elements are the necessary premises in order to repropose its  formal appearance16.

57In the case of mosaics, such an investigation is particularly important in as much as an inadequate attention to philological questions has been responsible, in the past, for considerable delays in the area of methods. Such a delay was considerable both on a theoretical ground and in the practice of restoration: a faulty knowledge of ancient manufacts leads to bad forms of conservation. We assume for instance, but it has not been the current practice to do so, that it is possibile to preserve the original setting bed, and that a practice of mimetic reconstruction and the substitution of the tesserae is not acceptable; even more important, workshop tests demonstrated that preservation in situ is definitely possibile. Therefore, we thought it might be useful to investigate into the relationship between restoration methods and a philological perspective concerning the works in question, so that the original connection between a manufact and its original context is understood in the most complete way- this is our contribution in the founding of a conservative procedure upon a correct theoretical basis.

58We believe it is most important to investigate into the “kunstwollen” (the wish for art) of the author of a manufact, as we are persuaded that this might help in its restauration: in this field too, techniques must be guided as much as possible by philological principles when conservation practices are deviced, and by the knowledge of  ancient civilisations. Such an investigation naturally has its limitations, as we tried to keep in mind and take into account all along.       

59In this particular case, in considering interstitial spaces we tried to understand to what extent it was possible to establish the relevance of such spaces. Indeed a mosaic should to be considered as an organic system, in which interstices between tesserae are a relevant element, and not simply a sort of net meant to add vibrations and charm to an image. Are they indeed a structural element in the representation and a part of the manufact’s history, with all its irreguralities and deformations? We believe one should not neglect such an aspect of a manufact in the context of both restoration theory and its practice, if we are aiming at a considered and respectful  action.

60A mosaic should be considered as a mixture of empty parts and full parts in which interstices are the empty parts and the tesserae the full ones, the interstices being not mere lines; they take some space and have their colour; hence they have a certain weight in the general economy of an image. This is why we considered the colour element of the interstitial mortar as having a function in the visual perception of a manufact; we took into account a number of specimens of interstitial mortars representing the variety of mortars used in preceding restaurations, in order to verify to what extent their chromatic characteristics may have influenced the perusal of the work. Such an influence was considered relevant in connection with the colours of the single tesserae.

61On the practical side we wish to make clear that the aim of restoration is not a sort of cosmetic treatment, but a support of the chromatic values of the mosaic, along with the avoidance of arbitrary new solutions; one must be careful in avoiding a vaguely intended wish for improvement as well as a general effect of cleanlyness and tidyness. There are no good or bad mortars—it is various interactions which produce different effects.

62One of the main considerations to keep in mind is naturally the colours of the tesserae and their interplay: it is fundamental to take into account a sort of chromatic hyerarchy in  the mosaic itself. For this reason the colours one is to select for a reintegration should belong to the range of those already existing. Through the knowledge of the phisical charachter of colour, one might extend the chromatic gamut to colours having the same shade as those of the manufact, although with a different  saturation and brightness.

63We noticed moreover how another characteristic of mosaics which influences their overall appearance is their type of texture; in turn this is related with the shape of the tesserae and the nature of  the setting bed. These elements have a capital part in the nature of a manufact, and concur in determining the final reading of a mosaic and; last but not least they bear witness of the conservative history of the manufact.

64One must not forget to consider the height and the width of interstices, as illustrated in the images  of the virtual simulations here proposed; the former in its interaction with light can lead to a variable perception of colours; the latter must be considered in connection with the amount of surface which it occupies. Their variations produce considerable consequences.

65Another consideration we made is regarding the type of perusal of a manufact; this, as we noticed, in our tests, is liable to modify considerably the perception of it: a part of our research was devoted to  investigating into the chromatic variations deriving from the distance of the observer.  Certain optical phenomena are related with the distance of the observer, i.e. a physiological merging of certain colours, at a given distance. The process of vision implied in these phenomena is bound to our physiology and it too, depends on a number of  causes. We must not forget that in certain cases correct premises may lead to unexpected results. To be sure our investigation was centered upon the characteristics of the Dionysos mosaic and is therefore referring to a limited range of colours.

66Similarly in our experiment we considered a limited number of mortars even if representative of the most frequent. A wider analysis might supply more specific directions and easier technologies. We only aimed at supplying the specialist with a means to quickly identify the elements upon which  to found his investigation. Again we do not indend to offer a systematic handbook of chromatic combinations but only a tool which may help in interpreting the available data in a correct way.

Haut de page


ALBERS, J., Interazione del colore: esercizi per imparare a vedere, Milano, 2005.

ARDOVINO, A.M., Problemi di restauro e di conservazione del mosaico, in: Atti del I Colloquio AISCOM, Ravenna 29 Aprile-3 Maggio 1993, Ravenna, 1994.

ARDOVINO, A.M., “Problemi di filologia e sviluppo della tecnica del restauro e della conoscenza delle tecniche antiche sul mosaico”, in: Atti del convegno di Bressanone, Bressanone, 2002, p. 23-31.

BALMELLE, C., DARMON, J.-P., “L’artisan-mosaïste dans l’antiquité tardive, Réflexions à partir des signatures”, in Artistes, artisans, et production artistique du Moyen Age, colloque international de la récherche scientifique, Rennes 1983. Vol. Les Hommes, Paris 1986, p. 235-253.

CASAGRANDE, F., Percezione visiva del colore “un caso di analisi del colore in ambito musivo”, Tesi di laurea in Conservazione dei Beni Culturali, Università di Bologna, a/a 2007-2008.

CORSO, A., ROMANO, E., Vitruvio De Architectura, Torino, 1997.

DONDERER, M., Die Mosaizisten der Antike und ihre wirtschaftliche und soziale Stellung, Eine Quellenstudie, Erlangen 1989.

DOPPELFELD, O., Das Dionysos-Mosaik am Dom zu Cologne, Cologne, 1962.

DUNBABIN, K. M. D., The mosaics of the roman North Africa, Oxford,  1978.

DUNBABIN, K. M. D., Mosaics of the greek and roman world, Cambridge, 1999.

EAA 1967- Enciclopedia dell’Arte Antica, Roma 1967.

FONDELLI, M., Trattato di fotogrammetria urbana e architettonica, Roma, 1992.

FREMERSDORF, F., Das Roemisches Haus mit dem Dionysos-Mosaik vor dem Südportal des Cologneers Domes, Berlin, 1956.

FROVA, A., Luce colore visione, perchè si vede ciò che si vede, Milano, 2004.

GOETHE, J. W., La teoria dei colori, Milano, 1981.

GREGORY, R. L., Occhio e cervello. La psicologia del vedere, Milano, 1998.

ITTEN, J., Arte del colore: esperienza soggettiva e conoscenza oggettiva come vie per l'arte, Milano, 1983.

MOORE, R. E. M., “A Newly observed Stratum in Roman Floor Mosaics”, in American Journal of Archaeology 72, 1968, p. 57-68.

PALAZZI, S., Colorimetria: la scienza del colore nell'arte e nella tecnica, Firenze, 1995.

PARLASCA, K., “Neues zur Chronologie der roemischen Mosaiken in Deutschland”, in  Colloque International du Centre National de la Récherche Scientifique, Paris, 1963, p. 77-80.

Haut de page


1  This article is a spinn-off of our degree research in Mosaic restoration, discussed on December 2009 at the Istituto Superiore di Conservazione e Restauro of Rome. The colorimetric measurements and the photogrammetric testings were carried out by the Physics lab of the ISCR of Rome with the help of dott. Fabio Aramini.

2  About the story of Dyonisos Mosaic see Fremersdorf 1956, p. 29; Dunbabin 1999, p. 82- 85; Parlasca 1963, p. 109-128.

3  See Dunbabin 1999, p. 79-91.

4  About the controverso on the datation see Dunbabin 1999, p. 81.

5  See Vitruvius, I , VII, ed. Corso 1997; Moore 1968, pp. 60-65. This composition would be diversified according to the raw materials available in different aereas: in central Itlay, for example, rather than rubble and pebbles, pozzolana sand was used as it was easy to find in that aerea and also because of  technical charachteristics of its hydraulicity. For the same hydraulicity necessities in other provinces of the Roman empire such as Gemania and Northern Africa ones, the sand was substitued by brick powder and potsherd fragments.

6  About the observation and study of this layer see Moore 1968, p. 60-65.

7  Fremersdorf 1956, p. 27; Doppelfeld 1962.

8  Ardovino 1994, p. 33-48; Ardovino 2002.

9  Ardovino 2002.

10  Frova 2004; Gregory 1998.

11  See Goethe 1981.

12  Albers 2005, p. 21; Casagrande 2008, Frova 2004, Gregory 1998.

13  Fremersdorf, describing the mosaic in the digging report writes: “ Seine Steinchen sind in einen roten Moertel eingebettet der, zwischen 2 cm (nach der mitte hin) und 4 cm (zum Rand hin) schwankt”. This information lets us presume that the mortar was composed by brick powder. See Fremersdorf 1956, p. 27.

14  See Palazzi 1995.

15  See Fondelli 1992.

16  Vitruvius, ed Corso1997; Balmelle-Darmon 1983, Donderer 1989, Dunbabin 1978, Dunbabin 1999, EAA 1967.

Haut de page

Table des illustrations

Titre Fig.1 The Dionysos mosaïc
Légende Ill. 1 The Dionysos Mosaic, Römisch-Germanisches Museum, Cologne.
Crédits Crédit: F. Moro   
Fichier image/jpeg, 52k
Titre Fig. 2 Samples of the modern tesserae used in our tests
Légende The tesserae chosen for the test were made by stones different from the original ones for origin, but similar in colours.
Crédits Crédit: F. Moro
Fichier image/jpeg, 16k
Titre Fig. 3 Test n.1 potsherds powder setting bed
Légende The tesserae are layed on a pink mortar, composed of lime and potsherds powder.
Crédits Crédit: F. Moro
Fichier image/jpeg, 44k
Titre Fig.4 Test n. 2 concrete setting bed
Légende The tesserae are layed on a cement setting bed.
Crédits Crédit: F. Moro
Fichier image/jpeg, 60k
Titre Fig. 5 Test n.3 white mortar
Légende The tesserae are layed on a white mortar setting bed composed by lime and yellow-gray sand.
Crédits Crédit: F. Moro
Fichier image/jpeg, 52k
Titre Fig. 6 Graphic
Légende The graphic shows the variation in perception of the colour of the tesserae laying on different setting beds.
Fichier image/jpeg, 32k
Titre Fig. 7  Virtual simulation
Légende This simulation of a mosaic laying on unusual colours shows relevant differences in perception of the colour of the mosaic, depending on the laying bed colour.
Crédits Simulation  F. Moro
Fichier image/jpeg, 112k
Titre Fig. 8 Photographic simulation
Légende This simulation of reintegration is meant to examine different possibilities of chromatic re-integration of the original manufact.
Crédits Simulation  F. Moro
Fichier image/jpeg, 52k
Titre Fig.  9 Virtual model
Légende This model is constructed to show the variation in perception of a mosaic depending on mortar thickness.
Crédits Simulation  F. Moro
Fichier image/jpeg, 60k
Titre Fig. 10 Perception of tesserae
Légende The perception of the tesserae  depends largely on the colour chosen for the setting bed mortar.
Crédits Simulation  F. Moro
Fichier image/jpeg, 104k
Titre Fig. 11: Lab test of mosaic detachment
Légende The detachment test was made using traditional detachment techniques applied on a small dimension specimen, to reproduce the stress suffered by a mosaic during a stripping procedure.
Crédits Credit: Fabiana Moro
Fichier image/jpeg, 44k
Titre Fig. 12 Variations of measures registered after the detachment of the mosaic.
Légende The test showed in a small scale the damages that a mosaic undergoes during a detachment procedure.
Crédits Credit: Fabiana Moro
Fichier image/jpeg, 85k
Haut de page

Pour citer cet article

Référence électronique

Fabiana Moro et Angelica Pujia, « The newly applied mortars in mosaic restoration », CeROArt [En ligne], 1 | 2010, mis en ligne le 18 novembre 2010, consulté le 31 mars 2017. URL :

Haut de page


Fabiana Moro

Fabiana Moro graduated in Decorative Arts in 2003 from the Fine Arts Academy in Rome. In 2009 she obtained her second degree from the Istituto Superiore per la Conservazione e il Restauro in Rome, specializing in the restoration of stone, stuccoes and mosaics. At the moment she works as a restorer at Palazzo Barberini in Rome.

Angelica Pujia

Angelica Pujia obtained her Degree in Classical Archaeology from RomaTre University, in 2004. In 2009 she graduated from the Istituto Superiore per la Conservazione e il Restauro specializing in the restoration of stone, mosaics and stuccoes. At the moment she is working at mosaic restorations in Leptis Magna, Libya and Djemila in  Algeria.   

Haut de page

Droits d’auteur

Licence Creative Commons
CeROArt – Conservation, exposition, restauration d'objets d'arts est mis à disposition selon les termes de la licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International.

Haut de page
  • Directory of Open Access Journals
  • Les cahiers de