THE MARITIME ATLASOak Island Data method

DATA METHOD

Oak Island

How the map, the clock and the record on this page were made, what each can and cannot show, and where every fact comes from.

Snapshot: 15 October 2026. Nothing on the page comes from a database. Every moment, point, dig and source sits in a small data file beside it (data/), checked by a script (tools/build-data.mjs) that fails if a moment has no source, a fact has no source the moment cites, or a point has no basis. The facts number 62: 55 checked for the film (f001 to f055) and 7 of the page's own (p01 to p07), each with the source it rests on (data/facts.v1.json). Each moment carries the numbers of the facts it uses.

Three kinds of record

The page's one rule is to keep apart what was written down at the time and what was told later. Each of the 29 moments in data/events.v1.json is one of three kinds.

  • A record of its time, read (a solid amber dot; 12 moments): a letter, newspaper, map, report or data set made within a few years of what it reports, read for this page. The 'Patrick' letter in the Nova Scotian (1861); three printings of the 1866 account from Chester; A. F. Church's county map (1883); the Oak Island Treasure Company's prospectus (1893); Ainslee's Magazine (1899); Collier's (1906); the New York Times (1909); Golder Associates' borehole report (1971); the Woods Hole Oceanographic Institution's report (1996); CBC News (2014); the University of New Brunswick (2018); and Natural Resources Canada's laser survey (2016). The prospectus is a seller's document: it is a record of the company of 1893, and its history of the island is a story told to investors.
  • A witness, writing later (a half dot; 2 moments): someone who was there, writing years afterwards. J. B. McCully of the Truro Company wrote in 1862 of the dig of 1849. John Brown's account of 1867 and H. L. Bowdoin's of his own search of 1909, written in 1911, are of this kind too.
  • A later history (a hollow grey-blue dot; 15 moments): county histories, chronologies, fan reference sites, Wikipedia and the television series. They gather the earlier accounts and often copy one another.

"One source" marks a moment that rests on one account (5 moments). The later histories draw on the same few accounts of the 1860s and on the prospectus of 1893, so their agreement is not counted as more than one source.

Quotations are no longer than a phrase. Facts from news articles, Wikipedia (CC BY-SA 4.0), the later histories and the two engineering reports are used as facts, in the page's own words.

The clock

The clock runs 60 seconds, from 1795 to the end of 2026, and it is not even. The years before 1849 take a tenth of it; 1849 to 1900, when the story was first written down, about a third; 1900 to 1955 about an eighth; 1955 to 1972, the years of the deep drilling, about a seventh; 1972 to 2013 under a tenth; and the years of the television series, from 2013, about a fifth. The tick marks under the slider show the real years.

A moment known only by its year sits at 1 July of that year; one known by its month sits at the first of the month. Every dig and hole in our table appears in the middle of its first year.

The map

The projection is the atlas's: a spherical Lambert conformal conic with standard parallels at 44° and 46° N, central meridian 63° W and origin 45° N, so that shapes match the other pages. North is up to within about one degree at Oak Island.

The shore is today's, in three levels of detail that replace one another as the map moves in. The province is the atlas's own copy of Statistics Canada's 2021 Cartographic Boundary Files (Open Government Licence – Canada), simplified to about 0.003°, with New Brunswick, Prince Edward Island and Maine around it. Mahone Bay is asked from the same files at 0.0001° and simplified to 0.00005°, about 5 m. The island is the Nova Scotia Topographic DataBase at 1:10,000 (Open Government Licence – Nova Scotia), with its roads. It is generalized for a screen. It is not the shore of 1795 and it is not a chart.

The ground is Natural Resources Canada's 2016 laser survey of Lunenburg County (HRDEM, the 1 m bare-earth model; Open Government Licence – Canada): the shape of the ground with the trees and buildings taken away. It is drawn as shaded relief in the atlas's land colour, with contour lines every 2 m and a stronger line every 10 m, smoothed and simplified to about 1 m. The island's highest ground in the model is 16.4 m. The model shows the island of 2016, with two centuries of digging in it: it is not the ground of 1795. The image is in degrees and is fitted onto the projection by its three corners; over the island the difference this makes is well under a metre. The shore of the topographic database and the edge of the laser survey can differ by a few metres.

The four views (Nova Scotia, Mahone Bay, the island and its east end) are windows chosen for the story. They are not positions of anything. Places outside the province (New York, Boston, New Jersey, Texas, Michigan and Victoria, British Columbia) are drawn as arrows at the edge of the province view, pointing the way, with their Wikidata coordinates (CC0) behind them.

Recorded and placed

Every point on the map is in data/places.v1.json, with its status, its error in metres and the sources behind it.

  • Recorded: the Money Pit. The page uses the pit area as the modern work used it: the mean of two positions that agree to 5.9 m. One is the Woods Hole survey by satellite of 28 July 1995, at the reference well of the pit area. The other is Golder's borehole plan of 1970, fitted to today's shore (below). The map draws a ring of 8 m (26 ft) round it. This is not the shaft of 1795: nobody surveyed it before the later shafts were dug beside and through it, and Golder wrote in 1971 that its exact location had been lost.
  • Recorded: Borehole 10-X, from the same Woods Hole survey; the swamp, the middle of the topographic database's swamp area; the causeway, the middle of its bridge line; and the towns, each the province's official name point (Nova Scotia Geographic Names, GeoNAMES). A name point marks a community, not a building.
  • Placed from Golder's plan (a dashed ring). Golder's Figure 1 of 1971 is a borehole plan drawn at 1 inch to 30 ft with a north arrow. Its scale comes from the sheet, its turn from the north arrow, and its position from three fits to today's shore (the topographic database's outline and the laser survey's 0 m and 0.5 m lines); the plan's shore then lies within about 5 m of today's, typically, and within 13 m in nine places out of ten. From it: Golder's Borehole 204, where the void of 1970 was found (to about 10 m), and Smith's Cove (about 20 m). Golder's Figure I-1, drawn at 1 inch to 20 ft after R. V. Harris's book of 1958, is tied to Figure 1 through eight boreholes drawn on both. From it: the Cave-in Pit (about 20 m) and the Chappell, Hedden and Restall shafts.
  • Placed from a tourism map (a dashed ring): the stone triangle and Nolan's Cross, from the markers of the Oak Island Interactive Map at oakislandmap.com, to about 40 m (130 ft). They are one editor's placements on a satellite photograph, not a survey. No survey or plan of either was found. The triangle's marker lies 8 m from the south shore, where the report of 1897 puts the triangle.

On the map the Chappell and Hedden shafts are drawn with the Money Pit, 5 m and 3 m from it, and the Restall shaft with Smith's Cove, 8 m from it. At any scale the map shows, they would sit on top of each other.

Smith's Cove is the east-facing beach of the island's east end, as the diggers, Golder's plan of 1970 and the Woods Hole report name it. The province's gazetteer and topographic database give the name "Smiths Cove" to a different bay, on the south shore. The page follows the diggers.

The digs and holes

Our table (data/digs.v1.json) counts 153 digs, shafts, drill holes, cofferdams and tests from 1795 to the seasons of 2025 and 2026. It is built from Golder's report of 1971, the Woods Hole report of 1996, the histories in the timeline and the series' episodes as their guides describe them. It counts what the sources record, not what is in the ground: treat 153 as a floor.

Measured: from a survey or a gazetteer point20
From Golder's 1970 borehole plan, fitted to today's shore32
Approximate: at the site the source names90
No place recorded; not on the map11
In all153

A solid dot is measured or fitted; a hollow dot is approximate. The series gives no coordinates, so its digs are drawn at the site each episode names, which is why most of them are hollow. Rows that share a site are spread within their stated error, at most 30 m, in a fixed pattern, so that each can be seen; the spread is drawn, not recorded. Depths are as the source reports them, in feet.

The dig groups in "Who dug" come from the research table of companies, with each group's deepest reported shaft or hole. The usual count of deaths in the search is six (Wikipedia; Fortune). The page gives them as totals, with their years.

Distances

These are straight lines between the points above, worked out on a sphere. No source prints them except where the table says so.

The Money Pit to Smith's Cove137 m (450 ft). Golder's Figure I-1 has the drains about 450 ft away.
The Money Pit to Borehole 10-X45 m (146 ft). Fortune says 160 ft; Golder 180 to 200 ft.
The Money Pit to Golder's Borehole 204, the void of 197082 m (269 ft). Golder gives about 270 ft.
The Money Pit to the Cave-in Pit108 m (355 ft)
The Money Pit to the stone triangle76 m (250 ft)
The Money Pit to Nolan's Cross337 m (1,107 ft)
The Money Pit to the south shore83 m (271 ft). Golder gives "some 300 ft".
Oak Island to Chester5.1 km (3.2 miles)
Oak Island to Halifax57.7 km (35.8 miles)
Oak Island to Truro124 km (77 miles)
Oak Island to Sydney368 km (229 miles)
The island, on today's outline without the causewayAbout 55 ha (136 acres); 1,342 m (0.83 miles) on its longest straight line. Other sources give other figures.

Pictures

  • The county map of 1883 is A. F. Church's Topographical Township Map of Lunenburg County, from the David Rumsey Map Collection's scan as it is published on Wikimedia Commons, in the public domain. It is cropped to Oak Island and the Western Shore and resized.
  • The film's poster is the cover of Joel Hirtle's film. Its map is drawn from the 2016 laser survey.
  • Not shown: pictures from The Curse of Oak Island (the producers'); photographs of the lead cross; the figures of the Golder and Woods Hole reports, whose licence is not stated, which are used for positions and facts only; the maps in Collier's of 1905 and 1906; and any picture from the Nova Scotia Archives.

Each picture has a file beside it (photos/*.json) with its source, licence, credit and what was changed.

Where the record disagrees

  • The year 1795. No paper of the time records the first dig. The accounts of 1862 and 1863 give no year; 1795 first appears in print in the prospectus of 1893.
  • Captain Kidd. The link is the settlers' story of a dying sailor who said he had sailed with Kidd. No record ties Kidd to the island.
  • What lay every 10 ft. Oak timber (McCully, 1862); "marks" (the prospectus, 1893); planks (Collier's, 1905); platforms (Bowdoin, 1911). DesBrisay (1895) has a tier of logs at about 12 ft and a second some 15 ft lower.
  • The carved stone. At 80 ft in the account of 1862; at about 90 ft from 1893 on. Its reading, "two million pounds" ten feet below, is one unnamed expert's in the prospectus of 1893, which says the company does not claim it is right.
  • The water. Soft earth at 90 ft and more water at 93 ft (McCully and the printings of 1866); no water until 95 ft (the prospectus, 1893). By the next morning, 60 ft (18 m) of water (DesBrisay) or water "as high as the tide level" (McCully).
  • The Truro Company's years. 1849 (McCully, who worked the auger); 1848 (DesBrisay); 1849 to 1851 (Wikipedia, the Compendium); 1849 to 1853 (Fortune).
  • The boring of 1849. Through a platform at 98 ft, oak, "metal in small pieces" known by its sound, more oak and clay, by McCully, the one witness. The three small links he says were gold rest on him alone.
  • The flood tunnel. From 1861 the diggers said the sea came in by a tunnel. Its length is 365 ft in 1893, about 400 ft in 1905 and over 600 ft in 1911. It has never been surveyed.
  • The drains. Five drains found about 1850 (Wikipedia says 1851). The letter of 1862 speaks of drains at the shore; the count of five and the name Smith's Cove first appear together in 1893.
  • The fibre. "Like the husk of a cocoanut" in the accounts; dated to about 1,100 years by Woods Hole in 1995, which says how it came to be there is unclear.
  • The parchment. From 156 ft in the report of 1899; 153 ft in others. A boring reported in 1906 struck cement at about 153 ft.
  • The 1866 papers. Three printings credit the New York Herald; the Victoria printing credits the New York Times.
  • The first printed account. Wikipedia names the Liverpool Transcript of October 1862. The 'Patrick' letter in the Nova Scotian is a year older, and a notice of 1857 is known but was not read.
  • The death of 1861. A pump boiler burst and one worker died, by the later histories. No record of 1861 was read for it.
  • Gilbert Hedden's years. From mid-1935 (Wikipedia); 1936 to 1937 (Fortune); 1936 to 1938 (the Compendium); 1934 to 1938 (oakislandmystery.com).
  • Borehole 10-X. 160 ft from the pit (Fortune), 180 to 200 ft (Golder), 141 ft between Woods Hole's two points of 1995. Drilled in 1969 (Golder, Fortune); Wikipedia and the Compendium date its 235 ft shaft to 1971. The camera of 1971 recorded possible chests, remains and tools, by the owners' account; the images were unclear and nothing has been confirmed independently. In 2016 divers found no artifacts at the bottom (Wikipedia); Fortune says only one diver made it down.
  • Nolan's Cross. Six boulders by the tourism map's marker; five by other sources. No source read gives the year it was found.
  • Where the first shaft was. The surveys fix the pit area the modern work used, to about 8 m. Golder wrote in 1971 that the pit's exact location had been lost.
  • The island's size. 55 ha (136 acres) measured on today's outline; 57 ha (CBC; Wikipedia); about 130 acres (53 ha; Golder, 1971). The shore has moved since 1795, and the swamp was once a cove.
  • The series. The first episode aired on 5 January 2014 (CBC) or 6 January (Wikipedia). The date of season 14, 10 November 2026, rests on one source.

What was not read

  • The notice of 1857 in the Liverpool Transcript, said to be the first mention in print.
  • The New York Herald of 2 September 1866 itself. Three printings of its account were read: Louisville, Columbia and Victoria.
  • The Nova Scotian of 1861 and the Liverpool Transcript of 1862 on paper. Both letters were read in a transcription (Critical Enquiry).
  • R. V. Harris's book of 1958 (in copyright). It is used only through Golder's Figure I-1, which was drawn after it.
  • The episodes of the series. Its digs and holes come from the episode guides, not from watching the episodes.
  • Golder's and Woods Hole's final reports, if any. The page read the draft of each, as they are published on oakislandmystery.com: Golder's of April 1971 and Woods Hole's of 8 April 1996.

The page follows the treasure hunt. It does not tell the island's other history: its Mi'kmaw history, and the families who farmed it.

Every source

The list is written by the build script from data/sources.v1.json, so it cannot drift from the page.

RECORDS OF THEIR TIME, READ FOR THIS PAGE

WITNESSES, WRITING LATER

LATER HISTORIES AND RETELLINGS

THE MAP