In 2026, conservators at the Royal Observatory Greenwich extracted 38 years of daily latitude fixes from the inside wall of a 1794 Arnold box chronometer—data never logged in any ship’s journal. Salt-laden air, gunpowder fumes and decades of palm-oil lubrication had created a nanometre-thick chemisorption film in which mercury droplets, once used to poise the escapement, acted as mobile sensors. Each droplet left a dendritic trace whose branching density encoded the Earth’s magnetic inclination at noon. By scanning the brass bore with a helium-ion microscope and running a mercury-diffusion inverse solver, researchers replayed an unbroken GPS-like track from Portsmouth to Botany Bay and back, the first “positional palimpsest” ever recovered from a pocket-sized instrument.
Marine chronometers were never intended as data stores; their job was to keep Greenwich time. Yet the mercury-filled Earnshaw spring (used for temperature compensation) sloshed micro-litres of Hg across the brass substrate with every pitch and roll. Mercury forms an amalgam with copper only above 120 °C, well beyond cabin temperature, but chloride ions from sea spray catalyse a room-temperature reaction: 2 Hg + CuCl₂ → Hg₂Cl₂ + Cu. The product, calomel, grows as a 2–5 nm passivation layer whose porosity is inversely proportional to the local magnetic field (field aligns chloride diffusion). Over decades the cumulative porosity map becomes a latent image of geomagnetic latitude, accurate to ±0.3°.
Reading the signal starts by sectioning the cylindrical movement ring under dry nitrogen. A 5 × 20 mm coupon is ion-milled to electron transparency, then transferred to a helium-ion microscope equipped with a time-of-flight secondary-ion mass spectrometer (ToF-SIMS). Each pixel yields a full mass spectrum; the ratio ³⁵Cl⁻/⁶³Cu⁻ is proportional to porosity and hence to magnetic field intensity. Spatial resolution is 0.5 nm laterally and 0.1 nm in depth, sufficient to resolve individual mercury menisci that sat for ~24 h each. A 360° helical scan produces a cylindrical projection; unwrapping it gives a 1-D magnetic latitude trace sampled daily at noon, locked to the chronometer’s winding schedule.
Clock recovery is built into the alloy. Brass expands 19 µm m⁻¹ K⁻¹; seasonal temperature cycles create 12 µm striations that act as annual rings. Counting stripes gives elapsed years; interpolating between them yields calendar days. One anomalous 0.5 mm gap corresponds to April–May 1805, the exact period when HMS Calcutta (the instrument’s last known ship) was laid up in Simon’s Town for coppering—validation that the chronometer was ashore and subject to different humidity, visible as a chloride spike.
Error correction exploits geomagnetic physics. The recovered inclination I must satisfy tan I = 2 tan λ where λ is geographic latitude. Segments violating this to >1 % are flagged as mercury slosh during storms and replaced by cubic-spline interpolation. After filtering, the RMS difference between recovered latitude and Royal Navy logged latitude is 18 km, better than 18th-century dead-reckoning by an order of magnitude.
Storage capacity is modest but unique. A 40 mm bore stores ~200 kB of inclination data, equivalent to 38 years of daily fixes. Across the estimated 4,200 surviving Arnold & Earnshaw chronometers, the global archive could reach 840 MB of high-resolution 18th-century ship tracks—enough to refine geomagnetic field models and rewrite narratives of exploration, mutiny and contraband.
Restoration is non-destructive; the coupon is re-inserted into a newly machined brass ring, returning the movement to museum display with no visual change. Legal title follows UK heritage law: the brass remains Crown property; the track, being non-tangible, is released under Open Government Licence.
For maritime archaeologists the lesson is clear: every tarnished chronometer is a latitude logger, every mercury film a magnetic tape. With the right ion beam and the right diffusion model, the age of sail will navigate again—one nanometre of calomel at a time.
