Cartographic Style in the First Urban Maps of Cadiz, Spain: A Technique in Transition

ABSTRACT

This article deals with the cartometric analysis of various seventeenth-century urban maps of the city of Cadiz (Spain), from among which the so-called Vista Arámburu and the map belonging to the atlas of the Marquis of Heliche, discovered in the Krigsarkivet (Military Archive) of Stockholm, stand out for their uniqueness. These hitherto relatively unknown documents present evidence of an evolution of cartographic style towards greater topographic accuracy and hence cannot just be considered as simple drawings. In this seventeenth-century period of transition, the cartography of the city evolved from sixteenth-century aerial-view perspectives to the exhaustive planimetric maps of the eighteenth century, made by Spanish and French Military Corps of Engineers. These documents hold great historical value, not only due to the importance of Cadiz during the Modern Age but also because these maps constitute a graphic testimony of the fortification and growth of the city in this period.

KEYWORDS:

• Cartographic style
• seventeenth century
• Cadiz
• historical maps
• cartometric analysis
• planimetric accuracy

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on the contributor

Gabriel Granado-Castro is a professor at the Department of Engineering Graphics at the University of Seville. He received his PhD at the University of Seville. His teaching and research interests are in survey/topography and cartography and his two main areas of research are: maps of the city of Cádiz (Spain) in the Modern Age and the work of military engineers in the representation of territory. Using laser scanners and GIS have been particularly important in the development of his historical research.

ORCID

Gabriel Granado-Castro http://orcid.org/0000-0002-6386-0265

Joaquín Aguilar-Camacho http://orcid.org/0000-0003-1022-1061

Notes

1 In this respect, there are other similar cases of significant maps of strategic European ports, such as Dover, Chatham, Plymouth and Portsmouth in the British Isles (Smith, 2010: 54).

2 As a sample of this intense project activity carried out in this period, multiple drawings and partial representations of the city are currently preserved (Cámara Muñoz, 2014).

3 A particular example is the representation of Milan of 1475 by Pietro del Masai (Hidalgo García and García Doménech, 2017: 132).

4 One of the causes that can explain this lack of complete maps of the city can be found in the secrecy surrounding the development of this type of document, to which the State attributed great importance in the defence of its territories.

5 In the present paper, we have opted for the use of the term urban map to define the maps of Cadiz herein analysed, given that their characteristics bar them from being associated with the term topographic map, which is typically employed in the representation of broader land extensions where the primary purpose involves the representation of the topographic relief. In addition, the maps analysed lack certain elements that define a topographic map, such as scale bar, north arrow, type of map projection and geodetic coordinate system.

6 Cristobal de Rojas was one of the most important engineers of the Spanish empire who worked in the service of Felipe II and Felipe III. He is also noted for the development of one of the most notable fortification treaties of the era. Much of his work was carried out in Cadiz (de Mariátegui, 1985).

7 The schema of this representation corresponds to that used by engineers during the sixteenth and seventeenth centuries, in which the fortified perimeter was represented in ground plan and the buildings of the city were represented in three dimensions. This urban map of the very late sixteenth-century marks the trend that would be found in the maps of the city during the seventeenth century.

8 Joris Hoefnagel and Anton Van den Wyngaerde were two Flemish painters, both practically of the same period, who created collections of Spanish landscapes during the second half of the sixteenth century. The Hoefnagel collection belonged to his famous work Civitates Orbis Terrarum, promoted by Georg Braun, and the Wyngaerde collection was commissioned directly by Philip II.

9 For example, in the perspective view of 1513, the author did not hesitate to clearly alter the volume of the old medieval village in order to highlight it in the whole drawing (de Navascués y de Palacio, 1996).

10 The file is accompanied by a report describing the city and its defences, dated 11th January 1609, and made by the engineer Luis Bravo. It remains unknown whether this engineer also developed the map, a circumstance that we consider probable.

11 We refer to ‘oblique cavalier projection’ (also called ‘Elevation Oblique’) in those oblique drawings in which the actual elevation is projected in an angle or upward. In an elevation oblique drawing, the orthographic elevation view of an object or building retains its true size and shape. Instead we refer to ‘military projection’ (also called ‘Plan oblique’) in those oblique drawings in which the actual plan is projected upward, downward or in an angle. In a plan oblique drawing, the plan view of an object or building retains its true size and shape. Finally in a ‘trimetric axonometry’, the object is turned so that all three axes make different angles with the picture plane. In this case, orthographic plans and elevations cannot be used (Saleh Uddin, 1997).

12 Gaspar de Haro y Guzman, Marquis of Heliche, son of Luis Méndez de Haro, second valid of Felipe IV and great-nephew of the Count-Duke of Olivares, was one of the most prolific art collectors of his time. It appears that the commission of this atlas was for propaganda purposes at the service of his family (Sánchez Rubio et al., 2004: 21).

13 Within the classification that Fernando Marías makes of the chorographic views (Marías Franco, 2002: 102–106), we would place it clearly in the military views or plans with military elevation. In this case, as the only metric alteration, a possible change in the width of some streets, usual at the time, is detected in order not to overlap certain façades.

14 Although the analysis of positional accuracy has been made from a relative point of view and not in absolute terms (referring to the positioning of a map in a global coordinate system), it is interesting to analyse the orientation in each of these maps, since its study can help clarify its possible purpose (pictorial or cartographic).

15 The picture was divided into four parts photographed using a Panasonic Lumix DMC-GX8 digital camera, previously calibrated with the help of PhotoModeler software. With the support of a Leica FlexLine TS02 total station and this software, the idealized images were obtained (without the optical aberrations introduced by the lens). Finally, with AutoCAD 2016, the composition of the four shots was created as a single image (without changing the scale), on which this investigation is based.

16 This software offers three transformation options between the pairs of control points identified on the old and recent maps: Euclidean (four parameters or Helmert) and two Affine transformations (five or six parameters). This last option, according to Beineke (2001: 16–19) and Niederöst (2005: 70–73), has proven to supply the most reliable results in the assessment of the positional accuracy of historical maps since the differential bias of scale and orthogonality usually affect their supports and traces.

17 A list of more than one hundred publications on MapAnalyst applications in this field can be found at http://mapanalyst.org/press.html.

18 These visualization utilities are: distortion grids, displacement vectors and Mekenkamp circles and scale and rotation isolines. For more information, see Jenny et al. (2007).

19 The use of JPEG (Joint Photographic Experts Group), PNG (Portable Network Graphics) or GIF (Graphics Interchange Format) formats are recommended.

20 For the assessment of positional accuracy in historical urban maps, the mapping used by MapAnalyst provided by OpenStreetMap (OSM) is not the most appropriate since it lacks sufficient definition in urban environments (in that it fails to distinguish between pavements and buildings) and uses the Mercator projection that considerably distorts areas in medium and high latitudes (Jenny, 2010: 179). For more information on the quality of the mapping provided by OSM, see Haklay (2010).

21 For more information about this software, see http://www.aptop.com/.

22 The use of French system of scales was frequent in Spanish cartography during the Modern Age, due to the influence of the engineers that arrived from Flanders and France, particularly during the eighteenth century.

23 The average rotation experienced by the mesh when projected on the old map is obtained from a Euclidean transformation (of four parameters or Helmert) between the control points identified in the historical and the contemporary map (UTM-29 / ED50, Cadiz 2006).

24 Historical values estimated according to the International Geomagnetic Reference Field (IGRF12) model, from the Magnetic Field Calculators of the National Geophysical Data Center (National Centers for Environmental Information, NCEI) of the US Department of Commerce. Accuracy for the magnetic declination generally lies within 30 minutes (0.5°) of arc.

25 Particularly scarce are the reports published in the literature that deal with the assessment of the accuracy of the orientation of a historical map, mainly due to the limitations and difficulties that this entails. In this regard, see Pimentel Cintra (2010: 340–342) and the proposal made by Ballard (1982: 173–174) to date historical maps by means of magnetic variation.

26 A good example of this occurs when, in the analysis of the 1609 map (Figure 14), eight points located on the fortification of Santa Catalina (the five-pointed castle identified with the letter ‘L’ on the right-hand side of the map) are not included in the transformation calculations. In this case, the global RMSE of the map decreases to 4 ± 3 m. If the same analysis of the map is now reproduced with only the eight control points of Santa Catalina castle, then the RMSE is 12 ± 8 m.

27 In the literature, several tools have been postulated for the study and visualization of geometric accuracy (distortion grids, displacement vectors, isolines of scale and rotation, Mekenkamp circles and differential distortion analysis). However, their true usefulness is determined by the redundancy and representation of the number of ground control points identified on the map, used to calculate the parameters of a mathematical transformation model. For more information, see Mekenkamp and Koop (1986); Mekenkamp (1990); Jenny et al. (2007); Claeys Boùùaert et al. (2016).

28 However, the technique used for the drawing in perspective of the urban fabric sought to avoid the visual overlap between the adjacent buildings, which surely distorted the original planimetric accuracy of the inner plot in light of the need to exaggerate the real width of the streets for this purpose.

29 In addition, the deformation of the distortion grid and the length of the displacement vectors have also been verified in each case.

30 The image of the city portrayed on this map corresponds to the physiognomy of Cadiz between 1703 and 1705, during the War of Spanish Succession.

31 The haste with which Leonardo de Ferrari had to conclude and deliver the atlas to the Marquis of Heliche in the year 1655 is justified by John H. Elliott (Sánchez Rubio et al., 2004: 16–17) in the wake of the conflict that arose during that same year between the kingdoms of Spain and England, and by the urgent need of the Marquis de Heliche to ascertain the state of the border defences and ports of the Kingdom, due to the imminent war. In the unfinished atlas, there are striking omissions from the territory of Flanders, most of the Portuguese border, Spanish plazas as significant as those in Barcelona and Cadaques, and Santo Domingo or Havana in the West Indies.

32 The positional quality is employed in the assessment of the accuracy of the geographic information represented on a cartographic product based on its scale and purpose. Many countries establish the limit of visual perception (0.2 mm) as the reference standard to determine the quality of contemporary cartographic production (Baiocchi and Lelo, 2010: 98).

33 In the Spanish scale system of the time: 1 vara (0.835905 metres) equals 3 feet; 1 foot equals 12 inches; and 1 inch is 12 lines (Vila de Macabeo, 1855: 73, 77–78).

34 In the French scale system of the time: 1 foot (0.324839 metres) equals 12 inches (Martini, 1883: 466).

35 This was probably in order to better portray the buildings located inside and highlight the most important defensive element of the city during the seventeenth century. Furthermore, this castle is represented with an additional rotation (30–35° anti-clockwise) with respect to the rest of the map elements, possibly in order to better fit its scale increase within the composition of the picture.

36 This was usual practice in urban maps of Cadiz until the beginning of the nineteenth century, except for those made by French cartographers.

37 On the 1609 map, the compass rose identifies the cardinal points with the prevailing local winds in the region: Tramontana (T), Levante (L), Oeste (O) and Sur (S).

38 The phenomenon of magnetic declination had already appeared in a parchment map (1436) of Andrea Bianco (Cerezo Martínez, 1994: 12).