Antique door knobs are identified by documenting the full assembly, finding maker or patent marks on concealed surfaces, and matching the complete pattern to a period manufacturer catalog; patents date the design, never the individual knob.
Two knobs that look identical under a magnifier can be a century apart. One carries a threaded shank from a late-1800s patent; the other is a 1980s cast-brass reproduction poured from the same-looking mold. The difference hides in surfaces nobody photographs, in construction details, and in catalog pages you have to hunt down.
Whether a set of old hardware came out of a farmhouse door or off a flea market table, the same sequence works. Document the whole assembly first. Look for marks where they actually live. Then let the strongest evidence—not the fanciest ornament—set the date.
Where Do Makers and Patent Marks Hide on Old Hardware?
Marks rarely sit on the knob face. Check the shank, the reverse of the rose or escutcheon, the lock-case interior, the faceplate, the spindle hub, and the key.
- Maker names, initials, trademarks, and catalog numbers
- Pattern or design names stamped by the foundry
- Patent dates and patent numbers, sometimes faintly
Worth knowing which names to search: Russell & Erwin, P. & F. Corbin, Sargent, Yale & Towne, Reading, Norwalk, and Branford. An absent mark proves nothing either way—plenty of genuine period pieces left the factory unmarked.
Before any cleaning, photograph both knobs, roses or backplates, lock case, faceplate, strike, spindle, screws, keys, and escutcheons. Record dimensions, finish, and whether the lock is a surface-mounted rim lock or a mortised lock recessed into the door edge. Polish now and you erase the finish and wear evidence you’ll need later.
Reading Construction and Material Without Getting Fooled
Construction dates better than ornament, and the spindle attachment is the single most informative detail on the set. Record whether the knob uses a threaded shank, a transverse pin, a retaining ring, or another system.
- Hand-forged iron: hammer marks, irregular surfaces, uneven weight
- Cast components: mold seams, substantial weight, deeper relief
- Stamped metal: thin, hollow-backed, shallow relief
- Machine-made pieces: uniform repetition across the set
For material, weight, color, and magnet response narrow the field—magnetic points to iron or steel, nonmagnetic to brass, bronze, and other alloys. These tests are preliminary only. They do not identify the alloy and they do not date the piece. Manufacturers also ran the same casting in several finishes, from Berlin bronze to oxidized copper to brushed brass to plain “antique,” and finish codes shifted over time.
One trap catches almost everyone: a cast-brass reproduction that resembles an old pattern. Resemblance is not evidence. Catalog match plus construction is.
Here is the dating evidence, ordered by how much weight it carries:
| Evidence Type | Example | What It Proves |
|---|---|---|
| Dated catalog match | Corbin’s illustrated catalog, 1870 | Pattern existed by that year |
| Patent record | US 459,444 “Knob attachment,” 1891 | Mechanism existed by that date |
| Construction method | Threaded shank vs. transverse pin | Narrows the manufacturing era |
| Stylistic period | Gothic, Eastlake, Art Nouveau ornament | Broad range only, never an exact year |
| Set consistency | Matching wear and finish on all parts | Whether components belong together |
Search the whole combination—knob, rose, shank, lock—against scanned manufacturer catalogs, not just a similar-looking knob face. Catalog entries often give the pattern name, company, year, material, dimensions, finish, and item number. Reproductions that mimic a listed pattern but fail the construction test are the most common misidentification in the field. If you’re replacing a missing knob on an otherwise complete set and need period-correct matches rather than a single unobtainable original, this roundup of tested vintage antique door knobs lists options worth comparing.
How Do Patent Records Set a Date Without Overstating It?
Search the patent number or its exact wording in Google Patents and read the filing and publication dates. Then treat those dates for what they are: benchmarks for the design, not stamps on your hardware.
A clear illustration is US Patent No. 459,444, “Knob attachment,” published September 15, 1891. You can pull the original record through its Google Patents page for the 1891 knob attachment and see the mechanism described in the claims. Popular patterns stayed in production for decades, so a knob can postdate its patent by thirty or forty years and still be period-correct.
Some useful anchors along the timeline:
- 1877: US Patent No. 191,363, May 29, covering knob-to-spindle attachment
- 1887: US Patent No. 355,311, “Knob-lock,” January 4
- 1942: US Patent No. 2,293,857, “Doorknob fastener,” tied to Schlage, August 25
None of these records will tell you what a knob is worth today. They establish identification and chronology only.
What Separates a Convincing Date From a Guess?
Build the date from the strongest evidence you actually have, not the strongest evidence that exists in general, and accept a range when that’s the honest answer.
- Photograph and measure the complete set before any cleaning.
- Classify the ornament—Gothic, Romanesque, Italian Renaissance, Louis XVI, Eastlake, Art Nouveau—for a broad period, not a year.
- Inspect every concealed surface for marks.
- Record construction and spindle attachment.
- Match the full assembly to a dated catalog entry.
- Verify any patent through the primary record.
- Date the weakest part honestly; knobs, roses, screws, and locks get replaced independently, so one component should not date the whole set.
Do that, and the verdict holds up to a second opinion. Skip it, and you’re dating a style and calling it a year.
Common Questions
Is a patent date the same as the manufacture date?
No. A patent date marks when the design or mechanism was protected, not when your particular piece was made. Popular patterns stayed in production for decades, so an 1891 patent could cover hardware built in 1920. Use the patent as the earliest possible date for the design, then confirm the actual production window through catalog evidence.
Does an unmarked knob mean it’s a modern reproduction?
Not at all. Many genuine antique knobs left the factory without any maker’s mark, initials, or patent stamp at all. Absence of a mark tells you nothing about age on its own. Rely on construction, spindle attachment, material, and a catalog match instead of the presence or absence of a stamp.
Should I clean or polish an old knob before researching it?
Don’t, not yet. Polishing, stripping, or chemically cleaning removes the original finish and the wear patterns that support a date. Photograph every component first and record dimensions, finish, and construction. Once the documentation is complete, cleaning is a separate decision from identification.
Sources
- Google Patents. “US Patent No. 459,444 — Knob attachment.” Primary record for the 1891 knob-attachment mechanism cited in the dating section.