Cartographic analysis of two centuries of map printing using copperplates – examples from the Czech Republic and Malta collections

Research into old maps often focuses on manuscript or printed maps or atlases, which are the most frequently preserved cartographic products. From a process point of view, maps represent the final product following a series of connected activities and procedures. The initial phase covers data acquisition, processing, evaluation, and cartographic interpretation. The result is a manuscript map, or a model for print reproduction, which is usually considered an original work by the author. In the next phase, an engraver becomes involved, reproducing an original draft and influencing the original work with their style. The significance of these interventions is complicated to capture or quantify. Engravers were not just technicians; they were often skilled artists. Their role involved interpreting and translating the cartographer’s or mapmaker’s drawings and designs onto printing plates, typically made of copper. This process required a high degree of artistic skill and precision to accurately reflect the geographical information and artistic elements of the map (Woodward 1996).

A small number of these original drawings have survived (Čada & Vichrová 2009). Original printing plates, although still quite rare, can be traced more often. There are several options for examining this specific cartographic product, from the primary bibliography to the measurement of quantitative parameters to the examination of differences in the printed product. The presented study aims to provide a comprehensive inventory of preserved printing plates that the authors discovered in collections in the Czech Republic and Malta. Basic physical parameters were measured, the purpose of which was to compare their properties in relation to their different time and geographical origins. This study has the potential for further research into the technology used or the quality of the input material.

The analysis of copperplates used in historical map-making holds significant relevance for both geography and the history of science. Copperplates were extensively used for printing maps before the advent of modern printing techniques. By studying these plates, geographers can trace the evolution of map-making techniques. Analysis of old copperplate maps can reveal the accuracy of geographical understanding in different historical periods. The content and style of maps etched on copperplates can shed light on the cultural and political contexts. This includes understanding the interests and biases of mapmakers. Studying copperplates can also provide insights into the technological advances in geographical exploration and cartography. Copperplate illustrations often depict an intersection of geography, biology, astronomy, and other sciences. This can provide a holistic view of how different scientific disciplines influenced each other and evolved together. The artistic aspects of copperplate engravings, especially in scientific illustrations, demonstrate the historical interplay between art and science.

There is only a limited number of preserved map printing copperplates (see further in this paper) held in several archives and museums. As it is only possible to work with the original printing plates in a limited manner, many studies, therefore, focus on them indirectly. Based on the analysis of the prints, conclusions are drawn about the number of printing plates used for individual maps and the time interval of their use, as in the case of the Nuremberg cartographer Johann Baptista Homann (1664–1724) (Heinz 1993). The classification of engraving changes, the identification of printing plates, and the designation of the number of plates used and the condition they were found in is crucial for a correct carto-bibliographic description (Verner 1965, Campbell 1989, Baynton-Williams 2007).The procedures proposed in this paper were refined on the basis of the study of other maps, such as those produced by the Dutch publishing house Covens & Mortier (Egmond 2002). Also, the extent of plate scratches can be used to identify the different map states (Woodward 1994). Other contributions pay attention to the issue of the complexity of the engraving preparation of the printing plate in terms of finances and time (Dörflinger 1983). Therefore, the present article builds on previous efforts to achieve a uniform description and to provide a better understanding of their history, techniques of origin, engraving, and publishing practice (Inventory of surviving copper plates 1994).

The methodology was based on detailed measurements (in situ) of the dimensions of the plates as well as the inner and outer parts of the map frame (vertical vs. horizontal dimension in the middle of the plate edge, in millimetres). These data can be used to study the quantitative properties of the printing paper – such as changes in its size due to precipitation or poor drying after printing. These are the basic calibration values for cartometric analyses, which, in most cases, are performed on printed maps for objective reasons. The distance from the edges to the outer frame indicates the efficiency of the use of the input blank plate in relation to the future map format.

Another investigated quantity is the thickness of the plates. This was measured at the individual corners using a calliper micrometer at a distance of about 2 cm from the corner of the plate and then averaged (to the nearest tenth of a millimetre). Likewise, the average thickness was calculated for each set of plates, including indexes. Individual plates were weighed to the nearest gram. Each set as a whole, then, represents the average weight of 1 cm² of the plate expressed in grams and its standard deviation.

In order to obtain comparable data, the measured values were further standardised in the form of the surface area per weight (cm²/g or cm²/kg) or the bulk density of copper in grams per cm³. The obtained data allow for a comparison of the quality of copper used for the production of plates, from different time periods and places of origin, with regard to technological sophistication. The measurements are given in appendices 2 and 3.

For individual copperplate identification, a unique system was created. The unique abbreviations consist of two, three or four digits:

First letter from author’s surname (e.g. M = Müller, P = Petroschi)

The classification system and secondary use system is based on the work of Marc Hameleers (Hameleers 1989).We distinguish between maps for commercial purposes and maps intended for the needs of institutional cartography.

Commercial production emphasised financial profitability and marketability. In this case, the engravings were used to the very end of their life. In the case of irreversibly damaged content, the input raw material was usually melted down and reused. However, in order to maintain the availability of the product on the market, some publishers had one or more plates of the same map engraved, which were then used in parallel (Heinz 1993).

In the category of non-commercial producers of maps, we mainly include institutions connected with the ruling power or city authorities.

Based on the research, we identified interventions on the reverse side of the plates. Besides the cartographic materials, portraits can also be identified. A completely unique group is represented by oil painting on copper, for which old plates with a cartographic theme were demonstrably used as well.

Technology originating in Italy spread throughout Europe during the 16th century, thanks to the greater availability of metals. Oil paintings are also documented on copperplates with an engraving of portrait graphics (Schirò 2014).

Several dozen surviving printing plates were identified, including a fragment of three etched copperplates of the original fifteen-leaf plan of London, dating from about 1553–1559.

Two of them are owned by the Museum of London (Moorfields, Eastern City) and one by the Anhalt Art Gallery in the German town of Dessau (Western City).

Other printing plates are owned by the Plantin-Moretus Museum in Antwerp.

According to the museum’s printed guide, its collections contain around 3,000 copper engravings (etchings), mostly book illustrations.

In total 64 available copper printing plates from the collection in the Czech Republic and seven from the collections in Malta were analysed (for a detailed description, see appendices 2 and 4).The analysed material is divided into two sets according to place of origin (see Fig. 1).

Figure 1.

In set 1, we find products of commercial and non-commercial map production. MM, MBl, MBs and MCh are topographic maps designed primarily for the economic and administrative needs of the nobility, the monarch or the city council, who also covered the costs of their acquisition. We can therefore consider them as publishers. The military purpose of the maps was triggered by a future sequence of events (The War of the Austrian Succession (1740–1748), The Seven Years War (1756–1763)).

Several copies of the map with handwritten notes and an outline of the deployment of military regiments in Moravia, which are probably related to the War of the Austrian Succession, have been preserved at the Österreichische Nationalbibliothek in Vienna.

In the case of HMd, the production of the map was also commissioned from a supplier from the commercial sphere; however, only the engraving work and possible printing was commissioned. The map had already been sold by the client himself to potential customers from church circles. A completely unique finding is probably the trial ?Mt, which is registered in one collection together with HMd.

The printing plates from set 2 can be classified as commercial production. Most of them were engraved for the purpose of book illustration.

The dimensions of the individual printing plates plotted on the scatter plot (see Fig. 3) show the expected variability in width and height. Logically, the format of the printing plate depends on the size of the displayed area and the selected scale. The investigated set includes small-format maps for book illustrations (IV, IMa) as well as separate large-format maps (PaMa, MM, HMd, ?Mt, MCh or MBl). In particular, MBs and MM printing plates form a compact size file without significant deviations. MBl and HMd show higher variability. The position of the values plotted in the graph shows the deviation between the registers and the map part of MM. The discrepancy in both sets is also confirmed by other measured quantities (see figures 4 and 5). The obtained values indicate different times of origin and quality of production technology.

Figure 2.

Figure 3.

A linear relationship between the weight and the area of the printing plates is observable across the examined set (see Fig. 4). It is worth noting the distribution of MBs values, which form a relatively compact set in terms of the surface area of the individual printing plates. Unlike the graph in figure 3, they are not clustered around the centroid but stretched horizontally. The reason is the different weights of the individual printing plates. A similar phenomenon can be observed in MBl and, to a lesser extent, in MM and HMd. Here, however, greater variability is evident in the direction of the y-axis – namely, in the values of the area of the printing plates. The position of AMa and MMr indicates a slight deviation from the imaginary correlation line, which is caused by their lower mass in relation to the area similar to HMd and ?Mt.

Printing plates originating north of the Alps show less diversity in thickness than Mediterranean specimens (see Fig. 5). The measured values range from approximately 1.5 to 2.5 mm. In contrast, the thickness of some Mediterranean printing plates varies from 1 to 3.1 mm.

The scatter plot (see Fig. 5) shows further significant variability in MBs. Both quantities have a rather linear dependence. In contrast to the dimensions, there are significant differences in both the weight (approx. 3/4 kg) and the thickness of the plates (approx. 0.6 mm). The variability may indicate, above all, the primitiveness of the production technology, producing qualitatively different outputs, or differences in the original input products, using semi-finished plates. Sections I and VI, which have been completely overwritten over time, do not show measurable deviations from the rest of the set. A similar trend is noticeable for HMd between a four-leaf and a single-leaf overview map. Despite the minimal size difference, the overview map plate is noticeably heavier and thicker than the rest of the file. On the contrary, ?Mt is close in size, weight and thickness to a set of detailed HMd plates. Leaving aside the overall dimensions, MCh is very close to MBl in quality. The set of MBl printing plates shows a considerable variance of values in their thickness. The difference between the thinnest and thickest plate is similar to the 0.6 mm as seen in the case of MBs. However, the dependence on weight is not as clear-cut as with MBs, as even heavier plates have a low thickness. This relationship can be explained by the variation in the size of the individual plates (see Fig. 3). Graphs 3, 4, and 5 show the difference in the indicators between a set of printing plates from the Mediterranean and the northern areas, whether it is the overall size, thickness or weight. The only exception is PMa and PR (Rome), which is qualitatively similar to MBs (Augsburg). From this, it can be deduced that the input material is the same for two geographically different localities, which could be commercially interconnected.

The distance from the edge of the printing plate to the outer frame of the map drawing is very similar in height and width for both MBl and MBs files (see Table 3). The difference can only be seen in the distribution of deviations, which are more consistent with MBs. The overall composition of the map consists of five horizontal and five vertical map sheets. Their format is worth attention. For MBl, the height of the map field is always about 1.5–2 cm greater for plates from the first and last (fifth) row. The map field for the plates from the second, third and fourth rows is practically the same height. However, within each row, the map field is always about 1.5 cm wider at the first and last plate. At MBs, we observe the same trend, with the difference being of about 1 cm both in the height of the map field of the first and last row and in the width of the first and last plate in each row. For both the MBl and MBs with the largest number of files, there is an apparent shift in the engraver’s practical experience in planning the format of the map field with respect to the size of the printing plate. The total material losses in the form of an unused edge are practically the same. However, they differ significantly from piece to piece in MBl, while the individual MBs plates have a consistent ratio of edge distance from the outer frame (see Table 3).

Figure 4.

Figure 5.

The box plot diagram shows the values of the ratio of the area of the printing plates in cm²/kg (see Fig. 6). The values of monolithic maps are also displayed for comparison across the entire file over time. In the case of multiple sets of printing plates, the variability within the MBs is evident, which corresponds to previous findings. The asymmetrical form of the graph indicates the possibility of using more input materials of different quality and origin. To a lesser extent, this is also the case for HMd and MBl. The mean HMd and ?Mt are almost identical. It can be rightly assumed that even the test engraving, without any further identification, comes from Homann’s publishing house in addition to HMd. The position of the MM and MMr graphs is completely different. The measured values correspond to the archival records of the different acquisition times of the map printing plate and the topographical register. AMa is completely out of the range of the rest of the set. From its position, it is clear that it is a very thin printing plate, which is confirmed by the thickness measurement itself (c.1 mm). Leaving aside the position of the MMr in the set of printing plates from the area north of the Alps, the position of the mean values is again less variable than in the case of the Mediterranean specimens. The only difference is the PMa and PJ values, which are practically identical to MBs or MM. In the case of these specimens of different geographical origin, it is possible to look for a common source of input material. MCh and MBl have similar values. The variability within a set of Mediterranean printing plates cannot be reliably explained; perhaps it is due to varying sources of input raw material or the inconsistent quality of copperplate production technology.

Figure 6.

The distribution of the frequency of plate area values per kilogram is shown in Figure 7. The distribution is not completely symmetrical. It stretches in the direction of the maximum values. Values in the range of 551–600 cm²/kg are most often represented. The graph shows the highest concentration of values around the average, with measurements occurring even above 700 cm²/kg. The area-to-weight ratio of AM (c.1,200 cm²/kg) can be considered as a complete outlier.

The achieved results were primarily conducted using information from the Maps as Prints… publication (Woodward 1996). The search conducted by the authors did not reveal other sources that would at least partially work with the physical parameters of the preserved printing plates to this extent.

Qualitative differences in the examined set of printing plates are not the result of rapid technological progress. The change in the technology of copperplate production has been recorded since the middle of the 16th century when rolling copperplates replaced hand-beaten ones (Woodward 1996). The reason for the differences can be found more in the local aspects of the quality of the production process and in the source raw material, including its availability.

From the point of view of the dimensions of the printing plates, it is obvious that the produced formats of copperplates should be compared. In many cases, the engraver used the area of a standard copperplate practically to its full extent. If we consider the maximum 70 × 50 cm format (imperial size), MM, MCh, HMdand ?Mt used this format. MBl printing plates are also of similar dimensions (Woodward 1996). Here, even at the cost of more waste, the engraver reduced one side by about 10 cm. The smaller available format of copperplates c.29.5 × 43.4 cm was used by the engravers MBs, PMa, and PR (Woodward 1996). Some of the examined plates have atypical dimensions, especially those adapted for book illustrations–AMa, PJ, IMa, IV or PaMa. In these cases, the input copperplate was cut to the required dimensions. For multi-leaf maps MBl and MBs, even due to differences in the size of the engraving and the printing plate, it is possible to observe the natural development of the engraver’s skills and experience, which he underwent within six years of their creation.

The issue of accurate plate thickness measurement is debatable from several angles. The selected method finds limits in the device used. In order to be able to consider the measured thickness values as representative, it is appropriate to convert them to another indicator –for example, bulk density. Unlike the density of the metal, this assumes a certain degree of material inconsistency in the form of pores, cavities, and so on. However, this indicator is sufficient for a basic comparison. The density of copper corresponds to c.8.96 g/cm³. The average value of the measured set isc.9 g/cm³. This value can also be validated thanks to the preserved printing plate of a map of a part of South America by G. Sanson and A. H. Jaillot from 1674 (Davis-Allen & Reinhartz 1990). Its bulk density is c.8.5 g/cm³. Therefore, with the exception of a few outliers, the measured thickness values can be considered relevant (see appendices 2 and 3). With regard to geographical aspects, this printing plate appears to have a more compact set of plates from the area north of the Alps. The standard deviation corresponds to approximately 1 g/m³ compared to c.2.7 g/m³ in the Mediterranean area. The smallest values of bulk density were achieved for printing plates that were engraved in Valletta (PaMa, IMa, IV), geographically furthest from the source of the raw material. The mines in Hungary (today’s Slovakia) produced some of the main copper ore deposits, but the Italians also used other sources, such as from the Tyrol area (Woodward 1996). This finding offers space for additional research into the geographical correlation between the quality of the copperplates used and the place of extraction of the input raw material. It is necessary to take into account the fact that the cited data work within the time period of the 16th and 17th centuries. The comparison of the two largest sets of printing plates, MBl and MBs, which were prepared by the Augsburg engraver Michael Kauffer over a period of six years, brings interesting results. MBl reaches an average bulk density of 8.5 g/cm³, MBs 9.8 g/cm³. The size of the standard deviation does not exceed 1 g/cm³ for both equally large groups. It is worth noting the noticeable change in the quality of the material in the form of copper purity in the very short term. As early as the 16th century, most of the copper trade was controlled by the Bavarian cities of Nuremberg and Augsburg (Woodward 1996). It is a matter of discussion whether the cause can be found in a qualitatively different source of copper ore or in the shortcomings of the metallurgical process of its extraction. However, from the achieved results, it is not possible to deduce the differences between the two cities, Nuremberg and Augsburg, in terms of preference for the materials used and their resources. The occurrence is rather accidental, as the same engraver (Michael Kauffer) worked in the case of MBl and MBs with qualitatively different materials in a relatively short period of time (1720–1726).

In addition, in the case of MBs, the measured parameters (thickness, surface area/weight) indicate the variable quality of the copperplates within the set. It is less likely that the engraver himself would prepare a plate for engraving from copper ingots. This activity fell under the responsibility of a specialised metalworker (Woodward 1996). It is also possible to produce an alloy of copper and other metals. Such a connection can be found more for printing plates PaMa, IMa, IV or lower MMr, whose bulk density reaches values between 4.1 and 6.5 g/cm³. However, in this case, the influence of the quality of their thickness measurement is representative of the values obtained. MM (10.2 g/cm³) also achieves similar values as MBs, of which it is known only that the engraving was prepared by the Brno engraver Johann Leidig, originally from Swabia (the capital of which is Augsburg). The HMd file from Homann’s Nuremberg publishing house has an average value of 8.7 g/cm³. This is a very similar value to Augsburg’s MBl production, only with a time span of less than 40 years.

Determining the problematic thickness can be replaced by measuring more reliable indicators, namely the dimensions and weight of printing plates. When converting the surface area of the printing plate to one kilogram, c.86% of all values lie in the range of 501–700 cm²/kg (see Fig. 7). When values of less than 500 cm²/kg are taken into account, it is already about 90%. The maximum values are interesting, most of which are in the range of approximately 700–800 cm²/kg, and one complete outlier value of 1,200 cm²/kg for AMa. A large deviation of MBs confirms the inconsistency of the file. In the case of HMd, the higher deviation is caused by a different value for the four-leaf and overview map of the diocese. A slight difference is also evident in other indicators (see figures 4 and 5). The MCh value is different compared with other plates (HMd, ?Mt) originating from Nuremberg. It is closer to the Augsburg MBl and MBs, which date from about the same time period. The discussed source reports the measurement of the same quantity in seven plates from the Atlantic Neptune atlas (Woodward 1996). The values range from 589 to 773 cm²/kg. In the case of our group, the average value of 613 cm²/kg is closer to the lower limit of this interval. Values above 700 cm²/kg tend to have the character of remote measurements. Qualitative differences are evident from this basic comparison. As with the work of J. Ch. Müller, 64 printing plates have been preserved from the Atlantic Neptune atlas. Both works roughly cover the same time period. There is potential for comparing two comprehensive sets of different geographical origins.

The performed research confirms earlier conclusions about the different period of origin of MM and MMr. Assertions based on archival materials can be supported by the results of the exact measurements of the physical properties of both files. However, the position of MMr on the individual graphs is so different from other printing plates that it is not possible to perform even indirect time identification.

The chosen methodology for measuring the thickness of printing plates can be considered satisfactory, and the obtained values are representative. While maintaining the quality of measurement accuracy, it is not absolutely necessary to approach more sophisticated technologies (e.g. 3D scan), which are not always easily accessible in connection with the storage of archival material.

The conclusion of this research into copperplates encompasses a range of important insights into the cartographic practices over two centuries, particularly focusing on collections from the Czech Republic and Malta. The study involved a comprehensive analysis of 63 copper printing plates from the Czech Republic and seven from Malta, including measurements of their dimensions, thickness, and weight. This provided valuable data for understanding the variations in printing paper and the overall efficiency of the map printing process.

A key finding of the study is the notable variability in the dimensions and quality of copperplates used in historical map-making. This variability was influenced not only by technological advancements but also by local production quality and the availability and quality of raw materials. The research highlighted significant geographical differences in the characteristics of the plates, with those from north of the Alps showing less thickness diversity compared with their Mediterranean counterparts, indicating regional variations in copperplate production processes and raw materials.

The study also distinguished between commercially produced maps, which emphasised profitability and marketability, and non-commercial maps produced by institutions connected with ruling powers or city authorities. This distinction is crucial for understanding the historical context and purpose behind the maps. Additionally, the research uncovered how copperplates were used and reused, including the melting down of damaged plates and the use of the reverse side for other illustrations, such as portraits or oil paintings.

Figure 7.

A unique system was developed for identifying individual copperplates, taking into account factors such as the author’s surname, the area covered, and specific notes. This system aids in the classification and categorisation of the plates, which is essential for correct carto-bibliographic description. The detailed measurement of physical properties such as plate thickness, weight, and surface area allowed for comparisons across different time periods and origins, shedding light on the technological sophistication used in map printing.

The findings of this research open avenues for further exploration into the technology of map printing, the quality of input materials, and the influence of geographical factors on the production of copperplates. They also exemplify the use of detailed physical measurements as a methodological approach in cartographic history research, providing an example for future studies in this field.

Overall, this research significantly contributes to the understanding of historical map printing techniques, particularly the use of copperplates, and provides a foundation for further exploration in the field of cartographic history.