MELISSA BLOCK, HOST:
A team of researchers is poring over a map at Yale's Beinecke Library. It's more than 500 years old. It's a map of the world as the world was understood around the year 1491. And it's not just any map. There's ample evidence that Christopher Columbus may have consulted it as he planned his first voyage across the Atlantic.
The map was made by the cartographer Henricus Martellus, and it includes not just oceans in green and landmasses in tan, but words - lots of them. Many of them are hidden or obscured. Even in tests of the map in the 1960s using ultraviolet light, vast portions remained illegible. But now, as we read this week in Wired Magazine, the researchers are using multispectral imaging to reveal more of what the map says. Chet Van Duzer is leading that team.
CHET VAN DUZER: It's extremely difficult to read the text on the map. And in fact, if you look at the map under natural light, you can see some hints that there is a lot of text on the map, but you can't read it. If you look at the map with, for example, ultraviolet light, you start to perceive that yes, there is a lot of text on this map.
And the map was acquired by Yale University in the early 1960s, and there was an initial burst of enthusiasm about the map. It was called a missing link in our understanding of the development of cartography in the late 15th century. But then when people started to look at the map, I think they were stymied by the fact that you just can't read most of the text on it.
BLOCK: Well, you have been trying to get images off of the map - the text that's hidden in this map that you can't see with the naked eye. You've been using multispectral imaging. How does that work? What are you trying to find?
DUZER: Well, what one does is take multiple images of the same - in this case, part of a map - at different frequencies of light. And what often happens is that each frequency of light will reveal additional information about the object. And multi-spectral imaging is a way to combine the additional information revealed at each frequency of light into one image. And it's particularly useful for revealing text that's beneath another layer.
BLOCK: Do you have a sense of what the text will indicate? In other words, is it text that says the names of countries, names of mountain ranges - rivers? What would it be?
DUZER: It's all of those things. It's the names of places, but it's also descriptive text. For instance, in eastern Asia, a lot of that information will come from Marco Polo.
BLOCK: What are some of the things that show up in those UV images from the 1960s?
DUZER: Well, in northeastern Asia there's a text about a people that live there, and they live without corn and without wine. And they mostly eat deer, and they also ride the deer like horses.
BLOCK: That's actually written on the map?
DUZER: That is written on the map. There's also text about the siren. There's a text about a sea monster that has one eye in the middle of its face.
BLOCK: Tell me about the siren.
DUZER: The text about the siren is very short. It just says, at this point in the Indian Ocean, there is the horrible monster called the siren.
BLOCK: Really?
DUZER: Yeah.
BLOCK: This is not stuff we think about as being included in maps to say the least.
DUZER: Well, in maps of this period it's not so unusual. On this map there are no images sea monsters, but there are these descriptive texts.
BLOCK: Is it a frustrating thing to look at the Martellus map, know that there's text in there that you just can't read - know that it's there, but you don't have access to it yet?
DUZER: It is tremendously tantalizing, and particularly when one knows what an important map it is.
BLOCK: Well, Chet Van Duzer, thanks for talking to us about the Martellus map of 1491.
DUZER: Thank you so much. It's been a pleasure.
BLOCK: Chet Van Duzer is leading the team that scanned the map using multispectral imaging. They're now processing and studying those images in search of lost words. Transcript provided by NPR, Copyright NPR.