Gold in Dartmoor A research project

The four systems

To understand where gold sits here, it helps to see the region as four overlapping geological stories rather than one.

1. The Devonian basement (~419–359 million years ago). Slates, sandstones and volcanic rocks laid down in a shallow sea and around volcanic centres. South of Dartmoor these form the Dartmouth Group and, further south still, the metamorphic Start Complex. These rocks are locally pyritic and locally mineralised, and they are chemically reactive, which turns out to be the whole point.

2. The Variscan granite (~280 million years ago). The Dartmoor batholith, one lobe of a much larger granite mass that surfaces again at Bodmin, St Austell, Carnmenellis, Land’s End and the Scillies. Its intrusion drove the hydrothermal system that produced Cornwall and Devon’s tin, copper, tungsten and arsenic. It is spectacularly productive, and essentially barren of gold.

3. The Permian red beds (~299–252 million years ago). After the mountains built by the Variscan orogeny began to erode, Devon became a hot desert basin. Red breccias, sandstones and mudstones accumulated in half-grabens such as the Crediton Trough, with alkaline basalt and lamprophyre lavas erupting through them.

4. Everything since. Erosion stripped most of the Permian cover off south Devon, periglacial processes smeared solifluction “head” deposits across the valleys, and the present rivers cut down through the lot, sorting and re-sorting whatever heavy minerals they encountered.

Why the granite fails

The granite makes tin and tungsten because those metals partition into the late, volatile-rich fluids that a cooling granite exsolves. Gold does not behave that way in this system. When Brammall assayed Dartmoor granite in the 1920s the best result was 8.6 grains per ton at Merrivale Quarry, roughly 0.55 g/t, with most localities at trace or nil Legendary Dartmoor 2016.

For scale, the Cononish deposit in Scotland which supports a real mine runs at 15.9 g/t British Geological Survey. Dartmoor granite is thirty times poorer than that, at its very best, before you account for the fact that you would have to crush and process it.

The model that actually works: Permian redox gold

The interesting discovery came from the other direction entirely.

When BGS panned streams across the South Hams in the late 1980s, the gold grains they recovered were odd. Many were dendritic: branching and fragile, incapable of surviving transport Leake 1992. Many carried palladium, in intricate internal growth zones, with some rims reaching 13% Pd Leake 1990. Some grains were potarite (PdHg), a genuinely rare platinum-group mineral. Inclusions inside the grains were selenides, and there were no sulphides at all.

That assemblage is diagnostic. It does not come from a granite. It comes from low-temperature oxidising brine.

The model, worked out through the 1990s and formalised in the 2000s, runs like this Leake 1991–2004:

  1. Permian red beds accumulate over the eroded Devonian surface. They are oxidised, iron-rich, and interleaved with alkaline volcanic rocks.
  2. Hot, saline, strongly oxidising groundwater circulates through them at around 86 ± 13 °C, leaching gold and palladium out of the immature sediments and volcanics.
  3. That brine descends through fractures deep into the underlying Devonian rocks.
  4. Where it meets reduced, pyritic slate or limestone, the redox potential collapses and the dissolved precious metals drop out into low-temperature carbonate veins.
  5. Erosion later strips the Permian cover, exposes the soft altered zone, and the modern streams wash the grains out.
The Permo-Triassic redox gold model Cross-section showing Permian red beds overlying an unconformity cut into folded Devonian slates. Oxidising saline brine descends from the red beds through a fault, and deposits gold and palladium in carbonate veins where it meets reduced pyritic slate. A later panel shows erosion stripping the Permian cover and a modern stream washing gold grains out of the altered zone. A · Permian, about 280 million years ago unconformity fault oxidising saline brine, ~86 °C reduced pyritic slate redox front Au + Pd drop out Permian red beds Devonian slate B · Today outlier altered zone soft, oxidised, carbonate-veined modern stream grains in the gravel this is what you pan Why this matters for a prospector 1 · The gold source is the Permian cover, not the granite. Look for red beds and the unconformity beneath them. 2 · Deposition needs reactive, reduced rock (pyritic slate or limestone) under that cover. 3 · The altered zone is soft, so it erodes preferentially and feeds nearby streams. 4 · Grains are fragile and dendritic, so they have not travelled far. Sample small tributaries.
Schematic only: vertically exaggerated, not to scale, and drawn for this project from the published description of the model rather than copied from any figure. Temperature from fluid-inclusion work on the south Devon carbonate veins.

Where that puts you on the map

South of the moor: the South Hams

The best-documented ground anywhere near Dartmoor. Panned concentrates from 44 of 450 drainage sites exceeded 0.5 ppm Au Leake 1992. Anomalies occur across the whole Lower Devonian sequence and the Start Complex, and are less frequent over the Middle Devonian rocks in the north-west.

The standout is a site about 2 km south-west of Holbeton, where the dendritic gold and associated Pt- and Pd-rich grains, around 5% of the total, are abundant enough that BGS drilled four boreholes at Brownstone Farm to chase the source. They found an east–west zone of intense oxidation, brecciation and carbonate veining reaching 380 ppb Au Leake 1992.

BGS also identified four localities with an earlier, epithermal-style polymetallic phase: Wadham Rocks, Churchill Farm near Marlborough, east of Chivelstone, and the ground around and west of the old Loddiswell mine, where native gold occurs in quartz–carbonate veins alongside antimony Stanley 1990.

North of the moor: the Crediton Trough

BGS took the south Devon model, applied it to the Permian basin running from Hatherleigh east to the Exe, and it worked: gold was confirmed in drainage for the first time at numerous localities on the Permian outcrop, with grain chemistry closely matching south Devon Leake 1994.

Rock samples returned up to 1.8 ppm Au in alkaline basalts and 42 ppb in Permian sedimentary breccias. At Smallbrook, near Newton St Cyres, the Late Permian Newton St Cyres Breccia appears to be a fossil placer, an ancient gold-bearing gravel, now lithified, being eroded by a modern stream Leake 1994. That is about as good a set-up as a panner could ask for. The Bow Breccia plays the same role at Deckport and Solland near Hatherleigh.

Interestingly, gold anomalies in the Crediton Trough all lie on or near lineaments picked out from Landsat imagery, usually a north-east trending set Leake 1994.

The moor itself

Trace gold in alluvial tin ground, documented from the medieval period onwards but never systematically recovered. Real, historically attested, and quantitatively negligible.

A note on scale and honesty

None of the above is a deposit. BGS’s own conclusion is that south Devon mineralisation is “likely to be locally rich, but possibly of limited vertical extent” Leake 1992, and nobody has developed a gold mine on the back of it in the thirty years since. What the work establishes is that gold is genuinely and widely present in the drainage around Dartmoor, at levels a patient person with a pan can find, in places that are almost entirely privately owned.

That is the honest state of the science, and everything on the targets page follows from it.