US 632572

I'll run targeted searches on the literal patent number 632572 across USPTO and CAFC-related sources.

Initial searches returned nothing on point. Let me try more specific queries.

Let me verify the assignee/ownership record and check the CAFC docket angle once more.

US 632,572 — Search Results and Summary

Scope note on the literal identifier

Per your instruction I interpreted "632572" literally, and it resolves to a single U.S. patent grant: US632572A. I did not auto-correct it to any seven-digit modern number (e.g., US 6,325,726) or to any other patent whose number merely ends in "572." If you in fact intended a different document, please confirm — but nothing in the searches indicates that 632572 maps to anything other than the 1899 grant below.

One caution: hits mentioning a "'572 patent" exist in unrelated litigation (e.g., Markem Corp. v. Zipher, Ltd., D.N.H., construing a "'572 patent"; Robert Bosch LLC v. Pylon Mfg. Corp., 632 F. Supp. 2d 362 (D. Del. 2009)). Those are a different patent and a volume/page citation respectively, not US 632,572. I did not conflate them.

Bibliographic data (from the granted document)

Abstract

There is no abstract. The abstract as a statutory requirement did not exist in 1899 (mandatory abstracts began with the 1952 Patent Act). The "Definitions" paragraph shown on the Google Patents page is an auto-generated summary, not part of the original grant. The document instead opens with a "To all whom it may concern" specification preamble.

Object of the invention (from the specification)

1. A centrifugal pump, chiefly for hydraulic dredging, drawing water at both sides of the disk at equal flow and force so pressures are automatically balanced, maximizing induction and reducing wear/obstruction from debris.
2. Adjustable disk/wing mounting to take up wear and keep the blade edges close to the chamber walls.
3. Automatic priming — venting air from the suction pipe on startup without external primer, jet pump, or strainer; the water head below the main valve is lifted to fill the suction pipe.
4. A main valve assembly removable through an opening in the valve chamber without disturbing the suction pipe, all working parts adjustable above water and in sight.

Plain-language overview of the claims

The patent carries 10 claims; all are drafted as independent "combination" claims (typical of the era — there is no dependent-claim structure). Note that the OCR of the granted text is broken in the middle of claims 7–8, so those two are partly reconstructed from context.

Prior-art keywords / cited-by

Google Patents lists prior-art keywords: valve, pump, disk, chamber, water. Documents citing US632572A (all much later, no bearing on validity): US4247250A (Allis-Chalmers, fabricated pump casing), US5762551A (freeze-free vent), WO2005035979A3 / US6942448B1 (Fast Flow LLC pump), US10138891B2 and US11105333B2 (Pemberton Patents LLC double-suction pumps). None of these appear in the original 1899 examination record.

CAFC 2026 docket

No CAFC 2026 docket or appeal referencing patent 632572 was found in the searches I ran. That is consistent with the record: the patent expired September 5, 1916, and it has no recorded assignment, no published litigation, and no inter partes review (an IPR would be impossible — the patent predates the AIA and has been expired for over a century). Searches for "632572" combined with "CAFC," "Federal Circuit," and "2026 docket" returned only unrelated matters.

Uncertainty / caveats

• No abstract exists; any "abstract" for this patent would be a modern reconstruction, not the original text.
• Claims 7 and 8 are partially garbled in the digitized document (a mid-claim OCR/transcription break). My plain-language reading of them is an interpretation, not verbatim text.
• The description text in the source has numerous OCR errors ("hearing" for bearings, "ind notion" for induction, "month" for mouth, "tho" for the), which I have not silently corrected in quoted material.
• I could not access PACER/CM-ECF or the CAFC's own docket system directly; the "no 2026 docket" conclusion rests on web-indexed sources, which is strong but not a substitute for an official docket query. Given the 1916 expiration, I have high confidence no live 2026 proceeding exists.

I'll search for litigation involving US patent 632572.

Let me try more targeted searches on litigation databases.

Let me check additional litigation-specific sources.

Litigation Involving U.S. Patent No. 632,572

Result: No litigation is known to exist involving U.S. Patent No. 632,572.

I searched for litigation records tied specifically to this patent number, including general litigation queries, Unified Patents' litigation portal, CourtListener/RECAP, and searches pairing the patent number with its inventor (Edmund M. Ivens) and subject matter ("centrifugal pump"). Every hit returned was a false positive — unrelated patents (e.g., U.S. 6,625,858; U.S. 6,292,974; provisional application 60/632,572; U.S. 4,591,191), or instances where US632572 appears only as a cited prior-art reference in the prosecution of later patents (e.g., U.S. 6,942,448 and U.S. 2019/0093660 A1 both cite it as background art). No case caption, docket number, or party involving US632572 appeared in any source.

Why no litigation record exists (contextual reasons)

1. The patent is long expired. Per the authoritative Google Patents record, US632572 was filed December 18, 1897, granted September 5, 1899, and shows an anticipated expiration of September 5, 1916 ("Expired – Lifetime"). A patent that lapsed over a century ago cannot be asserted in any live infringement suit.
2. It predates modern litigation infrastructure. The Federal Circuit was not created until 1982; PACER was not deployed until the late 1980s/1990s; CourtListener, Unified Patents, and comparable databases index essentially nothing from the 1899–1916 era. Any infringement action that might theoretically have been filed during the patent's ~17-year term would, if it existed, be recorded only in historical print reporters and would not appear in these systems.
3. No current assignee asserts it. The Google Patents record lists the assignee as "Individual," with no assignment history and no party currently holding enforceable rights.

Important caveats

• I cannot rule out the possibility of an undisclosed historical infringement action from the 1899–1916 period that never entered electronic databases. I found no evidence of one, but absence of an electronic record is not affirmative proof that none was ever filed. Confirming that would require manual review of historical federal court records (e.g., the relevant circuit/docket books for the Eastern District of Louisiana), which the sources I searched do not index.
• I intentionally did not report litigation for similar-looking numbers. Matching by pattern alone would be an error — e.g., "632572" appears as a provisional application number (60/632,572) in EP 1 825 261 B1, which is not U.S. Patent 632,572 and is not litigation. Per your instruction, I am treating the identifier strictly literally.

Summary Table

Conclusion: There is no known litigation involving U.S. Patent No. 632,572 ("Centrifugal pump," Edmund M. Ivens, granted Sept. 5, 1899). The patent expired in 1916, and no docketed case — historical or current — is discoverable in the litigation databases searched.

I'll verify the structured data against public sources before writing this up.

Let me run a few more targeted checks before concluding.

One more check on the modern patents that cite this one, in case it surfaces as art in a live proceeding.

PTAB Proceedings — U.S. Patent No. 632,572

Note on scope: The canonical list in the structured "PTAB proceedings on file" block is empty — the USPTO ODP API returns no AIA trial proceedings for this patent. I ran targeted web searches against PTAB/PTACTS material, PTAB decision sources, and Federal Circuit docket material for proceeding numbers, petitioner names (Ivens, and the modern assignee/family names), and the patent number in both 632,572 and 632572 forms. Nothing surfaced. The per-proceeding template below is therefore omitted rather than filled with fabricated docket numbers — there is nothing to report.

Proceedings overview

Total PTAB proceedings: 0. No IPRs, no PGRs, no CBMs, no derivation proceedings, and no Federal Circuit appeals of anything PTAB-related are on file for US 632,572 as of 2026-10-01 — and that number is almost certainly permanent. This is not a "hardened patent survived two IPRs" signal and it is not a "claims 1–5 canceled, troll has no case" signal. It is a dead-patent signal: US 632,572 issued 1899-09-05, was filed 1897-12-18, and its term expired 1916-09-05 (per the structured data's "Anticipated expiration" entry). A defendant does not need a PTAB strategy here — the patent left the enforceable universe 110 years ago.

No proceedings to detail

There is no proceeding number, no petitioner, no panel, no institution decision, no Final Written Decision, and no appeal to report. I verified this rather than assumed it:

• The structured ODP block is authoritative and lists no trials.
• Web searches for 632572 IPR, "US632572" PTAB, and "632572" ptacts.uspto.gov returned no petition, trial, or decision referencing this patent as the challenged patent.

Why there is no activity (and why that will not change):

• Expiration. The patent expired 1916-09-05. An expired patent cannot be asserted in an infringement action, so no petitioner has the practical trigger for an IPR — being accused of infringing it. (IPR is technically available against an expired patent under 35 U.S.C. § 311(a), which contains no live-patent requirement, but that is a theoretical point with no fee-paying petitioner.)
• PGR is categorically unavailable. Post-grant review under § 321 reaches only patents subject to the first-inventor-to-file provisions. An 1897 application is far outside that window.
• CBM is inapplicable. A covered business method review requires a claim directed to a financial-services business method plus a prior infringement suit. A centrifugal pump for hydraulic dredging satisfies neither.
• The patent's only modern relevance is as prior art. The Google Patents page shows it is cited by later patents in the Pemberton Patents LLC double-suction pump family (US 10,138,891; US 11,105,333), which claim 2015–2016 priority. That makes the '572 disclosure relevant to other people's validity fights — it is not a proceeding on the '572 itself, and I found no PTAB petition in which it was the challenged patent.

Strategic summary

Claim status: all claims UNTESTED by the PTAB — and all claims expired. The patent carries 10 claims (claims 1 and 2 directed to the adjustable two-section disk in a diametrically-opposed-inlet chamber; claim 3 to the set-screw adjustment; claim 4 to the automatic priming devices; claim 5 to the eccentrically-journaled disk; claim 6 to the dual key-slots on the shaft; claim 8 to the wedge-portion water-column splitter with the air-operated supplemental valve; claim 9 to independent keys; claim 10 to the valve-inserting opening and removable valve-carrying member — note the OCR of the printed claims in the source is partly garbled, so treat this mapping as approximate). None of that matters operationally. No IPR has ever invalidated a claim of this patent, but equally, no claim has been sustained, and the absence of PTAB activity tells you nothing about claim strength. The controlling fact is that US 632,572 has no remaining term.

Estoppel landscape: none, and none will arise. Section 315(e)(2) estoppel attaches only to a petitioner that obtained a final written decision on a claim. With zero proceedings, there is no estoppel running against anyone, and no ground is foreclosed. If you somehow needed to challenge this patent, every § 102/§ 103 ground based on patents and printed publications would remain available — but an IPR on an expired patent is an exercise in spending money to change the labels on a public-domain reference.

Pattern signals: none. No repeat petitioners (there are no petitioners), no PTAB appeals by the owner (there is no owner of record — the structured data lists "Individual" with the inventor Edmund M. Ivens having long since passed into history), and no defensive aggregator such as Unified Patents in the chain. There is no assertion campaign to defend against. The "Cited By" trail into the Pemberton patents is the only live thread, and it flows outward from the '572 (as prior art), not inward.

Recommended next steps

1. If you received a demand letter citing "US 632,572" — do not file an IPR. The correct response is a short letter pointing out that the patent issued 1899-09-05 and expired 1916-09-05, that it is public domain, and that any demand premised on it is meritless. Demand letter link it to the Google Patents record showing the 1916-09-05 anticipated-expiration entry: https://patents.google.com/patent/[US632572](/patent/US632572)/en.
2. Verify the number before you spend anything. A seven-digit transposition (e.g., a "6,325,572"-style number) points to a completely different, modern patent that may well have live claims and real PTAB exposure. If the demand letter's number has six digits, it is this 1899 patent; if it has seven, you are in a different dispute. Do not let a plausible-looking citation to a 19th-century patent substitute for checking the actual asserted number.
3. No PTAB milestones to track. There are no institution deadlines, no oral hearings, and no statutory one-year FWD dates, because there is no trial. Any calendaring around this patent is wasted effort.
4. Treat the "no PTAB activity" signal correctly. Ordinarily, an empty PTAB docket on an actively asserted patent is a meaningful — sometimes suspicious — data point, because well-asserted patents attract IPRs. Here the emptiness is fully explained by expiration, not by the patent being robust. Do not read this as "hardened."

Bottom line for a defendant: there is no PTAB story on US 632,572. The patent has been in the public domain since 1916-09-05, and its only contemporary significance is as a cited prior-art reference against modern double-suction dredge pump patents. If it appears in a demand letter, the issue is the letter's credibility, not the patent's validity.

I'll verify the assignment record before writing, checking both the USPTO Assignment Center index and general sources for any recorded conveyance of this patent.

Inventors

• The grant names a single inventor. The specification preamble reads: "Be it known that I, EDMUND M. IVENS, of New Orleans, in the parish of Orleans and State of Louisiana, have invented a new and Improved Centrifugal Pump…" — there is no "assignor to" recital, no joint inventor, and no corporate party anywhere in the front matter or body text.
• Employer/secondment pattern: not present. With a single named inventor and no corporate recital, the standard "all inventors departed the assignee within 12 months" pre-fire-sale tell is structurally inapplicable — there was never an assignee to depart from.
• Context worth flagging for the record (but not evidence of anything): the stated field of use is hydraulic dredging, and the contemporaneous pump trade in New Orleans was active — the Menge Pump Company of New Orleans was advertising centrifugal-type pumps to Louisiana plantations in the 1890s (see the Library of Congress HABS narrative for Laurel Valley Plantation, https://cdn.loc.gov/master/pnp/habshaer/la/la0100/la0197/data/la0197data.pdf). I found nothing tying Ivens to Menge or to any other manufacturer, and I am not inferring one.

Original assignee

No assignee. The patent issued to the inventor personally.

• Google Patents records Original Assignee: Individual and Current Assignee: Individual, with no assignment history field populated. The "Assignee" column that shows "Edmund M Ivens" in third-party citation tables (e.g., the citation list on https://patents.google.com/patent/[US6942448](/patent/US6942448)) is a rendering of "no corporate assignee," not a recorded conveyance.
• Product embodying the claims: none documented. The specification describes a hydraulic-dredging pump but discloses no manufacturer, no licensee, and no commercial embodiment. Whether Ivens or anyone else actually built the FIG. 1 machine is not answerable from the patent record.
• Primary line of business: unknown for the individual inventor. There is no corporate entity whose status could be "operating / acquired / dissolved / in bankruptcy."
• Current status: not applicable — there is no assignee entity. Any rights that ever existed expired 1916-09-05 (the anticipated-expiration legal event on the Google Patents record), and the patent has been public-domain subject matter for over a century.

Assignment timeline

The USPTO Patent Assignment Search / Assignment Center has no records for this patent.

Nothing is recorded against US 632,572 — no Assignment, no Security Agreement, no Merger, no Change of Name, no License, no Release, no Correction. Consequences of that finding:

• No reel/frame exists to cite. I am not supplying a reel/frame, a conveyance type, an assignor/assignee pair, or a correspondent, because inventing any of those would be fabrication. The "Correspondent" analysis you asked for — the repeat-player attorney that is the usual tell on anonymous assignee LLCs — is unobtainable here for the simple reason that no recording was ever filed, by anyone.
• No chain, therefore no "current owner" other than the public. The Google Patents legal-events block shows only the intrinsic prosecution lifecycle events: filed 1897-12-18, granted 1899-09-05, anticipated expiration 1916-09-05, status "Expired – Lifetime." There is no post-issuance transfer of any kind on the face of the record.
• This is a coherent historical result, not a data gap. Recorded-assignment practice in the 1897–1899 period was far less routine than today, and the overwhelming majority of individual-inventor grants of that era — especially outside large corporate research programs like Bell, Westinghouse, or GE — were never assigned at all. An unassigned individual grant is the expected state, not an anomaly.

Per the conditional instruction, I am stopping here. The Assignment Center has no records, so the Timeline diagram, NPE/troll-pattern signals (1–8), and the Verdict sections are omitted rather than filled with content that would have to be inferred from a non-existent chain. For completeness on the verdict question: with no recorded assignments and no assignee entity, the only defensible verdict would be Insufficient data — but the more accurate statement is that the category does not apply, since the patent has no owner-side history to analyze.

Verification links:

• USPTO Assignment Center: https://assignmentcenter.uspto.gov/
• USPTO Patent Assignment Search (legacy index): https://assignment.uspto.gov/patent/index.html
• Google Patents record with the full legal-events block: https://patents.google.com/patent/[US632572](/patent/US632572)/en

Consistency check against the earlier sections: this result is fully consistent with the prior Patent summary ("Assignee: None recorded — patent text names only the inventor") and with the Litigation summary (no assignee of record to assert the patent). No contradiction to flag. One practical note carried forward from the prior work: because there is no corporate assignee chain, none of the NPE-family cross-references (Acacia, Marathon, IV, Wi-LAN/Conversant, Unified, RPX, etc.) can attach to this patent — a shell-entity or cascading-transfer signal requires at least one recorded conveyance, and there are zero.

I'll verify the citation record for the literal number 632572 before analyzing prior art.

Confirming the citation direction and checking the pre-1897 candidates from the "Similar Documents" list.

Threshold finding: US 632,572 has no prior-art citation record

Before mapping references claim-by-claim, one determination drives the entire analysis: there is nothing in the record of US 632,572 that qualifies as a citation of prior art.

Google Patents' page for the literal number 632572 (https://patents.google.com/patent/[US632572](/patent/US632572)/en) contains only three citation-type tables, and none of them is a prior-art list:

Antecedent U.S. grants from 1899 were not printed with a front-page reference list, and the authoritative full text you supplied contains no examiner citation, no "references cited," and no prior-art discussion. So the honest answer to "look at each patent citation for 632572" is: the only patent citations attached to this number are forward citations, and forward citations cannot be prior art. I set that out explicitly below rather than manufacturing a § 102 mapping onto references that do not exist in the record.

Part 1 — The forward citations (each one, with § 102 analysis)

These are the six documents Google Patents records as citing US632572. Every one post-dates both the 1897-12-18 filing and the 1899-09-05 grant, so none can anticipate any claim of '572 under any subsection of § 102. They are relevant only in the reverse direction: '572 is potential § 102(b)/§ 102(a)(1) art against them.

Additional forward citation found in search but not in the structured list: US 2019/0093660 A1 (published 2019-03-28; Pemberton family) lists US632572A in its "Family Cites Families" table. Same conclusion — forward citation, not prior art. Source: https://patents.google.com/patent/US20190093660A1/en

Note on the 5-vs-6 discrepancy: the authoritative text prints both a five-row "Cited By (5)" table and a six-row "Cited By (6)" table; the difference is US6942448B1, present in the second. This looks like a Google Patents rendering artifact. It does not change the analysis, but flagging it per your instruction to surface contradictions.

Part 2 — The technically relevant candidates (not of record)

Since the record supplies no cited art, the only defensible thing I can do is identify pre-1897 documents that (a) predate the '572 and (b) are topically capable of bearing on its claims. These come from the algorithmic "Similar Documents" list — they were not cited by the examiner and there is no evidence they were considered. I label them candidate art accordingly.

Source-metadata warning: Google Patents lists US351129A under the assignee string "Dale and john fraser," but the document text itself names H. Salomo of Melbourne and identifies the invention as an "Improved Rotary Pump and Blower." The assignee field in the source is unreliable for this record — do not rely on it in any citation.

Part 3 — Framing correction you should carry forward

Two points that materially affect any § 102 assignment here, and that I state explicitly rather than gloss:

1. § 102 did not govern this grant. US 632,572 was filed 1897-12-18 and granted 1899-09-05. Novelty was then measured under Rev. Stat. § 4886 (the Act of 1870, as amended), with the defenses of Rev. Stat. § 4920. The "two-year on-sale/public-use bar," "§ 102(e) prior applications," "§ 102(g) interference," and "§ 119 foreign priority" concepts you would apply to a 1952-Act patent are anachronisms here. A PTO reexamination or modern validity analysis of an 1899 patent would apply the 1952 Act's § 102/§ 103 to the pre-AIA evidence, and pre-AIA § 102(e) does not reach applications that were not published — which matters because the cited-by family (items 4–6 above) asserts against '572, not the reverse.

2. "Patent citations" in a Google Patents table are not automatically prior art. The tables are unlabeled as to direction on that page. Only the forward direction exists for '572. If a demand letter or filing asserts that US 10,138,891 or US 11,105,333 "cites" the '572 as prior art, that is correct in direction but useless to establish anticipation of '572 — it establishes the opposite relationship.

Bottom line

Most relevant prior art for US 632,572: none is identifiable from the patent's own citation record, because the record contains no prior-art citations at all. All six documents Google Patents associates with the number are later documents citing the '572 — chronologically and legally incapable of anticipating its claims under § 102.

Claim-by-claim § 102 determination:

What I could not verify, stated plainly:

• I could not reach US 405,294 A or US 126,729 A content before my search budget was exhausted. Their descriptions above are title-level only. If you need a real anticipation opinion, those two plus US 351,129 A are the only pre-1897 leads worth pulling, and I would need to read them in full.
• I found no recorded examiner citation, no reexamination file, and no third-party submission for this patent. Given the 1916-09-05 expiration, I have high confidence none exists in electronic form — but I cannot prove a negative against paper-only 1899–1916 Patent Office records.
• Consistent with the prior sections: no live enforcement, no PTAB record. I have not repeated that analysis here.

I'll first verify what the "Prior Art" material on this page actually contains, since the obviousness analysis depends entirely on it.

Useful context. Let me pin down the pre-1897 state of the art more precisely, since the obviousness date is December 1897.

Obviousness Analysis — U.S. Patent No. 632,572 (Ivens, "Centrifugal Pump")

Bottom line up front

US 632,572 cannot be analyzed under 35 U.S.C. § 103 as a matter of law, because § 103 did not exist when it was examined. Read as a forensic proxy under the modern Graham/KSR framework, the great majority of the claims would likely fall: every element traces to long-settled 19th-century practice in double-inlet centrifugal pumps, and the pre-1897 literature itself articulates both the problem and the design trade-offs Ivens addressed. The genuinely defensible residue is narrow — the self-contained, automatic priming subcombination of claims 4/8/10, in which a vent valve is carried by and seated in the suction-pipe main valve and operates on the pressure differential without external primer or jet pump. I flag below that I could not verify a pre-1897 reference for that specific subcombination, and a rigorous § 103 challenge would have to find one.

1. Threshold problem: § 103 is anachronistic here

• US 632,572 was filed 1897-12-18 and granted 1899-09-05. 35 U.S.C. § 103 was enacted by the Patent Act of 1952, Pub. L. No. 82-593, 66 Stat. 792, effective 1953 — roughly 54 years after this grant.
• Prior to 1952, patentability rested on R.S. § 4886 (novelty/utility) plus the judicially created "invention" requirement from Hotchkiss v. Greenwood, 52 U.S. 248 (1850) — the difference from the prior art had to exceed what "an ordinary or skillful mechanic" would supply. The legislative history of the 1952 Act states that § 103 codified that pre-existing judge-made condition, not a new one (S. Rep. No. 82-1979; H.R. Rep. No. 82-1923 (1952)); the Supreme Court confirmed this in Graham v. John Deere Co., 383 U.S. 1 (1966).
• Practical consequence: because § 103 codifies Hotchkiss, applying the Graham factors here is a reasonable proxy — but note that in the 1899–1916 window courts were drifting toward the stricter Cuno Engineering "flash of genius" formulation (Cuno Eng'g Corp. v. Automatic Devices Corp., 314 U.S. 84 (1941)). A sitting court of that era would have been more inclined to sustain these claims than a modern § 103 analysis implies. That asymmetry should be stated in any opinion letter.

Date note: the framing of this session gives today as 2026-10-01 in one place and 2026-04-26 in the task header. Nothing turns on it — the patent expired 1916-09-05 and has no enforceable term either way. This is also consistent with the earlier-generated sections of this file.

2. What the page's "Prior Art" section actually supplies — and what it does not

This is the weakest link in the record, and I want to be explicit rather than paper over it. The fetched Google Patents page contains no "Citations" / backward-reference block at all. Its "prior art" material reduces to three things:

Critical caveat: I have not verified the disclosures of US351129A or US405294A. They appear on the page only as similarity hits. I will not assert what they teach. Any invalidity contention built on them requires pulling the full documents first.

Because the page is essentially barren of pre-1897 art, the analysis below leans on pre-1897 printed publications that I located independently (clearly marked), plus the '572's own specification, which is itself an admission of the state of the art. Where I lack a verified pre-1897 reference, I say so.

One useful data point does come from the record: USPTO/Google classified this patent under Y10T137/7837 (direct-response/check valves) and Y10T137/7841 — "One valve carries head and seat for second valve." That classification is an administrative acknowledgment that the '572's claim-4/8/10 subcombination is a recurring, recognized genus in the check-valve art — which supports treating it as a known element, though a classification is not itself prior art.

3. The 1897 person having ordinary skill in the art (PHOSITA)

A competent pump engineer/mechanic at a dredging or drainage works, familiar with: volute and spiral casings; open and enclosed impellers with curved (Appold-type) vanes; clearance adjustment via renewable wearing rings; standard keyed hub-and-shaft assembly; bolted sectional casings; and standard suction-line check/foot valves.

4. Pre-1897 evidentiary base I was able to verify

These are documentation of the art, not patents, and each pre-dates 1897-12-18:

1. "Double Inlet Pumps," The American Engineer and Railroad Journal, September 1894 (Vol. 68). A full trade-press discussion of double-inlet centrifugal pumps as an established, if contested, construction. It enumerates the objections to the double-inlet arrangement: "impossibility of securing easy-entering curves"; "inconvenience of access and expense of fitting"; "the spindles passing through and obstructing the inlet pipes"; "reducing the inlet areas by dividing them into two parts with a sharp ledge to catch obstructions"; and, critically for a dredging application, that "double-inlet pumps are liable to obstruction by title [tule] roots, bean straw, weeds and other kinds of debris." URL: https://upload.wikimedia.org/wikipedia/commons/8/86/American_engineer_and_railroad_journal_%28IA_americanengineer68newy%29.pdf
2. Id. (same volume) — records J.S. Gwynne's 1850/1851 balancing arrangement, and states that "the adoption of open wheels or impellers… called for a double or balanced suction," describing the balance-by-double-suction question as "a cycled experiment running over a period of 40 years."
3. "Pumps" (historical survey) — "The admission of water on both sides of the centrifugal pump, as practised in the Massachusetts pump of 1881… was not Andrews's nor was it Gwynne's… Mr. Lloyd and Mr. Appold had already done so." URL: https://forgottenbooks.org/pt/download/Pumps_10272122.pdf
4. A Catechism of the Steam Engine (Carrett, Marshall & Co. Appold pump description) — "there are two suction pipes, one for each side of the fan." URL: https://upload.wikimedia.org/wikipedia/commons/0/09/…
5. John Richards, Centrifugal Pumps (1894) — index terms include "double inlets," "balancing piston," "cut-off disc," "encased impellers," "Appold pump," "Bessemer," "Easton & Anderson." This establishes both double-inlet practice and cut-water/dividing-plate practice as textbook material before the filing date.
6. Contemporaneous pump patent practice — a mid-1890s-era U.S. grant (No. 495,760) discloses a casing with a "false bottom… in the form of a trap door, so that stones or other obstructions can be removed and wear and tear of the bottom renovated," and uses set screws for adjustment. (Low confidence on the exact grant date — flagged for verification before reliance.) URL: https://patentimages.storage.googleapis.com/d0/5b/e6/a0961820b4d57f/[US495760](/patent/US495760).pdf

5. Claim-by-claim obviousness combinations

Combination A = double-inlet pump art (items 1–5) + casing-access practice (item 6 and routine split/bolted casings).
Combination B = A + clearance-renewal practice (renewable/adjustable impeller-to-casing running clearance).
Combination C = B + keyed hub/shaft practice.
Combination D = A/B + suction-line check valve with air-relief.

Threshold novelty point (worth flagging separately)

Much of Combination A + B is not merely obvious but arguably anticipated: double-inlet casing + rotating disk + two suction pipes, one per side, was public before 1881. Ivens' contribution over that base is the adjustability, not the duplex geometry. Any § 103 opinion should therefore lead with § 102 exposure on claims 1, 5 and 7/8, using § 103 only as the fallback.

6. Where a § 103 challenge is weakest (do not concede these lightly)

1. Claims 4 and 10 — the self-contained automatic primer. The recited cooperation is nontrivial: the vent valve is carried by the main valve, above the normal water head, and the sequence runs on the pressure differential alone (vacuum above the main valve lifts an atmospherically-driven vent plug; the rising water then lifts the main valve). The specification expressly frames the point as dispensing with the jet pump and external primer that "high lift" practice otherwise required, and explains why: "On a high lift the elasticity of the air under the main valve would prevent the main valve opening." That is a stated, non-trivial problem with a stated mechanism. I found no verified pre-1897 reference teaching a vent valve seated in and carried by a suction-pipe check valve. All the priming references surfaced on the page post-date 1897 by decades (US868192A/1907; US2073733A/1937; US2400434A/1946; US2660956A/1953). Absent a pre-1897 hit, a KSR challenge on claims 4/10 rests on "obvious to try" and design-incentive reasoning — defensible but contestable.
2. The combination/functional-claiming style. All ten claims are single-paragraph combination claims closing "substantially as specified." Under 19th-century practice these were read narrowly and, in some cases, as means-plus-function-style language keyed to the specification. That cuts against invalidity — a narrow reading makes it easier to find a difference over the art.
3. Secondary considerations. The specification asserts advantages in a specific, hard environment (hydraulic dredging — the 1894 article confirms this art was plagued by tule roots, weeds and debris). Evidence of long-felt need and commercial uptake in that niche is the classic Graham counterweight, and it was treated as decisive in the 19th-century case law (Smith v. Goodyear Dental Vulcanite Co., 93 U.S. 486 (1876)).
4. Wrong-era standard. As noted in § 1, an 1899–1916 tribunal would have applied a stricter test than § 103.

7. Verification gaps — stated plainly

• The page's "Prior Art" section contains zero backward citations; I have invented none.
• US351129A (1886) and US405294A (1889) are the only plausibly relevant pre-filing patents on the page; their contents are unverified by me.
• I have not verified the grant date of patent 495,760.
• The '572's printed claims 7 and 8 are interleaved/garbled in the digitized text, so the element mapping for those two is a reconstruction. This is consistent with the earlier-generated sections of this file, which flagged the same defect.
• All "Cited By" items are forward citations and are legally incapable of being prior art; anyone treating them as such (e.g., citing US6942448B1 or US11105333B2 against this patent) has the polarity backwards.
• Consistent with the earlier sections: US632572A is confirmed as the sole document matching the literal string "632572," the patent expired 1916-09-05, and it has no live PTAB or CAFC presence. An obviousness opinion on it is therefore academic — except where the patent is cited as prior art against modern double-suction dredge-pump claims (the Pemberton-family citations), in which case its content, not its validity, is what matters.