Mercator- Bearing straight (IIIa)

Revealing Replicas

Few historic maps demonstrate as dramatically as Gerard Mercator's 1569 world map that size is both an asset and a liability. Printed on eighteen separate sheets and measuring 80 by 49 inches (202 by 124 cm) when fully assembled, its abundant space easily accommodates the detailed coasts and continents that proved a valuable source for smaller, less detailed world maps by Ortelius and Hondius, among others. Its suitability as a wall display, vulnerable to light and dirty fingers, partly explains the small number of surviving copies—a mere four complete sets if you include one lost in World War II—as well as its rare appearance as a facsimile illustration in books about old maps and cartographic history. When reduced to a black-and-white page-size halftone with noticeable lineations where its sheets meet, the famous chart becomes a disappointingly drab demonstration of its illustrious author's skill as a cartographer.

Several illustrators devised clean, book-friendly replicas by transcribing the map's key elements at a smaller, more manageable scale and adding labels describing its larger blocks of text. The resulting map of a map (so to speak) not only affords a concise graphic summary of its prototype's geography but also allows for further reduction, as in figure 4.1, which reproduces at a still smaller size the version drafted by famed British engraver Emery Walker for the eleventh edition of Encyclopaedia Britannica, published in 1911. The warped frame of the Britannica image, which measures 7 inches wide, suggests it might have been picked up photographically from an earlier illustration, perhaps in another publication. Despite this flaw, Walker's diagram captures the essence of Mercator's grid, layout, continents, and obvious belief in a southern continent (Antarctica) and northeast and northwest passages across the Arctic. Similar renderings with English translations of legend labels illustrated a 1969 science article commemorating the great map's four hundredth anniversary and a 1947 guidebook on map projection (The Round Earth on Flat Paper) from the National Geographic Society. Although these small quasi facsimiles lack the authenticity of Mercator's intricate engraving, they illustrate nicely the progressive poleward separation of his projected parallels.

Full-size, eighteen-sheet facsimiles afford a more realistic impression of the famous mapmaker's attention to detail. The oldest is part of Les monuments de la géographie, a collection of twenty-one facsimile maps published between 1842 and 1862 by Edmé-François Jomard (1777–1862), a Parisian geographer who had served with Napoleon's 1798 Egyptian expedition. In 1828 Jomard founded what eventually became the Bibliothèque Nationale's Department of Maps and Plans, which owns one of the surviving copies. The municipal library in Breslau (now the Polish city of Wroclaw) furnished the originals for the Berlin Geographical Society's 1891 Drei Karten von Gerhard Mercator, which includes the 1569 world map as well as Mercator's 1554 fifteen-sheet map of Europe and his 1564 eight-sheet map of the British Isles. Destroyed during World War II, the Breslau copy also served as the prototype for a commemorative edition published in 1931 by the International Hydrographic Bureau, headquartered in Monaco. The third known copy, in the Maritime Museum at Rotterdam, was reproduced in 1961 (a year before Mercator's 450th birthday) and distributed as a supplement to Imago Mundi, the principal scholarly journal for historians of cartography. A fourth copy, at the University Library of Basel, Switzerland, appeared in Imago Mundi in 1955 as a much-reduced 20 by 13 inch (51 by 33 cm) foldout.

Life-sized reproductions of Mercator's 1569 world map are not widely available in libraries and map collections. A joint search of the RLIN (Research Libraries Information Network) online catalogue and the OCLC (Online Computer Library Center) FirstSearch database, which focus on university libraries in the United States, failed to turn up a single copy of either the Jomard facsimile or its 1931 Monaco counterpart, although the National Union Catalog, printed in the late 1960s, found ten of the former but none of the latter. Many American libraries apparently did not bother to add older materials to their electronic catalogues. By contrast, my Web search uncovered twelve copies of the Berlin Geographical Society's 1891 version in American libraries—one more than the printed bibliography. Oddly, the Library of Congress (which apparently does not share all its holdings with RLIN or OCLC) owns both the Jomard and the Monaco editions but lacks the 1891 Berlin reprint. The 1961 Rotterdam facsimile fares much better: the Library of Congress has one, the National Union Catalog found nineteen more, and an RLIN/OCLC search uncovered another sixty copies in North American and Europe. The list would no doubt be longer had Imago Mundi automatically sent its Rotterdam "supplement" to all subscribers. Syracuse University, where I teach, has a run of Imago Mundi that starts with volume 1 (1935), but lacks a full-size facsimile of Mercator's 1569 world map. Yet the SUNY College at Cortland, forty miles south, has a copy, which I now know of thanks to FirstSearch.

If you think all "full-size" facsimiles look alike, think again. Although Gerard Mercator engraved eighteen separate plates and printed his 1569 world map on eighteen separate sheets of paper, the 1961 Rotterdam reprint consists of a large, atlas-like portfolio with fourteen huge pages that reformat the planet into geographically coherent chunks. A complex diagram (fig. 4.2) in the accompanying sixty-nine-page guide describes the scheme. The solid lines and large, bold numbers,1 through 18, represent Mercator's original layout. The dashed lines represent the reformatted "sheets," also numbered 1 through 18. Some of the new sheets are smaller than others and share a page in the facsimile atlas with another sheet. Note too that Mercator's original plates were in portrait format (longer vertical axis), whereas the facsimile is in landscape format.

If this reformatting seems needlessly confusing, consider the inconvenient boundaries between Mercator's original plates. Anyone familiar with U.S. Geological Survey topographic maps knows the problem of quadrangle boundaries that invariably partition our area of interest, however small, among two, three, and sometime four map sheets. Mercator's 6 by 3 grid validates this cartographic variant of Murphy's Law with horizontal plate boundaries that separate England from Scotland and chop off the southern tip of Africa while vertical boundaries slice through east Africa and what is now the eastern United States. Instead of retaining Mercator's original layout—a benefit only if you're obsessed with authenticity or eager to decorate a wall—a benevolent editor chose to preserve the integrity of continents and other large regions. Thus a historian interested in medieval Europe need only examine facsimile sheet 10, reformatted from plates 4, 5, 10, and 11 to include all of Europe (including Scandinavia) and the entire Mediterranean coastline. Similarly, an Africanist can concentrate on sheet 1 1, reformatted from plates 10, 1 1, 16, and 17 to include all of Africa and the Arabian Peninsula. This useful redundancy excuses the omission of tiny portions of Antarctica and northern North America, where equivalent details are missing or inconsequential, as well as the museum's admission that "the size of the maps in the reproduction is slightly reduced."

If you think it unseemly for anyone to muck around with the great mapmaker's original layout, you'll not be surprised that the meddlesome editor of the Rotterdam copy was none other than Gerard Mercator, who willingly sacrificed three prints of some plates in reformatting his huge wall map into an atlas of coherent pages, not arbitrary quadrangles. To provide a meaningful geographic scope for each sheet, Mercator cut out appropriate portions of prints covering the region in question and mounted them on large, folio-size sheets of paper. Although the guide ignores the composition and condition of Mercator's adhesive, lines between adjoining pieces are readily apparent in the 1961 reprint. Repackaged as a portfolio atlas, the Rotterdam copy is notably different from its intact, eighteen-plate cousins in Basel and Paris, and the Maritime Museum owns the only copy so arranged.

A handwritten letter from Mercator found with the atlas indicates he assembled it around 1578 at the request of Werner von Gymnich, a wealthy patron. Sometime later a von Gymnich married a von Mirbach, and the atlas moved to the Mirbach estate at Harff Castle, near Cologne, where it evaded the public gaze until the late nineteenth century. Discovered in the course of an inventory of the Mirbach family library, the atlas was exhibited briefly in 1894, written up in a Frankfurt newspaper in 1902, and mentioned in academic articles in 1911 and 1930. Pieced together from printed map sheets, it earned recognition in Leo Bagrow's seminal History of Cartography as "the first printed sea-atlas."

Priceless objects become disposable assets during hard times. In 1932, amid the economic and political turmoil that eased the Nazi rise to power, the atlas appeared in the catalog of an auction house in Lucerne and caught the attention of the Maritime Museum's director, J. W. van Nouhuys. Eager to acquire a rare work of immense importance to cartography and navigation, van Nouhuys decided to attend the sale. On the way he stopped at Basel to examine the copy in the university library. His hopes rose when an overly optimistic auctioneer opened the bidding at 6,000 Swiss francs (3,000 guilders), raised the price twice, and failed to find a buyer. Convinced he had a chance, van Nouhuys returned home, found two wealthy contributors each willing to match the museum's 1,000 guilders, and mailed in the winning bid. The museum got its atlas.
On at least one other occasion Mercator repackaged a wall map as an atlas. The evidence is a bound collation of regionally reconfigured cutouts from his maps of Europe ( 1554) and the British Isles (1564) as well as portions of the 1569 world map and thirty additional sheets from his friend Abraham Ortelius's 1570 world atlas. Dutch schoolmaster Thomas Varekamp, who discovered it serendipitously in a Belgian used-book shop in 1967, reckons that Mercator assembled the atlas in 1571 for his patron Werner von Gymnich, who undertook a lengthy tour of Europe the following year. As with the Rotterdam sea atlas, Mercator willingly cannibalized his wall maps to get the right framing. The nine sheets he assembled by cutting up four copies of his European wall map are especially rare because the last known intact copy was destroyed in 1945, during the siege of Breslau, along with the municipal library's copy of the 1569 world map. A pair of manuscript maps of northern Italy—the only surviving examples of maps hand-labeled by Mercator—makes the atlas triply unique. Robert Karrow, who considers it "the most important Mercator discovery of the twentieth century," once lamented the atlas's purchase in 1979 by the British Rail Pension Fund, which persistently refused scholars' requests for close inspection. But not any longer: a 1997 grant from the Heritage Lottery Fund enabled the British Library to purchase what is now known as Mercator's Atlas of Europe and endorse a facsimile reprint, which appeared a year later.

If you want your own copy of the 1569 world map and are willing to settle for less than full size, Respree.com, a Los Angeles dealer in posters and reproductions, advertises two poster versions at its Web site, one 24 by 31 inches and the other 37 by 54 inches. Reproduced in colour from an unidentified original, Respree's posters make attractive wall decorations but lack the detail of Mercator's lines and labels, better captured by a full-size black-and-white facsimile.
If you crave fine details, don't mind low-resolution scanned images, and can appreciate a well-illustrated online celebration of Mercator's projection, all in German, check out the "Ad Usum Navigantium" page of Wilhelm Krücken's Web site, "Ad maiorem Gerardi Mercatoris gloriam" (www.wilhelmkruecken.de). In addition to an incisive exploration of the projection's mathematical properties, Krücken provides individual screen-size images of all eighteen plates. Although the map's Latin inscriptions are barely discernible, its geography and artistry are clearly apparent. Anyone with a high-speed Internet connection can move from plate to plate more readily than a library patron handling (carefully, I hope) eighteen oversize map sheets. It was during a cursory interactive examination of Krücken's facsimile that I first understood the great mapmaker's appreciation of modest amounts of redundancy. Unlike contemporary cartographers who give you the whole world only once, Mercator cleverly extended his left-hand plates a few degrees west of his 180th meridian and his right-hand plates a few degrees east to provide dual, alternative images (fig. 4.3) of what he considered the most likely position of the north magnetic pole.

For scholars concerned with a map's lines and labels, an accessible black-and-white facsimile is often more valuable than a rare hand-coloured print ensconced in a distant library. That's clearly the view of Swedish scientist-explorer Adolf Erik Nordenskiöld (1832–1901), who inspired map historians with his Facsimile Atlas to the Early History of Cartography, published in 1889 and reprinted in paperback in 1973. Unable to include Mercator's 1569 world map because of its huge size, he reported an analysis of its progressively spaced parallels. One column of a table contains distances from the equator calculated for a spherical earth according to the mathematical formula for the Mercator projection, and a second column lists the corresponding distances as measured on the Jomard reprint. The two columns are nearly identical up through 200, but a discrepancy apparent at 300 grows progressively larger with increased latitude. In attributing this discrepancy to "the imperfection of the mathematical resources of the map-constructors in the middle of the 16th century," Nordenskiöld suggests that Mercator employed a mathematical approximation that should yield a closer correspondence between the calculated and measured distances. Remaining differences, he argues, "can be explained by engraving-errors or by stretching's in the paper"—a persistent source of uncertainty when working with facsimiles of old maps.

Paper shrinkage is less troublesome in exploring Mercator's reliance on other mapmakers. In announcing the 1931 Monaco reprint by the International Hydrographic Bureau, Britain's Geographical Journal noted several dubious debts to Ptolemy, including a Niger River that connects with the Nile. Bert van 't Hoff, who prepared the guide accompanying the 1961 Rotterdam facsimile, listed other prominent influences, which are equally apparent on original copies and reprints. Especially noteworthy is Abraham Ortelius's 1564 world map, on which inscriptions nearly identical to their counterparts on Mercator's 1569 world map suggest that Mercator and Ortelius exchanged information or consulted the same sources. Impressed with an original print of the Ortelius map in the library at Basel, van 't Hoff observed that "this beautiful and remarkable map deserves to be reproduced [and] also compared in detail with Mercator's map." He also recommended looking at Diego Gutierez's 1562 map of South America (fig. 4.4, left), the likely source of Mercator's erroneous westward diversion of the coastline for what is now southern Chile (see fig. 4.1).

The straighter, more accurate rendering of the Pacific shoreline on Mercator's 1538 world map (fig. 4.4, right) confirms the adage that new is not always better.

One map certain to warrant a facsimile reprint—if it's ever found—is a larger version of Erhard Etzlaub's small maps of Europe and North Africa on what looks like a Mercator projection (see fig. 1.7). As I noted in chapter 1, two tiny maps produced in Nuremberg in 1511 and 1513 exhibit progressively spaced parallels, which suggest a deliberate attempt to straighten out rhumb lines. Facsimiles of these maps, each engraved on the cover of a portable sundial, have made map historians wonder what Etzlaub was up to and whether his tiny maps had influenced Mercator. In 1918, in a short note titled "Who Originated Mercator's Projection?" the Geographical Journal reported the opposing views of two German professors, Drecker and Hammer. Convinced that Etzlaub's maps were not flukes, Drecker believed the Nuremberg instrument maker had merely reproduced miniature versions of a much larger map, yet to be discovered. Hammer, who questioned Etzlaub's understanding of principles underlying the Mercator projection, wondered why no sixteenth-century geographer or mathematician had ever discussed his alleged innovation.

Curious about current views among European map historians, I put the question of Etzlaub's influence on Mercator to Ingrid Kretschmer, who teaches the history of cartography at the University of Vienna. She began by noting that the University of Duisburg—the duke's idea was eventually fulfilled—celebrated the four hundredth anniversary of the great mapmaker's death in 1594 with three symposia, in 1992, 1993, and 1994. The series stimulated an intense reanimation of the Mercator's work, including a careful comparison of his and Etzlaub's maps by mathematician Wilhelm Krücken, who maintains the Mercator Web site mentioned earlier. According to Kretschmer, a detailed examination of graticules convinced Krücken that the two cartographers had applied different principles. Renewed interest in Etzlaub's influence failed to uncover a larger version of his tiny sundial maps—the several, larger scale road maps he published are all laid out on an equirectangular grid. More telling is Etzlaub's apparent failure to tout his accomplishment in a published article or private correspondence. In Kretschmer's opinion, it is "rather unlikely that a famous instrument maker and cartographer like Erhard Etzlaub would not have mentioned [his development of ] a new projection."

Although Etzlaub's influence on Mercator remains a mystery, the great mapmaker might well have been inspired by the work of Pedro Nuñes (1502–78), a Portuguese astronomer and mathematician who described loxodromic spirals in 1537. In pointing out that a direct course is usually not the most easily followed course, Nuñes criticized globe makers for confusing great circles (orthodromes), which are easily delineated on a globe by a taut thread, with rhumb lines (loxodromes). Unless aligned with a meridian or parallel, a rhumb line is a comparatively complex corkscrew curve.

Mercator knew about loxodromic spirals as early as 1541, when he included a multitude of these curved lines of constant direction on his famous terrestrial globe (fig. 4.5). How he did this invites speculation: Mercator never described his method, and scholars have yet to uncover an earlier prototype. Dutch map historian Johannes Keuning concluded that Mercator simply drew his loxodromes graphically, "with the aid of metallic loxodromic triangles, made by himself." Using thin metal triangles or templates crafted to ensure constant bearings seems both obvious and ingenious. I'm surprised that Keuning, after proposing a graphic, ad hoc solution for placing loxodromes on a globe, did not dispute Nordenskiöld's assumption that Mercator used a mathematical approximation in laying out his 1569 world map.

Although mathematics provides a theoretical foundation for map projection, not all solutions are numerical. It's easy to see how a map-maker sufficiently clever to delineate curved loxodromes on globe gores could have straightened them out graphically on a flat map. Quite simple, in fact, according to Robert Karrow, who summarizes a likely Mercator recipe in a single sentence: "By following his curved rhumb lines and noting the longitude at which these lines crossed the various parallels, then transferring these coordinates to flat paper using straight rhumb lines with the same bearing, he would have obtained the basic framework of his projection." As William Warntz and Peter Wolff point out in Breakthroughs in Geography, Mercator's genius lay in believing a solution existed. It might well have happened, they note, "that the requirement that rhumb lines should be straight lines could not be satisfied on any chart or map."

If copying is a clue, other mapmakers appreciated Mercator's genius less readily than his scholarship. Although no one adopted his projection for nearly three decades, his 1569 world map was a key source for his friend Abraham Ortelius's influential Theatrum Orbis terrarum, published in 1570. According to map historian Peter Meurer, Ortelius based eight of his plates on Mercator's map. A few writers have interpreted these similarities as blatant plagiarism, but most note that Mercator and Ortelius exchanged information and consulted similar sources. Ortelius was not the only mapmaker to rely on Mercator's geography. Features and place names from the 1569 wall map (but not its projection) are readily apparent on the world map accompanying the Polyglot Bible, published in Antwerp by Benito Arias Montanus in 1572, and the "Planisphere," a world map published in Antwerp and Amsterdam by Petrus Plancius (1552–1622) in 1592.
Mercator's 1569 world map proved a convenient source for his son and grandsons, who compiled continental maps for the posthumous 1595 edition of his famous Atlas, which also includes a world map his son Rumold had published separately in 1587. Laid out on a pair of hemispherical projections, Rumold's map is a smaller-scale generalization of his father's much larger 1569 chart. Paradoxically, not a single map in the epic 1595 Atlas is on a Mercator projection.

Another paradox is the full-size 1574 woodcut reprint of the 1569 world map by Bernard van den Putte, an Antwerp engraver. With the apparent approval of Mercator, van den Putte reengraved all eighteen plates by cutting out non-inked areas on blocks of wood analogous to massive rubber stamps. Although coastlines and other features left standing on a woodblock are less elegant than lines cut into a copper plate, a wood engraver could add place names or descriptive text by merely cutting a rectangle into the wood and inserting pieces of metal type. That only a single sheet survives suggests that van den Putte's version was less commercially successful than Mercator's copperplate edition. Even so, this mechanical facsimile, which acknowledged Mercator's authorship, contributed to the projection's growing prominence and could have inspired one or more of three substantially smaller Mercator maps published around 1595. Nordenskiöld included one of them in his Facsimile Atlas: a 9 by 12 inch (22 by 32 cm) copperplate engraving from Matthew Quad's Geographic Handbook, published in Cologne between 1594 and 1608. Although Quad's map (fig. 4.6) lacks a graticule, its origin is plainly apparent in both its title ("... ad imitationem universalis Gerhardi Mercatoris") and the telltale shapes of its continents.

If multiple maps by diverse authors are a reliable indicator, Mercator's projection became the cartographic expression of a hot idea in the late 1590s, when Jodocus Hondius and Edward Wright offered their own versions of Mercator's world. Hondius's map predates Wright's, but as the next chapter points out, the Dutch cartographer relied heavily on the English mathematician, who developed a mathematical description as well as tables showing how to position the parallels. Although Mercator demonstrated the projection's look and use, Wright made the secret of its construction readily available to other mapmakers.
Mercator- Bearin

To be followed next week



  LMB-BML 2007 Webmaster & designer: Cmdt. André Jehaes - email andre.jehaes@lmb-bml.be