File

How it works

Each tooth is a small chisel-edged ridge pointing toward the tip of the file, so that the file cuts when pushed forward and slides when drawn back. The Nicholson File Company’s treatise of 1878 says this directly: the points of the teeth are formed so that the file can only cut on the ordinary or advancing stroke, and equal pressure on the back stroke damages the points of the teeth.

The teeth are made by cutting the surface of a steel blank with a chisel. A single-cut file has one course of parallel cuts set at an angle of 5 to 20 degrees to the center line of the file, according to Nicholson. A double-cut file has two courses crossing each other, the first (the over-cut) at 35 to 55 degrees and the second (the up-cut) at 5 to 15 degrees, leaving the surface covered with pointed teeth. Nicholson’s grades of coarseness, from coarse to smooth, are applied across both kinds.

The coarser files remove stock quickly and leave deeper scratches; the finer remove the scratches of the one before. Moxon described this sequence in 1703: the rough file takes off the unevenness left by the hammer, the bastard file removes the deep strokes of the rough, the fine file those of the bastard and the smooth file those of the fine, after which the work may be made smoother still with emery or tripoli.

Anatomy

Body (blade): The cut portion. Nicholson’s plates give cross-sectional shapes: flat, hand, mill, half-round, round, square, three-square, pillar, pitsaw and others.
Faces and edges: Faces are cut; edges may be cut or left smooth. Nicholson defines a safe edge or side as one left uncut so the file can be presented to the work without injury to the adjoining surface.
Heel and shoulder: The portion of the file to which the tang is attached; Nicholson advises forcing the handle up to the shoulder.
Tang: The spike for the handle. Nicholson says the tang is never included in the length of a file.
Point: The end opposite the heel, according to Nicholson.
Handle: Ordinarily wood, driven onto the tang. Nicholson says that of the many special handles devised none had yet bettered the plain wooden handle for most uses, if properly fixed.

Variants

Flat, hand and mill files: Rectangular sections. Nicholson’s flat bastard is one of the most common files in use; Nicholson lists the hand file separately as a heavier section with one safe edge.
Half-round, round, square and three-square files: Named for the section. Nicholson says the half-round bastard, with the round, square and flat bastards, meets the large majority of the machinist’s requirements.
Warding files: Thin files made parallel in thickness and tapered on the edges, used by jewelers, machinists and locksmiths for the ward notches in keys, according to Nicholson.
Saw files (handsaw taper, pitsaw, hook-tooth, bandsaw): Triangular or curved sections for sharpening saws. Nicholson says the handsaw taper is probably the most widely distributed of any file, entering the household as well as the workshop.
Cabinet and wood files: Wider and thinner than the half-round, used by cabinetmakers, pattern and last makers, gunstockers and other woodworkers, according to Nicholson.
Rifflers: Small bent files used by carvers in wood, metal and stone, and in shaping irregular pattern work, according to Nicholson.
Small Swiss and French files: Files for dentists, watchmakers and similar uses, which Nicholson says were said to be machine cut; Disston says Raoul’s machine of 1800 made only watchmakers’ files.

Specifications

Length · in: Measured without the tang; Nicholson says length bears no fixed proportion to width or thickness and gives customary lengths for each section.
Cut types · single, double, rasp: Nicholson’s three classes of tooth form.
Coarseness · rough, coarse, bastard, second cut, smooth, dead smooth: Nicholson’s six terms for the coarseness of teeth.
Single-cut angle · degrees: 5–20Obliquity of the single course of cuts to the center line of the file (Nicholson).
Double-cut angles · degrees: 35–55 (over-cut), 5–15 (up-cut)The two courses of a double-cut file (Nicholson).

Made to

ISO 234-1
ISO 234-2
DIN 7261

Safety and care

Safety

Nicholson advises that a file be properly handled and the handle forced up to the shoulder of the tang, not driven halfway, and that where heavy pressure is used the file be held firmly in a handle or holder, as a means of safety to the workman against the sharp point of the tang in case the file suddenly becomes detached.

Care

Nicholson advises cleaning a file with a file card or file brush, or by rubbing chalk across the teeth, during use and carefully before laying it aside.
Nicholson says files thrown loosely into a drawer with chisels, hammers and other tools are damaged because only slight rubbing on a hard surface dulls the teeth, and recommends separate provision for them.
Nicholson describes re-cutting, the working over of worn files by annealing, grinding out the old teeth, re-cutting and hardening, as a practice some shops used two or three times but whose economy “is very much questioned in some of the best appointed shops.”

Tips

Work the forward stroke only
Because the teeth point toward the tip, Nicholson says pressure should be relieved on the back stroke. It calls this a matter of economy.1
Go through the cuts in order
Moxon sets out a sequence of rough, bastard, fine and smooth files, each removing the scratches left by the one before. He also says to keep as much of the file’s length as possible ranging along the length of the work, so that the file rides on the rises and slips over the hollow between them until the rises are brought level with it.1
Use the safe edge to protect a shoulder
A file with one edge left smooth can be presented to the work without cutting the adjoining surface. Nicholson says the hand file has one edge cut and the other left safe, which makes it particularly useful where the flat file would not answer; it also lists square blunt files made with one or two safe sides.1
Draw-file for a fine finish
Nicholson describes holding a file at each end and moving it sideways across the work, as with a spokeshave, to lay the marks of turned work lengthwise or to give a final fit to a shaft. The teeth then cut with a shearing action and leave finer scratches, provided the file is held so the faces of the teeth meet the work at a cutting angle.1
Set the vise at elbow height
For general filing Nicholson puts the top of the vise jaws level with the worker’s elbow, which it says averages 40 to 44 inches from the floor, with 42 inches an average. Delicate work should sit higher, and heavy work lower so the body’s weight can be thrown on the file.1
Chalk the teeth to stop pinning
Nicholson says rubbing chalk or charcoal across the teeth of a new file and brushing it off removes oil, which is desirable on glassy cast-iron surfaces, while good workmen sometimes fill the teeth with oil and chalk on new files for narrow, harder fibrous metals. On fine files, oil reduces the tendency to pin and scratch the work.1
Take the scale off castings first
Nicholson says the outer skin and chilled edges of iron castings can be harder than the file itself, and that scale should be removed by pickling and the chilled surfaces ground before filing.1
Keep a convex face for flat work
Nicholson says a file with some convexity from point to heel gives control and is needed to touch the exact spot indicated by a straightedge or surface plate on broad surfaces, because a file that is straight or hollow cannot be so directed.1

History

Earliest documented

c. 1500 BC (bronze file from Crete, as dated by Disston); iron and bronze files from the Hallstatt period

Crete; Hallstatt, Austria; Thebes

A 1922 pamphlet published by Henry Disston & Sons says that among the earliest abrading instruments of metal for which a date can be fixed is a bronze file, 3-1/2 inches long, with a rounded back and a flat face, dug up in Crete by a University of Pennsylvania expedition and believed to date from about 1500 BC; that bronze files and one iron file were found in a prehistoric cemetery at Hallstatt; and that an Assyrian-period set of iron armourers’ tools found at Thebes by Flinders Petrie, probably left in 666 BC, included a rasp and a file. The pamphlet is a company publication that reproduces earlier reports; the excavation reports and museum records themselves have not been read.1

The file is among the oldest metal tools in the record, but the record for it is secondhand. The Disston pamphlet of 1922 collects reports of a bronze file from Crete dated to about 1500 BC, bronze and iron files from the Hallstatt cemetery, an Assyrian file of the seventh century BC and the armourers’ set that Petrie found at Thebes. The pamphlet adds that Schliemann found no trace of files at Troy or Mycenae, and explains that iron files left few remains because iron corrodes and such tools were so ordinary that writers rarely mentioned them. The Roman file the pamphlet pictures, from Silchester, was cranked and had teeth five to the inch, filed in.

Until the nineteenth century the file was a hand-cut tool. Nicholson’s treatise of 1878 describes the hand cutter seated before an anvil with the blank strapped down by the feet, striking a chisel with a peculiarly shaped hammer to throw up a barb across the surface, then sliding the chisel forward to the barb for the next blow. It concludes that hand-cut files have teeth of some irregularity depending on the skill of the workman, and that machine cutting aimed to approach a regular irregularity by deliberate variation.

The machine for cutting files has a long and disputed history. Disston dates the first authentic record to Leonardo’s drawing of about 1505, and notes later attempts by Duverger, Fardonet and others in the eighteenth century, and that James Watt failed because his machine lacked a pressure foot to hold the file in place. Nicholson says only that Duverger’s invention is the first account of such an attempt, dated 1699, and that Raoul’s machine of 1800 produced watchmakers’ files. Disston calls the Bernot machine, introduced into Great Britain about 1860 and patented in the United States in 1860, the first really practicable one, and reports that W. T. Nicholson invented a machine in 1865 that is still widely used.

American file making grew up in the 1860s. Disston reports that before 1840 manufacture was practically unknown in the United States, that a small works began at Matteawan, New York, about 1845, and that Europe continued to supply most files until about 1864. Nicholson’s company was incorporated in 1864; in its 1878 treatise it reports that most American files were by then machine cut, and it argues that the extreme regularity of the first machine-cut files made them chatter and cut channels in the work, so that cutters worked toward deliberate irregularity in tooth spacing.

Moxon’s account of 1703 shows how little the practice of using files changed. He describes the sequence of rough, bastard, fine and smooth files and the order in which they are applied to a forged piece; Nicholson’s 1878 hints on filing add advice on grasping the file, on the height of the vise and on cleaning. The file remains in wide use: O*NET’s Tools Used table lists files by commodity, as flat hand, half-round, mill saw and bastard cut files, in 7 of the 24 occupations mapped to this archive’s trades, including machinists, welders, plumbers and automotive technicians.

Timeline

c. 1500 BC
A bronze file with a rounded back and flat face is dug up in Crete, according to the Disston pamphlet.1
7th century BC
Assyrian-period iron armourers’ tools at Thebes include a file, according to the Disston pamphlet quoting Petrie; the pamphlet also pictures an Assyrian file of this century.1
c. 600 BC
A combined flat and round file from Hallstatt, Austria, is shown in the Disston pamphlet, which dates it to about 600 BC.1
c. AD 300
A cranked file from the Roman remains at Silchester, with teeth five to the inch, is illustrated in the Disston pamphlet.1
c. 1505
Leonardo da Vinci draws a file-cutting machine, the first authentic record of such a machine according to the Disston pamphlet; the Science Museum Group’s model of the machine follows a Codice Atlantico folio it says was written between 1483 and 1518.1 2
1699
The French mechanic Duverger presents a file-cutting machine to the Académie des Sciences, according to the Journal des Savants as cited by Disston, and as stated by Nicholson.1 2
1703
Moxon’s Mechanick Exercises describes the square, flat, three-square, half-round, round and thin files, in rough, bastard, fine and smooth cuts, and the order in which they are used.1
1800
A Frenchman named Raoul builds a machine that produces watchmakers’ files reported to equal the best hand-made English files.1 2
1812
Morris B. Belknapp of Greenfield, Massachusetts, patents an American file-cutting machine that did not prove commercially successful, according to Disston.1
1860
The Bernot file-cutting machine of Paris is patented in the United States on 24 July, and Disston calls it the first really practicable machine.1
1864
The Nicholson File Company of Providence is incorporated, according to its 1878 treatise; Disston reports that until about 1864 Europe supplied most files used in America.1 2
1865
W. T. Nicholson of Providence invents a file-cutting machine that Disston says was improved and is still widely used in the United States.1
1878
Nicholson states that much the larger proportion of files made and used in the United States are machine cut, and that machine cutting of the larger files was surmounted only within the past fifteen or twenty years.1

Where the sources disagree

When did Leonardo conceive the file-cutting machine?

Some time before 1505, while he served as engineer to the Duke’s army in 1503.1
Henry Disston & Sons
The design is in a Codice Atlantico folio written between 1483 and 1518.1
Science Museum Group

When did Duverger build his file-cutting machine?

1699, the year of the invention, as stated by Nicholson in 1878.1
Nicholson File Company
About 1735, according to what Disston says “has been stated”; Disston itself says the Journal des Savants of the Académie des Sciences shows the machine was presented to the Académie in 1699.1
Henry Disston & Sons (reporting both)

Gaps in the record (7)

The Science Museum Group holds a 1926 replica of a small ancient Roman file (object 1926-567, credited to Elkington and Company), but its record gives no date, findspot or description of the original, so the Roman form is known here only from the Disston pamphlet’s illustration.
The Cretan, Hallstatt, Assyrian and Silchester files are known through a company pamphlet of 1922; no excavation report or museum record has been read, so dates and findspots are the pamphlet’s own.
Greek, Roman and medieval file development, and the development of hardening, is not sourced.
Sheffield and the English file trade, which both Nicholson and Disston treat as the world center until the mid-nineteenth century, are not examined here.
No current standard for file dimensions, cuts or grades has been read; the standard slugs are guesses.
No independent test of files is cited, so buying is null.
Nicholson and Disston are manufacturers; their statements about the merits of machine-cut files are attributed to them and not used to rank products.