In Search of Lost Islands

Three books look at islands that never were.

We expect maps to tell the truth; indeed we need them to on a fierce and primal level. “I believe cartography enjoys an enviable position of credibility and confidence among the people who see it. If you see it mapped, you believe,” wrote Charles Blow last fall; he was writing in response to Trump’s petty defacement of a hurricane forecast map with a marker. The reaction to Trump’s stunt, was, I thought, revealing. It’s part and parcel with what Matthew Edney refers to as the ideal of cartography: striving toward a universal, unbiased and perfect map.

When a map has a mistake on it, when it’s wrong, it does something funny to our heads. We obey our phones and dashboard GPS navigators even when they send us off a cliff. We concoct nutty theories about ancient civilizations because a 16th-century portolan chart had a funny bend on a coastline. We wonder, because someone wrote “here be dragons” on a map, whether dragons were actually real. We make brain pretzels trying to force maps to be truthful even when they are manifestly wrong.1

Maps have to tell the truth. They simply have to. Maybe that’s why stories about mistakes on the map, and the havoc those mistakes cause, fascinate us so much. Which brings me to three books, all published for the first time in 2016, that talk about map errors of an older kind: islands and other features that appeared on maps, sometimes for centuries, that in the end turned out not to exist.

Long before we got this funny idea that maps had to be truthful, before Edney’s ideal of cartography took hold, maps were full of conjectures, rumours, mistakes in surveying and even some outright frauds.

Reproduction of the 1558 Zeno Map from Henrich Peter von Eggers, Priisskrift om Grønlands Østerbygds sande Beliggenhed, 1793. Wikimedia Commons.

Take, for example, Frisland. A hoax perpetuated by the 14th-century Zeno brothers of Venice, or possibly their 16th century descendent: the latter published a book of the Zeno brothers’ correspondence in which they described their travels to Frisland, a large island south of Iceland in the North Atlantic with a Latin-speaking ruler. (Many phantom islands of the era seem to be full of previously undiscovered Christian realms where Latin is spoken: they’re a westerly variant of the Prester John legend.) The story was swallowed whole, and Frisland appeared on many maps; England claimed it. It took centuries for the Frisland myth to disappear completely. (Previously: The Invention of Frisland.)

Detail showing Bermeja from Henry S. Tanner, “A Map of the United States of Mexico,” 1846. David Rumsey Map Collection.

Or for a more recent example, the island of Bermeja in the Gulf of Mexico. First sighted in the 16th century (but not since), it remained on maps of the region into the 20th century. In 2009 a Mexican aerial survey determined the non-existence of the island, which led to some conspiracy theories that it had been destroyed by the Americans: its position might have been important in determining who owned the subsurface oil exploitation rights in the Gulf of Mexico, and an agreement with the U.S. had just been completed on that very issue.

“[A]s the story of Bermeja demonstrates, a fascinating characteristic of many of these misbeliefs is their remarkable durability,” writes Edward Brooke-Hitching in The Phantom Atlas (Simon & Schuster UK, Nov 2016; Chronicle, Apr 2018). Indeed, as all three of the books under review today demonstrate, phantom islands continue to be “un-discovered” into the present day.

But where do phantom islands come from? “Among the multitude of non-existent islands that have appeared on maps over the past few centuries,” writes Malachy Tallack in The Un-Discovered Islands (Birlinn, Oct 2016; Picador, Nov 2017), “the vast majority are the result of mistakes. They are accidental phantoms, caused by imperfect navigation, optical illusions or poor recording by mariners and cartographers. Sometimes, though, there is no accident at all. Islands are invented deliberately, often creating inordinate confusion as a result.” To that list Brooke-Hitching adds mythology and religious dogma, which surely must have been at play with not just Frisland, but Hy-Brasil and Saint Brendan’s Island too; as well as volcanic destruction, because that can be a thing; and, because The Phantom Atlas isn’t just talking about islands, copyright traps.

Lexikon der PhantominselnIn the end, the solution to a phantom island is more exploration: repeated voyages and surveys. Of course, establishing that something doesn’t exist—proving a negative, in other words—is much more difficult than suggesting that it existed in the first place. “Often, the process of refuting the existence of an island is more exciting, but also more complicated and dangerous, than its discovery,”2 writes Dirk Liesemer in Phantom Islands, first published in Germany in 2016 as Lexikon der Phantominseln (Mare), now translated into English by Peter Lewis and published, last October, by Haus.

So, three books, with the same premise, covering the same territory, often using the same examples, and in much the same way, published at more or less the same time. Must have been something in the water.3 These books are more similar than not. It’s tempting to treat them as a whole. So I will.

Each is a collection of short chapters explaining how an island was added to the map, and how it was found out not to exist. I’m glossing over a lot in that sentence: there are some truly fascinating stories in these books. The Un-Discovered Islands covers twenty-four of them (with another ten briefly mentioned), arranged by theme; Phantom Islands covers thirty, in alphabetical order. The Phantom Atlas has sixty chapters, also arranged alphabetically, and goes beyond islands to other geographical features, and indeed to more intangible subjects, with chapters on the various monsters found on maps, and the ideas of a flat earth and an earthly paradise.

Naturally there is some overlap: a total of 11 mythical islands are covered by all three books, for example. (These are, for the record, Antilla, Atlantis, the Aurora Islands, Bermeja, Buss Island, Crocker Land, Frisland, Hy-Brasil, Saint Brendan’s Island, Sandy Island and Thule.)

The Phantom AtlasBoth Phantom Islands and The Un-Discovered Islands are relatively short, at 160 and 144 pages respectively. They’re elegantly designed but more illustrated than mapped, if you follow me. The Phantom Atlas is nearly twice as long and has two to three times the chapters (it also costs twice as much); it fills that space with reproductions of maps and art and other illustrations. (I read the ebook version, which in hindsight was a mistake: you get the images, but not the page layout.) The Phantom Atlas has its eye set on the coffee table: it’s the kind of map book you look at as much as you read it. The other two not so much, but they make up for it with stronger prose: each of these little books make for an afternoon’s pleasant reading, with The Un-Discovered Islands being a little more slight, and Phantom Islands having a somewhat different focus owing to its originally being written for a German audience.

In the end it depends on what you’re looking for. Information on phantom islands is readily available online, but these books spin their tales better, and The Phantom Atlas has better pictures. And these books’ overlap (see above) is not so much that you’d be wasting your time or money by reading all three. Particularly if you find this subject fascinating.

I received a review copy of Phantom Islands from the publisher. I bought the other two as ebooks.

The Phantom AtlasThe Phantom Atlas: The Greatest Myths, Lies and Blunders on Maps
by Edward Brooke-Hitching
Simon & Schuster UK, Nov 2016 (U.K. edition)
Chronicle, April 2018 (U.S. edition)
Amazon | Apple Books

Phantom IslandsPhantom Islands
by Dirk Liesemer
translated by Peter Lewis
Haus, Oct 2019
Amazon | Apple Books

The Un-Discovered IslandsThe Un-Discovered Islands
by Malachy Tallack
Birlinn, Oct 2016 (U.K. edition)
Picador, Nov 2017 (U.S. edition)
Amazon | Apple Books


  1. How to Lie with Maps, in which Mark Monmonier asks us to approach maps with “a healthy skepticism,” is in its third edition, but its message doesn’t seem to be getting through.
  2. Dirk Liesemer, Phantom Islands (Haus, 2019), p. 10.
  3. Not to say that these were the first, or the only, books about phantom islands. A cursory search led me to a number of other titles, such as Henry Stommel’s Lost Islands: The Story of Islands that Have Vanished from Nautical Charts (UBC Press, 1984), which, it just so happens, was reprinted in 2017. Or Donald S. Johnson’s Phantom Islands of the Atlantic (Walker, 1996). Or, going even further back, Raymond H. Ramsay’s No Longer on the Map (Viking, 1972).

Author: Jonathan Crowe

Jonathan Crowe blogs about maps at The Map Room. His nonfiction has been published by AE, The New York Review of Science Fiction, the Ottawa Citizen and His sf fanzine, Ecdysis, was a two-time Aurora Award finalist.