BERNARD LEVINE'S KNIFE LORE   NUMBER 58   MAY 1993

     (c)1993 Bernard Levine, exclusive to the NKCA



PERRET ON POCKETKNIFE PATTERNS

     In the January and March installments of this column we met 

18th century Parisian master cutler Jean Jacques Perret.  We did 

not actually meet him in person, of course.  Rather, we read a 

translation into English of several chapters from his three-

volume masterwork, L'Art du Coutelier (The Art of the Cutler), 

published in 1771.  

     In January Monsieur Perret explained to us how he made twist 

damascus steel, a strong and attractive material for which he 

found such diverse uses as table knives and executioners' swords.  

That 222 year old explanation should (but won't) put to rest the 

silly lies of such pseudo-historians as Raymond Thorp, who 

claimed that the art of forging damascus was somehow lost, and 

needed to be rediscovered.  Ever since 1771, when the making of 

both real and fake damascus was commonplace, the method of 

forging damascus has been available to anyone with a library 

card.  Indeed, damascus steel blades were made in Europe, without 

any interruption, from long before Perret's time right through 

1945, and are made there still today.  

     In March Perret showed us some of the hand tools that could 

be found on his workbench.  He explained in detail how those 

tools were made -- on the principle that anyone competent enough 

to make knives and scissors, was certainly competent enough to 

make drills, scrapers, and pliers.  He also explained, in general 

terms, how those tools were used.  

     This month we shall return to The Art of the Cutler. In this  

installment, Perret shares with us one of the most important but 

least understood elements of the cutler's art, the making and 

using of patterns.  

     Some contemporary hand knifemakers claim that no two of 

their knives are ever alike.  This claim can be a boast of 

artistic freedom, as was the case with William Wales Scagel 

(1873-1963), but more often it is really a confession of 

technical incompetence.  Much like an abstract painter who cannot 

draw, a knifemaker who cannot repeat a successful design has not 

really mastered his craft.  

     The key to repeatable standard designs for a knifemaker is a 

set of durable patterns.  This is especially important for makers 

of folding knives, because each folding knife must not only look 

like its prototype, it must also operate mechanically just as 

smoothly as its prototype.  A springless folder of the simple 

"penny knife" type can be individually tuned with a hammer, but a 

knife with blades, springs, liners, and locking devices is a 

precision machine with no more margin for error than the 

proverbial gnat's eyelash.  

     This fact about pocketknives as precision machines was just 

as true in 1771 as it is today.  Anyone who tells you that knives 

were "primitive" back then is either ignorant, or lying, or both.  

In fact the best knives then were more mechanically complex, and 

sometimes more ornate, than almost any knives being made today.  

Of course plain and simple knives were also made then, just as 

plain and simple knives are made today, but even those plain 18th 

century knives were very far from primitive.  

     Perret achieved consistent appearance and consistent 

mechanical action in his folding knives by the careful use of 

patterns.  Here is his explanation, from L'Art du Coutelier.  




Chapter 15

PATTERNS, AND HOW ONE MAKES THEM

     by Jean Jacques Perret


     Nothing contributes more to the regularity of one's work, 

and to speeding up one's execution, than does making use of 

patterns.  This is especially true of master part patterns which, 

having been made with precision, serve as a basis for all similar 

pieces which one plans to work upon.  

     Each pattern is made of steel forged out to a length that is 

appropriate to the item one wishes to make.  Their thickness is 

around that of a 24 sols piece [an old French quarter].  One need 

not temper patterns precisely, but it is important to harden them 

to a blue color, about that of water, for two essential reasons.  

     First, if they were not hardened at all, one might readily 

damage them with even the lightest touch of a file "escaping" 

from the workpiece.  Second, if they were made too hard, they 

would be very brittle, and would be at risk of shattering into 

many pieces when they are clamped in the vise with the workpiece.  

One can avoid the latter sort of accident by tempering them to a 

cherry red color, after hardening them to a color like water.  

     For the rest, each pattern does not need to be filed [and 

finished], except along its edges.  It is sufficient merely that 

they be forged uniformly with the hammer, and that they be what 

one might call well planed.  

     What is of ultimate importance, however, is that they be 

precisely and squarely laid out, with exact dimensions of length 

and width.  This is because if a pattern has even one little 

flaw, that same flaw will be repeated in every single workpiece 

made with that pattern.  

     Even though patterns are a great help in speeding up one's 

work, nonetheless there really is no such thing as a general 

purpose pattern for cutlery making [as contrasted to tools, most 

of which do serve a variety of purposes].  Now we are going to 

detail some of the [various types of] patterns which can be 

useful.  

           PLATE XIX GOES SOMEWHERE NEAR HERE



DINNER KNIFE PATTERNS

     The first pattern is the one for table knives.  Private 

citizens ordinarily buy a matching set of a dozen table knives, 

or at least six.  It is necessary that all of these knives be 

equal in length and width.  This equality cannot be achieved 

efficiently, except by filing all their blades against a pattern.  

     Figure 1 shows a pattern for a pointed table knife [what we 

would today call a steak knife].  Figure 2 shows another style, 

with a rounded tip.  For each style it is necessary to make 

patterns for a variety of lengths and widths.  



POCKETKNIFE PATTERNS

     Figure 3 shows the pattern for the blade of a spring knife.  

Figure 4 shows the pattern for its backspring.  For these styles 

it is also necessary to make a variety of different sized 

patterns, such as from 3 to 6 [French Royal] inches.  [A French 

inch was 1/12 longer than an English inch, so this size range was 

actually 3-1/4 to 6-1/2 of our inches.]  

     Note that the method of filing the parts of a spring-back 

knife against their patterns is only applicable to those with a 

square tang, and without any "secret" mechanisms.  Look at the 

tang (A) of Figure 3.  One may, by means of a well-made pattern, 

file two pieces at a time, by putting one on each side, and the 

pattern in the middle.  One joins the two tangs by means of a 

"false" [or temporary] rivet, and one uses the hand-vise to hold 

the two points together.  Then, without any risk, one can file 

both pieces right down to the pattern.  

     Figure 5 represents the pattern for a bistoury [a surgical 

knife similar to a scalpel] with a convex cutting edge.  Figure 6 

is another bistoury, with a concave cutting edge.  Figure 7 is 

one with a straight cutting edge.  [Folding bistouries did not 

have backsprings.  The stubby projection of the tang served as a 

blade stop.]  

     Figure 8 is the pattern for an "oat grain" lancet [a blood-

letting instrument whose point was shaped like an oat grain].  

Figure 9 is the pattern for a "pyramidal grain" lancet.  Figure 

10 is the pattern for a "barley grain" lancet.  

     Both with lancets and with bistouries, one files two pieces 

at a time, but they should only be held together with a "false" 

rivet [not at the points].  

     A "false" rivet is a rivet made of steel, which tapers 

almost imperceptibly from one end to the other.  One rounds it 

carefully.  Then one smoothes it.  Then one hardens it, and draws 

it to blue.  See Nos. 1, 2, 3 at the bottom of Plate XX [below].  

[A modern name for Perret's "false rivet" would be "drift pin."]  

     Figure 11 shows the pattern for the backspring of a one-

blade quill knife; figure 12 is the pattern for its blade.  

Figure 13 is the pattern for the spring of a double-ended quill 

knife [with the second blade shown below Figure 12].  

     Figure 14 [far right] shows the pattern for the handle of 

the spring-back quill knife with one blade.  Figure 15 [to the 

left of Fig. 14] is the one for the two-bladed quill knife.  


SQUARE AND CLEAN JOINTS

     Figure 16 shows the pattern for [the handle of] a spring-

back knife with a square tang.  One must pay careful attention to 

this example, because it provides the basis and standard for all 

the others [which follow].  

     I will demonstrate that in locating the rivet holes 

geometrically, one will readily succeed in making a knife with 

all the correct proportions required of a square tang.  By 

contrast, it is a lot of trouble to accomplish this when one 

has located the holes haphazardly.  

     [As I explain in Levine's Guide III, on page 274, under the 

picture of four folders reproduced from Perret, the square and 

clean joint was common practice (and a mark of quality) on 18th 

century French and English spring-back pocketknives.  In this 

elegant construction, the square tang exactly filled the joint in 

all three positions: open, half-open, and closed.]  

     Figure 16 is divided into four equal parts, the long way.  

On the center line, numbered (1), is located the first hole, for 

the [head] end of the spring.  On line (2), at (e) [not marked] 

one locates the center hole for the spring.  And halfway between 

lines (1) and (3) is placed the hole for the blade [pivot].  Can 

you see that the tang of the is blade well squared [Fig. 3], and 

that plenty of width is allowed for the spring?  That is why the 

handle is divided into four equal parts.  One allows three parts 

for the width of the tang, and one part for the width of the 

spring.  

     What's more, the center hole is not shown in the sketch.  

[To locate this hole] it is necessary to divide the length of the 

pattern into five equal parts, and to allow three parts between 

(e) [the location of the center hole] and (g) [the head or non-

blade end].  

     [This seems to be a typographical error, because the center 

hole in the spring (Fig. 4) is two-fifths of the way from the 

end, not three-fifths.]  

     To make sense of this observation, it is necessary to 

consider that the flexible part of the spring is between (e) [the 

center hole] and (f) [the blade end, not marked], while the stiff 

part is between (e) and (g) [the head end].  Now, if the 

flexible part is not one-third longer than the stiff part, the 

spring will not be able to take the strain that would be required 

of it.  It will break at (h) [not shown, but probably in front of 

the center hole].  

     If one gives more than three parts [of five] to part (e-f), 

one will be obliged to give more length to the spring, without 

which it would become too feeble to "return" [make the blade snap 

into place].  This larger sized spring would produce a 

destructive force, in that the edge of the blade would be caused 

to strike against the backspring, and would be chipped every time 

one wanted to close the knife.  

     Alternately one could abandon the rules of the [cutler's] 

art, and make the blade narrower than the handle.  Thus, this is 

the principle: when one reaches the stage [in the art] where one 

would like to make a square tang knife, it is necessary to follow 

the divisions of lines and the dimensions of holes, as they are 

shown on the pattern in Figure 16.  



DOUBLE-END KNIVES

     Figure 17 shows the pattern for a two-blade knife with 

square tangs [on both blades].  Although it appears to have been 

laid out differently [from Figure 16] the final result is the 

same.  Much as Figure 16 is divided into four parts, Figure 17 is 

divided into eight parts.  And it is by the latter [example] that 

I shall allow you to see clearly the exact proportions of the 

square tang, which occupies six parts of the handle; the hole 

finds itself placed naturally in the center [of the tang]; and 

two parts are reserved for the width of the spring.  

     The bottom of the pattern (B) is divided in the same manner, 

because it also holds a square tang, such as those drawn in 

Figures 18, 19, & 20.  The first of these [Fig. 18] is the 

pattern for a cork-pulling hook.  [Corkscrews had only recently 

come into use in 1771, and they would not entirely replace the 

much less efficient cork-hooks in France for another generation 

or two.]  The second [Fig. 19] is the pattern for a small punch.  

The third [Fig. 20] is the pattern for a long punch, suitable for 

wagon drivers.  [This style of eyed punch was primarily designed 

for mending broken harness.]  

     Figure 21 [center] shows the pattern for the spring of a 

double-end knife.  



COMPASS JOINT

     Figure 22 [left of center] shows the pattern for a knife 

called "compass joint" (a tete de compas).  The layout for this 

style is different from the others.  The tang of the blade cannot 

be made to stop halfway open [because it is round].  The blade is 

held open and shut by a little hook practically at the end of the 

spring, as shown at (C) in Figure 23 [Plate XX].  That hook 

engages one of the two notches made in the tang of the blade at 

(D), in Figure 24.  So, to give this tang the required strength, 

it is necessary to divide the width of its pattern into five 

parts, allowing four parts for the tang, and only one for the 

spring.  

           PLATE XX GOES SOMEWHERE NEAR HERE


     Drill the hole in the center of the "compass."  Then file 

the head round, as drawn.  Leave the square at (K) for the spring 

to strike.  ["K" is not marked, but is probably the "run up" of 

the tang; see LG-III page 32.]  The spring is one-fifth the width 

of the pattern.  Then one will have a precise pattern, lacking 

only the two additional holes, which may be located by following 

the same directions as accompany Figure 16.  


LOCKBACK FOLDERS

     Figure 25 shows the pattern for a "button knife" (couteau a 

mouche).  This pattern is also divided into five parts.  One 

allows four parts for the tang, and one part for the width of the 

spring.  Also, part of the width of the spring is divided into 

two halves, of which one half is cut away to form the "channel" at 

(P), which accommodates the button on the end of the spring [not 

shown; the button is lifted to release the lock].  The other half 

[of the spring] is notched, so that it can engage [the notch in] 

the tang of the blade, much like a door latch, as indicated at 

(q) in Figure 26.  

     The button knife may be made either square-tang or compass-

joint.  For the square tang version, the handle should extend to 

the corners (r r), Figure 26.  One would leave the [rounded] 

extension to (O) for the compass joint version.  

     Figure 27 shows the pattern for a type of button knife 

called a [plug] bayonet knife, because it can be placed in the 

end of a rifle.  This pattern is laid out the same as the 

previous one.  Note that on these button knives one can only file 

the handle to the pattern.  The [ends of the] blade and the 

spring should not be filed [all the way down] to the patterns, 

because of the delicate adjustments required for the "secret" 

lock mechanisms.  

     Figure 28 shows a straight pattern, which can serve for 

shaping several styles of handles.  Examples include "pump" 

knives [with a lock release stud in the back; see LG-I page 274, 

Fig. 43], split-spring rosette secret knives [LG-III page 286, 

bottom right], military knives [LG-I page 274, Fig. 41], etc.  

All these different constructions will be detailed in later 

chapters.  


OTHER FOLDING KNIVES

     Figure 29 shows the pattern for the handle of a pruning 

knife that should never have a true square and clean joint.  It 

is necessary to leave 3 or 4 lignes [1 ligne = 1/12 French Royal 

inch] of additional height [shown] at (r), to give solidity to 

the blade, which it will need.  [The true square and clean joint 

is elegant, but it is not strong enough for such hard use as 

pruning trees.]  Only pruning knife handles can be shaped with a 

pattern, because [gardeners'] taste in blade shapes varies so 

much.  If you try to make a pattern for each one, you will soon 

find yourself with more pruning blade patterns than you can 

count.  

     Figure 30 shows the pattern for the blade of a gressoir 

[horticultural grafting knife].  Figure 31 shows its handle, and 

Figure 32 its spring.  The grafting knife handle pattern is laid 

out following the example in Figure 16, as far as the holes for 

the blade pivot and the center of the spring are concerned.  

However, two holes are required at the head end (x x), to hold a 

piece of ivory that one calls the ecusson [bark lifter], which is 

represented by the dotted line around (Z).  [In modern French an 

ecussonoir is a budding knife.]  

     Figure 33 shows the pattern for the handle of a bistoury 

with concave cutting edge.  [The blade pattern is shown in Figure 

6.]  Figure 34 is for making the handle of a straight-edged 

bistoury; Figure 35 for the handle of convex-edged type.  

     Figure 36 is the pattern for a lancet handle.  Figure 37 is 

the pattern for a razor handle.  Figure 38 is the pattern for the 

handle of a [surgeon's] folding lithotome.  

     Figure 39 shows the pattern for the handle of a spring-back 

quill knife.  Its handle is thick and fills the hand, much like 

the handle of a fixed blade pen knife.  Figure 40 shows the 

pattern for the blade of this folding pen knife, while Figure 41 

is the pattern for its spring.  

     Figure 42 shows the pattern for [the handle of] a split 

spring knife.  Figure 43 [in Plate XIX, bottom left] shows the 

pattern for [the handle of] a "pump" knife, also called a passe-

par-tout.  

                          *     *     *

     Mr. Bill Adams of Atlanta Cutlery Corporation has donated a 

fascinating collection of old Sheffield cutlers' tools to the 

NKCA Museum.  The collection includes a large variety of steel 

patterns, most of them used for making elaborate multi-bladed 

horseman's and sportsman's knives.  Many of the patterns are 

stamped with the mark of the cutler who made and used them.  I 

plan to discuss these tools in detail in a future issue of the 

magazine.  Here is a photo of a few of these steel patterns.  

                       PHOTO 1 GOES HERE

                          *     *     *