If you want to find lode gold in Pennsylvania, you go to one place: the southeastern part of the state. York County, Lancaster County, Lebanon County, Berks County, Chester County. There's a reason for that. A very specific geological recipe had to come together in exactly that part of the state, at exactly the right time, to make the rocks there gold-bearing. Miss any one ingredient and you get iron mines with no gold, copper prospects with no gold, or just plain old rocks with no gold.
Here is the recipe, in one sentence: Triassic diabase magma intruded into Cambro-Ordovician limestone, drove contact metamorphism, and the resulting sulfide minerals carried gold.
Three ingredients, all necessary:
The heat source AND the metal source. Tholeiitic basalt — a dark, dense, iron-rich liquid from the deep earth — pushed up into the crust in massive sheets during the Triassic-Jurassic boundary, about 201 million years ago. Each sheet was 100 to 600+ meters thick and extended 30 to 60 kilometers along strike. That's more than 1,000 cubic kilometers of magma per sheet. Big.
The chemical trap. Cambro-Ordovician limestone — laid down 500+ million years ago in a shallow tropical sea — sat in the rock column waiting. When the hot diabase intruded into it, the heat and the chemistry of the contact zone turned the limestone into a kind of geological pressure cooker. The diabase provided the heat and the metals (iron, copper, sulfur, gold). The limestone was the reactive host that made the ore drop out of solution.
The gold carrier. The hot, chloride-rich hydrothermal fluids that circulated through the contact zone dissolved iron, copper, sulfur, and gold from the cooling diabase. As the fluids cooled (below ~500°C), the iron dropped out as magnetite, and the sulfur combined with copper and iron to form sulfide minerals — chalcopyrite, pyrite, pyrrhotite. Gold is locked inside the crystal lattice of those sulfides, invisible, sub-micron, refractory. Not visible gold. Not nuggets. Gold you can't see until you process the ore.
Take away the diabase, you get no heat and no metals. Take away the limestone, you get no chemical trap. Take away the sulfides, you get iron ore with no gold. SE PA is the only part of Pennsylvania where all three ingredients came together in the right combination at the right time. That's why the gold is there and not in the other 90% of the state.
To understand why SE PA is gold country, you have to understand what happened deep underground 201 million years ago. Because the story starts with the breakup of a supercontinent.
Two hundred and one million years ago, all the world's continents were jammed together into one giant landmass called Pangaea. That supercontinent was about to come apart. The Atlantic Ocean was just beginning to open. The crust in what would become eastern North America was being stretched, thinned, and torn. From the Carolinas to Nova Scotia, the continent was breaking.
As the crust stretched, tholeiitic magma from the deep mantle pushed up through the fractures. This is a special kind of basalt — high in iron, low in silica, very fluid when molten. The magma didn't always reach the surface. In many places, it spread sideways between layers of rock, forming sills — flat-lying sheets of igneous rock that intrude parallel to the bedding of the rocks above and below.
These sills were huge. A single diabase sheet might be 200 meters thick and extend for 50 kilometers along strike. They represent more than 1,000 cubic kilometers of magma each. The volume is hard to overstate.
The diabase magma was extremely hot — 1,100°C or more. The limestone it intruded into was sitting there in the rock column, having been laid down in a Cambrian or Ordovician sea half a billion years earlier. When 1,100°C magma pushes into limestone, things happen.
The contact zone — the band of rock right next to the diabase — gets heated to 500 to 700°C and pressurized to about 1,500 bars. That's hornfels facies, or pyroxene hornfels. The limestone doesn't melt (it would decarbonate first at those temperatures), but it does recrystallize into marble. The shale around it turns into hornfels. New minerals form: cordierite, garnet, andalusite, pyroxene, scapolite.
As the diabase slowly cooled, it released hot, chloride-rich hydrothermal fluids. These fluids were supercharged with iron, copper, sulfur, gold, cobalt — all the goodies that were dissolved in the magma. The fluids circulated through the contact zone, through fractures and pore spaces in the cooling rock.
According to the Eugster & Chou (1979) model, these fluids had to carry up to 30,000 parts per million of dissolved iron to form a deposit the size of Cornwall. That's a LOT of iron in solution. The fluid was basically liquid ore.
As the hydrothermal fluids cooled below about 500°C, things started to precipitate. First the iron dropped out as magnetite (Fe₃O₄) — tons and tons of it, replacing the limestone in massive ore bodies. Then the sulfur combined with copper to form chalcopyrite (CuFeS₂) and with iron to form pyrite (FeS₂) and pyrrhotite (Fe₁₋ₓS).
And locked inside those sulfide crystals — invisible, sub-micron, dissolved in the crystal structure like sugar dissolved in water — was the gold. Trapped in the sulfide lattice at the atomic level. You can't see it. You can't pan it. You have to chemically extract it from the chalcopyrite concentrate.
Now here's where the prospector comes in. Over millions of years, the magnetite ore bodies were exposed at the surface. Rain, groundwater, and oxygen worked on the sulfides. The sulfides oxidized and broke down. The gold — which doesn't oxidize — was released as tiny free particles. Those particles washed into creeks, settled into gravel, concentrated behind boulders and in the cracks of the bedrock.
That's the placer gold in the SE PA creeks. It came from the same lode sources — Cornwall, Grace, Dillsburg, Jones, and their cousins — and it didn't have far to travel. That's why SE PA placer gold is coarser than the flour gold you find up in the northern tier. The source is local, the transport distance is short, and the gold didn't get ground to dust by a thousand miles of ice transport.
The Triassic diabase intrusions that made SE PA gold-bearing have been precisely dated. Zircon U-Pb geochronology from Rutgers gives us:
That's the Triassic-Jurassic boundary — the moment the dinosaurs were about to take over the world, the supercontinent Pangaea was cracking apart, and molten rock was pushing up into Pennsylvania limestones. The ages are essentially identical, which means the multiple sheets were emplaced in a geological instant — a few hundred thousand years, maybe less. Not millions of years apart.
CAMP is what made SE PA gold-bearing. The same magma in New Jersey didn't produce Cornwall-type gold deposits, even though it's the same chemistry and the same event. The reason is the host rock. In SE PA, the diabase intruded Cambro-Ordovician limestone — carbonate rock, chemically reactive, full of iron and the right chemistry to make ore drop out of solution. In New Jersey, the diabase intruded Triassic basin fill — sandstone, shale, lake sediments — the wrong host rock. No limestone, no magnetite ore bodies, no gold.
It's not the magma. It's the rock the magma intrudes into.
The Triassic diabase in SE Pennsylvania isn't one big sheet. It's at least four distinct sheets, each one a separate pulse of magma. The H.R. Naslund 1998 NYSGA field trip guide B3 mapped all four, and they correspond to the different Cornwall-type mining districts:
| Sheet | Location | Major Mines / Districts |
|---|---|---|
| Gettysburg Sheets | Gettysburg Basin (Adams/York/Franklin Cos.) | Dillsburg District, Rossville, York Haven |
| Harrisburg Sheets | West of the Susquehanna (Dauphin/northern York Cos.) | Cornwall Mine, several smaller |
| Quakertown Sheets | Upper Bucks / Lehigh / Northampton Cos. | Several small prospects, mostly iron |
| Morgantown Sheets | Berks Co. south of Hopewell | Grace Mine, Jones Mine, French Creek district |
Each sheet was emplaced in essentially the same time window — a few hundred thousand years — but they are physically separate intrusions. Each one has its own contact zones, its own hydrothermal systems, and its own potential for Cornwall-type ore.
Smith, Rose, and Lanning (1975) established in their GSA Bulletin paper that the PA diabase falls into at least three compositionally distinct types — York Haven type (high-TiO₂, quartz-normative tholeiite, the most common in the Gettysburg Basin), Rossville type (low-TiO₂, more primitive, like island-arc tholeiite), and Quarryville type (olivine tholeiite, a separate magma pulse, named for the Lancaster County area).
For the prospector, the important thing is that the diabase didn't intrude random spots. The diabase dikes and sills in PA parallel Precambrian and Paleozoic dike trends, which means they followed pre-existing zones of crustal weakness. The same basement weaknesses that guided the diabase also controlled Paleozoic deformation and the location of the Triassic basins. If you can map the dike trends, you can predict where mineralized contact zones are most likely to occur.
At the contact between diabase and limestone, three things happen at once:
Pure thermal recrystallization. Limestone turns into marble. Shale turns into hornfels, with new minerals like cordierite, garnet, andalusite, pyroxene, scapolite. Sandstone turns into quartzite. No chemical change — just heating and recrystallization.
This is where the ore forms. Hot, chloride-rich hydrothermal fluids — carrying iron, copper, sulfur, gold, cobalt, silica — react with the limestone. The iron precipitates as magnetite, replacing the limestone. Sulfur combines with copper to form chalcopyrite. Sulfur combines with iron to form pyrite and pyrrhotite. The calcium from the limestone goes into actinolite, chlorite, garnet, and epidote — the gangue minerals that surround the ore.
The volume of fluid required to make a Cornwall is staggering. Eugster and Chou (1979) calculated that the Cornwall orebody required fluids with up to 30,000 parts per million dissolved iron. That's a fluid that is essentially a hot, dense, iron-bearing brine. And it had to circulate through the contact zone for a long time to deposit that much magnetite.
The calcareous rock is altered to a greenish pyroxene-garnet-epidote mass called tactite (or skarn in the broader sense). The Dillsburg tactite, described by Hotz (1950), is "pyroxene, garnet, and lesser amounts of epidote and calcite" with relict conglomerate textures. The tactite is the host rock for the magnetite ore — when you see greenish, heavy, garnet-studded rock in a Cornwall-type district, you're looking at the ore host.
Multiple authors, multiple methods, all converging on roughly the same range:
That's the practical range to remember: 500–700°C at about 1,500 bars. Hornfels facies. Gold-bearing sulfides form at the lower end of this range as the system cools.
Gold is not in the magnetite. Gold is in the sulfide minerals that co-precipitate with the magnetite. Specifically:
The 67,000 ounces of gold recovered at Cornwall from 1908 to 1973 was extracted from chalcopyrite concentrates. The chalcopyrite is mined for copper, and the gold comes along for the ride as a byproduct. The chalcopyrite crystals contain gold locked in their structure at the atomic level — invisible to the naked eye, recoverable only by chemical processing.
Common at Dillsburg and other Cornwall-type deposits. Pyrite concentrates can carry gold, though typically at lower concentrations than chalcopyrite.
A common accessory mineral at these deposits. Hotz (1950) confirmed pyrrhotite at Dillsburg alongside chalcopyrite.
Gold occurs as invisible (refractory) gold within the sulfide crystal lattice and as sub-micron inclusions. It is not visible to the naked eye in the ore — you don't see gold in Cornwall iron ore until you process the chalcopyrite. The largest documented gold nugget ever found in Pennsylvania (11+ ounces, recovered in 1938 in the York County region) was almost certainly weathered from a Dillsburg-type magnetite deposit. That nugget was the gold that got released from a sulfide crystal somewhere upstream and concentrated in a stream bed over decades of erosion.
Everything we've talked about — the recipe, the engine, the diabase, the contact metamorphism, the sulfides — was proven at one mine. The Cornwall Iron Mine, in Lebanon County, Pennsylvania. The type locality for what geologists call Cornwall-type deposits. The American cousin of Olympic Dam.
| Commodity | Production | Notes |
|---|---|---|
| Iron ore | 106–140 million tons (1742–1973) | Two main magnetite lenses 30–50 m thick, >1,000 m strike length |
| Gold | 67,000 oz (1908–1973) | Recovered from chalcopyrite concentrates. First refined 1908: 35 troy oz from initial batch. 1953 alone: 1,700 oz from ~1.4M tons ore. |
| Silver | 443,000 oz (1908–1973) | Byproduct of copper concentrates |
| Copper | ~0.4% average grade, 0.29% per Naslund 1998 | Chalcopyrite ore, byproduct |
| Cobalt | 400,000–600,000 lbs | First commercially produced 1941 for the WWII war effort |
Cornwall is in Cornwall Borough, Lebanon County, PA. Coordinates: 40.27072°N, 76.40700°W. The mine sits on the York Haven diabase sill where it intruded into the Cambro-Ordovician limestone. The famous "Big Hill" pit lake is still visible from US Route 322 — it's now a state park and a popular fishing spot.
Cornwall was discovered in 1732 by Peter Grubb, who noticed the heavy black boulders in the field while plowing. The first furnace was blown in 1742. Mining continued for 231 years, in three open pits (Big Hill, Middle Hill, and Grassy Hill, now merged into one water-filled pit), and processing continued until 1973. That's one of the longest continuous mining operations in American history.
The gold business was a side hustle. Cornwall was an iron mine. From 1908, when the first chalcopyrite concentrate batch was refined and yielded 35 troy ounces of gold, through 1973, the mine produced 67,000 ounces of gold as a byproduct of copper recovery. The gold never drove the operation. The iron drove it. The gold was a bonus.
This is where it gets interesting for prospectors. The Jones Mine, about 2 miles northeast of Morgantown in Berks County, sits on the same diabase sheet (Morgantown Sheet) and the same Cornwall-type geological model. But the Jones Mine has 1% to 6–7% copper — even higher than Cornwall's 0.4%. And it was never assayed for gold. The records just don't exist. Higher copper means higher gold potential — at Cornwall, the gold came from the copper concentrates. But Jones was never tested.
That's an open target, partner. More on this in Section 11.
Cornwall is the type locality and the biggest producer, but it's not the only Cornwall-type deposit in SE PA. There are several others, and the prospector should know about all of them.
Operated by Bethlehem Steel from 1958 through the late 1970s, Grace Mine is about 2 miles north of Morgantown, in Caernarvon Township, Berks County. The ore is magnetite, replacing Cambro-Ordovician limestone, with associated sulfides: pyrrhotite, chalcopyrite, marcasite, pyrite, goethite, hematite, covellite, digenite. Same Cornwall-type model — diabase intrusion, contact metamorphism, magnetite replacement, sulfides.
The Grace Mine produced "valuable byproducts of iron, copper, cobalt and GOLD" from sulfide concentrates, per the USGS MRDS record (#10067406). The Tsusue (1964) study "Mineralogy of the Grace Mine magnetite deposit" is the foundational reference. The Grace Mine confirms that the Cornwall-type gold-bearing model is not unique to Lebanon County — it works in Berks County too.
The Jones Mine is the most interesting target on this list. Why? Because nobody has tested it for gold.
Located 3/4 mile east of Joanna, about 2 miles northeast of Morgantown in Berks County, the Jones Mine has magnetite ore with 1% to 6–7% copper — even higher than Cornwall's 0.4%. The high copper is in chalcopyrite. The mine historically supplied iron ore to the Rebecca, Warwick, and Joanna furnaces in the 1800s. It was a working Cornwall-type mine.
But the modern era: the Jones Mine has never been assayed for gold. The records just don't exist. The same geological model as Cornwall (which produced 67,000 oz gold from similar chalcopyrite concentrates) plus higher copper content = untested but very high gold potential. This is an open target in the truest sense — nobody has even checked.
The Dillsburg district is a 10,000-foot-long, 1,500–3,000-foot-wide arc-shaped belt about 1.5 miles east of Dillsburg, in Carroll Township, NW York County. About 15 miles southwest of Harrisburg on US Route 15. The district had 30+ separate mines operating between 1855 and about 1915 — McCormick, Underwood, Longnecker, Bell, Jauss, Mumper, Grove, Butler, King, Reynolds, and many more. The largest areal extent of any Cornwall-type locality except Cornwall itself.
Production: about 1,500,000 tons of iron ore in that 60-year period. The geology: two diabase sheets (a thin upper sheet and a thick lower sheet) sandwich a 200–300 foot plate of Triassic sedimentary rock. Limestone conglomerate lenses within the sedimentary plate were replaced by magnetite. The lower diabase differentiated in place to form granophyre — a more evolved, iron-rich differentiate that is the interpreted source of the iron-bearing solutions.
The Bureau of Mines did WWII-era exploration here — July 1944 magnetic survey, 1945–1946 diamond drilling, 18 holes totaling 5,551 feet, ~2.5 square miles of magnetic survey. Part of the strategic minerals program during the war. A. F. Buddington supervised the Geological Survey cooperation. The drilling identified several additional ore bodies and a previously unknown magnetic anomaly.
Thomas Edison himself visited the Jauss Mine in August 1906 to evaluate the magnetite fields for his iron ore concentration venture. That's how important Dillsburg was in its day.
Sulfides at Dillsburg: pyrite, chalcopyrite, pyrrhotite. Magnetite is accompanied by pyroxene, chlorite, or garnet (metasomatic gangue). The sulfide assemblage is the same as Cornwall, which means the gold potential is the same as Cornwall. The Dillsburg district is the most likely source of all the placer gold in the York County creeks.
The largest documented gold nugget ever found in Pennsylvania — 11+ ounces, recovered in 1938 — came out of the York County region. That nugget was almost certainly weathered from one of the Dillsburg magnetite deposits. The gold was released from a sulfide crystal somewhere upstream and concentrated in a stream bed over decades of erosion. The Dillsburg district is the source. The creeks downstream are the destination.
Cornwall-type deposits don't exist in a vacuum. They are part of a global family of ore deposits called IOCG — Iron Oxide Copper Gold — and the most famous member of the family is the Olympic Dam deposit in South Australia.
Olympic Dam is, simply, one of the largest ore deposits on Earth. It contains:
That's an enormous amount of metal. Olympic Dam is the deposit that the entire IOCG family is named for — Iron Oxide (magnetite and hematite), Copper (chalcopyrite), Gold (in the sulfides). Same geological model as Cornwall. Same recipe. Just at a much larger scale.
Naslund (1998) places Cornwall-type deposits as a low-apatite variant of Kiruna-type (Sweden) iron oxide-apatite deposits, and notes that Kiruna/Cornwall-type deposits worldwide are now recognized as part of the IOCG family. Why this matters for Pennsylvania: the same geological setting that creates Cornwall, Grace, and Dillsburg also creates the conditions for gold-bearing sulfides. The diabase is the magma. The chlorine-bearing hydrothermal fluids are the transport medium. The limestone is the trap. The sulfides (with their gold content) drop out at the contact. This is a globally robust metallogenic model, not a Pennsylvania quirk.
So you understand the geology. Now what? How do you actually apply this knowledge in the field? Here are the things to look for.
The Cornwall-type gold deposits are at the contact between Triassic diabase and Cambro-Ordovician limestone. If you can map that contact — and a good geological map of SE PA will show you — you're looking at the right neighborhood. The four diabase sheets (Gettysburg, Harrisburg, Quakertown, Morgantown) are your guides. Where the diabase touches limestone is where the action is.
The contact zone is altered to a greenish pyroxene-garnet-epidote mass called tactite or skarn. This is the ore host. If you're walking a creek bed and you see a greenish, heavy, garnet-studded rock, you're looking at the right kind of country rock. Heavy minerals in the creek — magnetite, pyrite, chalcopyrite, garnet — are your indicators.
Chalcopyrite is the primary gold carrier. If you find chalcopyrite in a pan concentrate, you're potentially downstream of a Cornwall-type deposit. Even tiny amounts are significant. Pyrite and pyrrhotite are secondary indicators. Look for these in heavy mineral concentrates.
If you find gold in a SE PA creek, follow it upstream. The coarser the gold and the closer to bedrock, the closer you are to the source. The biggest nuggets come from the smallest drainage basins — the ones that head in the diabase-limestone contact zones. The 11-ounce York County nugget came from a small stream, not a major river.
Cornwall-type deposits generate distinctive stream sediment signatures: anomalous copper, anomalous cobalt, anomalous gold in the heavy fraction. Old USGS and PA Geological Survey stream sediment data exist for parts of SE PA. If you're serious about a specific area, the historical data is a starting point.
Almost all of SE PA is private land. Cornwall, Grace, Jones, and Dillsburg are all on or adjacent to patented mining claims or private property. Get permission before you sample. Stay off posted land. The gold isn't going anywhere — there's plenty of public-land analog to chase elsewhere first.
Based on the Cornwall-type model, the SE PA gold-bearing geology is not fully explored. Here are the highest-priority untested or under-tested targets, in rough order of promise:
Already discussed in detail. Same diabase sheet as Grace Mine, same Cornwall-type model, 6–7% copper (more than Cornwall), never assayed for gold. If you were looking for a place where the geological model says gold should be and modern exploration hasn't checked, this is it.
Historical mining at Jones supplied iron to local furnaces. The chalcopyrite is documented. The gold is the open question. The 67,000 ounces at Cornwall came from similar chalcopyrite. Multiply the higher copper content by the proven gold-per-copper ratio and the math gets interesting fast.
Coordinates: about 40.155°N, 75.892°W (2 mi NE of Morgantown, Berks County). Status: historically mined, currently inactive, on private land.
The Lockatong Formation is a sequence of lacustrine (lake-deposited) shales, mudstones, and siltstones in the Newark Basin. Famous for its red/black cycles. Some of the Lockatong contains carbonate beds. If CAMP diabase cuts the Lockatong (and it does in some places), and the Lockatong has carbonate, then the same Cornwall model could apply. This is a research-grade target, not a proven one, but the model says it's possible.
The Morgantown Sheets are the diabase source for the French Creek district mines: Elizabeth Mine, Susie Mine (both at St. Peters Village, closed 1928), Hopewell Mine, and others. Originally opened as copper mines. Magnetite ores with ubiquitous pyrite and chalcopyrite. This district is on the same diabase as Grace and Jones, and it has documented copper. Whether it has gold is undertested. Historical records exist but modern assays are rare.
The Connecticut Valley basin is the same Triassic rift as the Gettysburg and Newark basins. The Portland Formation (Jurassic) has local limestone interbeds. The CAMP diabase is present. If the diabase cuts the limestone, the Cornwall model might apply. This is a research target that would require significant field work to evaluate. Not a high-priority target compared to PA, but scientifically interesting.
The same South Mountain belt that hosts the Catoctin Furnace in Maryland extends into PA (Cumberland, Franklin, Berks counties). The Catoctin Furnace was iron, not gold, but the geological setting is similar to the South Mountain belt that hosts Virginia's gold-pyrite belt further south. The PA portion of South Mountain is mostly iron-bearing, but a comprehensive gold survey of the South Mountain extension in PA has not been done. Worth a research trip if you're in the area.
The single most useful research you could do on SE PA gold is a systematic chalcopyrite survey of the Cornwall-type districts. Sample chalcopyrite from each of the four diabase sheets. Assay for gold. Compare to Cornwall's 67,000 oz / 230 years baseline. If chalcopyrite from Jones, Grace, or Dillsburg assays as high as Cornwall's chalcopyrite, that's a publishable result and a serious exploration target. If it's lower, we learn something about why Cornwall was special.
This is the kind of work that could change how we think about Pennsylvania gold — and the kind of work that hasn't been done systematically in the modern era.
This guide is based on primary geological literature and authoritative field trip guides. Every claim is sourced. Where to read the originals: