As we saw last time, Columbus optimistically believed he was sailing west to Marco Polo's Cipangu (Japan), then thought of as part of the "Indies," which he estimated to be 2,300 miles from the Canary Isles. Had he known it was actually 7,800 miles, he might have had second thoughts. The main inspiration for assuming such a relatively short passage was almost certainly a 4-foot-by-6-foot map drawn in Florence, Italy in 1491 by the German cartographer Henricus Martellus.

As it was, of course, Columbus landed in the Bahamas, although he was convinced he'd made landfall near Japan, calling the natives he encountered Indios. Even after he was shown evidence of his mistake, he persisted in his belief, which is why we still have places named Indiana and the West Indies, and why both the U.S and Canada have many "Indian Acts" on their books. Thankfully, referring to a Native American as an "Indian" is going the way of "Negro" and other dated designations.

Despite its importance to Columbus and other navigators of the late 15th and early 16th centuries, the 1491 Martellus map has also been one of the most frustrating for historians. You can see the problem on the left side of the illustration: The text has faded to the point of near invisibility. About the time Yale University acquired the map in 1960, crude (by today's standards) ultraviolet imaging showed there was much more to the map than met the naked eye. In August of 2014, an intense 10-day program using modern multispectral imaging techniques, appropriately named the Lazarus Project, resulted in the detail shown on the right side.

First, the researchers photographed the map in 12 wavelengths of light from ultraviolet to infrared. The second phase of the project was manipulation of the four terabytes of data thereby obtained to reveal previously unseen details. (Four terabytes is roughly half of Wikipedia's primary SQL database.) If you've played with HDR photography, you'll have a rough idea of what's involved; the software identified the best combination of wavelengths for each section of the map. After weeks of massaging data, virtually all the text and geographical features became visible, and the map could finally be studied in detail.

The project turned out to be a bonanza for historians, laying bare hundreds of place names and descriptions and giving us a much clearer understanding than we previously had of Renaissance cartography. According to the project leader, independent historian Chet Van Duzer, the map is "a missing link in our understanding of people's conception of the world." The new analysis shows how much Martellus relied upon Ptolemy's Geography (published nearly 1,300 years earlier!) together with information brought back from China by Marco Polo and the 1488 discovery of the Cape of Good Hope by Portuguese explorer Bartholomew Diaz.

The influence of Martellus' map reached far and wide. Martin Behaim of Nuremburg drew heavily on it in preparing the oldest extant globe in 1492. So did Martin Waldseemüller, the German cartographer who first called the newly-discovered continents "America" on his 1507 map, the "birth certificate of America." Which is why, in 2003, the Library of Congress bought it for $10 million.

Barry Evans ([email protected]) is looking forward to the results of multispectral imaging of the Dead Sea scrolls.