Scratch Stock

identity

• name: The scratch stock (the common shop/craft term). Scratch stock, scratch beader, hand beader and beading tool are overlapping observed labels, not a fixed taxonomy: identify the actual body, fence, cutter, depth control and stroke rather than the name. The patent-office term of art in the records found is consistently "beading tool" or "hand-beader" (US 284,777; US 335,856; US 319,161); none uses "scratch stock" [SS1][SS11][SS12][SS34]. The "Windsor beader" lead resolves to US 319,161, Orlando E. Williams and Lawrence V. Poole of Windsor, Vermont, 1885 (see variations).
• family/category: A molding-profile hand tool defined by operation: a guided holder presents a shaped cutter at a high scraping angle to form a small repeated profile or groove. The cutter is filed or supplied with the reverse (negative) of the desired profile. It is a profile-forming cutter that works by scraping — "scraper, not cutter" is needlessly categorical. Mechanically related to the Card Scraper and Toothing Plane (high-angle scraping action), which is a functional and construction affinity, not documented descent [SS1][SS3][SS16].
• one-line: A small, fenced holder for a hand-filed steel cutter, scraped along an edge in light repeated passes to cut a custom bead, flute, quirk, or inlay groove that no molding plane or router bit happens to match.

⚠ Naming collisions and boundary cases — checked and flagged, not conflated. The Stanley No. 66 and the current Lie-Nielsen derivative are two-handled fenced beading tools using interchangeable flat profile cutters. Lie-Nielsen's current page and 2023 manual describe seven separate double-ended flat profile blades giving fourteen shapes, plus a blank; the blades scrape, generally on the pull, but pushing is also permitted [SS31][SS32]. They overlap mechanically with the scratch-stock family, though their cast body, adjustable fences and standardized cutters distinguish them from the simplest shop-made L-shaped stock. (An earlier version of this entry described the No. 66 as using "a rotary cutter disc with six selectable profiles" scraping one way only; that was wrong — the cutter disk belongs to the Williams–Poole hand-beader, US 319,161 [SS34].) A "coach router" / "quirk router" / "sash router" is a distinct tool in the same broad family: Peter Follansbee, comparing a period coachmaker's tool directly against his own scratch stock, notes the blade is "not perpendicular like a scratch stock, but more pitched like a true plane" [SS9]. In luthiery, a "purfling cutter" (e.g., the Ibex, sold by StewMac) is a purpose-built twin-blade tool for binding/purfling channels; forum practice shows the terms overlap in casual use, with a shop-made scratch stock offered as a DIY substitute on that task [SS27][SS28]. Not interchangeable by default: scratch stock vs. purfling cutter, chair devil, coach/quirk router, molding plane or powered router — overlap must be matched on geometry and operation.

how NOT to use it (the heaviest section)

not-for: long, straight production runs (too slow, no advantage over a router or molding plane at volume); deep/architectural moldings beyond the tool's reach without a purpose-built larger holder [SS6]. Softwood is a caution, not a prohibition — trial on scrap [SS16].

dangerous: No scratch-stock-specific injury account was located. That is a research result, not a risk assessment, and it does not establish low risk. Hand scraping is low-energy relative to powered routing, but the tool has an exposed sharpened cutter, can slip when the track or fence loses registration, and is often made by cutting, snapping, filing or grinding saw or scraper steel — which adds sharp sheet edges, flying fragments, abrasive-wheel, heat and eye hazards. Controls, from CCOHS and OSHA general hand-tool and eye-protection guidance and the makers' instructions [SS35][SS36][SS37]:

• using it with the work held in the free hand — secure the work mechanically so both hands control the stock; never hold a small or unstable piece in the free hand.
• skipping the pre-use check — inspect the body, the clamp/screw and the cutter before each run; verify retention; take a cracked body or slipping clamp out of service.
• excess cutter projection — set the minimum projection the cut needs [SS3].
• hands or body in the slip path — keep fingers and the body outside the line the stock would travel if the fence lost registration.
• sharp non-working edges and saw teeth — deburr every non-working edge of the blank; remove or isolate any saw teeth. David Ray Pine, on making a cutter from an old bandsaw blade: "make sure to file the teeth off (enough to prevent yourself getting cut) and use the back side of the blade for your profile, where the metal isn't hardened" [SS3].
• filing an unsecured blank — clamp the blank close to the vise jaws in a sound, securely mounted vise to reduce vibration and slipping [SS37].
• cutting, snapping, aggressive filing or grinding metal without eye protection — wear eye/face protection where flying particles are present [SS36]; when grinding or cutting metal, use the machine's guards and a compatible, correctly rated abrasive, and control hot or flying fragments [SS35]. Prefer purpose-made annealed or maker-supplied blanks for novices [SS22][SS33].
• unprotected storage — protect the cutter in storage and when reaching into a tool chest.
• repetitive force — many passes on a long run add up; take short sections and rest rather than forcing deeper cuts.
• Follow the exact maker's cutter and sharpening instructions; do not apply a generic edge rule across variants [SS32][SS33].

damaging: ★ Traut designed clearance behind the thin spring cutter of his 1886 beading tool to reduce its tendency to chatter — the patent says the cutter may "spring slightly, and thereby have less tendency to chatter" [SS12]. That proves chatter was a design concern for that patented tool; it does not establish the prevalence, cause or severity of chatter across all scratch stocks. In general use, chatter and tearing come from model-specific causes — dull or burred cutters, excess projection, loose retention, poor registration, stroke direction against the grain — and are corrected by sharpening to the maker's geometry, reducing projection, re-seating the cutter, steadier fence pressure, or reversing direction [SS4][SS3][SS33]. Dull blades tear rather than scrape cleanly, producing rough finishes [SS4]. Filing a hardened production blade (e.g., Lie-Nielsen A2 at Rc 60–62) to change its profile, instead of using a blank, ruins the blade and the file [SS32].

beginner-trap: ★ on a basic shop-made stock, the missing depth stop is the central trap: "Since there's no depth stop on a scratch stock, you'll have to pay attention to how deep you're cutting and stop when you reach the desired depth of the profile" [SS16] — add a stop block (as Hock specifies) or choose a model with a gauge if consistent depth matters [SS33]. The same source names uneven fence pressure on the first passes, before the cutter's track self-guides, and trying to work the whole length in one continuous pass rather than short sections [SS16]. A lower-confidence source adds that tilting the tool on a rounded profile distorts a circular arc toward an ellipse, which may not show until the piece is finished [SS2]. Using salvaged steel without checking which part is fileable — and leaving teeth on — is the cutter-making trap [SS3].

misuse by analogy to related tools: ★ two independently sourced patterns. First — treating it like the plain freehand card scraper and skipping the fence: "would be almost impossible to use without a fence and make a straight profile, unless there is already a profile started or by using a batten on the piece of wood as a guide" [SS2]. Second — holding it to the standard of its powered alternative, the router, and expecting router-table consistency or speed: "Don't expect a scratch stock to create a profile as sharp and consistent-looking as a router bit... your results will likely have some minor variations" [SS16]. (Earlier wording called these "ancestor" and "descendant"; per the reviewer, they are related tools, not a documented lineage.)

correct use

Filing your own profile. Scribe the desired profile onto the end of a fileable blank and shape it with needle files; most simple profiles (beads, quirks, flutes) take only a few minutes [SS6][SS19]. Clamp the blank close to the vise jaws for filing, in a sound, securely mounted vise [SS37]. Prefer purpose-made fileable blanks (Hock, a maker's unhardened blank) for a first cutter; if salvaging saw steel, use the non-toothed portion and remove teeth and burrs first [SS3][SS33].

Light, repeated passes; depth control depends on the model. The profile deepens over many light passes rather than being cut to depth in one stroke [SS16]. A basic shop-made scratch stock usually has no integral depth stop, so depth must be watched or controlled with an added stop block [SS16]. Some patented and commercial beading tools include or permit a depth gauge: Hock specifies an attachable stop block for fixed depth [SS33]; Traut's 1883 beading-tool patent includes a bottom gauge "for regulating the depth of the cut" [SS11]; Traut's 1901 fence-support patent describes setting a support to the target depth. On the No. 66, lower the blade progressively per the manual [SS32]. Keep cutter projection to the minimum the cut needs [SS3].

Registration. Secure the work mechanically so both hands control the stock [SS35][SS37]. A flat fence registers most solidly on straight stock; curved fences are needed for curved edges but demand more deliberate control [SS21]. Keep fence pressure even on the first passes, before the cutter's own track begins to guide it [SS16].

Grain and direction — model-specific. The high-angle scraping action is often less direction-sensitive than a low-angle slicing cut, but grain, stock, cutter geometry and registration still determine the preferred direction. Tearing or chatter signals a need to change stroke angle or direction [SS15][SS21]. Hock says its tool can be pushed but pulling is more satisfactory [SS33]; Lie-Nielsen's No. 66 is generally pulled but may be pushed [SS32]; Lee Valley's cast stock works push or pull [SS38]; the Williams–Poole patent moves the cutter to either side of the guide specifically to suit grain [SS34].

Sharpening — follow the exact cutter maker. Hock blanks: file the profile edge square (90°), polish it with a slipstone, and polish both faces to remove burrs [SS33]. Lie-Nielsen No. 66 blades: stone the flat of the hardened blade and work the profile only rarely with slipstones, preserving the supplied profile [SS32]. Shop-filed salvaged cutters: keep the filed profile and deburr after filing [SS19]. No family-wide convex-edge rule is supported; an earlier version's "slightly convex" instruction and its uncertain "[sic: waste]" quotation are removed.

matching (the workbench's first read)

• does: forms a small repeating decorative profile — bead, quirk-bead, reed, flute, shallow rabbet, or a narrow inlay/stringing groove — by repeated light scraping passes, not bulk material removal.
• suits (strong matches): one-off custom beads/reeds/flutes; stringing grooves; short, stopped or faded profiles; curved work compatible with the exact fence and body; repair and conservation copying an existing small molding (Judd's and Bishop's documented uses); restricted access; work where quiet, direct tactile control matters [SS6][SS14]. Conditional matches (depend on the exact tool, cutter, stock and setup): softwood; concave or convex edges; deep profiles; long runs; push stroke; use without a depth stop.
• kit-fit: pairs with the card scraper (similar steel and action) and with chisels/a block plane for cleanup [SS21]; a marking gauge for layout; a vise or clamps for workholding. Alternatives: molding plane; plow or router plane for appropriate grooves; purfling cutter; other scratch/beading-tool variants; and a powered router with a safe guided setup and suitable bit. Match a scratch stock where a custom small profile, a stopped/faded cut, a short run, restricted access, a quiet hand process, or antique irregularity makes router setup or cutter availability unfavorable — Highland markets the Hock stock as working "quietly, in the time it can take to dig through your drawer of router bits and collets" [SS22]. Routers remain suitable for many straight and curved edges (bearings, fences, templates) when a safe guided setup and matching bit exist; their limits are cutter geometry, minimum radius, access, setup, tearout and desired surface, not curvature as such. For long, straight, production-volume runs, a router table/shaper or molding plane is the better match.
• not-for / caution (project-keyed): long production runs (slow per linear foot, hand fatigue); deep or architectural moldings beyond the holder's reach — the basic shop-made tool reaches only about 1" from the edge before a larger blade/holder is needed [SS6]. Softwood is a caution, not an exclusion: best results are commonly reported in even, straight-grained hardwoods; softwoods may compress or tear — test the exact profile, stock, edge and direction on scrap [SS16].
• access: mixed, and not automatically beginner-friendly. A first-hand reviewer of the Lee Valley cast tool: "not an easy scratch stock to use. It is definitely not a first timer's choice... the curved fence offers very little registration" [SS21]. A first-timer needs guidance on: making or choosing the cutter safely (see dangerous), sharpening to the maker's geometry, setting projection with clearance rather than flush (flush setups clog), light multi-pass strokes, workholding, and reading grain as it emerges.
• revival: currently listed by multiple specialty makers and retailers (Lee Valley, Hock/Highland Woodworking), with a strong shop-made culture; Lie-Nielsen's No. 66 is made intermittently and was marked unavailable when checked (2026-07-30) [SS22][SS31][SS38]. A current listing proves present availability, not continuity. Modern makers pick it up for the reasons Garrett Hack names — unlimited custom profiles, matching an antique profile no bit reproduces, low cost of entry — and because hand scraping is quiet and typically low-dust (cutter filing or grinding still makes particles) [SS19].

anatomy

Anatomy varies materially across the family; match by model, not by the classic form.

• Classic shop-made stock (Woodsmith, Fine Woodworking, Canadian Woodworking pattern). An L-shaped wooden block: the beam holds the cutter near-vertical and the short leg forms the fence. Usually no sole and no integral depth stop [SS16][SS18][SS24].
• Hock Scratch Stock. Bamboo-plywood body radiused at one end for inside curves; spring-steel cutters held with light screw pressure, presented at 90° to the work; an attachable stop block sets a fixed depth; straight/convex and concave setups per the maker's instructions [SS22][SS33].
• Lee Valley Cast Scratch Stock. Cast body with convex sole/fence for straight and curved work, push or pull use, single-point, beading and blank cutters, and shallow-rabbet capability [SS20][SS38].
• Stanley No. 66 / Lie-Nielsen Iron Beading Tool. Two-handled cast body with adjustable straight and curved fences and interchangeable flat double-ended profile blades (Lie-Nielsen: seven profiled blades, fourteen shapes, plus a blank); blades lowered progressively [SS31][SS32].
• Williams–Poole hand-beader (US 319,161, 1885). Two-handled stock with an adjustable guide and a sliding cutter plate carrying a cutter disk; the cutter can be moved to either side of the guide to suit grain [SS34].
• The cutter — material varies; verify it. Three distinct kinds: (1) maker-supplied fileable blanks — Hock supplies spring-steel cutters meant to be filed; Lie-Nielsen's custom blank is unhardened A2 [SS22][SS32][SS33]; (2) hardened production cutters — Lie-Nielsen's profiled No. 66 blades are hardened A2 at Rockwell 60–62, not for filing [SS31][SS32]; (3) salvaged steel — handsaw, bandsaw or scraper stock, used in shop-made stocks [SS19][SS24]. Salvaged saw blades may have harder teeth and a softer back, and some scraper stock is difficult to cut or file: verify the material, use the non-toothed/fileable portion, and remove teeth and burrs before mounting [SS3].
• Retention. Shop-made versions typically capture the cutter in a saw kerf pinched by a screw, sized to the cutter's thickness; commercial versions use a clamp screw or lever cap. Check retention before each run [SS18][SS21][SS33].
• Fence face. Flat fences register more solidly on straight stock; curved/convex fences follow curved work but are harder to control — a first-hand reviewer of the Lee Valley tool: "the curved fence offers very little registration, and care must be taken to hold the body square" [SS21].
• The working edge. Cutters differ: single-point cutters, full-profile flat blades, beading cutters, and the Williams–Poole disk. On a filed shop-made cutter, Garrett Hack describes the cutting action coming from the sharp corner left on the filed profile [SS19]; on maker-supplied blades, follow the maker's edge geometry (see correct use — sharpening).

variations (living)

• Shop-made (classic form) — a wood block with a cutter filed from scrap steel, buildable in under an hour at near-zero material cost; described consistently across multiple independent how-to sources [SS18][SS24][SS25][SS3].
• Lee Valley/Veritas Cast Scratch Stock — a current-production reproduction of the Victorian-era Edward Preston & Sons "Patent Reeding, Rabbeting and Moulding Tool No. 1393S" (Preston founded 1825). Cast stainless steel, black powder-coat finish, made in Canada, roughly 5" wide × 1½" tall and only ⅝" front-to-back so it reaches into inside corners; ships with a single-point cutter, a beading cutter, and a blank cutter for custom profiles [SS20].
• Hock Scratch Stock — a modern commercial hybrid developed with blade maker Ron Hock: bamboo-plywood body (2" × 4¾", radiused on one end for inside curves), two spring-steel cutters (one precut with a 3/16" quirk-bead profile), designed to also accept user-filed custom profiles [SS22][SS30].
• Stanley No. 66 / Lie-Nielsen Iron Beading Tool — two-handled fenced beading tool with interchangeable flat double-ended profile blades (see the naming note above). Stanley made the No. 66 from 1886 to 1941 [SS17][SS26]. Lie-Nielsen status, dated: its product page (checked by the reviewer, 2026-07-30) marks both No. 66 versions currently unavailable and describes intermittent small-batch production — "made intermittently," not continuous production [SS31].
• Williams–Poole hand-beader (the "Windsor beader") — US 319,161, granted June 2, 1885, to Orlando E. Williams and Lawrence V. Poole of Windsor, Vermont: a two-handled hand-beader with an adjustable guide, a sliding cutter plate and a cutter disk, the cutter movable to either side of the guide to suit grain [SS34]. (An earlier version held this as an unverified Lawrence Poole attribution; the primary record resolves it and names both inventors.)

trades (where it appears)

Cabinetmaking and joinery (inlay borders, stringing grooves, small moldings); 17th–18th century joined-furniture reproduction (Follansbee's own working practice); furniture repair and restoration (Judd's professional use, Historic New England's institutional conservation use); coachmaking, via the related coach-router cousin; chairmaking, via the related-but-distinct chair devil; and, informally and at the boundary, luthiery, via the scratch stock's use as a DIY substitute for a dedicated purfling cutter.

lineage

Related tool — the card scraper (relationship, not documented descent). Modern shop-made scratch stocks often reuse card-scraper steel and share its high-angle scraping action, and one trade source describes them as "First made by users as a simple holder for a scraper blade" with "a fence arrangement to work a measured distance from an edge" [SS6]. That establishes functional and construction affinity. No documentary evidence of transmission from card scraper to scratch stock was located, and this entry's own history section leaves the tool's origin unresolved, so the relationship is stated as "mechanically related to a card scraper and often made from similar steel," not as ancestry.

Contemporaries / siblings.

• Card Scraper and Toothing Plane — the corpus's own entries already name this tool; this run returns the citation. All three use a high-angle scraping action, which is often less grain-direction-sensitive than a low-angle slicing cut — but grain, stock, cutter geometry and registration still decide the preferred direction; none of the three is grain-indifferent [SS3][SS16][SS33][SS34].
• Molding plane — the tool the scratch stock most often stands in for, and the clearest documented functional contrast in this entry (see "history," below).
• Coach router / quirk router / sash router — a genuinely distinct cousin used historically in coachmaking to inlay a contrasting strip into curved members; blade is pitched like a plane iron, not vertical like a scratch stock's [SS2][SS9].
• Chair devil — a related but distinct chairmaking tool (a housed scraper functioning "much like a spokeshave," per Lost Art Press's account of designs shared between chairmakers Peter Galbert and Rich Pallaria) — named here as a boundary case, not folded into this entry [SS7].
• Purfling cutter (luthiery) — see naming note above; a parallel, task-adjacent tool with a soft, forum-level boundary against the scratch stock rather than a formal industry distinction [SS27][SS28]. Built out to its own corpus entry at Run 531.

Related commercial forms and powered alternatives (not descendants). The nineteenth-century patented beading tools — Traut 1883 and 1886, Williams & Poole 1885 — are related commercial forms, contemporaneous with or historically parallel to an undated shop-made form; they are not shown to descend from it [SS11][SS12][SS34]. The power router with a profile bit is a powered alternative / functional successor for many of the same profiles; shared function is not lineage. One enthusiast source claims a router bit "cannot cut some of the more delicate profiles made by a simple scratch stock or beader tool" — uncorroborated, held LOW, and not used for matching [SS2]. "Descendant" is reserved in this entry for sourced historical transmission, and none was found.

history (Petroski lens — what failure did each evolution fix?)

★ A mechanical affordance, documented in modern practice — not an origin story. A short guided cutter with no long sole can start, stop, and fade a profile in a short run where a molding plane's sole may not reach full depth. Peter Follansbee, examining a joined New Haven, CT chest (reproduced from Victor Chinnery's Oak Furniture: The British Tradition), writes: "The molding fades in & out at the juncture between the horizontal rails and the vertical muntins... I think you can't do that with a molding plane — the length of the plane's sole would prevent you from reaching that full-depth in such a short run, while still fading out before the muntins" [SS9]. John Lloyd (Woodworkers Institute) makes the same mechanical point in modern technique writing on the mason's-mitre joint [SS8]. What this does not establish: Lloyd's statement that the scratch stock "was the main moulding weapon of choice" in the fourteenth or fifteenth century carries no period citation, and is retained only as modern explanation/reenactment of the mason's-mitre technique — not as proof of medieval prevalence. Nor does the affordance prove that it caused the tool's origin. Claims of medieval prevalence remain unratified until a period text, image or object is produced.

★ Earliest known form and period — an open historiographic question, named rather than resolved. Follansbee states plainly, of his own long-practiced technique: "I don't know the history of this tool. I do know I have never seen it by that name in any 17th-century records" [SS9]. The closest confirmed seventeenth-century analog is Joseph Moxon's Mechanick Exercises (1683), describing a "sweep" used to scrape moldings: "To this Tool is filed a Tooth of Steel with such Roundings and Hollows in the bottom of it, as I intended to have Hollows and Roundings upon my Work... even as they do in the Molding-Plains Joyners use" [SS9] — a primary description of the scraping-profile idea, which Follansbee calls "a rather specific tool," not proof of the modern stock, name or form. The earliest depiction resembling the modern tool is an engraving in Roubo's L'Art du menuisier (1774), which Follansbee reports secondhand via Ejler Greber's History of the Woodworking Plane and says he did not check against the original French — a lead awaiting inspection of Roubo's original plate and text, not a confirmed depiction [SS9]. The comment thread is comment-level evidence: on Follansbee's post, Drew Young asks where the notion that they existed comes from, having found nothing in Moxon or Holme; Follansbee replies "yes, they existed. no, we don't know what they looked like"; Robert Trent writes that use of scratch stocks "is I think documented for 18C curved or broken scroll cornice moldings," without a citation in the comment, and Young asks for documentation [SS10]. The page establishes the names and the exchange; it does not establish all participants' credentials or that they independently examined the same evidence. Bottom line carried into this entry: eighteenth-century use is plausible but incompletely documented in the sources checked; seventeenth-century existence and appearance are unresolved; medieval prevalence is unsupported.

No source connects Petroski's "form follows failure" framework to the scratch stock by name — checked explicitly, not found. The stopped/faded-profile affordance above is a documented mechanical advantage, not a sourced design-failure origin.

human side

Peter Follansbee — the single most substantive named authority in this entry. A 17th-century joined-furniture reproduction specialist (formerly associated with Plimoth Plantation), documents his own routine technique on his blog: cutting "crease moldings" on joined oak chests with a scratch stock in combination with a plow plane (grooving the center) and a round plane (removing waste before scraping), and checking the finished profile "by laying a scrap on edge across the work to see the shadow cast" [SS9][SS10]. His account is the entry's clearest bridge between the historical-uncertainty question above and a real, dated, hands-on modern practice.

Robert S. Judd, a professional furniture repair and restoration specialist, supplies this entry's clearest restoration-trade use case, distinct from new-build joinery: "In my repair and restoration business, I often need to duplicate broken or missing moldings... no power tool can match the irregularities of the handworked wood found in older pieces" [SS6].

Karen Bishop, Mellon Conservation Fellow at Historic New England, made a custom scratch stock in 2022 to reproduce trim on a replica of a 17th-century carved box, filing a steel blank "to match the negative space around the profile of the remaining original molding piece" — a dated, named, institutional conservation use documented on the museum's own blog [SS14].

Garrett Hack, a nationally known furniture maker writing for Fine Woodworking, gives the entry's clearest design-freedom account: "there's really no limit to the profiles you can create... they've also changed my design aesthetic, allowing me to play with edges to create custom profiles" [SS19].

Trades and adjacent communities. Coachmaking, via the coach-router/quirk-router cousin (see lineage) [SS2][SS9]; chairmaking, via the related-but-distinct chair devil, developed and shared between named makers Peter Galbert and Rich Pallaria [SS7]; and inlay/stringing work, noted in a 1994 patent's background language: "scratch stocks or hand beading tools have long been used by cabinetmakers for forming grooves or other molded shapes in work pieces... to receive inlaid strings and bands" — background statement, not proof of any origin date [SS5].

Forum testimony. A Sawmill Creek thread documents a bead on a curved hunt-board skirt: the poster built a scratch stock from 8/4 hard maple, a 5/16" knurled screw, and an old card scraper, and reported: "there's no magic involved... maintain steady pressure and repeat the cut until you're done" [SS15]. A Fine Woodworking forum thread confirms continued informal use for the narrow groove for line-and-berry stringing inlay [SS29].

Honest gap. No source ties a specific named individual to the earliest (17th–18th century) use of the tool — the human-side record for that period is entirely the historiographic dispute named above under "history," not biography. The living human record belongs to named 20th–21st century practitioners and restorers, not period makers.

proof of existence

Confirmed primary patents. US 284,777 ("Beading-Tool," Justus A. Traut, New Britain, CT, assigned to Stanley Rule and Level Co., granted Sept. 11, 1883) describes a bench-plane-class beading tool with a variable-width cutter clamp, fence/gauge, spurs and a bottom depth gauge [SS11]; it should not silently stand for the minimal L-shaped scratch stock. US 335,856 ("Beading-Tool," Traut, granted Feb. 9, 1886) supports the thin spring cutter and its specific anti-chatter clearance, and frames itself as an improvement on "hand beading-tools of the class which are similar to spokeshaves" — evidence the fenced hand-held beading-tool class was established enough by 1886 to have prior art [SS12]. US 319,161 (Orlando E. Williams and Lawrence V. Poole, Windsor, VT, June 2, 1885) is the two-handled hand-beader with a cutter disk [SS34]. A modern design patent, USD591580S1 (Lee Valley Tools, priority 2008), and Lee Valley's current product show modern design and sale; they do not alone prove uninterrupted lineage from Preston or an earlier shop-made form [SS13][SS38]. Checked absence: no patent found uses the literal term "scratch stock."

Institutional record — a dated, named, modern conservation use, not a period artifact. Historic New England documents a 2022 project by a named Mellon Conservation Fellow (Karen Bishop) making and using a scratch stock to reproduce 17th-century molding trim [SS14]. This is real, high-reliability institutional evidence of the tool's continued use in professional conservation practice — but it documents a modern reproduction tool, not a surviving period artifact. Checked absence: no specific catalogued scratch-stock object record was found at Colonial Williamsburg, Winterthur, or the Mercer Museum in this research pass, despite all three institutions holding general early American tool collections at a scale that plausibly includes one — an honest gap in this pass, not a confirmed negative, and a candidate for a deeper follow-up search against those museums' own object databases.

No archaeological evidence was found, and no dated 17th-century physical example — consistent with, and explained by, the open historiographic dispute named under "history."

references

(Reference kind: book | video | blog | forum | reddit | museum | patent | named_maker | manufacturer | trade_press | retailer | vendor_blog | archaeological)

The reviewer's corrections (17)

• BLOCKER — 1. Absence-of-incidents safety assurance — Exact claim: “Physical injury risk is low,” followed by “no scratch-stock-specific injury account, forum incident, or safety-body warning was found” and a “low-severity, low-documented danger profile.”
• SHOULD-FIX — 2. Stanley/Lie-Nielsen No. 66 mechanism — Exact claim: The No. 66 “uses a rotary cutter disc with six selectable profiles and scrapes in one direction only,” making it mechanically distinct from a flat-blade scratch stock.
• SHOULD-FIX — 3. Universal absence of depth control — Exact claim: “Light, repeated passes — no depth stop,” and “the tool’s complete lack of a depth stop is named directly as the central beginner trap.”
• SHOULD-FIX — 4. Williams–Poole patent left unresolved — Exact claim: The “Windsor beader” is attributed only to Lawrence Poole in 1885, with patent number, date, and inventor described as unconfirmed.
• SHOULD-FIX — 5. Direct descent from the card scraper — Exact claim: “Every technique source checked agrees the scratch stock is a direct, fenced, profiled evolution of the freehand card scraper.”
• SHOULD-FIX — 6. Router and No. 66 as “descendants” — Exact claim: The power router/router bit is a “descendant,” and the manufactured beading tool is a “parallel commercial descendant.”
• SHOULD-FIX — 7. Grain-direction insensitivity — Exact claim: Scratch stock, card scraper, and toothing plane are said to share “grain-direction insensitivity.”
• SHOULD-FIX — 8. Fourteenth-/fifteenth-century adoption stated as fact — Exact claim: John Lloyd’s article is used to say that the scratch stock was the “main moulding weapon of choice” in the fourteenth or fifteenth century and that the tool fixed the molding plane’s historical failure.
• SHOULD-FIX — 9. Follansbee evidence overpromoted — Exact claim: The Follansbee posts are described as a live dispute among “three named furniture historians working from the same primary sources,” and the eighteenth-century existence of the tool is called “near-certain.”
• SHOULD-FIX — 10. Chatter patent made universal — Exact claim: Traut’s 1886 patent is said to prove chatter as “a historically-recognized, engineered-against failure of a rigidly-mounted scratch-stock blade.”
• SHOULD-FIX — 11. Sharpening geometry is internally mixed — Exact claim: “Sharpened on a file and whetstone on each edge,” with an added instruction that the edge “should stay slightly convex” and an unexplained “[sic: waste]” quotation.
• SHOULD-FIX — 12. Cutter material generalized — Exact claim: The cutter is “a thin card of spring temper steel,” classically from handsaw, bandsaw, or scraper stock, “soft enough to file easily, hard enough to hold an edge.”
• SHOULD-FIX — 13. Softwood exclusion — Exact claim: Softwoods are included under “not-for.”
• SHOULD-FIX — 14. Router/molding-plane matching absolutes — Exact claim: Scratch stocks suit curved edges “a router or molding plane can’t easily follow,” and power routers cannot cut some delicate profiles a scratch stock can.
• MINOR — 15. Anatomy presented as universal — Exact claim: The body is an L-shaped block or cast metal beam, and the cutting point is a small filed arris rather than a continuous edge.
• MINOR — 16. “Dust-free” and access framing — Exact claim: Revival is attributed to “dust-free” work and current commercial continuity.
• MINOR — 17. Schema and confidence cleanup — Exact claim: Frontmatter says “in-progress — first full pass, ready for review,” reviewed_by remains null, and MEDIUM-HIGH confidence rests partly on current pricing and unverified historical leads.