How a lost island panicked astronomers

In 1932, the United States found that one of its claimed islands in the Pacific was missing. Meanwhile, huge hopes were laid on finding it again.

The upcoming total solar eclipse on 8 June 1937 was going to be a historic event, providing scientists with a significant 7-minute-long window to observe the phenomenon. In a major twist of fate, it was calculated that the eclipse would not be viewable from anywhere where telescopes could be installed.

A map of the world showing the path of the solar eclipse of 1937 over the Pacific Ocean — The path of the 1937 total solar eclipse (8 June). Source: timeanddate.com

The only exception where scientists could possibly erect an observatory, it seemed, was the island of Sarah Ann (or Sarah Anne) – a tiny speck of land that had appeared on old US naval charts but had not been reported on for at least 15 years.

Scientists raised the alarm in 1932 – could Sarah Ann be found in time for the 1937 eclipse? Multiple newspaper reports at the time spread the news of the island’s apparent disappearance and the scientists’ concerns.

Ludington Daily News, 16 Oct 1932. Vanishing of Sarah Ann, Tiny Pacific Island, Causes Scientists Much Worry Washington, Oct. 15 (AP). Sarah Ann, much to the consternation of the astronomical world, has disappeared and, it is feared, drowned. Just a pin-point freckle on the broad face of the Pacific, Sarah Ann was - and maybe is - the tiniest sort of an uninhabited island. This Cinderella of the sea was last reported alive some 15 years ago. The of no value to anyone and avoided by ship captains, she has become today a spot of great importance. If she can be located, many scientists want a date with her five years from now. The reason for the sudden popularity arises from the efforts of Professor A. Kopff, director of the Rechen-Institute of Berlin, and Professor James Robertson, director of the naval almanac office here, to calculated the course of the 1937 eclipse. They have found that on June 28, 1937, the sun will be eclipsed totally by the moon for about seven minutes. In all the Pacific, Sarah Ann was the only reported spot of land suitable for observation in the 5000-mile long, 120-mile wide band of totality. Recalling that the scientific world spent thousands of dollars to visit New England and Canada for the August eclipse that lasted a minute and 40 seconds, the value of a seven-minute observation can be easily understood. Kopff, who has gone back to Germany after conferring with Robertson, plans to ask German merchant ships to hunt for her, and other observatories may join in the search. — Ludington Daily News, 16 Oct 1932.

Tiny Island Lost to Science. Disappearance of "Sarah Ann" in Mid-Pacific complicates problem of eclipse observation in 1937. — Ludington Daily News, 16 Oct 1932.

Portrait photograph of Cmdr. Joseph S. Skerrett c. 1867. — Cmdr. Joseph S. Skerrett c. 1867.

The US Navy’s Pacific Fleet was asked to make a search, repeating one they had already made in the 1870s by the USS Portsmouth under the command of J.S. Skerrett.

If the island had merely sank a few metres below the waves, there would still be hope a platform for viewing the eclipse could be constructed. However, no new evidence for Sarah Ann Island was found, and the eclipse observations were instead successfully made from Canton Island, roughly halfway between Hawaii and Fiji.

Source: US Naval Observatory

The total Solar Eclipse of June 8, 1937

But why had Sarah Ann created so much confusion?

To start answering this, we must go back in time to the mid-19th century. In 1859, the coordinates for ‘Sarah Anne’ (4°0’N, 154°22’W) appeared among a list of islands containing “large quantities of guano” that formed an affidavit of discovery signed by Captain William Taylor.

A handwritten list of islands and their coordinates, signed by W.W. Taylor — Taylor’s original list of islands included in his affidavit of discovery, dated Feb. 12, 1859. (U.S. State Department)

The same list was later published by the New York Tribune. The islands were claimed under the American Guano Islands Act of 1856, which essentially enabled any US citizen to take possession of unclaimed and uninhabited islands containing guano deposits.

Guano Islands Discovered, in the New York Daily Tribune. March 08, 1859, Page 5. Washington, March 5, 1859. Noticing, a few days since, that Clipperton Island had been proclaimed to the world as belonging to the Emperor of France, and as this guano question has become one of the first moment, it has occurred to me it would not be uninteresting to our readers to know if any, and how many guano islands in the Pacific Ocean or elsewhere have become the property of citizens of the United States, and have been recognized by the Government as pertaining to its territories under the act of Congress approved August 18, 1856. The following is believed to be a correct list of said islands, and their several latitudes and longitudes, viz: Bakers, Jarvis, Howland, Maldens, Arthurs, Christmas, Caroline, Anns, Stavers, Flints, Baumans, Regeweins, Groninque, Frienhaven, Quiros, Low, Clarence, Favorites, Duke of York, Farmers, Birnies, Phenix, Marys, Enderburgs, Sydney, Penhryne, Pescado, Ganges, Riersons, Siderons, Humphreys, Frances, Flint, Nassau, Danger, Mary Letitias, Kemins, Wakers, Sarah Anne, America, Prospect, Samarang, Palmyros, Danger, Makin, Mathews, Davids, Barbers. The two first named Islands have been claimed by the American Guano Company, and the rest by the United States Guano Company, and other citizens of the United States. I understand these acquisitions are all to be surveyed and charted, and the quality and quantity of the guano thereon to be ascertained by competent analytical chemists and topographical engineers, and a report thereof made to Congress at the earliest practicable period. At some of these islands there are good harbors and safe anchorage, and at most of them there is good lee, which, coupled with the fact that most all of them are situated where storms are seldom known (the prevailing winds being from the East), makes them places of safe resort for ships. The quantity and accessibility of the guano on...

Guano Islands Discovered, in the New York Daily Tribune. March 08, 1859, Page 5. ...many of these islands is placed beyond doubt. What remains to be demonstrated is its quality, and whether that is such as to warrant its importation. On this point I am not competent to decide, because there are two theories which now divide the opinions of scientific men, viz: the ammoniacal and the mineral. The former advocated by Laws, Gilbert, Johnson, and others, and the latter by Liebig, Gale, and others, who claim that it is nutrition not stimulus which is the great desideratum. The guano from these islands comes under the last head; the Peruvian, Elide and Ichaboe under the first. Time will settle which theory is correct.

With its rich nitrogen and phosphates, guano was not just a resource, but a potential goldmine. It could be used to produce highly profitable agricultural fertilisers, and it was a source of saltpeter for gunpowder. The Guano Islands Act led to the US gaining control of around 94 islands. By 1903, 66 of these islands were recognised as territories of the US.

A group of miners stand at the base of a massive guano deposit, which has a series of steps dug into it. — Mining guano on the Chincha Islands, Peru. An illustration from Flore des serres et des jardins de l’Europe (1867-8).

Yet, a 1933 review of the Guano Islands Act found no evidence that Sarah Ann Island was ever mined for guano. Furthermore, the report noted that no island could be found at the reported coordinates of 4°0’N, 154°22’W, concluding

…it is evident that Captain Taylor [co-founder of the United States Guano Company]…,drew on his imagination, and upon old, inaccurate charts, for many of the islands.
Rogers, E.S. (January 9, 1933). The Sovereignty of Guano Islands in the Pacific Ocean (Report). Washington, D.C.: Department of State, Office of the Legal Advisor. (p.290)

We can’t know the real motives of Taylor, but one such inaccurate chart he likely referred to is this 1836 map of the Pacific Ocean by British chart maker JW Norie – the first to have plotted Sarah Ann at these coordinates. The map also includes many other phantoms Taylor had included in his claim for guano mining rights. Click the icon below to view a close-up view.

A New Chart of The Pacific Ocean (northern and southern Sheets) as mapped by J.W. Norie in 1836.

A New Chart of The Pacific Ocean (northern and southern Sheets) as mapped by J.W. Norrie in 1836, with a focus on the equator region around Sarah Anne Island.

A new Chart of The Pacific Ocean (northern and southern Sheets) (J.W. Norie, 1836).

But what information had Norie used? The source of the phantom Sarah Ann island appears to stem from a simple clerical error. Just a few years previously ‘Sarah Island’ had been plotted in a mirror position just south of the equator, within the vicinity of Malden Island.

South Sea Islands in Geographical Annual or Family Cabinet Atlas, T. Starling, 1933.

A close zoom view of Malden and Sarah islands on the map South Sea Islands in Geographical Annual or Family Cabinet Atlas, by T. Starling, 1833.

South Sea Islands in Geographical Annual or Family Cabinet Atlas, (T. Starling, 1833).

Photograph of a three-quarter length portrait of George Byron, 7th Baron Byron of Rochdale (1789-1868) seated, facing halfway towards the right. He turns his head towards the camera. He poses with his left arm resting on a side table. — Admiral George Anson Byron, 7th Baron Byron.

The discovery of Malden Island by Westerners is attributed to Lord George Anson Byron in 1825. However, the island had already been sighted several times before, including by Captain Samuel Bunker of the whaleship Alexander just a few months previously:

March 25th 1825 […] at 10 AM described land. It proved to be an Island seen by the Sarah Ann of London and the Independence of Nantucket Capt Whippey […] latitude and longitude of the island in 3-58S 154-30w
Log of Capt. Samuel Bunker

Sarah Ann and Malden islands were thus undoubtedly one and the same, but imprecise methods for measuring coordinates meant duplicate sightings were often recorded on maps. The additional repositioning of Sarah Island into the northern hemisphere, perhaps the result of a simple misreading of the island’s latitude by Norie, set in motion a chain of searches for this phantom that would span the next century and caused panic before the 1937 eclipse.

Further reading:

Dehner, Steve (2018). Supplements To Malden Island’s History: The Nantucket Connection II. Bad Tattoo Inc. – via Google Books.

Polk, L., 2021. The Question of Olohega – The Asia-Pacific Journal: Japan Focus. The Asia-Pacific Journal: Japan Focus 19.

How a lost island panicked astronomers

Be First to Comment

Leave a Reply Cancel reply

Don’t miss new posts!