US 816616

I'll verify the identifiers against live sources before summarizing.

Verification performed

Live searches were run for the exact identifier 816616 against patent and court-oriented sources. Findings below use the authoritative full text of US816616 supplied in my working context, cross-checked against Google Patents and the USPTO/Google facsimile PDF.

---

US 816616 — bibliographic data

Abstract: This is a 1906 grant; no formal abstract appears in the printed patent, and Google Patents supplies none. The document's opening statement of invention is the functional equivalent:

"The present invention relates to improvements in that class of furniture which can be converted to different articles for different uses. One of the objects is to provide a combined settee or sofa and table wherein the structure is such that the change from one to the other can be readily accomplished without the necessity of manually lifting any of the heavy members…"

The stated design context is that the structure is "particularly applicable for use in connection with convertible furniture, wherein a billiard-table constitutes a part," and one feature is leveling mechanism (screw-stems 13, knurled wheels 14, angle-brackets 15 with slots 16 and guiding-screws 17).

---

Plain-language overview of the claims

Claim 1 — core linkage geometry. A base having a seat and upstanding arms, plus a table that swings between (a) upright behind the base and (b) horizontal resting on the arms. An intermediate device (a link) is pivotally fixed at one end to the base below the arms and at the other end to the longitudinal central portion of the table. The table-side pivot sits higher than the base-side pivot, and when the table is in its rearmost position that table-side pivot sits at the rear upper corners of the arms. That arrangement lets the table be swung over the arms before the intermediate device has to move — i.e., the first part of the fold is essentially a free pivot.

Claim 2 — rock-shaft plus rigid links. A seat member with a seat and end arms; a table that rests flat on the arms over the seat and can move to upright behind the seat. A rock-shaft is journaled on the seat member near the rear, beneath the seat and between the arms, and links are rigidly attached to that shaft, with pivotal connections to the table. When the links are rearmost, those connections are at the upper rear corners of the arms, so the table can swing from upright to horizontal independently of link movement. This is the equalization feature: because both links are rigid on one common shaft, both ends of the table must travel together, avoiding the binding the specification says occurs "with independently-movable links."

Claim 3 — removable back covering the links. A seat member; a table that swings behind the seat member and can also rest on it; spaced links connecting seat and table; and a removable back with detachable connections to the links, the back covering the links when installed.

Claim 4 — removable back between the arms with spring clips. Adds structural detail: seat member with upstanding arms, table swinging on the arms and behind the seat, spaced links arranged between the arms, and a removable back that fits between the arms in advance of the links, secured by spring-clips connecting the back to the links (specification: spring-clips 30 engaging links 24).

Claim 5 — links as back supports. Seat member with arms; table swinging behind and resting on the arms; spaced links connecting the two and located between them; and a removable back fitted between the arms that rests against the links, the links themselves constituting the supports for the back.

In short: claims 1–2 cover the balancing/equalizing linkage; claims 3–5 cover the removable upholstered back and the way the linkage doubles as its mounting structure. The specification also describes a modified embodiment (Fig. 7) in which the supplemental back 28a itself is the longitudinal link connection between the table's central portion and the rear lower portions of the arms.

---

CAFC 2026 docket check

Result: no Court of Appeals for the Federal Circuit docket activity exists for this patent number, and none is possible. Two independent reasons:

1. The patent expired April 3, 1923 on the statutory term of that era — well before the Federal Circuit existed. The CAFC was created by the Federal Courts Improvement Act of 1982.
2. CAFC dockets are indexed by appeal/case number, not by patent number, so a "docket search on 816616" is not a meaningful query in that court's records.

I therefore cannot report any 2026 appellate proceeding, and I will not manufacture one. If you are chasing a real 2026 appeal, you likely have an application number, a case number, or a different patent number in hand.

---

Identifier-collision warnings (found during targeted searching)

Because you asked for the exact number only, these matched 816616 but are not US816616A — flagged so they are not conflated:

• DE 816616 — a German patent cited in EP 2 979 538 A1 and US 9,820,444 (lopping-shear pull-cable/transmission arrangement). Different jurisdiction, different subject matter.
• Batch No. 816616 — a MedChemExpress certificate of analysis (ART0380). Not a patent.
• Report 816616 — a Mira Inform credit report for "Sol Photovoltaik Solution." Not a patent.

---

Downstream citations of US816616A (for context)

Per Google Patents' "Cited By" listing: US 3,913,973 A (Mintz, Combination pool table and couch, 1975); US 7,621,818 B1 and US 7,828,668 B1 (Tippit, System for assembly and leveling of playing surface for billiards); US 11,065,169 B2 (Sinclair, Convertible examination table). These indicate the patent is treated as prior art in the pool-table/couch conversion art.

Confidence note: Bibliographic data, claim text, and specification quotes above are drawn from the authoritative full text supplied and corroborated by the Google Patents record and the USPTO facsimile PDF. The only genuine ambiguity I found is the application serial number (275,979 vs. 276,979 in OCR), which I have flagged rather than resolved by assumption.

I'll search for litigation involving US patent 816616.

Let me run additional targeted checks on litigation databases.

Answer: No litigation found for U.S. Patent No. 816,616

Bottom line: Based on my searches of patent-litigation and general web sources, I found no record of any litigation — district court, ITC, PTAB/IPR, or CAFC — involving U.S. Patent No. 816,616. I could not identify a plaintiff, defendant, jurisdiction, case number, filing date, or outcome, because no such case appears to exist.

Patent identified

• U.S. Patent No. 816,616 — "Convertible sofa and table"
• Inventor: Nathaniel B. Stone (Outlook, Washington)
• Application No. 275,979, filed Sept. 16, 1903 (renewed Aug. 26, 1905)
• Granted/issued April 3, 1906
• Assignee: Individual (never assigned to a corporate entity)
• Legal status: Expired – Lifetime, "Anticipated expiration" April 3, 1923
• Claim set: 5 claims (rock-shaft/links convertible sofa-billiard table; removable back with spring-clips)
• Source: https://patents.google.com/patent/[US816616](/patent/US816616)/en

Why litigation is essentially foreclosed

1. The patent expired on April 3, 1923, and no term extension or reissue/renewal appears in the record. An expired patent cannot be asserted for prospective infringement, which eliminates essentially all 20th- and 21st-century causes of action.
2. The patent issued in 1906 to an individual inventor, with no corporate assignee in the Google Patents family data (Family ID 2885098). There is no recorded assignment trail to a litigious operating company or NPE.
3. Google Patents' forward-citation table for US816,616 lists only three to four later patents that cite it as prior art (e.g., US 3,913,973 to Mintz; US 7,621,818 and US 7,828,668 to Tippit; US 11,065,169 to Sinclair) — these are citations, not litigation. I found no "Litigation" tab entry or case data attached to US816,616.

False positives I deliberately excluded (per the strict-ID rule)

Your instructions require literal interpretation, so I am flagging the "616" hits that are different patents or different subject matter entirely and must not be attributed to US816,616:

Caveats on completeness

• Searching was performed against publicly indexed web and patent sources (Google Patents, court document PDFs, ITC initial determinations, Federal Circuit and Supreme Court filings). Pre-1923 U.S. litigation is very poorly digitized; a 1906–1923 infringement suit in a state or federal court, if one ever occurred, may simply not be online. I cannot rule that out with absolute certainty, and I am stating that limitation explicitly rather than asserting a negative I cannot fully verify.
• I was unable to directly query the Unified Patents litigation portal (portal.unifiedpatents.com/litigation/caselist), PACER, or Docket Navigator in this session; those are paywalled or access-restricted. If you have credentials, those are the best places to confirm a negative. Note also that PACER coverage is generally post-1990s and would not capture litigation during the patent's 1906–1923 term.
• The one residual, non-infringement legal exposure worth noting: an expired patent like US816,616 could in theory have been cited as prior art in a later validity challenge, but that is a citation, not litigation involving the patent itself, and I found no such challenge either.

Conclusion: No known litigation involves US patent 816616.

Proceedings overview

Total AIA trial proceedings on file for US 816616: 0. The structured "PTAB proceedings on file" block (canonical, sourced from USPTO ODP) returns no IPR, PGR, or CBM proceeding for this patent — no active proceedings, no claims invalidated, no claims sustained, no settlements, no institution denials, and no Federal Circuit appeals. My independent web searches for US816616 PTAB inter partes review and "8,166,616" IPR petition patent trial and appeal board surfaced no petition, institution decision, final written decision, or appeal naming this patent; every hit was a different patent whose number merely contains "816" or "616" (see the confusion warning below).

Defensive posture: the usual defensive playbook does not apply, because US 816616 expired on 1923-04-03 — the statutory expiration date recorded by ODP, 17 years after grant on 1906-04-03 (pre-1995 term). A patent that has been expired for over a century cannot support a live infringement claim (the § 286 six-year damages lookback cannot reach 1923), and it therefore attracts no IPRs: there is no commercial incentive, and estoppel benefits under § 315(e)(2) are worthless against a dead patent. There is no "hardened patent" vs. "claims canceled" story here — the correct answer is that the entire patent is inert, and the PTAB docket is empty. If you have received a demand letter citing "US 8,166,616," you are almost certainly being cited the wrong patent number (see below).

Why the record is empty (not merely "not yet indexed")

• AIA trials are unavailable as a practical matter. IPR/PGR/CBM exist under 35 U.S.C. §§ 311–329, enacted by the AIA on 2011-09-16. A petition requires a petitioner with a live stake (a defendant in district court litigation or an ITC respondent). With the patent expired since 1923, there is no defendant and no litigation, so no petitioner.
• ODP confirms it. The structured block states ODP returns no AIA trial proceedings "as of the most recent ingest." Given the patent's age, this is not an indexing lag — it is the correct permanent state of affairs.
• No reexamination or interference either. The ODP/Google Patents record for this document shows no reexamination certificate, no post-issuance disclaimer, and no interference; the only post-issuance legal event is "Anticipated expiration" on 1923-04-03.

Sequencing caveat (why "court" enters here at all)

• 1903-09-16 — application filed (serial No. 275,979).
• 1905-08-26 — renewed (per the specification's filing data).
• 1906-04-03 — patent granted, 2 sheets / 5 claims.
• 1923-04-03 — term expired.
• 2011-09-16 — AIA trial provisions take effect (long after expiration). No petition has ever been filed under them against this patent.

⚠️ Number-confusion warning — do not let these contaminate your file

My searches surfaced several real PTAB proceedings against similarly numbered but unrelated patents. None of them involve US 816616, and none can be used as authority for or against it:

• U.S. Patent 7,822,816 — RPX Corporation v. MacroSolve, Inc., IPR2014-00140 (instituted; panel: Medley, Jefferson, Chen). Different patent; different owner.
• U.S. Patent 8,166,081 — challenged by Hyundai Motor America in IPR2021-01267 (Hyundai Motor America v. STRATOSAUDIO, Inc.) and by Volkswagen Group of America in IPR2021-00721, among others. This "8,166,081" is the single most likely source of a mis-cited "8,166,616."
• U.S. Patent 10,850,616 — subject of IPR2024-00630.
• U.S. Patent 7,290,616 — subject of IPR2020-01648.
• U.S. Patent 9,771,169 / 8,985,816 / 8,903,307 — separate Volkswagen and other proceedings that bubbled up in the same queries.

If your demand letter, complaint, or invalidity chart cites "the '616 patent" alongside a modern product, verify the number digit-by-digit. A demand premised on US 816616 would be meritless on expiration alone.

Strategic summary

Claim status for US 816616. There is no PTAB record of any kind on claims 1–5, so nothing is "canceled" and nothing is "sustained" in the AIA sense. All five claims — claim 1 (the combination claim to the base member, table member, and intermediate device with the pivotal-connection geometry), claim 2 (the rock-shaft-and-links species), claim 3 (spaced links plus removable back with detachable connections to the links), claim 4 (spring-clips connecting the back to the links), and claim 5 (links as supports for the back) — stand as issued, unamended and unadjudicated. "Untested" is technically accurate but misleading: they are untested because they expired unenforceable in 1923, not because the patent has been lucky. For practical purposes every claim is dead — by operation of 35 U.S.C. § 154's term, not by any FWD. No FWD exists to link to or quote.

Estoppel landscape. Section 315(e)(2) estoppel is inapplicable: it attaches only to a petitioner in an instituted IPR, and no IPR was ever instituted. There is no petitioner, no privy, no ground "raised or reasonably could have raised," and therefore no estoppel benefit and no estoppel exposure for anyone. Equally, there is no IPR-derived narrowing that would help you read claims off your product. Your invalidity position, if you ever needed one, would rest entirely on § 102/§ 103 art predating the 1903 filing date — an academic exercise, since the patent cannot be asserted.

Pattern signals. No petitioner has filed even one IPR against this patent, let alone multiple; there is no patent-owner appeal history because there is no patent owner of record beyond the original individual inventor, Nathaniel B. Stone of Outlook, Washington (the ODP record lists "Individual" as both original and current assignee). There is no defensive aggregator (Unified Patents, RPX, or similar) in the chain — Unified/RPX appear in my search results only in connection with other patents. The RPI-privity fights that dominate modern Unified-funded IPRs (cf. the American Patents v. Unified Patents briefing on U.S. 7,373,655) have no counterpart here. The Google Patents "Cited By" list shows only three later artifacts — Mintz (US 3,913,973, 1975), the Tippit billiards-leveling family (US 7,621,818 / US 7,828,668, 2009/2010), and Sinclair (US 11,065,169, 2021) — i.e., this 1906 patent is cited as prior art by modern furniture and billiards patents, not litigated or challenged itself.

Recommended next steps

1. If you are a defendant or a demand-letter recipient, verify the patent number before anything else. Confirm whether the assertion is actually against US 816,616 (1906, expired 1923) — in which case respond that the patent expired on 1923-04-03 per the ODP legal-status record and cannot support infringement liability, and consider fee-shifting exposure for the asserting party. If the intended patent is instead 8,166,081, 9,816,616, 10,850,616, or another modern "616," my analysis must be redone against that patent — re-run the PTAB check at PTAB E2E (PTACTS) and patents.google.com.
2. If you need an authoritative zero-result confirmation for a filing or opinion letter, pull the ODP record yourself and preserve a dated screenshot, plus a PTAB E2E party/patent search for "816616." Cite the ODP legal-status line: patent term expired 1923-04-03; status "Expired – Lifetime."
3. If an active AIA proceeding somehow exists that ODP has not ingested, the statutory clocks are: institution decision due within 6 months of the § 313 preliminary-response deadline or the petition's filing date, whichever is later (35 U.S.C. § 314(b)), and a final written decision within 12 months of institution, extendable up to 6 months for good cause (§ 316(a)(11)). Nothing in my searches suggests any such proceeding exists for this patent — I found no petition number, no institution decision, and no FWD, and I will not invent one.
4. Nothing to monitor. With no active proceeding, there is no oral hearing, no FWD due date, and no CAFC appeal docket to track. Absence of PTAB activity is normally a weak signal (well-asserted patents eventually attract IPRs); here it is a strong and permanent signal, because the patent's enforceable life ended ninety-nine years before the AIA trial regime even began.

Confidence note: I have high confidence that no AIA trial proceeding exists for US 816616 — the ODP structured block says so explicitly and two targeted web searches corroborate it. I have not verified the docket for a hypothetical proceeding I could not find, for the obvious reason that I will not fabricate a proceeding number. The proceedings I did surface (IPR2014-00140, IPR2021-00721, IPR2021-01267, IPR2020-01648, IPR2024-00630, IPR2023-01041) belong to other patents, and I have flagged them only so they are not mistaken for this one.

I'll verify the assignment record against live sources before reconstructing the chain.

Confirmed the Google Patents legal-events record so far. Let me check for any recorded transfer involving the inventor and any trace of an assignee entity.

Inventors

The front page of US 816616 names exactly one inventor and no corporate applicant. Outlook was a small Yakima Valley community (platted c. 1902), and there is no evidence in the record — and none surfaced in searching — that Stone assigned his rights to an employer, a furniture manufacturer, or a billiard-table company before or at filing. No joint inventors, no "all inventors departing within 12 months" pattern, and no employer-assignment pattern can be established, because there is only one inventor and no assignee.

⚠️ Name-collision warning (do not conflate). Searching for "Stone" plus early-1900s patents surfaces John Stone Stone, the radio engineer whose specifications (e.g., Nos. 716,995; 767,970; 768,002) were litigated in Marconi Wireless Telegraph Co. of America v. United States. That is a different person in a different technology, and none of that "Stone" art involves US 816616. Only Nathaniel B. Stone of Outlook, Washington is the inventor here.

Original assignee

There is no original assignee other than the inventor personally. Google Patents records both "Original Assignee" and "Current Assignee" as Individual, with no recorded corporate owner and no assignment of record.

• Product embodying the claims: No evidence that any commercial product was shipped. The specification is written in the present tense of a workshop-style build ("the article of furniture consists of two principal parts"), describes a convertible billiard table/sofa, and claims no manufacturer, no trade name, and no licensee. No catalog, advertisement, or trade-press mention of a manufactured "Stone" convertible sofa-billiard table was found.
• Primary line of business: Not determinable. The record does not disclose Stone's occupation; his residence (Outlook, W.A.) does not itself establish a business.
• Current status: The patentee is an individual inventor from a 1903 filing; the patent expired April 3, 1923 and the legal status is "Expired – Lifetime." No successor entity, merger, or dissolution is on record because no entity ever held title.

Assignment timeline

The USPTO Assignment Center has no assignment records for US 816616. I found no recorded assignment of any kind — no assignment, security agreement, merger, change of name, license, release, or correction — and therefore no reel/frame, assignor, assignee, execution date, recording date, or correspondent of record to report. I will not invent one.

This is consistent with the Google Patents legal-events timeline, which contains only four events — application filed (1903-09-16), priority (1903-09-16), granted (1906-04-03), publication (1906-04-03), anticipated expiration (1923-04-03) — and no assignment entries at all. There is consequently no correspondent attorney to assess for recurrence, and no repeat-player signal to compute.

Important verification caveat — the online database cannot prove a negative here. Per MPEP § 301.01(a)(2), assignment records for patents recorded before May 1, 1957 are not held in the USPTO's online assignment system; they are maintained by the National Archives and Records Administration (NARA). The online Assignment Center result of "no records" is a statement about the electronic index, not about the paper record. A 1903–1923 paper assignment from Stone to a furniture or billiard concern could exist on microfilm at NARA and never appear online. Establishing a true zero would require a NARA pull of the pre-1957 assignment records. I flag this rather than assert an absolute negative.

Two further record points, offered without over-reading them:

• The specification's only other named individuals are John H. Siggers and Ernest L. Harkness, listed as witnesses to the inventor's signature. They are not attorneys or agents of record, and nothing links them to a later conveyance.
• The application serial number is itself ambiguous in the source material (275,979 in Google Patents metadata vs. 276,979 in the OCR of the granted specification, as flagged in the earlier bibliographic section). Anyone verifying by application number at assignmentcenter.uspto.gov or assignment.uspto.gov should search both the serial and the patent number, since a serial-number mismatch will return a false "no records."

Timeline diagram

This is not an ownership chain — no ownership transfer exists to diagram. It records only the patent's legal events, which is the complete set of dated entries the record supports.

timeline
    title Legal events for US 816616
    1903 : Application filed by Nathaniel B Stone
    1906 : Patent granted to Stone as an individual
    1923 : Patent term expired
    2026 : No recorded assignment found

NPE / troll-pattern signals

Inverse-signal note. Because US 816616 is cited by modern patents — US 3,913,973 (Mintz, Combination pool table and couch, 1975), US 7,621,818 and US 7,828,668 (Tippit, billiards-leveling), and US 11,065,169 (Sinclair, Convertible examination table, 2021) — it functions in the modern record as prior art, not as an asserted asset. That is a defensive-neutralizing role the patent occupies by operation of law, not a defensive-aggregator acquisition.

Verdict

Insufficient data (no records, or only the original assignment).

Justification: The only ownership fact the record supports is that US 816616 issued on 1906-04-03 to Nathaniel B. Stone, an individual, of Outlook, Washington, with no assignment, no reel/frame entry, no assignee entity, and no correspondent of record anywhere in the electronic assignment index or the Google Patents legal-events timeline. Because the record contains a single ownership link — the original inventor — there is no chain to analyze for shell entities, cascading transfers, or repeat correspondents, and I decline to characterize an absent chain. The NPE verdicts are independently foreclosed on the merits as well: the patent expired 1923-04-03, so it cannot be asserted, licensed for revenue, or transferred as a live asset. Caveat driving the "insufficient" rather than "operating-company" call: pre-May 1, 1957 paper assignment records sit at NARA, not in the online USPTO index, so a 1903–1923 paper assignment cannot be ruled out from the online system alone.

Verify at: USPTO Assignment Center and USPTO Patent Assignment Search (search both patent number 816616 and application serials 275,979 / 276,979); patent record at https://patents.google.com/patent/US816616/en.

Confidence note: High confidence that no assignment is recorded in the USPTO's electronic index — the Google Patents legal-events record shows none, and three targeted searches surfaced none. I have not verified the pre-1957 NARA microfilm assignment records, which is the one place a contrary finding could realistically live, and I state that limitation rather than assert a complete negative. The serial-number discrepancy flagged in the earlier bibliographic section directly affects how a verification search should be run and is carried forward here rather than resolved by assumption.

I'll verify the citation record for US816616 directly against live sources before analyzing prior art, since the full text I have shows a "Cited By" block but no applicant-cited references.

Key finding: the Mintz patent's citation table surfaced a second Nathaniel B. Stone patent (US786693A) and revealed that US816616's own Google record has no "Patent Citations" block. Let me verify both points.

Two important findings: US816616's record has no "Patent Citations" block, and its sibling patent US786693A (same inventor, same title, filed 1904-10-28) has a known serial-number anchor. Let me pull the citation lists of the patents that cite US816616, plus the pre-1903 pre-grant art.

Bottom line

The requested search returns a null set: US 816616 has no patent citations. Its official record contains no "[56] References Cited" block, no "Patent Citations" table, and the specification itself cites no prior patent, publication, author, or document of any kind. I searched for "each patent citation for 816616" and there is nothing to list. Below I (a) document that negative finding with the evidence, (b) explain why the record is empty rather than merely unindexed, and (c) supply the closest available substitutes — with their evidentiary weight honestly graded — plus a §102 mapping framework and one genuine prior-art wrinkle that connects back to the serial-number ambiguity flagged in the earlier section.

---

1. Evidence for the null result

Contrast that proves the point: the very same database shows a citation table for a comparable document. Google's page for US 3,913,973 displays "Patent Citations (9)" and "Citations (9)." For US816616 the corresponding table is simply absent. This is a real difference in the records, not a fetch artifact.

---

2. Why the list is empty (not "not yet indexed")

1. Era of grant. US816616 issued April 3, 1906. The practice of printing an examiner's and applicant's reference list on the face of the patent (the modern [56] block) post-dates it. The citation-system proposal attributed to H. C. Harry and submitted to the USPTO is dated 1947 (secondary-source corroboration only — I am citing this as support, not as primary authority). A 1903-filed / 1906-granted patent therefore has no front-page art list to retrieve.
2. The specification cites nothing. I read the full specification text supplied to me. There is no citation of any kind. The only art-adjacent statements are generic field statements — "improvements in that class of furniture which can be converted to different articles for different uses," and the functional observation that "with independently-movable links there is a tendency for one end to move in advance of the other and bind." Neither identifies a document.
3. Any examiner citations would live only in the paper file wrapper (1903–1906 prosecution file), which is not digitized in USPTO PatentCenter for that era. That is the one place an actual citation list could exist, and it is not retrievable online.

So the correct answer to "look at each patent citation for 816616" is: there are zero, and any list I produced would be invention rather than retrieval. I will not do that.

---

3. The reference universe that does exist around US816616, graded by evidentiary weight

⚠️ Read the "Weight" column before using any of these. Four of the five categories below are not citations in US816616.

3A. Companion patent by the same inventor (same title, different mechanism)

3B. The only pre-1903 reference in the cluster

Screening note: claim 1 is the broadest claim and therefore the best §102 target; claims 2 and 4 are the narrowest and hardest to anticipate (claim 2 requires a rock-shaft with rigidly affixed links; claim 4 requires spring-clips). A generic "convertible lounge" reference is unlikely to reach claim 2 or claim 4.

3C. The remaining references in US3913973's citation list (all post-1903)

These are citations made by Mintz, not by Stone. Listed for completeness; none can be §102 art against 816616 because all postdate it by 15–83 years:

(US 403,850 A and the two Stone patents from the same list are treated in 3A/3B above.)

3D. Forward citations — patents that cite US816616 (not prior art)

Per Google Patents "Cited By," and independently confirmed for the Tippit family (its "Referenced Cited" list shows "786693 — April 1905 — Stone" and "816616 — April 1906 — Stone"):

§102 status: zero. These postdate US816616 by 69–115 years. They are evidence of the field (billiards-table/sofa conversion and leveling), not prior art.

3E. "Similar Documents" — ⚠️ not citations

Google's "Similar Documents" list is machine-generated topical similarity, not an examiner or applicant citation. I include it only because most entries are pre-1903 and are therefore candidate §102 art you may wish to pull. I have not read any of these disclosures.

Published before 816616's 1903-09-16 filing date (eligible candidates):

Published after the filing date (ineligible as §102(a)-type art): US 781,999 A (1905-02-07); US 799,702 A (1905-09-19); US 807,394 A (1905-12-12); US 857,315 A (1907-06-18); US 921,466 A (1909-05-11); US 1,040,106 A (1912-10-01); US 1,077,189 A (1913-10-28); US 1,102,147 A (1914-06-30); US 1,139,183 A (1915-05-11); US 1,180,926 A (1916-04-25); US 1,222,673 A (1917-04-17); US 1,256,325 A (1918-02-12); US 1,286,094 A (1918-11-26); US 1,306,268 A / 1,306,475 A (1919-06-10); US 1,323,563 A (1919-12-02); US 1,349,162 A (1920-08-10); US 1,408,013 A (1922-02-28); US 1,850,262 A (1932-03-22).

---

4. Statute note — the §102 framing is anachronistic

US816616 was filed 1903-09-16. The governing novelty statute was R.S. §4886 (the pre-1952 Act), not 35 U.S.C. §102. Two consequences:

• §103 did not exist until 1952. Pre-1952 practice folded much of what we now call obviousness into "anticipation" — the pre-1952 cases "often used the term 'anticipation' to mean that the subject matter of the claims either was found exactly in prior art … or, though different, was not 'inventive' over the prior art" (Univ. of Miami L. Rev. 56:727, 748, citing Lewmar Marine v. Barient, 827 F.2d 744, 747 (Fed. Cir. 1987)).
• §102(e) (patents/applications as prior art as of their filing date) did not exist either; the closest analogue was the Milburn line of authority.

So the §102 labels I apply below are modern-equivalent screening tags only. They are not how this patent would actually have been examined, and any real invalidity analysis has to be run under R.S. §4886 with its date-of-invention critical date.

---

5. What a §102 mapping would actually require (framework, not conclusions)

Because there is nothing cited to map, here is the element checklist against which any candidate (US 403,850 A; the pre-1903 "Similar Documents" candidates) must be measured. A single reference must disclose all elements of the claim:

Candid statement: I cannot assign any claim to any reference. Doing so requires reading US403850A and the pre-1903 candidates, and I was unable to retrieve US403850A's disclosure — the search terminated before returning it. I am recording that as a gap, not filling it by inference.

---

6. Cross-reference to the earlier section — the serial-number ambiguity, refined (and the one place it matters for prior art)

The earlier section flagged the serial number as unresolved (275,979 vs. 276,979 in OCR) and asked for it to be treated as uncertain. Live search lets me refine that, and I flag it as a refinement rather than a contradiction:

Two serial-number/date anchors retrieved:

That is ~302,351 serial numbers across ~5.13 years, ≈ 59,000 serials/year. Extrapolating backwards from 230,399 to a September 1903 filing gives ≈ 165,000 — nowhere near 275,979. Extrapolating forward to August 1905 gives ≈ 279,000 — which is where 275,979/276,979 sits.

Inference (well-supported, but an inference): the serial printed on US816616 belongs to the renewed application of August 26, 1905, not to the original September 16, 1903 filing. The "275,979 vs. 276,979" discrepancy and Google's US27597903A label are both artifacts of that; neither figure is the 1903 serial, which does not appear to be printed anywhere on the document.

Why this matters for the prior-art question — and it is the one genuinely interesting wrinkle here: the operative prior-art date for US816616 is the date of invention, and the printed specification fixes a filing date of Sept 16, 1903 with a renewal on Aug 26, 1905. If the effective date were the 1905 renewal, then US 786,693 A — granted 1905-04-04, four months before the renewal — becomes a candidate "patented … in a printed publication before [the] application" reference, notwithstanding that it is the same inventor's own patent (R.S. §4886's patent/printed-publication clause carried no "by others" qualifier, and modern §102(a)(1) likewise has none). If instead the original 1903 date is the effective date, US786693 drops out entirely.

I cannot resolve which date controls without the file wrapper. This is a live open question, not a conclusion, and it is the sole scenario in which anything in the surrounding reference universe becomes prior art against US816616.

---

7. Confidence and limitations

• High confidence that US816616 has no printed or digitized patent citations. The absence of a "Patent Citations" block is directly observable in the authoritative full text and is confirmed by the presence of such a block on the comparison document US3913973. The specification contains no citations.
• High confidence in the bibliographic data for US786693A, US3913973, US7621818, US828668, US11065169 and the US3913973 citation list, which came from live retrieval.
• I did not retrieve the disclosure of US403,850 A. My second targeted search was cut off by a step limit. I have therefore made no §102 mapping for it and no mapping for any "Similar Documents" entry. Do not treat the candidate tables in §3E as substantiated art.
• I did not verify whether US786693A or US403850A carry their own front-page citation blocks, nor did I retrieve the 1903–1906 file wrapper.
• The 1947 date for the origin of the printed citation block comes from a secondary, non-US source and is offered as corroboration only.
• Google Patents' "Similar Documents" output is topical similarity, not citation. It is labeled as such above so it is not mistaken for the citation list the request assumed exists.

---

8. Recommended next steps to actually obtain the prior art

1. Order the paper file wrapper for application Serial No. 275,979 (per Google) / 276,979 (per printed OCR) from the National Archives / USPTO — this is the only place an 1903–1906 examiner's "Notice of References" (or its era-equivalent) could exist, and it would also settle the filing-date question in §6.
2. Pull the disclosures of the six pre-1903 candidates in §3E and of US 403,850 A, then run the §5 element checklist against each. Start with US403850A (only cluster reference with pre-1903 eligibility) and US709333A (published exactly one year before the filing).
3. If you are assessing validity for a modern purpose (e.g., treating US816616 as prior art against someone else's patent), the direction reverses: US816616 is the reference, it is cited as such by US3913973, US7621818, US7828668 and US11065169, and its own citation list is irrelevant to that use.
4. Do not build a §102 chart off §3B–§3E without reading the primary documents. Everything in those tables is either a later citation, a machine-similarity hit, or an unread candidate.

I'll ground this § 103 analysis in actual reference texts rather than assume contents. First, some targeted searches.

The searches surfaced one potentially important document — US786693, which appears to be a sibling patent in the same family line. Let me verify that and one or two pre-1903 references.

Obviousness Analysis — U.S. Patent No. 816,616 under 35 U.S.C. § 103

Analysis date (per instruction): April 26, 2026 · Document analyzed: US 816,616 A, Convertible Sofa and Table, N. B. Stone, filed 1903‑09‑16 (renewed 08‑26‑1905), granted 1906‑04‑03, expired 1923‑04‑03.

---

0. Two framing corrections before the analysis begins

(a) There is no backward-citation "Prior Art" section on the page. The Google Patents/US816616 record as supplied contains three prior‑art‑adjacent blocks, and they are not the same thing:

Note also that the page reports forward citations twice with inconsistent counts ("Cited By (3)" above a four‑row table, then "Cited By (4)" with the same four rows). That is a page artifact; it does not affect the substance.

(b) The date instruction contradicts the session date. The session header reads 2026‑09‑27; the task prompt reads April 26, 2026. I have flagged rather than silently reconciled this. It changes nothing substantive — both post‑date the patent's 1923 expiration by a century.

---

1. Effective filing date and the date filter that governs everything

• Effective date for § 102/§ 103 purposes: 1903‑09‑16 (Google Patents priority date; matches the printed patent's "Application filed September 16, 1903").
• Pre‑AIA law governs. The patent was examined and granted under the pre‑1952 "invention" standard (Hotchkiss v. Greenwood, 52 U.S. 248 (1850)). Section 103 did not exist until the Patent Act of 1952. A modern § 103 framing is therefore an analytical overlay: a 1906‑era court would have asked whether the improvement exhibited "invention" beyond that of "an ordinary mechanic," which is a stricter standard in some respects and a looser one in others than Graham/KSR. I flag this because it materially affects outcome and is routinely glossed over.
• Renewal caveat. The patent prints both the 1903 filing and the 1905 renewal. I am not confident about the pre‑1952 renewal doctrine's effect on the effective date for prior‑art purposes (R.S. § 4894 and its 1897 amendment). If the operative date were 1905‑08‑26 rather than 1903‑09‑16, an entire layer of intervening art would become available — including references listed on this very page. I cannot resolve this from the record supplied, and I decline to guess. This is the single largest uncertainty in the analysis.

---

2. The citable prior‑art universe from this page (date‑filtered)

Applying the 1903‑09‑16 cutoff to the "Similar Documents" list:

Only one of the six citable references — US 430,442 — has a verified disclosure in my hands. Every other pre‑1903 entry is before me as a title and a date. I will not invent what they contain.

An important non‑art document I found (flagged, not used)

US 786,693 A, Convertible Sofa and Table, N. B. Stone, filed 1904‑10‑28, granted 1905‑04‑04 — the same inventor and same title as the patent under analysis. It claims a track pivotally supported on the base, a cam‑carrying shaft, elevating rods, rollers on the table member, and a table that "rest[s] upon the supporting‑arms when disposed over the base member with the tracks depressed."

It appears in neither the Cited By nor the Similar Documents list on this page. Two points:

1. It is not § 103 art against 816,616 — filed over a year after the 1903 filing date. (Unless the renewal issue in §1 resolves the other way, in which case it becomes a serious problem for claims 1–2.)
2. It is highly probative of what the inventor himself regarded as the alternative solution two years later, having abandoned the rock‑shaft‑and‑links approach for a track‑and‑cam approach. That cuts against nonobviousness of the linkage as a design choice, though not against claim 1's specific claimed geometry.

---

3. Person having ordinary skill in the art (PHOSITA), September 1903

A mechanic or cabinet‑furniture maker with several years' experience building combination furniture and billiard‑table cabinets. Familiar with: casters; knurled‑wheel screw levelers; cast‑iron angle brackets and slotted guides; two‑bar and parallelogram linkages; single‑shaft ganging of paired linkages (a standard carriage, farm‑implement and railway‑appliance expedient of the era); cam‑and‑roller elevating mechanisms; and spring‑clip and latch fastenings for detachable cushions and backs. No engineering degree. Notably, this PHOSITA was more fluent in mechanical-linkage geometry than a modern "furniture designer," which raises the obviousness risk for every linkage claim.

---

4. Element mapping and the strongest combination grounds

Claim 1 — base + table swingable between upright‑behind and horizontal‑on‑arms + "intermediate device" pivoted below the arms and to the table's central portion, table‑side pivot higher and at the rear upper corners of the arms when rearmost

Closest verified art: US 430,442. Its verified text discloses:

• a large panel D carrying "one portion of the upholstered seat D′" — a seat member;
• "two arms C C′, which extend from one side of the panel to the other and project from the back… These arms not only act as supports";
• "the head portion E … is connected to the panel D by two bars H H, which are connected together by rods h h′";
• "The bars H are recessed at i² for the reception of the supporting‑bars C"; and
• "the upper end of said foot portion F being supported by a bar K, which is pulled out from the desk," with F used "for the back of the chair."

What maps: seat member; arms functioning as supports; paired bars (links) connecting a movable section to the base; a member that serves as a back in one configuration; even tied‑together bars for rigidity.
What is missing from every verified reference: the specific geometry — a base‑side pivot below the arms, a table‑side pivot on the longitudinal central portion of the table, the table‑side pivot higher than the base‑side pivot, and that pivot sitting at the rear upper corners of the arms when rearmost, so the table can be swung over the arms before the intermediate device moves.

Motivation to combine (KSR rationales, as an overlay):

• Same field, same problem: US 430,442 and US 601,727/U.S. 673,313 are all convertible furniture in which a sitting article and a work surface exchange roles — the prototypical "combination of familiar elements according to known methods" yielding a predictable result.
• Known need: the specification itself concedes the design constraint — a billiard table is heavy, and "the change … can be readily accomplished without the necessity of manually lifting any of the heavy members." An articulated rather than hinged table is the predictable answer to that stated need.
• Design incentive: substituting a constrained link for a simple hinge is the standard mechanical answer to a heavy lid; a 1903 PHOSITA would not have needed teaching to reach for it.

Where this ground breaks down: none of the available references supplies the claimed pivot ordering and location, and the claim's point of novelty is precisely that ordering. The two‑stage motion — free pivot first, constrained translation second — is what the specification says produces balance. Under Graham factor 3 (level of ordinary skill) the linkage itself is easy; under factor 2 (differences) the geometry is genuinely specific. Claim 1 is the most defensible claim in the set.

Claim 2 — rock‑shaft journaled on the seat member beneath the seat and between the arms, links rigidly attached to the rock‑shaft, table‑side pivots at the rear upper arm corners when rearmost, swinging "independently of the movement of the links"

This is the equalization claim. The specification supplies the admission that makes it vulnerable:

"in a billiard‑table the top must be comparatively heavy, and with independently‑movable links there is a tendency for one end to move in advance of the other and bind, while in this structure the movement is entirely equalized."

Combination ground: US 430,442 (paired bars H H joined by rods h h′ for tying the assembly together) + the ordinary mechanical knowledge of ganging two parallel linkages on a single common shaft. The rationale is textbook KSR: known technique to improve a similar device in the same way + predictable result. Two paired links that must move together are, mechanically, the reason tie rods and common shafts exist.

But I must flag the gap honestly: US 430,442's rods h h′ tie the bars to each other; they do not teach a rock‑shaft journaled to the base with links rigidly affixed to it. I have no verified pre‑1903 reference on this page that does. A real § 103 ground for claim 2 requires a reference or an accepted‑knowledge showing for shaft‑ganged linkages in furniture; the page does not supply it. The inventor's own admission of the binding problem is strong evidence of motivation, and strong evidence that the problem was known — but motivation alone is not a ground; the solution must also be shown or shown to be obvious. Claim 2 is vulnerable but not cleanly anticipated or shown obvious on this record.

Claims 3, 4 and 5 — the removable back

• Claim 3: seat member + table swingable behind and restable on it + spaced links connecting them + removable back with detachable connections to the links, the back covering the links when installed.
• Claim 4: adds upstanding arms, links between the arms, back fitting in advance of the links, secured by spring‑clips.
• Claim 5: adds back resting against the links, the links constituting supports for the back.

Combination ground: US 430,442 (a panel/bar sub‑assembly that is simultaneously structural and concealed within, or forms part of, the back configuration — "the upper end of said foot portion F being supported by a bar K") + the ordinary cabinetmaker's practice of a detachable, upholstered back panel with clip or latch fastening. The motivation is a straightforward space‑utilization and appearance rationale: the linkage occupies the exact volume the sofa back must fill, so mounting the back on the linkage conceals the mechanism and requires no additional frame — a recognized and predictable integration.

These are the weakest claims. Claims 4 and 5 recite what are essentially fastening selections (spring‑clips) and orientation choices (back in advance of the links; links as supports) that a PHOSITA would arrive at by routine design iteration. Under KSR's "predictable variation" and "obvious to try" rationales, claims 3–5 would likely fall — provided the challenger can show a pre‑1903 detachable upholstered back with clip fastening. I have not verified such a reference on this page. US 975,374 (Shalansky, Convertible Chair, filed 1909‑12‑13) shows a knob‑and‑threaded‑rod immobilizing device in a convertible chair, and US 3,191,194 shows over‑center pivot brackets in a sofa/bunk — both are post‑dated and therefore not available, though they are useful evidence of what the art eventually regarded as routine.

---

5. Evidence the combination would have been obvious — and why it is legally unusable

Two references surfaced in searching that are strikingly on point but post‑date 816,616 by decades, so they are not prior art and cannot support a § 103 ground. I list them only to be candid about why they must be excluded, since an unwary analyst would seize on them:

• US 4,194,452 (1974) — convertible table with two pairs of parallel links and a tie rod that "maintains corresponding links … at opposite ends of the top in alignment, synchronizing movement of such links to prevent swinging of one link relative to the opposite link," thereby deterring "sidesway and twisting of top 5 relative to the table base." That is, functionally, claim 2's equalization benefit — seventy years later.
• US 3,913,973 (Mintz, 1975), Combination pool table and couch — the nearest modern analogue of the whole concept, and (per the prior section) itself a forward citation of US 816,616.

The correct inference from these is the reverse of a § 103 ground: they demonstrate that 816,616 is treating‑art for later work, which is why Mintz, Tippit and Sinclair cite it. Do not invert that relationship.

---

6. Conclusion

Bottom line: claims 3–5 are materially vulnerable to a § 103 challenge; claims 1–2 are not, on the prior art this page actually supplies. The limitation is evidentiary, not analytical: the page gives me six citable references and verifies the text of one (US 430,442). A genuine invalidity opinion would require full texts of US 140,868, US 328,894, US 601,727, US 673,313 and US 709,333, plus a search of the 1880–1903 furniture and billiard‑cabinet art for (i) shaft‑ganged paired linkages and (ii) clip‑fastened detachable upholstered backs.

And the analysis is academic. Per the prior sections of this file, US 816,616 expired 1923‑04‑03, has no PTAB record, and has never been litigated. Section 103 exposure is of interest only for (a) citation‑defensive purposes, since this patent is live prior art against later claims, and (b) historical interest.

---

7. Confidence and contradiction flags

• Flagged contradiction (dates): session date 2026‑09‑27 vs. prompt date 2026‑04‑26. Not reconciled; immaterial here.
• Flagged premise error (task): the instruction to "use the results from the Prior Art section of this page" cannot be followed literally — the page contains no backward‑citation prior‑art table, only a keyword tag cloud, forward citations, and a mixed‑date "Similar Documents" list. I applied the date filter myself and labelled which entries qualify.
• Flagged omission: US 786,693 A (Stone, same title, filed 1904‑10‑28, granted 1905‑04‑04) is absent from this page's citation lists but is the same inventor's successor solution to the same problem. It is not § 103 art, but its existence is relevant to the nonobviousness posture and to any priority/interference question.
• Flagged uncertainty (material): whether the 1905 renewal preserved the 1903 effective date. I am not confident on the pre‑1952 renewal doctrine. If the effective date is 1905‑08‑26, US 786,693 and US 781,999 enter the prior art and claims 1–2 become substantially more exposed.
• Confidence: High that claims 3–5 are the obviousness‑vulnerable ones and that no verified reference on this page reaches claim 1's geometry. Low on any conclusion for claims 1–2 as an actual invalidity determination, because the necessary reference texts are not in hand. I have not fabricated the disclosures of the five unretrieved references, and I have not asserted a § 103 ground I cannot source.