Cast Iron Skillet Identification: Foundry Marks, Gate Marks and Era Registry

A gate mark feels like a healed seam. Run a thumbnail across the base and it catches — the bar stands maybe half a millimetre proud, sometimes with a small stub where the sprue was snapped off. It is not a crack and it is not damage. Some owners find it and panic.

Foundries that used bottom-gated moulds fed iron in through the base of the pan. Side-gating moved that entry point to the rim, where the scar could be ground off during finishing, which is why post-1890 pans have clean bottoms and a faint ground band on the outer rim edge instead.

Two cautions. The cutover was not a single day and not every foundry moved at once, so a Southern jobbing foundry could still be gate-marking well into the 1900s. And a small number of specialist and reproduction pieces have been made with gate marks deliberately. Treat the gate mark as a strong indicator, not a certificate.

Mass is quoted for the ten-inch class — a No. 8 in pre-1960 numbering, a 10.25 in pan in modern sizing, and the closest equivalent for makers who never used the old numbers. Rim wall is measured 10 mm below the lip, which is the thinnest and most comparable section on any skillet. Era ranges follow the standard collector chronologies, and those chronologies genuinely disagree at the boundaries by a few years in both directions.

Read the surface column carefully, because it does more work than anything else on this page. Machine-ground means the cooking face was milled or ground smooth after casting. As-cast means it was not.

One column deserves staring at. A Wagner Ware No. 8 lands around 4 lb; a modern mass-market 10.25 in pan lands around 5.4 lb. Same nominal size, a pound and a third apart, and the difference is almost entirely wall section. That single fact explains most of what people mistake for technique.

The Finex row is the outlier that proves the point in reverse. At close to 8 lb for a ten-inch pan it holds heat like nothing else on the list and takes about twice as long to get there. Neither behaviour is better. They are different tools that share a name.

Pans went largely unmarked after the 1960s, and a great many pans that were marked have had the mark ground away by an over-enthusiastic restoration. Neither situation is hopeless. Work the tree below in order and stop when you have two independent findings that agree.

The number cast into the handle or the base is not a diameter. It is a stove-eye number, the size of the hole in the top of a wood or coal range that the pan was designed to drop into. That is why a No. 8 is roughly 10.5 inches across the rim rather than eight of anything.

Diameters shift by a fraction between foundries and between eras, so treat the figures below as the centre of a band rather than a specification.

Wall section drives thermal response, and thermal response is what people are describing when they argue about cast iron on the internet.

A pre-1960 pan with a 2.5 mm rim wall and a 4.1 lb mass comes up to temperature quickly, loses heat quickly when cold food hits it, and recovers quickly. A modern 4.0 mm-walled pan at 5.4 lb does the opposite on all three counts. Put four steaks in the vintage pan and the surface temperature drops further and faster; put them in the modern one and you get a slower, steadier sear. Neither is a defect.

Surface finish drives the other half. A machine-ground face has peaks and valleys measured in tens of microns. A sand-cast face has a visible pebbled texture you can feel with a fingertip. Both will cook eggs cleanly once the film is closed, but the sand-cast pan needs two to three more coats to get there because the seasoning has to fill the valleys before it presents a flat surface to the food.

This is the mechanism behind most cast iron folklore. Somebody follows a method that was developed on a smooth 4 lb pan, applies it to a rough 5.4 lb pan, gets a different result, and concludes the method is wrong. It usually was not.

Fakery in cast iron is narrow and predictable. It concentrates on a handful of high-value marks, and it nearly always fails on the same tells.

Recast fakes are made by taking a mould from an original pan. Iron shrinks as it cools, so a pan cast from a mould taken off a finished pan comes out around one to two percent small. A No. 8 that measures 10 3/8 in instead of 10 5/8 in is a recast until proven otherwise. The lettering is the second tell: on a recast, raised letters lose their crisp shoulders and look slightly melted, and the fine detail inside a logo goes soft.

Wapak reportedly produced pans carrying ERIE-style marks during its own operating years, which muddies the water at the top of the market and has done for a century. And the modern Wagner's 1891 Original reissues are not fakes at all — they are honestly marked reproductions from the 1990s that get resold as vintage by people who did not check the mark text.

One practical rule: weigh it, measure the rim, and compare against the registry row for the mark you think you are looking at. Fakes fail on dimensions far more often than they fail on appearance.

• ▸Thin vintage, machine-ground: preheat on medium for 4-6 minutes, not 10. Overshooting a 2.4 mm wall is easy and it is the fastest route to a scorched film.
• ▸Thick modern, as-cast: give it 8-12 minutes on medium and accept that the handle will be too hot to hold. Rushing this pan produces the cold-centre complaint that gets blamed on cast iron generally.
• ▸Any pan with a full heat ring: it will rock on a flat glass hob because the ring, not the base, is the contact surface. That is not warp, it is design intent for a stove that no longer exists in most kitchens.
• ▸Any pan under about 4.3 lb in the ten-inch class: treat it as fragile against thermal shock. Thin sections crack more readily than thick ones when cold water meets hot iron.
• ▸Enamelled cast iron: none of the seasoning logic applies to the glazed interior. The bare rim band is the only ferrous surface on the pan that needs oil.
• ▸Induction: every pan in the registry works on induction electrically, but a full heat ring or a warped base will cause the hob to cycle and buzz. Flatness matters more than mass here.

Does a rough bottom mean it is old? No — it usually means the opposite. Pre-1960 pans were finished on the cooking face and often left the exterior fairly rough; post-1965 pans are rough everywhere. Judge on the inside, not the outside.

Is an unmarked pan worthless? Not remotely. A late BSR with a notched heat ring is a genuinely good cooking pan at an IronGrade of 74 and costs a fraction of a marked Griswold that cooks the same. Collectors pay for marks. Cooks should pay for wall section and surface.

The mark is worn nearly flat — is it ruined? No. The registry is designed to work without it. Mass, rim wall and heat-ring type will get you to a class in under two minutes.

Can heat rings be added or removed? Removal happens, usually by someone flattening the base on a grinder, and it leaves a tell: the outer band is bright and dead flat while the rest of the underside has casting texture. That pan has lost material and probably some IronGrade with it.